Advances in Physical Organic Chemistry, Volume 30

Advances in Physical Organic Chemistry ADVISORY BOARD W. J. Albery, FRS University of Oxford A. L. J. Beckwith The Au...

Author: Donald Bethell

34 downloads 1166 Views 9MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Advances in Physical Organic Chemistry

ADVISORY BOARD W. J. Albery, FRS University of Oxford A. L. J. Beckwith The Australian National University, Canberra R. Breslow Columbia University, New York L. Eberson Chemical Center, Lund H. Iwamura Institute for Fundamental Research in Organic Chemktry, Fukuoka G. A. Olah University of Southern California, Los Angeles Z . Rappoport The Hebrew University of Jerusalem P. von R. Schleyer Universitat Erlangen-Nurnberg G. B. Schuster University of Illinois at Urbana-Champaign

Advances in Physical Organic Chemistry Volume 30

Edited by

D. BETHELL The Robert Robinson Laboratories Department of Chemistry University of Liverpool P.O. Box 147 Liverpool L69 3BX

ACADEMIC PRESS Harcourt Brace & Company, Publishers London San Diego New York Boston Sydney Tokyo Toronto

ACADEMIC PRESS LIMITED 24128 Oval Road London NWI 7DX

United States Edition published by ACADEMIC PRESS INC. San Diego, CA 92101 Copyright 0 1995 by ACADEMIC PRESS LIMITED All rights reserved

No part of this book may be reproduced in any form by photostat, microfilm, or any other means. without written permission from the publishers

A catalogue record for this book is available from the British Library ISBN 0-12-033530-1 ISSN 0065-3160

Typeset by Keyset Composition, Colchester. Essex Printed and bound in Great Britain by T.J. Press Ltd. Padstow, Cornwall

Contents

Preface

vii

Contributors to Volume 30

ix

Matrix Infrared Spectroscopy of Intermediates with Low Coordinated Carbon, Silicon and Germanium Atoms

1

VICTOR A . K O R O L E V

AND

OLEG M. NEFEDOV

1 2 3 4

Introduction 1 Carbenes and their silicon and germanium analogues 7 Free radicals 32 Conjugated organic radicals: allyl, propargyl, benzyl and cyclopentadieiiyl types 37 5 Unstable compounds with double-bonded silicon and germanium atoms (silenes, silanones, germanones, germathiones) 45 6 Conclusions 56 Acid-Base Behaviour in Macrocycles and Other Concave Structures

63

ULRICH LUNING 1 2 3 4 5 6

Introduction 63 Bases on the inside 65 Acidic centres on the inside 86 Macrocycles with both acidic and basic functionalities Hydrogen bonding 107 Closing remarks 110

103

Photodimerization and Photopolymerization of Diolefin Crystals

MASAKI H A S E G A W A 1 Introduction - history of topochemical [2 + 21 photoreactions V

117

117

CONTENTS

vi

+

2 Characteristic features of topochemical [2 2) photoreactions of diolefin crystals 121 3 Effects of wavelength of irradiating light and irradiation temperature 134 4 Kaleidoscopic topochemical behaviour of diolefin crystals 142 5 Topochemical reactions of mixed crystals, inclusion complexes and molecular complexes 162 6 Concluding remarks 167 Ionic Dissociation of Carbon-Carbon u Bonds in Hydrocarbons and the Formation of Authentic Hydrocarbon Salts

KUNIO OKAMOTO, KEN’ICHI TAKEUCHI TOSHIKAZU KITAGAWA

173

AND

1 Introduction 174 2 Stability of hydrocarbon ions 176 3 Ionic dissociation of the carbon-carbon u bond in hydrocarbons 184 Isolation of hydrocarbon salts 200 Physical properties of hydrocarbon salts 204 Chemical behaviour of hydrocarbon salts in solution 206 Control of hydrocarbon salt formation 212 Conclusions 216

Author Index

223

Cumulative Index of Authors

233

Cumulative Index of Titles

235

The thirtieth volume in the series marks the completion of over three decades of endeavour by more than 170 contributors from laboratories in 19 different countries to bring over 120 selected areas of progress in physical organic chemical research to the attention of readers worldwide. Its publication provides an occasion to thank them and also to reflect on the changing times and priorities of research in this field, and to contemplate the future. Victor Gold founded Advances in Physical Organic Chemistry in 1963 with the intention of providing a medium for the broad development of ideas concerning the quantitative study of organic compounds and their chemical behaviour. He aimed to do this by the publication of authoritative reviews of significant research areas, written as far as possible by leaders in those areas. This was an important development since, at that time, despite much research activity, the subject had no recognized specialist publications, no recognized review media and no national or international organizations to promote such research. This is no longer the case; IUPAC established its Commission on Physical Organic Chemistry in 1972 and this has worked to regularize matters of nomenclature and to organize a continuing series of highly successful biennial international conferences. Other international conference series have also developed which, while not always using the term in their titles, are devoted to physical organic chemistry. Successful specialist journals now exist for the publication of research papers. Since the establishment of Advances in Physical Organic Chemistry, the application of physical organic principles and methods has grown ever wider. The subject was always more than just the study of organic reaction kinetics and mechanisms, important though that continues to be. Now, however, ideas deriving from physical organic studies guide thinking in such diverse areas of scientific endeavour as pharmacology, organic synthesis and the design of materials and effect chemicals. The available research methods in the field have also been transformed. The ready accessibility of, in particular, nmr spectroscopy at increasingly high field, and of powerful computational methods, have wrought dramatic changes not only in what is being discovered but also in the questions that can be asked. Advances in the achievable resolution in both space and time. have transformed our ability to examine the fundamentals of organic chemical processes. Such developments bring with them new specialisms, but it is important that the power of new methodologies should be presented to a wider audience within the broader context of physical organic chemistry. VII

viii

PREFACE

With so much research activity that is physical organic in nature, the need for lively, yet careful critical discussion of aspects of the subject of current importance continues, not only for the health of the subject itself, but also in order to assist in the provision of well trained scientists possessing that important combination of skill in quantitative physical investigation with a feel for the relationship between the structure and behaviour of organic compounds. Advances in Physical Organic Chemistry will continue to play its part in satisfying that need. Consciousness raising is less important now. The principal aim of the series is, as it was from the outset, to facilitate access to areas of application of underlying physical chemical principles to the behaviour of organic substances in diverse situations of interest and potential value. The Editor and his Advisory Board continue to welcome offers and suggestions from the chemical community that will help the series to achieve that objective. D. BETHELL

Contributors to Volume 30

Masaki Hasegawa Department of Materials Science and Technoiogy, Faculty of Engineering, Toin University of Yokohama, 1614 Kurogane-cho, Midori-ku, Yokohama 225, Japan Toshikazu Kitagawa Division of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan Victor A. Korolev N. D . Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky prospect 47, 117 913 Moscow, Russian Federation Ulrich Luning Institut fur Organische Chemie, Christian-AlbrechtsUniversitat zu Kiel, Olshausenstrasse 40, D-24098 Kiel, Federal Republic of Germany Oleg M. Nefedov N. D . Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky prospect 47, 117 913 Moscow, Russian Federation Kunio Okamoto Meisei Chemical Works Ltd, 1 Nakazawacho, Ukyo-ku, Kyoto 615, Japan Ken'ichi Takeuchi Division of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan.

ix

This Page Intentionally Left Blank

Matrix Infrared Spectroscopy of Intermediates with Low Coordinated Carbon, Silicon and Germanium Atoms VICTORA. KOROLEVAND OLEGM. NEFEDOV

N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation

1 Introduction Principles of matrix isolation experiments Methods of generation of organic intermediates Spectroscopic methods of study of matrix-isolated molecules 2 Carbenes and their silicon and germanium analogues 3 Free radicals 4 Conjugated organic radicals: allyl, propargyl, benzyl and cyclopentadienyl types 5 Unstable compounds with double-bonded silicon and germanium atoms (silenes, silanones, germanones, germathiones) 6 Conclusions References

1 2 4 6 7 32 37 45 56 56

1 Introduction

One of the main tasks of physical organic chemistry is to study the mechanisms of chemical reactions by instrumental methods. The rapid development of various techniques and new spectroscopic methods in recent years has attracted attention to the investigation of elementary steps of reactions and the intermediates involved. In accordance with modern requirements, the description of reaction mechanisms should include the participation of relatively stable species. Direct detection and investigation of the intermediates are of great importance, not only for the solution of mechanistic tasks, but also for studies of their structure. As a rule these intermediates have unusual structures, open electronic shells, delocalized unpaired electrons and new types of chemical bonds. That is why their investigation sets new problems for the general theory of chemical structure. However, the typical intermediates of organic reactions (e.g. free radicals, carbenes, etc.) are highly reactive species and their concentration in reaction media under normal conditions is below the detection limits of usual 1 ADVANCES IN PHYSICAL ORGANIC CHEMISTRY ISBN 0-12-033530-1 VOLUME 30

Copvright 0I997 Academic Press Limited All rights uf repruducrion in anv Jorm reserved

2

V. A. KOROLEV AND 0 . M. NEFEDOV

techniques. There are some ways of overcoming these difficulties in the case of gas-phase reactions. One of the ways consists in carrying them out in vacuum (to prevent recombination) and using highly sensitive detection methods, e.g. mass spectrometry, photoelectron and laser spectroscopy. The other consists in freezing the reaction products at low temperatures until they accumulate in an amount sufficient to be investigated by standard techniques [electron spin resonance (esr), infrared (IR) and ultraviolet (UV) spectroscopy]. The very low temperatures (4-20K) used in this method allow reactive intermediates to be condensed on a target with an excess of an inert gas which can form a solid transparent matrix. Reactive species which are isolated in an inert matrix of this kind do not participate in secondary reactions and therefore persist. We shall discuss here results concerning the application of lowtemperature matrix IR spectroscopy to thermal and photochemical studies of reactions of organic and organometallic compounds leading to the formation of reactive intermediates such as free radicals, carbenes, and their silicon and germanium analogues, as well as compounds with double-bonded silicon and germanium atoms. But first, in order to provide for nonspecialists an outline of the methodology of matrix isolation and matrix spectroscopy, we briefly recount the basic principles of the method which has been well described by Cradock and Hinchcliffe (1975).

PRINCIPLES OF MATRIX ISOLATION EXPERIMENTS

During the last 40 years, matrix isolation has undergone considerable changes due to the modernization of experimental techniques. Numerous methodological variations using the new possibilities provided by modern spectrometers and cryogenic systems are available at present. These differ in the specific methods of generation and investigation of various chemical objects. However, there are some general principles which are basic for all experiments using matrix stabilization. These are: the group of inert gases which are used as the matrix substance; the materials used for targets; the regions of low temperature; the concentration which is needed for effective isolation of reactive species in a matrix. In comparison with hydrocarbon and polymeric matrices, which have their own absorptions in the IR and can react chemically with the intermediates, inert gas matrices are free of these shortcomings. Neon, krypton and xenon have been used as matrix substances in some studies. However, only argon and nitrogen matrices are widely adopted because of the availability of the pure gases and the fact that there is a variety of cryostats that can provide the optimal temperature conditions for the formation of rigid and transparent matrices from these elements. To obtain a rigid matrix, the temperature of the target should not be

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

3

above 0.3T, (where T , is the melting point of the matrix gas). However, decreasing the temperature below O.lST, leads to the formation of a matrix having a somewhat less regular crystal lattice and to the deterioration of the optical properties of the matrix. Thus, the choice of matrix substance is connected with the temperature to which a target is cooled, i.e. with the method of use of the cooling agent. In the initial period of development of the matrix isolation technique (from 1956 to the end of the 1960s), cryostats with liquid helium (4.2 K) or hydrogen (20 K) baths were used. However, in such cases it was difficult to increase the target temperature quickly enough (to permit diffusion in the matrix) and then to recool the matrix. Such “annealing of the matrix” is an important operation which is necessary for the spectral monitoring of the disappearance of reactive species as a result of their secondary reactions. Thus cryostats cooled with a continuous flow of liquid or gaseous helium which provide a smooth temperature variation were more convenient. But the most important period came only with the wide application of closed-cycle microrefrigeration systems. This equipment released scientists from the purchase, transportation, storing and filling up of cryostats with the expensive and rare helium, as well as from problems connected with the collecting and the utilization of the evaporated coolant. As to the concentration of unstable species in a matrix, long-term experience shows that the minimal dilution of the intermediates by the excess of matrix gas should be about 1000:1, although in some cases this value must be even higher. In an inert gas matrix, which has an atomic crystal lattice, molecules are not as well isolated as in vacuum. Spectral bands of molecules may change due to the influence of the matrix surroundings. In particular, the number, the position and the shape of IR peaks in gas-phase spectra may differ from those in matrix spectra. For example, the appearance in matrix IR spectra of several peaks located close to each other may be assigned to one and the same vibration (matrix splitting), as a result of molecules being placed in different site types in the crystal lattice of the matrix. The appearance of the new bands in matrix spectra may also be connected with a decrease of the molecular symmetry of the species. This phenomenon takes place when the molecule is frozen in an irregular matrix (e.g. at a high condensation rate). In this case some additional bands are assigned to fundamental modes which are inactive at higher symmetry. The bands of matrix-isolated molecules are frequently observed at the wavelengths which differ from those in gas-phase spectra. These matrix shifts are induced by the repulsive and attractive forces between the isolated molecules and the atoms which form the matrix site. Repulsions lead to small increases (1-15 cm-’) of * vibrational frequencies, and attractions decrease them. Matrix shifts depend on the type of matrix gas; they rise in the sequence from neon to xenon. In general, the shifts are positive (the

V. A. KOROLEV AND 0. M. NEFEDOV

4

vibrational frequency increases in the matrix) in the region below 1000 cm-', but are negative above 1000 cm-'. Nevertheless, the frequency values of free and matrix-isolated molecules usually differ negligibly. This fact indicates that there are minimal changes of their geometry and force fields on their isolation in inert gas matrices. The choice of cooled target depends on the kind of spectral irradiation employed. The plates of traditional monocrystals of alkali metal halides (plastic CsI is preferred) are used for the analysis in the IR region. Metallic mirrors, which reflect the IR beam to give a recurring passage through the matrix layer and thence to the detector of the spectrometer, are widely adopted at present. This method has two important advantages: the absence of a thermal gradient along the target, which has good thermal conductivity, and higher sensitivity due to double beam passage through the matrix. But in this case an additional reflection unit is needed in the sample chamber of the spectrometer. Similar reflection plates are used for recording ultraviolet-visible and Raman spectra of matrix isolated molecules, although the traditional beam path passing through transparent quartz windows is more frequently used in UV spectrometers. Sapphire rods, which are placed in the spectrometer cavity, are applied as targets in matrix esr studies.

METHODS OF GENERATION OF ORGANIC INTERMEDIATES

There are two basic ways of generating unstable species for matrix isolation studies. The first one consists in the formation of intermediates directly in a solid matrix. In the second, the reactive molecules are generated in the gas phase (at very low pressure) with subsequent stabilization by condensation in an inert matrix at 10-20 K. A few methods produce reactive molecules by reactions in solid matrices. The most widely used consists of irradiating already isolated precursors with UV light (including vacuum UV light at A < 200 nm), y- or X-rays. In this case, the fragments which are formed as products of the precursor's dissociation must not recombine in the matrix site. To achieve this effect, one of the fragments should be either chemically inactive [e.g. N,, CO,; see (la)] or able to diffuse easily from the site [e.g. hydrogen atoms as in (lb)]. Inert gas matrix

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

5

The intermediates may also be obtained in matrices by the reactions of atoms with stable compounds. This process is carried out in two stages photodecomposition of a starting molecule leading to formation of atoms, and interaction of the atoms with other stable molecules, as, for example, in (2). Inert gas matrix

Finally, it should be mentioned that the unstable species may be generated in a matrix by co-condensation of metal atoms evaporated from the effusion cells and stable molecules diluted with an excess of inert gas. Equation (3) represents an example. Inert gas matrix

- Li

cc14

Li

cc13

(3)

:cc12

The stabilization of reactive molecules frequently follows their initial generation in the gas phase. In this case the intermediates are formed as a result of the influence of an electric discharge or of high temperature on a stable precursor. Of all the types of discharge, the microwave one is most frequently used for the generation of intermediates for matrix isolation. Passing the substance through the discharge zone at low pressure leads to plasma formation, and all molecules are practically fully dissociated into atoms. Various particles, which arise as a result of recombination of the atoms beyond the discharge zone, can then be frozen in a matrix and studied by spectroscopic methods. To increase the selectivity of this process, the discharge is performed in a beam of pure inert gas which then is mixed with the starting compound before their condensation in the matrix. The unstable species are formed in the gas phase as a result of the collisions of excited inert gas atoms (usually argon atoms) with precursor molecules (4).

Ar

- 13,

Ar*

CHI

Inert gas matrix Ar+(CH4)*

-

CH3

(4)

Simpler methods of generation of intermediates in the gas phase are the thermal ones. The thermal reactions are carried out either in Knudsen effusion cells under equilibrium conditions or in flow reactors at very low pressure. Effusion cells are widely applied for the evaporation of monomeric

6

V. A. KOROLEV AND 0. M. NEFEDOV

molecules of inorganic compounds of low volatility. This method also allows the study of organic conformers frozen in matrices from their equilibrium mixtures obtained in the cell. However, the generation of more reactive intermediates ( e . g free radicals and carbenes) under these conditions is practically impossible because of their destruction and secondary reactions. Very-low-pressure pyrolysis (VLPP) is a traditional method of generation of reactive species in matrix studies. In contrast to pyrolysis, which is carried out at atmospheric pressure, VLPP is performed at 10-3-10-5 Torr. A t these pressures molecules are in the heating zone not more than 10-3-10-2 s and their mean free path lengths are comparable to the oven length. Energy transfer is brought about by collisions of the molecules with the reactor walls which provide the vibrational energy needed for the monomolecular dissociation of the starting compounds. Under high-vacuum conditions the real-time concentration of molecules is very low and bimolecular processes are minimized. That is why thermally generated reactive intermediates may exist in a molecular beam in a concentration which is sufficient for their spectroscopic study.

SPECTROSCOPIC METHODS OF STUDY OF MATRIX-ISOLATED MOLECULES

The basic methods of the identification and study of matrix-isolated intermediates are infrared (IR), ultraviolet-visible (UV-vis), Raman and electron spin resonance (esr) spectroscopy. The most widely used is IR spectroscopy, which has some significant advantages. One of them is its high information content, and the other lies in the absence of overlapping bands in matrix IR spectra because the peaks are very narrow (about 1cm-I), due to the low temperature and the absence of rotation and interaction between molecules in the matrix. This fact allows the identification of practically all the compounds present, even in multicomponent reaction mixtures, and the determination of vibrational frequencies of molecules with high accuracy (up to 0.01 cm-' when Fourier transform infrared spectrometers are usedj. The absence of overlapping of bands of various matrix-isolated compounds and the possibility of freezing highly reactive intermediates make this method very convenient for the direct study of reaction mechanisms. Additionally, direct IR spectroscopy of intermediates allows estimation of important structural parameters, e.g. valence force fields, which show the character of bonds in these species. Esr spectroscopy allows the detection of paramagnetic organic intermediates such as radicals, biradicals and triplet ground-state carbenes. However, there are very few examples of matrix esr studies of these species because there is a simpler procedure for recording esr spectra. Thus many radicals have been investigated by esr after irradiation of precursor molecules in glassy organic matrices at 77 K. Nevertheless, the use of the highly sensitive

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

7

esr method in matrix studies is very helpful because the esr data may confirm the presence of one or other radical among the reaction products frozen in inert gas matrices. Thus, it provides the possibility of making more reliable assignments of optical spectra to a specific radical. UV-vis spectra of matrix-isolated intermediates are not so informative as matrix IR spectra. As a rule, an assignment of the UV spectrum to any intermediate follows after the identification of the latter by IR or esr spectroscopy. However, UV-vis spectra may sometimes be especially useful. It is well known, for example, that the energy of electronic transitions in singlet ground-state carbenes differs from that of the triplet species. In this way UV spectroscopy allows one to identify the ground state of the intermediate stabilized in the matrix in particular cases. This will be exemplified below. Matrix Raman spectroscopy allows detection of some additional vibrations which are inactive in IR spectra (e.g. symmetrical vibrations v1 in AB3 molecules having D3,, symmetry) or which lie in the far infrared region. In practice, matrix-isolated organic intermediates have not been studied by Raman spectroscopy; the main objects of these investigations are inorganic molecules (A1Cl3, PbS, GeF2, SiO, etc.) which are evaporated from solids in effusion cells. Raman spectroscopy of matrix-isolated molecules carries some difficulties connected with the possibility of local heating of the matrix under laser irradiation. Besides, because of the relatively low intensity of Raman bands, higher concentrations of the species to be studied are needed in the matrix (the ratio of matrix gas to reagent = 100-500). As a result, the effective isolation of reactive intermediates is prevented. Thus, a more complete study of the spectral properties and the structure of intermediates frozen in inert matrices is achieved when the IR, Raman, UV and esr spectroscopic methods are mutually complementary. Since IR spectroscopy is the most informative method of identification of matrixisolated molecules, this review is mainly devoted to studies which have been performed using this technique. 2 Carbenes and their silicon and germanium analogues

The comparatively small size of the simplest carbene (methylene) ensures that it has a definite mobility in frozen inert matrices, which leads to the formation of dimerization products under these conditions. It became possible only in 1981 to detect in the spectra of the diazomethane photolysis products bands at 1115 cm-I (Ar matrix) and 1109 cm-' (Xe matrix) which were attributed to the deformation vibration of methylene in its ground triplet state (Lee and Pimentel, 1981). In contrast to methylene, a number of the simplest triatomic halocarbenes

8

V. A. KOROLEV AND 0. M. NEFEDOV

having a singlet ground state were successfully studied by IR spectroscopy in low-temperature matrices. Thus, the fluorocarbene and its deuterium and 13C isotope-labelled analogues have been generated under vacuum UV photolysis conditions from the corresponding trifluoromethanes frozen in argon or nitrogen matrices at 14 K. Two recorded IR bands of the carbene at 1406 and 1187.5 cm-' were assigned to the HCF bending mode and C-F stretching vibration, respectively (Jacox and Milligan, 1969). UV irradiation ( D 3 0 0 n m ) of an argon matrix containing tetrafluoromethane led to the formation of difluorocarbene CF2 (Milligan and Jacox, 1968a). It was shown that the IR spectrum of this species contains three bands at 1222 ( v l ) , 1102 (v3), and 668 (v2)cm-'. Some time later difluorocarbene was stabilized in a neon matrix at 4.2 K from the gas phase after vacuum flash pyrolysis (1300°C) of perfluoroethene (Snelson, 1970b). In this case the IR bands of CF2 differed from those in an argon matrix by less than 2 cm-'. The formation of the monochlorocarbenes CHCl and CDCl was observed when HCl or DCl molecules reacted with carbon atoms in argon or nitrogen matrices at 14K (Jacox and Milligan, 1967). Two IR bands attributed to CHCl were assigned to the C-Cl stretching (815 cm-') and deformation (1201 cm-') vibrations. Before our studies, high temperatures (>SOO"C) had usually been used to generate dichlorocarbene in the gas phase. Based on trapping experiments we have shown that the trihalomethyl mercury derivatives RHgCHa13, which were successfully used earlier as sources of dihalocarbenes in solution (Seyferth, 1972), are also convenient precursors of carbenes in the gas phase (Mal'tsev et al., 1971a,b). Low-temperature matrix stabilization (matrix temperature 8-10 K) of the vacuum pyrolysis (2O0-60O0C, lop3 Torr) products of compounds RHgCC13 (R = Ph, CCI3, C1) has indicated the presence of dichlorocarbene CC12 (vl 720, v3 746 cm-') and trichloromethyl radical CC13 (v3 898 cm-') together with the corresponding RHgCl (695 and 729 cm-' for PhHgC1, 733 cm-' for CC1,HgCl and 403-413cm-' for HgC12) and small amounts of C2C14 (vI1 779, v9 915cm-') and C2C10 (vl0 785cm-') (Nefedov et al., 1971). The warming of the matrix up to 35-40 K (with subsequent cooling to 10 K to record the spectrum) resulted in the disappearance of the 720, 746 and 896 cm-' bands, whereas absorptions of the stable molecules remained or even increased. At 200-500"C, the relative intensity ratio of y (CC12)Iv3 (CC13) varied from 1.1 to 4.0, depending on the substituent R in RHgCCI3. It increased when the pyrolysis temperature was reduced. These results suggest that carbene CCI2 and radical CC13 are formed in separate competing reactions (Scheme l ) , and that the formation of CC12 is preferred. This conclusion has been supported by pyrolytic mass spectrometry studies (Ujszaszy et al., 1980).

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

9

RHgCCI,Br3-, 45IMlX)"C

.CCI,Br3_,,+ [RHg.] R

Ph, CC13, CI

=

n = 1, 2 , 3

Scheme 1

Torr) of PhHgCC12Br in the gas phase Vacuum pyrolysis (250-450"C, is even more selective with regard to the formation of dichlorocarbene, which is formed along with PhHgBr (693.4 and 727.7cm-') (Mal'tsev et af., 1971d). The pyrolysis is complete at 360"C, and the band intensities of CC12 were at least 10 times stronger than those of the CC12Br radical. Bromochlorocarbene and PhHgBr are formed predominantly together with small amounts of CClBr2 radical upon the vacuum pyrolysis of PhHgCC1Br2 (Mal'tsev et al., 1971d). Besides thermolysis, the photochemical decomposition of solid trihalomethylmercury compounds RHgCC13, CF3HgOCOCF3 and Hg( OCOCF3)2 has been studied (Scheme 2). The irradiation of samples placed in an evacuated quartz tube, which was connected to a helium cryostat, was carried out at -50 to +lO"C. Thus, a desorption into the gas phase of the primary products of the photolysis occurred, and consequent lowtemperature matrix stabilization of them was made. As a result, the formation of only the radicals CC13 ( v3 898 cm-') and CF3 ( v1 1084, v2 702, v2 v4 1205, v3 1249 cm-') or of products of their secondary reactions was observed (Mal'tsev el af., 1974, 1975, 1977b). The selectivity in the vacuum pyrolysis of RHgCHa13 has led to the highly

+

hv

RHgOCOCF3

-50 10 t 10°C

R

=

CF3 + COz + [RHg]

CF3, OCOCF3

Scheme 2

V. A. KOROLEV AND 0 M. NEFEDOV

10

Table 1 Matrix IR spectra and structural data of the perhalogenated carbenes and radicals formed in pyrolysis reactions. Intermediate CCl?u.h.c

Isotope distribution

Vibration frequencieskm-’

C3”l2

c3~c1’7cI C”C12 CCIBr“ CC131i

C3’C1Br C”CIBr C3Q C3”I 37

c1 c35c137c12 c37c12 2

CC12Br“

719.5 ( ~ 1 ) . 745.8 ( ~ j ) 717.0 (vI), 744.0 ( ~ 3 ) 714.9 (v,), 741.7 ( ~ 3 ) 743.9 ( y ) 739.5 ( y ) 897.8 (Q) 896.4 ( ~ 3 ) 895.2 ( ~ 3 ) 893.9 (Q) 888.3 ( ~ 3 ) 886.4 ( ~ 3 ) 884.3 ( ~ 3 ) 856.5 ( ~ 3 ) 853.0 (Q) 1084 ( ~ 1 ) ~ 702 ( Q), 1205 ( V Z + ~ q ) , 1249 ( ~ 3 )

Bond angle/ degrees

106

611.qf 15.9

835. If 782Af 17%

“Mal’tsev ef a/. (1971~).hMal‘tsev ef al. (1971d). ‘Svyatkin ef al. (1977). dMaass et al. (1973); Kagramanov ef al. (1977). ‘Mal’tsev et al. (1977b). fC-Br stretch. gAngle between C-X bond and the X3 plane.

informative matrix IR spectra of carbenes CC12, CClBr and CBr2 (Mal’tsev et al., 1971c,d). The observed splitting of the bands, due to 35Cl and 37Cl isotopes, and their intensities are in accord with the number and the natural abundance of the C1 atoms in the intermediates studied. Based on the isotopic splitting, the bond angles in these species have been calculated (Table 1). According to the matrix IR spectra, the preferred formation of dichlorocarbene along with Sic& (v3 600-620 cm-’) or GeCI4 (v3 445-465 cm-’) has been also observed under vacuum pyrolysis (500-1000”C, 10-2-10-3 Torr) of trichloromethyltrichlorosilane and trichloromethyltrichlorogermane (Scheme 3). The CC13 radical was formed in substantially lower amounts (Svyatkin et al., 1977; Nefedov et al., 1976). The preference for formation of dihalocarbenes (but not the trihalomethyl radicals) upon thermolysis of trihalomethyl mercury, silicon and germanium derivatives seems to be a result of intermolecular coordination, of type [ 11, and of a thermodynamic preference for the carbene-forming pathway. The

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

11

.CC13 + [MC13] M

=

Si, Ge Scheme 3

same factors determine the pathways of Si2C16thermolysis, whereas they are excluded in the thermolysis of C2X6 (see below). R,M-CX2

\I

M = Hg, Si, Ge

X

PI Dibromocarbene CBr2 has been formed in inert matrices by two different procedures. The reaction of CBr4 with lithium atoms in an argon matrix as well as the irradiation (with vacuum UV light) of a matrix containing tribromomethane HCBr3 led to the appearance in the IR spectra of two bands of CBr2 at 595 and 641cm-'. These absorptions were assigned, respectively, to the symmetrical and asymmetrical C-Br stretches. Two methods were also used for the formation and IR spectroscopic study of the diiodocarbene. Only one band at 525cm-' assigned to y of the carbene has been found in the IR spectra of the reaction products of vacuum pyrolysis of tetraiodomethane C14 and its matrix reaction with atomic lithium. Besides bromochlorocarbene mentioned above, a number of mixed halocarbenes (mainly fluorohalocarbenes) frozen in rare gas matrices have also been studied by IR spectroscopy. The gas-phase reactions of corresponding halomethanes with metastable argon atoms excited by microwave discharge were used for the generation of these species (Prochaska and Andrews, 1980). In addition, fluorochlorocarbene has been obtained by vacuum UV photolysis of matrix-isolated CH2FCl and CD2FCl. For these carbenes, the stretching vibrations of the C-F bond lie within the range 1157-1133 cm-'. In contrast, the frequency values of the C-X (X=Cl, Br, I) stretches depend on the size of the X atoms and decrease from 742cm-' for CFCl through 656 cm-' for CFBr to 576 cm-' for CFI. In contrast to carbenes of the AX2 type, which contain three atoms, generation of carbenes with a more complex structure under photolysis or vacuum pyrolysis conditions may be accompanied by intramolecular rearrangements. Thus, the matrix isolation study of the vacuum pyrolysis

V. A. KOROLEV AND 0. M. NEFEDOV

12

products of Me3SiCH=N2 (730"C, lop4Torr) and bis(trifluoromethy1)diazirine (350-500°C, 10-'Torr) show that the carbenes Me3SiCH and (CF3)2C formed under these conditions isomerize completely into silene Me2Si=CHMe (Mal'tsev et al., 1980) and an olefin CF3CF=CF2 (Mal'tsev et al., 1985a), respectively. These compounds were identified by their matrix IR spectra. The carbene-olefin isomerization under these conditions takes place in less than 10-s-10-6 s. A photolysis of matrix-isolated methyldiazomethane also led to immediate conversion of the initially formed excited singlet methylcarbene to ethylene (Seburg and McMahon, 1992). Nevertheless, a more traditional approach to the stabilization of carbenes and the investigation of their spectral properties deals with the direct generation of carbenes in low-temperature matrices, e.g. by the photolysis of diazo-compounds or ketenes. The method allows stabilization of carbenes in their ground electronic state, prevents intramolecular isomerization and also facilitates direct spectroscopic monitoring of their chemical transformations in low-temperature matrices. Cyclopropenylidene [2] has been obtained from two different precursors, [3] and [4], by their thermal decomposition in the gas phase (Reisenauer et al., 1984; Maier et al., 1987, 1989a). The subsequent UV photolysis (A = 313nm and 254nm) of this carbene in an argon matrix led to its isomerization to propargylene [5] and further to vinylidencarbene [6]. It should be noted that the carbene [5] was independently obtained by

[41

-\ A

=

313, 254nm

HCECCH A = 313 nm

HC=CCHN2 [71

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

13

photolysis of ethynyldiazomethane [7] frozen in an argon matrix. The vibrational frequencies of the carbene [2] were calculated from its matrix IR spectrum and it was shown that it has a singlet ground state. The results agree well with data obtained by the esr method (Maier et al., 1987). The IR bands of carbenes [2], [5], and [6] have also been observed in the spectrum after vacuum UV photolysis of matrix-isolated methylacetylene (Huang and Graham, 1990). It was found that a fourth carbene propendiylidene [8] - was formed in this reaction as well. In accord with ab initio calculations, the first of two absorptions (3292 cm-' and 1960 cm-') has been assigned to v2 ( a ' ) of cis-[8] and the second one to v3 ( a ' ) of truns-[S].

Hr H

cis-[S]

NGC-CHN2

"rC:

H

trans-[8]

hv ( A P 350 nm) P Ar, 12 K

Cyanocarbene [9] and its D-, 13C-, and "N-labelled analogues have been obtained by UV irradiation (A>350 nm) of diazoacetonitrile [lo] frozen in argon or nitrogen matrices at 12 K (Dendramis and Leroi, 1977; Dendramis et al., 1978). A total of 17 IR bands for all isotopically labelled species has been recorded. The frequency values 1735 and 1179cm-' have been assigned to the stretching vibrations of the CCN backbone. Taking into account the linear structure of the carbene [9], a full normal coordinate analysis of the experimental spectra has been performed. Analysis of the calculated force field showed that both CC and CN bonds have significant double bond character. This fact, as well as the high interaction constant between these bonds, indicates a substantial delocalization of electron density and suggests the allene structure (.CH=C=N-) for the carbene [9]. Furthermore, the high frequency value of the C'-H stretch (3229 cm-') indirectly confirms this idea. The same method was used earlier for the generation of dicyanocarbene [ll] from matrix-isolated dicyanodiazirine (Smith and Leroi, 1969). The carbene [ l l ] was described by three IR bands at 1756 (vas CCN), 1158 (vs

14

V. A. KOROLEV AND 0 . M. NEFEDOV

NC hv

NC

Ar. 12K

Nc): NC

CCN) and 392 cm-' (6 CCN). One can see that the frequency values of the skeletal stretching vibrations of [ l l ] are close to those in carbene [9]. Ethynylhydroxy carbene [131 has been obtained by photoreaction (A>400 nm) of a triatomic carbon cluster with water in an argon matrix and studied by IR spectroscopy (Ortman et al., 1990). Five frequencies were measured for [13] and a vibrational band at 1999.8 cm-' has been assigned to the C=C stretch. This value is more than 100cm-' lower than the C=C stretching vibrations in acetylene derivatives, indicating that the C=C bond in the carbene [13] has lost some of its triple bond character. At the same

H20

12400nm

HC=CCOH

280 nm < A < 360 nm

0

HCEC-C,

//

\

H

time the vibrational band at 1061.1 cm-', assigned to the C-C stretch, is about 100 cm-' higher than usual for C-C bond stretches. The lower C = C and higher C-C stretching frequencies suggest that intermediate [13] has a carbene structure partially resonant with a diradical structure. The irradiation of the matrix-isolated [ 131 (280 nm< A <360 nm) led to isomerization of the carbene and formation of propynal. Irradiation (A = 334 nm) of dicyclopropyldiazirine [ 141, matrix isolated in N2 at 6 K, gives two major products, dicyclopropyldiazomethane [15] and dicyclopropylcarbene [ 161 (Ammann et al., 1992). Subsequent irradiation of the matrix at 475 nm destroyed the IR peaks of [16] and increased those of ethylene, cyclopropylacetylene [ 171, and 1-cyclopropylcyclobutene [18], which were initially present in minor amounts. The diazo compound [15] was stable at this wavelength. Confirmation of the assignment of the IR spectrum of [16] came from trapping experiments. Irradiation of diazirine [14] in an argon matrix doped with 2% CO produced peaks in the IR of the carbene [16]. Warming the matrix to 38 K resulted in a decrease of the IR bands of [16] with concurrent growth of dicyclopropylketene [19] bands. The UV-vis spectrum of [16] contains a broad band at 400-600 nm (A,, 490 nm), along with a stronger absorbance at A,, 230 nm. This result is the first spectroscopic characterization of a singlet dialkylcarbene [16] (triplet carbenes absorb well only below 200 nm).

MATRIX IR SPECTROSCOPY OF INTEHMEDIATES

I = 338 nm

I

15

Ar. 12 K

n Cyclopropylchlorocarbene [20] has been generated by UV photolysis (A = 335 nm) of cyclopropylchlorodiazirine [21] frozen in a nitrogen matrix at 1 2 K (Ho et af., 1989). IR and UV spectra of [20] have been recorded. The reaction of [20] with HCI resulted in the formation of (dichloromethy1)cyclopropane [22], and annealing of the matrix gave (dicyclopropy1)dichloroethene [23]. Subsequent irradiation (A = 450 nm) of the carbene [20] led to its isomerization to 1-chlorocyclobutene [24], which was partialy destroyed to give ethene and chloroacetylene. Ab initio calculations predict the existence of two carbene conformers, but attempts to distinguish them in IR or UV spectra were unsuccessful. An interesting carbene, 1-oxobutatrienylidene [25], having cumulated double bonds, has been found by IR spectroscopy in the photolysis (A>230 nm) products of matrix-isolated 1,2,3,4-pentatetraene-175-dione [26] (Maier et al., 1988) (in its turn the unstable dione [26] was generated by thermo- or photo-destruction of compound [27]). The second product was carbon monoxide. The linear structure of the carbene [25] has been suggested on the basis of two intense IR bands at 2222 cm-' and 1923 cm-' indicating respectively ketene and allene fragments. An attempt to measure the I R spectrum of the hydroxy carbene [28] after UV photolysis (A>220 nm) of formaldehyde isolated in an argon matrix was unsuccessful (Sodeau and Lee, 1978). Instead of [28] only hydroxyacetaldehyde resulting from carbene insertion into the C-H bond of the starting formaldehyde was found in the reaction products. Due to its small size, the

16

V. A. KOROLEV AND 0. M. NEFEDOV

hVIA

-Nz, -CO

carbene [28] is probably labile even in a rigid matrix and is able to diffuse from the matrix site. The UV irradiation ( D 2 0 0 nm, 12 K) of an argon matrix containing 0.1% diazirine [29] gave bis(trifluoromethy1)carbene [30] (IR bands at 1380, 1344, 1197, 1157, 965, 671 cm-’) and a small amount of (CF&C=N2 (Mal’tsev et al., 1985a).

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

17

,c=c=o \ F3C

[311

During photolysis of [29] in an argon matrix doped with 4% CO, bands belonging to the ketene [31] were observed along with those of the carbene [30]. Upon further warming to 40-45 K the carbene bands disappeared and were replaced with bands of the ketene [31], indicating a direct interaction of carbene (CF3)*C with CO. Photolysis of diazirine [29] in a matrix doped with 17% C O resulted in the appearance of only ketene [31] bands; the carbene bands were not detected in this experiment. Photolysis of [29] in a matrix doped with 3% C12 yielded (CF3)2C and (CF3)&CI2 [32] (1313, 1276, 942, 934, 908, 560cm-'). Upon warming to 40-45 K the intensities of the carbene bands were decreased simultaneously with an increase of the intensities of the bands of [32]. As was mentioned above, bis(trifluoromethy1)carbene [30] was not detected in matrix IR spectra of the products formed upon vacuum pyrolysis (500-1000"C, lop3Torr) of diazirine [29] or perfluoropropylene. Singlet halo(trifluoromethy1)carbenes [33a]-[33c], which were characterized by IR and UV spectroscopies and chemical trapping with HCl, have been generated from respective 3-halo-3-(trifluoromethyl)diazirines [34a][34c] frozen in an argon matrix at 1 2 K and irradiated with UV light at

""XIA > 320 nm

X

N

Ar, I 1 K

[34a]-[34c]

X F,C, >:

[3Sa]-[35c]

X'

[33a]-[33c] a:X=F b: X = CI c: X = Br

F

H CI

V. A. KOROLEV AND 0 . M. NEFEDOV

18

= 320 nm (O'Gara and Dailey, 1992). Extended photolysis (A>280 nm) converted the carbenes into the corresponding alkenes [35a]-[35c]. The vibrational spectrum for each carbene was calculated using ab initio molecular orbital calculations and after appropriate scaling (since ab initio calculations generally predict frequency values some 10-15% higher than experimental ones) was in reasonable agreement with the experimentally determined one. The calculated spectra for the singlet states showed better agreement with experimental spectra than those calculated for the triplet states, indicating that all three carbenes have singlet ground states. Indeed, the UV spectra calculated for the singlet carbenes [33a]-[33c] gave good agreement with their experimental spectra which were typical for the singlet species and exhibited absorptions at 435-665 nm and 235-270 nm. The photolysis of matrix-isolated (Ar, 10-15 K) aryldiazo compounds [36]-(411 has given interesting results. Irradiation by UV light (A>470 nm) led to formation of respective triplet carbenes [42]-[47] which were characterized by UV and esr spectroscopies (Chapman el al., 1978; Chapman and Sheridan, 1979; West et al., 1982; Chapman et al., 1984). Infrared spectra of these species have also been recorded, except for the carbene [44]. Irradiation (A>470nm) of the diazo compound [38] gave predominantly o-xylylene [51] via hydrogen atom transfer in the carbene [44], which was converted to benzocyclobutene [52] by irradiation of [38] at shorter wavelengths ( ~ 2 8 nm). 4 Nevertheless, the IR spectrum of the carbene [44] was measured later at a lower matrix temperature of 4.6K (McMahon and Chapman, 1987). Photolysis of the carbenes [42] and [43] led to aromatic ring expansion and formation of cycloheptatetraenes [48] and [49], unusual cyclic molecules having a strained allene fragment.

A

A>478nm Ar. 10 K

@c*

-

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

19

Similarly irradiation of the carbenes [46] and [47] gave azacycloheptatetraene [50].The cycloheptatetraenes [48]-[50] were observed by IR spectroscopy and each of them showed the bands at--1810 cm-', which are typical for the allene group in the rings. Compounds [49] were formed as sole photoproducts (A>416 nm) from [43]. In addition, cycloheptatetraene [48] was obtained by pyrolysis (500", lop4Torr) of phenyldiazomethane [36] and stabilized in an argon matrix after condensation from the gas phase. Both [46] and [47] were formed from [50]and irradiation (A>261 nm) of [40] gave [47], thus establishing the interconversion of [46] and [47]. The conversion of the intermediate carbene [45] to styrene was observed after irradiation or warming to 80 K of a xenon matrix.

V. A. KOROLEV AND 0 . M. NEFEDOV

20

Table 2 Infrared spectra of phenylhalocarbenes (argon matrix, 12 K).

Carbene

Vibrational frequencies/cm-'

Ph(CF3)C 1599, 1460, 1403, 1394, 1210, 1163, 1140, 1129, 1100, 1090, 979, 913, 751, 671,665,608,602 1593, 1480, 1445, 1320, 1304, 1284, 1245, 1226, 1206, 1169, 1023,998, 847,764,738,718,686, 667,609,567 1588, 1473, 1443, 1319, 1303, 1281, 1240. 1224, 1172, 1162, 1023, 998, 844. 833,763, 755, 686, 672, 658,604, 553

Ph(CI)C

Ph(Br)C

"Mal'tsev er al. (198%). 'Mal'tsev

el

Ref.

'' h h

al. (1987a).

The photolysis of phenyldiazirines [53a]-[53c] in an argon matrix was studied in a similar way. Intense bands of the corresponding carbenes [54a]-[54c] (Table 2) have been observed in all cases. The bands disappeared either when the matrix temperature was raised from 12 to 40-45 K (the corresponding stilbene bands appeared at the same time), or when a trapping agent (HCI, Clz) was doped into the matrix (Mal'tsev et al., 1985b, 1987a). It has been noticed that intense bands at 158&1599cm-' and 12101226 cm- are consistent with the presence of a strong electron-withdrawing group in conjugation with the aromatic ring (Ganzer et al., 1986). In all cases the formation of carbenes was accompanied by partial photoisomerization of the precursor diazirines [53] into the corresponding diazo compounds [55] which possess a higher photochemical stability than the diazirines. A number of substituted oxycarbenes [56a]-[56d] having singlet ground states were obtained by UV irradiation of corresponding diazirines [57a][57d] frozen in inert gas matrices and studied by IR and UV-vis spectroscopy. The assignment of observed IR bands to these carbenes has been confirmed by trapping experiments, studying their photoproducts and computational methods as well. For example, the matrix reaction of the carbene [56a] with HCl led to formation of dichloromethoxymethane (Kesselmayer and Sheridan, 1986a). Warming of the matrix to 35-38 K without a trapping agent led to dimerization of these species. The irradiation (A = 270nm) of the carbene [56a] in an argon matrix led to its rearrangement to acetyl chloride which partially decomposed to ketene and hydrogen chloride. Prolonged photolysis of the matrix-isolated carbene [56b] gave benzoyl chloride, chlorobenzene and C O (Kesselmayer and Sheridan, 1986b). According to experimental IR spectra of deuterium- and "0labelled species and ab initio calculations, the intense absorptions at about 1300 cm-' for the carbenes [56a]-[56d] (1300 cm-' [56a], 1315 cm-' [56b], 1320 cm-' [56c] (Sheridan et al., 1988)} and 1442 cm-' for [56d] (Du et al., 1990) have been assigned to C- 0 stretching vibrations. The higher

'

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

21

Ph \c/H

/

Ph

40K

X >: [53a]-[53c]

I

[54a]-[54c]

' Ph

Phx'

yxx Ph

\c'/c'

A1340nm

x'

Ph

a: X b: X C: X

>N=N X

c '1

= CI = Br = CF3

[55a]-[55c] ROCHCIX

, /

3SK

Lv, Ar or N2. 12 K

X

-Nz

[57a]-[57d] a: R = Me, X = C1 b: R = Ph, X = C1 c: R = Me, X = Me d: R = Me, X = F

RC \X

frequencies compared with the corresponding stretches in ethers (11001150cm-') indicate considerable C - 0 double bond character as a result of the interaction of oxygen electron pairs and the unoccupied p-orbital of the carbene centre. The increased multiplicity of this bond suggests the existence of two isomeric forms (cis and trans) of the carbenes [56a]-[56d]. Indeed, the analysis of the IR spectra of the carbenes [56a]-[56d] in different matrices, as well as the results of experiments involving their irradiation with UV light at various wavelengths, made it possible to pick out the spectral absorbances of each isomer. For,,instance, the band at 1300cm-' corresponds to vas(CO) of cis-[.56a] whereas this mode of trans-[56a] absorbs at 1310 cm-'. The UV spectra of the matrix-isolated carbenes [56a)-[56b] have absorption bands in the region 300-400 nm which is more typical for the singlet species.

V. A. KOROLEV AND 0. M. NEFEDOV

22

/"

+O

I:

C-

\

X

trans-[%]

cis-[561

One more interesting carbene [58], obtained as a photoisomer = 285 nm) of tricyclic ketone [59], was stabilized in a nitrogen matrix at 20 K and studied by IR and UV spectroscopies (Kesselmayer and Sheridan, 1987). The structure of this carbene was confirmed by trapping experiments with methanol, giving the corresponding acetal [60]. Subsequent irradiation of I581 at various wavelengths resulted either in reversible isomerization to the starting compound [59] or a conversion to acetylenic ether [61]. Similarly to the carbenes [56] described above, this one shows a band at 1369cm-' assigned to the stretching vibration of the partial C - 0 double bond. As previously mentioned, the carbenes having complex substituents may rearrange by migration of an atom or group adjacent to the carbene centre. Thus, the spectroscopic data for carbonyl carbenes RC(0)CR' were restricted to esr spectra for a long time. The spontaneous rearrangement of these carbenes to the isomeric ketenes (the Wolff rearrangement) did not allow their production in concentrations sufficient for their direct detection in matrices by IR spectroscopy. However, the rearrangement can be slowed down if an electron accepting group (e.g. perfluoroalkyl groups) is attached (A

MATRIX If? SPECTROSCOPY OF INTERMEDIATES

23

RR'C=C=O

A

> 335 nm

N2

[63a]-[63c]

[62al-[ 62c] R

a: R = R' = CF3 b: R = CF3, R' = C-,F5 C: R = C2F5, R' = CF-,

to the carbene atom. Indeed, perfluorosubstituted carbenes [62a]-[62c] have been stabilized in solid argon after irradiation of the corresponding starting diazocompounds [63a]-[63c] by UV light (A>335 nm) and their IR spectra have been recorded (Torres et af., 1983). The bands of ketenes [64a]-[64c] have also been observed. Subsequent irradiation of the matrices with h>210 nm light led to full conversion of the carbenes to the ketenes. When the photolysis of the diazo precursors [63] was carried out with 270 nm light, a number of new bands, which were attributed to the substituted oxirenes [65a]-[65c], appeared in addition to the spectra of the carbenes. Prolonged irradiation of the photolysate with 360 nm light resulted in the isomerization of the oxirenes to the ketenes as well. The thermal stability of the carbenes [62a]-[62c] and oxirenes was also briefly explored. Warming of the argon matrices to 35 K caused a substantial decay in the intensity of the spectra of the oxirenes, but those of the ketenes remained unaffected. Thus, thermally, the oxirenes are considerably less stable than their isomeric a-ketocarbenes. The IR and UV spectra of triplet a-ketocarbene [66] were obtained on UV photolysis (A = 365 k 8 nm) of the diazocompound [67] isolated in an argon matrix at 10K (Hayes et af., 1983). The chemical identity of the carbene [66] was confirmed by trapping with oxygen in argon giving 1,8-naphthalic anhydride [68] and with carbon monoxide in argon giving ketoketene [69]. Further irradiation (A = 625 f 8 nm) of the carbene [66] led to the Wolff rearrangement and the formation of the ketene [70]. The IR and UV spectra of the triplet cycloheptatrienylidene [71] were recorded after the UV photolysis (A>574 nm) of dia,zocycloheptatriene [72] in an argon matrix (McMahon and Chapman, 1986). This carbene interacts with the CO-doped matrix, forming the ketene [73], and it also dimerizes with formation of heptafulvalene [74]. Experiments have shown that [71] cannot be converted into the cycloheptatetraene [48] either photochemically

V. A. KOROLEV AND 0. M. NEFEDOV

24

//

( = J L

- 0: A > 574 nm

35 K

A2514nm

P I

[481

(A>574 nm) or thermally at temperatures up to 35 K. The presence in the electronic spectrum of [71] of an absorption maximum in the longwavelength region at 530 nm suggests a planar structure for this compound, which agrees well with esr data. The cyclopentadienylidene [75], which was obtained by irradiation (A>300 nm) of matrix-isolated diazocyclopentadiene [76], has been studied

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

25

by IR and UV spectroscopies (Baird et al., 1981). Annealing of the matrix containing the photolysate led to formation of the fulvalene [77] as a product of the dimerization of the carbene. At the same time, the ketene [78] was formed when CO was doped into the matrix. The carbene [75] is extremely reactive even at liquid helium temperature (4.2K) and at large matrix dilution (10 0OO:l). The high reactivity may be associated with the ability of [75] to rotate in the matrix site. This conclusion has been supported by photolysis with polarized light of the ketene [78] isolated in a CO matrix. In this case an appreciable reorientation of the initially oriented ketene molecules was observed due to the reversibility of this process. Direct IR monitoring has been successfully carried out on the [ 1 + 21cycloaddition reaction of cyclopentadienylidene [75] to ethylene in an argon matrix which gave the corresponding cyclopropane (Mal'tsev et al. , 1987b, 1989). UV irradiation (h>300 nm) of diazocyclopentadiene [76] (mole ratio [76]:Ar = 1:lOOO) at 12 K led to carbene [75] having a triplet ground state (bands at 1341, 1335, 1100, 1074, 700, 573 cm-'). The use of a harder matrix (Ar 10-15% N 2 0 ) prevented completely the formation of fulvalene [77] during the photolysis. However, [77] was formed upon warming a matrix containing carbene [75]. After the complete photolytic decomposition of [76] into [75] at 12 K in an argon matrix doped with 2% ethylene, the matrix temperature was raised to 40-50K. An increase of cyclopropane [79] concentration and simultaneous disappearance of carbene [75] bands were observed. Increase of the ethylene concentration in the argon matrix to a mole ratio [76]:C2H4:Ar = 1:150:1000 gave cyclopropane [79] as the only photolytic product. The bands of [75] and the dimer [77] in the IR matrix spectra were not observed in this case.

+

V. A. KOROLEV AND 0. M. NEFEDOV

26

It has been possible to record the IR and UV spectra of several derivatives of the carbene [75] - tetrachlorocyclopentadienylidene [80], indenylidene [81] and fluorenylidene [82] (Bell and Dunkin, 1985). These carbenes were formed by UV photolysis of the corresponding diazo precursors frozen in inert matrices and have a triplet ground state. The carbenes [80]-[82] react with CO in inert matrices at 30K, but exhibit a lower reactivity than the carbene [75]. Furthermore, they were stabilized in a pure CO matrix at 12K, whereas the free carbene [75] could not be detected under these conditions. The different reactivity towards CO between [75] and [80J-[82] may be associated with the different steric shielding of the carbene centres and with the different triplet-singlet gap as well. 4-0xo-2,5-cyclohexadienylidene[83] was generated in solid argon at 9 K by irradiation of diazo compound [84] with visible light (A>495 nm) (Sander ef al., 1988; Bucher and Sander, 1992; Bucher ef al., 1992). The IR, UV, and esr spectra of [83] were in accord with a structure having a triplet state with one delocalized electron. In the IR spectrum of the carbene [83] the v ( C 0 ) mode was found at 1496cm-', which indicates a bond order of the C - 0 bond considerably less than 2. The low-temperature reaction of carbene [83] with CO generated the keto-ketene [85]. Irradiation (A = 543 k 10 nm) of [83] led to its transformation into a very labile species, presumed to be [86], which rearranged back to [83] not only under UV or

A

o=@~ 4 1

S

> 495 nm

A>495nm

0

0

[861

:

.A

= 543 nm

~ 3 1

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

27

Ar. IOK

IR irradiation but even on standing in the dark (e.g. in Ar at 9 K , 165i-30h). The carbene [86] is a carbonyl compound [v(CO) 1720cm-’] with a singlet ground state (no esr signals) lying energetically above the triplet ground state of [83]. The slow intersystem crossing rate for the conversion of [86] to [83] was explained in terms of their different structures: “Whereas 1831 is best described as a . . . partially ‘aromatic’ ring system, [86] has a more localized cyclohexadienone structure.” In this case “the intersystem crossing requires a large geometric change of the molecule, although no covalent bonds are rearranged” (Sander et aZ., 1988). The five-membered cyclic biradical [88] has been generated from precursors [89] and [90] and stabilized in an argon matrix at 10K (Roth et al., 1987b). The triplet ground state of this species has been confirmed by esr spectroscopy. The structure of [88] was supported by trapping experiments with 02,SOz, and N-phenylmaleimide as well as its dimerization. The intense IR bands at 868 and 886cm-I were assigned to deformations of exo-methylene groups and strong absorption at 767 cm-’ to “out-of-plane” deformation of CH groups. Another triplet diradical [93] has been generated by irradiation (A = 313nm) of the carbonyl bicyclo compound [94] in an argon matrix (Roth et al., 1987a). Thermolysis of compound [95] resulted in formation of [93] as well. In accordance with the IR spectrum of the free ally1 radical, two intense absorptions at 854 and 810cm-’ in the IR spectrum of [93] were assigned to out-of-plane deformations of CH groups”. The formation of the species [93] was confirmed by trapping experiments with olefins and oxygen. Irradiation of the matrix isolated [93] in the absence of trapping agents led to intramolecular rearrangement of the biradical and formation of an intermediate [98] which was converted reversibly into [99]. t1/2=

28

V. A. KOROLEV AND 0.M .NEFEDOV

?',

A = 254nm

A number of close carbene analogues - silylenes and germylenes - have also been generated, stabilized in inert matrices and studied using the IR spectroscopic technique. Thus, the simplest silylene SiHz and its deuterium-substituted derivatives (SiHD and SiD2) were generated by vacuum UV photolysis of the silane frozen in an argon matrix at 4 1 4 K (Milligan and Jacox, 1970). Three IR bands at 2032, 2022 and 1088 cm-', attributed to SiH2 in the spectrum of the photolysis products, were assigned to fundamental modes of this intermediate - y , v1 and vz, respectively. Normal coordinate analysis of the vibrational spectra of the silylenes SiH2, SiHD and SiDz was performed and valence force fields were calculated. Later SiH2 was also obtained as a product of a matrix reaction between silicon atoms and hydrogen molecules (Fredin et al., 1985). It should be noted that the reactions of silicon atoms with H 2 0 and HF were used for the generation of silylenes HSiOH (Ismail et al., 1982a) and HSiF (Ismail et al., 1982b) for subsequent IR study in argon matrices. The experimental frequency values of the Si -H stretching vibrations of silylenes SiHz, HSiOH and HSiF lie in the region 2030-1850 cm-' and the calculated

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

29

values of force constants ( F ) of these bonds are (2.0-2.2) x lo2N m-'. These values are somewhat lower than those for silanes (cf. SiH4: v(SiH) 2160-2190 cm-', F(SiH) 3.5 X lo2 N m-'); therefore, in comparison with silanes, the Si-H bonds in silylenes are weakened. The matrix IR spectrum of HSiOH showed that this intermediate exists as cis- and trans-isomers. When the matrix was heated the cis-isomer was converted into the more stable trans-form. Vacuum UV photolysis of matrix-isolated GeH4 and GeD4 has also been used to generate germylenes GeH2 and GeD2 (Smith and Guillory, 1972) for spectroscopic study. The IR spectrum of GeH2 exhibited absorptions at 1887 and 1864 cm-', shifted to 1338 and 1325 cm-' in the spectrum of GeD2, which were assigned to antisymmetrical and symmetrical stretching vibrations of Ge-H bonds. The presence of the band at 920cm-' was evidence of the bent structure of GeH2. Less reactive inorganic silicon and germanium analogues of dihalocarbenes (SiF2, GeF2, SiCI2, GeCI2, SiBrz and GeBr2) have been obtained and stabilized in argon, neon, krypton and nitrogen matrices and their IR spectra recorded. Vacuum pyrolysis (600-1 1Oo"C, 10-3-10-4 Torr) of hexachlorodisilane ( 5 ) proceeds selectively with the formation of only SiC12 and SiC14 (Nefedov et al., 1974; Svyatkin et al., 1977).

