摘要:

ISSN 0306-0012 CSRVBR 25(3) 155-228 Chemical Society Reviews Volume 25 Issue 3 Pages 155-228 June 1996 Arene-catalysed Lithiation Reactions By Miguel Yus (pp. 155-1 62) The lithiation of different substrates in the presence of a catalytic amount of an arene [mainly naphthalene or 4,4’-di-rerr-butylbiphenyl (DTBB)I allows the easy preparation of organolithium compounds under very mild reaction conditions. Using this methodology, new routes for alkyllithium compounds from non-halogenated materials (allylic alcohols, or their 0-silyl or 0-mesyl derivatives, alkyl sulfates or phosphates, sulfides, sulfoxides or sulfones and nitriles) as well as oxygen-, nitrogen- and sulfur-containing functionalised organolithium compounds or polylithium intermediates from chlorinated materials or saturated heterocycles (three-, four- and six- membered systems) are developed.I I Chiral Discrimination by Modified Cyclodextrins By Chrisopher J. Easton and Stephen F: Lincoln (pp. 163-1 70) Naturally occuring cyclodextrins show only limited enantioselectivity in their interactions with chiral guests, because they form inclusion complexes in which there Is only limited interaction between chiral centres of the cyclodextrin and those of the guest. As the extent of interaction between these groups is increased, as a result of modification to the cyclodextrin, the stereoselectivity is often increased. Additional secondary bonding interactions, complexation to metallocyclodextrins and covalent host-guest interactions all lead to enhanced chiral discrimination.Through-bond and Through-space Models for Interpreting Chemical Reactivity in Organic Reactions By Keith Bowden and Edward J. Grubbs (pp. 171-178) The relative importance of the two mechanisms considered to be responsible for localised polar substituent effects. 0 i.e.the electrostatic field and inductive effects, has been disputed. There are three relatively simple model systems which can assess the reality of these effects. The results appear to be consistently explained by the electrostatic field effect, but not by the inductive effect. Other reactive consequences of the electrostatic field effect are reviewed and computational approaches il I ustrated. The Chemistry of Paper Conservation By Vincent D.Daniels (pp. 179-186) Collectables on paper contain a large number of components of varying degrees of stability. The factors that cause these to deteriorate include oxidation, acid hydrolysis, air pollutants and water. Deacidification is often used by +conservators and there are beneficial effects from the introduction of Mg2 and Ca2 +;however, excessive alkalinity must be avoided. There are many mechanisms for the development of discolouration in paper. Several examples of ageing effects of inks and pigments are considered. Diatomic Molecular Probes for Mid-IR Studies of Zeolites By A. Zecchina and C. Otero Arean (pp. 187-1 98) Mid-IR spectroscopy of adsorbed probe molecules is a very active research field which contributes to a deepening of our knowledge of the local structure of zeolite active sites, and of the physico-chemical properties of the intra- zeolite space.This has strong bearings on the many technological applications of zeolites and of related microporous materials. This review discusses the use of diatomic molecular probes for IR studies of Brqjnsted and Lewis acidity, structural defects and internal electric fields. Dual acid-base pairs are also considered. Solid State Metathesis Reaction for Metal Borides, Silicides, Pnictides and Chalcogenides. Ionic or Elemental Pathway By 1. P Parkin (pp. 199-208) Solid state metathesis reactions offer a rapid, low external energy route to a range of inorganic ceramic materials including nitrides, oxides, pnictides, chalcogenides, silicides and borides.The reactions can be filament or bulk thermally initiated and often proceed with a thermal flash or propagation wave. The reactions can follow either an ionic metathetical or a reductive recombination pathway. Modelling of Solvent Effects on the Diels-Alder Reaction By C. Cativiela, J. 1. Garcia, J. A. Mayoral and L. Salvatella (pp. 209-218) This review presents a summary of the different approaches used quantitatively to describe solvent effects on Diels-Alder reactions. Basically, two methodologies are considered: on the one hand, empirical models, based on solvent parameter scales and linear regression analyses, and on the other hand, theoretical models, based on quantum chemical and Monte Carlo calculations. It is shown that both approaches lead to similar conclusions, and provide complementary information on these processes. After the Actinides then what? By Simon A.Cotton (pp. 219-228) More sophisticated nuclear syntheses, together with more rapid means of separation and detection, have led to the identification of elements with atomic numbers up to 1 1 1. Chemical experiments have been carried out for elements 104 and 105 (and are in prospect for 106) indicating them to be members of a 6d transition series. Chemical properties have been predicted for heavier elements but it remains to be seen if stable ‘magic number’ nuclei with atomic numbers around 1 14 can be made. Articles that will appear in forthcoming issues include A Radical Reappraisal of Gif Reactions M.John Perkins On the Mechanism of the Gif Reaction Derek H. R. Barton INGOLD LECTURE: Reactive Intermediates: Carboxylic Acid Enols and Other Unstable Species A. J. Kresge Photoelectron Spectroscopy in a New Light: Zero Kinetic Energy (ZEKE) Photoelectron,Spectroscopywith Coherent Vacuum Ultraviolet Light John W. Hepburn The Changing Face of Arene Oxide-Oxepine Chemistry Derek R. Boyd and Narain D. Sharma New Approaches to Chemical Kinetics Barry Johnson and Stephen K. Scott Assembly and Encapsulation with Self-complementary Molecules Julius Rebek, Jr. Application of Fluorescence Microscopy to a Study of Chemical Problems R. S. Davidson Designing New Lattice Inclusion Hosts Roger Bishop Potential Energy Surface Crossings in Organic Photochemistry Fernando Bernadi, Michael Robb and Massimo Olivucci Specificity and Versatility in Erythromycin Biosynthesis Rembert Pieper, Camilla Kao, Chaitan Khosla, Guanglin Luo and David E. Cane Glutamate and 2-Methyleneglutarate Mutase: From Microbial Curiosities to Paradigms for Coenzymes B ,,-dependent Enzymes Wolfgang Buckel and Bernard T. Golding Nitrous Acid and Nitrile in the Atmosphere Gerhard Lammel and J. Neil Cape Environmentally Friendly Catalytic Methods James H. Clark and Duncan J. Macquarrie ‘Covalent’ Effects in ‘Ionic’ Systems Paul A. Madden and Mark Wilson Non-porphyri n Photosensiti sers in Biomedicine Mark Wainwright Scanning Transitiometry Stanislaw L. Randzio

ISSN:0306-0012    

DOI:10.1039/CS99625FP007

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Two of the most distinguished and widely respected names in chemical publishing, The Royal Society of Chemistry and the American Chemical Society have joined forces to co-publish this exciting new journal, which seems sure to become the first point of reference in its field. Organic Process Research & Development is a bi-monthly journal which will contain comprehensive reviews of developments IN EACH ISSUE YOU’LL FIND 0Substantive articles and reports from leading researchers 0Reports on original work in process chemistry 0Reports on scale up of process to pilot plant, with discussions of safety and environmental issues 0Special focus on the design of chemical processes, and their development and scale up from laboratory to manufacture 0Current R&D in the fine organic and speciality chemical industries relating to the batch and semi-batch chemical process industries, keeping you up-to-date with the very latest in fine organic chemical and speciality chemical industries including pharmaceuticals, electronics, agrochemicals, intermediates and speciality polymers -subjects never previously covered in depth by any existing science journal.Editor Trevor Laird Scientific Update East Sussex, UK Associate Editors john F Arnett j. F. Arnett & Associates West Chester, PA, USA Richard Pariza, C&P Associates North Zion, II, USA The first issue of Organic Process Research & Development will be available in January 1997. ORDER YOUR COPY TODAY TO order, please contact: THE ROYAL SOCIETY OF UK & Europe The Royal Society of Chemistry USA & Rest of World The American Chemical Society x G Turpin Distribution Services Ltd Customer Service and Sales Blackhorse Rd, Letchworth 11 55 Sixteenth Street, NW Herts SG8 1HN Wahington, DC 20036 USA Tel +44 (0) 1462 672555 Tel +1 202-776-8100 lriforrriativrr Fax +44 (0)1462 480947 Fax +1 202-872-6067 Services RSC & ACS Members are enttled to a discount.Please contact your society for further details m W PRODUETS DESlGlED AND PRODUCED VALID ANALYTICAL MEASUREMENT BY AMAlYSTS FOR ANALYSTS RAISING QUALITY THROUGH THE VAM INITIATIVE VAM (Valid Analytical Measurement) is a DTI initiative, spearheaded by the Laboratory of the Government Chemist (LGC), in conjunction with the DTI in order to improve the quality of analytical measurements made in the UK, and to facilitate the mutual acceptance of analytical data across international boundaries.This is done by: defining and disseminating best analytical practice which will enable laboratories to deliver reliable results every time; developing the tools which enable laboratories to implement best analytical practice; working with analysts in other countries to ensure the comparability of analytical measurements across international boundaries. The Royal Society of Chemistry is pleased to be working with the LGC to distribute VAM products -which include books, guidelines, videos and software -that are designed to be used by practising analysts, laboratory managers, analytical quality managers, and university lecturers and their students.GUIDELINES FOR ACHIEVING kQUALITY IN TRACE ANALYSIS M, Sargent h G. MacKay Presents a concise protocol that acts as a check-list of key issues for those involved in carrying out trace analysis, or who rely on the results. Softback book; 1995;ISBN 0-85404-402-7;f 12.50 I $22 GENERAL PRINCIPLES OF , , GOOD SAMPLING PRACTICE N.7: Crosby 8; /.Patel A welcome reference for researchers and professionals who need to access the important information on how to sample, in the form of an instant, thorough and reliable guide. Softback book; 1995;ISBN 0-85404-412-4;f 17.50 I $32 TRACE ANALYSIS: A STRUCTURED APPROACH TO OBTAINING RELIABLE RESULTS Edited by E.Pritchard, with G. MacKay & J. Points A highly practical and systematic bench guide which deals with the science rather than the paperwork of quality systems, taking the analyst step by step through the stages of any trace analysis. Softback book; 1996;ISBN 0-85404-417-5; f69.50 1$125 --*-, INTERNATIONAL GUIDE TO QUALITY IN 'c ANALYTICAL CHEMISTRY: AN AID TO ' ACCREDITATION A C/TAC* documenf, published by the LGC Provides laboratories with guidance on best practice for improving quality of the analytical operations they perform, and is cross referenced throughout to the related parts of IS0 Guide 25, IS0 9000 and OECD GLP Principles. * Cooperation on International Traceability in Analytical Chemistry Softback book; 1996;ISBN 0-94892-609-0;f30 I$55 PROFICIENCY TESTING IN ANALYTICAL CHEMISTRY RE.Lawn, M. Thompson h RE Walker Deals exclusively and comprehensively with the role of proficiency testing in the quality assurance of analytical data, addressing topics such as the organisation, effectiveness and costs and benefits of proficiency testing. Softback book; Due autumn 1996; f22.50 1$39 TRACING THE CHEMICAL LINKS An educational package considering the concepts of calibration and traceability in the context of analytical chemistry, and explaining how these are an essential part of achieving comparability for analytical measurements. Video plus illustrated booklet; 1995; f35 (inc VAT) I $75I Rest of World f40 CONFIDENCE IN ANALYSIS An educational package looking at the way -analysts establish confidence in the validity of their results.It follows through a complex multi-stage analysis from sampling to reporting and shows how the uncertainty can be estimated. Video plus illustrated booklet; 1995; f35 (inc VAT) / $75I Rest of World f40 fl A computer-aided learning package for the "AMSTAT analytical chemist. Includes techniques, practical examples and questions and covers: basic concepts; significance tests; handling non-normal data and outliers; ANalysis Of VAriance (ANOVA); quality of analytical data; regression and applications. Software package (IBM or compatible); Standard single user licence f90 (+VAT) / $160;Multi-user site licence f290 (+VAT) / $520;Discounts available for academic institutions.To order any of these products, or for more detailed information, please contact: Jenny McCluskey, Sales & Promotion Executive Tel +44 (0) 1223 420066 Royal Society of Chemistry Fax +44 (0) 1223 423623 CHEMISTRYThomas Graham House, Science Park E-mail: sales@ rsc.org Information Milton Road, Cambridge CB4 4WF, UK wwweb: http://chemistry.rsc.org/rsc/ Services Ref No 1313 Pesticides Developments, Impacts, and Control Edited by G.A. Best Clyde River Purification Board, East Kilbride, Glasgow, UK A. D. Ruthven Scottish Agricultural Science Agency, Edinburgh, UK Pesticides -Developments, Impacts, and Control covers the latest developments in pesticide usage, with particular emphasis on the regulations that safeguard users, consumers and the environment.It provides a comprehensive guide to the whole "story" of the pesticide business, from the efforts of manufacturers to produce products that are both effective and environmentally benign; the difficulties and hazards associated with the application of these substances; their environmental effects, particularly in water; and the control of storage, use and residue levels in non-agricultural habitat and in foodstuffs. It covers the use of chemicals in fish farming and weed control, and methods of detection of pesticide residues in ground water, rivers and the North Sea. To do so it draws on a wide range of expertise and experience, with contributors active at the various stages, demonstrating the value of good science in presenting relevant and reliable data on environmental impacts.Pesticides -Developments, Impacts, and Control explodes many of the myths about pesticides, and draws attention to less familiar problems and the means of addressing them. Thorough and up-to-date, it is core reading for undergraduates in environmental and agricultural science courses, researchers working with pesticides, and legislators and officials with health and safety responsibilities. Special Publication No. 174 Hardcover x + 180 pages ISBN 0 85404 785 9 1995 Price f49.50 RSC members Price f32.50 Ref No 1312 Crop Protection Agents From Nature Natural Products and Analogues Edited by L.G.Copping Critical Reports on Applied Chemistry, Volume 35 This book is a complete critical review of natural crop protection. It examines the complete range of potential crop protection agents from nature, and identifies those areas that show potential within the crop protection industry. Crop Protection Agents From Nature: Natural Products and Analogues examines the secondary metabolites that occur in nature including bacteria, fungi, algae, higher plants and animals. It evaluates their biological effects, and, wherever possible, identifies their mode of action. In addition, the book discusses opportunities for commercial exploitation and lists the regulatory requirements. Overall, this book is an excellent examination of all aspects of the natural world as a source of crop protection agents. It will be of great interest to researchers in bio- organic and agrochemistry, as well as professionals in the pesticides industry. Hardcover xxviii + 500 pages ISBN 0 85404 41 4 0 1996 Price f129.50 RSC members Price f89.50

ISSN:0306-0012    

DOI:10.1039/CS99625BP009

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

The Royal Society of Chemistry Chemical Society Reviews Editorial Board Professor H. W. Kroto FRS (Chairman) (University of Sussex) Professor M. J. Blandamer (University of Leicester) Dr. A. R. Butler (University of St. Andrews) Professor E. C. Cafstable (University of Basel, Switzerland) Professor T. C. Gallagher (University of Bristol) Professor D. M. P. Mingos FRS (Imperial College London) Consulting Editors Dr. G. G. Balint-Kurti (University of Bristol) Dr. J. M. Brown (University of Oxford) Dr. J. Burgess (University of Leicester) Dr. N. Cape (Institute of Terrestrial Ecology, Lothian) Professor B. T. Golding (University of Newcastle upon Tyne) Professor M. Green (University of Bath) Professor A. Hamnett (University of Newcastle upon Tyne) Dr.T. M. Herrington (University of Reading) Professor R. Hillman (University of Leicester) Professor R. Keese (University of Bern, Switzerland) Dr. T. H. Lilley (University of Sheffield) Dr. H. Maskill (University of Newcastle upon Tyne) Professor A. de Meijere (University of Gottingen, Germany) Professor J. N. Miller (Loughborough University of Tech no Iog y ) Professor S. M. Roberts (University of Liverpool) Professor B. H. Robinson (University of East Anglia) Professor M. R. Smyth (Dublin City University, Republic of Ireland) Professor A. J. Stace (University of Sussex) Chemical Society Reviews aims to foster current progress in the chemical sciences and related disciplines. The journal has the broad appeal necessary to enable scientists to benefit from recent advances made in research outside their immediate interests.In particular, students embarking on a research career should find Chemical Society Reviews a particularly Chemical Society Reviews (ISSN 0306-0012) is published bimonthly by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge, UK CB4 4WF. All orders accompanied by payment should be sent directly to The Royal Society of Chemistry, Turpin Distribution Services Ltd., Blackhorse Road, Letchworth, Herts., UK SG6 1HN. N.6.Turpin Distribution Services Ltd., distributors, is wholly owned by The Royal Society of Chemistry. 1996 annual subscription rate: EEA f120.00; Rest of World f123.00; USA $225.00.Customers in Canada will be charged the Rest of World price plus a surcharge to cover GST. Customers should make payments by cheque in sterling payable on a UK clearing bank or in US dollars payable on a US clearing bank. Second-class postage is paid at Jamaica, NY 1141-9998. Airfreight and mailing in the USA by Publications Editorial Staff Managing Editor Martin Sugden Editorial Production Peter Whittington Editorial Secretary Debbie Halls Editorial Office The Royal Society of Chemistry Thomas Graham House Science Park Milton Road Cam bridge UK CB4 4WF Telephone +44 (0)1223 420066 Facsimile +44 (0) 1223 420247 Electronic Mail (Internet) csr@rsc.org or sugdenm@rsc.org http://chemistry.rsc.org/rsc/ Advertisement sales Telephone +44 (0)171 287 3091 Facsimile +44 (0) 171 494 1134 Typeset by Servis Filmsetting Ltd.Printed in Great Britain by Black Bear Press Ltd. stimulating and instructive springboard to further reading. The Editorial Board encourages an international and interdisciplinary approach to science, which is reflected in the succinct, authoritative articles commissioned. The Board members welcome comments and suggestions; these should be directed to the Managing Editor Expediting Services Inc., 200 Meacham Avenue, Elmont, NY 11003, and at additional mailing offices. US Postmaster: send address changes to Chemical Society Reviews, c/o Publications Expediting Services Inc., 200 Meacham Avenue, Elmont, NY 11003. All despatches outside the UK by Bulk airmail within Europe and Accelerated Surface Post outside Europe. PRINTED IN THE UK. 0 The Royal Society of Chemistry, 1996. All rights reserved. No parts of this publication may be repro- duced, stored in a retrieval system, or transmitted in any form, or by any means, electronic, mechanical, recording, or other-wise, without the prior permission of the publishers.

