Advances in Physical Organic Chemistry, Volume 12

Advances in Physical Organic Chemistry

This Page Intentionally Left Blank

Advances in Physical Organic Chemistry Volume 12 Edited by

V. Gold Department of Chemistry King’s College University of London Associate Editor

D. Bethel1 The Robert Robinson Laboratories University of Liverpool England

1976

@

Academic Press London New York

San Francisco

A Subsidiasy of Harcourt Brace Jovanovich, Publishers

ACADEMIC PRESS INC. (LONDON) LTD 24-28 Oval Road, London NWI

U.S. Edition published by ACADEMIC PRESS INC. 111 Fifth Avenue, New York, New York 10003, U.S.A.

Copyright 0 1976 By Academic Press Inc. (London) Ltd

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

Library ofcongress Catalog Card Number :62-22 125 ISBN 0- 1 2-03 35 1 2-3

PRINTED IN GREAT BRITAIN B Y WILLIAM CLOWES & SONS LIMITED LONDON, COLCHESTER AND BECCLES

Contributors to Volume 12

L. Eberson, Division of Organic Chemistry, Chemical Center, University of Lund, Sweden J. F. Ireland, Department of Chemistry, The University of St. Andrews, Scotland

P. Neta, Radiation Research Laboratories and Department of Chemistry, Mellon Institute of Science, Carnegie-Mellon University, Pittsburgh, Pennsylvania, U.S.A.

K. Nyberg, Division of Organic Chemistry, Chemical Center, University of L u n d , Sweden P. A. H. Wyatt, Department of Chemistry, The University of St. Andrews, Scotland

V

This Page Intentionally Left Blank

Contents

Contributors to Volume 1 2

.

.

v

. .

2 4

Structure and Mechanism in Organic Electrochemistry - L. Eberson and K . N y berg

. 1. Introduction The Experimental Situation . 2. 3. Phenomenological Classification of Organic Electrode Reactions . 4. Mechanistic Problems . . 5 . Direct and Indirect Electrode Reactions Effect of Concentration Gradients . 6. 7. Nature of the Electroactive Species . . 8. Reaction Sequence Role of Adsorption . 9. The Electron Transfer Process . 10. . 11. Structure and Reactivity . Influence of the Electrode Material 12.

.

8 19 26 . 29 . 40 . 71 . 87 . 100 . 106 .lll

. .

Acid-Base Properties of Electronically Excited States of Organic Molecules - J. F. Ireland and P. A . H. Wyatt

1. Introduction . 2. Experimental Methods. . 3. Kinetics and Equilibria of Excited State Protonation Reactions . vii

. 132 . 136 . 144

...

Vlll

CONTENTS

4. Survey of Experimental Results . 5. Excited States and Acidity Scales 6. Applications .

. 158 .

.207 .212

Application of Radiation Techniques to the Study of Organic Radicals - P. Netu

1. Introduction . 2. Techniques of Radiation Chemistry . 3. Reactions of Organic Compounds with Transients from Water . 4. Optical Absorption Spectra of Organic Radicals . 5 . Electron Spin Resonance Studies . 6. Acid-Base Equilibria of Organic Radicals . 7. Kinetics and Mechanisms of Radical Reactions . 8. Concluding Remarks .

. 224 . 225 . 230 . 243 . 247 . 253 . 270 . 289

Structure and Mechanism in Organic Electrochemistry L. EBERSON and K. NYBERG Division of Organic Chemistry, Chemical Center, University of Lund, P.O.Box 740, S-220 07 Lund, Sweden

1. Introduction 2. The Experimental Situation

2

.

3. Phenomenological Classification of Organic Electrode Reactions . Electron Transfer . Conversion of Functional Groups . Substitution Addition . Elimination Coupling Cleavage . Miscellaneous Reactions 4. Mechanistic Problems . 5. Direct and Indirect Electrode Reactions . 6. Effect of Concentration Gradients . Bimolecular Reactions between Intermediates Locally High Acid or Base Concentrations Near the Electrode . Surface Concentration Effects . 7. Nature of the Electroactive Species . Solvent and Supporting Electrolytes . Chemical Modifications of the Substrate before Electron Transfer Some Case Studies . 8. Reaction Sequence . . Reactions of Radical Ion versus Doubly Charged Ion Disproportionation versus ECE Mechanisms . Coupling; via Radical Ions, Neutral Radicals, "or Reaction . between Radical Ion and Substrate? . 9. Role ofAdsorption . Concentration Effects in the EX . 1

4

8 9 10 11 13 15 16 18 19 19 26 29 30 34 40 40 41 47 52 71 73 76 82 87 89

2

L. EBERSON AND K. NYBERG

Orientation Effects Induced by the Electrode Surface Competing Reactions; the Role of Adsorption . Adsorption versus E 1 12 -values 10. The Electron Transfer Process 11. Structure and Reactivity . Kinetics of Electron Transfer at Electrodes . Half-wave Potentials as Reactivity Indices . Role of Electrochemical Parameters in Physical Organic Chemistry 12. Influence of the Electrode Material References .

.

.

. . . . . . *

95 98 99 100 106 106 108 111 111 116

1. INTRODUCTION The realm of phenomena connected with the interaction between organic compounds and electricity has vastly increased during the last few decades. Not only has there been a vigorous expansion of the classical areas of organic electroanalysis and its applications to physical organic chemistry (see, for example, Zuman 1967, 1969; Zuman and Perrin, 1969) and electrosynthesis and related mechanisms (see reviews and books: Adams, 1969, 1969a; Anderson et al., 1969; Baizer, 1969; Baizer et al., 1973; Baizer and Petrovich, 1970; Beck, 1972; Bewick and Pletcher, 1970, 1971; Brago et al., 1971; Casanova and Eberson, 1973; Chang et al., 1971; Conway and Vijh, 1967; Eberson, 1968; Eberson and Nyberg, 1973; Eberson and Schafer, 1971; Elving and Pullman, 1961; Fichter, 1942; Fleischmann and Pletcher, 1969a, 196913, 1971, 1973; Fry, 1972, 1972a; Gilde, 1972; Haufe and Beck, 1970; Hoijtink, 1963; Humffray, 1973; Lehmkuhl, 1973; Lund, 1970, 1970a, 1971; Mann and Barnes, 1970; McKillop and Korinek, 1971, 1972; Nagase, 1967; Peover, 1971; Popp and Schultz, 1962; Rifi and Covitz, 1974; Robertson, 1973, 1974; Swann, 1956; Tomilov, 1961; Tomilov et al., 1972; Utley, 1969, 1970; Weinberg and Weinberg, 1968; Wawzonek, 1967, 1971), but, in addition, a host of new fields has opened up during this period. These include research on such practical applications of organic electrochemistry as fuel cells (Piersma and Gileadi, 1966; Vielstich, 1965), high-energy batteries, corrosion inhibitors, electrodeposition of metals (Isserlis, 197 1) and paint (Cooke et al., 197 l), electrochemiluminescence (Hercules, 1971; McCapra, 1973), organic semiconductors (Ferraris et al., 1973; Gutmann and Lyons, 1967; Wudl and Southwick, 1974), liquid crystals for electronic display, organic photoelectric materials

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

3

(Daniels, 1972; Kuwana, 1966), and membranes. Mention should also be made of the fascinating perspective raised by Little’s prediction (Little, 1964, 1967) that it should be possible to synthesize organic materials which are superconducting at or even above room temperature (Coleman et al., 1973). Connected with these applications is fundamental research dealing with solid state chemistry and physics, homogeneous and heterogeneous electron transfer involving organic molecules, the structure of the electrified interface, adsorption, and electrocatalysis. In view of the interdisciplinary nature of this type of research, it is hardly surprising that a fair amount of controversy and misunderstanding has created some degree of confusion as to which meaning one should attach to the concept of mechanism when applied t o organic electrode processes. On the one side physically oriented chemists have been preoccupied with transport mechanisms for molecules to and from electrodes, adsorption, structure of the electrified interface, quantum aspects of electron transfer and electrode kinetics, whereas on the other side specialists in organic chemistry have utilized electrolytic methods for the synthesis of organic compounds and in most cases have been content to write down mechanisms in the usual manner of physical organic chemistry without bothering too much about the special features of electrochemical reactions. The fact that two new types of intermediates-radical anions and cations (the last ones to be found in organic chemistry?)-have been introduced into organic chemistry largely through the developments in organic electrochemistry has added to the complexity. Only now are we beginning to get an idea of the kind of chemistry that these species can undergo in homogeneous media (reviews: radical ions in general: Miller, 1971; Russell and Norris, 1973; radical cations: Bard et al., 1976; radical anions: Dorfman, 1970; Holy, 1974; Szwarc, 1968, 1969) and many electrode reaction mechanisms had previously been postulated essentially without a body of firm knowledge of radical ion chemistry. A contributing factor to this state of affairs has n o doubt been the relative neglect of electron transfer processes in organic chemistry, except for a few cases (see, for example: Andrulis et al., 1966; Bilevich and Okhlobystin, 1968; Dessau et al., 1970; Kochi et al., 1973; Ledwith and Russell, 1974; Norman et al., 1971, 1973; Sheldon and Kochi, 1973; Trahanovsky et al., 1974). This contrasts strongly with the situation in mechanistic inorganic chemistry (Basolo and Pearson, 1967; Reynolds and Lumry, 1966) and biochemistry (Bishop, 1971; Rabinowitch and Govindjee, 1969).

4

L. EBERSON AND K. NYBERG

This review will attempt to cover the most important aspects of structure-reactivity relationships and mechanisms in organic electrochemistry proper, i.e. those aspects relating to electrochemical reactions of organic compounds. The main emphasis will be on selective processes and thus much of the work on fuel cell electrochemistry (Piersma and Gileadi, 1966; Vielstich, 1965), with its aim to oxidize organic compounds rapidly and completely to water and carbon dioxide, will be left out. We have not aimed to provide an exhaustive review and would therefore like to apologize to those of our colleagues who will justifiably feel that some of their studies relevant to the subject have been omitted. Our only excuse is that organic electrochemistry has now grown so large that not even a series of volumes can cover the field exhaustively.

2. THE EXPERIMENTAL SITUATION The experimental prerequisites for running an electrochemical experiment with an organic system are simple: a solvent, capable of dissolving both the organic substrate and an electrolyte (the so-called supporting electrolyte) added to give a reasonably highly conducting' medium (the electrolyte solution), two electrodes (the anode and the cathode), made from metallic materials but sometimes also from semiconducting ones, and a source of electric power to apply across the electrolyte solution via the two electrodes. In order to follow the course of the electrochemical reaction of interest, provision is made for recording the potential of the working electrode (or rather the change in potential, since absolute potentials across interfaces cannot even in principle be measured) by a reference electrode (e.g. a calomel electrode) the tip of which is placed as near the working electrode as possible. The potential difference between the working and reference electrodes is measured by a voltmeter with a very high internal resistance. In this electrode arrangement, the third electrode is denoted the auxiliary electrode. The circuit connecting the For laboratory experiments, the requirement that the electrolyte solution should possess a high conductivity is not so crucial and hence one can use solvents with dielectric constants as low as between 5 and 10, e.g. acetic acid or methylene chloride. On a large scale, the problem of minimizing the voltage across the cell is very important and has to be tackled either by special cell designs (Beck, 1972, 1974; Eberson, 1974; Eberson et al., 1973; Fleischmann, 1974; Goodridge, 1974) or other means (Eberson and Helgee, 1974).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

5

electrolytic cell is also provided with an ammeter to measure the current flowing through the electrolyte. Let us assume that the electrode reaction under study is to take place at the anode, at which electrons flow f r o m the electrolyte solution t o the electrode (anodic oxidation), and hence that the potential difference between the anode and the reference electrode is the anode potential, E,. Also, let the first experiment be run with only solvent and supporting electrolyte (SSE) present in the electrolyte solution. The anode potential is now gradually changed toward more positive values by increasing the potential applied across the electrolyte solution (of course the cathode potential, E,, will move towards more negative potentials but this is of no other interest than to bring about the cathodic process necessary to keep up the electroneutrality principle), and at the same time readings of the current, I , are taken from the ammeter. Alternatively, the whole procedure can easily be adapted for automatic recording of the I versus E, curve. Such a voltammetric curve is shown in Fig. 1, curve

Figure 1. Voltammetric curves (for explanation, see text).

A; actually i, the current density (the current passed per unit surface area of the working electrode) is the quantity recorded, since it is current density which is the correct measure of the rate of an electrochemical process. Note, however, that the determination of It is often necessary to use a divided cell, in which a membrane, non-permeable to substrate and product but permeable to ions (usually cation exchange membranes are used), is inserted in the cell in order to protect substrate and/or product from reacting at the auxiliary electrode. Complications due to this arrangement are entirely of a practical nature.

6

L. EBERSON AND K. NYBERG

actual as opposed t o geometric surface areas is not a trivial problem' (Damaskin et al., 1971) except in the case of liquid metal electrodes such as mercury. Hence one often finds I used as a measure of electrochemical rate at solid electrodes. This is justifiable as long as the same experimental set-up and electrode pretreatment is used for all the electrochemical experiments t o be compared. The voltammetric curve in our imaginary experiment displays a wide region (between 0 and + 2.0 V versus the calomel electrode) in which no current, except for a very minute residual current, due mostly t o non-faradaic processes (i.e. processes not involving electrochemical transformations), flows through the electrolyte solution. Obviously, none of the possible candidates in the system, the solvent molecule or the anion of the supporting electrolyte, is electroactive in this potential range. However, at anode potentials above + 2.0 V, current starts to flow, normally with an exponential increase as long as no other factor controls the rate of the process. In the region above 2.0 V, either the solvent or the anion of the supporting electrolyte (or both) undergoes oxidation with formation of one or several new compounds. In the next experiment we add a small amount, e.g. an initial concentration of 0.001 M , of the organic substrate and record a second voltammetric curve ( B in Fig. 1). Still as a simulated situation, let us assume that the substrate is electroactive at a lower potential than that of the SSE alone. In such a case, the voltammetric curve will have a sigmoid shape, first with an exponential increase of i and then a gradual flattening out to a plateau value, ilim, at which the rate of transport of substrate molecules by diffusion to the electrode is rate-limiting. This is the region of diffusion control of the rate; under properly controlled conditions ilim is linearly related to the concentration of the electroactive compound. The potential at i l i m /2,El l 2 , is an important parameter in that it can be used as a relative measure of the oxidizability of different electroactive compounds (see Section 11). Voltammetric curves with plateaus are obtained only in stirred electrolyte solutions, whereas in unstirred solutions curves exhibiting a potential peak are obtained (curve C in Fig. 1) due to depletion of the electroactive species in the layer near the electrode. The peak potential, E,, or any other suitably defined potential on the peak voltammetric curve can be used in the same way as El 1 2 . The ratio between actual and geometric surface area is commonly called the roughness factor (see further Section 12).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

7

It should be fairly obvious by now that at a sufficiently high substrate concentration the voltammetric curve should look something like curve D in Fig. 1, almost the same shape as curve A but displaced toward lower potentials, since again our equipment does not allow for a current high enough to reach the plateau value. This situation is very common in electrosynthetic experiments. Also, in order to ensure that n o mixing of the electrode processes corresponding to curves A and D, respectively, should take place, the actual synthetic experiment should preferably be carried out at a constant potential, chosen somewhere near the beginning of or on the plateau of curve B. Then the electrolysis will take place at an initially high current density which exponentially decreases to a value near zero when all the substrate has been consumed. For electrolysis at a potential on the plateau, eqn (1) gives the current change with time t:

. = 21im . 10-(0.43DA t/V6 )

2

(1)

where D is the diffusion coefficient, A the electrode area, V the volume of the electrolyte solution, and 6 the thickness of the Nernst diffusion layer. This is assumed to be a stagnant layer near the electrode to the boundary of which molecules are brought by convection (stirring) and through which molecules travel by diffision up to the electrode. It is important to remember that the thickness of the Nernst diffusion layer is not constant under all conditions. It decreases with increasing stirring rate and becomes especially small if gas evolution takes place at the electrode, in which case the Nernst layer is stirred “from inside”. A very useful extension of the voltammetric technique is cyclic voltammetry (Adams, 1969; Cauquis and Parker, 1973) in which one scans the potential of the working electrode in an unstirred electrolyte solution in the anodic (cathodic) direction and records one or several peaks due to oxidation (reduction) of the substrate. At some suitable potential, the direction of the scan is reversed and peaks due to reduction (oxidation) of intermediates and/or products formed during the forward scan are observed. In the simplest case a linear increase (decrease) of the potential with time is employed (triangular cyclic voltammetry) with scan rates in the range 0-01- 1000 V s - l . It should be noted that cyclic voltammetry at scan rates above 1 V s-l requires the use of a differential cell to reduce the residual current due to charging of the electrified interface (see, for example, Peover and White, 1967). The theory of cyclic voltammetry has been

8

L. EBERSON AND K. NYBERG

thoroughly worked out (Nicholson and Shain, 1964, 1965). Using this technique one can rapidly obtain large amounts of useful information concerning the initial electron transfer step and subsequent reactions of the intermediate(s) formed (for an example, see Section 8). Another simple and useful technique is voltammetry at a rotating disc electrode (Adams, 1969). A more complicated version of this electrode is the rotating ring disc electrode (Adams, 1969).

3. PHENOMENOLOGICAL CLASSIFICATION OF ORGANIC ELECTRODE REACTIONS Electrode processes are conveniently classified according to the nature of the final product’ and its formal mode of formation, since then the interplay between nucleophile(s) or electrophile(s) , substrate, and loss or addition of electron(s) is best expressed. It is upon our ingenuity to choose the correct combination of electrolyte components that the practical success of an electrochemical reaction rests, and therefore the rather formalized classification system t o be outlined and exemplified below is the logical point of departure into the maze of mechanistic intricacies of electrode processes. From this point of view we can distinguish between the types of reactions listed below, Nu- and E + denoting a negatively and positively charged nucleophile and electrophileY2respectively, and thus R-E and R-Nu symbolizing respectively substrates for the following anodic and cathodic transformations: Pure electron transfer Conversion of functional groups Substitution Addition Elimination Coupling Cleavage. In the following we shall give a brief description, together with representative examples, of these reaction types. In all cases, both cathodic and anodic variants are known. It is only rarely that a single compound is formed in an electrochemical reaction, and hence “product” should actually read “desired product” or “product of interest”. The charges are introduced for simplicity only; should it be necessary to denote charged substrates, the symbols R-E- and R-Nu+ can be used.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

9

Electron Transfer Anodic and cathodic electron transfer is the elementary act in all electrochemical reactions of interest here (atom transfers are possible too, especially in electrocatalytic reactions: Piersma and Gileadi, 1966) and results in the formation of radical ions from neutral molecules [(Z) and (3)] and neutral radicals from charged species [(4) and (5)]. In the overwhelming majority of cases, radical ions R-E + R-E * + R-Nu + e -

+ e-

+ R-Nu * -

R - N ~ + e-

+ R-NU

(2) (3)

(5)

and neutral radicals are consumed as they are formed in very fast chemical follow-up reactions. Only relatively stable species can be detected experimentally e.g. by e.s.r. spectroscopy (Adams, 1966; Cauquis and Parker, 1973), cyclic voltammetry and other electrochemical techniques (Adams, 1966; Baizer et al., 1973; for recent applications, see: Bechgaard and Parker, 1972; Bechgaard et al., 1972; Geiger, 1973; Hammerich and Parker, 1973, 1974; Hammerich et al., 1972; Longchamp et al., 1974; R o n l h and Parker, 1974; Svensmark Jensen and Parker, 1974; Svanholm and Parker, 1972; Svanholm et al., 1974), or optical methods (Gruver and Kuwana, 1972; Cauquis and Parker, 1973). In some cases stable radical ion salts have been isolated in the solid form, e.g. the radical cation of dibenzodioxan (Cauquis and Maurey, 1968), thianthrene (Parker and Hammerich, 1972), and 9,lO-di-p-anisylanthracene(Hammerich and Parker, 1972). Certain experimental conditions greatly increase the stability of radical cations. These include the use of trifluoroacetic acid as a solvent or cosolvent (Bechgaard e t al., 1972; Bechgaard and Parker, 1972; Hammerich et al., 1972; Hammerich and Parker, 1974; RonlAn and Parker, 1974; Svanholm and Parker, 1972; Svanholm et al., 1974) or the exclusion of traces of nucleophiles, especially water, from the electrolyte solution by adding neutral alumina directly into the cell during the run (Hammerich and Parker, 1973, 1974). The more obvious way to stabilize intermediates, to decrease the temperature, has also been successfully tried (Byrd et al., 1972; Van Duyne and Reilley, 1972, 1972a, 197 2b).

10

L. EBERSON AND K. NYBERG

Equations (2)- ( 5 ) depict single one-electron transfers to give intermediates; two successive one-electron transfers are also known to give relatively stable intermediates if their structure is suitably chosen and the experimental conditions are favourable. Thus, relatively stable solutions of certain aromatic dications can be prepared [e.g. those derived from tetra-p-anisylethylene (Parker et al., 1969; Bard and Phelps, 1970; Stuart and Ohnesorge, 1971), tetraphenylethylene (Svanholm et al., 19 74a), hexamethoxytriphenylene (Bechgaard and Parker, 1972), 9,lO-disubstituted anthracenes (Hammerich and Parker, 1973, 1974), thianthrene and its 2,3,7,8tetramethoxy derivative (Glass et al., 19 73; Hammerich and Parker, 1973), biphenylene ( R o n l h and Parker, 1974), and a model compound of a-tocopherol (Svanholm et al., 1 9 7 4 ) l . A tripositive radical cation has even been observed in trifluoroacetic acid at - 5 O O C (Bechgaard and Parker, 1972). Dications from highly substituted N,N-dimethylaminoalkenes are also easily available (Fritsch et al., 1970; Kuwata and Geske, 1964). Certain aromatic hydrocarbons, such as 9,10-diphenylanthracene, give relatively stable radicals and cation radicals upon electrochemical reduction and oxidation, respectively. If one arranges to have the radical ions from both processes mixed, either by normal DC electrolysis in a suitably designed cell or by using an alternating current for the electrolysis, the phenomenon of electrochemiluminescence appears (Hercules, 1971; McCapra, 1973).

Conversion of Functional Groups This type of reaction is similar to its chemical counterpart, in that a functional group is reduced or oxidized to another one, e.g. as in eqns (6)-( 8): R-NOz

- +4e-

+4H+

R-COOH

RCONHz

R-NHOH

+2e-

+2H+

R-NHZ

+4eRCHzOH +4H+

__+

+4e-

+4H+

RCHzNHz

(6)

(7) (8)

It is difficult to give a general formula covering all ele ctr oche mical functional group conversions, but these examples should be suf-

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

11

ficient to illustrate such processes and their characteristics. Mechanistically, they present the same kind of problems as those encountered in the types of reactions t o be discussed in the remainder of this Section; in addition, they present ordinary mechanistic problems since suitably substituted molecules can undergo intramolecular electrophilic and/or nucleophilic reactions utilizing otherwise unstable, electrogenerated functional groups, giving rise t o heterocyclic compounds (Lund, 1970a).

Substitution

A general expression for anodic substitution reactions is shown in eqn (9). Here E is often hydrogen, but can also be another atom or group, e.g. t-butyl, OCH3, or COO- (this is the carbonium ion pathway of the well-known Kolbe reaction: Eberson, 1968, 1973a). Examples of anodic substitution reactions include acetoxylation R-E I- NU-

[eqn

(lo)] of

-+

R-NU + E+ + Ze-

(9)

aromatic compounds in the ring (Eberson and Nyberg, R-H + ACO-

HOAc

Pt or C

R-OAC+ H+ + Ze-

R = Ar, ArCHz, RCON(CH3)CH2 and

\

(10)

I

F=C--CH2

1964, 1966, 1973; Ross et al., 1964) or the &-position, if available (Eberson, 1967; Magnusson et al., 1971, Ross et al., 1964), of amides a t o nitrogen (Ross et al., 1966), and of olefins in the allylic position (Courbis and Guillemonat, 1966; Shono and Kosaka, 1968; Shono and Ikeda, 1972), and cyanation [eqn ( l l ) ]of aromatic compounds MeOH or CH3CN

Ar-E + CN-

Pt

+

Ar-CN + E+ + 2e-

(11)

E = H, CH3O

in the ring (Koyama et al., 1965, 1966; Eberson and Nilsson, 1968; Andreades and Zahnow, 1969) or of tertiary amines (Y to the nitrogen (Andreades and Zahnow, 1969). A case of an anodic substitution reaction involving a neutral nucleophile is acetamidation [eqn (12 ) ] , an important process in

12

L. EBERSON AND K. NYBERG

R-I + CH3CN

-

R-NHCOCH3

H2 0

RNHCOCH3

+ 412 + e-

(13)

organic electrochemistry due to the excellent solvent properties of acetonitrile and its resistance towards both oxidation and reduction (Billon, 1960). Here E can be hydrogen (Eberson and Nyberg, 1966a), COO- (Eberson and Nyberg, 1964a; Kornprobst et al., 1968, 1970; Muck and Wilson, 1970; Thomas, 1971), t-butyl (Popp and Reitz, 1972), and iodine (Miller and Hoffmann, 1967; Laurent e t al., 1973) in a formal one-electron transfer [eqn (13)]. Hydrogen in almost all kinds of situation can be substituted; non-activated positions as in saturated hydrocarbons (Clark et al., 1973; Fleischmann and Pletcher, 1968; Koch and Miller, 1973) and remote positions in esters of fatty acids (Miller and Ramachandran, 1974), allylic positions in olefins (Clark et al., 1972), a-positions in alkylaromatics (Eberson and Olofsson, 1969) and ring positions in aromatic compounds (Hammerich and Parker, 19 74a; Popp and Reitz, 1972; Matsuda et al., 1973). Cathodic substitution reactions conform to the general eqn (14). R-Nu + E+ + 2e-

+ R-E

+ Nu-

(14)

Here E+ in most cases is H+ or an equivalent proton donor, but it can also be a neutral molecule, such as COz or methylFromide. Nu is a leaving group of some kind (halogen, RSO, RSOz , NR3, etc.). The most important cathodic substitution reaction is the replacement by hydrogen of a substituent, especially halogen (Casanova and Eberson, 1973). Compounds containing 'onium (Homer and Lund, 1973), cyano (Arapakos and Scott, 1968;Manousek and Zuman, 1965; Rieger e t al., 1963; Volke and Kardos, 1968), and sulphone or sulphoxide groups (Horner and Neumann, 1965; Lamm and Samuelsson, 1970, 1970a; Lamm and Simonet, 1974) are also commonly used substrates. Among neutral electrophiles, halides have been reduced in the presence of carbon dioxide (Baizer and Chruma, 1972; Wawzonek et al., 1964) to give carboxylates and a-haloketones, and reduction in the presence of methyl bromide gives alkylation products (McDowell, 1967). The latter reaction might equally well be classified as a mixed coupling process (p. 17).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

13

Addition In an anodic addition process two nucleophile molecules are added across a double bond or system of double bonds with loss of two electrons, as shown for one double bond in eqn (15). Examples Nu Nu

R2C=CR2 + 2Nu-

+

I I R2C-CR2 + 2e-

(15)

include the addition of acetoxy groups to diphenylethylenes (Mango and Bonner, 1964), cyclo-octatetraene (Eberson et al., 1967), 1,3cyclohexadiene (Baggaley and Brettle, 1968) and to indene and its 3-alkyl derivatives (Bernhardsson et al., 1971; Eberson, 1974), of azido groups to alkenes (Schafer, 1970a) of carbon monoxide to naphthalene (ultimately leading to 1,4-dihydronaphthalene-1,4dicarboxylic acid: Conway et al., 1967), and of alkoxy groups to anthracenes (Parker, 1970; Parker et al., 1971), stilbenes (Inoue et al., 1967), and norbornene (to give the 2,7-addition product: Brettle and Sutton, 1974; Inoue et al., 1967; Shono and Ikeda, 1972). For hitherto unexplained reasons, addition of two cyan0 groups has only been successful with rather special substrates (tetraphenylpyrrole and 9,lO-dialkylanthracenes:Longchamp et al., 1974; Parker and Eberson, 1972) in spite of a great deal of experimentation with others (Nilsson, 1971). In a cathodic addition reaction two electrophile molecules add across a double bond or system of double bonds with addition of two electrons [eqn ( I S ) ] . The most important reaction here is

cathodic hydrogenation ( E + is a proton donor), a reaction that has much in common with the Birch reduction and other dissolving metal reductions (see, for example, Birch and Subba Rao, 1972). In fact, a dissolving metal reduction constitutes nothing but a cathodic process at a consumable electrode in a short-circuited electrochemical system. Cathodic hydrogenations have been performed with many types of substrates, e.g., aromatic hydrocarbons (Asahara et al., 1968; Avaca and Bewick, 1972; Benkeser and Kaiser, 1963; Benkeser et al., 1964, 1970; Misono et al., 1968; Osa, 1968, 1968a; Sternberg et al., 1969; Sternberg et al., 1967), alkynes (Benkeser and Tincher, 1968), and aryl-conjugated olefins (Horner

14

L. EBERSON AND K. NYBERG

and Roder, 1969). Again, carbon dioxide can function as a neutral nucleophile in the addition to aromatic hydrocarbons (Wawzonek and Wearring, 1959) or stilbenes (Dietz and Peover, 1968). A slightly more complex addition mode becomes possible under conditions allowing for a particular bond in the reagent to split during the transformation. Formally, the anodic variety of such a cleauageaddition reaction is written as in eqn (17) for a substrate with one

double bond. Examples of this practically interesting but far from optimized reaction type include processes with R-E chosen as R-COO- (Fioshin et al., 1963; Lindsey and Peterson, 1959; Schafer and Pistorius, 1972; Smith and Gilde, 1959, 1961), e.g., as in eqn ( 18) and RMgX (Schafer and Kunzel, 1970) ,e.g. as in eqn ( 19).

0- 0:

COOEt

2EtOOC-C00-+

+ 2 C 0 2 + 2e-

COOEt (18)

2BuMgBr + CH2=CH--CH=CHz

+

BuCH2CH=CHCH2Bu + 2MgBr'

+ 2e(19)

Cathodically, cleavage-addition is formulated as in eqn (20) for one double bond. Few clearcut examples are known, but an intraR R 2R-Nu + R;C=CRh

+ 2e-

I I

+ RkC-CRh

+ 2Nu-

(20)

molecular version is represented by the ring closure shown in (21)

ClCHzCHzCHzCl+

+2c1\

(21)

/

(Lipkin et al., 1963). If some atoms of the system of double bonds are different from carbon, additional possibilities open up, such as in the capture of the intermediate in the cathodic reduction of lY3-diketones(22) to give acetates of cyclopropanediols (Curphey et al., 1969).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

15

To complicate things further, one can formally write down a hybrid of eqns (16) and (20) and arrive at a mixed version [eqn (23)], exemplified by the extremely rich variety of reactions possible in the R E R-NU + E+ + R ~ C = C R +~ 2e-

I

I

+ R~C-CR;

+N ~ -

(23)

cathodic reduction of organic halides or 'onium compounds in the presence of activated olefins (Baizer, 1966; Baizer and Chruma, 1972; Wagenknecht and Baizer, 1966, 1967), e.g. as shown in eqn (24). C13C-Cl + CHz=CHCN + H+ + 2e-

+ C13CCHzCH2CN

+ C1-

(24)

Elimination Anodic and cathodic elimination is simply the reverse of cathodic

[eqn (16)] and anodic [ eqn (15)] addition, respectively. Important cases are anodic bisdecarboxylation, either in the 1,2- (Corey and Casanova, 1963; Radlick et al., 1968; Westberg and Dauben, 1968) or 1,3-fashion (Vellturo and Griffin, 1966), with the preparation of Dewar benzene and dimethyl bicyclobutane-2,4-dicarboxylateas the more prominent cases [eqns (25) and (26)], and cathodic dehalogenation of dihalides with the halogens in the 1,2- (Zavada et al., 1963), 1,3- (Casanova, 1974; Gerdil, 1970; Rifi, 1967, 1969), 1,4(Casanova and Rogers, 1974; Wiberg et al., 1974) and 1,6- (Covitz, 1967) positions. The synthesis of bicyclobutanes (27) and [2,2,2]propellane (28) bear witness to the usefulness of this reaction type.

16

ncoo-

L. EBERSON AND K. NYBERG

MeOOC

Q

MeOOC

___j

- ooc

COOMe

+ 2C02 + 2e-

COOMe

(26)

Coup ling In many respects, electrochemical coupling is the most useful of the reaction types described here, since many of its varieties have few or no counterparts in conventional laboratory practice. Bifunctionalization of dimeric carbon skeletons from monomers is perhaps the most attractive practical synthetic procedure among the myriads of possibilities. Two simple versions of anodic coupling exist formally, namely coupling-elimination (29) and coupling-addition (30). 2R-E + R-R + 2E' + 2e\

2Nu- + 2 ,C=C

/

\

-+

I l l 1 I I I I

Nu<-C-C-C-Nu

(29)

+ 2e-

(30)

The former type is best represented by the Kolbe reaction, without doubt the most extensively studied and still not very well understood electro-organic process (for reviews, see: Conway and Vijh, 1967; Eberson, 1968, 1973a; Eberson and Nyberg, 1976; Weedon, 1952, 1960). Another widely studied reaction of this type is biaryl and diphenylmethane coupling of aromatic compounds of which both intermolecular (Bobbitt, 1973; Nyberg, 1973, 1973a and references cited therein; RonlQn et al., 1973a, RonlQn and Parker, 1974a) and intramolecular cases (e.g. Falck et al., 1974; RonlQn et al., 1973, 1974; Svanholm et al., 1974a) are known. The Nu- in many applications of the latter reaction type is a stabilized carbanion (derived from e.g., 1,3-dicarbonyl compounds, nitriles, or nitro compounds: Schafer, 1969; Schafer and Steckhan, 1969), methoxide

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

17

ion (Belleau and Au-Young, 1969; Eberson and Schafer, 1971; Katz al., 1974; Schafer, 1970; Schafer and Steckhan, 1969), or acetate ion (Cedheim and Eberson, 1974). Moreover, it is easy to see that the complication represented by eqn (17), i.e. of using a nucleophile (R-E) that can split under the prevailing reaction conditions, can be introduced also here, Again (see p. 14), carboxylates (Brookes et

et

al., 1974; Chkir and Lelandais, 1971; Fioshin et al., 1963; Lindsey and Peterson, 1959; Schafer and Pistorius, 1972; Smith and Gilde, 1959, 1961) and Grignard reagents (Schafer and Kunzel, 1970) are the most obvious candidates for this reaction. The cathodic counterparts of eqns (29) and (30) are indicated in (32) and (33). While reaction (32) often accompanies cathodic 2R-Nu + 2e\

2E+ + ,C=C<

+= R-R

+ 2e-

+ 2NuI l l 1 I I I I

+= E-C-C-C-C-E

(32) (33)

substitutions [eqn (14), p. 121 as a side-reaction and is difficult to promote, reaction (33) is of great importance in cases where E+ is a proton. The general designation for this mode of coupling is electrohydrodimerization (for reviews, see: Baizer, 19 73; Baizer and Petrovich, 1970), the best known case being the flagship and grant swinger of electro-organic chemistry all over the world, Monsanto’s process (34) (Baizer, 1964; Danly, 1973) for making adiponitrile directly from acrylonitrile. ZCH,=CH-CN

+ 2H’ + 2e-

+- NC(CH2)4CN

(34)

No doubt the reader by now suspects that one can create the worst possible set of complications by mixing eqns (32) and (33) as done previously for cathodic addition [eqn (23)]. It is, however, highly improbable that such a reaction would proceed with anything near selectivity toward the formal product, and hence we refrain from considering it. As far as we are aware, no such case is known. Equations (30), (31), and (33) depict formation of “dimers” of the starting monomeric olefin. It should be noted that trimers and higher oligomers can also be isolated in many cases, e.g. in the addition of methoxy groups to butadiene (Katz et al., 1974 and references cited therein) and in the co-electrolysis of monoethyl

18

L. EBERSON AND K. NYBERG

oxalate with butadiene (Schafer and Pistorius, 1972, and references cited therein).

Cleavage Some of the previously discussed processes, especially cathodic substitution reactions according to eqn (14) with E + = H’, might also be considered as cleavage reactions. In fact, it is difficult to accommodate electrochemical cleavage reactions within a unique formal scheme (for a review within a rather broad definition, see Horner and Lund, 1973), and we are therefore content merely to show a few examples which conform reasonably well to the commonly accepted notion of cleavage. (i) Selective anodic cleavage of tyrosyl-peptide bonds (Cohen and Farber, 1967; Isoe and Cohen, 1968), as shown in reaction (35). NHCOR

NHCOR

CONHR’ +HzO

0

==lo + R’NHz + 2e- + 2H+

(35)

(ii) Anodic cleavage of 0-alkyl or 0-acyl bonds as in eqn (36) (Parker, 1969; cf. also Eberson, 1973a, and Utley and Yates, 1973

)$- 0

+2R++2e

OR

(36)

0 R = Me, Ac

who trapped acyl cations from this reaction by adding hydrocarbons). (iii) Cathodic cleavage of C - 0 or C-S bonds in optically active derivatives of atrolactic acid [eqn (37)] (Erickson and Fischer, 1970; Fischer and Erickson, 1973).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

19

R

I

H

X

I Ph-C-COOR' I

+ 2e- + H+

--f

I I

(37)

CH3

CH3

X, R = 0, PhCO X, R = S, Ph

+ RX-

Ph-C-COOR'

Racemization 5-9% Inversion

(iv) Cathodic cleavage of C-P and C-As bonds in optically active phosphonium and arsonium salts t o give optically active phosphines and arsines with retention [eqn (38)] (Homer e t al., 1971; Horner and Fuchs, 1962). + + R3P(As)CH2Ph + 2e- + H+

--f

R3P(As)

+ PhCH3

(38)

Miscellaneous Reactions The electrochemical synthesis of organometallics (Lehmkuhl, 1973, 1973a) by conducting the electrolysis at a consumable electrode, e.g. as in the Nalco process (39), is again not readily accommodated in a single formal scheme. 4R-MgX + Pb + R4Pb + 4e-

+ 4MgX'

(39)

4. MECHANISTIC PROBLEMS

After the preceding brief survey of electro-organic reactions, let us go back to the experimental situation and try to discern the nature of the mechanistic problems originating from the reaction conditions peculiar to electrochemical experiments. Figure 2 shows, in the direction of the arrows, the train of events involved in the electrochemical process and the associated mechanistic problems in the different layers of the electrolyte solution First, the different components of the bulk solution must be brought up to the electrode surface by the processes summarized under the heading mass transfer. Convection (movement of the solution relative to the electrode) takes molecules and ions t o the boundary of the Nemst layer (Section 2), through which diffusion takes them up to the electrode surface. Diffusion occurs due to the

20

L. EBERSON AND K. NYBERG

disappearance of a compound undergoing reaction at the electrode; by this process a concentration gradient of the compound is set up in the Nernst layer and its molecules will then move toward the electrode. A third type of mass transfer, migration in the electric field between the electrodes, operates for charged species. As molecules and ions arrive at the electrode suface, it becomes Nature of the electron transfer process Transition state structure Reactivity Effect of electrode material

I

I

Nature of substrate Structure and composition Adsorption of EI effects Orientation of substrate I

Transport of ions and molecules Concentration profiles

Transport of ions and molecules Form of substrate

I Electrified interface thickness: 10

I Nernst layer thickness: 0.5-0.01 mm

\ Nature of first and subsequent intermediates Sequence and nature of follow-up reactions Adsorption effects

products of bulk Same problem as for substrate

t Bulk solution thickness: mm-cm

Product in

Figure 2. Problems in mechanistic organic electrochemistry.

necessary to know which species undergoes the primary electron transfer process, i.e. whether it is the substrate which reacts (a direct process) or some other component (an indirect process). It is also necessary to know whether the reacting species has been somehow chemically modified, e.g. in the simplest case by undergoing an acid-base reaction in the bulk solution or at the electrode surface. The distinction between direct and indirect electrode pro-

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

21

cesses can obviously not be very sharp, as will be discussed in more detail later (Section 5). So far, all mechanistic problems except one have proved easily amenable to solution by different experimental techniques; voltammetric studies, sometimes in conjunction with product studies at controlled potential, usually suffice for identifying the reacting species. However, there are still cases which resist mechanistic clarification even on this fairly “trivial” level (p. 40). The exception mentioned above is connected with the concentration gradients set up in the Nernst layer (Fig. 2). Imagine a reaction that produces at the electrode a fairly high concentration of protons, as many anodic processes do, whereas the bulk solution is kept macroscopically neutral or even basic by suitable buffering. It is easy to understand that a certain ambiguity will always be attached to product studies if acid-catalysed rearrangements of primary products are demonstrable in blank experiments in a homogeneous medium (p. 34). Similarly, base gradients can be set up at the cathode. Once the nature of the reacting species has been established, the difficulties become more formidable. This is of course due to the heterogeneous nature of electrode processes which makes it necessary to take the composition and structure of the electrified interface (abbreviated EI) into account. Although this region is a very narrow layer, 10- 15 thick, all the important events occur here, and hence it is of central importance to try to understand how it can influence the outcome of the electrochemical reaction. EI’s are set up wherever there is an interface between two phases. In the case of a metal and an electrolyte solution, adsorption of solvent molecules and ions will create a layer of charge on the solution side of the interface and this will be compensated by the same amount on the metal side by movement of free electrons in the metal to or from the phase boundary. It is important to realize that the build-up of an EI takes place spontaneously in the interface and does not require an external charging of the metal, i.e. we need not connect the electrode to an external potential source. When an external potential is applied, changes in the composition and structure of the EI, including net electron transfer if the potential change is large enough, take place in order to compensate for the potential change. Adsorption is classified either as physisorption in which the forces between the surface atoms and the adsorbing particles are essentially of the van der Waals type, or chemisorption in which these forces are

a

22

L. EBERSON AND K. NYBERG

more akin to ordinary chemical bonds. A well-known case is the dissociative chemisorption of hydrogen on noble metal catalysts. In electrochemistry, dissociative chemisorption is the desired process in fuel cells but hardly in synthetic experiments, since it leads to complete combustion of the organic compound to carbon dioxide and water. Adsorption at electrodes is sometimes called electrosorption; excellent treatises on this phenomenon have been published (Damaskin et al., 1971; Damaskin and Frumkin, 1971; Gileadi et al., 1967). I

I

I

I

I

IHP

I

OkP

Figure 3. Schematic representation of the anode EI. For explanation of symbols, see p. 22.

It is a fair conclusion that our present knowledge of the composition and structure of electrode EI’s is at best very schematic. Nevertheless, it is illuminating to try to give a pictorial representation of how it might look in structural terms, following the lucid treatment given in “Modern Electrochemistry” (Bockris and Reddy, 1970), highly recommended further reading for those who want to master EI theory in more detail. Such a picture is shown in Fig. 3, which summarizes the most important features of an anode EI. At the electrode surface, represented as being completely flat and

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

23

devoid of structural detail,’ we find adsorbed solvent molecules, represented as dipoles with their negative ends pointing toward the positively charged surface. Also included are two negative ions, which are adsorbed in contact with the surface and free from their normal solvation shells. This kind of adsorption is called contact or specific adsorption, “specific” indicating that it is governed not only by electrostatic interaction between the ion and the electrode but also by a weak bond, characteristic of the particular ion involved. Anions and large cations, which both generally are weakly solvated, are most prone to become specifically adsorbed, since desolvation [eqn (40)] M(Solv),

+ A-(Solv),

f

M

*

’

* * * *

A- + (Solv),+,

is required before contact adsorption can occur. It is important to note-and, at least for an organic chemist, hard to accept the thought-that specific adsorption also can take place between an electrode surface and an ion of like charge. Figure 3 further shows, in the layer next to the surface, a cross-section of a hydrocarbon molecule, thought to be adsorbed via a bond closely resembling the familiar n-bond in n-donor-acceptor complexes (see for example: Barradas and Conway, 1961; Blomgren and Bockris, 1359; Blomgren et al., 1961; Bockris et al., 1964; Conway and Barradas, 1961; Dahms and Green, 1963). The second layer contains anions, adsorbed electrostatically in a non-specific manner and surrounded by their solvation shells, together with solvent molecules interspersed with the ions. A few cations are shown outside this layer. The planes indicated, being the loci of the centres of specifically and non-specifically adsorbed ions, respectively, are called the inner and outer Helmholtz planes (abbreviated IHP and OHP). Proceeding with the series of events in Fig. 2, the nature of the electron transfer between electrode and substrate and possible requirements in their relative orientation for the electron transfer process to occur next command our attention. Clearly, this is the central act in electrochemistry, and it is of utmost importance to give a description of it in terms that can be reconciled with known concepts in theoretical organic chemistry. The structure of the This is of course an oversimplification for all metal surfaces, except possibly that of liquid mercury. Electrode surfaces at solid metals and other solid materials are normally quite heterogeneous in the sense of exhibiting lots of cracks, crevices, comers, declivities, asperities, and edges, but this is for obvious reasons not possible to include at the present state of our knowledge (cf. Section 12).

24

L. EBERSON AND K. NYBERG

transition state will be crucial for our understanding of the electron transfer act. The problem of the reactivity of organic molecules at electrodes and its dependence on structural factors will logically enter at this stage. The most difficult problem of all, to eludicate the role of the electrode material in determining the rate of the electron transfer step, has to be tackled in the same context. Empirically, one often finds large effects of changing the electrode material on kinetic parameters [e.g. for reduction of H 3 0 + to hydrogen the electrochemical rate constant, defined as the equilibrium exchange current density (see p. 46), changes by a factor of almost 10" in going from palladium to mercury] and product distribution, but it is usually very hard, if not impossible, to interpret such results in terms of a consistent theory. The nature of the electrode material also affects the EI structure and composition by its adsorptive properties towards the components of the electrolyte solution. After the primary electron transfer, the nature of the first intermediate and its reactivity will be of critical importance in determining the reaction sequence leading to product(s). As an example, a neutral molecule, R, undergoing one-electron oxidation gives a radical cation, R + that can react further via a number of competing pathways [eqns (41)-(44) and (46)]. Equations (42)Second electron transfer to the anode

R'+ --f R+++ e-

Coupling

R"

+ +

+ R '++ R-R

(43)

Reaction with nucleophile (Nu can be equal to R itself) R"

+ Nu-

+ k-Nu

(44)

(43) lead to dications which in most cases (see, however, p. 10) will be trapped by nucleophiles present to give product(s). k-Nu in eqn (44) is in many cases easier to oxidize than R itself, so that the electrode potential is sufficient to oxidize it to a cation [reaction (45)], giving product(s) in a terminating step with a base/nucleo-

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

25

phile. Alternatively, electron transfer from R-Nu can take place in a homogeneous reaction with R - H + present in high concentration in or near the EI as shown in (46). R-H ' +

+ i-Nu

--f

;-Nu + R-H

(46)

In electrochemical nomenclature, reaction sequences are described in terms of E (electrochemical) and C (chemical) steps. Thus a succession of the primary electron transfer step, eqn (44), eqn (45), and the terminating chemical step would be an ECEC sequence. It is sometimes useful to si'gnify nucleophile or base action in the C step by subscripts N or B, e.g. ECBECN for sequence (47) (Parker and Eberson, 1969). ArCH3

-e-

GAc

ArCH3*+ + ArCHz-

- -

-e-

ArCHZ'

OAc

ArCHZOAc (47)

Reactions (41)- (46) have their exact counterparts in cathodic processes but these need not be listed here. If the initial electron transfer occurs with a charged substrate, the first intermediate will be a neutral radical, capable of undergoing the conventional reactions of radical chemistry such as coupling, disproportionation, hydrogen atom abstraction from 0-H bonds, cleavage, homolytic addition t o and substitution in added compounds. All of these processes are well known from homogeneous solution chemistry. In addition, the possibility exists that the first radical or any one formed subsequently can undergo electron exchange with the electrode to give a carbanion (at the cathode) or carbocation (at the anode). Thus we Scheme 1

-e-

R

R-

Radical cation chemistry

R+

k7

Carbocation chemistry

Radical chemistry

R+y R - y - R -

-

Carbanion chemistry

/

+e-

+e-

R

Radical anion chemistry

26

L. EBERSON AND K.

NYBERG

can summarize the spectrum of research fields that has t o be familiar to practitioners of electro-organic chemistry in Scheme 1. If someone is missing carbene chemistry, don’t worry; carbenes can also be generated electrochemically (Wawzonek and Duty, 1961). With the problem of the reaction sequence outlined, we are near the end of our list of mechanistic problems, since now only the transport of product to the bulk solution remains. If it can be ascertained that the product is not chemically transformed in the electrolyte (blank experiments) or electrochemically further transformed (runs at low conversion), we are actually at the end and can begin to go into more detail with mechanisms. In the following treatment, the mechanistic problems will be discussed under the following headings: Effect of concentration gradients Nature of the electroactive species Nature of the reaction sequence Role of adsorption and EI Nature of the electron transfer process Reactivity and structure Influence of electrode material. First, however, we shall dissect the concepts of direct and indirect electrode processes, since mechanistic problems are implicated in this mode of classification.

5. DIRECT AND INDIRECT ELECTRODE REACTIONS One need not go further back than about thirty years (Fichter, 1942) to find that most organic electrode reactions were viewed as indirect ones,’ i.e. that components other than the substrate were primarily transformed at the electrodes into oxidizing and reducing agents, respectively and that these then attacked the substrate. Reminiscent of the view is, for example, the hydrogen peroxide theory of the Kolbe reaction (Glasstone and Hickling, 1934; for a review, see Weedon, 1952) and the more general theory that organic electrode reactions were mediated by very reactive atomic hydrogen A notable exception is the “discharged ion theory” of the Kolbe reaction, dating back to 1891 (Brown and Walker, 1891). This remarkably farsighted suggestion is entirely in accordance with present-day theory of the Kolbe reaction, but was for a long time abandoned in favour of indirect mechanisms (e.g. the hydrogen peroxide theory).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

27

(cathodic reduction) and atomic oxygen (anodic oxidation). In the period between 1940 and 1965 these ideas slowly faded away and were replaced by other varieties of what still were indirect mechanisms, mostly expressed as attack on the substrate by reagents or intermediates generated anodically (cathodically) by electron transfer from the anion (cation) of the supporting electrolyte, a solvent molecule or sometimes an added reagent. Thus, it is not surprising that anodic halogenation of organic substrates in media containing C1-, Br-, and I- were considered as essentially homogeneous reactions (for a review, see: Allen, 1958) with electrogenerated halogen as the reagent [eqn (48)]. Nowadays, exceptions to this mechanism

are known but they are rare (p. 58). Equally understandable is that other anodic substitutions, formally identical, came to be regarded as radical substitutions, mediated by electrogenerated radicals as shown for acetoxylation in eqn (49) and for cyanation in eqn (50) (Koyama et al., 1965; cf. however Tsutsumi and Koyama, 1968). However, a

CN-

- -e-

CN*

Ar-H -H

Ar-CN

*

number of indirect mechanisms was later to be proven wrong or at least less likely, primarily because a host of new solvent/supporting electrolyte combinations with very high anodic and/or cathodic limits had become available, allowing for voltammetric investigation of very unreactive substrates. This led to a complete re-evaluation of the existing mechanistic ideas, such as done for anodic reactions by Weinberg and Weinberg in 1968. Today, there are still a few question marks on this demarcation problem (Section 7). It is also somewhat blurred by the fact that it is not always possible or simple to say which component of the electrolyte solution should be considered as the substrate. If, for example, we reduce a mixture of carbon tetrachloride and acrylonitrile together, it is fairly obvious that reaction (24) could occur via an electrogenerated carbanion from CC14 that adds to the olefin in a Michael reaction (51). It then becomes largely a question of se-

28

L. EBERSON AND K NYBERG

mantics to classify the mechanism as direct or indirect. The same ambiguity is inherent in all processes in which we use one organic compound to generate an intermediate-often after extensive structural change-which attacks a second organic component in a reaction that ultimately leads to the desired product(s). Other examples of this type already encountered are shown in eqns (17), ( Z O ) , (23), and (31) and in particular in eqns (18), (19), (21), and

(22).

However, leaving reactions of this type aside for a while, we can distinguish a number of processes which are clearly indirect, involving electrogenerated inorganic reagents. These are neglected in the following, not because they are less interesting than direct processes, but for the reason that their mechanisms conform to reactions in homogeneous solution (not that these are very well understood). Indirect processes via electrogenerated inorganic reagents (for reviews, see Dietz and Lund, 1973; Eberson and Schafer, 1971) include reactions mediated by halogen and pseudohalogen; metal ions or compounds in high oxidation states (sometimes called “mediators”), such as Mn04 -, 104-, O s 0 4 , Co(III), Mn(III), Ce(IV), Ag(II), Hg(II), and Pd(II), to act as oxidizing agents or catalysts; metal ions in low oxidation states, such as Al(I),’ Mg(I),’ Cu(I), Sn(II), and Cr(II), to act as reducing agents or catalysts; nucleophiles and bases, such as superoxide ion; hydrogen; solvated electrons (for a review, see Lund, 1973); amalgams (for a review, see Lund, 1973a). Of these, the last three are mechanistically not always easy to distinguish from direct mechanisms, and we shall have occasion to return to them later. Some electrogenerated organic intermediates so obviously participate in indirect mechanisms that they can also be left out in what follows: carbenes (Wawzonek and Duty, 1961), benzyne (Wawzonek and Wagenknecht, 1963), azobenzene anion (acting as a base in a Wittig reaction; see Iversen and Lund, 1969; cf. also Iversen, 1971, for an analogously induced Stevens rearrangement). On the other hand, the finding that electrogenerated anion radicals from aromatic hydrocarbons can reduce organic halides which either are more Use of this reagent constitutes an “anodic reduction”, since it is generated by anodic oxidation of a consumable metal electrode.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

29

difficult to reduce than the hydrocarbon or even impossible to reduce at the electrode (Sease et al., 1969; Lund et al., 1974) poses some very interesting problems with regard to the electron transfer mechanism and will therefore be discussed in Section 11. Finally, we may also draw attention here to the use of the electrogenerated naphthalene anion radical for the reduction of that important, electrochemically nonreducible inorganic compound, nitrogen, via a Ti(IV)/Ti(II) redox couple (van Tamelen and akermark, 1968).

6. EFFECT OF CONCENTRATION GRADIENTS The effect of concentration gradients in electrode reactions is really not a problem of mechanism but rather a troublesome source of possible systematic ambiguity in the interpretation of the product distributions observed, one of the tasks that lies close to the heart of the organic chemist. To see how this comes about, it is instructive to make the mental experiment that we generate acetoxy radicals by the Kolbe reaction of acetate ion in acetic acid [eqn (52)] at an electrode of 1 cm2 surface area, passing a current of 1 A during -e-

CH3COO- + CH3COO'

'

(52)

s. This generates lo-' moles of CH3COO-, and if we now for simplicity assume-remember it is a mental construction-that we can arrest the further reactions of CH3COO* and distribute it evenly in the first 100 layer, it is easy to calculate that [CH3COO*] would

a

Distance

-

Figure 4. Concentration gradients of substrate and electrogenerated intermediate.

30

L. EBERSON AND K. NYBERG

be 0-01 M . An experiment in a homogeneous medium designed to be comparable with this process would be the thermal decomposition of AczO2 in acetic acid to give acetoxy radicals. At 85"C, the rate s-l ,and it constant of this reaction (Levy et al., 1954) is 1.3 x is easy to see that the concentration of reaction-arrested CH3COO., produced during s from a 1 M solution is about eight powers of ten lower than in the electrochemical experiment. As we shall see below, this fact has significant consequences for the product distribution from Kolbe reactions. Thus, neglecting adsorption effects in the EI for a while, we can conclude that concentration gradients (see Fig. 4) in electrochemical experiments will have the following consequences for mechanistic studies: they will

( a ) affect product distributions as compared to mechanistically analogous homogeneous processes by favouring bimolecular reactions between intermediates, the differences being greater the more reactive the intermediate is, and ( b ) sometimes generate near the electrode high acid (anode) or base (cathode) concentrations from the reaction itself or from a side-reaction, even if the bulk solution is buffered to neutrality or alkalinity; this could then cause unexpected transformations of primary products, which would be difficult to simulate and might possibly be aggravated in dipolar aprotic solvents in which acids and especially bases exhibit extraordinary high reactivity. If we also consider the role of adsorption on the distribution of products, then we should note that surface concentrations of substrate and intermediate(s) must be taken into account, i.e. their concentrations in the inner Helmholtz layer (cf. for example, Wendt, 1973). One effect of this would possibly be to ( c ) increase the probability of encounters between adsorbed substrate molecules and intermediates, leading to coupling reactions and sometimes even to polymerization.

Bimoleculur Reactions Between Intermediates The very success of electrochemical reactions that give products resulting from coupling between two fragments originating from two molecules is an indication that locally high concentrations of intermediates must play an important role. Owing to the difficulty of

31

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

finding truly analogous homogeneous processes t o compare with, there are only few clear-cut cases corroborating this postulate. One of them is the Kolbe reaction of potassium propionate in anhydrous propionic acid at 1 OO°C, compared to thermal propionyl peroxide decomposition under as closely similar conditions as possible (Goldschmidt et al., 1952). Equation (53) shows the products

--coz/ (C2H5COo)Z

Iattaclton solvent

(53) ~

CzH6 + CH3CHCOOH I

CH3CHCOOH I CH3CHCOOH

CH3 \ CHCOOH

CZH5

/

formed via radicals; here we have immediately one source of ambiguity, in that the Kolbe reaction has a competing cationic pathway in which the ethyl radical is converted into the ethyl cation; this reaction has no counterpart in the peroxide decomposition. Neglecting this factor, the composition of radical-derived products in the two cases is shown in Table 1. From this we can immediately establish that bimolecular reactions, coupling and disproportionation, between ethyl radicals are favoured in the heterogeneous reaction, whereas attack on solvent is far more pronounced in the homogeneous medium. Both findings qualitatively agree with expectations based on concentration differences.’ Moreover, the fate of the a-carboxyethyl radical is significantly different in the two cases; at the electrode the concentration of ethyl radicals is relatively high, favouring formation of methylethylacetic acid, whereas in solution where the steady state concentration of ethyl radicals is extremely low, the resonance-stabilized a-carboxyethyl radical is allowed It is gratifying to find that the ratios (0.43 and 0.35, respectively) of the rate constants for disproportionation and coupling, equal to the ratio of the percentages of CzH4 and C4H10, are close to the ones reported for a range of ten solvents (acetic acid was not included, however). These values fall between 0.15 and 0-24 for ethyl radicals produced by photolysis of azoethane at 65OC (Stefani, 1968).

32

L. EBERSON AND K. NYBERG

enough life-time to undergo self-coupling with formation of 2,3-dimethylsuccinic acid t o a higher extent. Stabilization of Kolbe-generated radicals by -M substituents has the effect of decreasing the difference in concentration dependence between the electrochemical reaction and another Kolbe-duplicating TABLE I Organic Productsu Formed by Decomposition of Propionyl Peroxide and by Kolbe Electrolysis of Potassium Propionate in Propionic Acidb

Product CZH4 C2H6 (from disproportionatione) CZH6 (from solvent attackf) C4H10 Methylethylacetic acid 2,S-Dimethylsuccinic acid

Composition of Composition of product from the anode products from decomposition of electrolysis of 2 moles of 1 mole of peroxide potassium propionate

3.0' 3.0' 25.1' 6.9' 0.038 mole 0.106 mole

5*6d 5*6d 8 *4d 15.fjd 0.002-0*014 mole 0.00015-0.002 mole

The yield of carbon dioxide is similar in the two cases. Goldschmidt et uL, 1952. Percentage of total gaseous products. Percentage of anode gases. Estimated from the amount of ethylene,assumed to form via disproportionation only. f Residual ethane after subtraction of the percentage of ethane formedvia disproportionation.

'

homogeneous process, oxidation of carboxylate ion by sulphate ion radicals [eqns (54) and ( 5 5 ) ] . The substrates used were sodium

monoethyl t-butylmalonate, t-butylcyanoacetate, and t-butylmalonamate which all combine the desirable features of giving radicals that cannot disproportionate due to lack of @-hydrogensand are not further oxidized anodically to cations to any significant degree due to the presence of the -M substituent. Hence the absolute yields of coupling product can be compared directly as a measure of the similarity between the reactions. Table 2 shows that the yields are indeed similar in the two cases (Eberson et al., 1968).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

33

Another reaction variable, the current density, affects the product distribution from the Kolbe oxidation in an entirely predictable manner, as for example shown by the variation in the ratio between coupling and radical attack on C-H bonds in the Kolbe oxidation of TABLE 2 Yield of Coupling Products from Kolbe and Persulphate Oxidation of t-BuCH(R)COO- in Aqueous Medium' R

Kolbe oxidation at P t

Persulphate oxidationb

COOEt CONHz CN

41 55c 14

34 41 23

' Eberson et aL, 1968.

At 100°C; [R'COO-1 /[SzOi-] = 5:l. Run in methanol owing to filming problems in aqueous solution.

aqueous 1 M CH, COOH/CH3COONa at platinum, as defined by the ratio between ethane and methane formed' [eqn ( 5 6 ) ] (Shukla and

Currentdensity,mAcm-2 Ratio CH4/CzHb

2 2-9 4 10 35 200 0.75 0.28 0.18 0.05 0.03 0.01

Walker, 1931). This and analogous findings in other systems (see e.g. Haufe and Beck, 1970) have been explained in terms of an increasing concentration of radicals at the anode with increasing current density (Dickinson and Wynne-Jones, 1962). However, other types of coupling processes behave differently with respect to changes in current density. Thus the product yield in the anodic coupling of aromatic hydrocarbons in non-nucleophilic media to give biaryls and/or diphenylmethanes (Nyberg, 1970, 1971, 1971a, 1971b, 1971c, 1973a; Eberson et al., 1973a) is independent of current density but strongly dependent on substrate concenMethane has been shown conclusively to originate from methyl radical attack on the methyl group of CH3COOH (Clusiusand Schanzer, 1943).

34

L. EBERSON AND K. NYBERG

tration, as shown for biaryl coupling of mesitylene [reaction (57)]. This has been taken as evidence for a mechanism involving attack of

-

Ar-H

i-5 Ar-Ar

-e-

ArH*+

-2H+, - e -

Concentration of substrate (mesitylene) Current yield of biaryl (bimesityl)

(57)

Ar-Ar

0.2 M 1.0 M 2‘0 M 1 30 53

the cation radical on a substrate molecule. A chemical analogue, oxidation by anhydrous iron(II1) chloride, shows similar features (Nyberg, 1974a, b). Cathodic hydrodimerization of acrylonitrile at a mercury cathode in aqueous tetraethylammonium toluene-psulphonate solution [eqn (34): Baizer, 1964; Beck, 1972al shows the same characteristics, a strong dependence on substrate concentration and insignificant variation with current density. Mechanism (58), formally analogous to the one given in the upper half of

Concentration of substrate (W)

5

Molar ratio NC(CHZ)~CN/CH~CH~CN1

10 3.5

21 50

40 41

equation (57), has also been suggested here. As will be seen from the further discussion of this reaction, the mechanism is considerably more complicated (see pp. 82 and 93).

Locally High Acid or Base Concentrations Near the Electrode

To raise this problem is a favourite remark from members of the audience at meetings on organic electrochemistry, and the answer is a never-failing: “I don’t know, and I know of no way of finding out”. However, since pH-control in the layer near the electrode appears to be crucial in, e.g., cathodic metal deposition, several methods have actually been developed for the direct measurement of pH near

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

35

electrodes (see e.g. Kadyrov et al., 1971; Ovchinnikova et al., 1962). We shall not dwell here upon the nature of these methods, but rest content with establishing that even at low current densities the pH of the cathode region very rapidly increases after the onset of electrolysis of a mascroscopically neutral aqueous electrolyte solution. The increase is in the range 3- 5 for current densities between 0.1 and 10 mA cm-2, even the latter value being a rather low one for organic electrolyses. The difference in pH between the electrode layer and the bulk solvent decreases with increasing stirring rate and temperature. The alkalinization of the cathode region during electrodeposition of a bivalent metal has been treated theoretically (Harris, 1973); the calculations indicate that hydroxide ion is produced by the cathodic reduction of water and that the pH-gradient from the cathode to the bulk solvent can be as large as 6 - 7 pH units. Given these results, pertaining as they do t o inorganic systems in aqueous medium, one nevertheless wonders what they imply for electrolyses in organic systems, and how one shall avoid mechanistic pitfalls due to such gradients. Especially in dipolar aprotic media, acid- and base-catalysed reactions of products primarily formed constitute a likely source of mechanistic ambiguity. However, since the necessary blank experiments have only been performed in isolated cases and thus little experience with this problem has accumulated, we shall first examine a few electrode reactions in which pH-gradients might possibly be of some importance. As mentioned before, the Kolbe reaction has a cationic pathway (reviews, see: Eberson, 1968, 1973) which is strongly favoured by the use of carbon anodes (Koehl, 1964; see further Section 12). Skell and Reichenbacher (1968; cf. also 1968a; Keating and Skell, 1969) studied the anodic oxidation of aqueous potassium 4,4-dimethylpentanoate at a carbon anode and analysed for products from the different reaction modes of the 3,3-dimethyl-l-butyl cation formed as a function of electrolyte pH in the range of 6.6- 14.5. A rather drastic change was observed at a pH around 14, where products from the trapping of the unrearranged 3,3-dimethyl-l-butyl cation increased sharply at the expense of those from rearranged cations. Thus, it seems as if trapping by hydroxide ion replaces trapping by water at this pH. But is it actually at pH 14? Protons are formed at the anode from elimination and solvolysis reactions, and the actual pH in the region near the anode might be much lower. In this particular case, however, a homogeneous cation generating process, deoxidation, exhibited analogous behaviour around the same pH

36

L. EBERSON AND K. NYBERG

value, suggesting that pH gradients do not significantly affect the results in the anodic process. It is not often, however, that one has possibilities of comparing electrochemical and homogeneous processes as in this case. A reaction that has been discussed in terms of high acid concentrations is the nitration that takes place upon electrolysis of a suspended aromatic hydrocarbon in dilute (1-2 M ) nitric acid (for summaries, see Allen, 1958; Fichter, 1942; Tomilov, 1961; Tomilov and Fiochin, 1963). The proposal has been that nitric acid would be formed in high concentration in the anode region, thus causing nitration in an ordinary electrophilic substitution reaction. This is an entirely reasonable idea, as it has since been shown (Cauquis and Serve, 1970) that nitrate ion is oxidized to dinitrogen pentoxide using nitromethane as a solvent (59). In an aqueous solution this NO;

-

2e- + 1/202 +NO;

NO;

~~0~

(59)

would result in a high nitric acid concentration near the anode. We should like to suggest that a still simpler explanation is possible. Owing to proton release from the oxidation of water molecules in the aqueous medium, the high hydrogen ion concentration near the anode would cause the formation of nitrating species like H2NO: and possibly NO: (Hoggett et al., 1971, pp. 9-12). The same mechanism has been suggested for the formation of nitro-derivatives from the electrolysis of mesitylene and naphthalene in acetic acid/ tetrabutylammonium nitrate' (Nyberg, 1971d). Another phenomenon that might possibly be attributable to a steep pH gradient near the anode is what was originally postulated (Eberson and Olofsson, 1969; Nyberg, 1969, 1971e) to be an effect caused by a change in EI structure in anodic acetamidation [eqn (12)]. When water is added to the electrolyte, water and acetonitrile molecules compete for anodically generated cations, formed via an

R-H

ECE

H20

R+

R-CH20H

(60)

CH3CN R-NHCOCH3

Although less likely, a direct mechanism cannot be completely excluded. Voltammetric studies of anthracene in molten tetrabutylammonium nitrate at 15OoC indicate that the initial electron transfer is from anthracene (Woodhall and Davies, 1969). It is difficult, however, to formulate a satisfactory mechanism for direct nuclear nitration, since aryl nitrates are known to exhibit different chemical behaviour (cf. Nyberg, 1971d).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

37

ECE mechanism (Parker, 1969a; see p. 25). In this particular case, a series of benzylic cations were found to show anomalous reactivities, in that the less reactive ones were selective toward the weaker nucleophile, acetonitrile, contrary to predictions based on concepts Moreover, in a certain from homogeneous solution chemistry.' region of water concentrations the competition ratio between the ?*

' '\O \ \ '

\,a\ \

I \

\ I

\ \

', 0

?

Figure 5. Plot of the logarithm of the competition ratio (ratio of products formed via trapping by acetonitrile and water, respectively) versus percentage water in the solution. Filled circles, p-xylene; open circles, o-xylene; triangles, hexamethylbenzene. Solid lines, electrochemical reaction; broken lines, solvolysis of ArCHz OTs. These predictions were found to be correct for benzyl and o- and p-methylbenzyl cations, generated by solvolysis of the corresponding tosylates in bicarbonate buffered acetonitrilelwater mixtures (Cedheim and Eberson, 1969). With 3% (w/w) water present, essentially all of the product was the alcohol (Fig. 5 ) ; from these data it can be estimated that water is 40-50 times more reactive as a nucleophile than acetonitrile towards cations. It is pertinent to point out that the considerably more stable cation, diphenylmethyl cation, when generated in acetonitrile solution by treatment of benzhydryl azide with nitrosonium tetrafluoroborate, exhibits behaviour analogous to that shown in Fig. 5 when quenched with water or methanol after its generation (Doyle and Wierenga, 1972). This was explained in terms of irreversible formation of amide from N-diphenylmethylacetonitrilium ion competing with reversible reaction of the diphenylmethyl cation with water. This possibility appears less likely for the more reactive benzylic ions generated anodically but cannot be completely dismissed.

L. EBERSON AND K. NYBERG

38

two trapping reactions of eqn (60) was constant (Fig. 5). This was interpreted in terms of an EI composition that is independent of the bulk concentrations in the region of 3-5% (w/w) water; the behaviour of o- and p-methylbenzyl cation in an analogous homogeneous reaction is strikingly different. Later, it was found that the nature of the anion of the supporting electrolyte had a strong effect upon the amide/alcohol ratio (Nyberg, 1969, 1971a), a change from 95:5 to 5:95 being observed in the oxidation of hexamethylbenzene by substituting perchlorate ion for tetrafluoroborate in acetonitrile/ water (9: 1 molar ratio, 4.6% w/w). A similar behaviour was found in mixtures of acetonitrile and acetic acid, where the effect of fluoroborate as compared to perchlorate was to decrease the amide/acetate ratio (Table 3), analogously to the water case. The anion effect was also proposed to be an effect of having an EI composition different from that of the bulk solution. The better hydrogen-bonding tetraTABLE 3 Anodic Oxidation of Hexamethylbenzene in Mixtures of Acetonitrile and Acetic Acid" ~~

No. 1

2 3 4 5 6

7 8 ge 10 1 le 12 13 14 15 16 17 a

Supporting electrolyte, conc. ( M ) NaC104 (0-5) NaC104 (0.5) NaC104 (0.5) NaC104 (0.5) NaC104 (0.5) NaC104 (0.5) NaC104 (0*5),acidicC NaC104 (0.5) acidicd NaC104 (0.2) Bu4NC104 (0.19) Me4NBF4 (0.21) EtqNBF4 (0.19) B u ~ N B F(0.095) ~ B u ~ N B F(0.19) ~ B u ~ N B F(0.38) ~ Et4NOTs (0.19) Bu~NOTS(0.19)

Molar ratio acetonitrile/ acetic acid 50:50 67~33 9O:lO 95:5 99: 1 99.9:O.l 99:l 99:l 99:l 99:l 99:l 99:l 99:l 99:l 99:l 99:l 99:l

Molar ratio acetamide/ acetateb 1.1 2.2 4-2 3-2 5-2 16 8 4.9 0.33 0.23 0.54 0-52 0.18 0.23 0-16 0.81 0.69

Nos. 1-8,10, 12-17: Nyberg, 1969, 1971e; Nos. 9, 11: Mayeda and Miller, 1972. N-Pentamethylbenzylacetamide/pentamethylbenzyl alcohol. Perchloric acid, 0.4 mM. Sulphuric acid, 0-4 mM. A control expetiment revealed no catalytic effect due to acidity.