The bands of the Sic& radical (Milligan and Jacox, 1968b), have not been observed at all. Due to the high selectivity of its thermal decomposition within a wide temperature interval, Si2CI6 has become one of the most suitable precursors of dichlorosilylene in preparative chemistry (Chernyshev and Komalenkova, 1990). In contrast to the compounds C13MCC13 (M = Si, Ge) and Si2CI6, hexahaloethanes [1001 do not produce dihalocarbenes under vacuum pyrolysis. Instead, homolysis of the C-C bond takes place, giving the radicals, or dehalogenation of [ 1001 yields the corresponding tetrahaloethylenes (Nefedov et al., 1976; Svyatkin et al., 1977).

X = C1, Br

(22x4

+ xz

30

V. A. KOROLEV AND 0. M. NEFEDOV

Table 3 Bond angles (ao)in MX2 molecules.

MX2

(YO, ~~~

~

from IR matrix spectra

a', from electron diffraction

~

cc1*

106 +. 5* 102 +. 5'

SiCI? SiBr2 GeCIg GeBr2

109.2" 102.8-t O.@ 102.7 +. 0.3g 100.3 -t 0.4h 101.2 f 0.9'

99 -t 4f

"From the millimetre- and submillimetre-wave spectrum (Fujitake and Hirota, 1989). bMW spectral data give a value of 101.25" (Tanimoto et al., 1989). "The values of 94.6" and 97.7" have been obtained by X-ray diffraction of C4H902.GeCI, and Ph3P. GeCI, molecular complexes, respectively (Kulishov et al., 1970; Bokyi et al., 1975). dMal'tsev et al. (1971~).'Svyatkin et al. (1977). 'Mal'tsev et al., (1976a). RHargittai ef al. (1983). 'Schultz et al. (1979). 'Schultz et al. (1982).

Table 4 Bond lengths and M-CI bond force constants in MC12 species and MC4 molecules.

MClz M

10-2F(M-Clj/ N m-'

C Si Ge Sn

2.99" 2.29" 2.08" 2.1

MC14

M-Cl bond length/pm 171.57' 208.3c.d 218.6' 234.7

10-2F(M-CI)/ N m-'

M-CI bond length/pm

3.59 3.37 2.79 3.63

176.6 202.0 211.3 228.08

"Svyatkin et al. (1977). 'From millimetre- and submillimetre-wave spectra (Fujitake and Hirota, 1989). 'Hargittai et al. (1983). dSchultz et al. (1979). 'Schultz et al. (1982).

The high-resolution matrix IR spectra of dichlorosilylene SiC12 generated from Si2C16 (Svyatkin et al., 1977) and of dichlorogermylene GeC12 (Mal'tsev et al., 1976a) obtained by thermal depolymerization of (GeC12)xin vacuum display complex patterns due to many combinations of chlorine isotopes (35Cl, 37Cl) as well as Si (28Si, 2ySi, 30Si) or Ge (7"Ge, 72Ge, 73Ge, 74Ge, 76Ge) isotopes. Comparison of the experimentally obtained and the calculated frequencies allowed the determination of the force constants and bond angles of these species (Tables 3 and 4). The values of bond angles (around 100') are in accordance with a singlet ground state. The bond angles in the dichlorides SiC12 (Hargittai et al., 1983) and GeC12 (Schultz et al., 1982) which were determined by direct electron diffraction studies of these unstable species are close to those from the IR matrix technique (Table 3). The generation of the dichlorides MC12 and dibromides MBr2 (M = Si,

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

31

Me2 Si Me2Si’ ‘SiMe2

I

I

Me2Si, .,%Me2 SI A = 4XXnm

WI

Me

A = 2611 nm

MeSiH =CH,

[ 1031 Me

\

/

Me

Si

/N3

\

N3

[lo21

Ge) for the electron diffraction measurements was carried out by thermal reaction of silicon (1200”C, 1Torr) or germanium (620-66OoC, 1 Torr) with the corresponding tetrahalides MX4 or hexahalides M2X6 (Shultz et al., 1979; Schultz et al., 1982; Hargittai et al., 1983). To obtain the maximum concentration of MX2 species, the conditions for these reactions were first optimized by the pyrolytic mass spectrometry method (Kagramanov et al., 1983a). The contents of the mixture during the electron diffraction experiments were controlled by a quadrupole mass spectrometer. Both the bond angle (from 106“ to 99”) and force constant (from 2.99 X lo2 to 2.08 x 10’ N m-’) values decrease in the following order: CC12> SiC12> GeC12 (Tables 3 and 4). The force constants values F(M--1) in MC12 are appreciably smaller compared to those in the corresponding MC14. Thus, this fact excludes for MC12 species a significant contribution of resonance structures of the type Cl-Mf= C1- resulting from a possible interaction of the lone pairs on the chlorine atoms with a vacant p-orbital of the central M atom. This conclusion is in accord with direct electron diffraction studies of MC12 and MBr2 (M = Si, Ge) which show that the bond lengths in MC12 are usually longer compared with those in the corresponding MC14 (Table 4). The organic silylene Me2Si was generated in an argon matrix by irradiation (A = 254 nm) of dodecamethylcyclohexasilane [1011 (Arrington et al., 1984) and dimethyl(diazid0)silane [lo21 (Vancik et al., 1985; Raabe et al., 1986) and nine IR bands were attributed to this intermediate in the spectra of the photolysis products. Photolysis with 488nm light led to isomerization of Mez% to silene [103]. To assign experimental IR bands of Me2Si and [103], irradiation with polarized light of the matrix, containing Me2Si, was performed. As a result, six IR transitions of Me2Si and twelve IR transitions of [ 1031 were assigned as polarized “in-plane’’ and “out-ofplane”. Later, dimethylgermylene, Me2Ge, was obtained in a similar

32

V. A. KOROLEV AND 0. M. NEFEDOV

manner by photolysis (A = 254 nm) of matrix-isolated dimethyl(diazid0)germane [lo41 (Barrau et al., 1989b) and studied by IR and UV spectroscopies. Twelve IR absorptions of MezGe were detected in the matrix spectrum of photoproducts. CH;,

-

,N3 Ge

A

254 nm

Ar. 10 K

CH(

'N; ~

4

CH3, ,Ge: CH;

1

In contrast to diazido compounds [lo21 and [104], which throw off two azido groups and form silylene and germylene, photodecomposition of silyl azide [lo51 led to the generation of aminosilylene [lo71 via isomerization of initially formed nitrene [106] (Maier et al., 1989b). The IR spectrum of the H3Si-N3

A > 254 nm

[1051

[H3Si-N:] [lo61

-

H\ Si: NH~' [1071

A>3Wnm

T H N F S i + HZ A>254nm

[lo81

photolysate contained six bands attributed to the silylene [lo71 and the absorption at 1975cm-' was assigned to the stretching vibration of the Si-H bond. At the same time the bands at 3495, 3409 and 1563 cm-' were assigned to vibrations of the amino groups - v,,(NH2), vS(NH2), and 6(NHz), respectively. Subsequent irradiation of [ 1071 with wavelengths > 300 nm yielded silaisonitrile [lo81 and molecular hydrogen. This reaction may be reversed with 254 nm light. 3

Free radicals

After the first unsuccessful attempts to record a matrix IR spectrum of the methyl radical, reliable data were obtained by the use of the vacuum pyrolysis method. IR spectra of the radicals CH3 and CD3 frozen in neon matrices were measured among the products of dissociation of CH31, (CH&Hg and CD31 (Snelson, 1970a). The spectra contained three absorptions at 3162 (v3), 1396 ( y ) and 617 cm-' (vl) belonging to the radical CH3 and three bands 2381, 1026 and 463 cm-' assigned to the radical CD3. Normal coordinate analysis of these intermediates was performed and a valence force field calculated. In accordance with the calculations, methyl radical is a planar species having symmetry D3,,. Photolysis of symmetrical diacyl peroxides [lo91 was used for generation in inert matrices of a number of alkyl radicals (see Pacansky et al., 1991; Pacansky and Waltman, 1989, and references cited therein). Thus, ethyl,

33

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

hv

R

Ar. in^

2R+2CO

0

propyl, isopropyl, butyl, isobutyl, pentyl, neopentyl and cyclopentylmethyl radicals were studied by IR spectroscopy. Of all the radical vibrations only those of the groups CH2 and C-H at the radical centre are particularly interesting. For example, asymmetrical CH2 stretches appear at 31003115cm-' and symmetrical ones at 3015-3035cm-', i.e. in a region of ethylenic C-H stretches. This observation is supported by the high frequency values (1150-1180~m-~)of the C-C stretches between a and p carbon atoms [cf. for ethane .(C-C) is 1016 cm-'I. Out-of-plane deformations of the terminal CH2 groups also appear in a very narrow region at 530560cm-' as medium-intensity bands. At the same time the stretching vibrations of CH bonds at the p carbon atom have lower frequencies at 2810-2840 cm-' than those in alkanes, which indicates some weakness of these bonds in the radicals. An informative IR spectrum of the t-butyl radical, containing 18 bands, has been recorded after freezing of the products of vacuum pyrolysis of azoisobutane [110] and 2-nitrosoisobutane [ l l l ] in an argon matrix at 10 K (Pacansky and Chang, 1981). This spectrum is in agreement with a pyramidal structure of the radical (CH3)3C (symmetry C3v) which has elongated CH bonds in positions trans to the radical centre. On the basis of experimental vibrational frequencies and ab initio calculations of the radical geometry the enthalpy value [%(300)] of its formation has been calculated as 44 kJ mol-l. Phenyl radical, side by side with methyl radical, carbon dioxide and methyl benzoate, was also stabilized in an inert matrix as a product of UV photolysis of acetyl(benzoy1)peroxide [ 1121 (Pacansky and Brown, 1983). Of nine IR bands of the radical C6H5, intense absorption at 710cm-', which was shifted to 519 cm-' for the deuterium-labelled radical C6D5, has been assigned to out-of-plane C H deformation. The bands of the phenyl radical

34

V. A. KOROLEV AND 0. M. NEFEDOV

as well as the bands of the methyl radical disappeared when the matrix was annealed. Simultaneously an increase of intensity of the toluene absorptions was observed. The presence of methylbenzoate in the photolysate was explained as a result of the interaction of initially formed benzoyloxy radicals [113] with methyl radicals. At the same time the absence of phenyl acetate indicated that of the two primary photoproducts the benzoyloxy radical is a longer-lived intermediate than the acetyloxy radical [ 1141. An IR spectroscopic study of the radicals CF3, C2F5, C3F7 and i-C3F7 has been carried out. These radicals were formed as products of vacuum pyrolysis in a platinum reactor of the respective fluorinated iodoalkanes and were stabilized in argon matrices at 10-12K (Snelson, 1970b; Butler and Snelson, 1980a,b,c) as shown in (6).

IR absorptions of these species were assigned to fundamental modes by comparison with the spectra of stable perfluoroorganic compounds. Normal coordinate analysis of the perfluoroethyl radical was performed and the valence force field of C2F5was calculated (Snelson et al., 1981). Attempts have also been made to obtain the radicals (CF3),C and C6Fs as products of vacuum pyrolysis of (CF3)3CI and C6FsI (Butler and Snelson, 1980b). However, only perfluoroisobutene was observed in an IR spectrum of pyrolysis products of (CF3)&I. Thermolysis of C6FsI led to formation of CF,, CF3 and CF2 as a result of decomposition of the aromatic ring. This behaviour was explained as due to catalytic effects which take place on the platinum reactor surface. Alkyl radicals generated from azoalkanes as in (7) react with oxygen added to argon matrices giving alkylperoxy radicals. In this manner radicals RN=NR

n

-N?

R

=

R.

O? Ar. 12K

Me, Et, Pr', But

ROO.

(7)

35

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

MeOO (Ase et al., 1986), EtOO (Chettur and Snelson, 1987b), Pr'OO (Chettur and Snelson, 1987a), ButOO (Chettur and Snelson, 1987c) as well as the radical C F 3 0 0 (8; Butler and Snelson, 1979) were obtained when argon matrices containing pyrolytically generated alkyl radicals and 10% O2 were warmed to 30-35 K. Infrared spectra of 1602-, '*02and 160'80labelled peroxy radicals clearly showed a presence of two non-equivalent

oxygen atoms in these species. Subsequent irradiation (h>254 nm) of the alkylperoxy radicals resulted in their decomposition and formation of low molecular weight products CO, COz, H 0 2 , H 2 0 as well as carbonyl compounds R'R2C=0 (e.g. MeOO gave H,C=O, EtOO gave MeCH=O, and Pr'OO gave Me2C=O). Annealing of the matrices resulted in dimerization of the peroxy radicals giving unstable alkyltetroxides R ( 0 ) 4 R which were also characterized by IR spectroscopy. It should be noted that the 0-0 stretching mode frequencies of all the peroxy radicals studied lie in the narrow range 1092-1145cm-'. This is much higher than the 0-0 stretches of hydrogenated and perfluorinated alkyl peroxides (700800cm-'). At the same time the C - 0 stretching frequencies of the alkylperoxy radicals show considerable variation, decreasing overall by ~ 2 0 %(from 900 to 730 cm-') in the order primary>secondary>tertiary. This is opposite to the variation found for C - 0 stretching frequencies in aliphatic alcohols, which cover the range from 1000 cm-' (primary) to 1200 cm-' (tertiary). A normal coordinate analysis of peroxy radicals MeOO (Cyvin et al., 1986) and C F 3 0 0 (Snelson et al., 1981) has been performed and their valence force fields have been calculated. In particular, the values of the force constants of C - 0 and 0-0 bonds were estimated as -4.5 x lo2 N m-l and 4 . 8 5 x lo2 N m-', respectively. A number of hydrocarbon radicals having multiple bonds at the radical centre have also been trapped in inert matrices and studied by IR spectroscopy. Thus, ethynyl radical was obtained by vacuum UV photolysis (9) of matrix-isolated acetylene (Shepherd and Graham, 1987) as well as when acetylene and argon atoms excited in a microwave discharge were codeposited at 12 K (Jacox and Olson, 1987). An appearance of diacetylene bands was observed when the matrices were warmed up, while the absorptions of the radical C2H disappeared. Detailed isotopic studies of Dand 13C-labelledethynyl radicals showed a surprisingly low frequency of the C=C bond stretching vibration at 1846cm-' instead of c.2100cm-' for a true C=C triple bond (the band at 2104cm-' was attributed to the HC=CH

Ar' Or

VdCUUm

10 K

UV

HCrC.

(9)

36

V. A. KOROLEV AND 0. M. NEFEDOV

+

combination of v2 v3). This means that the bond order of the radical C=C bond is significantly lower than 3. On the other hand, abnormally high absorption value at 3610 cm-' assigned to the C-H stretching vibration (instead of c.3300 cm-' for terminal =C-H stretches) corresponds to C-H stretching vibrations at the C = C bond having a bond order higher than 3. In our opinion this assignment is not reasonable. Fourier transform isotopic (13C and D) studies of potential interstellar species - C4H.(butadiynyl radical) and C6H (hexatriynyl radical) - have also been carried out. The radical C4H was produced (10) by trapping of products from the vacuum UV photolysis of diacetylene (C4H2) or 1,3butadiene (C4H6) in solid argon at 1 0 K (Shen et al., 1990). Similarly the radical C6H was obtained (11) by vacuum UV photolysis of matrix-isolated

HC=CH

acetylene and 1,3-butadiene (Doyle et al., 1991). The bands at 3307 and 2084cm-' of the radical C4H have been assigned, respectively, to the v1 C-H stretching and y C=CH stretching modes. At the same time the terminal C=C stretches of both C4H and C6H radicals absorb, respectively, at 2024 and 1953cm-l. Thus, in contrast to the alkyl radicals mentioned above (p. 33) which have a stronger than ordinary C-'C bond at the radical centre, a tendency to weaken the C=C triple bond takes place in ethynyl radicals. The main difficulty in obtaining the vinyl radical is that the species easily loses the hydrogen atom and is converted into acetylene. Nevertheless, a very low concentration of the radical H2C= C H has been achieved (Shepherd et al., 1988) by vacuum UV photolysis of ethylene frozen in an argon matrix, and a Fourier transform IR study of this intermediate has been carried out. A variety of 13C- and deuterium-substituted ethylene parent molecules were used to form various isotopomers of vinyl radical. On the basis of its isotopic behaviour and by comparison with ab initio

37

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

calculations, a sole radical absorption observed at 900 cm-' has been assigned to the out-of-plane deformation mode p(CH2). When formic acid was codeposited at 14 K with a beam of excited argon atoms, formyl radical, HOCO, was produced (12) in sufficient yield for the IR detection of most of its vibrational fundamentals (Jacox, 1988). Detailed analysis of the matrix spectra of isotopically (D, 13C and "0) labelled formyl radical showed absorptions at 3603, 1844 and 1065 cm-', which correspond to the stretching vibrations of 0 - H , C=O and C - 0 bonds. H0 H-C,

Ar'or F.

OH

12 K

H0

.c,

OH

Normal coordinate analysis of the radical has been carried out and excellent agreement of experimental and calculated frequency values was obtained for the trans structure of HOCO. The sulfur analogue of formyl radical - dithioformyl, HSCS - was a predominant product of a reaction of hydrogen atoms and carbon disulfide (13) during their co-condensation in an argon matrix at 12K (Bohn et al., 1992). This reaction must have a surprisingly low activation energy because it proceeds spontaneously in solid argon at 18 K. Thus, IR bands, assigned H.+CS,

-2 Ar. I 2 K

.C 'SH

H.

* = 3 2 ~ 1 0 0 0nm

H-C,

2 SH

(13)

to the radical, were decreased in intensity by irradiation with UV light (A = 320-1000nm) and appeared again when the matrix was warmed up to 18 K to allow diffusion and reaction of trapped hydrogen atoms. Deuterium, I3C and 34S isotopic data as well as ab initio calculations of the structure and vibrational spectrum of HSCS were used to identify this species as trans-HSCS. Stretching vibrations of the intermediate exhibited absorptions at 2527 cm-' (S-H), 1275 cm-I (C=S) and 628 cm-' (C-S). The reaction of a second hydrogen atom with trans-HSCS in the matrix gave both cis- and trans-HC(S)SH, the sulfur analogue of formic acid. Owing to the large exothermicity of the reaction, the [cis]l[trans]ratio for dithioformic acid produced in the matrix cage was higher than the gas-phase thermal equilibrium ratio at room temperature. 4 Conjugated organic radicals: allyl, propargyl, benzyl and cyclopentadienyl types

Unsaturated conjugated organic radicals are another group of unstable molecules studied by matrix IR spectroscopy, pyrolytic mass spectrometry

V. A. KOROLEV AND 0. M. NEFEDOV

38

0

I

X = I, Br, CI, SiMe3 Scheme 4

and electron diffraction. These radicals have both a conjugated r-system and an unpaired electron, which should result in redistribution of the electron density affecting the bond character in these species. Therefore, the determination of their energetic and structural parameters is of great importance, especially in connection with structural investigations of r-ally1 and cyclopentadienyl metal complexes. The free allyl radical C3H5 [115] was obtained by vacuum pyrolysis (600-1000"C, 10-3-10-5Torr) of various precursors as shown in Scheme 4 (Mal'tsev et al., 1982a, 1983, 1984b; Bock et al., 1983; Maier et al., 1983). Increasing the pressure in the pyrolysis cell up to 10-'-1 Torr promoted secondary intermolecular transformations of the C3Hs radical, resulting in formation of propylene and 1,5-hexadiene (diallyl). Stabilization of the C3H5 radical in an argon matrix at 12 K enabled the determination for the first time (Mal'tsev et al., 1982a) of 17 IR bands of this radical (3107, 3051, 3040, 3019, 1602, 1477, 1463, 1389, 1317, 1284, 1242, 1182, 983, 973, 809, 801, 510 cm-'). A year later a similar matrix IR spectrum of the allyl radical was published by Maier, Bock and co-workers (Bock et al., 1983; Maier et al., 1983). In 1984 the matrix IR spectrum of C3D5 radical was obtained which displayed 11 bands (2285, 2214, 2209, 1263, 1062, 1018, 1007, 767,

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

39

Table 5 Vibrational assignment in IR spectra of free allyl and perdeuteroallyl radicals. ~

Vibrational frequencieskm-' CH2CHCH2 Vibration no. 1 13 2 3 14 15 4 16 17 5

18 6 10 8 11 7 9 12

Symmetry

Assignment

CD2CDCD2

Exp."

Exp.'

Calc.'

3105 3105 3048 3019 3019 1463 1463 1285 1477 1403 1389 1242 985

3107 3107 3051 3019 3019 1477 1463 1389 1284 1242 983 973

3109 3095 3056 3010 2998 1488 1466 1379 1275 1239 988 970 915 804 795 510 281 200

802 510

809 80 1 510

810

"Maier et at. (1983). "Mal'tsev et a f . (1982a). 'Mal'tsev

eta[.

Exp.'

2285 2214 2209 1062 1018 1007 1263 762 767 650 646

Calc.' 2312 2303 2263 2195 2194 1100 1053 978 1268 1151 744 815 757 634 596 420 196 151

(1984b).

762, 650, 646 cm-') (Mal'tsev et al., 1984b). Assignment of the bands in the IR spectrum was carried out (Table 5 ) and the valence force field was calculated. This analysis showed that the stretching frequencies of all the C-H bonds are in the region of ethylenic CH stretches (>3000 cm-'). The vibrational frequencies and the force constants of the carbon frame in the C3H5 radical [vas(CCC) 1284 cm-', v,(CCC) 1242 cm-', F(CCC) 5.8 X lo2 Nm-I lie between the corresponding values for a double C = C bond [.(C=C) 1640cm-', F(C=C) 9.0 x 102Nm-'] and a single C-C bond [.(C-C) 920 cm-', F(C-C) 4.5 x 102N m-'I. Later, successful determination of the molecular structure of the free allyl radical was achieved by high-temperature electron diffraction, augmented by mass spectrometry studies (Vaida et al., 1986). The structural parameters obtained for the allyl radical were: rcc 142.8 pm, r& 106.9 pm, act- 124.6", aCCH120.9". This was the first electron diffraction study of an unstable organic molecule. Comparison of IR and ED data for the C3H5 radical with IR and X-ray data of .rr-ally1 metal complexes shows that the formation of such complexes

40

V. A. KOROLEV AND 0. M. NEFEDOV

results in an increase of the v,,(CCC) frequency from 1284cm-' to 1380 or 1480cm-' and in shortening of the C-C bond from 142.8pm to 138.0pm. These changes may be explained by the transfer of electron density from the metal atom to a non-bonding orbital of the r-allyl system rather than to an anti-bonding orbital. The allyl radical [115] trapped in an argon matrix can be photolytically (A = 410 nm) converted into the cyclopropyl radical [116] (Holtzhauer et al., 1990). Dicyclopropane and cyclopropane were formed when the photolysed matrix was warmed from 18 to 35 K. The intermediate [116] was shown to be a v-type ( C , symmetry) and not a .rr-type (C2" symmetry) radical. Normal coordinate analysis of the radical [116] has been carried out and the IR band at 3118 cm-' has been assigned to the stretching vibration of the C-H bond at the radical centre. Perfluoroallyl radical, C3F5, was obtained by vacuum pyrolysis (850950°C, lop4Torr) of 175-perfluorohexadiene or of 3-iodopentafluoropropylene (14) and was studied by pyrolytic mass spectrometry (Kagramanov et al., 1983b) and by IR spectroscopy in an argon matrix (Mal'tsev et al., 1986). CF2= CFCF2CF2CF=CF2

WPC. IW'Torr , (CF2 -CFzCF2)'

CF2= CFCF2I

The ionization potential of C3F5 is 8.44 eV (cf. IP of C3H5 8.13 eV). Five bands were detected in the matrix IR spectrum of the C3F5 radical: 1498, 1350, 1215, 1008, 597 cm-'. A normal coordinate analysis has been carried out and force constants calculated. Comparison of the antisymmetric stretching frequency of the carbon skeleton [va,(CCC) 1498 cm-'1 of C3F5 and the force constant of its CC bonds [F(CCC) 5.0 x lo2 N m-'1 together with the corresponding values in C3H5 (1284 cm-I and 5.8 X lo2 N m-') and the frequency and the force constant values in perfluoropropylene (1795 cm-' and 7.8 X lo2 N m-') show a substantial weakening of the C-C bonds in C3F5 compared with perfluoropropylene, and in insignificant weakening compared with the allyl radical (Baskir, 1989). Propargyl radical was produced by vacuum pyrolysis (900-1O5O0C, lop4lop5Torr) of propargyl iodide or of dipropargyl oxalate (Korolev et al. , 1989) and it was frozen into an argon matrix at 1 2 K (15). Twelve bands were observed in the matrix IR spectrum of the C3H3 radical [only three bands were recorded earlier (Jacox and Milligan, 1974) for this radical]: 3307, 3111, 3026, 2080, 1440, 1369, 1061, 1017, 686, 618, 532, 482cm-'. Based on the experimental IR spectrum, a normal coordinate analysis has been carried out and a valence force field calculated (Table 6). Stretching frequencies of the carbon skeleton [v(C-C) 2080cm-' and .(C-C)

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

41

HCECCH~I

1017cm-'], as well as calculated values of the force constants [F(C=C) 15.05 X lo2N m-', F(C-C) 6.53 X 102N m-'1 show a weakening of the triple bond and strengthening of the C-C bond in the propargyl radical due to electron density delocalization in the conjugated system [cf. for methylacetylene HC-CCH3: v(C-C) 2142 cm-', F(C=C) 15.8 x lo2 Nm-'; .(C-C) 931 crn-l, F(C-C) 5.12 x lo2 N m-'I. Vacuum pyrolysis of iodoacetonitrile (16) proceeds similarly to that of propargyl iodide and leads to the corresponding cyanomethyl radical, C2H2N, which was studied by low-energy mass spectrometry in the gas phase and by IR spectroscopy in an argon matrix (Nefedov, 1991a,b). Contrary to one author, who observed only one IR band of this radical at 664cm-' (Jacox, 1979), we have succeeded in detecting seven bands: 3142, 3044, 2087, 1431, 1025, 1021, 664cm-l. The bands at 2087 and 1021cm-' were assigned to the stretching of the C-N bond and C-C bond, respectively. These frequencies and the calculated values of the force Table 6 Vibrational assignment of the IR spectra of the free propargyl radical" and methylacetylene.

H C s C - CH2'

HCsCCH3

vtcm-' Symmetry

Assignment

Exp.

Calc.

3307 3111 3026 2080 1440

3307 3111 3026 2080 1443

1062 1017 686 618 532 482

1061 1016 684 618 529 484 423

-

"Korolev et al. (1989).

10-2Fl N m-l

vlcm-' Exp .

10-2Fl N m-l

5.84

3334 3008 2941 2142 1440, 1382 1052 931 633

6.44

15.05

6.53

328

15.8

5.12

42

V. A. KOROLEV AND 0. M. NEFEDOV

Table7 IR spectra and force constants of the free cyanomethyl radical" and acetonitrile.

N=C-CH2'

NEC-CH~ v/cm-'

Symmetry

Assignment

10-2F/ N m-'

Exp.

Calc.

3142 3044 2087 1431

3142 3044 2086 1431

15.09 0.31

1025 1021 664

1025 1020 664 428 419

6.82 0.14 0.44 0.40

0.52

v/cm-' Exp .

3010 2954 2267 1388, 1412 1059 918 378

lO-'Fl N m-'

17.5 0.47

0.72 5.14 0.19

"Nefedov (1991a,h). Bold entries show a weakening of the CN bond and strengthening of the CC bond as a result of electron delocalization in the radical in comparison with acetonitrile.

constants [F(C=N) 15.09 x lo2 N m-', F(C-C) 6.82 x lo2 N m-l (Table 7)] show, as in the propargyl radical, a weakening of the C=N and strengthening of C-C bonds caused by delocalization of electron density [cf. for acetonitrile N=C-CH3: v(C=N) 2267 cm-', F(C=N) 17.5 x lo2 N m-'; .(C-C) 918 cm-', F(C-C) 5.14 X lo2 N m-'I. This conclusion is indirectly supported by the high values of the stretching frequencies of the CH bonds [vas(CHz)3142 cm-' and vs(CH2) 3044 cm-'1 which are found in the region of the C-H stretchings in ethylene. Benzyl radical, C6H5CH2, and its deuterated analogues (C6H5CD2, C6D5CH2)were obtained by vacuum thermolysis of dibenzyl derivatives or by pyrolysis of the corresponding benzyl bromides (17) (Baskir, 1989).

The following frequencies were recorded in the matrix IR spectra of benzyl radicals: C6HsCH2: 3111, 3069, 1469, 1466, 1409, 1305, 1264, 1015, 948, 882, 862, 710, 667, 466 cm-'; C6H5CD2:1468, 1440, 1289, 1030, 880, 752, 668 cm-I; C6DSCH2:1410, 1201, 821, 811, 759, 711, 519cm-'.

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

43

GFsCF~I 750°C. lO-'Torr

c,F~CF~

40 K

C,F~CF~CF,C~F~

Scheme 5

Warming the matrix up to 40K resulted in the disappearance of the radical bands and the appearance of the corresponding toluene or dibenzyl bands. Perfluorobenzylradical has been obtained similarly (Scheme 5) by vacuum pyrolysis of perfluorobenzyl iodide or of perfluorodibenzyl (Baskir et al., 1989). The matrix IR spectrum of C6F5CF2radical contained the following bands: 1597, 1501, 1485, 1312, 1267, 892 cm-'. Upon warming of the matrix from 1 2 K to 40K the disappearance of these bands and the appearance of perfluorobibenzyl bands were observed. Based both on the determined isotopic shifts and the comparison of the radical IR spectrum with the spectra of various substituted benzenes, the bands have been assigned to the normal modes and the force field of the benzyl radical calculated (Table 8). The increase of the exocyclic C-C bond stretching frequency from 1208cm-' in toluene to 1264cm-' in the benzyl radical and the simultaneous decrease of the C-C ring bond stretching frequencies (from 1494 and 1460 cm-' to 1469 and 1446 cm-', respectively) result from electron density delocalization in the benzyl system. Furthermore, the force constant value for the C-C bond in the C6H5CH2 radical (5.5 X 102Nm-') is between the values for the ordinary C-C bond (4.5 x lo2 N m-') and the double C = C bond (9.0 X lo2 N m-') and is close to the corresponding force constant in the ally1 radical (5.8 X lo2N m-'). Compared with perfluorotoluene the stretching vibration frequency of the exocyclic C-C bond in perfluorobenzyl radical increase (from 1237 to 1267 cm-') whereas the stretchings of benzene ring decrease (from 1657, 1525 and 1510 to 1597, 1501 and 1485 cm-'). Finally, the cyclopentadienyl radical, C5H5, was obtained by vacuum pyrolysis (970"C, Torr) of bis(cyc1opentadienyl)nickel (18) and it was frozen into an argon matrix at 1 2 K (Nefedov, 1991a,b; Korolev and

V. A. KOROLEV AND 0. M. NEFEDOV

44

Nefedov, 1993). Only three bands (3079, 1383, 661cm-I) of the four possible ones for this highly symmetrical molecule (D5,,) have been found in the IR spectrum. The band at 1383 cm-’ belongs to a stretching of the C-C bond in C5H5. Comparison of this band with the corresponding band in the IR spectra of the cyclopentadienyl ligand of .rr-complexes (1410-1435 cm-l) and of free cyclopentadienyl anion (1455cm-’) (Table 9) leads to the conclusion that contrary to olefinic systems, the C-C bond stretching frequencies increase in the order: C5H5 (radical)+.rr-C5H5 (ligand)+C5H5(anion) due to an increase in the aromaticity of the cyclopentadienyl system.

Table8 IR spectra of free benzyl and perfluorobenzyl radicals, toluene and perfluorotoluene. ~

vlcm-’ Assignment

PhCHza

v/cm-’

PhCH3

Assignment

C6F5CF/

C6F5CF3

1597 1501 1485

1657 1525 1510 1434

3111 3069 1469 1446 1409 1305 1264 1015 948 882 762 710 667 466

2952 3063 1494 1460 1380 1331 1208 1030

1312 1267

1040 893 728 -

695 464

892

1350 1237 1192 1154 1090 994 875 714

“Baskir (1989); Baskir et al. (1993). bBaskir et al. (1989). Bold entries signify results discussed in the text.

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

45

Table 9 IR spectra of free cyclopentadienyl radical,“ cyclopentadienyl ligand and free cyclopentadienyl anion.

Assignment 3079

3075-3100

3039

1383

1410-1435 1000-1010

1455

66 1

770-825

1003 710

“Nefedov (1991a,b); Korolev and Nefedov (1993).

5 Unstable compounds with double-bonded silicon and germanium atoms (silenes, silanones, germanones, germathiones)

Compounds containing double-bonded silicon and germanium atoms are the nearest analogues of olefins, ketones and thioketones. However, most of them are very unstable and highly reactive species. The first successful stabilization of a silene, Me2Si=CH2 [117], in an argon matrix was achieved in experiments on the vacuum pyrolysis of 1,l-dimethylsilacyclobutane[118] (Mal’tsev et al., 1976b). The IR spectra of the silene [117] and some of its deuterated analogues have been recorded

P181

[1171

and their full assignment made (Mal’tsev et al., 1976b, 1977a, 1982b, 1984a; Nefedov et al., 1980; Khabashesku et al., 1983). The bands at 642.9, 817.5, 825.2, 1003.5, 1251.0, 1259.0cm-’ were assigned to [117]. The band at 1003.5 cm-I was specifically assigned to the silicon-carbon double bond stretching vibration. The force constant for the Si=C bond turned out to be 5.6 x lo2N m-’. The value is considerably higher than 3.06 X lo2 N m-l for a silicon-carbon single bond and is close to that of well-studied phosphonium ylides, called “salt-free ylides” (5.59 x lo2 N m-’). This similarity between Si=C and P=C rather than between Si=C and C = C reflects a certain polarity of the silicon-carbon double bond. It was also confirmed by reactions of silenes with polar trapping agents an& by ab initio calculations (Ahlrichs and Heinzmann, 1977; Apeloig and Karni, 1984). According to Siebert’s rule the calculated order of the Si=C bond is 1.62 (Siebert, 1953). A first attempt in argon matrices to stabilize silenes H2Si=CH2 and C12Si=CH2, which do not have multiatomic substituents at the silicon atom,

46

V. A. KOROLEV AND 0. M. NEFEDOV

CF, [1191

R

=

H, D, C1, Me

was performed in 1979 (Mal'tsev et af., 1979). However, at 7O0-80O0C, the starting compounds, silacyclobutane and 1, l-dichlorocyclobutane, decomposed to low-molecular hydrocarbons (CH4, C2H2, C2H4, etc.) as well as SiC12 (in the case of the second parent molecule). Probably this occurred due to the lower thermal stability of H2Si=CH2 and C12Si=CH2 in comparison with Me2Si= CH2. Nevertheless, the silenes R2Si=CH2 (R = H, D, C1, Me) were stabilized in argon matrices at 10 K as products of pyrolysis (19) of 6-silabicyclo [2,2,2]octa-2,5-dienes [119] (Maier et al., 1981, 1984b,c; Reisenauer et af., 1982). In contrast to silacyclobutanes, the compounds [119] decompose at lower temperatures, about 500°C. The values of the IR and UV absorptions of the silenes H2Si=CH2, D2Si=CH2, C12Si=CH2, and MeSiH=CH2, which were obtained in this manner, have been reported. An assignment of the vibrational bands of the silenes H2Si=CH2, D2Si=CH2 and C12Si=CH2 has been made on the basis of ab initio calculations of the spectra of these molecules. In particular, it was found that Si = C double bond stretching vibrations exhibit absorptions in the region near 1000 cm-' (H2Si=CH2, 985 cm-'; D2Si=CH2, 952 cm-'; C12Si=CH2, 1008 cm-'). These conclusions were confirmed later (Maier et al., 1984b). Matrix IR spectra of various silenes are important analytical features and allow detection of these intermediates in very complex reaction mixtures. Thus, the vibrational frequencies of Me2Si=CH2 were used in the study of the pyrolysis mechanism of allyltrimethylsilane [120] (Mal'tsev et al., 1983). It was found that two pathways occur simultaneously for this reaction (Scheme 6). On the one hand, thermal destruction of the silane [120] results in formation of propylene and silene [117] (retroene reaction); on the other hand, homolytic cleavage of the Si-C bond leads to the generation of free allyl and trimethylsilyl radicals. While both the silene [117] and allyl radical [115] were stabilized and detected in the argon matrix, the radical SiMe3 was unstable under the pyrolysis conditions and decomposed to form lowmolecular products. The IR spectrum of another silene homologue, trimethylsilaethene [ 1211, Me2Si=CHMe, was recorded after matrix stabilization of the products of

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

47

+ Me2Si=CH2 ~171

+ .SiMe3 [ll5]

I

C H 3 , CH4, C2H2, C2H4, Si Scheme 6

UV photolysis (Chapman et al., 1976; Chedekel et al., 1976) and vacuum pyrolysis (Mal’tsev et al., 1980) of trimethylsilyldiazomethane [ 1221. The silene formation occurred as a result of fast isomerization of the primary reaction product, excited singlet trimethylsilylcarbene [1231 (the ground state of this carbene is triplet). When the gas-phase reaction mixture was diluted with inert gas (helium) singlet-triplet conversion took place due to intermolecular collisions and loss of excitation. As a result the final products [124] of formal dimerization of the triplet carbene [123] were obtained. Nowadays silenes are well-known intermediates. A number of studies have been carried out to obtain more complex molecules having Si=C double bonds. Thus, an attempt has been made to generate and stabilize in a matrix l,l-dimethyl-l-silabuta-1,3-diene[125], which can be formed as a primary product of pyrolysis of diallyldimethylsilane [1261 (Korolev et al., 1985). However, when thermolysis was carried out at 750-800°C the absorptions of only two stable molecules, propene and 1 , l dimethylsilacyclobut-2-ene [ 1271, were observed in the matrix IR spectra of the reaction products. At temperatures above 800°C both silane [126] and silacyclobutene [127] gave low-molecular hydrocarbons, methane, acetylene, ethylene and methylacetylene. A comparison of relative intensities of the IR Me3SiCHN2

A or hu -Nz

Me3Si-CHS

Me3Si -CHT

Ar, 10K

-

Me2Si =CHMe

Me,SiCH =CHSiMe3

48

V. A. KOROLEV AND 0. M. NEFEDOV

bands of these molecules showed intermediate formation of the silabutadiene [125] in these reactions. It is probable that this species easily converts to [127] at temperatures below 800°C7but above 800°C it is thermally unstable and is decomposed to low-molecular products. A matrix isolation IR study of cyclic siladienes was more successful (Khabashesku et al., 1992). At first, unstable l-sila~yclopenta-2~4-diene (1281 was generated by vacuum pyrolysis (800°C; 10-3-10-5 Torr) of 5-silaspiro[4.4]nona-2,7-diene [ 1291 or pyrolysis and photolysis (A = 248 nm) of 1,l-diazido-l-silacyclopenta-2,4-diene [130]; it has been studied by UV and IR spectroscopy in an argon matrix at 12K. The UV band at A,,, = 278nm and nine IR bands (including two sp3 Si-H stretching vibrations at 2175 and 2144cm-') have been recorded in matrix spectra of [1281. Reversible photochemical interconversion of [1281 with silacy-

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

[ 135a]-[ 135~1

49

[137a], [137c]

a: R = H, R' = AcO b: R = H, R' = CHz=CHCHz c: R = Me, R' = CHzCHCHz

clopentadienes [131] and [132] as well as 1-silacyclopenta-1,l-diyl[133] has been observed after irradiation of the matrix at various wavelengths. Thus, irradiating [128] with broad-band UV light (A = 260-390 nm) leads to [133] (UV bands at 250 and 480nm) and traces of [131] and [132]. Further irradiation of the matrix with 488 nm visible light resulted in gradual removal of the bands of [133] and appearance of an absorption at 296 nm which has been attributed to [132]. Finally, the band at 270nm assigned to (1311 was observed in the matrix UV spectra of 308 nm photolysis products of [128]. According to ab initio calculations at the 6-31G* level, the IR bands at 936 and 929 cm-' are attributable to the Si=C stretching vibrations in [131] and [132], respectively, and are lower than those in unconjugated silenes (cf. 989 cm-' in MeSiH=CH2). This, together with the red-shifted UV absorptions of [131] and [132] (in comparison with silenes), provides evidence of the presence of Si=C-C=C .rr-conjugation in [131] and [132]. Similar studies of molecules having a Si=C double bond in a silahexatriene ring have also been carried out. An analysis of matrix IR spectra of pyrolysis products of various parent compounds (l-silacyclohexa-2,5-diene [134] (Maier et al., 1982, 1984a) , l-acetoxy-l-silacyclohexa-2,4-diene [ 135al [ 135bl (Maier et al., 1980, 1984a) as and l-allyl-l-silacyclohexa-2,4-diene well as the composition of eliminated molecules (H2, AcOH and C3H6, respectively) led to the conclusion that the intermediate which was observed in these reactions is silabenzene [137a]. At the same time, silatoluene [137c] was generated by pyrolysis of l-methyl-l-allyl-l-silacyclohexa-2,4-diene [135c] (Kreil et al., 1980). Unfortunately normal coordinate analysis of silabenzene and silatoluene has not been made in spite of great interest in the changes of their vibrational frequencies in comparison with those of benzene and toluene. Nevertheless, UV-vis absorption spectra of silabenzene and silatoluene show red shifts of their bands relative to those of benzene. These shifts in the longer wavelength region are typical for other aromatic heterocycles (e.g. for phospha-, arsa- and stibabenzenes in comparison with pyridine). Thus, UV spectra of silabenzene and silatoluene are in agreement with the presence of a silicon atom in the cyclic aromatic system.

V. A. KOROLEV AND 0. M. NEFEDOV

50

UV irradiation (A>320 nm) of matrix-isolated silabenzene [137a] led to the disappearance of its absorptions and the appearance of bands of other unstable molecules, which were assigned to Dewar silabenzene [138]. This assumption is roasonable because sp3 hybridization of the silicon atom is preferable to sp2 hybridization and favourable to the stability of [ 1381. Besides, according to calculations the energy differences between monocyclic [137a] and bicyclic [138] structures decrease from C6H6 (314 kJ mol-') to silabenzene C5SiH6 (125 kJ mol-').

[ 137al

~381

A number of relatively stable silenes with bulky substituents are known at present, but stable silanones have not been isolated till now. Their instability, like that of silenes, is caused by a kinetic factor, according to various calculations. Thus, cyclooligomerization of silanones should proceed with zero activation barrier (Kudo and Nagase, 1985). The first inorganic silanones, C12Si=0 and F2Si=0, were obtained by UV irradiation of argon matrices containing silicon monoxide and halogens (Schnockel, 1978, 1980a,b). The first organic silanone, Me2Si=0, was detected by IR as one of the products of vacuum pyrolysis of the 6-oxa-3-silabicyclo[3.1 .O]hexanes [ 1401 (Khabashesku et al., 1986). The bands at 1244, 1240, 1210, 822, 798, 770, 657 cm-' in the matrix IR spectra were assigned to this intermediate, and a full assignment of various vibration modes was made on the basis of IR spectra of isotopically labelled species (Khabashesku et al., 1986, 1988a,b) (Table 10). As follows from the force field calculations, the Si =0 stretching vibrations for both isotopomers are very characteristic; the contributions of the potential energy of Si= 0 stretching to vibrations at 1210 cm-' (Me2Si=0) and 1215 cm-' (Me2Si=O-d6) are about 90%. Slightly later Me2Si=0 was produced in inert matrices by cophotolysis of Me2SiH2 and Me3SiH with ozone (Withnall mwwc, I O - ~ T O ~ Ar. 12K

/ /

Me Me ~401 R = H, Me

Me2Si=0 [I391

40K

(Me2SiO)3 ~411

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

Table 10 Vibrational assignment in dimethylsilanone-d6.'

IR

51

spectra of

(CH3)2Si=0 v/cm-' Symmetry

Exp.

dimethylsilanone and (CD&Si=O

v/cm-'

Calc.

PED/%b

Exp.

Calc.

1397 1395 1395 1395 1247

966(CH3) 96 S(CH3) 96S(CH3) 96 S(CH3) 54 W H 3 ) , 46 P(CH3) 56 S(CH3), 42 P(CH3) 89 v(Si=O) 57 v(Si-C), 41 P(CH1) 96 P(CH3) 96 P(CH3) 96 P W 3 ) 55 P(CH3), 44 v(Si-C) 86 4%-C)

1032

1007

1011 1008 1007 1007 976

995

973

1215 685

1217 638

67 1

611 594 592 773

B1

1240

1235

Al

Bl

1210 822

1218 848

B2 Al

798

A2 Bl

770

810 808 795 772

Al

657

657

712

PEDI%~

581

"Khabasheskuet a / . (1986, 1988a,b). bPotential energy distribution.

and Andrews, 1986). The frequencies obtained for Me2Si=0 were in good agreement. In each experiment an unstable dimer (Me2Si0)2 was also observed in addition to other pyrolysis products. The presence of (Me2Si0)2 was confirmed by mass spectroscopic study (Tamas et al., 1988). At a pyrolysis pressure higher than Torr the bands of Me2Si=0 were not observed, whereas the bands of (Me2Si0)2 were more intense, indicating that dimerization of dimethylsilanone took place. Dimethylsilanone has also been detected in the matrix IR spectra of the pyrolysis products of siloxy compounds [142] and [143] (Khabashesku et al., 1988~).

V. A. KOROLEV AND 0. M. NEFEDOV

52

Table 11 Thermal and kinetic stability of silanones. Thermal stability of silanones R2Si=O+ SiO + 2R Initial temperature"of silanone decompositionPC Kinetic stability of silanones Initial dimerization pressure/Torr

PhzSi=O

Me2Si=0

(MeO)$i=O

900

850

800

(MeOhSi=O >lo-'

PhzSi=O >10-2

Me2Si=0 >5 x 10-4

The calculated force constant for the Si=O bond in MezSi=O is 8.32 x 102N m-'. This agrees with similar parameters for X2Si=0 (X = C1, F) being equal to 9 x lo2N rn-l (Schnockel, 1978, 1980a,b), whereas the force constant of the S i - 0 single bond is roughly 5.3 x 102Nm-'. The calculated order of the Si=O bond is 1.45. In order to study the influence of substituents on the spectral behaviour and on the thermal and kinetic stability of silanones, two new organic silanones, dimethylsilicate [ 144al and silabenzophenone [ 144b], have been examined. These were obtained in the vacuum pyrolysis of the corresponding epoxides [145a] and [145b] (Khabashesku et al., 1988~).

-

R,Si=O

40K

(R2Si0)3

[144a], [144b] [145a], [145b] a: R = M e 0 b: R = Ph

It has been found that, among the silanones studied, silabenzophenone seems to be the most thermally stable while dimethyl silicate is the kinetically most stable (Table 11). An attempt to obtain and stabilize dimethylsilathione, Me2Si=S [146], in an inert matrix has been made by pyrolysis of hexamethylcyclotrisilatrithiane [ 1471 and 3,3-dimethyl-3-silathietane [148] (Gusel'nikov et al., 1983). However, thermolysis of [147] led to the formation of only the cyclodimer of Me2%= S, tetramethyldisiladithietane [149]. At the same time, the compound [ 1481 was decomposed giving 1,l-dimethylsilaethene (Me2Si=CH2), thioformaldehyde (H2C=S) and the dimer [149], which were detected in the matrix IR spectra of the pyrolysate. The bands of the silathione have not been found.

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

-

s'%Me2 .

Me&"

I

53

-

A

I

KSj/S Me2

[MezSi=S]

Me2Si-S

I 1

S--SiMe2

\

[I471

S

Me2Si=CH2

A

[

__t

H2C=S +

a i M e 2

[149]

r1iMe2]

Recently the first silathione C12Si=S has been generated (20) and stabilized in an argon matrix (Schnoeckel et al., 1989). hu

hu

SiS+ C12

C12Si=S

Ar. 1 2 K

SiCh + COS

(20)

v (Si=S) 805.6 cm-'

Some years ago the first inorganic germanone, F2Ge= 0, was produced by reaction of GeO with F2 in an argon matrix under photolysis, and v(Ge=O) was found to be 989.9cm-' (Schnoeckel, 1981). Recently three new inorganic germanones have been characterized (Withnall and Andrews, 1990). The first organic germanone MezGe =0 [ 1501 was successfully detected .O]hexanes [151] upon vacuum pyrolysis of the 6-oxa-3-germabicyclo[3.1 (Khabashesku et al., 1990a,b). Seven bands at 1241, 1231, 972, 796,606,524 and 465 cm-' were assigned to the germanone [150] (Table 12). The force constant for the G e = O bond has been calculated to be 7.2 x lo2 N m-' [cf. F(Ge=O) in GeO is 7.34 X 1O2Nrn-' (Ogden and Richs, 1970)l. The calculated order of the G e = O bond is 1.75 and its frequency is 972 cm-'.

800-850"C. 10'3Torr Ar, 12 K

Ge

/ /

Me

Me

[1511 R = H, Me

* MezGe = 0 [1501

40K

(Me2GeO)3

V. A. KOROLEV AND 0. M. NEFEDOV

54

Table 12 Vibrational assignment in the IR spectrum of dimethylgermanone Me2Ge=0.'

Vibrational frequencies/cm-' Exp.

Calc.

Potential energy distribution/%

1241 1231 972 796 606 524 465

1237 1233 972 796 606 526 455

53 a(CH3), 47 p(CH3) 53 a(CH3), 48 p(CH3) 97 v( Ge =0) 97 P(CH3) 88 v( Ge -C) 96 v( Ge -C) 100 4CGeC)

"Khabashesku et al. (1990a,b).

By using epoxysilacyclopentane [1521 and epoxygermacyclopentane [ 1531 as suitable precursors of Si=O and Ge=O containing species, attempts have been made to generate in the gas phase the monomeric S i 0 2 and G e 0 2 , produced earlier by reactions in inert matrices at 10-20 K (Bos et al., 1974; Schnoeckel, 1978). Monomeric Si02 has been stabilized in an argon matrix (band at 1419cm-'), but G e 0 2 could not be detected, although the band of monomeric GeO at 976 cm-' was found in the matrix IR spectra of the pyrolysis products. Recently the first germathione, Me2Ge= S, was obtained independently (Barrau et al., 1989a) and (Khabashesku et al., 1989, 1991) in an argon matrix by vacuum pyrolysis of (Me2GeS)3 (21). The IR bands assigned to Me2Ge=S by both groups are basically in agreement. As follows from the force field calculations (Table 13), the Ge=S stretching vibration is Me

/

70&11WC, lo-' Torr

O=Si=O

Ar, 12K

Me

/

8oo-9o(pc,10-3 TO^

Ar, 12K

Me

11531

+

[O=Ge=O]

-

(GeO&

+ GeO

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

55

(MezGeS)2+ MezGe=S

45 K

non-characteristic due to a strong mixing with the symmetrical Ge -C vibration, which results in splitting into two frequencies at 606 and 518 cm-', with the greatest contribution of v(Ge=S) stretching in the latter. Mass spectrometric study of the pyrolysis of (Me2GeS)3 has resulted in the determination of the ionization energy of Me2Ge=S (8.63 & 0.1 eV), practically coinciding with that obtained earlier by photoelectron spectroscopy (8.60 eV) (Guimon et al., 1985). Matrix IR and Raman spectra of tetramethyldigermethene, Me2Ge= GeMe2, which was obtained by pyrolysis (140-200°C) of compound [154], have been recorded (Bleckmann et al., 1984). Since this intermediate GeMe2

14C-20o"C Ar or

-

Nz.10 K

Me2Ge=GeMe2+ Ph

Table 13 Vibrational assignment in the IR spectra of dimethylgermathione Me2Ge=S .

Vibrational frequencies/cm-' Exp."

1390 1229 850 761 753 605 u(Ge =S ) 516

EX^.^

Calc.b

1407 1407 1392 1237

1382 1382 1382 1231

1231

1227

850 809 765 606

812 809 806 605

574 518

573 518

"Barrau et al. (1989a). bKhabashesku et al. (1989, 1991).

Potential energy distribution/% 99 a(HCH) 99 a(HCH) 99 a(HCH) 53 a(HCH), 49 P(HCGe) 51 a(HCH), 50 P(HCGe) 95 P( HCGe) 97 P(HCGe) 97 P(HCGe) 43 u(Ge=S), 52 u(GeC) 99 v(GeC) 56 v(Ge =S ) , 36 u(GeC)

56

V. A. KOROLEV AND 0. M. NEFEDOV

has a symmetrical Ge=Ge double bond, its vibration is inactive in the infrared. The frequency value (404 cm-') measured by Raman spectroscopy is significantly higher than that of a Ge-Ge single bond (270-300 cm-').

6 Conclusions The results described in this review show that matrix stabilization of reactive organic intermediates at extremely low temperatures and their subsequent spectroscopic detection are convenient ways of structural investigation of these species. IR spectroscopy is the most useful technique for the identification of matrix-isolated molecules. Nevertheless, the complete study of the spectral properties and the structure of intermediates frozen in inert matrices is achieved when the IR spectroscopy is combined with UV and esr spectroscopic methods. At present theoretical calculations render considerable assistance for the explanation of the experimental spectra. Thus, along with the development of the experimental technique, matrix studies are becoming more and more complex. This fact allows one to expect further progress in the matrix spectroscopy of many more organic intermediates.