ISSN:0306-0012    

DOI:10.1039/CS99625FX013

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0306-0012    

DOI:10.1039/CS99625BX015

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Arene-catalysed Lithiation Reactions Miguel Yus Departamento de Quimica Organica, Facultad de Ciencias, Universidad de Alicante, Apdo. 99, 03080 Alicante, Spain chromatographic means are necessary to know when the reaction 1 Introduction Organolithium compounds are very useful intermediates in syn-thetic organic chemistry mainly in carbon-carbon bond forming processes by reaction with carbon electrophiles.' Among the differ- ent methods to prepare this type of organometallic compound, the most versatile is probably via halogen-lithium exchange, bromine and chlorine being the most commonly used halogens. For this purpose commercially available lithium is in general reactibe enough to perform this transformation unless the reaction has to be carried out at low temperature; in this case it is necessary to activate the metal.ZOne way to get very active lithium is to dissolve the metal in a stoichiometric amount of an arene.3 almost always using tetrahydrofuran as solvent. As arenes, naphthalene (Np) and 4,4'-di-tert-butylbiphenyl (DTBB) are the most frequently used.Five years ago4 we found that the use of a catalytic amount of an arene in the lithiation of functionalised chlorinated precursors is a very powerful method to prepare unstable functional ised organolithium compounds5 under very mild conditions (Scheme 1 ). Some advantages of this methodology compared to the use of a sto- ichiometric amount of arene are: (a)yields are similar or better in the catalytic version; (b)reaction times are far shorter (1 instead of 8 h); (c)reactions are very clean, and form only the desired product (no byproducts resulting from the reaction of the arene radical- anion and the electrophile); (4the method avoids separation of sig- nificant amounts of the arene; (e)the reaction can be followed by a simple colour change: at the begining (before adding the substrate to be lithiated) the reaction mixture shows the colour of the lithium- arene (dark green for naphthalene and dark blue for DTBB), and after addition of the chlorinated material the colour disappears and the mixture becomes again coloured at the end of the lithiation step, when the substrate has been consumed.Thus no spectroscopic or Miguel Yus was born in Zaragoza in 1947, and received BSc (1969).MSc (1971) and PhD (1973) degrees from the University of Zaragoza. After spending two years as a postdoctorul fellow at the Max Planck lnstitut fir Kohlenforschung in Mulheim ad. Ruhr he returned to Spain to the University of Oviedo where he became assistant professor in 1977, being promoted to full professor in 1987 at the same university. In 1988 he moved to a chair in orgunic chemistry at the University of Alicante where he is currently the head of the Organic Chemistry Department. Professor Yus has been viAiting professor at different institutions suchs as ETH-Zurich and the universities of Oxford, Harvard, Uppsala, Marseille and Tucson. He is a member or fellow of the chemicul societies of Argentina, UK, Germany, Jupan, Spain, Switzerland and the United States of America.He is coauthor of about 170 papers mainly in the field of development of new methodolo- gieJ involving organometallic intermediates. His current re- search interest is focused on the preparation of very reactive functionalised organolithium compounds and their use in syn- thetic organic chemistry. finishes. To the best of our knowledge, before this finding only a few examples had been described in which a catalytic amount of an arene was used as an electron carrier: (a)preparation of a strained adamantanol,6rr (b)lithiation of esters, 6h (c) reductive scission of cycl~propylacetylenes~and (6)reductive opening of oxetane.8 In addition, in some particular cases, this technique has been applied to the activation of other metal^.^ In the present review, the synthetic possibilities of the arene- catalysed lithiation will be explored in order to prepare: (a) organolithiurn compounds starting from non-halogenated pre- cursors, (b)very unstable functionalised organolithium intermedi- ates and (c) polylithium synthons RCl RLi Li*ArH ArH [A~H=N~.DTBB....] ii Scheme I 2 Organolithium Compounds from Non- halogenated Mater ia Is 2.1 Reductive Carbon-Oxygen Cleavage Allylic or benzylic alcohols 1are transformed into the correspond- ing organolithium compounds 2 by successive deprotonation with n-butyllithium and DTBB-cataiysed lithiation; final reaction with different electrophiles gives, after hydrolysis, the expected products 3.Alternatively, the same reaction products are available starting from the corresponding 0-silyl derivatives 4, in this case the catal- ysed lithiation being performed in the presence of the electrophile (Barbier-type process) (Scheme 2).10"Recently, this methodology has been applied to the synthesis of olivetol and related com-pounds.Ioh 1) Bu"Li, 0°C 1) E+ ROH -f 1 R = CH,=CHCH,, CH,=CMeCH,, CH2=CHCHMe, PhCH,, PhCHMe, geranyl R'=Me, Ph E' = Me,SiCl, Pr'CHO, PhCHO, Et,CO, CH,[CH,],CO Scheme 2 Another indirect transformation of alcohols 1 into the corre- sponding organolithium derivatives consists of their conversion into mesylates 5. The naphthalene-catalysed lithiation of these materials in the presence of electrophiles yields, after hydrolysis, the expected products 3.The two-step process can be carried out at 155 CHEMICAL SOCIETY REVIEWS, 1996 expected alkyllithium compounds 2, which behave as usual towards electrophiles giving products 3 This reaction has been recently applied to the synthesis of a-silylated organolithium intermedi- ates l5 When the same procedure is used with phenyl sulfoxides 9I6(l or phenyl sulfones it is necessary to work under Barbier-type reaction conditions in order to avoid decomposition of the in situ generated organol ithium compound even at low temperatures (Scheme 6) Li,Np(l%)PhSR 8 -78°C PhSOR 9 1) Li,Np or DTBB (5-8%),E', -78 or O°C 2) H20 RE 3 [14-95%] PhS0,R10 1) h,Np(8%),E', -78 to 20°C 2) H,O I R = Me, Et. CH,=CHCH,, Prl.PhCH, , I E' H,O, Me,SiCI, Pr CHO, PhCHO, Et,CO,CH,[CH,],CO, Ph,CO Scheme 6 23 Reductive Carbon-Carbon Cleavage Nitriles 11 have been decyanated reductively using a DTBB- catalysed lithiation and working under Barbier-type reaction conditions at low temperature, so the intermediate organolithium compound of type 2 prefers to react with the electrophile present in the reaction medium instead of reacting with the starting nitrile (a-deprotonation or addition to the cyano group) (Scheme 70) " 1) LI.DTBB (5%).E', -78 or -30"C RCN * RE [21-63%]11 2) H20 3 R = Me, Et. c C,H,, Ph, PhCH, E-= Me,SiCI, PrICH0,n C,H,,CHO, Me,CO, Et,CO,C H,ICH,],CO,Ph,CO Scheme 7 3 Preparation of Functionalised Organolithium Compounds 3.1 Lithiation of Phenones and Phenone Imines Treatment of different phenones 12 with lithium and a catalytic amount of naphthalene yields the dianion 14, which by reaction with several electrophiles gives, after hydrolysis, the corresponding reac- tion products 15 (Scheme 8) When the corresponding imines 13 are used as starting materials the reaction has to be performed at lower temperature and under Barbier-type reaction conditions in order to avoid destruction of the corresponding intermediate of type 14, after hydrolysis, functionalised amines 16 are prepared (Scheme 8) Igh low temperature, so the corresponding solution of the organolith- ium compound 2 is obtained before the last reaction with the electrophile (Scheme 3) Ii 1) Li,Np(4%),E', 0°C 2) H2O R = CH,=CHCH,, CH,=CMeCH,, PhCH, , , E+ = (PhCH2S),.PrlCHO. PhCHO, Et,CO, CH,ICH,],CO, Ph,CO Scheme 3 As Schemes 2 and 3 show, the indirect transformation of alcohol derivatives into alkyllithium intermediates IS limited to allylic or benzylic systems This limitation has been overcome by working with the corresponding sulfates 6, in this case the process can be applied to aliphatic derivatives Thus, the two-step process (naph- thalene-catalysed lithiation followed by reaction with electrophiles) leads to the expected reaction products 3, through the correspond- ing organolithium compounds 2 (Scheme 4) lZnh When this methodology is applied to cyclic sulfates derived from 1,3-diols, the corresponding cyclopropanes are easily obtained after the lithiation step LiNp Np (4%) Ll R = Me, Et.Pr'. Bun. II C,H,,CHMe ,E* = (PhCH,S),, PrCHO, PhCHO, Et,CO, CH,ICH,I,CO, PrTOMe, Ph,CO Scheme 4 Another possibility to prepare organolithium compounds indi- rectly from all type of alcohols is the use of the corresponding phos- phates 7 as starting materials and working once again under Barbier-type reaction conditions Thus, using DTBB as the electron transfer catalyst, the expected products 3 are obtained. in which only one group is transferred from the starting phosphate Only in the case of the allyl derivative (it was not possible to prepare the tri- benzyl derivative to be tested) are the three allylic moieties con- verted into allyllithium The facility to form the alkyllithium intermediate, which could be established using mixed phosphates, follows the series allyl = benzyl > phenyl > primary alkyl >> -1 secondary alkyl, that is, in good agreement with the relative stabil- ity of the corresponding carbanionic intermediates (Scheme 5) I 1) Li, DTBB (5%),E'.-30°C2) H,O (I4) PhXRERE [31-90%](RO)3PO + [Y="]3 12Y=o 1)L1,Np (8%). E'. -78 to 20°C 15Y=O [I39623 = NR [21 75%]13Y=NR' 16 YR = Et, CH?=CHCH,, Pr', Bu", Ph 2) H,O E+ = Me,SiCI, PhMe,SiCI. PhCHO, Et,CO. PhCOEt Scheme 5 Finally, the naphthalene-catalysed lithiation has been used for the preparation of arylmethyllithium reagents from the corresponding methyl ethers (4 2.2 Reductive CarbonSulfur Cleavage Phenyl sulfides 8 react with lithium powder at low temperature in the presence of a catalytic amount of naphthalene to give the R = H, Me.Ph. 2 MeC,H,. 3 MeC,H,. 4 MeC,H,, 4 MeOC,H,, 2,4 Me,C,H, R = Me,Ph.c C,H,, E+, = Mel, , EtBr, PrCHO, PhCHO, Me,CO, Et,CO, CH2(CH21,C0, C H,[ CH,I,CO MeCN Scheme 8 3.2 Lithiation of Functionalised Chlorinated Materials Chlorinated phenols 17 or pivalanilides 20 are transformed into the corresponding dianions 18 and 21, respectively, by successive ARENE-CATALYSED LITHIATION REACTIONS- MIGUEL Y US deprotonation with n-butyllithium and naphthalene-catalysed lithiation, the subsequent reaction of these intermediates with difterent electrophiles yields the expected functionalised phenols 19 or anilides 22 (Scheme 9) l9 1) Bu'LI * 2) L,Np(3-12%).78 or 0°C 2) HzO 17Y=O 18Y=O 19Y= 0 (10-7846) 20 Y = BU'CON 21 Y = BU'CON 22~= BU'CON 160-9046] E' = PrICHO, Bu'CHO, EtZCO, CH21CH,l,C0, PhCN, PhCNO Scheme 9 a-Functionalised organolithium compounds, the so-called 'carbenoids,' are very unstable species owing to their tendency to undergo an a-el imination process However, intermediates of the type 24 can be prepared by a DTBB-catalysed lithiation of the cor responding chloroether 23 either in a two-step process at -90 "C or under Barbier-type reaction conditions, working in this case at 0 "C Following these two protocols functionalised ethers 25' are pre pared (Scheme 10) 2o LI, DTBB (5%).90°C 1) E+ -E~O"EEtO-Cl -23 E', 0°C 25 170 98%11) Li, DTBB(SW), E + = Bu"CH0. Bu'CHO, PhCHO. PrlCO, Bu;CO. C H21CH,I$0.cyclo hex 2 enone, PhCOMe. PhMe,SiCI, CO,. PhCN. PhCONMe,, cyclo C,H I I NCO, PhN=CHPh Scheme 10 Another type of very unstable a-functionalised organolithium compound are the acyllithium derivatives,2' namely the corre- sponding carbamoyl and thiocarbamoyl compounds, which have been prepared iri Jitu by naphthalene-catalysed lithiation of the chlorinated precursors 26 and 28, respectively Working at low temperature and under Barbier-type reaction conditions, the expected functionalised amides 27 or thioamides 29 may be pre- pared (Scheme 11) 22 v Y 1) LI,Np(3%), E+,-78T RZN lia 2) H20 * RZN 26Y=O 27 Y = 0 [62-79%] 28Y=s 29 Y = S [40-84%] R = Me.Pr' E' = EtCHO.PhCHO. MezCO. PhCOMe Scheme 11 P-Functionalised organolithium compounds of the type 31 with n = 1 are also very unstable owing to their easy decomposition through a p-elimination process In spite of that, these intermedi- ates and the corresponding y functionalised derivatives 31 with 17 = 2 can be prepared at low temperature by successive deprotonation and naphthalene-catalysed lithiation of the corre sponding chlorohydrins 30 yielding, after reaction with an electrophile and final hydrolysis, the expected alcohols 32 (Scheme 12) OLI OH1) Bu"LI,-78OC 2)L1,Np(l-2%).-78~C* R&y -R/I1Y;E 30 31 32 [68-89%1 R = Me,Ph ri = 1,2 E+ = H,O,PrCHO Scheme 12 ' bith cycloheu 2 enone as electrophile I 2 addition was the only process obsened 157 The stability of P-functionalised organolithium intermediates can be increased if the metal is attached to an sp2-hybridised carbon atom An example of this behaviour is the preparation of intermedi- ates 34 by naphthalene-catalysed lithiation of chlorinated allyl arnines 33 at low temperature, these species survive under these reaction conditions giving, by treatment with an electrophile fol-lowed by final hydrolysis, the expected functionalised allyl amines 35 (Scheme 13) However, the corresponding oxygen- and sulfur- containing derivatives (34 with X = OR or SR) suffer p-elimina- tion, even at low temperature and/or under Barbier-type reaction conditions, showing as expected, that oxygen- and sulfur-contain ing groups are better leaving groups than the corresponding nitro- gen-containing ones 23 33X=" 34 X=" 35 [20-7781 R,R' = Me, Ph, R-R'=ICH,],OICH,I, E' = Me,SICI, ButCH0,PhCH0.Me2C0.(cycloC,H,),C0.kH,ICH21,k0 Scheme 13 In contrast to the behaviour of systems of the type 33, the corre- sponding y-derivatives resulting from the naphthalene-catalysed lithiation of different oxygen , nitrogen-and sulfur-containing materials 36 can be trapped in utu with different electrophiles under Barbier-type reaction conditions giving, after hydrolysis, the corre- sponding products 37 (Scheme 14) 24 As the starting materials 36 are easily prepared from the corresponding dichlorinated precursor (36 with X = Cl), the preparation of products 37 represents a suc- cessive introduction of a nucleophile X and an electrophile E in the isobutylene skeleton XTCl 1) Li, Np(8%), E+,-78T 2) HzO 36 37 [21-7481 X = BullO, PhCH,O, morpholino, PhCH,NMe, PhCH,S E+ = Bu'CHO, Et,CO, CH,ICH21,C0 Scheme 14 y-Functionalised oxygenated organolithium compounds bearing a protected 1,2-diol moiety of the type 39 have been prepared in a racemic and enantiomerically pure form, depending on the starting material 38 Their low-temperature DTBB-catalysed lithiation can be carried out in a two-step process or in the presence of the electro- phile giving, after hydrolysis, products 40, which are isolated as a 2 1-1 I diastereoisomeric mixture when the electrophile is pro- chiral, so the observed asymmetric induction is very low (Scheme 15)25 R 1) E+ 39R-tti 38 1) LI,DTBB(5%).E'. 78'C I 40[16 92%]I 2) HzO R = R = Me. R-R = ICH,I, E+, = Me,3S~C1. PrlCHO. Bu'CHO. PhCHO. MeXO Et,CO, CH21CH21,C0, PhCOMe Scheme 15 Following the same strategy shown in Scheme 15, but combin- ing the effects of both catalysts with the temperature (naphthalene at low temperature and DTBB at room temperature), it is possible to prepare the very unstable sulfur-containing y-functionalised intermediate 42 from both brominated or chlorinated thioethers 41, which by reaction with electrophiles leads to products 43 Alternatively. the Barbier-type process performed at ambient tem- peratLre affords similar results (Scheme 16) The instability of intermediate 42 arises from the existence of acidic protons dt the CY position relative to the sulfur atom Compounds 43 can be easily transformed into the corresponding thiols by reaction with mercury(11) acetate-trifluoroacetic acid followed by treatment with hydrogen sulfide 26 h,DTBB (5%) Bu'S I -78°C Bu'Sj 42 I-""'U 41 I 1) Li.DTBB (5%). E* I 43 [45 82461 2) H,O X = Br,Cl E+ = H,O, D,O, Me&, CO,, Me,SICI. BuCHO, PhCHO. Me,CO, CH,[CH,],CO, cyclo-C,H,COPh Scheme 16 As mentioned above for intermediate 34 (Scheme 13), the pres- ence of a lithium atom attached to an sp2-hybridised carbon atom stabilises the species This is also the case for the y-functionalised intermediates 45, which can be prepared by low temperature DTBB-catalysed lithiation of the starting materials 44 and react with electrophilic reagents to give, after hydrolysis, products 46 The very low temperature used in the lithiation step is necessary to avoid partial decomposition of the chlorinated precursor 44 by dehydrochlorination The final deprotection of compound 46 has to be performed carefully to yield functionalised unsaturated ketones 47 and avoid decomposition of the final products (Scheme 17) 27 n Ll,DTBB (4%) n -90°C 44 45 n -47 [>95%] 46 i43-9081 R = Prn,Prl E+ = H,O, D,O, Bu'CHO.PhCHO, CH,ICH,I,CO, PhCOMe Scheme 17 Masked high-order lithium enolatest of the type 49 and 53 are interesting intermediates to transfer a remote carbonyl functionality to electrophilic reagents The naphthalene-catalysed lithiation of the corresponding chlorinated precursors 48 or 52 followed by reac- tion with different electrophiles yields, after under neutral hydro- lysis conditions, compounds 50 and 54, respectively, which could be easily transformed into functionalised ketones 51 and 55 under acidic conditions (Scheme 18) 28 Some of the products prepared are interesting for further synthetic manipulations, thus, when the electrophile is a carbonyl compound, the corresponding hydroxy carbonyl compounds are transformed into alcohols and cyclic ethers using a boron trifluoride-catalysed reaction with silyl deriva- tives 280b On the other hand, for 48 (a = 2) and using an O-pro- tected a-hydroxy carbonyl compound as electrophile, 6,8-dioxabicyclo[3 2 I ]octanes are prepared as final products, after acid hydrolysis, some of them (frontalin, brevicomins) being important biologically active molecules 28c Alternatively, inter- mediate 49 can also be prepared using lithium-naphthalene (stoi- chiometric amount of the arene) as lithiating agent 28d-f 33 Reductive Opening of Saturated Heterocycles The reductive opening of epoxides, which may be carried out with a stoichiometric amount of an a~ene,2~0 h can also be performed in a +These types of intermediates are masked lithium w enolates in which the lithium atom is attached to a carbon atom different from the corresponding one at the a position with respect to the carbonyl group CHEMICAL SOCIETY REVIEWS, 1996 48n=2 52 n= 3 R = H, Me, Et, Ph 1E+ = H,O, D,O, (PhCH,S),, PrCHO, Bu'CHO, PkCHO, Et,CO, C H,ICH,l,CO, PrTONICH, I,CH,, PhCON I CH,],CH,, PhCN, EtOCOCI, PhCH=NPh Scheme 18 catalytic fashion starting from chiral epoxides 56 and using DTBB as the electron-carrier agent Thus, via chiral intermediates 57, p-functionalised organolithium compounds, enantiomerically pure products 58 may be prepared (Scheme 19) As described above for intermediates 39 and owing to the high reactivity of this type of functionalised species, asymmetric induction was almost non-exis- tent with prochiral electrophiles Nevertheless, when products 58 are mixtures of diastereoisomers, they can be easily separated by flash chromatography, so giving both diastereoisomers enantiomer- ically pure This is a typical example of enantiomerically pure com- pounds (EPC) synthesis 29c This methodology has been applied to the synthesis of chiral polyols using a protected hydroxy epoxide (56 with R = MOMOCH,) as starting material OLl 1)E' EELi,DTBB (5%) -78°C R LL1 R 56 57 58 [58-69%] R = Me, MOMOCH, E' = CO,, BuCHO, PhCHO CH,(CH,],CO, PhCOMe Scheme 19 In contrast to the behaviour illustrated in Scheme 19 for epoxides, aziridines 59 cannot be opened by a lithium-arene reagent However, they suffer reductive opening using a catalytic amount of naphthalene, so P-nitrogenated organolithium intermediates 60 may be prepared, which by reaction with different electrophiles give the expected functionalised amines 61, after the final hydro- lysis$ A limitation of this reaction is that it works only if a phenyl group is attached somewhere on the aziridine ring Scheme 20 shows the reaction with substituted N-phenylaziridines 59 The process has been applied to chiral aziridines [easily prepared from enantiomerically pure (-)-ephedrine], so chiral products of the type 61 are accessible by this methodology 3O Ph PhNLl PhNHLI,DTBB (2 5%) 1) E+ R -78'C R LL1 R R = H,Me E = H20, D,O, Me,S,, MeI, CH,=CHCH,Br, Bu'CHO, PhCHO, .Me,CO, C H2[ CH, ],CO, (EtO),CO, CH,= CHCO,Me, PhCONI CH,] <CH,, PhCH=NPh Scheme 20 Azetidines suffer ring opening by means of a DTBB-catalysed lithiation but the process has to be performed at higher temperature in order to obtain a y-nitrogenated organ01 ithium intermediate, 4 Only conjugate additon was observed in the reaction of intermediate 60 with methyl acrylate as electrophile ARENE-CATALYSED LITHIATION REACTIONS- MIGUEL YlJS which by reaction with electrophiles followed by hydrolysis gives the final functionalised amines The process, which also needs the presence of a phenyl group on the azetidine ring, is exemplifed in Scheme 21 for the case of the phenyl-azetidine 62, which via the intermediate63 affords the final products 64 Ph PhNH ELI, DTBB (5%) PhNLl L1 1) E+0 -15OC U 62 63 64 [20-90961 E+ = H,O D,O CO, Bu'CHO PhCHO Me,CO CH,[CH,],CO PhCH=NPh Scheme 21 As was mentioned in the Introduction, oxetanes can be reduc- tivel y opened using a DTBB-catal ysed process Tetrahydrofuran itself has been reductively opened using lithium and a stoichiomet ric amount of DTBB in the presence of boron trifluoride 321r The cat alytic version of this process allows the transformation of tetrahydrofuran into the corresponding intermediate 65, which reacts with different carbonyl compounds to give, after hydrolysis, the expected 1,5-diols 66 Considering that these reaction products can be easily cycl ised to the corresponding tetrahydropyrans under acidic conditions, this methodolog) represents a homologation of the starting heterocycles (Scheme 22) 3*17 66 [35-80961 E+ PrICHO.Bu"CH0 Bu'CHO PhCHO, Et,CO. BuCOMe, PhCOMe Scheme 22 Substituted 2-phenyl- 1,3-dioxoIanes 67 are prone to be reduc tively opened using naphthalene as the catalyst giving dianions 68 and products 69 after successive reaction with electrophiles and hydrolysis (Scheme 23) 330 The same process has been applied to 2-vinyl- 1,3-dioxolanes, such as the 2-cyclopentenone derivative 70, affording products 72 in moderate yields via the dianion 71 (Scheme 24) 37c It is noteworthy that compounds 72 are the corre- sponding umpoled' variants when compared to the Michael-type addition of a nucleophile to cyclopent-2-enone R = H.Me,Ph E+ = H20,D,O, Me,CO, Et,CO, CH,[CH,],CO (n = 3,4,6)Scheme 23 Li.DTBB(2596) ~ ooc E 70 L1 72 l33-401 71 E+ = Bu'CHO, Et,CO, Pr;CO, MeCOEt, CH,[CH,]JO Scheme 24 The DTBB-catalysed lithiation of phthalan 7374aand isochro- man 7634haffords the corresponding dianions 74 and 77, respec-tively, which by treatment with different electrophiles followed by final hydrolysis leads to the formation of products 75 and 78, + The term Umpolung which was introduced by D Seebach (Angrw Chem Int Ed Engl 1979 18 239) means in a general context inversion of the reactivity respectively (Scheme 25) Diols derived from the reaction of inter- mediates 74 and 77 with carbonyl compounds are easily cyclised under acidic conditions to give 6-or 7-membered cyclic ethers, so the whole process represents a homologation of the starting materi- als 73 and 76 Similar results are obtained starting from thio deriv atives, namely thiophthalan 79 and thioisochroman 82, which are easily opened with DTBB as the arene catalyst yielding finally products 81 and 84, respectively, through the corresponding dian- ionic intermediates 80 and 83, respectively (Scheme 25) 74c Also in this case, the corresponding carbonyl derivatives are easily cyclised to yield 6-or 7-membered cyclic thioethers under acidic conditions Finally, and showing a parallel behaviour, 5-or 6-mem- bered nitrogen-containing heterocycles 85 and 88 suffer reductive opening giving dianions 86 and 89,respectively, the expected func- tionalised amines 87 and 90, respectively, being isolated after reac- tion with electrophiles followed by hydrolysi5 (Scheme 25) 34rf 73Y =O.n= 1 74Y =O,n= 1 76Y=O.n=2 77Y =O,n=2 79Y =S, n= 1 80Y=S,n=l 82Y=S.n=2 83 Y = S, n =2 85 Y =PhN, n= 1 86Y=PhN.n=1 88 Y = PhN,n= 2 89Y=PhN,n=2 75Y = 0, n= 1 [51-82961 78 Y = 0, n = 2 [44-89461 81 Y = S, n= 1 142-8946] 84Y = S, n= 2 [52-7146) 87 Y = PhN, n = I [32-70%] 90Y =PhN, n=2 [34 89961 E+ = H20, D,O, CO,, EtCH0, PrCHO, ByCHO, Bu'CHO, PhCHO, Me,CO, Et2C0, PrnCOMe, CH,[CH,J,CO, CH,[CH21,C0, PhCOMe Scheme 25 4 Preparation of Polylithiated Synthons Dichloromethane or dideuteriodichloromethane (91 with R = H,D) are lithiated under Barbier-type reaction conditions using DTBB as the arene catalyst and a carbonyl compound as electrophile to afford, after hydrolysis, the expected 1,3-diols 92 The same process may be applied to substituted trichloromethane 93 or even to tetra- chloromethane 95, in these cases chlorotrimethylsilane being the electrophile used, so persilylated compounds 94 and 96, respec-tively, are obtained (Scheme 26) 35 92 [21-61%] -40 or -78OCRCC13 [E+=M%SlCl] * RC(SiMe,),93 2) H2O 94 [51-98%] CCl, I 95 R = H, D, Me, Ph E+ = Me,SiCI, PrCHO, BulCHO, Bu'CHO, CH,[CH,],CO, CH,ICH,I,CO Scheme 26 Dichlorobenzene97 can be sequentially Iithiated, with naphtha- lene as the catalyst, so two different electrophiles can be introduced in the molecule However, only the metu-and para-derivatives give the expected products 98, the corresponding monosubstituted com- pound being the only product isolated starting from o-dichlorobenz- ene (Scheme 27) l9 In this last case the first intermediate, CHEMICAL SOCIETY REVIEWS, 1996 o-chlorophenyllithium decomposes under the reaction conditions used, probably taking a proton from the reaction medium, so after f the second lithiation the final reaction product arises from phenyllithium as the organometallic component 1) L1.Np (3%). -78°C 2) El+,-78OC 3) E2+.-78 to 20°C *6, 6,4) H,O 98 [42-96%) E' + = Bu'CHO, C H2[CH,],CO E2+ = H,O, D,O, Et,CO Scheme 27 In the case of the dichlorinated propene 992336a or isobutene 1012436h it is neccessary to work under Barbier-type reaction condi- tions in order to avoid decomposition of the intermediate dilithiated species Working with DTBB (for 99) or naphthalene (for 101) the expected products 100 and 102 are, respectively, obtained, using in all cases a carbonyl compound as electrophile (Scheme 28) 27 34 w, Diol derivatives 100 and 102 are easily cyclised to the correspond- ing cyclic ethers under acidic reaction conditions 1) Li, DTBB (5%). E+,0°C I 1) Lt, Np (6%),E+,-70°C r=E' PrCH,O, Bu'CHO, PhCHO, Me,CO, Et,CO, PrTOMe, PriCO, CH2[CH2],!C0(n = 3.4,6), PhCOEt Scheme 28 A ca 1 1 mixture of diastereoisomers of 1,3-dichloropropenes 103 is easily lithiated in the presence of a catalytic amount of DTBB and an electrophile to give, after hydrolysis, the corresponding ca 1 1 diastereoisomeric mixture of products 104,which in the case of the Z-diastereoisomer can be cyclised to the corresponding dihy- dropyrans (Scheme 29) 37 103 -,--4-104 [50-72%J E+ = Me,SiCI.Bu'CHO, Me,CO, Et,CO, CH2[CH21,C0, CH,ICH2],C0 Scheme 29 The DTBB-catalysed lithiation under Barbier-type reaction conditions of the three isomers of dichlorobutene 105,106and 107, gives, after reaction with an electrophile and final hydrolysis, the same mixture of 1,2- and 1,4-reaction products (108 and 109, respectively) and the same Z/Emolar ratio for 109, independently of the starting material used This fact would suggest that the inter- mediates derived from 105- 107 are the same (Scheme 30) Finally, I ,4-dichlorobut-2-yne 110 is lithiated under DTBB catalysis and in the presence of an electrophile giving the corre- sponding diaddition products 111 (Scheme 3 1) 39 Concerning probable mechanistic pathways involved in the lithiation of polychlorinated materials under Barbier-type reaction conditions described in this section, two possible routes may be involved (a)a sequence of lithiation reactions with the electrophile and (6) the formation of polylithiated intermediates In general, and owing to the high instability of polylithiated species, we think that the most probable pathway involves the route (a) In all cases the lithiated intermediate resulting from the first lithiation would decompose (by elimination processes or proton abstraction from the reaction media) if the electrophile were not present in the reaction ET E 108 lW E-EI 109I c1-p-Cl 107 E-= Me,SiCI, BuCHO.CH21CH2]$0 Scheme 30 C1 1) LI, DTBB (2 5%), E+,-4OOC 2)H20 110 111(27-7341 E' = Me,SiCI, Bu'CHO.Me,CO, Et,CO, CH2[CH,jllCO (n = 3,4,6) Scheme 31 mixture, so the Barbier-type reaction conditions are essential in order to get successful results 5 Conclusions From the results shown in this review it can be concluded that the arene-catalysed lithiation of different substrates (non-halogenated precursors, functionalised chlorinated materials, saturated hetero- cycles and polychlorinated compounds) is a powerful methodology for preparation of a wide range of very reactive or unstable organolithium intermediates, which are versatile species in syn-thetic organic chemistry A question remains to be answered why is the catalytic version more effective than the stoichiometric one'? A possible explanation could be that in the presence of a deficiency of the arene, the excess of lithium provokes the formation, at least to some extent, of the corresponding arene-dianion (from the ini-tially formed arene-radical anion) This dianionic species IS very powerful as a reduction (lithiation) agent For this reason, most of the reactions described in this review do not work with a lithium-arene mixture (stoichiometric ratio) under the reaction conditions described here, on the other hand, when the process was successful in the stoichiometric version, yields were lower and reac- tion times were significantly longer In addition, combining an arene-catalysed lithiation with Barbier-type reaction conditions (performing the reaction in the presence of the electrophile) the method is very effective in some cases, mainly when poly-chlorinated compounds are used as synthons for polylithium inter- mediates Acknowledgements I would like to thank all my coworkers cited in the references not only for their skilful experimental work but also for their initiative and intellectual contributions DGICYT of Spanish Ministerio de Educacion y Ciencia has generously sup- ported this project (nos PB91-075 1 and PB94- 1514) 6 References 1 B J Wakefield, Organolithium Methods, Academic Press, London, 1988 2 P Cintas, Activated Metals in Organic Synthesis, CRC Press, Boca Raton, 1993 3 N L Holy,Chem Rev, 1974,74,243 4 M Yus and D J Ramon, J Chem Soc Chem Commun , 1991,398 5 C Najera and M Yus, Trends Org Chem , 1991,2, 155 6 (a) H Choi A A Pinkerton and J L Fry, J Chetn Soc Chetn Commun , 1987,225, (6) R Kararnan and J L Fry.Tetrahedron Left. 1989,30,4935 7 A Maerker and U Girreser, Angew Chem Int Ed Engl ,1990,29,667 8 B Mudryk and T Cohen, J Org Chem ,1989,54,5657 ARENE-CATALYSED LITHIATION REACTIONS -MIGUEL YUS 9 (a)R D Rieke,Scrence, 1989,246, 1260, (6)for the most recent paper on this topic, see M V Hanson and R D Rieke, J Am Chem SOC, 1995 117.10775 10 (a) E Alonso, D Guljarro and M Yus, Tetrahedron, 1995,51, 114.57. (6)E Alonso, D J Ramon and M Yus. unpublished results 11 D GuiJarro, B Mancheiio and M Yus, Tetrahedron, 1992,48,4593 12 (a) D Guijarro. B Manchefio and M Yus, Tetrahedron Lett , 1992.33, 5597. (b) D Guijarro, G Guillena, B Manchefio and M Yus, Tetrahedron, 1994.50,3427 (c)D Guijarro and M Yus, Tetrahedron, I995,51,1144S 13 D Guijarro, B Mancheiio and M Yus, Tetrahedron, 1994,50,8551 14 U Azzena S Demartis, M G Fiori, G Melloni and L Pisano.Tetrahedron Lett , I995,36,5641 15 E J Corey and Z Chen, Tetrahedron Lett , 1994,35,8731 16 (a)E Alonso, D Guijarro and M Yus, Terrahedron, 1995,51,2699,(6) D Guijarro and M Yus, Tetrahedron Lett ,1994,35,2%5 17 D Guijarro and M Yus. Tetrahedron, 1994,50,3447 18 (a)D Guijarro, B Manchefioand M Yus. Tetrahedron, 1993,49,1327. (6) D Guijarro and M Yus. Tetrahedron, 1993,49,7761 19 A Guijarro, D J Ramon and M Yus, Tetrahedron, 1993,49,469 20 (a)D Guijarro and M Yus. Tetrahedron Lett. 1993, 34,3487, (6)A Guijarro.B Manchefio, J Ortiz and M Yus, Tetrahedron, 1996, 52, 1643 21 C Najera and M Yus, Org Prep Proced Int , 1995,27,383 22 D J Ramon and M Yus, Tetrahedron Lett ,1993,34,71 15 23 F F Huerta, C Gomez and M Yus, Tetrahedron, 1995,51,3375 24 D J Ramon and M Yus, Tetrahedron, 1993,49,10 103 25 F F Huerta, C Gornez and M Yus, Tetrahedron, in press 26 J Almena, F Foubelo and M Yus, Tetrahedron, 1995 51,11883 27 A Bachki F Foubelo and M Yus. Tetrahedron Lett, 1994, 35, 7643 28 (a)J F (311.D J Ramon and M Yus, Tetrahedron, 1993,49,4923,(6) J F (311,D J Ramon and M Yus, Tetrahedron. 1994, 50, 7307, (c) M Yus and D J Ramon, J Org Chem, 1992, 57, 750, (6) D J Ramon and M Yus, Tetrahedron Lett, 1990, 31, 3763, (e)D J Ramon and M Yus, Tetrahedron Lett ,1990,31,3767,(t) D J Ramon and M Yus, J Org Chem , 1991,56.3825 29 (a) E Bartmann, Angew Chem Int Ed Engf , 1986,25,653.(6) T Cohen.1 H Jeong,B Mudryk.M Bhupathyand M A Awad,J Org Chem ,1990,55, 1528, (c)D Seebach and E Hungerbuhler, in Modern Synthetic Methods, ed R Shefold, Salle & Sauerlander Verlag, Aarau, 1980, (6) A Bachki, F Foubelo and M Yus, Tetrahedron Asvmmetrv, 19956,1907 30 (a)J Almena,F Foubeloand M Yus, Tetrahedron Lett, 1993,34,1649, (6)J Almena, F Foubelo and M Yus, J Org Chem ,1994,59,3210 31 J Almena, F Foubelo and M Yus, Tetrahedron, 1994,50,5775 32 (a)B Mudryk and T C0hen.J Am Chem Soc , 1991,113,866,(6) D J Ramon and M Yus, Tetrahedron, 1992,48,3585 33 (a)J F (311, D J Ramon and M Yus, Tetrahedron, 1993.49,9535,(6) J F Gil. D J Ramon and M Yus, Tetrahedron, 1994,50,7307, (c)J F GII, D J Ramon and M Yus, Tetrahedron, 1994,50,3437 34 (a)J Almena, F Foubelo and M Yus, Tetrahedron, 1995,51,3351,(6) J Almena, F Foubelo and M Yus, Tetrahedron, 1995,51,3365, (c)J Almena, F Foubelo and M Yus, J Org Chem , I996,61, 1859, (4J Almena, F Foubelo and M Yus, Tetrahedron, in press 35 (a) A Guijarro and M Yus, Tetrahedron Lett, 1994,35, 253, (6)A Guijarro and M Yus, Tetrahedron, 1996,52, 1797 36 (a)A Guijarro and M Yus, Tetruhedron Lett , 1993,34,2011,(6)F F Huerta,C Gomez,A Guijarroand M Yus, Tetrahedron, 1995,51,3375 37 A Guijarro and M Yus, Tetrahedron, 1994,50. 13269 38 A Guijarro and M Yus, Tetrahedron, 1994,50,7857 39 A Guijarro and M Yus, Tetrahedron. 1995,51.23I