--

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

39

fluoroborate ion would be preferentially solvated by acetic acid (water) molecules, thus increasing the concentration of acetic acid (water) in the outer Helmholtz layer (Fig. 3) as compared with the more weakly hydrogen bonding perchlorate ion. Whatever the merits of this hypothesis are (and, viewed in a larger context, it seems to fit in with a lot of other results; see p. 89), the results pertaining to the acetonitrile/water experiments were later challenged on the basis of the electrolyte becoming macroscopically acidic during the run due to the liberation of protons at the anode which are not exactly balanced by base production at the cathode’ (Mayeda and Miller, 1970). The acidic medium would then convert pentamethylbenzyl alcohol to the thermodynamically stable acetamide in an acid-catalysed reaction, as was shown to be the case in control experiments. Acidic electrolytes were produced either by design (divided cell or use of lithium perchlorate as supporting electrolyte2 ) or in experiments run under considerably different conditions from those used by Nyberg (1969). The difference resides in the water concentration, Miller and Mayeda (1972) using a 0.4% (w/w) concentration against Nyberg’s 4.6%. At the lower water concentration, acidity builds up during electrolysis and can afterwards be shown to convert externally added alcohol to the acetamide in about 25 min, perchlorate electrolytes in most but not all cases being more effective than the ones based on fluoroborate salts. At the higher concentration, the electrolyte solution stays neutral during the run and does not afterwards affect externally added alcohol. Thus, while macroscopically high acid concentrations did not influence Nyberg’s experiments, one cannot easily dismiss possible effects from locally high concentrations near the anode. Values of “pH” in acetonitrile/water of about 2 are quite sufficient to bring about the acid-catalysed conversion of pentamethylbenzyl alcohol to its amide (Mayeda and Miller, 1972), and from what has been said earlier, this is the kind of pH that might prevail near the anode in a macroscopically neutral electrolyte solution. The acetonitrile/acetic acid experiments were also to some extent questioned by Mayeda and Miller (1972), as seen from Table 3. Conflicting results concerning the effect of acidity on the stability of the acetate were obtained (Nos. 7, 8, and 9); in addition the In this case, acid and base production should theoretically match each other exactly.

* Lithium metal precipitates at the cathode and causes a lag in base production due to slow reactivity towards water in the electrolyte.

40

L. EBERSON AND K. NYBERG

results found with sodium perchlorate as the supporting electrolyte were different (Nos. 5 and 9). This was explained (Mayeda and Miller, 1972) as being due to formation of acetate ion at the cathode, thus slowing down the acid-catalysed process “compared to that in aqueous acetonitrile”.’ We think this example suffices to show that the use of “neutral” electrolyte solutions based on dipolar aprotic solvents may raise a lot of difficult questions with respect to both macroscopic and local, microscopic acidity or basicity. Adequate control experiments should be carried out, as correctly urged by Mayeda and Miller (1972), but these are not always easy to design. How difficult the problem can be is best shown by the fact that it was possible to effect transacetalization of benzaldehyde diethyl acetal in alkaline methanol solution by oxidizing hydrogen in the solution at a platinum anode (Schafer, 1974). In this experiment protons liberated at the anode must act catalytically in the inner part of the Nernst layer.

Surface Concentration Effects As already noted (p. 30) effects of surface concentrations are really effects of adsorption, since it is the adsorption properties of the components of the electrolyte solution that influence the structure of the inner Helmholtz layer.

7. NATURE OF THE ELECTROACTIVE SPECIES The first mechanistic problem of importance is to find out the nature of the electroactive species for the electrode reaction of interest. Once this has been established, the way is open for mechanistic studies of a higher order, such as the reaction sequence, the role of adsorption, reactivity, and so on. In principle, it should be-and normally is-easy to determine which of the components of the electrolyte solution is the electroactive species. Voltammetric experiments of the type discussed in Apart from the fact that this statement pertains to the acetate which was never studied in aqueous medium, it seems inconsistent to allege cathodic base production in the presence of acetic acid and not of water. In both cases base production should theoretically match acid production exactly.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

41

Section 2 (Fig. 1, curves A and B), usually performed at the rotating platinum electrode (anode reactions) or the dropping mercury electrode (cathode reactions), should ideally suffice t o define the electroactive species and determine its half-wave potential. It may be that systems in which acid-base equilibria exist are somewhat more laborious to study due to the necessity of recording voltammetric curves over a wide pH range, but in most cases the task can be accomplished with some effort. Once the voltammetric characteristics are known, it remains to carry out preparative constant potential electrolysis (cpe) at a suitable potential in order to make sure that the electroactive species is connected with the reaction of interest.

Solvent and Supporting Electrolytes This sounds and is simple as long as systems displaying the ideal behaviour depicted in Fig. 1 are involved. However, many organic electrode reactions belong t o the less well-behaved category, in which the voltammetric curve with substrate added is identical or almost identical with the background curve (solvent + supporting electrolyte) and from which practically no mechanistic information can be extracted. In such cases, it is much more difficult to find out unambiguously the nature of the electroactive species. We shall discuss some cases of such behaviour, but let us first look into one aspect of the practical side of the problem, namely, the solventsupporting electrolyte system t o be used for voltammetric studies. A number of combinations are shown in Table 4; in view of the large number of new systems that have been developed in recent years, only a representative selection is shown (for additional information, see Lund and Iversen, 1973; Mann, 1969). Most of the solvent-supporting electrolytes (SSE) in Table 4 have been deliberately designed t o exhibit as wide an accessible potential range as possible in order t o render possible electrochemical studies of substrates which are difficult t o oxidize and/or reduce. The range covered in Table 4 is roughly between -3.5 and 3.5 V, thus providing driving forces for electrolytic reactions of the order of 3 eV or 70 kcal/mole (Fleischmann and Pletcher, 1969a). The ideal behaviour exhibited by, e.g., SSE no. 14 should, however, be contrasted with the non-ideal properties of, e.g. nos. ( 5 ) and ( 6 ) ; yet these latter ones are of considerable synthetic interest, since they

TABLE 4

b P

N

Accessible P o t e n t i a l Rangesa for Some R e p r e s e n t a t i v e S o l v e n t - s u p p o r t i n g E l e c t r o l y t e C o m b i n a t i o n s [in V versus t h e Saturated Calomel Electrode (SCE)b] No.=

1 2 3 4 5 6 7 8 9 10

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Solvent A. Protic solvents Water Hydrogen fluoride (liq.) Ammonia (liq.) Fluorosulphuric acid Methanol Methanol Formic acid Acetic acid Trifluoroacetic acid B. Aprotic solvents Acetonitrile Ace tonitrile Acetonitrile Acetonitrile Acetonitrile Ace tonitrile Acetonitrile Acetonitrile Acetonitrile Acetonitrile N, N-Dimethylacetamide N,N-Dimethylformamide N, N-Dimethylformamide N,N-Dimethylformamide N, N-Dimethylformamide Dimethylsulphoxide Dimethylsulphoxide Hexamethylphosphortriamild& Methylene chloride

Supporting electrolyte

B~4NC104 BF;,~ P F , ~ BQNI CH3COOH2F S 0 3 KOH LiC104 HCOONa CH3COONa C F 3 COONa

Anodic limit (electrode)

1'5 (Pt) 3.0 (Pt)

Cathodic limit (electrode)

-2'7 (Hg) -2.8 (Hg)=

>2.0 0.6 1.3 1.0 2.0 2.4

(Pt)f (Pt) (Pt) (Pt) (Pt) (Pt)

NaBF4 Et4NBF4 E t4NP F 6 Bu~NBF~ Et4NCF3S03

3.7 (Pt) 3.5 (Pt) 3.6 (Pt)

LiC104 NaC104 Et4 NC104 B~4NC104 MeEt3NOTs Et4NC104 Bu~NCF~SO~ Et4NC104 Et4NC104 Et4NC104 E t4NC104 Et4NC104 LiC104 Et4NC104

2.6 (Pt) 2.6 (Pt) 2.9 (Pt) 2.6 (Pt) 3.3 (Pt) 2.0 1-6 (Pt)

3.2 (Pt)

0.5 1.0 1.9 1.5 2.1 1.0

(Hg) (Pt) (Pt) (Pt) (Pt) (Pt) 0.8 (Hg)

-1.0 (Pt) -1.0 (Pt) -1.0 (Pt) -0.6 (Pt)

r m W

kl ..

2

*

2

U

?: -2.8 -2.6 -2.3 -3.2 -1.7 -2.9 -2.7 -2.2 -2.7 -2.5 -3.0 -4.0 -2.8 -2.8 -2.3 -3.4 -1.9

(Hg) (Hg); (Pt) (Pt) (Pt) (Pt) (Pt) (Pt) (Pt) (Hg) (Hg) (Hg) (Pt) (Pt) (Pt) (Pt) (Hg)

2

2

m

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

Methylene chloride Methylene chloride Nitrobenzene Nitromethane Propylene carbonate Propylene carbonate Pyridine Pyridine Sulpholane Sulpholane Sulpholane Tetrabutylammonium nitrate' Tetrahexylammonium b e n z o a t d Tetrahydrofuran Tetrahydrofuran Tetrahydrofuran Sulphur dioxide Antimony trichloridek (0.1 M i n AlCl3)

Et4 NC104 Et4NCF3S03 Pr4NC104 LiC104 Et4NC104 E t 4 NC104 Et4NC104 Et4NC104 Et4NC104 NaC104 Et4NC104 -

LiC104 Bu~NI Bu~NCF~SO~ Bu~NBF~ -

1% (Pt) 1% (Pt) 1.6 (Pt) 3.0 (Hg) 1.7 (Pt) 1.7 (Pt) 1.7 (Graphite*) 3.3 (Pt) 1.2 (Hg) 3.3 (Pt) 2.3 (Pt) 2.1 (Pt) 0.3 (Pt) 1.8 (Pt) -0.9 (Hg) 1.8 (Pt) 3.0 (Pt) 1.5 (Pt)

-1.7 -1.7 -0-7 -2.4 -1.9 -2.5

(Pt)

(Hg) (Pt) (Pt) (Pt)

(Hg)

-

-2.2 -2.3 -1.3 -2.9 -2.4 -1.2 -3.3 -3.7 -3.0

(Pt)

(Hg) (Pt)

(Pt) (Pt)

(Pt) (Pt) (Hg) (Hg)

-0.9 (Pt)

a The definition of the anodic and cathodic limit may vary somewhat with different sources, depending on the intended use of the electrolyte. For analytical purposes the limit is often taken to be the potential at which the current density exceeds 1 UAcm-2; for synthetic studies it is often put as high as 1 mA cm-2. No distinction between these cases has been made, however, since the d a b are only intended to give an illustrative view of the possibilities of choosing SSE systems in organic electrochemistry. For uniformity, all values are given on the sce scale, except in one case (for reference electrode comparisons used, see Mann and Barnes, 1970, pp. 26-27). These numbers also key the references to the literature (see below). Nature of the cation not stated. Formation of solvated electrons begins. f Versus the Pd/H2 electrode in the same solvent. g Water (0.3 M ) was added. Pyrolytic graphite with a6 plane exposed. Used as a melt at 150°C. Used as a melt at 93°C.

*

0

w

0

?i i;

*

References (1) Lund and Iversen, 1973. (2, 45) Doughty et al., 1972. (3, 22, 28, 29) Mann, 1969. (4) Bertram e t al., 197 1. (5) Weinberg and Reddy, 1968. (6, 17, 19, 2, 24, 25, 36) Andreades and Zahnow, 1969. (7) Ross et al., 1966. ( 8 ) Eberson and Nyherg, 1964, 1966. (9) Petit and Bessiere, 1971. (10) O'Donnell, 1965. (11, 12) Fleischmann and Pletcher, 1968; Osa et al., 1969. (13) House e t a l . , 1971. (14, 21, 30.44) Rosseau et QL, 1972. (15, 16, 18) Billon, 1960. (23) Breant et aL, 1963. (26) Courtot-Coupez and Le DCmkzet, 1967. (27) Duhois et al., 1966. (31) Marcoux e t al., 1967. (32) Cauquis and Serve, 1966. (33, 34) Nelson and Mams, 1967. (35) Turner and Elving, 1965. (37.38) Headridge et aL, 1967. (39) Coetzee e t aL, 1969. (40) Woodhall and Davies, 1969. (41) Swain et al., 1967. (42, 43) Peridon and Buvet, 1968. (45) Miller and Mayeda, 1970. (46) Baueret al., 1971; Bauer and Beck, 1971.

.p

w

44

L. EBERSON AND K. NYBERG

sustain alkoxylation reactions of added substrates (for a review, see Fiochin et al., 1973) and hence have been extensively-and controversially-studied from the mechanistic point of view. Many of the SSEs in Table 4 contain perchlorate salts. It should be stressed-and has been so repeatedly'-that this practice is now unnecessary in the overwhelming majority of cases and should be abandoned in favour of the use of fluoroborates, hexafluorophosphates, or trifluoromethanesulphonates as their tetraalkylammonium salts (Eberson and Olofsson, 1969; Fleischmann and Pletcher, 1968; Rosseau et al., 1972). These have excellent solubility properties and extreme limits on both the anodic and cathodic side.

TABLE 5 Accessible Potential Ranges in Dimethyl sulphoxide Containing Different Supporting Electrolytes (0.1 M) on Platinum o r Vitreous Carbon (in V versus scey Supporting electrolyte

Anodic limit

Et, NC104 LiC104 KC104 NaC104 m 0 3

KB F4 K2 s 2 0

8

LiCl Me4 NCl Bu4 NBr

2.1 2- 1 2.1 2.1 2-1 2-1 2.1 1-5 1-5 1.4

Cathodic limit

-2-3 -2.7 -2.3 -2.1 -2.3 -2.3 -2.3 -2.7 -2.4 -2.4

a Courtot-Coupez and Le DGmPzet, 1967.

Inspection of Table 4 reveals that the electrochemical processes limiting the use of a particular combination involve either solvent or supporting electrolyte. In, e.g., acetonitrile (nos. 10- 19) the anodic limit is dependent on the nature of the anion and the cathodic one on the nature of the cation. On the other hand, in a solvent like dimethyl sulphoxide the anodic limit is due to oxidation of solvent for anions which are difficult to oxidize (nos. 25, 26; cf. also Table 5) and of anions in cases of easily oxidizable ones, whereas the An explosion (appropriately enough, in a military laboratory) has actually been reported to have occurred during an electrolysis experiment in which a perchlorate w a s used as the supporting electrolyte (Titus, 1971). Whatever the chemistry involved, this accident merits attention from large-scale practitioners of organic electrolysis.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

45

cathodic limit is again governed by the nature of the cation. In nitrobenzene (31) solvent reactions appear to be the limiting ones on both the anodic and cathodic side. Thus one can order anions in a series that qualitatively shows the increasing degree of resistance toward anodic oxidation, viz., I-

< Br- < C1- < NO3 < CH3COO- < ClO, < CF3COO- < ClO, < CF3SO; OTs- < BF; PF; X

Similarly, cations of common use in organic electrochemistry can be ordered according to their resistance towards cathodic reduction, viz., Na+

< K+ < R4N+ < Li'

With regard to electrode material, it can be seen (Tables 4 and 5) that cathodic limits on mercury are displaced by a few tenths of a volt to more negative potentials than on platinum. On the anodic side, the number of practically useful electrode materials is limited to noble metals and different types of carbon; one case (anodic limit of pyridine: nos. 35 and 36) shows that the anodic limit is lower on graphite than on platinum, and this seems to be a general trend for the comparison of carbon based anode materials, except possibly for vitreous carbon (Table 5) and bright (smooth, polished or shiny) platinum. We have just mentioned that one reason for a limited range of potentials in a particular SSE is the reactivity of the components of the SSE toward oxidation and reduction. It is also obvious that the limiting cathodic process in protic solvents, nos 1-9 in Table 4, must be reduction of protons or the equivalent, the proton donor. The unfavourable cathodic limit for reduction of protons can, however, be vastly improved by the use of mercury as the cathode material and a tetraalkylammonium salt as SSE (nos. 1 and 3). The reason for mercury being such a favourable material is its large overpotential (see Section 10) for the reduction of protons (hydrogen evolution reaction). We have already commented (p. 24) on the fact that the reduction of protons occurs many orders of magnitude faster on certain metals than on others, and this manifests itself by the ov@rpotentiuZ, i.e., in order to make the reaction go at a measurable rate one has to increase the electrode potential from the equilibrium potential. Table 6 shows overpotentials for hydrogen evolution and

46

L. EBERSON AND K. NYBERG

the corresponding equilibrium exchange current density, i, , the current density at the equilibrium potential. At anodes, aqueous electrolytes behave similarly with an overpotential for oxygen evolution. What one sees in effect in Table 6 is that a suitable choice of metal can move the potential limit for a certain process, e.g. the hydrogen

TABLE 6 Overpotentials' and Equilibrium Exchange Current Densitiesb for Hydrogen Evolution and Overpotentials for Oxygen EvolutionC a t Different Metals in Aqueous Medium Hydrogen overpo t en tial, Metal

V

Palladium Gold Iron Bright platinum Silver Nickel Copper Cadmium Tin Lead Zinc Mercury

0.00 0.02 0.08 0.09 0.15 0.21 0.23 0.48 0.53 0-64 0.70 078

Exchange current density for hydrogen evolution, A cm-2

Oxygen overpotential,

10-3 4 x 10-6

0.43 0.53 0-25 0.45 0.4 1 0.06

10-6 9 x 10-6 2 x 10-7 3 x 10-11 10-12

V

0.43 0.3 1

5 x 10-13

'In aqueous sulphuric acid, 0 5 M, 5 x 10- '

(Allen, 1958). In aqueous sulphuric acid, 1 M (Bockris and Reddy, 1970, p. 1238). In aqueous potassium hydroxide, 1 M (Allen, 1958, p. 5).

evolution reaction at the cathode, over a considerable range, thus permitting the oxidation or reduction of otherwise "inert" substrates. This is a very important phenomenon in electrochemistry. It can also appear in the form that a low-potential process taking place at a certain electrode material at a certain potential can be slowed down and eventually almost completely inhibited by a second, high-potential process. The Kolbe reaction [e.g. eqn (53)] in aqueous medium is of this type, displacing oxygen evolution on platinum in the region of 1.5-1-8 V by carboxylate oxidation above 2-2 V. This phenomenon will be discussed in more detail in Section 12.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

47

Chemical Modifications of the Substrate Before Electron Transfer The following modifications of the substrate, apart from purely chemical transformations, have been observed: complexation with a r-donor (acceptor), complexation with a metal ion, ion pair formation (charged substrate), acid-base reactions. Of these, acid-base reactions are by far the most important and best studied cases, since cathodic processes in water-containing media are often strongly dependent on the pH. In superacidic media interesting consequences of protonation appear at the anode, too.

Complexation with a n-donor (acceptor) Charge transfer (CT) complexes are kept together by rather weak forces, and it is not to be expected that such forces should influence their electrochemical behaviour significantly. Thus, the CT complex between tetracyanoethylene and hexamethylbenzene has its halfwave potential for reduction shifted 0.039 V towards a more negative potential as compared to tetracyanoethylene itself (Peover, 1967) as is predictable from theoretical considerations of the formation of the CT complexes. Cation (anion) radicals of aromatic hydrocarbons should in principle be strong .rr-acceptors (n-donors) and it is interesting to speculate that CT complexes between substrate and radical ion might play a role in coupling reactions, such as the biaryl coupling mechanism shown in eqn (57). A CT complex of the [ArH--ArH] type would certainly be more difficult to oxidize than ArH. Thus ArH might be protected from oxidation, and coupling within the complex would take place instead. '+

Complexation with a metal ion One example of this kind of interaction is the anodic oxidation of propene which changes direction upon addition of mercuric ion

48

L. EBERSON AND K. NYBERG

(Clark et al., 1973a; Fleischmann et al., 1969, 1970) as shown in eqns (61) and (62). The latter process has other interesting characteristics, t o which we shall return later (p. 98).

1 M HC104

Hz0 CHzzCHCH3

____+

Hg+

-H+ lox. -2e-

CHzCH--CH3 I + ' Hg OH

CH3COCH3

(62)

CH3COOH + HCOOH

Interesting observations were made when the anodic oxidation of aromatic hydrocarbons was studied in the molten salt system AlC1, -NaCl-KCl, which can be conveniently handled at 150" (Fleischmann and Pletcher, 1970). The most intriguing finding was that the oxidation potentials of all compounds were shifted very strongly towards less positive potentials, as exemplified by benzene which oxidizes a t about 2.4 V versus sce in acetonitrile/tetrabutylammonium hexafluorophosphate (Osa et al., 1969) but was estimated to be shifted t o ca. 0-8 V in the melt. Two mechanisms were believed to be responsible for this behaviour: (i) the complex between aluminium chloride and the hydrocarbon might be easier t o oxidize than the hydrocarbon (less likely in view of what was said above), and (ii) assistance, seemingly, of the electron transfer step by the anion AlC1,. The latter hypothesis is difficult to reconcile with the fact that anions do not affect oxidation potentials to any large extent in other media. The authors did not consider the possibility of purely chemical reactions taking place to give more easily oxidized compounds under the very forcing conditions in the melt (containing 50 mole % AlCl,); even benzene is known t o polymerize at 100°C under very high pressures using A1C13 as the only catalyst (Gonikberg and Gavrilova, 1952). Moreover, the melt was exposed t o hydrogen chloride in the course of its preparation, thus creating conditions for the formation of HA1C14, an extremely strong acid (see e.g. Bauer and Foucault, 1972) and a good catalyst for benzene polymerization under oxidative conditions (Kovacic, 1963).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

These suggestions aluminium chloride' matic hydrocarbons malies being found compared t o more

49

are supported by the fact that nitromethane/ can be used for anodic voltammetry of aro(Bauer and Foucault, 19 72) without any anowith respect to the scale of potentials as normal SSEs (Table 7). Also, acetonitrile/ TABLE 7

Half-wave Oxidation Potentials of Some Aromatic Compounds in SSEs Containing Lewis Acidsa Compound 9,lO-Diphenylanthracene (0/1) 9,1 O-Diphenylanthracene (1/2) Perylene Anthracene 1,4-Dimethoxybenzene Naphthalene Ferrocenee Te tracene Coronene Pyrene

CH3N02/ CH3CN/ EhNC104d AlC13d LiC104d AlC13d SbC13boc

0.88

0.88

0.90

0.90

0.39

1.28 065 0.83

1.28 0.65 0.83

1.28 0.74 0.89

0.70 0.90

0-24 0.51

093 1.30 0

0.93 1.30 0

1-04 1-35 0

1.35 0

'

0.21 0-64 0.59

Data taken from Bauer and Foucault, 1972; Bauer et al., 1971. Referred to the Sb/SbCl,, satd. KCl in molten SbCI, , electrode. At 99°C. Referred to the ferrocene/femcinium couple. The E l l 2 of the ferrocene/ferricinium redox couple is assumed to be almost independent of the medium, thus enabling comparisons of this type. a

aluminium chloride appears to be a well-behaved SSE. Another melt of the Lewis acid type, antimony trichloride at 99OC containing 0.1 M AlCl,, has been used for anodic polarography (Table 7) without any apparent problems (Bauer et al., 1971; Bauer and Beck, 1971).

Ion pair f o r m a t i o n

Only a few cases of effects of ion pairing upon the substrate are known, simply because such effects are not easy t o demonstrate due 1

In this solvent mixture, nitromethane slowly undergoes a cleavage reaction to give an appreciable concentration (0.05 M) of NO' (Bauer and Foucault, 1972a). However, this species can be removed by cpe before the mixture is used as an SSE.

L. EBERSON AND K. NYBERG

50

to their small magnitude. One example is to be found in the Kolbe oxidation of acetate ion [eqn (56)] ,the half-wave potential of which is very dependent upon the cation of the supporting electrolyte (Fleischmann and Pletcher, 1973) viz.,

E l l 2 versus Ag/Ag+

Et4N'

K+

H+

1.2 (1.35)'

1.6

2.9

Although the substrate in the acidic SSE really is the acetic acid molecule and not an ion pair, the cation has a fairly pronounced effect on E l l 2 in the other two cases. Obviously, the looser ion pair should be the easier one to oxidize, as is also found. It would be interesting to see if complexing of the potassium ion by a suitable crown ether would lower the half-wave potential still further.

Acid-base reactions The effect of acid-base reactions preceding electron transfer is best known by far for cathodic processes, since aqueous and aqueous organic SSEs, often buffered, can in most cases be used for the cathodic reduction or organic compounds without interference from the background reaction (due to the high hydrogen overpotential of mercury and favourable influence of tetraalkylammonium ion upon the cathodic limit; see Tables 4 and 5 ) . Thus the role of the pH of the SSE in cathodic reduction is well understood and has been the subject of an exhaustive review (Zuman, 1969). The special case of quinones has been thoroughly treated recently (Chambers, 19 74; cf. also Parker, 1973). We shall therefore not try to cover here the vast amount of work done on electrochemical reactions in water-containing media. It is, however, pertinent to summarize a few important points:

( a ) Changes in pH may affect the voltammetric curve in three ways, viz., by changing the half-wave potential, the limiting current and/or the shape of the wave. If there is no variation of these parameters with pH, the substrate itself is the electroactive species. ( b ) If a substrate is not electroactive in itself or is at least very difficult to reduce at the cathode, it can be made more easily This value was obtained (Geske, 1959) with Pr4NOAc (0.3 mM) in acetonitrile/ tetrabutylammonium perchlorate (0.1 M).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

51

reducible by protonation. As the acid-base equilibrium is gradually shifted toward the side of the protonated form by lowering the pH, the half-wave potential will change towards less negative values until it reaches a constant value when the equilibrium has been driven completely to the side of the protonated form. Deprotonation of a substrate in a cathodic process will lead t o a form which is more difficult to reduce. (c) Conversely, a substrate that is difficult or impossible to oxidize can be made more easily oxidized by deprotonation (for example, see the Kolbe oxidation of acetate ion versus acetic acid molecule on p. 50). Protonation of a substrate at the anode should lead to a protonated form that is more difficult t o oxidize. Summarizing, for a substrate R-H the following orders of ease of electrochemical oxidation and reduction prevail as a general rule. Reduction ( b ) : RH; easier than RH easier than R-. Oxidation (c): R- easier than RH easier than RH;. This is of course to be expected from considerations of electrostatic interactions alone, all other factors being left aside. While examples corresponding to ( a ) , ( b ) and the deprotonation case of (c) can be counted literally in thousands, the effect of substrate protonation in anodic oxidation is less well documented. However, amines and other nitrogen compounds have been thoroughly investigated on this point (Adams, 1969) and found to behave normally, but some recent work on anodic reactions in superacidic media has revealed a theoretically interesting exception to the rule. This concerns the anodic oxidation of alkanes and cycloalkanes in fluorosulphuric acid (Table 4, no. 9) with varying concentrations of added base, potassium fluorosulphate and/or acetic acid (Bertram et al., 1971, 1973). The overall chemistry of this reaction is shown for cyclohexane in eqn (63) and the interesting part of the reaction sequence in eqn (64). As is known from the chemistry of organic compounds in CH3 1

R = alkyl, cycloalkyl

52

L. EBERSON AND K. NYBERG

superacid media (for a review, see Brouwer and Hogeveen, 1972) protonated alkanes cleave spontaneously to form hydrogen and R + under similar conditions, but this possibility was excluded in the electrochemical experiments by proper adjustment of the base concentration. It was then found that E , 1 2 increased with increasing base concentration, indicating that the protonated species is the electroactive form and thus easier to oxidize than the alkane, contrary to what the general rule predicts. This observation was explained in terms of the unusual electronic properties of pentaco-ordinated cations (Kollmar and Smith, 1970). The CH: species is best described as a hydrogen molecule with an abnormally long bond, 0.94 to which a deformed trivalent carbonium (carbenium) ion is coordinated. The energy of CH: was calculated to be 47 kcal mol-' lower than that of H2 + CH; (gas phase value) and two of the C-H bonds only of half the strength of a C-H bond in methane. This electronic structure, perhaps best being described as closely related to protonated hydrogen, is probably the reason for the reverse order of oxidation potentials observed. For aromatic hydrocarbons which form carbenium ions upon protonation, the order of oxidation potentials conforms to the rule. The peak potential for the proton adduct in methylene chloride/7% CF3S03H was found (Hammerich and Parker, 1974) to be displaced about 1.5 V towards more positive potentials than that of the parent hydrocarbon for a series of 9,lO-disubstituted anthracenes. We can use this potential difference to estimate E,12 for protonated hydrogen at about 1.5 V versus the Pt/H2 electrode and, accordingly, a methyl cation coordinated to a hydrogen molecule (ECH: ) would be expected to have its half peak potential somewhere around this value. For protonated hydrocarbons, half peak potentials fall in the region of 1.9- 2.2 V versus the Pd/H2 electrode.

'

a,

Some Case Studies

Doubts as to which component of the electrolyte solution is the electroactive species seem to arise predominantly in anodic reactions; this is quite natural in view of the range of useful anions, spanning a range of oxidation potentials between 0 and 3-5 V, available for E p p is the so-called half peak potential, measured on peak voltammograms as the potential at half the peak current.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

53

making up SSEs. For cathodic processes we are more limited in our choice of cations, alkali metal or ammonium ions being by far the most commonly used ones, and since these are reduced at rather negative potentials, the electroactive species is easily identified in the presence of these ions. Proton reduction is also no problem (p. 46). However, the cathode has its particular problems, and these have to do with the generation of “active hydrogen”, solvated electrons, and amalgams.

Anodic substitution As a prelude to the discussion of some anodic reactions, Table 8 shows the present state of knowledge with respect t o the electroactive species in a number of representative anodic reactions of different types. Perusal of Table 8 reveals that the problem of identifying the electroactive species is not solved for a number of reactions (nos. 3, 5, 23, 27, 28, 29) and that recent work utilizing modern electrochemical techniques has revived the discussion of some important ones (nos. 14, 15) and paved the way for renewed attempts at mechanistic discrimination. We shall deal with some of these reactions in the following discussion. Before we proceed, we shall anticipate the treatment of reactivity in Section 11 by mentioning that most of the organic substrates studied in reaction nos. 1-41 (Table 8) have E l 1 2 for oxidation in the region between 1.0 and 2-0 V versus sce (see Table 21 for examples). A few classes of compounds, notably alkanes and simple olefins, have higher E l l 2 values, and a few have lower ones due to extensive conjugation by oxygen, sulphur, and/or nitrogen functions.’ An average substrate of interest for synthetic purposes would typically have a half-wave potential for oxidation of 1.6-1.7 V. With this in mind, it is obvious from the discharge potential data of Table 8 that a number of nucleophiles ( 3 , 11, 14, 15, 19, 34, 35) are oxidized at potentials far below 1-6 V, an E l l 2 corresponding to substrates which have been successfully used in all the reactions and yet for which direct mechanisms have been established. Intuitively, one would like to think that a high-potential process would not take The

most

extreme

case

appears

to

be

Ell2

for

(Me2N)2C=C(Me)C(Me)=C(NMe)2, -0.9 V versus sce (Fritsch et al., 1970).

oxidation

of

54

L. EBERSON AND K. NYBERG

TABLE 8 Nature of the Electroactive Species in Some Anodic Reactions of Different Types

No.

Reaction type

Nucleophile

Discharge potentiala of nucleophile versus sce

A. A d d i t i o n reactions [eqn ( 1 5 ) ] 1 Addition of fluorine HzF;

2 3

4

5

groups Chloroacetamidation Addition of methoxyl groups Addition of acetoxy groups Addition of aroyloxy groups

6 Addition of thiocyano groups 7 Addition of azido groups 8 Addition of pyridine groups 9 Addition of nitromethyl groups 10 Addition of enolate groups

1 1 Addition of cyano

> 2.0'

C1CH30-

l.oe

CH3COO-

2.0'

0.6

PhCOO-

Not known

SCN-

0.55;' 0.3m

N3 -

1.00

C.jH5N

CH~NO~

> 3.34 < 0.5'

Enolate ionsf

0.8- 1.2'

CN-

0.9"

Mec hanismb

Directd Indirect via Clze Direct in most casesf;. .g, Directs1 Not settled suggested rt t o be indirect via PhCOO. Indirect via (SCN)2 * Indirectp Direct' Indirect via OzNCHz ' S Indirect via enolate radicalsU DirectX

groups B. Cleavage-addition reactions [eqn ( 1 7 ) ] RCOOradicals RMgX

12 Addition of alkyl

C. Substitution reactions [eqn ( 9 ) ] F- (in liq. HF)

13 Fluorination

2.0'

> 2.5

14 Chlorination

c1-

l*oe

15 Bromination

Br-

0.7dd

16 Iodination

112, CH3CN

17 Iodination

-

Indirect via R Indirect via R *

J'

Indirect via NiF3aa Assumedbb t o be indirect via Clz, b u t an exception is knowne,'' Indirect in many cases,bb b u t exceptions have been demonstrated" Indirect via l z b b Indirect via N-iodoacetonitrilium ion and/or N-iodoace tamidegg

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

55

TABLE 8-continued

No.

Reaction type

Nucleophile

Discharge potential" of nucleophile versus sce

18 Hydroxylation

19 20 21 22 23

Methoxylation Formyloxylation Acetox ylation Acetoxylation Acetoxylation

24 Trifluoroace toxyl. ation 25 Propionoxylation

CF3COOH

3.PS

CzHSC00-

1*7uu

26 Acylox ylation, intramolecular

Ar-Alk-COO-

Not known

27 Aroyloxylation

ArCOO-

Not known

28 Intramolecular aroyloxylation 29 Substitution by nitrate ion 30 Thiocyanation 31 Ace tamidation 32 Azidation 33 Pyridination 34 Cyanation

Ar- ArlC00-

1.6

NO; CH3CN CN-

> 3.5 0.gv

D. Coupling-addition reactions [ e q n ( 3 0 ) ] 35 Coupling with addition CH300-6 of methoxy groups 36 Coupling-addition of nitromethyl and enolate groups

E. Coupling-elimination reactions [ e q n (29)j 37 Coupling of anions, such as Kolbe coupling of RCOO- and coupling of anions of active methylenc compounds

Mechanismb Direct and indirect mechanisms have been discussed;bb.ii one case has been shown to occur via anodically generated PbOzJJ See reaction no. 3 Indirect via HCOO- kk See reaction no. 4" Direct"" A blend of two mechanisms has been suggested; one direct and one indirect via NO3 ' Directn Claimed to be indirect via EtCOO * uu Both directvv and indirectxx mechanisms proposed Indirect mechanisms via ArCOO has been proposedyy Direct mechanism has been proposed'' See reaction no. 23 See reaction no. 6 DirectSee reaction no. 7 See reaction no. 8bbb Directv,ccc See reaction no. 3Prddd See reactions nos. 9 and 10

56

L. EBERSON AND K. NYBERG TABLE 8-continued

No.

Reaction t y p e

38 Intermolecular biaryl coupling (2ArH + Ar-Ar + 2H+ + 2e-) 39 Intramolecular biaryl coupling 40 Diphenylmethane coupling (2ArCH3 + H3CAr’CHZAr + 2H’ + 2e-)

Nucleophile

Discharge potential‘ of nucleophile versus sce

ArH

Changes w i t h A r H

Ar-Ar)

Changes with ArAr’ Direct=

ArCH3

Changes w i t h ArCH3 Directfff

F. Cleavage-coupling-addition reactions [ e qn ( 3 1 ) ] 41 Coupling-addition of RCOOalkyl radicals

Mechanismb Directffl

See reaction no. 12

RMgX

a This concept, sometimes called the foot potential, is only an approximate one; it is taken as the potential at which the nucleophile begins to be discharged at an appreciable rate e.g., at 1-5 mA ern-'. Classified as direct or indirect (Section 5). O’Malley, 1973. Ludman et al., 1972. Faitaet al., 1970. This reaction is potential dependent, following an indirect mechanism between 0-7 and 1.2 V and a direct one at high potential, 2.1 V (see reaction no. 14). f Weinberg and Reddy, 1968. gBaggaley and Brettle, 1968; Parker, 1970a; Parker et al., 1971; Ross e t al., 1969; Weinberg, 1968; Weinberg and Belleau, 1973. /I Originally postulated to be homolytic via methoxy radicals. Eberson and Nyberg, 1964,1966. I Mango and Bonner, 1964; Parker, 1970a, 1970b; Parker e t a L , 1971. kKoyamastal., 1968. I Chang et al., 1971. Cauquis and Pierre, 1968. DeKlein, 1973. Ward and Wright, 1964. p Schiifer, 1970. Andreades and Zahnow, 1969. Lund, 1973; for more recent work, see: Blount, 1973; Parker, 1973; Svanholm and Parker, 1973. Schiifer, 1969, 1970. Diethoxycarbonyl-, diacetyl-, and acetylethoxycarbonylmethanideion. Schafer and Azrak, 1972. Parker and Burgert, 1965. * Longchamp et al., 1974; Parker and Eberson, 1972; Yoshida et al., 1971. J’ Fiochin et al., 1963; Lindsey and Peterson, 1959; Schafer and Pistonus, 1972; Smith and Gilde, 1959, 1961; for reviews, see: Eberson, 1968, 1973a. Schafer and Kiinzel, 1970, aa The reaction is run at a nickel anode (Burdon et al., 1972; Burdon and Tatlow, 1960; Nagase, 1967). bb Allen, 1958; Tomilov, 1961. cc Allylic chlorination takes place at high potential via a direct mechanism (see also reaction no. 2).

* ”

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

57

TABLE 8-continued dd Kolthoff and Coetzee, 1956.

ee Millington, 1969. ff Oxidation of iodine. gg Miller, 1968; Miller et al., 1970. In aqueous sulphuric acid at a PbO, anode. Weinberg, 1974. Additive hydroxylation proceeds via a direct mechanism (Parker, 1970c, 1970d, 1970e; Soda, 1968). ii Nilsson, A., et al., 1973. kk Ross et al., 1964a, 1966. I1 Cf. also Leunget al., 1965; Ross et al., 1964. rnrn Fleischmann and Pletcher, 1973. nn The anion of the SSE has been ClO;, BF; or OTs- (Eberson, 1967; Magnusson et al., 1971; Ross et al., 1970). O0 Ross et al., 1970. pp Nyberg, 1970a. 44 Formaro et aL, 1973; Rao et aL, 1970. rr Formaro et al., 1973; Nyberg, 1970a; Ross et al., 1967, 1970, 1972. Clark e t al., 1973. rt Eberson et al., 1973a; Nyberg and Trojanek, 1975; Svanholm and Parker, 1972. “ Kunugi et al., 1970. Eberson, 1968, 1973a. xx Bonner and Mango, 1964; Bunyan and Hey, 1962; Koehl, 1967. J’y Aniskova et aL, 1973; Koyama et aL, 1966, 1968. C’ also: Matsuda et aL, 1973; Wilshire, 1963. zz Eberson and Nyberg, 1966. aaa Clark et al., 1972, 1973, 1973a; Coleman et al., 1968; Eberson and Nyberg, 1964, 1966a; Parker and Burgert, 1968; Parker, 1969. bbb Cf. also: Blackburn and Will, 1974; Ikenoya et al., 1974; Masui and Ohmori, 1973. ccc Andreades and Zahnow, 1969; Eberson and Nilsson, 1968; Nilsson, 1973; Tsutsumi and Koyama, 1968. Reactions nos. 3 and 35 compete normally, no. 35 being favoured at graphite anodes and no. 3 at Pt (Belleau and Au-Young, 1969; Eberson and Schafer, 1971; Katz et al., 1974; Schafer and Steckhan, 1969). For reviews, see: Eberson and Nyberg, 1976; Nyberg, 1973. fffEberson etal., 1973a; Nyberg, 1970, 1971, 1971a, 1971b, 1971c, 1971e; Nybergand Trojanek, 1975; R o d i n e t al., 1973a; Ron& and Parker, 1974a. ggg Falck et al., 1974; Ronlin et al., 1973, 1974; Svanholm et al., 1974a.

’’ ””

precedence over one occurring at lower potential when they are run together. But this is what happens (cf. the Kolbe reaction, p. 98, a very clearcut example) and what makes electrochemistry exciting. One can always be optimistic before a new experiment! To establish a direct mechanism in such a case is fairly straightforward. One carries out cpe at a series of potentials and analyses for products at each potential. If the mechanism is direct, relatively high currents will pass at all potentials, but the products derived from the substrate do not appear until the potential is high enough t o discharge the substrate, the E l of which one has to determine in an

58

L. EBERSON AND K. NYBERG

SSE with a high enough anodic limit (Table 4). In this way, the direct mechanisms of anodic methoxylation (nos. 3, 11, 19 and 35), and cyanation (nos. 11 and 34) have been established. For chlorination (no. 14) and bromination (no. 15) i t remains to be seen whether the cases of the direct mechanism found hitherto can be generalized. Aroyloxylation (nos. 5, 27, and 28) both additive and substitutive, has been proposed to be an indirect reaction, as has also formyloxylation (no. 21). On the other hand, the other acyloxylation reactions taking place in the presence of oxidation-resistant anions (nos. 21, 22, and 24) have been definitely shown to be direct processes. Unfortunately there is no recent value for the discharge potential of benzoate ion; an old investigation (Bose, 1898) places it at about 1.5 V versus sce in aqueous solution, but then it should be noted that this value pertains to benzoic acid dissolved in water. Thus the value for benzoate ion might be still lower.' So it is really not easy to refute off-hand the proposal that aroyloxylations proceed via aroyloxy radicals that attack the substrate in a homolytic reaction, followed by a second electron transfer from the intermediate to the electrode and proton loss. Contrary to the behaviour of alkoyloxy radicals, aroyloxy radicals are fairly stable toward decarboxylation, as seen from Table 9, which gives estimated values for the rate constants at 20°C for a number of such decarboxylations. Hence there will be enough time for an aroyloxy radical to react before cleavage. Characteristically enough, it is only with great reluctance that an aromatic carboxylate will undergo Kolbe coupling (for a summary, see: Eberson and Nyberg, 1976). In homogeneous reactions (thermal decomposition of diaroyl peroxides) aroyloxy radicals have life-times long enough to give substitution products from, e.g. benzene (Williams, 19 60). Clearly, anodic aroyloxylations involving substrates more easily oxidized than benzoate ion should give a better background for judgements on the mechanistic pathways possible. The rate constants of the alkoxycarbonyl radicals in Table 9 have been estimated to provide an idea of how stable a radical of this type can be and yet react with a substrate before decarboxylation. Alkoxycarbonyl radicals enter reactions of the type shown in eqns Note that neat tetrahexylammonium benzoate (SSE no. 41, Table 4) has a reported anodic limit of about 0.3 V versus SCE. This value is, however, given for currents of the order of /.A and cannot be taken as the discharge potential for benzoate ion. Even traces of impurities might interfere at this current level, bearing in mind that the benzoate ion concentration is about 2 M.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

59

TABLE 9 Estimated Rate Constants for the Decarboxylation of Acyloxy and Alkoxycarbonyl Radicals a t 20°C

Radical CH3+2OO* C2H5-C00. C6H5-cOO. C~HS-COO. t-BU-OCO. Me-OCO.

Rate constant from literature, s-l ("C)

1.6 x l o 9 (60)a 1.6 x lo9 104 (80)~ 1 O ' O (356)g

Ea

Estimated rate constant a t 2OoC ( s - l )

6.6' 6-6b 15-4d 18e

4 x 108 4 x 108 102

7.7f

105

lo-@

106

( k c d mole-')

1d.e

a Braun et al., 1962. Assumed to be the same as for the acetoxy radical (cf. Pryor et al., 1972).

DeTar, 1967. Calculated on the basis of a reasonable A-factor of 101 3 . 5 . Cook and Depatie, 1959; cf. Eberson, 1963. = (Griller and Roberts, 1971). g Solly and Benson, 1969.

(17) and (31), as exemplified in eqn (65) by butadiene and Kolbe-generated ethoxycarbonyl radicals. It is difficult to imagine Et00C-COO-

-e-

-c02

EtO-60

CH2=CH-CH=CH2 b

70% yield, based on RCOO-

EtOOCCHzCH=CHCH2COOEt + (EtOOCCH2CH=CHCH2)2

+ branched isomers

any other mechanism for the formation of the products obtained in this process than by attack of the ethoxycarbonyl radical on butadiene. Thus, a radical with a rate contant for decarboxylation of 105-106 s-l survives intact to a very high extent; certainly the benzoyloxy radical should have ample chance to react before decarboxylation. A second conclusion is that a radical with a rate constant for decarboxylation somewhere in the region 106-4 x l o 8 s-' must constitute a borderline case. Acetoxylation (and probably also propionoxylation, no. 25; cf. Table 9) is definitely a direct process, and the products obtained with substrates that are too difficult to acetoxylate by a direct mechanism are invariably derived from the corresponding alkyl radicals [eqns (17) and (31)]. Yet, considering these closely spaced limits, one cannot completely rule out the possibility that anodic acetoxylation via acetoxy radicals can contribute to some extent, thus partly reviving the indirect mecha-

60

L. EBERSON AND K. NYBERG

nism originally proposed by the discoverers of the reaction (Linstead et al., 1952). In the case of alkoyloxylation, inference from the homogeneous solution chemistry of alkoyloxy radicals is of little help. Although such reactions are abundant (Kochi et al., 1973; Rawlinson and Sosnovsky, 1972, 1973; Sheldon and Kochi, 1973; Williams, 1960), there seems to be n o established case of a reaction that is mediated by alkoyloxy radical attack on the substrate. The anodic oxidation of aromatic hydrocarbons (Nyberg, 19 70a; Ross et al., 1967, 1970), propylene (Formaro et al., 1973), and N,N-dialkylamides and -sulphonamides (Ross et al., 1966a, 1972) using nitrates in glacial acetic acid as the SSE (no. 23 in Table 8) has received a great deal of attention, partly because of the change in product distribution as compared to that obtained using the commonly employed electrolyte in acetic acid, sodium acetate, and partly because of the higher current efficiency sometimes observed in nitrate electrolytes. The change in product distribution is illustrated in eqn (66) for mesitylene (Nyberg, 1970a), a typical borderline case OAc

Product distribution/ supporting electrolyte NaOAc

Trace

Me4NN03

18

NaOAc/Me4NNO3,10:1

7

82 4 18

18

-

50

28

56

19

[Ar = 3,5-Dimethylphenyl]

in view of the very similar oxidation potentials of mesitylene and nitrate ion. The most pronounced effect of nitrate ion is to direct acetoxylation almost entirely to side-chain (a-)substitution(the small percentage of nuclear acetate found in the nitrate run may be due to the lack of buffer capacity in this type of electrolyte, making possible a slight “alkalinization” from the cathode process; see p. 34). This a-directing effect is not unique for nitrate ion but is shared by inert ions of the type TsO-, ClO,, BF, (Eberson, 1967; Magnusson et al., 1971; Ross et al., 1967, 1970). Thus only electrolytes containing acetate ion sustain nuclear aromatic acetoxy-

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

61

lation. Another interesting feature of the data given above is the small amount of nitrate ion, only 10% of the acetate ion concentration, that suffices to change the ratio of nuclear to a-acetoxylation by a factor af about 20.' Similar results from product studies were observed with toluene as the substrate (Ross et al., 1970). Voltammetric curves from substrates with relatively high oxidation potentials were in all cases markedly displaced towards higher potentials upon addition of the toluene, a phenomenon ascribed to blocking of anode sites by considerably adsorption of toluene. Hexamethylbenzene, with E lower than the discharge potential of nitrate ion, gives a mixture of the a-nitrate and acetate upon oxidation at 1.1 V versus sce, demonstrating at least in this case that these can be products of a direct oxidation mechanism. In all cases small amounts of bibenzyl derivatives were formed, which of course necessitates the intervention of benzyl-but not nitrate-radicals at some stage of the mechanism. An investigation of the anodic oxidation of mesitylene in nitrateion based electrolytes but with aprotic solvents revealed little more to illuminate the mechanistic picture (Nyberg, 1971d). Again, a very pronounced shift of the voltammetric curve was observed upon addition of the substrate when platinum was the anode material, whereas on graphite a small shift toward less positive potentials was noted. Product distributions are shown below eqn (67). The forma-

Product distribution/ solvent (anode)

CH,CN(Pt)

8 5 79

CH,CN(C)

1

CH3N02 (Pt) CH3NO2 (c)

39

56 1

92 20

99 [Ar = 3,5-Dimethylphenyl]

In one experiment with only 1% nitrate relative to acetate, the factor was about 10; since this experiment had to be run at very low conversion, the accuracy of this number is fairly poor.

62

L. EBERSON AND K. NYBERG

tion of the nitro-compound probably occurs via an anodically generated nitrating agent (see p. 36) and is not relevant to the mechanism at hand. However, the predominance of 2,4,6-trimethylphenol from the -runs at the graphite anode is of significance, since this compound is probably formed by hydrolysis of the nuclear nitrate. Aryl nitrates seem to be very unstable compounds, none having been described in the literature as far as we have been able to ascertain. An attempt to prepare phenyl nitrate from silver nitrate and phenyl chloroformate resulted in the formation in high yield of o-nitrophenol, which was proposed to arise by rearrangement of phenyl nitrate (Chaney and Wolfrom, 196 1). Significantly enough, anodic oxidation of t-butylbenzene at platinum in nitromethane/ tetrabutylammonium nitrate gave at least three isomeric t-butyl-onitrophenols among the products. Propylene, a substrate with E , l 2 for oxidation at considerably higher potential than nitrate ion, gave the products indicated below eqn (68) on anodic oxidation in acetic acid containing a perchlorate or nitrate salt (Formaro et al., 1973). Both nitrates were postulated as originating from nitrate radical attack upon either an allylic hydrogen or the terminal carbon of the double bond.

-

CH3COOH, Pt CHjCHzCH2

C H ~ = C H C H ~ O A+CCH2=CHCH20N02 + CH3CH2CH2ON02 (68)

supporting electrolyte

current yield

LiC104

19

-

-

LiN03

-

17

2

Summarizing, it is difficult at present to rule out either of the mechanisms shown in eqns (69) and (70) for reactions taking place in the presence of nitrate ion. NO;

- -e-

NO3'

attack on RCH3

HOAc or

-e-

HNO3 + RCH2'

+

RCHf

RCH2OAc

+

A

(69)

RCHzONOz

-e-

RCHJ

RCH3*+

base attack

RCHZ'

-e-

R = vinyl, aryl

RCHf

HOAc or

_T

RCH~OAC

+ RCH2ONO2

(70)

TABLE 10 Isomer Distributions for Cyanations Initiated via Cation Radicals and Cyano Radicals' Anodic cyanationb i n CH30H/NaCN a t P t Compound

44

Anisole Chlorobenzene Biphenyl Toluene Ethylbenzene Isoprop ylbenzene t-Butylbenzene Naphthalene Nitrobenzene

53 50 24 40 4 41 90

m(0) 0.1 0-5 0-4 8 9 13 10

Photolysis of CN - /M eOHC

P

4.1

40)

47 50 76 52

53

0.2

30

3

Photolysis of

I-CN~

P

401)

47

58 27 44 48

67

Diazotization of CN-NH2

4-4P

4a)

40) P

c4

14

44 41 54 50 14

15 9 29

2

27 28 31

28 46 28 21

27 10

43 50 17 23 76

87 46 90

10

61 22

39 63

60

15

40

m r m n c3

a

Nilsson, 1973. Known to be a direct process by electrochemical evidence. Probably initiated by photoionization to give a radical cation (Letsinger and McCain, 1966). Homoiytic via CN'. Presumably homolytic via CN' (Eberson e t al., 1972).

64

L. EBERSON AND K. NYBERG

It is more doubtful whether a nitrate radical would attack at a ring hydrogen of an aromatic ring and thus start a sequence leading t o the (admittedly not positively identified) aryl nitrate, since aryl radicals would be very difficult t o oxidize t o aryl cations. On the other hand, it is not possible t o formulate a reasonable direct mechanism for the formation of propyl nitrate from propene. Thus, a blend of mechanisms may operate, as indeed was concluded earlier (Nyberg, 1970; Ross et al., 1972). Anodic cyanation has been shown t o be a direct process by electrochemical methods in conjunction with the analysis of products from cpe experiments (nos. 11 and 34, Table 8). In addition, cyano radicals can be generated in homogeneous solution, and a comparison of processes initiated by radical cation and cyano radical initiated processes reveals the indiscriminate nature of the latter towards aromatic substrates (cf. also Williams, 1960). This is in contrast t o the electrophilic nature of the radical cation process. Even perchlorate ion, with its very high anodic limit (Table 8), can be a source of ambiguity in mechanistic studies. In connection with studies on the anodic oxidation of aliphatic sulphides in acetonitrile/ NaC104 under extremely dry conditions it was found (Cottrell and Mann, 1969) that perchlorate ion must be a source of oxygen in the oxygen-containing products, in spite of the fact that cpe was performed far below the anodic limit of the SSE. A closer examination by ultraviolet spectroscopy indicated that both C1, O7 and C10, were formed during electrolysis, and this was suggested t o occur via anodically generated protons (p. 34) in the anolyte of the divided cell [eqn (71)]. This proposal was later substantiated by e.s.r. 2HC104

+ HzO + C1207

(71)

evidence for the C102 formed (Cauquis and Serve, 1970a; Glass and West, 1972). Moreover, Cauquis and Serve were able to show that C10, was formed both at and substantially below the anodic limit, presumably by one or both of reactions (72) and (73). The HC104

-

+ C1207 -+ CIOz

-e-

ClO,

C104'

+ ClOZ

+ o*

(72) (73)

formation of heterocyclic compounds in the anodic oxidation of CH3CN/perchlorate salt might also be due t o the intervention of these species (Eberson and Olofsson, 1969; Fleischmann and Pletcher, 197 3 ) .

65

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

Another supposedly very inert anion, BF,, has been shown to cause complications by acting as a fluoride ion donor (Koch et al., 1973). We can confirm that this sometimes occurs in anodic coupling reactions too (Nyberg, 1971f), albeit only as a minor side-reaction, and suspect that partly hydrolysed tetrafluoroborate ion might be responsible for this phenomenon.

Cathodic reactions

As pointed out above, the cathodic limit of an SSE can often be extended toward such negative potentials that the problem of identifying the electroactive species never becomes acute. However, three types of possible complication deserve mentioning, one mainly due to the nature of the electrode material and the others to the nature of the SSE. The first type is connected with cathodic reduction in a protic medium, in which we can distinguish between the usual direct mechanism (73) in which the substrate molecule is the electroactive species and mechanism (74) involving formation of adsorbed hydrogen atoms. The latter reaction is obviously very similar to ordinary +e-

R-H --+

-

Proton donor

R-H'-

E step(s)

Product(s)

(74)

Favoured at cathode materials of high overpotential (Table 6 : Hg, Sn, Zn) +e-

H30+ --+

Metal

* . *

H

R-H

Product(s)

(75)

Favoured at cathode materials of low overpotential (Table 6: Pt, Pd, Ni) and with substitutes of low electron affinity

catalytic hydrogenation, in which surface adsorbed hydrogen atoms are formed by dissociative adsorption of hydrogen; at the cathode adsorbed hydrogen atoms are formed by discharge of hydroxonium ion. Hence we are actually dealing with electrocatalytic hydrogenation, no different in principle from ordinary catalytic hydrogenation, and we shall therefore not go into detail with this process here. The reader is referred to reviews for further information (Dietz and Lund, 1973; Sokol'skii, 1971). As we proceed through the region of cathode potentials, two processes are possible at the negative extreme of a highly inert SSE, either injection of solvated electrons into the solution or reduction of the cation to give the metal (or, in case of the commonly used

66

L. EBERSON AND K. NYBERG

mercury electrode, an amalgam). These possibilities create mechanistic problems when the substrate is reduced near or at the cathodic limit of the SSE. To begin with the solvated electron, we think that the very concept of esOIv,emanating from the fascinating studies on the hydrated electron, eaq (Hart and Anbar, 1970), has in some ways created a semantic problem at least for practitioners of organic electrochemistry. Almost any neutral compound that accepts an electron from an electron donor, be it an electrode surface, a metal in a low valence state, or even from esolv itself, is denoted a radical anion, as is illustrated for some representative compounds below in the middle column. Although at times used as part of the electrochemical jargon, nobody uses the nomenclature in the upper part of the right-hand column and yet this is exactly what is implied in a consistent use of the solvated electron terminology: Electron acceptor

Radical anion nomenclature

Solvated electron nomenclature

Benzene

Benzene radical anion

Benzenated electron

Naphthalene

Naphthalene radical anion

Naphthalenated electron

Anthracene

Anthracene radical anion

Anthracenated electron

Hexamethylphosphortriamide (HMPA)

HMPA radical anion

Solvated electron

Ammonia

Ammonia radical anion

Solvated electron

Water

Water radical anion

Hydrated electron

We cannot see any significant conceptual difference between the hydrated and, e.g. naphthalenated electron, and we would therefore urge the reader to think of solvated electrons as radical anions of solvent molecules; in some environments the solvent radical anions are stable (ammonia, amines, HMPA), in some not (protic solvents), and that is exactly the same kind of behaviour exhibited by more conventional radical anions (Szwarc, 1968, 1969). Thus, in lithium chloride/HMPA the process responsible for the cathodic limit of the SSE is reduction of the HMPA molecule to give esolv in the normally used terminology. the HMPA radical anion We then encounter exactly the same mechanistic problem with respect to the electroactive species as earlier for anodic processes; is the process a direct reduction of the substrate (76) or an indirect one (77) mediated by the solvent (denoted S) radical anion? Dealing with the semantic problem does not change the nature of the problem,

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

R-H

+e-

---+

R-H'-

Proton donor

E step(s)

Product(s)

67 (76)

though. As we have described the EI (Fig. 3, p. 22), the inner Helmholtz layer consists of solvent molecules and adsorbed ions (large cations and even anions). Thus, conditions are present for at least a consideration of the possibility that the electron is first transferred t o an adsorbed species which then transfers it to the substrate. In this sense, many reactions might be indirect ones and it would be difficult t o find out whether they are or not. Obviously, this is also partly a question of semantics, and we shall define this particular situation as a direct mechanism, unless experiments tell a different story. Experiments designed to elucidate the role of S ' - in cathodic reduction tend t o be just as ambiguous as their anodic counterparts, unless certain precautions are taken. The possible intervention of S * in the reduction of aromatic hydrocarbons (Asahara et al., 1968; Benkeser and Kaiser, 1963; Benkeser et al., 1964; Sternberg et al., 1963, 1966, 1967, 1969) in SSEs made up of amines or HMPA (to which up to 65% ethanol can be added without impairing the stability of HMPA'- too much) as compared to the possible direct processes taking part in protic solvents illustrates the problem. Table 11 shows some representative results from the cathodic reduction of some aromatic hydrocarbons. These include cases with near the cathodic limit or in the discharge region of the SSE (benzene, toluene) and cases with E l / 2 at considerably more positive potential (naphthalene, anthracene; again we must anticipate the discussion of reactivity and refer t o Table 21). Reactions nos. 1, 2, 6, and 7 immediately demonstrate one difficulty with such studies in that the catholyte of a divided cell becomes strongly basic as electrolysis progresses. In sufficiently basic medium, the initial product, a 1,4-dihydro derivative (cf. the Birch reduction: Birch and Subba Rao, 1972), will rearrange t o a conjugated system which, in contrast t o the 1,4-dihydro derivative, is further reducible to the tetrahydro product (nos. 1 and 6). In a non-divided cell the acid production at the anode balances the base production and thus only a little rearrangement occurs. It is therefore not a trivial problem to find out if the tetrahydro product is formed from the conjugated dihydro product, formed directly or by rearrangement [eqn (78)].

cn

TABLE 11

00

Results from the Cathodic Reduction of Aromic Hydrocarbons under Different Conditions

No. Compound

1 Benzeneb 2 Benzeneb

Cell type

SSE (cathode)

Reference electrode

Cathode Current Current Product distribution, % potential, density, yield, - _ _ _ _ _ - - - ~ Vversus ref. mAcm-2 % DHa THa HHa OHa Ra 0

100

95

5

64

17

70

71

93

7

95

23

10

Divided

CH3NH2/ L E I Pt)" Non-divided CH3NH2 /LiC1

r

(Wd

n

3 Benzenee

Divided

H2NCH2CH2NH2/ LiCl(Pt)f

4 Benzend

Divided

Diglyme-H2O Sce (9%)/Bu4NBr(Hg)

5 Benzeneh

Divided

6 Tolueneb

Divided

HMPA-EtOH (67 mol %) LiCl (Al)d CH3NH2/LiCl(Pt)"

7 Tolueneb

Non-divided CH3NH2/LiC1(Pt)d

8 Tolueng

Divided

Diglyme-H2O Sce (9%)/Bu4NBr(Hg)

-3.3

9 Naphthalenei

Divided

Diglyme-H2O Sce (12*5%)/Bu4NBr (Hg)' HMPA-MeOH Ag wire (50% v/v)/LiCl (Pt)d,

-2.4

10 Naphthalenek Divided

3'

Zn(Hg)/ZnC12/ SSE

Ag wire in SSE

--

1.1

0.49

--3.3 --2.4

-1.7

91

2-5

10

0

100

94

6

61

90

10

79

96

4

95

88

6

13

m

M

E0 2

67

6

T A B L E 11-continued

Cathode No. Compound

Cell t y p e

11 Naphthaleneh Divided

SSE (cathode) HMPA-EtOH

Reference electrode

potential, density, yield, V v e r s u s ref. m A c m - 2 %

Ag wire in SSE

-

P r o d u c t distribution, %

Current C u r r e n t

2.4

91

5

96

OH^

~a

s

22

11

40

E

-

7

20

(Wd Divided

HMPA-HOAc ( 1 M ) / Li/LiCI (sat. in LiCl (vitreous C)d HMPA)

+0.9

13 Anthracene"

Divided

HMPA-HOAc ( l M ) / Li/LiCl (sat.) LiCl (vitreous ~ ) d ,

-1.0

DH = dihydro, TH = tetrahydro, HH = hexahydro, OH = octahydro, R Benkeser and Kaiser, 1963; Benkrser e t al., 1964.

= higher

>125P

2

ci C

2!U

Ei

100 57

ci

DHa THa HHa

(67 mol %)/LiCl

12 Anthracene"

(II

i2 3

88

7

hydrogenated product.

Blue colour pervaded the entire electrolyte during electrolysis with substrate prrsent. Blue colour visible near catholyte during electrolysis with substrate present. Sternberg et al., 1966. f In blank runs at about -1.1 V (3.5-7 mA/cm2) dark blue lithium globules were visible in the layer near the cathode surface. g Osa, 1968. Sternberg et al., 1969. Misono et al., 1968. I With THF as the solvent similar results were obtained. Acidic proton donors, such as phenol, did not sustain reduction of naphthalrnr. Asahara et al., 1968. In a blank run dark blue lithium globules were observed at the cathode surface. Current yields with 0, 10, 20, 30, and 40%(v/v) of HMPA gave current yields of reduction products of 0, 20, 27, 41, and 39%,respectively. Avaca and Bewick, 1972. Blue colour pervaded the entire catholyte at very high current density (1000 mA cm-2). p Estimated by us.

z

E 2 0

E>

E

m r m n ci 0

i EE H

z

CD

70

L. EBERSON AND K. NYBERG

I

Base

1

+2e-

+2H'

This is true even in cases where proton donors have been deliberately added (nos. 1 2 and 13) because of a possible base gradient near the cathode (the blue colour of this zone indicates at least that we have a gradient of mIPA.-). However, it is still obvious that the product distributions from reactions nos. 3, 5 , 11, and 1 3 represent different reaction paths from the remaining ones. The reaction proceeds beyond the tetrahydro stage and shows all indications of being conducted under more forcibly reducing conditions. Cathode potentials are well in the region where discharge of the SSE takes place, and the blue colour of S' - is clearly visible. Thus these cases certainly represent indirect mechanisms proceeding via S ' -, as demonstrated especially well by reactions 1 2 and 13. Here one can note that a t the more anodic potential (no. 12), the direct reduction gives only the dihydro derivative of anthracene (in fact only the 9,lO-isomer is formed). The current density, 5 mA c m - 2 , represents ilim for the particular concentration of anthracene used. Pushing the cathode potential into the SSE discharge region t o a value at which the current density exceeds 125 mA c m - 2 , one can be certain that the anthracene molecules diffusing toward the cathode surface never come so close as t o undergo a direct reduction since they have been reduced by S ' long before. Moreover, a good proton donor is present so that the dihydro product is formed rapidly and reduced further by S ' - etc. Yet we can note that there is still another mechanistic possibility in at least reactions nos. 3 and 10. In both cases blank runs revealed that lithium metal precipitated on the cathode at very negative potentials, and in one instance (Sternberg et al., 1966) it was suggested that the reaction takes place via solvated lithium metal, i.e. an indirect reaction analogous t o the Birch reduction. One could even say that the reaction takes place at a lithium cathode. This example may suffice to illustrate the complications due to participation of the solvent in cathodic reductions. It may also

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

71

convince the reader that solvated electrons are nothing but solvent anion radicals. The last problem, to elucidate the role of amalgams formed by discharge of the SSE cation at mercury cathodes, is actually a variant of the problem with the lithium precipitation mentioned above. Since in such a case we are running the electrolysis a t an amalgam electrode, even though the amalgam structure may only be a few atomic layers thick, the problem is one of electrode material and its role for the electrochemical reaction (for a review of amalgam reductions, see Lund, 1973a).

8. REACTION SEQUENCE Once the electroactive species has been identified, a far more exacting task is to uncover the sequence of elementary steps leading t o stable products. In discussing and exemplifying this problem, we shall deal only with direct processes since the indirect ones in principle have reasonably well understood counterparts in homogeneous systems. It is an important postulate of electrochemistry that electrons are always transferred one by one (Semenov, 1958), and hence the first discrete intermediate from a direct electron transfer (79) will be a radical cation or anion. R-H-

Cathodic reduction +e-

R-H

Anodic oxidation -e-

R-H'+

(79)

It has already been pointed out (Section l ) that the chemistry of radical ions, having hardly penetrated t o the textbook level in organic chemistry, is little known, and we shall therefore briefly summarize their reactions here (for more detailed treatments, see Bard et al., 1976; Dorfman, 1970; Szwarc, 1968, 1969). Shown below are the elementary acts (denoted C or E; cf. p. 25) in which radical ions have been shown to participate: It should be noted that radical cations have a dual reactivity toward a nucleophile, dependmg on the properties of the latter as a base. It is possible t o distinguish between these alternatives by cyclic voltammetry in cases of relatively stable radical cations (Parker and Eberson, 1969a) and by their reactivity toward pyridine nucleophiles of varying steric demands. The principle is illustrated in Fig. 6, which

72

L. EBERSON AND K. NYBERG

A . Cation radicals

-

Disproportionation (C): 2R-H'+ =$ R-H2++ R-H

(80)

-e-

Oxidation (E or C): R-H'+

(81)

R-H2+

+e-

+ R-H

Reduction (E or C): R-H"

(82)

Coupling (C): 2R-H'+ -+ H-k-k-H Reaction with Nu- (C): R-H" (base action)

+ Nu-

(83)

+R *

Reaction with Nu- (C): R-H" + Nu(nucleophile action)

+ NuH

(84)

HkNu

(85)

=+R-HZ- + R-H

(86)

--f

B. Anion radicals Disproportionation (C): 2R-H '+ Oxidation (E or C): R-H ' -

-e-

+ R-H

(87)

i--

Reduction (E or C): R-H'- + R2Coupling (C) : 2 R-H ' -

-+

Reaction with E+: R-H * -

(88)

H-R-R-H

(89)

+ E+

(90)

-+

HKE

shows a cyclic voltammogram of 9,lO-dimethylanthracene (DMA) without any nucleophile added (curve A) and with increasing added concentrations of 2,6-lutidine, a hindered nucleophile (curves B- F). Curve A (note that the anodic scan is from right to left and the

U

E

F

Figure 6. Oxidation of 9,lO-dimethylanthracene (1.0 mM) in the presencc of 2,6-lutidine (A, 0 mM; B , 0-25 mM; C, 0.50 mM; D 0.75 mM; E, 1.0 mM; F, 1.25 mM). Reprinted by courtesy of Pergamon Press Ltd. (Parker and Eberson, 1969a).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

73

anodic current is on the negative y-axis) shows that DMA'+ is a relatively stable species on the timescale involved, since almost the same current due t o reduction of DMA" is observed on scanning in the cathodic direction (from left t o right, cathodic current on the positive y-axis). Adding increasing concentrations o f 2,6-lutidine decreases the cathodic current due to DMA'+ until it has almost disappeared at a ratio of [DMA]/[2,6-Lu] of 1:l-25. Obviously DMA" is consumed in a reaction with the nucleophile. A series o f lutidines, allowed to react in the same way with radical cations of different structures, established the reactivity orders shown in Table 1 2 (given as the ratio of [Lu] /[substrate] necessary to reduce the cathodic current t o Lero). This set of data shows that the two radical cations which cannot engage in a base reaction (from 9,lO-diphenylanthracene and 1,4-dimethoxybenzene) show considerably decreased reactivity toward the more hindered lutidines, whereas DMA '+,which is capable of undergoing proton abstraction at the unhindered hydrogens of the methyl groups, exhibits the same reactivity toward all three bases.

TABLE 1 2 Relative Reactivitiesa of Radical Cations Toward Different Lutidinesb Radical cation from

2,6-Lutidine

9,lO-Diphenylanthracene 1,4-Dirnethoxy benzene 9,lO-Dimethylanthracene

1.0 2.1

0.75

2,5-IAutidine

3,j-Lutidine

6.5 35 1.0

37 64 1.3

a Expressed as the [Lu] /[substrate] ratio necessary to reduce the cathodic current to zero at the same initial [substrate]. Parker and Eberson, 1969a.

Reactions of Radical Ion versus Doubly Charged Ion Returning to the problem of establishing the reaction sequence, it was already mentioned in Section 3 that the radical cation (anion) formed in the first one-electron step can undergo a second step with formation of a dication (dianion). In a medium of sufficiently low nucleophilicity (electrophilicity) this is clearly indicated by the voltammetric curve exhibiting two successive waves, each corre-

74

L. EBERSON AND K. NYBERG

sponding t o the transfer of one electron' (see Table 1 3 ) . Thus we have the possibility that the product distribution of an electrochemical process is potential-dependent due to differing reaction modes of the radical ion and the doubly charged ion (see below). Just t o illustrate the rather complex situation involved, a reaction scheme for the anion radical/dianion reactions that are possible with an added electrophile is presented in reaction sequence ( 9 1 ) in which M denotes the substrate. Note that M is also assumed t o have electrophilic properties and thus can engage in dimerization and further polymerization processes, ultimately giving living polymers unless the growing chain is terminated (Szwarc, 1968). M

To orient the reader in this maze of reactions, all electron transfer steps have been arranged vertically and C steps horizontally. Furthermore we have not indicated whether the electron transfer steps are of the E or C type, another factor to take into account. The somewhat complex appearance of the scheme should not, however, obscure the fact that suitable manipulation of the experimental variables can give either of the two products, E-M-E or E-M-M-E in nearly quantitative yield, as for example in the cathodic hydrodimerization (34) of acrylonitrile. A dependence of the product distribution on potential is observed in cases where the radical ion and doubly charged ion follow differing reaction paths. Thus, 4,4'-dimethoxystilbene shows two anodic waves in acetonitrile/liC104 at 0 - 9 0 and 1-15 V versus The number of electrons corresponding to a particular wave is easily determined by, e.g. comparison of an unknown ilim with that of a known one-electron process.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

75

sce; cpe in CH,CN-HOAc/NaOAc at 0.90 V gave exclusively a one-electron oxidation product from dimerization of two radical cations [eqn (92)], whereas at 1.35 V only the two-electron oxidation product from solvolysis of the dication [eqn (93)] was

-2e-

ZRCH=CHR

RCH=CHR

+

.