References Ahlrichs, R. and Heinzmann, R. (1?77). J. Am. Chem. SOC. 99, 7452 Ammann, J., Subramanian, R. and Sheridan, R. S. (1992). J. Am. Chem. SOC. 114, 7592 Apeloig, Y. and Karni, M. (1984). J. Am. Chem. SOC. 106, 6676 Arrington, C. A., Klingensmith, K. A., West, R. and Michl, J. (1984). J. Am. Chem. SOC. 106, 525 Ase, P., Bock, W. and Snelson, A. (1986). J. Phys. Chem. 90, 2099 Baird, M. S., Dunkin, I. R., Hacker, N., Poliakoff, M. and Turner, J. J. (1981). J . Am. Chem. SOC.103, 5190 Barrau, J., Balaji, V. and Michl, J. (1989a). Organometallics 8, 2034 Barrau, J., Bean, D. L., Welsh, K. M., West, R. C. and Michl, J. (1989b). Organometallics 8, 2606 Baskir, E. G. (1989). Ph.D. Thesis. Institute of Organic Chemistry. Russian Academy of Sciences, Moscow Baskir, E. G., Korolev, V. A., Mal'tsev, A. K., Ujszaszy, K., Tamas, J. and Nefedov, 0. M. (1989). Izv. Akad. Nauk SSSR. Ser. Khim. 818 Baskir, E. G., Korolev, V. A., Mal'tsev, A. K., Khabashesku, V. N. and Nefedov, 0 . M. (1993). Izv. Akad. Nauk. Ser. Khim. 1499 Bell, G. A. and Dunkin, I. R. (1985). J. Chem. SOC., Faraday Trans. 2 81, 725 Bleckmann, P., Minkwitz, R., Neumann, W. P., Schriewer, M., Thibud, M. and Watta, B. (1984). Tetrahedron Lett. 25, 2467 Bock, H., Mohmand, S., Hirabayashi, T., Maier, G. and Reisenauer, H. P. (1983). Chem. Ber. 116, 273.

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

57

Bohn, R. B., Brabson, G. D. and Andrews, L. (1992). J. Phys. Chem. 96, 1582 Bokyi, N. G., Struchkov, Yu. T., Kolesnikov, S. P., Rogozhin, I. S. and Nefedov, 0. M. (1975). Izv. Akad. Nauk SSSR. Ser. Khim. 812 Bos, A., Ogden, J. S. and Ogree, L. (1974). J. Phys. Chem. 78, 1763 Bucher, G. and Sander, W. (1992). Chem. Ber. 125, 1851 Bucher, G., Sander, W., Kroka, E. and Cremer, D. (1992). Angew. Chem. 104, 1225 Butler, R. L. and Snelson, A. (1979). J. Phys. Chem. 83, 3243 Butler, R. L. and Snelson, A. (1980a). J. Fluorine Chem. 15, 89 Butler, R. L. and Snelson, A. (1980b). J. Fluorine Chem. 15, 345 Butler, R. L. and Snelson, A. (1980~).J. Fluorine Chem. 16, 33 Chapman, 0. L., Chang, C. C., Kolc, J., Jung, M. E., Lowe, J. A., Barton, T. J. and Tumey, M. L. (1976). J. Am. Chem. SOC. 98, 7844 Chapman, 0. L. and Sheridan, R. S. (1979). Prepr. Div. Pet. Chem. A m . Chem. SOC. 24, 130 Chapman, 0. L., Sheridan, R. S. and Le Roux, J. P. (1978). J . A m . Chem. SOC. 100, 6245 Chapman, 0. L., McMahon, R. J. and West, P. R. (1984). J . A m . Chem. SOC.106, 7973 Chedekel, M. R., Skoglund, M., Kreeger, R. L. and Shechter, H. (1976). J . A m . Chem. SOC. 98, 7846 Chernyshev, E . A. and Komalenkova, N. G . (1990). Uspekhi Khimii 59, 918 Chettur, G . and Snelson, A. (1987a). J. Phys. Chem. 91, 913 Chettur, G. and Snelson, A. (1987b). J. Phys. Chem. 91, 3483 Chettur, G. and Snelson, A. (1987~).J . Phys. Chem. 91, 5873 Cradock, S. and Hinchcliffe, A. J. (1975). Matrix Isolation. Cambridge University Press, London Cyvin, B. N., Cyvin, S. J. and Snelson, A . (1986). Z. Anorg. Allg. Chem. 542, 193 Dendramis, A. and Leroi, G. E. (1977). J . Chem. Phys. 66, 4334 Dendramis, A., Harrison, J. F. and Leroi, G. E. (1978). Ber. Bunsenges. Phys. Chem. 82, 7 Doyle, T. J., Shen, L. N., Rittby, C. M. L. and Graham, W. R. M. (1991). J. Chem. Phys. 95, 6224 Du, X. M., Fan, H., Goodman, J. L., Kesselmayer, M. A., Krogh-Jespersen, K., La Villa, J. A., Moss, R. A., Shen, S. and Sheridan, R. S. (1990). J. A m . Chem. SOC. 112, 1920 Fredin, L., Hauge, R. H., Kafafi, Z. H. and Margrave, J. L. (1985). J. Chem. Phys. 82, 3542 Fujitake, M. and Hirota, E. (1989). J. Chem. Phys. 91, 3426 Ganzer, G. A., Sheridan, R. S. and Liu, M. T. H. (1986). J. A m . Chem. SOC. 108, 1517 Guimon, C., Pfister-Guillouzo, G., Rima, G., El-Amine, M. and Barrau, J. (1985). Spectroscopy Lett. 18, 7 Gusel’nikov, L. E., Volkova, V. V., Avakyan, V. G., Nametkin, N. S., Voronkov, M. G., Kirpichenko, S. V. and Suslova, E. N. (1983). J. Organomet. Chem. 254, 173 Hargittai, I., Schultz, G., Tremmel, J., Kagramanov, N.,D., Mal’tsev, A. K. and Nefedov, 0. M. (1983). J. A m . Chem. SOC. 105, 2895 Hayes, R. A., Hess, T. C., McMahon, R. J. and Chapman, 0. L. (1983). J . Am. Chem. SOC.105, 7786 Ho, G.-J., Krogh-Jespersen, K., Moss, R.A., Shen, S., Sheridan, R. S. and Subramanian, R. (1989). J. A m . Chem. SOC. 111, 6875

58

V. A. KOROLEV AND 0. M. NEFEDOV

Holtzhauer, K., Cometta-Morini, C. and Oth, J. F. M. (1990). J. Phys. Org. Chem. 3, 219 Huang, J. W. and Graham, W. R. M. (1990). J. Chem. Phys. 93, 1583 Ismail, Z. K., Hauge, R. H., Fredin, L. and Margrave, J. L. (1982a). J. Chem. Phys. 77, 1617 Ismail, Z . K., Fredin, L., Hauge, R. H. and Margrave, J. L. (1982b). J. Chem. Phys. 77, 1626 Jacox, M. E. (1979). Chem. Phys. 43, 157 Jacox, M. E . (1988). J . Chem. Phys. 88, 4598 Jacox, M. E . and Milligan, D. E. (1967). J. Chem. Phys. 47, 1626 Jacox, M. E. and Milligan, D. E. (1969). J. Chem. Phys. 50, 3252 Jacox, M. E. and Milligan, D. E. (1974). Chem. Phys. 4, 45 Jacox, M. E. and Olson, W. B. (1987). J. Chem. Phys. 86, 3134 Kagramanov,". D., Kazansky, V. B., Mal'tsev, A. K., Nefedov, 0. M., Shelimov, B. N. and Schteinschneider, A. Ya. (1977). Dokl. Akad. Nauk. SSSR 237, 140 Kagramanov, N. D., Mal'tsev, A. K., Dubinsky, M. Yu. and Nefedov, 0. M. (1983a). Izv. Akad. Nauk SSSR. Ser. Khim. 536 Kagramanov, N. D., Ujszaszy, K., Tamas, J., Mal'tsev, A. K. and Nefedov, 0. M. (1983b). Izv. Akad. Nauk SSSR. Ser. Khim. 1683 Kesselmayer, M. A. and Sheridan, R. S. (1986a). J. A m . Chem. SOC. 108, 99 Kesselmayer, M. A . and Sheridan, R. S. (1986b). J. A m . Chem. SOC. 108, 844 Kesselmayer, M. A. and Sheridan, R. S. (1987). J. A m . Chem. SOC. 109, 5029 Khabashesku, V. N., Baskir, E. G., Mal'tsev, A. K. and Nefedov, 0. M. (1983). Izv. Akad. Nauk SSSR. Ser. Khim. 238 Khabashesku, V. N., Kerzina, Z. A., Mal'tsev, A. K. and Nefedov, 0. M. (1986). Izv. Akad. Nauk SSSR. Ser. Khim. 1215. Khabashesku, V. N., Kerzina, Z. A., Mal'tsev, A. K. and Nefedov, 0. M. (1988a). In Silicon Chemistry (ed. Corey, M. T., Corey, E. R. and Gaspar, P. P.). Ellis Horwood Ltd, Chichester Khabashesku, V. N., Kerzina, Z. A., Baskir, E. G., Mal'tsev, A. K. and Nefedov, 0. M. (1988b). J. Organomet. Chem. 347, 277 Khabashesku, V. N., Kerzina, Z. A. andNefedov, 0. M. (1988~).Izv. Akad. Nauk SSSR. Ser. Khim. 2187 Khabashesku, V. N., Boganov, S. E., Kerzina, Z. A., Nefedov, 0. M., Tamas, J., Gomory, A. and Besenyei, I. (1989). 36th International Conference on Organometallic and Coordination Chemistry of Germanium, Tin and Lead Compounds, Brussels. Abstr. P.37 Khabashesku, V. N., Boganov, S. E. and Nefedov, 0. M. (1990a). Izv. Akad. Nauk SSSR. Ser. Khim. 1199 Khabashesku, V. N., Kerzina, Z. A., Boganov, S. E. and Nefedov, 0. M. (1990b). IX International Symposium on Organosilicon Chemistry, Edinburgh .' Abstr. P.8.25 Khabashesku, V. N., Boganov, S. E., Zuev, P. S., Nefedov, 0. M., Tamas, J., Gomory, A. and Besenyei, I. (1991). J. Organomet. Chem. 402, 161 Khabashesku, V. N., Balaji, V., Boganov, S. E., Bashkirova, S. A., Matveichev, P. M., Chernyshev, E. A., Nefedov, 0. M. and Michl, J. (1992). Mendeleev Commun. 38 Korolev, V. A. and Nefedov, 0. M. (1993). Izv. Akad. Nauk SSSR. Ser. Khim. 1497 Korolev, V. A., Mal'tsev, A. K. and Nefedov, 0. M. (1985). Izv. Akad. Nauk SSSR. Ser. Khim. 711 Korolev, V. A., Mal'tsev, A. K. and Nefedov, 0. M. (1989). Izv. Akad. Nauk SSSR. Ser. Khim. 1058

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

59

Kreil, C. L., Chapman, 0. L., Burns, G . T. and Barton, T. J. (1980). J . Am. Chem. SOC. 102, 841 Kudo, T . and Nagase, S. (1985). J . Am. Chem. SOC. 107, 2589 Kulishov, V. I . , Bokyi, N. G., Struchkov, Yu. T., Nefedov, 0. M., Kolesnikov, S. P. and Perlrnutter, B. L. (1970). Zh. Strukt. Khim. 11, 71 Lee, Y. P. and Pirnentel, G. C. (1981). J . Chem. Phys. 75, 4241 Maass, G., Mal‘tsev, A. K. and Margrave, J. L. (1973). J. Inorg. Nucl. Chem. 35, 1945 McMahon, R. J. and Chapman, 0. L. (1986). J . Am. Chem. SOC. 108, 1713 McMahon, R. J. and Chapman, 0. L. (1987). J . Am. Chem. SOC. 109, 683 Maier, G., Mihrn, G . and Reisenauer, H. P. (1980). Angew. Chem. 92, 58 Maier, G., Mihrn, G . and Reisenauer, H. P. (1981). Angew. Chem. 93, 615 Maier, G., Mihrn, G. and Reisenauer, H. P. (1982). Chem. Ber. 115, 801 Maier, G., Reisenauer, H. P., Rohde, B. and Dehnicke, K. (1983). Chem. Ber. 116, 732 Maier, G., Reisenauer, H. P., Baurngartner, R. 0. W. and Mihrn, G. (1984a). Chem. Ber. 117, 2337 Maier, G., Mihrn, G . and Reisenauer, H. P. (1984b). Chem. Ber. 117, 2351 Maier, G., Mihrn, G., Reisenauer, H. P. and Littrnan, D. (1984~).Chem. Ber. 117, 2369 Maier, G., Reisenauer, H. P., Schwab, W . , Carsky, P., Hess, B. A. and Schaad, L. J. (1987). J . Am. Chem. SOC. 109, 5183 Maier, G.,‘Reisenauer, H. P., Shaefer, U: and Balli, H. (1988). Angew. Chem. 100, 590 Maier, G., Reisenauer, H. P., Schwab, W . , Carsky, P., Spirko, V., Hess, B. A. Jr. and Schaad, L. J. (1989a). J . Chem. Phys. 91, 4763 Maier, G., Glatthaar, J. and Reisenauer, H. P. (1989b). Chem. Ber. 122, 2403 Mal‘tsev, A. K., Mikaelian, R. G. and Nefedov, 0. M. (1971a). Izv. Akad. Nauk SSSR. Ser. Khim. 199 Mal‘tsev, A. K . , Mikaelian, R. G . and Nefedov, 0. M. (1971b). Dokl. Akad. Nauk SSSR 201, 901 Mal’tsev, A. K., Mikaelian, R. G., Nefedov, 0. M., Hauge, R. H. and Margrave, J. L. (1971~).Proc. Natl. Acad. Sci. USA 68, 3238 Mal‘tsev, A. K., Nefedov, 0. M., Hauge, R. H., Margrave, J. L. and Seyferth, D. (1971d). J. Phys. Chem. 75, 3984 Mal’tsev, A. K., Kagrarnanov, N. D. and Nefedov, 0. M. (1974). Zzv. Akad. Nauk SSSR. Ser. Khim. 1993 Mal’tsev, A. K., Kagrarnanov, N. D. and Nefedov, 0. M. (1975). Dokl. Akad. Nauk SSSR 224, 630 Mal’tsev, A. K., Svyatkin, V. A. and Nefedov, 0. M. (1976a). Dokl. Akad. Nauk SSSR 227, 1151 Mal’tsev, A. K., Khabashesku, V. N. and Nefedov, 0. M. (1976b). Izv. Akad. Nauk SSSR. Ser. Khim. 1193 Mal‘tsev, A. K., Khabashesku, V. N. and Nefedov, 0. M. (1977a). Dokl. Akad. Nauk SSSR. 233, 421 Mal’tsev, A. K., Kagrarnanov, N. D. and Nefedov, 0. M. (1977b). Izv. Akad. Nauk SSSR. Ser. Khim. 1835 Mal’tsev, A. K., Khabashesku, V. N. and Nefedov, 0. M. (1979). Izv. Akad. Nauk SSSR. Ser. Khim. 2152 Mal’tsev, A. K., Korolev, V. A . , Khabashesku, V. N. and Nefedov, 0. M. (1980). Dokl. Akad. Nauk SSSR 251, 1166 Mal’tsev, A. K., Korolev, V. A. and Nefedov, 0. M. (1982a). Izv. Akad. Nauk SSSR. Ser. Khim. 2415

60

V. A. KOROLEV AND 0. M. NEFEDOV

Mal‘tsev, A. K., Khabashesku, V. N. and Nefedov, 0. M. (1982b). J. Organornet. Chem. 226, 11 Mal‘tsev, A. K., Korolev, V. A., Kagramanov, N. D. and Nefedov, 0. M. (1983). Izv. Akad. Nauk SSSR. Ser. Khim. 1078 Mal’tsev, A. K., ‘Khabashesku, V. N. and Nefedov, 0. M. (1984a). J . Organornet. Chem. 271, 55 Mal’tsev, A. K., Korolev, V. A. and Nefedov, 0. M. (1984b). Izv. Akad. Nauk SSSR. Ser. Khim. 555 Mal’tsev, A. K., Zuev, P. S. and Nefedov, 0. M. (1985a). Izv. Akad. Nauk SSSR. Ser. Khim. 957 Mal‘tsev, A. K., Zuev, P. S. and Nefedov, 0. M. (1985b). Izv. Akad. Nauk SSSR. Ser. Khim. 2159 Mal’tsev, A. K., Baskir, E. G., Kagramanov, N. D. and Nefedov, 0. M. (1986). Izv. Akad. Nauk SSSR. Ser. Khim. 1998 Mal’tsev, A. K., Zuev, P. S. and Nefedov, 0. M. (1987a). Izv. Akad. Nauk SSSR. Ser. Khim. 463 Mal’tsev, A. K., Zuev, P. S., Tomilov, Yu. V. and Nefedov, 0. M. (1987b). Izv. Akad. Nauk SSSR. Ser. Khim. 2202 Mal‘tsev, A. K., Zuev, P. S . , Tomilov, Yu. V. and Nefedov, 0. M. (1989). Tetrahedron Lett. 763 Milligan, D. E. and Jacox, M. E. (1968a). J . Chem. Phys. 48, 2265 Milligan, D. E. and Jacox, M. E. (1968b). J . Chem. Phys. 49, 3130 Milligan, D. E. and Jacox, M. E. (1970). J . Chem. Phys. 52, 2594 Nefedov, 0. M., Mal’tsev, A . K. and Mikaelian, R. G. (1971). Tetrahedron Lett. 4125 Nefedov, 0. M., Mal’tsev, A. K. and Svyatkin, V. A. (1974). Izv. Akad. Nauk SSSR. Ser. Khim. 958 Nefedov, 0. M., Mal’tsev, A . K. and Svyatkin, V. A . (1976). Izv. Akad. Nauk SSSR. Ser. Khim. 1901 Nefedov, 0 . M., Mal’tsev, A. K., Khabashesku, V. N. and Korolev, V. A. (1980). J . Organomet. Chem. 201, 123 Nefedov, 0. M. (1991a). Pure Appl. Chem. 63, 231 Nefedov, 0. M. (1991b). Izv. Akad. Nauk SSSR. Ser. Khim. 2425 O’Gara, J. E . and Dailey, W. P. (1992). J. Am. Chem. SOC. 114, 3581 Ogden, J. S. and Richs, M. J. (1970). J . Chem. Phys. 52, 352 Ortman, B. J., Hauge, R. H., Margrave, J. L. and Kafafi, Z. H. (1990). J . Phys. Chem. 94, 7973 Pacansky, J. and Chang, J. S. (1981). J . Chem. Phys. 74, 5539 Pacansky, J. and Brown, D. W. (1983). J. Phys. Chem. 87, 1553 Pacansky, J. and Waltman, R. J. (1989). Spectrosc. Lett. 22, 739 Pacansky, J . , Koch, W. and Miller, M. D. (1991). J . Am. Chem. SOC.113, 317 Prochaska, F. T. and Andrews, L. (1980). J . Chem. Phys. 73, 2651 Raabe, G., Vancik, H., West, R. and Michl, J. (1986). J. Am. Chem. SOC. 108, 671 Reisenauer, H. P., Mihm, G. and Maier, G. (1982). Angew. Chem. 94, 864 Reisenauer, H. P., Maier, G., Riemann, A. and Hoffmann, R. W. (1984). Angew. Chem. 96, 596 Roth, W. R., Langer, R., Bartmann, M., Stevermann, B . , Maier, G., Reisenauer, H. P., Sustmann, R. and Miiller, W. (1987a). Angew. Chem. 99, 271 Roth, W. R., Kowalczi, K. U., Maier, G., Reisenauer, H. P., Sustmann, R. and Miiller, W. (1987b). Angew. Chem. 99, 1330 Sander, W., Miiller, W. and Sustmann, R. (1988). Angew. Chem. 100, 577 Schnockel, H. (1978). Angew. Chem. 90, 638

MATRIX IR SPECTROSCOPY OF INTERMEDIATES

61

Schnockel, H. (1980a). Z . Anorg. Allg. Chem. 460, 37 Schnockel, H. (1980b). J . Mol. Struct. 65, 115 Schnockel, H. (1981). J. Mol. Struct. 70, 183 Schnockel, H., Gloecke, H. J. and Koeppe, R. (1989). Z . Anog. Allg. Chem. 578, 159 Schultz, G., Tremmel, J., Hargittai, I., Berecz, I . , Bohatka, S . , Kagramanov, N. D., Mal’tsev, A. K. and Nefedov, 0. M. (1979). J . Mol. Struct. 55, 207 Schultz, G . , Tremmel, J., Hargittai, I . . Kagramanov, N. D., Mal‘tsev, A. K. and Nefedov, 0. M. (1982). J. Mol. Struct. 82, 107 Seburg, R. A . and McMahon, R. J. (1992). J . A m . Chem. Soc. 114, 7183 Seyferth, D. (1972). Acc. Chem. Res. 5 , 65 Shen, L. N., Doyle, T. J. and Graham, W. R. M. (1990). J . Chem. Phys. 93, 1597 Shepherd, R. A. and Graham, W. R. M. (1987). J. Chem. Phys. 86, 2600 Shepherd, R. A., Doyle, T. J. and Graham, W. R. M. (1988). J . Chem. Phys. 89, 2738 Sheridan, R. S., Moss, R. A., Wilk, B. K., Shen, S., Wlostowski, M., Kesselmayer, M. A., Subramanian, R., Kmeicik-Lawrynowicz, G. and Krogh-Jespersen, K. (1988). J . A m . Chem. SOC. 110, 7563 Siebert, H. (1953). Z . Anorg. Allg. Chem. 273, 170 Smith, D. V. and Leroi, G. E. (1969). Spectrochim. Acta 25A, 1917 Smith, G. R. and Guillory, W. A. (1972). J . Chem. Phys. 56, 1423 Snelson, A. (1970a). J . Phys. Chem. 74, 537 Snelson, A. (1970b). High Temp. Sci. 2 , 70 Snelson, A., Cyvin, B. N. and Cyvin, S. J. (1981). Z . Anorg. Allg. Chem. 482, 133 Sodeau, J. R. and Lee, E. X. C. (1978). Chem. Phys. Lett. 57, 71 Svyatkin, V. A., Mal’tsev, A. K. and Nefedov, 0. M. (1977). Izv. Akad. Nuuk SSSR. Ser. Khim. 2236 Tamas, J . , Gomory, A., Besenyei, I . , Nefedov, 0. M., Khabashesku, V. N., Kerzina, Z . A , , Kagramanov, N. D. and Mal‘tsev, A. K. (1988). J . Orgunomet. Chem. 349, 37 Tanimoto, M., Takeo, H., Matsumura, C., Fujitake, M. and Hirota, E. (1989). J . Chem. Phys. 91, 2102 Torres, M., Bourdelande, J. L., Clement, A. and Strausz, 0. P. (1983). J. A m . Chem. SOC. 105, 1698 Ujszaszy, K., Tamas, J., Kagramanov, N. D., Mal’tsev, A. K. and Nefedov, 0. M. (1980). J . Anal. Appl. Pyrolysis 2, 231 Vaida, E., Tremmel, J . , Rozsondai, B., Hargittai, I., Kagramanov, N. D., Mal’tsev, A. K. and Nefedov, 0. M. (1986). J . A m . Chem. Soc. 108, 4352 Vancik, H., Raabe, G., Michalczyk, M. J., West, R. and Michl, J. (1985). J . A m . Chem. SOC.107, 4097 West, R. P., Chapman, 0. L. and Le Roux, J. P. (1982). J. Am. Chem. SOC. 104, 1779 Withnall, R. and Andrews, L. (1986). J . A m . Chem. SOC.108, 8118 Withnall, R. and Andrews, L. (1990). J . Phys. Chem. 94, 2351

This Page Intentionally Left Blank

Acid-Base Behaviour in Macrocycles and Other Concave Structures ULRICHLUNING Institut fur Organische Chemie, Christian-Albrechts-Universitatzu Kiel, Federal Republic of Germany

1 Introduction 2 Bases on the inside Polyazamacrocycles and cryptands Pyridines 1,lO-Phenanthrolines 3 Acidic centres on the inside Macrocycles containing intra-annular carboxylic acids Macrocycles containing intra-annular sulfinic or sulfonic acids Intra-annular phenols 1,3-Diketones 4 Macrocycles with both acidic and basic functionalities General remarks One intra-annular acid-base function Intra-annular acids and bases only 5 Hydrogen bonding 6 Closing remarks References

63 65 65 73 83 86 86 95 97 101 103 103 104 105 107 110 112

1 Introduction

Soon after the first syntheses of macrocyclic ligands (Curtis, 1960; Pedersen, 1967), functional groups were attached to or incorporated into macrocycles. The synthesis of macrocycles boomed in the 1970s (Patai, 1980; Gokel and Korzenowski, 1982) and almost all conceivable functional groups were introduced. In many cases these functions were acids or bases, but usually the connection was not an incorporation but an attachment. In addition, the functional groups were attached in a very flexible way. Therefore the properties of the acid-base groups of these substituted macrocycles hardly differed from those of the same functional group'' in a non-macrocyclic molecule. If the functional group is on the inside, however, the situation is different because the surrounding macrocycle can have an influence on the functional group. In this review, we will discuss the acid-base properties of acids or 63 ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 30 ISBN 0-12-033530-1

Copyright 0I995 Academic Press Limited All rights of reproduction in any form reserved

u. LUNING

64

bases within a macrocycle, i.e. of intra-annular proton donors and acceptors. This review is focused on the word “in”. As we shall see, most molecules with in-functional groups are macrocyclic but they need not be so. Other geometries like clefts may also be used. However, as will become obvious in this review, concave geometries have been studied with macrocycles almost exclusively. Because acid-base properties on the inside of molecules will be discussed, those macrocycles (and related compounds) which possess functional groups hanging on the outside of the ring, e.g. lariat ethers or substituted polyazamacrocyles (Kaden, 1984; Gokel, 1991, 1992), will be mentioned only briefly. From the large group of azamacrocycles only a small fraction will be discussed in detail because the basic functionality rarely is located in a defined in-position. An in-geometry can be ensured by appropriate substitution of the building block which carries the acid-base functionality, for instance by using 2,6-disubstituted aromatic compounds like pyridines, 2,6-disubstituted benzoic acids or other 2,6-disubstituted phenyl derivatives (see Scheme 1). The use of 2,6-disubstituted arenes is sometimes called the “1,3-xylyl trick” and assures an intra-annular orientation.

COOH

x

Scheme 1

The interaction of macrocycles and other hollow molecules with guests has been reviewed very often (Lindoy, 1989; Vogtle, 1989, 1991; Izatt et al., 1991, 1992; Dietrich et al., 1993; an excellent list of other relevant literature can be found in Schneider, 1991). In this review, the focus will be on acid-base reactions (Sections 2 to 4) and on hydrogen bonding (Section 5). In water, the acidity of an acid usually is given as its pKa. The basicity of a base is defined as the acidity of its conjugate acid. An analogous pK, (solvent) can be defined when the measurements are undertaken in other solvents, but often the acid-base properties are given as log K values. While pKa implies the dissociation of the acid and is defined as the negative logarithm of the dissociation constant, in many cases log K values are given which correspond to an inverse pK,,. These log K values are the logarithms of the equilibrium constants for the protonation reaction rather than the dissociation. Therefore the numbers are the same but negative. In this review, the notation given in the original literature (log K, pKa, etc.) will be used.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

65

2 Bases on the inside POLYAZAMACROCYCLESAND CRYPTANDS

Countless macrocycles containing basic functionalities have been synthesized and, for a large number of compounds, pK, values have been determined (Patai, 1980; Gokel and Korzenowski, 1982; Kaden, 1984; Kimura, 1985; Lindoy, 1989; Vogtle, 1989, 1991; Gokel, 1991, 1992; Izatt et al., 1991, 1992; Dietrich et al., 1993). Usually, the basic function in these macrocycles is a nitrogen atom; the compounds are (po1y)azamacrocycles. But in azacrowns like [l] or cyclams [2] the in-orientation of the lone pair at the nitrogen atom is not assured. Murakami et al. (1991) synthesized the

R

PI polyazapolymacrocycle [3], but, even when protonated, it is capable of complexing hydrophobic guests. This indicates that, even in this polymacrocyclic molecule, the protonation and the solvation of the positive charges take place from the outside. Therefore most polyazamacrocycles behave as normal amines, a situation which has for instance been recognized by Lindoy (Anderegg el al., 1980), who has investigated closely related diazadioxamacrocycles [4] at 25°C in water and in 95% methanol. The ionization constants for the mono- and di-protonated macrocycles [4] were comparable to non-macrocyclic secondary amines (see Table 1). Many azamacrocycles, like cyclam [2] for instance, contain more than one nitrogen atom and therefore possess more than one pK,. Therefore, in the discussion of the basicity of these compounds, it is not only the in- and out-orientation of the lone pair or the proton, respectively, that has to be considered. There is also an influence of the first on the second, third, etc. protonation. The developing Coulomb repulsions are very sensitive to the charge separation. In 15-membered rings this phenomenon has been investigated in detail (Arnaud-Neu et al. , 1979).

66

u. LUNING

Compound [5a] is the open-chain analogue of [5b], but, if [5a] is diprotonated, the two positive charges can avoid each other. Therefore log K1 and log K2 values only differ by 0.8 units (see Table 2). For statistical reasons, the pKal and pKaZof a diacid with identical acidic functions differ by 0.6 p K , units. Thus in [5a] there is hardly any influence of the first on the second protonation.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

67

Table 1 Equilibrium constants log K for the protonation of macrocyclic diamines [4] in water and methanol at 25°C.

Solvent 9.4-10.1 9.2-10.1

Water 95% methanol

6.0-8. la 5.0-8.1"

"log K2 depends heavily on the ring size (see below, discussion of Coulomb forces).

Table 2 Mono- and diprotonation of macrocyclic and non-macrocyclic diamines [5].

9.84 9.06

8.86 5.21

8.60 7.55

8.76 8.04

In macrocycle [5b] the difference between log K1 and log K2 is 3.6, reflecting the strong Coulomb repulsion between two neighbouring positive charges. When the nitrogen atoms have a larger separation as in the isomer [5c] the difference is smaller (l.l), and if the sulfur atoms are replaced by oxygen atoms it is even lower (0.7 in [5d]). The conformational problems (in versus out) in relation to ring size

I

Me

I

Me [5a1

u. LUNING

68

Table 3 pK, values in cryptands [6] in different solvents.‘

In water

In H20/MeOH

Cryptand[6] [X.Y.ZI

PKal

[1.1.1 [2.1.1 [2.2.1 [2.2.2 [3.2.2 [3.3.3 [2.2.C 3IC

-b

-

-

-

7.85 7.5 7.3 7.3 7.0

10.6 10.5 9.6 8.5 7.7

6.6 6.6 6.6 6.55

11.0 10.4 9.85 9.1

-

6.6

9.3

-

6.7

9.3

[l](R

=

-

Me)

-

~Ka2

PKd

-

~Ka2

-

“Lehn and Sauvage (1975); Lehn and Montavon (1978). bNo pK, values could be measured under comparable conditions (see discussion below). “One chain of the bimacrocycle is not a polyether chain but a (CH,), chain.

variation and the differing Coulomb interactions can be observed in most polyazamacrocycles (Fabbrizzi et al., 1986; Fenton et al., 1987). Cryptands (Lehn and Sauvage, 1975; Lehn and Montavon, 1978; Dietrich et al., 1993) and similar bimacrocycles’ show the same behaviour (Bencini et al., 1993). Table 3 lists pK, values of some cryptands [6]. When the cryptands [6] of Table 3 are compared with one another and with analogues it becomes obvious that the pK,, values depend very little on the structure but the pKa2 values do. Here again Coulomb repulsions lead to higher pKa2 values for smaller cryptands [6].

Two cryptands [6] show special behaviour: [2. . l l an1 [ 1.1.11. In contrast to the larger cryptands where a fast proton exchaige takes place between

’

The expression bimacrocycle (rrirnacrocycle or polymacrocycle) is used instead of macrobicycle (macrotricycle, etc.) because polymacrocycles can contain further, non-macrocyclic ring systems (e.g. aromatic rings).

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

69

the protonated and unprotonated species, the protonation and deprotonation kinetics of [2.1.1] and [1.1.1] are slow. With cryptand [2.1.1], the first protonation step is fast, while the second protonation is slow (Cox et al., 1982). In the crystalline state, the dication has an in&-conformation with hydrogen bonds between the oxygen atoms of the chains and the in-oriented protons on each bridgehead. For cryptand [1.1.1], no pKa values are listed in Table 3 because the diprotonated cryptand can only be deprotonated under extreme reaction conditions. Treatment for 18 days in 5N-KOH yielded less than 5% of monoprotonated [l.l.l].HCl (Cheney and Lehn, 1972). Only a long contact time with a hydroxide-loaded tetraalkylammonium resin could generate the monoprotonated cryptand [l.l.l].H+ from the diprotonated cryptand [1.1.1].2H+. The proton of this monocation exchanges quickly between the two nitrogen atoms within the cavity, but exchange with external D 2 0 is slow. The overall kinetics of the mono- and double-protonation of cryptand [1.1.1] are even more complicated because the two ions [as for any other bimacrocyclic diamine (Alder, 1990)] may exist as different conformers. Scheme 2 shows the protonated forms in five different conformations:* Monoprotonated: Diprotonated:

(o+i): one out-NH+ and one quickly inverting N (i+i): one in-NH+ and one quickly inverting N (o+o+):two OU~-NH+ (i+o+): one in-NH+ and one out-NH+ (i+i+): two in-NH+

The rates for protonation and deprotonation depend on the conformation and in some cases they are so small that no equilibrium constants could be measured properly. But in those cases where rate constants have been determined, some equilibrium constants have been calculated (Smith et al., 1981). While the protonation of an out-nitrogen atom occurs at a rate comparable to the protonation of other amine nitrogen atoms, the protonation of the in-nitrogen atoms is extremely retarded. The activation barriers for the [from (ii)] and (i+i+) [from (i+i)] conformers are formation of the )i'( c.110 kJ mol-l. Almost the same value was determined for the deprotonation of (i+i+) to give (i+i). From these rates, the pKa values for the protonation of the in-mono- and in-diprotonated forms were calculated to

* The kinetic isotope effect of the protonation kHlkD= 3.9 suggests that an in-nitrogen atom is protonated directly rather than conformational changes exposing the lone pair of a nitrogen atom to the outside prior to protonation. It is assumed that a protonated nitrogen does not invert. Inversion is only possible by a deprotonation-inversion-reprotonation sequence (Kjaer ef al., 1979).

u. LUNING

70

ii

\

+OH-

p'q

+N-H

I

H-N

\+

o+o+ (pK, = c.1)

o+i (pK, = 7.1) 1.2

+ /

+H :-

+H+

i+i (pK, 2 18) kl = < 7 X lo-''

M-'S-l

k2 = 2.3 X 10-4s-'

k4 = 1.4 X lo-* M-'S-l

ks = 3.1 X

i+o+ (pK, = c.0) k3 = 3.8 X 10-3s-' M-~S-~

Scheme 2

be: pKa23 18 )i'( and pKa12 8 (i+i+). Thus cryptand [1.1.1] is thermodynamically a strong base but it is kinetically extremely slow. This means that, in cryptand [1.1.1], a strong macrocyclic effect exists which is expressed in slow kinetics if the protonation occurs in. The out-protonation, however, shows a normal behaviour. For larger cryptands [6] (Cox et al., 1978), the protonation/deprotonation kinetics have also been measured. Table 4 lists the kinetic and the equilibrium data for such cryptands. When compared to the neutralization of protonated tertiary amines by OH-, the reaction of the second smallest protonated cryptand [2.1.1].H+ is lo5 to lo7 times slower (Cox et al., 1978), indicating a strong shielding and possibly an in-orientation of the proton. For the [2.2.1] cryptand, no kl and k P l values could be calculated, probably due to a fast pre-equilibrium between in&- and in,out-conformations. An extremely low rate of neutralization was found for the hexapyridinotetraazatrimacrocycle [7] (Takemura et al., 1991). The monoprotonated form [7].H+ could be isolated as its hydroxide! The neutralization of the in-located proton took half a year, forming a complex in which a water molecule is bound on the inside.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

71

Table4 Kinetic and equilibrium data for the protonation of some cryptands [6]. ki

+

[6] H20

[6]*H+OHk-i

pK, in water Cryptand

1.2"C

25°C

-2.1.1 '2.2.1 '2.2.2

11.84 11.78 10.66 10.49

11.17 10.91 9.86 9.69

kl" 1.59

-

731 329

k-lb 1.1 x 103 -

1 x 107 6.7 x lo6

"Rate constant for protonation/s-I. 'Rate constant for deprotonatiodl mol-' s-' 'One central ethylene group is substituted by a 1,2-phenylene group.

In other bimacrocyclic bases, intramolecular hydrogen bonds can also lead to increased basicities (Bencini et al., 1993). Scheme 3 shows bimacrocycles [8]. While for most chains X, a value of logK1 of approximately 12 was found, for three bimacrocycles (X = CH2CH2, CH2CH2NHCH2CH2 and CH2CH20CH2CH2),log K1 values in water were larger than 14 (Table 5). For the bimacrocycle [8b] shown in Scheme 4, an X-ray analysis showed intramolecular stabilization of the monoprotonated compound by hydrogen bonds; this could explain the higher basicity (Ciampolini et al., 1986). Two classes of stiff polyazamacrocycles with intra-annular basic functions should not be forgotten: porphyrins and phthalocyanines. In these molecules, the perimeter is fixed. The pK, value may be influenced only by substituents but not by variation of the ring size of the cavity. In this context we will therefore not discuss porphyrins and phthalocyanines.

u. LUNING

72

Me

I

PI Scheme 3 Table 5 Logarithms of the basicity constants log K 1 and log K2 for the bimacrocycles [8] in water at 25°C.

[Sa] [Sb] [SC] [Sd1 [Se] [Sf] [8g] [Sh] [8i]

CH2CH2 CH2CH2NHCH2CH2 CH2CH20CH:CH2 (CH2)5 CH2CH2SCH2CH2 CH2CH2NMeCH2CH2 CH2NHCH2 CH2NMeCH2 CH2NBzCH2

>14 >14 >14 12.0 11.9 11.8 12.5 11.8 11.8

7.8 8.4 11.2 7.9 8.8 9.5 9.1 10.0 8.3

This quick glance at the acid-base properties of some (po1y)azamacrocycles already suggests which parameters will determine the pK, of macrocyclic and related acids and bases. Hydrogen bonds will probably be very important and in polyions Coulomb interactions have to be taken into consideration. But the geometry of the acid-base function has to be defined. In Sections 2, 3 and 4 we shall therefore focus on compounds with intra-annular acid-base functionalities (the 1,3-xylyl trick),

Scheme 4

ACID-BASE BEHAVlOlJR IN CONCAVE STRUCTURES

73

PYRIDINES

An average azamacrocycle like [ 11 or [2] contains tetrahedral sp3-hybridized secondary or tertiary nitrogen atoms. Due to the easy inversion at the nitrogen atom, the geometry of these bases is not fixed (see discussion on cryptands, pp. 68-70). By contrast, in pyridine the nitrogen atom is sp2 hybridized and therefore planar; no inversion may occur. If in addition the pyridine is 2,6-disubstituted7 an intra-annular orientation of the lone pair is guaranteed (for references, see Bell and Sahni, 1991). Many macrocycles containing one or more such pyridine units have been investigated in acid-base reactions.

Macrocycles containing one pyridine unit The pKa values of two classes of pyridinocrown ethers [9a] and [9b] with ring sizes varying from 15 to 33 ( n = 2-8) have been measured in different solvents by Reinhoudt and his co-workers (Grootenhuis et a l . , 1984, 1986; van Staveren et al., 1988). The data are summarized in Table 6. Table 6 shows that the pKa values of pyridinocrowns [9] vary with the ring size and with the solvent. In every solvent, the pKa values of the larger rings were more similar to those of open-chain analogues [lo]-[12] than the pKa values of the smaller rings. And in all systems, small rings were much more basic than the larger ones. When water was present, the largest basicity was always found for the 18-membered system. In pure methanol, however, a gradual drift from a very high basicity in the case of a 15-membered pyridinocrown ( n = 2) to smaller basicities for larger rings or open-chain molecules was measured. The high basicities of the smaller rings may be explained by intramolecular hydrogen bonds which stabilize the protonation pyridinium forms (see Scheme 5 ) . In X-ray analyses obtained for crystals grown from water, hydrogen bonds between the pyridinium ion, water molecule(s) and ether

R

[9a] R = Ph [9b] R = H

74

(I. LUNING

R

R

Scheme 5

oxygen atoms of the rings have been found (van Staveren et al., 1984; Grootenhuis et al., 1987a). The best fit exists in the 18-membered ring, explaining the highest basicities for these compounds (Grootenhuis et al., 1986). When the ring is enlarged, either the fit is less good (21-membered) or two molecules of water may be included (24-membered). Table6 pKa values of pyridinomacrocycles [9] of varying ring size and analogues [lo14121 in different solvents at 25°C.

[9a] in H20“ n Ring size PK,

3 18 5.12

4 21 4.25

5 24 3.97

6 27 3.74

7 30 3.70

[9b] in H20”96*C n 2 3 4 5 6 7 8 Ring size 15 18 21 24 27 30 33 4.88 4.95 4.16 3.95 3.70 3.53 3.36 6f;comparison: pK, (pyridine): 5.23, pK, [lo]: 3.36, pKa [Ill: 3.53, pKa [12]: 4.40 [9b] in MeOH‘ n 2 3 4 5 Ring size 15 18 21 24 PKa 6.33 5.98 5.51 5.30 For comparison: pK, [lo]: 4.08; pK, [ll]: 4.38

6 27 5.07

7 30 4.80

8 33 5.18

[9b] in 52.1 wt% MeOH/H20‘ n 2 3 4 5 Ring size 15 18 21 24 PK, 4.46 4.94 3.92 3.72 For comparison: pK, [lo]: 2.40; pKa [Ill: 2.55

6 27 3.38

7 30 2.98

8 33 2.65

[9b] in 85.4 wt% EtOH/H20‘ n 2 3 4 5 Ring size 15 18 21 24 PK, 4.08 4.57 3.62 3.39 For comparison: pK, [lo]: 2.15; pK, [ll]: 2.52

6 27 3.09

7 30 2.95

8 33 3.17

%an Staveren et al. (1988). bGrootenhuis et al. (1984). ‘Grootenhuis et a/. (1986).

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

[9b], m

75

k1

Scheme 6

If the incorporation of water molecules explains the high basicities, this also rationalizes the non-existence of a peak-basicity for an 18-membered ring in methanol. No water molecules may “relay” the hydrogen bond between the pyridinium ion and the ether oxygen atoms here. For a direct hydrogen bond between the pyridinium proton and ether oxygen atoms the best geometry exists in the 15-membered ring systems. When the pyridine ring bears no substituent in the 4-position [9b], the cavity of the larger rings is big enough for a flip of the pyridine nitrogen atom to the outside. The cavity is then filled by the pyridine unit itself (self-complexation, see Scheme 6). Protonation, and solvation of the resulting pyridinium ion, can take place from the outside. The minimum ring size for this process is 24 (Uiterwijk et al., 1986; Grootenhuis et al., 1987b; van Staveren et al., 1988) or 27 (Grootenhuis et al., 1984). In the case of the picrate of pyridino-27-crown-9 [9b, rn = 21, a hydrogen bond between the pyridine nitrogen atom on the outside and a picrate anion was found by X-ray analysis. In this case, the orientation of the pyridine unit becomes comparable to that in non-macrocyclic analogues which explains the similar pK, values between larger pyridinocrowns [9] and the analogues [lo]-[12] (Grootenhuis et al., 1984). But the difference in the orientation of the pyridine units in small and large rings is not the only reason for the difference in basicity because the ring-size dependence of the pK, values for the 4-phenyl substituted pyridinocrowns [9a] and the unsubstituted ones [9b] is very similar although the phenyl-substituted compounds [9a] cannot self-complex analogous to Scheme 6. In concave pyridines [13], the position of the pyridine nitrogen atom is even more defined. A large number of bimacrocyclic concave pyridines [13] were synthesized, and their relative basicities were determined in ethanol; these are given in Table 7 (Luning, 1987; Luning et al., 1990, 1991a, 1993). As expected, Table 7 shows that the basicities of concave pyridines [13], can be influenced by substitution in the 4-position (R = H, OMe, NEt,).

u. LUNING

76

Table 7 Relative basicities (log K)re, of concave pyridines [13] and related compounds [14]-[16] in ethanol.

Compound

R

x (X')

~3a1 p3b1 13c] :13d] .13e] 13f]

:

:w .13h] -13i]

IWl

:13k]

Y

(log a r e l a

(CH2110 +0.6 (CH2)10 -1.6 (CH2)10 -0.3 (CH2)10 -1.4 (CH2)7 -0.3 (CH20CHz)z -0.8 (CH20CH2)3 +0.4 (CH20CH2)3 +0.2 (CH2)10 > +1.4 (CH2)10 +1.0 (CH2110 c.4

For comparison: H CH2(CH20CH2)2CH2 ~ 4 4 H CH2(CH20CH2)3CH2 ~4b1 H n-Bu,n-Bu [ w H n-Bu,n-Bu ~5b1 H n-Bu,n-Bu [ W H (CH2)20Me,MeO(CH2)2 [16bI Pyridine 2,6-Dimethylpyridine 2,6-Di-t-butylpyridine 2,6-Diphenylpyridine

Me Me Me Me (CH2)lO (CH20CH2)3 Me Me Me Me

-0.8 -0.3 -1.8 -1.0 -1.3 to -1.6 -1.3 C.0 +1.3 -1.2 0.2

"The basicities were determined by photometric titrations in ethanol relative to thymol blue [(log K),, = 01 (Liining and Muller, 1989).

The more electron donating the substituent in the 4-position, the more basic is the concave pyridine [13], but the basicities also vary with the ring size. A comparison between concave pyridines [ 131 monomacrocyclic compounds [14] and [15] and open-chain analogues [16] shows a macrocyclic effect. The open-chain analogues [ 161 possess very small basicities [(log K),I = -1.3 to -1.61. In the monomacrocyclic systems, those which contain ether oxygens, [14] and [15b], had larger (log K),e, values. The largest (log K)rel values, however, were found for bimacrocyclic compounds [13], but in this class, a large dependence of the basicity on the ring size was found. Large basicities were found when a polyether chain was short and the other chain was long, specifically when a resulting pyridinocrown part of the concave pyridine became 15-membered as in [13a] and [13g]. Thus the highest basicity of a 4-unsubstituted concave pyridine was found for [13a] with (logK),eI = 0.6. This may again be explained by an intramolecular stabilization of the positive charge of the pyridinium ion by hydrogen bonds (see Scheme 5).

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

77

X' = Bu, (CH&OMe r151

[I61

As observed for the pKa values of pyridinocrowns [9] in methanol (see Table 6), no peak basicity of the (logK),,, values was found for 18membered concave pyridines [ 131 in ethanol. Whether a peak basicity would exist in water cannot be ascertained due to their low solubility. Concave pyridines [131 are comparable to pyridine-containing cryptands (Wehner and Vogtle, 1976). Table 8 compares the pKa of a pyridinecontaining cryptand [17] with an analogous cryptand [6]. In contrast to the concave pyridines [13], the nitrogen bridgeheads of [17] are the most basic sites in the bimacrocycle [17]. Therefore it is not surprising that the pK1 values of [17] and [6] differ by only 0.6 units. But the gap between the first and the second protonation is larger for the cryptand [6]. It may be that the

u. LUNING Table 8 pK, for a pyridine-containing cryptand [17] and an analogous cryptand [6] ([2.2.1]).

diprotonated form of [ 171 receives extra stabilization by additional hydrogen bonds to the pyridine nitrogen atom. Another class of bimacrocyclic 2,6-disubstituted pyridines is the bissulfonamides [MI. They were synthesized in an analogous manner to the concave pyridines [13] and their basicities were also measured relative to thymol blue in ethanol (Luning et al., 1991a). The data are listed in Table 9. When the concave pyridinebissulfonamides [181 are compared to comparable concave pyridinebislactams [13], it is obvious that the sulfonamides are less basic [A(logK),,, = 0.8-2.41 (Luning et al., 1991a). Two possible reasons may be responsible for this; firstly a sulfonamide group possesses a larger electron-withdrawing effect than a carbonamide group, and secondly the geometry of the sulfonamide group differs from the geometry of a carbonamide (tetrahedral instead of planar).

When sulfonamides [18] with the same substituents in the 4-position (H, OMe or NEt,) are compared with one another, those bimacrocycles containing two polyether chains (X and Y) are the more basic compounds. Presumably, more oxygen atoms may stabilize the positive charge by more hydrogen bonds. Again, as in the concave pyridines [13] and the pyridinocrown ethers [9], very high basicities are found when 15-membered ring systems are present. The 2,6-disubstitution pattern of a pyridine can also be found in (formally) 4-hydroxy-substituted pyridinocrowns [191 or [20] (Nakatsuji et al., 1985; Vogtle et al., 1991). Basicity measurements have also been carried

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

79

Table 9 Relative basicities (log K)re, of concave bissulfonamides [18] in ethanol.

Compound

R

Y

X

For comparison: [18m1 H

@rela

-1.8 -0.4 -1.8 -1.6 0.2 0.5 -1.0 0 -0.4 2.0 2.5 2.0

H H H H OMe OMe OMe OMe OMe NEt, NEt2 NEt2

'18a'18b' '18~: '18d :18e' 18f.

(1%

CH2(CH2OCH&CH2

Me Me

-1.1

"The basicities of [18] were determined by photometric titrations in ethanol relative to thymol blue [(log K),, = 01 (Liining and Miiller, 1989).

Table 10 Comparison of the pK values of (formally) 4hydroxypyridinocrowns [19] and [20] which may exist as 4-hydroxypyridines or 4-pyridones.

PK1 PK2

~ 9 1

1201

[211

1.7 8.5

3.1 11.0

3.3 11.1

out for these compounds, but the possible tautomerization of the 4hydroxypyridines to form the corresponding 4-pyridones (shown in [20]) complicates the situation. In a 4-pyridone, protonation will take place on the carbonyl oxygen which lies outside the cavity. The most basic lone pair no longer has an intra-annular location. Table 10 compares two pyridinocrowns [ 191 and [20] both of which may in principle exist as 4-hydroxypyridine or as 4-pyridone tautomers. Which form is prevalent depends on the nature of the substituents in the 2,6-positions. Electron-withdrawing groups like the ester groups in [19] lead to a larger population of the 4-hydroxypyridine form, whereas the oxymethylene groups in [20] stabilize the 4-pyridone tautomer. The comparison of the pK values of [19] and [20] with 4-hydroxypyridine which exists as its 4-pyridone tautomer [21] is in agreement with structural assignments based on X-ray data (Nakatsuji et al., 1985). Owing to the existence of tautomers, the intra-annular orientation of the proton is not guaranteed. In addition, the protonated form is an ambident acid. Either the N-H or the 0-H proton may dissociate.

u. LUNING

80

OH

(9 I

H

The acidity of a smaller 4-pyridone-containing macrocycle [22] has also been measured (Shukla et al., 1988). Its pK, value at 25°C in 70% dioxane/water was 13.7, but, due to the use of a different solvent, no comparison with other 4-pyridones is possible. Also comparable to the pyridine [20] is the macrocyclic 4-pyridone [23] in which one ring oxygen atom is replaced by a sulfur atom. The resulting pK, values (Wu et al., 1991) are similar to those of the corresponding all-oxygen macrocycle [20]. In Scheme 7 the 4-pyridone structure of [23] and its deprotonated form [24] are shown. The protonation of the anion can take place at the nitrogen or at the oxygen atom. Macrocycles containing more than one pyridine

In the preceding sub-section it was shown that one reason for increased pK, values in pyridinocrowns [9] was the possibility of intramolecular hydrogen bond formation. In macrocycles containing more than one pyridine unit, such stabilization is even more likely. Table 11 lists pK, values for a number of polypyridinocrowns [25]-[26].

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

81

0-

pKal = 3.03 pKa;?= 10.13 Scheme 7

Table 11 pK, values of polypyridinocrowns [25]-[26] in water at 25°C."

Compound [9b] (n = 3) 1251 [26a] ( n = 2) [26b] ( n = 1) [26c] (n = 3) Pyridine 2,4,6-Trimethylpyridine

pKa (monoprotonated)

pKa (diprotonated)

4.9 5.3 5.3 7.9 4.8 5.1 7.4

3.6 3.7 <3 >3

"Newcomb ef al. (1974).

Cram (Newcomb et af., 1974, 1977a; Bell et af., 1982) has compared the basicities of (po1y)pyridine analogues of 18-crown-6 [25]-[26]. As already listed above, pyridino-18-crown-6 [9b] (n = 3) has a pK, similar to pyridine itself and some open-chain analogues. In Table 11, two systematic variations have been made: the number of pyridine units in 18-crown-6 has been increased from [9b] via [25] to [26a], and the ring size of the compounds [26] has been varied from n = 1 to n = 3. While the pK, values of the 18-crown-6 derivatives did not vary very much with changing numbers of pyridine units, much larger pK, changes were found when the ring size was altered. For the smallest ring [26b] (12-membered) the largest pK,, 7.9, was found. Again intramolecular hydrogen bonds are probably the reason for the stabilization of the protonated form as could be shown in CPK-models

82

u. LUNING

(Newcomb et al., 1977a). For larger rings such hydrogen bonds are very unlikely. The second pKa in these polypyridinocrowns [26] again is distinctly lower due to Coulomb repulsion between the positive charges (see p. 67). The larger the ring, the smaller is this interaction: pKa2 (dipyridinocrown-12 [26b]): <3, pKa2 (tripyridinocrown-18 [26a]): 3.7, pKa2 (tetrapyridinocrown24 [26c]): >3. Related to dipyridinocrown-12 [26b] is the pyridinocyclophane [27]. However, in this macrocycle no cooperativity between the two pyridine functions was found when [27] was protonated (Vogtle et al., 1989). Indeed [27] had the same basicity as 2,6-dimethylpyridine, and the introduction of a methoxy group in the 4-position gave the usual basicity increase of one pKa unit (see Table 7). The reason for the different behaviour of [27] in comparison to the pyridinocrowns [26] may be its geometry. The two pyridine nitrogen atoms are too close to each other to form a hydrogen bond. They are twisted (Vogtle et al., 1989) and each pyridine acts separately when protonated. The small value of pKal is again caused by Coulomb repulsion forces.

83

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

Table 12 pKa values of pyridinocyclophanes [27] in comparison with other pyridines.

Compound [27] (R = H) [27] (R = OMe) Pyridine 4-Methoxypyridine 2-Ethylpyridine 2,6-Dimethylpyridine

PKal

pKa2

1.2 3.0

6.7 7.7 5.2 6.6 5.9 6.7

Another macrocyclic compound containing more than one pyridine is the trispyridino-cryptand [28] (Newkome et al., 1981). The ring size in this [3.3.3]-analogous cryptand is 24. Therefore it is not surprising that an X-ray structure shows one pyridine which has turned around to fill the cavity in a self-complexing way (see Scheme 6, p. 75). Thus one free electron pair is located on the outside of the cryptand. Furthermore, the bridgeheads (in-oriented in the X-ray structure) of this bimacrocyclic compound are nitrogen atoms which should be more basic than the pyridine itself anyway. No pK, values have been measured. 1,lO-PHENANTHROLINES

Although 1,lO-phenanthrolines are closely related to pyridines, the number of macrocycles containing 1,lo-phenanthroline units is much smaller and even fewer compounds of this class have been investigated in acid-base

84

u. LUNING

reactions. Analogous to the concave pyridines [13] (see pp. 78-83), concave 1,lO-phenanthrolines [29] have been synthesized and their relative basicities determined in ethanol (Liining and Miiller, 1989). For the 1,lOphenanthrolines [29] also, the basicities were found to depend on the number and nature of the chains X and Y (see Table 13). Again, the more oxygen atoms present in the molecule and the shorter one polyether chain was, the higher was the basicity. But the variation of the chains X and Y in the concave 1,lO-phenanthrolines [29] had a smaller influence on their basicity than changes in the chains X and Y of the concave pyridines [13] (Liining and Miiller, 1989). Probably the concave 1,lO-phenanthrolines [29] are geometrically less flexible due to the extended aromatic system. In contrast the concave pyridines [ 131 are more flexible and probably can adopt other geometries leading to a larger variation in basicity. Basicities have also been estimated for a second class of concave 1,lO-phenanthrolines [30] (Liining and Miiller, 1990; Miiller, 1991; Liining et al. , 1994). These concave 1,lo-phenanthrolines [30] are approximately two orders of magnitude more basic than the bislactams [29]. Their basicities also depend on the nature of the chains X (polyether or polymethylene) [estimated (log K)relvalues between 1.5 and 2.51, but due to uncertainties in the measurements the difference cannot be discussed in detail. On top of the macrocyclic effect, Sauvage (Cesario et al., 1986) has recognized a change of basicity by topological factors in the [2]-catenane [31] which contains two 1,lO-phenanthroline units, one in each ring. When compared to the non-macrocyclic analogue [32], the monoprotonated form of the catenane [31] is extremely stable (pK, = 8.5 instead of 5.1, see Table 14). On the other hand, a second protonation of the catenane [31] is very difficult; the basicity of the second 1,lo-phenanthroline is lo7 times smaller. The reason for the large pKaZ and the small pKal was found in an X-ray analysis of the protonated catenane [31].H+. The proton is hydrogen bonded to the nitrogen atoms of both 1,lo-phenanthroline units, the structure of the H + complex being similar to that of a Cu+ complex. In

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

85

[301 (X = polyether, polymethylene)

X = CH2(CH20CH2)4CH*

addition to the hydrogen bonds, .rr-stacking between the oxysubstituted aryl rings of one macrocycle and the 1,lO-phenanthroline system of the other ring stabilizes the complex. These forces stabilize the monocation and have to be overcome if a second protonation is to take place.

u. LUNING

86

Table 13 Relative basicities (log K)rel of bimacrocyclic (concave) and monomacrocyclic 1,lO-phenanthrolines [29].