ISSN:0306-0012    

DOI:10.1039/CS9962500155

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Chiral Discrimination by Modified Cyclodextrins Christopher J. Easton" Research School of Chemistry Australian National University Canberra ACT 0200 Australia Stephen F. Lincoln Department of Chemistry University of Adelaide South Australia 5005 Australia 1 Introduction The naturally occurring a ,p and y cyclodextrins 1-3 are cyclic oligosaccharides. consisting of six seven and eight a 1.4 linked D glucopyrdnose units. respectively Interest in these compounds stems from the fact that they act as host niolecules to foim inclusion complexes with a wide variety of guests (Scheme 1) I The cyclo- dextrins each exist as a single enantiomer with the consequence that when they act as host molecules interaction with a racemic gued may lead to the formation of diastereoisomeric complexes of differ ing thermodynamic stability This chiral discrimination by uninod ified cyclodextrins has been intensively studied and extensively exploited most notably through the work of Armstrong et ul ? in the development of cyclodextrin based chromatographic systems The extent of chiral discrimination displayed by the naturally occurring cyclodextrins is typically quite modest however with efficient rewlution of racemates only resulting from repeated inter actions with a cyclodextrin as is the case with cyclodextrin based 1 2 Cyclodextnn Host Guest inclusion Complex Scheme 1 Inclusion complex association constant K = Iinclusion complex I/( Icyclodextrin host 1 [guestI ) chromatography The low enantioselectivity may be attributed to the inherent symmetry of the cyclodextrins with each having an axis of symmetry In addition. inclusion complex formation often occurs principally as a result of interaction of the hydrophobic annulus of the cyclodextrin with an achiral hydrophobic portion of OH HO Ho+H 'OH HO /O 1 x=a 2 x=P 3 x=Y HO' 3 A truncated cone is often used to represent the torus of a cyclodextrin A substituent drawn at the narrow end of the cone indicates that it replaces one of the C 6 hydroxy groups in the cyclodextrin while a substituent drawn at the wide end of the cone indicates that it replaces either a C 2 or a C 3 hydroxy group Chris Easton was horn in I955 and raised in the McLuren Vale wine-producing urea oj South AuJtraliu He is ugraduate of the Flinders Universitv of South Australia and oj the Universitj of Adelaide where he conipleted his PhD under the supervision of Athel Beckwith Afrer post-doctoral studiec at Harvard Utuverut) with Jeremy Knowles he held appointments at the Australian National University the University of Canterburr in Yew Zealund arid the University of Adelaide before returning to the Australian National UniLersity arid his current position His re yearch intere,tJ include amino acid and peptide chem isto and biochemistry and molecular recognition in host-guest c omplexes Stephen Lincoln is heud of the department of chemistry at the Univeryrtv of Adelaide where he holds a personal chair He was horn in SufSolk and graduated BSc (Tech) withjrst class honours jrom UMIST in 1962 He then decided to see the world and graduuted with a PhDfiom the University of Adelaide in 1966 He went on to a post-doctorul position at Wushington State University returning to Adelaide us a lec turer in 1968 He wus uwurded a DSc by the Univerrity of Manchester in I984 His reseurch interests include inor ganic and bininorganic mecha nismJ and molecular recognition cherni stry 164 CHEMICAL SOCIETY REVIEWS 1996 a guest and there is little interaction between the chiral centres of aggregate formation rather than inclusion complex formation It the cyclodextrin and those of the guest It follows that increased has been noted previously that little direct correlation exists chiral discrimination can be expected with modified cyclodextrins between the retention times of molecules on cyclodextrin-based where through the modification the degree of asymmetry of the chromatography columns and the thermodynamic stability of the cyclodextrin has been increased and there is the possibility of inclusion complexes formed in solution between those molecules greater interaction between chiral portions of the cyclodextrins and and cyclodextrins Spectroscopic discrimination does not neces- those of the guests sarily correlate with thermodynamic discrimination so examples of Modifying cyclodextrins and their complexing characteristics the former are only discussed where they have been used to measure usually involves substitution of one or more of the C-2 C-3 and C-the thermodynamics of inclusion complex formation Since our aim 6 hydroxy groups The modifications may be divided into two cat- is to compare the chiral discrimination displayed by the natural and egories In one the hydroxy substituents are substituted in a modified cyclodextrins we have only included details of symmetric fashion to give a single modified cyclodextrin (e g all enantioselectivity shown by natural cyclodextrins where compar- the hydroxy groups may be substituted) or at random to give a ative data with cyclodextrin derivatives are available complex mixture of cyclodextrins in which the average effect is that The values for cyclodextrin-guest association constants given of a symmetric substitution As we will show this tends not to alter herein are quoted directly from the primary literature It should be the symmetry of the cyclodextrin or the enantioselectivity that it noted that these data arise from work in various laboratories with displays With the other type of modified cyclodextrin either a the result that a range of experimental conditions has been used For single substituent or a specific combination of substituents ISintro-this reason key experimental parameters are indicated to show the duced This may induce substantial changes in the asymmetry of the limits to which results from various studies are directly comparable cyclodextrin and result in additional and more specific interactions Nevertheless there is remarkable consistency between the various between the chiral area of the guest and the asymmetry of the host experiments with most studies being carried out in aqueous solu- which restrict the geometry of binding leading to greater tion at or near 298 K Most importantly identical conditions pre- enantioselectivity The additional interactions between the cyclo- vailed in all cases where comparisons are made between dextrin substituent and the host may be subdivided into secondary diastereoisomeric pairs of host-guest complexes bonding interactions metal complexation and covalent attachment Again we will show that as the extent of the interaction between the cyclodextrin substituent and the guest increases the magnitude of 2 Effect of Additional Secondary Bonding chiral discrimination often becomes greater Interactions In choosing examples to illustrate this review we have restricted As mentioned above symmetrically substituted cyclodextrins tend our selection to those for which thermodynamic and/or kinetic para- to show no greater chiral discrimination than the naturally occurring meters of the homogeneous solution-phase interaction between the analogues This holds even where the modification results in more cyclodextrin and each enantiomer of the guest have been reported favourable interactions between the racemic guest and cyclodextrin We have not included results from heterogeneous systems on the host as reflected in much higher association constants for the basis that they may depend on factors such as phase solubility and diastereoisomeric inclusion complexes For example as shown in other medium and surface effects and gues? or cyclodextrin Table 1 (entries 1 -12) the association constants of the inclusion Table 1 Association constants of cyclodextrin inclusion complexes Entry Cyclodextrin Guest KRldm3 mol I K,ldm3 rnol I K,IK; Ref 1 1 4 + H' 77203 82203 0 94 4 2 7 4 + H' 54 2 3 59 T 4 0 92 5 3 1 4-H+ 215204 225204 0 96 4 4 7 4 -H+ 49 % 3 55 2 3 0 89 5 5 1 5 144201 146k01 0 99 4 6 7 5 451 % 7 434 2 7 104 5 7 1 5 -H+ 131205 141 %05 0 93 4 8 7 5 -H+ 80 % 3 77 2 3 1 04 5 9 1 6 83203 83203 100 4 10 7 6 142 % 6 155 2 6 0 92 5 I1 1 6 -H+ I24 2 03 106 2 04 117 4 12 7 6 -H+ 143 2 6 153 2 6 0 93 5 13 1 10 27 2 3 1724 1 59 7 14 2 10 1090 2 30 1010 2 40 1 08 6 15 7 10 220% 10 207 2 8 1 06 7 16 8 10 129 2 5 1702 10 0 76 7 17 2 10 -H+ 63 2 8 52 2 5 121 6 18 9 10 -H+ 36 -t 6 13 2 7 2 77 6 19 13 16 14 7 10 8 136 II 20 14d 16 540276 425273 1 27 10,lI 21 15' 16 455 2 82 345257 132 10 22 18 17 295 2 3 6292 10 0 47 13 23 19 17 160 2 36 83 2 28 I 93 12,13 24 20 17 I39 2 24 231 245 0 60 12,13 These ratios substantiate the trends referred to in the text but it should be noted that standard deviations in the association constants of the diastereoisomeric pairs of inclusion complexes limit the reliability of the data Although a range of conditions has been used in mea-suring the association constants cited herein in the text comparisons are only made of data recorded under similar conditions Experimental conditions were as follows refs 4and 5 solvent 10% aqueous D,O I = 0 10 rnol dm-3 T= 295 5 K refs 6 and 7 solvent H,O 1 = 0 10 mol dmP3 T = 298 2 K refs 10and 1 I solvent H,O N%B,O (I5 4 X lop3 rnol dmP3) and H3B0 (34 6 X lop3 rnol drnp3) T = 298 K refs 12 and 13 solvent H,O 0 066 rnol dm-3 phosphate T = 298 K Compound 13 is a mixture of the 6A,6B-isomers in which the primary hydroxy groups of two adjacent glucose residues of the cyclodextrin have been substituted The association constants are the same for each isomer within experimental error lo I1 Compounds 14 and 15 are 6A,6B-isomers in which the primary hydroxy groups of two adjacent glucose residues of the cyclodextrin have been substituted The structures of the 6A,6B-isomers 14 and 15 may be the reverse CHIRAL DISCRIMINATION BY MODIFIED CYCLODEXTRINS-C J EASTON AND S F LINCOLN 4 5 7 X=a,m=12,n=6 9 8 ~=P,m=14,n=7 (3HC02H Me 10 complexes of the variously protonated and deprotonated fluorinated amino acid derivatives 4-6 with termethylated a-cyclodextrin 7 are substantially greater than those formed with the parent a-cyclodextrin 1 yet the enantioselectivity shown by the modified cyclodextrin 7 is little different from that displayed by a-cyclodex- trin 1 Similarly the extent of chiral discrimination displayed by the termethylated cyclodextrins 7 and 8 in the formation of inclu- sion complexes with the (R)-and (S)-enantiomers of 2-phenyl- propanoic acid 10 is not much different from that exhibited by the natural cyclodextrin analogues 1 and 2 (Table 1 entries 13-16) It is worth noting that the methyl substituents of the modified cyclo- dextrins 7 and 8 increase their flexibility as hosts This flexibility allows conformational change to occur more easily to accommo-date a guest and increase complex stability but it is unlikely to favour chiral discrimination Conversely lack of flexibility of the host and specific host-guest interactions should lead to increased enantioselectivity but this is likely to correlate with the formation of less stable complexes The association constants of the com- plexes of the enantiomers of the anion of 2-phenylpropanoic acid 18 with P-cyclodextrin 2 and the corresponding amine 9 (Table I entries 17 and 1Q6 provide a pertinent illustration The enantioselectivity displayed by the modified cyclodextrin 9 is sig- nificantly greater but the association constants are lower indicating a specific and unfavourable effect of the amino substituent of the host 9 on complexation of the propanoate guest The enhanced stereoselectivity displayed by the amino-substituted cyclodextrin 9 in the formation of inclusion complexes is reflected by an increase in asymmetric induction in reactions of included guests While the sodium borohydride reduction of benzoylfomic acid 11 in the presence of P-cyclodextrin 2 gave the (R)-enantiomer of the alcohol 12 in 4%)enantiomeric excess a 13% excess was obtained when the reaction was performed in the presence of the amino-substituted cyclodextrin 9 The effect of the modified cyclodextrin 9 was attrib- uted to electrostatic interaction between the amino substituent of the cyclodextrin 9 and the carboxy moiety of benzoylformic acid 11 (=&HCO.H OH 11 12 With an increase in the number of interactions between the guest and substituents introduced on to the modified host greater chiral discrimination by the host could be expected Tabushi et a1 ,lo I' synthesised the modified cyclodextrins 13-15 having both posi- tively and negatively charged substituents and investigated their behaviour as chiral artificial receptors for tryptophan 16 (Fig 1) Each of the modified cyclodextrins 13-15 displayed a modest degree of enantioselectivity (Table 1. entries 19-21) The stability 13 14 H 16 15 Figure 1 Schematic representation of the complexation of tryptophan 16 by the modified cyclodextrin 14 constants of the complexes were found to be larger in the cases of the cyclodextrins 14 and 15 than those observed with the analogue 13 and this was attributed to greater polar interactions between the guest and host when the host substituents were in a relatively non- polar environment The greater polar interactions were not reflected in enhanced chiral discrimination however as the enantioselectiv- ity displayed by the cyclodextrins 13-15 was quite similar An alternative facet of enantioselective guest complexation by a modified cyclodextrin was reported by Takahashi et (I/ I* l3 Amino acid-substituted cyclodextrins formed diastereoisomeric complexes with the N-dansylphenylalanine anion 17 in the case of the tyrosine CHEMICAL SOClETY REVIEWS 1996 Q p2 H 17 NHCO-C--H ’NHCHO 19 derivative 18 their association constants differed by a factor of 2 13 (Table 1 entry 22) In this case where the substituent of the mod- ified cyclodextrin is chiral the cyclodextrin annulus probably serves mainly to bind the guest and contributes little towards the enantioselectivity Instead stereoselectivity probably results from interactions between the chiral substituent of the cyclodextrin and chiral portions of the guest Support for this interpretation comes from the observation that the enantioselectivity displayed bq the modified cyclodextrin diartereorsonier5 19 and 20 in complexing the N-dansylphenylalanme anion 17 is similar in magnitude though reversed in terms of absolute stereochemistry (Table 1 entries 23 and 24) l2 I3 3 Metallocyclodextrins The examples given above show that secondary bonding interac- tions between included guests and substituents of modified cyclo- dextrins can lead to greater stereoselectivity in the formation of inclusion complexes Nevertheless the association constants of the diastereoisomeric inclusion complexes differ by no greater than a factor of three and generally by much less Through metal complexation which further increases the extent of interaction between the cqclodextrin and the guest the diastereoselectivity can be further improved This involves the coordination of both the cyclodextrin substituent and the guest to a metal in the host-guest complex as a result of which the binding geometry can be quite restricted The tenfold chiral discrimination displayed by the nickel(I1) complex 22 (M = Ni) of 6A-(3-aminopropylamino)-6A-deoxy-P-cyclodextrin 21 in the formation of inclusion complexes with the 21 OH 18 I CH2 ‘NHCHO 20 enantiomers of the anion of tryptophan 16 (Table 2 entry 3) is the largest reported for a metallocyclodextrin l4 I5 Comparison of the association constants of the inclusion complexes of the metallocy- clodextrin 22 (M = Ni) with those of the complexes formed in the absence of a metal and with the parent P-cyclodextrin 2 (Table 2 entries 1-3) provides an insight into the origin of this enantioselectivity There is no chiral discrimination in the formation of the diastereoisomeric inclusion complexes of the enantiomers of the anion of tryptophan 16 with P-cyclodextrin 2 or with the amino- propylamino-substituted cyclodextrin 21 although the thermo- dynamic stability of the complexes is greater with the modified cyclodextrin 21 The thermodynamic stability of the ternary complex of each enantiomer of the anion of tryptophan 16 with the metallocyclodextrin 22 (M = Ni) is even greater showing the pres- ence of even more favourable interactions By comparison with the complexation constant for the interaction between the anion of tryp- tophan 16 and nickel(I1) (Table 3 entry 2) the ternary complexes are less stable however indicating that the cyclodextrin annulus disrupts coordination of the anion of tryptophan 16 to nickel(1r) The extent of these unfavourable interactions appears to depend on the chirality of the anion of tryptophan 16 thus affecting the enan tiosel ec tivity The adverse effect of the cyclodextrin on the thermodynamic stability of the ternary complex is also apparent though less marked in the interaction of the anion of tryptophan 16 with the cobalt(1r) and copper(r1) complexes 22 (M = Co) and 22 (M = Cu) of the aminopropylamino-substituted cyclodextrin 21 (Table 2 entries 4 and 5,Table 3 entries 4 and 6) I5 These metallocyclodex- trins also display enantioselectivity but to a lesser extent than that displayed by the nickel(1r) complex 22 (M = Ni) By contrast the 22 23 CHIRAL DISCRIMINATION BY MODIFIED CYCLODEXTRINS-C. J. EASTON AND S. F. LINCOLN 167 Table 2 Association constants of metallocyclodextrin inclusion complexesa Entry Cyclodextrin Guest log (KR/dm3 mol-I) log (Ksldm3 mol-I) KdKS Ref. I 2 16 -H+ 2.33 2 0.06 2.33 t0.08 I .oo 14,15 2 21 16 -H+ 3.41 2 0.05 3.40 -+ 0.07 1.oo 14,15 3 22 16 -H+ 4.1 +-0.2 5.1 2 0.2 0.10 14.15 (M = Ni) 4 22 16 -H' 4.04 -t 0.03 4.32 +-0.05 0.53 15 (M = Co) 5 22 16 -H+ 7.85 ? 0.07 8.09 +-0.05 0.58 15 (M = Cu) 6 22 16 -H+ 5.3 t0.1 5.3 -+ 0.1 1.oo 15 (M = Zn) 7 22 23 -H' < 3.6 4.4 t0.1 <0.16 16 (M = Ni) 8 22 23 -H' 3.6 2 0.2 3.69 2 0.06 0.8 1 16 (M = Co) 9 22 23 -H' 7.2 -+ 0.1 6.9 2 0.1 2.oo 16 (M = Cu) 10 22 23 -H+ 4.7 2 0.1 4.7 ? 0.1 I .oo 16 (M = Zn) In H,O / = 0.10 mol dm-7 T= 298.2 K. which the cyclodextrin disrupts the binding of the guest to the Table 3 Metal complexation constantsa metal. The discrimination displayed by the nickel(I1) and cobalt(l1) Entry Metal Ligand log (K/dm3 mol -I ) Ref. metallocyclodextrins 22 (M = Ni) and 22 (M = Co) favours 1 Ni2+ 21 5.2 2 0.1 14,15 binding of the (S)-enantiomers of the anions of tryptophan 16 and 2 Ni2+ 16 -H+ 5.42 -t 0.03 14,15 phenylalanine 23. The discrimination of the copper(i1) metallo- 3 co2+ 21 4.22 +-0.02 15 cyclodextrin22 (M = Cu) favours binding of the (S)-enantiomer of 4 co2+ 16 -H+ 4.41 +-0.05 15 the anion of tryptophan 16 and the (R)-enantiomer of the anion of 5 cu*+ 21 7.35 + 0.04 15 phenylalanine 23. 6 cu2+ 16 -H+ 8.11 2 0.03 15 While the work carried out to date with the metal complexes of +7 Zn2 21 4.96 2 0.08 15 the aminopropylamino-substituted cyclodextrin21 has been mostly 8 Zn2+ 16 -H+ 4.90 2 0.04 15 limited to studies with the anions of tryptophan 16 and phenylala- 9 Ni2+ 23 -H+ 5.09 + 0.05 16 nine 23 as guests a more extensive range of amino acids has been 10 co2+ 23 -H+ 4.19 ? 0.03 16 11 cu2+ 23 -H+ 7.8 +-0.1 16 used to investigate chiral discrimination by the copper(r1) com- 12 Zn2+ 23 -H+ 4.59 -+ 0.04 16 plexed histamine-monofunctionalised P-cyclodextrin 24 (Table 4).17,'8 In this case the metallocyclodextrin 24 displayedIn H20 / = 0 10 rnol dmP3 T= 298.2 K. enantioselectivityin the complexation of the anions of the aromatic amino acids tryptophan 16 phenylalanine23 and tyrosine 25 with the stability constant of the complex of the (R)-enantiomer being the diastereoisomeric ternary complexes 22 (M = Zn) of the anion of tryptophan 16,zinc(I1) and the modified cyclodextrin 21 are thermo- Table 4 Association constants of copper(1r) ternary complexes of dynamically indistinguishable (Table 2 entry 6) but more stable the cyclodextrin 24 with amino acid anionsa than the binary complexes of zinc(r1) with the modified cyclodextrin Entry Amino log (KRI log (Ks I KRIK Ref.21 and of the anion of tryptophan 21 with the metal ion alone (Table 3 entries 7 and 8). It seems that enantioselectivity only results from acid dm3 mol-I) dm3 mol-I) unfavourable interactions in the ternary complexes which restrict 1 16 -H+ 16.47 -+ 0.02 16.12 IT 0.01 2.23 17,18 the geometry of binding. 2 23 -Hf 15.85 2 0.01 15.68 2 0.02 1.48 18 Analogous effects were observed in the formation of ternary 3 25 -Hf 15.22 2 0.01 14.82 2 0.01 2.51 18 26 -Hf 15.51 2 0.02 15.53 2 0.04 O.% 17,18complexes of the metallocyclodextrins 22 with the anion of phenyl- 4 27 -Hf 14.87 -+ 0.05 14.80 2 0.02 1.17 18alanine 23 (Table 2 entries 7-10; Table 3 entries 9-12).16The 6 5 28 -H+ 14.96 -+ 0.02 14.89 ? 0.02 1.18 18enantioselectivity was greatest with the nickel(i1) metallocyclodex- In H,O I = 0.10 mol dm-3 T = 298 K.trin 22 (M = Ni) decreasing in the order nickel(l1) > copper(r1) = cobalt(r1) > zinc(1r). Again this order correlates with the extent to HoeCH2 -C;H3* / c02-24 25 c02-Me c02-Me\ co2-Me-CH &'CH-<H CH-CH~-<H NH3' NH3+ Me/ N H3+ 26 27 28 larger in each case By comparison the diastereoisomeric pairs of ternary complexes of the anions of the aliphatic amino acids 26-28 showed only small differences in thermodynamic stability In this work calorimetric studies were carried out in order to examine the factors contributing to the enantioselectivity The overall complexation process for each of the amino acids was found to be enthalpically and entropically favoured For the complexes of aro- matic amino acids however the enthalpy contribution was found to be more favourable for the (R)-enantiomers while the entropy factor was less favourable This indicates that the geometry of complexation of the (R)-enantiomers is more restricted but the binding interactions in the complexes are stronger and is consistent with a model in which the complexation of the (R)-enantiomers is favoured by the preferential inclusion of their aromatic side chains in the cyclodextrin cavity The histamine-substituted metallocyclodextrin 24 also dis- played spectroscopic and chromatographic chiral discrimination in the complexation of amino acid anions and the extent of chro- matographic discrimination for various amino acids paralleled the thermodynamic enantioselectivity l7 l8 Interestingly the iso- meric metallocyclodextrin 29 showed even greater enantioselectivity when used in chromatography with the anion of tryptophan 1619 but no thermodynamic data for this dis- crimination have been reported The copper(I1)-complexed aminoethylamino-substituted cyclodextrin 30 also displayed chromatographic and spectroscopic discnmination in complexing the anion of tryptophan 16 but there was no thermodynamic enantioselectivity in this case 2o Again this illustrates the lack of correlation between thermodynamic and chromatographic and spectroscopic effects In this regard while the spectroscopic dis- crimination displayed by lanthanide-cyclodextrin complexes2' and the enantiodiscriminating oxygenation of a-pinene using a porphyrin-substituted cyclodextrin22 are interesting examples of exploitation of the enantioselectivity displayed by metallo-cyclodextrins they are difficult to evaluate further in the absence of thermodynamic data 29 30 Although only a limited number of studies of chiral discrimina- tion by metallocyclodextrins have been reported they are sufficient to support the hypothesis stated above that coordination of both the cyclodextrin and the guest to a metal which increases the extent of interaction between the cyclodextrin and the guest will generally increase the enantiodiscrimination It is likely that even greater stereoselectivity can be expected where the substituent attached to the cyclodextrin and coordinating the metal is chiral ,thus increas ing the asymmetry of the complex though this has yet to be tested 4 Covalent Interactions An alternative form of interaction between cyclodextrins and guests which also leads to enhanced enantioselectivity involves the formation of a covalent bond between the host and guest in the inclusion complex The hydrolysis of esters by cyclodextrins has been intensively studied as a model of covalent catalysis by enzymes 23 The process involves the formation of a host-guest complex between a cyclodextrin and an ester then transesterifica- tion between host and guest followed by hydrolysis of the acylated cyclodextrin The interest in cyclodextrins as enzyme mimics stems from the fact that they enhance the rates of reaction of included esters and they show enantioselectivity in the case of chiral deriva- tives 24-32 In principal the chiral discrimination could arise either CHEMICAL SOCIETY REVIEWS 1996 from stereoselectivity in the formation of the host-guest complexes or from different reactivities of the guests in the diastereoisomeric complexes or from a combination of these processes In practice more substantial stereoselectivity has usually arisen from differ- ences in the reactivity of the complexed species 25-30 This is illus trated by the association constants for complexation of the phenylpropionates 31 by a-and P-cyclodextrin 1and 2 and the rate constants for the reactions of the complexed species (Table 5) 26 An overall enantioselectivity of 19 0 was observed for the interaction of the ester 31b with P-cyclodextrin 2 that figure comprising factors of 1 2 for the complexation and 155 for the reactions of the complexed species (3HC02R Me 31 Table 5 Thermodynamic parametersu for interaction of the esters 31 with cyclodextrins26 Cyclodextrin Ester KR/Ksh kcR/kci (kcRKR)/(kcRKS) 1 31a 133 12 16 1 31b 1 07 87 93 2 31a - 95 - 2 31b 122 15 5 19 0 In H,O 0 2 X mol dmP3 sodium carbonate buffer T =298 K Ratio of the association constants for the enantiomers Ratio of the rate constants for the reactions of the complexed species It has been clearly demonstrated that the enantioselectivity dis played by the cyclodextrin depends on the extent to which the geom- etry of binding and transesterification has been restricted Trainor and Breslow28 showed that freezing out residual rotational degrees of freedom in the acylation transition state increased the enantioselectivity shown by the cyclodextrin The enantiomers 33 and 34 correspond to one of the preferred conformers of the ester 32 and P-cyclodextrin 2 was found to accelerate their rates of reaction to extents approximately ten times and one half respectively of that observed with the ester 32(Table 6) The esters 35 and 36 correspond to the enantiomers of the other preferred conformer of the ester 32 and the enantioselectivity observed in their reactions with P-cyclodextrin 2 was much less A further minor modification to the geometry of the cyclodextrin acylation in the reactions of the esters 37 and 38 resulted in a 62-fold enantioselectivity (Table 6) 29 This is the largest reported for hydrolysis of an ester by a cyclodextrin 33 35 CHIRAL DISCRIMINATION BY MODIFIED CYCLODEXTRINS-C ~~ ~ ~ Table 6 Rate accelerations for reactions of esters complexed by P-cyclodextrin 2” Ester k,iku (X Ref 32 36 28 33 16 28 34 320 28 35 10 28 36 66 28 37 590 29 38 95 29 ‘I In 60%Me2S0 / 40% H,O (L/L ) T= 303 K Frequently studies of the interactions of cyclodextrins with esters have concentrated on the formation of the host-guest complexes and the subsequent transesterification and the possibility of diastereo- selective hydrolysis of the acylated cyclodextrins has often not been examined Deacylation of the cyclodextrin 40 was investigated as part of a study of the reaction of the ester 39 with a-cyclodextrin 1 0’ 39 40 OCO-~~CH2CHMe23=:41 CICO-!aCH2CHMe2H 42 The reaction occurred without diastereoselectivity as was the case with formation of the inclusion complexes between the ester 39 and a-cyclodextrin 1 although the rates of reaction of the included enantiomers of the ester 39 to give the acylated cyclodextrin 40 dif-fered by a factor of 7 More recently we reported a tenfold diastereo- selectivity in the hydrolysis of the cyclodextrin 41 31 32 The synthesis of the ester 41 through reaction of the Ibuprofen acid chlo- ride 42 with P-cyclodextrin 2 afforded a 5 1 mixture of the diastereo- isomers in favour of the isomer derived from (/?)-Ibuprofen and that diastereoisomer was also the most readily hydrolysed Consequently the overall stereoselectivity for the two-step reaction of the acid chloride 42 is ca 50 1 The complementary nature of the diastereoselectivity of the synthesis and hydrolysis was attributed to J EASTON AND S F LINCOLN I69 similarities between the reaction transition states The contrast in diastereoselectivity in the reactions of the esters 40 and 41 is not sur- prising given the differences between these systems The acyl sub- stituents of the esters 40 and 41 are bound via secondary and primary hydroxy groups respectively In addition the acyl group of the Ibuprofen derivative 41 is more hydrophobic and is more likely to interact with the cyclodextrin annulus Stereoselectivity has also been observed in a variety of other reactions where the guests become covalently bound to the cyclo- dextrins Enantioselectivity has been found in the acylation of cyclodextrins with 5(4H)-o~azolones,3~ in reactions which are 34 mechanistically quite similar to those of esters interacting with cyclodextrins A variety of Schiff base derivatives of cyclodextrins has been synthesised and studied as models of pyridoxal phosphate- dependent enzymes Breslow et a1 ,3s reported the synthesis of the pyridoxamine derivative 43 and showed that in the reaction of this compound with phenylpyruvic acid 44 phenylalanine 23 was pro- duced as a 5 1 mixture of the (S)-and (R)-enantiomers With the related cyclodextrin derivative 45 Tabushi et a1 ,I‘ reportedj6 much higher stereoselectivity in the reactions of ketoacids pro- ducing the (S)-isomers of phenylalanine 23 tryptophan 16 and phenylglycine 46 each in at least 90% enantiomeric excess 43 44 45 46 Recently we reported high enantioselectivity in the reactions of 2-phenylethylamine 47 with the iodocyclodextrin 48 to give the diastereoisomers of the amine 49 37 In further experiments aimed to elucidate the thermodynamic parameters of those interactions we have now found that the extent of the stereoselectivity IS highly irregular however and is generally much less than was observed originally Currently we are examining the possibility that ternary complexes may be involved in these processes 47 48 Me2CH0 0 Me/ \F 50 49 Me2CH0 0 Me,CHO 0\p4 Me’ ‘oC&-p-N@ Me’ ‘SCH2CH2NMe 51 52 170 With Sarin 50 the compound used recently in terrorist attacks in Japan the reaction with a-cyclodextrin 2 proceeds by inclusion complex formation followed by phosphonylation of the cyclo- dextrin and each of these processes is stereoselective (Table 7) 38 39 The reactions of a-cyclodextrin 1 with the related phosphonate 51 and phosphonothioate 52 are also highly stereoselective (Table 7) 39 40 The high enantiomeric selectivity reported in the cleavage of organophosphates may be attributed to the fact that the reaction takes place directly at the chira! centre further supporting the hypothesis deveioped throughout this review that higher stereo- selectivity will result from a more intimate interaction between the chiral centres of the cyclodextrins and the guests Table 7 Thermodynamic parametersa for interaction of a-cyclodextrin 1 with the organophosphorus compounds 50-5238-40 Guest K,/ K sb kcdh; (kcKK#kc& 50 0 15 35 0 52 51 0 38 2 76 2 29 52 191 > 100 > 191 ‘I In H,O I = 0 10 mol dm-j T = 298 K Ratio of the association constants‘ for the enantiomers Ratio of the rate constant5 for the reactions of the complexed species 5 Conclusion In summary it IS apparent from the work reviewed here that the nat- urally occurring cyclodextrins show only limited enantioselectivity in their interactions with chiral guests because they form inclusion complexes in which there is on14 minimal interaction between chiral centres of the cyclodextrin and chiral substituents of the guests As the extent of interaction between these groups is Increased as a result of modification to the cyclodeutrin the stereo- selectivity is often increased The immediate result of this improved stereoselectivity is that whereas separation of racemic guests using the naturally occurring cyclodextrins requires multiple interactions between the host and guest more efficient practical and larger- scale resolutions should be possible with the modified cyclo-dex trins 6 References 1 M L Bender and M Komiyama Cvclodexrrin Chemistrv Springer Verlag New York 1978 J Szejtli Cvclodenrrins and Their lriclu~ion Complexes Akademiai Kiado Budapest 1982 R J Clarke J H Coates and S F Lincoln. A~L,Carbohvd Chetn Biothern . 1989 46. 205 2 W L Hinze TE Rieh D W Armstrong W DeMond A Alak and T Ward And Chem 1985 57 237 D W Armstrong T Ward. R D Armstrong and T E Beesley. Science 1986,232. I132 3 V Schurig and H P Nowotny. Angeu Chem .lnt Ed Engl 1990.29. 939 4 S E Brown,J H Coates,S F Lincoln,D R CoghlanandC J Easton J Chern SOL.,Furuduv Tram 1991,87.2699 5 S E Brown,C J Easton and S F Lincoln J Chem Res (S). 1995.2 6 S E Brown. J H Coates. P A Duckworth S F Lincoln C J Easton and B L May J Cheni Soc Furadq Trany 1993,89,1035 7 K Hendrickson,C J Easton and S F Lincoln,Ausr J Chem ,1995.48 1125 8 K Hattori. K Takahashi and N Sakai Bull Chern Soc Jpn 1992,65 2690 CHEMICAL SOCIETY REVIEWS 1996 9 K Takahasi and K Hattori. J Incl Phenorn Mol Recog Chem . 1994. 17,l 10 I Tabushi Y Kuroda and r Mizutani. J Am Chem Soc 1986. 108. 4514 11 I Tabushi. Pure Appl Chew 1986.58 1529 12 K Takahashi. Y Ohtsuka and K Hdttori. Chetn Lrrt . 1990.2227 13 K Takahashi. Bull Chern Soc Jpn . 1993.66,550 14 S F Brown J H Coates C J Easton S J van Eyk S F Lincoln B L May. M A Stile C B Whalland and M I Wil1iams.J Chrrn Soc Chetn Cornmuti 1994.47 15 S E Brown J H Coates C J Easton and S F Lincoln. J Chcrn Soc Farudav Trans I994,90,739 16 S I3 Brown C A Haskard C J Easton and S F Lincoln. J Chern Soc Furuduv TrunJ . 1995.91 1013 17 G Impellizzeri G Maccarrone E Rizzarelli. G Vecchio R Corradini and R Marchelli. Angrw Chem lnr Ed Engl . 1991,30.1348 18 R Corradini A Dossena G lmpellizzeri G Maccarrone. R Marchelli E Rizzarelli. G Sartor and G Vecchio. J Am Chrni Soc 1994 116. 10267 19 V Cucinotta F D’Alessaiidro. G Impellizzeri and G Vecchio J Chern Soc ,Chern Cotntnun . 1992 I743 20 R P Boiiomo. V Cucinotta F D’Allesxmdro G lmpellizzeri. G Maccarrone. E Rizzarelli and G Vecchio. J Incl Phenom Mol Recog Chern 1993.15. 167 21 ?’ J Wenzel M S Bogyo and E L Lebeau J Am Cherii SOL. 1994. 116,4858 22 C M Spencer J I-Stoddart and R Zdrzycki J Chetn Soc Pcvkin Trans 2 1987.1323. L Weber I Imiolczyk G Haufe. D Rehorek and H Hennig J Chetn Soc Chern Comtnun 1992.301 23 R L van Etten J F Sebastian. G A Clowes and M L Bender. J Am Chem Soc 1967,89 3242,3253 N Heinrich and F Cramer. J Am Chern Soc 1965.87 1121. M F Czdrniecki and R Breslow. J Arn Chrrn Soc ,1978 100.771. I Tabushi. Arc Chenz Res 1982.15.66 24 F Cramer and W Dietshe. Chetn Ber 1959. 92. 378 R Ueoka Y Mdtsumoto K Hdradd. H Akahoshi Y Ihara and Y Kato. J Am Chern SOL 1992 114,8339. Y Kitdurd and M Bender Bioorg Chetn 1975,4.237 R Breslow. M F Czarnircki J Emert and H Hamaguchi J Arn Chem Soc . 1980 102.762. lJ ronellato. Bid1 Chon Soc Ft . 1988.277 25 Y Ihara E Nahanishi. M Nanpo and J Koga. Bull Chern SOL Jpn . 1986.59 1901 26 R Fornasier. F Reniero P Scrimin and LJ Toiiellato J Chem Soc Perkiri Trati\ 2. 1987 193 27 R Fornasier F Reniero P Scrimin and U Tonelldo J Cherii Soc PerXiti Tram 2 1987 1221 28 G L Trainor and R Breslow. J Atn Cherii Soc . 1981,103 154 29 R Breslow. G Tiainor and A Ueno. J Am Chern Soc . 1983. 105 2739 30 K Flohr. R M Paton and E T Kaiser. J Am Chern Soc 1975. 97. 1209 31 J H Coates C J Easton S J van Eyk B L May P Singh and S F Lincoln J Chern Soc Chern Cotnrnun . 1991.759 32 J H Coates. C J Easton. N L Fryer and S F Lincoln. Chem Lett . 1994. 1153 33 V Ddtfe and J Fastrez J Am Chern Soc 1980. 102,3601 34 V Daffe and J Fastiez J Chetn Soi Perkin Trunr 2. 1983,789 35 R Breslow. M Hammond and M Lauer,J Atn Chern Soc .1980 102. 421 R Breslow and A W Czarnik J Atti Chern Soc . 1987. 105 1390 36 I Tabushi. Y Kuroda. M Ydmada and H Higashimura. J Am Chrm Soc . 1985,107,5545 17 C J Easton. S F Lincoln and D M Schliebs. J Chetn Soc Chrrn Comrnuti . 1995. 1 167 38 C van Hooidoiik and J C A E Breebaart Hansen Hetl 7rai Chinz Pub r Bus. 1970,89,289 39 C van Hooidonk and C C Groo$.Red Trui Chirn Pats BUJ.1970.89. 845 40 C van Hooidonk. Recl TraL Chiin PUIr Bus 1972.91 I 103