2RCH-CHR

-2e'

+

--+

+

RCH-CHR

AcO-

-+

RCH(OAC)CH(OR')R

H2O

(93)

(R = p-methoxyphenyl; R ' = acetyl or hydrogen)

obtained (Eberson and Parker, 1970; Parker and Eberson, 1969b; cf. also O'Connor and Pearl, 1964; Sainsbury, 1971). A similar potential-controlled radical cation versus dication mechanism was observed in the oxidation of 9,lO-dimethylanthracene (Parker, 1969b). On the cathodic side numerous potential-dependent reactions have been described (Fry, 1972; Baizer et al., 1973), but the dependence on potential is mostly due to the possiblity of further reduction of an initially formed product and not t o the radical ion/doubly charged ion dichotomy. One difficulty might be that the potential region in which dianion formation occurs is negative enough for SSE interference. In fact, the whole body of electrochemical data concerning the reduction of radical anions of aromatic hydrocarbons to dianions was called into question due to the fact that cyclo-octatetraene (COT) and two of its benzo-fused derivatives were shown not to undergo any second one-electron transfer in rigorously dried media. It was then proposed that the second wave observed in SSEs containing a proton donor could be due to the reduction of the protonation product of COT [eqn (94)] (Thielen and Anderson, 1972; Anderson and Paquette, 1972). However, even if this interpre-

l e wave

wave

76

L. EBERSON AND K. NYBERG

tation is correct,' COT and HCOT. do not represent a case analogous t o the corresponding species in the reduction of alternant aromatic hydrocarbons. For these, simple HMO theory places the reduction potential of ArH, a t a more positive value than that of ArH (Aten and Hoijtink, 1959; Hoijtink, 1954, 1957; Hoijtink e t al., 1954), while this may not necessarily be true for COT and HCOT. which are both nonplanar. Moreover, it was recently shown that the technique of adding neutral alumina t o the electrolyte solution to keep trace impurities of nucleophiles under control could be extended t o cathode processes too (Svensmark Jensen and Parker, 1974, 1974a). Thus, cyclic voltammetry of anthracene at -30" in DMF/Me4 NBr showed reversible behaviour (see Section 10) both for radical anion and dianion formation with the E,lz values (values within parentheses taken from earlier work in 96% dioxan-H20 and DMF; Hoijtink et al., 1954; Pointeau, 1962) at 1.92 (1-98; 1-94) and 2.66 (2.44; 2.52) V versus SCE, respectively. The mechanism of the reduction of aromatic hydrocarbons was actually established early by Hoijtink and his co-workers (Hoijtink, 1970) as an ECE mechanism (p. 25) (see also Given and Peover, 1960; Santhanam and Bard, 1966). The two one-electron waves due t o formation of anion radical and dianion in an aprotic medium change in a characteristic way upon addition of incremental amounts of a proton donor; the height of the first wave increases at the expense of the second one until at sufficiently high concentration only a single two-electron wave is obtained. This behaviour in combination with the HMO calculations referred to above clearly show that the radical anion is protonated to a neutral radical which is reducible at a less negative potential than the substrate [reaction Ar-H

+e-

--+

Ar-H.

-

H+

--+ ArH.

+e-

H' ArH- --+ ArHz (95)

(98)] . Many other cathodic reductions show analogous behaviour.

Disproportionation versus ECE Mechanisms

The possibility of the heterogeneous electrochemical transformation of a radical ion into a doubly charged ion [eqns (81) and (88)l In all probability it is not correct, since an earlier investigation has shown that the first COT wave can be resolved into two closely spaced one-electron transfers (Huebert and Smith, 1971; cf. also Section 9 and Table 20).

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

77

has not been seriously disputed. However, the possible existence of a homogeneous disproportionation equilibrium [eqns (80) and (86)J as the source of a kinetically active doubly charged ion in the reaction with the nucleophile was raised for radical cations by kinetic results obtained with thianthrene (Th) radical cation in homogeneous solution (Shine and Murata, 1969; Murata and Shine, 1969). These showed the solvolysis of Th" to be second order in [Th"] and inverse first order in [Th] and led t o the suggestion that the disproportionation equilibrium of eqn (80) preceded the ratedetermining step, with Th2+as the kinetically active species. On the electrochemical front, a study (Manning et a / . , 1969) of the pyridination of 9,10-diphenylanthracene (DPA; reaction no. 8, Table 8) using RDE voltammetry (p. 8) unequivocally established an ECE mechanism for this reaction. In this technique, voltammetry is performed at a rotating disc electrode, the rotation rate ( a )of which can be changed. If the radical ion formed in the initial step reacts infinitely fast with a nucleophile present, i l i m will be independent of w and have a value corresponding to a two-electron process (since RHNu will be oxidized at a lower potential than RH, analogously t o the cathodic case just discussed). On the other hand, if no nucleophiles are present, one-electron oxidation is observed, again with n o change with a.At intermediate values for the rate constant of R-H" with the nucleophile, ilimwill be dependent on w since R-H'+ can either be "spun off" from the electrode (oneelectron) or react with the nucleophile and undergo the two-electron reaction. Hence the number of electrons transferred ( a o bs ) varies with w between 1 and 2. Dependent on the reactivity of the nucleophile a set of different curves of nabs versus w1/ 2 was obtained for DPA. Rate constants for the RH'+/Nu reaction could be obtained by a digital simulation technique (Feldberg, 1969) and were found to lie in the region between 3 and 1000 s-l for a range of nucleophiles chosen between 4-cyano- and 4-methylpyridine in basicity. As can be seen from Fig. 7, the order of reactivities is that predicted from the basicities of the amines. Since this paper ended with a claim of rather universal validity of the ECE mechanism in anodic addition and substitution reactions, cf. the mechanism proposed for the solvolysis of Th" in homogeneous medium, namely, disproportionation, it started a controversy. We shall not go into detail with this story (Hammerich and Parker, 1972; Marcoux, 1971, 1972; Parker, 1972; Parker and Eberson, 1970b), since it encompasses a lot of rather involved argumentation which

78

L. EBERSON AND K. NYBERG

Figure 7. Rate of DPA+ interactions with pyridine nucleophiles: A, 0.50 mM, 9,lO-DPA

+ 25 mM 4-methylpyridine; 0, 0.50 mM 9,lO-DPA + 25 mM pyridine; 0, 0.50 mM 9,lO-DPA + 25 mM 4-acetoxypyridine; 0,0.50 mM 9,lO-DPA + 25 mM 4cyanopyridine. All solutions were 0.20 M tetraethylammonium perchlorate in acetonitrile plus the constituents given above. Reprinted by courtesy of the American Chemical Society (Manning et al., 1969).

does not easily lend itself t o a discussion of pros and cons in a limited amount of space (cf. Bard et ul., 1976). However, we can look at some recent experimental results on the position of the equilibrium of the disproportionation process, since such data are of prime importance for discrimination between the two mechanisms. Table 1 3 gives log K d values, calculated from U ,the difference in E l l 2 for the first and second wave, for a number of model compounds. The high degree of reliability of this method was established independently (Svanholm and Parker, 197 3a). Two trends are noticeable, namely that ( u ) U-values increase as the SSE nucleophilicity decreases, and ( 6 ) U-values decrease as the structure of the substrate is modified by substituents that stabilize positive charge. At one extreme, we actually arrive at systems where the second electron transfer can be estimated to occur at lower potentials than the first one (nos. 10, 11; no. 8 is a borderline case). The first trend is the expected one, since the change from less to inore ideal SSEs means that chemical reactions of the dication especially become more and more suppressed. The effect of structural variation is entirely consistent with the known ability of oxygen and nitrogen functions t o accommodate positive charge. Thus, to go back to where the controversy arose, thianthrene has Kd = 2.5 x a number that together with a value of k 2 k d of

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

79

about 0.2 M - ' s-' gives k , = 10' M - ~s - l ( k , is the second order rate constant for the reaction between dication and water; Murata and Shine, 1969). This rate constant is one order of magnitude lower s - l ) and than that of diffusion controlled reactions ( l o 9 - 10' M therefore compatible with the disproportionation mechanism. However, electrochemical studies based on pulse relaxation methods of the T h ' + - H 2 0 system in acetonitrile seem t o rule out the disproportionation mechanism (Kuwana, 1973; Broman et al., 1973). For 9,lO-diphenylanthracene (DPA), with a Kd-value of l o - * in acetonitrile, k 2 K d is not known, but it seems reasonable that k 2 for reaction with nucleophiles will be below the limit for diffusion controlled reactions, again making a disproportionation mechanism possible at least in principle. However, kinetic data obtained for the homogeneous reaction between pyridine and DPA * + in acetonitrile (Svanholm and Parker, 19 73) established a first-order dependence for both reactants, thus favouring the ECE mechanism (96). Spectroelectrochemical studies, in which the concentration of DAP * + formed by cpe of DPA was monitored by a special technique using a transparent electrode, were in agreement with these kinetics (Blount, 1973; Blount and Kuwana, 1970), but a "half-regeneration mechanism" (97) was found to fit better with the experimental data (see also Sioda, 1968). In this, the second E step of the ECE mechanism is replaced by a step in which the neutral radical is oxidized by the radical cation [cf. also eqn (46)] :

-'

ECE: DPA

- - - -e-

Half regeneration: DPA

DPA'+

PY

PY

-e-

+ DPA

DPAPY'+

'+

-e

DPAPy

D P A P Y ~ + (96)

DPA"

'+

DPAPy*+ (97)

It is a moot point whether we really should distinguish between the ECE and the half regeneration mechanism (Adams, 1969; Marcoux, 1972; Hammerich and Parker, 1974) as independent mechanisms. That disproportionation can be an alternative is shown by the case of tetraphenylethylene, the cyclization of which presumably occurs via the dication according t o eqn (80) (Svanholm et al., 1974a). This is a somewhat disconcerting conclusion, since it will now be necessary t o work through a fair number of systems in order t o find out the structural factors favouring either type of process. + Finally, it should be mentioned that the elusive species R-Nu of eqn (45) has been identified in the cyclic voltammogram of a series

co 0

TABLE 13 Disproportionation equilibrium constantsa for 2RH.+ =+R2++ R as determined from E1/2 d a t a

~Acetonitrile _ _ Nitrobenzene _ _ No. 1 2 3 4 5 6 7

Compound 4,4’-Dimethoxybiphenyl‘ Thianthrenee 9,lO-Diphenylanthracenef 9,lO-Di-p-tolylanthracenef 9,10-Dip-anisylanthracenee,f Tetraphenylethylend 2,3,7,8-Tetramethoxythianthreneh

8 Tetra-(p-anisyl)ethylenei 9 (Me2N)2C=C(NMe2)2”

AE,v

-1ogKd

0.27 0.51 0.46 0.44 0.22

4.6 8.6 7.8 7.4 3.7

0.25 0.007 0.14 0.23

4.2 0.1 2.3 3.9

10 (Me2N)2C=CHPCHC(NMe2)2 -0.08’ 11 (MezN)2C=C(Me)- C(Me)= (NMe2)zk <-0-24’ 12 Cetrathiotetracene“ 13 2,3,7,8-Tetramethoxybiphenylenen

AE,v -1ogKd

AE,v

0.34 0.51 0-51 0.50 0.26

0.37 0-50

6.3 8.5

0-52 0.30

8.8 5.1

0.09

1.5

0.49

8.3

5.8 8.6 8.6 8.5 4.4

BC

Methylene chloride Ab -1ogKd

AE,V

-1OgKd

0.41 0.62 0.60 0.56 0.32 0.36

7-0 10.4 10.5 9-5 5.4 -6

0.41 0.69

7.0 11.7

0.53

9.0

-1.4

<-4

m

Cd

A E , V -1OgKd

0.33

r

5.6

AE,

v

W

m -1OgKd

0

z z

*

0.63 0.55

10.5 9.3

U

’

z

2 E0

a Given as -log Kd; calculated from the expression 23.06 x AE = -2-3 R T log Kd, where aE is the difference in E 112 for formation of RH.+ and Nz+, respectively. Methylene chloride/trifluoroacetic anhydride/trifluoroacetic acid (45:5: 1). Trifluoroacetic acid/trifluoroacetic anhydride (9: 1). Methylene chloride/fluorosulphuric acid (100:1 ) at -34°C. F Hammerich and Parker, 1973. H f Hammerich and Parker, 1974. c Svanholm et al., 1974a. Glass et al., 1973. 9 f Bard and Phelps, 1970; Parker et al., 1969. z 1 Kuwata and Geske, 1964. U Fritsch et al., 1970. Calculated from an independently estimated value of Kd; the voltammogram shows one reversible two-electron wave.

2

co E

s-s I

5n

,

0000

;Geiger, 1973. I

s-s 8

,

Ronlin and Parker, 1974.

z

0

E

82

L. EBERSON AND K. NYBERG

of 9,lO-diphenylanthracenes. In this particular case, this species could also be generated chemically by treatment of the corresponding 9,lO-dihydroxy- or dimethoxy-derivatives of 9,lO-dihydroanthracene with trifluoroacetic acid (Hammerich and Parker, 19 74; cf. also Dietz and Larcombe, 1970).

Coupling; via Radical Ions, Neutral Radicals, or Reaction b e t w e e n Radical Io n and Substrate?

The problem outlined in the heading is in most cases an important one for coupling reactions of the types schematically depicted in eqns (29), (30), (32), and (33). We shall use as an example the electrohydrodimerization (EHD) of activated olefins (for reviews see Baizer, 1973; Baizer and Petrovich, 1970; the situation with respect to mechanism has recently been summarized: Lamy et al., 1973) illustrated in eqn (98) by acrylonitrile [cf. eqn (34)].

+2H+

NC-CH-CH2-CH2-EH-CN

-

NC(CH~)~CN

(984

\tH2-~H--CN

+eCHz=CH-CN

[CHz-eH-CN

4

JCH2=CH-CN

6H2-CH-CN]

-

H+

-

+ CH2CH2CN

(98b) ,/CH,CH2

CN

+e-

NC-EH-CH2

-CH2 -6H-CN

+2H+

NC (CH2)4CN

(98c)

To begin with, let us look at some product distributions from a number of mixed EHD reactions, i.e. ones in which two monomers, M1 and M 2 , are electrolysed together giving varying amounts of HMI MI H, HMl M2 H, and HM, M2 H, since these results triggered a debate on the respective merits of the mechanisms exemplified by eqns (98a) and (98c). Table 1 4 shows some illustrative examples of mixed EHD processes. It is evident that if two compounds with E l l 2 -values differing widely enough are coelectrolysed at a potential where only one of the compounds is reducible, the symmetrical hydrodimer of the other compound is missing (no. 2 ) but the mixed hydrodimer is formed. If the two half-wave potentials are close together, all three coupling products are formed (cf. also Petrovich e t al., 1969). The result of reaction no. 1 (Table 14) is highly significant, since it shows that at a potential where only one of the compounds

TABLE 1 4 9

Yields of Symmetrical and Mixed Coupling Products from Some Representative EHD Reactions'

z

0

Product distribution, mol % No.

Monomer MI

V)b

Monomer M2 (El,,, V)b

Cpe a t E , V b

HMlMlH

HMlM,H

HM2M2H

- 1.4

6 87

-( 1.4-1.5) -( 1.8-1.9)

39

58 13 63 37

36 0 37 24

5 i92 2

1 2 3 4

CH,=CHCOOEt (-1.8) EtOOCCH=CHCOOEt (-1.35) 4-Vinylpyridine (-1.5) CH,=C-CN (- 1.8)

I

CHpzCHCN (-1.9) CH,=CHCN (- 1.9) CH2=CHCOCII, (-1.4) CH,=CHCN (-1.9)

CH2CHzCN

' Anderson et al., 1965; Baizer, 1964a; Baizer and Anderson, 1965. b Versus sce.

Claimed to be formcd but no yield stated.

- 1.85

C

2

2 0

P

0 9

2m

r m n F

*

x

00

w

84

L. EBERSON AND K. NYBERG

undergoes electron transfer, it still is possible to isolate the mixed but not the second symmetrical hydrodimer. This immediately suggests the formation of a nucleophilic species from the most easily reducible compound 'which then attacks either of the two electrophilic monomers present in a Michael type reaction, i.e. the radical anion/substrate reaction of eqn (98c). Subsequent voltammetric studies on a number of bis-activated olefins, which correspond to more stable radical anions and thus are more amenable t o electrochemical investigation were in agreement with this coupling mechanism (Petrovich e t al., 1969a). EHD has a cyclic variety in which a suitably substituted bisactivated diolefin can cyclize t o give three- to seven-membered rings [eqn (99)]. The E , i2-values of the cyclization reaction were all at CH=CH-COOE t

-22 ( +2H+

/ (CH2)n

\

CH-CH2COOEt

CH=CHCOOE t

C G q

\

(99)

CH-CH2COOEt

Yield of cyclized product, %:

1.61(1), 1-69(2),1*65(3),1*68(4),1.79(5) 98 41 100 81 10

- E l p for linear reaction (n):

1.93(1), 1*86(2),1*97(3),1*87(4)

-El/*

for cyclization (n):

less negative potentials than those of the linear dimerization process, the fractional height of the first wave being directly proportional to the macroelectrolysis yields of cyclization product. A comparison with E l l 2 , -1.82 V, of a model simple activated olefin, ethyl crotonate, shows that those diolefins which give good yields of cyclization product, have E l / 2 -values slightly displaced toward less negative potentials, A E being 0-15-0.20 V. Since these data indicate that both double bonds might be involved in the potential-determining step, a concerted electron transfer step involving transfer of the electron simultaneously with the formation of the intramolecular C-C bond was proposed for cyclic EHD (Petrovich et al., 1966, 1966a); by analogy, the proposal was extended to the linear process too. We shall return to the problem of concerted mechanisms for electron transfer later, since it requires special attention, and continue with other, mostly electrochemical evidence bearing on the EHD mechanism. A chronoamperometric technique employing a double potential step was applied to the EHD reaction in an attempt to distinguish between five possible mechanisms. These included mechanisms corre-

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

85

sponding to eqns (98a) and (98c), the two most likely alternatives under the conditions employed (Childs et al., 1971 ; Puglisi and Bard, 1972). In brief, the potential of the cathode is first stepped to E l where R + e- +. R'- occurs at a diffusion controlled rate. At a time t F , the potential is stepped to E , where only the process R'- + R + e- occurs. The current is measured just before t F ( I F ) and just before 2tF(Is). The ratio I B / I F has a value of 0.293 in the absence of any kinetic perturbations, independently of the length of t F . If, however, R ' - reacts chemically to form a non-oxidizable intermediate, I B/ I F will be a function of t F . By suitable choices of t F and concentrations, the variations in I B / I F can be used to calculate the order and rate of the perturbing chemical reaction. This is done by digital simulation of the different mechanistic cases and comparison with appropriately normalized functions of the experimental variables. In the case of the anion radical coupling and anion radical/substrate coupling mechanisms, it is a fortunate fact that the simulated curves differ significantly at long t F times. Applying this technique to a series of four substrates, the kinetic data of Table 15 were obtained. Also included are data from homogeneous solution studies on the radical anion of 1,1-diphenylethylene. For the two esters, radical anion coupling seems to be the preferred reaction mode under these conditions, whereas both mechanisms operate for the two nitriles and 1,l-diphenylethylene, k , being 10- 100 times larger than k i in these cases. Looking back at the preparative results, we are in a position to assess the validity of the mechanism (9 8c) originally proposed versus the radical anion coupling mechanism [eqn (98a)l. In the macroscale TABLE 1 5 Kinetic Data for Two Different EHD Coupling Mechanisms

Substrate Diethyl fumaratea,b Dimethyl fumarate b,c Cinnamonitrilebsc FumaronitrilebrC 1,l-Diphenylethylened

k 2 , M-'s-l R'- + R m 2 --+ - R - R34 110 880 7 x lo5 los

k ; , M-'s-l R.- + R + - R - R*

60 1o4

1o4

Childsetal., 1971. In DMF. Pugiisi and Bard, 1972. Staples et al., 1969. These values were obtained in HMPA and are believed to pertain to free ions, not ion pairs. For the ion pair with Na', k , is of the order of l o 6 M sC1. a

86

L. EBERSON AND K. NYBERG

electrolyses, it is common and indeed often necessary (see p. 34; Baizer, 1964; Beck, 1968) to employ substrate concentrations in excess of 1 M , thus favouring radical anion/substrate reaction. On the other hand, the voltammetric experiments are typically conducted at the 1- 10 mM level of substrate concentration. Given a concentration ratio of 1 O2 - l o 3 , it is entirely reasonable that the radical anion/ substrate mechanism should predominate under preparative conditions, and the other mechanism under analytical conditions. Anodic coupling reactions of the biaryl and diphenylmethane type (Table 8, nos. 3 8 - 4 0 ) seem to be of the radical cation/substrate coupling type, at least when judged from results obtained in macroscale electrolyses with high substrate concentrations (see pp. 33-34). The other mechanism has been proposed in a case of an intramolecular biaryl coupling and also extended to the general case, but this suggestion seems to need further substantiation (Ronlin et al., 1973). The system investigated is shown in eqn (100). In our

Me0

OMe M

e

o

w

e

@OMe f

d

I

I

OMe

0.03 mA cm-2 100 mA cm-2

OMe

62% 19%

opinion, the preparative results are as expected for a radical cation/ substrate mechanism. Thus a high probability of intermolecular coupling would be expected when the radical cation is generated at a very low rate in an environment containing a relatively high substrate concentration, but some probability of intramolecular coupling (really not a favourable reaction in this case) might remain if the radical cation is generated at a diffusion controlled rate and has little substrate around to react with intermolecularly.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

87

9. THE ROLE O F ADSORPTION The role of adsorption (p. 21) in organic electrode reactions has been one of the more controversial problems' in modern organic electrochemistry (p. 3). Electrochemical and other methods designed specifically to study electrosorption phenomena have given the physical chemist a wealth of empirical knowledge regarding adsorption and some insights into the structure of the EI (p. 22; Payne, 1970); yet is has been consistently difficult to trace any significant effects of adsorption upon products and their distribution, the analysis of which is one of the preoccupations of organic chemistry. In this section we shall try t o examine available evidence on this problem, bearing in mind that it is only possible to discuss a minor part of the vast body of phenomena in this area. Adsorption from solution is really a competition reaction. Looking at the EI structure of Fig. 3 , let us for a moment assume that the solvent molecules are water molecules. Since the electrode surface is positively charged, the negative end of the water dipole is directed toward the surface for simple electrostatic reasons. This is a so-called flop-down dipole (Bockris and Reddy, 1970), and this will change its direction t o the opposite one, a flip-up dipole, if the surface is made negatively charged by changing the electrode potential. Molecules and ions that adsorb will have t o displace a certain number o f water molecules from the surface, as shown in, e.g., eqn (40). Now it is obvious that somewhere between a positively and negatively charged surface there will be a certain potential at which the surface has zero charge (potential of Lero charge = pzc). At this potential the water molecules will not have any definite direction with respect to the surface and the adsorptive interaction will be at minimum. (This is the flip-flop model o f water adsorption.) Adsorption of nonpolar molecules, being a displacement reaction involving adsorbed water molecules, will accordingly have its maximum value at-or near, since the world is never ideal-the pzc. The pzc of some representative materials is shown in Table 16 (Tomilov et al., 1972), from which it can be seen that adsorption taking place via the water displacement mechanism is at its maximum at potentials of relatively little interest t o the electrochemist, since few compounds oxidize or reduce at the potentials shown. This state of affairs has been immortalized in a song giving full coverage t o the role of adsorption in electroorganic chemistry (Fleischmann et al., 1971b).

88

L. EBERSON AND K. NYBERG

As a rule anions and cations will tend to be adsorbed more strongly at respectively, positively and negatively charged surfaces, unless specific adsorption (p. 23) occurs. The maximum in the adsorption of organic compounds near or at the pzc has been verified TABLE 16 Potentials of Zero Charge of Different Electrode Materials Electrode material

Pzc, V versus sce

Lead dioxide Oxidized platinum Graphite Gold Platinum

1.8 0.4-1.0 0.25 0.23 0.18

Electrode material Tin Nickel Lead Zinc Cadmium

Pzc, V versus sce -0.35 -0.36 -0.60 0.65 -0.68

in many cases. Table 1 7 shows representative adsorption data for organic compounds, predominantly aromatic ones. i t should first be noted that all values but one (no. 7) refer to aqueous electrolytes and that the forces involved are rather weak, -(3-8) kcal mol-' in terms of free energy differences. i n one case, benzene (no. l ) , an estimate of the number of water molecules displaced per hydrocarbon molecule gives a number of 9 k 2 for flat orientation; larger molecules should displace a proportionately larger number. in many cases 7r-bonding, similar t o the interaction within n-donor-acceptor complexes, has been suggested as the interactive force, thereby also implying that the distance between the surface atom(s) and the (Fig. 3). Note also adsorbed molecule should be something like 3.5 that in some cases a change from a flat to a perpendicular orientation has been deduced and that this occurs mostly for molecules with dipoles. Apparently the n-bonding predominates in one potential region and dipole-charge interaction in another; for a n-donor one would expect the flat orientation t o prevail at an electron-deficient surface, e.g. a positively charged one. The 7r-bonding hypothesis is further supported by calculations of the differences in intrinsic free energy of adsorption, i.e., A G a d s , with the effect of displaced water molecules subtracted. For a series of butyl, phenyl and naphthyl derivatives these were found to be -6.4, -8-9, and -12.4 kcal mole-', respectively. Adsorption of neutral molecules from nonaqueous media has only been studied in isolated cases (Payne, 1970). As far as one can judge

a

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

89

from these data, organic adsorption from such solvents should be appreciably weaker than from water, presumably due to the stronger interaction between organic solvents and solutes. Thus, AGad would presumably be of the order of - ( l - 2 ) kcal mole-’ in organic solvents. What then are the effects of adsorption in organic electrode processes? We have already touched upon a few effects, but it is appropriate to try to summarize them at this point: Concentrations of ions and molecules in the EI might not be the same as in the bulk electrolyte. The structure of the EI might influence the stereochemistry of the reaction, e.g. due to an orientation effect of the electrode surface on the substrate molecule. One process might “outrun” another one, although the order of discharge potentials would seem to indicate otherwise. E l /2-values might change in a homologous series as the chain length increases. In the following we shall discuss some possible cases of adsorption effects, remembering that alternative explanations are possible and not distinguishable at the present state of knowledge. Concentration Effects in the E I

In principle it is possible to measure the adsorption properties of all the components of an electrolyte solution and thus construct a series of models for the EI at different potentials; this would show the composition and, possibly, the positions of ions and molecules in the EI. However, a little thought dismisses this procedure as impossibly laborious, and therefore resort has t o be made to educated guesses about the possible effect of EI composition. Earlier we discussed (p. 36, Table 3 ) product distributions of concurrent acetamidation and acetoxylation in the presence of different anions as possibly being due t o an adsorption effect, the better H-bonding anion bringing relatively more acetic acid into the EI and thus favouring acetoxylation to a disproportionately high degree. It could not be entirely ruled out, however, that acid catalysed processes in the Nemst layer might be responsible for this phenomenon. Looking a t another acetamidation reaction, oxidation of RCOO- in acetonitrile, we find a similar puzzling phenomenon under conditions which do not allow for the acid catalysed

TABLE 17

W

0

Adsorption Characteristics of Organic Compounds a t Electrodes

No.

Compound

Electrode material

Aqueous electrolyte solution (M)

AG& kcal mol-'

Estimated' orientation (charge o n surface)

1 Benzene*

Pt (plated)c H2SO4 (0.5)

-6.5

Flat

2 Benzene' 3 Cyclohexane'

Au Au Au Ni

H2SO4 (0.5) H2SO4 (0.5) H2SO4 (0.5) NaC104 ( 1 )

-6.1 -2.8 -8.0 -6.0

Flat

Pt Hg

NaC104 (1) NaC104 (0.1) in DMF

-8.4

Hg Hg Hg C paste

H2SO4 (0.05) Na2S04 (0.05) Na2S04 H2SO4 (1)

1:lat Perpendicular (pzc); flat(+) Flat Perpendicular(+), flat(-) Perpendicular (-- ), flat(+) Flat (+)

4 5

Naphthalene' Naphthalenef

6 Naphthalenef 7 Naphthalene, biphenylg 8 9 10 11 12 13 14 15 16 17 18

Phenolh Pentafluorophenoli CoumarinJ Diphosphate of 1,4-dihydroxy-4methylnaphthalenck Anilinez C~HSNH;" ~,~-(CH~)ZC~" H~ Anilinen Pyridine" QuinolineO Denaturated DNA

Hg Hg NHg H; Hg Hg Hg Hg

KC1 (1) HCI (0.1) HCI (0.1) KCI ( 1 ) KC1 ( 1 ) K N 0 3 (1) McIlvaine buffer, pII 5.90

-6.0 -8.3 -4.5 -3.3

Flat Flat

Perpendicular( ), flat(+) Flat (pzc) Flat (piic) Flat Flat Flat (+ and -) Bases flat o n the surface

Assumed interaction

*

Covalentid 9 2 H 2 0 molecules displaced per benzene molecule adsorbed n-bonding n-bonding n-bonding ( t o some degree) n-bonding

P

n-bonding n-bonding n-bonding n-bonding n-bonding n-bonding n-bonding

a By determination of the coverage (molecules cm-' of actual surface area), one can estimate thr surface area occupied by each molecule and hence jud e its orientation. Heiland et al., 1966. Catalytic electrode; actual surface area 50-100 times the geometrical area. Probably loss of aromatic character. Dahms and Green, 1963. f Bockris et al., 1964. R Kaganovich et al., 1570. Damaskin et al., 1964. f Damaskin et al., 1967. 1 Partridge ct al., 1966. Meier and Chambers, 1969. Damaskin e t al., 1964a. Blomgren and Bockris, 1959. Barradds and Conway, 1961; Conway and Barradds, 1961. Bordi and Papeschi, 1969. p Valenta and Nurnberg, 1974.

%

9 2 U

'

U i

92

L. EBERSON AND K. NYBERG

rearrangement (buffering effect of RCOO- ; large concentration of water). These results (Muck and Wilson, 1970) are summarized in eqn (101) for the case of valerate; three products actually form from the nitrilium ion (Kornprobst et al., 1968, 1970; Thomas, 1971), but for simplicity we denote them as “amides”. Two facts are not explicable in terms of the concentrations in the bulk electrolyte; (a) the efficient capture of the carbonium ion by acetonitrile in

-

I

-2e-

C4HgCOO-

-c02

H2O (v/v)/% 15 20 25 30 35 13 23 39 23 44

40 26%

C4Hs

10 21

C4H90H

0

0

0

0

43

22

Amides 69

81

76

60

34

34 26%

48%

C4H9 CH CN H20

(101)

competition with water, a stronger nucleophile by factor of 40- 50, and (b) the absence of the alcohol at water concentrations of up to roughly 15 M . The rationalization of this behaviour, that carboxylate anions would block the surface for access by water molecules, seems entirely reasonable, especially since it was found for the same reaction that long-chain carboxylates underwent predominantly coupling. Adsorption, reinforced by a stacking effect of the long alkyl chains, was again invoked to explain this phenomenon. Another variation on the same theme is the predominant formation of alcohols in the oxidation of carboxylates in mixtures of pyridine and water. To take but one example, the oxidation of a 3-bicyclot 3,l ,O] hexanecarboxylate in water-pyridine ( 3 :1) gave isolated yields of alcohols of up to 60% (Gassman and Zalar, 1966), although we generally think of pyridine as a far stronger nucleophile than water by a factor of perhaps l o 3 - l o 4 . And, in these two examples, why does the carboxylate itself not function as a nucleophile (as can be found in other cases under different conditions)? The competition between the radical and cationic pathway in the Kolbe reaction is strongly influenced by the presence of forei‘gn anions, e.g. (210, and F-, as has been known for a long time (for reviews, see Eberson, 1968, 1973). Exactly how strong this influence is was recently shown for the oxidation of phenylacetate ion; the addition of only one hundredth molar proportion of perchlorate relative to phenylacetate ion completely suppressed formation of the

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

93

dimer, bibenzyl. In the absence of perchlorate ion, the dimer yield was 68% (Coleman et al., 1971). Preferential adsorption of perchlorate ion was believed t o cause the effect, favouring further oxidation of the radical t o the carbonium ion (a monomolecular reaction) at the expense of radical coupling (a bimolecular reaction). In cathodic reactions, adsorption effects would be expected to arise predominantly from adsorbed cations. Perhaps the most interesting phenomenon ascribable to adsorption of cations is the asymmetric induction observed when cathodic reduction of a prochiral substrate is performed in the presence of a chiral cation (Gourley et al., 1967; Homer and Degner, 1948). Table 18 summarizes a few selected cases of this reaction type (for reviews see Homer et al., 1972; Eberson and Horner, 1973) just t o illustrate the scope of the reaction. The structural variables affecting the optical yield are unfortunately too complex t o cover here. Characteristic of this process is the low concentration of chiral cation necessary to induce optical activity (10% of the substrate concentration is sufficient) and that the chiral salt is recovered mostly unchanged (except for no. 3). Quaternary ammonium ions work as well as substituted ammonium ions, thereby making less probable the mechanism first suggested (Gourley et al., 1967, 19 70), namely, hydrogen atom transfer from an ammonium radical R3fiH to the substrate. Today it seems reasonably well established that adsorption of the chiral cation, possibly as a complex with an intermediate is connected with the asymmetric induction observed (Horner et al., 1972; Kariv et al., 1973a). Proton availability or nonavailability in the EI plays an important role in many cathodic reactions. In fact, the exclusion of proton donors from the EI by quaternary ammonium ions in aqueous

+e-

CH2=CHCN

CH2-kHCN

+e-

CHjCHzCN

1

Et4NOTs

88

Ph4POTs

88

2

CaOTs

47

50

Et3hH

35

60

NaOTs

66

30

LiOTs

15

84

(102)

TABLE 18

W k P

Asymmetric Induction during Cathodic Reduction of Prochiral Compounds in t h e Presence of Chiral Cations ~~

l

'

q

o

CH3

2b

+2e-

-+ +2H

PhCOCH3

3

Chiral Cation

Reaction

No.

~ H

Sparteine-H+ Emetine-H+ Narcotine-H+ CM3

PhCH(OH)CH3

Sameasno. 2

II

4d

PhC-CH3

5'

Brucine-H+

NHCH2Ph

+2e- PhCH-CH3 I

PhCOCOOH

+2e+ -

+2d

PhCH(0H)COOH

r

PhCH(OH)CH(CH,)G(CH3)3 ( l R , 2 s ) PhCH(OH)CH(CH,)k(CH3)3 ( l S , 2 R ) Quinidine-H+,Ca t

+=+

NCH2Ph

Optical yield, % (configuration)

1.70 V -1.75 V - 1.80 V -1.90 v - 2.00 v - 2.10 v ~

M

W

m !a

4-5 (S) 13- 1 6 (S) 16-1 7 (S) 14-15 (S) ll-l3(S) 8-1 0 (S)

PhCH(OH)CH(CH3)k(CH,), (1R, 2s) PhCH(OH)CH(CH3)i(CH3)3 ( l S , 2 R )

7.0(R)' 6*6(S4

S try chnine-H+

22(R)g

Gourley et al., 1967, 1970. Horner and Degner, 197 1. Since quinidine is itself reducible in this potential region, these runs were carried out with prereduced chiral salt; the resulting product mixture contained chiral material too (Kariv et al., 1973, 1973a). Horner et al., 1972. At -10°C; the optical yield decreased with increasing temperature. f At 0°C; the optical yield decreased with increasing temperature. 8 Jubault et al., 1973. a

'

0"z >

3

F z

s

M

P

0

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

95

electrolytes is crucial for the successful operation of the EHD process with acrylonitrile (Baizer, 1964). Two reagents, water and acrylonitrile, compete for the radical anion as it is formed (Beck, 1968), as shown in eqn (102). It is difficult to explain these results unless one assumes that adsorbed tetra-alkylammonium ions create a hydrophobic environment in the EI and thus keep water molecules from reacting with the radical anions. In the presence of alkali cations, high concentrations of DMF o r DMSO have the same effect, presumably by dehydrating the cations (for a discussion, see Beck, 1972a). Anomalous proton donor effects were also noted in reaction (103) (Fry and Reed, 1971, 1972). Here added acids, such as a phenol or

\

&:

Br

(103)

c1

H+ donor/% Et4 N+

62

38

H2O ( 5 M ) Et3NH+

80

HOAc (1 M)

20 8 32

92 68

2,4,5-Trimethyl-

40

60

phenol (1 M)

acetic acid, are less efficient proton donors than the trialkylammonium ion, indicating that they never reach the site of reaction in the EI.

Orientation Effects Induced b y the Electrode Surface The suggestion that 7-r-bonding is responsible for the interaction between aromatic compounds and the electrode surface, particularly

96

L. EBERSON AND K. NYBERG

a positively charged one, immediately raises the problem of possible steric control of the reaction by the electrode surface. The substrateelectrode complex can be regarded as being severely sterically hindered to approach by reagents from the electrode side. A planar molecule designed t o have different steric requirements on each of its faces, would then be expected t o be adsorbed with its least hindered side toward the electrode, thus making possible stereoselectivity towards a thermodynamically less stable isomer (Dirlam et al., 1972; Dirlam and Eberson, 1972; Eberson and Sternerup, 1972; Sternerup, 1974). This principle is illustrated in (104) for a-acetoxylation of a 2-a1kylindan . OAc -

e

R

ECE +

-

Free cation

T

P

67

OAc

I

,OAc

Table 19 shows results from the a-acetoxylation of two 2-alkylindans under different conditions; analogous homogeneous reactions are included for comparison. Table 19 demonstrates that, at most, a weak effect is noticeable in the case of 2-t-butylindan, in that anodic oxidation on platinum, a strongly adsorbing metal, gives a cis/truns ratio that is about 10 times that observed from a 2-t-butyl1-indanyl cation generated by solvolysis. Ratios on the weaker adsorbing electrode materials, carbon and lead dioxide, fall in between. For a smaller substituent (R = methyl) no effect is observable. This idea has been extended to 1-t-butylacenaphthene(Dirlam and Eberson, 1972), predicted to be a more rigid molecule and a better

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

97

n-donor, but the results were difficult to interpret due to experimental difficulties in finding suitable reference reactions. Additive acetoxylation to 3-alkylindenes [eqn (1O S ) ] gives essentially the

i,

H R

same stereochemistry as that observed for homogeneous acetoxylation using cobalt(II1) acetate (Cedheim and Eberson, 19 74; Eberson, 1974). Orientation effects by the electrode surface have been invoked in a large number of other cases (for reviews of the stereochemistry of electrode processes, see Eberson and Horner, 1973; Fry, 1972a), for example in halide reduction (for a review, see Casanova and Eberson, 1973), formation of dimers in the Kolbe reaction (Hawkes et al., 1973), reduction of systems with double and triple bonds (Horner and Roder, 1969), and anodic coupling of 1-methylcorypalline TABLE 19 CislTrans Ratios in the a-Acetoxylation of 2-Alkylindans 2-Alkyl group t-Bu'

Reaction and reaction conditions

Cisltrans ratio

Anodic oxidation in HOAc/NaOAc, at Pt

16:8 4

t-B@

Anodic oxidation in HOAc/NaOAc, at C

5:95

t-Bua

Anodic oxidation in HOAc/NaOAc, at PbO2

5:95

t-Bu'

Oxidation of 2-t-butylindan by lead tetraacetate in H O A C ~

5:95

t-Bu'

Solvolysis of 1-p-nitrobenzoyloxy)-2-tbutylindan' in HOAc/NaOAc

CH3d

Anodic oxidation in HOAc/NaOAc, at Pt

51:49

CH3

Anodic oxidation in HOAc/NaOAc, at PbO2 or C

45:55

CH3d

Oxidation by Co(II1) acetate in HOAc

37:63

CH3'

Solvolysis of 1-p-nitrobenzoy1oxy)2-methylindanc in HOAc/NaOAc

43:57

a Eberson and Sternerup, 1972. Heterogeneous reaction. Both isomers. Sternerup, 1974.

'

<2:98

98

L. EBERSON AND K. NYBERG

(Bobbitt, 1973; Bobbitt et al., 1971). Although electro- and photopinacolization of aromatic aldehydes and ketones show practically identical stereochemistry in spite of one being heterogeneous and the other homogeneous (Stocker and Jenevein, 1968, 1968a, 1969, 1971; Stocker and Kern, 1968), it has been shown that adsorption must play a role in this cathodic coupling reaction (Bewick and Cleghorn, 1973: Puglisi et al., 1969).

Competing Reactions; the Role of Adsorption In discussing Table 8, we have already noted that many anodic processes involve discharge of a substrate that is more difficult to oxidize than the SSE; experimentally this manifests itself as an anodic shift of the voltammetric curve when the substrate is added. Reactions displaying this behaviour are, among others, methoxylation, cyanation, possibly acetoxylation in the presence of nitrate ion, and, most pronounced, the Kolbe reaction (p. 16). The latter process takes place at potentials above 2.3 V versus sce in aqueous medium, thereby completely suppressing oxygen evolution in the region of 1.5-1.8 V. The transition region from one process to the other one is characterized by a steep potential rise, until the critical potential (about 2.2 V) for the Kolbe reaction is reached. The transition region is believed to involve a build-up of an adsorption layer of acetate ion which gradually displaces water from the surface. Some authors have instead claimed that a layer of platinum oxides is the prerequisite for the Kolbe reaction; whatever it is, there is a general a'greement that platinum(0) sites are only available in the low-potential region (below 0.8 V; this is the potential range of interest in fuel cell chemistry). For a more detailed discussion of this problem, the reader is referred to reviews of the different viewpoints of the Kolbe reaction (Conway and Vijh, 1967; Eberson, 1968, 1973; Eberson and Nyberg, 1976). The previously discussed oxidation of olefin-Hg(11) complexes (p. 48) in aqueous perchloric acid behaves similarly t o the Kolbt reaction, oxygen evolution between 1.4 and 1.8 V being displaced by the organic process at a potential above 2.3 V. In general, competition reactions between two substrates instead of between substrate and SSE components can show similar features as those discussed above. As an example, methyl iodide reduces at a potential 0.1 V more anodic than ethyl iodide at a lead cathode and

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

99

hence should be reduced t o give lead tetramethyl preferentially when electrolysis of a mixture of both is carried out a t the foot potential of the methyl iodide wave. However, it turns out that the ethyl radical is formed simultaneously in appreciable quantities, and mixed lead alkyls are formed. A catalytic cycle (106) involving surface lead alkyl species was invoked (Flrischmann et al., 1971a). Pb

Me1 + PbMe -e-

-I

EtI

PbEtMeI

4-e-

PbMeEt,etc.

-1-

(106)

-

Another type of -mutual interaction between two substrates in competition reactions is the phenomenon that one substrate can influence the product distribution from another one (Eberson and Wilkinson, 1972). Thus, 5 mole % of pyrene added to an electrolyte solution for acetoxylation of anisole (HOAc/NaOAc) changed the o / p ratio from 2-2t o 1.3. Similar behaviour was observed for other mixtures.

Adsorption uersus E , 1 2 -values The mere lengthening of an alkyl chain would not be expected to change the bond or group involved in oxidation or reduction t o any

I

5

10

15

Number of C atoms

Figure 8. Plots of E l l 2 for different classes of compounds versus number of carbon atoms in the alkyl group. A, Alkyl bromides (Lambert and Kobayashi, 1960); B , N,N-dimethylalkanecarboxamides (Mann and Barnes, 19 70, p. 290); C, trialkylamines (Mann and Barnes, 1970, p. 279); D, alkyl chloroacetates (Smirnov and Markova, 1970).

100

L. EBERSON AND K. NYBERG

significant extent. Yet we can see from Fig. 8 that the lengthening of the R groups in the classes of compounds illustrated lowers or increases the E l Iz -values by more than 0.2 V from the methyl to the C9- C1 derivative. It is reasonable to ascribe this change to stronger adsorption of the longer alkyl chains as compared to the shorter ones.

10. THE ELECTRON TRANSFER PROCESS A rather dramatic and eloquent view of the electron transfer act has been given by Bockris and Reddy. It pertains t o reduction of protons but could relate to any electroactive species. “The scene is the electrified interface. The actors are the metal, teeming with free electrons in a hierarchy of levels, the proton, whose conversion into a hydrogen atom is the theme of the play, and the water molecules, whose association with the proton earns for it an important role. Electrons are too weak in energy to climb the potential-energy hill at the metal-solution interface. Nevertheless, their quantum qualities permit the electrons to sneak through the barrier in ghostlike fashion provided there are hospitable acceptor states for the electron, i.e. welcoming vacant energy levels in the hydrated hydrogen ions.” It is really difficult t o improve upon this description, and yet we shall try t o give a simple, perhaps even nai‘ve, account of the electron-transfer act in terms familiar t o practitioners of organic chemistry. To start with, we postulate that electron transfer will take place from the highest occupied molecular orbital (HOMO) or to the lowest unoccupied orbital (LUMO), of the electroactive species; transfer t o or from other orbitals will be energetically very unfavourable. Thus, the situation at an electrode potential at which no electron transfer can take place is illustrated in Fig. 9a. Here the energy of the free electrons in the metal is denoted E F , the Fermi energy, and the energy level occupied by the free electrons is called the Fermi level, the energy of which is fairly narrowly defined, t o f kT at ordinary temperatures. The MO of interest in the electroactive species is the LUMO, and we are interested in transferring one electron from the metal t o this orbital, i.e. an act of reduction. As the levels are placed in Fig. 9a, electron transfer between the two energy levels cannot take place, since there is n o way to bring about the exact matching of the two levels required for electron transfer to

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

101

-LUMO

Figure 9. Different situations of EF, the Fermi energy, with respect to the LUMO of the substrate: (a) No electron transfer possible, ( b ) electron transfer possible, and (c) electron transfer possible due to excitation of the molecule, depicted as occurring along a Morse curve.

occur.' The Fermi energy EF is constant, and there is n o way for the molecule t o decrease the energy of the LUMO further. By changing the electrode potential across the EI we can raise the energy of the Fermi level t o accomplish the required exact matching of the two levels (Fig. 9b). This situation allows for the quantum mechanical tunnelling of the electron in the Fermi level through the energy barrier separating it from the LUMO of the molecule. Let us continue t o change the electrode potential towards more negative potentials and accordingly the Fermi level to a value higher This is the same situation as in electron transfer mechanisms in homogeneous medium, where we have to take into account two transition states, identical in energy but differing with respect to the position of the electron (Reynolds and Lumry, 1966).

102

L. EBERSON AND K. NYBERG

than that o f the LUMO in Fig. 9b (Fig. 9c). In this situation, the exact matching o f energy levels is seemingly not fulfilled, and the question is: can electron transfer take place now? It certainly can, and this is achieved by changes in the molecule to which the electron is to be transferred, changes that raise the LUMO to a higher level, such as desolvation, bond lengthening, bond angle deformation, adsorption. Thus, the energy of the LUMO will be raised until it again fits with the Fermi energy of the free electrons in the metal. The effect of the rise in the electrode potential will be an increase in the rate of electron transfer, since now more molecules will have a chance to fulfil the energy requirement for electron transfer. Now, to go to the experimental situation, what happens as we insert a metal electrode into an electrolyte solution without connecting it to an external electron source? As we have discussed before (p. 22), an EI is built up and hence a certain potential is established across the interface region. At this potential, charge transfer between electrode and electroactive species takes place, but, since no net current flows, the rates of electronation and de-electronation are identical. The system has reached the equilibrium potential at which 9 the current density zcfor electronation is equal to the current density of de-electronation i. This current density is designated io, the equilibrium exchange current density (cf. Table 6), given by the expression:

-+

c

Here F is the Faraday, k , and k c the heterogeneous rate constants for electron transfer in either direction, C M and c M - are the concentrations of the electroactive species M and its electronated form M-, 01 is the transfer coefficient, and E,, is the potential difference across the interface at equilibrium. This equilibrium potential is given by the Nernst equation,

RT E,, = E o +---In---,

F

CM CM-

where E o is the standard potential of the redox couple M/M-. In principle it is possible to calculate E , from thermodynamic data or sometimes to measure it.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

103

The equilibrium exchange current density, i, , can vary over a wide range (see Table 6). For a relatively large value, the electron transfer reaction is said t o be reversible; for small values the designation irreversible is used. This terminology has created a lot of confusion due to the differing meanings of reversibility in electrochemistry and thermodynamics, and hence it has been suggested that the terms fast and slow should be more appropriate t o use. Sometimes io is converted into a heterogeneous rate constant k o (unit = cm s - ' ) , valid at the standard electrode potential E,, using the expression

When we connect the electrode t o an external electron source and start changing the potential across the EI toward more negative values, a net current i will start to flow. This nonequilibrium current density is given by the Butler-Volmer equation,

where 7 now designates the overpotential (p. 45), the difference between E,, and the actual electrode potential. The slower the electron transfer reaction is and hence the smaller i, is, the larger r ) must be in order for the reaction t o occur at a measurable rate. It is easy to see that for a value of r ) > 0.12 V, the Butler-Volmer equation reduces to

The logarithmic form of this equation is the so-called Tafel equation. This is valid at sufficiently high overpotential and under conditions where mass transfer is negligible, and in principle allows the determination of i,. Thus i, appears as a fundamentally interesting constant for an electrochemical reaction, in that it is a measure of the rate of the electron transfer reaction at equilibrium, and hence it determines the appearance of the i versus E curve of the process. To go back t o Fig. 9c, the situation illustrated there obviously must have some meaning in terms of transition state structure. The molecule has changed its energy from the ground state in order to match the Fermi energy in the metal, e.g. by bond lengthening, bond angle deformation, desolvation or at least rearrangement of the solvent shell, and adsorption. In this state, when the energy levels match, the electron tunnels through the energy barrier during a time

104

L. EBERSON AND K. NYBERG

interval of about lo-’ s; hence Franck-Condon restrictions apply and we can consider the atoms of the molecule as stationary during the electron-transfer act. The problem of estimating the structure of the transition state of an electrochemical process is especially important with respect t o possible concerted reactions, i.e. reactions at which the electron transfer coincides with bond breaking or formation. The reduction of a protonated water molecule has been described in exactly those terms which an organic chemist would use for an ordinary concerted mechanism.

H

__________

H

+/

0 \

H

10-l6s

+

0 ’

\

H

Before electron transfer

(106)

H

After electron transfer

As shown in eqn (106), one 0-H bond of the hydroxonium ion is lengthened before electron transfer, whereas afterwards this bond is shown as having been broken and replaced by a lengthened bond between the hydrogen atom and a metal atom in the electrode surface. If ever a mechanism was described as concerted, this is it. One of us suggested some years ago (Eberson, 1963) that the Kolbe reaction might be a concerted reaction according to eqn (107). The ensuing debate (Conway and Vijh, 1967, 1967a; Eberson, 1967, 1969; Reichenbacher et al., 1968, 1968a) never really managed t o

(107)

-0 Before electron transfer

clarify what some of the participants implied by the concept of a concerted mechanism. Here we should like to repeat and elucidate what the original description of the concerted Kolbe mechanism was. A lengthening of the C-COO- bond, synchronous with bond angle deformation in the COO- group of a possibly adsorbed carboxylate ion was envisaged so as t o make it more “carbon dioxide-like”. When the electron transfer has occurred from such a modified species, we are left with a deformed COz molecule and a radical R.; RCOO. is not formed as a discrete species. We still maintain-in fact reinforced in our belief by the mechanism given in eqn (106)-that the electron

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

105

transfer step of the Kolbe reaction might well be described in terms of such a concerted process and that certain experimental facts are difficult to account for by other mechanisms (Eberson and RydePetterson, 1973; Muck and Wilson, 1970). Numerous concerted electron transfer/bond breaking or formation mechanisms have been suggested. We have already touched upon one (p. 84). the concerted electron transfer/coupling mechanism for EHD. If this mechanism is correct, it should have stereochemical consequences in a cyclic EHD process involving a conjugated system, such as that shown in eqn (108). A concerted step would be

-I

H

6

disrotatory

Phv

II

0

expected t o obey the Woodward-Hoffmann rules, i.e. ring closure should occur in a disrotatory fashion. The test of this idea has s o far been prevented by the impossibility of reducing the fulgenic acid derivative of eqn (108) to a cyclobutane derivative (Andersson and Eberson, 1974). The facile formation of cyclopropane derivatives from 1,3-dihalides was originally formulated as a concerted mechanism (109) (Rifi, 1967, 1969). It has been shown more recently, however,

that the ring closure of both m e ~ oand dl forms of substituted 1,3-dihalides is nonstereospecific (Fry and Britton, 1973) and that carbanionic intermediates can be trapped by protonation (Casanova, 1974). A final example of a proposed concerted mechanism (110) is the anodic oxidation of the olefin-Hg(I1) complex (Fleischmann et al., 1969) previously mentioned on p. 48. H

I

CH3-C-CH2

-

CH?COCH?

+ Hg2+ + H+

(110)

106

L. EBERSON AND K. NYBERG

Whether the examples given above really are cases of concerted electrochemical mechanisms or not, we think that the concept in itself is an entirely reasonable one and should be thoroughly tested. There is an intellectual difficulty in the description of the electron transfer act given above and the derived consequences concerning transition state structure. Given that we can adjust the electrode potential at will over a fairly wide range, do we not then have to assume that there is a continuous range of transition states of different structure? This interpretation has actually been given in a paper on the cathodic reduction of cyclo-octatetraene COT (Allendoerfer and Rieger, 1965): “The si
11. STRUCTURE AND REACTIVITY So far we have used Ei 1 2 as a measure of overall electrochemical reactivity. In this Section, we shall examine some aspects of this practice, keeping in mind that this vast field has been thoroughly covered previously (Zuman, 1967). We shall first briefly examine the kinetics of the electron transfer process itself, however. Kinetics o,f Electron Transfer at Electrodes Table 20 gives a representative selection of kO-values (p. 103) for electron transfer reactions of organic molecules at electrodes. It

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

107

T A B L E 20 R a t e C o n s t a n t s for Electron Transfer Between Organic C o m p o u n d s a n d t h e Mercury Electrode Compound Naphthaleneb AnthraceneC (Anthracene)' - c Tetraceneb PeryleneC (Perylene) 'truns-S tilbeneb cis-? tilb,eneb

Medium"

E,,

v versus sce

k o ( c m s-1)

(0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1)

-2.49 -1.95 -2.55 -1.58 -1.67 -2.26 -2'15 -2.18

1.0 >4 0.009 1.64 >4 0.009 1.22 0.42

D M F / T B A I (0.1) D M F / T B A P (0.1) D M F / T B A P (0.1) D M F / T E A P (0.1) D M F / T B A I (0.1) D M F / T B A I (0.1) DMF/TBAI (0.1) DM F / T B A I (0.1 ) D M F / T B A I (0.1)

-2.46 -1.62 -1.86 -2.0 -1.73 -1.14 -1.38 -1.42 -0.80

0.1 8 0.002 0.12 0.5 >4 2.7 0.28 0.1 5 5.0

DMF/TBAI DMF/TBAI DMF/TBAI DMF/TBAI DMF/TBAI DMF/TBAI DMF/TBAI DMF/TBAI

2,2 ,4,4 ,6,6'-Hexamethyl-truns-stilbeneb Cyclooctate traened (Cyc1ooctatetraene)'Azobenzenee BenzophenoneC p-Nitro toluenef Nitrornesitylenef Nitrodurenef m-Dinitrobenzenef

"

TBAI = tetrabutylammonium iodide; TBAP = tetrabutylammonium perchlorate; TEAP = tetraethylammonium perchlorate. Dietz and Peover, 1968. Aten and Hoijtink, 1961. Huebert and Smith, 1971. Aylward et u L , 1967. f Peover and Powell, 1969.

should be noted that these measurements refer t o electrochemically well-behaved systems, in that n o complicating consecutive reactions occur in the systems studied. The reactions shown in Table 20 belong to the category of fast or reversible electrochemical reactions (p. 103), the borderline between fast and slow or irreversible processes being placed at about lo-' cm s-'. Theoretical calculations (Hale, 1971) on some of the fast systems have shown that the electroactive species can actually be assumed t o be located outside the OHP (p. 23) or something like 1 5 away and yet give good ageement with experiment. A comparison between rate constants for electron transfer of some stilbene derivatives at the mercury electrode and an analogous homogeneous electron transfer process revealed a parallelism between the two sets of constants (Dietz and Peover, 1968). R e p - d i n g COT and its radical

a

108

L. EBERSON AND K. NYBERG

anion, we may note that the formation of the dianion is faster than the formation of the radical anion (p. 75), a nice confirmation of the Huckel rule. Interesting problems with respect t o the mechanism of electron transfer between electrode and molecule or between molecules are raised (cf. p. 28) by the observation that naphthalene radical anion, easily generated at the mercury cathode at about -2.5 Vversus sce, can be used for the reduction of compounds otherwise electroreducible only with great difficulty or not at all, e.g. alkyl chlorides and even fluorides (Garst and Barton, 1974; Lund e t al., 1974; Sease e t al., 1969). Thus naphthalene can be considered as an added catalyst in these electrochemical reactions. The results indicate that the requirements for suitable orbital overlap between two reactants are easier to fulfil between molecules than between a molecule and an electrode surface. Experiments designed t o find analogous reactions on the anodic side have so far failed [e.g. attempts t o oxidize RCOO- by (p-BrC6H4)3N-+t o simulate the Kolbe reaction (Eberson and Helgke, 1971)].

Half wave Potentials as Reactivity Indices Half-wave potentials have been used extensively as measures of electrochemical reactivity. Since one assumes that anodic oxidation (cathodic reduction) takes place by removal (addition) of an electron from (to) the electroactive species, it is logical to try t o correlate half-wave potentials for oxidation (reduction) with the energy of the HOMO (LUMO) of the electroactive species. Such correlations belong to the most successful corroborations of MO theory that exist (Gleicher and Gleicher, 1967; Hoijtink, 1955, 1958; Neikam and Desmond, 1964; Pirkinyi and Zahradnik, 1965; Pysh and Yang, 1963; Zahradnik and Pirkinyi). Criticism has been voiced (Mark, 1968) of the use of E,/2-values obtained in media containing nucleophiles/proton donors-deliberately added or present as impurities-for correlation with MO parameters since it could be shown that small amounts of such reagents could affect the El / 2 -value. However, the change observed, ca. 0-05 V in E / 2 for the addition of 10 mM phenol in the reduction of an aromatic hydrocarbon, can hardly affect the gross appearance of these correlations. The B , / 2 -values quoted for the reduction of aromatic hydrocarbons in the presence of suspended alumina (p. 76) show that the difference from values obtained in 96% dioxane/H20 is a t most 0-2 V larger

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

109

than that mentioned above, but again this has insignificant consequences for the gross appearance of correlations with MO parameters. Table 21 contains E , l2-values for the oxidation and reduction of a number of organic compounds, selected to show the gross effects of structure. Excellent empirical correlations with gas phase ionization potentials (oxidation) and electron affinities (reduction) were soon found in limited series of compounds, e.g. aromatic hydrocarbons (see, for example, Pysh and Yang, 1963; Briegleb, 1964), but it was later shown to be possible t o combine results for covering all types of compounds for at least oxidation (Fig. 10;

l1

t

E1/2 Figure 10. Plot of vertical ionization potential versus half-wave potential for oxidation (data taken from Miller et al., 1972).

Miller et al., 1972). In general one notes that electron-donating substituents ease oxidation and electron-attracting ones reduction and vice versa. We have previously commented on the fact that E , / 2 -values might be influenced by adsorption properties, e.g., of a long alkyl group (Fig. 7). While this factor would be expected to be most pronounced for slow electrochemical reactions, it should be taken into account whenever correlations involving substrates with widely differing adsorption properties are discussed.

110

L. EBERSON AND K. NYBERG

TABLE 21 Half-wave Potentials for Oxidation and Reduction of Organic Compoundsa

Compound Pentane Octane 2-Methylpentane 2,2-Dimethylbutane Ethylene 1-Butene 2-Butene 1,3-Butadiene 1,4:Cyclohexadiene Methyl chloride Methyl bromide Methyl iodide Alcohols Dimethyl sulphide Dimethyl sulphoxide Trimethylamine Carboxylic acids Carboxylates 2-Nitro-2-methylpropane Benzene Toluene Mesitylene Hexamethylbenzene Anisole Phenol Aniline Bromobenzene Biphenyl Nitrobenzene p-Dinitrobenzene Acetophenone Benzaldeh yde Phthalic anhydride Naphthalene Anthracene Phenanthrene Tetracene Coronene Pyridine Quinoline Thiophene

Ell2 for oxidation, versus sceb 3.1 3.0 3.3 3.5 3.1 5 3.03 2.46 2-28 1.85 -

2-37 >2.5 1.60 1.98 1.07 2-5 0.1-1.7 -

2.29 2.2 1 1.78 1.20 1.65 1-29 0.95 2.23 1.73 -

-

1.59 1.09 1.48 0.78 1.18 2.07 1.98 1-95

E l 2 for reduction, V versus sceC -d

-2.7 -1.96 -1.63

-1.64

-2.4 -2.58 -1.08 -0.54 -1.99 -1.80 -1.36 -2.53 -1.95 -2.44 -1.58 -2.07 -2.6 1 -2.13 -

Taken from published compilations (Mann and Barnes, 1970). In acetonitrile at platinum. In DMF at mercury. Signifies that the compound is non-reducible (= non-oxidizable)below the cathodic (anodic) limit. For ROH and RCOOH, proton reduction takes place of course. a b

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

111

R o l e of Electrochemical Parameters in Physical Organic Chemistry Since electrochemical data with proper precautions can be converted into thermodynamic quantities, the determination of E l l 2 -values and similar electrochemical parameters can often give very useful thermodynamic information that cannot easily be obtained in any other way. To discuss these applications here would be outside the scope o f this article, but we should like t o draw attention to the use of electrochemical parameters for the study of such important problems in physical organic chemistry as the following: Aromaticity (see e.g. Fry et al., 1974; 0 t h et al., 1972). Antiaromaticity (see e.g. Breslow and Chu, 1973; Breslow and Mazur, 1973; Rieke and Hudnall, 1973). Conformational analysis (see e.g. Wilson and Allinger, 1961 ; Zavada et al., 1963). Ringstrain effects (Rieke et al., 1971). Cyclopropyl conjugation (Baizer e t al., 1970). Nonbonded intramolecular interactions (Shono e t al., 1972). Mechanisms of redox processes, especially those involving metal complexes and organic substrates. Nucleophilic reactivity (Edwards, 1954). Mechanism of the S N 2 reaction (which possibly has an electron transfer component; see e.g. Bilevich and Okhlobystin, 1968; Bank and Noyd, 1973).

12. INFLUENCE O F THE ELECTRODE MATERIAL The electrode material plays an important, although little understood role for the outcome or organic electrosyntheses. Innumerable reports (see e.g. Swann, 1956) bear witness to much painstaking work on electrode preparation and pretreatment, sometimes t o an extent that one despairs of ever getting any order in this vast, amorphous body of know-how. Just t o take one example, it has been reported that the temperature at which a certain solid electrode was cast had a marked effect upon product distribution (Swann et al., 1966)! The choice of electrode material has been considered most critical for cathodic processes, which is possibly a reflection of the fact that there really are lots of cathode materials from which to choose. Due

112

L. EBERSON AND K. NYBERG

to corrosion problems, anode materials are severely limited t o noble

metals like platinum and iridium, metal oxides like lead dioxide and magnetite, and different kinds of carbon, from vitreous carbon to carbon felt. Even so, anodic processes can be dependent on the choice of anode material. To continue with cathodic processes, we think that the fact that most cathodic reductions have t o be carried out in the presence of a proton donor is responsible for much of the trouble in finding a suitable cathode material. In the period before non-aqueous solvents were used, this meant that the SSE consisted of something like an alcohol/water/sulphuric acid mixture in which the electrolysis of the substrate was carried out at constant current. Given a certain composition of the SSE and a certain, often unknown, half-wave potential of the substrate under the prevailing conditions, the choice of cathode material was critical insofar as it should have a suitable overpotential for hydrogen evolution; if this were too small, only hydrogen evolution would take place while if it were too high, the substrate would indeed be reduced but perhaps to a greater extent than was intended. A correctly chosen material would sustain the substrate reduction at a potential which was “controlled” by the concurrent hydrogen evolution process. In fact, one was running a constant potential electrolysis of the substrate under constant current conditions, with the hydrogen evolution reaction acting as a “potential buffering” device. Looking at the last 15 years’ applications of cpe at high hydrogen overpotential electrodes in nonaqueous solvents, it is difficult t o escape the conclusion that the problem of cathode material had this origin to a large extent. One reservation is appropriate, however. On low hydrogen overpotential electrodes we might be dealing with electrohydrogenation reactions (p. 65), and these are of course very much dependent on the nature of the cathode material. Anodic reactions at Pt have been claimed t o be dependent upon the surface state of the platinum. The Kolbe reaction is perhaps the best known case (for a review, see Conway and Vijh, 1967) for which a change in the surface composition has been held responsible and indeed necessary for the reaction to occur. Thus, at a low potential, < 0 * 8 V, acetate in aqueous solution is completely oxidized t o carbon dioxide and water on “pure” platinum sites (i.e. we have in effect a fuel cell electrode). On raising the potential, PtO and adsorbed oxygen begin to cover the surface and oxygen evolution takes place in the range between 1.2- 1.8 V. A further increase in the

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

113

potential brings about a change in the oxide composition t o a higher oxide (Table 22), which is the prerequisite for the switch to the Kolbe reaction. These deductions (Fleischmann et al., 1965, 1965a) have been very nicely reinforced by ESCA studies of platinum surfaces' under anodic polarization in aqueous perchloric acid (1 M ) TABLE 22 Surface Composition of Platinum Anodes in Aqueous Solution Composition

(X)

Potential

Pt

Ptoads

PtO

PtO,

+@7 V

56 39 34

39 37 24

5 24 22

0 0 20

+1.2 v +2.2 v

(Kim et al., 1971). These studies were carried out in aqueous medium, and it is therefore pertinent t o ask whether a similar surface modification of platinum will take place in a non-aqueous solvent. Considering the fact that water is easily present at the 1 mM level in such solvents, one is forced to conclude that the conditions for surface oxide formation are favourable under most imaginable conditions. To continue with the Kolbe reaction, it has been shown that carbon anodes strongly favour the carbonium ion pathway (Koehl, 1964) a t least for simple alkanecarboxylic acids. Also, for phenylacetic acid and 1-methylcyclohexylacetic acid the same tendency towards carbonium ion formation on carbon anodes was observed, the phenomenon being explained as due to the presence of paramaLgnetic centres in carbon. These would bind the initially formed radicals, impede their desorption and hence promote the formation of carbonium ions via a second electron transfer (Ross and Finkelstein, 1969). However, cases of Kolbe oxidations in which no dependence on anode material was noticeable have been found more recently (Brennan and Brettle, 1973; Eberson and Nilsson, 1968a; Sat0 et al., 1968). Actually, the nature of the carbon material determines the yield of products formed via the radical versus carbonium ion pathway (Breman and Brettle, 1973). Yields of the 1 A field-ion microscope study of anodic film formation on platinum in 0.05 M sulphuric acid between +0.5 and +2.2 V gave similar results (Schubert et al., 1973).

114

L. EBERSON AND K. NYBERG

coupling product, dodecane, from the electrolysis of heptanoate at different anode materials (Table 23) illustrate this. The results in Table 22 were discussed in terms of the roughness (p. 6) of the surfaces of the different materials, and it was suggested that the real current density at a porous anode surface, TABLE 23 Yields (%) of Dodecane from Electrolysis of Heptanoate at Different Anodes and in Different Solvents'

Solvent/anode material

Graphite

Methanol

1

H2 0

2

DMF (very low c.d.) Neat acid

0 1

a Brennan and Brettle,

Vitreous carbon

Baked carbon

24-33 45-53 1-3.5 33

30

Platinum 52 45 66

1973.

such as graphite, would be much smaller than that at platinum. This would favour oxidation of radicals t o carbonium ions (monomolecular reaction) at the expense of coupling between radicals (bimolecular reaction). Note the analogy between this explanation and that offered for the effect of added perchlorate ion (p. 93). Knowing that, e.g. pyrolytic carbon has a roughness factor of about 20 (Epstein et al., 1971), it is obvious that we can have large differences in real current density between electrolysis at carbon and platinum anodes, and that such a difference can be responsible for a shift in product distribution merely due to the differences in concentration
115

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

bibenzyl (one-electron radical pathway) and toluene (predominantly formed via the two-electron carbanion pathway) in DMF as shown in Table 24. TABLE 24 Dependence of Product Proportional o n Electrode Material in the Cathodic Reduction of Benzyltriethylammonium Iona Cathode/product mol %

Toluene

Bibenzyl

A1 Mg Ta C( “spectroscopic”) Hg Pt

55 64

45

a

76 100 100 100

36 24 0 0 0

Ross et al., 1970a.

The high yield of bibenzyl at the aluminium cathode was later explained as being due t o the peculiar nature of the aluminium surface (Brown and Gonzalez, 1972). This is largely covered by an insulating oxide layer with a small proportion of breaks acting as active centres for the reduction process. Thus the real current density is high, permitting very high local concentrations of benLyl radicals. We are well aware that this discussion of the electrode matcrial and its role in electro-organic chemistry is severely limited, both with respect to the number of examples quoted and to the number of possible factors involved. However, we think that the mere fact that hardly any synthetic investigations dealing with the problem have adequately characterized the surfaces of the clectrode materials used gives us a good excuse for refraining from such a discussion. Let us first get the experimental facts on this point!

ACKNOWLEDGEMENTS We wish t o thank Drs. Vernon D. Parker, University of Copenhagen, Alvin Ronlin, University of Lund, Sidney D. Ross, Sprague Electric Company, and Norman L. Weinberg, Hooker, for their kindness in disclosing unpublished work and providing immensely helpful manuscripts of their books before publication.