"The basicities were determined by photometric titrations in ethanol relative to thymol blue [(log K),,, = 01 (Liining and Miiller, 1989).

Table 14 pKa values of catenane [31] in comparison to the 1,lo-phenanthroline [32] in CD2C12/CD3CN (70 :30)."

Compound

PKal

pKa2

[311 [321

c.1.5

8.5 5.1

"Cesario et al. (1986).

3 Acidic centres on the inside

In general, two classes of acids have to be discussed: Lewis and Bransted acids. In macrocycles, only a few Lewis acid centres have been incorporated, e.g. tin (Newcomb et al., 1987; Newcomb and Blanda, 1988; Blanda and Newcomb, 1989; Blanda et al., 1989) and boron (Reetz et al., 1991). Here we will discuss Brmsted acids, starting with carboxylic acids.

MACROCYCLES CONTAINING INTRA-ANNULAR CARBOXYLIC ACIDS

Macrocycles containing one COOH group Macrocycles with intra-annular carboxylic acid groups [33] have been synthesized by Cram and Reinhoudt. As for the analogous pyridinocrowns [9], the pK, values of the acids were measured in water at 22°C and at 25°C (Newcomb et al., 1977b; Aarts et al., 1986). As for the pyridines [9] (see p. 73), the acidity of the acids [33] also depended on the ring size. The 18-membered ring [33b] had the largest pK,, but the pK, values of the 1 5 , 21- and 24-membered rings were also quite

87

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

Me

WI

Me

[35a], n = 1; [35b], n = 2

large. The large values for [33a] and [33b] can again be understood in terms of intramolecular hydrogen bond stabilization (Newcomb et al., 1977b). For [33c] and [33d], the values were lower but still larger than for the larger analogues [33e]-[33g]. An X-ray analysis (Aarts et al., 1986) of the 24-membered compound [33d] suggests that the intermediate pK, value of [33c] and [33d] is caused by a water molecule which acts as a relay between the COOH function and the ring oxygen atoms. This is again comparable to the pyridine series [9]. In an X-ray analysis of [33b] (Goldberg, 1976), intramolecular hydrogen bonds between the COOH group and ether oxygen atoms were found. However, in addition to the formation of hydrogen bonds, steric factors can also contribute to the higher pK, values of [33a] and [33b]. In the smaller rings the intra-annular orientation of the COOH group is more pronounced. Therefore the solvation of the anionic carboxylate "group should be less favourable and this should lead to a larger pK, (Newcomb and Cram, 1975). In another study (Adamic et al., 1986a,b), the influence of a solvent change on the pK, values has been investigated for macrocyclic and non-macrocyclic acids [33] and [35] (see Table 16). Again the largest

u. LUNING

88

Table 15. pK, values of crown ethers with intra-annular carboxyl groups [33] and of their analogue [34].

In H 2 0 at 25°C" n Ring size PK,-

2 [33a] 3 [33b] 4 [33c] 5 [33d] 6 [33e] 7 [33f] 8 [33g] 15 18 21 24 27 30 33 5.31 5.71 4.38 4.06 3.80 3.94 3.93

In H 2 0 at 22"Cb n 2 [33a] 3 [33b] 4 [33c] Ring size 15 18 21 PKa 4.8 4.8 3.8 For comparison: pK, [34]: 3.3

3.4

aAarts et al. (1986). *Bell er al. (1982); Newcomb er al. (1977b); Newcomb and Cram (1975).

Table 16 Comparison between the pK, values of macrocyclic acids with [33b] or without [35] intra-annular carboxylic groups and other carboxylic acids.

pK, in:

H20

MeOH/H20 (80120 w/w)

WI WbI Wl

4.59 4.89 4.80 4.76 4.20 3.87

5.90 6.72 7.76 6.71 6.46 4.92

MeCOOH PhCOOH EtOCH2COOH

MeOH/H20 K,(H20) (99 : w/w) K,(MeOH/H20 80 :20) 8.27 -

10.32 -

20 68 912 89 182 11

influence was found when the acid had the possibility to form intramolecular hydrogen bonds. In water hardly any differences between the acids could be found due to the strong hydrogen bond capacity of water. But in a solvent mixture with a lower hydrogen bond capacity, the intra-annular acid [33b] became to a larger degree less acidic than the other acids [35]. The protonated (neutral) form of [33b] is stabilized by intramolecular hydrogen bonds (Goldberg, 1976) and the acidity drops three orders of magnitude. The different behaviour upon solvent change can best be seen by looking at the quotient K,(H20)IKa(MeOHIH20 80 : 20). This ratio is much larger for the intra-annular acid [33b] than for the acids [35] or the analogues. It seems clear that the acidity of an intra-annular acid depends on its ability to form intramolecular hydrogen bonds. By how much the pK, values are changed by hydrogen bonds depends strongly on the medium and whether it can offer alternative hydrogen bonds or not. Therefore in even less polar media, an intra-annular acid should be even less acidic than the analogues.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

89

Table17 pK, (ethanol) values for concave acids [74a]-[74c] and [75] and for analogues.

Pal PbI

WI P I

Benzoic acid Acetic acid

11.2 11.8 >11.3 9.95

10.25 10.4

So far, we have discussed monomacrocyclic intra-annular acids, but bimacrocyclic concave benzoic acids [74a]-[74c] and [75] have also been synthesized and their acidities determined at 25°C by photometric measurements in ethanol (Wangnick, 1991). Table 17 compares the pK, (ethanol) values. No general trend could be found for the acidity of concave benzoic acids [74] and [75]. While [75] with 3,5-disubstituted outer phenyl rings was slightly more acidic than benzoic or acetic acid, [74] with 2,6-disubstituted outer phenyl rings was much less acidic than the non-macrocyclic analogues. In most other acids (see above), a low acidity is usually caused by the stabilization of the acidic form by intramolecular hydrogen bonds. However in the case of the acids [74] it is much less likely that the low acidity is caused by the stabilization of the acidic form because no ether oxygen atoms

90

u. LUNING

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

91

for hydrogen bonds are present. If the protonated form is not stabilized, the anion must be destabilized to explain the large pK, values. Presumably the solvation of the anion is hindered in the acids [74] due to their concave shielding. In [75] the substitution pattern is different and it is a matter of speculation whether anion solvation is easier for this compound. More than one COOH group The pK, values of two systems with two intra-annular COOH groups have been measured. In Cram's study (Bell et ul., 1982), the macrocyclic diacid [36] and an open-chain monomeric analogue [37] had almost identical pK, values (see Table 18). In contrast, Gennari's compound [38], which contains two intra-annular COOH groups and in addition two"ethy1ester groups, has a different acidity from that of the analogues [39] and [40], as shown in Table 19 (Gennari et al., 1992). But in both systems, the difference ApK, between pKal and pKa2 was comparable (ApK, [36] = 1.7, ApK,[38] = 1.5). The increase from pKal to pKa2 may occur for two

u. LUNING

92

Table 18 pKa values of macrocycle [36]

containing two intra-annular carboxylic groups and an open-chain analogue [37] in DMSO/H20 (2: 1v/v).

PKa 1

6.3

~Ka2

8.0

6.65

Table 19 pKa values for a macrocycle containing two carboxylic acid and two carboxylic ester groups [38] and for analogues [39] and [40] in MeOH/H20 (1:1).

PKa I pKa2

[381

[391

[401

6.5 8.0

4.9

5.1

reasons: (i) as stated above, Coulomb interactions disfavour the deprotonation of the second COOH group; (ii) the mono-anion may be stabilized by an intramolecular hydrogen bond, a stabilization which is not possible for a dianion (see [41]). The acids [36] and [38] themselves also have the possibility to form hydrogen bonds. For [36], X-ray analysis (Goldberg, 1981) shows the formation of a carboxylic acid dimer within the cavity as a result of hydrogen bonding. If the extent of stabilization by intramolecular hydrogen bonding is comparable for the diacid and the mono-anion, it is not surprising that [36] has a pKa value similar to the analogue [37]. The analogue [37] cannot form intramolecular hydrogen bonds as acid or as anion but in the diacid both forms have this possibility. In the diacid [38] however, the ester functions offer additional hydrogen bond acceptor sites. This can be the reason for the higher pKa when compared to the analogues [39] and [40].

93

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

1401

0

0

-6'1 -

\\

0--H-0

7-

1411 Table20 pKa values for convergent diacids [42] with spacers X of different lengths n.*

Compound

n

PK1

PK2

APKa

3 3

4.8 5.1 5.1 5.5

11.1 7.7 7.7 7.5

6.3 2.6 2.6

5 7

2.0

"Rebek et al. (1986).

In the context of the macrocyclic diacids [36] and [38], other diacids [42] which are not macrocyclic but contain convergent carboxylic groups should be discussed (Rebek et al., 1986). The geometry in these acids is brought about by the use of Kemp's triacid; two COOH groups face each other and are forced to form hydrogen bonds. The distance between the carboxylic groups is determined by the number of atoms n in the spacer X. Depending on the distance, drastic differences between pK1 and pK2 were found in ethanol/water, as shown in Table 20. The maximum difference ApK, is 6.3 when the carboxylic acids are very close to each other [42a]. The distance

u. LUNING

94

Scheme 8

between the oxygen atoms of the carboxyl groups is only 3 A and hydrogen bonds can easily be formed. There is, however, a second parameter influencing the acidity: the Coulomb repulsion in the dianion. It is especially large in the anion of [42a] because of the short distance between the carboxylate groups. In [42c] and [42d], the central 1,3-phenylene unit is replaced by longer spacers leading to smaller Coulomb repulsions, whereas in [42b] the missing methyl groups in the 4,6-positions allow the dianion to rotate and to adopt a conformation with remote negative charges (Scheme 8). When the acid-base properties of macrocyclic pyridines and carboxylic acids are compared to one another, it becomes obvious that in both classes the formation of hydrogen bonds is very important for the acidity. In both classes hydrogen bonds stabilize the protonated forms. But when two pyridines or two carboxylic groups are present in one molecule, macrocyclic dipyridines and macrocyclic dicarboxylic acids behave differently. A monoprotonated macrocyclic pyridine can stabilize itself by intramolecular hydrogen bonds in the same way as a macrocyclic mono-deprotonated carboxylic acid. But in the neutral form the two classes of molecules are different. While uncharged dicarboxylic acids have hydrogen atoms and can form stabilizing hydrogen bonds, the dipyridines cannot form such stabilizing hydrogen bonds in the neutral state.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

95

A

0 I

0 I

MACROCYCLES CONTAINING INTRA-ANNULAR SULFINIC OR SULFONIC ACIDS

Besides macrocyclic benzoic acids [33] (see above), investigations have also been carried out on compounds with intra-annular sulfur-based acidic groups, sulfinic acids [43] (Skowronska-Ptasinska et aI., 1987) and sulfonic acids [45](van Eerden et al., 1989). Analogous to the carboxylic acids [33], the sulfinic acids [43] showed an increasing acidity with increasing ring size (see Table 21). Furthermore, the 18-membered system [43b] had the highest pK, value and the pK, values of the larger rings were identical with an open-chain model [44]. Again intramolecular hydrogen bonds can explain the acid-base behaviour of the sulfinic acids. A hydrogen bond is supported by CPK (space-filling) models and an X-ray structure of [43b] (Skowronska-Ptasinska et al., 1988) although the sulfinyl group is disordered.

u. LUNING

96

Table 21 pKa values of crown ethers containing sulfinic acid functions [43] and of an analogue [44].‘

n Ring size PKa

2 [43a] 15 3.0

3 [43b] 18 3.3

4 [43c] 21 2.7

5 [43d] 24 2.7

6 [43e] 27 2.5

7 [43f] 30 2.5

For comparison: [441 2.5 “In water at 25°C; Skowronska-Ptasinska ef al. (1987).

The second class of crown ethers bearing intra-annular sulfur-based acidic groups are the sulfonic acids [45]. The ring size of these macrocycles has been varied from 15 to 33 ( n = 2 to 8). pKa values were determined in water (25°C) and found to be surprisingly low when compared to benzenesulfonic acid. The intra-annular fixation of a sulfonic acid group into a macrocycle [45] leads to a drop in acidity of c.5 orders of magnitude (pK, of benzenesulfonic acid, -2.8; [45], +2.0 to +2.5). Furthermore, in contrast to the pyridinocrowns [9] (see Section 2) and the analogous carboxylic acids [33] (see above), there was almost no influence of the ring size on the pK, values. A possible explanation comes from X-ray analyses of the sulfonic acids [45]. All X-rayed crown ether crystals contained water and the sulfonic acid moiety was dissociated. Therefore in crystals of [45], macrocyclic benzenesulfonate anions and hydronium ions (sometimes hydrated) are present. The ions are bound to each other by hydrogen bonds. The size of the included water-hydronium ion cluster (varying by the number of solvating water molecules) depends on the ring size. In the 15-membered ring, H30+ was found, whereas in a 21-membered ring H502 and in the 27-membered ring H70; were present. This means the sulfonic acid functions in [45]are still strong acids. They dissociate (l), but the resulting ions form a tight ion pair within the cavity and are held together by hydrogen bonds. The microacidity of the sulfonic acid functions in the crown ethers [45] is high,

97

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

but only a small amount of H30+ is released (2); thus the free H 3 0 + concentration in solution is low and the overall pK, values are larger. This difference between the microacidity and the actual observed pK, value may be of importance for the understanding of proteins/enzymes where highly perturbed pKa values have also been found (see Section 6; Kokesh and Westheimer, 1971).

INTRA-ANNULAR PHENOLS

The 1,3-xylyl trick was also used for the incorporation of phenols into crown ethers. Three classes of phenols [46a]-[46c] have been investigated. They differ by their substituents in the 4-position. In Table 22 the pK, values of different macrocyclic phenols [46a]-[46c] are compared. The data obtained for the phenol-containing crowns [46a] and [46b] show very little evidence for a macrocyclic effect. No extra stabilization of the protonated (acidic) form by a macrocycle of appropriate ring size was found. The acidities of the macrocyclic phenols [46a] and [46b] were independent of the ring size and comparable to non-macrocyclic analogues. However, the azo-substituted crowns [46c] showed a difference of 0.8 pK, units which was not expected from the pKa values of [46a] and [46b]. This different behaviour of [46c] is not yet understood.

G:&

M e 4 M e

0

1461

a: X = H

b: X = NOz

OH

A

Me0

WI

OH

[481

c: X = -N=N

NO2

OMe

u. LUNING

98

Table 22 pKa values for macrocyclic phenols [46a]-[46c] substituents in the 4-position and analogues [47] and [48]. [46a] in water, 20°C 2 Ring size 15 PKa 10.8

3 18 10.6

4 21 10.5

[46b] in water, 20°C"3b 2 Ring size 15 PK," 6.8 PK,b 6.9

3 18 6.6 6.8

4 21 6.5

[46c] in water, 25°C' 2 Ring size 15 PKa 7.05

3 18 6.7

n

n

n

Pal 10.7

bearing different

Phenol 10.0

4-Nitrophenol 6.9-7.2 6.9

148~1 6.2

"Browne et al. (1985). bCassol et al. (1990). 'Nakashima ef al. (1983, 1986).

Table23 pKa values of a bimacrocyclic phenol [50] in comparison to a monomacrocyclic phenol [49], to diaza-18crown-6 [l] and to cryptand [2.2.2] [6]. Compound

PKal

pKa2

8.0 7.5 7.6 7.4

9.2 9.0 9.95 9.5

~Ka3 10.2 11.0

The acidity of a bimacrocyclic phenol [50] has also been measured and has been compared with analogues (Czech et af., 1988). Values of pKa are given in Table 23. In the bimacrocyclic phenol, no diversion from the expected pKa values could be found. The observed acidities were comparable to those of the analogues [l], [6] and [49], although a comparison is difficult due to the fact that, again, Coulomb forces probably play an important role. A huge group of macrocycles which contain 2,6-disubstituted phenols are the calixarenes [51]-[53]. Their conformation has been investigated intensively (Gutsche, 1989, 1991). In most conformations, however, the phenolic rings are oriented almost vertically in relation to the plane of the macrocyclic ring. Therefore the OH functions are not oriented in an intra-annular fashion. Nevertheless the pKa values of calixarenes [51] differ from those of other comparable phenols. The reason for this is the

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

99

OH

I

R [Sl], n = 4; [52], n = 6; [53], n = 8 [51a] R = S03Na [51b] R = S02N(CH2CH20H)2 [51c] R = N=N-p-C6H4-NMe; C1[Sld] R = NO2

OH

OH

OH

R

R

S03Na

[54],rn

=

0; [55], rn = 2

[561

interaction between the OH groups of adjacent rings as shown in Scheme 9. If one OH is deprotonated, the resulting anion is stabilized by hydrogen bonds from the neighbouring phenol groups. In further deprotonated calixarene-polyanions, Coulomb repulsions and fewer stabilizing hydrogen bonds lead to higher pK, values. Therefore the first pKa of a calixarene is much lower than the pK, of a related phenol (Shinkai, 1993) (see Table 24). But, due to the limited solubilities of most calixarenes, pK, values have only been determined for a few of them. In Table 24, the pK, values of calix[4]arenes [51] and related molecules are compared.

u. LUNING

100

Table 24 Comparison of pKa values of calixarenes [51] and related phenols [54]-[ 561. ~

R

In H20, 25"Cb [5lala [561a

3.3 8.9

11.8

12.8

c. 14

[51b1 P4bI

1.8 8.3 4.7

9.7

12.5

>14

8.3

11.6

0.5

2.0

10.0

c.13

10.9

12.3

>14

10.6

12.5

WI

In H 2 0 , 20°C' [51~1 [ W

c. 8.0

In 85.4 wt% EtQH/H20b [51di [ W

WI

p-Nitrophenol

2.9 8.7 3.6 8.7

0

"The pK, values are listed for the sodium salt. The pK, of the strong acidic sulfonic acid group is not listed. bShinkai (1993). 'Shinkai et al. (1989).

H'

H.,

-n+

,H"

-H ',

+H+

b Scheme 9

It is not only p-substituted phenols that have been connected by methylene groups; an extended calixarene based on naphthalene units [57] has also been synthesized and its acidity constants have been measured (see Table 25; Poh and Lim, 1989, 1990). In contrast to the calix[4]arenes, the pKa of the expanded calixarene [57] did not differ much from the pKa of the analogue [58]. This may have two explanations: (i) the conformation of the expanded calixarene is different from phenol-based calix[4]arenes and a special stabilization of the anion does not occur, (ii) this stabilization occurs but it is also present in the analogue. A stabilization by the neighbouring

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

101

Table 25 pK, values for an extended, naphthalene-based calixarene P71.

Compound

PK1

PK2

PK3

PK4

5.8 6.5 5.5

8.8

10.5

11.5

“Poh and Lim (1989). bPoh and Lim (1990).

[57], n = 4

1581

peri-OH group is conceivable for the anions of [57] and [58] and this would result in comparable pKa values. In the past years, the bridging of calixarenes has been realized (Alfieri et a f . , 1983; Beer and Keefe, 1987; Reinhoudt et al., 1987; Goldmann et al., 1988; Gutsche et af., 1988; van Loon et al., 1991). If the bridge contains an acid or a base function according to the “1,3-xylyl trick”, a concave orientation of this function should be possible. Such a bridged (or capped) calixarene [59] has already been prepared (Berger et al., 1992), but its geometry was too flexible and the additional phenol group was not located on the inside of [59]; pKa values have not been determined.

1,3-DIKETONES

1,3-Dicarbonyl functions have been built into macrocyclic structures, and pKa values for the resulting macrocycles [60] have been determined (Alberts and Cram, 1979). When the open-chain model [62] is compared with the macrocycles [60], identical first pKa values were found (pK, = 8.6). Thus for the diketones [60], no macrocyclic effect is noticeable. But for the dissociation of a second proton from the mono-anioils of [60] much higher pKa values are found. To a certain extent, Coulomb repulsions (see Section 2) are probably the reason for this behaviour, but the large difference in the pKa values (ApKa = 2.9, see Table 26) argues for a special stabilization of the mono-anion. Again hydrogen bonds are not unreasonable.

HO

10 1

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

103

Table 26 pKa values for macrocycles containing 1,3-diketones [60] and analogues [61] and (621 in waterldioxane (1 :1 v/v) at 24°C.

Compound

PKd

PKa2

APKa

[60aI [60bI

8.6 8.6 9.5 8.6

11.5 11.5

2.9 2.9

WI

[621

4 Macrocycles with both acidic and basic functionalities GENERAL REMARKS

Many polyazamacrocycles have been alkylated with chloroacetic acid leading to compounds which contain both amine and carboxylic acid functions, e.g. [63] (Stetter et al., 1981). These compounds are similar to EDTA (ethylenediamine tetraacetate). Although the ring size of this class of macrocycles and the number of N-CH2-COOH groups has been modified

the pKa values for this class (water, 20°C) are surprisingly homogeneous (first pK,: 3.2-4.4, last pKa: 10.0-11.5A2.3). A closer look reveals that ring size changes have only a small influence on the pKa values if compared to the changes brought about by alkyl substitution of carbon atoms in the ring. The comparability of the last pKa with EDTA suggests, too, that no special effects exist which are caused by the macrocycle. The compounds seem to react on their outside like any other non-macrocyclic acid or base. Some macrocyclic compounds with both acidic and basic functions have, however, been synthesized where at least one of the acid or base groups is located on the inside.

u. LUNING

104

Br I

a: Y = H b: Y

=

NOz

ONE INTRA-ANNULAR ACID-BASE FUNCTION

In the polyazamacrocyclic phenols [64], an acidic function, the phenol group, has an intra-annular orientation while the basic units, the nitrogen atoms of the macrocycles, have no defined orientation. In water at 25”C, the pK, values of the phenols were measured and compared with those of other macrocyclic and non-macrocyclic phenols (Kimura et al., 1987a,b). Because the nitrogen atoms and the phenol function both possess acid-base properties, more than one pKa value could be measured. By the use of UV measurements, the pKa values of the phenol group could be distinguished from those of the amines. When the pKa values of [64a] and [64b] are compared with those of the analogues it becomes obvious that [64b] behaves differently (see Table 27). The pKa value of the phenol function in [64b] is much smaller (4.8) than for the analogue 4-bromophenol (8.8), while the pKa of [64a] is only 0.5 pK, units smaller than the one of 4-bromophenol. The extreme behaviour of (64b) is presumably caused by Coulomb interactions. The dissociation of the phenol group takes place at the stage where the macrocycle is triply charged while for instance [64a] carries only two positive charges. The higher positive charge (protonation of three amine nitrogen atoms) facilitates the dissociation because of the charge reduction. Therefore the phenol group in [64b] is four orders of magnitude more acidic than 4-bromophenol. The difference between [64a] and [64b] therefore is probably not caused by a macrocyclic effect but by Coulomb forces. The resulting phenolate ion is a betaine. When the neutral macrocycles [64] were dissolved in solvents other than water, equilibria between the neutral forms and betaine structures were also found. In ethanol, the equilibrium between a phenol and a betainic

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

105

Table 27 pKa values for polyazamacrocyclic phenols molecules.

Compound [64a]" [64b]" 4-Bromo phenol [65aIb Y=H [65b]" Y = NO2 p-Nitrophenol [46a] (n = 2)d [46a] (n = 3)d

[64] and comparable

PKa1

PKa2

PKa3

PKa4

PK,S

3.7

4.2

8.3 (phenol) 4.8 (phenol) 8.8 8.9 (phenol) 6.4 (phenol) 7.1 10.8 10.6

10.5

10.7

9.6

10.3

10.5

11.8

10.4

11.8

<1

t 2

<2

€2

<2

c2

PKa6

11.3

"Kimura et al. (1987a). bKimura et al. (1985); in the work of Kimura et al. (1987a,b) are listed the pK, values of related compounds with different sizes of the cyclam ring showing that changing ring size has only a marginal influence on pK, for the phenolic group (kO.1). 'Kimura et al. (1987b). 'Browne et al. (1985).

Br

Br I

I

Scheme 10

phenolate is different for [64a] (1:l) and [64b] (1:0.25). A very small solvent dependence ([64b], 1:0.75 in CHC13) indicates that the betaine is formed rather than that the phenol dissociates (see Scheme 10). In water at pH 7 [64b] picks up two protons. INTRA-ANNULAR ACIDS AND BASES ONLY

There are some examples where acids and bases are both incorporated in intra-annular fashions in the same macrocycle (Cram and Trueblood, 1981; Bell et al., 1982; Uitenvijk et al., 1986). In Table 28 such a macrocycle [66] is compared with compounds discussed in Sections 2 and 3.

u. LUNING

106

Table 28 pKa of macrocycle [66] containing an intra-annular acidic and an intraannular basic functionality in comparison to related compounds.

In water, 22°C" 1661

[371

[33bI

PKa1 3.6

3.3

4.8

pKa2 5.8

PbI

[lo1

4.8

4.9

(n = 3)

In DMSO/H20 (2 :1 v/v) P71

[361 PKa1 6.30

6.65 "Bell et al. (1982).

[9b], n = 3

POI

~Ka2 8.00

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

107

In molecules with both an acidic and a basic function, there is in principle the possibility of an intramolecular proton transfer leading to betaines (see above). Due to the pK, values of the functional groups, pyridine and carboxylic acid, in [66] no intramolecular proton transfer takes place and no betaine is formed. A comparison of the pK, values of the macrocycle [66] with open-chain and macrocyclic analogues show differences indicating that in [66] hydrogen bonding is also an important factor. In an X-ray analysis of [66], intramolecular hydrogen bonds were found (Goldberg and Rezmovitz, 1978). Thus the structures of [66] and [36] are related [see above (Goldberg, 198l)l. The acidity of the COOH groups is decreased by the intramolecular hydrogen bonds. 5 Hydrogen bonding

The acid-base properties of the intra-annular acids and bases of Sections 2-4 usually have been investigated using strong acids or bases as counterparts. However, when weaker acids or bases are used, the formation of ions (proton transfer) is not the only possible reaction. The Bronsted acid and the base may also react to form a complex which is held together by hydrogen bonds. Whether ion formation or hydrogen bonding takes place depends on the pK, values of the acids and bases involved. As an example, macrocyclic pyridine derivatives may be protonated with strong acids (see Section 2) but, with weaker acids like phenols, they form hydrogen bonds (see below). While in protonations and deprotonations of a molecule only one proton is added or is subtracted, hydrogen bonds can only exist between two molecules: between a hydrogen bond donor and a hydrogen bond acceptor (when intramolecular hydrogen bonds are ignored). Hydrogen bonding therefore means association. And if the hydrogen bond formation takes place on the inside of a molecule, one molecule will be bound inside another. Therefore macrocycles with the ability of forming hydrogen bonds on their inside are potential structures for host-guest systems. The hostguest complex formation by hydrogen bonds has been reviewed quite often (Rebek, 1987; Lindoy, 1989; Vogtle, 1989, 1991; Izatt et al., 1991, 1992; see also Schneider, 1991; See1 and Vogtle, 1992; Dietrich et al., 1993). Therefore we will only highlight those examples where the acid or the base is constructed according to the 1,3-xylyl trick and where a possible ionization is conceivable. For example, guanidinium ions (Gu+) have been bound in pyridinocrowns [9b] (van Staveren et al., 1984; Grootenhuis et al., 1987a). Due to the difference In pK, [pyridinocrowns [9b] (n = 6-8): 3.4-3.7, guanidine: 13.61 no proton transfer was observed but hydrogen bonds were formed between the guanidinium ions and macrocycles of matching size. An

u. LUNING

108

[68a] R = H [68b] R = C1 [68c] R = NMe2

X-ray analysis of the complex [67] between 2,6-pyridino-l7-crown-9 [9b] (n = 6) and a guanidinium ion showed that one hydrogen bond is formed between the lone pair of the pyridine nitrogen atom and a hydrogen atom of Gu+. The other five hydrogen atoms are bound to ether oxygens in their vicinity. The association constants for the reaction of some macrocycles of type [9b] ( n = 6-8) with Guf were comparable (log K 1.18-1.44). The pK, values of pyridines and acidic phenols are such that good

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

109

"r"

r

.A

L

3

hydrogen bonds are formed. Whitlock built receptors [68] for phenols (Neder and Whitlock, 1990; Whitlock and Whitlock, 1990) and showed that acidic phenols like p-nitrophenol were bound by hydrogen bonds within the cavity. A series of different phenols has been complexed by [68]. The complex stability depended on the basicity of the pyridine ring. A p-value of c. -2.3 was calculated when different 4-substituted pyridines were used. The p-value for their protonation by p-toluenesulfonic acid was estimated to be -6 (Neder and Whitlock, 1990). Another class of phenol-binding pyridines has been synthesized by Vogtle (Ebmeyer and Vogtle, 1989; See1 and Vogtle, 1992). In the bimacrocycles [69], those phenols were especially well bound which form hydrogen bonds to all three bipyridine units, for example 1,3,5-trihydroxybenzene derivatives. Much rarer than the inclusion of phenols is complex formation between alcohols and pyridinomacrocycles. In some combinations (selected macrocycles and alcohols) complexes could be isolated (Weber and Vogtle, 1980). The hydrogen bond formation between pyridines and alcohols is the basis for an application, the additions of alcohols to ketenes catalysed by concave pyridines [13] (Luning et al., 1991b; Schyja, 1995). Another class of molecules which are easily bound to hosts by hydrogen bonds are nucleic bases and related molecules like barbiturates (Hamilton and Little, 1990; Chang et al., 1991). The receptor arrays to bind such molecules often contain 2-acylaminopyridine units or related structures such

u. LUNING

110

R I

R:C=C=O

+ R20H

RiCH-COOR’

as naphthyridines (Hamilton and Pant, 1988). The principle of the binding of a nucleotide into an acylaminopyridine macrocycle is shown in Scheme 11. Receptors of this type which bind dicarboxylic acids have also been synthesized (Garcia-Tellado et al., 1990). Also tris-endo-brenzcatechin macrocycles have been used to complex nucleic bases and analogues by hydrogen bonding (See1 and Vogtle, 1991, 1992).

Scheme 11

6 Closing remarks

In the previous Sections, the properties of acids and bases in macrocycles and other concave structures have been compared. A number of factors have been recognized which influence the acidity or basicity of an acid or base: (i) hydrogen bonds, (ii) hindered solvation (exclusion of solvent), (iii) formation of tight ion pairs (high microacidity but low overall acidity), and (iv) Coulomb forces when polyanions are formed. A fifth influence, (v) steric hindrance, still has to be discussed.

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

111

Table 29 pK, values for 2,6-substituted pyridines [70] in 50% ethanol/water.*

R' H Me Et Pr' But

R2

H

Me

Et

Pr'

But

a: 4.38

b: 5.05 f 5.77

c: 4.93

d: 4.82

e: 4.68 g: 5.52 h:5.36 j: 5.13 k:3.58

-

-

-

i: 5.34

"Kanner (1982).

Steric factors and especially hindered rotations may change the pKa of an acid by up to two pK, units as has been found for 2,6-disubstituted pyridines: 2,6-di-t-butylpyridine [70k] is an unusual (Kanner, 1982) nonnucleophilic base which has a surprisingly low pKa value. When [70k] is compared to other 2,6-dialkylsubstituted pyridines [70] it is found to be the only disubstituted pyridine with a smaller pKa than pyridine itself (see Table 29). The exceptional behaviour of [70k] has been investigated intensively

and it has been recognized that the low basicity is caused by entropic factors, by hindered mobilities of the t-butyl groups and solvating molecules (Hopkins and Ali, 1977; Kanner, 1982; Hopkins el al., 1984; Abboud el al., 1988). However, for most acids and bases of this review, the dominating influence of the concave environment on the acid-base properties is the formation of stabilizing intramolecular hydrogen bonds. The other factors have been found only in a few cases [microacidity for [45] (van Eerden et al., 1989) and perhaps destabilization of the anion solvation for [74] (Wangnick, 1991) and for [33] (Newcomb and Cram, 1975)l. While the formation of hydrogen bonds and tight ion pairs is stabilizing, steric hindrance is a destabilizing effect, but both effects can alter the acid-base properties. Drastic hindrance of rotation and solvation as found for [70k] has not yet been found for macrocyclic and related acids and bases. Understanding the acidity and/or basicity of acids and bases in macrocyclic systems has an important connection to biological problems. It should help us to understand enzymes. In many proteins, drastically varying pKa values have been found for functional groups in the active site; for example, the

112

u. LUNING

pKa of the active site lysine in acetoacetate decarboxylase is 6.0 while lysine in water has a pKa of 10.7 (Kokesh and Westheimer, 1971). One possible reason for such drastic changes may be the exclusion of water from the active site. For the intra-annular acids and bases discussed in this review, such an exclusion of solvent from the acid-base function will probably not be as strict as in enzymes. However, the trends found in macrocyclic systems should help to further understanding of acid-base properties in enzymes.

References Aarts. V. M. L. J.. van Staveren, C. J.. Grootenhuis, P. D. J.. van Eerden, J.. Kruise, L., Harkema, S. and Reinhoudt, D. N. (1986). 1. Am.' Chem. SOC. 108; 5035 Abboud, J.-L. M., Catalan, J., Elguero, J. and Taft, R. W. (1988). J. Org. Chem. 53, 1137 Adamic, R. J., Lloyd, B. A., Eyring, E. M., Petrucci, S. and Bartsch, R. A. (1986a). J. Phys. Chem. 90, 1659 Adamic, R. J., Lloyd, B. A., Eyring, E. M., Petrucci, S . , Bartsch, R. A., Pugia, M. J., Knudsen, B. E., Liu, Y. and Desai, D. H. (1986b). J. Phys. Chem. 90, 6571 Alberts, A . H. and Cram, D. J. (1979). J. Am. Chem. SOC. 101, 3545 Alder, R. W. (1990). Tetrahedron 46, 683 Alfieri, C., Dradi, E., Pochini, A., Ungaro, R. and Andreetti, G . D . (1983). J. Chem. SOC., Chem. Commun. 1075 Anderegg, G., Ekstrom, A., Lindoy, L. F. and Smith, R. J. (1980). J . Am. Chem. SOC. 102, 2670 Arnaud-Neu, F., Schwing-Weill, M.-J., Louis, R. and Weiss, R. (1979). Inorg. Chem. 18, 2956 Beer, P. D. and Keefe, A . D. (1987). J. Incl. Phenom. 5, 499 Bell, T. W. and Sahni, S. K. (1991). In Inclusion Compounds, vol. 4 (ed. Atwood, J. L., Davies, J. E. D. and MacNicol, D. D.). Oxford University Press, Oxford, New York, Tokyo Bell, T. W., Cheng, P. G . , Newcomb, M. and Cram, D. J. (1982). J. Am. Chem. SOC.104, 5185 Bencini, A., Bianchi, A., Bazzicalupi, C., Ciampolini, M., Dapporto, P., Fusi, V., Micheloni, M., Nardi, N . , Paoli, P. and Valtancoli, B. (1993). J. Chem. Soc., Perkin Trans. II 115 Berger, B., Bohmer, V., Paulus, E., Rodriguez, A. and Vogt, W. (1992). Angew. Chem. 104, 89; Angew. Chem. Int. Ed. Engl. 31, 96 Blanda, M. T. and Newcomb, M. (1989). Tetrahedron Lett. 30, 3501 Blanda, M. T., Horner, J. H. and Newcomb, M. (1989). J. Org. Chem. 54, 4626 Browne, C. M., Ferguson, G., McKervey, M. A., Mulholland, D. L., O'Connor, T. and Parvez, M. (1985). J. Am. Chem. SOC.107, 2703 Cassol, A., Di Bernardo, P., Zanonato, P. L., Tolazzi, M. and Tomat, G. (1990). Inorg. Chim. Acta 171, 217 Cesario, M., Dietrich, C. O., Edel, A., Guilhem, J., Kintzinger, J.-P., Pascard, C. and Sauvage, J.-P. (1986). J. Am. Chem. SOC. 108, 6250 Chang, S.-K., Van Engen, D., Fan, E. and Hamilton, A. D . (1991). J . Am. Chem. SOC. 113, 7640

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

113

Cheney, J. and Lehn, J . M. (1972). J . Chem. SOC., Chem. Commun. 487 Ciampolini, M., Micheloni, M., Vizza, F., Zanobini, F., Chimichi, S. and Dapporto, P. (1986). J. Chem. SOC., Dalton Trans. 505 Cox, B. G., Knop, D. and Schneider, H. (1978). J. Am. Chem. SOC. 100, 6002 Cox, B. G., Murray-Rust, J., Murray-Rust, P., van Truong, N. and Schneider, H. (1982). J. Chem. SOC., Chem. Commun. 377 Cram, D. J. and Trueblood, K. N. (1981). Top. Curr. Chem. 98, 43 Curtis, N. F. (1960). J . Chem. SOC. 4409 Czech, A., Czech, B. P., Bartsch, R. A., Chang, C. A. and Ochaya, V. 0. (1988). J . Org. Chem. 53, 5 Dietrich, B., Viout, P. and Lehn, J.-M. (1993). Macrocyclic Chemistry, VCH Weinheim Ebmeyer, F. and Vogtle, F. (1989). Angew. Chem. 101, 95; Angew. Chem. Int. Ed. Engl. 28, 79 Fabbrizzi, L., Kaden, T. A . , Perotti, A., Seghi, B. and Siegfried, L. (1986). Inorg. Chem. 25, 321 Fenton, D. E., Murphy, B. P., Leong, A. J., Lindoy, L. F., Bashall, A. and McPartlin, M. (1987). J . Chem. SOC.,Dalton Trans. 2543 Garcia-Tellado, F., Goswami, S., Chang, S.-K., Geib, S. J. and Hamilton, A. D. (1990). J . Am. Chem. SOC. 112, 7393 Gennari, C., Molinari, F., Bartoletti, M. and Potenza, D. (1992). Gazz. Chim. Zfal. 112, 279 Gokel, G. W. (1991). In Inclusion Compounds, vol. 4 (ed. Atwood, J. L., Davies, J. E. D. and MacNicol, D. D.). Oxford University Press, Oxford, New York, Tokyo Gokel, G . W. (1992). Chem. SOC. Rev. 21, 39 Gokel, G. W. and Korzenowski, S. H. (1982). Macrocyclic Polyether Synthesis, vol. 13 in Reactivity and Structure Concepts in Organic Chemistry. Springer-Verlag , Berlin Goldberg, I. (1976). Acta Cryst. B 32, 41 Goldberg, I. (1981). Acta Cryst. B 37, 102 Goldberg, I. and Rezmovitz, H. (1978). Acta Cryst. B 34, 2894 Goldmann, H., Vogt, W., Paulus, E. and Bohmer, V. (1988). J . Am. Chem. SOC. 110, 6811 Grootenhuis, P. D. J., van Staveren, C. J., den Hertog, Jr, H. J., Reinhoudt, D. N., Bos, M., Uiterwijk, J. W. H. M., Kruise, L. and Harkema, S. (1984). J. Chem. SOC., Chem. Commun. 1412 Grootenhuis, P. D. J., Uiterwijk, J. W. H. M., Reinhoudt, D. N., van Staveren, C. J., Sudholter, E. J. R., Bos, M., van Eerden, J . , Klooster, W. T., Kruise, L. and Harkema, S. (1986). J . Am. Chem. SOC. 108, 780 Grootenhuis, P. D. J . , Van Der Wal, P. D. and Reinhoudt, D. N. (1987a). Tetrahedron 43, 397 Grootenhuis, P. D. J., van Eerden, J . , Sudholter, E. J. R., Reinhoudt, D. N., Roos, A., Harkema, S. and Feil, D. (1987b). J . Am. Chem. SOC. 109, 4792 Gutsche, C. D. (1989). Calixarenes. Royal Society of Chemistry, Cambridge Gutsche, C. D. (1991). In Inclusion Compounds, vol. 4 (ed. Atwood, J. L., Davies, J. E. D. and MacNicol, D. D.). Oxford University Pre$,s, Oxford, New York, Tokyo Gutsche, C. D., Iqbal, M., Nam, K. C., See, K. A. and Alam, I. (1988). Pure Appl. Chem. 60, 483 Hamilton, A. D. and Little, D. (1990). J . Chem. SOC., Chem. Commun. 297 Hamilton, A. D. and Pant, N. (1988). J . Chem. SOC., Chem. Commun. 765 Hopkins, Jr, H. P. and Ali, S. Z. (1977). J . Am. Chem. SOC. 99, 2069

114

u. LUNING

Hopkins, Jr, H. P., Jahagirdar, D. V., Moulik, P. S., Aue, D. H., Webb, H. M., Davidson, W. R. and Pedley, M. D. (1984). J. Am. Chem. SOC. 106, 4341 Izatt, R. M., Pawlak, K., Bradshaw, J. S. and Bruening, R. L. (1991). Chem. Rev. 91, 1721 Izatt, R. M., Bradshaw, J. S., Pawlak, K., Bruening, R. L. andTarbet, B. J. (1992). Chem. Rev. 92, 1261 Kaden, T. A. (1984). Top. Curr. Chem. 121, 157 Kanner, B. (1982). Heterocycles 46, 683 Kimura, E. (1985). Top. Curr. Chem. 128, 113 Kimura, E., Koike, T. and Takahashi, M. (1985). J. Chem. SOC., Chem. Commun. 385 Kimura, E., Kimura, Y., Yatsunami, T., Shionoya, M. and Koike, T. (1987a). J . Am. Chem. SOC. 109, 6212 Kimura, E., Koike, T., Uenishi, K., Hediger, M., Kuramoto, M., Joko, S., Arai, Y., Kodama, M. and Iitaka, Y. (1987b). Znorg. Chem. 26, 2975 Kjm, A. M., Skensen, P. E. and Ulstrup, J. (1979). J. Chem. SOC., Chem. Commun. 965 Kokesh, F. C. and Westheimer, F. H. (1971). J. Am. Chem. SOC. 93, 7270 Lehn, J.-M. and Montavon, F. (1978). Helv. Chim. Acta 61, 67 Lehn, J.-M. and Sauvage, J. P. (1975). J. Am. Chem. SOC. 97, 6700 Lindoy, L. F. (1989). The Chemistry of Macrocyclic Ligand Complexes. Cambridge University Press, Cambridge Luning, U. (1987). Liebigs Ann. Chem. 949 Luning, U. and Muller, M. (1989). Liebigs Ann. Chem. 367 Luning, U. and Muller, M. (1990). Chem. Ber. 123, 643 Luning, U., Baumstark, R., Peters, K. and Schnering, H. G. v. (1990). Liebigs Ann. Chem. 129 Luning, U., Baumstark, R. and Muller, M. (1991a). Liebigs Ann. Chem. 9tZ7 Luning, U., Baumstark, R. and Schyja, W. (1991b). Liebigs Ann. Chem. 999 Luning, U., Baumstark, R., Wangnick, C., Muller, M., Schyja, W., Gerst, M. and Gelbert, M. (1993). Pure Appl. Chem. 65, 527 Luning, U., Muller, M., Gelbert, M., Peters, K., Schmering, H. G. V. and Keller, M. (1994). Chem Ber. 127, 2297 Muller, M. (1991). Ph.D. Thesis, Universitat Freiburg Murakami, Y., Ohno, T., Hayashida, 0. and Hisaeda, Y. (1991). J. Chem. SOC., Chem. Commun. 950 Nakashima, K., Yamawaki, Y., Nakatsuji, S., Akiyama, S., Kaneda, T. and Misumi, S. (1983). Chem. Lett. 1415 Nakashima, K., Nakatsuji, S., Akiyama, S., Kaneda, T. and Misumi, S. (1986). Chem. Pharm. Bull. 34, 168 Nakatsuji, Y., Bradshaw, J. S., Tse, P.-K., Arena, G., Wilson, B. E., Dalley, N. K. and Izatt, R. M. (1985). J. Chem. SOC., Chem. Commun. 749 Neder, K. M. and Whitlock, H. W. (1990). J. Am. Chem. SOC. 112, 9412 Newcomb, M. and Blanda, M. T. (1988). Tetrahedron Lett. 29, 4261 Newcomb, M. and Cram, D. J. (1975). J . Am. Chem. SOC. 97, 1257 Newcomb, M., Gokel, G. W. and Cram, D . J. (1974). J. Am. Chem. SOC. 96, 6810 Newcomb, M., Timko, J. M., Walba, D. M. and Cram, D. J. (1977a). J. Am. Chem. SOC. 99, 6392 Newcomb, M., Moore, S. S. and Cram, D. J. (1977b). J. Am. Chem. SOC. 99, 6405 Newcomb, M., Horner, J. H. and Blanda, M. T. (1987). J. Am. Chem. SOC. 109, 7878

ACID-BASE BEHAVIOUR IN CONCAVE STRUCTURES

115

Newkome, G. R., Majestic, V. K. and Fronczek, F. R. (1981). Tetrahedron Lett. 22, 3035 Patai, S. (1980). The Chemistry of Functional Groups, Supplement E, Part 1 and 2, The Chemistry of Ethers, Crown Ethers, Hydroxyl Groups and their Sulphur Analogues. Wiley, Chichester, New York, Brisbane, Toronto Pedersen, C. J. (1967). J. Am. Chem. SOC. 89, 7017 Poh, B.-L. and Lim, C. S. (1989). Tetrahedron Lett. 30, 1005 Poh, B.-L. and Lim, C. S. (1990). Tetrahedron 46, 3651 Rebek, Jr, J. (1987). Science 235, 1478 Rebek, Jr, J., Duff, R. J., Gordon, W. E. and Parris, K. (1986). J. Am. Chem. SOC. 108, 6068 Reetz, M. T., Niemeyer, C. M. and Harms, K. (1991). Angew. Chem. 103, 1517; Angew. Chem. Int. Ed. Engl. 30, 1474 Reinhoudt, D. N., Dijkstra, P. J., in’t Veld, P. J. A., Bugge, K. E., Harkema, S., Ungaro, R. and Ghidini, E. (1987). J. Am. Chem. SOC. 109, 4761 Schneider, H.-J. (1991). Angew. Chem. 103, 1419; Angew. Chem. Int. Ed. Engl. 30, 1417 Schyja, W. (1995). Ph.D. Thesis, Universitat Freiburg Seel, C. and Vogtle, F. (1991). Angew. Chem. 103, 433; Angew. Chem. Znt. Ed. Engl. 30, 442 Seel, C. and Vogtle, F. (1992). Angew. Chem. 104, 542; Angew. Chem. Int. Ed. Engl. 31, 528 Shinkai, S. (1993). Tetrahedron 49, 8933 Shinkai, S., Araki, K., Shibata, J., Tsugawa, D. and Manabe, 0. (1989). Chem. Lett. 931 Shukla, J. P., Jeon, E. G., Knudsen, B. E., Pugia, M. J., Bradshaw, J. S. and Bartsch, R. A. (1988). Thernzochim. Acta 130, 103 Skowronska-Ptasinska, M., Telleman, P., Aarts, V. M. L. J., Grootenhuis, P. D. J., van Eerden, J., Harkema, S. and Reinhoudt, D. N. (1987). Tetrahedron Lett. 28, 1937 Skowronska-Ptasinska, M., Aarts, V. M. L. J., Egberink, R. J. M., van Eerden, J., Harkema, S. and Reinhoudt, D. N. (1988). J . Org. Chem. 53, 5484 Smith, P. B., Dye, J. L., Cheney, J. and Lehn, J.-M. (1981). J . Am. Chem. SOC. 103, 6044 Stetter, H., Frank, W. and Mertens, R. (1981). Tetrahedron 37, 767 Takemura, H., Shinmyozu, T. and Inazu, T. (1991). J. Am. Chem. SOC. 113, 1323 Uiterwijk, J. W. H., van Staveren, C. J., Reinhoudt, D. N., den Hertog, Jr, H. J., Kruise, L. and Harkema, S. (1986). J. Org. Chem. 51, 1575 van Eerden, J., Skowronska-Ptasinska, M., Grootenhuis, P. D. J., Harkema, S. and Reinhoudt, D. N. (1989). J. Am. Chem. SOC. 111, 700 van Loon, J.-D., Groenen, L. C., Wijmenga, S. S., Verboom, W. and Reinhoudt, D. N. (1991). J. Am. Chem. SOC. 113, 2378 van Staveren, C. J., den Hertog, Jr, H. J., Reinhoudt, D. N., Uiterwijk, J. W. H. M., Kruise, L. and Harkema, S. (1984). J. Chem. Soc., Chem. Commun. 1409. van Staveren, C. J., Aarts, V. M. L. J., Grootenhuis, P. D. J., Droppers, W. J. H., van Eerden J., Harkema, S. and Reinhoudt, D. N. (1988). J. Am. Chem. SOC. 110, 8134 Vogtle, F. (1989). Supramolekulare Chemie. Teubner, Stuttgart; (1991) Supramolecular Chemistry. Wiley, Chichester Vogtle, F., Korsgen, U. U., Puff, H. and Reuter, H. (1989). Chem. Ber. 122, 343

116

u. LUNING

Vogtle, F., Breitenbach, J. and Nieger, M. (1991). J. Chem. SOC., Chem. Commun. 860 Wangnick, C . (1991). Ph.D. Thesis, Universitat Freiburg Weber, E. and Vogtle, F. (1980). Angew. Chem. 92, 1067; Angew. Chem. Znt. Ed. Engl. 19, 1030 Wehner, W. and Vogtle, F. (1976). Tetrahedron Lett. 2603 Whitlock, B. J. and Whitlock, H. W. (1990). J. Am. Chem. SOC. 112, 3910 Wu, G., Liang, W., Lamb, J. D., Bradshaw, J. S. and Izatt, R. M. (1991). J . Am. Chem. SOC. 113, 6538

Photodimerization and Photopolymerization of Diolefin CrystaIs MASAKIHASEGAWA Department of Materials Science and Technology, Faculty of Engineering, Toin University of Yokohama, Yokohama, Japan

1 Introduction - history of topochemical [2+2] photoreactions 2 Characteristic features of topochemical [2+2] photoreactions of diolefin crystals Typical four-centre-type photopolymerization and properties of the resulting polymers Polymerization mechanism based on topotaxy Non-topochemical factors in topochemical [2+2] photoreactions 3 Effects of wavelength of irradiating light and irradiation temperature Control of molecular chain growth by the wavelength of the irradiating light Effect of photoirradiation temperature 4 Kaleidoscopic topochemical behaviour of diolefin crystals “Even-numbered’’ degree of polymerization behaviour “Topochemical induction” into syndiotactic structures Absolute asymmetric synthesis and amplification of asymmetry Preparation of cyclophanes 5 Topochemical reactions of mixed crystals, inclusion complexes and molecular complexes Enhancement of photopolymerizability by complex formation Topochemical copolymerization of diolefin mixed crystals A perfectly ordered polymer composite by four-centre-type photopolymerization of a molecular complex 6 Concluding remarks References

1 Introduction

117 121 124 126 131 134 135 138 142 145 147 151 156 162 163 164 166 167 169

- history of topochemical [2+2] photoreactions

Various types of organic reactions in the crystalline state, so-called “topochemical reactions”, have been reported in this century, such as photodimerization of olefins, isomerization (Pincock et al., 1967; Wilson and Pincock, 1975), ring-opening polymerization (Okamura et al., 1962), formation of cage molecules (Yates and Singh, 1969), nucleophilic substitution (Sukenik et al., 1975, 1977), and acyl group migration (Gordan, 1967). 117 ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 30 ISBN 0-12-033530-1

0

Copyright 1995 Academic Press Limited All righfs of reproduction in any form reserved

118

M. HASEGAWA

Among these topochemical organic reactions, [2+2] photodimerization of olefin crystals is one of the most intensively investigated subjects. Formation of a-truxillic and p-truxinic acids from two types of cinnamic acid crystals has been known for a long time (Liebermann, 1889), and was interpreted in terms of a crystal lattice controlled reaction by Bernstein and Quimby (1943). The hypothesis was first visualized by Schmidt and co-workers, who correlated the crystal structures of several olefin derivatives with [2+2]photoreactivity and with chemical structures of the products; the reacting double bonds of the molecules must be located within about 4.2A and aligned parallel to each other (Cohen and Schmidt, 1964; Schmidt, 1971). The result established the topochemical principle in which the reaction tends to occur with a minimum of atomic and molecular motion. On the other hand, in all examples at that time, the crystal structure of the starting olefin compound had not been correlated with that of the resulting cyclobutane derivative because the crystallinity of the products was not high enough for crystal structure analysis. Crystalline state polymerization through thermal processes, or by photoor yray irradiation, had been a vast area of research in polymer chemistry between about 1950 and 1980. Vinyl-type and ring-opening polymerizations via chain reaction mechanisms in the crystalline state had been extensively investigated around 1960 by several groups worldwide. According to the crystallographic results on photoreactive olefin crystals, Hirshfeld and Schmidt (1964) had suggested the possibility of topochemical polymerization systems if suitable contact between the reactive centres of neighbouring molecules in the crystal could be maintained during the polymerization. Along with their postulate, Lahav and Schmidt (1966) had attempted several types of crystalline state polymerizations such as that of the dimethyl ester acid to yield cyclobutane derivatives. of all-truns-hexatriene-1,5-dicarboxylic In the polymerization systems investigated, however, no clear-cut example of a topochemical polymerization, in which all elementary steps proceed under a crystal lattice controlled process, had been found before the discovery of the photopolymerization (1) of 2,5-distyrylpyrazine (2,5-DSP) crystal in 1967 (Hasegawa and Suzuki, 1967). 2,5-DSP was initially prepared by Franke (1905), and in 1958 Koelsch described briefly in an article on pyridine N-oxide derivatives that “2,5distyrylpyrazine becomes white on exposure to uv-radiation, turning into an insoluble, possibly polymeric substance with a melting point of 331-333°C” (Koelsch and Gumprecht, 1958). It was independently confirmed that a brilliant yellow crystal of 2,5-DSP was converted into powdery white substance (“a very fine crystalline substance”) under the action of sunlight in the course of a preparative study of pyrazine-2,5-dicarboxylic acid from 2,5-DSP (Suzuki et ul., 1965; Hasegawa, 1966). This phenomenon was reinvestigated, and it was concluded that a linear high-molecular-weight

PHOTODIMER IZATlON AND PHOTOPOLYMERlZATl ON

119

[2,5-DSP]

polymer crystal ([4> 3.0 in CF,COOH solution)' with recurring cyclobutane units in the main chain had been produced from 2,5-DSP crystal by the action of sunlight (Hasegawa and Suzuki, 1967; Hasegawa et al., 1969). Crystallographic study of 2,s-DSP and poly-2,5-DSP demonstrated that the polymerization proceeded with retention of the space group (Pbca) of the starting 2,5-DSP crystal (Sasada el al., 1971; Nakanishi et al., 1972a). The result was the first evidence of an organic reaction which proceeded in the crystal lattice. Direct transformation from 2,5-DSP to poly-2,5-DSP through 2,5-DSP oligomer in the crystal is shown in Scheme 1. This new reaction was named a four-centre-type photopolymerization. As well as being the first example of a topochemical reaction in a pure sense, the four-centre-type photopolymerization of 2,5-DSP crystals was the first example of photopolymerization via a step-growth mechanism. After the first report on the crystal structure correlation between 2,5-DSP and poly-2,5-DSP crystals, a different crystallographic result was reported on a poly-2,5-DSP crystal (Meyer et al., 1978). It was reconfirmed, however, that the first structural analysis was correct (Nakanishi et al., 1979a). 2,5-DSP, crystallized from benzene solution, is highly photoreactive (aform), while the same compound, sublimed at a rather high temperature

' Solution viscosity is empirically related to molecular weight for linear polymers. Intrinsic viscosity ([q])is (q,dC),=,, where qspis (qr - 1) and C is a concentration of the polymer in solution. The quantity qc represents q/q,, where q and l), are the viscosity of the polymer solution and pure solvent, respectively. Inherent viscosity (qinh)is In(qJC).