ISSN:0306-0012    

DOI:10.1039/CS9962500163

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Through-bond and Through-space Models for Interpreting Chemical Reactivity in Organic Reactions Keith Bowden Department of Chemistry and Biological Chemistry University of Essex Wivenhoe Park Colchester Essex C04 3S0,UK Edward J. Grubbs Department of Chemistry San Diego State University San Diego California 92 182 USA 1 Introduction Polar substituent effects have been the Subject of many investiga tions It has been generally agreed that the major influences can be considered to be localised effects i e the electrostatic field and/or inductive effect5 and the delocalised effect i e the resonance effect The relative importance of the two mechanisms described as the electrostatic field and inductive effects has been disputed Both represent approximations but can be clearly distinguished in designed models The inductive effect relates to through bond transmission by successive polarisation of the bonds between a dipolar oi charged substituent and the reaction site The effect is attenuated by each bond in a ratio described as the transmission coefficient The summation of this effect over all possible paths can be made using either a simple relation or a more complex topolog ical approach The electrostatic field approach attempts to calcu late the Coulombic interaction within the molecule The more sophisticated approach requires a knowledge of the distance between the reaction site dnd the monopole or dipole the dipole moment the angles between the vector relating to the dipole and the distance described above and the effective relative permitti\ ity Westheimer in a retrospective5 has stated that the theory devel oped by Kirkwood and himself4 'has stood the test of time for more than half a century ' Some excellent organic texts in\ idious to iden tify refer almost exclusively to inductive transmission of polar sub rtitutent effects. rather than electrostatic forces Others refer to an inductive transmission acting through space which is by definition not possible Kirkwood and Westheimer-l derived equations (I) and (2) to esti mate the Gibbs free energy required for the removal of an acid proton in the presence of a monopolar or dipolar substituent respectively In these equations r p 6 and E are the distances between the ionisable proton and the monopole or dipole. the dipole moment the angle between the vectors relating to the dipole and Keith Bowden was horn in London England in 1936 he received his BSc and PliD (N B Chapman and J Shorter) from the Llniversit) of H~illIn I959 and 1962 followed hv hi 5 DSc in 1973 Following a post doctoral fellowship (Ross Stewart) at the Universitv of Britidz Colurnhia he joined the faculty ofthe Universit\ of Essex where he ir now professor cfchetniw-v His research interests in phtsic a1 organic chemistrt include rntra t?ioleculur catalvsi and neigh houring group parriapation the transtnission of polar effects stronglv basic systems weak carbon ac ids and ring-chain tautorneri\m Another area of research is medicinal c hemistrb structure-activin relations arid the mode of action ofhiologicallt active compound\ 171 distance r and the effective relative permittivity respectively The effective relative permittivity E is a function of the geometry and relative permittivity of the molecular cavity E and the solvent medium E (E~being the relative permittivity of vacuum) On the other hand the inductive effect involves propagation by successi\ e poldrisation of the bonds by dipolar or charged substituent between the substituent and the reaction site and will diminish with chain length A quantitative ticatment was suggested by Branch and Calvin" and developed further by McGowan A simple rela tion is that shown in equation (3) below where 4 and n are the transmission coefficient for inductive effects and the number of intervening bonds respectively A more sophisticated relation has been suggested by bxner and Fiedler,3 in which the substituent effect on the n th atom of a straight chdin molecule is expressed by the value QJThe value n for any non terminal atom is related to the values of those of the neighbouring atoms by being the arith metic mean This tredtment can be applied to any atom in a branched or cyclic system summing over all neighbouring atoms as in equation (4) log (KIK,,) = Z,Z,e'l2 3 RTr (47~5E,) (1) log (KIK,,)= Z,epcostU2 3 KTr '(47i-5)~~) (2) log (KIK,))x 2dY (3) tz=' En (4) These are only simplified relations and must be applied to real and complex molecular systems Thus there is no implicit role for solvent dependence in the inductive effect treatments described above Furthermore this treatment cannot take account of the angular dependence of substituent effects However the calculation of the effective relative permittivities for the electrostatic field treat ment presents problems Molecular cavity shapes that are required Edward Gruhhs was horn in Los Angeles Cal$orniu in I934 He received his BA at Occidental College in 1956 He earned the PhD at MIT in I959 working under the direction of Herbert House He worked with David Curtin (University of Illinois) as a postdoctoral associate jrom I960 to 196 I He joined the facultv at San Diego State Universitv in September I96 I and is currently professor of chemistri His research interests have included carbocation rearrangement thermal de compositions and rearrangement via radical intermediates geo metm isomerizution5 of imine derivatives modelr oj transmis sion of dipolar rub5tituent efft.c ts electtmtutic effects in elimination reaction) and intramolec ular based induced elimination\ 172 to be assumed are often both unreal and unrealistic Further 'macroscopic,' rather than 'microscopic' or molecular values for relative permittivities are used these being the only available values 2 Examination of the Theories using Model Systems There are three relatively simple model systems that test the reality of the inductive and/or electrostatic field effect theories and require relatively rigid molecules with known and fixed stereochemistry to enable calculations to be performed The first type of model systems has employed studies that contrast between the number of links available for an inductive transmission and the distance r and E for an electrostatic field effect transmission An ideal system would be one in which it was possible to hold one of these constant while changing the other The second type of model systems has employed studies of the dependence of the transmission on solvent effects For an electrostatic field effect model the dependence on the relative permittivity of the solvent can be predicted both in terms of its type and relative magnitude The third type of model systems has employed studies of the angular dependence of substituent effects upon the angular orientations of dipoles The extreme of this model would be that originally suggested by Roberts and Carboni Molecules could be designed in which 'normal' acid-strengthening and electron-withdrawing substituents could generate a dipole which would oppose the ionisation rf an electrostatic field effect were operating They suggested the design of a compound to reverse a dipolar substituent effect and suggested the molecules 1 and 2 as examples The inductive effect would be expected to be identical in 1 and 2 whereas the electrostatic field effect would be expected to be the reverse in 2 of that in 1 Thus such model systems could test the prediction from the electrostatic field theory that a dipolar substituent effect can be modified by its geometrical orientation to be either 'normal,' diminished absent or even 'reversed ' COOH 1 2 Grob et a1 lo have made a study of the effect of positive nitrogen polar substituents on the ionisation of aliphatic carboxylic acids as in 3,4 and 5 + Me3N(CH2),C02H Me2 Me 4 5 The dpK values of the systems corrected for steric effects esti- mated from uncharged model systems are shown in Table 1 As the links are increased from one to two to three a very modest increase in the ApK value is observed This is quite impossible to relate to an inductive effect model whereas an electrostatic field effect accounts for the results realistically The reaction constants p have been obtained' I for the ionisation of substituted acetic 6 tram-4-substituted cyclohexane- I -carb-oxylic 7,4-substituted bicycle[ 2 2 2)octane- 1-carboxylic 8 and 4-substituted bicyclol2 2 1 Iheptane-1-carboxylic acids 9 The p values for the substituted acetic acids are insensitive to the medium,'* unlike the behaviour of the 4-substituted bicycle[ 2 2 2 ]octane- 1-carboxylic and metu-/para-substituted ben- zoic acids as would be expected from the electrostatic field theory The model calculations for the transmission factor p,/pphave been made for 7,8 and 9 The results for the inductive effect using 4 = 0 36 and for the electrostatic field effect using the chloro sub- CHEMICAL SOCIETY REVIEWS 1996 Table 1 ApK values for the systems 3,4 and 5 in 50% aqueous ethanol at 25 'CIO System ApK; 3,n-I 3 44 n=2 1 96 n=3 1 17 4,2 3 75 3 2 08 4 1 57 5,2 3 89 3 2 26 4 2 03 Values corrected for steric effects X flco2HXCH2C02H X-(3-C02H X-@202H 6 7 8 9 stituted acids are shown in Table 2 The inductive models do not show general agreement between the experimental and calculated values of p,/pp whereas the Kirkwood-Westheimer electrostatic field effect calculations result in good agreement Table 2 Transmission factors p,/p,O,for the systems 7,s and 9" Calcd Exptl Electrostatic fieldX 50% aq lnductive3 System Water ethanol [eqn (3)l Water 50% aq ethanol 7 0 18 024 0O9,(02li) 0 16 028 8 021 034 0 14(036)" 0 19 032 9 0 25 (0 32,)h 0 22,(0 41 ,)n 0 24 (0 40)h 'IInductive [eqn (4)l SO% aq methanol Hammett upvalues derived from the ionisation of para-substi- tuted benzoic acids 10 can be used successfully in the correlation of tram-and cis-3-siibstituted acrylic 11 and 12 and ortho-substi- tuted benzoic acids 13 l2 For the latter two systems 12 and 13,com-pounds having 'bulky' ortlzo substituents and the parent unsubstituted compounds must be excluded from the correlation x 10 11 12 13 Table 3 Hammett reaction constants p for the ionisation of the systems 11,12 and 13 in various solvents at 25 "CI2 System Water 50% aq ethanol 80%aq MCS" 10 1 00 1 47 166 13 2 43 2 21 2 40 12 2 25 -2 24 11 2 45 -3 62 MCS = 2 methoxyethanol As shown in Table 3 the p values obtained for the ortho-substituted benzoic and cis-3-substituted acrylic acids are almost independent of the medium and comparable to each other whereas those for the para-substituted benzoic and trctns-3-substituted acrylic acids are strongly dependent on the nature of the medium This is the result expected from an electrostatic field effect theory For the systems 13 and 12 the solvent transmitted effect is comparable from both 'ends' THROUGH-BOND AND THROUGH-SPACE MODELS-K BOWDEN AND E J GRUBBS of the dipole unlike the situation in systems 10 and 11 in which the distant 'end' of the dipole is less effective in transmitting the effect via the high relative permittivity medium Thus the transmission in water is ca 2 4 times as effective from the ortho- as from the para-position in benzoic acids as shown in Table 4 Hammett reaction constants p for the ionisation of systems 14 in water at 25 "CI3 P P Z metulpara ortho None 100 2 43 ICH I 0 24 0 26 OCHl 0 34 0 48 trans CH=CH 0 42 0 56 Table 3 In Table 4 are shown the p values for a series of meta-lpara- and of ortho-substituted systems in which the phenyl and car- boxylic acid group are separated by an intervening link 14 i e [CH,] OCH tram-CH=CH l7 If an inductive effect is para- mount the effectiveness of ortho-relative to para-substituent effects would be expected to remain at ca 2,4 whereas for an electrostatic field effect the ratio would be expected to tend to unity as 'distance' between substituent and reaction site increases The results are clearly in accord with the latter X 14 A study has been made by Beugelmans-Verrier et a1 l4 of the ion- isation of a series of 3-arylnorbornene-2-carboxylicacids 15 16 and 17 These results reveal the effects of para-substituents in the phenyl groups on the ionisation of the endo trans-15,exo trans-16 and endo cis-17 carboxylic acids The rigid framework allows the effects to be studied with a precise knowledge of the &C02H QC02H X X 15 16 17 endo trans exo trans endo cis Kirkwood-Westheimer r and 8 values and permits a useful comparison with similarly substituted 3-phenylpropionic acids The Hammett p values for these systems are shown in Table 5 The rel- ative transmission coefficients plp have been calculated based on an inductive transmission leqn (3)j and an electrostatic field trans- mission ([eqn (2) for para-chloro systems] as shown in Table 5 The results are in better agreement with the electrostatic field effect than with an inductive effect theory Table 5 Hammett reaction constants p for the ionisation of 3- arylnorbornene-2-carboxylicacids in 50% (vlv) aqueous ethanol at 25 "Ci4 PlP! Calcd System P Exptl Inductive Field endo truns 15 0 44 10 10 10 em trans 16 0 49 11 10 10 endo CLS17 0 68 1 55 10 1 35 3-Phenyl propionic 0 37 0 84 0 94 0 8,p anit Conformer The first unequivocal example of a reversed dipolar substituent effect was considered by Golden and Stocki5 to be found in their study of a 9,lO-bridged anthracene-1 -carboxylic acid system Acevedo and BowdenI6 more recently studied a closely related series of 5-substituted triptycene-7-carboxylic acids and their esters 18 Their reactivities are illustrated in Table 6 The occurrence of Table 6 The reactivity of 5-substituted triptycene-7-carboxylic acids and their corresponding methyl esters 1816 41kal I ne hydro1 y sis Ionisation Esterification with of methyl esters 80%aq MCS" PK k,ll mol DDM I min I. 70% aq dioxane X,/1 mol I min I. Substituent at 25 "C MCSUat 30 "C at 304 "C H 6 88 0 627 0 266 CI 7 04 0 983 0 181 CN 7 03 0 188 0 387 OMe 7 16 0 583 0 204 See Table 3 18 reversed dipolar substituent effect is clear in the ionisation reaction but the results for the esterification with diazodiphenylmethane (DDM) and the alkaline hydrolyses are more complex Grubbs and his coworkersi7 have studied 9,10-bridged 1-and 2- anthroic acids with chloro substituents in the bridges The ApKa values of the isomeric dichloroethano- and chloroetheno-bridged 1-and 2-anthroic acids 19,20 and 21 are shown in Table 7 While the Table 7 ApK values for the 9,lO-bridged 1-and 2-anthroic acids in 50% aq ethanol at 25 "Ci7 JPK1 Compound Exptl Calc 19b 0 58 19c 0 18 20b 0 47 0 36 20c 0 02 0 01 21b 0 22 0 29 21c 0 04 0 13 Kirkwood-Westheimer theory with Tanford modihcation inductive effect does not predict any difference between 19b and 19c or 20b or 20c the agreement between experimental and calcu- lated ApK values according to equation (2) is very good The angular orientations of the dipoles in 20c and 21c are responsible for the 'vanished' dipolar substituent effects (2 2]Paracyclophane is a very good template for the evaluation of the angular dependence of dipolar substituent effects With the reac- tive group in one ring and the substituent in the other the four isomers represented by 22 can be studied As shown in Table 8 the ApK values of isomeric pseudo substituted bromo-4-carboxyl2 2jparacyclophanes are given An inductive effect theory would predict a value of ca -0 1 for ApK of all isomers The electrostatic field effect calculations shown in Table 8 are in good agreement with the experimental results which have been obtained for a wide range of pseudo substituted 4-carboxyl2 2jparacyclo- phanes more recently 19 Roberts and Carboni' had suggested that an ideal model for the observation of reversed dipolar substituent effects would be the 1-(8-substituted 1-naphthyl)propynoic acids 23 which were 19a X=H,Y=H 20a X=H,Y=H 21a X=H,Y H b X=H,Y=CI b X=H,Y=CI b X=CI,Y=H c X=CI,Y=H c X=CI,Y=H c X=H,Y-CI -I-Br 22 Table 8 The ApK values of the bromo-4- carboxy12 2 lparacyclophanes 22 in 80% aqueous MCS” at 25 “C APK Substituent Expt Calcd pseudo gem 0 40 0 73 pseudo ortho 0 16 0 06 pseudo metu -031 -033 pseudo para -035 -031 See Table 3 Kirkwood-Westhermer spherical point dipole theorj subsequently prepared and studied by Bowden and Hojatti 2o The ApK values for the 8-bromo or chloro acid 23 of ca 0 4 pK unit in 80%aqueous 2-methoxyethanol cdn be compared with a value of ca -0 5 pKa unit for the 3-halo substituted propynoic acids 24 in water 21 The reversal of the dipolar substituent effect is almost com- plete The results agree with the electrostatic field effect whereas an inductive effect would be estimated to give an effect of ca -0 2 pK unit X-CZC-CO2H 24 CO;IH23 IQ 25 I Br 26 A ‘simple’ system is the 8-substituted 1-phenylnaphthalene system in which the 1-phenylnaphthalene will be ‘locked’ in a conformation with the two aryl groups almost orthogonal The 4-(8-bromo-1-naphthyl)benzoic acid 25 can be considered to be a ‘reversed’ model of 4-bromobenzoic acidZ2 26 The ApKa values in 80% aqueous MCS for 25 and 26 are 0 49 and -0 53 pKa unit respectively This is an almost complete reversal of the dipolar substituent effect and is in full accord with the electrostatic field effect 3 Other Reactive Consequences of Electrostatic Field Effects Acidity measurements and other reactive properties of carboxylic acids and their derivatives have traditionally served as the princ~pal probes of the transmission of substituent effects However there is a growing array of reaction phenomena which is revealing the importance of through-space charge-charge and charge-dipole CHEMICAL SOCIETY REVIEWS 1996 C02H C02H 278 X=Z=CI W=Y=H b X=Z=H W=Y=CI c X=Y=H W=Z=CI d X=Y=CI W=Z=H effects in a variety of chemical systems Thus Grubbs and cowork- ers*? have studied the regioselectivities in the base-induced eliminations of HCI from the four isomeric acids 27a-d The selectivities were quantitatib ely accounted for in terms of differ ences in charge-charge repulsions dominated by interactions between the carboxylate anion and the partially charged alkoxide ion In the competing transition states Similarly the high degree of regioselectivity in the rut-butoxide-induced dehydrochlorinations of isomers 28a and 28b (29/30 from 28a = 98/2 29/30 from 28b = 95/5) could be attributed to related charge-charge repulsive energy differences in competing transition states And for the two homologues of 28a and 28b in which the carboxylate is attached directly to the bridgehead carbon (CH removed) the based-induced eliminations lead exclusively to the formation of the homologue of 29 as a result of minimizations of charge-charge repulsive interac- tions y 2 y 2 C02H COZH 28a X=H Z=CI 29 30 b X=CIZ=H Stereoselectivities of electrophilic additions to substituted 7-iso- propylidenebenzonorbornenes have been studied by Paquette et a1 24 The so-called ‘weak’ electrophiles such as rnetn-chloroperbenzoic acid (HI-CPBA) and N-bromosuccinimide (NBS) exhibit preferen- tial anti (to the aromatic ring) attack They bear lone-pair electrons and have high electron densities on their ‘surfaces ’The moderate (dichlorocarbene) to strong (CH,CO+ and HCO+) electrophiles have modest or large positive charges at the attacking atom and approach preferentially from the ,\n direction Thus for 7-iS0- propylidenebenzonorbornene (9-isopropyl idenemethano- I ,4-dihy-dronaphthalene) the reaction with in-CPBA or NBS leads to anti syn ratios of 83 17 and 81 19 respectively By contrast :CC12 and CH,CO+ (from AcCI-AICI,) give svn anti product ratios of 65 35 and 100 0 respectively Although differing from the explana- tion originally offered Houk and coworkers*’ argue convincingly that what they call ‘negative’ electrophiles (such as m-CPBA and NBS) give anti addition because of the electrostatic repulsion between the electrophile (I e the nonbonding electrons of the attack- ing atom) and the electron-rich aromatic ring in the svn transition structure This differs from the earlier24 explanations since it empha-sizes that syn addition is disfavoured not that anti addition is stabi- lized by ‘backside’ involvement of the aromatic relectrons With a positive electrophile rvn attack 1s believed to be favoured by both torsional effects and electrophile-.rr-electron-region attractions Additional experimental support for the primacy of electrostatic effects in the .rr-facial selectivity of nucleophilic substitutions in cyclohexanone derivatives comes from the recent work of Adcock and coworkers 25 Their investigations were of a series of S-substi- tuted adamantyl derivatives 31 and included (carbonyl carbon) chemical shifts r-facial selectiv~ties in NaBH reductions and 0 31 I75THROIJGH BOND AND THROIJGH-SPACt MODELS-K BOWDEN AND E J GRUBBS inethyllithium additions and correlations with polar field parame- ter\ They concluded that electrostatic field interactions exert a major effect in determining these n-facial competitive reactions Similar electrostatic effects appear important in other related systems Okada and coworkersZh reported that LiAIH reductions of 32 in tetrahydrofuran (THF) yielded 62% \\tz addition of hydride when X and Y are H but this increased to 92% and 100% when X and Y are C1 and E respectively By contrast the atiti addition IS dightlq favoured when X is methoxy The same general charge-dipole phenomena appear to account for the preferred direc- tion of hydride ddditions to 2,2-dicirylcyclopentanones” 33 and the wbstituted norbornanone2x 34 0 x % CIS NO 79 33 0-30 X NH 36 32 34 It should be noted with regard to stereofacial selectivities that there remains 50me controversy regarding the validity of various stereoelectronic theories employed in the analyses of these pro- cesses A detailed review of alternative theories has been pre- sented 2‘) However the results appear to be predominantly accommodated by an electrostatic field theory 24 4 Computational Approaches Ab rtzitio calculations of transition structures are attracting growing interest in probing electrostatic effects in a kariety of reactive systems Several examples will illustrate the general approach Houk et a1 21 have investigated the electrostatic potentials of 7- methy lenenorborene as well as energies for 7-methylenebenzonor- bornene interacting with a point charge as a model for electrophilic additions For methj lenenorbornene the electrostatic potentials range from -27 kcal mol I (1 cat = 4 184 J) over the nsystem to + 23 kcal mol I near the hydrogens (calculations at the 6-31G* level) The surface potentials indicate that charged species should attack preferentially from the J\rt (to cyclopentene double bond) direction due to electrostatic effects Negatively charged species will preferentially attack the anrr face to minimize repulsive inter- actions 1hus the results of computations of energies (again at the 6-3 I G* level) for 7-methylenenorbornene possessing an added charge (representing an attacking electrophile or nucleophile) of 0 2 e placed 2 0 A ‘above or below’ C 7 are shown with structure 35 This indicates that the cyclopentene double bond stabilizes a positively charged dttacking species and destabilizes a negatively charged one Similar conclusions are drawn for attacking charged specie5 interacting with 7-methylenebenzonorbornene Charge E,,.,/kcal mol Q(SYn) QOVI~)+ 0 2 -5 9 q,=CH,Q(utltr) + 0 2 -2 7 Q(,\.n) -0 2 33 &ant/)PQ(~nt1)-0 2 -0 3 35 H-\I H-+I X+=b Y I X 36 37 (calculations performed with Pople’s GAUSSIAN suite programs) The calculated axial preference (structure 36 X = Y = H) over equatorial hydride transfer (structure 37) specifically I 8 kcal mol I (when calculations are performed with the 6-3 I G* basis set on the 3-2 1G geometries) is in qualitative agreement with experi- mental observations of an axiai-equatorial hydride transfer ratio of about 90 10 for the lithium aluminium hydride reduction of 4-tert-butylcyclohexanone When a substituent is situated equatorial at the 4-position the calculations predicted very small variations in stereoselectivity When the substituent IS axial (X = H) the calcula tions indicated a sub5tantial additional stabilization for transition structure 36 in ever) case Thus when Y = F. the dE value favour ing 36 is 2 5 kcal mol I when Y = CI the difference is 2 9 kcal mol I (calculated Ah values for Y = OH or NH are conforma- tionally dependent with respect to rotation about the C-0 and C-N bonds but similar in magnitude when the lone-pair electrons are in the conformationally more stable orientation ‘inside’ a vert- cal projection of the ring) An experimental system for which these calculations serve as a model is the borohydride reduction of 4-sub- stituted tranr-decalones the results of which are shown in Table 9 Table 9 Stereoselectivities in the reduction of 4-substituted tratic- decalones with excess sodium borohydride in methanol at 2s “C 7o Comp X %39 %a40 38a H 60 40 38b ey OH 61 39 38c eq OAc 71 29 38d eq Br 66 34 38e 38f eq C1 ux OH 71 8S 29 1s 38g ux OAc 83 17 38h u.r CI 88 12 3% ux F 87 13 38 39 40 Thus as with the computational cyclohexanone models axial electron-withdrawing substituents at C-4 in decalone Cause larger increases in the axial borohydride addition than when in equatorial positions Houk and coworkers $0argue that electrostatic effects are a principal cause of these observations As illustrated in 41 and 42 equatorial attack by the hydride is destabilized by electrostatic repulsive interaction with the axial substituent c-6-Y 6-Y 41 42 Williams and Paddon-Row?’ have computed energy differ- ences between anti and svn directions of attack by LiH on the nor- bornenones 43 The results show a preferred svn attack for all Houk and coworkersxo have also explored transition substituents The electrostatic basis for these calculated selectiv- structures for the axial and equatorial additions of lithium hydride ities (iepulsive interactions between the hydride component of to cyclohexanone with full optimization using the 3-2 1G basis set LiH and the C=C double bond) was supported by conducting CHEMICAL SOCIETY REVIEWS 1996 X = SiH3 Me H CH,OH CN Ix X 43 point-charge calculations in which the LiH moiety was in each case replaced by a point negative charge located at the position vacated by the hydride (and assigned a charge equal to the Mulliken charge of the vacated hydride) The anti-syn energy differences [HF/6-3 1G(d) computations] paralleled those for the real transition structures Related computational studies28 32 have been reported for .rr-facial selectivities for hydride (LiH) attack on 43a where X=CO,CH or the corresponding dialdehyde 43b and on 44 including the systems X=O and X=CH Computational studies (based upon PM3-derived molecular electrostatic poten- tials) reveal significant electrostatic field contributions to stereo- facial selectivities in the reaction of dichlorocarbene with 45,33 stereochemical selectivities in nucleophilic additions to vinylic ~ulfoxides,~~electrophilic reactions with allylic alochols and 44 45 46 ethers ,35 and to the regiochemistry and stereochemistry of electrophilic additions to other variously substituted allylic double bonds 36 5 NMRStudies The results from I9F NMR studies37 are more complex Thus the I9F NMR substituent chemical shifts of 4-substituted F F x x 47 48 bicycle[ 2 2 2 loctan-1 yl fluoride 47 were considered to be depen- dent on both the electrostatic field and electronegativity effects The latter effect was considered to be identical to the u-inductive effect and not significant after the first atom of attachment Then the electronegativity contribution was ascribed to a ‘through-three-bond’ electron delocalization mechanism This was supported by a study of the 4-substituted bicyclol2 2 1Jheptan-1-yl fluorides 48 which revealed an electronegativity contribution of the opposite sign to that observed for 47 What is very clear is that IYFNMR sub- stituent effects involve both through-bond and through-space effects However the effects are not always the same as those asso- ciated with studies of reactivity 6 Conclusions In this account we have given an overview of a variety of chemical reactivity phenomena The results appear to be consistently explained in terms of an electrostatic field effect but not by the clas- sical inductive effect The Kirkwood-Westheimer theory even with the modifications available is not entirely satisfactory A compar-ative study3* of the gas-phase acidities of substituted quinucl idin- ium and bicyclol2 2 2]octanylammonium ions indicated a field effect transmission but the fall-off observed between the two series is considerably less than that expected from a simple electrostatic theory However it appears likely that future generations of theoret ical treatments will allow calculations for molecules even more complex than those discussed here The following quotation seems highly relevant to the present position of the classical inductive effect regarding reactivity studies ‘Is there any other point to which you would wish to draw my attention?’ ‘To the curious incident of the dog in the night-time ’ ‘The dog did nothing in the night-time ’ ‘That was the curious incident,’ remarked Sherlock Holmes 39 Acknowledgements We are grateful to our research students and colleagues for numerous discussions and insights regarding this problem over many years 7 References 1 A R Katritzky and R D Topsom J Chem Educ 1971,48,427 2 L M Stock,J Chem Educ 1972,49,400 3 0 Exner and P Fiedler Collect Czech Chem Commun 1980 45 1251,O Exner and Z Friedl Progr Phvs Org Chem 1993,19,259 4 J G KirkwoodandF H Westheimer,J Chem Phys ,1938,6,506,F H Westheimer and J G Kirkwood J Chem Phys 1938,6,513 5 F H Westheimer Tetrahedron 1995,51,3 6 G E K Branch and M Calvin The Theory of Organic Chemistry Prentice Hall New York 1941 7 J C McGowan J Appl Chem 1960,10,312 8 K Bowden and E J Grubbs Progr Phys Org Chem 1993,19,183 9 J D Roberts and R A Carbon] J Am Chem Soc 1955,77,5554 10 C A Gr0b.A KaiserandT Schweizer,Hefv Chim Acta 1977,60,391 11 M Charton Progr Phys Org Chem ,1981,13 1 19 12 K Bowden M Hardy and D C Parkin Can J Chem ,1968,46,2929 K Bowden and G E Manser Can J Chem 1968 46,2941 K Bowden J Chim Phvs Chim Biof 1992,89 1647 K Bowden Can J Chetn ,1965,43,3354 13 K Bowden and D C Parkin Can J Chetn . 1968,46,3909 14 M Beugelmans Verrier L Nicolas A Gaudemer and J Parello Tetrahedron Lett 1976,361 15 R Golden and L M Stock J Am Chem Soc 1966 88 5928 R Golden and L M Stock J Am Chem Soc 1972,94,3080 16 S Acevedo and K Bowden J Chem Soc Perkin Trans 2,1986,2051 17 E J Grubbs and R Fitzgerald Tetrahedron Lett 1968 4901 E J Grubbs R Fitzgerald R E Phillips and R Petty Tetrahedron. 1977,27 935 C T Wang and E J Grubbs J Org Chetn 1977,42,534 E J Grubbs,C T Wangand L A Deardurff,J Org Chem ,1984,49,4080 18 S Acevedo and K Bowden J Chem Soc Chem Cornmun ,1977,608 S Acevedo and K Bowden J Chem SOC Perkin Trans 2.1986,2045 19 M G Siegel C L Liotta and D J Cram J Am Chem Soc 1982,104 1387 20 K Bowden and M Hojatti J Chem Soc Perkin Trans 2 1990,1197 21 Y D Wu,Y L1.J NaandK N Houk,J Org Chem,1993,58,4625 22 K Bowden and K D F Ghadir J Chem Soc Perkin Truns 2 1990 1329 23 E J Grubbs S P Schmidt C T Wang M H Goodrow R M Lewis L A Deardurff and D Coffey Jr ,J Am Chem Soc ,1983,105,4115 E J Grubbs,S P Schmidt,C T Wang,Z Chen,A A Hamed,E S Soliman,P Nunez M H Goodrow,R M Lewis,L A Deardurffand D Coffey Jr ,J Org Chem ,1985,50,2886 E J Grubbs unpublished results 24 L A Paquette,L W Herte1,R GletterandM Bohm,J Am Chem Soc 1978 100,6510 L W Hertel and L A Paquette J Am Chem Soc . 1979,101,7620 L A Paquette L W Hertel R Gleiter M Bohm M A Beno and G G Christoph J Am Chem Soc ,198 1,103,7106 L A Paquette,F Klingerand L W Hertel J Org Chetn 1981 46,4403 L A Paquette F Bellamy G J Wells M C Bohm and R Gleiter J Am Chem Soc 1981,103,7122 25 W Adcock J Cotton and N A Trout J Org Chem 1994,59 1867 G Mehta F A Khan and W Adcock J Chetn Soc Perkin Trans 2 1995,2189 26 K Okada S Tornita and M Oda Tetrahedron Lett 1986.27.2645 27 R L Halterman and M A McEvoy J Am Chem Soc ,1990,112.6690 28 G Mehtaand F A Khan,J Am Chem Soc . 1990 112,6140 J Chem Soc Chem Corntnun 1991 I8 29 H LI and W J le Noble Reel Truv Chim Pays Bas 1992 111 199 30 Y D Wu J A Tucker and K N Houk J Am Chetn Soc 1991 113 5018 31 L Williams and M N Paddon Row J Chem Soc Chem Comtnun 1994,343 THROUGH-BOND AND THROUGH-SPACE MODELS-K BOWDEN AND E J GRUBBS 177 32 M N Paddon Row,Y D Wuand K N Houk,J Am Chem Soc ,1992 37 W Adcock and A N Abeywickrema J Org Chem ,1982,47,2957,W 114 10638 Adcock A N Abeywickrema and G B Kok J Org Chem ,1984,49 33 B Halton and S G G Russell J Org Chem ,1991,56,5553 1387 W Adcock and V S Iyer Tetrahedron Lett ,1984,25,5209,W 34 S D Kahn K D Dobbs and W J Hehre J Am Chem Soc 1988 Adcock and V S Iyer J Org Chem 1988,53,5259 110.46202 S D Kahn and W J Hehre J Am Chem Soc ,1986,108 38 W Adcock F Anvia G Butt A Cook P Duggan C A Grob S 7399 Marriott J Rowe M Taagepera et a1 ,J Phys Org Chem 1991,4 35 S D Kahn and W J Hehre J Am Chem Soc 1987,109,666 353 36 S D Kahn. C F Pau. L E Overman and W J Hehre J Am Chem 39 A C Doyle Memoirs of Sherlock Holmes Silver Blaze Newnes Soc 1987,109.650 London 1893