116

L. EBERSON AND K. NYBERG

REFERENCES Adams, R. N. (1966). “Electrochemistry at Solid Electrodes.” Dekker, New York. Adams, R. N. (1969).Accounts Chem. Res. 2, 175. Allen, M. J. (1958). “Organic Electrode Processes.” Chapman & Hall, Ltd., London. Allendoerfer, R. D., and Rieger, P. H. (1965).J. A m e r . Chem. S O C . 87, 2336. Anderson, J. D., Baizer, M. M., and Prill, E. J. (1965). J. Org. Chem. 30, 1645. Anderson, J. D., Petrovich, J. P., and Baizer, M. M. (1969). Adv. Org. Chem. 6, 257. Anderson, L. B., and Paquette, L. A, (1972). J . Amer. Chem. S O C . 94, 4915. Anderson, J., and Eberson, L. (1974). Unpublished results. Andreades, S., and Zahnow, E. W. (1969).J. Amer. Chem. S O C . 91,4181. Andrulis, P. J., Dewar, M. J. S., Dietz, R., and Hunt, R. L. (1966). J. Amer. Chem. S O C . 88, 5473. Aniskova, L. V., Dubinin, A. G., Mirkind, L. A., and Fiochin, M. Ya. (1973).J. Org. Chem. USSR (Engl. Transl.) 9,428. Arapakos, P. G., and Scott, M. K. (1968). Tetrahedron Lett. 1975. Asahara, T., Seno, M., and Kaneko, H. (1968). Bull. Chem. SOC. Japan 41, 2985. Aten, A. C., and Hoijtink, G. J. (1959). Z. Phys. Chem. (Frankfurt) 21, 192. Aten, A. C., and Hoijtink, G. J. (1961). Adv. Polarography, 777. Avaca, A., and Bewick, A. (1972).J. Chem. Soc. Perkin IZ, 1709. Aylward, G. H., Garnett, J. L., and Sharp, J. H. (1967). Anal. Chem. 39,4571. Baggaley, A. J., and Brettle, R. (1968).J. Chem. S O C . (C) 2055. Baizer, M. M. (1964).J. Electrochem. S O C . 111, 215. Baizer, M. M. (1964a).J. Org. Chem. 29, 1670. Baizer, M. M. (1966).J. Org. Chem. 31. 3847. Baizer, M. M. (1969). Naturwissenschuften 56, 405. Baizer, M. M. (1973). “Organic Electrochemistry”, Chap. 19 (M. M. Baizer, ed.). Dekker, New York. Baizer, M. M., and Anderson, J. D. (1965).J. Org. Chem. 30, 1357. Baizer, M. M., and Chruma, J. L. (1972).J. Org. Chem. 37, 1951. Baizer, M. M., and Petrovich, J. P. (1970). Progr. Phys. Org. Chem. 7 , 189. Baizer, M. M., Chruma, J. L., and Berger, P. A. (1970).J. Org. Chcm. 35, 3569. Baizer, M. M. Pt al. (1973). “Organic Electrochemistry” (M. M. Bazier, ed.). Dekker, New York. Bank, S., and Noyd, D. A. (1973).J. Amer. Chem. SOC. 95, 8203. Bard, A. J., and Phelps, J. (1970).J . Electroanal. Chem. 25, App. 2. Bard, A. J., Ledwith, A., and Shine, H. J. (1976). A d v . Phys. Org. Chem. In press. Barradas, R. G., and Conway, B. E. (1961). Electrochim. A c t a 5 , 349. Basolo, F., and Pearson, R. G. (1967). “Mechanisms of Inorganic Reactions”, 2nd edn. Wiley, New York. Bauer, D., and Beck, J. P. (1971). Coll. Czech. Chem. C o m m . 36, 323. Bauer, D., Beck, J. P., and Texier, P. (1971). Coll. Czech. Chem. Comm. 36, 940. Bauer, D., and Foucault, A. (1972).J. Electroanal. Chem. 39, 385. Bauer, D., and Foucault, A. (1972a). J. Electroanal. Chem. 39, 377. Bechgaard, K., and Parker, V. D. (1972).J. A m e r . Chem. SOC.94,4749. Bechgaard, K., Hammerich, O., Moe, N. S., Ronlin, A., Svanholm, U., and Parker, V. D. (1972). Tetrahedron Lett. 2271.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

11 7

Bechgaard, K., Parker, V. D., and Pedersen, C. Th. (1973).J. Amer. Chem. SOC. 95,4373. Beck, F. (1968). Ber. Bunsenges. 72, 379. Beck, F. (1970). Chem. Zng. Techn. 42, 153. Beck, F. (1972). Angew. Elektrochem. 26, 215. Beck, F. (1972a). Angew. Chem. 84, 798. Beck, F. (1974). “Applied Electrochemistry”. XXIVth International Congress, 111, IUPAC. Butterworths, London. Belleau, B., and Au-young, Y. K. (1969). Can. J . Chem. 47, 2117. Benkeser, R. A., and Kaiser, E. M. (1963).J. Amer. Chem. SOC. 85, 2858. Benkeser, R. A., andTincher, C. A. (1968).J . Org. Chem. 33, 2727. Benkeser, R. A., Kaiser, E. M., and Lambert, R. F. (1964).J . A m e r . Chem. SOC. 86, 5272. Benkeser, R. A., Watanabe, H., Mels, S. J., and Sabol, M. A. (1970). J. Org. Chem. 35, 1210. Bernhardsson, E., Eberson, L., Nyberg, K., and Rietz, B. (1971). A c t a Chem. Scand. 25, 1224. Bertram, J., Fleischmann, M., and Pletcher, D. (1971). Tetrahedron Lett. 349. Bertram, J., Coleman, J. P., Fleischmann, M., and Pletcher, D. (1973). J. C. S. Perkin IZ 374. Bewick, A., and Cleghorn, H. P. (1973). J. C. S. Perkin ZZ 1410. Bewick, A., and Pletcher, D. (1970, 1971). “Specialist Periodical Reports of the Chemical Society-Electrochemistry”, (G. J. Hills, ed.). Bilevich, K. A., and Okhlobystin, 0. Yu. (1968). Russian Chem. Rev. 37, 1. Billon, J. P. (1960).J. Electroanal. Chem. 1, 486. Birch, A. J., and Subba Rao, G. (1972). Adv. Org. Chem. 8, 1. Bishop, N. I. (1971).Ann. Rev. Biochem. 40, 197. Blackburn, G. M., and Will, J. P. (1974). Chem. C o m m . 67. Blomgren, E., and Bockris, J. O’M. (1959).J. Phys. Chem. 63, 1475. Blomgren, E., Bockris, J. O’M., and Jesch, C. (1961). J. Phys. Chem. 65, 2000.

Blount, H. N. (1973).J. Electround Chem. 42, 271. Blount, H. N., and Kuwana, T. (1970).J. Electroanal. Chem. 27,464. Bobbitt, J. M. (1973). Heterocycles 1, 181. Bobbitt, J. M., Noguchi, I., Yagi, H., and Wiesgraber, K. H. (1971). J. Amer. Chem. SOC.93, 3551. Bockris, J. O’M., and Reddy, A. K. N. (1970). “Modem Electrochemistry”, Vol. 2. Plenum Press, New York. Bockris, J. O’M., Green, M., and Swinkels, D. A. J. (1964).J. Electrochem. SOC. 111,743. Bonner, W. A., and Mango, F. D. (1974).J. Org. Chem. 29,430. Bordi, S., and Papeschi, G. (1969).J. Electroanal. Chem. 20, 297. Bose, E. (1898). Z. Elektrochem. 5, 153. Brago, I. N. et a1 (1971). “Progress in Electrochemistry of Organic Compounds” (A. N. Frumkin and A. B. Ershler, eds.). Plenum Press, New York. Braun, W., Rajbenbach, L., and Eirich, F . R. (1962).J . Phys. Chem. 66, 1591. Breant, M., Bazoin, M., Buisson, J., Dupin, M., and Rebattu, J.-M. (1968). Bull. SOC. Chim. Fr. 5065. Brennan, M. P. J., and Brettle, R. (1973).J. C. S. Perkin 1 2 5 7 . Breslow, R., and Chu, W. (1973).J. Amer. Chem. SOC.95,411. Breslow, R.. and Mazur, Z. (1973).J. Amer. Chem. SOC. 95, 584. Brettle, R., and Sutton, J. R. (1974). Chem. C o m m . 449. Brieglieb, G. (1964). Angew. Chem. 76, 326.

118

L. EBERSON AND K. NYBERG

Broman, R., Heineman, W. R., and Kuwana, T. (1974). Disc. Faraday SOC. 5 6 , 1 6 . Brookes, C. J., Coe, P. L., Owen, D. M., Pedler, A. E., and Tatlow, J . C. (1974). Chem. C o m m . 323. Brouwer, D. M., and Hogeveen, H. ( 1 9 7 2 ) .Progr. Phys. Org. Chem. 9 , 179. Brown, 0. R., and Gonzalez, E. R. (1972). J . Electroanal. Chem. 35, 13. Brown, 0. R., and Harrison, J. A. ( 1 9 6 9 ) . J . Electroanal. Chem. 2 1 , 3 8 7 . Bunyan, P. J., a n d Hey, D. H. ( 1 9 6 2 ) . J . Chem. Soc. 324. Burdon, J., and Tatlow, J. C. (1960). Adv. Fluorine Chem. 1 , 129. Burdon, J., Parsons, I. W., and Tatlow, J. C. (1972). Tetrahedron 2 8 , 4 3 . Byrd, L., Miller, L. L., and Pletcher, D. ( 1 9 7 2 ) . Tetrahedron Lett. 2419. Casanova, J. (1974). Private communication. Casanova, J., and Eberson, L. ( 1 9 7 3 ) . “The Chemistry of the Carbon-Halogen Bond”, Part 1 (S. Patai, ed.). Wiley, London, 979. Casanova, J., and Rogers, H. R. (1974). J . Amer. Chem. SOC. 9 6 , 1942. Cauquis, G., and Maurey, M. (1968). Compt. Rend. Acad. Sci. Paris 266 (C), 1021. Cauquis, G., a n d Parker, V. D. (1973). “Organic Electrochemistry”, Chap. 2 (M. M. Baizer, ed.). Dekker, New York. Cauquis, G., and Pierre, G. ( 1 9 6 8 ) . Compt. R e n d , Acad. Sci. Paris 266 (C), 883. Cauquis, G., and Serve, D. (1966). Bull. SOC. Chim. Fr. 302. Cauquis, G., and Serve, D. ( 1 9 7 0 ) . Compt. Rend. Acad. Sci. Paris 271 (C), 324. Cauquis, G., and Serve, D. (1970a). J. Electroanal. Chem. 2 i , App. 3. Cedheim, L., and Eberson, L. ( 1 9 6 9 ) . Unpublished results. Cedheim, L., and Eberson, L. (1974). Unpublished results. Chambers, J. Q. (1974). “The Chemistry of t h e Quinonoid Compounds”, Chap. 1 4 (S. Patai, ed.). Wiley, London. Chaney, A., and Wolfrom, M. L. ( 1 9 6 l ) . J . Org. Chem. 2 6 , 2 9 9 8 . Chang, J., Large, R. F., a n d Popp, G. ( 1 9 7 1 ) . ‘‘Physical Methods of ChemistryPart IIb”, Chap. 1. (A. Weissberger, and B. W. Rossiter, eds.). Wiley, London. Childs, W. V., Maloy, J. T., Keszthelyi, C. R., and Bard, A. J. ( 1 9 7 1 ) . J. Electrochem. SOC. 118, 874. Chkir, M., and Lelandais, D. (1971). Chem. C o m m . 1369. Clark, D. B., Fleischmann, M., and Pletcher, D. (1972). J. Electroanal. Chem. 36, 137. Clark, D. B., Fleischmann, M., and Pletcher, D. (1973). J. C. S. Perkin ZI 1578. Clark, D. B., Fleischmann, M., and Pletcher, D. (1973a). J. Electroanal. Chem. 4 2 , 133. Clauson-Kaas, N., Limborg, F., and Glens, K. (1952). A c t a Chem. Scand. 6, 531. Clusius, K., a n d Schanzer, W. ( 1 9 4 3 ) . Z. Phys. Chem. (Leiprig) 192A, 273. Coetzee, J. F., Simon, J. M., and Bertozzi, R. J. (1969). Anal. Chem. 41, 766. Cohen L. A., and Farber, L. (1967). “Methods i n Enzymology”, Vol. 9 , 299 (C. H. W. Hirs, ed.). Academic Press, New York. Coleman, A. E., Richtol, H. H., and Aikens, D. A. (1968). J. Electroanal. Chem. 18, 165. Coleman, J. P., Utley, J. H. P., and Weedon, B. C. L. (1971). Chem. C o m m . 438. Coleman, L. B., Cohen, M. J., Sandman, D. J., Yamagishi, F. G., Garito, A. F., a n d Heeger, A. J. ( 1 9 7 3 ) . Solid State Commun. 12, 1125. Conway, B. E., and Barradas, R. G. (1961). Electrochim. A c t a 5 , 319. Conway, B. E., and Vijh, A. K. (1967). Chem. Rev. 67, 623. Conway, B. E., and Vijh, A. K. (1967a). Electrochim. A c t a 1 2 , 102.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

119

Conway, B. E., Loveland, W., and Neikam, W. C. (1967). Canad. Patent 766 741. Cook, C. D., and Depatie, C. B. (1959). J. Org. Chem. 24, 1144. Cooke, B. A., Ness, N. M., and Palluel, A. L. L. (1971). “Industrial Electrochemical Processes”, (A. T. Kuhn, ed.). Elsevier, Amsterdam. Corey, E. J., and Casanova, J. (1963).J. Amer. Chem. SOC. 8 5 , 165. Cottrell, P. T., and Mann, C. K. (1969).J. Electrochem. SOC.116, 1499. Courbis, P., and Guillemonat, A. (1966). Compt. R e n d . Acad. Sci. Paris 262 (C), 1435. Courtot-Coupez, J., and Le Dkmkzet, M. (1967). Bull. SOC. Chim. Fr. 4744. Covitz, F. H. (1967).J. A m e r . Chem. SOC.8 9 , 5403. Curphey, T. C., Amelotti, C. W., Layloff, T. P., McCartney, R. L., a n d Williams, J. H. (1969).J. Amer. Chem. SOC. 91, 2817. Dahms, H., and Green, M. (1963).J. Electrochem. SOC. 110, 1075. Damaskin, B. B., and Frumkin, A. N. (1971). “Reactions of Molecules a t Electrodes” (N. S. Hush, ed.). Wiley, London. Damaskin, B. B., Gerovich, V. M., Gladkikh, I. P., and Kaganovich, R. I. (1964). Russ. J . Phys. Chem. 38, 1351. Damaskin, B. B., Mishutushkina, I. P., Gerovich, V. M., a nd Kaganovich, R. I. (1964a).Russ,J. Phys. Chem. 38, 976. Damaskin, B. B., Andrusev, M. M., Gerovich, V. M., and Kaganovich, R. I. (1967). Soviet Electrochem. 3, 585. Damaskin, B. B., Petrii, 0. A., and Batrakov, V. V. (1971). “Adsorption of Organic Compounds.” Plenum Press, New York. Daniels, F. (1972). Biophys. J . 1 2 , 723. Danly, D. (1973). “Organic Electrochemistry” (M. M. Baizer, ed.). Dekker, New York. DeKlein, W. J. (1973). Electrochim. A c t a 1 8 , 4 1 3 . Dessau, R. M., Shih, S., and Heiba, E. I. (1970).J. A m e r . Chem. SOC.92, 412. DeTar, D. F. (1967).J. A m e r . Chem. SOC. 8 9 , 4 0 5 8 . Dickinson, T., and Wynne-Jones, W. F. K. (1962). Trans. Faraday SOC. 5 8 , 400. Dietz, R., and Larcombe, B. E. (1970).J. Chem. SOC. (B) 1369. Dietz, R., and Lund, H. (1973). “Organic Electrochemistry” (M. M. Baizer, ed.). Dekker, New York. Dietz, R., a n d Peover, M. E. (1968). Disc. Faraday SOC. 45, 154. Dirlam, J . P., and Eberson, L. (1972). Acta Chem. Scand. 26, 1454. Dirlam, J. P., Eberson, L., and Sternerup, H. (1972). Chem. Zng. Techn. 44, 178. Dorfman, L. M. (1970). Accounts Chem. Res. 3 , 224. Doughty, A., Fleischmann, M., and Pletcher, D. (1972). Quoted from: Fleischmann a nd Pletcher, 1973. Doyle, M. P., and Wierenga, W. (1972).J. Amer. Chem. SOC. 94, 3901. Dubois, J.-E., Lacaze, P.-C., and Ficquelmont, A. M. (1966). Compt. Rend. Acad. Sci. Paris 262 (C), 181 Eberson, L. (1963). Acta Chem. Scand. 1 7 , 2004. Eberson, L. (1967).J. A m e r . Chem. SOC.89, 4669. Eberson, L. (1967a). Electrochim. A c t a 1 2 , 1473. Eberson, L. (1968). “Chemistry of Carboxylic Acids and Esters”, Chap. 2 ( S . Patai, ed.). Wiley, London. Eberson, L. (1969).J. Amer. Chem. SOC.9 1 , 2402. Eberson, L. (1973). “Organic Electrochemistry”, Chap. 26 (M. M. Baizer, ed.). Dekker, New York. Eberson, L. (1973a). “Organic Electrochemistry”, Chap. 1 3 (M. M. Baizer, ed.). Dekker. New York.

120

L. EBERSON AND K. NYBERG

Eberson, L. (1974). “Applied Electrochemistry”, X X Z V t h International Congress, IUPAC/Butterworths, London, p. 137. Eberson, L., and Helgke, B. (1971). Unpublished results. Eberson, L., and Helgke, B. (1974). Chem. Scipta 5 , 47. Eberson, L., and Horner, L. (1973). “Organic Electrochemistry”, Chap. 27 (M. M. Baizer, ed.). Dekker, New York. Eberson, L., and Nilsson, S. (1968). Disc. Faraday Soc. 45, 242. Eberson, L., and Nilsson, S. (1968a).A c t a Chem. Scand. 22, 2453. Eberson, L., and Nyberg, K. (1964). Acta Chem. Scand. 18, 1568. Eberson, L., and Nyberg, K. (1964a). A c t a Chem. Scand. 18, 1567. Eberson, L., and Nyberg, K. ( 1 9 6 6 ) . J .A m e r . Chem. SOC.88, 1686. Eberson, L., and Nyberg, K. (1966a). Tetrahedron Lett. 2389. Eberson, L., and Nyberg, K. (1973). Accounts Chem. Res. 6 , 106. Eberson, L., and Nyberg, K. (1996). “Encyclopedia of the Electrochemistry of the Elements (Organic)” (A. J. Bard, and H. Lund, eds.). Dekker, New York, in press. Eberson, L., and Olofsson, B. (1969). A c t a Chem. Scand. 23, 2355. Eberson, L., and Parker, V. D. (1970). A c t a Chem. Scand. 2 4 , 3 5 5 3 . Eberson, L., and Ryde-Petterson, G..( 1973). A c t a Chem. Scand. 27, 1155. Eberson, L., and SchSfer, H. (1971). Fortschr. Chem. Forsch. 21, 1 . Eberson, L., and Sternerup, H. (1972). Acta Chem. Scand. 26, 1431. Eberson, L., and Wilkinson, R. G. (1972). A c t a Chem. Scand. 26, 1671. Eberson, L., Nyberg, K., lstein, M., Petersen, R. C., Ross, S. D., and Uebel, J. J. (1967).J. Org. C Eberson, L., Granse, S sson, B. (1968). A c t u Chem. Scand. 22, 2462. Eberson, L., Nilsson, S., and Rietz, B. (1972). A c t a Chem. Scand. 26, 3870. Eberson, L., Nyberg, K., and S t e e e r u p , H. (1973). Chem. Scripta 3 , 12. Eberson, L., Nyberg, K., and Ste ). Acta Chem. Scand. 27, 1679. Edwards, J. 0. (1954). J . Amer. Ch Elving, P. J., and Pullman, B. (1961).Adu. Phys. Chem. 3 , 1 . Epstein, B. D., Dalle-Molle, E., and Mattson, J. S. (1971). Carbon 9 , 609. Erickson, R. E., and Fischer, C. M. (1970).J. Org. Chem. 35; 1604. Faita, G., Fleischmann, M., and Pletcher, D. (1970). J. Electroanal. Chem. 25, 455. Falck, J . R., Miller, L. L., and Stermitz, F . k (1974). Tetrahedron 30, 931. Feldberg, S. W. (1969). “Electroanalytical Chemistry”, Vol. 3 (A. J. Bard, ed.). Dekker, New York. Ferraris, J., Cowan, D. O., Walatka, V., and Perlstein, J. H. (1973). J. Amer. Chem. SOC.9 5 , 9 4 8 . Fichter, F. (1942). “Organische Elektrochemie.” Steinkopf, Dresden. Fiochin, M. Ya., Kamneva, A. I., Mirkind, L. A., Salmin, L. A., and Kornienko, A. G. (1963). Chem. Abstr. 5 8 , 11205. Fiochin, M. Ya., Mirkind, L. A., and Zhurinov, M. Zh. (1973). Russian Chem. Rev. 42, 293. Fischer, C. M., and Erickson, R. E. ( 1 9 7 3 ) . J . Org. Chem. 3 8 , 4 2 3 6 . Fleischmann, M. ( 1 974). “Applied Electrochemistry”, X X Z V t h Jnternational Congress, IUPAC/Butterworths, London, p.1. Fleischmann, M., and Pletcher, D. (1968). Tetrahedron Lett. 6255. Fleischmann, M., and Pletcher, D. (1969a). R.Z.C. Rev. 2, 87. Fleischmann, M., and Pletcher, D. (196913). Platinum Metals Rev. 1 3 , 4 6 . Fleischmann, M., and Pletcher, D. (1970). J. Electroanal. Chem. 25, 449.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

12 1

Fleischmann, M., and Pletcher, D. (1971). “Reactions of Molecules a t Electrodes” (N. S. Hush, ed.). Wiley, London. Fleischmann, M., and Pletcher, D. (1973). Adu. Phys. Org. Chem. 10, 155. Fleischmann, M., Mansfield, J. R., and Wynne-Jones, W. F. K. (1965). J. Electrounul. Chem. 10, 511. Fleischmann, M., Mansfield, J. R., and Wynne-Jones, W. F. K. (1965a). J. Electrounul. Chem. 10, 521. Fleischmann, M., Pletcher, D., and Race, G. M. (1969). J. Electrounul. Chem. 23, 369. Fleischmann, M., Pletcher, D., and Race, G. M. (1970). J. Chem. SOC.(B) 1746. Fleischmann, M., Pletcher, D., and Vance, C. J. (1971a). J . ElectrounuZ. Chem. 29, 325. Fleischmann, M. and sundry authors who prefer t o remain anonymous (1971b). Euchem Meeting, Ronneby Brunn, Sweden. Formaro, L., Firoi, G., and Colombi, M. ( 1 9 7 3 ) .J . Electrounul. Chem. 44, 255. Fritsch, J. M., Weingarten, H., and Wilson, J. D. ( 1 9 7 0 ) . J . Amer. Chem. SOC. 9 2 , 4038. Fry, A. J. (1972). “Synthetic Organic Electrochemistry.” Harper & Row, New York. Fry, A. J. (1972a). Fortschr. Chem. Forsch. 34, 1. Fry, A. J., and Britton, W. E. ( 1 9 7 3 ) . J . Org. Chem. 3 8 , 4 0 1 6 . Fry, A. J., and Reed, R. G. (1971).J. Amer. Chem. SOC. 9 3 , 553. Fry, A. J., and Reed, R. G. (1972). J. Amer. Chem. SOC.9 4 , 8475. Fry, A. J., Chung, L. L., and Boekelheide, V. (1974). Tetrahedron Lett. 445. Garst, J. F., and Barton, F. E. ( 1 9 7 4 ) . J . Amer. Chem. SOC. 9 6 , 523. Gassman, P. G., and Zalar, F. V. ( 1 9 6 6 ) . J . Amer. Chem. SOC. 88, 2252. Geiger, W. E. ( 1 9 7 3 ) . J . Phys. Chem. 77, 1862. Gerdil, R. (1970). Helu. Chim. A ct u 53, 2100. Geske, D. H. (1959). J. Electround. Chem. 1, 502. Gilde, H.-G. (1972). “Methods in Free-Radical Chemistry” Vol. 3 , Chap. 1 (E. S. Huyser, ed.). Dekker, New York. Gileadi, E. e t ul. (1967). “Electrosorption” (E. Gileadi, ed.). Plenum Press, New York. Given, P. H., and Peover, M. E. ( 1 9 6 0 ) . J . Chem. SOC.385. Glass, G. E., and West, R. (1972). Inorg. Chem. 11, 2847. Glass, R. S., Britt, W. J., Miller, W. N., and Wilson, G. S. ( 1 9 7 3 ) .J . Amer. Chem. SOC.95, 2375. Glasstone, S., and Hickling, A. ( 1 9 3 4 ) . J . Chem. SOC.1878. Gleicher, G. J., and Gleicher, M. K. (1967). J. Phys. Chem. 71, 3693. Goldschmidt, S., Leicher, W., and Haas, H. (1952). Ann. 577, 153. Gonikberg, M. G., and Gavrilova, A. E. ( 1 9 5 2 ) . J. Gen. Chem. USSR 22, 1384. Goodridge, F. (1974). “Applied Electrochemistry”, X X I V t h International Congress, IUPAC/Butterworths, London. Gourley, R. N., Grimshaw, J., and Millar, P. G. (1967). Chem. Comm. 1278. Gourley, R. N., Grimshaw, J., and Millar, P. G. ( 1 9 7 0 ) . J . Chem. SOC.(C) 2318. Griller, D., and Roberts, B. P. (1971). Chem. Comm. 1035. Gruver, G. A., and Kuwana, T. ( 1 9 7 2 ) .J. Electrounul. Chem. 36, 85. Gutmann, F., and Lyons, E. (1967). “Organic Semiconductors”. Wiley, New York. Hale, J. M. (1971). “Reaction of Molecules at Electrodes” (N. S. Hush, ed.). Wiley, London. Hammerich, O., and Parker, V. D. ( 1 9 7 2 ) . J . Electrounul. Chem. 3 6 , App. 11. Hammerich, O., and Parker, V. D. (1973). Electrochim. Ac tu 18, 537.

122

L. EBERSON AND K. NYBERG

Hammerich, O., and Parker, V. D. ( 1 9 7 4 ) . J . Amer. Chem. SOC.9 6 , 4 2 8 9 . Hammerich, O., and Parker, V. D. (1974a). Chem. Comm. 245. Hammerich, O., Moe, N. S., and Parker, V. D. (1972). Chem. Comm. 156. Harris, L. B. (1973). J. Electrochem. SOC.120, 1034. Hart, E. J., and Anbar, M. (1970). “The Hydrated Electron”. Wiley, New York. Haufe,J., and Beck, F. (1970). Chem. Zng. Techn. 4 2 , 1 7 0 . Hawkes, G. E., Utley, J. H. P., and Yates, G. B. (1973). Chem. Comm. 305. Headridge, J. B., Pletcher, D., and Callingham, M. (1967). J. Chem. SOC. (A) 684. Heiland, W., Gileadi, E., and Bockris, J. O’M. (1966). J. Phys. Chem. 70, 1207. Hercules, D. M. (1971). “Physical Methods of Chemistry-Part I1 B” (A. Weissberger and B. W. Rossiter, eds.). Wiley, New York, p. 257. Hoggett, J. G., Moodie, R. B., Penton, J. R., and Schofield, K. (1971). “Nitration and Aromatic Reactivity”. Cambridge University Press. Hoijtink, G. J. ( 1 9 5 4 ) .Rec. Trav. Chirn. 73, 895. Hoijtink, G. J. (1955). Rec. Trav. Chirn. 74, 1525. Hoijtink, G. J . (1957). Rec. Trav. Chim. 76, 885. Hoijtink, G. J. (1958). Rec. Trav. Chim. 77, 555. Hoijtink, G. J. (1963). Chem. Zng. Techn. 35, 333. Hoijtink, G. J. (1970). “Advances in Electrochemistry and Electrochemical Engineering”, Vol. 7 (P. Delahay and C. W. Tobias, eds.). Interscience, New York. Hoijtink, G. J., van Schooten, J., De Boer, E., and Aalbersberg, W. I. (1954). Rec. Trav. Chim. 73, 355. Holy, N. L. ( 1 9 7 4 ) . Chem. Rev. 74, 243. Homer, L., and Degner, D. (1971). Tetrahedron Lett. 1245. Homer, L., and Fuchs, H. (1962). Tetrahedron Lett. 203. Homer, L., and Lund, H. (1973). “Organic Electrochemistry”, Chap. 21 (M. M. Baizer, ed.). Dekker, New York. Horner, L., a n d Neumann, H. ( 1 9 6 5 ) . Chem. Ber. 9 8 , 1 7 1 5. Homer, L., a n d Roder, H. ( 1 9 6 9 ) .Ann. 732, 11. Horner, L., Winkler, H., Rapp, A., Mentrup, A., Hoffmann, H., and Beck, P. (1971). Tetrahedron Lett. 161. Homer, L., Degner, D., and Skaletz, D. (1972). Chem. Zng. Techn. 44, 209. House, H. O., Feng, E., and Peet, N. P. (1971). J. Org. Chem. 36, 2371. Huebert, B. J., and Smith, D. E. (1971).J. Electroanal. Chem. 31, 333. Humffray, A. A. (1973). “Modern Aspects of Electrochemistry”, Vol. 8, 121 0. O’M. Bockris and B. E. Conway, eds.). Butterworths, London. Ikenoya, S., Masui, M., Ohmori, H., and Sayo, H. (1974). J. Chem. SOC.Perkin ZI, 571. Inoue, T., Koyama, K., Matsuoka, T., and Tsutsumi, S. (1967). Bull. Chem. SOC. Japan 40, 162. Isoe, S., and Cohen, L. A. (1968). Arch. Biochem. Biophys. 127, 522. Isserlis, G. (1971). “Industrial Electrochemical Processes”, Chap. 9 (A. T. Kuhn, ed.). Elsevier, Amsterdam. Iversen, P. (1971). Tetrahedron Lett. 55 Iversen, P., and Lund, H. (1969). Tetrahedron Lett. 3523. Jubault, M., Raoult, E., and Peltier, D. (1973). Compt. Rend. Acad. Sci. Paris 277 (C), 583. Kadyrov, M.Kh., Goiuber, A. I., Koshechkina, L. I., and Shekhvatov, M. S. (1971). Soviet Electrochem. 7, 85. Kaganovich, R. I., Damaskin, B. B., and Kaisheva, M. K. (1970). Soviet Electrochem. 6 , 1313.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

123

Kariv, E., Terni, H. A., and Gileadi, E. (1973).J. Electrochem. SOC. 120, 639. Kariv, E., Terni, H. A., and Gileadi, E. (1973a). Electrochim. A c t a 1 8 , 433. Katz, M., Saygin, O., a n d Wendt, H. (1974). Electrochim. A c t a 1 9 , 193. Keating, J. T., and Skell, P. S. (1969).J. Amer. Chem. SOC. 9 1 , 695. Kim, K. S., Winograd, N., and Davis, R. E. (1971). J . A m e r . Chem. SOC. 93, 6296. Koch, V. R., and Miller, L. L. (1973).J. Amer. Chem. SOC. 9 5 , 8631. Koch, V. R., Miller, L. L., Clark, D. B., Fleischmann, M., Joslin, T., a n d Pletcher, D. (1973)../. Electroanal. Chem. 43, 318. Kochi, J . K.,-Tang, R. T., and Bernath, T . (1973). J. Amer. Chem. SOC. 9 5 , 71 14. Koehl, W. J. (1964).J. Amer. Chem. SOC. 8 6 , 4 6 8 6 . Koehl, W. J. (1967).J. Org. Chem. 32, 614. Kollmar, H., and Smith, H. 0. (1970). Chem. Phys. Lett. 5 , 7. Kolthoff, I. M., and Coetzee, J. F. (1956).J. Amer. Chem. SOC. 79, 1852. Kornprobst, J.-M., Laurent, A., a n d Laurent-Dieuzeide, El. (1968). Bull. SOC. Chim. E’r. 3657. Kornprobst, J.-M., Laurent, A., and Laurent-Dieuzeide, E. (1970). Bull. SOC. Chim. Fr. 1490. Kovacic, P. (1963).J. Amer. Chem. SOC. 8 5 , 454. Koyama, K., Susuki, T., and Tsutsumi, S. (1965). Tetrahedron Lett. 627. Koyama, K. Susuki, T., and Tsutsumi, S. (1966). Tetrahedron 23, 2675. Koyama, K., Susuki, T., and Tsutsumi, S. (1968). Bull. Chem. Soc. Japan 4 1 , 2668. Kunugi, A., Shimuzi, T., and Nagaura, S. (1970). Bull. Chem. SOC. Japan 43, 1298. Kuwana, T. (1966). “Electroanalytical Chemistry”, Vol. 1 , 1 9 7 ( A . J . Bard, ed.). Dekker, New York. Kuwana, T. (1973). Ber. Bunsenges. 77, 858. Kuwata, K., and Geske, D. H. (1964).J. Amer. Chem. Soc. 8 6 , 2101. Lambert, F . L., and Kobayashi, K. (1960).J. Amer. Chem. SOC. 8 2 , 5324. Lamm, B., and Samuelsson, B. (1970). Acta Chem. Scand. 24, 561. Lamm, B., and Samuelsson, R. (1970a). Chem. C o m m . 1010. Lamm, B., and Simonet, J. (1974). A c t a Chem. Scand. (B) 28, 147. Lamy, E., Nadjo, L., and Saveant, J. M. (1973). J. Elrctroanab Chem. 42, 189. Laurent, A., Laurent, E., and Tardivel, R. (1973). Tetrahedron Lett. 4861. Ledwith A., and Russell, P. J . (1974). J . C. S. Perkin ZZ 582. Lehmkuhl, H. (1973). Synthesis 3 7 7 . Lehmkuhl, H. (1973a). “Organic Electrochemistry” (M. M. Baizer, ed.). Dekker, New York. Letsinger, R. L., andMcCain, J. H. (1966).J. Amer. Chem. SOC. 8 8 , 2884. Leung, M., Herz, J., and Salzberg, H. W. (1965). J . Org. Chem. 30, 310. Levy, M., Steinberg, M., and Szwarc, M. (1954).J. Amer. Chem. SOC.76, 5978. Lindsey, R. V., and Peterson, M. L. (1959). J . Amer. Chem. SOC.81, 2073. Linstead, R. P., Lunt, J. C., Weedon, B. C. L., and Shephard, B. K. (1952). J . Chem. SOC. 3621. Lipkin, D., Galiano, F. R., and Jordan, K. W. (1963). Chem. Znd. (London) 1657. Little, W. A. (1964).Phys. Rev. 134, A1416. Little, W. A. (1967).J. Polym. Sci. (C), 17, 3. Longchamp, S., Caullet, C., and Libert, M. (1974). Bull. SOC. Chim. Fr. 353. Ludman, C. J., McCarrou, E. M., and O’Malley, R. E’. ( 1 9 7 2 ) . J . Electrochem. SOC. 119, 874.

124

L. EBERSON AND K. NYBERG

Lund, H. (1957). Acta Chem. Scand. 11, 1323. Lund, H. (1967). Osterreich. Chem.-Z. 6 8 , 4 4 , 152. Lund, H. (1970). “Chemistry of the Carbon-Nitrogen Double Bond” (S. Patai, ed.). Wiley, London. Lund, H. (1970a). Adv. Heterocyclic Chem. 12, 213. Lund, H. (1971). “Chemistry of the Hydroxyl Group”, Chap. 5 (S. Patai, ed.). Wiley, London. Lund, H. (1973). “Organic Electrochemistry”, Chap. 25 (M. M. Baizer, ed.). Dekker, New York. Lund, H. (1973a). “Organic Electrochemistry”, Chap. 23 (M. M. Baizer, ed.). Dekker, New York. Lund, H., and Iversen, P. (1973). “Organic Electrochemistry”, Chap. 4 (M. M. Baizer, ed.). Dekker, New York. Lund, H., Michel, M.-A., and Simonet, J. (1974). Abstracts, 1 5 t h Scandinavian Chemistry Meeting, Tammerfors, June. Magnusson, C., Olofsson, B., and Nyberg, K. (1971). Chem. Scriptu 1 , 57. Mango, F. D., and Bonner, W. A. (1964).J. Org. Chem. 29, 1367. Mann, C. K. (1969). “Electroanalytical Chemistry”, Vol. 3, 58 (A. J. Bard, ed.). Dekker, New York. Mann, C. K., and Barnes, K. K. (1970). “Electrochemical Reactions in Nonaqueous Systems”. Dekker, New York. Manning, G., Parker, V. D., and Adams, R. N. (1969). J. Amer. Chem. SOC.91, 4584. Manousek, O., and Zuman, P. (1965). Chem. Comm. 158. Marcoux, L. S. (1971).J . Amer. Chem. SOC.93, 537. Marcoux, L. S. (1972).J. Phys. Chem. 76, 3254. Marcoux, L. S., Fritsch, J. M., and Adams, R. N. (1967). J. Amer. Chem. SOC. 89, 5766. Mark, H. B. (1968). Record Chem. Progr. 29, 217. Masui, M., and Ohmori, H. (1973). J. C. S. Perkin ZZ 1112. Matsuda, Y., Kimura, K., Iwakura, C., and Tamura, H. (1973). Bull. Chem. SOC. Japan 46,430. Matthews, D. B., and Bockris, J. O’M. (1971). “Modern Aspects of Electrochemistry”, 242 (J. O’M. Bockris, and B. E. Conway, eds). Butterworths, London. Mayeda, E. A., and Miller, L. L. (1972). Tetrahedron 28, 3375. McCapra, F. (1973).Progr. Org. Chem. 8 , 231. McDowell, C. S. (1967). Dissertation, Indiana University, Bloomington, USA. McKillop, T. F. W., and Korinek, K. (1971). Ann. Reports 68B, 297. McKillop, T. F. W., and Korinek, K. (1972). Ann. Reports 69B, 291. Meier, E. P., and Chambers, J. Q. (1969) Anal. Chem. 41,914. Miller, L. L. (1968). Tetrahedron Lett. 1831. Miller, L. L. (1971).J . Chem. Educ. 48, 168. Miller, L. L., and Hoffmann, A. K. (1967).J. Amer. Chem. SOC.89, 593. Miller, L. L., andMayeda, E. A. (1970). J. Amer. Chem. SOC.92, 5818. Miller, L. L., and Ramachandran, V. (1974). J. Org. Chem. 39, 369. Miller, L. L., Kujawa, E. P., and Campbell, C. B. (1970). J. Amer. Chem. SOC. 92, 2821. Miller, L. L., Nordblom, G. D., and Mayeda, E. A. (1972). J. Org. Chem. 37, 916. Millington, J. P. (1969).J. Chem. SOC.(B), 982. Misono, A., Osa, T., and Yamagishi, T. (1968). Bull. Chem. SOC.Japan 41, 2921. Muck, D. L., and Wilson, E. R. (1970). J . Electrochem. SOC.117, 1359.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

125

Murata, Y., and Shine, H. J. (1969).J. Org. Chem. 34, 3368. Nagase, S. (1967). Fluorine Chem. Rev. 1, 77. Neikam, W. C., andDesmond, M. M. (1964). J . A m e r . Chem. SOC.88,4811. Nelson, R. F., and Adams, R. N. (1967).J. Electroanal. Chem. 13, 184. Nicholson, R. S., and Shain, I. (1964). Anal. Chem. 36, 706. Nicholson, R. S., and Shain, I. (1965). Anal. Chem. 37, 178. Nilsson, A., Ronlin, A., and Parker, V. D. (1973). J. C. S. Perkin I 2 3 3 7 . Nilsson, S . (1971). Dissertation, University of Lund, Sweden. Nilsson, S. (1973). A c t a Chem. Scand. 27, 329. Norman, R. 0. C., Thomas, C. B., and Willson, J. S. (1971). J. Chem. SOC. (B) 518. Norman, R. 0. C., Thomas, C. B., and Ward, P. J. (1973). J. C. S. Perkin I 2914. Nyberg, K. (1969). Chem. C o m m . 774. Nyberg, K. (1970). A c t a Chem. Scand. 24, 1609. Nyberg, K. (1970a). Acta Chem. Scand. 24, 473. Nyberg, K. (1971). A c t a Chem. Scand. 25, 534. Nyberg, K. (1971a).A c t a Chem. Scand. 25, 2499. Nyberg, K. (1971b).A c t a Chem. Scand. 25, 2938. Nyberg, K. ( 1 9 7 1 ~ )A. c t a Chem. Scand. 25, 3770. Nyberg, K. (1971d).A c t a Chem. Scand. 25, 3246. Nyberg, K. (1971e). Dissertation, University of Lund, Sweden. Nyberg, K. (1971f). Unpublished results. Nyberg, K. (1973). “Organic Electrochemistry” (M. M. Baizer, ed.). Dekker, New York. Nyberg, K. (1973a). A c t a Chem. Scand. 27, 503. Nyberg, K. (1974a). Chem. Scripta 5 , 115. Nyberg, K. (197413). Chem. Scripta 5 , 120. Nyberg, K., and Trojanek, A. (1975). Coll. Czech. Chem. C o m m . 40, 526. O’Connor, J. J., and Pearl, J. A. (1964). J. Electrochem. SOC. 111,335. O’Donnell, J. F. (1965). Quoted from Mann (1969) p. 65. O’Malley, R. F. (1973). Private communication. Osa, T. (1968). Bull. Chem. SOC.Japan 41, 69. Osa, T. (1968a). J. Electrochem. SOC. 115 226. Osa, T., Yildiz, A., and Kuwana, T. (1969).J. Amer. Chem. SOC.91, 3994. Oth, J. F. M., Baumann, H., Gilles, J.-M., and Schroder, G. (1972).J. A m e r . Chem. SOC.94,3498. Ovchinnikova, T. M., Taran, L. A., and Rotinyan, A. L. (1962). Russ. J . Phys. Chem. 36, 1031. PCkinyi, C., and Zahradnik, R. (1965). Coll. Czech. Chem. C o m m . 30,4287. Parker, V. D. (1969). Chem. C o m m . 610. Parker, V. D. (1969a).J. Electroanal. Chem. 21, App. 1. Parker, V. D. (1969b). Chem. C o m m . 848. Parker, V. D. (1970a). A c t a Chem. Scand. 24, 3455. Parker, V. D. (L970b). A c t a Chem. Scand. 24, 3151. Parker, V. D. ( 1 9 7 0 ~ )Acta . Chem. Scand. 24, 2757. Parker, V. D. (1970d). A c t a Chem. Scand. 24, 2775. Parker, V. D. (1970e). Dissertation, University of Lund, Sweden. Parker, V. D. (1972). J. Electroanal. Chem. 36, App. 8. Parker, V. D. (1973). Electrochim. A c t a 18, 519. Parker, V. D., and Burgert, B. E. (1965). Tetrahedron Lett. 4065. Parker, V. D., and Burgert, B. E. (1968). Tetrahedron Lett. 2411. Parker, V. D., and Eberson, L. (1969). Tetrahedron Lett. 2843.

126

L. EBERSON AND K. NYBERG

Parker, V. D., and Eberson, L. (1969a). Tetrahedron Lett. 2839. Parker, V. D., and Eberson, L. (196913). Chem. C o m m . 340. Parker, V. D., and Eberson, L. (1970). Acta Chem. Scund. 23, 3542. Parker, V. D., and Eberson, L. (1970b).J. Amer. Chem. SOC. 92, 7488. Parker, V. D., and Eberson, L. (1972). Chem. C o m m . 441. Parker, V. D., Nyberg, K., and Eberson, L. (1969). J . Electroanal. Chem. 22, 150. Parker, V. D., Dirlam, J. P., and Eberson, L. (1971). A c t a Chem. Scund. 25, 341. Partridge, L. K., Tansley, A. C., and Porter, A. S. (1966). Electrochim. Acta 11, 51 7. Payne, R. (1970). “Advances in Electrochemistry and Chemical Engineering”, Vol. 7, 1 (P. Delahay, ed.). Interscience, New York. Peover, M. E. (1967). “Electroanalytical Chemistry”, Vol. 2, 1 (A. J. Bard, ed.). Dekker, New York. Peover, M. E. (1971). “Reactions of Molecules at Electrodes”, 259 (N. S. Hush, ed.). Wiley, London. Peover, M. E., and Powell, J. S. (1969).J. Electroanal. Chem. 20,427. Peover, M. E., and White, B. S. (1967). J . Electroanal. Chem. 13, 93. Peridon, J., and Buvet, R. (1968). Bull. Soc. Chim. Fr. 1279. Petit, G., and Bessikre, J. (1971). J. Electroanal. Chem. 31, 393. Petrovich, J. P., Baizer, M. M., and Anderson, J. D. (1966). J. Org. Chem. 31, 3 890. Petrovich, J . P., Anderson, J. D., and Baizer, M. M. (1966a). J. Org. Chem. 31, 3897. Petrovich, J. P., Baizer, M. M., and Ort, M. R. (1969).J. Electrochem. SOC. 116, 749. Petrovich, J. P., Baizer, M. M., and Ort, M. R. (1969a).J. Electrochem. SOC.116, 743. Piersma, B . J., and Gileadi, E. (1966). “Modern Aspects of Electrochemistry”, Vol. 4, Chap. 2 0 . O’M. Bockris, ed.). Plenum, New York. Pointeau, R. (1962). Ann. Chim. (Paris) 7, 669. Popp, F. D., and Schultz, H. P. (1962). Chem. Rev. 62, 19. Popp, G., and Reitz, N. C. (1972).J. Org. Chern. 3 7 , 3646. Pryor, W. A., Morkved, E. H., and Bickley, H. T. (1972).J. Org. Chem. 37, 1999. Puglisi, V. J., and Bard, A . J . (1972).J . Electrochem. SOC. 119, 829. Puglisi, V. J., Clapper, G. L., and Evans, D. H. (1969). Anal. Chem. 41, 279. Pysh, E. S., and Yang, N. C. (1963).J. A m e r . Chem. Soc. 85, 2124. Rabinowitch, E., and Govindjee (1969). “Photosynthesis”. Wiley, New York. Radlick, P., Klen, R., Spurlock, S., Sims, J. J., Van Tamelen, E. E., and Whitesides, T. (1968). Tetrahedron Lett. 5117. Rao, R. R., Mulliken, S. B., Robinson, S. L., and Mann, C. K. (1970). Anal. Chem. 42,1076. Rawlinson, D. J., and Sosnovsky, G. (1972). Synthesis, 1. Rawlinson, D. J., and Sosnovsky, G. (1973). Synthesis, 567. Reichenbacher, P. H., Liu, M. Y., and Skell, P. S. (1968). /. A m e r . Chem. SOC. 90, 1816. Reichenbacher, P. H., Morris, M. D., and Skell, P. S. (1968a). J. Amer. Chem. SOC. 90, 3432. Reynolds, W. L., and Lumry, R. W. (1966). “Mechanisms of Electron Transfer”. Ronald Press Co., New York. Rieger, P. H., Bernal, I., Reinmuth, W. H., and Fraenkel, G. K. (1963).J. Amer. Chem. SOC. 85,683. Rieke, R. D., and Hudnall, P. M. (1973).J. Amer. Chem. SOC. 95, 2646.

STRUCTURE AND MECHANISM IN ORGANIC ELECTROCHEMISTRY

127

Rieke, R. D., Rich, W. E., and Ridgway, T. H. (1971).J. Amer. Chem. SOC. 93, 1962. Rifi, M. R. (1967). J. Amer. Chem. SOC.89, 4442. Rifi, M. R. (1969). Tetrahedron Lett. 1043. Rifi, M. R., and Covitz, F. H. (1974). “Introduction t o Organic Electrochemistry”. Dekker, New York. Robertson, P. M. (1973, 1974). “Specialist Periodical Reports of the Chemical Society-Electrochemistry.” Ronlin, A., and Parker, V. D. (1974). Chem. Commun. 33. Ronlin, A., and Parker, V. D. (1974a). J . Org. Chem. 39, 1014. Ronlin, A., Hammerich, O., and Parker, V. D. (1973).J. Amer. Chem. Soc. 95, 7132. Ronlin, A., Bechgaard, K., and Parker, V. D. (1973a). Acta Chem. Scand. 27, 2375. Ronlbn, A., Coleman, J., Hammerich, O., an d Parker, V. D. (1974). J. Amer. Chem. SOC. 96, 845. Ross, S . D., and Finkelstein, M. (1969). J. Org. Chem. 34, 2923. Ross, S. D., Finkelstein, M., and Petersen, R. C. (1964).J. Amer. Chem. Soc. 86, 4139. Ross, S. D., Finkelstein, M., and Petersen, R. C. (1964a). J . Amer. Chem. Soc. 8 6 , 2745. Ross, S . D., Finkelstein, M., and Petersen, R. C. (1966).J. Org. Chem. 31, 128. Ross, S. D., Finkelstein, M., and Petersen, R. C. (1966a). J . Amer. Chem. SOC. 88, 4657. Ross, S . D., Finkelstein, M., and Petersen, R. C. (1967).J. Amer. Chem. SOC. 89, 4088. Ross, S . D., Finkelstein, M., and Uebel, J. J. (1969).J. Org. Chem. 34, 1018. Ross, S. D., Finkelstein, M., and Petersen, R. C. (1970).J. Org. Chcm. 35, 781. Ross, S. D., Finkelstein, M., and Petersen, R. C. (1970a). J. Amer. Chem. Soc. 92, 6003. Ross, S. D., Finkelstein, M., and Rudd, E. J. (1972).J. Org. Chem. 37, 2387. Rousseau,K., Farrington, G. C., and Dolphin, D. (1972). J. Org. Chem. 37, 3968. Russell, G. A., and Norris, R. K. (1973). “Organic Reactive Intermediates” (McManus, S. P., ed.), Academic Press, New York and London. Sainsbury, M. (1971).J. Chem. SOC. (C), 2888. Santhanam, K. S. V., and Bard, A. J. (1966).J. Amer. Chem. SOC.88, 2669. Sato, N., Sekine, T., and Sugino, K. (1968).J. Electrochem. SOC.115, 242. Schafer, H. (1969). Chem. Zng. Techn. 41, 179. Schafer, H. (1970). Chem. Zng. Techn. 42, 164. Schafer, H. (1970a).Angew. Chem. 82, 134. Schafer, H. (1974). Private communication. Schafer, H., and Arzak, A. A. (1972). Chem. Ber. 105, 2398. S c h s e r , H., and Kunzel, H. (1970). Tetrahedron Lett. 3333. Schafer, H., and Pistorius, R. (1972). Angew. Chem. 84, 893. Schafer, H., and Steckhan, E. (1969). Angew. Chem. 81, 532. Schubert, C. C., Page, C. L., an d Ralph, B. (1973). Electrochim. Actu 18, 33. Sease, J. W., Reed, R. G., and Fry, A. J. (1969). Quoted from Brown and Harrison (1969). Semenov, N. N. (1958). “Some Problems of Chemical Kinetics and Reactivity”, Vol. 1, 198. Princeton University Press, Princeton, N.J. Sheldon, R. A., and Kochi, J. K. (1973). Oxidation and Combustion Rev. 5, 135. Shine, H. J., and Murata, Y. (1969). 1. Amer. Chem. SOC. 9 1 , 1872. Shono, T., and Ikeda, A. (1972). J . Amer. Chem. SOC. 9 4 , 7892.

128

L. EBERSON AND K. NYBERG

Shono, T., and Kosaka, T. (1968). Tetrahedron Lett. 6207. Shono, T., Ikeda, A., and Hakozaki, S. (1972). Tetrahedron Lett. 4511. Shukla, S . N., and Walker, 0.J. (1931). Trans. Faraday SOC.27, 35. Sioda, R. E. (1968).J. Phys. Chem. 72, 2322. Skell, P. S., and Reichenbacher, P. H. (1968).J. Amer. Chem. SOC. 90, 2309. Skell, P. S., and Reichenbacher, P. H. (1968a).J. Amer. Chem. S O C . 90, 3436. Smirnov, V. A., and Markova, A. V. (1970). Soviet Electrochem. 6,1767. Smith, W. B., and Gilde, H.-G. (1959).J. Amer. Chem. SOC.81, 5325. Smith, W. B., and Gilde, H.-G. (196l).J. Amer. Chem. SOC. 83,1355. Sokol’skii, D. V. (1971). “Progress in Electrochemistry of Organic Compounds” (A. N. Frumkin and A. B. Ershler, eds.). Plenum Press, London. Solly, R. K., and Benson, S. W. (1969). Znt. J. Chem. Kinetics 1, 427. Staples, T. L., Jagur-Grodzinski, J., and Szwarc, M. (1969). J . Amer. Chem. SOC. 91, 3721. Stefani, A. P. (1968).J. Amer. Chem. SOC.90, 1694. Sternberg, H. W., Markby, R., and Wender, I. (1963).J. Electrochem. SOC.110, 425. Sternberg, H. W., Markby, R. E., Wender, I., and Mohilner, D. M. (1966). J. Electrochem. SOC. 113, 1060. Sternberg, H. W., Markby, R. E., Wender, I., and Mohilner, D. M. (1967). J. Amer. Chem. SOC. 89, 186. Sternberg, H. W., Markby, R. E., Wender, I., and Mohilner, D. M. (1969). J. Amer. Chem. SOC.91,4191. Sternerup, H. (1974). Acta Chem. Scand. B28,969. Stocker, J. H., and Jenevein, R. M. (1968).J. Org. Chem. 33, 294. Stocker, J. H., and Jenevein, R. M. (1968a).J. Org. Chem. 33,2145. Stocker, J. H., and Jenevein, R. M. (1969).J. Org. Chem. 34,2807. Stocker, J. H., and Jenevein, R. M. (1971). Coll. Czech. Chem. Comm. 36, 925. Stocker, J. H., and Kern, D. H. (1968).J. Org. Chem. 33, 291. Stuart, J. D., and Ohnesorge, W. E. (1971).J. Electroanal. Chem. 30, App. 11. Svanholm, U., and Parker, V. D. (1972). Tetrahedron Lett. 471. Svanholm, U., and Parker, V. D. (1973). Acta Chem. Scand. 27, 1454. Svanholm, U., and Parker, V. D. (1973a).J. C. S. Perkin ZZ 1594. Svanholm, U., Bechgaard, K., and Parker, V. D. (1974).J. Amer. Chem. SOC.96, 2409. Svanholm, U., RonlPn, A., and Parker, V. D. (1974a). J. Amer. Chem. SOC.96, 5108. Svensmark Jensen, B., and Parker, V. D. (1974). Chem. Comm. 367. Svensmark Jensen, B., and Parker, V. D. (1974a). Private Communication. Swain, C. G., Ohno, A., Roe, D. K., Brown, R., and Maugh, T. (1967).J. Amer. Chem. SOC.89,2648. Swann, S. (1956). “Technique of Organic Chemistry”, 2nd ed., Vol. 2, 385 (A. Weissberger, ed.). Interscience, New York. Swann, S., Eads, D. K., and Krone, L. H. (1966).J. Electrochem. SOC.113, 274. Szwarc, M. (1968). “Carbanions, Living Polymers, and Electron Transfer Processes”. Interscience, New York. Szwarc, M. (1969). Accounts Chem. Res. 2, 87. Thielen, D. R., and Anderson, L. B. (1972).J. Amer. Chem. S O C . 94, 2521. Thomas, H. G. (1971). Angew. Chem. 83,579. Titus, J. A. (1971). Chem. Eng. News. June 7, 6. Tomilov, A. P. (1961). Russ. Chem. Rev. 30, 639. Tomilov, A. P., and Fioshin, M. Ya. (1963). Russ. Chem. Rev. 32, 30. Tomilov, A. P., Mairanovskii, S. G., Fioshin, M. Ya., and Smirnov, V. A. (1972). “Electrochemistry of Organic Compounds.” Halsted Press, New York.

STRUCTURE AND MECHANISM I N ORGANIC ELECTROCHEMISTRY

129

Trahanovsky, W. S., Cramer, J., and Brixius, D. W. (1974). J. A m e r . Chem. SOC.96, 1077. Tsutsumi, S., and Koyama, K. (1968). Disc. Faraday SOC.45, 247. Turner, W. R., and Elving, P. J. (1965). Anal. Chem. 36,467. Utley, J. H. P. (1969, 1970).Ann. Reports o f t h e Chemical Society (B). Utley, J. H. P., and Yates, G. B. (1973). Chem. C o m m . 473. Valenta, P., and Nurnberg, H. W. (1974). Dechema Meeting, “Electrochemical Processes”, Abstracts, Frankfurt, February. Van Duyne, R. P., and Reilley, C. N. (1972). Anal. Chem. 44, 142. VanDuyne, R. P., and Reilley, C. N. (1972a). Anal. Chem. 44, 153. Van Duyne, R. P., and Reilley, C. N. (1972b). Anal. Chem. 44, 158. van Tamelen, E. E., and Akermark, B. (1968). J. Amer. Chem. SOC.90, 4492. Vellturo, A. F., and Griffin, G. W. (1966). J. Org. Chem. 31, 2241. Vielstich, W. (1965). “Brennstoffelemente.” Verlag Chemie, Weinheim/Bergstr. Volke, J., and Kardos, A. M. (1968). Coll. Czech. Chem. Comm. 33, 2560. Wagenknecht, J. H., and Baizer, M. M. (1966). J. Org. Chem. 31,3885. Wagenknecht, J. H., and Baizer, M. M. (1967). J . Electrochem. SOC. 114,1095. Ward, G. A., and Wright, C. M. (1964). J. Electroanal. Chem. 8, 302. Wawzonek, S. (1967). Science 155, 39. Wawzonek, S. (1971). Synthesis 285. Wawzonek, S., and Duty, R. C. (1961).J. Electrochem. SOC. 108, 1135. Wawzonek, S., and Wagenknecht, J. H. (1963). J . Electrochem. SOC. 110, 420. Wawzonek, S., and Wearring, D. (1959). J. Amer. Chem. SOC.81, 2067. Wawzonek, S., Duty, R. C., and Wagenknecht, J. H. (1964). J. Electrochem. SOC. 111,74. Weedon, B. C. L. (1952). Quart. Rev. (London) 6, 380. Weedon, B. C. L. (1960).Adv. Org. Chem. 1, 1. Weinberg, N. L. (1968). J. Org. Chem. 33,4326. Weinberg, N. L. (1974). (ed.) “Techniques of Chemistry V: Technique of Electroorganic Synthesis, Parts I and 11,” Wiley, New York. Weinberg, N. L., and Belleau, B. (1973). Tetrahedron 29, 279. Weinberg, N. L., and Reddy, T. B. (1968). J. Amer. Chem. SOC. 90, 91. Weinberg, N. L., and Weinberg, H. R. (1968). Chem. Rev. 68, 449. Wendt, H. (1973). Chem. Ing. Techn. 45, 1303. Westberg, M. H., and Dauben, H. J. (1968). Tetrahedron Lett. 5123. Wiberg, K. B., Epling, G. A., and Jason, M. (1974). J. A m e r . Chem. SOC. 96, 912. Williams, G. H. (1960). “Homolytic Aromatic Substitution.” Pergamon Press, Ox f or d Wilshire, J. F. K. (1963).Aush. J. Chem. 16, 432. Wilson, A. M., and Allinger, N. L. (196l).J. Amer. Chem. SOC. 83, 1999. Woodhall, B. J., and Davies, G. R. (1969). Abstracts, Meeting of the Society for Electrochemistry, Thornton, England. Wudl, F., and Southwick, E. W. (1974). Chem. C o m m . 254. Yoshida, K., Saeki, T., and Fueno, T. (1971).J. Org. Chem. 36, 3673. Zahradnik, R., and Phkinyi, C. (1965). Talanta 12, 1289. Zavada, J., Krupicka, J., and Sicher, J. (1963). Coll. Czech. Chem. C o m m . 28, 1664. Zuman, P. (1967). “Substituted Effects in Organic Polarography.” Plenum Press, New York. Zuman, P. (1969). Fortschr. Chem. Forsch. 12, 1. Zuman, P. (1969a). “The Elucidation of Organic Electrode Processes.” Academic Press, New York. Zuman, P., and Perrin, C. L. (1969). “Organic Polarography.” Wiley, New York.

.

This Page Intentionally Left Blank

Acid-Base Properties of' Electronically Excited States of Organic Molecules J. F. IRELAND and P. A. H. WYATT Department o f Chemistry, The University of St. Andrews, Scotland

1. Introduction . 2. Experimental Methods

3.

4.

5. 6.

Forster Cycle Determinations . Fluorescence Titrations . Triplet- Triplet Absorption Titrations Lifetime Measurements . Indirect Methods . Kinetics and Equilibria of Excited State Protonation Reactions Singlet Decay . The pH Dependence of Fluorescence Intensities . Triplet Decay Chemical Complications . Physical Complications (1): Kinetic Behaviour During the Approach to the Steady State . Physical Complications ( 2 ): Adiabatic and Diabatic Protonation Reactions The Rate Constants of Radiationless Transitions Survey of Experimental Results . TheTables Limitations of the Forster Cycle . Some Aspects of Recent Work Relative Ordering of pK(So), pK(Sl), and pK(T1) Values and Theories of pK-Change Excited States and Acidity Scales Applications . References

.

.

. .

131

132 136 136 138 141 143 143 144 145 148 152 154 154 157 158 158 165 167 189 204 207 212 215

132

J. F. IRELAND AND P. A. H. WYATT

1. INTRODUCTION This subject was initiated by Forster (1949), who took up Weber’s earlier observation (Weber, 1931) that the fluorescence of l-naphthylamine-4-sulphonate changes in wavelength at a markedly different pH from the absorption spectrum. Forster’s work was developed by Weller and was reviewed by him in 1961, and by others later, viz., Vander Donckt (1970), Schulman and Winefordner (1970), and Schulman (1971a). A section on the subject also appears in new photochemical textbooks, among which we mention a clear exposition of its kinetic and chemical aspects by Parker (1968). Weller’s review (1961) is not confined t o acid-base reactions but deals with the kinetics of excited state reactions in general. Vander Donckt (1970) covers developments of the acid-base section of the Weller field but pays more attention t o physical organic aspects, such as applications of resonance theory to the interpretation of pK-shifts upon excitation and the application of linear free energy relationships. The reviews by Schulman and Winefordner (1970) and by Winefordner et al. (197 l a ) are directed towards possible analytical applications. In the last few years more information on excited state pK-values has accumulated and the present review contains extensive reference tables of the experimental results in the literature available up to August 1974. We have confined our attention throughout to Brdnsted acids and bases, though work on Lewis acid (donor-acceptor) systems continues (Weller, 1961; Birks, 1970; Ottolenghi, 1973) and may prove directly relevant to a deeper understanding of prototropic reactions, which must often be preceded by the formation of hydrogen-bonded complexes. Such interactions also play a role in solvent effects upon the absorption frequencies of acid and base molecules. Where chemical processes such as protonation are concerned, it is the general rule that only the first excited singlet state ( S , ) and the first excited triplet state ( T , ) are involved. This is closely related to Kasha’s rule for radiation emission (Kasha, 1950): fluorescence always occurs from the lowest excited singlet state and phosphorescence from the lowest triplet. Since experimental conditions are often arranged so that protonation is in competition with emission, i.e. s o that their rates are similar, these rules are easily understood in terms of the much shorter time ( & l o - ’ s) required for the S 1 state to be reached from the higher states produced immediately on absorption

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

133

than for the radiative and chemical changes t o take place ( l o T 8 s or longer). Exceptions do occur; it is well known that azulene fluoresces from the S2 level, although the protonated forms of related compounds fluoresce from S, (Dhingra and Poole, 1968), and other cases of S2 emission have been reported recently (Geldof et al., 1969; Easterly et al., 1970, 1973, 1974). Occasionally, simple chemical changes, e.g. electron transfer t o a solvent, are believed t o be rapid enough t o involve higher states (Lesclaux and Joussot-Dubien, 1973). The common pattern for all the compounds to be discussed can be summarized as follows. In the ground state the B and BH+ species have all their electrons paired in the lowest molecular orbitals and are therefore in their lowest singlet states, So. Upon absorption of radiation, an electron is pictured as being promoted, without change of spin, t o higher levels, S, , S, , etc., which are also usually vibrationally excited in a way that depends upon the relationship of the geometry of the higher electronic state to So, i.e. upon the FranckCondon principle or the wave-function overlap between the states. Within about l o - ' s the molecule settles down (at room temperature) in the lowest vibrational level of the S1 state by internal conurnion. From S 1 it either emits fluorrscence or, in the absence of any sufficiently rapid chemical changes, undergoes further internal conE Vibrationally excited S p { Vibrationally

=A-

s3

{, s2

l+

TI

T

I/ I

t

Figure 1. Radiative and radiationless decay processes in a polyatomic molecule (from Henry and Siebrand, 1973).

J. F. IRELAND AND P. A. H. WYATT

134

S2 S1 , S, S, , etc., because of the much larger electronic energy difference between S1 and S o ) . There is also usually time for the electron in the excited orbital to change its spin and thus t o produce a vibrationally excited form of T1 (or T 2 ,if low enough in energy) by intersystem crossing. After that, phosphorescence or radiationless conversion t o the ground state (or further chemical reaction) may occur, but the rates of these spin-forbidden processes are generally much slower than the corresponding S1 changes. For this reason triplet states ficgure much more prominently than excited singlets in photochemical discussions. These primary photophysical processes are summarized in Fig. 1, which is taken from the useful recent review of radiationless processes by Henry and Siebrand (1973). That changes in acid and base strength upon excitation can be very important is evident upon considering the simple indicator equilibrium (1). For the base B to act as an indicator, it must absorb at a -+

-+

different frequency after protonation. Suppose that BH' absorbs at a higher frequency than B. If one then imagines a solution of pH equal t o the ground state pK, s o that [BH'] = [B] , then upon excitation BH' will find itself at a higher energy level than excited B and will then (apart from relatively minor entropy effects) exhibit a strong tendency to change to B. In other words, it will become a stronger acid. If B absorbs at a higher frequency than BH+, then B will become a stronger base on excitation. At 300 nm, near which many molecules of interest absorb, the frequency is 10' s - l , corresponding to -400 kJ mole-'. Thus a 30 nm shift in spectrum between B and BH' corresponds to -40 kJ mole-' which makes a change of 7 units in pK. Since changes of 30 nm or more are common upon protonation, it is quite usual to find that the acid dissociation constant of a protonated compound changes by between 6 and 10 powers of ten after absorption of light. Our knowledge of such processes does not rely entirely upon absorption spectroscopy however; with the development of luminescence spectrophotometry and flash photolysis techniques it is now possible to study protonation equilibria directly in excited states. The relationship suggested above between the frequency shift accompanying protonation (or any other chemical reaction) and the change in equilibrium constant upon excitation is formalized in the Forster cycle (Forster, 1950), illustrated in Fig. 2. Proceeding from

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

BH+

135

14" BH+ k, B +H+

Figure 2. Forster's relationship of enthalpy changes t o electronic transitions.

LhvBH+-k A H *

=

LhVB + A H

1.e.

AH*

- A H = Lh(VB - VBH+)

(2)

,ground state BH' t o excited B by two routes we are led to the equality (2), where AH* and AH refer to the enthalpy changes in the excited and ground states respectively, and us and v B H + t o the frequencies o f the lowest absorption bands of B and BH'; h is Planck's constant and L the Avogadro number. If the solutions are dilute enough for AH to approximate to the standard value AH2 or a t least for the difference between AH* and AH t o approximate the difference AH*0- A H 5 and if AS++ does not change appreciably upon excitation so that AH*0- A H e c a n be replaced by A G * 6 - AG-,

AHe'- A c e = 2.303 RT pK

(3)

If the temperature is 298"K, insertion of numerical values of constants gives (4).

ApK = 0 * 0 0 2 0 9 ( v-~ vBH+)/Cm-'

(4)

When, as above, the excited state is not specified, it is convenient to use the symbol pK*, but when a closer specification is required we can write pK(S1) or p K ( T I ) , and ApK will then stand for pK(S,) - pK(So) or p K ( T , ) - pK(So). Early results seemed t o

136

J. F. IRELAND AND P. A. H. WYATT

version to So by a radiationless process (at a rate which is slower than indicate that pK(T1) always lay between pK(So) and pK(Sl) and much closer to the former than to the latter, as in the classical case of 2-naphthol (Jackson and Porter, 1961); but as further results have accumulated it has become clear that pK(T1) may be nearer t o p K ( S l ) than to pK(So), as in anthroic acids (Vander Donckt and Porter, 1968a), and may even lie outside the pK(S1) - pK(So) range, as in xanthone and benzophenone (Ireland and Wyatt, 1972, 1973). In view of the clear relationship between pK-changes and absorption spectra, a study of' the influences of substituents and other consitutional changes upon such spectra has a very direct bearing upon the field of acid-base properties in excited states. For example, the -OH and -0- groups function as different substituents a t the 2-position in naphthalene. Any theory which accounts for their different effects upon the naphthalene transitions therefore automatically also explains the change in the naphthol-naphtholate equilibrium upon excitation. The search for linear free energy relationships in electronic spectra will therefore continue t o impinge upon this field. 2. EXPERIMENTAL METHODS Forster Cycle Determinations Excited state pK-values are most easily accessible through the use of the Forster cycle which has been described in the introduction. To perform this calculation for a particular molecule it is necessary to know the ground state equilibrium constant for the reaction in question and to have some measure of the energy difference between the lowest vibrational level of the ground and the excited state in both the B and BH' forms. Thus to calculate pK(S,) we need the 0-0 energy of the So-Sl transition and for pK(T1) that of the So- T I transition. Methods of obtaining pK(So)-values are well documented (e.g., Albert and Sargent, 1962). Since the molecules of interest in excited state acid-base studies absorb at different wavelengths in the B and BHf forms, absorption spectroscopy is commonly used in the relevant ground-state pK-determination. In proceeding to the excited state pK, it is not always easy to obtain a good value for the energy of the 0-0 transition except for

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

137

those cases in which the molecule shows vibrational fine structure in its absorption, fluorescence, or phosphorescence spectrum. Unfortunately, for most substituted aromatic compounds in solution, the absorption and emission spectra are characterized by broad, structureless long-wavelength bands, and the best that can be done is to use a maximum (or an obvious shoulder) of the longest wavelength feature. In either case, whether vibrational features are clear or the maximum of a broad band has to be used, the emission spectral data used in these calculations should always be obtained from the corrected spectra. The B and BH+ forms will emit at different frequencies and the spectral correction factors can differ significantly. Methods of obtaining corrected emission spectra have been reviewed by Parker (1968) and b y Aqgauer and White (1970). In practice, handling So- T I transitions is often simpler (though less satisfactory), since absorption spectra are not normally available and the maxima of the B and BH’ phosphorescence bands must then be used in the Forster cycle. However So-Tl absorption spectra can sometimes be obtained, especially if perturbation methods can be used to enhance the singlet t o triplet transition probability. For example, Grabowska and Pakula (1966) induced So-T1 absorption in a series of nitrogen-containing heterocyclic compounds by the oxygen perturbation method of Evans (1957). Hence, for these compounds, by combination of absorption and phosphorescence spectral results, the 0-0 transitions could be located more accurately. If the compound does not fluoresce, pK(S1)-values can be calculated from the absorption maxima but they are then mainly of theoretical interest since the lifetime o f the S1 state is likely t o be too short to allow the protolytic reaction ever to reach equilibrium. For compounds which fluoresce there are three approaches to obtaining a 0-0 energy approximation: ( u ) the use of absorption maxima, (6) the use of fluorescence maxima, and (c) the averaging of the absorption and fluorescence maxima of each form (Weller, 1952; Wehry and Rogers, 1965a). There has been much discussion as t o which of these is the “best” method of obtaining the energy values needed for the Forster cycle, but in any case the sensitivity of the calculated energy to small errors in the location of the 0-0 band can be appreciated when it is recalled that, at 300 nm, an error of 4 nm corresponds to 1 unit of pK. The effects of solvation on 0-0 energies are further discussed in Section 4. In practice it is sometimes possible to make use of a special characteristic of the acid or base form of a particular compound and even to estimate the 0-0

138

J. F. IRELAND AND P. A. H. WYATT

transition energy differently for the two species. For example, in the case of several benzophenones (Ireland and Wyatt, 1973) the So-S1 absorption of the base form in aqueous solution is obscured by the more intense So- S2 transition but, since the phosphorescence excitation spectrum shows the vibrational spacing of the first transition, it is not difficult to get a good 0-0 value for the B form. On the other hand, the protonated forms of the benzophenones have broad, intense, long-wavelength absorption bands, but because they fluoresce, unlike the B forms, a good estimate of the 0-0 energy can be obtained by the averaging technique. It is clear that the Forster cycle will only give precise pK*-values in special cases, but it will at least indicate the direction of the pK-shift and in many cases will give a good approximation to the maLgnitude of the change. (When ApK is small, even the direction of the change may occasionally be incorrect, e.g. Ballard and Edwards, 1964a; Winefordner et al., 1971b).