M. HASEGAWA

120

hv

hv c

-

Scheme 1 Table 1 Crystal data of 2,5-DSP((u) and poly-2,5-DSP.”

Space group 2,5-DSP(a) poly-2,5-DSP

Pbca Pbca

alA

blA

20.638 9.599 18.36 10.88

CIA

Z

Dx

7.655 7.52

4 4

1.244 1.257

DistanceiA (>C=C<) 3.939 -

“Date from Sasada ef a/. (1971); Nakanishi ef a/. (1972a).

under reduced pressure, forms a different crystal structure and is photostable (yform). In the photostable 2,5-DSP ( 7 ) crystal, molecules form a characteristic layer-type packing without any overlap of adjacent molecules (Nakanishi et al., 1976). Such a different reactivity, between the same molecules with different crystal structures, clearly reveals that the crystalline state photopolymerization of 2,5-DSP (a)proceeds under the strict control of the reacting crystal lattice. Since then, several review articles on topochemical phenomena in the

PHOTODIMERIZATIONAND PHOTOPOLYMERIZATION

121

crystalline state [2+2] photoreaction have been published (Hasegawa et al., 1972; Thomas et al., 1977; Hasegawa, 1983; Ramamurthy and Venkatesan, 1987). Of further particular interest was that the crystallographic results on 2,5-DSP and poly-2,5-DSP had pointed out a very important future possibility that an “absolute” asymmetric synthesis could be achieved if any prochiral molecule, e.g. an unsymmetrical diolefin derivative, could be crystallized into a chiral crystal and if the reaction of the chiral crystal proceeded in the same manner as the 2,5-DSP crystal with retention of the crystal lattice (Wegner, 1972, 1973). Such types of absolute asymmetric synthesis with a high enantiomeric yield have now been performed by topochemical [2+2] photoreaction of unsymmetric diolefin crystals (Addadi et af., 1982; Hasegawa et al., 1990; Chung et al., 1991a,b). At present, based on the topochemical rule of Schmidt, the configuration of photoproducts, as well as photoreactivity, can be precisely predicted from the crystal structure of the starting olefin compounds with certain exceptions. A few examples of regioselective behaviour which cannot be explained by the topochemical rule have been discussed in terms of steric and electronic factors in the crystal (Hasegawa et al., 1992; Hasegawa, 1993). Recently the topochemical behaviour of mixed crystals and molecular complexes have been investigated focusing on copolymerization and on the reactivity- and product-controlling factors. In this chapter the topochemical [2+2] photoreactions of diolefin crystals are reviewed from the viewpoints of organic photochemistry, analysis of reaction mechanism, and crystallography as well as in terms of synthetic polymer chemistry and polymer physics.

2 Characteristic features of topochemical [2+2] photoreactions of diolefin crystals

After the discovery of the photopolymerization of 2,5-DSP crystals, several types of photoproducts were found, not only the linear polymers, but some other derivatives, e.g. the V-shaped dimer or cyclophane (Hasegawa and Hashimoto, 1992). The photopolymerization occurs in a step-growth mechanism by cyclobutane formation between the excited olefin and the olefin in the ground state. Typical four-centre-type photopolymerization of diolefin crystals, in which molecules are superimposed in the direction of the long molecular axes displaced by about half a molecule (a-type crystal packing), e.g. 2,5-DSP, ethyl 4-[2-(2-pyrazyl)ethenyl]cinnamate [l:OEt], proceeds very rapidly even in sunlight to give a linear high-molecular-weight polymer with an extremely high degree of crystallinity in nearly quantitative yield (2). On the other hand, diolefin crystals, in which molecules are superimposed without

M. HASEGAWA

122

[ 1 :OEt]

[2 :OPr-Cyclo]

displacement in the direction of the long molecular axis (P-type crystal packing), generally give amorphous dimer and oligomeric products in a moderate yield. Only a few crystals with P-type packing have been reported to give highly crystalline photoproducts, e.g. quantitative formation of [2.2] paracyclophane as shown in (3) from propyl a-cyano-4-[2-(4pyridy1)ethenyllcinnamate [2:OPr(P)], but only when it is recrystallized from EtOH/H,O solution (Maekawa et al., 1991b). For the diolefin crystals, including unsymmetrical diolefin crystals, each packing of the a- and P-types is further classified into translation- and centrosymmetry-type packings. Of the photoproducts derived from unsymmetrical diolefins, the cyclobutane ring which has the same substituent on a ring is called a homo-adduct, and that which has different substituents is called a hetero-adduct. Corresponding to the molecular arrangements of these diolefin crystals, four types of photoproducts (a-and P-types, and homo- and hetero-adducts) are expected to be formed based on the topochemical principle, as shown in Scheme 2.

I

cp

c L

L

L

L

o

e

124

M. HASEGAWA

The behaviour, which is not controlled by the topochemical rule but is greatly influenced by non-topochemical factors, is discussed in Section 2 in terms of molecular mobility, stabilization energy by orbital interaction and energy transfer in the crystals.

TYPICAL FOUR-CENTRE-TYPE PHOTOPOLYMERIZATION AND PROPERTIES OF THE RESULTING POLYMERS

As a typical example of topochemically prepared polymers, the 'H nmr spectrum of the polymer derived from ethyl 4-[2-(2-pyrazyl)ethenyl]cinnamate [l:OEt] crystals by reaction (2), and the X-ray diffraction patterns of the same monomer and polymer are illustrated in Figs 1 and 2 (Hasegawa et al., 1989a). = 8.19 dl g-' in hexafluoro-2-propanol, HFIP, soluThe polymer (qnh tion) in Figs 1 and 2 is prepared on photoirradiation by a 500W super-high-pressure Hg lamp for several hours and subjected to the measurements without purification. The nmr peaks in Fig. 1 (8 9.36, 8.66 and 8.63, pyrazyl; 7.35 and 7.23, phenylene; 5.00, 4.93, 4.83 and 4.42, cyclobutane; 4.05 and 1.10, ester) correspond precisely to the polymer structure which is predicted from the crystal structure of the monomer. The outstanding sharpness of all the peaks in this spectrum indicates that the photoproduct has few defects in its chemical structure. The X-ray patterns of the monomer and polymer in Fig. 2 show that they are nearly comparable to each other in crystallinity. These results indicate a strictly crystal-lattice controlled process for the four-centre-type photopolymerization of the [l:OEt] crystal. Crystal melting points of the polymers thus obtained are very high (>3OO0C) and generally higher than the temperature of thermal cleavage of the cyclobutane rings in the polymer main chain. Therefore, on heating, it is not the crystal melting point but topochemical thermal depolymerization of these polymers that is seen by means of continuous changes of X-ray diffraction pattern and differential scanning calorimetry-thermogravimetry (DSC-TG) curve in the course of thermal depolymerization (Hasegawa et al., 1975, 1978). By thermal treatment under reduced pressure, almost all of the polymers afford crystals of the original monomer as a sublimation product in high yield. Similar thermal treatment of the polymer crystals derived from unsymmetrical diolefin crystals gives not only the original monomer by symmetric thermal cleavage of the cyclobutane ring, but other diolefin monomers by asymmetric cleavage, as exemplified by the poly-2,5DSP derivative in Scheme 3 (Hasegawa et al., 1988a). The thermal stability of the polymers is remarkably dependent on their molecular weight; that is, the higher the molecular weight, the lower is their thermal stability. Such peculiar thermal behaviour has been interpreted as a

PHOTODIMERIZATIONAND PHOTOPOLYMERIZATION

I lb

I

I

125

-

ppm I

1

I

5

I

I

I

I

I

0

Fig. 1 400MHz 'H nmr spectrum (in CF3 COOD solution) of the polymer (poly-1:OEt) obtained by lyophilizing the polymer HFIP solution.

DIFFRACTIONANGLE 29 (")

Fig. 2 X-ray diffraction patterns of (a) 1:OEt and (b) poly-1:OEt.

M. HASEGAWA

126

Qy\

COOR

COOR

COOR

2.5-DSP

Scheme 3

common characteristic of rigid rod-shaped polymer chain structures (Hanamura, 1987). The polymer film cast from the solution is amorphous and difficult to recrystallize, either by thermal annealing of the amorphous film or by slow removal of the solvent from a polymer solution. To the touch, poly-2,5-DSP films are rather similar in stiffness to polystyrene films. Another peculiar property of amorphous poly-DSP is rapid photo-oxidation (4); the film becomes powdery within a few days under the action of sunlight in the air and gives cinnamonitrile dimer as the main product (Hashimoto et al., 1991).

POly-DSP (amorphous)

Sunlight in thc .iir

, NC

'3

CN

+

Photodegraded products

(4)

However, except for two reports of preliminary results on poly-DSP (Fujishige and Hasegawa, 1969; Kanetsuna el al., 1970), there are few reports so far which describe the film and solution properties of the polymers obtained by four-centre-type photopolymerization. POLYMERIZATION MECHANISM BASED ON TOPOTAXY

In all the photopolymerizable a-type crystals, nearly planar molecules are piled up and displaced in the direction of the molecular longitudinal axis by

PHOTODIMERIZATIONAND PHOTOPOLYMERIZATION

127

about half a molecule to form a parallel plane-to-plane stack. The periodicity in the stack is about 7 A . The shortest intermolecular distance between the double bonds in photopolymerizable crystals is about 4 A. When the double bonds in the stack react to form a cyclobutane ring by [2+2] photodimerization, the polymer chain should grow in the direction of the stack, as shown in Scheme 1. Rotation and Weissenberg photographs of poly2,5-DSP crystals (polymerized to complete conversion), which were taken around the axis along the direction of chain growth and of the c-axis of the monomer crystal, are shown in Figs 3(a) and (b), with those of the monomer (2,5-DSP) in Figs 4(a) and (b) (Nakanishi et al., 1972a). As shown in Figs 3(a) and (b), the polymer is three-dimensionally oriented. The lengths of the reflections of the polymer in the direction of the c-axis hardly increase from those of the monomer, as seen in their rotation photographs, Figs 3(a) and 4(a). On the other hand, the reflections in the Weissenberg photograph of poly-2,fi-DSP [Fig. 3(b)] are considerably elongated compared with those of the monomer [Fig. 4(b)] in the direction of the abscissa (w-direction), but not the ordinate (28-direction). These facts indicate that the polymer crystallites are regular in the direction of the c-axis but have some disorder in the (001) plane. After polymerization of 2,5-DSP, the c-axis (direction of chain growth) has contracted by 1.8% and the density has increased by about 1.0%. In the

[P2VB]

case of 1,4-bis-[a-pyridyl-(2)-ethenyl]benzene(P2VB), the crystal structure is very similar to that of 2,5-DSP and the c-axis has elongated by 3.0% and the density has decreased by 1.6% (Nakanishi et al., 1972b). The relative orientation of the crystals of 2,5-DSP and poly-2,5-DSP was determined from rotation and Weissenberg photographs of a partially polymerized crystal which were taken around the c-axis of the monomer crystal (Nakanishi et al., 1972a). In the rotation photograph somewhat diffuse spots of the polymer in layer lines almost overlap the layer lines of the monomer. In the Weissenberg photograph, diffraction streaks of the polymer crystal superimpose on the monomer reflections, and the a* and b* (reciprocal lattice of a and b , respectively) of the polymer agree with the corresponding axis of the monomer, suggesting that both the lengths and directions of the three crystal axes hardly change during the polymerization. Therefore, it was concluded that the a- and the b-axes of the polymer were

128

M. HASEGAWA

Fig. 3 X-ray diffraction photographs of poly-2,s-DSP. (a) Rotation photograph along the c-axis (oscillation angle 42"); (b) Weissenberg photograph of hkO zone.

PHOTODIMER IZATION AND PHOTOPOLYMERlZATlON

129

Fig. 4 X-ray diffraction photographs of 2,5-DSP. (a) Rotation photograph along the c-axis (oscillation angle 10'); (b) Weissenberg photograph of hkO zone.

130

M. HASEGAWA

[3:OR]

[4:OPr]

parallel to those of the monomer. The hOO reflections of the polymer shift slightly outwards from those of the monomer, while those of OkO move inwards, as expected from the unit cell dimensions of the polymer and the monomer crystals (see Table 1). In the 2,5-DSP(a) and P2VB crystals, the directions of three axes of the polymer coincide with those of the monomer whereas a different type of relative orientation is seen in diphenyl- and diethyl 1,Cphenylene diacrylates (3:OPh and 3:OEt), and dipropyl 1,4-(2-~yano)-phenylenediacrylate (4:OPr). In the latter type, the direction of the unique axis (b-axis) of the polymer coincides with that of the monomer while the directions of the other two axes do not. In the case of 3:OMe none of the directions of the axes of the polymer coincide with those of the monomer. However, the temperature effect on the reaction behaviour (see Section 3) and the continuous change of the X-ray diffraction pattern indicate a typical diffusionless crystal-lattice controlled mechanism (Hasegawa et al., 1981). Thus, the monomer crystal lattice control of the whole process from the monomer to the polymer is generally established for the four-centre-type photopolymerization of conjugated diolefin compounds. Topotaxies in the four-centre-type photopolymerization of several diolefins are shown in Table 2 Topotaxies in the four-centre-type photopolymerization of conjugated diolefins."

Group 1 2 3

Coincidence of crystal symmetry between monomer and polymer Crystal system, space group and directions of three axes Crystal system, space group and direction of unit axis Crystal system and space group

"Data from Nakanishi et nl. (1977).

Monomer

Crystal system

Space group

2,5-DSP P2VB 3:OEt 3:OPh 4:OPr 3:OMe

Orthorhombic Orthorhombic Monoclinic Monoclinic Monoclinic Triclinic

Pbca Pbca P2Ja P2,lc P21/a pi

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

131

Table 2 (Nakanishi et al., 1977). These results represent the first evidence that these reactions occur in the crystal lattice. Similar results and discussions have been introduced in other literature (Nakanishi et al., 1980; cf. Nakanishi et al., 1972a, 1977). NON-TOPOCHEMICALFACTORS IN TOPOCHEMICAL [2+2] PHOTOREACTIONS

Recently several examples of diolefin crystals in which the reaction behaviour deviates from the topochemical rule have been observed. For example, in the photoreaction of methyl cr-cyano-4-[2-(4-pyridyl)ethenyllcinnamate (2:OMe), the first reaction occurs exclusively at the pyridyl side although the distance between the ethylenic double bonds on the pyridyl side is exactly the same as that between the ethylenic double bonds on the ester side (4.049A), as shown in Fig. 5 (Maekawa et al., 1991a). A few other unsymmetrical diolefin compounds display the same regioselective behaviour (Hatada, 1989).

2:OMe

2:OMe-dimer

b

Fig. 5 Topochemical behaviour and crystal structure of methyl a-cyano-4-[2-(4pyridy1)ethenyllcinnamate (2:OMe).

132

M. HASEGAWA

Methyl 4-[2-(ethylthiocarbonyl)ethenyl]cinnamate (3:SMe) crystallizes into a typical a-translation-type packing structure in which the distances between the ethylenic double bonds are 3.988 8, and 4.067 A, respectively. However, the 3:6Me crystal is entirely photostable even though it should be photoreactive based on the topochemical rule (Sukegawa, 1991). Several examples of exceptionally photostable diolefin crystals have been found in compounds having a thioester moiety. Such anomalous behaviour of crystals such as 2:OMe and 3:SMe cannot be explained simply in terms of the topochemical rule since this rule involves only the positional relationship between the reactive olefin pair. Although photoquantum yield has been measured on some diolefin crystals in the course of photopolymerization, the explanation of the yield does not yet seem to be satisfactory either (see Section 3; Table 3). The regioselective behaviour of the 2:OMe crystal has been discussed from the viewpoints of steric and electronic factors: (i) the free cavity around two olefins of each monomer in the crystal, (ii) the change of packing potential energy during the dimerization within the stack, and (iii) the stabilization energy induced from orbital interaction between the reacting olefins (Hasegawa et af., 1992; Hasegawa, 1993). (i) The free cavity around two olefins of each monomer of 2:OMe in the crystal was calculated by the program “CAVITY” (Ohashi, 1980), using crystallographic data. The result reveals that there are not many cavities around the molecules in the direction of the movement necessary for the dimerization, and that the cyano group is particularly closely packed in the crystal of 2:OMe. The latter is a factor responsible for inhibiting the movement of the olefin having the cyano group. It follows that the selective [2+2] reaction may occur at the double bond on the 4-pyridyl side, in agreement with experimental observation. (ii) The change of packing potential energy within the stack during the dimerization of 2: OMe, was calculated by the original program “MOLALL”, using the same data as for the CAVITY calculation. For the calculation of atomic coordinates of the molecules during the photodimerization in the stack of the monomer crystal, the two olefins on either side of two reacting molecules approached each other by the same distance in the direction of the centre of the two olefins. The potential functions and parameters used in this work is the non-bonding atom-atom van der Waals potential of the Buckingham-Lennard-Jones type (Mirsky, 1978). (iii) The stabilization energy resulting from orbital interactions between two reacting molecules was calculated by the original program “MOLSTA” using the results of molecular orbital calculation by the AM1 method (Dewar et af., 1985) installed in the MOPAC (version 4.01). [2+2] Photodimerization between a monomer in the first excited singlet state (S,) and the neighbouring monomer in the ground state (So) was assumed. Molecular orbitals (MO) for the Franck-Condon S1 are obtained through

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

133

configuration interaction (CI) by putting two electrons into the highest occupied T MO (HOMO), the lowest unoccupied T MO (LUMO) and the next T LUMO (Fukui et al., 1960). According to the calculation, the intrinsically higher reactivity of the double bond on the 4-pyridyl side, helped by the steric factor in the crystal, results in the regioselective photodimerization of 2:OMe at the pyridyl side double bond. The perturbation energies for the selected frontier MO (FMO) were also calculated. From such calculations, a non-synchronous concerted reaction mechanism has been supported by several quantum chemical results (Bernardi et al., 1987). The analysis of the regioselective reactivity of olefins in identical topochemical environments by three computational methods concludes that both steric factors (cavity and potential energy) and electronic factors (perturbation energy from orbital interactions) play important cooperative roles in determining which C-C double bond in a molecule reacts first in [2+2] photodimerization. The steric factor is considered to be effective in the movement of olefins at an early stage of the reaction, whereas the electronic factors are effective in the adduction of olefins at a later stage of the reaction. In previous studies in the author’s laboratory it has become clear that, in most diolefin derivatives, replacement of the oxygen atom of an ester moiety by a sulfur atom is possible without changing the photopolymerizable molecular arrangement, and that all of the thioester derivatives and even mixed crystals of the ester containing a small amount of thioester derivatives of 1,4-phenylene diacrylate (PDA) are photostable (Hasegawa et al., unpublished data). In a comparison of fluorescence spectra between the ester and thioester derivative crystals of PDA, the ester crystal shows a strong emission whereas the thioester crystal fluoresces much more weakly. For example, the intensity of a PDA methyl thioester crystal is about one-thousandth of that of a PDA methyl ester crystal. Furthermore, the fluorescence lifetime of mixed crystals which consist of a large amount of PDA methyl ester and a small amount of the corresponding thioester moiety is much shortened, compared to the lifetime of pure PDA methyl ester crystals. In quenching experiments in solutions of PDA ester, the fluorescence of the PDA ester is dramatically quenched by thioacetate. Similar behaviour has been obtained with several types of diolefin derivatives having a thioester moiety, where crystal structures are isomorphous with the corresponding ester derivatives. From the results of the fluorescence spectroscopic study it is concluded that excitation energy at the lowest excited state of a PDA derivative having a thioester moiety is localized at the thioester group; intra- or intermolecular energy transfer from the conjugated system of the PDA to the thioester moiety must have occurred in the crystalline state to afford a photostable crystal (Hasegawa et al., unpublished data).

M. HASEGAWA

134

3 Effects of wavelength of irradiating light and irradiation temperature

As the topochemical [2+2] photoreaction proceeds by a step-growth mechanism and since the photoreactions of diolefin crystals observed so far are restricted to molecules in which two reacting olefin bonds are conjugated to each other through an aromatic moiety,' the reaction conversion can be readily controlled by the irradiation interval and by the wavelength of the irradiating light. Furthermore, in order to avoid excessive thermal vibration of the crystal lattice, the irradiation temperature generally should be considerably lower than the melting point of the starting crystals. For a-type crystals, the molecular weight and its distribution in the final polymer can be widely controlled by these three factors (the photoirradiation interval, wavelength, and temperature). For /3-type crystals, in an extreme case, even the chemical structure of the topochemical product varies strikingly with irradiation temperature. Photochemical quantum yields depend on the wavelength and also on the irradiation temperature.

>280 nm at - 10 "C 4

+

254 nm at 2 "C

(5:OMel

It is well known that the [2+2] photodimerization of diolefinic compounds is allowed to occur photochemically but not thermally, whereas the cyclobutane cleavage reaction occurs both photochemically and thermally. The cleavage reaction occurs with irradiating light of shorter wavelength Although [2+2] photopolymerization of crystals of glycol bis-cinnamate has been reported (Miura el al., 1968), it has not proved possible, despite several attempts in the author's laboratory, to confirm that the polymer formed contains cyclobutane rings. In the monomeric molecule, the olefinic bonds are not conjugated to each other.

PHOTODIM ER lZATl0 N AND PHOTO POLYME R lZATl0 N

135

than that for the photodimerization because cyclobutane compounds generally absorb at shorter wavelengths than the original olefinic compounds. Consequently, the wavelength of the irradiating light and the irradiation temperature give rise to various significant effects on topochemical [2+2] photoreaction behaviour. By such a control technique a clear reversible topochemical photoprocess has been realized between methyl 4-[2-(4-pyridyl)ethenyl]cinnamate [5:OMe] and corresponding polymer crystals under the reaction conditions given in (5) (Hasegawa, 1986; Hasegawa et al., 1990; Chung and Hasegawa, 1991). However, a reversible photochemical process between olefin bonds and a cyclobutane ring is generally not observed so clearly in the crystalline state, because the crystal lattice favours the polymerization but depresses the photodepolymerization.

CONTROL OF MOLECULAR CHAIN GROWTH BY THE WAVELENGTH OF IRRADIATING LIGHT

Since all the photopolymerizable monomers (A) contain two conjugated double bonds, the T-T* electronic transition of a dimer and a molecule larger than a dimer (B) is shifted to a higher energy level than that of A. The reaction scheme is as shown in (6)-(11) (Tamaki et al., 1972), where A* and B* represent the species A and B, respectively, in the T-T* excited state. Equation (8) represents a dimerization reaction and (9)-( 11) represents growth reactions. On photoirradiation at the longer wavelength edge of the monomer A ( h q ) , e.g. 430nm for 2,5-DSP, only reactions (6), (8) and (9) proceed to give the oligomer exclusively. The degree of polymerization at this stage is determined by the reaction velocity ratio of growth step (9) to the dimerization step (8). For 2,5-DSP the pentamer on average is obtained. By a similar technique the formation of 2,5-DSP trimer on average was reported by Wegner and co-workers (Braun and Wegner, 1983). On successive photoirradiation of the as-prepared oligomer crystals with the wavelength of light which excites B ( h y ) , e.g. 350nm for 2,5-DSP oligomers, high-molecular-weight polymer crystals are produced by step (1 1) which is a growth reaction of the terminal unit in the growing chain. By such a selective excitation technique, step (10) actually does not occur because all the monomer molecules (A) have already been exhausted in the course of oligomerization. The ultraviolet (UV) absorption of 2,5-DSP crystal as measured in a thin layer deposition on a quartz plateis shown in Fig. 6. The quantum yields, the number of olefinic double bonds consumed to form cyclobutane per absorbed quantum, of the oligomerization and polymerization of 2,5-DSP, P2VB and 3:OMe have been measured by using monochromatic light. The quantum yield (@) is defined by the equation

M. HASEGAWA

136

RI

A:

\

RIIl

/

C=CH- Ar-CH=C \ / RII RIV

Photoexcitation processes A B Photocyclodimerization processes

hui

hv2

A*

B*

A*+A

B

A*+B-

B

B*+A-

B

B*+B-

B

(7)

@ = (dc/dt)/Zab, where dcldt is the rate of disappearance of the olefinic double bonds per unit volume and Zabs the rate at which the incident light is absorbed per unit volume of the KBr pellet containing the sample. The rates of disappearance of the olefinic double bonds during oligomerization and polymerization were monitored by infrared (IR) spectroscopy. The initial quantum yields for the oligomerization and the polymerization of 2,5-DSP, P2VB, and 3:OMe are summarized in Table 3. The quantum yields of oligomerization and polymerization of 2,5-DSP and 3:OMe are between 0.7 and 1.6. These quantum yields indicate that these photoreactions belong to a class of single photon reactions in which the theoretical maximum value is equal to 2, and that these reactions proceed very efficiently. On the other hand, the quantum yield of oligomerization of P2VB is only 0.04 and that of polymerization is very small and not measurable by the same technique, even though the crystal structures of 2,5-DSP and P2VB are isomorphous. Recently Eckhardt and co-workers explained the difference of quantum yields between 2,5-DSP and P2VB in terms of energy trapping caused by the m* t n excitation-phonon coupling which was observed only in 2,5-DSP (Peachey and Eckhardt, 1993; Stezowski et al., 1993). On the basis of a concerted X-ray crystallographic and polarized, single-crystal Raman scattering study of a partially reacted

PHOTODIMERlZATlON AND PHOTOPOLYMER IZATION

137

I

1

I

250

300

I

1

I

350 400 Wavelength (nm)

450

Fig. 6 UV spectra of (a) 2,S-DSP crystal, (b) 2,5-DSP oligomer crystal, and (c) poly-2,S-DSP crystal.

Table 3 Quantum yields (a) of oligomerization and subsequent polymerization of 2,5-DSP, P2VB, and 3:OMe."

2,S-DSP P2VB 3:OMe

Wavelength used in irradiation (o1igomerization)hm

CP

436 405 365

1.2 0.04 1.2

Wavelength used in irradiation (po1ymerization)hm

CP

365

1.6

313

0.7

-

-

"Data from Tamaki et al. (1972).

2,5-DSP crystal, a detailed picture of the lattice motion and related displacements was constructed and related to the topochemical postulate and the mechanism of phonon assistance. Holm and Zienty (1972) have measured the quantum yield for the overall polymerization process of (~,(~'-bis(4-acetoxy-3-methoxybenzylidene)-p-benzenediacetonitrile (AMBBA) crystals in slurries and reported it to be 0.7 on the basis of the disappearance of two double bonds (a = 1.4 if assigned on the basis of the number of double bonds consumed).

(AMBBA]

138

M. HASEGAWA

EFFECT OF PHOTOIRRADIATION TEMPERATURE

As the topochemical reaction is intimately related to molecular dynamics during the react.ion process, thermal motion of reacting molecules in the crystal must be taken into account in the reaction. In typical four-centretype photopolymerization of a-type diolefin crystals it is generally observed that, with decreasing photoirradiation temperature, a higher final degree of polymerization is achieved while the apparent reaction rate is depressed. That is, the four-centre-type photopolymerization of diolefin crystals proceeds in good order, providing the photoirradiation temperature is retained within the range in which a rigid crystal lattice is maintained. For example, 3:OEt (melting point 96°C) photopolymerizes quantitatively into a linear higher-molecular-weight polymer crystal on photoirradiation at a temperature below 0°C (Nakanishi et al., 1973), even when the temperature is as low as that of liquid helium (4.2 K) (Gerasimov et al., 1974). On the other hand, when the irradiation is performed at a temperature above room temperature, the same crystal does not give a linear high-molecularweight polymer but gives fairly low-molecular-weight oligomers, as is shown in Scheme 4. The dimer is produced on photoirradiation at 6045°C (30% conversion) and isolated as fine crystals from the oligomeric photoproducts (Nakanishi et al., 1974). In contrast to the a-type crystal, the photoreaction of P-type diolefin crystals at a very low temperature sometimes does not occur at all, or sometimes proceeds but levels off at a low conversion, suggesting that the photoreaction of P-type diolefin crystals requires an appropriate thermal motion oi the reacting molecules. It should be emphasized that, in all the topochemical photoreactions without exception, an apparent reaction rate at the initial stage increases with increase in the irradiation temperature, as long as the temperature is sufficiently low to maintain the molecular orientation in the crystal. In order to rationalize such characteristic kinetic behaviour of the topochemical photoreaction, a reaction model has been proposed for constant photoirradiation conditions (Hasegawa and Shiba, 1982). In such conditions the reaction rate is assumed to be dependent solely on the thermal motion of the molecules and to be determined by the potential deviation of two olefin bonds from the optimal positions for the reaction. The distribution of the potential deviation of two olefin bonds from the most stable positions in the crystal at OK is assumed to follow a normal distribution. The reaction probability, which is assumed to be proportional to the rate constant, of a unidimensional model is illustrated as the area under the curve for temperature T2 between 6 and 6 + W in Fig. 7. k = A exp { - (RT,,,

+ E,)/RT}

a,

2

5

.-OI

a

+ 5 0

8

..

m

I

C

T -t-

140

M. HASEGAWA

Fig. 7 A model for topochemical reaction probability.

The rate equation for the two-dimensional model is then given by (12), where A = PT,,,/T and /3 is a constant related to the individual reaction, and Topt represents the temperature where the reaction occurs most efficiently. Of the P-type crystals, the irradiation temperature affects not only the rate and final conversion, but sometimes it causes rearrangements of the molecular skeleton in the final product. Such a product control technique by variation of the photoirradiation temperature can be seen in the reactions of a few P-type diolefin crystals. When propyl a-cyano-4-[2-(4pyrimidyl)ethenyl]cinnamate (6:OPr) crystals are irradiated with A > 410 nm at room temperature, they give a monocyclic dimer nearly quantitatively while when the same crystals are irradiated with A>300nm at lower temperature, they give the highly strained [2.2] paracyclophane in yields of 6% and 27% at irradiation temperatures of -40°C and -78”C, respectively (Chung et al., 1991a). The observed correlation between the yield of cyclophane and the irradiation temperature suggests the existence of a highly strained “non-isolable” intermediate dimer molecule that is frozen at the lower temperature. The whole reaction scheme is shown in Scheme 5. The modification of molecular conformation from the highly strained “non-isolable” dimer molecule to the V-shaped dimer molecule (6:OPrdimer) is explained in terms of relaxation of the strain energy due to the bond angle in the “non-isolable” dimer, which accumulated during the cyclobutane formation. Therefore, strictly speaking, the process going from the “non-isolable” dimer into the V-shaped dimer (6:OPr-dimer) is not a

::;i6")",,I '

.

COOPr CN

.

N,N

COOPr

6:OPr-dimer

PrOOC

J

6:OPr "

-78

c

N-N

COOPr

( 27 Yo)

Non-isolable" dimer 6: OPr-cyclo Scheme 5

142

M. HASEGAWA

topochemical process but a thermal rearrangement under moderate control by the crystal lattice. Similar behaviour has been observed in the photoreaction of methyl a-cyano-4-[2-(2-pyridyl)ethenyl]cinnamate (7:OMe) crystals in which the yield of [2.2] paracyclophane reached 65% on irradiation at -78°C (see Scheme 10; p. 153) (Hasegawa et al., 1989b). From the crystal structure analysis of the same type of [2.2] paracyclophane, which is topochemically derived from alkyl a-cyano-4-[2-(4-pyridyl)ethenyl]cinnamatecrystals, a highly strained molecular shape is confirmed in which two phenylene rings are severely bent (Maekawa et al., 1991b). These results are significant in understanding the photochemical behaviour of organic crystals at low temperature and in the development of a new synthetic route to highly strained [2.2]paracyclophane derivatives. 4 Kaleidoscopic topochemical behaviour of diolefin crystals

At present, based on the topochemical rule, the configuration of photoproducts, as well as photoreactivity, can be precisely predicted from the crystal structure of the starting olefin compounds, with certain exceptions. On the other hand, the crystallization process of diolefin compounds often plays a significant role in determining their topochemical behaviour, by changing their crystal structure or by forming solvent inclusion complexes. Furthermore, topochemical photoreactions of crystals with P-type packing are accompanied by thermal processes under moderate control by the reacting crystal lattice (see p. 140). These factors seriously complicate the whole reaction scheme. Such “kaleidoscopic” topochemical behaviour of 2:OPr is exemplified in Scheme 6 (Hasegawa, 1992). 2:OPr crystallizes with a-type packing [2:OPr(a)] from 1-propanol solution but into the P-type [2:OPr(P)] from a mixture of ethanol and water. On photoirradiation using a filter (>410 nm), the a-type homo-adduct dimer (2:OPr-a-dimer) and the P-type heteroadduct dimer (2:OPr-P-dimer) are produced from 2:OPr(a) and 2:OPr(P), respectively, both in nearly quantitative yield. The dimer crystal 2:OPr-adimer as-prepared is entirely stable on photoirradiation whereas, on further photoirradiation without a filter, 2:OPr-/3-dimer is converted nearly quantitatively into a highly strained [2.2] paracyclophane (2:OPr-cyclo). The photostable 2:OPr-a-dimer crystal is, however, transformed into a highly photoreactive crystal complex (2:OPr-a-dimer.PrOH) if 2:OPr-a-dimer is recrystallized from 1-propanol solution. The crystal (2:OPr-a-dimer-PrOH) photopolymerizes into a crystalline a-type homo-adduct linear polymer. Judging from their X-ray diffraction patterns, recrystallized P-type dimer (2:OPr-P-dimer) has a different crystal structure from that of the crystal

+ -*

2:OPr(a)

(photostable)

I

2:OPr-a-dimer recryst. from PrOH

recryst. from PrOH

hv

2:OPr-a-Dimer-PrOH - N

COOPr

2:0prCN recryst. from EtOH I H20

\

2: 0 Pr(B) 2:OPr-B-dimer(as-prep.)

2: 0 Pr-cyclo

hv

2: OPr-B-dimer(recryst)

Scheme 6

oligomer

M. HASEGAWA

144

2:OPr-P-dimer (as-prepared). On photoirradiation, 2: OPr-P-dimer (recrystallized) does not afford 2:OPr-cyclo at all, but rather an oligomer. All these types of behaviour are reasonably interpreted by crystallographic analysis of these compounds, based on topochemical consideration. In this section, various types of topochemical behaviour such as the “even-numbered’’ degree of polymerization mechanism, topochemical induction into the syndiotactic structure, stereo- and enantio-selective reactions, and the formation of highly strained cyclophanes are described.

“EVEN-NUMBERED” DEGREE OF POLYMERIZATION BEHAVIOUR

Ethyl a-cyano-4-[2-(4-pyridyl)ethenyl]cinnamate (2:OEt) crystallizes with photoreactive a-centrosymmetric-type packing. Upon irradiation of 2:OEt crystals with a 1OOW super-high-pressure Hg lamp at room temperature, an a-homo-type linear polymer (M,= 3100) is produced with the accumulation of one type of dimer at the intermediate stage (Scheme 7) (Hasegawa, 1986). On the other hand, exclusive photoexcitation of the monomer with wavelengths longer than 410nm results in a quantitative formation of the dimer (2:OEt-dimer). The gel permeation chromatography (GPC) profile at an early stage of the photoreaction (2h) of 2:OEt crystals (Fig. 8) shows that the products consist of molecular species with only an “even-numbered” degree of polymerization and which have only pyridyl ethenyl groups at growing terminals. The even-numbered degree of polymerization behaviour of 2:OEt can be interpreted from its crystal structure (Fig. 9), in which the molecule is related to its neighbouring molecules by two different inversion centres to make a plane-to-plane stack (Maekawa et al., 1991a). The ethylenic double

1

I

20 30 Elution volume (ml)

Fig. 8 GPC profile at the early stage of the photoirradiation of 2:OEt crystal.

g\ 0 $ 8 z

M. HASEGAWA

146

Fig. 9 The crystal structure of 2:OEt.

bonds on the ester side, related by one inversion centre, are separated by 3.758 A, whereas the ethylenic double bonds on the pyridyl side, related by the other inversion centre, are separated by a distance of 4.868A; the former ethylenic double bonds can react predominantly according to the topochemical principle. The GPC profile and the crystal structure show that the monomer reacts only with the monomer and that further chain growth occurs between two dimer molecules. Between the dimers the intermolecular pyridyl ethenyl groups must have come closer to each other and within the reactive distance (-4 A). Accordingly, the “even-numbered polymerization” mechanism can be explained on the basis of the difference in topochemical environment; the polymerization proceeds through a single type of dimer, accumulated spontaneously during the photoirradiation, to give an even-numbered polymer species, as shown in (13). During the photoreaction of ethyl methyl 1,4-~henylenediacrylatecrystals (3:OEtMe), in which the distances of intermolecular double bonds are 3.891 8, between two methyl cinnamate groups and 4.917 A between two M+M

- hv

Dimer

hv

M+Dimer

Trimer

M+Trimer

Tetramer

Polymer (M-M),, (13)

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

147

Fig. 10 The crystal structure of 7:OMe-dimer.

ethyl cinnamate groups, again only a single type of dimer having ethyl cinnamate terminal groups is accumulated as indicated by nmr spectroscopy (Hasegawa et al., 1986a). However, a GPC curve at an early stage of photopolymerization does not show a clear even-numbered degree of polymerization behaviour, indicating the occurrence of the reaction of the monomer with other molecular species to a certain extent.

“TOPOCHEMICAL INDUCTION” INTO SYNDIOTACTIC STRUCTURES

Crystals of methyl a-cyano-4-[2-(2-pyridyl)ethenyl]cinnamate(7:OMe) with P-type packing are highly photoreactive and are converted into the P-type dimer (7:OMe-dimer) in high yield upon photoirradiation (>410 nm, -40°C).The dimer (7:OMe-dimer), on further photoirradiation (> 300 nm, -40°C) after recrystallization from methanol/l,2-dichloroethane, gives a = 12000) !I,,via a tetramer (7:OMe-tetramer) with retention of polymer ((& high crystallinity. The ‘H nmr analysis of 7:OMe-tetramer suggests that it is an a-homo type formed by the cycloaddition of ethylenic double bonds on the ester side of 7:OMe-dimer (Hasegawa et al., 1989b). Finally, the polymer (poly-7:OMe) has 7:OMe-tetramer as a repeating unit according to the ‘H nmr spectrum of poly-7:OMe. On the basis of the geometry of the two pyridyl side olefins of (7:OMe-dimer) (Fig. lo), the olefinic pair is expected to yield an &-type(or

148

M. HASEGAWA

&type) cyclobutane ring.3 Thus, the polymer structure should have a unique repeating unit in which three types of cyclobutane structures are incorporated in a sequence,,of [ ~ P E P ~ P - ’ E P - ’(or ] [a/31jpaP-1&3-1]),where P and P-’ are of opposite absolute configuration. Considering the stereochemistry of the cyclobutane rings, the polymer should be “double syndiotactic” (Scheme 8) (Chung et al., 1991b). From the viewpoint of synthetic polymer chemistry, although the formation of stereospecific polymers (isotactic and syndiotactic) is very popular, the present polymer is the first example having a double syndiotactic structure. In addition, the polymer consists of an alternating zigzag-linear main chain structure. Ethyl a-cyano-4-[2-(2-pyridyl)ethenyl]cinnamate(7:OEt) also crystallizes with P-centrosymmetric-type packing yielding photoreactive crystals and, upon photoirradiation (>410 nm), is converted into the p-hetero-type dimer (7:OEt-dimer) nearly quantitatively. The 7:OEt-dimer (space group Pi, triclinic) has the structure which is predicted from the crystal structure of 7:OEt (space group Pi, triclinic). In the crystal of 7:OEt, two molecules form a molecular pair as is the case in the crystal of 7:OMe. Considering the intermolecular distances between the ethylenic double bonds (3.714 and 3.833A within the pair, and 4.734 and 4.797A between the pairs), each molecule can react only with its partner in the molecular pair and not with any molecule of another pair. Since paired molecules are related by centrosymmetry, two pairs of facing ethylenic double bonds should be equal in photoreactivity, affording two

3

Designated by Annet (1962).

0 ..

s

.i

a,

L

.-E U

.. r-

PHOTODIMERlZATlON AND PHOTOPOLYMERlZATlON

151

enantiomeric cyclobutanes. However, after the formation of 7:OEt-dimer in one molecular pair followed by the thermal process (modification of the molecular conformation from the “non-isolable” dimer into the V-shaped dimer) (see p. 140), the photoreactivity of the two pairs of ethylenic double bonds in neighbouring molecular pairs should no longer be equal; the ethylenic double bonds in the pair of 7:OEt, adjacent to the cyclobutane side of the 7:OEt-dimer, is less reactive due to the steric repulsion of the substituents that protrude after dimerization. Upon further photoirradiation, 7:OEt-dimer gives the tetramer (7:OEttetramer) with a certain amount of octamer and higher oligomers. The tetramer (7:OEt-tetramer) contains three cyclobutane rings of P-hetero-, a-homo- and P-hetero-type structures. In addition, on the basis of the geometry of the two ethylenic double bonds on the ester side of the 7:OEt-dimer, the polymer should have a unique repeating unit in which three types of cyclobutane structures are incorporated in a sequence of [pap-’a] where P and P-’ are enantiomeric (“syndiotactic”) (Scheme 9). At the same time, as is the case for 7:OMe, the repeating unit consists of an alternating zigzag-linear chain structure as seen in Figs ll(a) and (b). The photochemical behaviour of 7:OEt is the first example in which the reaction of achiral molecules in an achiral crystal packing does not occur at random but stereospecifically, resulting in a syndiotactic structure. As no external chiral catalyst exists in the reaction, the above result is a unique type of ‘Ltopochemicalinduction”, which is initiated by chance in the formation of the first cyclobutane ring, but followed by syndiotactic cyclobutane formation due to steric repulsions in the crystal cavity. That is, the “syndiotactic” structure is evolved under moderate control of the reacting crystal lattice. The complete reaction processes of 7:OMe and 7:OEt crystals are shown in Scheme 10. The concept of topochemical induction was first assumed in the topochemical formation of a cyclic dimer from 1,4-dicinnamoylbenzene (1,4DCB) crystal (see p. 157) (Hasegawa et al., 1985).

ABSOLUTE ASYMMETRIC SYNTHESIS AND AMPLIFICATION OF ASYMMETRY

Along with the guidepost (Wegner, 1972, 1973) based on the crystal-tocrystal transition from 2,5-DSP to poly-2,5-DSP, absolute asymmetric synthesis has been achieved by the topochemical reaction of a chiral crystal of an achiral diolefin compound in the absence of any external chiral reagents. Ethyl 4-[2-(4-pyridyl)ethenyl]cinnamate (5:OEt) crystals (P-type packing) gives an optically active dimer through a topochemical [2+2] photocycloaddition (enantiomeric effect > 90%). The asymmetric induction is ex-

M. HASEGAWA

152

1R,ZS, 3R,4R B

n B

1S,2R,3S,4s

Fig. 11 (a) Schematic polymer structure of poly-7:OEt. Phenylene rings are omitted in order to simplify. (b) Molecular model of repeating structure of poly-7:OEt. Four chiral centres on each of two cyclobutane rings in both sides are enantiomeric to each other.

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

153

7:OR C N hv

1

< -40 "C

"on- isolable-

NC MeOOC

-

\

N/

7: OM e-cyclo

7:OEt-tetra hv

I

$.

hv

polymer ( i n = 12 000)

I

oligomer

Scheme 10

hv

(+) - 5:OEt-dimer (+)

5:OEt

- crystal

I I I

.

I

--

(-) 5:OEt-dimer

(The signs of optical rotations are arbitrary.)

Scheme 11

PHOTODIMER IZATION AND PHOTOPOLYMER IZATlON

155

Table 4 Examples of the crystallization of 5:OEt with or without seeding and their photoreaction into the chiral dimer.“ Photoirradiation: Example

Crystallization of 5:OEt

Time/h

Temp./”C ~~

1 2 3 4 5

With (+) seeding With (-) seeding Without seeding Without seeding Without seeding

10 7 1 1 1

Chemical yield e.e.b of dimer/% % ~~

-40

-40 rtc rt rt

~

63 45 20 21 32

(+I 92 (-1 95 (-1 95 (-1 90 (+I 94

“Data from Hasegawa et al. (1992). *Determined by hplc on an optically active phase. ‘rt = room temperature.

plained by the formation of a chiral arrangement of achiral 5:OEt molecules (space group :P212121),followed by a subsequent topochemical photoreaction (Scheme 11) (Chung and Hasegawa, 1991). In the crystal of 5:OEt, there are two crystallographically independent molecules; these form a molecular pair in a @-typearrangement in which no stack for polymerization exists. The intermolecular distances of the two facing ethylenic double bonds in each molecular pair are approximately within the normal photoreactive distances (3.802 and 4.387 A for one pair of facing bonds, and 3.829 and 4.123 A for the other). In addition, since every molecular pair is related only by a 21 screw axis, enantiomerically homogeneous cyclobutanes arise from the topochemical photoreaction of one single crystal. As expected, a large quantity of 5:OEt crystals having the same chirality is afforded by seeding with a fine crystalline powder, prepared from a single crystal of 5:OEt, during recrystallization from ethanol. Furthermore, surprisingly, it is observed that the crystals of 5:OEt obtained from each recrystallization batch without seeding always give by chance one or other enantiomeric dimer in large excess (Table 4) (Hasegawa, 1992). Such growth of chiral crystals without seeding is regarded as “amplification of asymmetry”. That is, the seeding crystals .only cause crystal formation having the same chirality, but the amplification of asymmetry always occurs without seeding with an equal probability. This phenomenon may be a key stage in the generation of molecules having a single chirality in large excess if the assumption is made that the diffusion of chiral molecules in nature is much faster than the next occurrence of amplification of asymmetry. Thus the entire topochemical process, including the crystallization process, could be a model for the generation of chiral homogeneity in the prebiotic era in nature.

156

M. HASEGAWA

PREPARATION OF CYCLOPHANES

In the plane-to-plane stack of diolefin crystals with P-type packing, two facing molecules are sometimes oriented identically with respect to neighbouring upper and lower molecules. On the other hand, the two molecules sometimes make one molecular pair separated relatively far from neighbouring pairs. Moreover, in very rare cases, molecules do not make successive plane-to-plane stacks, but two molecules make a molecular pair independently, e.g. a molecular arrangement of ethyl and/or propyl a-cyano-4-[2-(4pyridyl)ethenyl]cinnamate(2:OEt and/or 2:OPr) crystals. In conclusion, as shown in Scheme 12, a P-type crystal could afford a cyclophane derivative consisting of two monomer molecules and/or a zigzag-type polymer through the V-shaped dimer molecule. So far several types of cyclophanes have been prepared from diolefin crystals with P-type packing whereas only a few well-defined high-molecular-weight polymers have been reported. This is probably because the topochemical reaction of these crystals is often accompanied by a thermal process under moderate control by the reacting crystal lattice, which may cause deterioration of the topochemical environment to a certain extent and which makes photoproducts amorphous (see pp. 140 and 142).

Fig. 12 The crystal structure of 1,CDCB projected on to the (100) plane. (From Hasegawa et al., 1985, with permission.)

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

157

0

0 1,4-DCB

./

\hv.

n

-

Scheme 12

In the crystal of 1,Cdicinnamoylbenzene (1,CDCB) (see Fig. 12), the distances between the intermolecular photoadductive carbons are 3.973 and 4.086 A for one cyclobutane ring, and 3.903 and 3.955 8, for the other. The two topochemical pathways may occur competitively in a single crystal of 1,CDCB at the initial stage of reaction. Then, both intramolecular photodimerization and intermolecular photopolymerization of the diolefinic mono-cyclobutane intermediate occur competitively to give tricyclic dimer 21,22,23,24-tetraphenyl-1,4,11,14-tetraoxo-2(13),12( 13-diethanol, [4.4] paracyclophane or oligomers (Hasegawa el al., (1985). On photoirridation at room temperature the 1,CDCB crystal gives >90% of the tricylic

158

M. HASEGAWA

cyclophane (isolated in 58% yield after recrystallization) and a small amount of oligomers (Scheme 12). The formation of a &type cyclobutane ring was predicted from the crystal structure of 1,6DCB and confirmed by 'H nmr analysis. The total conversion and the photoproduct ratio of the cyclophane to the oligomers varies considerably with the reaction temperature, suggesting that each elementary process in the 1,CDCB crystal is influenced to a different degree by thermal vibration of the crystal lattice. At -17°C the reaction proceeds at a moderate rate. On the other hand, it almost ceases after irradiation for 28 h, to give yields of about 70% of cyclophane, about 10% of oligomers, and about 20% of l,CDCB, respectively. On further irradiation of the above as-prepared photoproducts at 28°C for 30 h, the reaction continues to give the final photoproducts containing about 90% of cyclophane and about 10% of oligomers with a trace amount of 1,4-DCB. In contrast with a typical topochemical photodimerization, such a striking decrease of reactivity of 1,4-DCB at a moderately depressed temperature suggests that more vigorous thermal movement is required for the photoexcited species to react at a later stage. Assuming that the reaction probability of all the elementary processes is equal in the reaction of 1,4-DCB crystals, the calculated yields of unreacted 1,6DCB, cyclophane, and oligomer by simulation, should be 1.8, 37.7, and 60.5% by weight, respectively. Furthermore, if all the photoexcited species of the monocyclic dimer are assumed to be converted into cyclophane, these yields should become 6.9, 65.6 and 27.5%. It is, therefore, rather surprising that in an extreme case of the experiment the yield of cyclophane is more than 90% while the amount of unreacted 1,CDCB is less than 2%. One plausible mechanism to explain this result is that the first formation of cyclophane induces the successive formation of cyclophane so as to enhance its final yield. If such an induction mechanism plays an appreciable role, an optically active cyclophane zone may be formed, at least in a micro spot surrounding the first molecule of cyclophane, as illustrated in Scheme 13. The assumption of an induction mechanism was verified later in the photoreaction of 7:OMe crystals (see p. 151). The crystal of 2:OPr recrystallized from EtOH/H20 solution, and the mixed crystal of the same ethyl and propyl cinnamate derivatives (2:OEt and 2:OPr), on photoirradiation for 2 h at room temperature with a 500W super-high-pressure Hg lamp, afforded the highly strained tricyclic [2.2] paracyclophane (2:OEt.2:OPr-cyclo) crystal quantitatively (Maekawa et al., 1991b). A crystal structure analysis was carried out of a single crystal of the complex of 2:OEt.2:OPr-cyclo with HFIP (recrystallization solvent) in a 1:2 molar ratio. Fig. 13 shows the molecular structure of 2:OEt-2:OPr-cyclo viewed along the phenylene planes. The short non-bonded distances and deformation of the benzene rings, as seen in Fig. 13, are common to those of [2.2] paracyclophanes, as previously reported (Hope et al., 1972a,b).

PHOTODIMERlZATlON AND PHOTOPOLYMERlZATlON

Random mechanism

159

Induction mechanism Scheme 13

It is of great interest that the topochemical driving force is sufficient to give such a highly strained molecule 2:OEt.2:OPr-cyclo quantitatively, even though a great amount of internal energy should be qxhausted in twisting two phenylene rings during its formation from 2:OEt.2: OPr-dimer. In the mixed crystal, the reactive molecules, which are related by a pseudocentre of symmetry, make a pair and are superimposed along the [ O l l ] direction without any displacement of the molecular long axis. The double bond on the pyridyl side in one molecule and the double bond on the

160

M. HASEGAWA

Fig. 13 Molecular structure of 2:OEt-2:OPr-cyclo viewed along the phenylene planes. Both of the ester alkyl moieties are depicted by a propyl group.

A

B

C

D Fig. 14 Crystal structure of 2:OEt.2:OPr. Two molecules making a pair (B and C) and two neighbouring molecules in the other pairs (A and D) viewed along the phenylene groups in molecules B and C. All the ester alkyl moieties are depicted by an ethyl group.

PHOTODIMER IZATION AND PHOTOPOLYMERlZATlON

glide plane

1 1

\

\ (a)

161

\ (b)

Q

view

Fig. 15 Schematic molecular arrangements of (a) a normal p-type crystal and (b) 2:OEt.2:OPr crystal viewed along the molecular long axis.

ester side in the other molecule are within the distance topochemically allowed for [2+2] photocycloaddition. The quantitative formation of the highly strained 2:OEt.2:OPr-cyclo, on photoirradiation even at ordinary temperatures, is reasonably interpreted in terms of the crystal structure of 2:OEt*2:OPr. Fig. 14 shows the molecular arrangement of the reacting paired molecules (B and C) as well as the molecules in neighbouring pairs (A and D). As is obvious in the diagram, the irradiation of 2:OEt*2:OPr should give a P-type dimer having a hetero-type cyclobutane ring. On the other hand, the molecular arrangement in the crystal is unusual in that molecules A and B (also C and D) come into contact with each other at the van der Waals distances. Therefore, the modification of the molecular conformation into the Vshaped monocyclic dimer does not occur in a reacting crystal of 2:OEt*2:OPr even at ordinary temperature. Thus, after the first cycloaddition between two molecules in one pair, the residual olefins in the monocyclic dimer 2:OEt-2:OPr-dimer would be forced to stay within the reactive distance by virtue of repulsion with the neighbouring molecules. Figs 15(a) and (b) show two typical molecular arrangements in a crystal with P-type packing viewed along the direction of the molecular long axis. Each line in Fig. 15 depicts the average plane of a diolefin compound. In the usual cases of photoreactive diolefin compounds with P-type packing reported so far, the aromatic planes of the neighbouring molecules on both sides are parallel and superimposed to make a plane-to-plane stack [Fig. 15(a)]. In contrast, in the case of 2:OEt-2:OPr, which corresponds to the crystal type in Fig. 15(b), a pair of reacting molecules is isolated from the neighbouring pairs by a glide plane. Thus, the residual double bonds in the resulting monocylcic dimer 2:OEt.2:OPr-dimer could not be parallel to the double bonds in any neighbouring pair. Consequently, a second

M. HASEGAWA

162

-

rn

hv(>300 nm)

in crystalline state

X

X

X = CI 2.3-DSP:CI X = CN 2.3-DSP:CN

X = CI 2.3-DSP:CI-dimer (76 YO yield) X = CN 2.3-DSP:CN-dimer (58 % yield)

Scheme 14

cycloaddition should occur only between intramolecular double bonds to give 2:OEt.2:OPr-cyclo and not between intermolecular olefins. The photoreaction of 2,3-distyrylpyrazine (2,3-DSP) derivatives has been investigated in the crystalline state (Tsutsumi et al., 1991; Takeuchi et al., 1993). On photoirradiation, 2,3-di(4-chlorostyryI)pyrazine and 2,3-di(4cyanostyry1)pyrazine crystals give eight-membered [2.2]orthocyclophanes in 76 and 56% yields, respectively (Scheme 14). On the other hand, 2,3-DSP and 4-nitro or 4-methyl derivatives of 2,3-DSP are photostable in the crystalline state. These results are reasonably interpreted on the basis of X-ray crystallographic analysis of these crystals. 5 Topochemical reactions of mixed crystals, inclusion complexes and molecular complexes

It is of great interest to extend the study of four-centre-type photopolymerization into the area of copolymerization. Although there have been a few preliminary reports on the formation of mixed diolefin crystals, none of them described topochemical copolymerization behaviour (Nakanishi et al., 1979b; Addadi et al., 1982). Two other papers had described topochemical copolymerization; one was related to diacetylene (Enklemann, 1984 and references cited therein) and the other to diolefin derivatives (Hasegawa et al., 1989~).However, well-defined copolymers had not been reported before the four-centre-type photocopolymerization of a mixed crystal consisting of two alkyl 4-[2-(2-pyrazyl)ethenyl]cinnamate derivatives (Maekawa et al., 1991~).Rather more recently several diolefin mixed crystals, inclusion complexes which include solvent molecules or molecular complexes of two different diolefin species in separate layers, have been studied. These mixed crystals and molecular complexes are formed not only by cocrystallization from a solution of two diolefin compounds, but may be generally formed even by simple grinding of two crystals. Mechanistic interpretations have been proposed to account for such complex formation by simple grinding (Kinbara et al., 1993).