ISSN:0306-0012    

DOI:10.1039/CS9962500171

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

The Chemistry of Paper Conservation V. D. Daniels Department of Conservation The British Museum London WCIB 3DG UK 1 Introduction There are probably more collectable objects based on paper than any other material books prints. drawings watercolours stamps. banknotes matchbox labels etc. Not surprisingly owners of such objects want them to remain in good condition despite the fact that many of them were only intended to have a lifetime of as little as a few hours. e.g. tickets. However some papers and the images upon them are inherently unstable and others deteriorate quickly when stored under the wrong conditions. The study of paper conservation can be split into four areas (i) the chemistry of the deterioration of paper and the writing or images thereupon. (ii)the physics and chemistry of processes that can be performed to enhance stability and appearance (iii) the practical skills needed to carry out active conservation processes e.8. bleach-ing and Fvashing and (iv)the management and storage of a collec-tion. This review will concentrate on the first two of these areas as these \iill be of more interest to chemists. There are two ways in which paper objects can be prevented from deteriorating; these are sometimes known as passive and active methods. Passive methods are those which involve creating good storage and climatic control so that damage by handling incident light. thermal degradation. air pollution and excessively high or low relative humidity is minimised. Active methods involve treating objects to clean and stabilise them examples of this are washing deacidification and repair. 2 The Components of Paper Paper is essentially a mat of cellulose fibres formed into a sheet by draining a suspension of fibres in water through a mesh. On drying hydrogen bonds play a large part in bonding the fibres together. Many other materials can be added to sheets of paper e.g. fillers water repellents wet strength additives pigments. dyes and adhe- sives but often the main interest of a paper object is writing or surface decoration in the form of pencil chalk pastel watercolour paint or printing and writing inks. With the huge variety of materi- als involved the study of the conservation of paper artefacts becomes wide ranging. Cellulose occurs as a component of many plant cell walls. The amount of cellulose present varies between individual plants. Wood is 40-70% cellulose while the purest source cotton seed heads. contain up to 96%. These cellulose sources generally receive some sort of purification before they are made into paper Vincent Daniels obtained a BSc in cheniistry at Unirversity College. Cardifl (1970) and subsequentlj' obtained n PhD studying the reactions of ther-mally degraded PVC ( 1973). He joined the British Miiseirrn in 1974 to work on the conservation science of paper. Since then he has worked on n wide range of conservation problems and is currently working on the deteriorcition and conser\>ation of barkcloth arid black-dyed New Zealatid ,jar. thereby increasing the amount of cellulose present in the fibrous components. The cellulose molecule can be considered to be a linear polymer of P-ri-glucose. The average number of glucose molecules in the cellulose chain is indicated by the degree of polymerisation (DP) which varies with the plant source and the severity of any purifica- tion method and indeed the method of measurement. Measured by the ultracentrifuge method flax (Li~irni irtnsitissimi~tn) cellulose has a DP of 3500 cotton 1000-3000 and sujfite wood pulp 600-1000.' When two glucose molecules condense to form part of a cel-lulose chain one of them has to rotate bj 180" forming what many regard as the real repeat unit of cellulose cellobiose The cellulose molecule possesses a very high concentration of hydroxy groups and thus interacts strongly with water. Cellulose might be a water-soluble polymer if it did not have a strong tendency to associate to form crystalline regions. Natural wood. cotton and flax cellulose all show a strong X-ray diffraction pattern from a structure called cellulose I. The amorphous areas of a cellulose fibre are more hygroscopic and generally reactive than the crystalline areas because of their better accessibility to small molecules. The cellulose used for paper making is present as fibres which retain some of the shape of the cells from which they were derived and thus occur in a variety of shapes and sizes depending on the plant source. Mechanical treatments for paper making such as beating. enhance properties for paper-making but can destroy much of the natural structure and drastically decrease the crystallinity of the cellulose. Most European papers used to be made from recycled cellulose from cotton and linen (flax) rags. However as the demand for paper grew new sources of supply were looked for. Despite the use of some unusual sources for cellulose e.g. mummy wrappings a new plant source was needed; wood had the correct properties and from about 1870 until the present day most paper has been made from wood. Now cotton and linen papers are only used for speciality uses such as banknotes filters and artists' papers.2 Wood contains cellulose and two other main components other carbohydrates and lignin. The most permanent papers are usually those with a high proportion of cellulose but for economic reasons the amount of purification used on wood pulps is often minimal; the process can sometimes be little more than grinding the wood fol- lowed by bleaching. Papers from this type of pulp often known as ground wood papers are often used for cheap books and news- papers. The permanence of these papers is much better than when they were first produced at the turn of the century but still present a challenge to conservators as these papers progressively discolour and become acidic and brittle. Indeed the main conservation chal- lenge to large libraries is the preservation of huge quantities of this paper sometimes in the form of uninteresting low value books. Faced with the high cost of conservation and storage some libraries may choose to discard this type of material; however copyright libraries are obliged to keep it. I70 Besides the polymerised glucose in cellulose papers contain other polymerised five- and six-membered ring sugars polymers made from these are collectively known as hemicelluloses As they are not usually homopolymers crystallinity is low and their reactiv- ity is high compared to cellulose Hemicelluloses are thought to contribute to the flexibility of the cellulose fibre by having an effect on internal stress redistribution Compared with cellulose the natural degradation of hemicelluloses (and lignin) has not been examined in detail but the accelerating factors for cellulose proba- bly apply Lignin is present at a proportion of 17-32% in wood Its chem- istry is very complex and the structure of lignin varies between plant sources Essentially lignin is a three-dimensional polymer com- posed of building blocks similar to those shoun The units are linked together in an almost random manner by a variety of links p-coumaryl coniferyl sinapylalcohol alcohol alcohol CH&H CH20H CH20H I I I OH OH OH Monomer units which make up lignin These links may also be made to polysaccharides such as cellulose and hemicellulose Full or partial extraction occurs as part of the paper-making process and this further complicates the chemistry of the residual lignin For the paper conservator the presence of lignin in a paper is usually regarded as bad news as the initial strength of the paper is relatively low and it will be susceptible to development of acidity and discolouration However the bad reputation that lignin has is partly due to the associated acid pulping processes that leave aluminium sulfate and other acidic resides Modem processes can now make alkaline mechanical wood papers which are pre dicted to have good permanence and the paper industry is now anxious to prove that this is so Paper made only from plant fibres may have poor strength be extensively absorbent and have poor surface smoothness or optical properties A variety of mechanical operations can change these properties but additives such as fillers and surface coatings can be applied A common addition to most papers is that of a size tradi tionally this was made from gelatin applied as a hot solution. some- times with the addition of paper makers’ alum (aluminium sulfate) In the 19th century resin and alum were found to be cheap and effec- tive substitutes Sizes can be mixed into the vat of fibres or applied to the formed paper sheet 3 The Principal Chemical Mechanisms of Paper Degradation When the DP of weak brittle paper is measured it is found to be considerably lower than that of the new paper Thus the principal degradation mechanism is considered to be chain scission There are two principal routes for this acid-catalysed hydrolysis of the glycosidic bond and oxidation of the glucose ring and of these the first is the most important Absorbed air pollutants sulfur dioxide and nitrogen dioxide are important sources of acidity but acidity may also be introduced into the paper-making process by use of acidic components e g aluminium sulfate and resin Carboxylic acid groups on the cellu lose chain are another source of acidity and these can be produced by oxidation of the cellulose during the paper making process by the use of oxidising bleaches or subsequently in storage by oxida- tion with ozone or oxygen The rate of oxidation both during bleaching and on subsequent ageing is dramatically accelerated by the presence of transition metals such as iron and copper as they enhance the decomposition of peroxides formed during oxidation CHEMICAL SOCIETY REVIEWS 1996 Light especially ultraviolet also increases the rate of oxidative degradation Oxidised celluloses are considered to be unstable when the carboxy groups are in their free form Storage of paper in relative humidities of greater than about 70% encourages the growth of microorganisms which can rapidly cause destruction of paper 4 Acidity and deacidification For paper not exposed to excessively high levels of relative humid- ity heat or light the effects of acid hydrolysis are regarded as being the most serious degradative factor The obvious remedy for exces- sive acidity is deacidification Deacidification of paper can be as simple as immersing a sheet of paper in a bath of water or a dilute alkali However the hazards of such an aqueous process can be considerable possibly resulting in loss of image materials delamination or disintegration of the paper Thus any processes described are best carried out by a paper con- servator Readers interested in the practical skills of paper conserva- tion can find more information in the paper conservation journals The study of naturally old and artificially aged paper has shown that the introduction of calcium and magnesium carbonates is a good stabilising process However there are several other methods which can be used to give paper a neutral or alkaline pH We specify the pH of a water extract of a piece of paper to be in the range 6 0-8 5 for permanence Depending on the method adopted the pH is defined as the pH of an aqueous extract of 1 g of paper in 50 or 70 ml of cold or hot water 6 Water based deacidification treatments have the advantage that water-soluble acids and degradation products can be washed out The swelling of the cellulose fibre and the subsequent remaking of some hydrogen bonds in the drying process can change the proper- ties of the sheet usually for the better On air drying the deacid- ifying agents calcium and magnesium hydrogencarbonates and calcium hydroxide form carbonates in the fibres leaving an ‘alka- line reserve’ to deal with subsequent formation of acids The Ca2+ and Mg2+ ions can also exchange with H+ on carbonyl groups enhancing the stability of the cellulose against further deteriora tion 5 Acidity may be neutralised with sodium carbonate or hydro- gencarbonate but sodium ions have acquired a reputation for destabilising paper Unless carefully controlled sodium carbonate can produce an excessively high pH in paper alkalizing to over pH 10 eventually causes discolouration and oxidation and so alkaline earth compounds which do not cause these side effects are pre- ferred Novel deacidification agents are slow to be accepted for example French workers have reported that borax (sodium borate) can produce good results in accelerated ageing tests but conserva tors prefer to use tried and tested methods The bad effect that sodium hydrogencarbonate has on paper can be explained by the mechanism of alkaline hydrolysis In conditions of mild alkalinity the principal degradation mechanism is P-alkoxy elimination In this reaction a glucoside is eliminated which is in a position P to a carbon atom with an electron withdrawing group and an acidic hydrogen When the elimination occurs at the end of a cel- lulose chain it is called a ‘peeling’ reaction All straight cellulose chains have two ends one of these has reducing properties as it is a potential aldehyde and electron with drawing group If this is attacked by a hydroxy ion elimination takes place (Scheme 1) Other reaction mechanisms can be drawn for elimination due to electron withdrawing groups on other carbon atoms This degradation mechanism has received a great deal of atten- tion from the paper and textile industries as the bleaching of wood and textile fibres in alkaline conditions is of major importance One way of preventing this degradation is to remove metal ions known to catalyse it particularly cobalt iron and manganese However the addition of magnesium carbonate minimises the oxidative degrada- tion by either stabilising peroxides deactivating transition metals chelating with oxycellulose or complexing with hydroxy groups at C 2 and C 3 of the cellulose ring9 lothereby stabilising the cellulose against P-alkoxy elimination THE CHEMISTRY OF PAPER CONSERVATION-V D DANIELS CH20H cell-0 OH H- H- H OH 0 I glucoseend 11 fructose end p-elimination Scheme 1 While some single sheets of paper may sometimes be amenable to aqueous immersion treatments others are not and conservators have devised floatation and mist techniques enabling them to con- tinue to use aqueous methods However some objects cannot be exposed to water in any form and non-aqueous deacidification is necessary In particular wet books distort badly when wetted and even more so in redrying A great deal of ingenuity has gone into devising non-aqueous deacidification treatments One of the first was the use of barium hydroxide in methanol but this has now dropped out of use because of the toxicity of barium carbonate residues in the paper Magnesium methoxide in methanol is useful but the solution is par- ticularly unstable to moisture A great improvement in working properties is effected by saturating the solution with carbon dioxide the resulting product is called methylmagnesium carbonate and may be purchased from some of the conservation supply houses as an aerosol or solution Until recently CFCs were added to the methanol to dilute the solution II For mass treatments of the vast amounts of paper in libraries a vapour phase method has many attractions Ammonia is not suitable as the reaction product of the neutralisation of sulfuric acid (the essential reaction product of sulfur dioxide pollution) is ammonium sulfate which is still acidic (weak base + strong acid) Chemists in the USA have devised a treatment using diethylzinc gas (DEZ) The process has met with some success however under the wrong conditions DEZ reacts explosively with oxygen and water The process has to be performed in a vacuum chamber with heat-dried books I2 Recently two new processes have been devised using magnesium butoxy glycolate in a non-aqueous solvent' and a micro-dispersion of magnesium oxide particles (unpublished) The search continues for new methods which are inexpensive and easy to use 5 Discolouration of Paper Discolouration is a natural consequence of ageing as oxidation acid and alkaline degradation of the components take place and the chromophores formed start to absorb blueholet light Discoloured products can be obtained from any of the ingredients which make up paper and the range of chromophores that can be expected is correspondingly large Hon14 has reviewed the types of chromophore expected and his list includes furans quinones quinone methides and stilbenes He suggested that condensed furan derivatives produce most of the discolouration in paper and that cel- lulose containing aldehyde groups discolours most readily Aldehydes can result from several of the paper degradation mecha- nisms i e acid hydrolysis alkaline hydrolysis oxidation and pyro- lytic chain scission When paper is sized as old papers often were with gelatin then there are possibilities for Maillard reactions to take place producing brown products by reactions between proteins and reducing sugars The occurrence of brown spots on paper is known as foxing This type of discolouration is associated with storage of paper in high rel- ative humidity Two mechanisms have been identified one is the growth of fungi or bacteria in the paper the other is the corrosion of metal usually iron inclusions in the paper and both are caused by storage in excesively high relative humidity Often these mecha- nisms can be distinguished by the fluorescence caused by an ultra- violet lamp (UVA) Microbial spots fluoresce showing various colours while iron salts quench the fluorescence and appear black against the fluorescence of the paper around the spot In some papers copper alloy particles may corrode with the aid of chloride residues from bleaching eventually producing black dendritic spots composed of copper sulfide by reaction with hydrogen sulfide in the air Often an overall brown discolouration is acceptable on an old piece of paper If such a piece of paper is washed in water a lot of the discolouration can be removed as soluble products However intentional or accidental local wetting will cause migration and concentration of the stains and on drying an unsightly brown line is caused at the wet-dry boundary However localised wetting and subsequent drying can produce a line of brown or UV-fluorescing material even on a sheet of paper which is new or has recently been washed Such lines may be familiar to readers who have performed paper chromatography and are an indication of the position of the solvent front Such lines may be removed by washing but if left untreated may produce discolouration on ageing The chemistry of this phenomenon is still unclear15 but is associated with areas of enhanced oxidation Some types of paper discolouration may be caused by storage of an object next to an unsuitable material e g a piece of wood or acidic cardboard In dismantled picture frames and inside the covers of books are good places to look for these effects Often the grain pattern of wood transfers itself to paper This type of discolouration is usually attributed to the migration of acids which accelerate degradation and subsequent yellowing Acids are thought to migrate through the fibres by liquid diffusion or may be transferred through the vapour phase There has been very little research on the migration of chemical species but work in the paper industry has demonstrated that several components from resins in wood can migrate through air to deposit on adjacent objects causing water repellency Another possibility is that species from the degradation of paper can migrate from one sheet to another In particular during oxidation of paper hydrogen peroxide and oxygen-containing free radicals may be formed The production of hydrogen peroxide from autoxidation of paper and other cellulosic materials can be visual- ised on a photographic plate using the Russell effect In this process H,O reduces the alkaline photographic emulsion which is then developed to form an image such as the one shown in Figs 1 and 2 l6 Reactive oxygen-containing intermediates such as these may have been responsible for the image on the Turin shroud Other types of discolouration are produced by simpler mecha- nisms Traditionally printing inks were made from pigment (usually carbon black) mixed with a drying oil Sometimes the oil does not fully dry and it may transfer to an opposite page producing an offset image (Fig 3) On exposure to light most pigments and dyes eventually fade Paper also bleaches in these conditions but may instead discolour Lignified papers eg mechanical wood pulp are prone to dis- colouration as anyone who has left a newspaper in the sun for several weeks will know When a framed print is disassembled the area protected from illumination by the inner frame of card (matt) is sometimes seen to be a different hue In some cases the print itself will act as a type of photographic negative The paper behind the print will discolour in the areas where the light penetrates while areas behind black printing remain unaffected (Fig 4) Such phenomena are hopefully not produced in the museum or gallery environment in which both the illuminance (the amount of visible light) and the proportion of ultraviolet light in that light are usually controlled (passive conservation) Illuminance is usually CHEMICAL SOCIETY REVIEWS 1996 Figure 1 Discoloured end papers in a book (J Black s Experiments upon Magnesia Alba Quicklime err Pub1 1910) Figure 2 Russell effect image from the book cover in Fig 1 specified to be below 50-80 lux and the ultraviolet level to be below 75 p,W/lumen Ultraviolet light is more energetic than visible light and is more effective in promoting most photochem ical degradation mechanisms The use of flash lights for photo graphy IS prohibited in many exhibitions but calculations will easily show that photochemical changes are not the reason behind this ruling as the amount of light produced by the flash is equiva lent to only a few seconds of exposure to normal museum lighting A strange discolouration produced by light is that associated with zinc oxide pigment Zinc oxide (Chinese white) has been used as a white pigment since the 1830s In the presence of light and water a photochemical reaction takes place which produces hydrogen peroxide l7 Figure 3 The image IS offset from a print on the facing page and caused by non drying oils In the Ink The spots are caused by corrosion of iron par ticles ZnO + 0 -(ZnO) + 0hv 0 +HO-HO+HO 2HO-HO +O H 0 (ZnO) 22H0 Some or all of these reactive species can then react with the paper producing oxidation products At the time of illuminance no visible effect is produced but maybe a year later a yellow/brown discolour ation can be seen around the area of zinc oxide Figs 5 and 6 show THE CHEMISTRY OF PAPER CONSERVATION-V D DANIELS Figure 4 A negative image of a print caused by light induced discoloura tion of the backing board a backing board from a Japanese print the areas of discolouration correspond to areas of zinc oxide on the print in front This and many other discolouration reactions on paper have been reviewed by Daniels I* 6 Elimination of Discolouration Mdny types of uniformly distributed or localised discolouration in paper can be removed by washing in water Often a few drops of a non ionic surfactant or ethanol are added to the water to aid wetting Immersion of a piece of paper in a dish of water is often used but flotation of a sheet on the surface of the water or placement on wet blotting paper are favoured when the image inaterials might be mobile or the paper s wet strength is low Two recent advances have aided washing and have eliminated the necessity of using more chemically exotic techniques One of these techniques 15 the use of a suction table The paper is placed on a perforated table and solvent usually water based is drawn through by a low vacuum The other is the use of ultrasonically generated water mist in which objects are bathed for several hours without becoming too wet Stains which do not yield to treatment with cold water may be dissolved in warm water or water made alkdline with ammonia solution Proteinaceous or starch based stains may be removed with the aid of enzymes but oils may be removed by solvents such as morphol ine Metallic stains e g iron stains caused by rusting paper clips or staples can be eliminated by dissolving the iron with a complexing agent Some of the solutions that have been used include oxalic acid hydroxylamine hydrochloride the disodium salt of ethylene diaminetetraacetic acid ascorbic acid and mercaptoacetic acid Insoluble organic stains may be bleached out and there is a wide variety of treatments for this process The ideal bleach is one that does not harm the paper but most oxidising bleaches do damage the paper to some extent Although widely used up to ten years ago chemical bleaching agents are now out of favour however one green technique has achieved a degree of popularity that is light Conservators place paper In solutions Of a e g magnesium hydrogencarbonate and expose it to sunlight or a Figure 5 A modern Japanese print recently purchased by the British Museum The area on the top left hand corner has a high concentration of Figure 6 Discolouration of the backing board of the print in Fig 5 caused by light I84 bank of fluorescent tubes ultraviolet light being filtered out The discoloured areas are theoretically the only ones which absorb light and the only ones to be oxidised Fluidity measurements have shown that there is no degradation of the unstained paper and con servators feel that the paper has been generally improved by the process I9 Other bleaching methods owe much to the paper industry Only a few years ago the most popular bleaches were chlorine dioxide (in solution) and alkaline hydrogen peroxide Chlorine dioxide solution can be made by mixing sodium chlorite in water with formalde hyde owing to its toxicity it must be used in a fume cupboard For the home based conservator hydrogen peroxide or a hypochlori te bleach eg Chloramine T used to be popular Colour reversion after bleaching has been extensively studied by Burgess20 who found that some yellowing may return but that retention of Chloramine Tby paper sometimes resulted in further bleaching on ageing Reducing bleaches have never achieved any degree of popular ity but chemically they are attractive as reduction of carbonyl groups to C-OH results in enhanced stability to ageing Evolution of hydrogen bubbles by the reagents used e g sodium borohydride unfortunately enhances the removal of unstable image pigments and loose fibres 7 Methods for studying the Ageing of Paper The degradation of paper is slow under ambient conditions and measurable changes in chemical or physical properties often cannot be detected until after several decades of storage Two methods cur rently exist which may be used to study oxidation of paper under ambient conditions One of them the Russell effect uses specially prepared photographic film which develops a latent image by reac tion of the silver halide emulsion with hydrogen peroxide liberated during free radical oxidation processes such as HO + RH -R + H,O The technique may be used to compare samples when the rate of hydrogen peroxide evolution varies 16 The other technique chemi luminescence may also be used to monitor the rate of oxidation Cycling relative humidity has been found to produce an increase in the apparent rate of oxidation In this case chemiluminescence is produced by reaction between peroxide radicals Most laboratories use accelerated ageing tests to predict the effects of dark and light ageing and each has its own preferred conditions For dark ageing various ageing regimes exist all are based on the principals encapsulated in the Arrhenius equation Dry oven ageing at 100 "C for three days was once widely used however humid conditions and lower temperatures are now more popular Heat ageing in sealed containers keeps the water content reasonably constant but produces different results as the rates of degradation can be up to eleven times faster than expected Light ageing is usually performed by exposing samples to greatly increased illuminance sometimes with an enhanced ultraviolet content The light sources range from sunlight to mercury and xenon lamps but in most experiments heating is an unwanted side effect The aged paper can be characterised chemically by the methods used in the paper industry e g reducing power (copper number) DP (fluidity) pH FTIR etc Physical methods can distinguish between the strength of the fibres and the strength of the fibre to fibre bonds by the use of both standard tensile testing and zero span tensile testing A test which involves both bending and tensile strength is the folding endurance test All these methods reveal something different about the paper but it is impractical to perform them all Most workers use one or two of these A complicating factor in paper conservation research is the enor mous variety of paper types encountered A conservation treatment found to be ideal for one type can be damaging for another Ideally any conservation study should include over ten different papers however the scale of such projects can soon become overwhelming and the choice of papers has to be restricted CHEMICAL SOCIETY REVIEWS I996 8 Ideal Storage Conditions do they exist? The deterioration of the components of paper requires photochem ical or thermal energy and reactants in the form of water and oxygen and air pollutants Storage in an oxygen free environment is not usually considered worth the effort as oxidation is not the principal degradation mech anism Storage under dry conditions enhances stability consider ably but there is associated shrinkage and embrittlement Two pieces of paper joined together have different hygroexpansivities and storage at low relative humidity will produce a stress which may produce creep in the paper and subsequent loss of flatness on a return to ambient conditions Storage at low temperatures is also effective in prolonging the life of paper but bears a high energy cost Both low relative humidity and low temperature storage present problems of acclimatisation to ambient conditions when objects need to be removed from the store for examination The storage conditions generally accepted are within the range 14-20 "C with a tolerance of +2 "C within the range and a rela tive humidity within the range 45-55% these are a compromise between cost human comfort and practicability It is recommended that colour prints should be stored at 30-50% relative humidity preferably below 40% and at less than 2 "C 22 9 Effect of some Image Materials 9.1 Iron-gall ink Prior to this century many writing inks were either based on carbon or iron gallotannate black pigments Inks made from iron gallo tannates now present the conservator with challenges as they can cause so much damage to the paper that the degraded paper falls out of the page Preparation of these inks usually involved the use of iron sulfate and oak galls rich in tannins Reaction of these ingredi ents produced a blue/black iron-tannin complex and sulfuric acid Ageing of these inks produces fading as the tannin component degrades and the complex is destroyed with subsequent loss of colour intensity The degradation of the ink pigment may release the iron in a form which can catalyse the oxidative degradation of paper The sulfuric acid accelerates hydrolysis of the cellulose The standard treatment for stabilisation of these inks is deacid ification Recently Neevel has shown that deacidification using mag nesium hydrogencarbonate coupled with complexation of free iron(i1) ions with phytic acid is particularly effective 23 in this way both acid hydrolysis and oxidative degradation are slowed down An example of iron gall ink damage is shown in Fig 7 where the ink has completely degraded the paper in parts of a Guercino drawing 9.2 Verdigris Artists throughout the world have used a great variety of pigments Most have no accelerating effect on the ageing of the paper sub strate but the one that has proved most troublesome to the paper conservator is that known as verdigris a blue/green pigment of vari able composition but which contains copper and acetate ions When painted onto paper usually in a gum binder the copper ions cause catalysed oxidation of the paper the simultaneous release of acetic acid also accelerates deterioration by stimulating acid catalysed hydrolysis As with iron gall inks the paint can cause destruction of the paper This pigment was often used in Indian and Persian miniature paintings and was often applied as a rectangular border round a painting The interesting but unfortunate separation of the painting from its border often results Magnesium hydrogencarbonate solution as a deacidifying agent and stabiliser for oxidative degradation has been shown to prove effective in accelerated ageing tests If the paper has not deterio rated too much a strengthening consolidant of water soluble hydroxypropylcellulose has been recommended 24 Fig 8 shows one of the finest of all Moghul miniature paintings Princes of the House of Timur The roof of the pavilion has been almost entirely destroyed by verdigris only a cotton backing applied to strengthen the painting remains THE CHEMISTRY OF PAPER CONSERVATION-V D. DANIELS Figure 7 Iron-gall ink bum on a Guercino (1599-1666) drawing. The holes are the result of iron-catalysed degradation and acid-catalysed hydrolysis. 93 Mercury Scarlet An unusual pigment sometimes found in the Victorian watercolour box is based on the scarlet mercury(1r) iodide Hg,I,. It was used to paint flowers. Unfortunately the pigment has a significant vapour pressure and unless painted thickly or sealed into a binder has evap- orated from many pictures leaving nothing behind.25 9.4 Vermilion This pigment mercury(i1) sulfide HgS is found in nature as cinnabar but is also synthesised. Sometimes the pigment blackens particularly on wall paintings. Recent research has shown that the blackening is accelerated by the presence of halide anions and that the reaction responsible is probably solubilisation and reprecipita- tion as the amorphous HgS or a black crystalline form meta- cinnabarite. The solubilisation is aided by the halide ions forming an HgXi-ion and by increased relative humidity.26 95 Arsenic sulfides The pigment orpiment As,S was widely used in oriental and Egyptian paintings. It is however unstable to light and heat. In some instances the pigment bleaches to form the corresponding oxide and causes flaking of the pigment possibly due to destruction of the binder. The liberated sulfur may combine with other inor- ganic pigments containing lead and copper e.g. basic lead carbon- ate lead oxide lead chromate copper carbonate etc.to form black sulfides. Fig. 9 shows two adjacent sheets of papyrus bearing stripes of initially yellow orpiment. The stripes on the right-hand sheet have completely faded because of exposure to light. Figure 8 Detail of the Moghul miniature painting Princes of the House of Timur. The combined effects of copper-catalysed oxidation and acid- catalysed hydrolysis have destroyed the roof of the pavilion 9.6 Lead White A white pigment made by the corrosion of lead metal is basic lead carbonate 2PbCO;Pb(OH),. It was widely used as a pigment before being replaced by zinc oxide and later by titanium oxide. It has one great drawback it reacts with atmospheric hydrogen sulfide to form black lead sulfide. In paintings affected white areas appear grey or black and in drawings the once white highlights become shadows giving a peculiar photo- graphic ‘negative’ appearance to a picture. The cure for this afflic- tion is the use of an ethereal solution of hydrogen peroxide. Diethyl ether and 20 volume hydrogen peroxide are shaken together and some of the peroxide dissolves in the ether. The ether layer is brushed onto the blackened areas and the sulfide is oxidised to white sulfate restoring the original appearance?’ Fig. 10 shows a drawing in which white highlights have been black- ened. 9.7 Metals Some manuscripts carry metal foils; one particular problem is silver foil which tarnishes forming a layer of silver sulfide It is rare that the silver is sufficiently thick to polish so the change is permanent. In the case of finely divided silver in a paint medium blackening is similarly permanent. The conservation of photographs on paper has not been covered in this review but the black and white image is made of silver particles which are likewise sulfided resulting in fading of the image or silvering of the surface; a good review has been published by Kodak.28 CHEMICAL SOCIETY REVIEWS 1996 Figure 9 Orpiment (AslS,) on adjacent papyrus sheets of a book of the dead. The sheet on the right has been exposed to a lot of light. 10 Conclusions The many types of paper that have been made and the variety of images on the paper ensure that the study of the deterioration and conservation of paper is a vast subject. The principal deterioration mechanisms for papers and associated materials have been scientif- ically studied and many useful conservation methods have been produced. However many objects will remain untreated because of a lack of resources. Despite all the efforts of conservators and chem- ists deterioration can only be slowed down not halted. 11 Bibliography At present there is no text exclusively or even partly devoted to paper conservation science. Interested readers are directed to the two prin- cipal paper conservation journals Restaurator and The Paper Conservator. General conservation journals such as Studies in Conservation and The Conservator have articles related to paper conservation. Conference proceedings carry a large proportion of the published material; the American Chemistry Society’s Advances in Cheinistry Serif5 No. 164 193,212 and 410 are a good introduction. 12 References K. E. Britt Handbook of Pulp and Paper Technology. Van Nostrand New York 1970. D. Hunter Papennuking. Dover Publications New York 1974. H. L. Spiegelberg Tappi. 1966,49,388. J. Mills and R. White. The Organic Chemistry of Museum Objects,. Butterworth-Heinemann 1994. J.C. Williams,C. S.Fow1er.M. S.LyonandT. L.Merril,in Preservation of Paper and Te.rti1e.y of Historic and Artistic Value ed. J. C. Williams American Chemical Society 1976 pp. 37-57. Recommendations for repair and allied processes for the conservation of documents BS 4971 Pt.1,Dec. 1973 British Standards Institution. Figure 10 Blackened lead white highlights on a drawing from the Dal Pozzo Collection (ca. 17th century). 7 M. Hey The Paper Conservator 1979.4 66. 8 A. Meller Tappi 1965,48 231. 9 J. Defraye H. Driguez and A. Gardelle in Proc. 8th Cell. Conf. .. 111 General Papers,ed.T.E.Timel1 Appl. Poly. Symp. No. 28 Wiley. New York 1976. 10 B. Ericsson. B. 0.Lindgrew and 0.Theander Sv. Papperstidn.. 1971. 74,759. 11 G. B. Kelly L. C.Tang and M. K. Krassan. in vol. 5. pp. 62-7 1. 12 A. Lienardy Restaurator 1991 12,75. 13 R. Wedinger Chem. Br. 1992.898. 14 D. N. S. Hon Preservation of Paper and Textiles of Historic and Artistic ValueI1,ed.J. L. Williams,AmericanChemical Society 1981,pp. 119-141. 15 J. K. Hutchiss. J. Am. Inst. Cons. 1983.22,57. 16 V. Daniels Phil. Trans. Roy. Soc. 1986 A320,285. 17 R. Ranby and J. F. Rabek. Photodegradation. Photoxidation and Photostabilisation of Polymers Wiley London. 1975. 18 V. Daniels The Paper Conservator 1988,12.93. 19 T. T. Schoeffer M. T. Baker V. Blyth Hill and D. Van der Reyden.J. Am. Inst. Cons. 1922,31,289. 20 H.Burgess Conservation of Library and Archive material^ and the Graphic Arts ed. G. Petherbridge. SOC. Archivists-Inst. Pap. Cons.. Butterworth 1987 pp. 57-70. 2 1 V. D. Daniels and L. E. Fleming in Conservation of Historic und Artisric Works on Paper ed. H. D. Burgess. Canadian Conservation Institute. Ottawa 1994 pp. 155-162. 22 IS0 11799 International Standards Organisation in press. 23 J. G. Neevel. Restaurator 1995,16,143. 24 G. Banik H. Stachelberge and 0.Wachter in Science and techno log^ in the Service ofConservation,ed. N. S.Bromelle and G. Thomson Int. Inst. Cons London 1982 pp. 75-78. 25 V. D. Daniels in Recent Advances in the Conservation and Analysis of Artefacts ed. J. Black University of London. 1987. p. 283. 26 Ref. 25 pp. 280-282 and errata slip. 27 V. Daniels and D.Thickett.in Conference Papers 1992,ed. S. Fairbrass. Inst. Paper Cons. Manchester 1992 pp. 109-1 15. 28 Conservation of Photographs Eastman Kodak Company. New York. 1985.