Fluorescence Titrations

The changes in molecular fluorescence with acidity give information about the protolytic behaviour of the excited singlet state of a compound. A pK(S, )-value calculated from the Forster cycle gives an indication of the acidity range in which a fluorescence change is expected, but, since no account is taken of the rekvant rates of the particular processes involved, the Forster cycle does not indicate whether or not proton transfer in the excited state is kinetically feasible. Improved instrumentation has made fluorescence techniques more available and extended the range of compounds accessible t o the method because of improved sensitivity. The techniques of fluorescence spectroscopy are well known (Udenfriend, 1962; Parker, 1968; Argauer and White, 1970). To investigate the excited state behaviour of a particular compound the fluorescence intensity of the B and BH’ forms (or that of B or BH+if only one of these fluoresces) is recorded at various values on the pH or other acidity scale (e.g. Ho ). If the fluorescence spectra of the acid base pair overlap, then the and B (9’) must be measured fluorescence intensities of BH+ (9) corrected for the intensity component due t o the other form. The true fluorescence intensities (4 and 4’) are related t o the measured intensities according to the equations:

ACIDBASE PROPERTIES OF ORGANIC MOLECULES

139

where k and k’ are the overlap ratios of the BH’ and B forms respectively. To obtain the overlap ratios, measurements are made on solutions containing only one species in the excited state. Thus k is obtained from solutions showing only the characteristic fluorescence of the protonated form and is the ratio of the fluorescence intensity of BH’ (measured at the analytical wavelength for B ) to the intensity at the wavelength where BH’ emission is measured. Similarly k’ is obtained from measurements in a solution having B as the only fluorescent species. Rearranging eqns ( 5 ) and (6) we can obtain the true fluorescence intensities in terms of Y,Y’, k and k ‘ :

@’

=

9’- k 9

1 - kk’ In cases uncomplicated by quenching effects (h, and 4; values can be taken as the limiting value of the true fluorescence intensities at 2 or 3 pH units away from the half way point of the fluorescence change. The relative intensities (@/& and @‘/&) in the intermediate region can then be plotted against acidity. Figure 3 gives a plot of this type for the cation and zwitterion of 3-hydroxyquinoline

HO

PH

Figure 3. Acid dependence of 3-hydroxyquinoline fluorescence intensity: cation (0); zwitterion ( 0 )(Haylock et al., 1963).

J. F. IRELAND AND P. A. H. WYATT

140

(Haylock et al., 1963). Where complications due to non-equilibrium or quenching effects complicate the curve of one form so that $o or 4; is not directly measurable (see e.g. Fig. 4), one may be determined from the other via the relationship (9).

+I$o +@'I& = 1

(9) The inflection point of the fluorescence intensity curves against acidity gives a first approximation t o the pK(Sl )-value but this involves the assumption that the protolytic equilibrium is established within the lifetime of the S l state and that the fluorescence lifetimes of B and BH+ are equal. These assumptions are less likely to hold when only one form is fluorescent and Lasser and Feitelson (1973)

I

0.5

1.0

I

1-5

I

I

2.0

2.5

3.0

PH Figure 4. Acid dependence of 1-naphthamide fluorescence intensity: protonated form (0);unprotonated form ( 0 )(Watkins, 1972b).

have concluded that fluorescence against pH curves do not give good pK(S,)-values in such cases. It can be shown (see Section 3) that for excited state equilibrium to be attained the relationships (10) hold, where rb and ro are the

fluorescence lifetimes of B and BH+ respectively, k , is the rate of dissociation of BH+ and k 2 is the rate constant for protonation. It is not surprising that the system reaches equilibrium when the rate of fluorescence decay is considerably less than the protolytic rate constants. In these cases it is easily shown that eqn (11) holds, where

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

141

pH = pK( S ) - log T~ /T; pH is the acidity a t which the relative intensity curves show an inflection. To obtain corrected pK(S1 )-values we must therefore have a measure of excited state lifetimes. Triplet-Triplet Absorption Titrations

This method for obtaining pK(T, )-values was introduced b y Jackson and Porter (1961). It is even more time-consuming compared to the Forster cycle than the fluorescence titration method and relatively few direct determinations have been made of pK(T, ). The experimental techniques have recently been described by Chibisov (1970) and Labhart and Heinzelmann (1973). Developments in instrumentation, including the introduction of laser excitation, have been reviewed by Porter and West (1973). For this technique to be applicable, one or both forms of an acid-base pair must show a T-T absorption. If, as in the case of xanthone (see Fig. 5 ) , T-T absorption spectra are obtained from both forms, the observed optical densities must be corrected for any overlap in a manner analogous to that used for fluorescence inten-

460

500

540

500

620

Wave length /nm

Figure 5. Triplet-triplet absorption of xanthone at three pH values: (a) p H = 7.0, (b) pH = 2.8, ( c ) pH = 2.1 (Ireland and Wyatt, 1972).

142

J. F. IRELAND AND P. A. H. WYATT

sities. From plots of the relative optical densities against pH, pK(T, )-values can be obtained. For example, from Fig. 6 a value of 3.0 is obtained for pK(T, ) for xanthone. The subsequent decrease in optical density of the BH’ triplet around pH 1 as the acidity is increased, is caused by a protolytic reaction in the S , state. The BH’ singlet form is a more efficient fluorescer than the B form, whence intersystem crossing is diminished and the BH+ triplet state is not produced in such a high concentration as at lower acidities. This gives 10

? .

-~~ \ -8 B 0 6 4

,-o-04[y

r*-*-*

i ‘i

02

1 ;

>O\

,

2

4

6

0-0

-2

0 Ho

O

0

lo-

8

PH

Figure 6. Relative optical densities of triplet-triplet absorption of protonated and unprotonated xanthone as a function of acidity: 0,@/Oo at 530 nrn (protonated form); 0, O’/& at 595 nm (unprotonated form) (Ireland and Wyatt, 1972).

confirmation of the pK(Sl )-value obtained by fluorescence titration, subject however t o the same lifetime and equilibrium qualifications. Owing to the longer lifetime of the triplet state it is expected that the protolytic reaction will usually reach equilibrium within the lifetime of the state. Unlike the fluorescence titration method for pK( S 1) described above, the triplet-triplet absorption technique leads directly to pK(T,) without the necessity for a knowledge of lifetimes. Phosphorescence “titration” studies, on the other hand, will involve the lifetime term log rO/rbjust as for fluorescence. The use of laser flash photolysis has extended this type of work to shorter-lived species and has enabled the initial triplet concentration to be monitored against pH (e.g. Rayner and Wyatt, 1974). There is a drawback at present in that the molecule under examination must absorb at a frequency dictated by the laser, but this restriction will probably become less severe as tunable lasers are developed.

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

143

One common complication in flash photolysis studies is the production from the excited state of radicals which may be confused with or inhibit observation of the triplet state. The radical species, which usually has a longer lifetime than that of the triplet state, may be identified by separate experiments (Jackson and Porter, 1961); but such species may themselves show interesting acid base properties (e.g. Lindqvist, 1960; Lindqvist and Kasche, 1965; Simic and Hoffman, 1972; Neta, 1975). Their optical density contributions may therefore change with acidity in the range studied.

Lifetime Measurements

Two techniques, phase and pulse fluorometry, are used for the direct measurement of fluorescence decay rates, and their principles are described by Birks and Munro (1967), Parker (1968), and Birks (1970). The photon sampling method has proved useful and versatile. This is an iterative technique in which single photons are counted as a function of the time at which they appear after excitation and a complete decay curve is built up. (For recent references see e.g. Zimmerman et al., 1973, 1974). Wider use of the photon sampling technique will increase the precision of lifetimes obtained and extend the range of compounds studied to those with shorter lifetimes or very low fluorescence yields. If no direct measurement of the fluorescence lifetime is available the relations between the radiative lifetime and the fluorescence and absorption spectra can be used in conjunction with the quantum yield to obtain an indication of the fluorescence lifetime. Birks and Munro (1967) have reviewed the methods of calculating the radiative lifetime. In general these methods are limited to specific groups of compounds. For example, Favaro et al. (1973) applied Stickler and Berg’s (1962) formula to the spectral data obtained from an excited state acid-base study of some styrylpyridines and found a lack of quantitative agreement between the measured and calculated lifetimes.

Ind ir ec t Methods Although fluorescence and T-T absorption titration methods are the most commonly used techniques for obtaining direct experimental evidence of p K ( S I ) and pK(T,), other less direct approaches

144

J. F. IRELAND AND P. A. H. WYATT

have been tried. Rosebrook and Brandt (1966) measured the potential between one illuminated electrode and one dark electrode in a solution of a base of interest. Changes in the “photopotential” with pH were related to the pK-value of the excited singlet state. The results obtained for naphthylamines and 3-hydroxypyridine were in agreement with those from fluorescence titration and Forster Cycle calculation. This method does not appear to have been extended, perhaps because of a limitation in the types of compound to which it may be applied. Avigal et al. (1969) investigated the quenching effects of a series of carboxylate ions upon the fluorescence of substituted phenols. Rate constants for the quenching process, obtained from SternVolmer plots, satisfied the Br$nsted general base catalysis law and extrapolation to the base strength of water then leads to an estimate of the rate of reaction between ROH* and H, 0. Assuming the reverse reaction to have a rate constant of 5 x 10’ dm3 mole-’ s-’ , they were able to calculate values of pK(Sl). This does not seem to be very different in principle from the method used by Weller (1957a) to determine pK(Sl) for acridine. Since acridine is not protonated rapidly enough by H 3 0 + in the required region of pH, he determined the forward and reverse rate constants for protonation with a different acid, NH:, .and used the pK-value of NH: to obtain the pK(S,) of acridine in the H 3 0 + - H 2 0 system. The reactivity of the triplet state in photoreduction reactions was used by Nakamaru et al. (1969) to investigate the triplet state basicity of acridine. It should be possible to extend this method to any compounds in which an excited state reaction is affected by protonation.

3. KINETICS AND EQUILIBRIA O F EXCITED STATE PROTONATION REACTIONS In the earlier work (Forster, 1951; Weller, 1952) when the principal experimental information involved fluorescence intensities, the most useful algebraic expressions were those relating relative quantum yields of the conjugate acid base pair to the solution acidity. In favourable cases such expressions were used to obtain rate constants for the forward and reverse reactions, and hence equilibrium constants, though it was always necessary to make allowance

145

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

for the different lifetimes of the excited states of the species. With the development of laser flash and photon sampling techniques however, much greater possibilities have been opened up for the direct measurement and use of excited state lifetimes, and the equations derived for the rates of decay of luminescence or transient absorption intensities will therefore find increasing application. The groundwork for the solution of the types of problem encountered was laid by Forster (1951) and Weller (1952, 1954, 1957a, 195713, 1958a, 1958c, 1961) and some examples will now be given to illustrate their method of approach.

Singlet Decay To simplify the initial presentation, very rapid local effects, sometimes encountered during the approach to the steady state, will be omitted. Further, only the simplest proton transfer reaction will be considered, along with allowances for quenching, emission and intersystem crossing. Such other possibilities as solvated electron formation or the secondary formation of radicals by hydrogen abstraction etc. will be ignored initially. The two principal types of derivations can then be illustrated directly in terms of the singlet decay scheme, which represents the processes which occur after a molecule absorbs light and reaches the first excited singlet state. The protolytic reaction, radiation transition, and fluorescence by the two excited species are all allowed for in (12). The concentrations of

BH+(s~)

BH+(s~)

+hv

BH+(T~)

B(T1)

WSo)

WSo)

+hv'

BH+(S,) and R ( S l ) , the protonated and unprotonated base species in the S 1 state, will be represented by x and y respectively, and the rate constant suffixes denote transitions to the ground state by solvent and other quenching (q), and by fluorescence (f), or transitions by intersystem crossing to the triplet state (T). Primed constants refer to the base species B(S,) and unprimed to the conjugate acid BH+(Sl). In the protonation equilibrium the rate

146

J. F. IRELAND AND P. A. H. WYATT

constant for the dissociation of B H + ( S l ) is denoted by k , and the recombination rate is characterized by the effective first-order constant k , [H+], which is contracted to kk for convenience ([H'] being constant in a given solution). The rates of change of the BH+(S1)and B(S,) concentrations can now be represented by the relations (13) and (14), where a = k , + kf dx dt

-+ u x =

-++'y dY dt

k;y

=K,x

+ k , + k T and a' = k ; + k ; + kb + k > . These equations can sometimes be used directly on experimental results, as shown by Loken et al. (1972), who derived rate constants in this way which agreed well with those from the integrated solutions (15) and (16) (for which see also Weller, 1958a; Birks, 1970; Ofran and Feitelson, 1973; Rayner and Wyatt, 1974):

where m = (a - a ' ) , + 4k,k;

a = (a + a' +.\Tm)/2

p = (a + a' - f i ) / 2 and x o and y o are the values of x and y at t = 0. Clearly, in cases in which only BH' or B is excited (and no instantaneous conversion takes place before the steady state is set up), (15) and (16) are considerably simplified by setting x o = 0 or y o = 0. In specific cases the use of justifiable simplifications about the magnitudes of the components of a and a' can simplify the expressions still further. For example, if y o = 0 and the pH is high enough for k ; to be very small compared with all the other rate

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

147

constants in a and a ' , (15 ) and (16) become (17) and (18), where

7=

x = x o e-t/T

(17)

y = [ ~ , x , / ( a- a ' ) ] ( ~ ~ 1 e" - t / T ) ,

(18)

l / ( k l + k f + k , + k T ) is the lifetime of B H + ( S I )at e.g. neutral pH [as distinct from its lifetime ( k f + k , + k T ) - ' at low enough pH for only BH+(Sl ) to be present] and 7' = l/(kk + kh + k k ) is the lifetime of B at high pH. Equations (17) and (18) were used, for example, by Loken et al. (1972) and Ofran and Feitelson (1973) in their studies of 2-naphthol lifetimes. The form initially excited dies away exponentially from x o , while the other form grows from zero at a rate which initially depends only on k xo (since for small values of t (18) reduces to y = k x , t ) and then passes through a maximum, just as would be expected (Fig. 7). For molecules of interest, the excitation pulse and the instrumental response time will often be comparable to the lifetimes being measured and deconvolution methods will become necessary, so that

0

10

20

30

40

50

Time/ns

Figure 7. 2-Naphthol fluorescence decay curves at pH = 3.43: (a) flash profile; (b) 2-naphthol fluorescence intensity at 360 nm; (c) 2-naphtholate fluorescence intensity at 450 nm ( x 2.5 relative to ( b ) ) (from Loken et al., 1972).

(17) and (18) are not immediately applicable t o the experimental results as they stand (see e.g. Knight and Selinger, 1971; Demas and Adamson, 1971; Zimmerman et al., 1974). While the general form of the time dependence is quite recognizable in the 2-naphthol study mentioned above (Fig. 7), new features can, however, emerge when the curve-fitting is carried out carefully. Ofran and Feitelson (1973)

J. F. IRELAND AND P. A. H. WYATT

148

found that they could only fit their curve properly if they assumed that some of the naphtholate form was produced in the S 1 state so rapidly after excitation as t o justify introducing a y o value different from zero. This point will be referred to later, but at this stage we may note that effects of this kind can certainly affect the interpretation of the curves as late as 20 or 30 ns after excitation. The most important outcome is, however, that a study of the changes in 7 and 7’ with changes in experimental conditions can lead to the determination of k and k , and hence directly to k , / k , , the excited state equilibrium constant.

,

The pH-Dependence of Fluorescence Intensities When only fluorescence intensities in steady state conditions have t o be handled, some of the mathematical manipulation for the complete solution of (13) and (14) can be by-passed. Weller (1952) then focuses attention upon the probability, p , that a molecule will still be in the excited state at time t after excitation. If the molecule concerned is the initial absorber of the exciting radiation, po = 1 at t = 0: in this case p at some other time t will be equivalent to x/xo in the above treatment. In these terms (13) and (14) become (19) and (20), while the fluorescence quantum yields 6 and 6‘are given by dP dt

= &p’

-+Up

I dp‘ dt (21) and (22). Since these yields are

-+ a p = k , p

6=

60 = kf/(kf + k ,

k f p dt

+k

~= k) f 7 0 (21)

0

1 kkp’ m

4’

=

dt

(22)

0

and $I; = kk/(kk + k b + k k ) = k ; ~ ;respectively when only one species is present in the excited state, the quantum yields relative to $Io and $I; are given by (23) and (24).

149

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

These two equations then allow substitution of JF p dt and JF p‘dt in the integrated forms of (19) and (20) by T ~ @ / @ O and ~ b @ ‘ / @ band, p o = -po and S;; dp’ = -pb, yield finally eqns (25) with S;; dp = p,

and (26). Two special cases will illustrate the application of these equations. Case I : the compound is in the BH’ form in the ground state and this is the only form excited. Then po = 1, pb = 0, and (25) and (26) reduce to the equations used originally by Weller (1952) t o describe the variation with pH of the ultraviolet and blue fluorescence intensities of 2-naphthol, shown in Fig. 8 (see also Parker, 1968).

-0 \ -8

-8

0.5 0 -8

\

-8

0

2

4

6

8

10

PH

Figure 8. Acid dependence of 2-naphthol fluorescence: (a) 2-naphthol; (b) 2-naphtholate (Weller 19 58a).

At low pH any excited naphtholate ion formed from excited 2-naphthol is quickly reprotonated; but as the pH increases the reprotonation rate constant k ; becomes smaller (and @’ increases from zero) until beyond pH3 reprotonation becomes negligible and @/$o settles down to the plateau value 1/(1+ k l T ~ and ) @’/& to

150

J. F. IRELAND AND P. A. H. WYATT

k T~ /( 1 + k T ~ ) whence , k T~ is calculable directly. Along the plateau the fraction of excited 2-naphthol converted to naphtholate simply depends upon the ratio of the dissociation rate to the sum of all the rates for its disappearance, viz. k l / ( k l + k f + k , + k T ) or k /(k + 1 / ~ ~ At) pH . values near the ground state pK of 2-naphthol (9-5), further excited naphtholate appears because of its presence in the ground state, while at pH values just below the plateau region, where k ; 7; becomes comparable t o k T ; , the analysis shows k 2 7; to be determinable also, e.g. from a plot of (4/Qo)/(@'/&)against [H+] using (27). The Brqnsted dependence of k 2 upon the ionic strength,

,

I , can be allowed for by plotting [H+]1 0 Z B f ( ' ) instead of [H'], Z B being the charge on the base B and f(1) a function such as that of - 0.31, which reduces t o g at low ionic Davies ( 1 9 6 2 ) , f l / ( l +fl) strengths.) From the fluorescence intensity measurements it is therefore possible to determine

but the excited state pK can only be obtained if rb and ro are known. The deprotonation of protonated 3-hydroxyquinoline is of the Case I type like 2-naphthol, but no plateau is observed (Fig. 3 ) , presumably because k l T~ % 1 so that k , /(kl + 1 / ~is ~ almost ) unity. Since pK(Sl) = -0.2 (Haylock et al., 1963), it follows that k 10' s-' and the condition k T~ 3 1 is likely to be satisfied. Case 11:the compound is in the B form in the ground state and this is the only form excited. Then po = 0, p; = 1 and the equation corresponding to (27) is (29). Here the form of the fluorescence

-

intensity curves expected is always that of Fig. 3 (perhaps modified by quenching effects as in Fig. 4) and never of the 2-naphthol type with a plateau. Case I1 also differs from Case I in that variation of the pH cannot separate k 2 7 b and ) be only k 2 7 b / ( 1 + k l ~ o can

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

151

determined. If k l ~ % O 1, the same result as for Case 1 becomes accessible, viz. (28); but extraneous information, such as an estimate of k 2 (Watkins, 1972a, b) coupled with values of T~ and ~ b will , generally be necessary to separate k and k 2 . The pH at the half-way point on full “titration” curves of the Fig. 3 type, or on the similar curves approaching the plateau in Fig. 8, gives a simple indication of the pK(S1) in favourable cases. For compounds of either Case I or Case I1 type, the pH at the half-way point is log1 0 [k27;/(1 + k1 T ~ ) ]which , is only the same as pK(S, ) if 7; = 7o and k T~ 9 1. Apart from the complication of the lifetime ratio, the half-way point on an emission “titration” curve can only be expected to correspond to the excited state pK-value if equilibrium is achieved in the protolytic reaction; i.e. k , and kk must be large compared with the rates of the decay processes, which sum to 1 / and ~ 1~/ ~ respectively, ; whence k , T~ 9 1 and k b ~ 9 b 1. [The 1 in the denominator, or numerator, of equations like ( 2 5 ) , (26), (27), or (29) always takes care of the extent t o which failure to achieve equilibrium is important.] If k T~ < 1, the half-way point of a fluorescence “titration” curve must occur at a pH of log, 0 k 27;, which typically has a value -2, corresponding to k 2 10, mole-’ dm3 s-, and 7; lo-’ s. Thus reliance on the mid-point of “titration” curves is misleading for compounds with pK(S, )-values much above 2 or 3, which may all appear to have pK(Sl )-values near 2 unless their B and BH+ species have long lifetimes. Even if equilibrium is achieved in the excited state, the lifetime effect always remains in emission intensity measurements. If the experimentally determined quantity on the left of (28) is regarded as the approximate pK(S,)-value, the compound will appear t o be a weaker acid than it really is i.e. too much BH+(S1)will seem t o be present if BH+(Sl) has the shorter lifetime, or a stronger acid if B(Sl) has the shorter lifetime. The rule is that the species of shorter lifetime produces a greater emission than would be expected from the equilibrium concentrations, because its concentration is depleted more rapidly and is therefore replenished all the time from the longer-lived species in an attempt t o maintain the equilibrium. Despite the fact that emission intensity measurements require supplementary lifetime determinations, they have the advantage over laser flash absorption techniques that they by-pass complications due to absorption by secondary products, such as radicals or solvated electrons (see e.g. Klaning et al., 1973).

,

-

-

152

J. F. IRELAND AND P. A. H. WYATT

TripLet decay When triplets are produced by conventional flash photolysis, with a flash on the ps timescale, so many singlet lifetime cycles are passed through and the timescale is relatively so long that both BH+ and B may be found to some extent in the triplet state whichever form is present overwhelmingly in the ground state. Even in laser studies on the ns timescale, there will generally be some competition between the approach to the S 1 protolytic equilibrium and intersystem crossing. Thus in transposing the singlet scheme described above to the triplet reaction, it is even more necessary to allow for concentrations of both B and BH+ forms to be present initially. The lifetimes of the triplets are usually so much longer than those of the singlets that these initial concentrations of BH+(Tl) and B(T1 ) can be taken as arbitrarily given, the fast initial processes involved in their build-up being ignored. At certain acidities the singlet lifetimes may nevertheless be sufficiently long for these initial triplet concentrations to reflect the equilibrium singlet concentrations of the acid and conjugate base forms; but when pK(Tl ) differs from pK(S1 ) the approach to the triplet equilibrium from the initial concentrations has still to be allowed for, even though the triplet lifetimes may be sufficiently long for the triplet equilibrium to be achieved over a much wider pH range. Ignoring for simplicity other complications, such as radical formation, triplet-triplet annihilation, etc., the triplet scheme can therefore often be represented as (30). With triplet concentrations denoted by

X and Y , the corresponding eqns to (13) and (14) are (31) and (32), gdt+ A X = k ; Y

153

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

where A = ( k , + k , + k q ) and A' = ( k ; + kg + k b ) . The solutions are just like (15) and (16) but with X, Y, A , A ' , in place of x , y , a, a ' . (Rayner and Wyatt, 1974). When the rate of acid dissociation is large compared with the rate of phosphorescence and radiationless processes, i.e. k , % ( k , + k q ) or ( k ; + h i ) , neglect of small terms in the expansion

fi = ( k , + k ; ) + ( k ,

+k, -kb

-

kb)(k,

~

k;)/(k, +A;).

..

permits the simplification of the exact solutions to (33) and (34),

where

These equations are now easily interpreted. The first terms, in exp [ - ( k , + k ; ) t ] , show the changes in X and Y due to the approach to the triplet protonation equilibrium, while the second terms, in , the decay of the triplet population in terms of the exp ( - t / ~ )describe lifetime T. From (35) it can be seen that at high acidities, when k ; is ~ ( k , + k q ) , while at low acidities it approaches large, 1 / approaches (kg + k b ) , just as expected. Measurements of T over the intermediate range of acidity can then lead to a determination of K ( T , ) , i.e., of k , / k , . For example, eqn (35) can be thrown into the form (36), in

which the coefficient of K(T,) is determinable by experiment, the factor 10- B f(') allowing for the Br$nsted dependence of k , on the ~ between ( k , + k q ) ionic strength. As the signs show in (36), 1 / lies and ( k b + k b ) and, if these should happen to be equal, (35) requires that 1 / will ~ not alter with pH and pK(T, ) will not be determinable in this way.

154

J. F. IRELAND AND P. A. H. WYATT

Chemical Complications The above treatment gives some idea of a common pattern of the argument and of the scope of intensity and lifetime measurements. When other chemical processes have to be taken into account they can generally be incorporated into the schemes in a straightforward way and the resulting differential equations can be handled similarly. One of the simplest modifications arises in the treatment of acid-base reactions other than those involving the solvated proton. Thus Weller was able to investigate pK(S,) of acridine through its protonation by NH; at high pH, where protonation by H 3 0 + was negligible within the S1 lifetime (Weller, 1957a). Hydrogen abstraction reactions, like that involved in the formation of naphthoxyl from naphthols, must also be expected as a complication in the interpretation of triplet transient spectra: thus a strong absorption at 465 nm due to the naphthoxyl radical has to be distinguished from the triplet absorptions when 2-naphthol is flash photolysed Wackson and Porter, 1961). Electrons can be ejected by ultra-violet light from aromatic molecules in solution (for a review, see Lesclaux and JoussotDubien, 1973). Recently Klaning et al. (1973) have reported that solvated electrons are formed from 2-naphthol by a 337.1 nm pulse from a nitrogen laser, and deduce that the rate of electron formation is higher than the rate of vibrational relaxation to the lowest excited singlet state. Since the electrons are relatively long-lived and take no further part in the ns time-scale reactions, the only effect they have upon the acid-base investigation is to necessitate a correction to the absorption spectra of the transient species under observation.

Physical Complications ( 1 ) :Kinetic Behaviour During the Approach t o the Steady State Many examples have now accumulated of chemical rate constants which have been found to agree well with the values predicted by Debye's equation for ionic encounters in solution (Debye, 1942), which, as is well known, requires encounter rate constants to be of the order 10" dm3 mole-' s-l in water at 25°C. A review of the theory of encounter rates in general, with some reference to quenching mechanisms, has been given by Noyes (1961), and of

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

155

encounter rates with particular reference to acid-base excited state reactions by Weller (1961), who also gives a full account of tests of the application of Brjbnsted’s theory of primary salt effects to such reactions. Several investigators have given attention to the time dependence of rate parameters to be expected before steady-state conditions are established after a rapid perturbation. Forster (195 1) pointed out that, when quenching of fluorescence simply depends upon the formation of an encounter complex between the quenching species and the excited state molecule, there will necessarily be some difference in behaviour between those molecules which find themselves in the vicinity of a quencher immediately upon excitation and those which do not. Those excited molecules which have a quencher within their “diffusion volume” are quenched so much more rapidly than those which do not that they may be omitted from the stationary state scheme for the dependence of fluorescence intensity upon quencher concentration; and the effect will show up because of the increasing importance of the correction with increasing concentration of quencher. At sufficiently low concentrations the chance of finding a quenching molecule in the diffusion volume of the excited species is negligible. Weller has explored the implications of this effect upon acid-base equilibria, in which the diffusion-controlled nature of many protonation reactions obviously parallels that of quenching (Weller, 1957b, 1958a, 1961). He finds that the increased rate of the initial stages of the reaction can be accommodated by dividing the excited molecules into two parts, according to the presence or absence of a proton (or other reactant, such as OH- etc.) in their diffusion volume, and treating them quite separately according to equations like (35) and (36). The separate contributions to and $’$; are then finally summed. He gives two distinct examples of the application of this idea (Weller, 1958a): (u)

Reactions of the type (37). Here the excited HA* molecules are HA* + B

+

A*-+ HB+

(3 7)

divided into one fraction, W B , in the diffusion volume of which no B molecule is found, to which (25) and (26) are applied with po = W B and p; = 0; and another fraction 1 - WB , in which the reaction is so fast that all HA* molecules belonging t o this fraction can be regarded as converted initially to A*-, i.e. po = 0 and pb = 1 - M i B .

156

J. F. IRELAND AND P. A. H. WYATT

( b ) Reactions of the type (38). In this case the argument is slightly

different. All the HA* molecules are surrounded by H 2 0 and are therefore at liberty to react at their normal rate, but those which have an H 3 0 + ion within their diffusion volume are assumed to be reconverted rapidly t o HA* after dissociation and t o have a negligible chance of dissociating again within their lifetime (so that, for this fraction, k , = 0). The fraction of HA* with no H 3 0 + in the diffusion volume is then treated as before with po = WH+ and pb = 0; while the other fraction differs in having po = 1 - WH+ and pb = 0 and, since k , = 0, q5/$o = 1 - W H + and $'/q5; = 0 also for this fraction. Notice that pb = 0 for the whole population in this case. Although this work was published some years ago, it is recalled here in order t o point up a difference from a recent finding which at first sight appears to overlap. Ofran and Feitelson in their laser study of 2-naphthol (1973) found that their decay curves could only be made consistent with the known lifetime of 2-naphthol and 2-naphtholate if some 2-naphtholate was assumed to be present in the excited state initially, even though only the naphthol form could be excited from the ground state. There are two points of difference from the foregoing account: (i) the 2-naphthol reaction is of type (b) above, for which po was set at zero for both fractions of excited HA*, while Ofran and Feitelson find pb f 0; (ii) in the Weller treatment the two parts of the reaction are dealt with quite separately, often with different assumptions about k etc., and the $/$o contributions are summed at the end; the same result is not obtained in general by simply substituting values other than 1 or 0 for po or pb into (25) and (26) for the reaction as a whole. Ofran and Feitelson do not actually compare their findings with the Weller treatment of initial reaction effects, but suggest that the explanation may lie in an initially greatly enhanced acidity of the 2-naphthol molecule when it finds itself surrounded by the ground state solvent arrangement before relaxation to that appropriate to the excited state. Evidently we may look forward to more penetrating insights still into all these effects as the laser and single-photon counting techniques become further refined. It will obviously be very convenient if it always proves possible to justify the use of (35) and (36), and (15) and (16), with reasonable choices of p o , pb, x o , and y o .

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

157

Physical Complications ( 2 ) :Adiabatic and Diabatic Pro to na tion R eact ions

It has long been recognized (Jaffi. and Jones, 1965; Schulman, 1973b) that there might be a difficulty in realizing excited state equilibria unless the states of the base and its conjugate acid are of the same type, when the proton can be attached or detached “adiabatically” (without the loss of excitation). It is clearly quite possible to imagine “diabatic” transitions, in which the electronic reorganization during the transfer of a proton to an excited base necessarily results in the production of the ground state protonated species. Indeed protonation of a molecule in the vicinity of a quencher with immediate de-excitation of the protonated form could sometimes constitute a reasonable quenching mechanism (Umberger, 1968). Latterly Forster has examined these possibilities from a theoretical point of view (Forster, 1970, 1973). In general the relative possibilities of retention and loss of excitation during a proton transfer might be expected t o depend upon several factors, among which the types of state involved may be important; but there seems no compelling reason t o exclude the possibility of adiabatic proton transfers between states of different types, and there is some experimental indication that such transfers do take place (Ireland and Wyatt, 1973). From the practical point of view, however, effects of this kind may be accommodated by quenching terms in the foregoing algebra. The proportion of excited molecules which undergo diabatic proton transfers, ending up in the ground state, would count as being quenched either by protons when an [H’] dependent term appears in k.:, (which then reads k b o + K’,+[H+]), or by the solvent (or other base) when the effect is included in k.6 in the BH+(S, ) or BH+(T1)dissociation schemes. Cases of fluorescence quenching in acid solutions are quite common, but care is needed in their interpretation since at least part of the quenching effect may be due to the anions present (as can easily be demonstrated by adding salts to vary the anion concentration at a fixed acidity). Anion quenching effects have been recorded for some time (see e.g. Forster, 1951; Harriman and Rockett, 1973; Watkins, 1973) and are currently being interpreted in terms of the proximity of the excited state to higher charge-transfer states involving the anion (Watkins, 1973), thereby bringing them into line with charge-transfer mechanisms proposed for other auenchers (e.n. Weller, 1961: Carroll et ul., 1973). In some cases

158

J. F. IRELAND AND P. A. H. WYATT

however proton quenching, even of B*H+ species, appears to occur (Umberger, 1968; Hussain and Wyatt, 1972).

T h e Rate Constants o f Radiationless Transitions Much more is becoming known about the rates of the physical processes in competition with proton exchmge reactions in excited states. (For an excellent review see Henry and Siebrand, 1973.) The factors which determine the rate constants ( k ) for internal conversion and intersystem crossing are neatly summarized in the “Golden Rule” of time-dependent perturbation theory:

Here h is Planck’s constant, pm is the density of states in the manifold to which the transition is occurring, and lHLn 1’ arises from the overlap of the wavefunctions of the initial and final states and depends upon both electronic and nuclear (vibrational) effects. Thus the much more rapid conversions between higher singlet states than from S 1 to So are related to the large difference in the electronic energy gaps involved, conversions being more rapid the smaller the energy gap; and the lengthening of the lifetime of aromatic molecules by deuterium substitution is explained in terms of the poorer vibrational overlap in the CH bonds with the heavier isotope. Other useful simple rules which emerge are that transitions should be more rapid from S 1 to a triplet state than from S1 to the higher vibrational levels of the ground state (i.e., intersystem crossing from S 1 should be faster than internal conversion) and that intersystem crossing should be more rapid between nn-* and n-n-* states than between states of the same type. Further, since the energy gap between S1 and T, nn-* states is much less than that between those of rn-* states, intersystem crossing is faster between nn-* than n-r*states (unless T2 also is involved).

4. SURVEY OF EXPERIMENTAL RESULTS Reports of pK-values have appeared in increasing number since Weller’s review (1961): and the relationship of this subject to analytical chemistry and biochemistry has led to a wide scattering of

159

ACIDBASE PROPERTIES OF ORGANIC MOLECULES

information throughout the literature. Nevertheless, an attempt has been made here to gather together a reasonably comprehensive collection of pK(Sl)- and pK(T,)-values. It has proved difficult in many TABLE 1 Direction of t h e p K Change o n Excitation for t h e Dissociation Reaction of Various Functional Groups PK(S1) -PK(So)

Reaction

-C02H

+ H+

-S03Hi+-S03H+H+

1-Naphthol pK(S0) = 9.2, pK(S1) = 2.0

Negative

2-Naph thylamine pK(S0) 14, pK(S1)

12.2

Negative

2-N ap h t h ylamine pK(S0) = 4.1, pK(S1) = -1.5

Positive

Xanthone pK(S0) = -4.1, pK(S1) = 1.0

C

1-Naphthoic acid pK(S0) = 3.7, pK(S1) = 7.7

d

1-Napthoic acid pK(S0) = -7.7, pK(S1) = 2.0

e

Positive

1-Naphthalenesulphonic acid pK(S0) = -10'6, pK(S1) = -3.7

f

Positive

Phenylphosphonic acid pK(S0) = -6.3, pK(S1) = -2.2

g

Phenylaxsonic acid pK(S0) = -5.9, pK(S1) = -2.1

s

Nitrobenzene pK(S0) = -11'3, pK(S1) = 2.3

h

Acridine pK(S0) = 5.5, pK(S1) = 10.6

2

Fluorene pK(S0) = 20.5, pK(S1) = -8.5

j

Positive

Positive Positive Positive Negative Positive Positive

Weller, 1958a. Forster, 1950. Ireland and Wyatt, 1972. Vander Donckt and Porter, 1968a. Watkins, 1972a. f Yakatan and Schulman 1972. a

Ref.

Negative

Positive -C02Hl*

Example

Naphthalene pK(S0) = -4.0, pK(S1) = 11.7 Azo benzene pK(S0) =-2*90, pK(S1) = 13.7 g Liedke and Schulman, 1973b.

f 1

'

Winefordner et al., 1971a. Weller, 1957. Vander Donckt et al., 1969b. Vander Donckt el al., 1970. Ellerhorst and Jaffe, 1968.

k 1

TABLE 2.1 Ground and Excited state pK-Values for Substituted Benzenes PK(S1)

Molecule Benzene Toluene Phenol 4-Chlorophenol (see Table 2.2 for substituted phenols) Methoxybenzene 1,2-Dirnethoxybenzene Benzoic acid

2-Chlorobenzoic acid Benzoic acid-4-sulphonate Salicylic acid -H+ Methyl salicylate -H+ 3-Hydroxybenzoic acid -H+ 3-Methoxybenzoic acid -H+ 4-Hydroxybenzoic acid -H+ 4-Methoxybenzoic acid 4-Methoxybenzoic acid -H+ Methyl-4-hydroxybenzoate Methyl-4-hydroxybenzoate -H+

Protonation (P) or Deprotonation (D)

pK(S0)

Forster cycle calculations

P P D

-10.2 -6.8 10.0

5 10 3.6

D P P D P P P D D D D D D D D D D D

9.4 -6.8 -6.5 4.2 -7.67 -7.3

3.2 -5.0 -2.4 6.0 3.18 0 +2, -2 9.1 8-1

3.7 2.9 -8.0

-8.0 -8.4 -8.8

-7.6 4.3 -7.6

8.2 -8.2

Fluorescence intensity measurements

pK(TI) Forster cycle calculations Ref. a U

b 8.0

C

d d C

f

1-5

e

-4.9

g h

h

-7.0 -6.8 -1.0 1*9 -1.1 6.6

0.2 4-6 0.6

2

Benzamide Salicylamide Salicylamide -H+ Salicylanilide -H+ N-Methylanthranilic acid N,N-Dimethylanthranilic acid Diphenylamine 2-Nitroaniline (see Table 2.3 for substituted nitroanilines) 4-Methoxyaniline Phenylphosphonic acid Phenylarsonic acid Benzaldehyde (see Tables 7.1 and 7.2 for acetophenones and benzophenones) Ph thalide Nitrobenzene Azobeniene (see Table 2.4 for substituted azo- and aioxybenzenes) Mason and Smith, 1969. Bartok et al., 1962. Wehry and Rogrrs, 196513. Smith, 1969. Vander Donckt and Porter, 1968b. f Hopkinson and Wyatt, 1967. &' Paul and Schulrnan 1973. Wehry and Rogcrs, 1966. Kovi eta/., 1972b. a

P D D D D P D P P

-2.0 8.3 -2.6 -4.1 4.66 2-61 6 4 0.78

P P P P

-0.23 5.2 -6.3 -5.9

-4.86

P P P

-7.4 -7-98 -11.3

-3.5 1.27 2-3

P

-2.90

13.7

-0.4, -1.3

8-5 -15 10 -1 1.5 -10.18

-0.3 2.1 -5.3 -6.6

1.5

5.3

s k k k 1

I 9 -7.5

1 1

.f

-4

P

a

z% m a

!=

j

n

m

-0.8 -2.2 -2.1

Paul and Schulrnan, 1974. Schulrnan e t al., 1973c.

Trarnrr, 1970. Bridgrs and Williams, 1968. Idoux and Hancock, 1968. Liedkr and Schulrnan, 1973b. p Paul e t al., 1973. Wincfordncr et al., 1971a. Ellerhorst andJaffC, 1968.

-3.3 0

0 0

1.2 -9.3

?-4 r

E

2

v,

0

0 0

P

2

n

5

r W

n

5m w

c,

m

c

)"

c n

TABLE 2.2

N

Ground and Excited-state pK-Values for Substituted Phenols PK(S1)

Substituent

H

3-F 4-F 2x1 3-C1

4-C1 3-Br 4-Br 2-CH3 3-CH3 4-CH3 2-CzH5

Pro tonation (P) or Deprotonation (D) D D D D D D D D D D D D D D D D D D D D D D D

pK(S0)

9.99 10.0 10.0 9.2 9.8 9.91 8-5 9.1 9.13 9.4 9.42 9-39 9-03 9.3 9-36 10.28 10.09 10.09 10-26 10.26 10-16 10.2

Forster cycle calculations

Fluorescence intensity measurements

3.62 4.0

pK(T1) Forster cycle calculations Ref. a

8.5 3.7

C

8.5

3% 3.5 4.4 3.3 4.0 3.0 3.5 3.2

b b a

8-7

b a a

7-6

b a

8.0

2-6

C

7.8

2% 2.9 3.1 5.3

b

b a

7.7

b b a

4.2 4.0 4-1 4.3

8.7

b a

8.6 3-7

4.5

b C

a

3.3

C

3-OCH3

D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D

9.9 10.07 10.0 10.21 9.98

4.5 4.1 4-3 4.3 5.2

9.7 9.65

2.7 4-6

a

b a

b a

4.4

C

a

8-4

3 -4 10.20 10.21

4.7 5.6

C

a

8.6 4-1

9.54 10-13 9.92

4.4 5.3 2.9 3.0

9.82 9.4 9-4 10.0

3.0

9.03 8.35 9.1 10.5 9.0 8.67 9.13 9-37 9-28

2.4 1.7 -0.6

C

a C

a

3 -3 3.8 3-1 3.4

2.1 2.3 2.6 2-3

b

b a

3.4

0.0 -4.0

b C

8.5 8 *4 3 -3

9.83

b

C C

C C

1.0 8 -0 9-2 7.3 7.2 7.6 7.9 7.9

b b d d d e e e e

TABLE 2.2 -continued PK(S1)

Substitusnt

4-SOCH3 3-SOCH3 4-Sc6H~ 3-SC6H.j 4SCH3 3-SCH3 4S(CH,);Cl3-S (CH3)lC14-CHzOCH3 2-OCH3 -4-CH20H 2-OCH3-4-CH2OCH3 4-CHOHCH3 4-CHOCH3CH3 2-OCH3-4-CHOHCH3 2-OCH3-4-CHOCH3CH3 2-OCH3-4-CHOHCzHS a

Bartok et al., 1962. Wehry and Rogers, 1965b. Avigal et nl., 1969.

Protonation (P) or Deprotonation (D) D

D D D D

D D D D D D D D D D

D

pK(S0)

Forster cycle calculations

9.36 9.64 10-16 10-46 10.53 10.37 8.84 9.22 9.63 9.84 9-79 9.80 9.79 9.84 9-75 9.82

2.4 2.8 4.2 4.4 4.4 4-4 2.0 2-2 3.1 49 4.3 2.7 3-2 4.8 4.4 4.8

Winefordner et al., 1971a. Wehry, 1967. f Konschin et al., 1973.

Fluorescence intensity measurements

pK(TI ) Forster cycle calculations Ref.

7.6 7-9 8.8 9 -2 9.1 9.0 7.6 7.6

e e e e e

e e e

f f

f

165

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

cases to assess the experimental reliability of all the values quoted, and little discrimination has therefore been applied at this stage. Although some figures will therefore be subject t o revision, their inclusion here may stimulate further work. Table 1 is included for quick reference and shows the expected direction of pK-change on excitation t o S, for various functional groups. Molecules with a negative ApK [i.e., p K ( S I ) - pK(S,)] become stronger acids (weaker bases) in the excited state and those with a positive ApK stronger bases (weaker acids).

The Tables Tables 2- 8 contain pK-values for a range of aromatic compounds. The compounds are mostly classified according to their ring systems, but carbonyl compounds are grouped together for convenience. It has proved impossible t o include some material in tabular form and a few molecules with more than one ionizable site have not been included (see e.g. Hercules and Rogers, 1959; Ellis and Rogers, 1964; Mayer and Himel, 1972). Where such polyfunctional molecules are included, the tables may be ambiguous as to the position of protonation o r dissociation to which the quoted pK-values refer. For example, the excited state dissociation of protonated hydroxyTABLE 2.3 Ground and Excited-state pK-Values for Substituted 2-Nitroanilines'

Substituent

Protonation (P) or Deprotonation (D)

PK(S0)

PK(S 1 1 Forster cycle calculations

P P P P P P P P P P P P

-0.29 -0.44 -1.05 -2.25 -4.33 0.43 0.77 0-74 -1.48 -1.48 -2.49 -0.09

-4.86 -5.08 -3.75 -7.39 -13.28 -2.96 -2'05 - 1.40 -3.15 -3.38 -7.86 -2.56

H 4-F 4-Br 4-CF3 4-NO2 4-CH3 4-OCH3 4-OC4Hg 5-C1 5-Br 5-NO2 5-CH3

'Idoux and Hancock, 1968.

166

J. F. IRELAND AND P. A. H. WYATT TABLE 2.4

Ground and Excited-state pK-Values for Substituted Azo- and Azoxybenzenes'

Substituent

Protonation (P) or Deprotonation (D)

Azobenzene

4-OC2H5 4-OCH3 4-CH3 3-CH3 4-Br 3-Br 4-COCH3 4-CN 3-NO2 4-N+OH(CH3)2 4-OH Azoxybenzene CX-4-CH 3

ol-4-OCH3 CX-4-OCzHs CX-4-NO2 e-4-Br &-4-N+OH(CH3)2 P-4-CH3 P-4-OCH3 P-4-CI P-4-Br P-4-N+OH(CH3)2 P-4-N+H(CH3 ) 2

P P P P P P

P P P P P P P P P P P P P P P P P P P P

PK(S0 1

-2.90 -1.28 -1.36 -2'35 -2.70 -3.47 -3.83 -3.98 -4'52 -4.63 -4.70 -4.65 -1.02 -6.45 -6.04 -6.10 -6.04 -9.83 -7.01 -8.41 -6.16 -6.1 5 -6.96 -6.94 -8.00

-8.02

PK(S1) Forster cycle calculations

13.7 13.9 14.0 12.8 12.5 12.8 11.4 10.0 9.6 10.6 8.0

10.1 14.0 3-8 2.4 4-6 3'3 -7.3 2-4 1-5 1-6 5.0 4.3 2.6 2.0 1.8

'Ellerhorst and Jaffb, 1968. quinolines involves loss of a proton from the hydroxy-group to form a zwitterion, the nitrogen remaining protonated. This is not immediately obvious from the tables but is discussed in the text. To avoid confusion, for most hydroxyquinolines the cation-zwitterion equilibrium is the only one for which pK-values are quoted; for the other equilibria recourse t o the original paper is necessary. Since the Forster cycle provides the simplest and least timeconsuming route to pK, it is not surprising that most values have been determined in this way, nor that there are many more pK(SI) entries than pK(T,). In all cases pK-values are quoted as they appear in the original papers, although determinations of this kind cannot be reliable to better than -0-2 pK units (corresponding t o an error of

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

167

1 nm at about 300 nm). Directly determined pK(T, ) values are still rare because of the relative complexity of the techniques involved. No distinction has been made in the tables between pK(S,)-values estimated from the mid-point of fluorescence titrations and those which depend upon a more detailed analysis of the rate processes involved.

Limitations of the Forster Cycle

The Forster cycle is described briefly in the introduction and the experimental measurements necessary to perform the calculation are outlined in Section 2. It is obvious from the tables that agreement with the directly measured pK(S,)-values is not always good. The shortcomings of the Forster cycle have often been discussed (Jaffi and Jones, 1965; Wehry and Rogers, 1965a; Vander Donckt, 1970; Winefordner et al., 1971b; Schulman, 1971a; Rosenberg and Brinn, 1972; Pace and Schulman, 1972) but should be kept in perspective: like many other rough rules in physical organic chemistry, it can serve as a good guide even when it is not absolutely reliable quantitatively. The most obvious experimental difficulties in its application are the location of the 0-0 bands of the lowest transitions and the allowance for any alteration in the equilibrium solvation sheaths of the molecules upon excitation. An indication of solvation effects is often obtained from a marked Stokes shift of the fluorescence band, which increases the spacing between absorption and fluorescence maxima and removes their overlap in the 0-0 region. The contribution of this phenomenon in the Forster cycle can sometimes be taken into account by comparing the spectra of both the BH' and the B forms in polar and nonpolar solvents, a mixture of an acid such as CF3CO2H and hexane or other nonpolar solvent being used to obtain the BH+ spectrum (Winefordner et al., 1971b; Schulman and Capomacchia, 1972; Paul and Schulman, 1974); but most commonly it is assumed that averaging the frequencies of the absorption and fluorescence maxima will eliminate such solvation effects. It seems reasonable to expect at least some compensation in the averaging procedures because, when solvent relaxation occurs after both excitation and emission, it must in both cases be accompanied by a decrease in free energy. Furthermore, the two situations are not independent: a ground state solvent configuration around an excited

TABLE 3 Ground and Excited-state pK-Values for Naphthalene Derivatives ~~

PK(S1)

Molecule Naphthalene 1-Naphthol 2-Naphthol

2-Chloro-1-naphthol 4-Chloro-1-naphthol 1-Naphthol-2-sulphonate 1-Naphthol-4-sulphonate 1-Chloro-2-naphthol 1-Bromo-2-naphthol 6-Bromo-2-naphthol 6-Methyl-2-naphthol 7-Methyl-2-naphthol 2-Naphthol-5-sulphonate 2-Naphthol-6-sulphonate 1,2-Naphthalenediol 1,5-Naphthalenediol 1,7-Naphthalenediol 1&Naphthalenediol

~

PK(T1)

_

_

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref. P P D D D D D D D D D D D D D D D D D D D D D D

-4.0 -4.6 9.23 9-20 9.46 9.5

11.7 9.0

9.45

3.0

7.76 8,75 9.6 8.3 7.97 7.89 9-23 9.70 9-64 9.18 9-10 8.1* 8.1* 10.3* 9.8*

0.7 2.9 0-4 -0.1 1-7 1.4 3.1 4.4 3.7 0.53 1.6

2.0 2-1 2.5 3.1

2.5

a

b

2.5 G1.8 2.8 I

C

d C

7.7 3.0 2.8 5 2.8 G1.66 G1.43 2.9 2.9 1.3 1.9 3.49 3-24 0.73 1.66 2.4 2.9 3.1 2.9

8.1

e

f b R

d d

7.9 7.2

7.5 7.1

h h d d d d d 1

C

j j

i i

2,3-Naphthalenediol 2,6-Naphthalenediol 2,7-Naph thalenediol 1-Naphth oic acid

1-Naphthoic acid

2-Naphthoic acid

2-Naphthoic acid

Ethyl-1-naphthoate Ethyl-2-naphthoate 1-Naphthalenesulphonic acid 2-Naphthalenesulphonic acid 1-Naphthylamine 2-Naphth ylamine

2-Naph thylamine-6-sulphonate

D D D D D D D P P P P D D D D P P P P P P P P D D D D P P P P P P

9*9* 9.3* 9.2* 3.7 3.7 3.73 3.7 -6.9 -7.70 -7.7 -7.9 4.2 4.2 421 4-2 -6.9 -7.68 -7.68 -8.0 -8.6 -8.2 -10-6* -8.3*

14 4.07 4.1 4.08 4.1 3.72

3.2 3.4 2.5 4.6

10-12 7.7 10.0 9-6 1.5

-0.2 -0.2 - 0.8 10-12 6.6 11.5

8.2 1.5 -0.35 -0.35 1.5 -0.4 -3,4

3-8

i i i

e k 1

m n

0 2.0 -0.1, 1.6

0

P 4.2

4-0

m e

k 1 m n 0

0

0.7 1.4 -1.0, 0.9 -2.5, 0.9 -3.7 -5.1 13.5 12.2 12.3

P m 4

4 r

11-86 -1.5

-2 0.6 -8.1 0.0

.o

-2.9 0.64

-0.02

3.1

3.3

e S

U

cd

;a

gM

2 m

v)

0 I !

0 ;a

n > 3 n

TABLE 3-continued PK(S1)

PK(T1)

Fluorescence Flash Protonation (P) o r Forster cycle intensity Forster cycle photolysis Molecule Deprotonation (D) pK(So) calculations measurements calculations measurements Ref. N,”-Dimethyl- 1-naphthylamine P 4.9 2.9 2.7 e 1-Naphthaldehy de P -6.63 3.8 1.o 0 P -6.6 1.2 2.2, -3.2 W 2-Naphthaldehyde P -7.08 3.58 1.0 0

P

-7.0

2.0

1-Naphthamide 8-Naphthamide

P P P P

-6.22 -6.2 -6.16 -6.3

5.79

Methyl 2-naphthyl ketone

P P P

Methyl 1-naphthyl ketone

-2.35

1.2 5.27 2.9

2.47, 2-87

P

2.3 1.5

W

1.7, -2.9 1.5

W

Vander Donckt et al., 1970. Mason and Smith 1969. Weller, 1958a. Rosenberg and Bnnn, 1972. Jackson and Porter, 1961. f Stryer, 1966. g Ofran and Feitelson 1973. Henson and Wyatt, 1974. Weller, 1952. J Derkacheva, 1960i. Vander Donckt and Porter, 1968a. Wehry and Rodgers, 1966. Kovi and Schulman 197313.

0

2.3

W

2.7 2-74

X

Y

2.5

0

a

2.79

Y

?

* Calculated from pK(S1)-valueand ApK(S1 - SO)obtained using the Forster cycle. a

0

Weller and Urban, 1954. Hopkinson and Wyatt, 1967. p Watkins, 1972a. Kovi and Schulman, 1972. Yakatan and Schulman, 1972. Rosebrook and Brandt, 1966. Forster, 1950. Seliskar and Brand, 1971. Schulman and Capomacchia, 1972. Capomacchia et al., 1973. Watkins, 1971. Watkins, 1972b.

TABLE 4

Ground and Excited-state pK-Values for Anthracene Derivatives

Molecule

PK(S1) PK(T1) ____ Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref.

Anthracene 1-Anthroic acid 2-Anthroic acid 9-Anthroic acid 1-Aminoanthracene 2-Aminoanthracene 9-Aminoanthracene 9-Anthraldehy de 9-Anthrol et al., 1970. Vander Donckt and Porter, 1968b. Schulman et al., 1973a. Wemer and Hercules, 1970. Pace and Schulman, 1972.

a Vander Donckt

P D D P D P D D P P P P P P P 1)

17.0 3.3 6.9 3.7 8.3 3.7 -2.6, 1.6 -8.5 6.6 4.2 -6 to -10 0.6 3.0 6.5 3.0 6-2, 3.5 --7.4 -0.5 -5.5 3.3 -5.4 3.5 -4.4 3.4 -6.1 2.7 -4.8 1 12.2 -5.1 1.8

10.3

a

5.6

b

'md %m

b C

1.7

C

6.0

b C

4.2 -1.4

b d e

3-5 3.6 3.0

2.1 3.2 3-3

f,g ft g b

f h

1-3 0.6

f Grabowski etal., 1966. 'k Rotkiewicz and Grabowski, 1969. Hopkinson and Wyatt, 1967. Young and Schulman 1973.

2

i

0

P

52 2 n

TABLE 5 Ground and Excited-state pK Values for other Polycarbocyclic Aromatic Compounds and their Derivatives

Molecule Phenanthrene Chrysene Tripheny lene 1,2,5,6-dibenzanthracene Fluorene 9-Phenylfluorene 9 -Ethoxy carbonyl flu ore ne 9-Cy anofluorene Azulene 1-Aminopyrene 4-Aminopyrene 3-Acetylaminop yrene5,8,1O-trisulphonate 3-Methylaminopyrene5,8,10-trisulphonate 1-Hydroxypyrene 3-Hydroxypyrene5,8,1O-trisulphonate Pyrene-1-carboxylic acid 2-Phenanthrylamine 3-Phenanthrylamine 4-Phenanthr ylamine 9-Phenanthrvlamine et al., 1970. Vander Donckt et al., 196913. Hopkinson and Wyatt, 1967.

a Vander Donckt

-

PK(S1) PK(T1) .~ Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0 ) calculations measurements calculations measurements Ref. P P P P D D D D P P P P P

- 3-5 -1.7 -4.6 2.2 20.5 18.6 10.0 11.4 -0.8 -0.92 -1.1 2.8

D

13.9

6.7

P D

8.7

3.2

D D P P P P

7.3 4 4.06 3.90 3.18 3.50

19.5 8.2 0.6 -2.4 -8.5 -10.7 -11.8 -12.4 9-9 -31.2

-5.0 -10.2 --3.9 -10.5 5 4.2 6.0 5.0

U U

a a

b b b b C

d e

-5-74 -5.8

3.5

-2

f
6-9 -6

Mason and Smith, 1969. Hagen et ul., 1967. f Vander Donckt et al., 1969a.

1.0 8.1 -2.5 -1.6 -4.4 -3.2

i

1.38

R Weller and Urban, 1963. Weller, 195813. Weller. 1958a.

5.2 0.7 1.7 -0.3 -0.9 j Tsutsumi et al., 1971.

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

173

molecule and an excited state configuration around a ground state molecule may well exhibit a similar tendency t o change. In such cases, provided that entropy effects are not large, the enthalpy relationships can be represented schematically as in Fig. 9. The quantity required in the pK application is the difference between the relaxed ground and excited states (i.e. between bottom left and top right of the figure), which is then clearly obtained from the average of the absorption and fluorescence transitions. Schulman and his co-workers emphasize, however, that the energy of relaxation in the two states may not be the same and are therefore sceptical of the validity of the averaging procedure. When a molecule increases in dipole moment upon excitation (Lippert, 1957), relaxation in the excited state with a net ordering of polar solvent molecules in the new electric field might involve a reduction in both energy and entropy. Relaxation in the ground state after emission, though still requiring a free energy decrease, might be accompanied by an increase in entropy when the solvent molecules are no longer constrained in their orientation towards the charge centres. Entropy effects, though probably small, could therefore

relaxation

absorption

fluorescence

I I

reloxo t ion

Figure 9. Diagram showing the relationship of the enthalpy difference between relaxed ground and excited states when the enthalpies of relaxation are similar.

work in opposite directions upon relaxation in the excited and ground states and hence spoil the free energy compensation expected from averaging even if the enthalpy changes were as in Fig. 9. The basic assumption of the Fiirster cycle that AG** - A c e = AH*O- AHe can be checked in favourable cases by modern

TABLE 6.1 Ground and Excited-State pK-Values for Nitrogen-containing Heterocyclic Compounds and their Derivatives PK(S1)

Molecule 2-Aminop yridine

2-Amin~pyridine-H~ 3-Aminopyridine

3-Aminop yridine-H' 4-Aminopyridine 4-Aminopyridine-H+ trans-2-St yrylpyridine cis-2-Styr ylpyridine trans-3-St yr ylpyridine trans-4-St yrylpyridine (See Table 6.2 for substituted styrylpyridines) 2-(2’-Thieny1)pyridine 2-(3‘-Thieny1)pyridine 3-(2’-Thieny1)pyridine 3-(3’-Thieny1)pyridine 4-(2 ‘-Thieny1)pyridine 4-(3’-Thieny1)pyridine 2-Phenylpyridine

PK(T1)

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref. P P P P P P P P P P P P P P

6.86 6-9 -7.6 5.98 6.0 -1.5 9.2 -6.3 5.3 4.9 8 5.20 4-76 5.02 5.73

P P P P P P P

3-73 4.72 4.4 7 4-87 5.50 5.65 4-55

8.95 8.8 -19.8 11.2 10.8 -15.8 11.9 -8.6 10.67 11.4 12.04 12.4 11.76 13.3 9.5 11.9 12.9 14-0 12.1 17.4 10.0

a b b a

-8.5

b b b b C

d C

12-3

d C

d

e e e e e e e

3-Phenylpyridine 4-Phenylpyridine 3-Hydroxypyridine

P

P D P P

3-Hydroxy-N-methylpyridine

P

2-Hydroxypyridine 2-Hydroxy-N-methylpyridine Pyridoxol Pyridoxal (hemiacetal) Pyridoxamine Pyridoxal-5-phosphate (aldehyde hydrate) Indole

P P P P P

1-Methylindole 2-Methylindole

3-Methylindole Carbazole N-E thylcarbazole Indolisine Quinoline (See Tables 6.3 and 6.4 for substituted quinolines) Acridine

1-Azaphenanthrenet

P D P P D P D P D P P P P P

4.85 5.38

12.2 14.5

5-37

-2.24

4.96 0.75 0.32 5.05 4.20 3.32

-2.29 -2.52 -3.48 -2.36 -2'53 -4.25 -3.28

-3.5 -2.3 2 -0.26 -4.5 5 21.1* <-6 <-6 3.94 5.1 4.94 5.45 5-5 5.5 5-1 5.2 4-75

10.3 10-6 11.5 10.5

e e

f

4.37 -3.0 9.5 -3.0 -3.3 -4.0 -2.2 -3.3 -4.1 <-2.5 12.3 2-1 1.8 12.6 1.2 12.0 3.6 11.9 -1.3 -1.3 -6 5.8

1

10.65

m

5.4 6.4 6-8 7 -4

5.6

k n n

7.0

0

P

5 W v,

F

Crr

TABLE 6.1 --continued PK(S1)

Molecule 2-Azaphenanthrene 3-Azaphenanthrene 4-Azaphenanthrene 9-Azaphenanthrene 10-Phenyl-1-azaphenanthrene 10-Phenyl-3-azaphenanthrene Pyrazine Pyrazine-H+ Pyrimidine Purine (See Table 6.5 for substituted purines) Phenazine Phenazine-H+ Dipyridy 1 1,5-Diazaphenanthrene 1&Diazaphenanthrene 4,5-Diazaphenanthrene (0-phenanthroline)

PK(T1)

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref. P P P P P P P P P P P P P P D P

4.6 4.75 5-38 4.2 4-6 3.91 4.6 4-04 4.08 0.6 0-65 -5.8 -5.8 1.2 8-93 2-39

1.21 -4-3 4.5 1 4.0 4-0 4.85 4-85 4-85

9.4 10.5 9.5 9.6 11.1 9.6 15.0 14-9

4.6

0

6.2 7.2 7.2 5.7 6.1 5.7 1.4 11.5 -1.2 -1 1.1 --3.0 5-1 5.1

6-0 4.1

12.7

4.0 5.7 9.50 4.6 5.3 6.0 6.7

P P

n

5.7 5.7

0

P 0

P P

n

4 n 4

n r r

4

7

4,5-Diazaphenanthrene-H+ 9,lO-Diazaphcnanthrene

lO-Nitr0-4,5-diazaphenanthrenc 1O-Methyl-4,5-diazaphenanthrene 9,1O-Dimethyl-4,5-diazaphenanthrene 3,6-Dimethyl-4,5-diazaphenanthrene 4,lO-Diazachryzenc 4,1O-Diazachryzene-H+

P P P P P P P P P P

7.36

5.05 -1.4 -1.4

t -3.8

-2.2

2.2 3.23 5.1 1

9.5 5.50

5.44 5.77 3.22 1.44

8.42 8.8 1

-4.3 - 0.07

9

V

E!

W

7.92

3.8 2.1

U

:

t t

%

t

P

Ir!

w

t

$

3.6

X

2-1

X

2

m

F;

* Calculated from pK(SI)-valuc and ApK(S1 --SO)obtained using the Forster cycle. t Azaphenanthrenes are numbered using the IUPAC numbering of phenanthrenr. Weisstuch and Testa, 1968. Schulman ct a]., 1971a. Doty et al., 1969. E’avaro ctal., 1973. Bouwhuis and Janssen 1972. f Rosebrook and Brandt, 1966. ,x Bridges rial., 1966. Vandrr Donckt, 1969b. I Capomacchia and Schulman, 1972b. j Mason and Smith, 1969. Jackson and Porter, 1961. Capomacchia and Schulmdn, 1973. a

‘ h

‘

Weller, 1957. Grabowska and Pdkula, 1966. Vandrr Donckt et al., 1969a. P Favdro rial., 1971. Grabowska and Pdkula, 1967. Aaron and Wincfordncr, 1973. Grabowska and Pakula, 1969. Aditya and Lahiri, 1971. Brincn ct al., 1963. Winrfordncr c t al., 1971b. Ballard and Edwards, 1964a. Bulska c t al., 1972.

‘ ”

m

% 0

P

0 9

2

n

5r m

n C r

5

TABLE 6.2 Ground and Excited-state pK-Values of Substituted Styrylpyridines PWl)

Molecule

Fluorescence intensity measurements

Protonation (P) or Deprotonation (D)

PK(S0)

Forster cycle calculations

P P P P P P P P P P P P P P P P

5.3 5.20 6-40 5.52 5.72 5.78 5.60 5-44 5.31 5-17 5.13 5.19 5.07 5.13 4.90 4.5 6

10-67 12.04 17.00 19-12 13.44 17.25 12-15 15-14 11.26 13-49 11.01 10.78 11.17 10.35 9.10 4.05

a a a a a a a a a a a a a a a a

P P

5.97 5.65

12.22 10-79

a

Kef.

2-St yrylpyridines substituents 4'-NMe2 ( ctrans is

4'-Me

{

3'-Me trans 4'-NHCOMe trans 4'-C1 trans 3'-OMe trans 3'-C1 trans 4'-NO2 trans trans-5-Ethyl-2-styrylpyrydines substituents 4'-OCH3 4'-CH3

H 4'-N02

P P

5-62 4-96

10.18 3-86

a a

ij

trans-3-St yry lpyridines

a

cm

UJ

substituents

H 4'-CH3 4'-OCH3 4:-C1 4 -Br 4'-I 4:-N02 4 9 H 3 12 4 -NH2 4'-OCH3 4'-CH3 3'-CH3 H 4'-C1 3'-OCH3 3'-C1 4'-N02

b

P P P P P P P P P P P P P P P P P

4.76 4-77 4.73 4.74 4.74 4-73 4.44 5.85 5.62 5.1 7 5.17 4.99 5.02 4-88 4.93 4.68 4.47

12.4 12.6 12.9 12.4 12.5 12-3 1.6 17.10 15.44 13-55 12.82 11.98 11.76 11.20 13.02 11-66 5.64

12.3 12-5 13-0 12.3 12-4

b b b b b b b

a a a a

;1 m

2m

v,

0

w

0

>

z

a

n

a

0

a a a

a

x

r m n

5

W UJ

Doty et al., 1969. Favaro et al., 1973. Y

co

TABLE 6.3

0

Ground and Excited-State pK-Values and Quinoline and its Derivatives

___ PK(S1) Molecule Quinoline Isoquinoline 2-Methylquinoline 4-Methylquinoline 6-Methylquinoline 8-Meth ylquinoline 6-Methoxy quinoline 8-Methoxyquinoline 8-Methoxyquinoline-Hi 6-Nitroquinoline 8-Nitroquinoline 2-Aminoquinolinc 2-Aminoquinoline-H' 3-Aminoquinoline 3-Aminoquinoline-H+ 2-Methyl-4-aminoquinoline 2-Methyl-4-aminoquinoline-H+ 5 -Aminoquin oline

PK(T1)

___--

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref. P P P P P P P P P D P P D P P D P P P D P P P

5-1 4.94 5.40 5.43 5.34 4.89 4.65 5.1 8 5.14 6.6 2.10 1.97

5.8 5.8, --1.6 6-1 5.9, -1.0 5.8, -1.4 5.1, -0.8 4.6, 9.7 5.2 5.1 --3.9 4.4

7.27 -10.5

6.9

5.0 4.9 6 -0.4

<5.0 10.3 -4.6

5.46 5.63

10.9 18.0

6.0

4 a

b b b b b b b b ,c

-7.0 2.1 2.4

C

d d

12.7

f

-5.7 12.1

f

e

.?

R e

-5.6 12.6 -3.4

; f h e

?"

Er

9 Z

U

9

2

U

5-Aminoquinoline-H+

P P P 6-Aminoquinoline 6-Aminoquinoline-H+ P 7-Aminoquinoline P 7-Aminoquinoline-H+ P 8-Aminoquinoline P P 8-Aminoquinoline-H+ P Pamaquine-(H+)* P 2-Methyl-6-dimethylaminoquinoline P 2-Methyl-6-dimethylaminoquinoline-H+P 8-Amino-6-methoxyquinoline-H+ P Quinoline-2-carboxylic acid-€I+ P P Meth y lquinoline-2-carboxy late-H+ Quinoline-8-carboxylic acid-H+ P Methylquinoline-8-carboxylate-H P Quinoline-4-carboxylic acid-H* P Methylquinoline-4-carboxylate-H+ P 2-Phenyl uinoline-4 carboxylic P acid-H Me thyl-2-phenylquinoline P 4carboxylate-H+ 8-Mercaptoquinoline-H' D 2-H ydroxyquinoline D 2-Hydroxyquinoline-H+ D 3-Hydroxyquinoline I1

0.66 0.49 5.6 1.2 6.7 0.0 3.99 3-92 -0.1 3 -1.3 6.8 1.9 --0.4

--16.6 -16.1 12.5, 12.7 -11.2, -15'4 14.1, 10.7 -14.5, -12.8 10.0 12.1 --12.6

-

4-Hydroxyquinoline-H+ 5-llydroxyquinolinc.-II+ 6-Hydroxyquinoline

D D D D D

15.2 -1 0 --6.4 --7.0 -3.5 -2.8 -6.3 -7.1

9

3-Hydroxyquinoline-H+

11.9

-8.0

19.0, 12.9 16.0. -17.0

+

1)

7.7 1.2

-5.4 -6.3 -1.5 12.9 -2.2

2.1 8.03 8.03 5.52 5.52

3.6 4.5 -0.3, --0.5 -0.1, -0.7

6.49 8.87

-3.6 4.0, -1.9

tm

71

t-

0

-0.2 --0'2 0.6 -6.0

:

1 rn n

G L

E

TABLE 6.3-continued PK(S1)

Molecule

PK(T1)

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(So ) calculations measurements calculations measurements Ref.

6-Hydroxyquinoline-H+ 7-Hydroxyquinoline

7-Hy droxyquinoline-H+ 8-H ydroxyquinoline-H+ 5-Nitro-8-hy droxyquinoline 7-Nitro-8-hydroxyquinoline (See Table 6.4 for other substituted 8-hydroxyquinolines) 5-Hydroxyisoquinoline-H+

D D D D D D D P D

7.03 7-03 8.67 6.01 6.0 1 6-60 6-45 3.20 6.40

D

6.84

-2.7, 0.8 -1.8 3.5, -2.7 12.8, -1.4 -1.5 13-3

4

-4.5 -3.2

F

-2-2 -2.1 -6.5 4-1 5.3 6.4

1.0

-4.7

'Jackson and Porter, 1961.

j Abate and Schulman, 1972.

Capomacchia and Schulman, 1973. Ballard and Edwards, 1964b. Winefordner e t al., 1971a. Schulman, 1971c. f Kovi ct al., 1972a. I: Capomacchia and Schulman, 197%. Schulman and Sanders, 1971. Abate et al., 1973.

Schulman et al., 1972. Zalis etal., 1973. rn Schulman, 1972. Schulman et al., 1971b. Haylock etal., 1963. p Mason et al., 1968. 9 Fernando and Schulman, 1968. Aaron e t a[., 1972.