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

163

ENHANCEMENT OF PHOTOPOLYMERIZABILITY BY COMPLEX FORMATION

A 2:OPr crystal is quantitatively converted into the corresponding a-type dimer crystal by intermolecular [2+2] photoreaction between double bonds on the ester side with the retention of its space group (Pi) (Hasegawa et al., 1988b). The as-prepared dimer crystal is photostable since the intermolecular distance between the residual double bonds on the pyridyl side is 5.066& although they are parallel. However, the dimer forms a highly photoreactive inclusion complex with the solvent when recrystallized from alcoholic solutions, such as in ethanol. For example, on photoirradiation, = 3000) as the complex, 2:OPr-a-dimer-PrOH gives a linear polymer [(Rn already illustrated in Scheme 61 (Hasegawa, 1992). Molecular arrangements in the as-prepared dimer and 2:OPr-adimer.PrOH crystals are shown schematically in Figs 16(a) and (b), respectively. In the complex, propanol molecules are hydrogen-bonded to the pyridyl groups and, as expected from its high photoreactivity, the

5.066 A

I I

I

I

I I

I(

I

>I

4L = 23.727A

8

I<

4L = 26.945 A

Fig. 16 Molecular arrangements of 2:OPr-a-dimer in the crystals of (a) 2:OPr-adimer, as-prepared and (b) 2:OPr-a-dimer-PrOH, viewed perpendicular to the direction of the polymer chain with the length of two repeating units of the dimer.

164

M. HASEGAWA

intermolecular distance between residual double bonds in the complex becomes shorter from 5.04 8, in the original dimer to 3.835 A on one side and 4.208 8, on the other side in the complex. Moreover, since the repeating unit in a-type linear polymers is about 7.5 A, according to a molecular model evaluation, the stacks in the starting crystals prefer to have repeating units of a similar length ( L ) . However, L (5.932 A) in 2:OPr-a-dimer crystals is apparently shorter than the favourable distance (-7.5A). On the other hand, in the 2:OPr-a-dimer.PrOH complex, reactant dimer molecules are separated from each other by a more suitable distance ( L = 6.736 A). For example, the same dimer complex which contains two molar equivalents of ethanol, underwent photoreaction to give a higher molecular weight polymer (M,= 12000). The formation of such inclusion complexes with the solvent is rather generally observed with similar types of dimers formed with alcohols and some other solvents and, consequently, this enhances photopolymerizability. Such complex formation with a solvent may be one of the promising techniques that can be used for diolefin compounds in order to obtain polymers with high molecular weights. TOPOCHEMICAL COPOLYMERIZATION OF DIOLEFIN MIXED CRYSTALS

The preparation of mixed crystals of olefinic compounds has been investigated from the viewpoint of topochemical copolymerization. It is well known that some pairs of isomorphous compounds form mixed crystals (Kitaigorodsky, 1984) and that most a-type diolefin crystals photopolymerize topochemically into crystalline linear polymers (Hasegawa, 1983). On the basis of these facts, a pair of two diolefin compounds, which are isomorphous with each other in a-type packing, should form a mixed crystal with high probability and, subsequently, give a crystalline linear copolymer. Though one report on the formation of a mixed crystal of 2,5-DSP and P2VB has appeared, its topochemical behaviour has not been mentioned (Nakanishi et al., 1979b). When a mixture of equimolar amounts of 1:OEt and the corresponding S-ethyl derivative (1:SEt) is recrystallized from methanol solution, microcrystals comprising 1:OEt and 1:SEt in a molar ratio of 4555 are deposited. The mixed crystal has a structure isomorphous to those of pure crystals (Hasegawa et al., 1989a; Sukegawa, 1991), and upon irradiation for 24 h at room temperature with a 500 W super-high-pressure Hg lamp, the mixed crystal gives a crystalline polymer (qinh= 1.9 dl g-') having the same main chain as the homopolymers (see Scheme 15) (Maekawa et al., 1991~). A dimer portion can be prepared exclusively from the mixed crystal by controlling the wavelength of irradiating light (>410 nm) and isolated by preparative thin-layer chromatography. The high-performance liquid chro-

Ti

t, z

p" Zf

+ I

..

7

h

hl

Y

=

A

A

7

9

0

u-

166

M. HASEGAWA

matogram (hplc) of the isolated dimers shows three peaks; two of these are attributed to dimers from 1:OEt (00)and from 1:SEt (SS), and the other one is a mixture of two mixed dimers, 0s and SO. The existence of 0s and SO dimers indicates the formation of a mixed crystal, resulting in the formation of a copolymer of the 0-ethyl and the S-ethyl compounds. Furthermore, the existence of 00 and SS dimers indicates that the molecular arrangement of 1:OEt and 1:SEt is not alternating but random in the mixed crystal because the alternating arrangement of these two diolefin molecules would give only dimers 0s and SO without either 00 or SS. This is the first example of topochemical copolymerization in which the copolymer formation is completely substantiated on the basis of direct evidence from spectroscopic and preparative techniques. The same mixed crystal is also obtained through crystal-crystal contact by simple grinding of two monomers with an amalgamator. Irradiation of the resulting mixed crystal gives a crystalline linear copolymer (vinh= 0.6 dl g-l) having the same structure. Such mixed crystal formation by simple grinding has been widely observed on several pairs of diolefinic and diacetylenic compounds (Maekawa et al., 1991c; Hasegawa et al., 1W1, 1993). The formation of mixed crystals from diolefinic compounds has been ascertained by the change in the X-ray diffraction pattern upon grinding and by 'H nmr and hplc analyses of the photoproducts. According to several reports, it has been found that a mechanical treatment is necessary for the formation of the mixed crystal phase (Kinbara et al., 1993). The kinetic data concerning the analysis of the topochemical photoproducts reveal that the subsequent growth of the mixed crystal phase is triggered by each of the mixed crystal nuclei and that the growth of the mixed crystal phase is accelerated by either thermal annealing or exposure to humidity. During the early stage of mixed crystal growth, the growth rate is nearly proportional to the surface area of the nuclei of the mixed crystal. From these experimental results, it has been suggested that the whole processes of mixed crystal formation is kinetically analogous to the photographic processes; the grinding time and the subsequent thermal annealing correspond to the exposure time and development process, respectively. A PERFECTLY ORDERED POLYMER COMPOSITE BY FOUR-CENTRE-TYPE

PHOTOPOLYMERIZATION OF A MOLECULAR COMPLEX

Several pairs of isomorphous diolefin compounds have been reported to form photoreactive mixed crystals, whereas formation of the mixed crystal, comprising two olefin compounds which are not isomorphous to each other, is rare. 2,5-DSP and 1:OEt each gives a crystalline polymer upon UV irradiation in the crystalline state, respectively (Hasegawa et al., 1969, 1989a). On the

PHOTODIME R lZATl0N AN D PHOTOPOLYMER IZATlON

167

other hand, when an equimolar mixture of 2,5-DSP and 1:OEt is recrystallized from benzene, yellow crystals, comprising 2,5-DSP and 1:OEt in a molar ratio of 1:2, deposit. In the DSC curve of this crystal, a single endothermic peak is observed at 166”C, which is different from the melting point of either 2,5-DSP (223°C) or 1:OEt (156°C). Furthermore, the X-ray powder diffraction pattern of the crystal is quite different from those of the homocrystals 2,5-DSP and 1:OEt. Upon irradiation the cocrystal 2,5DSP-1:OEt affords a crystalline polymer (qinh= 1.0 dl 8-l: in trifluoroacetic acid). The ‘H nmr spectrum of the polymer coincides perfectly with that of a 1:2 mixture of poly-2,5-DSP and poly-1:OEt. In the dimer, only 2,5-DSPdimer and 1:OEt-dimer are detected by hplc analysis, but the corresponding cross-dimer consisting of 2,5-DSP and 1:OEt is not detected at all (Hasegawa et al., 1993). These observations by ‘H nmr and hplc indicate that the photoproduct obtained from the cocrystal 2,5-DSP.l:OEt is not a copolymer but a mixture of poly-2,5-DSP and poly-1:OEt in the ratio 1:2. From the result of the X-ray crystallographic analysis, in the cocrystal (2,5-DSP.l:OEt) both 2,5-DSP and 1:OEt form columns separately along the c-axis with sheets perpendicular to the a-axis related by a glide mirror plane. This result indicates that cocrystal is not a mixed crystal with a disordered molecular arrangement but a “molecular complex” with an ordered molecular arrangement. The distances between the neighbouring carbon-carbon double bonds in the same layer are 3.934 8, in the column of 2,5-DSP and 4.013 and 3.976A in the column of l:OEt, respectively, and are within the limit allowed for topochemical [2+2] photocycloaddition. Taking into account the distances between the neighbouring carbon-carbon double bonds and preservation of the crystallinity during the photoreaction, the polymerization would proceed along the a-axis upon photoirradiation to give a perfectly ordered (alternatively layered) crystalline polymer composite; in other words, the polymer crystal should comprise an alternating monolayer of poly-2,5-DSP and bilayer of poly-1:OEt, as shown schematically in Fig. 17 on p. 168. This is the first example of a topochemical reaction of a molecular complex of a perfectly ordered polymer composite. Complex 2,5-DSP.l:OEt is also obtained by simple grinding of homocrystals 2,5-DSP and l:OEt, as is observed for the pair of diolefinic compounds described on p. 166. 6 Concluding remarks

Topochemical [2+2] photoreactions of diolefin crystals”has been reviewed. The reactions clearly depart from typical solution chemistry; crystal-lattice control offers a unique synthetic route into photodegradable polymers, highly strained [2+2] paracyclophanes, stereoregular polymers, and “absolute” asymmetric synthesis. However, achieving the desired type of crystal

M.HASEGAWA

168

I

I

Fig. 17 (a) Molecular arrangement of 2,5-DSP and 1:OEt in the molecular complex (2,5-DSP.l:OEt). (b) Schematic representation of the photoproduct of the molecular complex (2,5-DSP.l:OEt). Black and white rods represent single polymer chain of poly-2,5-DSP and poly-l:OEt, respectively. Three arrows represent the direction of the a-, b-, and c-axes of the original complex before irradiation.

structure in any given case is still difficult because the factors that control the crystal packing are not yet satisfactorily understood; not only a slight change in the chemical structure of the starting compounds but, even for the same compound, a different crystallization procedure causes a drastic variation in the packing arrangement of the molecules and in the topo-

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

169

chemical photoreaction mode. In addition, discussion of non-topochemical factors has started recently in the field of topochemical photoreactions. Much progress in the field of computational methodology and crystal engineering is required for this endeavour. On the other hand, it has been recently demonstrated that the formation of mixed crystals by simple grinding of two kinds of organic crystals is a rather general phenomenon and that this technique could extend the present limited topochemical reactions into a wide variety of organic reactions such as regio- and enantio-selective reactions, optical resolution, Grignard reactions and so on. The result is a promising new synthetic technique for solvent-free organic reactions in the future.

References Addadi, L., van Mil, J. and Lahav, M. (1982). J . Am. Chem. SOC. 104, 3422 Annet, R. (1962). Can. J . Chem. 40, 1249 Bernardi, F., Olivucci, M., McDouall, J. J. W. and Robb, M. A. (1987). J . Am. Chem. SOC.109, 544 Bernstein, H. I. and Quimby, W. C. (1943). J. A m . Chem. SOC. 65, 1845 Braun, H.-G. and Wegner, G. (1983). Makromol. Chem. 184, 1103 Chung, C.-M. and Hasegawa, M. (1991). J . A m . Chem. SOC.113, 7311 Chung, C.-M., Nakamura, F., Hashimoto, Y. and Hasegawa, M. (1991a). Chem. Lett. 779 Chung, C.-M., Kunita, A., Hayashi, K., Nakamura, F., Saigo, K. and Hasegawa, M. (1991b). J. A m . Chem. SOC. 113, 7316 Cohen, M. D. and Schmidt, G. M. (1964). J. Chem. SOC. 1969 Dewar, M. J. S., Zoebisch, E. G., Healy, E. F. and Stewart, J. J. P. (1985). J . Am. Chem. SOC. 107, 3902 Enklemann, V. (1984). Polydiacetylenes Adv. Polym. Sci. 63, 91-136 Franke, R. (1905). Chem. Ber. 38, 3727 Fujishige, S. and Hasegawa, M. (1969). J . Polym. Sci., A 7 , 2319 Fukui, K., Morokuma, K., Yonezawa, T., Nagata, C. and Nagata, C. (1960). Bull. Chem. SOC. Jpn. 33, 963 Gerasimov, G. N., Mikova, 0. B., Kotin, E. B., Nekhoroshev, N. S. and Abkin, A. D. (1974). Dokl. Akad. Nauk SSSR 216, 1051 Gordan, A. J. (1967). Tetrahedron 23, 863 Hanamura, E . (1987). Solid State Commun. 63, 1097 Hasegawa, M. (1966). Ph.D. Thesis, The University of Tokyo Hasegawa, M. (1983). Chem. Rev. 83, 507 Hasegawa, M. (1986). Pure Appl. Chem. 58, 1179 Hasegawa, M. (1992). Proc. Jpn. Acad. 68, Ser. B, 9 Hasegawa, M. (1993). Reactivity in Molecular Crystals (ed. Ohashi, Y.), ch. 5, pp. 203-236. Kodansha, Tokyo Hasegawa, M. and Hashimoto, Y. (1992). Mol. Cryst. Liq. Cryst. 104, 3422 Hasegawa, M. and Shiba, S. (1982). J. Phys. Chem. 86, 1490 Hasegawa, M. and Suzuki, Y. (1967). J. Polym. Sci. B 5 , 813 Hasegawa, M., Suzuki, Y., Suzuki, F. and Nakanishi, H. (1969). J . Polym. Sci., A1 7, 743

170

M.HASEGAWA

Hasegawa, M., Suzuki, Y., Suzuki, F. and Nakanishi, H. (1972). Four-Center Type Photopolymerization in the Crystalline State, vol. 5, pp. 143-203. Kodansha, Tokyo Hasegawa, M., Nakanishi, H., and Yurugi, T. (1975). Chem. Lett. 497 Hasegawa, M., Nakanishi, H. and Yurugi, T. (1978). J. Polym. Sci., Polym. Chem. Ed. 16, 2113 Hasegawa, M., Yurugi, T., Namiki, S., Shiobara, T., Nakanishi, H. and Nakanishi, F. (1981). J. Polym. Sci., Polym. Chem. Ed. 19, 1291 Hasegawa, M., Saigo, K., Mori, T., Uno, H., Nohara, M. and Nakanishi, H. (1985). J . Am. Chem. SOC. 107, 2788 Hasegawa, M., Kato, S., Yonezawa, N. and Saigo, K. (1986a). J. Polym. Sci., Polym. Lett. Ed. 24, 153 Hasegawa, M., Harashina, H., Kato, S. and Saigo, K. (1986b). Macromolecules 19, 1276 Hasegawa, M., Katsumata, T., Ito, Y., Saigo, K. and Iitaka, Y. (1988a). Macromolecules 21, 3134 Hasegawa, M., Kato, S., Saigo, K., Wilson, S. R., Stern, C. L. and Paul, I. C. (1988b). J . Photochem. Photobiol., A , Chem. 41, 385 Hasegawa, M., Aoyama, M., Maekawa, Y. and Ohashi, Y. (1989a). Macromolecules 22. 1568 Hasegawa, M., Kunita, A . , Chung, C.-M., Hayashi, K. and Sato, S. (1989b). Chem. Lett. 641 Hasegawa, M., Endo, Y., Aoyama, M. and Saigo, K. (1989~).Bull. Chem. SOC.Jpn. 62, 1556 Hasegawa, M., Chung, C.-M., Muro, N. and Maekawa, Y. (1990). J. A m . Chem. SOC.112. 5676 Hasegawa, M., Ikeda, T. and Noguchi, H. (1991). Chem. Lett. 771 Hasegawa, M., Hashimoto, Y. and Chung, C.-H. (1992). Radiation Curing in Polymer Science and Technology (ed. Frouassier, J. P. and Rabek, J. F.) ch. 10, DD. 341-366. Elsevier Amlied Science. New York Hasegawa, M., Kinbara, K:,&Adegawa,Y. and Saigo, K. (1993). J . A m . Chem. SOC. 115, 3820 Hashimoto, Y., Adegawa, M., Hasegawa, M., Onodera, K., Sakuragi, H. and Tokumaru, T. (1991). Bull. Chem. SOC. J p n . 64, 1043 Hatada, S. (1989). Graduation thesis, Faculty of Engineering, The University of Tokyo Hirshfeld, F. L. and Schmidt, G. M. J. (1964). J. Polym. Sci., Part A A2, 2181 Holm, M. J. and Zienty, F. (1972). J. Polym. Sci., AZ 10, 1311 Hope, H., Bernstein, J. and Trueblood, K. N. (1972a). Acta Crystallogr. 1328 Hope, H., Bernstein, J. and Trueblood, K. N. (1972b). Actu Crystallogr. 1733 Kanetsuna, H., Hasegawa, M., Mitsuhashi, T., Kurita, T., Sasaki, K., Maeda, K., Obata, H. and Hatakeyama, T. (1970). J. Polym. Sci., A2 8, 1027 Kinbara, K., Adegawa, Y., Saigo, K. and Hasegawa, M. (1993). Bull. Chem. SOC. J p n . 66, 1204 Kitaigorodsky, A. I. (1984). Mixed Crystals. Springer-Verlag, Berlin Koelsch, C. F. and Gumprecht, W. H. (1958). J . Org. Chem. 23, 1603 Lahav, M. and Schmidt, G. M. J. (1966). Tetrahedron Lett. 26, 2957 Liebermann, C. (1889). Chem. Ber. 22, 124, 782 Maekawa, Y., Kato, S., Saigo, K., Hasegawa, M. and Ohashi, Y. (1991a). Macromolecules 24, 2314 Maekawa, Y., Kato, S. and Hasegawa, M. (1991b). J. A m . Chem. SOC.113, 3867 Maekawa, Y., Lim, P.-J., Saigo, K. and Hasegawa, M. (1991~).Macromolecules 24, 5752

PHOTODIMERIZATION AND PHOTOPOLYMERIZATION

171

Meyer, W., Lieser, G. and Wegner, G. (1978). J . Polym. Sci., Polym. Phys. Ed. 16, 1365 Mirsky, K. (1978). Proceedings of an International Summer School in Crystallographic Computing, Twente, p. 168 Miura, M., Kitami, T. and Nagakubo, K. (1968). Polym. Lett. 6, 463 Nakanishi, H., Hasegawa, M. and Sasada, Y. (1972a). J . Polym. Sci. A 2 10, 1537 Nakanishi, H., Ueno, K., Hasegawa, M. and Sasada, Y. (1972b). Chem. Lett. 301 Nakanishi, H., Nakanishi, F., Suzuki, Y. and Hasegawa, M. (1973). J. Polym. Sci., Polym. Chem. Ed. 11, 2501 Nakanishi, F., Nakanishi, H. and Hasegawa, M. (1974). Seni Kobunshi Zairyo Kenkyusho, Kenkyu-Hokoku 103, 49 Nakanishi, H., Ueno, K. and Sasada, Y. (1976). Acta Crystallogr. B32, 3352 Nakanishi, H . , Hasegawa, M. and Sasada, Y. (1977). J . Polym. Sci., Polym. Phys. Ed. 15, 173 Nakanishi, H., Sasada, Y. and Hasegawa, M. (1979a). Polym. Lett. 17, 459 Nakanishi, H., Jones, W. and Parkinson, G. M. (1979b). Actu Crystallogr. B35, 3103 Nakanishi, H., Jones, W., Thomas, J. M., Hasegawa, M. and Rees, W. L. (1980). Proc. R. SOC.London, Ser. A 369, 307 Ohashi, Y. (1980). CAVITY program, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo, Japan Okarnura, S., Hayashi, K. and Kitanishi, Y. (1962). J. Polym. Sci. 58, 925 Peachey, N. M. and Eckhardt, C. J. (1993). J. Am. Chem. SOC.115, 3519 Pincock, R. E., Wilson, K. R. and Kiovsky, T. E. (1967). J. Am. Chem. SOC.89, 6890 Ramamurthy, V. and Venkatesan, K. (1987). Chem. Rev. 87, 433 Sasada, Y., Shimanouchi, H., Nakoanishi, H. and Hasegawa, M. (1971). Bull. Chem. SOC.Jpn. 44,1262 Schmidt, G. M. J. (1971). Pure Appl. Chem. 27, 647 Stezowski, J. J., Peachey, N. M., Goebel, P. and Eckhardt, C. J. (1993). J . Am. Chem. SOC.115, 6499 Sukegawa, M. (1991). MS Thesis, Faculty of Engineering, The University of Tokyo Sukenik, C. N., Bonapace, J. A. P., Mandel, N. S., Bergman, R. G., Lau, P.-Y. and Wood, G. (1975). J. Am. Chem. SOC.97, 5290 Sukenik, C. N., Bonapace, J. A. P., Mandel, N. S., Lau, P.-Y., Wood, G. and Bergman, R. G. (1977). J. Am. Chem. SOC.99, 851 Suzuki, F., Suzuki, Y. and Hasegawa, M. (1965). Seni Kogyo Shikensho Kenkyu Hokoku 72, 11 Takeuchi, A., Korniya, H., Tsutsumi, T., Hashimoto, Y., Hasegawa, M., Iitaka, Y. and Saigo, K. (1993). Bull. Chem. SOC.Jpn. 66, 2987 Tamaki, T., Suzuki, Y. and Hasegawa, M. (1972). Bull. Chem. SOC.Jpn. 45, 1988 Thomas, J. M., Morsi, S. E. and Desvergne, J. P. (1977). Adv. Phys. Org. Chem. 15, 63 Tsutsumi, T . , Takeuchi, A., Hashimoto, Y., Hasegawa, M. and Iitaka, Y. (1991). Chem. Lett. 1533 Wegner, G. (1972). Polym. Preprints 13, 470 Wegner, G. (1973). Adv. Chem. Ser. 129, 255 Wilson, K. R. and Pincock, R. E. (1975). J. A m . Chem. SOC.97, 1474 Yates, P. and Singh, P. (1969). J . Org. Chem. 34, 4052

This Page Intentionally Left Blank

Ionic Dissociation of Carbon-Carbon u Bonds in Hydrocarbons and the Formation of Authentic Hydrocarbon Salts KUNIOOKAMOTO, KEN’ICHITAKEUCHI AND TOSHIKAZU KITAGAWA Division of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Kyoto, Japan

Introduction Bonding in hydrocarbons Hydrocarbon salts and the related hydrocarbons Stability of hydrocarbon ions Inorganic salts of stable hydrocarbon ions Some stability scales for hydrocarbon ions Stable all-hydrocarbon cations Stable all-hydrocarbon anions 3 Ionic dissociation of the carbon-carbon CT bond in hydrocarbons Radical dissociation Ionic dissociation and cleavage of organic molecules Ionic dissociation of genuine hydrocarbons Free energies of ionic dissociation Steric assistance of ionic dissociation Isolation of hydrocarbon salts Formation of carbocation-carbanion salts Isolation of authentic hydrocarbon salts Physical properties of hydrocarbon salts Some physical properties Electrical conductivity in DMSO Chemical behaviour of hydrocarbon salts in solution Hydrocarbon salts in chloroform Hydrocarbons which exist solely in solution A solution in which ionic, radical and covalent species coexist in equilibria Control of hydrocarbon salt formation Change from a covalent compound to a salt by a slight difference in carbocation stability Difference in electron affinity of the carbocation Conclusions Acknowledgements References

174 174 175 176 176 178 179 181 184 184 186 192 194 198 200 200 202 204 204 204 206 206 209 209 212 212 214 216 217 217

173 ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 30 ISBN 0-12-033530-1

Copyright 0 1995 Academic Press Limited All rights of reproduction in any form reserved

174

K. OKAMOTO ETAL.

1 Introduction BONDING IN HYDROCARBONS

The name “organic” is a relic of the days when many chemists believed that the origin of organic compounds must be in living organisms, and that consequently they could never be synthesized from inorganic materials. Ever since those times, chemical compounds have been divided into two classes, organic and inorganic, regardless of their origin. Thus, the distinction between organic and inorganic compounds has been realized on the structural basis that organic compounds always contain carbon, a fact stated early in 1848 by L. Gmelin. Carbon compounds are most commonly constructed using covalent bonds, and the nature of the bond is of prime importance in all organic compounds. Among the compounds of carbon, hydrocarbons, which are made up of only hydrogen and carbon, are the simplest known group of organic compounds having exclusively covalent bonds.

Such recognition of bonding in hydrocarbons has had to be modified, however, because of the isolation of a salt [1+2-] consisting solely of carbon and hydrogen (Okamoto et a f . , 1985). The isolation of an authentic hydrocarbon salt has shown the existence of ionic bonding in a genuine carbon-hydrogen system. Thereafter, the structural principle, that the entire hydrocarbon system is constructed exclusively of covalent bonds, has been demolished, and as a natural consequence the usual bonding feature, i.e. covalent and ionic bonds respectively for organic and inorganic compounds, has also been eroded.

IONIC DISSOCIATION OF C-C u BONDS IN HYDROCARBONS

175

HYDROCARBON SALTS AND THE RELATED HYDROCARBONS

The synthetic procedures for isolation of the salt appear to be rather simple. First, one prepares a solution in which the carbocation and carbanion coexist free from any combination reactions. Then, the hydrocarbon cation-anion salt is isolated after separation of the concomitant inorganic salt and evaporation of the solvent. For the purification of the crude salt recrystallization or reprecipitation with proper solvents is used. Sufficient stability of the hydrocarbon ions, as the salt or in the solution, is an obvious prerequisite for these procedures, and, in practice, selecting or designing the stable ions and choosing a proper solvent are tasks of primary importance. As an ordinary stability index for the ions, thermodynamic scales referred to the water molecule, i.e. pKR+and pK, values, are chosen for the carbocation and carbanion, respectively. In the course of the salt synthesis, it was found that a hydrocarbon t3-21, which was formed by an unfavourable cation-anion combination reaction, dissociates into the original carbocation and carbanion in a polar aprotic solvent (Okamoto et af., 1985) (1). This was the first example of ionic dissociation of the carbon-carbon u bond in genuine hydrocarbons, although a few cases of heterolytic dissociation of carbon-carbon u bonds had been reported by Arnett (Arnett et af., 1983; Troughton et af., 1984; Arnett and Molter, 1985) for compounds bearing cyano and nitro groups, e.g. [4-6] and [5-6] as in (2). This chapter focuses on authentic hydrocarbon systems. First, the

DMSO 7

r2-I

[3-21

R

CN

-

NC

MeCN

f--

CN

R [4-61 R=Me [5-61 R=Ph

R [4+] R=Me [5+] R=Ph

NC

16.1

K. OKAMOTO ET AL.

176

structural factors necessary for carbocation and’ carbanion stability are discussed briefly. The mechanistic problems of dissociation of the carboncarbon (T bond are then discussed, and finally the physicochemical nature of hydrocarbon salts and the related novel hydrocarbons. 2 Stability of hydrocarbon ions

INORGANIC SALTS OF STABLE HYDROCARBON IONS

The stability of hydrocarbon ions is discovered intuitively by observing whether the hydrocarbon ion can be isolated as a salt, for example, a sodium salt of the carbanion or a tetrafluoroborate salt of the carbocation. Conversely, a single hydrocarbon ion produced in the gas phase is obviously an unstable and short-lived species. Thus, many of the aliphatic carbocations in the gas phase are merely observable species but are not usable for synthesis. The free t-butyl cation [7+] in the gas phase is nothing more than a species detectable by the electron impact method (Ye0 and Williams, 1970). However, it is not only an observable species by nmr studies in SbF5/FS03H (Olah et al., 1964), but can be isolated from the solution in the form of its SbF; or SbzFfi salt (Olah and Lukas, 1967a,b; Olah et al., 1973; Yannoni et al., 1989). The crystal structure shows that this ion is planar and its carbon-carbon bonds are shortened to 144.2 pm (Hollenstein and Laube, 1993). Its particular electronic stabilization among aliphatic carbocations is attributed by physical organic chemists to the operation of both inductive and hyperconjugative effects in the (T bond system. As another example, the tropylium ion [3+], which is stabilized by virtue of the 67r electrons spread over a heptagonal sp2 hybridized carbon framework [Huckel’s (4n + 2)7r rule with n = 11, is also unstable in the gas phase. Its formation from toluene or the benzyl cation has been a long-standing problem in organic mass spectrometry, and the reaction mechanism and energetics have recently been exhaustively discussed (Lifshitz, 1994). It was, however, isolated as the bromide salt by Doering and Knox (1954, 1957), and was the first “non-benzenoid” aromatic carbocation. Me

I Me-C’ I

r7+1

Me Me-C’

I I

SbF6-

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

177

[3+] Br-

r3+1

Thereafter the interest of many synthetic and physical organic chemists was aroused in a search for a “novel aromatic system” in the field of hydrocarbon chemistry. Thus, putting aside the detection or observation of hydrocarbon ions in the gas phase, synthetic organic chemists have been trapping hydrocarbon ions as the stable salts of inorganic gegenions; anions of strong inorganic Brmsted acids or alkali-metal ions have been chosen for the stabilization. Three significant hydrocarbon ions, [8+], [9-] and [lo+], should be mentioned as additional examples of the formation of stable salts of the hydrocarbon ions. The triphenylmethylium ion [8+], as a prototypical aromatic carbocation, reacts with the chloride ion to give the chloride, a

[8+]C104-

K+ [9-1

[lo+]SbF6-

covalent compound. However, it can exist as the perchlorate salt, tetrafluoroborate salt, and so on. Notably, the covalent chloride can dissociate in an aprotic solvent such as sulfur dioxide to give the cation [8+] and the chloride ion. The cyclopentadienide ion [9-] can be isolated as the potassium salt. It is a classical carbanion, the stability of which can be understood in terms of the operation of Huckel’s (4n+2)7r rule (n = 1). Thus, the Huckel (4n + 2 )system, ~ as in [3+] and [9-], is one of the most significant structural features in previously isolated stable hydrocarbon ions. A bridged carbocation with a two-electron, three-centre bond was proposed as early as 1939 (Nevell et al., 1939) for the 2-norbornyl cation [lo+] as a reactive intermediate in the solvolysis of 2-norbornyl system (see also Winstein and Trifan, 1949). It has now been isolated as the SbF6- salt and the bridged structure is accounted for using solid-state 13C nmr studies

K. OKAMOTO ET AL.

178

of the salt crystals in the range 5-200 K (Yannoni et al., 1982a,b; Myhre et al., 1990).

SOME STABILITY SCALES FOR HYDROCARBON IONS

When physical organic chemists discuss the stability of a hydrocarbon ion, they use a stability scale, but they can define the scale only in relative terms. Thus, it is necessary to choose a reference species from which the ion can be derived or to which it can revert (Bethel1 and Gold, 1967). When one chooses a radical as the reference system, the stability of a carbocation or carbanion can be defined by the free-energy change for discharging the ion in vacuum, and the change can be approximately described by the classical Born equation (3) (Born, 1920), provided that the ion is represented by a conducting sphere on which the charge is located. AG

= - NZ2e2/2r

(3)

In (3) N is Avogadro’s constant, r the radius of a conducting sphere, Z the charge number, and e the elementary electric charge. Equation (3) indicates that the electrostatic potential of an ionic sphere is lowered as the radius of the sphere increases. Consequently, the ion in a vacuum turns out to be more stable when there is wider delocalization of the charge. Thus, as far as the electronic stability is concerned, a clue for the synthetic problem of how to make a stable carbocation or carbanion should be to design a system in which the charge does not localize on a single carbon atom but is dispersed over the whole carbon system. Such delocalization of the charge results from the inductive or hyperconjugative shift of the v electrons, as mentioned for the t-butyl cation [7+], and from the conjugative shift of the T electrons, as illustrated for the ions [3+], [8+] and [9-], depending on the structural feature of such hydrocarbon ions. The oxidation potential of carbanions, E,,, or the reduction potential of carbocations, Ered,could be a practical scale of stability as defined by (3). These potentials can be measured by voltammetry, although the scale is subject to assumptions regarding elimination of the diffusional potential and solvation effects. As thermodynamic stability indexes for the hydrocarbon ions, pKR+ and pK, values [(4) and (5)] have been widely applied for the carbocation and carbanion, respectively, in solution. Here KR+ stands for the equilibrium constant for the reaction (6) of a carbocation and a water molecule; K, stands for the equilibrium constant for the reaction (7) of a hydrocarbon with a water molecule to give the conjugate carbanion. The equilibrium constants are given by (8) and (9) for dilute aqueous solutions. Obviously, the reference system for the pKR+ scale is the corresponding alcohol, and

179

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

for pKa scale is the conjugate hydrocarbon. It is pertinent to mention that the pK scales are defined for the hydrocarbon ions in an aqueous solution.

KR+ = [ROH] [H30+]/[R+] K , = [R-] [H30f]/[RH]

The larger pKR+ index indicates a smaller KR+ value and in turn a higher stability of the carbocation; analogously, the smaller pKa index denotes a larger Ka value and a higher stability of the carbanion.

STABLE ALL-HYDROCARBON CATIONS

Besides the carbocations mentioned above, numerous highly stable carbocations have been isolated as the salts of inorganic anions. Azulene analogues of triphenylmethylium ion [11+]-[20+], [212+], [222'] and [233+]; trisubstituted cyclopropenylium ions [1+] and [24+]-[26+] ; cyclopropyl-substituted tropylium ions [27+]-[30+]; tropylium ions annelated with bicyclic frameworks [31+]-[36+] and [372+]-[392+]; fulvene-substituted cyclopropenylium [40+] and tropylium [41+] ions; a tropylium ion condensed with ~ [43+] have acenaphthylene [42+]; and a cation containing a 1 4 periphery

R / /

R3

[ll'] [12+] [l3+] [14+] [15+] [16+]

R' H H H H Me H

R2 But H But Me H H

\

R3 But But H H H H

[17+] R=Me [18+] R=H

\ R

' ',

\

[19+] R=Me [20+] R=H

K. OKAMOTO ETAL.

180

[212+]

p i + ] R=H [32+] R=Me [33+] R=Ph

[222+]

w3+1

IONIC DISSOCIATION OF C-C

(T

BONDS IN HYDROCARBONS

[36+1

[382']

181

[372']

[3g2']

all been synthesized. These cations show pKR+values higher than 7. The pKR+values are listed along with their reduction potentials, Ered,in Table 1.

STABLE ALL-HYDROCARBON ANIONS

Kuhn's carbanion, the all-hydrocarbon anion tris(7H-dibenzo[c,g]fluorenylidenemethy1)methanide ion [2-] (Kuhn and Rewicki, 1967a,b), is a stabilized system with T electrons widely spread over the sp2 hybridized carbon framework, and was isolated as the potassium salt. It also appears in DMSO solution by dissolving the parent hydrocarbon, resulting in a deep green colour. The pK, value of the precursor hydrocarbon [2]-H is 5.9 in aqueous HCl-DMSO (Kuhn and Rewicki, 1967a,b), and its enormous stability, as compared with cyclopentadiene [9]-H, pK, 18 in DMSO

K. OKAMOTO ETAL.

182

Table 1 Some stable all-hydrocarbon cations with pKR+ values 2 7. Carbocation salt

PKR+~

[1+]clo; [11+]PFg [12+]PFc [13+]PF; [14+]PFc [ 15+]PF; [16+]PF; [17+]PFg [18+]PF, [19+]PF, [2O+]PF; [212+]2PFc [222+]2PFg [233+]3PFg [24+]BF;

>10 14.3 13.7 13.2 11.4 13.4 11.3 10.8 10.5 8.7 7.3 11.9 ll.2f 9. lg 10.0

[25+]BF; [26+]ClO, [27+]ClO, [28+]BF; [2 9 7 BF, [30+]BF, [31'IS bFg [3 2 7SbF; [33+]SbF; [34+]SbFg [35+]BF; [36+]ClO; [372']2SbF, [382+]2SbFc [392+]2SbF; [40+]C10, [41+]ClO, [42+]BF; [43+]BF;

7.09 7.29 9.10 8.72 7.56 7.63 13.0 12.4 12.0 7.80' 7.80' 8.67 7.6 11.9 l0.d 8.9h 7.49" 8.7" 28.4"

EredNb - 1.42'

-0.91d -0.85d -0.83d -0.82d -0.75" -0.78d -0.70d -0.66d -0.62" -0.58" -0.59' -0.26" -0.71" -2.20h -1.525' -1.583 -0.938h -0.856h -0.760h -0.76oh -1.120 -1.094 -1.15 -0.762' -0.765' -0.405k -0.52' -1.01 -1.171 -1.48h

Reference Agranat and Aharon-Shalom (1976) Ito et al. (1994a) Ito et al. (1994a) Ito et al. (1994a) Ito et al. (1991) Ito et al. (1994a) Ito et al. (1991) Ito et al. (1991) Ito et al. (1991) Ito et al. (1991) Ito et al. (1991) Ito et al. (1992) Ito et al. (1992) Ito et al. (1994b) Komatsu et al. (1980); Moss and Munjal (1980); Moss et al. (1986) Moss et al. (1986) Takeuchi et al. (1993) Komatsu et al. (1982b) Komatsu et al. (1982b) Komatsu et al. (1982b) Komatsu et al. (1982b) Komatsu et al. (1988, 1991) Komatsu et al. (1988, 1991) Komatsu et al. (1994) Thummel and Chayangkoon (1983) Nakazawa et al. (1980) Komatsu et al. (1979) Kagayama et al. (1994) Komatsu et al. (1994) Komatsu et al. (1994) Komatsu et al. (1977) Fleming (1973); Kayama et al. (1973) Yamamoto and Murata (1976) Murata et al. (1974)

"In 50% aqueous acetonitrile. bReduction peak potential versus Ag/Ag+ measured by cyclic voltammetry in acetonitrile. 'Okamoto et al. (1990). dIto et al. (1995). 'Prof. T. Asao, private communication. 'pKR+ value for the dication. gpKR+value for the trication. hOkamoto et al. (1983). 'Takeuchi et al. (1993). 'Komatsu et al. (1987). *In dichloromethane. 'Komatsu (1993). "In 10% aqueous ethanol. "In 20% aqueous acetonitrile.

IONIC DISSOCIATION OF C-C

(T

BONDS IN HYDROCARBONS

183

r4-1

BC\ RF+j /

H H

\

(Bordwell, 1988), contributed to the successful isolation of the first hydrocarbon salt [1+2-], when it was combined with Agranat's cation, tris(3-guaiazulenyl)cyclopropenylium ion [1+] (Agranat and AharonShalom, 1976). Kuhn's carbanion analogues, [44-], [45-] and [46-], have recently been synthesized, and the precursor hydrocarbons [44]-H, [45]-H and [46]-H dissociate into the respective anions in DMSO to show deep blue colours without any added base (Kinoshita et al., 1994). A fullerene anion, B u ' G [47-], has also been obtained as a stable carbanion (Fagan et al., 1992); its lithium salt has been isolated in the form Li+[47-].4MeCN or Li+[47-].34THF. Several stable all-hydrocarbon anions of precursor hydrocarbons with low pK, values (G7) are listed in Table 2, along with their oxidation potentials, Eox.

K. OKAMOTO ETAL.

184

Table 2 Some stable all-hydrocarbon anions derived from hydrocarbon acids with pK, values <7. R-

Reference 3.2,' 5.9d 3.3 4.7 6.3' 5.7'

-0.18 -0.20 -0.33 -0.34 -0.25

(rev)' (irrev) (rev) (irrev)' (rev)'$

Kuhn and Rewicki (1967a,b) Kinoshita et al. (1994) Kinoshita et al. (1994) Jutz (1965) Fagan et al. (1992)

"In 90% aqueous DMSO. bMeasured vs. Ag/Ag+ in acetonitrile; rev: reversible, irrev: irreversible. 'Kinoshita et al. (1994). '% DMSO-water (molar ratio 85:15). 'In DMSO. 'The potential determined vs. ferrocene/ferrocenium was converted to the value vs. Ag/Ag+ by adding +0.083 V (Komatsu et ul., 1982a).

3 Ionic dissociation of the carbon-carbon cr bond in hydrocarbons RADICAL DISSOCIATION

The

radical

dissociation

of

the

"Gomberg

dime?,

[3-(diphenyl-

methylidene)-6-(triphenylmethyl)-l,4-cyclohexadiene] [48], is familiar to organic chemists as the original observation of carbon-carbon u bond dissociation in a solution (Gomberg, 1900; Lankamp et al., 1968). The yellow colour of the triphenylmethyl radical in the benzene solution should have been an observation convincing synthetic organic chemists of the stable existence of the triphenylmethyl radical [8.].

/ \

Q

Another means of observation of such radical dissociation of hydrocarbons, to a lesser extent, is achieved in modern esr spectroscopic studies. Dissociation of bitropyl [49]into the tropyl radical [3.] was observed in m-xylene solution (11). The radical is a long-lived species but only at low concentration, 10-8-10-9M (Vincow et al., 1969). The radical dissociation (12)of 7-(triphenylmethyl)-1,3,5-cycloheptatriene[50]has been found to be more facile than that of bitropyl (Okamoto et al., 1970).

IONIC DISSOCIATION OF C-C u BONDS IN HYDROCARBONS

R [w] R = B ~ [52*] R=Ph Introduction of bulky substituents often increases the kinetic stability of the radical against dimerization. Thus, trisb-t-butylpheny1)methyl [51.] (Colle et al., 1975; Dunnebacke et al., 1989; Eisenberg et al., 1991), tris(bipheny1-4-y1)methyl [52-] (Dunnebacke et al., 1989) and heptaphenyltropyl [53.](Battiste, 1962) are reported to be fully dissociated in solution. Besides such dissociation into long-lived radicals in solution, numerous examples are known of radical cleavage in the gas phase into unstable or reactive radicals. Two factors, the strain of hydrocarbon molecules and the stability of the radicals, are suggested as the major controlling factors for radical fission (Ruchardt and Beckhaus, 1980, 1986). The cleavage of carbon-carbon u bonds is the main reaction in thermal cracking of naphtha, one of the commonest processes in the hydrocarbon conversion industry. Thus, radical dissociation or cleavage of carbon-carbon (T bonds has been quite familiar to synthetic and physical organic chemists.

186

K. OKAMOTO ETAL.

IONIC DISSOCIATION AND CLEAVAGE OF ORGANIC MOLECULES

In contrast to the abundant examples of radical cleavage, only a few proposals of ionic cleavage of carbon-carbon (+ bonds have been put forward in the long saga of mechanistic studies on heterolytic cleavage reactions. Carbocation-carbanion zwitterionic intermediates were proposed for the thermal cleavage of several cyclic compounds. In most of these reactions the ionically dissociating bond belongs to one of four strained ring systems, i.e. cyclopropane (13), cyclobutane (14), cyclobutene (15) or norbornadiene (16). The mechanism is distinguished from the formation of a diradical intermediate through homolysis in terms of solvent and substituent effects

A 0 U

+ ___.)

U

ml U

-

-

n

+

-

ril-

+

/& and reaction with ion- and/or radical-trapping reagents. For the heterolytic cleavage to prevail over the “normal” homolytic one, the zwitterion should be sufficiently stabilized by solvation and by delocalization of its positive and negative charges to electron-donating and electron-withdrawing substituents. The cyano-, carbomethoxy- and phenylsulfonyl-substituted cyclopropanes [54] (Cram and Ratajczak, 1968), [55] (Yankee and Cram, 1970a; Yankee et al., 1973a) and [56] (Yankee and Cram, 1970b,c; Howe et al., 1973; Yankee et al., 1973b) were epimerized or were solvolysed to give an open-chain methyl ether; a carbocation-carbanion zwitterion was proposed as the intermediate.

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

187

[541

OMe

OMe OMe

M e z C b g r NC CN

NCfi:N NC CN

[611

R' R2

NC NC

CN

cyc-Pr cyc-Pr Me H

cyc-Pr Me cyc-Pr cyc-Pr

R

R' R', R3, R4: CN or C02Me

[631

Thermal isomerization and solvolysis of cyclobutanes [.57]-[61], obtained from [2+2] cycloaddition of an alkyl vinyl ether or l-morpholino-2methylpropene with tetracyanoethylene or diphenylketene, has been extensively studied by Huisgen (Feiler and Huisgen, 1969; Huisgen et al., 1969, 1974a,b; Huisgen and Steiner, 1973; Karle et al., 197.5; Schug and Huisgen, 1975; Huisgen, 1977a,b) and a mechanism involving heterolytic cleavage of the cyclobutane ring to give a zwitterionic intermediate was proposed. Similar heterolytic mechanisms were proposed for the ring opening of related cyclobutanes [62] (Hoffmann and Schafer, 1972), [63] (Shimizu and Nishida, 1978; Kataoka et al., 1980) and [64] (Gotoh et al., 1986). In the thermolysis of cis-fused cyclobutenes [6.5]-[70], obtained by [2+2]

K. OKAMOTO ETAL.

188

cycloaddition of cyclic enamines or a thiophene to an acetylenedicarboxylate, possible formation of zwitterionic intermediates that are stabilized by the 1-pyrrolidinyl and alkoxycarbonyl/cyano groups attached to the positive and negative carbons, respectively, has been suggested (Hall et al., 1982; Visser et al., 1982; Reinhoudt et al., 1984).

C02Me

r651

n

m2Me C02Me

C02Me

CN

The facile thermal isomerization (17) of norbornadiene derivatives [71]-[77] to cycloheptatrienes in polar solvents has been explained in terms of the initial heterolytic cleavage of the strained C(l)-C(7) bond (Hoffmann and Hauser, 1965; Lemal et al., 1966; Hoffmann, 1971, 1985; Lustgarten and Richey, 1974; Hoffmann et al., 1986; Bleasdale and Jones, 1993). The resulting zwitterion intermediates are stabilized by the cation-stabilizing groups attached to C(7) and the cyclohexadienyl-type delocalization of the negative charge. Takahashi et al. (1970, 1980; see also Arnett and Molter, 1985) reported that a quinonoid compound [78] having a cycloheptatrienylidene and a dicyanomethylene group readily forms cyclic oligomers [79] when its solution is concentrated or cooled. This oligomerization is reversible at lower concentrations or at high temperatures, regenerating the monomer [78]. A twisted zwitterionic intermediate [80] consisting of tropylium ion and malononitrile anion moieties has been suggested.

IONIC DISSOCIATION OF C-C

w BONDS IN HYDROCARBONS

R = Ph, OBut, OMe, OCH2CH2CH2CH3

R', R ~ electron: donating groups

[711

189

~721

&

Me0

OMe

COzMe

C02Me

[751

In addition to the cyclic substrates described above, several open-chain compounds are known to undergo heterolytic cleavage of a carbon-carbon w bond. A carbanion can be a good leaving group in S,l-type solvolyses when the anion is highly stabilized by strong electron-withdrawing substituents. Mitsuhashi reported some examples of such reactions for substrates which eject an anion stabilized by cyano (and nitro) groups [81] (Mitsuhashi, 1986), [82] (Mitsuhashi and Hirota, 1990) and [83] (Hirota and Mitsuhashi,

K. OKAMOTO ET AL.

190

Me

CN

'

CN

Me

CN

Me

el Me-C-C-NOZ I t

I

CN

1831

'C02Me

C02Me

R = Ph2CH, MeOCH2, But, 1 -adamantyl, p-MeOC6H4CH2

R = MeOCH2, But, 1 -adamantyl, p-MeOC6H4CH2

WI

R = MeOCH2, p-MeOC6H&H,

[861

1851

R = Ph2CH, MeOCH2, I -adamantyl. p-MeOC&CH2 1871

1881

1990). Kinetic studies showed that the leaving ability is increased in the order C(CN)*N02 < C(CN)3 C C(CN)2N=NC6H4N02-p. Warkentin, Jones and their co-workers have shown that a stepwise ion pair mechanism prevails over concerted [1.51 sigmatropic rearrangement in the migration of an alkyl group around the five-membered ring of 3-alkyl-3H-pyrazoles [84] and [85] (Majchrzak et af., 1990; Jefferson and Warkentin, 1992, 1994a), a 4-alkyl-4H-pyrazole [86] (Jefferson and Warkentin, 1992), a 5-alkylcyclopentadiene [87] (Jefferson and Warkentin, 1994b) and a 1-alkyl indene derivative [88] (Bleasdale and Jones, 1993). These compounds give a highly stabilized cyclic 6 7 ~anion whose negative charge is further delocalized on to methoxycarbonyl group(s) attached to the fivemembered ring. In the search for the possible occurrence of an ionic, stepwise Cope rearrangement, Wigfield el af. (1972a,b, 1974) found that compound [89] rearranges rapidly and quantitatively under mild conditions to give a more stable isomer [91], which appears to be formed by a fragmentationrecombination pathway (19). The product expected from the concerted [3,3] sigmatropic rearrangement [92] was not detected. The presence of the ionic

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

191

f

f

intermediate [90] was confirmed by trapping experiments with sodium borohydride, which showed the formation of 1-(p-methoxypheny1)propene. Unlike such unstable intermediates, the first, rare example of reversible dissociation of a carbon-carbon (+ bond into a stable carbocation and carbanion was reported for a nitro-dicyano compound (20) prepared from trimethyl- and triphenyl-cyclopropenylium tetrafluoroborate ([4+]BF; and [5+]BFy) with the potassium salt of p-substituted-phenylmalononitrile anions (Arnett et al., 1983; Troughton et al., 1984; Arnett and Molter, 1985). Other ionically dissociative malononitrile derivatives have been prepared from such carbocations as the tropylium [3+] (Arnett and Troughton, 1983) and the trisb-methoxypheny1)methylium [93+] (Arnett and Troughton, 1983) ions.

R ' = NO2, CN, C1, H, Me

[4+] R=Me

[5+] R=Ph

K. OKAMOTO ETAL.

192

OMe

Me0

NC I

Qj"0 +

-&yo2 /

NC:

-

OMe

Ohle

(93.61

[93+1

IONIC DISSOCIATION OF GENUINE HYDROCARBONS

In view of the observations of the ionic dissociation of nitro-cyano compounds, it is hardly surprising that even a hydrocarbon could dissociate ionically into a stable carbocation and carbanion, provided that the medium is polar enough to prevent the recombination reaction and to ensure equilibration. The first observation of ionic dissociation of a carbon-carbon (T bond in a hydrocarbon was reported for 7-(2,4,6-cycloheptatrien-l-yl)-7-[3-(7Hdibenzo[c, g]fluoren-7-ylidene)-2-(7H-dibenzo[c, g]fluoren-7-ylidenemethyl)l-propenyl]-7H-dibenzo[c,g]fluorene[3-21. The hydrocarbon [3-21 forms a green solution in DMSO, indicating the dissociation into green Kuhn's carbanion [2-] and colourless tropylium ion [3+] as shown in (1) (Okamoto et al., 1985). Because of the dependence of the dissociation on the polarity of the solvent medium, in the less polar acetone solvent the dissolution of [3-21 does not give rise to the green colour of the Kuhn's carbanion [2-] but simply the pale yellow colour of the hydrocarbon [3-21. However, when pyrene, which forms a charge-transfer complex with the tropylium ion (Dauben and Wilson, 1968), is added to the acetone solution, it turns green, indicating that the dissociation is induced by pyrene and that the equilibrium is shifted to the ionic side (Okamoto et al., 1985). Following the first observation, eight substituted tropylium ions [94+][98+], [30+], [34+] and [35+] (Scheme 1) were combined with Kuhn's carbanion [2-] to make ionically dissociative hydrocarbons [94-21-[98-21, [30-21, [34-21 and [35-21. The structures of these hydrocarbons were determined as indicated in (23) on the basis of their 'H and 13C nmr spectra. Compounds [94-21, [95-21, [98-21 and [35-21 were mixtures of two positional isomers, in which [2-] is connected to different positions of the sevenmembered ring. Positions that formed a carbon-carbon bond with [2-] are indicated by dots in Scheme 1. Spectrophotometric studies revealed that these hydrocarbons reversibly dissociate in DMSO and in 5:95 3-methylsulfolane-sulfolane (Miyabo et al.,

IONIC DISSOCIATION OF C-C

(T

BONDS IN HYDROCARBONS

[3-21, [94-21-[98-21 [30-21, [34-21, [35-21

193

r2-I

(23)

[30+1

Scheme 1

1992a, 1993) (23). Conductometric examinations also indicated the reversible dissociation of the hydrocarbons [3-21, [96-21 and [97]2] in DMSO (Miyabo et al., 1993). A plot of the equivalent conductances against the square root of the concentrations (Onsager's plot) gave curved lines, which are typical for weak electrolytes (Fig. 1). Similarly, three dissociative hydrocarbons [25-21, [99-21 and [ 100-21, whose structures were determined from their nmr spectra, have been prepared from substituted cyclopropenylium ions: phenyl- [25+], mmethylphenyl- [99+] and m,m'-dimethylphenyl- [loo+] dicyclopropylcyclopropenylium ions (Scheme 2). These hydrocarbons reversibly dissociate (24) into their ionic moieties in DMSO (Takeuchi et al., 1993). Another example of the heterolysis of a carbon-carbon CT bond has been

K. OKAMOTO ETAL.

194

a 4 -

3 -

2 -

1 -

o ! 0

I

I

1

10

1

20

Id x c ” ~ / M ” ~ Fig. 1

plot for [3-21, [96-21 and [97-21 in DMSO at 25°C.

reported for a fullerene (C&) derivative, 1-t-butyl-4-(3,6-dicyclopropyl[30-471. Partial dissociation 2,4,6-cycloheptatrien-l’-yl)-l,4-dihydrofullerene of [30-471 into [30+] and [47-] was observed spectrophotometrically in 4:l (v/v) DMSO-CS2 (Kitagawa et al., 1995) (25).

FREE ENERGIES OF IONIC DISSOCIATION

Based on quantitative determination of the degree of dissociation in DMSO or in sulfolane, the magnitude of the free energy change in carbon-carbon heterolysis, AGOhet(R-R’), was calculated for dissociation equilibria of three series of hydrocarbons: tropyl series (Scheme 1, Miyabo et al., 1992a, 1993), cyclopropenyl series (Scheme 2, Takeuchi et al., 1993) and the fullerene derivative [30-471 [(25), Kitagawa et al., 19951. Values are given in Table 3. The dissociation is much more favoured in DMSO than in sulfolane, in spite of the magnitude of their dielectric constants which are almost identical, i.e. 46.68 and 43.3, respectively. The favourable dissociation,could be ascribed to extraordinarily strong solvation by DMSO around the cations

IONIC DISSOCIATION OF C-C u BONDS IN HYDROCARBONS

195

p

Z

[25-21, [99-21,[IOO-21

[25+],[99'], [100+]

A [25+]

[99+1

Scheme 2

[30+1

as compared with sulfolane, and this is borne out by their donor numbers, 29.8 and 14.8, respectively (Gutmann, 1978). Alcohols are heterolysed into carbocations and water in the presence of the hydronium ion (6). From the equilibrium constant the free energy of heterolytic dissociation of the carbon-oxygen u bond, AGi,,(ROH + H 3 0 + ) , can be calculated. The AG",,,(ROH + H 3 0 + ) value is related to the pKR+ of the carbocation by (26). AG"h,,(ROH

+ H30')

= -2.303 RT p K R +

(26)

K. OKAMOTO ETAL.

196

Table 3 Free energy changes in the ionic dissociation of carbonxarbon v bonds in some hydrocarbons at 25°C.”