ISSN:0306-0012    

DOI:10.1039/CS9962500179

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Diatomic Molecular Probes for Mid-IR Studies of Zeolites A. Zecchina Dipartimento di Chimica Inorganica Chimica Fisica e Chimica dei Materiali Universita di Torino Via F! Giuria 7 10125 Torino Italy C. Otero Arean Departamento de Qui'mica Universidad de las lslas Baleares 0707 I 1 Introduction Zeolites are tecto-aluminosilicates which can be described by the general formula M';Lz[(A10,)r(Si02),jr-.zH20,where M can be a metal cation or a proton. The Si/AI ratio in synthetic zeolites varies considerably; limiting extremes being 1:1 (lower ratio for zeolite X) to near infinity 1 (in silicalites). This provides a means to modulate the ionicity of the material which increases with decreasing Si/Al ratio. The framework of every zeolite is constructed from tetra- hedral building blocks TO where T is a tetrahedrally coordinated atom (i.e.Si Al) as depicted in Scheme 1. An isolated SiO group would carry a formal charge of -4 but in a solid having an O/T ratio of 2 (as found for all zeolites) the SiO unit is neutral because each oxygen atom is part of a bridge between two T atoms. However the net formal charge of the AIO units is -1 so that the zeolite framework is negatively charged. r Extraframework cation Q 1'-M TO,building units Zeolite framework Scheme 1 The net negative charge is balanced by MI*+cations or by protons in the acidic form of the zeolite. These ions are not a part of the zeolite framework and under the right conditions they can be exchanged by other cations. Such an exchange has little effect on crystal structure which depends on the way in which the TO units arrange themselves; it does affect however other relevant proper- ties of the zeolite such as acidity and internal electric fields. Brcansted acidity arises from bridging Si(0H)Al groups in the proto- nic form of zeolites. Extraframework (charge-balancing) cations act at Lewis acid centres in the broad sense since they are electron acceptors. A different (and more important) source of Lewis acidity concerns structural defects and extraframework aggregates. Adriano Zecchina was educated at the Universiv ofTurin (Italy) where he is Ordinary Professor of Physical Chemistry. He has also spent a year at Bath UniversiQ (UK) doing research on surface spectroscopy. Prof. Zecchina is a foreign member of the Faraday Division Council and has acted as President of the Associazione ltaliana di Chimica Fisica. His main research interests focus on the study of surface chemistry using spectroscopic techniques. He has co-authored over 150 sci-entiJc articles in that jeld including three reviews. sensors. Carlos Otero Arean Palma de Mallorca Spain Charge-balancing cations are also the main source of intrazeolite electric fields which have a strength of several V nm-I the actual value depending on cation charge and radius. A characteristic feature of zeolites is their structural porosity. The framework of all zeolites defines regular systems of intracrystalline voids and channels of discrete size usually in the nanometre range accessible through apertures of well-defined molecular dimensions. This is a feature that differentiates zeolites from other microporous materials like amorphous carbon or silica gel (which have irregu- lar pore systems) and which places them in the same class as other molecular sieves. Zeolites containing rings consisting of 8 10 and 12 oxygen atoms are denoted as small- medium- and large-pore materials. The corresponding pore openings are about 0.4,0.55and 0.73 nm in size respectively. Recently other zeotypes having extra- large pores have been synthesized among them are cloverite (a gallophosphate molecular sieve) AIPO-8 VPI-5 and JDF-20 with present pore openings up to 1.2-1.4 nm in diameter.'-3 Zeotypes are materials structurally related to zeolites in which Si or Al atoms have been replaced by other atoms e.g. Ga for A1 or P for Si. For a detailed account see e.g. ref. 4. Their microporous framework structures wide range of chemical composition and surface acidity and the possibility of tuning inter- nal electric fields by appropriate choice of extraframework cations are key factors which render zeolites (and zeotypes) versatile materials for an increasing number of technological applications. Paramount among these is the use of zeolites as catalysts for the petrochemical industry pollution control and the synthesis of speciality chemicals. They also serve as ion exchangers and mole- cular sieves. A more recent perspective is the proposed use of zeo- lites as host materials for host-guest composites. These are a kind of advanced materials where zeolites (or zeotypes) act as hosts for encapsulating and organizing molecules crystalline nano-phases and supramolecular entities inside their pores. Space confinement and host-guest (electrostatic) interaction result in a type of com- posite materials with novel properties .5.6 Potential applications are expected in a number of technological fields such as photochem- istry optoelectronics semiconducting devices and chemical was educated at the University of Madrid (Complutense) and obtained PhD degrees in chemistry from the universities of Bath (UK)and Madrid. He has carried out post- doctoral research at the French CNRS (Orlians) and at the University of Oxford (ICL). Currently he is Professor of Inorganic Chemistry at the Universidad de las lslas Baleares (Spain) where his main research interests are in several aspects ofsolid state and surface chemistry. 187 Zeolite and zeotype cavities can be considered as nanoreactors where adsorbed molecules are guided to react following specific paths dictated by (1) the electrostatic forces acting inside the cav- ities (II) the distribution of sites on the internal surface (111) the spatial restrictions imposed by the dimension and shape of the void space and (iv)limitations on diffusion paths imposed by the regular organization of (intersecting) channels All of these points can be summed up under the synthesizing concept of host-guest interac- tions so fruitfully used to understand many properties of supramol- ecular and enzyme-substrate systems More specifically electrostatic fields operating inside cavities and channels can lead to the formation of internal adducts characterized by a profound deformation of the electron distribution of perturbed molecules with simultaneous polarization and appearance of new nucleo-philic and electrophilic regions and ultimately of new chemical properties In the case of proton-exchanged zeolites the internal adducts can be more properly classified as hydrogen-bonded com- plexes which under suitable conditions can evolve towards proto- nated species and initiate the chain of Brgnsted acid-catalysed reactions that (together with shape selectivity) make acidic zeolites so important in petrochemistry Precise characterization of the solid is a prime requirement for understanding intra-zeolite processes and it should also assist in designed synthesis of new materials To this end crystallographic methods are efficiently complemented with spectroscopic tech- niques mainly IR and MAS NMR spectroscopy which have the potential to furnish information about the nature of surfaces and of adsorbed species Mid-IR spectroscopy (4000-400 cm-') using appropriate probe molecules is a powerful technique for studying the internal surface of zeolites The method is indirect in the sense that it depends on a suitable choice of spectroscopic probes which render the zeolite cavities IR-readable Probe molecules most often used are CO N and H although many others have been utilized The observables are the changes in vibrational frequency and intensity and the induced vibrational modes of the probe molecule as a result of interaction with the zeolite (internal) surface Induced changes of skeletal and 0-H vibrations can also be highly informative A related aspect is the use of vibrational spectroscopy to follow chem- ical processes taking place in the intra zeolite space (reaction dynamics) Time-resolved (sub-second) FTIR spectroscopy has the potential to detect intermediate species in a reaction chain for example in a study of the interaction of acetylenes with an H-ZSM 5 zeolite,7 prior formation of an H-bonded complex and subsequent transformation into a protonated species followed by polymeriza- tion could be analysed This subject however is beyond the scope of the present review We shall focus on the use of diatomic probe molecules for mid-IR studies of (I) OH groups and Brprnsted acidity (II) internal electric fields and (iu)extraframework species Cation location can also (in some instances) be discerned by both mid- and far-IR spectroscopy the latter subject was recently reviewed by Baker et a[,*while Ward9 has given a detailed account of early work on mid-IR characterization of zeolites 2 General Background 2.1 Molecular Probes The information which can be obtained from IR spectra using mol- ecular probes concerns (1) Surface hydroxy groups both silanols and brigded Si(0H)AI groups which are the source of Brprnsted acidity Ideally acid strength should also be quantified (ii) Lewis acidity which depends mainly on framework structural defects and on extraframework species The latter are very small metal oxide particles generated during thermal treatment of the zeolite (iii) Other types of structural defects such as hydroxy nests or occluded material (iv)The nature and location of extraframework cations (v) Internal electric fields Changes undergone by the molecular probe can also be mforma- tive of the nature of chemical processes expected to take place CHEMICAL SOCIETY REVIEWS 1996 within the intra-zeolite space This aspect however cannot strictly be considered as a part of zeolite characterization Ideally molecular probes for IR studies should satisfy the fol- lowing requirements (I) small size so that the probe has unimpeded access to channels and cavities (ii) detectable spectroscopic response triggered by interaction with the zeolite surface this response should discriminate between surface sites of different nature and should allow quantitative determinations to be per- formed (111) extinction coefficients for characteristic vibrational modes should be high so as to lead to optimum detection sensitiv- ity (iv) the probe molecule should be responsive not only to the surface site but also to its close environment this is particularly rel- evant for detecting bifunctional surface sites and (v) in order to avoid alterations of the material under test probe-surface interac- tion should be weak the foregoing considerations on spectroscopic response notwithstanding Many molecular probes have been used to characterize different aspects of zeolite surfaces among them H N CO N,O pyridine and several hydrocarbons None of them however satisfies entirely all of the above requirements Concurrent use of two (or more) different probes can help to resolve specific problems CO N and H have kinetic diameters of 0 376 0 364 and 0 289 nm respec- tively so that they all satisfy the size requirement for acting as probe molecules in most zeolites Interaction modes and consequent spectroscopic response will now be briefly outlined Carbon monoxide has a small dipole moment (p = 0 112 D) which varies during vibration thus rendering the C-0 stretching mode IR-active For the free molecule the fundamental v(C-0) value is 2143 cm I which is expected to change little on physical adsorption Interaction with an ionic surface is expected to produce a more significant negative shift of the CO stretching frequency for adsorption via carbon at an anion site or via oxygen at a cation site Large positive shifts result from adsorption via oxygen to an anion site or via carbon at a cation site lo The reason for the reversed sign of the frequency shift is that the dipole moment derivative dp/dr changes sign when the orientation of the dipole axis with respect to the external electric field is reversed Extensive experimental evi- dence proves that CO interacts via the carbon end with cations in ionic metal oxides and halides leading to hypsochromic shifts in the 10-70 cm I range the actual value depending on cation size and charge For do cations having a net charge =S 2+ interaction with CO is dominated by electrostatic effects ion-dipole (and quadru- pole) interactions and CO polarization and the frequency shift of the C-0 stretching mode arises mainly from the vibrational Stark effect and from a 'wall effect' stemming from increased Pauli repul- sion when the CO molecule vibrates against the solid surface When cations having a net charge larger than 2+ or transition metal ions are involved chemical effects due to udonation (from the CO Saorbital) and d-7rback donation can lead to a more complex situa- tion It should also be noted that inside zeolite cavities multiple interactions can develop which must be carefully analysed Both dinitrogen and dihydrogen being homopolar molecules are IR-inactive when unperturbed However in presence of an electric field E,an induced dipole moment aE (where a is the polarizabil- ity) develops which varies during vibration thus rendering the fundamental transition IR-active For dinitrogen ab initio calcula tionsI2 I3 (both in the SCF approximation and with electron correla- tion) of the interaction with alkali metal cations showed that the optimum configuration corresponds to the linear arrangement M+ N=N (where M+ stands for the metal ion) This interaction leads to a decreased N-N bond length and to a positive shift of the v(N-N) stretching frequency with respect to the (Raman active) gas-phase value of 233 1 cm I The intensity of the field-induced IR absorption IS proportional to the second power of the electric field,I4 or more exactly to the square of the field component directed along the molecular axis Abiniriu calculations1 for the interaction between dihydrogen and alkali metal cations led to the geometry 1 depicted in Scheme 2 which results in lengthening of the H-H bond and a negative fre- quency shift from the position of the Q branch vibration of free H However when cation-anion (Lewis acid-base) pairs were consid ered,'6 dihydrogen was found to interact simultaneously with DIATOMIC MOLECULAR PROBES FOR MID-IR STUDIES OF ZEOLITES-A ZECCHINA AND C OTERO AREAN H-H M' IH IH 1 2 (bond lengths in pm) Scheme 2 both ions This is shown in Scheme2 (2) for an adsorption site mod- elled by the AI(OH) cluster with standard bond angles and bond lengths In species 2 adsorbed dihydrogen becomes slightly polar- ized which renders it IR-active with an H-H stretching frequency lowered with respect to the gas-phase value of 4163 cm I An important feature is that the adsorbed molecule acts as a probe for the surface acid-base pairs rather than only for the cation site 2.2 Procedures Mid-IR studies of zeolites are usually performed in khe transmission mode since transparency is generally good However in the high- frequency range scattering can sometimes be a serious problem particularly when dealing with materials which have been syn- thesized as relatively large crystallites Diffuse reflectance measure- ments can help to overcome this problem Studies concerning probe molecules make use of zeolite self-sup- porting wafers of 3-6 mg cm made by compacting the pow- dered material at a pressure of about 5000 kg cn? The lowest possible pressure should be used so that the crystal structure is not distorted Use of self-supporting wafers facilitates thermal activa tion of the zeolite sample prior to dosing with the appropriate gas (probe) Activation aims at removal of adsorbed water (and other atmospheric contaminants) and is usually performed by heating the zeolite wafer in a dynamic vacuum for several hours at a tempera- ture of about 650 K Higher temperatures are likely to cause severe alteration of the zeolite particularly dealumination with attendant formation of extraframework species and lattice defects IR cells are currently in use which facilitate in situ thermal activation dosing with the probe gas and recording the IR spectrum at low tempera- ture usually at a nominal value of 77 K sample wafer cooled with liquid nitrogen This last requirement is mandatory when dealing with spectroscopic probe molecules which show only a very weak interaction with the zeolite among them H and N 3 Structural Hydroxy Groups in Proton- exchanged Zeolites Silanols and Brsnsted Acid Sites Two main types of OH groups are present in the protonic form of zeolites silanols (3) and bridging hydroxy groups (4) they are depicted in Scheme 3 It should be noted that in species 4 Si-0 and Al -0 bond lengths are different d(Si-0) < d(Al-0) /" i 3 4 Scheme 3 Silanols are mainly found at external surfaces where they satu- rate silicon dangling bonds associated with framework truncation Nevertheless silanols also occur at internal lattice defects arising mainly from partial dealumination In this case a local cluster of silanols is formed which has been termed a silanol nest Isolated silanols on external surfaces are characterized by a sharp IR absorp-tion band in the 3750-3745 cm-I range When silanols occur at internal sites weak electrostatic perturbations cause a downward shift (and broadening) of the IR stretching band which then appears at 3720- 3700 cm- I Stronger hydrogen-bonded interactions (eg in hydroxy nests) result in very broad IR absorption bands in the 3650-3200 cm-1 range I8 Bridging OH groups which are the active sites in acidic catalysis are found in two different wavenumber ranges 3650-3600 and 3580-3530 cm The high-frequency range corresponds to OH groups vibrating in large cavities formed by greater than eight- membered rings while in smaller voids the low-frequency range is observed This can be typified by the faujasite-type zeolites (FAU) which show 0-H bands around 3650 and 3550 cm I the actual value depending on Si/Al ratio The FAU-type structure is depicted in Fig 1 there are four different oxygen atoms O(1)-0(4) Neutron diffraction studies19 have shown that preferred proton sites are near O( 1) and O(3) The proton at site 1 points into the super- cage while that at site 3 is directed into the sodalite unit where it interacts with nearby oxygen atoms Such an electrostatic inter- action (related to the inverse of the square of the H-0 distance) causes a bathochromic shift with respect to the unperturbed OH group at site 1 This explains the two IR absorption bands 3650 cm-1 for OH groups at site 1 and 3550 for those at site 3 In agreement with this assignment the high-frequency band is strongly perturbed (downward shifted) by adsorbed N or CO while the 3550 cm- I band remains unaffected because the sodalite cage is not accessible to these gases It should also be noted that the 0-H stretching wavenumber cannot be directly correlated with acid strength When considering only wavenumbers corresponding to vibra- tions of OH groups in cages (or channels) with dimensions exceed- ing those of an eight-membered ring a linear correlation was found between o(0-H) and the average (Sanderson) electronegativity of the zeolite framework With increasing Sanderson electronegativ- ity S which corresponds to increasing Si/AI ratio OH groups vibrating in large cavities shift to lower wavenumbers following the equation ~(0-H) = 4274- 158 576S as determined by Jacobs and Mortier 2o Since this empirical relation holds for a large number of zeolites irrespective of structure type it is inferred that small dif- ferences in Si(0H)AI bond angles have little influence on the 0-H stretching frequency Because of the electrostatic effects already mentioned hydroxy groups in eight-membered (or smaller) rings do not follow the above equation A different example is furnished by mordenite (MOR) This is a zeolite wich has a pseudo-unidimensional pore system (the main channels) with an elliptical cross section 0 65 X 0 70 nm in diam-eter defined by twelve-membered rings of TO tetrahedra The channel wall has side pockets circumscribed by eight-membered rings which are accessible through windows with a free diameter of 0 39 nm Accessible cation (or proton) sites are located on the walls of the main channels (sites B and C) and on the side pockets (site A) as depicted in Fig 2 A detail of the bridging OH spectrum is shown in Fig 3 The cor- responding IR absorption band has a maximum at 3609 cm I and is highly asymmetric on the low-frequency side Computer deconvolution showed two components centred at 36 I2 and 3585 cm-I respectively They were assigned2' to acidic hydroxy groups vibrating in the main channels (high-frequency component) and in the side pockets (low-frequency component) Note the larger half- width of the low-frequency component which results from electro- static perturbation of OH groups within side pockets Relevant parts of the FTIR spectra2'" 22 of CO and N adsorbed at liquid nitrogen temperature on H MOR are shown in Fig 4 Only the main points will be discussed here details are contained in the original references In the 0-H stretching region IR absorption bands are observed at 3747 (silanols) and at 3609 cm I (bridging OH groups) they are both affected by CO and N adsorption The silanol band is partially eroded by interaction with the probe mole- cules at the highest doses shown in the spectra Dinitrogen shifts the 3747 cm band down to 37 10cm I (Ac = -37 cm I) while for carbon monoxide the corresponding shift amounts to CHEMICAL SOCIETY REVIEWS 1996 Figure 1 (a)The faujasite type structure showing the 12 ring leading to the supercage and the 6 rings which define apertures into the smaller sodalite cage Extraframework cation sites (1-111) are depicted as well as the four different (1 -4) anion sites (b)Detail showing the orientation of the bridging OH groups at sites O(1) and O(3) -87 cm I Qualitatively this was the expected behaviour since cm-I Corresponding half-widths are 60 cm I for N and 175cm-I CO is a stronger Lewis base The IR absorption band correspond- for CO whlch should be compared wlth the half-wldth value of 40 ing to acidic OH groups (centred at 3609 cm 1) is more strongly cm I for the unperturbed 0-H stretching band at 3609 cm I affected by the adsorbed gases A downward shift of -109 cm I is Downward sh~fts increased band width and increased intensity are observed for interaction with N while for CO this shift is -294 precisely the effects predicted by the classical theory of hydrogen DIATOMIC MOLECULAR PROBES FOR MID-IR STUDIES OF ZEOLITES-A. ZECCHINA AND C. OTERO AREAN Figure 2 The mordenite structure (viewed along the main channels) showing cation sites and adsorbed CO molecules. Dots outline Connolly surfaces obtained using a probe molecule having a radius of 0.14 nm. (Further details can be found in ref. 7). I I . '*-'L-I-I 3700 3650 3600 3550 3500 wavenumber cm-' Figure 3 Computer deconvolution of the 3609 cm-l 0-H stretching band in H-MOR components at 3612 and 3585 cm-I. The solid line is the experimentally observed spectrum. Reprinted with permission from ref. 21(a). bonding and the three magnitudes are directly correlated. The observed frequency changes upon interaction of the zeolite OH groups with the probe molecules are summarized in Scheme 4. A further detail which deserves comment is that progressive erosion of the 3609 cm- I band by both CO and N is not uniform. The high- frequency component (3612 cm-I Fig. 3) is consumed faster than the low-frequency counterpart. This is in agreement with the expected higher difficulty for the probe molecules to gain access to the side pockets while OH groups in the zeolite main channels are readily available. The bathochromic shift Av(OH) of the OH stretching band upon interaction with probe molecules is often used as a quantita- tive measurement of Bronsted For carbon monoxide an empirical relationship was established by Paukshtis and Y~rchenko,~between proton affinity PA and the value of Av(0H) referred to the corresponding value for isolated silanols having adsorbed CO Av(Si0H). This relationship is as in eqn. (1). PAM rnol-' = 1390 -442.5logldv(OH)/d~(SiOH)I(1) The lower the PA value the higher is the Bronsted acidity of the OH group. This equation however should be used with caution when analysing the catalytic activity of zeolites in processes involv- ing protonation of a reactant molecule. Proton transfer leads to charged species which interact with the (charged) zeolite frame- work and the energy balance of the process can thus be significantly affected. For analogous reasons the above relationship between PA and AC(OH) can lead to dubious results when it is used to compare intrazeolite processes with those occurring in the gas phase or in solution. It should also be noted that Brmsted acidity can be strictly defined only in terms of the reactant molecule which acts as a proton acceptor. Besides frequency shifts the intensity of the perturbed OH band or its half-width can also be used to quantify interaction with probe molecules. However since the intensity increase is directly corre- lated with the frequency shift the shifted components of a complex OH band have different specific absorbance. Hence determination of the relative amounts of the various hydroxy groups present in the sample becomes possible*3c',*s only when extinction coefficients can be evaluated. Fig. 4 also shows the C-0 and N-N stretching regions of H-MOR with increasing amouts of the adsorbed molecular probes. For CO three IR absorption bands are observed at 21 38,2172 and CHEMICAL SOCIETY REVIEWS 1996 I 1 I I1.0 I I 1 1 I n 1.a v1 v1 L.