4

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

183

TABLE 6.4

pK(SI )-Values (Fluorescence Titration) for the Reaction

X

X

Y OH

H

Substituent H 5-F 5-C1 5-Br 5-1 5-SO3H 5-SCN 7-F 7-C1 7-Br 7 -I 7-SO3H 5-F-7-C1 5-F-7-Br 5-F-7-1 5-Cl-7-Cl 5-Cl-7-Br 5-C1-7-1 5-Br-7-C1 5-Br-7-Br 5-Br-7-1 5-1-7421 5-I-7-Br 5-1-7-1 5-1-7-SO3H 5-SO3H-7-SO3H a

Bratzel et al., 1972. Ballard and Edwards, 1964b.

experimental techniques in several temperatures. AG*e and their difference can be Where pK* has only been

H

PK(S1) -7.65 -7.70 -11 -9 -8.60 -7.8 -7.30 -6.7 -9.30 8.95 -9.40 -9.20 -8.65 -5.90 -9.15 -8.80 -8.60 -7.95 -8.75 -8.55 -8.35 -7 -8.40 -8.20 -7.60 -7.50 -7.10 -6.95 -6.40 -6.1 5 -9.60

Ref. a

b C C

a C

a C

a a a a

a a a

a a a a a a C

a

a C

a a a a a a

Fernando and Schulman 1967.

which the true pK*-value is measured at and AH*0 are then directly determined compared with that in the ground state. determined directly at one temperature,

184

J. F. IRELAND AND P. A. H. WYATT TABLE 6.5 Ground and Triplet-state pK-Values of Purine and its Derivativesa

x N

Y

k

-

j

AN

N%;

+H+

L

N -

-H+

Y

+H+LH AN x

AN

N

I

7 Y

I H

H

Molecule Purine 8.93 6-Aminopurine 9-83 6-Chloropurine 7.68 6-Bromopurine 7.34 6-Methylpurine 9.02 6-Benzylaminopurine 9.5 6-Methylmercaptopurine 8.74 2-Amino-6-Methylmercaptopurine 9.55 2,6-Diaminopurine 10.77 ~

5.1 7-1 6.0 4.4 5.2 7-0 8.1 7.4 8.0

2.39 4.25 3.35 2.6 4.0 0 3.5 5-09

5-1 6.5, 5*3b 2.1 4.5 6.9 -0.8 1.2 7.6

~

a Aaron and Winefordner, 1973; pK(Tl)-valuesare calculated using the Forster cycle. Longworth et al., 1966.

comparison with the B-BH' frequency spacing using (40) g'ives an indication of the change in AS0 upon excitation if the frequencies

AS*0

--

AS" = Lh(vB

-

vBH+)/T- 2.303 R(pK* - PK)

(40)

are sufficiently sharply defined. Using temperature effects, Weller (1952) showed that there was practically no difference in AS" for the ionizatiqn of 2-naphthol in the So and S, states, despite the very large change in pK and the obvious difference in solvation adjustment between naphthol and the naphtholate ion, as shown by the absorption-fluorescence spacing. A probable limit to the error introduced by the change in ASe upon excitation can be estimated in the following way. Bell (1959) has summarized some of the information available for ground state changes of the types (41) and (42), in which several acids are RCOzH + MeC0;

=+

RCO; + MeCOzH

(41)

TABLE 7 Ground and Excited-State pK-Values for Aromatic Carbonyl Compounds 1. Acetophenones PK(S 1 )

Substituents

H 4-CH3 4-CH30 4-OH 4NHz 4NH3 4-F 4-C1 4-Br

*

Protonation (P) or Deprotonation (D)

P P P P P P* P P P P

Refers to protonation of the amino-group. a Paul e t a l . , 1973. Ireland, 1972.

PKPo 1

-6.5 -6.1 -5.5 -4.8 -4.7 z0.5 -6.1 -6.5 -6'5

Forster cycle calculations

-4,7 -3.2 -4.4 -1.3 -2.3 1.8 pK(S0) + 0.7 -8.1 4.2 -3.6 ~

Fluorescence intensity measurements

-3.8 >-3.4

PK(T1) Forster cycle calculations

1.0 0.9 2.6 3 -4 2-0 -6.8 pK(S0) + 7.4 0.6 1.9 2'1

Ref. a

b b b b b b b b b

TABLE 7-continued Ground and Excited-State pK-Values for Aromatic Carbonyl Compounds 2. Benzophenones ~~

~~~

~~~

PK(S1)

Substituent

H 4-OH 2-OH 4-C6H5 4-NHz 4-NH; 4-Cl-4-Cl 4-CH30-4-CH30 2-OH-4-CH3 2-OH-5-CH3 2-OH-4-OH 2-OH-5-OH 2-OH-4-OCH3

PK(T1)

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref. P P D P D P P P* P P P D P D P P P D P

-5.7

-4.2

-0.4

a

1-5 6.5 -5.0 10.83 --6.27 X0.5 -6.5 -4.4 11.10 -6.09 11-30 -6.18 -4.98 -6.25 10.86 -5.1 6

4.8 0.3 3-78 1-43 pK(S0) + 7.4 -1.1 pK(S0) + 0.9 -3.0 4.0 4-20 0.6 1 3.70 3.64 2-12 2.6 7 2-70 2.69

1.5 2.3 4.13 -2.5 7 pK(S0) + 5.3

-6.5 pK(S0) + 5.7

2.1 5.5 5.13 -4.94 4.1 8 -0.7 3 -4.7 7 -6.46 5.41 -1.81

b a a C C

a a a a

a C

C C C

C C

C C

zU9

P

2-OH-5-OCH3 2-OH-4-Cl 2-OH-5-Cl 2-OH-4-NO2 2-OH-5-NO2 2-OH-5-COCH3 2-OH-4-OCOCH3

D P D P D P D P D P D D

* Refers to protonation of the amino-group. a Ireland and Wyatt, 1973.

Rayner and Wyatt, 1974. Hrdlovic et al., 1968; Kysel, 1969.

11.03 -6.31 9.50 -6.63 9.80 -6.68 8.23 -7.49 6.69 -7.41 8.15 9-35

5-59 0.7 1 0.9 3 2.63 2.14 4.09 -3.54 0.09 -8.36 4'22 -0.75 7-06

6-60 -2.54 4.68 -2.44 2-78 -0.6 1

a C

C C C C C

C

-7.49 2'50 -9.50 2.70

C C

z

%m

; m P

=!

E

C C C

0

t>?

3 5

E;1 n

s M

m

TABLE 7-continued

m

m

Ground and Excited-State pK-Values for Aromatic Carbonyl Compounds 3. Other Systems pK(S1)

Substituent

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref.

Xanthone Anthrone Acridone Flavone 4'-Me thoxyflavone Cfialcone 4 -Methoxychalcone 3-Methoxybenzanthrone Phenalenone Benzo [ b ] phenalenone Benzo [c,d] pyrene-5-one Benzo [c,d] pyrene-6-one 4,5-Dime thylphenalenone 4,5-Dimethoxyphenalenone 4,8-Dimethoxy phenalenone Ireland and Wyatt, 1972. Hopkinson and Wyatt, 1967. Kokubun, 1957 and 1958. d Derkacheva and Petukhov, 1973. e Reid et al., 1968. f Schulman et al.. 197313. a

PK(T1)

P P P P P P P P P P P P P P P

-4.1 -5.02 -0.6

-1.2 -0.8 -5.0 -4.25 -0.19 -1.45

0.08 -1.44 0-15 0.08 1.05

3-5 6.15 2.0 6.75 5.87 7-10 7.44 pK(So) + 5.2 5.9 6.3 6-2 7.8 4.2 7-0 4.7

0-96 0.92

6.6

3 -0

U

6,f C

b b b b d e e e

e e e C

ACIDBASE PROPERTIES OF ORGANIC MOLECULES

189

compared with acetic acid or the ammonium ion. At least in respect of their entropy effects upon the surrounding solvent, such reactions may serve as models for the equilibrium (43), the entropy change of which is AS*e - ASe. B*H++ B

+ B* + BH+

(43)

Among the carboxylic acids (the pK-values of which admittedly range over little more than 2 units), T A P (against acetic acid) within the aromatic family only varies from +3.73 kJ mole-' (+0-89 kcal mole-') for p-hydroxybenzoic acid to +8.12 kJ mole-' (+leg4 kcal mole-' ) for p-nitrobenzoic acid, benzoic acid itself having the value +5.35 kJ mole-' (+1.28 kcal mole-'). Though a tendency for T A P to correlate with pK may need further attention in view of the large change in pK upon excitation, these figures suggest that differences in TASO do not contribute more than 1 pK unit (corresponding t o 5.7 kJ mole-' or 1.36 kcal mole-' ). Even if an aromatic ring became so buckled on excitation that it gained the flexibility of an open chain system (which is most unlikely; see Murrell, 1963), the error would still not exceed 2 pK units; equilibrium (44) has T A P = -10.12 MezCH . CH2CHzCO2H + C ~ H S C O ~ MezCH -= . CHzCHzC02-+ C6HsC02H (44)

kJ mole-' (-1.14

1.28 kcal mole-'). A similarly extreme case among nitrogen bases, the reaction of n-C4H9NH3f with C6H, NH,, has TASe = 5-14 kJ mole-' (= -0.18 - 1.05 kcal mole-'), corresponding to a Forster cycle estimate of less than 1 unit too negative for an aromatic base. We conclude therefore that errors from the assumption that AG*e- AGQ= AH*e- AH* will rarely approach 1 unit of pK and their direction may even become more predictable as better ground state models are discovered. -

Some Aspects of Recent Work Substitnsent effects Attempts have been made to correlate displacements of the absorption bands of parent aromatic compounds by substituents with Hammett o-values derived from the effects of the substituents on ground state kinetic and equilibrium properties. For example,

TABLE 8 Ground and Excited-State pK-Values for Other Miscellaneous Organic Compounds PK(S1)

Molecule

PK(T1)

Fluorescence Flash Protonation (P) or Forster cycle intensity Forster cycle photolysis Deprotonation (D) pK(S0) calculations measurements calculations measurements Ref.

Coumarin 4-Methyl-7-hydroxycoumarin 4-Methyl-7-methoxycoumarin Esculin 7-Hydroxycoumarin 4-Hydroxycoumarin War farin-H+ Methaqualone-H+ Chloroquine Rib0 flavin Lumiflavin 9-E thylisoalloxazine 9-Methylisoalloxazine 6,7-Dichloro-9-methylisoalloxazine Tyrosine 3-Aminotyrosine Thionine-H+ Azur A-H+ Azur B-H+ Azur C-H+

P D P P P D D D P P P D P P P P P P P P P P P

-7.4 7.8

-5.3 2.2

a a

-4.1 -4.3

a a

pK(So) + 11.0

-5.0 -5.8

b b 1-2

4.1 -4.6 -10*8* -0.1 0.0 0.0 0.1 2.1 4.4 -0.7

C

1.1 0.4

-6.2 -8.5

-8.2 -8.9 -1.1

d d d e

-9.2 -7.2 11-7 1-7

f

s s s s s

1.7 1.7 1.7 1.3 1.8 -0.4

h

1

6.3 7.2 7.4 7 .O

i,k i i i

Methylene blue Fluorescein monoanion Fluorescein Dichlorofluorescein monoanion Bromodeoxyuridine Thymidine, 5-Bromo-2 -deoxyuridyl(3’ -+ 5’)-thymine Ortic acid Dehydraluciferin Thymine Uracil Yakatan etal., 1972a. Derkacheva and Petukhov, 1973. Fink and Koehler, 1970. Yakatan etal., 197213. Schulman e t al., 1974. f Schulman and Young, 1974. g Schulman, 1971 c. Fayet and Wahl, 1971. f Cowgill, 1971. 1 Faure et al., 1967. a

P P D D P D D D

D1

D2 D D D D

6.7 7.5 7.0

6.7 6.7

6.9 -2’1

2.2

5.0 7.9 9.8 8.1 10.1 1.8 8-7 10.0 9.5

5-2

9.2 10.3 4.6

Bonneau and Stevens, 1972. Kato e l al., 1964. Lindqvist, 1960. Leonhardt etal., 1971. Lindqvist, 1963. Peter and Drever 1971. 4 Herbert and Johns, 1971. Morton e t al., 1969. Burr e l al., 1972.

m n 0

ti

z%

m

n

P P P P

8-9 10.6

-1 3.2 3

I

i

4

cd

z!2 m

r

0,

s

0 P 0

5

>

5n

5r

m

n

s!2

J. F. IRELAND AND P. A. H. WYATT

192

Barker et al. (1961) found such a correlation for Malachite Green cations. It is not necessary, however, that the acid and conjugate base frequency shifts should separately be related to the substituent U-values in this way for a good correlation to be obtained between u and the difference in their frequency shifts, i.e. the difference in their frequency maxima and hence in pK(S,)-pK(S,). Figure 10 shows a plot of Av against u for substituted 5-phenylazotropolones, the acid and base forms of which separately showed very erratic

7 ' J'""

5.5 -

IE

5.0

CH3CO

0

m

0

5 4'5-

\

:/*

I

4.0

c

Figure 10. (BH+-B) wave number difference (b) for the SO-SI transition as a function of i'ammett @values for substituted phenylazotropolones (Griffiths, 1971).

variations with u (Griffiths, 1971; see also Yagupol'skii and Gandel'sman, 1965, 1967; Vander Donckt, 1970). Comparisons with pK( S ), rather than pK( S1) - pK( So), have met with less success, although modified u-constants have given some improvement (Jaffk et al., 1964a, 1964b; Bartok e t al., 1965; Wehry and Rogers, 1965b; Jaffk and Jones, 1965; Idoux and Hancock, 1968; Ellerhorst and Jaffi, 1968). The results in general indicate an increase in the electron attracting or releasing effects of the substituent in the excited state compared with the ground state. In most of these earlier cases pK(S, )-values obtained by Forster's method were used. More recently, using pK(S )-values from fluorescence titration, Favaro et al. (1973) report a better correlation with U + than with u for 4'-substituted 3-styrylpyridines, and Bratzel et al. (19 73) adopt the equation of Taft (1960a) to interpret the behaviour of 8-hydroxyquinolines. Related work includes that of Winefordner et al. (1971a), who conclude that charge transfer to the nitro-group in substituted

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

193

phenols and quinolines is much more important in the S , than in the So or T, states, since the nitro-group affects the acidity of S 1 to a greater extent than either So or T, (see also Aaron and Winefordner, 1973; Ireland and Wyatt, 1973). Further in the case of sulphone, sulphonium, and sulphoxide groups, increased interaction with the aromatic rings in excited states enhances the acidity of phenols (Wehry, 1967), whereas in the case of sulphonic, phosphonic, and arsonic acid groups it increases base strength (Yakatan and Schulman, 1972; Liedke and Schulman, 1973b).

Intramolecular p r o t o n transfer From a study of the absorption and emission spectra of salicylic acid and 2-methoxybenzoic acid, Weller (1956) concluded that the greatly increased Stokes shift of salicylic acid was due to intramolecular proton transfer (45) in the excited state:

On excitation the phenolic group becomes more strongly acidic while the carboxylic acid group becomes a stronger base and proton exchange then occurs between the two. In confirmation of this explanation, methyl salicylate was found to behave similarly, but methyl 2-methoxybenzoate, with no transferable proton, showed a normal Stokes shift. Quenching experiments demonstrated that at room temperature the proton transfer reaction reached equilibrium within the lifetime of the excited state. Salicylic acid and its methylated derivatives have been reexamined recently (Schulman and Gerson, 1968; Kovi et al., 197213) and the pH-dependence of the intramolecular proton transfer reaction has been investigated. In both chloroform and aqueous media, salicylic acid both in the uncharged and carboxylate forms undergoes intramolecular proton transfer in the excited state. In the case of methyl salicylate, the proton transfer occurs intramolecularly in chloroform but intermolecularly in aqueous media. By comparison with the behaviour of 2-methoxybenzoic acid and methyl salicylate,

194

J. F. IRELAND AND P. A. H. WYATT

the ionization sequences shown in Scheme 1 were postulated for the ground and excited singlet states: SCHEME 1 Ground state 0-

I

0I

OH I

OH I

OH

OH I

Lowest singlet state 0-

?H

p K ( S 1 ) = 16.0 I

Very similar behaviour was found for salicylamide and salicylanilide (Schulman et al., 1973c) but 3- and 4-hydroxybenzoic acids showed no intra- or intermolecular phototautomerism, possibly because of the very short lifetimes of the S1 states involved. Also, the ionization sequences in the ground and lowest excited singlet states appear to be the same for these molecules and their methyl esters (Paul and Schulman, 1974). Not surprisingly, similar behaviour occurs in o-hydroxynaphthoic acids. Hirota (1962) interpreted his results on 3-hydroxy-2-naphthoic acid in terms of the rapid attainment of equilibrium between the

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

195

prototropic forms, but Ware et al. (1971) found this hypothesis to be inadequate and concluded, from their fluorescence quenching and lifetime measurements in a variety of solvents, that bases may promote phototautomerism of 3-hydroxy-2-naphthoic acid in nonpolar solvents by parallel static and dynamic mechanisms. The effect of pH on the absorption and fluorescence spectra of 3-hydroxy-2naphthoic acid has been investigated by Kovi and Schulman (1973a). Although the ground-state behaviour was similar to that of salicylic acid, the molecules differed in the prototropic reactions of the S, state; phototautomerism occurred only partially in the anion and not at all in the neutral molecule. Similar experimental evidence has been obtained for 1-hydroxy-2-naphthoic acid and 2-hydroxy-1-naphthoic acid. In these cases the neutral molecules undergo intermolecular phototautomerism to form the zwitterion but, unlike 3-hydroxy-2naphthoate, the anions do not react in this way (Schulman and Kovi, 1973). Interesting excited-state prototropic behaviour is shown by 7-azaindole [ l ]. The monomer in aqueous solution undergoes excited state dissociation of the indolic proton, at the N( 1) position, to form a non-fluorescent anion (Ingram and El-Bayoumi, 1974). The pK(S,)-value for this reaction is 12.3, which is identical to that reported for indole itself (Vander Donckt, 1969) and indicates that aza-substitution in the six-membered ring has little effect on the excited state charge density at the N ( l ) position. In more acid solutions the decrease in fluorescence intensity of the neutral molecule parallels the ground state protonation reaction. In nonhydrogen-bonding solvents 7-azaindole exhibits two fluorescence bands. Studies of methylated derivatives and of the effect of temperature and concentration on these bands (Ingram et al., 1971) indicate that the second fluorescence band originates from a tautomer formed by excited state double proton transfer within a hydrogen-bonded dimer [reaction (46)] . The dimer formation, which is concentration dependent, is not necessary for a double

196

J. F. IRELAND AND P. A. H. WYATT

proton transfer t o take place; in ethanol solution the similar fluorescence bands, which are not concentration dependent, are attributed to double proton transfer (47) in the excited state of a 1:1 complex between ethanol and 7-azaindole.

A temperature dependence study of the dimer phototautomerism indicated that at 77°K proton tunnelling is the dominant mechanism for proton transfer in the excited state. Results obtained on a deuterium substituted compound were consistent with this mechanism. A similar type of phototautomerism has been observed in lumichromes and alloxazines (Song et al., 1974). In the singlet state the N( 1) proton is shifted to the N( 10) position and the catalytic effect of pyridine and acetic acid is explained in terms of 1:l complexes like that postulated in the 7-azaindole case.

Naphthylamine: proton quenching and deuterium isotope effects Reinvestigation of the excited state acid-base properties of 2-naphthylamine (Schulman and Capomacchia, 1972) showed that a reported change of hybridization from sp3 to sp2 on excitation had little effect on the entropy of protonation of nitrogen in the S 1 state and that therefore the Forster cycle was still applicable. A pK(Sl )-value, calculated from the fluorescence maxima of the B and BH' form, of -8.1 is in poor agreement with the value, -2, obtained from fluorescence titration measurements. From the acidity dependence of fluorescence intensity for 1- and 2-naphthylamine Liedke and Schulman (1973a) found that the decrease in emission of the B form occurred at lower acidities than the appearance of BH' fluorescence. [A similar titration curve for the fluorescence of the neutral molecule was obtained by Seliskar and Brand (1971), who obtained a value of 0.64 for pK(Sl) from the decrease of the

197

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

fluoresence of B but did not discuss the BH' fluorescence.] The non-reciprocity of the B and BH' fluorescence titration curves was explained by Liedke and Schulman as due to a third, non-fluorescent species intermediate in structure between the cation and neutral molecule, which they regarded as an exciplex formed between the neutral molecule and a hydronium ion. In a similar study of 2-naphthylamine, but including results for the D20 / D 2SO4 system, Forster (1972) also found gaps between the B and BH' titration curves. (See Fig. 11). At low acidities the quantum yield of 0.500

Figure 1 1 . Acid dependence of the fractional quantum yields of separate fluorescence 4:) components of 2-naphthylamine: (@$) molecular (violet) component in H2SO4 H2O; ( same in D2S04/D,O; (4;) cationic (UV) component in H2S04/H20; (4")same in DzS04/D20 (Forster, 1972).

(I

fluorescence of the neutral molecule in H,O (&) and in D 2 0 (6:) approach the same value. With increasing acidity &' decreases more rapidly than 4; and the fluorescence of the BH' form grows in more rapidly in D,O than in the aqueous system. For both systems the minimum quantum yield coincides with the spectral change of B to BH'. Forster argues that the large isotope effect in this region can only be explained by the participation of proton transfer in the quenching process, which must involve at least one non-emitting species. The results give no definite information as to the exact nature of this species, but several possibilities have been suggested. Forster has suggested that the non-fluorescent species may be an intermediate in the proton transfer process with the proton shared between the amine and the solvent. This is closely related to the 1:1 H3 O+-naphthylamine exciplex suggested by Liedke and Schulman (1973a). Another possibility proposed by Weller (Forster, 1972)

198

J. F. IRELAND AND P. A. H. WYATT

involves protonation at one of the carbon atoms in the neutral amine leading to a non-fluorescent species. Schulman and co-workers have invoked non-fluorescent H30 exciplexes to interpret similar breaks in the titration curves of 1- and 2-ethyl naphthoate (Kovi and Schulman, 1972), 1-naphthoic acid and methyl-1-naphthylketone (Capomacchia et al., 1973). It is difficult to explain why in some cases only the 1-isomer shows this behaviour. It has also been suggested (Capomacchia et al., 1973) that the fluorescence titration curve reported by Watkins (19 72b) for 2-naphthamide is an example of the same phenomenon although the B and BH+ curves seem to be complementary at lower acidities in this case (see Fig. 4). The kinetic isotope effects shown in Fig. 11 (Forster, 1972) resemble those reported for 2-naphthol by Stryer (1966). Like 2-naphthylamine, 2-naphthol shows an increased quantum yield and protonation of S 1 occurs at lower acidities in D,O than in H,O. For 2-naphthol, pK(S,)-values of 3.0 in H 2 0 and 3.4 in D2O are calculated from the measured excited state rate constants; in H2 0 k , = 5.29 x l o 7 s-' and k 2 = 5.5 x 10" dm3 mole-' s - ' , while in D 2 0 k , = 1.3 x l o 7 s-' and k 2 = 3.5 x 10" dm3 mole-' s - ' . These results confirmed the earlier pK(S, )-values calculated by Wehry and Rogers (1966) using the Forster cycle (Table 9), which show incidentally that the pK-values are closer by about 0.1 unit in the S1 state. +

TABLE 9 pK(S1)-Values for Aromatic Alcohols in HzO and DzO Obtained by the Forster Cycle Method Compound

PK(SO)H20

pK(S1 ) H , O

pK(SO)D20

pK(S1 ) D , O

Phenol 4-Phenolsulphonate 4-Bromophenol 3-Methoxyphenol 4-Methoxyphenol 4-Hydroxyphenyl trimethylammonium chloride 2-Naphthol

10.00 8-97 9.35 9.62 10.24

4.1 2.3 2-9 4.6 5.7

10.62 9.52 9.94 10.20 10.85

4.6 2.7 3.4 5.1 6.2

8.34 9.47

1.6 3-0

8.90 10.06

2.0 3-5

Recently, a similar deuterium isotope effect was noted for 3-( 3'-thienyl) pyridine (Bouwhuis and Janssen, 1972). In acid solutions in which the B and BH' forms both fluoresce, the ratio of the

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

199

BH+ to B fluorescence intensities decreased on changing from H,O to D20 as solvent.

Substituted quinolines The excited state pK-behaviour of quinoline derivatives has received considerable attention (see Tables 6.3 and 6.4). Since the heterocyclic nitrogen atom of quinoline is expected to become a stronger base in the excited state while the acidity of hydroxyl or amino-substituents increases, different ionization sequences can be obtained in the So and S 1 states. 3-Hydroxyquinoline is a typical example of this behaviour as shown in Scheme 2 (Haylock et al., 1963; Mason et al., 1968). SCHEME 2

At the extremes of the acidity range studied, the cation and the anion are the species found in both the ground and S, states. At intermediate pH values, the neutral molecule is the predominant species in the ground state. On excitation, however, the zwitterion is more stable, because the phenolic group beomes more acidic and the nitrogen more basic, and only the fluorescence of this species is observed. Fluorescence titration in the acid region gives a value of -0.22 for pK(S, ) of the cation-zwitterion equilibrium and, although the fluorescence curves in alkaline solutions are complicated by quenching effects, a pK(S1)-value of 12.2 can be obtained for the

200

J. F. IRELAND AND P. A. H. WYATT

zwitterion-anion reaction. Similar explanations, involving functional groups with excitation pK-changes in opposite directions, have been proposed for other hydroxyquinolines, amino-, carboxy-, and mercapto-quinolines (Tables 6.3 and 6.4) and hydroxycoumarins (Yakatan e t al., 1972a, 197213). The parent quinolinium cation unexpectedly shows two inflections, at 5 - 8 and -1.6, in the fluorescence intensity curve as the solution acidity is increased from neutral pH t o H , = -4; the absorption spectrum remains unchanged, however. Capomacchia and Schulman (1973) discount further protonation at a carbon atom as an explanation of this effect, on the grounds that there is no shift in the frequency of the band. They suggest that the nearly degenerate Laand Lb states may have very different acid strengths, an idea that must ultimately be reconciled, at least approximately, with the Forster cycle.

A n t h r o i c acids The pK-behaviour of 1-, 2- and 9-anthroic acids in the excited state was studied by Vander Donckt and Porter (1968a). Directly determined pK(T,)-values were found to lie nearer to the pK(S,)-values calculated using the Forster cycle than to pK(So). In a study of the fluorescence of 1- and 2-anthroic acids over a wide acidity range (Schulman et al., 1973a), it appeared that the deprotonation reactions did not come to equilibrium in the excited state. For protonation of the carboxyl group only 1-anthroic acid showed an excited state reaction and, as expected, it became more basic in the S state. Fluorescence lifetime measurements on the prototropic species derived from 1- and 2-anthroic acids help in understanding the failure t o reach equilibrium:

1-Anthroic acid

2-Anthroic acid

( [

cation (Ho = -0.7) neutral molecule (pH = 3.5) anion (pH = 8.0) cation (Ho = -10.0) neutral molecule (pH = 2-0) anion (pH = 8.0)

r/ns 12.2 13.8

9.1

< 1.6 < 1.6 14-5

The lifetime of the neutral 2-anthroic acid molecule is obviously too short t o allow dissociation and, although the 1-anthroic acid is longer-lived, the dissociation rate constant, - l o 4 s-l if

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

201

pK(Sl) = 6-6, is certainly too slow to compete with fluorescence. It is not immediately obvious why the short lifetime of the electrically neutral 2-anthroic acid molecule prevents the protonation of the carboxylic acid groups in the S1 state at such high acidities. Since the fluorescence of the neutral molecule is observed in solutions slightly less acidic than pK(So) (x- l o ) , a surprisingly low rate constant of excited state protonation must be responsible for the non-attainment of the S, equilibrium. The Forster cycle calculations on the dissociation of 9-anthroic acid (Vander Donckt and Porter, 1968a) have been criticized (Werner and Hercules, 1969) on the grounds that excited state rotation of the carboxyl group invalidates the equal entropy assumption of the Forster cycle. Pace and Schulman (1972) found that this dissociation equilibrium was not established in the excited-state but obtained a pK(S, )-value from fluorescence titrations for protonation of the carboxyl group. As expected its basicity was increased in the S1 state: a ApK-value of 6 units was recorded.

R a t e constants of excited state p r o t o l y t i c reactions

The rate constants for protonation of the excited singlet states of several compounds were determined by Weller (1961). Although the measurement of excited state equilibrium constants has become more common, there have been relatively few determinations of the rate constants involved. Trieff and Sundheim (1965) investigated the effects of solvent changes on the rates of protonation and deprotonation of 2-naphthol in the S, state. The dissociation rate constant decreased progressively with the addition of methanol or glycerol to the aqueous solution but the protonation rate constant varied in a more complex manner. As mentioned above, Stryer (1966) found both rate constants smaller in D20 than in H2 0. Recently the 2-naphthol system has been re-examined. From lifetime measurements at various pH values, Loken et al. (1972) and Ofran and Feitelson (1973) determined the rate constants (see Table 10). The effects of further ring substituents on the rate constants has been studied by Rosenberg and Brinn (1972) (see Table 10). Dissociation in the S, state is retarded by substituents which increase the electron density on the phenolic oxygen; the electron-donating methyl group reduces k while chloro- and bromo-substitution, as expected, increase it.

202

J. F. IRELAND AND P. A. H. WYATT TABLE 10 Rate Constants for the Excited State Dissociation Reaction of Naphthols kl

R*OH + H 2 0

R*O- + H3O+

k2 ~~

Molecule 2-Naphthol

1-Bromo-2-naphthol 6-Bromo-2-naphtho1 6-Methyl-2-naphthol 7-Methyl-2-naphthol 1-Chloro-2-naphthol 1-Naphthol 4-Chloro- 1-naphth01 4-Chloro-1-naphthol

kl11o7 s-l

k*/lO'o dm3 mole-' s-'

4.1 6.2 5.29 5.6 4-2 4.5 57 70 1.3 1.6 49 285 2160 2380

5.08 4.9 5.5 4.0 5.1 6 40 2.8 1.1 5.9 7.3 10

Ref. Weller, 1955 Trieff and Sundheim, 1965 Stryer, 1966 Rosenberg and Brinn, 1972 Loken et al., 1972 Ofran and Feitelson, 1973 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972 Rosenberg and Brinn, 1972

Watkins (1971, 1972a, 1972b) has studied the rate processes in the excited state protonation (47) of various naphthalenes having carbonyl containing substituents. kl

B*H+

k2

B* + H+

(47)

Methyl-2-naphthyl ketone k l = 0.70 x 10' s-' and k 2 = 3.8 x 10" dm3 mole-' s-l 1-Naphthoic acid k l = 1.7 x l o 8 s-l and k 2 = 1.7 x 10" dm3 mole-' s-' 2-Naphthoic acid k l = 4.0 x l o 9 s-l and k 2 = 2 x 10'' dm3 mole-' s-' 1-Naphthamide k1 = 8.6 x l o 7 s-l andk2 = 4.7 x 10" dm3 mole-' s-l 2-Naphthamide k1 = 7.7 x l o 7 s-' andk2 = 4.7 x 10" dm3 mole-' s-l

The last three values of k l could only be obtained after k , had been estimated from the diffusion coefficient. For the protonation of tyrosine in the S1 state, k = 5-8 x l o 8 s V 1 and k, = 3.6 x 10" dm3 mole-' s - l (Fayet and Wahl, 1971) and the rate constant for protonation of thionine in the triplet state was determined to be 0.83 x 10" dm3 mole-' s-' (Bonneau

203

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

and Stevens, 1972). Comparison of the thionine rate constant with the rate of protonation of other nitrogen bases suggested that in the triplet state the site of protonation is the ring N-atom. In a study of styrylpyridines Favaro et al. (1973) obtained the following rate constants ( k l ) for S1 protonation (48) by assuming StP*

+ €120

kl

StP*H+ + OH-

k2

kz

4’-Substituent

H

CH3

OCH3

c1

Br

k1/108 s-l

2.2

3.5

10.0

2.2

3.1

= 5.5 x

lo9 dm3 mole-’

s - l . Since the forward rate constant is

scarcely influenced by substitution the extent of equilibrium in the excited state is found to depend on the fluorescence lifetime of the base.

Carbon protonation Ground state protonation of aromatic hydrocarbons occurs in strongly acid media, but Fiirster cycle calculations indicate very large changes in pK on excitation, even bringing p K ( S l ) well into the ordinary pH range. For example, naphthalene has a pK(S,) value of -4.0 but pK(S,) and p K ( T , ) values of 11.7 and -2.5 respectively (Vander Donckt et al., 1970). It is doubtful whether S , reactions of this type can in practice be realized since protonation at carbon atoms is generally slow (Bell, 1959; Caldin, 1964;Jones, 1973). The possibility of observing these reactions should be greater in the triplet state because of the longer excited state lifetime involved. Colpa et al. (1963) calculated pK(S,)-values for a series of aromatic hydrocarbons, but could not detect fluorescence changes in the regions of acidity indicated by the Forster cycle, although fluorescence spectra attributed to “proton complexes” of 3,4-benzopyrene and 1,2-benzanthracene were observed in some solutions containing only the neutral molecule in the ground state. Flurry and co-workers (1963, 1966, 1967) have carried out theoretical and Forster cycle calculation on the excited state basicities of polymethylbenzenes and Kuz’min et a l . (1967) have also calculated pK(S1)- and pK(Tl )-values for polycyclic aromatic hydrocarbons: increases in base strength of from 7 t o 30 powers of ten were derived for S 1 .

204

.J. F. IRELAND AND P. A. H. WYATT

Mason and Smith (1969) found that for a series of mono- and bicyclic aromatic hydrocarbons the changes in the fluorescence spectrum with acidity reflected the ground state protonation reaction. The pK(Sl )-values calculated for benzene, toluene, naphthalene, azulene, and indolizine do not correspond t o observable processes since the rate of protonation is too slow to compete with deactivation of the S state. Photochemical deuterium and tritium exchange experiments in 1 mole dm-3 perchloric acid indicate that the radiative deactivation rate of an electronically excited aromatic hydrocarbon is faster than the rate of protonation by a factor > l o 5 . The cyclo-octatrienyl dianion is a stronger base in the excited state than in the ground state (Brauman et al., 1968) and on irradiation in the presence of weakly acidic proton donors the mono-anion is formed by proton exchange. Cyclo-octatrienes are formed by a second proton transfer. Although no proton removal from cyclooctatrienes occurs, either in the dark or on irradiation, deuterium isotope experiments showed that in these experiments the monoanion is deprotonated at a rate comparable to its further protonation. Similar experiments with naphthalene in CH, COPD showed that the positions of hydrogen exchange were the same in the ground and excited states, while phenanthrene is photodeuterated at positions 4 and 5 and not at 9 and 10 as in the dark (Vander Donckt et al., 1970).

Relative Ordering ofpK(S,), pK(S,) and pK(T,) Values and Theories of pK-Change In making a qualitative decision about the direction in which the acid-base properties of organic molecules will be affected by excitation, it is useful to apply the rough empirical rule that donor substituents (e.g. OH, NH,) will become stronger donors in the excited state while acceptors (e.g. COzH, CO) will attract electrons more strongly from the ring. (For a theoretical justification see e.g. Daudel, 1970.) In consequence the protons are less firmly bound in the S 1 state in hydroxyl and amino-groups which therefore become stronger acids (and weaker bases due to the loss of electron density on oxygen or nitrogen), while the opposite applies to the carboxyl groups, which become weaker acids or (like ketones and aldehydes) stronger bases (see Table 1).

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

205

When the first pK(T, )-values became available, though displaced in the same direction as pK(Sl) they were much nearer to pK(S,) than pK(SI) (Jackson and Porter, 1961). This phenomenon was again explicable in terms of the charge distributions within the molecules. Murrell (1963) pointed out that charge transfer states lie at considerably higher energy than any of the actual states considered but are nearer in energy to S1 than to T I , whence the acidity-enhancing effects of charge transfer structures like [ 21 will be

greater in the S1 state. Alternatively the similar effect of separation of the unpaired electrons in the triplet state has been pointed out (Porter, 1967). This system of inequalities, pK(So) > p K ( T l ) > pK(S,) or pK(So) < pK(T,) < pK(S,) with pK(T,) nearer pK(So) than pK(S1), is still widely quoted (Dewar and Trinajstic, 1971; Stenberg and Dutton, 1972; Daudel, 1973), although exceptions have been found. In a study of nitrogen heterocycles (Grabowska and Pakula, 1966), large shifts in the pK-value of the triplet state were indicated and Forster cycle calculations for phenazine showed that the shift in pK(T,) was at least as large as that in pK(Sl) (Grabowska and Pakula, 1969). Again directly determined pK(T,)-values for anthroic acids were found to lie nearer pK(S,) than pK(So) (Vander Donckt and Porter, 1968a). On the basis of theoretical calculations it was suggested (Daudel et al., 1969, 1970) that some substituents in certain positions in the ring might lead to a change in the pK-order and that the failure to find good triplet transients for the B and BH' forms of l-naphthol, analogous to those found for 2-naphthol, might be due to a difference in acid-base behaviour in the triplet states. Recent experiments with sulphonic acid derivatives of l-naphthol indicate, however, that the pK(T, )values of l-naphthol and 2-naphthol are probably not very different (Henson and Wyatt, 1975). A change in the pK-order has nevertheless emerged in a different class of compounds; xanthone and several substituted benzophenones have pK(TI )-values lying outside the range between pK(So) and pK(Sl) (Ireland and Wyatt, 1972, 1973). In these

206

J. F. IRELAND AND P. A. H. WYATT

molecules the lowest transition is of nn* type in the ketone and of

nn-* type in its conjugate acid. Since the Sl-T, energy difference is much smaller for nn* than for an* states, molecules of this type, which are already stronger bases in S , than So states, can become still stronger bases in the T l state. The relative energies of the states are shown schematically in Fig. 12. The pK-order is therefore pK(So) < P m l ) < PK(T1).

BH+

B+H+

Figure 12. Schematic representation of the energy levels of the ground and first excited states of xanthone and related compounds.

In the case of xanthone at least, this order is not only shown up in the Forster cycle estimates, but has been confirmed by observing the variation with pH of the optical densities of the triplet states of B and BH+ and comparing it with the fluorescence intensity behaviour (see Fig. 6). Confirmation that the pK order obtained using the Forster cycle is reliable in such cases is also found in a direct determination of pK(Tl ) of benzophenone by a laser technique; the value derived is consistent with earlier phosphorescence observations (Rayner and Wyatt, 1974). Ledger and Porter (1972) observed a marked decrease in the phosphorescence intensity of benzophenone near pH 5 , and the apparent discrepancy between this result and the pK(Tl )-value of 1.5 is due to the very large difference in lifetimes of B H + ( T l )and B(Tl). Since unprotonated benzophenone has a very shortlived S , state KT for the intersystem crossing alone in ethanol is 16.5 ps (Hochstrasser et al., 1974)], protonation in this state is unlikely. However, Forster cycle calculations indicate that the singlet state would be a weaker base than the triplet state. The realization that unprotonated benzaldehyde and acetophenone had T I states of the

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

207

nn* type while that of the protonated molecules was TIT* led Schulman (1973b) to discount his calculated values of pK(T, ), but the direct determinations so far show this approach to the Forster cycle t o be overcautious. Apart from the qualitative notions of charge transfer and electron pairing mentioned above, more detailed quantum mechanical considerations have been published (see Daudel, 1970, 1973). Coulson and Jacobs (1949) predicted that aniline should decrease in basicity on excitation, and Sandorfy (1951) showed that the total charge on the nitrogen atom of aniline, although negative in the ground state, becomes positive in the first excited state. The excited state pK-behaviour of aromatic hydrocarbons has also been studied (Colpa et al., 1963; Flurry and Lykos, 1963; Flurry, 1966), and for polymethylbenzenes a linear relationship was observed between the excited state localization energies and the FGrster cycle pK-values (Flurry, 1966). Cetina et al. (1967) carried their calculations far enough to derive pK(S1 )- and pK(T, )-values for acridine and quinoline which showed satisfactory agreement with the experimental results. Generally, however, a comparison is made between the qualitative changes in acid or base strength observed experimentally and the computed charge densities at the protonation site. A reasonable correlation has been obtained between these quantities for azo- and azoxybenzenes (Jafft. et al., 1964a) and for phenazine (Grabowska and Pakula, 1969; see also Grabowski et al., 1966; Tyutyulkov et al., 1967; Fernando and Schulman, 1968; Song, 1968; Mason et al., 1968; Kysel, 1969; Tyutyulkov and Hiebaum, 1969; Rotkiewicz and Grabowski, 1969; Sommer and Kramer, 1971; Rosenberg and Brinn, 1972).

5. EXCITED STATES AND ACIDITY SCALES For the dissociation of any acid BH', pK can be represented formally by expression (49), where a H +represents the hydrogen-ion

activity of the solution and f B and f B H + are activity coefficients of the species B and BH'. When comparing pK-values for different acids, or for the same acid in different electronic states, it is of

208

J. F. IRELAND AND P. A. H. WYATT

course necessary to relate all the activity coefficients t o the same standard conditions and for aqueous solutions they are conveniently taken as unity at infinite dilution of all solutes in water. The problem is then that the region of acidity in which [BH'] and [B] become comparable, so that their ratio is measurable with accuracy, may be very far removed from infinite dilution; some kind of extrapolation procedure is necessary. What is usually readily determined is the acid concentration at which the concentrations of protonated and unprotonated forms of a base become equal, so that the first term on the right hand side of (49) vanishes, leaving only the quantity -log1 0 (aH+fB/fB H + ) 1 / 2 with the suffix denoting half-protonation. A change in pK upon excitation to any excited state can then be expressed as in (50). The value of u H + for the excited state

equilibrium is in general different from that for the ground state since the acidity at half-protonation is usually different. In each case uH+ is characteristic of the solvent acid solution only, b u t f B / f B H + depends upon the structure of B (as well as the acid concentration) and could well be different for the same molecule in different electronic states, even at the same acid concentration. The f B / f B H + factors do not, however, contain the whole of the effect of the difference in molecular environment between the ground state and excited state species; only that part which depends upon the change in acid concentration from infinite dilution up to the solution at which [BH'] = [B] is included in each case. The remainder becomes incorporated into the pK-values (i.e. into AGO/RT) , partly in the AHO/RT and partly in the ASO/R difference, just as with any comparison of the protonation of a pair of bases not related by electronic excitation. When half-protonation of the ground state and of the S, and T , states all fall within the pH range 1-13, as in the classic studies on 2-naphthol (Weller, 1952; Jackson and Porter, 1961), there is little difficulty in fixing all the pK-values t o within at least k0.01 by standard electrochemical methods. Shifts in pK upon excitation are commonly so large, however, that, even if the ground state pK lies in the 1- 1 3 range, the S1 and T , values are likely to lie outside. There are now in any case hundreds of molecules known t o be halfprotonated well outside that range in the ground state. The need for

ACIDBASE PROPERTIES OF ORGANIC MOLECULES

2 09

a reliable acidity scale related to concentrations of the common strong acids is therefore apparent. It is well known that outside the pH range almost all pKdeterminations for unexcited molecules have been based on the Hammett indicator method (Hammett and Deyrup, 1932). Even in concentrated acid (or alkaline) solutions, where uncertainties in the value of aH+fB/fBH+ become very serious, it is easy to measure the ratio of protonated to unprotonated indicator concentrations spectrophotometrically when the absorption peaks are sufficiently resolved. The difficulty arises in trying to extrapolate beyond the measurable [BH+]/ [B] range to zero electrolyte concentration. Hammett and Deyrup assumed that the activity coefficient ratios of the type f B / f B H + were very similar for different indicators in the same acid solution. For the equilibrium (51) between two indicators A and B, a comparison of the concentration ratios [BH'] /[B] and [AH+]/[A] over an acidity range in which both could be measured would lead to a direct estimate of the difference in pK-values using (52) and (53). Beginning with 4-nitroaniline, which is about half-

BH+ + A +=AH+ + B

..

PKA - PKB 210gio [AH+]/[A1

-

logio [BH+I/[Bl

(51)

(53)

protonated at pH = 1, a pK-value was assigned to a rather more weakly basic substituted aniline which overlaps 4-nitroaniline at the acid end of its range, and then to further members of a chain of indicators, thus establishing an acidity scale by (54). If a truly HO

= pK

+

loglo [B1

= - loglO(aH'fB/fBH+),

(54)

universal scale had been set up in this way, the problem of applying (50) to excited state equilibria would be solved: the right side of (50) would simply reduce to the difference of two acidity function (H,) values assigned to the acid solutions at which half-protonation occurred in the excited and ground states. It is now known that consistent scales can only be hoped for if indicators of the same type are used along a series; thus several different acidity function scales can be set up using different families of compounds, such as nitroanilines (H, ), amides (HA), carbon acids

210

J. F. IRELAND AND P. A. H. WYATT

( H c ) , triarylmethanols ( H R), etc. Unfortunately these scales can differ by several units at the concentrated acid end. For recent reviews, the books by Rochester (1970) and Liler (1971) should be consulted. From the point of view of the acid-base properties of excited states the major question is therefore whether an S , or T, state of a substituted benzophenone, say, will behave like a groundstate benzophenone or like some other family of compounds with an electronic distribution, and hence polar character, rather more like that of the excited state in question.

20

40

60

80

wt % H2S04

Figure 13. Comparison with acidity functions of pK(So)-values calculated either from pK(S1) ( 0 )or from PW'I) (O).

As yet not enough information has been gathered to answer this question and most authors who have required an acidity scale for excited states have simply adopted the nitroaniline Ho function (Weller and Urban, 1954; Vander Donckt and Porter, 1968b; Haylock et al., 1963; Hopkinson and Wyatt, 1967; Bratzel et al., 1972). Those pK-values which have been fairly well established for excited states in the normal pH region, by direct observation of transient concentrations or by fluorescence titrations combined with lifetime measurements, have nevertheless been used to illustrate a form of test in Fig. 13. From the values of pK(S,) or pK(T,) in each

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

21 1

case, the Forster cycle has been used in reverse to estimate the value of pK(S,), which has then been plotted against the weight percentage of sulphuric acid at which the ground state is known by experiment to be half-protonated. The best that can be said at present is that the results obtained so far fall within the spread already found for the acidity function values, three examples of which are shown on the same graph. In the case of benzophenone (Rayner and Wyatt, 1974), the pK(So)-value so determined does at least deviate from H , in the direction of the latest benzophenone acidity function scale. There may therefore be some instances in which a single molecule on excitation can perform the function of a whole chain of indicators in spanning an acidity scale (Hopkinson and Wyatt, 1967). The departure of excited state pK-values from the Ho scale in tests such as that illustrated in the figure must, apart from the direct effect of the simple Forster approximation, reflect the same differences in local structure as have been invoked to account for the differences in ground state acidity functions amongst themselves. These differences have been shown to correlate well with the differences in reasonable estimates of the f B /fB H + ratios (Edward, 1964; Boyd, 1969). The most popular explanation at the molecular level has been hydration. The large overall drift in acidity is understandable in terms of successive hydration of the proton (Wyatt, 1957), or the specific formation of H,0: where there is sufficient water (Bascombe and Bell, 1957); but differences in the extents of hydration of the indicators themselves have to be invoked to explain differences between the various scales: (see e.g. Taft, 1960b). If this were the whole story, however, acidity functions of a given type in solutions of different strong acids would be a common function of the water activity. This appears to be the case for the nitroaniline indicators (Wyatt 1957; Yates and Wai, 1964), although even there the correlation may be quite fortuitous (Yates et al., 1973), but it is not so for the triphenylmethanols, for example (Hogfeldt, 1962). In the latter case special interactions involving the perchlorate ion may play a part (Bunton and Huang, 1972; Postle and Wyatt, 1972). Apart from all the important investigations in many solvents which are helping to shape a detailed theory of acid-base interactions at the molecular level, acidity functions are constantly being revised and viewed from new points of view. For example, Marziano, Cimino, and Passerini (1974) observe that the logarithms of activity coefficient ratios of the type ,fB f H + / f B H + are proportional for

212

J. F. IRELAND AND P. A. H. WYATT

different indicators over wide ranges of acidity and construct a reference function, M , = -log, 0fAfH+/fA H + , for aqueous sulphuric and perchloric acids, which they then use to estimate pK-values for weak bases. As with similar correlations, this approach requires some revision of many weak base pK-values. Again, a polarographic H G F scale based on the oxidation of ferrocene has been set up by Janata and Jansen (1972) and has been shown to correlate well with several other scales in estimating the hydronium ion activity in strongly acid media (Yates and McClelland, 1973). As more detailed studies of excited state pK-values accumulate and our understanding of acidic and basic solutions deepens, it should become possible to discover local effects which will explain consistently both the trends in the appropriate acidity scales and the spectral effects (absorption-fluorescence spacings) related to departures from the simple Forster calculation of the pK shift upon excitation.

6. APPLICATIONS Production, by an excited state protolytic reaction, of species not present in the ground state has been discussed as a method for obtaining population inversion for laser action (Derkacheva, 1963). Furthermore, since there is the possibility of obtaining several emitting species from one ground state molecule, excited state reactions can extend the spectral range of emission from a dye. Shank et al. (1970) obtained tunable laser emission from 385 to 574 nm with 4-methylumbelliferone by adjusting the solution pH t o obtain emission from the basic, neutral, and protonated species. Similarly three dyes were studied by Derkacheva and Petukhov (1973), who were able t o obtain laser emission over a range of 200 n m from the same dye solution. More generally, an increased awareness of excited state prototropic effects has led to its increasing consideration in the interpretation of organic photochemistry, (e.g. Gutsche et al., 1968, 1973). The acid-base equilibrium in the lowest excited singlet and triplet states was recognized (Godfrey et al., 1965) t o be of prime importance in determining the course and yield of photochemical reduction of aromatic ketones by hydrogen abstraction from the solvent. Even the case of 4-hydroxybenzophenone, which appeared at first t o be anomalous, could be explained in terms of a revised

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

213

pK(T, )-value (Ireland and Wyatt, 1973). The photoreduction of methylene blue (Kato et al., 1964) and phenazine (Bailey et al., 1968) have been interpreted similarly. The yields in photoisomerization reactions of 3-styrylpyridines, in the pH range, are related to the acid-base equilibria in the ground and singlet states (Bartocci et al., 1973) and the complex photoisomerization of eucarvone appears to proceed in acid media via a protonated triplet species (Hine and Childs, 1973). The protonated triplet states of nitrobenzene and nitronaphthalene have been postulated as the species responsible for their enhanced photoreduction in hydrochloric acid (Hurley and Testa, 1967; Trotter and Testa, 1970). Winefordner et al. (1971a), on the basis of the calculated pK(T,)-value for nitrobenzene, considered that the protonated triplet was unlikely to be involved, though the protonated singlet might be; but Wubbels et al. (1973) showed that the reduction was not enhanced when sulphuric acid was used in place of hydrochloric acid and suggested that electron transfer from chloride ion might be involved. Following the electron transfer the radical anion probably becomes protonated (Cu and Testa, 1974). Vinyl cations are highly reactive and easily hydrated to form ketones. Since in the ground state they are stable only in solutions of high acidity, precluding the use of many functional groups, Wooldridge and Roberts (1973) examined the possibility of their formation in an excited state. Phenylacetylenes which exhibited no reaction in the dark were found to undergo photohydration to ketones in dilute acid solution. Protonation of excited acetylenes followed by the addition of water is proposed as the mechanism for these reactions and an analogous scheme is proposed for the reported photoaddition of methanol and acetic acid to diphenylacetylene. Curves of the photohydration rates against pH for uracil and cytosine derivatives show inflection points near the pK(SI )-value of the molecules (Burr et al 1972). Further evidence for the participation of the S 1 state in the reaction is provided by the negative temperature coefficients for photohydration and fluorescence intensity. For the similar reaction, the photoinduced alcoholysis of 3,4-dihydroxycoumarins, the variation of quantum yield of product formation with pH correlates reasonably well with the pK(S, )-values (Gutsche and Oude-Alink, 1968). Though not concerned directly with protonation, the related charge density effects on the rates of 0, m-,and p-substitution in excited molecules have been investigated (see e.g. Zimmerman et al., 1963).

2 14

J. F. IRELAND AND P. A. H. WYATT

Several molecules which can undergo intramolecular proton transfer in the excited state have been found to be unusually photostable. Thus, for example, 2-hydroxybenzophenone is used as a photostabilizer in polymers while benzophenone itself is photoactive (Kysel, 1969). In crystalline 2-( 2'-hydroxyphenyl)benzothiazoleand its derivatives [3] a proton is transferred in the excited state from an

oxygen to a nitrogen atom (Williams and Heller, 1970) and the photochemical stability of the compounds is greatly improved relative to the N-methylated derivatives in which no hydrogen transfer can take place. The excited state acid-base behaviour of molecules has direct implications in the field of analytical fluorimetry and phosphorimetry. Since the emitting species can be changed by adjusting the acidity of the solution, the sensitivity or selectivity of an analytical procedure can be increased (see e.g. McCarthy and Winefordner, 1967; Argauer and White, 1970). Thus, for example, the limit of detection by phosphorimetry of 4-nitrophenol is much lower in solutions containing the anion (Schulman and Winefordner, 1970) and the long-wavelength fluorescence of warfarin in sulphuric acid allows its selective fluorimetric determination in the presence of other 4-hydroxycoumarins (Yakatan e t al., 1972). Also, in biochemistry, excited state protolytic behaviour is increasingly used both t o interpret results and to act as a probe (see e.g. Udenfriend, 1967; Chen, 1967; Morton et al., 1969; Lasser and Feitelson, 1971; Brand and Gohlke, 1971). Loken et al. (1972) have suggested that the rate of excited state proton transfers should give a quantitative measure of the environment of a probe molecule, since the rate constants depend o n the solvent and on the character of any possible proton donors and acceptors. For 2-naphthol-6-sulphonate adsorbed on bovine serum albumin, the changes in the rates of ionization were so great that no excited state ionization was observed at all; but using this method, it is argued, it should be possible to detect localized changes of the ionizing site, such as might be caused

ACID-BASE PROPERTIES OF ORGANIC MOLECULES

215

by a conformational change of a macromolecule (Loken et al., 1972; Bowie et al., 1973). Since coiled chains of proteins are known to uncurl because of ionic repulsions when ionization occurs, Reid (1957) suggested that excited state dissociation acts as a trigger in rapid biological processes. The 7-azaindole dimer, which undergoes photo-induced double proton transfer (see Section 4), has similarities t o the adenine-thymine and guanine-cytosine base pairs of DNA. Its excited state proton transfers have been proposed as possible mechanisms of mutagenesis (Ingram and El-Bayoumi, 1974). These examples show the widening scope of excited state acid-base applications. The necessary background of pK information is rapidly becoming consolidated but, in almost all the classes of compound studied in detail, significant inconsistencies in the interpretation call for further work. Without doubt exploitation of lifetime techniques will go a long way towards solving such problems.

REFERENCES Aaron, J. J., Winefordner, J. D., Schulman, S. G., and Gershon, H. (1972). Photochem. Photobiol. 16, 89. Aaron, J. J., and Winefordner, J . D. (1973). Photochem. Photobiol. 18, 97. Abate, K., Capomacchia, A. C., Jackman, D., Kovi, P. J., and Schulman, S. G. (1973). Anal. Chim. Acta 6 5 , 59. Abate, K., and Schulman, S. G. (1972).J. Pharm. Sci. 61, 1576. Adita, S., and Lahiri, S. C. (1971).Zndian J . Chem. 9 , 4 9 2 . Albert, A., and Sargent, E. P. (1962). “Ionization Constants of Acids and Bases.” Methuen, London. Argauer, R., and White, C. E. (1970). “Fluorescence Analysis.” Dekker, New York. Avigal, I., Feitelson, J., and Ottolenghi, M. (1969). J . Chem. Phys. 50, 2614. Bailey, D. N., Roe, D. K., and Hercules, D. M. (1968).J. Amer. Chem. SOC. 90, 629 1. Ballard, R. E., and Edwards, J. W. (1964a). Spectrochim. Ac ta 20, 1275. Ballard, R. E., and Edwards, J. W. (1964b).J. Chem. SOC. 4868. Barker, C. C., Bride, M. H., Hallas, G., and Stamp, A. (1961). J. Chem. SOC. 1285. Bartocci, G., Bortolus, P., and Mazzucato, U. (1973).J. Phys. Chem. 77, 605. Bartok, W., Lucchesi, P. J., and Snider, N. S. (1962). J . Amer. Chem. SOC. 84, 1842. Bartok, W., Hartman, R. B., and Lucchesi, P. J. (1965). Photochem. Photobiol. 4,499. Bascombe, K. N., and Bell, R. P. (1957). Disc. Faraday SOC. 24, 158. Bell, R. P. (1959). “The Proton in Chemistry.” Methuen, London.

216

J. F. IRELAND AND P. A. H. WYATT

Birks, J. B. (1970). “Photophysics of Aromatic Molecules.” Wiley, London. Birks, J. B., and Munro, I. H. (1967). Progr. Reaction Kinetics 4, 239. Bonneau, R., and Stevens, R. D. S. (1972). Chem. Phys. Lett. 13, 376. Bouwhuis, E., and Janssen, M. J. (1972). Tetrahedron Lett. 233. Bowie, L. J., Irwin, R., Loken, M. DeLucca, M., and Brand, L. (1973). Biochem. 12, 1852. Boyd, R. H. (1969). In “Solvent-Solvent I n t e r a c t i o n s ” ~98. . (J. F. Coetzee and C. D. Ritchie, eds.). Marcel Dekker, New York. Brand, L., and Gohlke, J. R., (1971).J. Biol. Chem. 246, 2317. Bratzel, M. P., Aaron, J . J., Winefordner, J. D., Schulman, S. G., and Gershon, H. (1972). Anal. Chem. 44, 1240. Bratzel, M. P., Aaron, J. J., Winefordner, J. D., Schulman, S. G., and Gershon, H. (1973b).J. Phys. Chem. 77, 1595. Brauman, J. I., Schwartz, J., and van Tamelen, E. E. (1968). J . Amer. Chem. S O C . 90, 5328. Bridges, J. W., Davies, D. S., and Williams, K. T. (1966). Biochem. J . 98, 451. Bridges, J. W., and Williams, R. T. (1968). Biochem. J. 107, 225. Brinen, J. S., Rosebrook, D. D., and Hert, R. C. (1963). J . Phys. Chem. 67, 2651. Bulska, H., Chodkowska, A., and Grabowska, A. (1972). Chem. Phys. Lett. 12, 508. Bunton, C. A., andHuang, S. K. (1972).J. Amer. Chem. S O C . 94,3536. Burr, J. G., Park, E. H., and Chan, A. (1972).J. Amer. Chem. S O C . 94, 5866. Caldin, E. F. (1964). “Fast Reactions in Solution.” Blackwell, Oxford. Capomacchia, A. C., and Schulman, S. G. (1972a). Anal. Chim. Ac ta 58,91. Capomacchia, A. C., and Schulman, S. G. (1972b). Anal. Chim. Acta 5 9 , 4 7 1 . Capomacchia, A. C., and Schulman, S. G. (1973).J. Amer. Chem. S O C . 95, 2763. Capomacchia, A. C., Kovi, P. J., and Schulman, S. G. (1973). Spectrosc. Lett. 6, 7. Carroll, F. A., McCall, M. T., and Hammond, G. S. (1973).J. Amer. Chem. S O C . 95, 315. Cetina, R., Jain, D. V. S., Peradejordi, F., Chalvet, O., and Daudel, R. (1967). Compt. Rend. 264,874. Chen, R. F. (1967). Fluorescence 443. Chibisov, A. K. (1970). Russ. Chem. Revs. 39, 891. Colpa, J. P., Maclean, C., and Mackor, E. L. (1963). Tetrahedron, Suppl. 19, 65. Coulson, C. A., and Jacobs, J. (1949).J. Chem. SOC. 1983. Cowgill, R. W. (1971). Photochem. Photobiol. 13, 183. Cu, A., and Testa, A. C. (1974).J. Amer. Chem. SOC. 96, 1964. Daudel, R. (1970). Adv. Quantum Chem. 5, 1. Daudel, R. (1973). “Quantum Theory of Chemical Reactivity.” Reidel, Holland. Daudel, R., B e r t h , J., and Chalvet, 0. (1969). Theoret. Chim. Acta 14, 1. Daudel, R., Bertran, J., Dannerberg, J. J., Leute, R., Ponce, C., and Chalvet, 0. (1970). Theoret. Chim. Acta 17, 249. Davies, C. W. ( 1962). “Ion Association.” Butterworths, London. Debye, P. (1942). Trans. Electrochem. S O C . 82, 265. Demas, J. N., and Adamson, A. W. (1971). J . Phys. Chem. 75,2463. Derkacheva, L. D. (1960a). Opt. Spectry. 9 , 110. Derkacheva, L. D. (1960b). Opt. Spectry. 9, 209. Derkacheva, L. D. (1963). Opt. Spectry. 15, 138. Derkacheva, L. D., and Petukhov, V. A. (1973). Soviet J . Quant. Electronics 2, 149.

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

217

Dewar, M. J. S., and Trinajstic, N. (1971).J. Chem. SOC. ( A ) 1220. Dhingra, R. C., and Poole, J. A. (1968).J. Phys. Chem. 72, 4577. Doty, J. C., Williams, J. L. R., and Grisdale, P. J. (1969). Can. J. Chem. 47, 2355. Easterly, C. E., and Christophorou, L. G. (1974). J.C.S. Faraday ZZ 70, 267. Easterly, C. E., Christophorou, L. G., Blaunstein, R. R., and Carter, J. G. (1970). Chem. Phys, Lett. 6, 579. Easterly, C. E., Christophorou, L. G., and Carter, J. G. (1973).J.C.S. Faraday ZZ 69, 471. Edward, J. T. (1964). Trans. R o y . SOC. Can. 2 , 313. Ellerhorst, R. H., and JaffC, H. H. (1968).J. Org. Chem. 33, 4115. Ellis, D. W., and Rogers, L. B. (1964). Spectrochim. Acta 20, 1709. Evans, D. F. (1957).J. Chem. SOC. 3885. Faure, J., Bonneau, R., and Joussot-Dubien, J. (1967). Photochem. Photobiol. 6, 331 Favaro, G., Mazzucato, U., and Masetti, F. (1971). Spectrochim. Acta 27A, 915. Favaro, G., Mazzucato, U., and Masetti, F. (1973).J. Phys. Chem. 77, 601. Fayet, M., and Wahl, P. (1971). Biochem. Biophys. A c t a 229, 102. Fernando, Q., and Schulman, S. (1967).J. Phys. Chem. 71, 2668. Fernando, Q., and Schulman, S. (1968). Tetrahedron 24, 1777. Fink, D. W., andKoehler, W. R. (1970). Anal. Chem. 42, 990. Flurry, R. L. (1966).J. Amer. Chem. S O C . 8 8 , 5393. Flurry, R. L., and Lykos, P. G. (1963).J. Amer. Chem. S O C . 85, 1033. Flurry, R. L., and Wilson, R. K. (1967).J. Phys. Chem. 71, 589. Forster, T. (1949). Naturwiss. 36, 186. Forster, T. (1950). Z. Electrochem. 54, 531. Fiirster, T. (195 1). “Fluoreszenz Organischer Verbindungen.” Vandenhoek und Ruprecht, Gottingen. Forster, T. (1970). Pure Appl. Chem. 24, 443. Forster, T. (1972). Chem. Phys. Lett. 1 7 , 309. Forster, T. (1973). Pure Appl. Chem. 34, 225. Geldof, P. A., Rettschnick, R. P. H., and Hoijtink, G. J. (1969). Chem. Phys. Lett. 4, 59. Godfrey, T. S., Porter, G., and Suppan, P. (1965). Dsic. Faraday SOC. 39, 194. Grabowska, A., and Pakula, B. (1966). Proc. Int. C o d . Lumin, 368. Grabowska, A., and Pakula, B. (1967). Chem. Phys. Lett. 1, 369. Grabowska, A,, and Pakula, B. (1969). Photochem. Photobiol. 9, 339. Grabowski, Z. R., Rotkiewicz, K., and Sadlej, A. J. (1966). Proc. Int. C o d . Lumin, 3 10. Griffiths, J. (1971).J. Chem. S O C . ( B ) 801. Gutsche, C. D,, and Oude-Alink, B. A. M. (1968). 1. Amer. Chem. S O C . 90, 5855. Gutsche, C. D., Oude-Alink, B. A.M., and Chan, A. W. K. (1973).J. Org. Chem. 38. 1993. Hagen, R., Heilbronner, E., Meier, W., and Seiler, P. (1967). Helv. Chim. A c t a 50,1523. Hammett, L. P., and Deyrup, A. J. (1932).J. Amer. Chem. SOC. 54, 2721. Harriman, A., and Rockett, B. W. (1973).J.C.S. PerkinZZ 1624. Haylock, J. C., Mason, S. F., and Smith, B. E. (1963).J. Chem. S O C . 4897. Henry, B. R., and Siebrand, W. (1973). In “Organic Molecular Photophysics”, Vol. I, 153 (J. B. Birks, ed.). Wiley, London and New York. Henson, R. M. C., and Wyatt, P. A. H. (1975).J.C.S. Faraday ZZ, 71, 669. Herbert, M. A., and Johns, H. E. (1971). Photochem. Photobiol. 14, 693.

218

J. F. IRELAND AND P. A. H. WYATT

Hercules, D. M., and Rogers, L. B. (1959). Spectrochim. A c t a 15, 393. Hine, K. E., and Childs, R. F. (1973).J. Amer. Chem. SOC. 95,6116. Hirota, K. (1962). 2. Physik. Chem. (Frankfurt) 35, 222. Hochstrasser, R. M., Lutz, H., and Scott, G. W. (1974). Chem. Phys. Lett. 24, 162. HGgfeldt, E. (1962). A c t a Chim. Scand. 16, 1054. Hopkinson, A. C., and Wyatt, P. A. H. (1967). J. Chem. SOC. ( B ) 1333. Hrdlovic, P., Bellus, D., and Lazar, M. (1968). Coll. Czech. Chern. Comm. 33, 59. Hurley, R., andTesta, A. C. (1967).J. Amer. Chem. SOC.89, 6917. Hussain, S. K., and Wyatt, ?. A. H. (1972). J.C.S. Faraday Z 68, 130. Idoux, J. P., and Hancock, C. K. (1968).J. Org. Chern. 33,3498. Ingram,K. C., andEl-Bayoumi, M. A. (1974).J. Amer. Chem. SOC.96, 1674. Ingram, K. C., El-Bayoumi, M. A., and Abu-Elgheit, M. (1971).J. Amer. Chem. SOC.93,5023. Ireland, J. F. (1972). Ph.D. Thesis, St. Andrews. Ireland, J. F., and Wyatt, P. A. H. (1972).J.C.S. Faraday I 6 8 , 1053. Ireland, J. F., and Wyatt, P. A. H. (1973).J.C.S. Faraday 1 6 9 , 161. Jackson, G., and Porter, G. (1961). Proc. R o y . SOC. Lond. A260, 13. Jaffd, H. H., and Jones, H. L. (1965).J. Org. Chem. 30,964. Jaffd, H. H., Beveridge, D. L., and Jones, H. L. (1964a). J. Amer. Chem. SOC.86, 2932. Jaffi, H. H., Jones, H. L., and Isaks, M. (1964b).J. A m e r . Chem. SOC.86, 2934. Janata, J., and Jansen, G. (1972). J.C.S. Faraday I 6 8 , 1656. Jones, J. R. (1973). “The Ionization of Carbon Acids.” Academic Press, London and New York. Kasha, M. (1950). Disc. Faraday S O C . 9, 14. Kato, S., Morita, M., and Koizumi, M. (1964). Bull. Chem. SOC.Japan 3 7 , 117. Klaning, U. K., Goldschmidt, C. R., Ottolenghi, M., and Stern, G. (1973). J. Chem. Phys. 59,1753. Knight, A. E. W., and Selinger, B. E. (1971). Spectrochim. A c t a 27A, 1223. Kokubun, H. (1957). Naturwiss. 44, 233. Kokubun, H. (1958). 2. Electrochem. 62,599. Konschin, H., Lunnala, R., and Sundholm, F. (1973). Suom. Kemistireuran, Tiedon 82, 8. Kovi, P. J., and Schulman, S. G. (1972). Spectrosc. Lett. 5 , 443. Kovi, P. J., and Schulman, S. G. (1973a). Anal. Chem. 45,989. Kovi, P. J., and Schulman, S. G. (1973b). Anal. Chim. Acta 63, 39. Kovi, P. J., Capomacchia, A. C., and Schulman, S. G. (1972a). Anal. Chem. 44, 1611. Kovi, P. J., Miller, C. L., and Schulman, S. G. (1972b). Anal. Chim. A c t a 61, 7. Kuz’min, M. G., Uzhinov, B. M., and Berezin, I. V. (1967). Zh. Fir. Khim. 41, 446. Kysel, 0. (1969). Kinet. Mech. Polyreactions, Int. Symp. Macromol. Chem. Prepr. 5,263. Labhart, H., and Heinzelmann, W. (1973). In “Organic Molecular Photophysics”, Vol. I, p. 297 0. B. Birks, ed.). Wiley, London and New York. Lasser, N., and Feitelson, J. (1973).J. Phys. Chem. 77, 1011. Ledger, M. B., and Porter, G. (1972).J.C.S. Faraday 1 6 8 , 539. Leonhardt, H., Gordon, L., and Livingston, R. (1971). J. Phys. Chem. 75,245. Lesclaux, R., and Joussot-Dubien, J. (1973). In “Organic Molecular Photophysics”, Vol. I, 457 0. B. Birks, ed.). Wiley, London and New York. Liedke, P., and Schulman, S. G. (1973a). 2. Physik. Chem. N.F. 84, 317. Liedke, P., and Schulman, S. G. (197313).Anal. Chim. A c t a 63, 197.

ACIDBASE PROPERTIES O F ORGANIC MOLECULES

219

Liler, M. (1971). “Reaction Mechanisms in Sulphuric Acid and Other Strong Acid Solutions.” Academic Press, London and New York. Lindqvist, L. (1960). Arkiv. Kemi 16, 79. Lindqvist, L. (1963).J. Phys. Chem. 67, 1701. Lindqvist, L., and Kasche, V. (1965). Photochem. Photobiol. 4, 923. Lippert, E. (1957). 2. Electrochem. 59, 962. Loken, M. R., Hayes, J. W., Gohlke, J. R., and Brand, L. (1972).Biochem. 11, 4779. Longworth, J. W., Rahn, R. O., and Shulman, R. G. (1966).J. Chem. Phys. 45, 2930. Marziano, N. C., Cimino, G. M., and Passerini, R. C. (1973). J.C.S. Perkin ZZ 1915. McCarthy, W. J., and Winefordner, J. D. (1967). In ‘ Fluorescence, Theory, Instrumentation and Practice” (G. G. Guilbault, ed.). Arnold, London. Mason, S. F., and Smith, B. E. (1969). J. Chem. Soc. ( A ) 325. Mason, S. F., Philp, J., and Smith, B. E. (1968).J. Chem. Soc. ( A ) 3051. Mayer, R. T., and Himel, C. M. (1972). Biochem. 11, 2082. Morton, R. A., Hopkins, T. A., and Seliger, H. H. (1969).Biochemistry 8, 1598. Murrell, J. N. (1963). “The Theory of Electronic Spectra of Organic Molecules”, p. 123. Methuen, London. Nakamaru, K., Nuzuma, S., and Koizumi, M. (1969). Bull. Chem. Soc. Japan 42, 255. Neta, P. (1975). Adv. Phys. Org. Chem. 12, 223. Noyes, R. M. (1961). Progr. Reaction Kinetics 1, 129. Ofran, M., and Feitelson, J. (1973). Chem. Phys. Letts. 19, 427. Ottolenghi, M. (1973). Accounts Chem. Res. 6 , 153. Pace, I., and Schulman, S. G. (1972).J. Phys. Chem. 76,1996. Parker, C. A. (1968). “Photoluminescence of Solutions.” Elsevier, London. Paul, W. L., and Schulman, S. G. (1973). Anal. Chem. 4 5 , 4 1 5 . Paul, W. L., and Schulman, S. G. (1974). Anal. Chim. Acta 69,195. Paul, W. L., Kovi, P. J., and Schulman, S. G. (1973). Spectrosc. Lett. 6, 1. Peter, H. H., and Drewer, R. J. (1971). Photochem. Photobiol. 14,561. Porter, G. (1967). In “Reactivity of the Photoexcited Organic Molecule”, p. 79, Interscience, London. Porter, G., and West, M. A. (1973). In “Techniques of Chemistry”, p. 367 (G. G. Hammes, ed.). Wiley, New York. Postle, M. J., and Wyatt, P. A. H. (1972).J.C.S. PerkinZZ474. Rayner, D. M., and Wyatt, P. A. H. (1974). J.C.S. Faraday Z 7 0 , 945. Reid, C. (1957). “Excited States in Chemistry and Biology.” Buttenvorths, London. Reid, D. H., Zahradnik, R., and Tichy, M. (1968). Tetrahedron 24, 3001. Rochester, C. H. (1970). “Acidity Functions.” Academic Press, London and New York. Rosebrook, D. D., and Brandt, W. W. (1966).J. Phys. Chem. 70, 3857. Rosenberg, J. L., and Brinn, I. (1972).J. Phys. Chem. 76,3559. Rotkiewicz, K., and Grabowski, Z. R. (1969). Trans. Faraday SOC. 65,3263. Sandorfy, C. (1951). Compt. Rend. Acad. Sci. 260, 13. Schulman, S. G . (1971a). Rev. Anal. Chem. 1, 85. Schulman, S. G. (1971b). Anal Chem. 43,285. Schulman, S. G . (1971c).J.Phurm. Sci. 60,371,628. Schulman, S. G. (1972). Anal. Chem. 4 4 , 4 0 0 . Schulman, S. G. (1973a). Spectrosc. Lett. 6 , 197. Schulman, S. G. (197313).Spectrosc. Lett. 6 , 5 1 3 . Schulman, S. G., and Capomacchia, A. C. (1972). Spectrochim. Acta 28A, 1.