Compound [91-21‘ [3-21 [95-2]‘ [96-21 [97-21 [98-21‘ [30-21 [34-21 [35-21‘ [25-21 [99-21 [ 100-21 [30-471

Solvent DMSO DMSO DMSO sulfolanef DMSO sulfolanef DMSO sulfolanef DMSO sulfolanef DMSO sulfolanef DMSO sulfolanef DMSO sulfolanef DMSO DMSO DMSO DMSO-CS2 (4 :1v/v)

K i e J AG iet(R-R’)/ lop6M~ kcal mol-’ 1.07 0.846 1.04 0.168 1.02 0.192 2.15 1.33 6.49 1.33 244 11.7 152 21.3 119 25.1 93.5 396 286 2.3

8.19 8.29 8.18 9.21 8.18 9.19 7.75 8.03 7.08 8.02 4.93 6.75 5.25 6.45 5.36 6.45 5.50 4.65 4.84 7.7

pKRcc

AGiet(ROH + H30’)/ kcal mol-ld

3.88 3.88 4.57

-5.29 -5.29 -6.23

4.90

-6.68

5.42

-7.39

5.76

-7.86

7.63

-10.4

7.80

-10.6

7.80

-10.6

7.09

-9.67

7.63

- 10.4

“Takeuchi et al. (1993); Miyabo et al. (1993). bCalculated from the concentration equilibrium constant Khel using equation KEel = Khelx y”. ‘pKR+ value of the cationic moiety in 50% aqueous acetonitrile. d-2.303RT pK,+ at 25°C. eMixture of two positional isomers. f5:95 methylsulfolane-sulfolane.

This expression makes possible the examination of extrathermodynamic correlation between the free energy changes for two dissociation equilibria, i.e. AGEe,,(R-R’) values of the hydrocarbons and AGg,,(ROH H30’) values of the alcohols. As shown in Fig. 2, the AGEet values of the hydrocarbons having a common anionic moiety [2-] measured in DMSO at 25°C are linearly correlated with the AGEe,,values of the alcohols in 50% aqueous acetonitrile at 25°C. It is also shown that, in terms of the free energy change, dissociation of the alcohols is more sensitive to structural variation than that of the hydrocarbons. Provided that the solvent and steric effects on the structural change are common, this difference in the sensitivity would most probably reflect the difference in the single-bond strength between carbonoxygen and carbon-carbon bonds. As regards enthalpy of dissociation AHhet(R-R’)and entropy of dissociation AShet(R-R’),only a few data are available (Table 4). Bordwell (1988;

+

IONIC DISSOCIATION OF C-C

(T

BONDS IN HYDROCARBONS

197

9

8 -

8 8 A?

7 -

k

3 6 -

5 -

4 f

I

- 12

I

1

- 10

I

I

-8

1

I

-6

-4

AGhe,(ROH+H30+)/ kcal mol-'

Fig. 2 Plot of the AGEet(R-R')in DMSO for dissociative hydrocarbons having the anionic moiety [2-] against the AG$,(ROH H30') in 50% aqueous acetonitrile for the corresponding alcohols. The numbers indicate the structure of the carbocation formed by the dissociation of C-C and C - 0 bonds.

+

Table 4 Enthalpy and entropy of dissociation for some hydrocarbons in DMSO.

Compound

AHh,,(R-R')/ kcal mol-'

&?h,,(R-R')/ cal K-' mol-'

Ere,@+)/ AHhomo(R-R')/ V" kcal mol-Ib Reference

13-21

3.06 (19.6')

-16.7

-0.510d

10.7

[25-21

-0.81 (11.9=)

-20.4

- 1.525

30.2

[loo-21

-0.38

-18.6

-1.566

31.6

Miyabo et al. (1993) Takeuchi et al. (1993) Takeuchi et al. (1993)

"Measured vs. Ag/Ag+ in acetonitrile. bCalculated by (27) with EOx([2-]) = -0.18 V vs. Ag/Ag+ in acetonitrile (Kinoshita et af., 1994). =Estimated from (28) by using pKR+= 3.88 (Okamoto et af., 1983) and pK, = 5.9 (cf. Table 2). dOkamoto et al. (1983). 'Estimated from (29) by using pKR+= 7.09 and pK, = 5.9 (cf. Tables 1 and 2).

K. OKAMOTO ETAL.

198

Bordwell et al., 1988, 1989) and Arnett (Arnett et al., 1990a,b, 1992; Venimadhavan et al., 1992) have employed thermodynamic cycles consisting of heterolysis of a molecule and redox processes of the resulting ions to evaluate homolytic dissociation energies of C-H, C-C, C-N, C - 0 and C-S bonds in solution. In a similar way, knowledge of the AHh,,(R-R') values allows determination of the heat of homolysis of carbon-carbon bonds [AHhom0(R-R')]using (27). The results are summarized in Table 4.

The AS&,(R-R') for the cyano-nitro compound [5-61 in (2) was reported to be OcalK-'mol-' in acetonitrile and -21 calK-'mol-' in 1,2dichloroethane. The dielectric constants of these solvents (37.5 and 10.36, respectively) indicate that the entropy of the dissociation becomes increasingly positive as the dielectric constant increases (Troughton et al., 1984). In contrast, the entropy values for the hydrocarbons, cited in Table 4, are much more negative than expected from the large dielectric constant (46.68) of DMSO. This would be also ascribed to strong solvation of DMSO [donor number 29.8 (Gutmann, 1978)] around the cation.

STERIC ASSISTANCE OF IONIC DISSOCIATION

A linear relationship of free energies is extrathermodynamic, and such a correlation is hardly a theoretical corollary which directly results from the axioms of thermodynamics alone. However, the slope of the straight line and departures from linearity can often suggest something physically meaningful concerning the chemical reactivity, as seen in Fig. 2. Another linear correlation between AH values and between AG values has been proposed to correlate the heats of heterolysis for the carbonvalues of the cations and pK, values of the carbon u bond with pK,, conjugate acids of the anions by Arnett et al. (1987a, 1990a). From the results of calorimetry for the coordination of resonance-stabilized carbocations and carbanions in sulfolane or acetonitrile, these workers demonstrated that (28) and (29), for secondary and tertiary cations, respectively, can be used for predicting heats of heterolysis of the carbon-carbon u bond. For secondary cations:

For tertiary cations: AHhet(R-R') = 8.895 - 0.648(pKR+)+ 1.294(pKa)

(29)

IONIC DISSOCIATION OF C-C

20

U

199

BONDS IN HYDROCARBONS

-I

/

15

10

I

0

5

0

-5

0

2

4

A pKp

6

8

+ B PK,

Fig. 3 Plot of AGg,,(R-R’) and AHh,,(R-R’) values against the Arnett equation scale. m: AHhe,(R-R’)for [4-1021 in acetonitrile. A: AHh,,(R-R’) for [3-21 and [25-21 in DMSO. 0 : AGie,,(R-R‘)for [101-1021 in acetonitrile. 0 : AGget(R-R’) for reaction (23) in sulfolane. 0 : AGgeI(R-R’) for reaction (23) in DMSO.

These empirical linear relations should be applied to groups of compounds with a common steric background. Thus, a negative deviation from these linear relations most probably indicates specific steric congestion assisting the dissociation. In this context, it is known that, in radical dissociation of carbon-carbon u bonds, steric congestion is a major enhancing factor (Ruchardt and Beckhaus, 1980). As shown in Fig. 3, the observed values of AHh,,(R-R‘) in DMSO for the hydrocarbons [3-21 and [25-21 are =15 kcalmol-’ less than the values predicted from (28) and (29), respectively, while AHhet(R-R’) for the nitro-cyano compound [4-1021 in acetonitrile (30) is close to the value predicted from (29). Therefore, this departure of the dissociative hydrocarbons from the predicted behaviour can be ascribed to steric congestion in the hydrocarbon, predominantly caused by the carbanion moiety [2-].

200

K. OKAMOTO ETAL.

R R+*

CN CN

=

R

[4-1021 R=Me [ioi-1021R = B ~

R

AR

[4+] R=Me [101+]R = B ~

NC

f

NC

(30) [102-1

The AGiet(R-R’)values observed in acetonitrile are almost identical with AHhet(R-Rf) values for the compounds [4-61 and [5-61 (Troughton et al., 1984). Thus, provided that this is the case also for a nitro-cyano compound, (tri-t-butylcyclopropenyl)(phenyl)malononitrile [101-1021, its AGi,,,(R-R’) value should be close to the AhH,,,(R-R’) value predicted from (29). In practice, as indicated in Fig. 3, the AGohet(R-R’)observed in acetonitrile is =6 kcal mol-’ less than that predicted (Miyabo et al., 1993). Hence this departure from linearity can be most probably ascribed to steric congestion, which is mainly derived from the cation moiety [101+] in [101-1021. On the basis of the results in acetonitrile, it might be reasonable to assume that the values for AHhet(R-R’) and AGhet(R-R’)are aFparently close to each other also in sulfolane, since the dielectric constant (43.3) and the donor number (14.8) of this solvent are close to those of acetonitrile (37.5 and 14.1, respectively). On the basis of this assumption, Arnett’s equation (28) was examined for reactions of type (23). For these reactions, except for [3-21, only the AGiet(R-Rf)values are available. As shown in Fig. 3, the values for this system are about 10 kcal mol-’ less than predicted from (28). The negative deviation can also be ascribed to steric congestion in these hydrocarbon molecules. The large negative deviations, similar to those observed in sulfolane, are also seen in Fig. 3 for the values of AGE,,(R-R’) in DMSO. To sum up, in addition to the electronic stabilization and solvation, classical steric congestion caused by either the cation or anion moieties effectively controls the ease of ionic dissociation of the carbon-carbon (+ bond in hydrocarbons.

4

Isolation of hydrocarbon salts

FORMATION OF CARBOCATION-CARBANION SALTS

A few observations of the isolation of carbocation-carbanion salt containing heteroatoms, such as oxygen and nitrogen, have been reported. Over 30 years ago LeGoff and LaCount (1963) reported the formation of a

201

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

0

Ph

702Me

702Me

Me02C@C02Me Me02C

Ph

C02Me

[5+103-1

[3 103-1 +

6, [3+10K]

[93’1 05-1 2-

Me02C

H+ C02Me

C02Me (Me2N)&

+

I

Et-6H

I

C02Me

-

702Me (Me2N)3C+ Et-6-

I

+

Me2NH

C02Me

(31) [ 1091-H

[108+109-]

carbocation-carbanion charge transfer salt by combining 1,2,3,4,5pentakis(methoxycarbony1)cyclopentadienide ion [103-] with carbocations, tropylium [3+] and triphenylcyclopropenylium [5+] ions, as acceptors. In the same year Matsumura and Set0 (1963) and Boyd (1963) reported the formation of similar salts [3+104-] and [93+105-], respectively. Carbocation-carbanion salts containing a dication or a dianion, such as [1062’(9-)2] (Maas et al., 1990) and [3+1072-]H+ (LeGoff and LaCount, 1964), are also known. Theoretical calculations by Mixon and Cioslowski (1991) suggested that a “push-pull” substituted ethane (H2N)3C- C(NO& exists as an ion pair (H2N)$+ -C(N02)3. Recently, a similar hexamethylguanidinium-

202

K. OKAMOTO ET AL.

carbanion salt, [108+109-], was prepared from highly CH-acidic ethylmalonic acid dimethyl ester, [109]-H, and tetrakis(dimethy1amino)methane (31). Its structural features have been clarified by an X-ray crystallographic analysis (Reetz et af., 1993).

ISOLATION OF AUTHENTIC HYDROCARBON SALTS

In view of the ease of dissociation of the (+ bond into resonance-stabilized hydrocarbon ions, it is expected that an authentic hydrocarbon salt could be isolated from an extraordinarily stable cation and anion. The first synthesis of a salt composed solely of carbon and hydrogen was achieved on the combination of tris(3-guaiazulenyl)cyclopropenylium ion [ 1+] and tris(7Hdibenzo[c,g]fluorenylidenemethyl)methanide ion [2-] (Okamoto et af., 1985). Four years after Doering’s discovery of the hygroscopic bromide of the tropylium ion [3+] (Doering and Knox, 1954), Vol’pin et al. (1957) reported that the tropylium ion (pKR+ 3.88) gives a covalent compound with acetate ion (pKb 9.24) but not with isocyanate ion (pKb 10.1). This suggests that an anion with PKb greater than 10.1 would give a salt with the tropylium ion. When a comparable examination was made with 1,3,5-tricyclopropylcyclopropenylium ion [24+] (pKR+ lO.O), the transition from covalent compound formation to no reaction in water was observed between the nucleophiles with the pKb values 4.79 (CN-) and 6.85 (p-nitrophenoxide ion) (Komatsu et al., 1980). As a corollary, it is expected that Kuhn’s anion [2-], whose PKb value (8.1) is greater than that of the p-nitrophenoxide ion, would give a stable salt when combined with the carbocations with a pKR+ value 9 or greater. So far, five hydrocarbon salts have been prepared in pure form from a series of trisubstituted cyclopropenylium and tropylium ions: [1+] (pKR+ 2 lo); [24+] (pKR+ 10.0); [26+] (pKR+ 7.29); [28+] (pKR+ 8.7); and [40+] (pKR+ 8.9) (cf. Table 1) and Kuhn’s carbanion [2-] (Okamoto et af., 1985, 1988, 1990; Komatsu et af., 1989; Takeuchi et af., 1993). Cation [31+] (pKR+ 13.0) also gave a stable salt with [2-], but it was obtained only as a monohydrate (Komatsu et af., 1991). The AHh,,(R-R’) value, which is derived from Arnett’s empirical linear (28) and (29), might be used as an index of feasibility of salt formation. The values calculated by use of (28) or (29) for the combination of [2-] with [1+], [24+], [26+], [28+] and [40+] are not less than zero but in the range of 10-18 kcal mol-’. These AHh,,(R-R’) values seem to be too large for salt formation. Therefore, the salt formation might be controlled not only by electronic factors, but also by steric hindrance to bond formation. The experimental procedure for the preparation of analytically pure hydrocarbon salts was as follows. First, the deep green solution of Kuhn’s

IONIC DISSOCIATION OF C-C

u BONDS IN HYDROCARBONS

203

anion [2-] was prepared by mixing potassium t-butoxide and the parent hydrocarbon [2]-H in THF (32). To prevent deterioration by exposure to oxygen and light, the preparation was conducted under argon in the dark. Then the solution of a carbocation salt, [l+]ClO,, [24+]BF;, [26+]ClO,, [28+]BF, or [40+]BF, in THF, THF-acetonitrile, or acetonitrile was mixed with an equivalent amount of the THF solution of K+[2-] under argon in the dark. The solution of Agranat's cation [1+] instantly gave black precipitates of mixed KC104 and [1+2-]. The crude hydrocarbon salt was isolated by recrystallization from DMSO as greenish black needles.

204

K. OKAMOTO ETAL.

On the other hand, the solutions of cations [24+], [26+], [28+] and [40+] gave homogeneous green solutions when they were mixed with K+[2-]. Evaporation of the solvents in vucuo or addition of pentane gave mixtures of the respective hydrocarbon salts and KBF4 or KC104 as dark green solids. The salts [28+2-] and [40+2-] were extracted with THF and reprecipitated by adding large amounts of pentane. In the case of [24+2-] and [26+2-], the crude salts were dissolved in chloroform in the form of the corresponding covalent hydrocarbons -[24-21 and [26-21 (see Section 6). The chloroform solutions were filtered and evaporated to regenerate dark green solids, which were dissolved in THF and reprecipitated by adding large amounts of acetonitrile.

5 Physical properties of hydrocarbon salts SOME PHYSICAL PROPERTIES

The molecular formula, crystalline forms, colour, the results of combustion tests and elemental analysis data are summarized for the isolated hydrocarbon salts [l+2-], [24+2-], [28+2-], [40+2-] (Okamoto et a f . , 1990) and [26+2-] (Takeuchi et a f . , 1993) in Table 5. The IR spectra (KBr disk) of the salts consist of absorptions of both the component cation, [1+], [24+], [26+], [28+] or [40+], and Kuhn’s carbanion [2-]. The ultraviolet-visible (UV-vis) spectra (DMSO solution) also agree with those of the component cation and anion superimposed, except that [26+2-] undergoes partial coordination into the covalent form [26-21 in DMSO, as indicated by the presence of only ~ 9 0 %of the theoretical amount of [2-] at the concentration of =lop4 M. Selected A,,, values are listed in Table 6. Figs 4 and 5 show the observed UV-vis and infrared (IR) spectra of [1+2-] as a representative.

ELECTRICAL CONDUCTIVITY IN DMSO

The complete dissociation of the hydrocarbons [1+2-], [28+2-] and [40+2-] in DMSO has been demonstrated by quantitative generation of both Kuhn’s carbanion [2-] and carbocations [1+], [28+] and [40+] as determined by UV-vis spectra (Table 6 and Fig. 4). However, since carbocation [24+] has no absorptions at a wavelength region longer than 220nm in the UV spectrum, there remained an ambiguity that this cation might have decomposed in the DMSO solution. A clue to this problem could be obtained by determination of the electric conductivity of DMSO solutions of hydrocarbon salts (Table 7) (Okamoto et af., 1990).

Table 5 Some physical properties of hydrocarbon salts prepared from Kuhn's anion [2-].a

Carbocation

P+l ~4+1

pKR+

>10 10.0

P6+1

7.29

P8+1

8.7

[40+1

8.9

Molecular formula

Crystalline form and colour

ClI5Hw Greenish black needles C79H54 Dark green powder Dark green powder CS3HSS Dark green powder G8Hm Dark green powder

Dec. pointPCb

Anal. obs. (calc.)

In air

In vacuo

Combustion test'

21gd

23gd

No ash

150

170

No ash

164

No ash

~145

i=

145

No ash

=160

=200

No ash

C%

H%

93.71 (93.84) 94.40 (94.57) 93.97 (94.64) 94.37 (94.46) 95.31 (95.11)

6.12' (6.16) 5.58 (5.43) 5.17 (5.36) 5.26 (5.54) 4.73 (4.89)

"Okamoto et al. (1990); Takeuchi et al. (1993). bNot liquefied; colour changed to grey or orange. 'Burned on a spatula. dChanged to greenish black liquid. 'CI = 0.00%. fF = 0.00%.

206

K. OKAMOTO ETAL.

Table 6 UV-vis spectra of hydrocarbon salts prepared from Kuhn’s carbanion [2-].a

Hydrocarbon salt [1+2-] [24+2-] [28+2-] [40+2-]

,,,A

in DMSOhm (log E )

303 (5.08),b2c 336 (4.92),b,‘ 350 (4.88),b*‘483 (4.72),d696 (5.17)b 303 (4.94),b353 (4.78),b696 (5.18)b 302 (5.01),b3‘ 351 (4.88),b7‘696 (5.17jb 304 (5.01),b353 (5.03),b7‘ 696 (5.18)

“Okamoto et al. (1990); Kitagawa et al. (1992). bAbsorptionof [2-]. ‘Overlapped by absorption of the cation. “Absorption of the cation.

Equivalent conductances at infinite dilution (A,, S cm2mol-l) for carbocations [1+], [24+] and Kuhn’s anion [2-] at 25°C were determined first by measuring A, values of [l+]ClO,, [24+]ClO, and H+[2-] and then subtracting reported A, values for C104- (24.52) (Bolzan and Arvia, 1970), or H + (14.6) (Gopal and Jha, 1974). Satisfactory Onsager plots for [l+]ClO,, [24+]ClO, and H+[2-] were obtained, giving A, values of 30.2, 36.5 and 19.9, respectively. From these values, A, [1+], A, [24+] and A, [2-] were estimated as 5.7, 12.0 and 5.3, respectively. These values afforded predicted A, values of 11.0 and 17.3 for [1+2-] and [24+2-], respectively. Observed conductance data for [1+2-] and [24+2-] gave linear Onsager plots, indicating that they are strong electrolytes in DMSO, and giving A, values of 10.9 and 18.2 at 25.OoC, respectively. Good agreement of the observed limiting equivalent conductances with the predicted values indicates that the component ions exist in DMSO without significant deterioration under argon. It was also shown that [1+2-] and [24+2-] are dissociated to more than 99% in DMSO over a concentration range 10-~-10-~M. 6 Chemical behaviour of hydrocarbon salts in solution HYDROCARBON SALTS IN CHLOROFORM

The five hydrocarbon salts [1+2-], [24+2-], [26+2-], [28+2-] and [40+2-] are generally unstable in chloroform. However, as far as the stage immediately after dissolution is concerned, their behaviour can be classified into two categories: (a) [1+2-], [28+2-] and [40+2-] afford a brown solution, which on immediate evaporation gives a reddish solid that generates [2-] only partly (
IONIC DISSOCIATION OF C-C u BONDS IN HYDROCARBONS

207

15

10

[1+] c104-

K+ [2-]

5

cf

2

L

w

O

--' I

I

I

I

,

I

,

I

300

400

500

600

I

I

_-

10

5

0

200

e 700

800

900

Wavelength/ nm Fig. 4 UV-vis spectrum of hydrocarbon salt [lf2-] in DMSO.

categories, the formation of brown solutions suggests that covalent hydrocarbons are formed in chloroform. The fact that [1+2-].and [40+2-], when dissolved in chloroform and then recovered, no longer regenerate [2-] in DMSO indicates that the reddish solids essentially consist of hydrocarbons that are difficult to ionize and whose cationic moiety may be different from the original structure.

K. OKAMOTO ETAL.

208

3000

2000

1200

1600

Fig. 5 IR spectrum (KBr disk) of hydrocarbon salt [1+2-]. 0 : absorptions of [2-].

0:

Table 7 Equivalent conductancesa at infinite dilution for hydrocarbon salts, [1+2-] and [24+2-] and related ionic species in DMSO at 25°C.

Salt or ion

Ao(obs.)

Ao(calc.)

10.9 18.2 30.2 36.5 19.9

11.0 17.3

~~

[ 1 +2-] [24+2-] [ 1 +po;= [24 ]ClO,' H+[2-]' [1+1 ~4+1

F-1

5.7 12.0 5.3

"ScmZrnol-'. See text for A, values of H+ and CIO;. 'Okamoto et al. (1990). 'See text for A, values of H+ and Clop.

800 cm-'

absorptions of [1+].

209

IONIC DISSOCIATION OF C-C (+BONDS IN HYDROCARBONS

HYDROCARBONS WHICH EXIST SOLELY IN SOLUTION

The brown colour of a chloroform solution of the salt [24+2-] suggests the formation of the covalent hydrocarbon [24-21 by cation-anion combination (33). The structure of [24-21 was determined by 'H and 13C nmr spectroscopy in CDC13 (Okamoto et ul., 1988, 1990). When the brown chloroform solution of the salt [24+2-] was kept for a few hours at c.20"C under argon and the solvent was evaporated in vucuo, the salt was quantitatively isolated as a greenish solid [24+2-]. The greenish salt was also precipitated by cooling the chloroform solution at -78°C. Thus, the covalent hydrocarbon [24-21 exists solely in solution. Most probably, great congestion in any hypothetical solid state of hydrocarbon [24-21 may be relieved in its salt form, being stabilized by the lattice energy. Such conversion may be seen as a mechanochemical phenomenon at the molecular level. Recently two additional examples, [26-21 and [28-21, of hydrocarbons existing solely in solution have been found in the chloroform solutions of hydrocarbon salts [26+2-] and [28+2-] (Table 5 ) (Komatsu et af., 1989; Takeuchi et al., 1993).

dissolution in chloroform

d

evaporation or refrigeration

[24+2-]

[24-21

(33) A SOLUTION IN WHICH IONIC, RADICAL AND COVALENT SPECIES COEXIST IN

EQUILIBRIA

Although redox equilibria via electron transfer between ions are familiar in inorganic systems, a comparable example in organic chemistry was first observed by Arnett and his co-workers only recently (Arnett et af., 1987b). Malachite green cation [1lo+] and tris(p-nitropheny1)methanide ion [ 111-1 have redox potentials favourable for single-electron transfer and are sterically hindered against covalent bond formation. As a result, when the tetrafluoroborate salt of [110+] and the sodium salt of [ill-] were mixed in THF, containing a crown ether (18-crown-6) to prevent ion pairing, the

K. OKAMOTO ETAL.

210

-

NO2

THF

Q-C

+

Sulfolane

Me2N

Me2N

[110+]

[110-]

No2

(34) [lll-]

single-electron transfer producing the corresponding radicals [110.1 and [111.1 was observed by spectrophotometric and electron spin resonance measurements. Reverse electron transfer to regenerate ions was observed by diluting the THF solution of radicals with sulfolane (34). If the cation moiety of the dissociative hydrocarbon has marked electronaccepting properties and the anion moiety is readily oxidized, a similar equilibrium is expected. An example has been found with hydrocarbon [28-21 (Komatsu et al., 1989; Okamoto et al., 1990). The hydrocarbon salt [28+2-] dissolves in DMSO without any reaction, giving a deep green solution. In contrast, when this salt is dissolved in chloroform, it affords an orange-brown solution, spectrophotometrically suggesting the formation of a hydrocarbon [28-21. The formation of the covalent hydrocarbon in chloroform is also supported by the fact that the anion [2-] is regenerated upon dilution of the freshly prepared solution with 10 volumes of DMSO. Evaporation under vacuum of the fresh chloroform solution does not afford the covalent hydrocarbon [28-21 but regenerates the salt [28+2-]. Hence the hydrocarbon [28-21 is unique in that it exists stably only in solution. As shown in Fig. 6, when the salt [28+2-] is dissolved in a variety of solvents, the anion [2-] is rapidly consumed by reaction with the cation [28+] and reaches an apparent equilibrium, the position of which depends on the solvent polarity. Clearly, the initial rate of the cation-anion reaction increases as the solvent polarity decreases. The distinguishing feature of the salt [28+2-] is most evident in THF. As shown in Fig. 7, the anion [2-] reacts rapidly with the cation [28+]. However, the concomitant formation of the radical [2.] in smaller amount than the consumed anion is observed spectrophotometrically. The difference between the amounts of consumed [2-] and of produced [2.] is attributed to the formation of the covalent hydrocarbon [28-21 in THF solution. Furthermore, the dimer [28-281 of 1,3,5-tricyclopropyltropylradical [28-] is isolated in 20% yield from the reaction mixture after 48 h at room temperature.

IONIC DISSOCIATION OF C-C

0

u BONDS IN HYDROCARBONS

4

2

6

21 1

8

10

Time/ day Fig. 6 Changes in concentration of anion [2-] determined from the visible spectrum for the reaction with cation [28+] in various solvents at 25°C. Initial concentration: [2-] = [28' -1 = 1 x 1 0 - ~ ~ .

("/.I 100

1. 50

0 0

1

2

3

4

5

6

Time/h Fig. 7 Changes in concentrations of anion [2-], radical [2.] and covalent compound [28-21 in THF at 20"C, as estimated from the visible spectrum of the reaction mixture of [2-] and [28+]. Initial concentration: [2-] = [28+] = 1 x 1 0 - 4 ~ .

212

K. OKAMOTO ETAL.

It is obvious from the graphical presentation in Fig. 7 that the cation [28+] and the anion [2-] exist in equilibrium with the covalent hydrocarbon [28-21 as well as the radical [28.] and [2-] as formulated in (35). In other words, the THF solution is a unique system in which one can observe four elemental species of organic compounds, i.e. covalent molecule, cation, anion and radical, at the same time. 7 Control of hydrocarbon salt formation CHANGE FROM A COVALENT COMPOUND TO A SALT BY A SLIGHT DIFFERENCE IN CARBOCATION STABILITY

It has to date been recognized that the breaking and forming of bonds in solution are in principle influenced by three major factors: electronic, steric and solvent effects. Thus, in a quantitative examination to differentiate covalent and ionic bond formation, it is necessary first to investigate the electronic effect alone, separate from steric and solvent effects. A series of l-aryl-2,3-dicyclopropylcyclopropenylium ions, [ 112+], [25+], [99+], [loo+], [26+] and [113+], allows progressive change in the electronic

IONIC DISSOCIATION OF C-C

U

BONDS IN HYDROCARBONS

A

[112+]

z= p-c1

Z Z [loo ] Z [26+1 Z [I13 ] Z [25+]

213

=

H

= m-Me =

m,m'-Me2

= p-Me = p-Me0

stability by changing the substituent on the aryl group, while the steric environment towards a carbanion is essentially unchanged (Miyabo et al., 1992b; Takeuchi et al., 1993). When a THF-acetonitrile solution of any one of the ClO; or BF; salts of cations [112+], [25+], [99+] and [loo+] was added to a deep green THF solution of Kt[2-] under argon in the dark, the colour immediately turned brownish orange, suggesting the formation of a covalent compound. After the isolation and purification of the product, the structures were determined to be the covalent compounds [112-21, [25-21, [99-21 and [loo-21, as shown in Scheme 3. In contrast, when a THF-acetonitrile solution of [26+]C10, or [ll3']ClO; was added to the THF solution of K+[2-], the deep green colour of carbanion [2-] persisted. After evaporation of the solvent, purification of the residue yielded a dark green salt, [26+2-] or [113+2-] (Scheme 31, the IR spectra (KBr disk) of which consisted of the absorptions of [2-] and [26+] or [113+] superimposed. The visible-near IR region

0

Z

126'2.1 Z = p-Me [ 1 13'2-1 Z = p - M e 0

Scheme 3

K. OKAMOTO ETAL.

214

Table 8 Eredand pKR+values of l-aryl-2,3-dicyclopropylcyclopropenylium ions.‘

Compound 112+]C10; 25+]BFT 99+]ClO, lOO+]ClO, 26’lClO;113 ]C104

Aryl group

EredNb

p-Chlorophenyl Phenyl m-Methylphenyl rn,m’-Dimethylphenyl p-Methylphenyl p-Methoxylphenyl

-1.412 -1.525 -1.532 - 1S66 -1.583 -1.632

pKR+ 7.09 k 0.05d 7.29 k 0.20e

Product‘ Covalent Covalent Covalent Covalent Salt Salt

“Takeuchi er al. (1993). bCV peak potential vs. Ag/Ag+ measured in acetonitrile with Bu,N+CIO; as the supporting electrolyte at 0.1 V precision 0.01 V. ‘Obtained from the reaction with [2-]. “In 50% aqueous acetonitrile (Moss et al., 1986). eEstimated from the methoxide exchange reaction between 3-methoxy-1,2,3-tri-t-butylcyclopropeneand [25+] or [26+]in CD,CN by the use of ‘H nmr spectroscopy (Takeuchi et al., 1993). s C ’ ;

electronic spectra of the solids [26+2-] and [113+2-] showed only the absorption of [2-] (A,,, 697 nm) in DMSO solution or in the solid state, and no charge-transfer band was observed up to 2500 nm. Hence, these salts are classified as charge-separated salts rather than charge-transfer complexes. It also turns out from the reduction potentials of [112+], [25+], [99+], [loo+], [26+] and [113+] (Table 8) that the product of the reaction, a covalent compound or a salt with [2-], changes abruptly from a covalent compound to a salt as the stability of the cation moiety is gradually increased. The borderline between the two types of reactions lies between [loo+] and [26+]. The difference (0.017 V) in the Ered values of these cations suggests that only a slight difference (0.4 kcal mol-’ or less) in electronic stability can completely switch the type of bond. DIFFERENCE IN ELECTRON AFFINITY O F THE CARBOCATION

The electron affinity of the carbocations as measured by Eredis a useful index for the stability of the carbocations. It is of great interest to correlate the occurrence of three principal reactions between a carbocation and a carbanion, i.e. covalent bond formation (36), single-electron transfer (37) and salt formation (38), with the magnitude of the Ered for the carbocations. R: R:

+ R:

+ R;

-+

-+

R,-R*

R1.

R:+R;-+R:R;

+ *R2

(36) (37) (38)

Accumulated data for the reactivity of a series of hydrocarbon cations towards the carbanion [2-] are summarized in Fig. 8 (Kitagawa et al., 1992).

\

-2.0

I

SALT

I

I

MeoMe / A

A

11'1

131'1

@

/ -1.0

1

I

U

Ph

l>h

0.0

1

Fig. 8 Reactions of various carbocations with Kuhn's anion [2-] as compared with their reduction potentials (peak potentials measured vs. Ag/Ag+ in acetonitrile by cyclic voltammetry; cf. Tables 1 and 8 and Okamoto et al., 1983). SALT, salt formation; COV, covalent bond formation; ET, single-electron transfer.

216

K. OKAMOTO ETAL.

The anion [2-] [E,, -0.18V vs. Ag/Ag+ in acetonitrile (Kinoshita et al., 1994)] gives products of all three types with resonance-stabilized hydrocarbon cations whose reduction potentials range from -0.54 to -2.20 V. As for the salt formation and single-electron transfer, thermodynamics for simple redox processes may be applied to predict their selectivity. As a first approximation, a cation with E red lower and higher than ~ 0 .V2 would give a salt and a radical pair, respectively, when combined with [2-].In practice, the cations which were found to give salts with [2-]have Ered values more negative than -0.8 V. On the other hand, quantitative single-electron transfer has been observed from [Z-]to the heptaphenyltropylium ion [53+], which is relatively unstable: pKR+ -0.54 in methanol (Battiste and Barton, 1968) and E r e d -0.30 V vs. Ag/Ag+ in acetonitrile (Kitagawa et a / . , 1992). As the cation becomes progressively more reluctant to be reduced than [53+], covalent bond formation is observed instead of electron transfer. Further stabilization of the cation causes formation of an ionic bond, i.e. salt formation. Thus, the course of the reaction is controlled by the electron affinity of the carbocation. However, the change from single-electron transfer to salt formation is not straightforward. As has been discussed in previous sections, steric effects are another important factor in controlling the formation of hydrocarbon salts. The significant difference in the reduction potential at which a covalent bond is switched to an ionic one around -0.8V for tropylium ion series and -1.6V in the case of l-aryl-2,3-dicyclopropylcyclopropenyliumion series - may be attributed to steric factors. 8 Conclusions

Considering the long saga of hydrocarbon chemistry, it is surprising that two new classes of hydrocarbon - ionically dissociative hydrocarbons and hydrocarbon salts - have been discovered in the last decade. The syntheses of authentic samples as analytically pure solids have revealed the very existence of such novel hydrocarbons in an unquestionable way, but the investigation of their basic features is just in the inchoate stage. The search for such novel hydrocarbons depends primarily on the synthesis and examination of highly stabilized hydrocarbon cations and anions. As mentioned above, until now such elaboration has been concentrated on the carbocation side, and examination of the carbanion moiety has only just started. Computational approaches remain almost untouched, and the main reason probably is the inevitable structural complication which is required of molecules and ions to gain sufficient resonance stabilization. Recently, a molecular orbital calculation has been applied to a simplified hydrocarbon salt model by Fujimoto et al. (1991).

IONIC DISSOCIATION OF C-C

(T

BONDS IN HYDROCARBONS

217

The details of structural features such as the length of the carbon-carbon bond that easily undergoes heterolysis and the spatial arrangement of ions in hydrocarbon salts are intriguing subjects. However, they have not been presented because of difficulties in X-ray crystallographic analysis. Electrical characterization of the hydrocarbon salts, though they are not yet stable enough t o air and light, might open unexpected delights as new functional materials. (T

Acknowledgements

One of the authors (K.O.) acknowledges his indebtedness t o his former mentor, Professor W. von E. Doering, for the encounter with stable carbocation chemistry. The authors are also grateful t o Professors Koichi Komatsu and Tomomi Kinoshita for their collaboration during the development and clarification of the problems discussed in this chapter.

References Agranat, I. and Aharon-Shalom, E. (1976). J. Org. Chem. 41, 2379 Arnett, E. M. and Molter, K. E. (1985). Acc. Chem. Res. 18, 339 Arnett, E. M. and Troughton, E. B. (1983). Tetrahedron Lett. 24, 3299 Arnett, E. M., Troughton, E. B., McPhail, A. T. and Molter, K. E. (1983). J. Am. Chem. SOC. 105, 6172 Arnett, E. M., Chawla, B., Amarnath, K. and Whitesell Jr, L. G. (1987a). Energy and Fuels 1, 17 Arnett, E. M., Molter, K. E., Marchot, E. C., Donovan, W. H. and Smith, P. (1987b). J. Am. Chem. SOC.109, 3788 Arnett, E. M., Amarnath, K . , Harvey, N. G. and Cheng, J.-P. (1990a). J . Am. Chem. SOC. 112, 344 Arnett, E. M., Arnarnath, K . , Harvey, N. G. and Venirnadhavan, S. (1990b). J. Am. Chem. SOC.112, 7346 Arnett, E. M., Venirnadhavan, S. and Arnarnath, K . (1992). J. Am. Chem. SOC. 114, 5598 Battiste, M. A. (1962). J. Am. Chem. SOC.84, 3780 Battiste, M. A. and Barton, T. J. (1968). Tetrahedron Lett. 2951 Bethell, D. and Gold, V. (1967). In Carbonium Zons, p. 117. Academic Press, London Bleasdale, C. and Jones, D. W. (1993). J. Chem. SOC., Perkin Trans. I 2441 Bolzan, J. A. and Arvia, A. J. (1970). J. Electrochim. Acta 15, 39 Bordwell, F. G. (1988). Acc. Chem. Res. 21, 456 Bordwell, F. G., Cheng, J.-P. and Harrelson Jr, J. A. (1988).'J. Am. Chem. SOC. 110, 1229 Bordwell, F. G., Harrelson Jr, J. A. and Satish, A. V. (1989). J. Org. Chem. 54, 3101 Born, M. (1920). Z . Phys. 1, 45 Boyd, R. H. (1963). J. Am. Chem. SOC. 85, 1555

218

K. OKAMOTO ET AL.

Colle, K. S., Glaspie, P. S. and Lewis, E. S. (1975). J. Chem. SOC.,Chem. Comrnun. 266 Cram, D. J. and Ratajczak, A. (1968). J. Am. Chem. SOC.98, 2198 Dauben Jr, H. J. and Wilson, J. D. (1968). Chem. Comrnun. 1629 Doering, W. von E. and Knox, L. H. (1954). J. Am. Chem. SOC.76, 3203 Doering, W. von E. and Knox, L. H. (1957). J. Am. Chem. SOC.79, 352 Dunnebacke, D., Neumann, W. P., Penenory, A. and Stewen, U. (1989). Chem. Ber. 122, 533 Eisenberg, D. C., Lawrie, C. J. C., Moody, A. E. and Norton, J. R. (1991). J. Am. Chem. SOC.113, 4888 Fagan, P. J., Krusic, P. J., Evans, D. H., Lerke, S. A. and Johnston, E. (1992). J. Am. Chem. SOC.114, 9697 Feiler, L. A. and Huisgen, R. (1969). Chem. Ber. 102, 3428 Fleming, I. (1973). J. Chem. SOC.,Perkin Trans. I 1019 Fujimoto, H., Denno, S. and Jinbu, Y. (1991). J . Phys. Chem. 95, 1612 Gomberg, M. (1900). J. Am. Chem. SOC.22, 757; Ber. 33, 3150 Gopal, R. and Jha, J. S. (1974). J. Phys. Chem. 78, 2405 Gotoh, T., Padias, A. B. and Hall Jr, H. K. (1986). J. Am. Chem. SOC.108, 4920 Gutmann, V. (1978). The Donor-Acceptor Approach to Molecular Interactions, ch. 2. Plenum Press, New York Hall, R. H., den Hertog Jr, H. J., Reinhoudt, D. N., Harkema, S., van Hummel, G. J. and Uiterwijk, J. W. H. M. (1982). J. Org. Chem. 47, 977 Hirota, H. and Mitsuhashi, T. (1990). Chem. Lett. 803 Hoffmann, R. W. (1971). Angew. Chem. Int. Ed. Engl. 10, 529 Hoffmann, R. W. (1985). Acc. Chem. Res. 18, 248 Hoffmann, R. W. and Hauser, H. (1965). Tetrahedron 21, 891 Hoffmann, R. W. and Schafer, W. (1972). Chem. Ber. 105, 2437 Hoffmann, R. W., Barth, W., Schuttler, R. and Mayer, B. (1986). Chem. Ber. 119, 3297 Hollenstein, S. and Laube, T. (1993). J. Am. Chem. SOC.115, 7240 Howe, N. E., Yankee, E. W. and Cram, D. J. (1973). J . Am. Chem. SOC. 95, 4230 Huisgen, R. (1977a). Acc. Chem. Res. 10, 117 Huisgen, R. (1977b). Acc. Chem. Res. 10, 199 Huisgen, R. and Steiner, G. (1973). J. Am. Chem. SOC. 95, 5054 Huisgen, R., Feiler, L. A. and Otto, P. (1969). Chem. Ber. 102, 3405 Huisgen, R., Schug, R. and Steiner, G. (1974a). Angew. Chem. Int. Ed. Engl. 13,80 Huisgen, R., Schug, R. and Steiner, G. (1974b). Angew. Chem. Int. Ed. Engl. 13, 81 Ito, S., Morita, N. and Asao, T. (1991). Tetrahedron Lett. 32, 773 Ito, S., Morita, N. and Asao, T. (1992). Tetrahedron Lett. 33, 3773 Ito, S., Morita, N. and Asao, T. (1994a). Tetrahedron Lett. 35, 751 Ito, S., Morita, N. and Asao, T. (1994b). Tetrahedron Lett. 35, 755 Ito, S., Morita, N. and Asao, T. (1994~).Tetrahedron Lett. 35, 3723 Ito, S., Morita, N. and Asao, T. (1995). Bull. Chem. SOC.Jpn. 68, 1409 Jefferson, E. A. and Warkentin, J. (1992). J. Am. Chem. SOC.114, 6318 Jefferson, E. A. and Warkentin, J. (1994a). J. Org. Chem. 59, 455 Jefferson, E. A. and Warkentin, J. (1994b). J. Org. Chem. 59, 463 Jutz, C. (1965). Angew. Chem. 77, 224 Kagayama, A , , Komatsu, K., Nishinaga, T., Takeuchi, K. and Kabuto, C. (1994). J . Org. Chem. 59, 4999 Karle, I., Flippen, J., Huisgen, R. and Schug, R. (1975). J. Am. Chem. SOC.97, 5285 Kataoka, F., Shimizu, N. and Nishida, S. (1980). J. Am. Chem. SOC.102, 711

IONIC DISSOCIATION OF C-C u BONDS IN HYDROCARBONS

219

Kayarna, Y., Oda, M. and Kitahara, Y. (1973). Chem. Lett. 1169 Kinoshita, T., Kirnura, H., Nakajirna, I . , Tsuji, S. and Takeuchi, K. (1994). J. Chem. SOC.,Perkin Trans. 2 165 Kitagawa, T., Takeuchi, K., Kornatsu, K. and Kinoshita, T. (1992). J . Synth. Org. Chem., Japan 50, 746; Chem. Abstr. 117, 170 467 Kitagawa, T., Tanaka, T., Takata, Y., Takeuchi, K. and Kornatsu, K. (1995). J. Org. Chem. 60, 1490 Kornatsu, K. (1993). Pure Appl. Chem. 65, 73 Kornatsu, K., Masurnoto, K. and Okarnoto, K. (1977). J . Chem. SOC., Chem. Commun. 232 Kornatsu, K., Takahashi, K. and Okarnoto, K. (1979). Tetrahedron Lett. 4747 Kornatsu, K., Tornioka, I. and Okarnoto, K. (1980). Tetrahedron Lett. 21, 947 Kornatsu, K., Masurnoto, K., Waki, Y. and Okarnoto, K. (1982a). Bull. Chem. SOC. Jpn. 55, 2470 Kornatsu, K., Takeuchi, K., Arirna, M., Waki, Y . , Shirai, S. and Okarnoto, K. (1982b). Bull. Chem. SOC.Jpn. 55, 3257 Kornatsu, K., Akarnatsu, H. and Okarnoto, K. (1987). Tetrahedron Lett. 28, 5889 Kornatsu, K., Akarnatsu, H., Jinbu, Y. and Okarnoto, K. (1988). J . Am. Chem. SOC. 110, 633 Kornatsu, K., Aonurna, S . , Takeuchi, K. and Okarnoto, K. (1989). J. Org. Chem. 54, 2038 Kornatsu, K., Akarnatsu, H., Aonurna, S . , Jinbu, Y., Maekawa, N. and Takeuchi, K. (1991). Tetrahedron 47, 6951 Kornatsu, K., Nishinaga, T., Maekawa, N., Kagayarna, A . and Takeuchi, K. (1994). J . Org. Chem. 59, 7316 Kuhn, R. and Rewicki, D. (1967a). Angew. Chem. Int. Ed. Engl. 6, 635 Kuhn, R. and Rewicki, D. (1967b). Justus Liebigs Ann. Chem. 706, 250 Lankarnp, H., Nauta, W. T. and MacLean, C. (1968). Tetrahedron Lett. 249 LeGoff, E. and LaCount, R. B. (1963). J . A m . Chem. SOC. 85, 1354 LeGoff, E. and LaCount, R. B. (1964). Tetrahedron Lett. 1161 Lernal, D. M., Gosselink, E. P. and McGregor, S . D. (1966). J . A m . Chem. SOC.88, 582 Lifshitz, C. (1994). Acc. Chem. Res. 27, 138 Lustgarten, R. K. and Richey Jr, H. G. (1974). J . Am. Chem. SOC.96, 6393 Maas, G., Weber, H. M., Exner, R. and Salbeck, J. (1990). J . Phys. Org. Chem. 3, 459 Majchrzak, M. W., Jefferson, E. and Warkentin, J. (1990). J . A m . Chem. SOC. 112, 2449 Matsurnura, S. and Seto, S. (1963). Chem. Pharm. Bull. (Tokyo) 11, 126 Mitsuhashi, T. (1986). J . Am. Chem. SOC.108, 2394 Mitsuhashi, T. and Hirota, H. (1990). J . Chem. SOC., Chem. Commun. 324 Mixon, S. T. and Cioslowski, J. (1991). J . Am. Chem. SOC.113; 6760 Miyabo, A., Kitagawa, T. and Takeuchi, K. (1992a). Chem. Lett. 679 Miyabo, A., Takeuchi, K., Kornatsu, K. and Kitagawa, T. (1992b). Chem. Lett. 683 Miyabo, A . , Kitagawa, T. and Takeuchi, K. (1993). J . Org. Chem. 58, 2428 Moss, R. A. and Munjal, R. C. (1980). Tetrahedron Lett. 21, 1221 Moss, R. A., Shen, S., Krogh-Jespersen, K., Potenza, J. A., Schugar, H. J. and Munial, R. C. (1986). J . Am. Chem. SOC. 108, 134 Murata," I . , Yarnamoto,'K. and Kayane, Y. (1974). Angew. Chem. Znt. Ed. Engl. 13, 808 Myhre, P. C., Webb, G. G. and Yannoni, C . S. (1990). J . Am. Chem. SOC.112, 8991

220

K. OKAMOTO ETAL.

Nakazawa, T., Niimoto, Y . , Kubo, K. and Murata, I. (1980). Angew. Chem. 92, 556 Nevell, T. P., de Salas, E. and Wilson, C. L. (1939). J. Chem. SOC. 1188 Okamoto, K., Komatsu, K., Kinoshita, T. and Shingu, H. (1970). Bull. Chem. SOC. Jpn. 43, 1901 Okamoto, K., Takeuchi, K., Komatsu, K., Kubota, Y., Ohara, R., Arima, M., Takahashi, K., Waki, Y. and Shirai, S. (1983). Tetrahedron 39, 4011 Okamoto, K., Kitagawa, T., Takeuchi, K., Komatsu, K. and Takahashi, K. (1985). J. Chem. SOC., Chem. Commun. 173 Okamoto, K., Kitagawa, T., Takeuchi, K., Komatsu, K. and Miyabo, A. (1988). J. Chem. SOC.,Chem. Commun. 923 Okamoto, K., Kitagawa, T., Takeuchi, K., Komatsu, K., Kinoshita, T., Aonuma, S., Nagai, M. and Miyabo, A. (1990). J. Org. Chem. 55, 996 Olah, G. A. and Lukas, J. (1967a). J. Am. Chem. SOC.89, 2227 Olah, G. A. and Lukas, J. (1967b). J. Am. Chem. SOC.89, 4739 Olah, G. A., Baker, E. B., Evans, J. C., Tolgyesi, W. S . , McIntyre, J. S. and Bastien, I . J. (1964). J. Am. Chem. SOC.86, 1360 Olah, G. A., Svoboda, J. J. and Ku, A. T. (1973). Synthesis 492 Reetz, M. T., Bingel, C. and Harms, K . (1993). J. Chem. SOC., Chem. Commun. 1558 Reinhoudt, D. N., Verboom, W., Visser, G. W., Trompenaars, W. P . , Harkema, S. and van Humrnel, G. J. (1984). J. Am. Chem. SOC. 106, 1341 Riichardt, C. and Beckhaus, H.-D. (1980). Angew. Chem. Znt. Ed. Engl. 19, 429 Riichardt, C. and Beckhaus, H.-D. (1986). Top. Curr. Chem. 130, 1 Schug, R. and Huisgen, R. (1975). J. Chem. SOC., Chem. Commun. 60 Shimizu, N. and Nishida, S. (1978). J. Chem. SOC., Chem. Commun. 931 Takahashi, K., Hirata, N. and Takase, K. (1970). Tetrahedron Lett. 1285 Takahashi, K., Takase, K. and Sakae, T. (1980). Chem. Lett. 1485 Takeuchi, K., Kitagawa, T., Miyabo, A , , Hori, H. and Komatsu, K. (1993). J. Org. Chem. 58, 5802 Thurnmel, R. P. and Chayangkoon, P. (1983). J. Org. Chem. 48, 596 Troughton, E. B., Molter, K. E. and Arnett, E. M. (1984). J. Am. Chem. SOC. 106, 6726 Venirnadhavan, S., Amarnath, K., Harvey, N. G., Cheng, J.-P. and Arnett, E. M. (1992). J. Am. Chem. SOC. 114, 221 Vincow, G., Dauben Jr, H. J., Hunter, F. R. and Volland, W. V. (1969). J. Am. Chem. SOC.91, 2823 Visser, G. W., Verboom, W., Trompenaars, W. P. and Reinhoudt, D. N. (1982). Tetrahedron Lett. 23, 1217 Vol’pin, M. E., Akhrem, I . S. and Kursanov, D. M. (1957). Khim. Nauka Prom. 2, 656; Chem. Abstr. 52, 429431. (1958) Wigfield, D. C., Feiner, S. and Taymaz, K. (1972a). Tetrahedron Lett. 891 Wigfield, D. C., Feiner, S. and Taymaz, K. (1972b). Tetrahedron Lett. 895 Wigfield, D. C., Feiner, S . , Malbacho, G. and Taymaz, K. (1974). Tetrahedron 30, 2949 Winstein, S. and Trifan, D. S. (1949). J. Am. Chem. SOC. 71, 2953 Yamamoto, K . and Murata, I . (1976). Angew. Chem. 88, 262 Yankee, E. W. and Cram, D. J. (1970a). J. Am. Chem. SOC.92, 6328 Yankee, E. W. and Cram, D. J. (1970b). J. Am. Chem. SOC.92, 6329 Yankee, E. W. and Cram, D. J. (1970~).J. Am. Chem. SOC.92, 6331 Yankee, E. W., Badea, F. D., Howe, N. E. and Cram, D. J. (1973a). J. Am. Chem. SOC. 95. 4210

IONIC DISSOCIATION OF C-C U BONDS IN HYDROCARBONS

221

Yankee, E. W., Spencer, B., Howe, N. E. and Cram, D. J . (1973b). J . Am. Chem. SOC.95, 4220 Yannoni, C. S., Macho, V. and Myhre, P. C. (1982a). J . Am. Chem. SOC. 104, 907 Yannoni, C. S., Macho, V. and Myhre, P. C. (1982b). J . Am. Chem. SOC.104, 7380 Yannoni, C. S . , Kendrick, R. D., Myhre, P. C., Bebout, D. C. and Petersen, B. L. (1989). J . Am. Chem. SOC.111, 6440 Yeo, A. N. H. and Williams, D. H. (1970). Chem. Commun. 737

This Page Intentionally Left Blank

Author Index Numbers in italic refer to the pages on which references are listed at the end of each chapter. Aarts, V. M. L. J., 73, 74, 75, 86, 87, 88, 95,96,112, I15 Abboud, J.-L. M., 111,112 Abkin, A. D., 138, I69 Adamic, R. J., 87, 112 Addadi, L., 121, 162,169 Adegawa, Y . , 126, 162, 166, 167,170 Agranat, I . , 182, 183,217 Aharon-Shalom, E., 182, 183,217 Ahlrichs, R., 45,56 Akamatsu, H., 182, 202,219 Akhrem, I. S., 202,220 Akiyama, S . , 98,114 Alam, I . , 101,113 Alberts, A. H., 101,112 Alder, R. W., 69,112 Alfieri, C., 101, 112 Ali, S. Z., 111,113 Ammann, J., 14,56 Anderegg, G., 65,112 Andreetti, G. D., 101,112 Andrews, L., 11,37,50,53,57,60,61 Annet, R., 148,169 Aonuma, S., 182, 202, 204, 205,206, 208,209,210,219,220 Aoyama, M., 124, 162, 166,170 Apeloig, Y . , 45, 56 Arai, Y., 104,105,114 Araki, K., 100,115 Arena, G., 78, 79, 114 Arima, M., 182, 197, 215,219,220 Armarnath, K., 198,217,220 Arnaud-Neu, 65,112 Arnett, E. M., 175, 188, 191, 198, 200, 209,217,220 Arrington, C. A., 31, 56 Arvia, A. J., 206,217 Asao, T., 182,218 Ase, P . , 35,56 Aue, D. H., 111,114 Avakyan, V. G., 52,57

Baker, E. B., 176,220 Balaji, V., 48, 54, 55,56,58 Balli, H., 15, 59 Barrau, J., 32,54,55,56,57 Barth, W., 188,218 Bartmann, M., 27,60 Bartoletti, M., 91, 113 Barton, T. J., 18, 47,49,57,59 Barton, T. J., 216,217 Bartsch, R. A., 80, 87,98,112,113,115 Bashall, A . , 68, I13 Bashkirova, S. A., 48,58 Baskir, 40,42, 43, 44, 45, 50, 51,56,58, 60 Bastien, I.J., 176,220 Battiste, M. A., 185, 216,217 Baumgartner, R. 0. W., 49,59 Baumstark, R., 75, 78, 109,114 Bazzicalupi, C., 68, 70, 112 Bean, D. L., 32,56 Bebout, D. C., 176,221 Beckhaus, H.-D., 185, 199,220 Beer, P. D., 101, 112 Bell, G. A., 26,56 Bell, T. W., 73,81,88, 91, 105, 106,112 Bencini, A., 68, 70, 112 Berecz, I . , 30,61 Berger, B., 101,112 Bergman, R. G., 117,171 Bernardi, F., 133,169 Bernstein, H. I . , 118, 169 Bernstein, J., 158,170 Besenyei, I., 51, 54, 55,58, 61 Bethell, D., 178,217 Bianchi, A., 68, 70, 112 Bingel, C . , 202,220 Blanda, M. T., 86,112,114 Bleasdale, C., 188, 190,217 Bleckmann, P., 55,56 ' Bock, H., 38,56 Bock, W . , 35,56 Boganov, S. E., 48, 53, 54, 55,58 Bohatka, S., 30, 61 Bohmer, V., 101,112,113

Badea, F. D., 186,220 Baird, M. S., 25,56 223

224

Bohn, R. B., 37,57 Bokyi, N. G., 30,59 Bolzan, J. A., 206,217 Bonapace, J. A. P., 117,171 Bordwell, F. G., 183, 196, 198,217 Born, M., 178,217 Bos, A . , 54,57 Bos, M., 73, 74, 75,113 Bourdelande, J. L., 23, 61 Boyd, R. H., 201,217 Brabson, G. D., 37,57 Bradshaw, J. S., 64, 65, 78, 79, 80, 107, 114,115,116 Braun, H.-G., 135,169 Breitenbach, J., 78, 115 Brown, D. W., 33,60 Browne, C. M., 98, 105,112 Bruening, R. L., 64, 65, 107, 114 Bucher, G., 26,57 Bugge, K. E., 101,115 Burns, G. T., 49,59 Butler, R . L., 34,35,57 Carsky, P., 12, 13,59 Cassol, A., 98, 112 Cataliin, J., 111, 112 Cesario, M., 84, 112 Chang, C. A., 98,113 Chang, C. C., 18,47,57 Chang, J. S., 33,60 Chang, S.-K., 109, 110,112,113 Chapman, 0. L . , 18, 23, 47, 49,57,59, 61 Chawla, B., 198,217 Chayangkoon, P., 182,220 Chedekel, M. R . , 47,57 Cheney, J., 69,113,115 Cheng, J.-P., 198,217,220 Cheng, P. G . , 81, 88, 91, 105, 106,112 Chernyshev, E. A . , 29,48,57,58 Chettur, G., 35,57 Chimichi, S . , 71, 113 Chung, C.-H., 121, 131, 155,170 Chung, C.-M., 121, 135, 140, 142, 147, 148, 155, 166, 169, 170 Ciampolini, M., 68, 70,71,112, 113 Cioslowski, J., 201, 219 Clement, A., 23, 61 Cohen, M. D., 118,169 Colle, K. S., 185,218 Cornetta-Morini, C., 40,58 Cox, B. G., 69, 70,113