-5 0.75 .-G 8 84 0.5-f! Eln s < 0.50 0.0 1 1 1 t 0.0 ! 31 0 3600 3400 3200 3000 3800 3700 3600 3500 3400 3300 wavenumber cm-’ wavenumber cm-’ O 07010.71 YI -.-$ 0.3E fElDU 0.25 t I I I I I 2370 2360 2350 2340 2330 2320 22 D wavenumber cm-I 0.0 2250 2200 2150 2100 2050 wavenumber cm-I Figure 4 (a)0-H stretching region of H-MOR outgassed at 673 K (spectrum I) and the effect of increasing doses of adsorbed CO (spectra2-9) equilib-rium pressure from 10-2 to 20 Torr (1 Torr = 133.3 Pa). (b)Effect of adsorbed N other details as above. (c)C-0 stretching region. (d) N-N stretching region. All spectra were taken at liquid-nitrogen temperature. Those in parts (c) and (6)were background-subtracted. Reprinted with permission from ref. 21(a) [parts(a)and (c)]and ref. 22 [parts(6)and (41. 2225 cm-1 [Fig. 4(c)]. The 2138 cm-1 band which only becomes species (Lewis acid sites) which will be considered in Section 5. prominent for comparatively high CO equilibrium pressures was The 2 172cm-I band corresponds to the C-0 stretching of carbon assigned2Iu to physisorbed (liquid-like) CO inside the zeolite monoxide interacting (viathe carbon end) with the proton in bridg-channels. Note its complex structure which arises from hindered ing Si(0H)AI groups. Note that although the band is asymmetric rotation.2’u,26 The 2225 cm-I band monitors extraframework in the low-frequency side the perturbation of the C-0 stretching DIATOMIC MOLECULAR PROBES FOR MID-IR STUDIES OF ZEOLITES-A ZECCHINA AND C OTERO AREAN .-' lH* ? I 3315em '(A6 =-294)t-3500em I (AQ =-109)7-Scheme 4 vibration is not sensitive enough to discriminate between the two types of bridging OH groups main channels and side-pockets For N the (complex) band corresponding to extraframework species is now found in the 2345-2360 cm I range OH N species give a complex N-N stretching spectrum with components at 2325,2330 and 2333 cm-I they correspond respectively to N molecules interacting with silanols and with OH groups in side- pockets (2330 cm-') and in the main channels (2333 cm I) The observed frequency range spans the (Raman active) gas-phase value of 233 1 cm-' (Section 2 1) However the Raman spectrum (taken in our laboratory) of dinitrogen adsorbed at 240 K on sil- icalite (a purely siliceous zeolite) gave a U(N-N) value of 2321(2 2) cm I Taking this as the reference value all the observed N-N stretching maxima [Figure 4(d)] are upward shifted as expected Note that N is a more sensitive probe than CO for monitoring OH groups Two main factors contribute to this improved performance (I) absence of a contribution from Iiquid- like species (because unperturbed N is IR-inactive) and (11) weaker interaction with hydroxy groups,which results in enhanced selectivity 4 Charge-balancing Cations and Electric Fields Adsorbed probe molecules can be used for IR monitoring of charge- balancing cations and their associated electric fields Alkali-exchanged ZSM-5 zeolites provide a good example ZSM-5 structure type MFI in IUPAC nomenclature is a silicon rich zeolite which has a three dimensional pore system consisting of two inter- secting sets of tubular channels (ca 0 55 nm in diameter) defined by ten-membered rings of TO tetrahedra as shown in Fig 5 IR spectra of CO adsorbed at liquid nitrogen temperature on M+ ZSM 5 (M = Na K Rb Cs) are depicted in Fig 6 They all show a main IR absorption band which for low CO doses was found to peak at 2178,2166,2162 and 2157 cm I for Na+ to Cs' respec-tively This cation-specific band has been assigned27 to the funda- mental C-0 stretching of carbon monoxide perturbed by the corresponding metal ion [Fig 5(b)J Observed wavenumbers are all higher than the 2143 cm I value for free CO as expected for an M+ -.-CO interaction (Section 2 I) A linear dependence was found between the C-0 stretching frequency and the parameter l/(RM + R,J2 where R is the cation radius and Rco = 0 21 nm By using calculated values of v(C0)for the molecule in the pres- ence of an axial electric field," the field strength due to charge balancing cations (and surrounding anions) has been experimentally determined from the corresponding IR frequen-cies 27~~The values obtained vary from 6 3 V nm-t for Na-ZSM 5 down to 2 4 V nm I for Cs-ZSM-5 Slightly smaller values were observed for alkali-exchanged mordenites 21u The whole set of results is summarized in Fig 7 Although the actual values reported here must be regarded as being only a first approximation mainly because of simplifications made in the direct use of the U(C0)vs electric field relationshipltl 27b it should be clear that IR molecular probes constitute a valuable means for experimental determination of intrazeolite electric fields A similar approach was taken by Cohen de bra et a1 1428rrand by Bose and ForsterZsh who determined the electric field in zeolites Na-A NaCa-A and Ca-A from the integrated IR intensities of adsorbed dinitrogen and carbon monoxide Values in the 5-8 V nm I range were obtained which are consistent with results from a theoretical calculation of the electrostatic field assuming an ionic model for the zeolite 28 Although more refined values are desirable the order of magnitude of intrazeolite electric fields seems to be well established and also their dependence on charge-balancing cation and zeolite structure For dJ1 cations more complex situations can arise as a result of chemical interactions involving the cation d orbitals (or hybrids) and cr/norbitals of the probe molecule Thus when Cu+-ZSM-5 was probed with CO an adduct stable at room temperature was found*9u which showed a CO IR absorption band at 2157 cm I this is significantly lower than the corresponding value of 2178 cm for CO/Na-ZSM-5 Since the cation radii are in the order r(Cu+) < r(Na+) it is clear that electrostatic interactions alone would not explain the observed spectroscopic results Dinitrogen was also found296c to form adducts with Cu+ ions in Cu+-ZSM-5 and Cu+-MOR These adducts which are stable at 298 K showed an N-N stretching frequency of 2295-2299 cm I i e down-ward shifted with respect to free N The whole set of results can be rationalized in terms of chemical interaction involving molec- ular orbitals (MO) of the probe molecule and suitable d s p orbitals of the metal cation For CO cr-donation through the weakly antibonding 5uMO raises V(CO) while back-donation to the antibonding 27 MO lowers it The net effect is a small hypso chromic shift as compared to free CO For dinitrogen the corre- sponding MOs are 3crg which acts as an electron donor and the empty 1 7rg which is the electron acceptor 3O In this case both the a-donor and the 7-acceptor interactions weaken the N-N bond thus explaining the observed bathochromic shift in the Cu+ -.-N adduct Appropriate IR probe molecules can be used to test not only charge balancing cations but also their local environment This was shown in a comparative studyJ7 of H and CO adsorption for a series of zeolites having increasing ionicity i e decreasing Si/AI ratio Na ZSM-5 Na MOR Na-Y Na-X and the sodium form of Linde 4A Diffuse reflectance IR spectra of dihydrogen adsorbed at liquid nitrogen temperature on these zeolites are shown in Fig 8 a com plete assignment of these spectra can be found in the original article Briefly Na-ZSM 5 presents only a single IR absorption band (at 41 10 cm I) consistent with a single type of cation site For Na MOR two absorption maxima are observed 4108 and 4125 cm I as expected for two different cation sites (Section 3) The faujasite-type zeolites X and Y also show complex spectra which were explainedt731 in terms of the different cation sites (Fig 1) and their relative population Finally two IR absorption maxima are observed for Na-A Structural data for this zeolite locate cations mainly near the centre of the six membered rings forming the sodalite cages CHEMICAL SOCIETY REVIEWS 1996 (4 Figure 5 (a)The structure of ZSM-5 viewed along the straight channels. The Connolly surface (defined by blue dots) obtained with a probe molecule 0.28 nm in diameter clearly shows the second set of (sinusoidal) channels. (b)Enlarged view down a straight channel showing extraframework cations (Na+) with adsorbed CO molecules. (band at 4075 cm-I) but other less populated positions are also However for adsorbed CO the corresponding hypsochromic shifts available which are responsible for the shoulder at 41 10 cm I. The were found to increase17 in the opposite sense highest for Na-ZSM- important fact to be noted is that when the low-frequency bands are 5 and lowest for Na-A. These results were rationalized in terms of considered bathochromic shifts of the H-H stretching frequency the interaction modes discussed in Section 2.1. Dihydrogen is polar- (from the 4163 cm-' value of free H2) increase with increasing ized by a cation and a neighbouring anion (Scheme 2) and the ionicity of the zeolite framework (from Na-ZSM-5 to Na-A). bathochromic shift of the H-H stretching frequency measures the DIATOMIC MOLECULAR PROBES FOR MID-IR STUDIES OF ZEOLITES-A ZECCHINA AND C OTERO AREAN 4110 I I 2200 2175 2150 2125 wavenumber cm ' Figure 6 IR spectra of CO adsorbed at 77 K on M+ ZSM 5 (M = Na K Rb Cs) The shift of the peak maximum from Na- to CS' is clearly seen Note. however that all spectra correspond to a CO equilibrium pressure of lo2Pa For smaller CO doses all peaks shift to higher wavenumbers Corresponding values are given in the text The shoulder marked with an asterisk corresponds to a small fraction of Na+ still present after ion exchange (Further details can be found in ref 27a) 2175-2170-E23& 2185-5 fj21m-. ZSMB2155-cs+ MOR Figure 7 C-0 stretching frequency and corresponding electric field versus l/(Rx + R,,)2 for alkali metal exchanged mordenites and ZSM 5 zeo lites R = cation radius combined Lewis acid-base strength of the ion pair while the electro- static field sensed by the CO molecule is smaller when the negative contribution from framework anion is maximized c e in the zeolite having highest ionicity Note that the local structure comprising a cation and a framework anion constitutes a dual acid-base site of the Lewis type which plays an important role in many (catalytic) chem- ical processes mediated by zeolites Characterization of these dual sites is another example of the potential of IR spectroscopy using probe molecules 5 Extraframework Species and Lewis Acidity During thermal dehydration or steaming the framework of zeolites (and zeotypes) tends to break up partially and extraframework material is generated which remains entrapped in the zeolite chan- nels The first stages of this process are represented in Scheme 5 Subsequent hydrolysis leads to detachment of T(OH) or TOOH species (T = Al Ga Fe etc ) with attendant formation of a silanol nest Figure 8 Diffuse reflectance IR spectra of H,adsorbed at 77 K on (a)Na-ZSM 5,(6)and (c)Na MOR (6)and (e),Na A cf)Na X and (g)and (h) Na Y P(H,) = 0 3 kPa for (a),(b),(e),cf)and (g) or 13 kPa for (c) and (d) Spectrum (h)was taken after outgassing (77 K 1 min) the Na Y zeolite sample with preadsorbed hydrogen Reprinted with permission from ref 17 T i H Scheme 5 The T(OH) species thus formed can migrate and ultimately form after clustering and dehydration very small particles of a T,O phase Owing to steric constraints imposed by the zeolite framework the dimensions of these clusters can be so small as to have a large proportion of coordinatively unsaturated T ions which behave as strong Lewis acid centres The high-frequency bands observed in Fig 4(c& correspond to CO and to N coordinated to Al'+ ions in extraframework species For CO a single band is observed centred at 2225 cm I while di- nitrogen shows a more complex spectrum maximum at 2348cm and shoulder around 2357cm I The 2348cm band should core spond to A13+ -..N adducts in extraframework AI,O while the 2357cm shoulder is tentatively assigned to the N-N stretching of N interacting with highly exposed T* ions (partially extraframe work) in Scheme 5 These species presumably exist only in a small concentration (because of ensuing hydrolysis) and carbon mon oxide fails to detect them The enhanced specific absorbance of N when subjected to high electric fields would explain its higher sensitivity (as compared to CO) for detecting strong Lewis acid centres Fig 9(a) shows the relevant section of the IR spectra of CO adsorbed at 77 K on a sample of H-[GaIZSM-5 (Si/Ga = 25)pre-viously activated at 823K so as to cause partial degalliation of the framework 32a H-[GaIZSM-5 is analogous to H-ZSM-5 with gallium substituting for aluminium For comparison the spectra of 196 CHEMICAL SOCIETY REVIEWS 1996 0.050 0 15 0 10Y =.-v) C=I5 0.025 Q)0 ae 5 na 005 0.000 0 00 2 1 2240 2230 2220 2210 2200 2190 2 D 2200 2150 2100 2050 wavenumber cm” wavenumber em ’ Figure 9 (a)IR spectra of CO adsorbed at increasing equilibrium pressure (from to 20 Torr) on extraframework ‘Ga,O,’ species formed by partial degalliation of H-[GaJZSM 5 previously activated at 823 K in vacuo Spectra taken at liquid nitrogen temperature (b)CO adsorbed on finely divided gallium oxide other details as in part (a) CO adsorbed on finely divided Ga,O activated at the same tem- perature are also shown [Fig 9(b)] The main IR absorption band for Ga203 appears at 2190 cm I with a minor shoulder at 2220 cm-i The 2190 cm-I band corresponds to the C-0 stretching of Ga3+ .-CO adducts formed on extended surfaces of the metal oxide while the 2220 cm-I shoulder (which saturates at low CO equilibrium pressure) is assigned to similar adducts with more exposed GaB+ ions eg in comers of Ga203 crystallites This assignment was made by companson with IR spectra of CO adsorbed on y-alumina32b It is noticeable that for the zeolite sample [Fig 9(a)]the 2220 cm-I band completely dominates the spectrum (the slight shift towards lower wavenumbers on increas- ing the CO doses is due to adsorbate-adsorbate interactions) thus proving the highly dispersed nature and consequently high Lewis acidity of the extraframework species which do not seem to have undergone appreciable sintering The same conclusion (regarding extraframework aluminium species) was reported by Gruver and Fr1piat,2~~who studied Lewis acid sites in partially dealuminated mordenites As pointed out by these authors the high dispersion degree of extraframework species should favour interaction with framework Brgnsted acid sites and consequent synergy in catalytic processes 6 Summary and Outlook The use of diatomic molecular probes for rendering zeolites and zeotypes IR-readable has been outlined Information attainable using this technique concerns Brgnsted and Lewis acidity struc- tural defects internal electric fields and the siting of extraframe- work cations Specific examples were discussed which involve the most commonly used diatomic molecules CO N and H Work with molecular oxygen was also recently reported 33 By choosing appropriate molecular probes dual acid-base pairs can also be monitored as shown for a series of Na-exchanged zeolites Detailed characterization of these dual sites should facilitate a better understanding of many (catalytic) chemical processes medi- ated by zeolites Mid-IR spectroscopy of adsorbed molecules is a very active research field which contributes extensively to deepening our knowledge of the local structure of zeolite active sites and of the physico-chemical properties of the intra-zeolite space This has strong bearing not only on heterogeneous catalysis and pollution control but also on the potential use of zeolites as host materials for a variety of advanced composites which could find application in a number of technological fields such as selective membranes mod- ified electrodes chemical sensors or low-dimensional electrical conductors to name only a few examples Advances in zeolite characterization using IR molecular probes require detailed experimental research and precise knowledge about the spectroscopic response triggered by interaction of the probe molecule with the zeolite surface this implies a close interplay between experimental work and theoretical studies The full power of quantum-chemical calculations and computer modelling tech- niques13 33 34 is currently being applied (with very fruitful results) in conjunction with spectroscopic and structural research and further developments are expected However it should be acknowledged that elucidating all aspects of the structure and physico-chemical behaviour of zeolites is not an easy task and concurrent use of several techniques IS often required IR spectroscopy is but one of them paramount among the others are X-ray and neutron diffrac- ti~n,’~ high resolution electron micro~copy,3~~~ 3sLr and MAS NMR spectroscopy 35e-r Major contributions are albo expected from techniques based on synchrotron radiation (XANES EXAFS PEXA and time-resolved X-ray diffraction) particularly since they allow in situ studies of zeolite catalysts to be carried out under oper ating conditions 36 Altogether the considerable efforts which are being made for an improved characterizaton of zeolites should result in a better understanding of their behaviour and this should boost the many potential technological applications of these versa- tile materials DIATOMIC MOLECULAR PROBES FOR MID IR STUDIES OF ZEOLITES-A ZECCHINA AND C OTERO AREAN 197 Acknowledgements We thank our collaborators cited in specific references who have contributed substantially to most of the experimental work discussed in this review The Spanish Ministerio de Educacion y Ciencia (DGICYT) is gratefully acknowledged for the award of the sabbatical stay of A Z at the Universidad de las Islas Baleares 7 References 1 M Estermann. L B McCusker C Baerlocher A Merrouche and H Kessler. Nature 1991,352,320 2 Q Huo R Xu S Li Z Ma J M Thomas R H Jones and A M Chippindale J Chetn Soc Chem Commun 1992,875 3 M E Davis,Acc Chem Res ,1993,26. 111 4 R Szostak Molecular Sieves Van Nostrand Reinhold New York 1989 5 G D Stucky Prog Inorg Chetn ,1992,40,99 6 (a)G A Ozin,A KupermanandA Stein,Angew Chem Int Ed Engl I989,28,359,(b)G A Ozin,Adv Muter 1992,4,612 7 S Bordiga G Ricchiardi. G Spoto D Scarano L Carnelli ,A Zecchina and C Otero Arean J Chem Soc Faraday Trans ,1993,89,1843 8 M D Baker G A Ozin and J Godber Catal Rev Sci Eng ,1985,27 591 9 J W Ward in Zeolite Chemistrj and Catalcsis ed J A Rabo ACS Monograph no 17 I American Chemical Society. Washington DC 1976 10 N S Hush and M L Williams,J Mu1 Spectrosc 1974,50,349 11 G Pacchioni G Cogliandro and P S Bagus Int J Quant Chetn ,1992 42,11 I5 12 V Staemmler.Chem Phys 1975,7 17 13 L Koubi M Blain E Cohen de Lara and J M Leclercq Chem Phys Lett 1994,217,544 14 E Cohen de Lara and Y Delaval ,J Chem Soc Faraday Trans 2,1978 74,790 15 W Kutzelnigg,V StaemmlerandC Hoheise1,Chetn Phvs ,1973,1,27 16 I N Senchenya and V B Kazansky Kinet Katal 1988,29,1131 17 S Bordiga E Garrone C Lamberti A Zecchina C Otero Arean V B Kazansky and L M Kustov J Chem Soc Faradav Trans 1994,90 3367 18 A Zecchina S Bordiga G Spoto L Marchese G Petrini G Leofanti and M Padovan J Phys Chem 1992,96,4991 19 M Czjzek H Jobic A N Fitch and TVogt J Phys Chem 1992,96 1535 20 P A Jacobs and W J Mortier Zeolites 1982,2,226 21 (a) S Bordiga C Lamberti F Geobaldo A Zecchina G Turnes Palomino and C Otero Arean Lungmuir 1995 11 527 (6) V L Zholobenko M A Makarova and J Dwyer J Phvs Chem 1993,97 5962 (c)F Wakabayashi. J Kondo K Domen and C Hirose J Phys Chern . 1993,97 10761 22 F Geobaldo. C Lamberti G Ricchiardi S Bordiga A Zecchina G Turnes Palomino and C Otero Arean J Phys Chem 1995,99,11167 23 (a)H Knozinger Proc Int Svmp Catal Sapporo 1988,ed K Tanabe et a1 ,VCH Weinheim 1989,146 (b)L Kubelkova S Beran and J A Lercher Zeolites 1989,9,539,(c)M A Makarova K M At Ghefaili and J Dwyer J Chem Soc Farudav Trans 1994 90 383 (d) V Gruver and J J Fripiat J Phvs Chem 1994 98 8549 (e) K M Neyman. P Strodel S P Ruzankin N Schlensog H Knozinger and N Rosch Catal Lett 1995,31,273 24 E A Paukshtis and E N Yurchenko React Kinet Catal Lett 1981 16,131 25 M A Makarova,A F 0j0,K Karim,M Hungerand J Dwyer,J Phys Chem ,1994,98,3619 26 S Bordiga E Escalona Platero C Otero Arean C Lamberti and A Zecchina,J Catal 1992,137,179 27 (a)A Zecchina S Bordiga C Lamberti G Spoto L Carnelli and C Otero Arean J Phys Chem 1994 98 9577 (6) C Lamberti S Bordiga F Geobaldo A Zecchina and C Otero Arean,J Chem Phys 1995,103,3158 28 (a) B Barrachin and E Cohen de Lara J Chem Soc Faraday Trans 2,1986,82,1953,(6)H BoseandH Forster,J Mol Struct ,1990,218 393 29 (a) G Spoto A Zecchina S Bordiga G Ricchiardi. G Martra G Leofanti and G Petrini Appl Catal 1994 3B 150 (h)G Spoto S Bordiga G Ricchiardi D Scarano A Zecchina and F Geobaldo J Chem Soc Faraday Trans 1995 91 3285 (c) Y Kuroda Y Yoshikawa S Konno H Hamano H Maeda R Kumashiro and M Nagao,J Phys Chem 1995,99,10621 30 P Pelikan and R Boca Coord Chem Rev 1984,55,55 31 L M Kustov and V B Kazansky J Chem Soc Faraday Trans ,1991 87,2675 32 (a)C Otero Arean G Turnes Palomino F Geobaldo and A Zecchina J Phys Chem 1996,100,6678,(b)A Zecchina E Escalona Platero and C Otero Arean J Catal ,1987,107,244 33 (a)F Jousse and E Cohen de Lara,J PhyJ Chem ,1996,100,233 (6) F Jousse A V Larin and E Cohen de Lara J Phys Chem ,1996,100 238 34 (a)G M Zhidomirov and V B Kazansky Adv Catal 1986,34 131 (b)V B Kazansky Catal Today 1988,3,367,(c)J Sauer Chem Rev 1989,89 199,(6)L Kubelkova S Beran and J A Lercher Zeolites 1989,9,539 (e) Modelling of Structure and Reactivity in Zeolites ed C R A Catlow. Academic Press London 1992 v) J 0 Titiloye S C Parker F S Stone and C R A Catlow J Phys Chem ,1991,95,4038 (g) S J Cook A K Chakraborty A T Bell and D N Theodorou,J Phy Chem 1993,97,6679 (h) A G Pelmenschikov and R A van Santen J Phys Chem 1993,97. 10678 (1) G J Kramer R A van Santen C A Emeis and A K Nowak Nature 1993,365,529,Q)J D Gale,C R A Catlow and J R Carruthers,Chem Phys Lett ,1993,216 155 (k) D M Bishop Int Rev Phvs Chem 1994 13,21 (1) D M Bishop and S M Cybulski Chem Phys Lett 1994,230 177 (m) R A van Santen and G J Kramer Chem Rev 1995,95 637 (n) A M Ferrari P Ugliengo and E Garrone J Chem Phvs submitted for publication 35 (a) L B McCusker Acra Crvstallogr Sect A 1991 A47,297 (b)G Vitale. L M Bull B M Powell and A K Cheetham J Chem Soc Chem Comtnun 1995,2253 (c)L M Bull. A K Cheetham B M Powell J A Ripmeester and C I Ratcliffe J Am Chem Soc 1995 117,4328 (d)0 Terasaki and T Ohsuna Cataf Today 1995,23,201 (e) J M Thomas Stud Surf Sci Catal 1989,49A,3 v) G Engelhardt and D Michel High Resolution Solid State NMR of Silicates and Zeolites Wiley Chichester 1987 (g)J Klinowski Annu Rev Muter Sci 1988 18 189 (h) G Mirth J Lercher M W Anderson and J K1inowski.J Chetn Soc Faraday Trans ,1990,86,1039 (1) H Koller B Burger A M Schneider G Engelhardt and J Weitkamp Microporous Muter 1995,5,2I9 36 J M Thomas and G N Greaves Science 1994,265,1675