220

J. F. IRELAND AND P. A. H. WYATT

Schulman, S. G., and Gerson, H. (1968).J. Phys. Chem. 72, 3297. Schulman, S. G., and Kovi, P. J. (1973). Anal. Chim. Acta 67,259. Schulman, S. G., and Sanders, L. B. (1971). Anal. Chim. Acta 56,83. Schulman, S. G., and Winefordner, J. D. (1970). Talanta 17, 607. Schulman, S. G., and Young, J. F. (1974). Anal. Chim. Acta 70, 229. Schulman, S. G., Capomacchia, A. C., and Rietta, M. S. (1971a). Anal. Chim. Acta 56, 91. Schulman, S. G., Capomacchia, A. C., and Tussey, B. (1971b). Photochem. Photobiol. 14, 733. Schulman, S. G., Kovi, P. J., and Miller, C. L. (1972). Anal Chim. Acta 62, 59. Schulman, S. G., Kovi, P. J., and Young, J. F. (1973c).J. Pharm. Sci. 62, 1197. Schulman, S. G., Kovi, P. J., McVeigh, H., and Torosian, G. (197313). Spectrosc. Lett. 6 , 77. Schulman, S. G., Rutledge, J. M., and Torosian, G. (1974). Anal. Chim. Ac ta 70, 455. Schulman, S . G., Capomacchia, A. C., Paul, W. L., Kovi, P. J., and Young, J. F. (1973a). 2.physik. Chem. N.F. 87, 308. Seliskar, C. J., and Brand, L. (1971).J. Amer. Chem. SOC. 93, 5405. Shank, C. V., Dienes, A., Trozzolo, A. M., and Myer, J. A. (1970). Appl. Phys. Lett. 16, 405. Simic, M., and Hoffman, M. Z. (1972).J. Phys. Chem. 76, 1398. Smith, B. E. (1969).J. Chem. S O C . ( A ) ,2873. Sommer, U., and Kramer, H. E. A. (1971).Photochem. Photobiol. 13,387. Song, P-S., (1968).Photochem. Photobiol. 7, 311. Song, P-S., Sun, M., Koziolowa, A., and Koziol, J. (1974).J. Amer. Chem. SOC. 96,4319. Stenberg, V. I., and Dutton, D. R. (1972). Tetrahedron 28,4635. Stickler, S . J., and Berg, R. A. (1962).J. Chem. Phys. 37, 814. Stryer, L. (1966).J. Amer. Chem. S O C . 88,5708. Taft, R. W. (1960a).J. Phys. Chem. 64, 1805. Taft, R. W. (1960b).J. Amer. Chem. S O C . 82, 2965. Tramer, A. (1970).J. Phys. Chem. 74, 887. Trieff, N. M., and Sundheim, B. R. (1965).J. Phys. Chem. 69,2044. Trotter, W., and Testa, A. C. (1970).J. Phys. Chem. 74, 845. Tsutsumi, K., Aoki, K., Shizuka, H., and Morita, T. (1971). Bull. Chem. SOC. Japan 44,3245. Tyutyulkov,N., and Hiebaum, G. (1969). Theoret. Chim. Acta 14,39. Tyutyulkov, N., Fratev, F., and Petkov, D. (1967). Theoret. Chim. Acta 8, 236. Udenfriend, S. (1962). “Fluorescence Assay in Biology and Medicine.” Academic Press, New York. Udenfriend, S. (1967). Fluorescence, 359. Umberger, J. Q. (1968).J. Phys. Chem. 72, 1350. Vander Donckt, E. (1969). Bull. SOC. Chim. Belges 73, 69. Vander Donckt, E. (1970). Progr. Reaction Kinetics 5,273. Vander Donckt, E., and Porter, G. (1968a). Trans. Faraday S O C . 64,3218. Vander Donckt, E., and Porter, G. (1968b). Trans. Faraday SOC. 64,3215. Vander Donckt, E., Draimaix, R., Nasielski, J., and Vogels, C. (1969a). Trans. Faraday SOC. 65, 3258. Vander Donckt, E., Nasielski, J., and Thiry, P. (196913). Chem. Comm. 1249. Vander Donckt, E., Lietaer, D., and Nasielski, J. (1970). Bull. SOC. Chim. Belges 79,283. Ware, W. R., Shukla, P. R., Sullivan, P. J., and Bremphis, R. V. (1971).J. Chem. Phys. 55,4048.

ACID-BASE PROPERTIES O F ORGANIC MOLECULES

221

Watkins, A. R. (1971). Z. physik. Chem. N.F. 75, 327. Watkins,A. R. (1972a). J.C.S. F a r a d a y Z 6 8 , 28. Watkins, A. R. (197213). Z. physik. Chem. N.F. 78, 103. Watkins, A. R. (1973).J. Phys. Chem. 77, 1207. Weber,K. (1931). Z. Phys. Chem. B15, 18. Wehry, E. L. (1967).J. A m e r . Chem. SOC.8 9 , 4 1 . Wehry, E. L., and Rogers, L. B. (1965a). Spectrochim. A c t a 21, 1976. Wehry, E. L., and Rogers, L. B. (1965b).J. A m e r . Chem. S O C . 87,4234. Wehry, E. L., 2nd Rogers, L. B. (1966).J. A m e r . Chem. SOC.88, 351. Weisstuch, A., and Testa, A. C. (1968).J. Phys. Chem. 72, 1982. Weller, A. (1952). 2. Elektrochem. 56, 662. Weller, A. (1954). 2. Elektrochem. 58, 849. Weller, A. (1955) 2. physik. Chem. N . F., 3, 238 Weller, A. (1956). 2. Elektrochem. 60, 1144. Weller, A. (1957a). 2. Elektrochem. 61,956. Weller, A. (1957b). 2. physik. Chem. N.F. 13, 335. Weller, A. (1958a). 2. physik. Chem. N.F. 15, 438. Weller, A. (195813). Z p h y s i k . Chem. N.F. 17, 224. Weller, A. ( 1 9 5 8 ~ )Z. . physik. Chem. N.F. 18, 163. Weller, A. (1961). Progr. Reaction Kinetics 1, 189. Weller, A., and Urban, W. (1954). Angew. Chem. 66, 336. Weller, A., and Urban, W. (1963). Ber. Bunsen Gesellschaft Phys. Chem. 67, 787. Wemer, T. C., and Hercules, D. M. (1969).J. Phys. Chem. 73, 2005. Wemer, T. C., and Hercules, D. M. (1970).J. Phys. Chem. 74, 1030. Williams, D. L., and Heller, A. (1970).J. Phys. Chem. 74, 4473. Winefordner, J. D., Schulman, S. G., and Sanders, L. B. (1971a). Photochem. Photobiol. 13, 381. Winefordner, J. D., Schulman, S. G., Tidwell, P. T., and Cetorelli, J. T. (1971b). J . A m e r . Chem. S O C . 93, 3179. Wooldridge, J., and Roberts, T. D. (1973). Tetrahedron L e t t . 4007. Wubbels,G. G., Jordan, J. W., and Mills, N. S. (1973).J. A m e r . Chem. S O C . 95, 1281. Wyatt, P. A. H. (1957). Discuss Faraday S O C . 24, 162. Yagupol’skii, L. M., and Gandel’sman, L. Z. (1965).J. Gen. Chem. (U.S.S.R.) 35, 1259. Yagupol’skii, L. M., and Gandel’sman, L. Z. (1967). J. Gen. Chem. (U.S.S.R.) 37,1992. Yakatan, G. J., and Schulman, S. G. (1972).J. Phys. Chem. 76,508. Yakatan, G. J., Juneau, R. J., and Schulman, S. G. (1972a). Anal. Chem. 44, 1044. Yakatan, G. J., Juneau, R. J., and Schulman, S. G. (1972b). J. Pharm. S c i 61, 749. Yates, K., and Wai, H. (1964).J. A m e r . Chem. S O C . 86, 5408. Yates, K., and McClelland, R. A. (1973).J. A m e r . Chem. S O C . 95, 3055. Yates, K., Wai, H., Welch, G., and McClelland, R. A. (1973).J. A m e r . Chem. S O C . 95,418. Young, J. F., and Schulman, S. G. (1973). Talanta 20,399. Zalis, B., Capomacchia, A. C., Jackman, D., and Schulman, S. G. (1973). Talanta 20,33. Zimmerman, H. E., Werthemann, D. P., and Kamm, K. S. (1973). J. A m e r . Chem. SOC. 95,5094. Zimmerman, H. E., Werthemann, D. P., and Kamm, K. S. (1974). 1. Amer. Chem. SOC.96, 439.

This Page Intentionally Left Blank

Application of Radiation Techniques to the Study of Organic Radicals P. NETA Radiation Research Laboratories and Department of Chemistry, Mellon Institute of Science, Carnegie-Mellon University, Pittsburgh, Pennsylvania 15213, U.S.A.

1. Introduction 2. Techniques of Radiation Chemistry . Product Analysis by Liquid Chromatography . Pulse Radiolysis . Kinetic Spectroscopy Conductometry . Polarography . Electron Spin Resonance . 3. Reactions of Organic Compounds with Transients from Water Radiolysis of Water and Manipulation of Primary Radicals Reactions of Organic Compounds with Hydrated Electrons Reactions of Organic Compounds with Hydrogen Atoms Reactions of Organic Compounds with Hydroxyl Radicals Comparison of Reactions of eiq,, H, OH, and 0 - and their Application for Radical Production . 4. Optical Absorption Spectra of Organic Radicals . . 5. Electron Spin Resonance Studies 6. Acid-Base Equilibria of Organic Radicals . . Methods of Determination of Dissociation Constants Comparison of pK-Values for Radicals and Parent Molecules Correlation of Structure and Dissociation Constants Kinetics of Acid Dissociation . 7. Kinetics and Mechanisms of Radical Reactions . Radical- Radical Reactions Radical Reactions With Solutes . Radical Reactions with Solvent and Intramolecular Reactions 8. Concluding Remarks . . References .

..

223

2 24 225 225 226 227 228 228 229 230 230 233 234 236 237 243 247 253 253 255 266 269 270 271 277 283 289 291

224

P. NETA

1. INTRODUCTION

Although radiation chemistry has been mainly concerned with investigating the chemical reactions taking place in irradiated solutions many studies have been carried out from the point of view of the physical organic chemist. The understanding of the primary processes in many irradiated systems is sufficiently advanced to allow the use of such systems for studies in other fields of chemistry. The radiation techniques are no longer tools for the study of radiation chemistry alone. Recent applications of radiation techniques to biochemistry and biology, to inorganic chemistry, to polymer and physical organic chemistry have proved very fruitful. These techniques can therefore be considered as additional advanced tools for research. The purpose of the present review is to familiarize organic chemists with the methods of research used by radiation chemists and to point out how these methods have been and can be used for studies in physical organic chemistry. Examples which characterize the various approaches are taken mostly from the literature of the last four to five years. Summaries of earlier studies have appeared in several books and reviews on the various aspects of radiation chemistry . The review by Fendler and Fendler (1970) on the application of radiation chemistry to mechanistic studies in organic chemistry covers most of the relevant literature prior to 1970. Recently published books include “The Hydrated Electron” by Hart and Anbar (1970), “The Radiation Chemistry of Water”, by Draganik and Draganid ( 19 71), “Principles of Radiation Chemistry”, by O’Donnell and Sangster (1970), “Introduction to Radiation Chemistry” by Swallow (1973), and “Einfuhrung in die Strahlenchemie” by Henglein et al. (1969). Reviews can be found in the series Advances in Radiation Chemistry, edited by Burton and Magee (since 1969), Current Topics in Radiation Research, edited by Ebert and Howard (since 1965), Radiation Research Reviews, edited by Phillips and Cundall (since 1968), and in other review series more familiar to chemists. In order to introduce the various radiation techniques without detailing all the basic principles, recent developments will be summarized with a brief mention of the now classical approaches. The chemistry itself will concentrate on aqueous solutions because most of the work has been done on such systems. The primary radiation

APPLICATION OF RADIATION TECHNIQUES

225

chemical processes are better understood in aqueous solutions than in any other system and they can therefore, be better controlled and utilized. The first part of the chemistry discussion will deal with the reactions of organic compounds with transients from water (H, OH, e,,), their rate constants and mechanisms, and their utilization for the production of organic radicals. The later chapters will deal with the properties or organic radicals, their absorption spectra, their dissociation constants and various protonation processes, any structural rearrangements they undergo, and their reactivity towards each other and towards organic compounds.

2. TECHNIQUES OF RADIATION CHEMISTRY The most commonly used sources of radiation are the 6 o C o gamma source for continuous irradiation and pulsed high-energy (>1 MeV) electron beams for fast kinetic studies. Detailed descriptions of several such sources and accelerators are given in numerous books, as are the various methods used by radiation chemists for dosimetry, sample preparation and irradiation, and common product analysis. Several new developments in the analytical procedures, both in the determination of final products and in the direct observation of transient species, will be discussed below.

Product Analysis b y Liquid Chromatography Permanent chemical changes in irradiated solutions have -1ially been deduced from overall changes in optical absorption or froill simple analytical determinations of products, such as hydrogen peroxide, hydrogen, nitrogen, or other inorganic products. Determination of halide ion yields has received greater attention following the recent development of highly sensitive ion-selective electrodes. While absorption spectroscopy has been used, very few organic products have been determined directly in solution, and in most cases separation is necessary. Gas or paper chromatography and related methods have often been used, but these are usually quite tedious in that the solvent has to be removed, or the products have to undergo chemical changes before separation. Moreover, the limited sensitivity dictates relatively high conversions, resulting in lower accuracy in the

226

P. NETA

determination of the initial yields. Newly developed high-speed liquid chromatographic methods overcome many of the problems in the previous techniques and have been successfully used in several studies (Bhatia, 1973; Bhatia and Schuler, 1973a, b). For analysing products in irradiated aqueous solutions, eluents can be used which contain water as their main component so that the irradiated solution can be directly injected into the stream without prior treatment. Products, of course, do not have to be volatile. In many instances, particularly those involving systems containing aromatic or heterocyclic rings, the products can be easily detected by their optical absorption at very low conversions. For example, for the case of phenol accurate determination was possible even at M (Bhatia, 1973), and other systems have been studied at even lower levels (Schuler et al., 1974). Other detection systems can also be applied, such as conductivity or counting of radioisotope-labelled compounds. However, the sensitivity of the latter systems has not been thoroughly tested yet. Liquid chromatography can also be applied advantageously to the examination of reactant disappearance. For example, it has been used t o determine the disappearance of 5-bromouracil in solutions which contain excess uracil (Bhatia and Schuler, 1973b). Such a determination is, of course, impossible by straight spectrophotometry.

Pulse Radio lysis This technique allows monitoring of short-lived species produced as a result of a very short pulse of radiation and can utilize various detection methods. Recently, with the application of on-line computer methods, this technique has received renewed attention in several laboratories. Patterson and Lilie (1974) have developed a system in which the computer is used not only for storage and treatment of data (see e.g. Aldrich et al., 1972), but also for controlling the actual experiment. In this system the experiment is defined through the computer, which then controls and properly sequences the monochromator setting, the analysing light pulsing, the accelerator pulse, and the detection system. Data on the transient signal are then rapidly reduced to optical absorbance and stored for subsequent manipulation. The signal-to-noiseratio can be enhanced by averaging over a number of pulses. The rapid treatment of the data by the computer allows one to obtain all the relevant information, a

APPLICATION O F RADIATION TECHNIQUES

227

great part of which might have been lost in manual calculations. For example, “three-dimensional” spectra can be obtained which show the changes with time of the optical absorption over a wide range of wavelengths. Furthermore, the immediate availability of partly processed data allows the investigator to plan and carry out further experiments without delay. Another aspect of pulse radiolysis which has been improved is the pulse duration. For most experiments of interest to the physical organic chemist the common machines with pulse durations of s are quite satisfactory, though for certain reactions, such as those involving protonation, examination on a shorter time scale can be of value. Several accelerators which supply nanosecond pulses are currently in use, but they are employed mostly with microsecond detection systems. Work in the l o - ’ 2 - lo-’ s region has recently become possible by the stroboscopic technique utilizing the fine structure pulses from a linear accelerator (Bronskill et al., 1970). More recently, a system which produces a single pulse of 40 picoseconds has been constructed (Ramler et al., 1975) and utilized for the observation of hydrated electrons at very short times (Jonah et al., 1973). Various detection methods can be used with pulse radiolysis, and the recent developments in these methods are discussed below.

Kinetic Spectroscopy Until recently, kinetic spectrophotometry has been carried out by passing the analysing light beam through a monochromator and observing the transient formation or decay kinetics at one wavelength at a time. Spectrography with photographic plates did not allow time resolution. A new technique, developed independently in two laboratories (Gordon et al., 1975; Pagsberg and Hansen, 1975), gives time-resolved spectra after a single pulse of radiation. The analysing light is passed through the solution, then a spectrograph, and on to an image-converter tube. The spectra at different times are displayed on the fluorescent screen which is scanned by a TV camera and the information fed into a computer. Although it probably has lower wavelength resolution as compared to the use of monochromators, this method saves time and chemicals and allows the automation of pulse machines capable of single pulses only.

228

P. NETA

Co nduc to m e try Many radiation-induced reactions are accompanied by changes in conductance, mainly through formation or neutralization of H+ or OH- ions, and these changes can be monitored as a means of following such reactions. The method is especially sensitive in determining the state of protonation of transient species. In the original technique of pulse conductivity (Schmidt and Buck, 1966; Beck, 1969) changes in the d.c. conductance were observed. This approach is satisfactory for work with solutions having very low initial conductivities and is in general restricted to observations at pH 4- 6 and pH 8- 10. A more generally useful approach employs an a.c. method (Lilie and Fessenden, 1973) and is applicable to solutions between pH 2 and pH 12, or solutions containing other ionic solutes at concentrations < l o W 2M . A high frequency (10 MHz) is used so that the time resolution is close to 1 ps. The sensitivity of the a.c. conductivity is somewhat lower than that of the d.c. method. Accurate measurements of changes in hydrogen ion concentration M with the a.c. and M with the can be achieved down to d.c. method. Both techniques have been applied to various chemical problems. Summaries of the various techniques and some of their applications are available (Asmus, 1972; Schmidt, 1972; Lilie and Fessenden, 197 3).

Po larograp hy The combination of polarography with pulse radiolysis has only recently been developed (Lilie et al., 1971; Lilie, 1972; Gratzel et al., 1972) and a large volume of data is accumulating (see e.g. Gratzel and Henglein. 1973; Gratzel et al., 1973a,b; Bansal et al., 1973; Bansal and Henglein, 1974; and references therein). So far this technique has been used to monitor the polarographic behaviour of many types of radicals and it gives information on the oxidation and reduction of these short-lived species at the electrode. Half-wave potentials for many radicals have been derived and some insight into the mechanism of reaction of these radicals with other molecules and with themselves has been achieved. With further data available in the future one will be able to gain information on the chemistry in the irradiated solution from the polarograms of the radical.

APPLICATION OF RADIATION TECHNIQUES

2 29

Electron Spin Resonance Electron spin resonance observation of organic radicals during in situ radiolysis of solutions was initiated by Fessenden and Schuler (1960) (their major work on hydrocarbon solutions was published in 1963). A review on the e.s.r. spectra of radiation-produced radicals summarizes the literature up to 1968 (Fessenden and Schuler, 1970). However, e.s.r. observation of radicals during irradiation of aqueous solutions was not achieved until 1968 (Eiben and Fessenden, 1968; Avery et al., 1968). Two e.s.r. techniques are currently in use, namely the steady-state and the time-resolved methods. With the steady-state method one records the e.s.r. spectrum while the solution flowing through a cell located in the e.s.r. cavity is continuously irradiated with a d.c. electron beam (see the description of Eiben and Fessenden, 1971). Under such conditions one may observe spectra of many radicals produced in solution, whether they are of major or minor importance in the primary processes or result from secondary reactions. These studies have proven to be very valuable in determining the structure of radicals and the chemical reactions taking place in many systems. One of the main factors determining the intensity of the spectrum, and whether a radical is observed at all, is the steady-state concentration of the radical which in turn depends on the rates of production and disappearance. Another factor is the number of magnetic nuclei in the radical which cause splitting of the resonance line with a corresponding reduction in intensity. Line-broadening due to proton exchange or anisotropy results also in decreased signal height. These factors can cause a radical produced by a minor reaction of the solute or by a secondary reaction of a product to be observed instead of the major primary radical. This limitation is being overcome by development of pulsed methods giving e.s.r. spectra resolved in time (Fessenden, 1973; Verma and Fessenden, 1973; Avery et al., 1968; Smaller et al., 1971; Nucifora et al., 1972). Although they have not yet reached the sensitivity of the steadystate method, these techniques have, at the moment, a signal-to-noise ratio only an order of magnitude poorer. The advantage of time resolution can, however, be masked by CIDEP (chemically induced dynamic electron polarization) effects, i.e. non-equilibrium population of the electron-nuclear spin states, which results in initially abnormal line intensities changing with time towards equilibrium conditions.

230

P. NETA

The great advantage of the e.s.r. technique over all the previous methods lies in the specific nature of the e.s.r. spectra, i.e. they give the number of magnetic nuclei in the radical, and with the accumulated knowledge in the field one can determine the identity and position of these nuclei with a high degree of certainty. Radicals with chemically similar structures might show identical behaviour when studied by spectrophotometry, polarography, or other similar techniques. The e.s.r. spectra, however, will show up even minimal difference in their structure, such as geometrical isomerism. For example, the three isomers formed upon addition of OH to benzoate have been simultaneously determined by e.s.r. (Eiben and Fessenden, 197 1) while optical pulse radiolysis shows one spectrum only (Wander e t al., 1968). Further details on the steady-state and time-resolved e.s.r. methods and considerations of the design of experiments, along with some representative results, were summarized by Fessenden (1975) at a recent conference.

3. REACTIONS O F ORGANIC COMPOUNDS WITH TRANSIENTS FROM WATER The primary effect of radiation on aqueous solution is the decomposition of the water, followed by reactions of the transients from water with the solutes present. Direct effect of radiation on the solute is practically negligible up to concentrations of about 1 M , and because we are usually dealing with reasonably dilute solutions the discussion will be restricted to these “indirect” effects. It is, therefore, important to know what radicals are produced in irradiated water, how they react with the organic solutes, and how they can be manipulated for the production of certain organic radicals. These subjects have been studied very thoroughly and are sufficiently well understood to enable us to use the radiolysis of aqueous solutions for studies of diverse chemical problems.

Radiolysis of Water and Manipulation of Primary Radicals The overall process of water radiolysis is usually formulated as in eqn (1). In a simplistic way the radiation can be considered to H20

--

OH, H, eiq, H’, HzOz,

H2

(1)

23 1

APPLICATION OF RADIATION TECHNIQUES

produce radicals in “spurs”, some of which combine within the spur to yield the “molecular” products, hydrogen and hydrogen peroxide, with the rest diffusing out to react with solutes. At low concentrations ( < W 3M ), solutes react principally with the diffusing transients in the bulk of the solution and do not significantly affect the yield of the molecular products. At higher concentrations, however, they may interfere with the early stages of reactions by scavenging radicals from the “spurs” (see e.g. Balkas e t al., 1970, and references therein) and decreasing the molecular yields. The yields are expressed by G-values, defined as the number of atoms or molecules produced or destroyed by 100 eV of energy absorbed in the solution. At low solute concentrations the generally accepted G-values are 2.8 for OH, 0.6 for H, 2.8 for eiq , 0.7 for H 2 0 2 , and 0.4 for H,. Detailed evaluation of these numbers and the methods of their determination can be found in radiation chemistry books. In order to study a certain desired process it is convenient to know how the initial transients can be manipulated. By choosing the right conditions one can convert one transient into another or separate one reaction from another. The use of radical “scavengers” is, therefore, very popular and the important ones are discussed below. The hydrated electron and the hydrogen atom can be considered the basic and acid forms of the reducing species produced in the irradiation of water. Interconversions are possible by reaction (2) eiq + H30’

-+

H + H20

which has a diffusion controlled rate (2 x 10” M reaction (3) which has a rate constant of only -lo7 H + OH-

-+

eiq + H2O

(2) s-’) and by M s - l . The

-’

(3)

reverse reactions of both (2) and (3) are very slow; for the latter case s - l . It is thus possible to the rate constant was found to be 16 M convert all eaq into H by operating in acid solutions and to convert H into e,, by using high pH. The conversion of e;, into H can also be achieved in neutral solution by the use of other proton donors, such

-’

ei, + H2P04

-+

H + HPOa-

(4)

as H2P0, in (4), although the rate constant for this reaction -5 x lo6 M s - l is not nearly as high as that of reaction (2). The hydrated electron can be conveniently converted into a hydroxyl radical by reaction with nitrous oxide [reaction ( 5 ) ] .

-’

eaq + N 2 0 + H2O

+

OH + OH- + N2

(5)

232

P. NETA

Nitrous oxide is relatively soluble in water (2 x lo-' M at atmospheric pressure), it reacts very rapidly with eiq (-7 x l o 9 M - l s-') and is practically inert toward OH and H (for the latter, k < l o 4 M - ' s-'). Because of partial scavenging of eaq from the spurs G(0H) in N,O-saturated solutions is 6.0. Hydrogen peroxide also converts e,, into OH very efficiently [reaction (6)] , but it has eiq + H202

+ OH

+ OH-

(6)

only a limited usefulness because it reacts also with H, OH, and organic radicals, though at lower rates. Conversion of OH into H (or eiq in alkaline solutions) can be carried out by reaction with hydrogen [reaction (7)]. However, the OH+H2 + H + H 2 O

(7)

-'

relatively low rate constant ( k , = 5 x l o 7 M s - l ) and the limited solubility of H,(-10-3 M at 1 atm) make this process useful only under very high pressures of hydrogen (-100 atm) and as a result only a limited number of experiments have been carried out by this method. The more common way of eliminating OH in order to study selectively the reactions of H or eiq is to scavenge it with t-butyl alcohol. In general t-butyl alcohol is used as a selective scavenger of OH (k = 5 x l o 8 M s-'); it leaves Hbehind ( k = 8 x l o 4 M s-l) to react with other added solutes. A relatively inert radical, kH,C(CH,),OH, is produced. Other alcohols can be used for scavenging both OH and H [reactions (8) and (9)] without affecting

-'

-'

RH + H -+

k + H~

(9)

e i q . Isopropyl alcohol scavenges OH and H quite effectively and, moreover, the resulting radical (CH,), COH may enter into electron transfer reactions to reduce the solute in a fashion similar to reduction by e i q . Ethylene reacts rapidly with both OH and H by addition to produce radicals which do not readily reduce other solutes. Many other scavengers have been used by radiation chemists but the above examples are the most important and the most commonly used ones and are sufficient for demonstrating the general approach.

APPLICATION OF RADIATION TECHNIQUES

233

Reactions of Organic Compounds with Hydrated Electrons The development of pulse radiolysis in 1960 and the observation of the intense transient optical absorption of eaq in 1962 led to a large number of studies concerning the rate constants and the mechanism of reactions of e i q . By the end of that decade a huge amount of information had been accumulated and summarized in a book by Hart and Anbar (1970). The reactions of organic compounds with hydrated electrons were specifically reviewed by Anbar (1969). The basic experiment in many of these studies involved the measurement of the lifetime of the transient eiq absorption (Amax = 715 nm, E = 18,500 M cm-') in the presence of varying concentrations of a substrate. By 1967 Anbar and Neta had compiled some 600 rate constants. More recently, the rates for some 700 compounds measured mostly by pulse radiolysis, have been summarized (Anbar et al., 1973). The reactivity of an organic compound toward eiq depends on its functional groups because the main hydrocarbon chain 11s nonreactive. Aliphatic alcohols, ethers, and amines are also nonreactive ( k < lo6 M s-'), although alkylammonium ions show a slight reactivity and can transfer a proton to the hydrated electron. Isolated double bonds are practically nonreactive, for ethylene s - ' , but conjugated systems or double bonds k < 2.5 x l o 6 M with an electron withdrawing group attached to them are very reactive. For example, butadiene and acrylic acid react with practiM s-' ). cally diffusion controlled rates (-10' Aldehydes and ketones are all very reactive toward eaq. However, when the carbonyl goup is attached to substituents other than alkyl the reactivity changes considerably. A correlation of the rate constants with the sum of Taft's U* values for substituents adjacent to the carbonyl group showed that the reactivity decreases when the substituents are more strongly electron withdrawing. It has been suggested, therefore, that the effect of such groups is to shorten the C=O bond length which results in increased 7r-electron density and thus decreased tendency to accept an additional electron. In the case of esters and amides, however, the correlation with Taft's u* values showed enhancement by electron withdrawing groups. The mechanism of reaction was suggested to involve addition of the electron to a positive centre rather than to the 7r-system (Anbar, 1969; Hart and Anbar, 1970, and references therein).

-'

-'

-'

-'

234

P. NETA

Carboxylic acids are generally nonreactive in their anionic form. The reactivity of the acid form is strongly enhanced by electron withdrawing groups (Peter and Neta, 1972; Midit and Markovid, 1972). It was concluded, therefore, that the electron adds predominantly to a positive centre on the OH oxygen and not to the carbonyl double bond. The reactivity of nitriles toward e i q is affected by substituents similarly to the effect in carboxylic acids (Draganit et al., 1973). The reactivity o f halogen compounds, of course, strongly depends on the halogen. Fluoroaliphatic compounds are nonreactive unless they contain another reactive functional group. Chloro-compounds s - l , and their reactivity increases in are fairly reactive, 12 > l o 8 M the presence of neighbouring electron withdrawing groups. Bromoand iodo-compounds are more reactive in that order. Thiols and disulphides are very reactive, whereas thiol anions and thioethers are only fairly reactive. Nitro- and nitroso-compounds are very reactive toward e i q . The rate constants for the reactions of eiq with aromatic compounds depend very strongly on the substituent. Benzene itself has a s-’ , phenol and aniline relatively low reactivity, k = 1.4 x l o 7 M are even less reactive, but substituents such as CN, Br, and NO2 increase the rate constant almost to the diffusion controlled limit. The rate constants for several series of mono- and disubstituted benzenes were correlated with Hammett’s substituent constants u and showed good linear relationships. A few heterocyclic compounds have also been studied. Discussion of these compounds and a detailed treatment of the systems mentioned above together with several others can be found in the reviews cited (Anbar, 1969; Hart and Anbar, 1970).

-’

-’

Reactions o f Organic Compounds with Hydrogen Atoms A brief review of the reactions of hydrogen atoms in aqueous solutions has been published (Neta, 1972a). Rate constants for these reactions have been measured b y several techniques and a compilation of the data is available (Anbar et al., 1974). Many relative rate constants have been determined by “classical” competition kinetics and product analysis, usually measuring G(H2) or G(H, )/G(HD). Pulse radiolysis enabled the establishment of an absolute scale for all the previous relative rates, but has been Esed directly with only a

235

APPLICATION O F PADIATION TECHNIQUES

limited number of compounds for this purpose. The recent development of an e.s.r. technique which permits determination of rate constants by direct observation of the H atom signals under steadystate conditions (Neta et al., 1971) gave rise to a number of studies on the rates of H with organic compounds (for summary see Neta, 1972a). The rate constants for the reactions of H with organic compounds depend not only on the functional groups, as in the case of e i q , but also on the carbon chain because many of the reactions involve hydrogen abstraction from it. Hydrocarbons can be relatively unreactive; the rate constant for methane is < l o 5 M s - l . However, their reactivity increases with chain length and with branching and reaches a plateau at -10' M-' s - ' . Methyl groups show little reactivity, while CH2 groups have partial rate constants -1 x 1O7 M s - l , and CH groups -2 x l o 7 M - ~s - l . The ratio of reactivities of primary : secondary : tertiary hydrogen is approximately 1:15:6O. Hydrogen atoms appear to react with organic compounds more selectively than organic radicals and also more selectively in water than in organic solvents. The reaction of H with alcohols, esters, amides, amines, and &mino acids takes place on the aliphatic chain (hydrogen abstraction) and the rate is somewhat affected by the substituent. Specifically, the rate of abstraction from a certain CH bond is greatly retarded by a neighbouring ammonium group and slightly enhanced by amino- and hydroxyl groups. Abstraction from SH-groups is very rapid. Olefinic and aromatic compounds are also among the most reactive ones; they s - l . Addition takes react by addition with rate constants >lo9 M place also to carbonyl and cyano-groups but the rates are only moderate. Bromo- and iodo-aliphatic compounds react quite rapidly; k los and l o 9 M s - l respectively, and the reaction is predominantly halogen abstraction. Chlorine abstraction is a slower process and 1O6 -1 O7 usually takes place concurrently with H-abstraction, k M s-l. Fluoro-compounds undergo only H-abstraction at relatively low rates depending on the aliphatic chain structure. The effect of substituents on the rate constant for reaction of H with benzene was also examined. Correlation with Hammett's substituent constants was possible and showed that the hydrogen atom behaves as a slightly electrophilic reactant (Neta, 1972a, and references therein; see also Brett and Gold, 1971, 1973; and photochemical studies by Pryor et al., 1973).

-'

-

-'

-'

hr

236

P. NETA

Reactions of Organic Compounds with Hydroxyl Radicals The reactions of OH, like those of H, involve mainly addition to unsaturated sites and abstraction from saturated compounds. Oxidation or reduction by OH or H, respectively, involving charge transfer has not been definitely established with any organic compound. While the relative reactivities of M and OH follow similar patterns, the absolute rate of reaction with OH is generally higher than that with H. Relative rate constants for reactions of OH have been measured by competitive methods in y-irradiated solutions where product formation or reactant destruction have been monitored. These methods have generally been of low accuracy and can sometimes be misleading because of the possible complications in the processes between the initial reaction and the final products. Several competitors that allow the competition at the initial step to be followed became available for use with the pulse radiolysis technique in 1965 (Adams et al., 1965). Most of the rate constants for OH reported in the literature have been determined by this method. Numerous rates have also been determined by pulse radiolysis in an absolute way, i.e. by directly observing the kinetics of the formation of transient absorption or of the decay of the parent compound absorption. Direct observation of OH (or of H) by pulse radiolysis cannot help in obtaining reaction rates, because the absorption is in the far ultraviolet and one can observe only the tail of this absorption which at -200-250 nm has a very low extinction coefficient (-500 M cm-') (Pagsberg et al., 1969). A review of the experimental methods and summary of the rate constants of OH reactions has recently been published (Dorfman and Adams, 1973) and another compilation of rate constants is currently being prepared (Farhataziz and Ross, 1975). Aithough only a few hydrocarbons have been studied it appears that most of them react with OH with a rate constant of ca. 109 M-1 s-1 . Methane is about 4 times less reactive than this value, and cyclopentane and cyclohexane about 5 times more reactive. Alcohols, amines, ethers, and many esters also fall in the same range. Carboxylic acids and carbonyl compounds seems to be to a certain degree less reactive. Lower reactivity is also found for the protonated forms of amines and amino acids. Direct reaction of OH with the substituent is usually unimportant except for a few cases such as thiols, where H is easily abstracted from the SH, or nitroso com-

-'

APPLICATION OF RADIATION TECHNIQUES

237

pounds, which are readily oxidized by OH. Reports on partial abstraction from OH and NH, groups do not suggest that these processes are very important. Addition of OH to olefinic and aromatic compounds is also very s - l . Despite the high rates, a rapid; usually k = 10’- 10’ M reasonable correlation with Hammett’s substitutent constants was obtained for the series of substituted benzenes and demonstrated the electrophilic behaviour of OH in its addition to aromatic compounds (Neta and Dorfman, 1968). In strongly alkaline solutions the hydroxyl radical is transformed into its basic form, the oxide monoanion radical [reaction ( l o ) ] . The OH + OH-

3 0-+ HzO

(10)

pK-value for this equilibrium is 11.9 (see Dorfman and Adams, 1973, for details and references). The reactions of 0 - differ from those of OH mainly in their rate constants. Abstraction reactions are generally slower with 0 - than with OH by about a factor of two. Addition reactions, however, are over three orders of magnitude slower with 0- than with OH. As a result compounds which contain sites for both addition and abstraction reactions, undergo mainly addition with OH but mainly abstraction with 0- (Neta e t al., 1972: Simic e t al., 1973). Abstraction by 0 - of benzylic and allylic hydrogens is an efficient process, and even abstraction of vinylic hydrogens has been demonstrated. Another unique reaction of 0 - is the oxidation of the phenoxide ion to phenoxyl radical (Neta and Schuler, 1975). This oxidation has been shown to involve an electron transfer mechanism, while the production of phenoxyl from the reaction of OH with phenol takes place via addition followed by water elimination (see Section 7). Comparison of Reactions of e i s , H, OH, and 0 - and their Application f o r Radical Production The hydrated electron is obviously a nucleophile and its reactions are affected by substituents correspondingly. The hydroxyl radical is expected to behave as an electrophile and this behaviour was, indeed, demonstrated with aromatic compounds. The low reactivity of 0 toward aromatic and olefinic ?r-systems suggests that this species behaves as a nucleophile because of its charge. The behaviour of hydrogen atoms is not easily predictable; the effect of substitution in benzene demonstrated a slight electrophilicity.

238

P. NETA

Hydrogen atoms and hydroxyl radicals react with aliphatic compounds mainly by H-abstraction from the chain, although reactions with certain substituents are also important. With hydrated electrons the functional group is the only site of reaction and its nature determines the reactivity. The reactions of hydrated electrons are by definition electron transfer reactions. The rate of reaction of a certain substrate will depend on its ability to accommodate an additional electron. For example, in an unsaturated compound the rate may depend on the presence of a site with a partial positive charge. Thus acrylonitrile and benzonitrile are three orders of magnitude more reactive toward eiq than are ethylene and benzene. On the other hand, this large difference does not exist in the case of addition of H and OH. Hydrogen atoms can formally reduce functional groups but no evidence is available to support an electron transfer mechanism. In those cases where functional groups are reduced by H the mechanism is believed to be one of addition, followed possibly by proton dissociation, depending on the pH. This mechanism is valid, for example, for reduction of carbonyl or nitro-groups by H. Chloro-, bromo-, and iodo-compounds, both aliphatic and aromatic, undergo efficient dehalogenation by eiq [reaction (1l ) ].

Their reaction with H, on the other hand, is an atom abstraction process (12) rather than electron transfer. In the case of chloroRX

+ H - + K + HX

-+

K + H+ + X-

(12)

aliphatic compounds this process is accompanied by a certain amount of H-abstraction, and in the case of all haloaromatic compounds H addition to the ring is the major reaction (Brett and Gold, 1971, 1973; Peter and Neta, 1972). Hydroxyl radicals do not abstract halogen directly, but oxidative dehalogenation takes place (within <1 ps) when OH adds to the carbon bearing the halogen [reactions (13) and (14)]. Similar elimination of a nitro-group was also demonstrated. For specific examples of halo- and nitrocompounds see Bansal et al. (1972), Eiben et al. (1971), Fendler and Gasowski (1968), Greenstock et al. (1973), Koster and Asmus (1973), Neta (1972b), Neta and Greenstock (1973), and SchulteFrohlinde et al. (1973).

APPLICATION O F RADIATION TECHNIQUES

239

x:

OH

In order to demonstrate the differences in the rates and mechanisms of reactions of organic compounds with the primary radicals of water radiolysis several representative examples are collected in Table 1. The original sources of these data can be found in the rate constant compilations mentioned above). When the production of a specific radical is desired, advantage must be taken of differences in the reactions of the primary species with organic substrates, as has successfully been done in many cases. In the study of radicals produced by reaction of the substrate with one of the primary transients it is usually necessary to scavenge the other transients. If they are not scavenged and they happen to be unreactive toward the substrate they can still interfere by reacting with the radicals or stable products formed from this substrate. If they react with the substrate itself, complications may arise in the observed transient spectra or kinetics. A practically one-radical system can be obtained by the conversion of eiq into OH using N,O [reaction ( 5 ) ] .In other cases it is often possible to produce the same radical by reactions of the various primary species with different substrates. This can be demonstrated in the study of hydroxyalkyl or ketyl radicals. In this case a solution containing both the alcohol and the corresponding carbonyl compound is irradiated and all the primary radicals produce the same species by reactions similar t o (15)-(18). Another example of production of a single species is (CH3)zCHOH + OH + (CH3)2&0H + H2O

(15)

TABLE 1

10

Rate Constants for Reactions of eiq, H, OH, and 0- with Selected Organic Compounds

0

+P

Rate constant (M-' s - l ) and products -

Reactant

eaq

H

CH3CH3

Nonreactive

2.5 x.106 CH3CHz

CH3CHzOH

Nonreactive

2.6 x. l o 7 CH?CHOH (+ CHzCHzOH)

Nonreactive

0x

CH3CO;

Nonreactive

4~ l o 5

CH3C02H

2 x lo8 CH3C0; + H

CH2COzH

CH3CHzCO;

Nonreactive

ClCHzCO2H

7 x lo9 CHzCOzH

BrCH2CO;

6 x lo9 CHzCO; 2 x lo6 9 x lo6 4~ l o 9 CHzCOzH + NH3 and'NH3CHzCO; + H

lo4 CHz C(CH 3) 2 OH x

lo4

1.8 x lo7 3 5 lo5 ClGHCOZH + CH2COzH 3 x lo8 CHzC0; -lo6 9 x lo4 8 x lQ4 'NH3CHCOzH

OH

1 x 1.09 CH3m2 1.8 X. l o 9 CH CHOH (+ & C H ~ O H )

!jx l o 8 CHzC(CH3)zOH 7 x lo7

0-

1.2 x l o 9

3.3 x

lo8

5 x lo7

? x lo7 CH2COzH 8 x lo8 5 5 lo7 ClCHC02H

3.3 x l o 8

5 X. l o 7 BrCHCO; 3x

lo9

2 x lo7 2 x 197 +NH3CHCOzH

5.6 x

lo8

CH3COCH3

Nonreactive

-02CC=CC0~ CH3N02 CH3CN

3 x lo6 (cp3)ZkoH + CH2COCH3 3 x 1.0~ CH3CH2

1 x lo7 CH2COCH3

2 x lo9 HOCH2kH2 4 x lo9 HOCH2kHCONH2 >lo9

2 x 1o'O (CH2CHCONH2)'-

2 x 1.0'0 CH3CHCONH2

-3 x 1 o ' O

-1 x

2 x 1.0'0 CH3NO; 3 x lo7 (+ CHjCH=&')

4 x 1.07


CH3N02 H

(-10"

1.5 x lo6 CIl3CH=fi + CHzCN

?-2 x 10' CH2 CN

1x

lo9

lo9

1 5 lo9 HC6H5 CO,; (3 isomers) 1.2 x 1.8 x 10' 4 . 0 ~l o 6

lo9

<5 x lo7

for CH2=NOi)

2-2 x l o 8

242

P. NETA

shown by reactions (19)-(21). In these cases the relative concenCH3CO; + OH + CHzCO; + HzO

(19)

BrCHzCOi + H + kH2CO; + HBr

(20)

+ eiq + 6HzCO; + Br-

(21)

BrCHzCO;

trations of the two solutes have to be chosen so that the desired reactions will predominate. A variation of reactions (19) and (21) has been used in the study of radicals from, e.g. propionic acid. The reaction of OH (and H) produces two radicals, whereas reactions of p ,

CH3kHC02H

+ HzO

CH3CHzCOzH + OH

(22)

CH~CH~CO~H

with the halo compounds produce only one radical in each case [reactions (23) and (24)]. These reactions have been used to study

ea-q

CH3CHC1CO2H + eiq + CH36HC02H + C1-

(23)

CICH2CH2COZH+ eiq + kHzCHzC02H + C1-

(24)

the distribution of OH abstraction from the various sites. By the same method the site of addition can be studied, e.g. reaction (25) can be compared with reactions of el, with the halo-compounds

CHz=CHCOzH + OH

/

HOCH?~ H C H O ~

(25) kH2CHOHC02H

which produce the corresponding radicals. In certain cases 0- can be utilized to produce radicals which cannot be produced by other methods, because 0- can abstract hydrogen from molecules to which OH tends to add, e.g. reactions (26) and (27). The disadvantage of this method is that it can be CH,CH=CHCO;

+ 0- -+ 6HzCH=CHCO; + OH-

C6H5CH3 + 0-

+ C6H56H2

+ OH-

(26)

(2 7)

utilized only in strongly alkaline solutions where most of the OH radicals have been converted into 0 - and their addition to the solute is thus prevented.

APPLICATION OF RADIATION TECHNIQUES

243

A one-radical system can sometimes be achieved via intermediate radicals. For example, the nitrobenzene radical anion can be produced by reduction with elq [reaction (28)], while H and OH can be c 6 H ~ N 0 2+ eiq

--f

C6H5iO;

(28)

made t o react with isopropyl alcohol [reactions ( 1 5 ) and ( I S ) ] to produce another reducing radical which then reacts with nitrobenzene via electron transfer [reaction ( 2 9 ) ] . c 6 H ~ N 0 2+ (CH3)zCOH + C ~ H S T ~+O(CH3)2CO ~ + H+

(29)

Another tool recently applied to radiation studies is the use of oxidizing species produced from OH (and which have different selectivities). These include radical ions such as Cl;, Br;, or (CNS); which are formed from the halides or pseudohalides by reactions such as (30) and (31). Such species react with organic compounds Br-

+ OH +. Br + OH-

(30)

+ Br- 2 Br;

(31)

Br

more slowly and more selectively than OH. Their main advantage is that they can sometimes oxidize a molecule by an electron transfer where OH radicals tend t o add. A typical example is ascorbic acid; oxidation by Br; produces the dehydroascorbate radical ion only, whereas reaction with OH produces both that and another adduct radical (Schoneshofer, 1972; Fessenden, 1974). Another example is the oxidation of NADH, where it has been reported that Br; converts NADH quantitatively into NAD* radicals while OH attacks unspecifically on both the nicotinamide and the adenine rings (Land and Swallow, 1971). The SO, radical is another oxidizing species currently being studied in irradiated solutions. It is produced by reaction of eiq with S z O z - and its main difference from other oxidizing radicals lies in its capability to cause decarboxylation of various acids.

4. OPTICAL ABSORPTION SPECTRA O F ORGANIC RADICALS A large variety of organic radicals have been produced in irradiated aqueous solutions and monitored by their optical absorption spectra. Pulse radiolysis techniques allow the recording of such spectra and

244

P. NETA

TABLE 2 Optical Absorption Spectra of Radicals in Aqueous Solutions

<210 <210 <220 225 <230 <250

>850 >700 >2000 900 >1300 >850

<220

>700

250

920

242

9000

240

6500

233 -250 269 235 320 350 335 <240 340 245 255 250 320 400 <260 280 310 313 400 290 42 5 400 300 318 307 258 545 350 285 275 410 410

8800 -9000 -40,000 3000 650 800 950 >600 1000 5100 5800 5400 800 1050 >2400 1900 3300 -3000

n

4000 7300

Stevens et al., 1972 Simic et al., 1969a Simic et al., 1969a Simic et al., 1969a Simic et al., 1971 Stockhausen and Henglein, 1969 Soylemez and Schuler, 1974 Schuler and Patterson, 1974 Soylemez and Schuler, 1974 Schuler and Patterson, 1974 Neta and Schuler, 1975 Lilie and Henglein, 1969 Neta and Schuler, 1975 Neta et al., 1969 Neta et al., 1969 Neta et al., 1969 Neta et al., 1969 Neta et al., 1969 Simic et al., 196913 Simic et al., 196913 Simic et al., 1969b Neta et al., 1970 Simic and Hayon, 1971 Hayon e t al., 1970 Simic and Hayon, 1973 Asmus and Taub, 1968 Michael and Hart, 1970 Neta and Dorfman, 1968 Land and Ebert, 1967; see also Neta and Schuler,1975 Adams and Michael, 1967 Christensen, 1972

4700 25,000 17,600 14,000 9250 3700 3700

Mittal and Hayon, 1972; see also Christensen et al., 1973 Adams and Willson, 1973

Asmus et al., Asmus e t al., Asmus e t al., Asmus e t al.,

1966b 1966b 1967 1967

245

APPLICATION OF RADIATION TECHNIQUES

TABLE 2-continued Radical

C6H560;-

~,,(nm)

emax(M-' cm-I )

445

322

8ooo 27,000

I

Reference Simic and Hoffman, 1972

the kinetics of their formation and decay. These spectra, although not sufficiently characteristic to identify the radicals, can play an important role in understanding the chemistry in many systems. For example, the distribution of reaction of OH with propionic acid to abstract hydrogen from either the alpha- or the beta-positions has been determined by use of the difference in the optical spectra of the two resulting radicals recorded independently (Neta et al., 1969). Another example is the reaction of H with nitroaromatic compounds, where addition to either the ring or the nitro-group may occur. The difference between the spectra of the two resulting radicals is sufficiently large (h,,, for C 6 H 5 N 0 2 His 275 and for Hk6H5N02 is 410 nm) to allow the determination of the mode of reaction. In a similar way the protonation of the electron adduct of benzoate has been suggested to take place on the carboxylate group and not on the ring (Simic and Hoffman, 1972). An important tool for such studies is a compilation of previously determined spectral parameters. Fortunately, such a compilation has been published by Habersbergerova et al. (1968, 1972) in the form of a list of wavelengths of absorption maxima and representative extinction coefficients for several hundred radiolytically produced transients. and e m a x for simple aliphatic and Some selected values of A,, aromatic radicals are presented in Table 2. Many of them have appeared in the above mentioned compilations but some have been taken from the more recent literature.

246

P. NETA

The recording of absorption spectra has also been applied t o studies of acid-base equilibria, since radicals with dissociable functional groups usually show spectral shifts with pH which can be utilized for the determination of pK-values. Most of the reported values of dissociation constants have, in fact, been determined in this manner. Several examples of spectral shifts can be found in Table 2 and a detailed discussion of acid-base equilibria of organic radicals is given in Section 6. Mos't substituted alkyl radicals have absorption maxima well below 250 n m so that the peaks have not been observed (see Table 2). The in the a-carboxyalkyl radicals show a distinct absorption with A,, region of 300-350 nm and E < 1000 M-' cm-', in addition to another band with a maximum below 200 nm. Similar situations appear also when the carboxyl group is bound in an ester or amide. Hydroxyl and amino-groups bound at the radical site of a-carboxyalkyl radicals diminish the absorbance in the 300- 350 nm region and increase that at lower wavelengths up to E 5000 M cm-' at 250 nm. A considerable difference has been observed between the spectrum of cyclohexyl and that of the cyclopentyl radical, the former cm- . exhibiting a pronounced shoulder at 250 nm with E = 920 M Cyclohexenyl and cyclopentenyl radicals show a much stronger absorption with definite maxima at -240 nm. These are allyl type radicals and like the allyl radical itself they show extinction coefficients of 7000-9000 M - ' cm-'. The optical spectrum of the allyl radical is greatly affected by unsaturated substituents which conjugate with the allylic 1 and 3 positions. These positions bear all the spin density and their interaction with carboxyl groups, for example, shifts A, to -270 nm with extinction coefficients of 20,000-40,000 M-' cm-' (Neta and Schuler, 1975). A carboxyl group attached to the central carbon of allyl has only a minimal effect on the absorption. Radicals of the cyclohexadienyl type show moderately intense peaks at -300-350 nm ( E 3000-4000 M cm-I). Radicals of this type are produced by addition of H or OH to benzene and its derivatives, and their spectral parameters are affected b y the substituent. In general, it is found that both the OH and H adducts have very similar absorption spectra and extinction coefficients. The electron adducts usually absorb at somewhat lower wavelengths. The effects of substituents on the absorption maxima, as expressed by the bathochromic shifts [ ( V A H - V A x ) / v A H where v is the wave

-

-'

-'

-

-'

'

APPLICATION OF RADIATION TECHNIQUES

247

number in cm-' 3 , have been correlated with the values for the parent molecules (Chutny, 1967). The ratios between the bathochromic shifts for various substituted cyclohexadienyl and hydroxycyclohexadienyl radicals and those for the parent benzene molecules were found to be -0.9 in most cases studied. The correlation for the electron adducts did not give consistent results. Phenoxyl and anilino-radicals, produced from the OH adducts of phenol and aniline as the result of elimination of water, show absorption maxima at 400 and -300 nm, distinctly different from the maxima at -340 nm for the initial adducts (Land and Ebert, 1967, Christensen, 1972). The only correlation of absorption maxima with theoretically calculated energies has been reported on radicals in which the unpaired electron is conjugated with a double bond or with a whole conjugated system (Lilie and Henglein, 1969). Transition energies, calculated by LCAO, gave a good linear correlation with the wave numbers of the absorption maxima for some twenty radicals mostly derived by addition of eas to conjugated carbonyl or nitrocompounds. Similar calculations combined with the available data can potentially help in distinguishing between possible radical structures.

5 . ELECTRON SPIN RESONANCE STUDIES

The importance of e.s.r. in this field is mainly in achieving a detailed characterization of radical structures. Because the resonance line due t o the unpaired electron is split by magnetic nuclei in the radical in a very characteristic way, the technique actually counts the various types of nuclei and allows in many cases an accurate determination of their respective position and geometrical orientation. Patterns of relationships between structure and e.s.r. parameters, based on many experimental results, have been developed and applied to new systems successfully. Details of e.s.r. methods are given in many books and reviews on the subject (see especially Fessenden and Schuler, 1970; Wertz and Bolton, 1972). Although much of the e.s.r. work has utilized radical production methods other than radiation, such as photolysis or rapid-mix redox reactions which can give mostly similar data, we shall limit our discussion here to those topics treated by the combined e.s.r.radiation chemical technique in aqueous solutions. Most of these studies have been carried out by the steady-state method developed by Eiben and Fessenden (1971) and have been primarily concerned

248

P. NETA

with identifying radicals produced by the reaction of organic compounds with the transients from watkr radiolysis, e i q , H, and OH. The identification was based on the hyperfine splitting constants of the naturally abundant H, 4 N , P, and the halogen atoms. With relatively long-lived radicals C hyperfine splittings could also be measured at the natural abundance level of -1% (Laroff and Fessenden, 1971; Laroff et al., 1972; Schuler et al., 1973), and where this information is available very detailed knowledge of the radical structure is achieved. In all of these studies the spectra observed were isotropic because the radicals were relatively simple (molecular weights usually much below 300) and could rapidly tumble in solution. Larger or polymeric radicals have not been studied. Some radicals have been studied by trapping them to form a long-lived radical, such as by addition to fumarate (Neta, 1971) or to the aci-form of nitromethane, CH, =NO; (Behar and Fessenden, 1972a). Such a trapping technique has been especially helpful for the study of radicals which could not be observed directly, e.g. because of line broadening (see Fessenden, 1974). In order to demonstrate the capabilities of the steady-state in situ radiolysis-e.s.r. technique some of the results obtained by this method are discussed below. These results usually supplement radiolysis experiments performed with other detection methods by providing positive radical identification. The reaction of OH with saturated aliphatic compounds is known to take place mainly by hydrogen abstraction from the alkyl groups. For example, amines undergo a reaction such as (32). However,

'

'

(CH&NH + OH + 6H2NHCH3 + H2O

(32)

abstraction from amino groups has also been suggested. The e.s.r. spectra observed when OH reacts with di- and trimethylammonium ions show that the corresponding aminium radicals are formed (Fessenden and Neta, 1972) possibly by reaction (33). In strongly (CH&NH;

+ OH + (CH3)zkH' + H2O

(33)

alkaline solutions abstraction from the amino-group was also indicated (Neta and Fessenden, 1971) but it is not clear whether this abstraction is accomplished by OH or by 0 -. The reaction of OH radicals with benzoate has been found to produce the three isomeric hydroxycyclohexadienyl radicals (Eiben and Fessenden, 1971). Moreover, addition to each para- and metaposition was found to take place with similar rate, whereas addition to each ortho-position proceeds at only about half that rate.

APPLICATION OF RADIATION TECHNIQUES

249

Addition of OH has been found in many cases to be followed by elimination of water, e.g. with phenols and anilines (Neta and Fessenden, 1974) or with pyrroles and imidazoles (Samuni and Neta, 1973a). The reaction of OH with ascorbic acid at pH < 5 has been found to involve addition at both position 2 and 3 to yield different radicals (Laroff et al., 1972), only one of which undergoes immediate elimination of water. Addition of OH has also been shown in certain cases to be followed by ring opening, as in the case of furans (Schuler et al., 1973) or uracils (Neta, 1972d). Moreover, with substituted furans several geometric isomers of the resulting butenedial anion radicals have been recorded. With chloro- and nitroheterocyclic compounds, rapid elimination of HC1 or HN02 has been found to follow addition of OH to the substituent position (Neta 1972b; Neta and Greenstock, 1973; Greenstock et al., 1973a). In contrast with the addition of OH and many other radicals to the aci-form of nitromethane [reaction (34)], it has been found that CH,=NO;

+ OH + H O C H ~ ~ O ;

(34)

radicals such as Br;, I;, and (CNS); react mainly by electron transfer [reaction (35)] rather than addition (Behar and Fessenden, 1972a). CH2'NO;

+ Bri

-+

kH2N02

+ 2Br-

(35)

The resulting radical, kH2 NO2, then adds to another molecule of nitromethane. It was concluded that the reaction path depends not only on the oxidation potentials but also on the energy of the bond formed during addition. Radiolytic e.s.r. studies of the reactions of SO,, Cl;, and Br; radicals with organic compounds are currently being carried out in this laboratory by Fessenden et al. One of the interesting findings is the selective decarboxylation by SO, radicals of certain aliphatic and aromatic carboxylic acids, whereas earlier studies with OH had shown that decarboxylation is not important in such cases. For example, it has been reported that the reaction of OH with malonic acid results mainly in hydrogen abstraction, with only 10% decarboxylation in acid solution and <0.5% in alkaline solution (Behar et al., 1973). In the study of radicals produced by reaction of OH with polycarboxylic or hydroxycarboxylic acids, very intense spectra have been observed in alkaline solutions, where the radicals are multiply charged so that their decay rates are slowed down by electrostatic

25 0

P. NETA

repulsion. I n such cases C hyperfine splittings could be observed at the natural abundance level. From a detailed analysis of the e.s.r. parameters it has been concluded that radicals of the type RC(O -)Cog are planar at the radical site, as opposed to the case of a-hydroxy or a-carboxyalkyl radicals (Laroff and Fessenden, 197 1). C Hyperfine splittings have also been measured for OH adduct radicals from benzene polycarboxylate and found t o be very useful in the determination of spin density distributions around the ring (Eiben and Schuler, 1975) in a way analogous to the use of proton hyperfine constants. E.s.r. studies have been helpful in demonstrating dechlorination and deamination by hydrated electrons (Eiben and Fessenden, 1971; Neta and Fessenden, 1970a). Many other reactions of eiq have also been studied. Nitriles are reduced by the electron to (RCN)- which rapidly protonates on carbon to form RCH=h even in strongly alkaline solutions (Neta and Fessenden, 1970b); a similar radical is also formed upon direct addition of H t o the cyano group. Other electron adducts which protonate on carbon are (CH?=CHCO;) and ( - O z C . C-C. C0;)- t o yield respectively CH,CHCO; and -OzC. CH=k . CO; (Neta and Fessenden, 1972). These protonations take place in alkaline solutions and must utilize H,O for this process. On the other hand, radical-anions such as those obtained by addition of an electron to fumarate or maleate do not protonate readily on carbon but undergo the normal acid-base equilibria with all the protons on the carboxyl oxygens, e.g. to yield HOz CCH=CHk(OH), in acid solution. Similarly, the radicals from aromatic carboxylic acids undergo all the protonations on the carboxyl groups (Neta and Fessenden, 1973). One-electron reduction of pyridines has been found to be followed by immediate protonation on nitrogen even at pH 14 (Neta, 1 9 7 2 ~ ) .With pyrazines the electron adducts can undergo double protonation on both nitrogens (Fessenden and Neta, 1973). It is not clear yet what determines the position of protonation of radical-anions, although resonance stabilization of conjugated systems and local negative charges are undoubtedly important. Another phenomenon observed with the radical-anions is the enhanced strength of intramolecular hydrogen bonds. The electron adducts of maleate and phthalate have been found to exist in a protonated form, with a proton bridging between the two carboxyi groups even at pH 14, while the adducts to fumarate or terephthalate are completely dissociated at p H > 11 (Neta and Fessenden, 1972 and 1973). These and other examples are discussed in Section 6.

APPLICATION OF RADIATION TECHNIQUES

25 1

The 0 - radical abstracts hydrogen from aliphatic compounds relatively rapidly (but about a factor of two more slowly than OH) while it adds to olefinic and aromatic compounds very slowly, if at all. E.s.r. experiments have demonstrated that the reaction of 0 with toluene or crotonate leads predominantly t o the radicals formed by abstraction and not by addition (Neta et al., 1972) as shown by reactions (26) and (27) above. These findings have since been used for the production of radicals by abstraction from compounds which tend normally to add. For example, a series of substituted benzyl radicals have been produced by reaction of 0 - with substituted toluenes (Neta and Schuler, 1973). An attempt t o correlate Hammett’s substituent constants with the e.s.r. parameters of this series of substituted benzyl radicals has been made. The variations in the hyperfine constants were found t o be very small, and although they are far beyond the experimental error no linear correlation with substituent constants could be obtained. Electron-donating and -withdrawing groups as strong as OH and CN cause changes of
252

P. NETA

pH units is obtained. Dissociation constants measured from the inflection point by this technique were found to be in good agreement with those measured by variations in optical absorption or conductance. Compared to the latter methods the e.s.r. measurement of pK-values is more time-consuming, but it is the only method which relates the two forms of the radical directly and leaves no doubt that the changes measured are due to acid-base equilibria. Moreover, in certain cases the e.s.r. technique allows measurements of the dynamics of the process as exhibited by the line width. Line broadening can be caused both by dissociation and by proton exchange. Several possible types of behaviour have been discussed b y Laroff and Fessenden (1973). We shall only mention here that for the hydroxylalkyl radicals, having pK-values -10- 12, the forward rate of the equilibrium (36) was found t o be diffusion controlled, kOH + OH- 2 KO-

-'

+ HzO

(36)

-10' M s-l as can be predicted from Eigen's work (1964). Electron spin resonance spectra recorded under steady-state irradiation have also been utilized for quantitative determination of radical concentrations under carefully controlled conditions of production rates and thus used for kinetic studies. Some discussion of the validity of these experiments has been given by Fessenden (1975). One of the most useful applications of this approach was the study of rate constants for reactions of hydrogen atoms with organic compounds (Neta et al., 1971). Strong e.s.r. signals of the hydrogen atom can be observed in irradiated acid solutions because of nonequilibrium population of the electron-nuclear spin states or what has become t o be known as CIDEP (Chemically Induced Dynamic Electron Polarization; see Fessenden, 1973; Verma and Fessenden, 1973). These signals decrease with time on the microsecond time scale because of relaxation processes. Further decrease can be caused by chemical reaction. By observing the H atom signal height under steady-state conditions in the presence of increasing concentrations of a substrate, the rate constant for reaction of H can be derived, provided that the secondary reactions affect the CIDEP similarly. As mentioned above, many rate constants have been determined by this method (see Neta, 1972a, for summary). Direct recording of the kinetics on an absolute time scale can be achieved by the time-resolved e.s.r. method. However, special attention has to be paid t o CIDEP effects which may distort the observation of chemical kinetics. When the rate of radical decay is

APPLICATION OF RADIATION TECHNIQUES

253

relatively slow, there is little polarization of the e.s.r. signals. Thus it was possible to follow by time-resolved e.s.r. the decay of the malonate radical and also to confirm the mechanism of reaction of OH with ascorbic acid (Fessenden, 1975). Some rates of reaction of radicals from pyrimidine bases have also been reported (Nucifora et al., 1972). The above examples are only selected problems that have been treated by e.s.r. Other examples are available in the literature, mostly dealing with mechanistic studies based on identification of radical structures. Several of these studies will be discussed in Section 7 in conjunction with results obtained by other techniques.

6. ACID-BASE EQUILIBRLA O F ORGANIC RADICALS The presence of an unpaired electron in the vicinity of a functional group which can undergo acid dissociation is expected to exert a considerable effect on the equilibrium constant. The unpaired electron can act as an electron-withdrawing centre or can affect resonance stabilization, causing the pK-value of a radical to be very different from that of a molecule with a similar structure. The most striking example is the case of alcohols, where the pK-values for the parent compounds are -16 while those for the radicals resulting from an alpha-hydrogen abstraction are -10-12. So far over 150 dissociation constants have been determined by several investigators (for a recent summary see Hayon and Simic, 1974). We shall mention here representative examples to facilitate discussion.

Methods o f Determination o f Dissociation Constants Several methods have been applied to determinations of dissociation constants of radicals. Most of them are based on measuring a parameter which at any pH is a weighted average of the parameters of the acid and basic forms present at that pH. A plot of the observed parameter versus pH gives the typical sigmoidal curve with an inflection point where pH = p K . Several parameters can be different for the acid and basic forms to allow pK-measurements. The most commonly used method utilizes differences in the optical absorption spectra of these two forms of radicals. Monitoring by

254

P. NETA

pulse radiolysis the optical density at a wavelength where the difference between the spectra of the two forms is large, an accurate pK-value can be obtained. It must however, be, ascertained that no other causes exist for the change in optical density. The majority of the pK-values reported in the literature have been determined by this method and only a few of them have been later proven incorrect by other techniques. Another method widely used is based on conductometric measurements at short times after an irradiation pulse. This method determines the electrical charge of the species studied and is thus useful for radicals where several dissociations can take place. This technique can be complicated by buffering effects if the parent compound itself undergoes acid-base equilibria in the region of interest for the study of the radical (see e.g. Bhatia and Schuler, 1973b). It is therefore, a prerequisite to know all the other pK-values which may be involved. Another limitation of this technique lies in the fact that only pK-values between about 2 and 1 2 can be studied and high background conductivity decreases the sensitivity of the measurements. Pulse polarography can also be utilized, but it has been used only in a limited number of cases. Another method which is not widely used involves differences in the kinetic behaviour of the acid-base forms of a radical. The radical-radical decay kinetics are in most cases different for the various forms and are affected t o a large extent by the electric charges. The reaction rate of a radical with another molecule can also depend very strongly on the state of dissociation. It is found for example, that the rate of electron transfer from an a-hydroxyalkyl radical to a nitro-compound is greatly enhanced by proton dissociation of the radical (Adams et al., 1968). These kinetic methods, however, have not been applied to many cases because they are more time-consuming (in data analysis) and slightly less accurate than the methods mentioned previously. The electric charge of a radical can also be determined by examining the effect of ionic strength on the kinetics. Such measurements did, in fact, establish that the major reducing species in water radiolysis is negatively charged. Last but not least is the e.s.r. method discussed in Section 5 . The accuracy of the e.s.r. method itself is very high, and the overall accuracy is limited by that of the pH measurements, whereas in the previous techniques, the determination of other parameters is less accurate than the pH determinations. As mentioned above, the e.s.r.

APPLICATION OF RADIATION TECHNIQUES

255

method is the only one which relates the two forms of a radical directly and can confirm beyond doubt that the phenomenon observed is an acid-base equilibrium. This method is also independent of radical yields, side reactions, or impurities, which might complicate the measurements by other techniques. These advantages of the e.s.r. technique are valid when the acid and basic forms undergo rapid exchange by themselves or in the presence of buffers (see e.g. Laroff and Fessenden, 1973). When exchange is slow, and one has to compare signal intensities of independent spectra of the two radicals, the accuracy of this method becomes somewhat limited.

Comparison of pK- Values for Radicals and Parent Molecules

When a hydrogen atom is abstracted from a carboxylic acid to produce a radical, the functional group remains intact and its pK-value in the radical can be compared with the value in the original molecule. A similar situation exists for hydroxyl- or amino-groups when abstraction takes place from a position other than at the substituent site. In fact, in all these cases only small changes in the dissociation constants are observed. However, abstraction from alpha-positions of alcohols or amines or addition of an electron to carbonyl or unsaturated compounds change the oxidation state of the functional group and result in a dramatic change in pK. As an illustration let us take the case of ethanol, where hydrogen abstraction leads mainly to the CH3CHOH radical. For this radical, pK is 11.5, very different from the value for ethanol itself (-16) and from pK for protonation of acetaldehyde (--8). For further discussion selected pK-values are summarized in Tables 3- 5. It can be seen from Table 3 that hydrogen abstraction from acetic and other simple acids does not affect the pK-value of the carboxyl group significantly. In the presence of an a-OH group, however, the pK-value of the carboxyl is higher in the radical than in the parent molecule by about one unit, but it is not higher than the usual values for unsubstituted acids. The effect of an amino-group is even larger. Thus pK for the carboxyl group of glycine increases from 2-2 in the molecule to 6.6 in the radical. It should be pointed out, however, that pK for the amino-group is greatly reduced, from 9.8 to <1 in the radical, so that the dissociation of a proton from the carboxyl group takes place from an electrically neutral radical rather than from the positively charged molecule. This difference in charge may

256

P. NETA

TABLE 3 Dissociation Const?nts of Carboxyl Groups in Organic Radicals (RC02H= RCO; + H’) pK, of parent compound

Radical (acid form)

pK,

Reference

COzH CH2G02H CH3CHCP2H CH~CHZCHCO~H CH3)26COzH LH 2C(CH3)2CO2H kH(QH)C02H CH 3C (OH)C02 H CH(NH2)COzH HOC~HSCO~H

1-4 4.5 4.9 4.8

Buxton and Sellers, 1973 Neta e t al., 1969 Neta e t al., 1969 Neta et al., 1969 Neta et al., 1969 Neta et al., 1969 Simic e t al., 196913 Simic e t al., 1969b Paul and Fischer, 1971 Simic and Hoffman, 1972

3.7 4-7 49 4.8 4.8 5-0 3.8 3.7 2.2 42

Greenstock et al., 1973

2.1

Neta and Patterson, 1974

4.9

Neta and Patterson, 1974

4.7

0 ‘-/’

5-8

4.8 4.6 5.3 6.6 4.4

6.3

H

be the main cause of the increased pK-value. A similar effect of charge can be seen in the case of 5-nitro-2-furoic acid. The pK-value of the parent molecule is 2.06 and that of the carboxyl group in the anion radical is increased to 3.77, probably because of the negative charge. On the other hand, addition of OH to benzoic acid does not change the charge of the molecule, and, despite changes in the electron distribution on the ring, the pK-value of the carboxyl group remains practically unchanged. The pK-value of the carboxyl radical has been the subject of many studies and is not yet known with great certainty. The value of 1.4 given in Table 3 is the latest determination. The previously determined value of 3.9 (Fojtik et al., 1970) was shown to be incorrect by recent

APPLICATION O F RADIATION TECHNIQUES

25 7

results obtained in this laboratory. The pulse conductivity experiments have been reported using the a.c. conductivity set-up and the results show no protonation of CO, down t o pH 2. Observation of the e.s.r. signal has shown that the pK-value is below 1.5. The reportedvalue of 1.4 can be taken as the upper limit and is probably very close to the real value. Table 4 summarizes pK-values for hydroxyl groups, mostly alpha to the radical site. The pK-values for P-hydroxyalkyl radicals are only slightly lower than those of the parent alcohols (usually by -1) and are close to 14 (Kirino, 1975). On the other hand, primary, secondary, and tertiary (at the radical site) a-hydroxyalkyl radicals have pK-values about 10.7, 11-5 and 12.1 respectively. Compared to methanol (pK FZ 15), ethanol (16), and 2-propanol (17) these pK-values show a decrease of 4-5 units caused by the unpaired electron at the alpha position. The odd electron acts, therefore, as an electron-withdrawing group stronger than CF, . Additional electronwithdrawing groups attached to the alpha position further decrease pK. This is demonstrated by the radicals containing carboxyl, acetyl, hydroxyl, trifluoromethyl, or conjugated double bonds attached t o the radical site. These effects seem to be additive and pK-values as low as -0.45 have been observed. The latter value has been observed for the radical from ascorbic acid, and it is interesting t o note that it is 4-5 pK units lower than that of ascorbic acid, a decrease similar t o that found for simple alcohols. On the other hand, pK for (CF3)2kOH is 1.7, which is -8 units below pK for hexafluoroisopropyl alcohol (Laroff and Fessenden, 1973). The a-amino-a-hydroxyalkyl radicals shown in Table 4 are produced by addition of an electron to the corresponding amides. The dissociation constants are also affected by electron-withdrawing groups similarly to the previous cases. No parent compounds are available in this case for comparison. The pK-value for the hydroxycyclohexadienyl radical formed by addition of OH to an aromatic ring is above 14, similar t o P-hydroxyalkyl radicals as might be expected from the electron distribution. Side-chain aromatic radicals of the type Ph-eR-OH cannot be produced efficiently by reaction of OH with the alcohol because OH tends to add to the aromatic ring more efficiently. They are, therefore, produced by addition of an electron to the carbonyl compound. The pK-values for the acetophenone and benzophenone ketyl radicals are -9.5 and they are affected by substituents on the aromatic ring as well as by those directly at the radical site.