AUTHOR INDEX

Cradock, S. 2,57 Cram, D. J., 81, 82, 86, 87, 88, 91, 101, 105, 106, 111, 112,113,114, 186,218, 220,221 Crerner, D., 26,56 Curtis, N. F., 63, 113 Cyvin, B. N., 34, 35,57, 61 Cyvin, S. J., 34, 35,57, 61 Czech, B. P., 98,113 Czech, N. F., 98,113 Dailey, W. P., 18, 60 Dalley, N. K., 78, 79, 114 Dapporto, P., 68,70,71,112,113 Dauben, H. J. Jr., 184,218,220 Davidson, W. R., 111, 114 de Salas, E., 177, 220 Dehnicke, K . , 38, 39,59 den Hertog, H. J. Jr., 73, 74, 75, 105, 107,113,115, 188,218 Dendrarnis, A., 13,57 Denno, S., 216,218 Desai, D. H., 87, 112 Desvergne, J. P., 121,171 Dewar, M. J. S . , 132,169 Di Bernando, P., 98, 112 Dietrich, B., 64,65, 66, 68, 107,113 Dietrich, C. O., 84, 112 Dijkstra, P. J., 101,115 Doering, W. von E., 176,202,218 Donovan, W. H., 209,217 Doyle, T. J., 36, 57, 61 Dradi, E., 101,112 Droppers, W. H. J., 73, 74, 75, 115 Du, X. M., 20,57 Dubinsky, M. Yu., 31,58 Duff, R. J., 93, 115 Dunkin, I. R . , 25, 26,56 Diinnebacke, D., 185,218 Dye, J. L . , 69, 115 Ebmeyer, F., 109,113 Eckhardt, C. J., 136,171 Edel, A . , 84, 112 Egberink, 95, 115 Eisenberg, D. C., 185,218 Ekstrom, A . , 65,112 El-Amine, M., 55,57 Elguero, J., 111,112 Endo, Y., 162, 170 Enklemann, V., 162, 169 Evans, D. H., 183, 184,218

AUTHOR INDEX

Evans, J. C., 176,220 Exner, R., 201,219 Eyring, E. M., 87, I12 Fabbrizzi, L., 68, 113 Fagan, P. J., 183, 184,218 Fan, E., 109, 112 Feil, D., 75, 113 Feiler, L. A , , 187, 218 Feiner, S . , 190, 220 Fenton, D. E., 68,113 Ferguson, G., 98, 105, 112 Fleming, I . , 182,218 Flippen, J., 187, 218 Frank, W., 103,115 Franke, R., 118,169 Fredin, L., 28,57,58 Fronczek, A. D., 83, 114 Fujirnoto, H., 216,218 Fujishige, S . , 126, 169 Fujitake, M., 30,57 Fukui, K., 133, 169 Fusi, V., 68, 70, I12 Ganzer, G. A . , 20,57 Garcia-Tellado, F., 110,113 Gelbert, M., 75, 84, 114 Gennari, C., 91, I13 Gerasirnov, G. N., 138,169 Gerst, M., 75, 114 Ghindini, E., 101, 115 Gieb, S. J., 110,113 Glaspie, P. S., 185,218 Glattthaar, J., 32, 59 Gloecke, H. J., 53, 61 Goebel, P., 136,171 Gokel, G. W., 63, 64, 65, 81,113,114 Gold, V., 178,217 Goldberg, I . , 87, 88, 92, 107, 113 Goldrnann, H., 101,113 Gornberg, M . , 184,218 Gornory, A . , 51,54, 55,58, 61 Gopal, R., 206,218 Gordan, A. J., 117, 169 Gordon, W. E., 93,115 Gosselink, E. P., 188, 219 Goswarni, S., 110, I13 Gotoh, T., 187,218 Graham, W. R. M . , 13,35, 36,57,58, 61 Groenen, L. C., 101,115

225

Grootenhuis, P. D. J., 73, 74, 75, 86, 87, 88,95, 96, 107, 111,112, 113, 115 Guilhern, J., 84, 112 Guillory, W. A , , 29, 61 Guirnon, C., 55,57 Gurnprecht, W. H., 118, 170 Gusel’nikov, L. E., 52,57 Gutrnann, V., 195,218 Gutsche, C. D., 98, 101, 113 Hacker, N., 25,56 Hall, H. K. Jr., 187,218 Hall, R. H., 188,218 Hamilton, A. D., 109, 110,112, 113 Hanarnura, E., 125, I69 Hargittai, I . , 30, 39, 57, 61 Harkerna, S., 73, 74, 75,86, 87, 88, 95, 96, 101, 105, 107, 111,112,113,115, 188,218 Harms, K., 86, 115 Harms, K., 202,220 Harrelson, J. A. Jr., 198,217 Harrison, J. F., 13, 57 Harvet. N. G., 198,217,220 Hasegawa, M., 118, 119, 120, 121, 122, 124, 125, 126, 127, 130, 131, 132, 135, 137, 138, 140, 142, 145, 147, 148, 151, 155, 157, 158, 162, 163, 164, 166, 167, 169, 170, 171 Hashirnoto, Y . , 121, 126, 132, 140, 155, 162, 170,171 Hatada, S., 131, 170 Hatakeyama, T., 126, I70 Hauge, R. H., 9, 10, 14, 28,57,58,59, 60 Hauser, H., 188,218 Hayashi, K., 117, 142, 147, 148, 166, 169,170, 171 Hayashida, O., 65, 114 Hayes, R. A . , 23,57 Healy, E. F., 132, 169 Hedinger, M., 104, 105, 114 Heinzrnann, R., 45,56 Hess, B. A . Jr., 12, 13,59 Hess, T. C., 23,57 Hinchcliffe, A . J., 2, 57 Hirabayashi, T., 38,56 Hirata, N., 188,220 Hirota, E., 30,57 Hirota, H., 189,218,219 Hirshfeld, F. L . , 118, 170 Hisaeda, Y . , 65, 114

226

Ho, G.-J., 15, 57 Hoffrnann, R. W., 12,60,187, 188,218 Hollenstein, S . , 176,218 Holm, M. J., 137, I70 Holtzhauer, K., 40,58 Hope, H., 158,170 Hopkins, H. P. Jr., 111,113,114 Hori, H., 182, 193, 194, 196, 202,204, 205,209,213,214,220 Homer, J. H., 86,112,114 Howe, N. E., 186,218,220,221 Huang, J. W., 13,58 Huisgen, R., 187,218,220 Hunter, F. R., 184,220 Iitaka, Y . , 104, 105, 114, 125, 162, 170, I 71 Ikeda, T., 166,170 in’t Veld, P. J. A , , 101, 115 Inazu, T., 70, 115 Iqbal, M., 101, 113 Ismail, Z. K., 28,58 Ito, S . , 182,218 Ito, Y . , 125, I70 Izatt, R. M., 64,65, 78, 79, 80, 107,114, 116 Jacox, M . E., 8, 28, 35, 37,40, 41,58, 60 Jahagirdar, D. V., 111, 114 Jefferson, E. A , , 190,218,219 Jeon, E. G., 80,115 Jha, J. S., 206,218 Jinbu, Y . , 182, 202, 216,218,219 Johnson, E., 183, 184,218 Joko, S . , 104, 105,114 Jones, D. W., 188,190,217 Jones, W., 131, 162,164,171 Jung, M. E., 18,47,57 Jutz, C., 184,218 Kabuto, C., 182,218 Kaden, T. A , , 64,65, 68,113,114 Kafafi, Z . H., 14,28,57,60 Kagayarna, A., 182,218,219 Kagrarnanov, N. D., 8, 9, 10,30, 31, 38, 39,40, 46, 51,57,58,59,60,61 Kaneda, T., 98,114 Kanetsuna, H., 126, I70 Kanner, B . , 111, I14 Karle, I . , 187,218 Karni, M . , 45,56

AUTHOR INDEX

Kataoka, F., 187,218 Kato, S . , 122, 131, 145, 147, 158, 163, 170 Katsurnata, T., 125, 170 Kayarna, Y., 182,219 Kayane, Y . , 182,219 Kazansky, V. B . , 10,58 Keefe, A . D., 101, 112 Keller, M., 84, I14 Kendrick, R. D., 176,221 Kerzina, Z . A . , 50, 51, 52, 53, 54, 55, 58,61 Kesselrneyer, M . A . , 20, 22,58, 61 Khabashesku, V. N., 12,44,45,46,47, 48,50, 51, 52,53, 54,55,56,58,59, 60 Kirnura, E., 65, 104, 105,114 Kirnura, H., 183, 184, 197,216,219 Kirnura, Y . , 104, 105,114 Kinbara, K., 162, 166, 167,170 Kinoshita, T., 182, 183, 184, 197, 202, 204,205, 206, 208, 209, 210, 214, 216, 219,220 Kintzinger, J.-P., 84, 112 Kiovsky, T. E., 117, 171 Kirpichenko, S. V., 52,57 Kitagawa, T., 174, 175, 182, 192, 193, 194, 196,200,202,204,205, 206,208, 209,210,213,214,216,219,220 Kitahara, Y . , 182,219 Kitaigorodsky, A. I . , 164,170 Kitarni, T., 134, 171 Kitanishi, Y., 117, 171 Kjaer, A . M., 69,114 Klingensmith, K. A., 31,56 Klooster, W. L . , 73, 74,113 Krneicik-Lawrynowicz, G . , 20, 61 Knop, D., 70,113 Knox, L. H., 176,202,218 Knudsen, B . E., 80, 87,112,115 Koch, W., 32,60 Kodarna, M . , 104, 105, 114 Koelsch, C. F., 118, 170 Koeppe, R., 53, 61 Koike, T., 104, 105,114 Kokesh, F. C., 96, 112,114 Kolc, J., 18, 47, 57 Kolesnikov, S. P., 30,59 Komalenkova, N. G., 29,57 Kornatsu, K., 182, 184, 192, 193, 194, 196, 197, 202, 204, 205,206, 208, 209, 210,213, 214, 215, 216,218,219,220

AUTHOR INDEX

Komiya, H., 162,171 Korolev, V. A . , 12, 38,39, 40, 41, 43, 44,45,46, 47,56,58,59,60 Korsgen, U. U., 82, 115 Korzenoski, S. H., 63, 65,113 Kotin, E. B., 138,169 Kowalcki, K. U., 27, 60 Kreeger, R. L., 47,57 Kreil, C. L., 49,59 Krogh-Jespersen, K., 15, 20,57,61, 182, 214,219 Kroka, E., 26,57 Kruise, L., 73,74, 75, 86, 87, 88, 105, 107,112,113,115 Krusic, P. J., 183, 184,218 Ku, A. T., 176,220 Kubo, K., 182,220 Kubota, Y . , 182, 197, 215,220 Kudo, T., 50,59 Kuhn, R., 181,184,219 Kulishov, V. I . , 30,59 Kunita, A., 142, 147, 148, 166,169,170 Kuramoto, M., 104, 105,114 Kurita, T., 126,170 Kursanov, D. M., 202,220 LaCount, R. B., 200,201,219 LaGoff, E., 200,201,219 Lahav, M., 118, 121, 162,169,170 Lamb, J. D., 80,116 Langer, R., 27, 60 Lankamp, H., 184,219 Lau, P.-Y., 117, 171 Laube, T., 176,218 Lawrie, C. J. C., 185,218 Le Roux, J. P., 18,61 Lee, E. X. C., 15,61 Lee, Y. P . , 7,59 Lehn, J.-M., 64,65, 66,68,69, 107,113, 114,115 Lemal, D. M., 188,219 Leong, A. J., 68, I13 Lerke, S. A., 183, 184,218 Leroi, G . E., 13,57,61 Lewis, E. S., 185,218 Liang, W., 80,116 Liebermann, C., 118, 170 Lieser, G . , 119,171 Lifshitz, C., 176,219 Liitle, D., 109,113 Lim, C. S., 100, 101,115 Lim, P.-J., 162, 164, 166, 170

227

Lindoy, L. F., 64, 65,68, 107,112,113, 114 Littman, D., 46,59 Liu, M. T. H., 20,57 Liu, Y., 87, 112 Lloyd, B. A . , 87,112 Louis, R., 65, 112 Lowe, J. A . , 18, 47,57 Lucas, J., 176,220 Luning, U., 75, 76,78, 84, 86, 109, 114 Lustgarten, R. K., 188,219 Maas, G . , 201,219 Maass, G., 10, 59 Macho, V., 178,221 MacLean, C., 184,219 Maedak, K., 126,170 Maekawa, N., 182,202,219 Maekawa, Y . , 121, 122, 124, 131, 135, 142, 145, 158, 162, 164, 166,170 Maier, G., 12, 13, 15, 27,32, 38, 39,46, 49,56,59, 60 Majchrzak, M. W., 190,219 Majestic, V. K., 83, 114 Malbacho, G., 190,220 Mal’tsev, A . K., 8, 9, 10, 12, 15, 20,25, 29,30,31, 38, 39,40, 41, 43, 44,45, 46,47,50, 51,56,57,58,59,60, 61 Manabe, O., 100,115 Mandel, N. S . , 117,171 Marchot, E. C., 209,217 Margrave, J. L., 9, 10, 14, 28,57,58,59, 60 Masumoto, K., 182, 184,219 Matsumara, S . , 201,219 Matveichev, P. M., 48,58 Mayer, B., 188,218 McDouall, J. J. W., 133,169 McGregor, S. D., 188,219 McIntyre, T. S . , 176,220 McKervey, M. A . , 98, 105,112 McMahon, R. J., 12, 18,23,57,59,61 McPartlin, M., 68, 113 McPail, A. T., 175,217, 191,217,220 Mertens, R., 103,115 Meyer, W., 119,171 Michalczyk, M. J., 31, 61 Micheloni, M., 68,70,71,112,113 Michl, J., 31, 32, 48, 54, 55,56,58,60, 61 Mihm, G., 46,49,59,60 Mikaelian, R. G., 8, 10,59, 60

228

Mikova, 0. B . , 138,169 Miller, M. D., 32, 60 Milligan, D. E., 8 , 2 8 , 4 0 , 5 8 , 6 0 Minkwitz, S., 55, 57 Mirsky, K . , 132, 171 Misumi, S., 98,114 Mitsuhashi, T., 126, 170 Mitsuhashi, T., 189,218,219 Miura, M., 134, I71 Mixon, S. T., 201,219 Miyabo, A . , 182, 192, 193, 194, 196, 200,202, 204, 205,206,208, 209, 210, 213,214,219,220 Mohmand, S . , 38,56 Molinari, F., 91, 113 Molter, K. E., 175, 188, 191, 200, 209, 217,220 Montavon, F., 68, 114 Moody, A. E . , 185,218 Moore, S. S., 86,87,88,114 Mori, T., 151, 155,170 Morita, N., 182,218 Morokurna, K . , 133, I69 Morsi, S. E., 121, 171 Moss, R. A , , 15, 20,57,61, 182, 214, 219 Moulik, P. S . , 111, 114 Mulholland, D. L., 98, 105,112 Miiller, M., 75, 76, 78, 84, 86, 114 Miiller, W., 26, 27,60 Munjal, R. C., 182,214,219 Murakami, Y., 65,114 Murata, I . , 182,220 Murata, Y . , 182,219,220 Muro, N., 121, 135,170 Murphy, B . P., 68, I13 Murray-Rust, J., 69,113 Murray-Rust, P., 69,113 Myhre, P. C., 176, 178,219,221 Nagai, M., 182, 202,204,205, 206, 208, 209,210,220 Nagakub, K . , 134,171 Nagase, S., 50,59 Nagata, C., 133,169 Nakajima, I . , 183, 184, 197, 216,219 Nakamura, F., 140, 148,169 Nakanishi, F., 130, 138,170 Nakanishi, H., 119, 120, 121, 122, 125, 127, 130, 131, 138, 151, 155, 162, 164, 166,169,170,171

AUTHOR INDEX

Nakashima, K . , 98, 114 Nakatsuji, S., 98, 114 Nakatsuji, Y . , 78, 114 Nakazawa, T., 182,220 Narn, K. C., 101,113 Narnetkin, N. S . , 52,57 Namiki, S . , 130,170 Nardi, N., 68, 70,112 Nauta, W. T., 184,219 Neder, K. M., 109,114 Nefedov, 0. M., 8, 9, 10, 12, 15, 20,25, 29, 30, 31, 38, 39,40,41,42, 43, 44, 45,46, 47, 48, 50, 51, 52, 53, 54, 55, 56,57,58,59, 60,61 Nekhoroshev, N. S., 138,169 Neumann, W. P., 55,57, 185,218 Nevell, T. P., 177, 220 Newcomb, M., 81,82,86,87, 88, 91, 105, 106, 111,112,114 Newkorne, G. R., 83,114 Nieger, M., 78,115 Nierneyer, C. M., 86,115 Niirnoto, Y . , 182,220 Nishida, S., 187,218,220 Nishinaga, T., 182,218,219 Noguchi, H . , 166,170 Nohora, M., 151, 155,170 Norton, J. R., 185,218 O’Gara, J. E., 18,60 O’Connor, T., 98, 105,112 Obata, H., 126, 170 Ochaya, V. O., 98,113 Oda, M., 182,219 Ogden, J. S., 53, 54,57,60 Ogree, L., 5 4 , 5 7 Ohara, R., 182, 197, 215,220 Ohashi, Y . , 124, 131, 132, 145, 166,170, 172 Ohno, T., 65,114 Okamoto, K . , 174, 175, 182, 184, 192, 197, 202, 204, 205,206,208,209, 210, 215,219,220 Okarnura, S., 117, I71 Olah, G. A . , 176,220 Olivucci, M., 133, 169 Olson, W. B . , 35,58 Onodera, K . , 126,170 Ortrnan, B . J., 14,60 Oth, J . F. M., 40,58 Otto, P., 187,218

AUTHOR INDEX

Pacansky, J., 32, 33, 60 Padias, A. B . , 187,218 Pant, N., 110,113 Paoli, P., 68, 70,112 Parkinson, G. M., 162, 164,171 Parris, K . , 93, 115 Parvez, M., 98, 105,112 Pascasd, C., 84,112 Patai, S., 63, 65,115 Paul, I . C., 162,170 Paulus, E., 101,112,113 Pawlak, K . , 64,65, 107,114 Peachy, N. M., 136,171 Pedersen, C. J., 63,115 Pedley, M. D., 111,114 Penenory, A., 185,218 Perlmutter, B. L., 30,59 Perotti, A. M., 68,113 Peters, K., 75, 84, I14 Petersen, B. L., 176,222 Petrucci, S., 87,112 Pfister-Guillouzo, G., 55,57 Pimentel, G. C., 7,59 Pincock, R. E., 117, I71 Pochini, A., 101,112 Poh, B.-L., 100, 101,115 Poliakoff, M., 25,56 Potenza, G., 91,113 Potenza, J. A., 182, 214,219 Prochaska, F. T., 11, 60 Puff, H., 82,115 Pugia, M. J., 80, 87,112,115 Quimby, W. C., 118,169 Raabe, G., 31, 60, 61 Ramamurthy, V., 121,171 Ratajczak, A., 186,218 Rebek, J. Jr., 93, 107,115 Rees, W. L . , 131,171 Reetz, M. T., 86,115, 202,220 Reimann, A., 12,60 Reinhoudt, D. N., 73,74, 75, 86, 87, 88, 95,96, 101,105, 107,111,112,113, 115, 188,218,220 Reisenauer, H. P., 12, 15, 27, 32, 38, 39, 46,49,56,59, 60 Reuter, H., 82,115 Rewicki, D., 181, 184,219 Rezmovitz, H., 107, 113 Richey, H. G. Jr., 188,229 Richs, M. J., 53, 60

229

Rima, G., 55,57 Rittby, C. M. L., 36,57 Robb, M. A., 133,169 Rodriguez, A., 101,112 Rohde, B., 38, 39,59 Roos,,A., 75,113 Roth, W. R., 27,60 Rozsondai, B., 39,61 Riichardt, C., 185, 199,220 Sahni, S. K . , 73,112 Saigo, K . , 125, 131, 145, 147, 148, 151, 155, 162, 163, 164, 166, 167,169,170, 171

Sakae, T., 188,220 Sakuragi, H., 126, 170 Salbeck, J., 201,219 Sander, W., 26, 27,57,60 Sasada, Y., 119, 120, 127, 130, 131,171 Sasaki, K . , 126,170 Satish, A. V., 198,217 Sato, S., 142, 147, 166,170 Sauvage, J. P., 68, 84,112,114 Schaad, L. J., 12, 13,59 Schafer, W., 187,218 Schmidt, G. M. J., 118,169, 170,171 Schneider, H.-J., 64, 69, 70, 107,113, 115 Schnering, H. G. V., 75, 84,114 Schnockel, H., 50,52,53,54,60,61 Schriewer, M., 55,57 Schteinschneider, A. Ya., 10,523 Schug, R., 187,218,220 Schugar, H. J., 182, 214,219 Schultz, G., 30,57, 61 Schiittler, R., 188,218 Schwab, W., 12, 13,59 Schwing-Weill, M.-J., 65, 112 Schyja, W., 75,78, 109,114,115 Seburg, R. A,, 12, 61 See, K. A., 101,113 Seel, C., 107, 109, 110,115 Seghi, B., 68,113 Seto, S., 201,219 Seyferth, D., 8, 9, 10,59, 61 Shaefer, U., 15,59. Shechter, H., 47,57 Shelimov, B. N., 10,58 Shen, L. N., 36,57,61 Shen, S., 15,20,57, 61, 182, 214,219 Shepherd, R. A., 35,35,61

230

Sheridan, R. S., 14, 15, 18, 20, 22,56, 57,58, 61 Shiba, S . , 138,169 Shimanouchi, H., 119, 120,171 Shibata, J . , 100, I15 Shimizu, N., 187,218,220 Shingu, H., 184,220 Shinkai, S., 99, lOC, 115 Shinmyozu, T., 70, I15 Shionoya, M., 104, 105,114 Shirai, S., 182, 197, 215,219,220 Shukla, J . P . , 80, 115 Siebert, H., 45,56 Siegfried, L., 68, 113 Singh, P., 117, 171 Siobara, T., 130,170 Skoglund, M., 47,57 Skowronka-Ptasinska, M., 95, 96, 111, 115 Smith, D. V., 13, 61 Smith, G. R., 29, 61 Smith, P. B . , 69,115 Smith, P., 209,217 Smith, R. J., 65,112 Snelson, A., 8,32,34,35,56,57,61 Sodeau, J. R., 15,61 S0rensen, P. E., 69,114 Spencer, B., 186,221 Spirko, V., 12,59 Steiner, G., 187,218 Stern, C . L., 162,170 Stetter, H., 103,115 Stevermann, B . , 27,60 Stewart, J. J. P., 132,169 Stewen, U., 185,218 Stezowski, J. J . , 136,171 Strausz, 0. P . , 23, 61 Struchkov, Yu. T., 30,59 Subramanian, R., 14, 15, 20,56,57, 61 Sudholter, E. J. R., 73, 74, 75, I13 Sukegawa, M., 132,171 Sukenik, C . N., 117,171 Suslova, E. N., 52,57 Sustmann, R., 26,27,60 Suzuki, F., 118, 119, 121, 166,171 Suzuki, Y., 118, 119, 121, 135, 137, 138, 166,171 Svoboda, J. J., 176,220 Svyatkin, V. A., 10, 29,30,59, 60,61 Taft, R. W., 111, 112 Takahashi, K . , 174, 175, 182, 188, 192, 197,202,215,219,220

AUTHOR INDEX

Takahashi, M., 105,114 Takase, K . , 188,220 Takata, Y . , 194,219 Takemura, H., 70,115 Takeuchi, A . , 162, 171 Takeuchi, K., 174, 175, 182, 183, 184, 192, 193, 194, 196, 197, 200,202, 204, 205,206,208,209, 210, 213, 214, 215, 216,218,219,220 Tamaki, T., 135, 137,171 Tamas, J . , 8, 40, 43,44, 51, 54,56,58, 61 Tanaka, T., 194,219 Tarbet, B . J., 64,65, 107,114 Taymaz, K . , 19,220 Telleman, P., 95, 115 Thibud, M., 55,57 Thomas, J. M., 121, 131,171 Thummel, R. P., 182,220 Timko, J . M., 81, 82,114 Tokumaru, T., 126,170 Tolazzi, M., 98, I12 Tolgyesi, W. S., 176,220 Tomat, G., 98, 112 Tomilov, Yu. V., 25,60 Tomioka, I . , 182,202,219 Torres, M., 23, 61 Tremmel, J., 30, 39,57, 61 Trifan, D. S., 177,221 Trompenaars, W. P., 188,220 Troughton, E. B . , 175, 191, 200,217, 220 Trueblood, K. N., 105,113, 158,170 Tse, P.-K., 78, 79,114 Tsugawa, D., 100,115 Tsuji, S., 183, 184, 197, 216,219 Tsutsumi, T., 162,171 Tumey, M. L., 18,47,57 Turner, J. J . , 25,56 Uenishi, K., 104, 105,114 Ueno, K., 120, 127,171 Uitenvijk, J. W. H. M., 73,74, 75, 105, 107,113,115, 118,218 Ujszaszy, K . , 8,40, 43, 44,56,58, 61 Ulstrup, J . , 69, I14 Ungaro, R., 101,112,115 Uno, H., 151, 155, I70 Vaida, E., 39, 61 Valtancoli, B . , 68, 70,112 Van Der Wal, P. D., 74, 107,113 Van Engen, D., 109,112

231

AUTHOR INDEX

van Eerden, J., 73, 74, 75, 86, 87, 88, 95,96, 111,112,113,115 van Hummel, G. J., 188,218 van Loon, J.-D., 101,115 van Mil, J., 121, 162,169 van Staveren, C. J., 73,74,75,86,87, 88, 105, 107,112,113,115 van Truong, N., 69,113 Vancik, H., 31, 60,61 Venimadhaven, S., 198,217,220 Venkatesan, K . , 121,171 Verboom, W., 101,115, 188,220 Vincow, G., 184,220 Viout, P., 64, 65,68, 107,113 Visser, G. W., 188,220 Vizza, F., 71,113 Vogt, W., 101,112,113 Vogtle, F., 64,65, 77,78, 82, 107, 109, 110,113,115,116 Vol’pin, m. E., 202,220 Volkova, V. V., 52,57 Volland, W. V., 184,220 Voronkov, M. G., 52,57 Waki, Y., 182, 184, 197,215,219,220 Walba, D. M., 81, 82,114 Waltham, R. J., 32, 60 Wangnick, C., 75,89, 111,114 Warkentin, J., 190,218,219 Watta, B . , 55,57 Webb, G. G., 178,219 Webb, H. M., 111,114 Weber, E., 109,116 Weber, H. M., 201,219 Wegner, G., 119, 121, 135, 151,169,171 Wehner, W., 77,116 Weiss, R., 65, 112

Welsh, K. M., 32,56 West, R. C . , 31, 32,56, 61 West, R. P., 18, 57, 61 Westheimer, F. H., 96, 112,114 Whitesell, L. G. Jr., 198,217 Whitlock, B . J., 109,116 Whitlock, H. W., 109,114,116 Wigfield, D. C., 190,220 Wijmenga, S . S., 101,115 Wilk, B. K . , 20,61 Williams, D. H., 176, 221 Wilson, B. E., 78, 79,114 Wilson, C . L., 177,220 Wilson, J. D., 192,218 Wilson, K. R., 117,171 Wilson, S. R., 163,170 Winstein, S . , 177,220 Withnall, R., 50, 53, 61 Wlostowski, M., 20, 61 Wood, G., 117,171 Wu, G., 8,116 Yamamoto, K . , 182,219,220 Yamawaki, Y., 98,114 Yankee, E. W., 186,218,220,221 Yannoni, C. S., 176, 178,219,221 Yates, P., 117, 171 Yatsunami, T., 104, 105,114 Yeo, A . N. H., 176,221 Yonezawa, T., 133, 147,169,170 Yurugi, T., 125, 130,170 Zanobini, F., 71,113 Zanonto, P. L., 98,112 Zienty, F., 137,170 Zoebisch, E. G . , 132,169 Zuev, P. S., 12, 16,20,25,54,55,58,60

This Page Intentionally Left Blank

Cumulative Index of Authors

Ahlberg, P., 19, 223 Albery, W. J., 16, 87; 28, 139 Allinger, N. I., 13, 1 Anbar, M., 7, 115 Arnett, E. M., 13,83; 28, 45 Ballester, M., 25, 267 Bard, A. J., 13, 155 Baumgarten, M., 28, 1 Bell, R. P., 4, 1 Bennett, J. E . , 8, 1 Bentley, T. W., 8, 151; 14, 1 Berg, U., 25, 1 Berger, S., 16, 239 Bernasconi, C. F., 27, 119 Bethell, D., 7, 153; 10, 53 Blandamer, M. J., 14, 203 Bowden, K., 28, 171 Brand, J. C. D., 1,365 Brandstrom, A., 15, 267 Brinkman, M. R., 10,53 Brown, H. C., 1, 35 Buncel, E., 14, 133 Bunton, C. A., 22, 213 Cabell-Whiting, P. W., 10, 129 Cacace, F., 8, 79 Capon, B., 21,37 Carter, R. E., 10, 1 Collins, C. J., 2, 1 Cornelisse, J., 11, 225 Crampton, M. R., 7, 211 Davidson, R. S., 19, 1; 20, 191 Desvergne, J. P., 15, 63 de Gunst, G. P., 11, 225 de Jong, F., 17, 279 Dosunmu, M. I., 21,37 Eberson, L., 12, 1; 18, 79 Emsley, J., 26, 255

Engdahl, C., 19, 223 Farnum, D. G., 11, 123 Fendler, E. J., 8, 271 Fendler, J. H., 8, 271; 13,279 Ferguson, G., 1,203 Fields, E. K., 6, 1 Fife, T. H., 11, 1 Fleischmann, M., 10, 155 Frey, H. M., 4, 147 Gilbert, B. C., 5, 53 Gillespie, R. J., 9, 1 Gold, V., 7, 253 Goodin, J. W., 20, 191 Gould, I. R., 20. 1 Greenwood, H. H., 4,73 Hammerich, O., 20, 55 Harvey, N. G., 28,45 Hasegawa, M., 30, 117 Havinga, E., 11, 225 Henderson, R. A , , 23, 1 Henderson, S., 23, 1 Hibbert, F., 22, 113; 26, 255 Hine, J., 15, 1 Hogen-Esch, T. E., 15, 153 Hogeveen, H., 10,29, 129 Huber, W., 28, 1 Ireland, J. F., 12, 131 Iwamura, H., 26, 179 Johnson, S. L., 5, 237 Johnstone, R. A. W., 8, 151 Jonsall, G., 19, 223 JosC, S. M., 21, 197 Kemp, G., 20, 191 Kice, J. L., 17, 65 Kirby, A. J., 17, 183; 29, 87 Kitagawa, T., 30, 173 Kluger, R. H., 25,99 Kochi, J. K., 29, 185 Kohnstam, G., 5, 121 233

Korolev, V. A., 30, 1 Korth, H.-G., 26, 131 Kramer, G. M., 11, 177 Kreevoy, M. M., 6, 63; 16, 87 Kunitake, T., 17, 435 Kurtz, H. A , , 29, 273 Ledwith, A., 13, 155 Lee, I., 27, 57 Le Fevre, R. J. W., 3, 1 Liler, M., 11, 267 Long, F. A., 1, 1 Luning, U., 30,63 Maccoll, A., 3, 91 Mandolini, L., 22, 1 McWeeny, R., 4, 73 Melander, L., 10, 1 Mile, B., 8, 1 Miller, S. I., 6, 185 Modena, G., 9, 185 More O’Ferrall, R. A., 5, 331 Morsi, S. E., 15, 63 Mullen, K., 28, 1 Nefedov, 0. M., 30, 1 Neta, P., 12, 223 Nibbering, N. M. M., 24, 1 Norman, R. 0. C., 5,33 Nyberg, K., 12, 1 Okamoto, K., 30, 173 Olah, G. A., 4,305 Page, M. I., 23, 165 Parker, A. J., 5, 173 Parker, V. D., 19, 131; 20,s Peel, T. E., 9, 1 Perkampus, H. H., 4, 195 Perkins, M. J., 17, 1 Pitt,man, C. U. Jr, 4, 305 Pletcher, D., 10, 155 Pross, A., 14, 69; 21, 99 Ramirez, F., 9, 25 Rappoport, Z., 7, 1; 27, 239

234

Reeves, L. W., 3,187 Reinhoudt, D. N., 17, 279 Ridd, J. H., 16, 1 Riveros, J. M., 21, 197 Robertson, J. M., 1, 203 Rose, P. L., 28,45 Rosenthal, S. N., 13, 279 Ruasse, M.-F., 28,207 Russell, G. A., 23, 271 Samuel, D., 3, 123 Sanchez, M. de N. de M., 21,37 Sandstrom, J., 25, 1 Saveant, J.-M., 26, 1 Savelli, G., 22, 213 Schaleger, L. L., 1, 1 Scheraga, H. A., 6, 103 Schleyer, P. von R., 14, 1 Schmidt, S. P., 18, 187 Schuster, G. B., 18, 187; 22,311 Scorrano, G., 13,83

CUMULATIVE INDEX OF AUTHORS

Shatenshtein, A. I., 1, 156 Shine, H. J., 13, 155 Shinkai, S., 17, 435 Siehl, H.-U., 23, 63 Silver, B. L., 3, 123 Simonyi, M., 9, 127 Sinnott, M. L., 24, 113 Stock, L. M., 1,35 Sustmann, R., 26,131 Symons, M. C. R., 1,284 Takashima, K., 21, 197 Takeuchi, K . , 30,173 Ta-Shma, R., 27,239 Tedder, J. M., 16,51 Tee, 0. S., 29, 1 Thatcher, G. R. J . , 25, 99 Thomas, A., 8, 1 Thomas, J. M., 15,63 Tonellato, U., 9, 185 Toullec, J., 18, 1 Tudos, F., 9, 127

Turner, D. W., 4, 31 Turro, N. J., 20, 1 Ugi, I., 9 , 25 Walton, J. C., 16, 51 Ward, B., 8, 1 Watt, C. I. F., 24, 57 Westheimer, F. H . , 21, 1 Whalley, E., 2, 93 Williams, A., 27, 1 Williams, D. L. H., 19, 381 Williams, J. M. Jr, 6, 63 Williams, J. O., 16, 159 Williams, R. V., 29, 273 Williamson, D. G., 1, 365 Wilson, H., 14, 133 Wolf, A. P., 2, 201 Wyatt, P. A. H., 12, 131 Zimmt, M. B., 20, 1 Zollinger, H., 2, 163 Zuman, P., 5, 1

Cumulative Index of Titles Abstraction, hydrogen atom, from 0 - H bonds, 9, 127 Acid solutions, strong, spectroscopic observation of alkylcarbonium ions in, 4, 305 Acid-base behaviour in macrocycles and other concave structures, 30, 63 Acid-base properties of electronically excited states of organic molecules, 12, 131 Acids and bases, oxygen and nitrogen in aqueous solution, mechanisms of proton transfer between, 22, 113 Acids, reactions of aliphatic diazo compounds with, 5, 331 Acids, strong aqueous, protonation and solvation in, 13, 83 Activation, entropies of, and mechanisms of reactions in solution, 1, 1 Activation, heat capacities of, and their uses in mechanistic studies, 5, 121 Activation, volumes of, use for determining reaction mechanisms, 2, 93 Addition reactions, gas-phase radical directive effects in, 16, 51 Aliphatic diazo compounds, reactions with acids, 5, 331 Alkyl and analogous groups, static and dynamic stereochemistry of, 25, 1 Alkylcarbonium ions, spectroscopic observation in strong acid solutions, 4,305 Ambident conjugated systems, alternative protonation sites in, 11, 267 Ammonia, liquid, isotope exchange reactions of organic compounds in, 1, 156 Anions, organic, gas-phase reactions of, 24, 1 Antibiotics, p-lactam, the mechanisms of reactions of, 23, 165 Aqueous mixtures, kinetics of organic reactions in water and, 14, 203 Aromatic photosubstitution, nucleophilic, 11,225 Aromatic substitution, a quantitative treatment of directive effects in, 1, 35 Aromatic substitution reactions, hydrogen isotope effects in, 2, 163 Aromatic systems, planar and non-planar, 1, 203 Aryl halides and related compounds, photochemistry of, 20, 191 Arynes, mechanisms of formation and reactions at high temperatures, 6, 1 A-SE2reactions, developments in the study of, 6, 63 Base catalysis, general, of ester hydrolysis and related reactions, 5, 237 Basicity of unsaturated compounds, 4, 195 Bimolecular substitution reactions in protic and dipolar aprotic solvents, 5, 173 Bromination, electrophilic, of carbon-carbon double bonds: structure, solvent and mechanisms, 28, 207 I3C NMR spectroscopy in macromolecular systems of biochemical interest, 13, 279 Captodative effect, the, 26, 131 Carbanion reactions, ion-pairing effects in, 15, 153 Carbene chemistry, structure and mechanism in, 7, 163 Carbenes having aryl substituents, structure and reactivity of, 22, 311 Carbocation rearrangements, degenerate, 19, 223 Carbon atoms, energetic, reactions with organic compounds, 3,201 Carbon monoxide, reactivity of carbonium ions towards, 10,24 Carbonium ions (alkyl), spectroscopic observation in strong acid solutions, 4, 305 Carbonium ions, gaseous, from the decay of tritiated molecules, 8, 79 Carbonium ions, photochemistry of, 10, 129 Carbonium ions, reactivity towards carbon monoxide, 10, 29 Carbonyl compounds, reversible hydration of, 4, 1 235

236

CUMULATIVE INDEX OF TITLES

Carbonyl compounds, simple, enolisation and related reactions of, 18, 1 Carboxylic acids, tetrahedral intermediates derived from, spectroscopic detection and investigation of their properties, 21, 37 Catalysis by micelles, membranes and other aqueous aggregates as models of enzyme action, 17, 435 Catalysis, enzymatic, physical organic model systems and the problem of, 11, 1 Catalysis, general base and nucleophilic, of ester hydrolysis and related reactions, 5, 237 Catalysis, micellar, in organic reactions; kinetic and mechanistic implications, 8, 271 Catalysis, phase-transfer by quaternary ammonium salts, 15, 267 Cation radicals in solution, formation, properties and reactions of, 13, 155 Cation radicals, organic, in solution, kinetics and mechanisms of reactions of, 20, 55 Cations, vinyl, 9, 135 Chain molecules, intramolecular reactions of, 22, 1 Chain processes, free radical, in aliphatic systems involving an electron transfer reaction, 23,271 Charge density-NMR chemical shift correlation in organic ions, 11, 125 Chemically induced dynamic nuclear spin polarization and its applications, 10, 53 Chemiluminescence of organic compounds, 18, 187 Chirality and molecular recognition in monolayers at the air-water interface, 28, 45 CIDNP and its applications, 10, 53 Conduction, electrical, in organic solids, 16, 159 Configuration mixing model: a general approach to organic reactivity, 21, 99 Conformations of polypeptides, calcuations of, 6, 103 Conjugated, molecules, reactivity indices, in, 4, 73 Cross-interaction constants and transition-state structure in solution, 27, 57 Crown-ether complexes, stability and reactivity of, 17, 279 Crystallographic approaches to transition state structures, 29, 87 Cyclodextrins and other catalysts, the stabilization of transition states by, 29, 1 D20-H20 mixtures, protolytic processes in, 7, 259 Degenerate carbocation rearrangements, 19, 223 Diazo compounds, aliphatic, reactions with acids, 5, 331 Diffusion control and pre-association in nitrosation, nitration, and halogenation, 16, 1 Dimethyl sulphoxide, physical organic chemistry of reactions, in, 14, 133 Diolefin crystals, photodimerization and photopolymerization of, 30, 117 Dipolar aprotic and protic solvents, rates of bimolecular substitution reactions in, 5, 173 Directive effects in aromatic substitution, a quantitative treatment of, 1, 35 Directive effects in gas-phase radical addition reactions, 16, 51 Discovery of the mechanisms of enzyme action, 1947-1963, 21, 1 Displacement reactions, gas-phase nucleoph~lic,21, 197 Double bonds, carbon-carbon, electrophilic bromination of: structure, solvent and mechanism, 28, 171 Effective charge and transition-state structure in solution, 27, 1 Effective molarities of intramolecular reactions, 17, 183 Electrical conduction in organic solids, 16, 159 Electrochemical methods, study of reactive intermediates by, 19, 131 Electrochemistry, organic, structure and mechanism in, 12, 1 Electrode processes, physical parameters for the control of, 10, 155

CUMULATIVE INDEX OF TITLES

237

Electron donor-acceptor complexes, electron transfer in the thermal and photochemical activation of, in organic and organometallic reactions, 29, 185 Electron spin resonance, identification of organic free radicals by, 1, 284 Electron spin resonance studies of short-lived organic radicals, 5, 23 Electron storage and transfer in organic redox systems with multiple electrophores, 28, 1 Electron transfer in the thermal and photochemical activation of electron donoracceptor complexes in organic and organometallic reactions, 29, 185 Electron-transfer reaction, free radical chain processes in aliphatic systems involving an, 23,271 Electron-transfer reactions in organic chemistry, 18, 79 Electron-transfer, single, and nucleophilic substitution, 26, 1 Electronically excited molecules, structure of, 1, 365 Electronically excited states of organic molecules, acid-base properties of, 12, 131 Energetic tritium and carbon atoms, reactions of, with organic compounds, 2, 201 Enolisation of simple carbonyl compounds and related reactions, 18, 1 Entropies of activation and mechanisms of reactions in solution, 1, 1 Enzymatic catalysis, physical organic model systems and the problem of, 11, 1 Enzyme action, catalysis by micelles, membranes and other aqueous aggregates as models of, 17, 435 Enzyme action, discovery of the mechanisms of, 1947-1963,21, 1 Equilibrating systems, isotope effects on nmr spectra of, 23, 63 Equilibrium constants, NMR measurements of, as a function of temperature, 3, 187 Ester hydrolysis, general base and nucleophilic catalysis, 5, 237 Ester hydrolysis, neighbouring group participation by carbonyl groups in, 28, 171 Exchange reactions, hydrogen isotope, of organic compounds in liquid ammonia, 1, 156 Exchange reactions, oxygen isotope, of organic compounds, 2, 123 Excited complexes, chemistry of, 19, 1 Excited molecules, structure of electronically, 1, 365 Force-field methods, calculation of molecular structure and energy by, 13, 1 Free radical chain processes in aliphatic systems involving an electron-transfer reaction, 23, 271 Free radicals, identification by electron spin resonance, 1, 284 Free radicals and their reactions at low temperature using a rotating cryostat, study of, 8, 1 Gaseous carbonium ions from the decay of tritiated molecules, 8, 79 Gas-phase heterolysis, 3, 91 Gas-phase nucleophilic displacement reactions, 21, 197 Gas-phase pyrolysis of small-ring hydrocarbons, 4, 147 Gas-phase reactions of organic anions, 24, 1 General base and nucleophilic catalysis of ester hydrolysis and related reactions, 5, 237 H20-D20 mixtures, protolytic processes in, 7, 259 Halides, aryl, and related compounds, photochemistry of, 20, 191 Halogenation, nitrosation, and nitration, diffusion control and pre-association in, 16, 1 Heat capacities of activation and their uses in mechanistic studies, 5, 121 Heterolysis, gas-phase, 3, 91

238

CUMULATIVE INDEX OF TITLES

High-spin organic molecules and spin alignment in organic molecular assemblies, 26, 179 Homoaromaticity, 29,273 Hydrated electrons, reactions of, with organic compounds, 7, 115 Hydration, reversible, of carbonyl compounds, 4, 1 Hydride shifts and transfers, 24, 57 Hydrocarbons, small-ring, gas-phase pyrolysis of, 4, 147 Hydrogen atom abstraction from 0-H bonds, 9, 127 Hydrogen bonding and chemical reactivity, 26, 255 Hydrogen isotope effects in aromatic substitution reactions, 2, 163 Hydrogen isotope exchange reactions of organic compounds in liquid ammonia, 1, 156 Hydrolysis, ester, and related reactions, general base and nucleophilic catalysis of, 5, 237 Interface, the air-water, chriality and molecular recognition in monolayers at, 28, 45 Intermediates, reactive, study of, by electrochemical methods, 19, 131 Intermediates, tetrahedral, derived from carboxylic acids, spectroscopic detection and investigation of their properties, 21, 37 Intramolecular reactions, effective molarities for, 17, 183 Intramolecular reactions of chain molecules, 22, 1 Ionic dissociation of carbon-carbon a-bonds in hydrocarbons and the formation of authentic hydrocarbon salts, 30, 173 Ionization potentials, 4, 31 Ion-pairing effects in carbanion reactions, 15, 153 Ions, organic, charge density-NMR chemical shift correlations, 11, 125 Isomerization, permutational, of pentavalent phosphorus compounds, 9, 25 Isotope effects, hydrogen, in aromatic substitution reactions, 2, 163 Isotope effects, magnetic, magnetic field effects and, on the products of organic reactions, 20, 1 Isotope effects on nmr spectra of equilibrating systems, 23,63 Isotope effects, steric, experiments on the nature of, 10, 1 Isotope exchange reactions, hydrogen, of organic compounds in liquid ammonia, 1, 150 Isotope exchange reactions, oxygen, of organic compounds, 3, 123 Isotopes and organic reaction mechanisms, 2 1 Kinetics and mechanisms of reactions of organic cation radicals in solution, 20,55 Kinetics of organic reactions in water and aqueous mixtures, 14, 203 Kinetics, reaction, polarography and, 5, 1 P-Lactam antibiotics, the mechanisms of reactions of, 23, 165 Least nuclear motion, principle of, 15, 1 Macrocycles and other concave structures, acid-base behaviour in, 30,63 Macromolecular systems of biochemical interst, 13CNMR spectroscopy in, 13 279 Magnetic field and magnetic isotope effects on the products of organic reactions, 20, 1 Mass spectrometry, mechanisms and structure in: a comparison with other chemical processes, 8, 152 Matrix infrared spectroscopy of intermediates with low coordinated carbon, silicon and germanium atoms, 30, 1

CUMULATIVE INDEX OF TITLES

239

Mechanism and structure in carbene chemistry, 7, 153 Mechanism and structure in mass spectrometry: a comparison with other chemical processes, 8, 152 Mechanism and structure in organic electrochemistry, 12, 1 Mechanisms and reactivity in reactions of organic oxyacids of sulphur and their anhydrides, 17, 65 Mechanisms, nitrosation, 19, 381 Mechanisms of proton transfer between oxygen and nitrogen acids and bases in aqueous solutions, 22, 113 Mechanisms of reaction in solution, entropies of activation and, 1, 1 Mechanisms of reaction of p-lactam antibiotics, 23, 165 Mechanisms of solvolytic reactions, medium effects on the rates and, 14, 10 Mechanisms, organic reaction, isotopes and, 2, 1 Mechanistic applications of the reactivity-selectivity principle, 14, 69 Mechanistic studies, heat capacities of activation and their use, 5, 121 Medium effects on the rates and mechanisms of solvolytic reactions, 14, 1 Meisenheimer complexes, 7, 211 Metal complexes, the nucleophilicity of towards organic molecules, 23, 1 Methyl transfer reactions, 16, 87 Micellar catlysis in organic reactions: kinetic and mechanistic implications, 8, 271 Micelles, aqueous, and similar assemblies, organic reactivity in, 22, 213 Micelles, membranes and other aqueous aggregates, catalysis by, as models of enzyme action, 17, 435. Molecular recognition, chirality and, in monolayers at the air-water interface, 28,45 Molecular structure and energy, calculation of, by force-field methods, 13, 1 Neighbouring group participation by carbonyl groups in ester hydrolysis, 28, 171 Nitration, nitrosation, and halogenation, diffusion control and pre-association in, 16, 1 Nitrosation mechanisms, 19,381 Nitrosation, nitration, and halogenation, diffusion control and pre-association in, 16, 1 NMR chemical shift-charge density correlations, 11, 125 NMR measurements of reaction velocities and equilibrium constants as a function of temperature, 3, 187 NMR spectra of equilibrating systems, isotope effects on, 23, 63 NMR spectroscopy, 13C, in macromolecular systems of biochemical interest, 13,279 Non-planar and planar aromatic systems, 1, 203 Norbornyl cation: reappraisal of structure, 11, 179 Nuclear magnetic relaxation, recent problems and progress, 16, 239 Nuclear magnetic resonance, see NMR Nuclear motion, principle of least, 15, 1 Nuclear motion, the principle of least, and the theory of stereoelectronic control, 24, 113 Nucleophilic aromatic photosubstitution, 11, 225 Nucleophilic catalysis of ester hydrolysis and related reactions, 5, 237 Nucleophilic displacement reactions, gas-phase, 21, 197 Nucleophilicity of metal complexes towards organic molecules, 23, 1 Nucleophilic substitution in phosphate esters, mechanism and catalysis of, 25,99 Nucleophilic substitution, single electron transfer and, 26, 1 Nucleophilic vinylic substitution, 7, 1

240

CUMULATIVE INDEX OF TITLES

0-H bonds, hydrogen atom abstraction from, 9, 127 Oxyacids of sulphur and their anhydrides, mechanisms and reactivity in reactions of organic, 17,65 Oxygen isotope exchange reactions of organic compounds, 3, 123 Perchloro-organic chemistry: structure, spectroscopy and reaction pathways, 25, 267 Permutational isomerization of pentavalent phosphorus compounds, 9, 25 Phase-transfer catalysis by quaternary ammonium salts, 15, 267 Phosphate esters, mechanism and catalysis of nucleophilic substitution in, 25, 99 Phosphorus compounds, pentavalent, turnstile rearrangement and pseudorotation in permutational isomerization, 9,25 Photochemistry of aryl halides and related compounds, 20, 191 Photochemistry of carbonium ions, 9, 129 Photodimerization and photopolymerization of diolefin crystals, 30, 117 Photosubstitution, nucleophilic aromatic, 11,225 Planar and non-planar aromatic systems, 1, 203 Polarizability, molecular refractivity and, 3, 1 Polarography and reaction kinetics, 5, 1 Polypeptides, calculations of conformations of, 6, 103 Pre-association, diffusion control and, in nitrosation, nitration, and halogenation, 16, 1 Principle of non-perfect synchronization, 27, 119 Products of organic reactions, magnetic field and magnetic isotope effects on, 30, 1 Protic and dipolar aprotic solvents, rates of bimolecular substutition reactions in, 5, 173 Protolytic processes in H20-D20 mixtures, 7, 259 Protonation and solvation in strong aqueous acids, 13, 83 Protonation sites in ambident conjugated systems, 11, 267 Proton transfer between oxygen and nitrogen acids and bases in aqueous solution, mechanisms of, 22, 113 Pseudorotation in isomerization of pentavalent phosphorus compounds, 9, 25 Pyrolysis, gas-phase, of small-ring hydrocarbons, 4, 147 Radiation techniques, application to the study of organic radicals, 12, 223 Radical addition reactions, gas-phase, directive effects in, 16, 51 Radicals, cation in solution, formation, properties and reactions of, 13, 155 Radicals, organic application of radiation techniques, 12, 223 Radicals, organic cation, in solution kinetics and mechanisms of reaction of, 20, 55 Radicals, organic free, identification by electron spin resonance, 1, 284 Radicals, short-lived organic, electron spin resonanace studies of, 5, 53 Rates and mechanisms of solvolytic reactions, medium effects on, 14, 1 Reaction kinetics, polarography and, 5, 1 Reaction mechanisms in solution, entropies of activation and, 1, 1 Reaction mechanisms, use of volumes of activation for determining, 2, 93 Reaction velocities and equilibrium constants, NMR measurements of, as a function of temperature, 3, 187 Reactions in dimethyl sulphoxide, physical organic chemistry of, 14, 133 Reactions of hydrated electrons with organic compounds, 7, 115 Reactive intermediates, study of, by electrochemical methods, 19, 131 Reactivity indices in conjugated molecules, 4, 73 Reactivity, organic, a general approach to: the configuration mixing model, 21, 99 Reactivity-selectivity principle and its mechanistic applications, 14, 69

CUMULATIVE INDEX OF TITLES

241

Rearrangements, degenerate carbocation, 19, 223 Redox systems, organic, with multiple electrophores, electron storage and transfer in, 28, 1 Refractivity, molecular, and polarizability, 3, 1 Relaxation, nuclear magnetic, recent problems and progress, 16, 239 Selectivity of solvolyses in aqueous alcohols and related mixtures, solvent-induced changes in, 27, 239 Short-lived organic radicals, electron spin resonance studies of, 5, 53 Small-ring hydrocarbons, gas-phase pyrolysis of, 4, 147 Solid-state chemistry, topochemical phenomena in, 15, 63 Solids, organic, electrical conduction in, 16, 159 Solutions, reactions in, entropies of activation and mechanisms, 1, 1 Solvation and protonation in strong aqueous acids, 13, 83 Solvent-induced changes in the selectivity of solvolyses in aqueous alcohols and related mixtures, 27, 239 Solvent, protic and dipolar aprotic, rates of bimolecular substitution-reactions in, 5, 173 Solvolytic reactions, medium effects on the rates and mechanisms of, 14, 1 Spectroscopic detection of tetrahedral intermediates derived from carboxylic acids and the investigation of their properties, 21, 37 Spectroscopic observations of alkylcarbonium ions in strong acid solutions, 4, 305 Spectrocopy, I3C NMR, in macromolecular systems of biochemical interest, 13, 279 Spin alignment, in organic molecular assemblies, high-spin organic molecules and, 26, 179 Spin trapping, 17, 1 Stability and reactivity of crown-ether complexes, 17, 279 Stereochemistry, static and dynamic, of alkyl and analogous groups, 25, 1 Stereoelectronic control, the principle of least nuclear motion and the theory of, 24, 113 Stereoselection in elementary steps of organic reactions, 6, 185 Steric isotope effects, experiments on the nature of, 10, 1 Structure and mechanisms in carbene chemistry, 7, 153 Structure and mechanism in organic electrochemistry, 12, 1 Structure and reactivity of carbenes having aryl substituents, 22, 311 Structure of electronically excited molecules, 1 , 365 Substitution, aromatic, a quantitative treatment of directive effects in, 1 , 35 Substitution, nucleophilic vinylic, 7, 1 Substitution reactions, aromatic, hydrogen isotope effects in, 2, 163 Substitution reactions, bimolecular, in protic and dipolar aprotic solvents, 5, 173 Sulphur, organic oxyacids of, and their anhydrides, mechanisms and reactivity in reactions of, 17, 65 Superacid systems, 9, 1 Temperature, NMR measurements of reaction velocities and equilibrium constants as a function of, 3, 187 Tetrahedral intermediates derived from carboxylic acids, specttoscopic detection and the investigation of their properties, 21, 37 Topochemical phenomena in solid-state chemistry, 15, 63 Transition states, the stabilization of by cyclodextrins and other catalysts, 29, 1 Transition-state structure in solution, cross-interaction constants and, 27, 57 Transition-state structure in solution, effective charge and, 27, 1

242

CUMULATIVE INDEX OF TITLES

Transition state structures, crystallographic approaches to, 29, 87 Transition-state theory revisited, 28, 139 Tritiated molecules, gaseous carbonium ions from the decay of, 8, 79 Tritium atoms, energetic reactions with organic compounds, 2, 201 Turnstile rearrangements in isomerization of pentavalent phosphorus compounds, 9, 25

Unsaturated compounds, basicity of, 4, 195 Vinyl cations, 9, 185 Vinylic substitution, nucleophilic, 7, 1 Volumes of activation, use of, for determining reaction mechanisms, 2, 93 Water and aqueous mixtures, kinetics of organic reactions in, 14, 203

This Page Intentionally Left Blank