ISSN:0306-0012    

DOI:10.1039/CS9962500187

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Solid State Metathesis Reaction For Metal Borides, Silicides, Pnictides and Chalcogenides: Ionic or Elemental Pathways 1. P. Parkin Department of Chemistry, University College London, Christopher lngold Laboratory, 20 Gordon Street, London WCIH OAJ I Introduction A key focal point in inorganic and materials chemistry is the development of new routes to materials The conventional 'heat and beat' approach, which has been utilised since antiquity, requires starting materials (precursors) to be heated in a fire or oven followed by grinding and further reheating until the desired phase is pro- duced This method works extremely well and forms the basis for most of the ceramics industry New approaches that aim to alleviate the time and energy costs involved in conventional preparations include sol-gel and hydrothermal synthesis,' molecular precursors and self propagating high temperature synthesis (SHS) * Some of these methods aim to reduce the external energy input for the reac- tion by introducing mixing of the required elements on the atomic scale, and thereby overcoming the solid-state diffusion barrier This review will concentrate on a modification of the SHS process, termed a solid state metathesis reaction or SSM2 It will first address in sections 1 1 -1 5 the general features of SSM reactions initiation conditions, product purity, effect of diluents and thermal profiles It will then discuss in sections2 1-2 5 the range of materi-als obtained from metathesis reactions borides, silicides, pnictides and chalcogenides Finally the evidence collated from sections 2 I -2 5 will be interpreted in terms of a reaction mechanism in section 3 Future directions and conclusions will be summarised in section 4 Table 1 SHS and SSM reaction parameters Parameter Typical combustion temperature.TC/"C Velocity route, VC/cms Initiation power, Plcal cmz s Initiation duration, t/s SHS SSM 1400-4000 700-1400 0 1-10 0 1-10 10- 100 10-100 0 5-5 0 5-5 SHS and SSM reactions have a number of features in common (Table 1) Energy for the SHS process comes from the inbuilt chem ical energy of the reaction and relies on outside energy only for initiation The process is very rapid and reaches temperatures in the range 1000-4000 "C in a matter of seconds, cooling rates are dependent on reaction size -with larger scale reactions being better Ivan Parkin (born Portsmouth 1965) obtained his BSc (1986)and PhD (1989) from Imperial College His thesis entitled "Metalla SulfurlSelenium Nitrogen Heterocycles Reactions in Liquid Ammonia I' was obtained under the guidance of J D Woollins He then spent two years at Indiana University with M H Chisholm FRS on a NATO Fellowship before taking up a lectureship at the Open University Since I993 he has been a lecturer at University College London He has wide- spread research interests from SSM, SHSreactions, and CVD to solid state chemistry espe- cially frustrated magnetism 199 insulated against heat loss, however, compared to the 'heat and beat' approach, cooling rates are rapid SHS processes can be used to make a range of materials, although control of the temperature and formation of specific phases can be difficult In a typical SHS reaction, a fuel source and an oxidiser are required and the reaction is normally of a reduction-oxidation type This is exemplified by the well known thermite reaction (l), which has commercial application for the formation of welds (railway tracks) and in met allurgical separation Fe,O, + A1 -Fe + A1,0, (1) In this reaction iron is the oxidising agent and aluminium the fuel source The AH,,, is -850 kJ mol I per mole of Al and the maximum adiabatic combustion temperature (Tad) reached is in excess of 3000°C It has been found empirically that SHS reactions will proceed with a propagation wave, spreading out from a point of ignition, if the calculated Tadis greater than 1530 "C I Solid-state metathesis reactions are a variant of the SHS process, reactions are just as rapid, energy for the process comes from the reaction enthalpy, and as will be highlighted in this review, a range of materials can be obtained In a solid-state metathesis reaction an alkali metal or alkaline earth metal pnictide, chalcogenide, silicide or boride is reacted with a metal halide [reaction (2)] M,Ey + M'Xp -M'E4 + pMX (2)(M = Li,Na,K, Mg,Ca, Sr, Ba, E = B, Si, N, P, As, Sb, Bi, 0,S, Se.Te, M' = transition. main group or actinide metal X = halogen) Three specific examples of these reactions are given below [reac-tions (3-5)] TiCI, + Na,P -TIP + 3NaCI 2TiC1, + Ca,N, -2TiN + 3CaCI, (3) (4) TiCI, + 2Li20-+ Ti02+ 4Lic1 (5) The driving force is the co-formation of alkali or alkaline earth metal halide, which accounts for up to 90%of the reaction enthalpy Co-formation of a salt also has important consequences for the maximum reaction temperature and hence the phases of material obtained It should be pointed out that removal of co-produced salt is relatively straightforward and involves trituration with an appropriate solvent, such as water, tetrahydrofuran or acetone 1.I Maximum Reaction Temperature The maximum adiabatic reaction temperature, Tad,for a solid state metathesis reaction can be determined by calculation of the reaction enthalpy using Hess' law and correlating this with the heat capac ities of the reagents and products It is found that despite large reac tion enthalpies, Tadnormally correlates with the boiling point of the co-produced salt, thus limiting the maximum reaction temperature (1400 "C for LiCl co-production) In practice, although the reac- tions are relatively fast, some heat losses do occur to the surround- ings Salt mediation and thereby lower maximum reaction temperature for SSM compared to SHS reactions implies that a different range of materials are accessible CHEMICAL SOCIETY REVIEWS, 1996 (a) (4 (4 Figure 1 Photographs of the reaction of lithium nitride and titanium tetrachloride (a)before addition of Li,N, (6) 1 s after addition, (c) 3 s after addition Figure 2 Photographs of the filament initiated reaction of VCl, with a mixture of 1 Ca,N,-2Mg,N2 after (a)0 7, (b) 1 4,(c) 3 5,(d)4 9 s (Reproduced with permission from Chem Muter, 1995,7, 1728") 1.2 Initiation of the Reactions SSM reactions require some energy input to initiate the reaction In some cases this may come from addition of reagents at room tem- perature Grinding reagents together: touching reagents with a heated filament: heating of the bulk sample in a furnace9 and heat from a chemical reaction such as partial hydrolysis (addition of a drop of water) have all proved successful in starting the reaction In many cases initiation is accompanied by a brief thermal flash of 2-3 s duration Initiation is dependent on the ease by which the solid state diffusion barrier is overcome and consequently is related to reagent melting points Addition of lithium nitride to tita nium tetrachloride (liquid) at room temperature produces a spontaneous reaction on mixing (Fig 1) For low melting point solids the energy supplied by grinding of the reagents is often suf- ficient to initiate reaction, for example, reaction of GaCl, with Na,P to form Gap7 The two most common way5 to start an SSM reaction are filament initiation (often in a sealed bomb type reactor) and oven heating in a sealed evacuated glass ampoule Filament initiation is propagated by the passage of a synthesis wave through the solid compactg (Fig 2) In some instances a reaction can be initiated at the tip of the filament but fails to proceed through the whole of the solid lo I1 The alternative way to initiate the reaction is to heat bulk material in a sealed ampoule, in many of these cases a thermal flash occurs indicating a fast SSM reaction In others, notably the less exother mic reactions, such as formation of metal oxides, no thermal flash is produced and heating at 500-800 "C for 2-12 h is sufficient to drive the reaction Care must be exercised with any SSM reaction SOLID STATE METATHESIS REACTION FOR METAL BORIDES-I.P. PARKIN 201 (4 Figure 3 SEM micrograph of the product from the reaction of TiCl and Li ,N: ((I) before and (6)after trituration with water. in a sealed container as large, potentially explosive.gas pressures may be built up. 13 Product Formation The solid formed in SSM reactions is a composite of salt and product. Some salt -especially lithium chloride -will sublime during the course of the reaction; however, the morphology of the particles is consistent with formation of agglomerates of crystallites with a surface covering of the salt. Trituration of the material with water, tetrahydrofuran or methanol has been utilised to remove the adherent salt completely (to < 0.01%). Trituration can alter the nature of the material, especially with solvent-sensitive products.9 A typical SEM micrograph before and after trituration is shown in Fig. 3. Energy-dispersive X-ray analysis (EDXA) has proved to be a good technique for the analysis of the surface (top 1 -S pm) elemental composition.Fig. 4 shows the EDXA trace before and after trituration indicating ready removal of the co-produced salt. It is not always possible to determine crystallite size from SEM data. An average crystallite size is obtainable from the linewidths of the X-ray powder diffraction patterns via the Scherrer equation.' These indicate crystallite sizes typically in the range 300-700 A. Crystallite sizes are remarkable considering the short timescales of the reactions (2-3 s) and are due to high reaction temperatures, in some cases possibly helped by the co-formed alkali metal halide.'* Reaction scale and enthalpy are important factors in determining the crystallite sizes, with larger scale, high-temperature reactions favouring larger crystallites.1.4 Reaction Thermal Profile. The maximum temperature reached in SSM reactions has been determined by optical pyrometry and by direct thermocouple measurements.l Direct measurements involve placing a thermo- couple at the centre of a mix of powders and recording the reaction temperature with time. The trace obtained (Fig. 5) is similar to the 202 CHEMICAL SOCIETY REVIEWS. 1996 I ClI Ti 2oor--------// i500 1400 1 1300 i $ ' keV i keV Figure 4 EDAX trace from the reaction of TiCI, and Li,N (u)before tri turation (b)after trituration with water Tma,930-940 "C --f :: 1'""" :: 4800 -200 60°C--IllllllIII -036 28 20 12 4 0 tls Figure 5 Thermocouple trace from the reaction of HfCI, and Na,P thermal explosion mode observed in SHS reactions, in that the reaction is rapid A 0 5 g scale reaction reaches its maximum tem perature after ca 0 5 s and cools to room temperature in ca 12-20 s This rapid heating and cooling leads to high concentrations of defects in the material and should make it readily sinterable Optical pyrometry assesses reaction temperature by means of the colour of the thermal flash associated with the reaction This often varies from a yellow-white ( 1400"C) to a red-orange flash (800-1000"C),it works best for reactions that are most exothermic Use of differential scanning calorimetry (DSC) enables heat flows Figure 6 DSC trace from the reaction of TiCI, and Li,N axis heat flow in mW r axis time in minutes temperature can be extrapolated from the diagonal line ,..,.. .... .... .. 421 0 42 5 43 0 4$5 201O Figure 7 X Ray powder diffraction of the TIN (2 0 0) line (a)no added LiCl (b)with 200 mg of LiCl (c)with 400 mg of LiCl (reaction scale 100 mg of Li,N) associated with the reaction to be monitored (Fig 6) In most SSM reactions, a very sharp exotherm is observed corresponding to initiation and propagation In some SSM studies the reactions show a small but reproducible exotherm prior to the main reaction l4 15 The Effect of Diluents Reaction of anhydrous metal halides and alkali metal or alkaline earth metal pnictides and chalcogenides can be modified by the addition of an inert heat sink Addition of lithium chloride to a reac- tion mixture of TiCl, and Li,N reduces the crystallite size of the product, TIN (Fig 7) The diluent can absorb some of the reaction enthalpy and also increase the path length for diffusion of reacting components, thereby reducing overall reaction temperature and product crystallinity Ease of initiation can also be modified by use of a diluent, I e reaction of MoCI, and Li,N is initiated at room tem perature on mixing the components but requires heating to ca 120"C In the presence of 1000mol% of LiCl I6 One interesting case of a dilution experiment modifying the reaction IS that of WCI, and Li,N In the absence of dilution, only tungsten metal was observed, whilst on subsequent dilution significant amounts of WN are detected in the product by XPS (Fig 8) The lower overall reaction SOLID STATE METATHESIS REACTION FOR METAL BORIDES-I P PARKIN 60/i 5 40 0" (I)c.I I I I z 20 m I I I I I I - 0 44 42 40 38 36 34 32 30 Binding Energy/ eV would be expected (confirmed by the experiments) to thermally decompose to the metal and dinitrogen This phenomenon has been called the 'chromium enigma ' For group 3 and 4 metals a single phase of binary nitride, MN, is formed in the reaction, whilst for group 5 metals a mixture of MN and M,N is formed For chromium a mix of CrN and Cr,N is produced and for manganese, metal and a trace of Mn,N are obtained Isolated phases conform to the most thermally stable metal nitride and in no instances have nitrogen-rich phases, such as M,N, (M = Hf, Zr) or Ta,N,, been observed Instead an amount of nitrogen gas was also produced Ireaction (6)j Changing the nitriding agent from lithium nitride to calcium or magnesium nitride does not affect the phases of material produced, except that single phases of VN and V,N12 have been isolated 28 26 3HfCI, + 4Li,N -3HfN + 12L1C1+ 1/2N, (6) Oven-initiated reactions of sodium azide and anhydrous transi-tion metal halides also produce metal nitrides in good yield, with the same products as from comparable Li,N reactions 16 A comparison by X-ray powder diffraction (Fig 9) reveals broader line7 and hence less crystallinity for the NaN, reactions The increased amount of N, produced with NaN, [reaction (7)j tends to spread the product rather than forming a fused mass as in the Li,N reactions, thus cooling is more rapid and a less crystalline material is obtained It is also interesting to note that the degree of nitridation for small-scale reactions (ca 0 5 g) that are oven-initiated in sealed ampoules is not increased by using sodium azide Contrastingly, Kaner et a1 l50I-?-44 42 40 38 36 34 32 30 28 26 Binding Energy/ eV '\ I \b 20t 1I 1.L 44 42 40 38 36 34 32 30 28 26 Binding Energy/ eV Figure 8 Tungsten 4f region of the X ray photoelectron spectra for prod ucts of WCI, and Li,N with increasing dilution ((I) undiluted. (b)with 300 mg of LiCl added, (c) with 600 mg of LiCl added All reactions on a scale of 50 mg of Li,N temperature obtained by using a diluent enables the thermally sen-sitive tungsten nitride (decomp 650 "C) to be isolated 2 Preparation of Materials 2.1 Metal Nitrides A range of transition metal, lanthanide, actinide and main group nitrides have been made by solid-state metathesis reactions (Table 2) Starting materials have included Li,N, Mg,N,, Ca,N, or NaN, as nitrogen sources Reactions with lithium nitride have been the most extensively studied I5 Reaction of anhydrous transition metal halides with lithium nitride produces either transition metal nitride or elemental metal Thermal stability of the metal nitride determines the nature of the product Early transition metal nitrides, MN (M = Ti, Zr, Hf, V, Nb, Ta, Cr), all have decomposition points above 1900"C, well above the maximum temperature in the reaction (1400 "C) and hence nitrides can be obtained l7 Later transition metal nitrides (groups 7-12) have decomposition points below 1000"C Hence, even if a transition metal nitride were formed in these cases, it have found that filament-initiated reactions conducted in a modified bomb calorimeter on a larger scale (c-a 5 g) do lead to better product nitridation For example, reactions that form a mixture of Ta,N and TaN with Li ,N can be forced to form a single phase of product, TaN, by employing NaN, Presumably in a sealed bomb reactor, very high nitrogen pressures can be generated using NaN, that favour increaqed product nitriddtion l5 HfCI, + 4NaN, -HtN + 4NaCI + 11/2N, (7) The reaction of lithium nitride and anhydrous lanthanide and actinide chlorides in sealed ampoules produces metal nitrides in good yields l5 These reactions proceed with a thermal flash and require heating to ca 400 "C to induce initiation The lanthanide nitride phases produced were consistent with the formation of LnN, ,(x = 0-0 1) iao 0 Figure 9 Top trace X ray powder diffraction pattern from the reaction of Li,N and TiCI, Bottom trdce X ray powder diffraction pattern of the material from the reaction of NaN, and TiCI, (on the same scale as the top trace) Stick pattern is the standard pattern for TIN,Osbornite synthesis (Reproduced with permission from Polvhedron, 1995,14, 913Ih) Solid-state metathesis reactions yielding metal nitrides can be adapted to form solid solutions of mixed metal nitrides, providing there is some degree of lattice matching between the individual nitride phases For example, solid solutions of Ti,V,N (x + \.= 1 ) could be obtained6 (Fig 10) Similarly, making nitride solid solu-tions of lanthanides was possible if the lanthanides had similar ionic radii but it was difficult if that ionic radii were dissimilar-despite all lanthanide nitrides having a cubic lattice It was found CHEMICAL SOCIETY REVIEWS, 1996 Table 2 Materials isolated from solid state metathesis reactions. Where two phases are indicated the first listed phase was the most abundant, phases shown in brackets were present at less than 10% Data are included for both filament- and ampoule-initiated reactions Metal chacogenidel pnictide Li,Ny l5 Na,NI5 Ca,N," Na,PX IR Na,AsIX Na,Sbix Na,BiiX LI,O~~)*' Na,S" 24 Zh Li,Sezh Li,Tezh Mg2Siz9 Mg3BZi0 Metal halide TiCI, TIN TIN TIN TIP TIAS TiSb,,TiSb TI,BI Ti0, -" --TISI, -TiCI, TIN --TIP TiAs TiSb Bi Ti0, TIS,~ --Ti,Si, -ILi,TiO,I ZrCI, ZrN ZrN -ZrP ZrAs -Zr.Bi ZrO, ZrSzh --ZrSI, ZrB, Zr,S I HfCI, HfN HfN HfN HfP HfAs --HfO, HfSZh --HfSi, HfB, Hf2Si VCI , VN,[V,NI VN,V,N VN,[V,NJVP,[VP,J VAs --vo, VSzh ---VB, IVAS, I IL1 ,VO,l VCI, VN VAs -Bi ------NbCI, NbN, NbAs, NbSb, -LiNbO, NbSlh --NbSi,, -I NbJ I NbAs, Nb,Si, TaCI, TaN , TaN, T&N,TaN Tap TaAs, TaSb,, Ta$i LiTaO, TaS,b --Ta,Si, -ITa,N 1 ITaAs,l I Ta,S b, 1 -CrCI, Cr,N,Cr CrP.ICr,,P, I CrCI, Cr,Cr,N -CrAs MoCI, Mo MOP Mo,As,. MoAs MoCI, ---MoS,,MoS MoSe,, Mo,Te;, --Se( MoTe ,Ted WCI, W WP WAS, WS,? WSe,,Se< WTe,< W,WSi,,-WSSI.3 MnCI, Mn,N, ----MnO Mn FeCI, Fe FeP -FeSi -COCI, co COP NiCI, Ni Ni,P Ni0 NIS,~ -CUCI," cu--cu,o --ZnCl, Zn Zn,P, Zn,Sb, -ZnO ZnS ZnSe ZnTe Zn,Si --CdCI, Cd ---CdO CdS CdSe CdTe -HgCI, Hg -Hg,Sb -HgO HgS HgSe HgTe --LnCI,I LnN LnP LnSbL LaBi Ln,O, NaLnS, --LnSi,, LnB ILnsS1, I AcCli, AcN, Ac,P,, AC,Sb, -AcO, AcS, AcSe, AcTe, --AcP, 3 AcP 4N.3 AII, AIN A1 P AIAs,[Alj AlSb -GaX, Ga GaP GaAs GaSb,Sb -LiGaO, Ga,S3' GqSe,' Ga,Te,' -InCI, In InP InAs 1nSb.In -In,O, In,S, In,Se, In,Te, --SnI, Sn ----SnO, SnS, SnSe2 ----PbCI, Pb ----Pb,O,CI PbS PbSe PbTe -No data available h Solution phase metathesis Na,Se or Na,Te used ' Reaction dependent on conditions WCI, ' With Na,O, only Fe,O, formed 4 K,NiF, and Na,S, Note CuO formed by Na,O, or KO, with CuCl or CuCI, Ln = Y, La, Ce, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm Ac2' = Th, U For Ln = Yb, Eu, Eu,Sb, and Yb,Sb, found Li,S, Li,Se or Li,Te used difficult to form solid solutions where the individual nitrides had most cases a single phase of metal phosphide was formed with no different crystalline modifications Thorough premixing of the traces of elemental transition metal [reactions (8)-( lo)] reagents was also found to be important in forming solid solutions and was best achieved by using miscible liquid metal halides MCI, + Na,P -MP + 3NaCI (M = Y, La, TI, V, Mo) (8) 3MC1, + 4Na,P -3MP + 12NaCI + P (M = Zr, Hf, W) (9) 2.2 Metal Phosphides, Arsenides and Stibides" 3MC1, + SNa,P -3MP + 15NaCI +2P Reaction of anhydrous transition metal halides and Na,E (E = P, (M = Mo, Nb,Ta) (10) As, Sb) produces a range of metal phosphides, arsenides and stibidesi8 (Table 2).Reactions were typically conducted in heated sealed ampoules at Metal phosphide reactions all involved some reduction of the 500 "C, however, for the more volatile metal halides, TaCI,, NbCI, transition metal and formation of some elemental phosphorus In and MoCI,, the reaction was found to be initiated on light grinding SOLID STATE METATHESIS REACTION FOR METAL BORIDES-I 28P Figure 10 X Ray powder diffraction pattern of Ti,V,N formed from the reaction of xTiCI, and bVCI, with Li,N Standard patterns for TIN (-1 andVN( 1 with Na,P and did not require thermal assistance A similar pattern of products and initiation on mixing was found for reactions involv- ing Na,As, with MAS the most common phase For reactions with Na,Sb initiation was less facile and MSb,, transition metal and anti- mony were the major products Reaction of transition metal halides and sodium bismuthide did not form a single binary metal bis- muthide, only metal and bismuth The reactivity of sodium pnic- tides with transition metal halides thus alters on progressing down the pnictogen group, with formation of the elements more likely and metal pnictide phases less likely with the heavier congeners Lanthanide halides readily react with Na,E (E = P, As, Sb) at 500 “C in sealed glass ampoules to form LnE in good yield l8 A range of solid solutions, from mixed lanthanide compounds LnLn’E (Ln, Ln’ = lanthanides, E = P, As, Sb) through to lanthanide mixed pnictides LnEE’ (EE’ = PAS, PSb, AsSb) could also be formed in good yields Reactions of some main group halides with sodium phosphide were often spontaneous on mixing the powders and enabled formation of high purity GaP and InP8 23 Metal Oxides17 Solid-state metathesis reactions have been util ised for the formation of metal oxides One of the first examples was by Heppert and Wille in the preparation of mixed metal ferrites by the reaction of metal halides and lithium ferrite” [reaction (I I)] Recently reactions have concentrated on use of Li,O, Na,O,, Na,O and KO, with transition metal and lanthanide metal halides20 (Table 2) Most reactions have been studied in sealed evacuated ampoules and heated in an oven In these reactions a thermal flash has not been observed In some instances, notably the filament-initiated reactions of Na,O, with CuF, the reaction will proceed in the propagation mode with a syn- thesis wave of velocity 3 mm s I Both filament-initiated and oven-initiated reactions afforded products which were normally single phase metal oxides Products are quite predictable in that the metal maintains its oxidation state throughout [reactions (12)- (14~ MCI, + Ll20 -MO + 2L1X (M = Zn.Cd, Hg.Co. Ni, Mn) 2MCI, + 3Li20 -M,O, + 6LlCI (M =Y, La, Pr, Nd, Eu, Dy, Ho, Er,V) MCI, + 2L1,O -MO, + 4LiCI (M = Zr, Hf. W) (14) In all the reactions (12)-( 14) lithium was not observed to be incorporated in the product after trituration Reactions which involved tantalum and niobium, however, generated the technolog- ically important ferroelectric materials LiNbO, and LiTaO, Other reactions involving vanadium and titanium starting materials, espe- cially when using TiCI, and VCI,, produced minor phases such as Li,TiO, and Li,VO, in addition to the expected Ti0, and VO, If reactions of Li,O and MClfl are carried out in air, then the degree of lithium incorporation becomes significantly greater, such that Li,MO, (M = V, Nb, Ta) and Li,MO, (M = Ti, Zr, Hf) are P PARKIN 205 formed exclusively In the latter cases, it is difficult to determine if the process is a direct metathesis or a partial oxidation in air fol- lowed by lithium incorporation 22 Thermal studies by DSC for formation of metal oxides from Li,O and MCI, show a single exotherm around 290-340 “C dependent on the sample l4The exotherm is of 5 minutes duration and is thus atypical of a fast metathesis reacfion, but correlates well with the predicted reaction enthalpy Formation of mixed metal solid solu- tions using Li,O have had only limited success, probably as the reaction is not very exothermic Wide-scale application of metathesis routes to metal oxides is unlikely in preference to the number of other methods available for their synthesis, especially as the reactions have no obvious advan- tage over more traditional routes It could prove useful for those instances where an aerobic environment or use of a large heat source is not possible Additionally, SSM reactions have been found to produce materials with high surface areas 2.4 Metal Sulfides, Selenides and Tellurides Solid-state metathesis reactions used to form metal chalcogenides have concentrated primarily on formation of sulfide materials A range of transition metal ,23 lanthan~de?~actinide25 and main group metal sulfidesZh have been synthesised from the reaction of sodium sulfide (Na,S) with anhydrous metal halides (Table 2) Reactions can be thermally initiated inside a sealed ampoule or by the propagation mode by means of a heated filament The most active area of this research by R B Kaner et nf has been directed to the preparation of MS, (M = Mo, W) materials, layered dichalco- genides, which have important applications as lubricants and cathode materials 23 A range of mixed metal disulfides Ireaction ( 15)1 and metal dichalcogenides [reaction ( 16)I have also been formed from analogous reactions xMoC1, + YWCI,+ Na,S -MoxW,S, + NaCl (15) MoCI, + Na,S,,Se4 -MoS,Se, + NaCl (16) One of the most notable features of filament-initiated reactions which produce metal mixed dichalcogenides (MSISe,) is that the stoichiometry of the product does not reflect that of the starting materials, with enhancement of the more volatile sulfur relative to selenium in the product Filament-initiated reactions to form MoTe, and WTe, proved unsuccessful, however, the reaction could be ther- mally driven by heating in an evacuated quartz tube at 1000 “C for 7 days These findings have important consequences for determin- ing the reaction mechanism Reactions of lanthanide halides with sodium sulfide in some cases have formed metal sulfides, Ln,S,, as well as NaLnS, These reac- tions have been hampered somewhat by oxygen contamination 24 2.5 Metal Borides and Silicides Solid-state metathesis reactions have been used to form a range of transition metal and lanthanide borides and silicides lo29 Silicide reac- tions (from Mg,Si and MCIJ produce a range of phases from MSI, to M,Si They cannot be initiated by a heated filament and have required oven thermolysis at 800-900°C for 2-24 h Products are often con- taminated with the elements and can be directly related to the thermo- dynamically most stable phase Indeed, in this case, the reactions offer little advantage over conventional synthesis by direct combination of elemental powders Metal boride reactions (Mg,B, and MCI,) can be initiated in the filament mode I* 3 Reaction Mechanism Two extreme mechanistic pathways for solid-state metathesis reac- tions have been proposed a direct ionic metathesis [reaction ( 17)I and a reductive recombination route [reaction (18) 1 In direct ionic metathesis the reaction intermediates are ions, that interdiffuse In the ‘NaCl’ melt In a reductive recombination pathway the compo- nents are reduced to the elements followed by recombination promoted by the high reaction temperatures 206 MCl,] + NallE-Mn++ E + nNaCl -ME + nNaCl (17) MClll + NanE -M + E + nNaCl -ME + nNaCl (1 8) The actual mechanism could be intermediate between the two extremes The initially formed product could further decompose as a result of thermal instability, this is illustrated by the formation of HfN from HfCl, and Li,N as shown via an ionic route (I 9) and an elemental route (20) 3HfC1, + 4L1,N -3Hf4+ + 4N3 + 12L1C1 -(rherrwll\ ririrtuhlr) Hf,N, -3HfN + 1/2N, (19) (redrrtrrLe \rep) (re(omhination trep) 3HfC1, + 4LI,N -3Hf + 4N + 12LiC1-3HfN + 1/2Nz (20) The speed of fast metathesis reactions, which are over in ca 2-10s,makes it difficult to study the reaction directly and isolate inter- mediates Mechanistic discussions have been built up from end-product analysis In no reactions involving the lanthanide halides with lithium nitride, sodium phosphide, arsenide and anti- monide has any elemental metal been detected in the product If the reaction were to proceed by an elemental pathway of reduction fol- lowed by recombination it would be expected that some elemental metal would be detected Further, the lanthanide ions are stable (E" = -2 778 V vs Pt) in the presence of a strong reducing agent (E3 ) indicating that an ionic pathway is probably preferred 8 31 1I I .. . , . . .'.I,'.'!. . , . . . . : '1. . . , . . .,ib. I. . , . . . .-. , . . . . 25 30 35 q0 45 50 55 bB 65 78 281 Figure 11 X Ray powder diffraction patterns of the crude products of reac tions of (a)GdCl, with Mg,N, heated for 2 h at 500°C. (6)GdCl, with Mg,N2 heated at 900 "C. (c)TbCI, with Mg,N, heated at 900 "C Stick pattern represents Gd,CI <N, bold circles represent MgCI?, indexed Iines for LnN (Reproduced with permission from Chem Mater, 1995,7. 1728") Reaction of lanthanide halides with magnesium and calcium nitride produces lanthanide nitrides after heating at 900 "C for 1 h If the reaction is stopped prior to completion then it is possible to isolate intermediates of the form Ln,Cl,N (Fig ll), further heating of the mix sequentially lowers the Ln,CI,N concentration and increases the amount of LnN Detection of such intermediates strongly suggests an ionic mechanism Contrastingly, for transi- tion metal nitride reactions, using Li,N and NaN, as an alterna- tive nitrogen source and conducting the experiments in a bomb type container, significantly extra nitridation is introduced for the azide reaction This is to be expected if the reaction proceeds via a reductive recombination pathway because of the increased nitrogen pressures generated by the azide [reactions (6) and (7)] It is also worth noting that in filament-initiated transition metal nitride reactions some residual metal (ca 1%) is often found in the product, whilst for ampoule reactions in an oven, a negligible amount of metal is detected This is probably because the oven reactions are somewhat self-annealed and hence do not cool to room temperature as quickly Detection of metal powders in the product does point to a reductive recombination pathway for the transition metal nitride reactions, however, one caveat is that the metal halide precursor does not thermally decompose before it reacts For example, TiCI, is known to disproportionate at CHEMICAL SOCIETY REVIEWS, 1996 elevated temperatures to TiC1, and TiCI, (500"C), and the TiC1, to disproportionate further to Ti metal and TiCI, on additional heating It may be that these decomposition events lead to the traces of metal observed in the reaction and that the process is properly considered as an ionic metathesis Reactions to form metal silicides, antimonides, bismuthides and tellurides always seem to proceed by a reductive recombina- tion pathway, elemental metal and element are observed in high concentrations and the phases of material formed are determined by thermodynamics Further, reactions are thermally driven (12-24 h, at 800 "C) and do not proceed by a thermal flash Reactions of MCl,? and Na2(SrSe,) formed M(Sx+ Se, )2 (M = Mo, W, x+v = I) in good yield The products normally showed a measurable enhancement of the content of the more volatile sulfur compared to selenium and compositional gradients Both of these facts are best explained by a reductive recombination pathway Solid-state metathesis reactions can thus be thought of as pro- ceeding via either a reductive recombination pathway or an ionic metathesis route or both Some points are clear-lanthanide halides prefer the ionic metathesis route, whilst transition metal halides can prefer either, sodium salts of the heavier pnictides and chalcogenides (Sb, Bi, Te) and magnesium silicide seem to follow a reductive recombination pathway exclusively A full and detailed understanding of the reaction mechanism will require some sort of zn situ probe to monitor the reaction, for fast metathe- sis reactions this is obviously very difficult as the process is com- plete in only a few seconds One experiment that has been tried is addition of a different powdered elemental metal to a reaction mixture to note if that element has become involved in the reac- tion Incorporation would favour a reductive recombination pathway Analysis of reaction enthalpy and correlation with the melting point of the coproduced salt has allowed reaction propagation to be determined in the filament-initiated mode Molten co-produced salt is important in breaking down the solid-state diffusion barrier If the reaction enthalpy is not sufficient to melt the salt then spread of the reaction to intervening layers would be impeded and propagation would fail Propagation occurs when the calculated adiabatic tem- perature for the reaction, Tadb, is sufficient to melt a fraction of the coproduced salt lo It was found that reaction of VCI, with Mg,N, Has a Tad of 78 "C and no propagation occurred, reaction of VCl, and Ca,N, has a Tad of 1338 "C and propagation was rapid, whilst a mixture of 2/3 Mg,N2- 1/3 Ca,N, with VCI, had a calculated T,, of 7 14"C, the melting point of MgCl,, and the reaction just propa- gated (Fig 2) I 4 New Directions in Metathesis Reactions One of the most interesting new directions in solid-state metathesis reactions has been shown by Wells and co-workers 27 They per- formed reactions in the presence of a solvent in which the alkali metal pnictide had limited solubility, and the metal halide (GaCI,, InCl,) was soluble and coordinated by a range of different solvent and co-ligating molecules The metathesis reaction proceeded in much the same way as in the solid state, except that reactions required 2-12h at reflux and, most importantly, products Gap, InP and GaAs were nanoparticulate with a relatively narrow size distri- butions The size of the nanoparticles could be altered simply by changing the solvent or co-ligating molecules Nanoparticles have a very large surface area to size ratio and can show quantum confinement effects that could be utilised in a new generation of electronic devices It has also been possible to develop new reactions that are inter- mediate between solid-state metathesis and a molecular precursor approach to ceramics Reaction of anhydrous metal halides and lithium amide at 300-350 "C induced a reaction with a thermal flash This produced a material that was rich in nitrogen and after gentle thermolysis produced a metal nitride [reaction (2 I)\ 3m-350"C 700°C 2h MClll + LiNH, -MN,H,CI, + LiCl +MN SOLID STATE METATHLSIS REACTION FOR METAL BORIDES-I These reactions are less exothermic than those of lithium nitride.and nitrogen-rich phases such as Ta,N, have been isolated, as well as thermally sensitive phases such as Zn,N, Solid-state metathesis reactions have also been extended to the synthesis of electroceramtc materials such as barium titinate, from the reaction of barium peroxide and TiCI, Further work has also allowed the formation of metal borides, main group nitrides, e BN, and composite materials such as BN/TiN Products of SSM reactions are usually very pure with little chem- ical contamination from the walls of the reaction vessel, as reactions are fast compared to traditional ceramic techniques The inbuilt heat brake afforded by the co-produced alkali metal halide regulates the maximum reaction temperature and allows for isolation of phases not accessible from conventional SHS reactions SSM reactions can also be initiated by means of a heated filament, which means that access to high-temperature furnaces is not required, this should prove useful tor small laboratory-scale preparations of binary ceramic materials The SSM reactions are important not only because of the speed of the reaction but because they can engender metastable productv of unusual inicrosti ucture A full and detailed investigation of the product microstructure has not so far been achieved and scope for extensive TEM work exists Average crystallite size has been deter mined by the Scherrer equation but a detailed analysis including size distributions and domain structure has yet to be achieved Particulate morphologies have been hinted at by SEM which nor- mally reveals agglomerates of crystallites In this context the work of Wells, in both determining microstructure by high resolution TEM and in manipulating crystallite six by use of a solvent heat sink.IS particularly noteworthy 27 In 5ome cases SSM reactions can provide access to new phases of materials and to better purity than conventional methods For example. the ieaction5 of TaCI, with Li3N and NaN, produce the meta5table high-temperature cubic phase TaN rather than the expected hexagonal TaN Traditionally cubic TaN require\ high nitrogen pressures (20-160 atm), long time periods and tempera- tures in excess of 1600°C for it to form The reactions also have future potential to make composite materials by reactions with layers of different composition Solid state metathesis reactions offer an alternative, relatively straightforward and in some cases advantageous synthetic route to a vast range of inorganic ceramic materials 5 References 1 lnrightr rtito Spec iuliti Inorgunit Chetnicul\.ed D Thompson. Royal Society of Chemistry. Cambridge. 1996 2 H C Yi and J J Moore, J Muter Sc i . 1990.25,1189. V I Yukhvid. Plrre App1 Chetn . 1992, 64 977 J B Wiley. P R Bonneau, R E Treece. R F Jdrvis. E D Gillan. L Rao and R B Kaner. ACS Sitnp Ser .1992.499 369 P PARKIN 207 3 B Z Shakhashirt .Chetnicul Detnon5trutionJ, uHundhook for Teacher5 ofChetnistr\. vol 1, University of Wisconsin Press, London, 1983, p 85 4 CRC Hundhook of ChenziJtrb and PhwcJ, ed R C West, CRC, Boca Raton, Florida, 1983 5 J B Wiley and R B Kaner, Science, 1992,255.1093, R E Treece, G S Macala. L Rao. D Franke. H Eckert and R B Kaner. Inorg Chem . 1993,32,2745 6 A L Hector and I P Parkin. J Chetn Sot-. Chetn Comtnun . 1993, 1096 7 R E Treece. G S Macala and R B Kaner. Chern Muter, 1992,4,9 8 R E Treece, J A Conklin and R B Kaner, ltiorg Chern , 1994,33, 5701 9 J C Fttzmaurice, A L Hector and I P Parkin, J Chetn Soc Dalton Tram . 1993.2435,Polyhedron. 1993.12,1295 10 L Rao, E G Gillan and R B Kaner, J Murer Rer .1995,lO.353 11 A L Hector and I P Parkin, Chern Murer . 1995.7. 1728 12 R E Treece, E G Gillan and R B Kaner. Cotntnentr lnorg Chetn , 1995,16.313 13 A L Hector. G Henshaw. A V Komdrov and I P Parkin, Pmceedrnqs ofthe lnternutionul Conference rn Adwinces in Muteriulr rind Muteriult Protersing Technolog\. vol 1, Dublin City IJniverstty Press, 1995.n 153 14 A T Rowley and I P Parkin. Imrg Chiin Am. 1993,211,77 15 E G Gillan and R B Kaner, Inorg Chrtn 1994,33. 5693. J C Fitzmaurice, A L Hector. A T Rowley and I P Parkin, Polihedron. 1994.13,235.R E Treece. G S Macala. L Rao, D Franke, H Eckert and R B Kaner. /not-g Chern . 1993.32.2745,L Rao and R B Kaner. Inorg Chetti , 1994,33,3210 16 A 1, Hector and I P Parkin, Pohhedron.1995, 14,913 17 F A Cotton and G Wilkinson Adiuriced Inorgunit Chetnirrri, 5th edn , Wiley. Chichester, 1988 I8 A L Hector and I P Parkin. Z Nuturforwh, TeiI B. 1994,49,477.A L Hector and I P Parkin. J Muter Chein , 1994. 4. 279. J C Fitzmaurice, I P Parkin and A T Rowley. J Muter Cherrr I 1994, 4, 285.A T Rowley and I P Parkin, J Muter Chetii , 1993.3.689 19 S Heppert and A Wille.2 Phi5 Chrtn . 1932,18B. 291 20 A L Hector and 1 P Parkin, Polvhedron, 1993, 12,I855 21 J B Wtley, E G Gillan and R B Kmer. Muter Re\ Bull, 1993.28, 893 22 G Coombe and I P Parkin, unpublished results 23 P R Bonneau. R F Jarvis and R B Kaner. Inorg Clwtn . 1992. 31. 2127.P R Bonneau. R K Shi bao and R B Kaner, Inorg Chetn , 1990. 29,251I,P R Bonneau,R F JarvisandR B Kaner,Natiirr.1991,349, 510. J C Fitzmaurice. A L HeLtor and I P Parkin, Muin Group Metul Chern . 1994,17, 537. R R Chianelli and M B Dines. lnorg Chern , 1979,10,2758 24 J Cotter, J C Fitzmaurice and I P Parkin, J Muter Chetn .1994,4,1603 25 J C Fitzmaurice and I P Parkin, Nru J Chern , 1994,18,825 26 I P Parkin and A T Rowley. Polihedron. 1991, 12,2961. J C Fitzmaurice and I P Parkin, Muin Group Mtml Chetn . 1994 17.7 27 S S kher and R L Wells Chetn Muter. 1994,6.2056 28 I P Parkin and A T Rowley.Ad1 Muter. 1994.6.780,I P Parkin and A T Rowley. J Muter Clwtn I995.5,909 29 J C Fitzmaurice, A L Hector. I P Parkin and A T Rowley. Phorphorii' Sulfur Silicon 1995.101,47

ISSN:0306-0012    

DOI:10.1039/CS9962500199

出版商:RSC    

年代:1996

数据来源: RSC