P. NETA

25 8

TABLE 4 Dissociation Constants of Hydroxyl Groups in Organic Radicals

k0- + H')

(kOH Radical (acid form)

PK,

~ H ~ ~ H ~ O H CH3CIjCHz OH CH30CHCHzOH OzCCHCHzOH -O%CC(CH3)CH20H NCSHCHzOH vOCHCH20H' CH2OH

14.7 14-6 15-1 14.6 14.9 13.3 0.74 10-7 10-71 11.6 11.51 11.5 11.5 11.6 11.3 12.2 12.03 12.1

CH3ChOH CH3CHzkHOH CH 3CH2 CIjz CHOH (CH3)2CyCHOH (CH3)3qCHOH (CH3)ZCOH CH2 (CH2)4kOH CH2=CH6HqH CH 3CH=CH CH P H CH3 (CH=CH)3CHOH CH3COCH=CHC(CH3)0H HOCYOH 02CCHOH -02CC(CH3)0H -OzCC(CH3)0H

I

0;. CH3.COC (OH)CH3 Q=COH N=CHQH (CF3)2COH OH

9.6 9-9 9.0 5.2 9.5 8.8 9.8 5.2 4.4 1.4 -5 1.70

Reference Kirino, 1975 Kirino, 1975 Kirino, 1975 Kirino, 1975 Kirino, 1975 Kirino, 1975 Bansal et al., 1973 Asmus e t al., 1966a Laroff and Fessenden, 1973 Asmus et al., 1966a Laroff and Fessenden, 1973 Asmus et al., 1966a Asmus e t al., 1966a Asmus e t al., 1966a Simic et al., 1969a Asmus e t al., 1966a Laroff and Fessenden, 1973 Simic et al., 1969a Lilie and Henglein, 1969 Lilie and Henglein, 1969 Lilie and Henglein, 1969 Lilie and Henglein, 1969 Stockhausen and Henglein, 1971 Simic et al., 1969b Simic et al., 1969b Hayon and Simic, 1973c Lilie e t al., 1968 Buxton and Sellers, 1973 Behar and Fessenden, 1972b Laroff and Fessenden, 1973

2.0

Laroff et al., 1972

-0.45

Laroff e t al., 1972

HO HO+OH

?H

CH3k (OH)NH2

>13.5

Simic and Hayon, 1973

259

APPLICATION O F RADIATION TECHNIQUES

TABLE 4-continued Radical (acid form)

PK,

H2NCOCH2CH2k(0H)NH2 H2 NCOCH2.C (OH)NH2 HzNCOI!THC(OH)NHz H2NCOC(Ov) NH2 OCCH2CHzC(OH)NH

11.3 9'8 7.3 -3.7 8.4

OCCH=CHk(OH)NCH2CH3

2.85

-OzCCH=CHk (OH)NHCHz CH3 10.4

> 14

@ - C H O H

Reference Simic and Simic and Simic and Simic and Simic and

Hayon, Hayon, Hayon, Hayon, Hayon,

1973 19 7 3 1973 1973 1973

Simic and Hayon, 1973 Simic and Hayon, 1973 Simic and Hoffman, 1972

9.9

Lilie, 1971

9.2

Lilie, 1971

13-5

Samuni and Neta, 1973a

12

Samuni and Neta, 1973a

8.4

Lilie and Henglein, 1969

9.6 10.0

Adams and Willson, 1973 Hayon et al., 1972

9.2

Adams and Willson, 1973

6.95

Adams and Willson, 1973

9.2 9.25

Adams and Willson, 1973 Hayon et al., 1972

OH

NC*T-CH, OH

P. NETA

260

TABLE 4-continued Radical (acid form)

PKa

-

Reference

OH

@&-OCHaI

6.3

Hayon et al., 1972

5-5

Hayon et al., 1972

5.5

Simic and Hoffman, 1972

7.7

Hayon et al., 1972

1.4

Neta and Patterson, 1974

5.3

Simic and Hoffman, 1972

OH

@d-NHz

I OH

@-OH

I OH

-0.2

Neta and Patterson, 1974

OH

@*I

12.0

Simic and Hoffman, 1972

6.3

Neta and Patterson, 1974

4-2

Nelson and Hayon, 1972 (see text)

OH

APPLICATION OF RADIATION TECHNIQUES

261

TABLE 4-continued Radical (acid form)

PK,

Reference

5-9

Nelson and Hayon, 1972

4-0

Adams and Michael, 1967

5.1

Willson, 197 1

4.1

Willson, 197 I

4.8

Rao and Hayon, 1973

OH

OH

H3c@cH3 H3C OH

CH3

I

OH

I

@OH 0

OH

The radicals formed upon addition of an electron to benzoylpyridines exhibit two dissociations with pK-values of 4.2 and 12.0 for the para-isomer and somewhat similar values for the other isomers. The question arises here as to whether the oxygen or the

262

P. NETA

nitrogen atom protonates first. The pK-value for the hydroxyl group should be lower than that observed for benzophenone (9.2) because the protonated nitrogen withdraws electrons. A cyano-group on the rings lowers the pK-value by almost 3 units and a stronger effect is expected for NH+ in the ring. Therefore, it is reasonable to assign the pK-value of 4.2 to the oxygen site. Moreover, the N+-CH, derivative shows only one pK-value of 5-9, again supporting the assignment. The second pK-value observed with the electron adduct of p-benzoylpyridine can be assigned to the nitrogen. This assignment seems reasonable when the radical is compared with electron adducts of pyridines (Table 5 ) . In the latter case pK is above 14 as compared to 5.2 for pyridine itself. With benzoylpyridine the change is not expected to be as large because a smaller portion of the spin density resides on the nitrogen atom and a pK-value of 12 is therefore reasonable. The acid-base equilibria can be formulated by (37) +

pK = 4.2

9

6

=====s

pK = 12.0

H

instead of the suggestion of the original authors by which the first dissociation is at the NH site (Nelson and Hayon, 1972). Another radical with two pK-values which were initially difficult to assign is that produced from glycine, +H,NkHCO,H. It is now clear that in the intermediate pH region of 1- 5 the radical exists as H, NkHC0, H and not in the zwitterionic form. Tables 4 and 5 include several electron adducts of aromatic and olefinic carboxylic acids. The dissociation constants of these radicals are generally much higher than those of the parent acids because of the additional charge. It appears that one should not compare the value of 12.0 for the electron adduct (C6HsC0,H)'- with pK = 4-2 for C6H5COzH. Instead, a comparison of the first pK = 5.3 for the conjugate acid of the electron adduct (C, H5CO, H, )' seems more suitable. Similarly pK for the conjugate acid of the electron adduct of the C6HsC(OH)OCH3 ( 5 . 5 ) is comparable to that of the acid or

APPLICATION OF RADIATION TECHNIQUES

263

TABLE 5 Dissociation Constants of Miscellaneous Radicals Radical (acid form)

@-$OH

CH~T~O~H

PK,

Reference

<1

Neta and Fessenden, 197 1

-7

Fessenden and Neta, 1972

11.7

Asmus et al., 1966c

4-4

Asmus et al., 1966d

3.2

Asmus et al., 1966b

3.6

Grunbein et al., 1969

2-55

Grunbein et al., 1969

Ho2coGo2 1.22

““E:@ H

Greenstock et al., 1973

>14

Neta, 1972c Neta and Patterson, 1974

>14

Fessenden and Neta, 1973

H

9.2

Fessenden and Neta, 1973

2 64

P. NETA

TABLE 5 -continued Radical (acid form)

PK,

12.0

Reference

Nelson and Hayon, 1972 (see text)

5.5 (SH)

Chan and Bielski, 1973

4.5 (NH)

Lilie, 1971

.10.1

Lilie and Fessenden, 197 3

Neta and Fessenden, 1973

Eiben and Fessenden, 1971 and unpublished data

CHCOzH [-02C!H

.-

]

10.7

Lilie and Fessenden, 1973

APPLICATION OF RADIATION TECHNIQUES

265

TABLE 5-continued Radical (acid form)

Reference

11.5

Hayon and Simic, 1973b

2.8

Hayon and Simic, 1973b

4.8

Hayon and Simic, 1973b

to the first pK of the conjugate acid of its electron adduct. Fumaric acid and its esters show a similar behaviour. An interesting observation made with these electron adducts concerns the formation of very strong intramolecular hydrogen bonds. Such bonds can exist, for example, in maleic and phthalic acids and they cause the last (i.e. second) pK-values for these acids to be higher by 1- 2 units than those for fumaric and terephthalic acids. In the electron adducts the effect is much larger. Whereas the last pK-values for the conjugate acids of fumarate and terephthalate electron adducts are 10.7 and 10.1 respectively, the corresponding values for maleate and phthalate are both >14. These findings are

266

P. NETA

explained by the structures 2-

0

.2-

in which the stability of the hydrogen bridge is enhanced by the delocalized spin density, a portion of which residing on this bond. A similar situation exists with the electron adduct of o-nitrophenol as opposed to the meta- and para-isomers. Some effect is also observed with the diesters of fumaric and maleic acids. With respect to the electron adducts it should be mentioned that, although many are seen to protonate in equilibrium processes, some undergo irreversible protonation on carbon. This process was observed for the electron adducts of acrylate and acetylenedicarboxylate (Neta and Fessenden, 1972),benzenesulphonate (Simic and Hoffman, 1972), and several other compounds. This type of process if further discussed in Section 7.

Correlation of Structure and Dissociation Constants The dissociation constants presented in Tables 4 and 5 above were seen to depend strongly on the structure of the radical. They are affected by the charge and by resonance stabilization of the various acid-base forms. They are also expected to be affected by substituents and, therefore, to behave according to Hammett’s or Taft’s relationships. Several correlations of these and other types have been reported in the literature. The first such correlation (Asmus et al., 1966a) for the pK-values of several simple a-hydroxyalkyl radicals using Taft’s u* constants for the various alkyl groups has now been extended to a range of pK-values about five times as wide (Fig. 1). Radicals of the type R1 R, COH are included and u* values for the R1 R, CH moieties are used. As was found originally, the a-radicals from simple alcohols show a good fit with the straight line. Ketyl radicals from acetophenone and crotonaldehyde also fit quite well. With more strongly

APPLICATION OF RADIATION TECHNIQUES

267

electron-withdrawing groups the data become more scattered but the general trend holds reasonably well. Biacetyl anion has a pi(-value lower than predicted by the plot in Fig. 1. The difference becomes

V*

Figure ?.Correlation of the pKa-values.for k O H with Taft’s Q* substituent constacts. 1, (CH3)zCOH; 2, CH2CH2CH2CH2CH270H; 3, CH3CH2CH2CHOH; 4, (CH3)zCHCHOH;

5, C H ~ C H Z ~ H O 6, H ;CH3CHOp; 7, CH,OH; 8, C HstOHCH3; 9, CH3Cy=CH6HOH; 10, C6HsCHOH; 1 1 , (C6H5) COH, 12, HOJHOH: 13, CH3COC(OH)CH3; 14, C,jHsC(OH)OCH3; 15, C ~ H S $ ( O H ) ~ 16, ; (CF3)zCOH. The o* values were taken from compilations by Hine (1962) and Wiberg (1964). For radical 14 the sum of the ol. values of the two groups was used. For radical 16 the U* value for CF3CH2 was increased by a factor derived from comparisons of XCH2 versus X2CH groups.

even greater when the trioxo radical anion (see TabIe 4, p. 258 for formula) from ascorbic acid is considered (pK = -0.45 and assuming u* as the sum of two acetonyl groups, i.e. 1.2). In these cases the acidity of the radical seems to be enhanced by the resonance stabilization of the anion, where the spin and the negative charge are delocalized over two or three carbonyl groups. h l i e and Henglein (1969) obtained a good linear correlation of the pK-values for C,H, $!OH, C6H, fi0, H, C, H, CHOH, CH3COk(OH)CH,, CH3COCH=CHC(OH)CH3, CH2=CHCHOH, and several other conjugated radicals with values of AH,,, (the difference between the sum of occupied energy levels in the ground state of the acid form and the sum of those of the basic form) calculated using the LCAO method. A linear correlation was shown to hold also between the pK-values for semiquinone radicals and the redox potentials of the corresponding quinones (Rao and Hayon, 1973).

268

P. NETA

Dissociation constants for a series of substituted nitrobenzene anion radicals have been successfully correlated with the electronic energy changes calculated by the Extended Huckel method (carsky et al., 1972). The linear plot was found to have a slope similar to that observed for the parent compounds. A correlation of the pK-values of nitrobenzene anions with the substituent constants CJ has also been reported (Grunbein et al., 1969). Figure 2 shows this type of correlation for the nitro-anions

-0 4

0.4

0.0

0.8

U

Figure 2. Correlation of the pKa-values f:r XC6H4fi02H and for XC&C(OH)CH3 with the substituent constants u. 1-18, XC6H4N02H. 1,p-OH; 2,p-OCzHs; 3,p-OCH3; 4,p-CH3; 5, m-CH3; 6, H; 7, m-OH; 8, m-OCH3; 9, p-C1; 10, p-CHO; 1 1 , p-C02R; 12, m-CHO; 13, m-CO2y; 14, n-CO2R; 15, Q-COCH3; 16, m-CN; 17, p-CN; 18, m-NO2; 19-23, XC6&COHCH3. 19, H; 20,p-F; 21, P-CI; 22, p-Br; 23, PCN.

studied by Asmus et al. (1966b), Grunbein and Henglein (1969), Grunbein e t al. (1969) and by Adams and Wilson (1973). Also included in the figure is a plot for the ketyl radicals studied by Adams and Willson (1973). The plot for the nitro-anions shows that except for the case of the two nitrobenzaldehyde anions the agreement is good. The line for the ketyl radicals, although based on four points only, has a much higher slope than that for the

APPLICATION O F RADIATION TECHNIQUES

269

nitro-anions. It should be also noted that the pK-values in the latter case are much lower. These facts are related to the spin distribution between the ring and the dissociating group. The e.s.r. parameters show that the spin density on the ring of the nitrobenzene anions is only about half that found in the case of the ketyls (see e.g. Eiben and Fessenden, 1971, and Neta et al., 1972). Therefore, the effect of substituents on the ring is expected t o be stronger for the ketyls. The electron adduct of p-nitroacetophenone (Adams and Willson, 1973) may be considered either a ketyl or a nitroanion radical. It is seen in Fig. 2 that the pK-value of 2.6 observed for this radical agrees very well with the plot for the nitro-anions using the u value for p-COCH,. It does not agree at all with the plot for the ketyls using the u value for p-N02. It is, therefore, clear that the radical can be considered a nitro-anion with only a small portion of the spin density on the carbonyl group. An intermediate case appears t o exist with p-cyanoacetophenone. The pK-value of the anion is lower than expected from the line for the ketyls using u for p-CN. This radical is probably an intermediate between a ketyl and a cyano-anion.

Kinetics of Acid Dissociation The kinetics of acid dissociation of radicals has received attention only t o a limited extent. This results from the fact that most of the kinetics of radical reactions have been studied by the pulse radiolysis technique in the microsecond range, while the process of acid dissociation is usually too rapid to be followed and equilibrium is attained before observations begin. Some kinetic measurements have been carried out on the protonation of uracils but these studies do not involve radicals (Greenstock et al., 197313) and will be discussed elsewhere. The e.s.r. approach of deriving rates of dissociation from line widths has been described by Laroff and Fessenden (1973) and used with the CH3kHOH and (CH3)2kOH radicals. The forward rates for equilibria of the type (38) have been found to be 7 x l o 9 f

kOH+OH-

-'

? kO-+HzO

(38)

and 9 x l o 9 M s-l for the two radicals respectively. The reverse rates can be calculated from the forward rates and the dissociation constants and are 4.1 x l o 5 and 1.8 x l o 6 M - l s - l for CH,kHOand (CH, ) 2 60- respectively. These findings agree with Eigen's

270

P. NETA

(1964) prediction that the forward rates are diffusion-controlled if pKk 0 H is lower than PKH 0 . Most radicals undergoing equilibria such as (38) will deprotonate with diffusion-controlled rates (38f). Similarly if k O - radicals with a high pK-value are produced in the acid region they will protonate with diffusion-controlled rates (39r), as will happen, for example, with the electron adduct of acetone. f

kOH

a r

kO-+H+

(39)

The reverse situation holds for radicals with low pK-values when (39f) becomes diffusion-controlled. Buffers present in solution also enter these proton transfer reactions and affect the rate of protonation and deprotonation according to their pK-values. Buffers are therefore used to increase the rate of exchange between acid-base forms at the pH range where these forms may exchange slowly, and as a result e.s.r. lines may be narrowed by the effect of buffers (see Section 6). Measurements on hydroxycyclohexadienyl radicals have shown that whereas the unsubstituted radical undergoes reaction (38f) with s - ’ , the multiply carboxylated a rate constant close t o 10’ M radicals react with OH- more slowly, with rate constants of the order of l o 9 M s - l (Schuler and Fessenden, 1975). Unfortunately, few measurements have been made on other organic radicals. The rate constant for dissociation of the OH radical in alkaline solution, reaction ( l o ) , has been determined by Buxton (1970) and found to be 1.2 x 10’ M s-’. Both this rate constant and the pK-value for OH are similar to those for the (CH,),kOH radical. These radicals are considered to have “normal” behaviour. Radicals may have abnormally low rates of reaction with OH- if their protons are protected either by an intramolecular hydrogen bridge or by neighbouring negatively charged groups. Examples of these two cases are found in the electron adducts of phthalate, maleate, and o-nitrophenol (see Table 5 ) and in the radical HOC(CO,), from tartronic acid. In all of these cases the rates of reaction with OH- are < l o 6 M s - l (Laroff and Fessenden, 1973).

-’

7. KINETICS AND MECHANISMS O F RADICAL REACTIONS Organic free radicals formed in irradiated aqueous solutions react rapidly with each other or with solutes. The relative importance of these two processes depends strongly on the radiation dose rate, i.e.

APPLICATION OF RADIATION TECHNIQUES

271

the rate of production of radicals. Under the commonly applied experimental conditions in pulse radiolysis the rate of production of radicals is very high (of the order o f l o 2 to l o - * M s - l ) . Because the duration of the pulse is usually shorter than the mean radical lifetime at steady-state, one can consider the total radical concentration produced during the pulse as the initial concentration which determines the subsequent reactions. This radical concentration is usually M . A typical radical-radical reaction with 2k = lo9 M s-l results in a practically complete decay of the radicals within a millisecond. With solute concentrations of the order of millimolar, the reaction of radical with solute becomes important only if the rate constant is higher than l o 7 M s - l . As a result only the relatively rapid reactions of radicals with solutes have been detected or measured by pulse radiolysis experiments. On the other hand, the rate of production of radicals in steady-state radiolysis experiments is typically of the order of l o T 7 to M s - l . The radical lifetime becomes much longer (10- 100 ms) so that much slower reactions of radicals with solutes can be detected. In the following discussion we shall deal separately with the kinetics and mechanisms of these two categories of reactions.

-

-'

Radical- Radical Reactions

The rate constants for radical-radical reactions have been determined in most cases by pulse radiolysis by following the decay of the transient optical absorption. This is possible with most organic radicals because their absorption spectrum is shifted toward higher wavelengths compared to the parent molecules, and even radicals which do not have any absorption maximum above 200 nm can usually be followed by observing the tail of the spectrum in this region. In order to observe a pure second-order decay of the radicals, normally on a time scale of lo-' s, the formation reaction has to be completed within a few microseconds after the pulse; higher solute concentrations are therefore generally used for the determination of decay rates than for measuring formation rates. In order to confirm the second-order kinetics it is common to record the decay at various pulse doses. Several representative values obtained by this technique are summarized in Table 6. Most of the rate constants in Table 6 have been determined using solutions saturated with N, 0, i.e. under conditions where almost all

272

P. NETA

TABLE 6 Rate Constants for Radical-Radical Reactions Radical

2k ( M - l s - l )

Reference

1.2 x l o 9 9.6 x l o 8 2.0 x lo9

Stevens et al., 1972 Stevens e t al., 1972 Soylemez and Schuler, 1974

1.5 x lo9

Soylemez and Schuler, 1974

2.4 x l o 9 9 x 10' 1.4 x l o 9 4 x lo8 1.4 x lo9 1.8 x lo9 1.5 x lo9 1.8 x lo9 1.0 x lo9 1.1 x lo9 8.5 x 10' 1.5 x l o 7 -1 x l o 9 5.7 x 1 0 7 1.5 x 10' 5-8 x lo6 2.3 x lo5

Simic et al., 1969a Simic et al., 1969a Simic et al., 1969a Simic et al., 1969a Simic et al., 1969a Simic et al., 1971 Neta e t al., 1969 Neta et al., 1969 Neta et al., 1969 Simic et al., 1969b Simic et al., 1969b Simic et al., 1969b Simic et al., 1969b Simic et al., 1969b Simic e t al., 1969b Simic et al., 1969b Simic et al., 196913 Hayon et al., 1970 Neta et al., 1970 Stockhausen and Henglein, 1971 Stockhausen and Henglein, 1971 Cercek, 1968 Wander e t al., 1968 Wander et al., 1968 Asmus e t al., 1967 Christensen, 1972 Wigger et al., 1967 Asmus et al., 1966b Asmus et al., 196613 Simic and Hoffman, 1972 Christensen e t al., 1973 Adams and Michael, 1967 Adams and Michael, 1967 Neta and Patterson, 1974 Neta and Patterson, 1954 Neta and Patterson, 1974

2.2 x 1.2 7.2 4.5 9.1 1.2 5 1.2 7 2.5 6

lo9

x lo9 x 10'

x 10' x lo8 x lo9 x 10'

x

lo9

x 10' x lo6 x 10'

-lo4

1.4 x 10' 3.1 x l o 9 1.1 x lo9 1-7 x 10'

<2 x

lo5

1 x 10' I x lo9

the radicals are produced by reaction of OH with the solute. The decay observed is therefore a measurement of the rate for the reaction of two identical radicals. Another method widely used to achieve a one-radical system is the reduction of a solute by both eaq

APPLICATION OF RADIATION TECHNIQUES

273

and a reducing radical such as CO; or (CH,),kO- produced from the H and OH (see Section 3). The achievement of a system containing only one radical is essential for the measurement of decay rate constants. Even though one may be able to follow the decay at a wavelength where only one radical absorbs, the reaction of this radical with another radical, present in the solution but not observed, affects the kinetics and makes the decay appear faster than in the pure system. Simple uncharged organic radicals decay by second-order processes with rate constants mostly in the region approaching the diffusions - l . Even radicals which controlled limit, viz., 2K = (1-3) x l o 9 M are usually considered more stable, such as the benzyl radical or the allylic cyclopentenyl radical, decay about as rapidly as the alkyl radicals. It can be seen from Table 6 that charged radicals decay more slowly than their neutral analogs. In general, the acid dissociation of a neutral radical to a singly charged one results in a decrease in the decay rate constant by a factor of -2 as has been found, for example, for the OH adduct of benzoate. This decrease is in good agreement (Wander et al., 1968) with the factor introduced by the coulombic interaction in the Deb ye equation. The equation predicts a decrease by a factor of -10 caused by the introduction of a second charge and this factor is in fact observed in several cases (Simic et al., 1969b). In addition t o electrostatic repulsion there are other effects. A comparison of the rate constants of the doubly charged species -O,C.CH,kHCO, (8-5 x l o 7 ) , -O,C.eHCO, (5.7 x l o 7 ) , and (1.5 x l o 7 ) or of the triply charged radicals -0kHCO; -02C.CH(OH)CO-60, (2.3 x l o 5 ) and -O,C~HC(OH)(CO;)CH&O; (3.5 x l o 6 ) (Simic et al., 1969b) reveals that resonance stabilization and the distance of the charge from the radical site also have considerable effect on the rate constant. These effects are also evident in other cases in Table 6. In particular, the case of the 4-carboxy-N-methylpyridinyl radical should be mentioned (last item in Table 6). Examination of several substituted pyridinyl radicals has shown a large difference in decay rates between 4-substituted radicals and their 3-substituted counterparts, the latter being more reactive (Neta and Patterson, 1974). Moreover, the state of protonation has a strong effect on the decay rate, far beyond any expected effect of charge. The greater reactivity of the acid form as compared to the basic form has been explained by the difference in resonance stabilization.

-'

274

P. NETA

An attempt has been made to correlate the second-order decay rate constants for a series of substituted hydroxycyclohexadienyl radicals with the substituent U-values (Cercek, 1968). A straight line corresponding to p = -0-75 has been drawn. However, if the results for the OH and NH, substituted radicals are omitted because of complicated decay (see below) the slope can be changed dramatically to give p = -0.2. Most of the rate constants are in the range of (4.6 ? 1.0) x 10' M - l s-l and the Hammett plot is inaccurate. A p-value of 0 to -0.2 appears to be better than the suggested value of -0.75, which does not fit the data for radicals with the substituents H, CH,, and C2H 5 , because it is more likely that these substituents should give a good fit rather than OH and NH, . It should be pointed out that in this and other studies the OH or H adducts of aromatic compounds are actually mixtures of concurrently formed isomeric radicals. E.s.r. observations can be helpful in determining the relative concentrations of the isomers, but in the steady-state approach one has to assume similar decay rates in order to relate these concentrations to the relative production rates. When two radicals R, and R, are present in the solution the contributions of the reactions R, + R1 , R, + R,, and R, + R, can be expected t o b e r e l a t e d a c c o r d i n g t o k R l + R z ~ ~ 2 k RXl +2 k~K1z + R z if the radicals combine with each other and the rate constants for all three processes are somewhat similar. However, when the radicals are of different types and tend to disproportionate, the rate constant for R, + R, may be much larger than those for R, + R, and R, + R2 so that the former process may be predominant. This complicates the determination of decay rates of electron adducts which are produced in solutions containing t-butyl alcohol as an OH scavenger. When the adducts are relatively stable by themselves, the presence of the radical from t-butyl alcohol may cause an increase in the decay rate by the R1 + R, reaction. Although this reaction changes the secondorder nature of the radical decay, it may be experimentally difficult to distinguish between pure second-order and mixed-order decays. Therefore, decay rate constants determined in such systems cannot be considered a true measurement even when the observation is made at a wavelength where only one of the radicals absorbs. An extreme example of this effect has been observed with 4-carbamoylpyridinyl radicals in neutral solution (Neta and Patterson, 1974). When the radicals are produced by reduction with eis and CO; or (CH,),kOH the true decay is measured and it is very slow ( < l o 5 M - l s -l ). However, when t-butyl alcohol is used as the OH-scavenger, the

275

APPLICATION OF RADIATION TECHNIQUES

-'

decay of the radical appears to have a rate constant > l o s M s-l because of the cross reaction. Another example was observed in the study of a-bromotetronic acid, which reacts with OH to yield two different radicals. The reaction of these radicals with each other was found to be much faster than the reactions of each radical with an identical one (Schuler et al., 1974). This type of effect may be observed also when R, and R, are the acid and base forms of the same species. For example, Wigger et al. (1967) monitored the decay of C6H5fiOH as a function of pH and found the highest rate at the pK-value of this radical. They concluded that the reaction of R, + Rz is about an order of magnitude faster than either that of R, + R, or R, + R,, where R , and R, are the acid and basic forms of C, H, ROH. The mechanism of radical-radical reactions involves either combination or disproportionation. In the case of a-hydroxyalkyl radicals the following reactions have been suggested from product analysis. In methanol solutions the radicals predominantly combine to form ethylene glycol (40); in the ethanol system both processes take place Z ~ H ~ O +H HOCH~CH~OH

(40)

concurrently (4 1), but with isopropyl alcohol disproportionation

PCH3kHOH

/

CH3CHOHCHOHCH3 (41)

\

m -.

CH3CHzOH + CH3CHO

(42) is the important process (see e.g. Cohen and Lam, 1971). From Z(CH3)zkOH -+ (CH3)zCHOH + (CH3)zCO

(42)

other cases studied it appears that, in general, primary radicals tend to dimerize and tertiary radicals tend rather to disproportionate. Other examples of radicals which dimerize are CH,, CH,kH,, and CH, CO, H. Cyciohexadienyl type radicals undergo both dimerization and disproportionation (Asmus et al., 1967) and so is the case with hydroxycyclopentyl (Soylemez and Schuler, 1974). Disproportionation is predominant with radicals which can be readily reduced or oxidized, such as C6 H, fiO;, c6H5kOH, c6H, 6, or heterocyclic radicals. It appears to be important also with bulky radicals where the expected dimers are sterically hindered.

276

P. NETA

Disproportionation is usually fomulated as a hydrogen transfer, e.g. (43). However, the same result can be obtained by electron 2CH36HOH

-+

CH3CH20H + CH2CHOH(-+CH3CHO)

(43)

transfer (44) followed by protonation of the anion (45), while the

CH3CHOH + H+ (or H2O)

-+

CH3CH20H

(45)

carbonium ion either deprotonates (46) or is neutralized by the CH36HOH -+ CH3CHO + H+

solvent (47). It is difficult t o determine which of the disproportionCH3&HOH + OH- (or H 2 0 )

-+

CH3CH(OH)2

-+

CH3CHO

(47)

ation mechanisms actually takes place. Possibly, different radicals can disproportionate by different mechanisms. It is reasonable to assume that radicals which are readily reduced or oxidized may disproportionate by electron transfer. Moreover, many cases are known where radicals have been oxidized by electron transfer from an inorganic oxidant such as CU*+or Fe(CN)z-, which indicates that similar oxidation by organic species is possible. On the other hand, when hydrogen abstraction from the radical and by the radical appears likely, disproportionat ion may then involve hydrogen transfer. Bansal and Henglein (1974) have suggested the use of the polarogram of the radical as a criterion for the possibility of electron transfer. When the radical shows a steep polarographic wave with an anodic part immediately followed by a cathodic part, this radical can be both reduced and oxidized at the same potential, and consequently can be assumed to disproportionate by electron transfer. Such polarograms have been observed for several hydroxycyclohexadienyl radicals. Although no definite evidence is available, some suggestions have been made to support the case of electron transfer with radicals from some pyrimidines (Haysom et al., 1972; Zemel and Neta, 1973). This was based on the observation of a pH effect on the product distribution which was explained as an effect on the parallel reactions (48) and (49). The production of hydroquinone and

APPLICATION OF RADIATION TECHNIQUES

PyrH+

/

T

277

PyrH (OH)

(48)

PYr

(49 1

catechol from phenoxyl radicals also supports the electron transfer mechanism (50) in this case (Neta and Fessenden, 1974). All the 2Ph6

-

PhO-

I

H+ PhOH

+

I

PhO+

(50)

OH- (Hz0)

Ph(0H)z

above observations have been made with aqueous solutions where the polarity of the solvent is expected t o facilitate a mechanism involving charged species. It is not clear, however, whether the positive and negative ions remain loosely bound in a complex until they are neutralized by water or corresponding ions. Kosower et al. (1973) have suggested that for pyridinyl radicals such a complex exists in equilibrium with the radicals. These authors have observed a pH effect on the decay of the 1-methyl-4-carbamidopyridinylradical which they explained as an effect on the separation of the positive and negative ions in the complex. Neta and Patterson (1974) have found, however, that the effect of pH on the decay rate constant was totally explicable by protonation of the radical. Because the protonated form decays more rapidly than the neutral form by more than four orders of magnitude, the effect of protonation on the decay was evident even when less than 0.1% of the radicals were protonated. A shift in the optical absorption which allows the determination of pK can be observed only when a much larger portion of the radicals are protonated. It appears, therefore, that even if the radicals disproportionate via a complex of the positive and negative ion, this complex is not necessarily in equilibrium with the parent radicals.

Radical Reactions with Solutes The type of reaction of radicals with solutes which has been investigated most is electron transfer from the radical. Another well studied type is the reaction with oxygen. Measurements exist also on

278

P. NETA

TABLE 7 Rate Constants for Radical Reactions with Substrates Reaction

k(M-'s-l)

Reference

4.7 x lo9 7.7 x l o 8 5-0 x 10' 2.9 x lo6 3-5 x l o 7

Thomas, 1967 S tockhausen and Henglein, 197 1 Dorfman et al., 1962 Cercek, 1968 Stevens et al., 1972

4.0 x lo4 6-0 x l o 9

Burchill and Thompson, 1971 Thomas, 1967

3.8 x 2x 6x 3.4 x 4.9 x 1.2 x 3.4 x 8x 9.2 x 4.0 x 5.8 x

lo4

lo8

Burchill and Wollner, 1972 Thomas, 1967 Thomas, 1967 Thomas, 1967 Thomas, 1967 Thomas, 1967 Bullock and Cooper, 1970 Bullock and Cooper, 1970 Bullock and Cooper, 1970 Bullock and Cooper, 1970 Bullock and Cooper, 1970

2.5 x

lo5

Chamberset al.. 1970

lo2 10'

lo3 lo3

lo6 lo5

lo3 lo4 lo7

3.3 x 10'

Burchill and Thompson, 19 7 1

4.3 x l o 2

Burchill and Wollner, 1972

2.9 x l o 7 2.0 x l o 8 7 x lo7

Nucifora e t al., 1972 Nucifora et al., 1972 Cercek and Kongshaug, 1970

21 x

lo5

3 x lo7 4.0 x lo9 1 x lo7 1.0 x l o 9 6-6 x lo9

Packer et al., 1971 Bhatia and Schuler, 1973a Asmus et al., 1966c Adams et al., 1968 Adams and Willson, 1973 Willson, 137 l b


Asmus et al., 1966b

2,7 x lo9 6.1 x l o 9 1.6 x lo9 3.0 x l o 9 7.8 x lo8

Asmus et al., 196613 Willson, 1971b Asmus et al., 1966b Asmus et al., 1966b Adams et al., 1968

APPLICATION OF RADIATION TECHNIQUES

279

TABLE 7-continued Reaction

Reference

k(M-'s-l)

(CH3)2CO- + C ~ H ~ C O C ~ H S (CH3)zCOH + p-CH3COC6H4N02 (C6H5COCH3)'- + C6H5COC6H5 (C6H5COCH3)'+ p-CH3COC6H4NO2 ( C 6 ~ ~ C O C H 3 ) ' -C6HsN02 CH3NO; + C(NO2)4 CH3N02 + C(N0.2); + NO2 CH3N02H + C(NO2)4 (thymine).. +p-CH3COC6H4N02 (thymine)- + C6H.jCOC6H5 (thymine)- + orotic acid (thymine)- + p-benzoquinone

1.2 x lo9 3.8 x lo9 7.8 X lo8

Adams et al., 1968 Adams and Willson, 1973 Adams and Willson, 1973

5.2 x 2.1 x

lo9 lo8

Adams and Willson, 1973 Adams et al., 1968

1.2 x 1.1 x 5.0 X 3'8 x 1.5 x 6.0 x

lo9 lo8 lo9 lo9 lo9 lo9

Asmus et al., 1966d Asmus et al., 1966d Adams et nl., 1972 Adams et al., 1972 Adams et al., 1972 Willson, 1971b

-+

several addition and abstraction reactions. The rate constants for selected examples are given in Table 7. Many radicals have been shown t o react rapidly with oxygen. The reaction is usually formulated as an addition to produce a peroxy radical [reactions (51) and (52)] and this mechanism is probably kH3 + 0

2

-+

CH36.2

(51)

correct for most cases. However, with several radicals electron transfer has been mentioned as another possible mechanism [reaction (53)]. These processes have been reviewed by Czapski (1971). 6 2

/ (CH3)2kO-

1 -

(CH3)2CO

(52)

(CH3)zCO + 0,

(53)

0 2

Organic radicals react with hydrogen peroxide more slowly than with oxygen. The reaction is known to produce a hydroxyl radical and thus results in a chain process as has been observed in several cases, e.g. in the oxidation of alcohols [reactions (54) and ( 5 5 ) ] k H 2 0 H + H202 OH + CH30H

+ CH2(OH)2

--f

+ OH

kH2OH + HzO

(54) (55)

(Burchill and Ginns, 1970a, b). Other types of chain reaction develop, for example, with isopropyl alcohol and N 2 0 in alkaline

280

P. NETA

solutions [reactions (56) and (57)] (e.g. Burchill and Wollner, 1972) (CH3)zkO- + NzO

+ (CH3)zCO

0- + (CH3)zCHOH

-+

+ Nz + 0-

( 56 )

(CH3)zkO- + H2O

( 5 7)

or with isopropyl alcohol and T13+ [reactions (58) and (59)] (CH3)2kOH + T13+

--f

(CH3)2C0 + T12++ H+

(58)

T12++ (CH3)zCHOH -+ T1+ + H+ + (CH3)zkOH

(59)

(Burchill and Hickling, 1970). Organic radicals have been shown to be oxidized by a variety of other inorganic ions. The most commonly used oxidants were C u 2 + and Fe(CN)%- (see e.g. Garrison, 1968; Haysom et al., 1972; Bhatia and Schuler, 1974). However, evidence has been presented also for the reduction of such radicals by Fe2+ and Ti3 + (Behar et al., 1973) and the process has been suggested to involve intermediate formation of a carbon-metal bond [reactions (60) and (61)] . Intermediates containing carbon-metal bonds have kH2OH + Ti3++ Ti3+- CH2OH

(60)

Ti3+- CH20H + H+ + Ti4++ CHJOH

(6 1)

also been proposed in the reaction of several organic radicals with radiolytically produced Ni+ (Kelm et al., 1974). In this case the rate constant for the k + Ni+ reaction is of the order of l o 9 M - l s - l and 1 s or longer). the resulting species react slowly with water ( t l/ 2 The reaction k + Cr2 has been thoroughly studied by Cohen and Meyerstein (1972, 1974) who observed characteristic spectra for C-Cr bonds and measured the rates of formation (-lo8 M s-l) and decomposition ( t o l o 2 s - l for various groups R) of a series of R-Cr intermediates in water. Hydrogen abstraction b y organic radicals is very slow compared to abstraction by H or OH and, therefore, only a few reaction rates have been measured (Table 7). Abstraction by 6 F 3 is -30 times more rapid than that by 6H3 and -200 times slower than that by H atoms. It has also been noted that the reactions of 6 H 3 are more rapid in water than in the gas phase (Thomas, 1967) and this difference is probably a result of solvation rather than caging. Rate constants for several addition reactions have also been reported. As expected these reactions are faster than the abstraction reactions; addition to butadiene is faster than that t o ethylene, and addition of k F 3 faster than that of 6 H 3 . Few other addition rates

-

+

-'

APPLICATION OF RADIATION TECHNIQUES

28 1

have been measured although many such reactions have been observed to take place efficiently: addition of phenyl radicals t o iodobenzene (Cercek and Kongshaug, 1970), addition of k O ; , k H 3 , kOH, RCHCO; and other radicals t o the aci-form of nitromethane (Behar and Fessenden, 1972a), t o fumarate (Neta, 1971) and t o other olefinic and aromatic compounds. Some radicals have been found t o react with substrates both by addition and by electron transfer and it is sometimes difficult to determine the relative importance of the two processes by the pulse radiolysis technique used for the kinetic studies. In cases where the radical products of the two reactions have very different optical absorptions the relative contributions of addition and electron transfer to the overall kinetics can be determined. Because the formation of only one of the radicals is followed while both reactions take place concurrently the observed kinetics are somewhat complex (Greenstock and Dunlop, 1973). It is seen in Table 7 that 60, transfers an electron efficiently to many compounds, the rate constants for the transfer being usually about an order of magnitude lower than those for the direct reactions of these compounds with e i q . In certain cases, however, electron transfer from has not been detected (i.e. k < l o 7 M s - l ) although the compound under consideration reacts with eiq at a diffusion-controlled rate, e.g. the case of benzonitrile (Chutny and Swallow, 1970) or of thymine (Loman and Ebert, 1970). This behaviour can be predicted from the half wave potentials determined by pulse polarography (Lilie et al., 1971). The half-wave potential for the oxidation of k0, is -1.34 V and of k 0 2 H -1.13 V. These radicals can be expected to reduce nitrobenzene, carbon tetrachloride, and biacetyl, but not chloroform or acetaldehyde (Lilie et al., 1971). The potentials for the oxidation of the a-hydroxyalkyl radicals CH,OH, CH,kHOH, and (CH3)2kOH have been found to decrease in that order and the potentials for the dissociated radicals are all lower. The kinetics of electron transfer from these radicals behave accordingly, e.g. the rate constants for their reactions with nitrobenzene are: CH20H, < l o 7 ; CH,kHOH, 3.3 x 10'; (CH3)2kOH, 1.6 x l o 9 ; k H 2 0 - , 2.7 x l o 9 ; CH,kHO-, 3.1 x l o 9 ; (CH3),kO-, 3.0 x l o 9 M - l s - l (Asmus et al., 196613). It should be pointed out that in many cases the pulse polarography technique measures the potential of an irreversible process and not the actual redox potential. Nevertheless, it appears that these measurements are often useful in predicting whether an electron transfer process will take place efficiently. An attempt has been

-'

282

P. NETA

made to use electron transfer rate constants for the determination of redox potentials for organic radicals (Rao and Hayon, 1974). Rates of electron transfer from a radical to a series of acceptors with known redox potentials can be expected to give an estimate of the redox properties" of this radical. However, because the potentials used as references were mostly those for two-electron and not oneelectron transfer, the reported results are not accurate and are probably shifted toward more positive values. Moreover, the experimental conditions were such that in many cases the partial transfer observed is not a result of equilibrium but a kinetic limitation, for example when radical-radical reactions compete with radical-acceptor reaction. One should, therefore be very cautious in applying such pulse radiolysis observations to the estimation of redox potentials. The necessary precautions have been outlined by Meisel and Czapski (1975) who have also correctly measured one-electron redox potentials for several semiquinone radicals. The values measured by pulse radiolysis under established equilibrium conditions include, for example, the one-electron reduction potentials for 1,4-benzoquinone, +O-1 V; 2,5-dimethyl-l,4-benzoquinone,-0.05 V; 0 2 ,-0.14 V; menaquinone, -0-20; and duroquinone, -0.23 V; all values for the quinone/semiquinone or 02/0;couples at pH 7 (D. Meisel, private communication). A large number of electron transfer reactions have been reported, including some studies on cascade electron transfer processes (Adams et al., 1968). The data given in Table 7 represent only a small fraction of those reported in the literature. The last four values in Table 7 have been determined in conjunction with studies of radiation sensitization in which electron transfer processes are believed to play a major role. Some electron-transfer reactions have been suggested to take place although their rate constants are too slow to be measured directly. A typical example is the reduction of organic halides by substituted alkyl radicals. Halide ions are formed in a process such as (62). In the

case of a-hydroxyalkyl radicals reaction (62) can also be formulated as a bromine abstraction to produce BrCH,OH which then hydrolyses. However, similar reactions have been found to take place with radicals such as CH, kHCO; where only the electron transfer mechanism can explain the results (Anbar and Neta, 1967). It

APPLICATION OF RADIATION TECHNIQUES

283

should be pointed out that in irradiated solutions containing a halogen compound and an alcohol a chain reaction may develop through reaction (62) followed by (63). Such a chain process has been 6HzCO; + CHJOH + CH3CO; + k H z 0 H

(63)

observed with several haloaliphatic compounds (Anbar and Neta, 1967; Sorensen et al., 1960) and with halouracils (Bansal et al., 1972; Zimbrick e t al., 1969). Other types of chain reactions which propagate by hydrogen abstraction have also been observed. The increased yield of acetaldehyde in irradiated aqueous solutions of ethylene glycol with increase in solute concentration has been explained by a chain involving reactions (64) and (65) (von Sonntag and Thorns, 1970; Burchill and Perron, 197 1; Schulte-Frohlinde and von Sonntag, C H ~ O H ~ H O H ~ H ~ C H+ O H ~ O -+

kHzCHO + CHzOHCHzOH

-+

CH3CHO + CHzOHkHOH

(64) (65)

1972). The radiation-induced reduction of p-toluenediazonium ions by methanol (Packer e t al., 1971) involves a chain reaction propagating by (66) and (67). k 6 h C H 3 + CHJOH -+ C ~ H S C H+ ~ k H 2 0 H + CH3C6H4Na + CHzO + H+ + Nz + k6H4CH3

(66)

(67)

Radical Reactions w i t h Solvent and Intramolecular Reactions Organic radicals produced in aqueous solutions may undergo structural change before reacting with a solute molecule or with another radical. Under the broad term structural change we include hydrolysis or irreversible protonation, elimination of water or other small molecules, ring openings, and any change in radical structure occurring before the radicals decay. Many such changes have been observed although only a limited number of rate constants have been determined. The values reported in the literature are summarized in Table 8. Elimination of a water molecule takes place with various types of radicals and is catalysed by acids and bases. This has been observed with the radicals from ethylene glycol [reaction (64)] (von Sonntag and Thorns, 1970; Burchill and Perron, 1971) and other glycols, and

284

P. NETA TABLE 8 Rate Constants for Some Radical Rearrangements Reaction

k (s-l),

<

bHOHCF2OHf bH2CHO + H+ + H2O CHxOHCHO- + CH2FHO + OHHOC6HsOH + C ~ H S O + H2O In the presence of 2 x M phosphate buffer HOC6H4(0H)2 + c,jH4(0H)b + H2O In the presence of 2 x M phosphate buffer HOG~HSNH -+ ~C6H&H + H2O HOC~HSCH ~C ~ H S C H + + ~H2O In the presence of 1 x M H+ In the presence of 1 X l o - ' M H+

Reference

8.6 x l o 5 Bansal et al., 1973 > l o 5 Bansal et al., 1973 G lo3 Land and Ebert, 1967 2.9 x lo4 Land and Ebert, 1967 4-6 x l o 4 Adams and Michael, 1967 8.8 x lo4 1 . 4 l~o 5
Adams and Michael, 1967 Christensen, 1972 Christensen et al., 1973 Christensen et al., 1973 Christensen et al., 1973

the detailed mechanism and kinetics have been studied (Bansal et al., 1973). It is seen in Table 8 that once the radical CH,(OH)CHOH undergoes protonation o r deprotonation the rate constant for elimination becomes > l o 5 s-'. Water elimination occurs also with several types of aromatic OH adducts. Rate constants have been measured for OH adducts of various phenols [e.g. reaction (68)] (Land and Ebert, 1967; Adams and Michael, 1967; see also Chrysochoos, 1968), OH

anilines (69) (Christensen, 1972), and toluene (70) (Christensen et

APPLICATION OF RADIATION TECHNIQUES

285

(69)

al., 1973). Reactions (68)-(70) are shown for the para-adducts, but they hold for the ortho- and meta-isomers as well. In fact, all isomers are produced concurrently with different yields and the rate constants reported are the weighted average for all isomers formed. These reactions are all acid-catalysed. The effects of H,O+ and H 2 P 0 i on the rate constants for several radicals are shown in Table

8. It should be pointed out that the radicals produced by reactions (68)- (70) can be also produced directly from the parent compounds by reaction with 0- radicals. The phenoxide ion is oxidized by 0- to the phenoxyl radical (Neta and Schuler, 1975), toluenes undergo hydrogen abstraction by 0- to produce benzyl radicals (Neta et al., 1972; Neta and Schuler, 1973), and aniline is converted to the anilino radical either by hydrogen abstraction or by electron transfer. Elimination similar to (68) and (69) has also been observed by electron spin resonance but rate constants have not been determined. Observations have been made with several phenols and anilines (Neta and Fessenden, 1974), ascorbic acid and similar compounds (Laroff et al., 1972; Kirino and Schuler, 1973), pyrroles [e.g. reaction (71)] and imidazoles (Samuni and Neta, 1973a).

Glycol phosphates have been found to eliminate either water or phosphate depending on the radical structure (Samuni and Neta, 1973b). For example hydrogen abstraction from glycerol-lphosphate yields three different radicals which behave differently. The 3-phosphate radical eliminates water at a moderate rate with

P. NETA

286

base catalysis [reaction (72)] , the 2-phosphate radical eliminates CH2 0PO:-

I CHOH I

CH2OPO:__t

.CHOH

I I

.CH

+ H2O

(72)

CHO

phosphate very rapidly even in neutral solution (73), and the 1-phosphate radical is relatively stable. CHzOPOZ-

I *COH I

CHzOH

-

CH2

I I

CO

+HPO:-

(73)

CHzOH

A different type of elimination has been observed following OH addition at a carbon atom bearing a halogen or a nitro-group [reaction (13), X = F, C1, Br, NO2 , and others]. Such eliminations of hydrogen halide or nitrous acid have been observed with several halouracils (Bansal et al., 1972; Neta, 1972b), fluorinated benzenes (Koster and Asmus, 1973), chloroethylenes (Koster and Asmus, 1971), nitrouracils (Neta and Greenstock, 1973), nitrofurans and bromofurans (Greenstock et al., 1973a), and nitrophenols (Eiben et al., 1971). In all cases studied by pulse radiolysis this 1,l-elimination of hydrogen halide or nitrous acid was found to be very rapid and to take place within less than 1 psec. This type of oxidative dehalogenation or denitration has been found useful in preparing certain types of radicals (Kirino et al., 1974; Schuler et al., 1974). With X = NH2, reaction (13) involves mainly elimination of water with a small contribution of elimination of NH3 (Neta and Fessenden, 1974). A reaction somewhat similar to this 1,l-elimination is that resulting in ring opening as observed for several heterocyclic compounds (Lilie, 1971; Schuler et al., 1973; Neta, 1972d), e.g. reactions (74)-(76). Some rate constants are given in Table 8 and the

287

APPLICATION OF RADIATION TECHNIQUES

measured equilibrium constants were summarized in Table 4. This type of ring opening is in fact a 1,l-elimination and has been observed with 0-,S-, and N-heterocyclic rings following addition of OH at the position adjacent to the heteroatom. Several radicals have been found to hydrolyse in aqueous solutions before decaying into final products. Chlorodifluoro- and dichlorofluoro-methyl radicals hydrolyse very rapidly; reactions (7 7) and (78) have t l l 2 = 15 psec (Balkas et al., 1971) and further hydrolysis kF2C1+ H2O + 6 F 2 0 H + H+ + C1-

(77)

kFC12 + H2O + kFClOH + H+ + C1-

(78)

of the resulting radicals to yield more halide ions takes place on the millisecond timescale. Trifluoromethyl radicals, on the other hand, have been found to be longer-lived in aqueous solutions, the reaction period being -30 psec (Lilie et al., 1972). Hydrolysis has also been suggested with phosphoalkyl radicals [reaction (79)] (von Sonntag et al., 1972). kHzOPO(OCH3)2

+ H2O + 6 H 2 0 H + (CH30)2PO; + HC

(79)

Several anion radicals have been found to undergo protonation on carbon by water. Steady-state esr studies on electron adducts in water have shown that the adducts of acrylate and acetylenedicarboxylate protonate on carbon rapidly whereas the adducts of fumarate and maleate do not (Neta and Fessenden, 1972). A more recent study by pulse techniques has shown that the differences between the various adducts are not qualitative but present differences in the rate of protonation. It has been found that the acid forms of the acrylate electron adduct protonate slowly on carbon whereas the basic form reacts much more rapidly [reaction (SO)] [CH2=CHC02H]

'- + OH-

\

k = 7.7 x lo4 s-l

[CH2=CHCO;]

'- +

HzO

(80) pK'

1201

7

106

s-l

CH3kHCOi

(Fessenden and Chawla, 1974, see also references therein). Similar behaviour has been observed for the fumarate electron adduct but

288

P. NETA

with lower rate constants, i.e. [-O,CCH=CHCO,]'reacts with water on the millisecond time scale to produce -0, CeHCH, CO; whereas [-02CCH=CHC02H]'- reacts much more slowly (Chawla and Fessenden, 1975). Aromatic anion radicals can protonate on the ring to yield cyclohexadienyl radicals. This process has been observed with the electron adduct of benzene [reaction ( S l ) ] (Michael and Hart, 1970)

and of benzenesulphonic acid (Simic and Hoffman, 1972) but hasnot been detected with the benzoate adduct which apparently reacts at a lower rate. The electron adduct of benzonitrile has also been found to be converted into a cyclohexadienyl-type radical (Chutny and Swallow, 1970). Aliphatic nitriles, on the other hand, protonate rapidly on the CN carbon following reduction [reaction (82)] (Neta and Fessenden, 1970b).

-

RCN+~;~

(RCN)'-

H2O

RCH=~

(82)

Among other anion radicals studied are those of halouracils (Bansal et al., 1972; Bhatia and Schuler, 197313). In these cases the electron adducts can either protonate on carbon o r eliminate a halide ion. It has been found that the 5-fluorouracil anion radical protonates on carbon, the 5-bromo derivative eliminates bromide, whereas the 5-chloro radical undergoes both processes (83) with equal rates at pH 5-2 (Bhatia and Schuler, 1973b). The radical formed upon

)I,

+ c1-

3;

or

H

debromination of a-bromotetronic acid by eiq has also been observed to protonate rapidly on carbon (Schuler et al., 1974)

APPLICATION OF RADIATION TECHNIQUES

28 9

Recent results in this laboratory have demonostrated the conversion of ortho- and para-, but not meta-hydroxyphenyl radicals into the phenoxyl radical. Furthermore, an internal electron transfer from the nitro-group to the bromide of p-nitrobenzyl bromide radical anion, t o produce p-nitrobenzyl radical and bromide ion, has been observed.

8. CONCLUDING REMARKS This review has dealt mostly with the production of free radicals and their identification and chemical behaviour. However, radiation techniques can also be utilized to the study of non-radical intermediates, which may be formed, for example, upon the decay of radicals. Only few such studies have been reported in the literature. Simic et al. (1969a) have followed the second order decay of the (CH3)2kOH radical and observed the formation of a sharp UV absorption band which then decays by a first order process. The decay of this intermediate was found to be very slow in neutral solutions ( 2 s - l ) and higher by 3-4 orders of magnitude in acid and base, suggesting a catalysed elimination of water. The authors suggested that disproportionation of the (CH, ) 2 6OH radicals, possibly by an electron transfer mechanism, produced as intermediate the hydrate of acetone. Recent photochemical CIDNP experiments (Laroff and Fischer, 1973) suggest, however, that the intermediate is the enol of acetone. Disproportionation of radicals by electron transfer to yield the negative and positive ions, and the subsequent reactions of the cation to either add OH- or lose H', has been discussed in the previous section. Several aromatic and heterocyclic ions have been suggested to be intermediates in such processes, and the lifetime of these intermediates appears to be long enough to allow an effect of pH on their subsequent reaction (e.g. Haysom et al., 1972). Another type of intermediate has been observed following reduction of nitroaromatic compounds (Grunbein et al., 1970) and subsequent disproportionation of the radicals [reaction (84jl. These 2 XC6H4fio2H + X C 6 h N 0 2 + XC6H4N(OH)2

(84)

nitrosobenzene hydrates have been found to undergo protonation

290

P. NETA

[reaction (85)l followed by rapid dehydration (86). The rate of XC&4N(OH)2 + H+

XC6H4N

,OH 'OH;

,OH XCs&N,OH;

+ H 2 0 + Ht XCBH~NO

(85)

(86)

dehydration depends only slightly on the substituent X. The equilibrium constant K for reaction (85) depends more strongly however, on the nature of X and a Hammett plot has been successfully drawn yielding log K/Ko = 2.20. The pK-values determined are in the region of 0.7 t o -1.3. In the course of pulse radiolysis studies of purine and pyrimidine bases, Fielden et al. (1970) and Greenstock et al. (197313) have been able t o follow the kinetics of deprotonation of these compounds by OH- produced in the irradiated aqueous solution. The observation is made possible by the difference in ultraviolet absorption between the neutral and basic forms. The rate constants for deprotonation were s - l and those for the protonation of found to be (1-2) x 10" M the anion by H+, 4 x 10" M - ' s-'. Radiation chemistry of aqueous solutions has also been applied to the study of micellar systems. Considerable micellar effects on the yield of radiolytic products and on rates of radical reactions have been observed by several authors (Gebicki and Allen, 1969; Fendler and Patterson, 1970; Bansal et al., 1971; Patterson et al., 1971, 1972; Fendler et al., 1972; Wallace and Thomas, 1974; Gratzel et al., 1974). These observations led t o conclusions on the permeability of micelles to various radicals and on the location of substrates in micelles. Recent experiments have also demonstrated a very efficient trapping of eiq by positively charged micelles even when chemical reaction between them did not take place (L. K. Patterson, personal communication). Radiation techniques have been used extensively with nonaqueous systems as well. Much work has been done on aromatic molecule cations and anions and on electron transfer processes involving these species (see the review by Dorfman, 1970). These and other studies on radical ions, on excited states, and on chargetransfer complexes have been reviewed by Fendler and Fendler (1970). Although much of the early work in radiation chemistry was done for the purpose of understanding the action of radiation on the

APPLICATION OF RADIATION TECHNIQUES

29 1

systems studied, most of the recent investigations cited in this review have been carried out with detailed prior knowledge of the action of radiation and with the purpose of solving chemical problems. It is clear that the presently available radiation techniques can in the future make a major contribution to studies of radical reactions, both kinetics and equilibria, and to structural studies. They can open new approaches t o the chemistry of metal-organic compounds. Radiation techniques have become a tool for the chemist and are being applied to various other fields. Radiobiological applications are almost as old as radiation chemistry itself (see e.g. the review by Adams, 1972; and other reviews in the series “Advances in Radiation Biology”). Very recently the radiation chemistry of water has also been fruitfully applied to studies in biochemistry (e.g. Bielski and Chan, 1973; Klug-Roth et al., 1973). Further work along these lines is under way. The present article is an attempt to illustrate the usefulness of the methods developed by radiation researchers to studies in physical organic chemistry. It is hoped that the attention of physical organic chemists will be drawn to the potential of these methods so that further applications will develop.

ACKNOWLEDGEMENTS I wish to thank Professors R. H. Schuler, R. W. Fessenden, and J. H. Fendler for helpful discussions and the U.S. Energy Research and Development Administration for partial support.

REFERENCES Adams, G. E. (1972). In “Advances in Radiation Chemistry”, Vol. 3, p. 125 (M. Burton and J. L. Magee, eds.). John Wiley, New York. Adams, G. E., and Michael, B. D. (1967). Trans. Faraday SOC.63, 1171. Adams, G. E., Boag, J. W., and Michael, B. D. (1965). Trans. Faraday SOC.6 1 , 1417. Adams, G. E., Michael, B. D., and Willson, R. L. (1968). Adv. Chem. Ser. 81, 289. Adams, G. E., Greenstock, C. L., van Hemmen, J. J., and Willson, R. L. (1972). Radiat. Res. 4 9 , 8 5 . Adams, G. E., and Willson, R. L. (1973).J.C.S. Faraday Trans. I 6 9 , 719. Aldrich, J. E., Foldvary, P., Hunt, J. W., Taylor, W. B., and Wolff, R. K. (1972). Rev. Sci. Instr. 43, 991.

292

P. NETA

Anbar, M. (1969). A d v . Phys. Org. Chem. 7, 115. Anbar, M., Bambenek, M., and Ross, A. B. (1973). “Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution. I. Hydrated Electron.” National Bureau of Standards Report No. NSRDS-NBS 43. Anbar, M., Farhataziz, and Ross, A. B. (1974). “Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution. 11. Hydrogen Atom.” National Bureau of Standards Report No. NSRDS-NBS 51. Anbar, M., and Neta, P. (1967).J. C h e m SOC.( A ) ,837. Anbar, M., and Neta, P. (1967). Znt. J . Applied Radiat. Isotopes 18,493. Asmus, K.-D. (1972).Znt.J. Radiat. Phys. Chem. 4, 417. . Bunsenges. Asmus, K.-D., Beck, G., Henglein, A., and Wigger, A. ( 1 9 6 6 ~ )Ber. Phys. Chem. 70,869. Asmus, K.-D., Cercek, B., Ebert, M., Henglein, A., and Wigger, A. (1967). Trans. Faraday SOC.63, 2435. Asmus, K:-D., Henglein, A., and Beck, G. (1966d). Ber. Bunsenges. Phys. Chem. 70.459. Asmus, K.-D., Henglein A., Wigger, A., and Beck, G. (1966a). Ber. Bunsenges. Phys. Chem. 70, 756. Asmus, K.-D., Mockel, H., and Henglein, A. (1973).J. Phys. Chem. 77, 1218. Asmus, K.-D., and Taub, I. A. (1968).J. Phys. Chem. 72, 3382. Asmus, K.-D., Wigger, A., and Henglein, A. (196613). Ber. Bunsenges. Phys. Chem. 70,862. Avery, E. C., Remko, J. R., and Smaller, B. (1968).J. Chem. Phys. 49, 951. Balkas, T. I., Fendler, J. H., and Schuler, R. H. (1970).J. Phys. Chem. 74, 4497. Balkas, T. I., Fendler, J. H., and Schuler, R. H. (1971). J. Phys. Chem. 75,455. Bansal, K. M., Gratzel, M., Henglein, A., and Janata, E. (1973). J. Phys. Chem. 77, 16. Bansal, K. M., and Henglein, A. (1974).J. Phys. Chem. 78, 160. Bansal, K. M., Patterson, L. K., Fendler, E. J., and Fendler, J. H. (1971). Znt. J . Radiat. Phys. Chem. 3,321. Bansal, K. M., Patterson, L. K., and Schuler, R. H. (1972). J. Phys. Chem. 76, 2386. Beck, G. (1969). Znt. J. Radiat. Phys. Chem. 1,361. Behar, D., and Fessenden, R. W. (1972a).J. Phys. Chem. 76, 1710. Behar, D., and Fessenden, R. W. (1972b).J. Phys. Chem. 76, 3945. Behar, D., Samuni, A., and Fessenden, R. W. (1973). J. Phys. Chem. 77, 2055. Bhatia, K. (1973a). Anal. Chem. 45, 1344. Bhatia, K., and Schuler, R. H. (1973a). J. Phys. Chem. 77, 1356. Bhatia, K., and Schuler, R. H. (1973b).J. Phys. Chem. 77, 1888. Bhatia, K., and Schuler, R. H. (1974). J. Phys. Chem.78, 2355. Bielski, B. H. J., and Chan, P. C. (1973). Arch. Biochem. Biophys. 159, 873. Brett, C. L., and Gold, V. (1971). Chem. C o m m . 148. Brett, C. L., and Gold, V. (1973).J.C.S. PerkinZZ 1437. Bronskill, M. J., Taylor, W. B., Wolff, R. K., and Hunt, J. W. (1970). Rev. Sci. Znstr. 41, 333. Bullock, G., and Cooper, R. (1970). Trans. Faraday SOC. 66, 2055. Burchill, C. E., and Ginns, I. S. (197Oa). Can. J . Chem. 48, 1232. Burchill, C. E., and Ginns, 1. S. (1970b). Can. J. Chem. 48, 2628. Burchill, C. E., and Hickling, G. G. (1970). Can. J. Chem. 48, 2466. Burchill, C. E., and Perron, K. M. (1971). Can. J. Chem. 49, 2382. Burchill, C. E., and Thompson, G. F. (1971). Can. J. Chem. 49, 1305. Burchill, C. E., and Wollner, G. P. (1972). Can. J. Chem. 50, 1751.

APPLICATION OF RADIATION TECHNIQUES

293

Burton, M., and Magee, J. L. (eds.). (1969 to present). “Advances in Radiation Chemistry.” John Wiley, New York (Review series). Buxton, G. V. (1970). Trans. Faraday SOC.66, 1656. Buxton, G. V., and Sellers, R. M. (1973). J.C.S. Faraday Trans. I 6 9 , 555. &rsky, P., Fojtik, A., and Zahradnik, R. (1972). Coll. Czechosl. C h e m . C o m m u n . 37,2515. Cercek, B. (1968).J. Phys. Chem. 72, 3832. Cercek, B., and Kongshaug, M. (1970).J . Phys. Chem. 74,4319. Chambers, K. W., Collinson, E., and Dainton, F. S. (1970). Trans. Faraday SOC.66, 142. Chan, P. C., and Bielski, B. H. J . (1973).J. A m e r . C h e m . SOC.95, 5504. Chawla, 0. P., and Fessenden, R. W . (1975).J. Phys. Chem. 79, 76. Christensen, H. (1972). Int. J. Radiat. Phys. Chem. 4, 311. Christensen, H. C., Sehested, K., and Hart, E. J. (1973). J. Phys. Chem. 77, 983. Chrysochoos, J. (1968). Radiat. Res. 33,465. Chutny, B. (1967). Nature 213, 593. Chutny, B., and Swallow, A. J. (1970). Trans. Faraday SOC. 66, 2847. Cohen, H., and Meyerstein, D. (1972). Chem. C o m m . 320. Cohen, H., and Meyerstein, D. (1974). Znorg. C h e m . 13, 2434. Cohen, S. G., and Lam, F. L. (1971). Radiat. Res. 45,462. Czapski, G. (1971). Ann. Rev. Phys. Chem. 22, 1 7 1 . Dorfman, L. M. (1970). A c c o u n t s C h e m Res. 3, 224. Dorfman, L. M., and Adams, G. E. (1973). “Reactivity of the Hydroxyl Radical in Aqueous Solutions.” National Bureau of Standards Report No. NSRDSNBS 46. Dorfman, L. M., Taub, I. A., and Buhler, R. E. (1962). J. Chem. Phys. 36, 305 1. Draganit, I. G., and DraganiC, Z. D. (1971). “The Radiation Chemistry of Water.” Academic Press, New York. Draganit, I., Draganid, Z., PetkoviC, Lj., and Nikolit, A. (1973).J. A m e r . Chem. SOC.95, 7193. Ebert, M., and Howard, A. (eds.). (1965 to present). “Current Topics in Radiation Research.” John Wiley, New York (Review series). Eiben, K., and Fessenden, R. W. (1968).J. Phys. Chem. 72, 3387. Eiben, K., and Fessenden, R. W. (1971). J. Phys. Chem. 75, 1186. Eiben, K., and Schuler, R. H. (1975).J. Chem. Phys. 62, 3093. Eiben, K., Schulte-Frohlinde, D., Suarez, C.. and Zorn, H. (1971). Int. ,I. Radiat. Phys. Chem. 3 , 4 0 9 . Eigen, M. (1964). Angew. Chem. 3,l. Farhataziz and Ross, A. B. (1975). “Selected Specific Rates of Reactions of Transients from Water in Aqueous Solution. 111. Hydroxyl Radical and Perhydroxyl Radical and Their Radical Ions.” National Bureau of Standards Report No. NSRDS-NBS (in press). Fendler, E. J., and Fendler, J. H. (1970). Prog. Phys. Org. Chem. 7, 229. Fendler, J. H., Fendler, E. J., Bogan, G., Patterson, L. K., and Bansal, K. M. (1972). Chem. C o m m . 14. Fendler, J. H., and Gasowski, G. L. (1968).J. Org. Chem. 33, 1865. Fendler, J. H., and Patterson, L. K. (1970). J. Phys. Chem. 74,4608. Fessenden, R. W. (1973).J. Chem. Phys. 5 8 , 2489. Fessenden, R. W. (1975). In “Fast Processes in Radiation Chemistry and Biology”, p. 60 (G. E. Adams, E. M. Fielden, and B. D. Michael, eds.). Institute of Physics, London, and Wiley Interscience, London and New York. Fessenden, R. W., and Chawla, 0. P. (1974).J. A m e r . C h e m . SOC.96, 2262. Fessenden, R. W., and Neta, P. (1972).J. Phys. Chem. 76, 2857.

294

P. NETA

Fessenden, R. W., and Neta, P. (1973). Chem. Phys. Lett. 18, 14. Fessenden, R. W., and Schuler, R. H. (1960).J. Chem. Phys. 33, 935. Fessenden, R. W., and Schuler, R. H. (1963).J. Chem. Phys. 39,2147. Fessenden, R. W., and Schuler, R. H. (1970). In “Advances in Radiation Chemistry”, Vol. 2, 1 (M. Burton and J. L. Magee, eds.). John Wiley, New York. Fielden, E. M., Stevens, G. C., Phillips, J. M., Scholes, G., and Willson, R. L. (1970). Nature (London) 225, 632. Fojtik, A., Czapski, G., and Henglein, A. (1970).J. Phys. Chem. 74, 3204. Garrison, W. M. (1968). I n “Current Topics in Radiation Research”, Vol. 4, 43 (M. Elbert and A. Howard, eds.). North-Holland Publishing Co., Amsterdam. Gebicki, J . M., and Allen, A. 0. (1969).J . Phys. Chem. 73, 2443. Gordon, S., Schmidt, K., and Martin, J. (1974). Rev. Sci. Znstr. 45, 552. Gordon, S., Schmidt, K., and Martin, J. (1975). In “Fast Processes in Radiation Chemistry and Biology”, p. 16 (G. E. Adams, E. M. Fielden, and B. D. Michael, eds.). Institute of Physics, London, and Wiley Interscience, London and New York. Gratzel, M., Bansal, K. M., and Henglein, A. (1973b). Ber. Bunsenges. Phys. Chem. 7 7 , l l . Gratzel, M., and Henglein, A. (1973). Ber. Bunsenges. Phys. Chem. 77, 1. Gratzel, M., Henglein, A., and Bansal, K. M. (1973a). Ber. Bunsenges. Phys. Chem. 77, 6. Gratzel, A., Henglein, A., Lilie, J., and Scheffler, M. (1972). Ber. Bunsenges. Phys. Chem. 76,67. Gratzel, M., Thomas, J. K., and Patterson, L. K. (1974). Chem. Phys. Letts. 29, 393. Greenstock, C. L., and Dunlop, I. (1973).J. Amer. Chem. SOC.95, 6917. Greenstock, C. L., Dunlop, I., and Neta, P. (1973a).J. Phys. Chem. 77, 1187. Greenstock, C. L., Shragge, P. C., and Hunt, J. W. (1973b). J. Phys. Chem. 77, 1624. Grunbein, W., Fojtik, A., and Henglein, A. (1969). Z. Naturforsch. 24b, 1336. Griinbein, W., Fojtik, A., and Henglein, A. (1970). Monatshefte fiir Chemie 101, 1243. Griinbein, W., and Henglein, A. (1969).Ber. Bunsenges. Phys. Chem. 73, 376. Habersbergerova, A., Janovsky, I., and Kourim, P. (1972). Radiat. Res. Rev. 4, 123. Habersbergerova, A., Janovsky, I., and Teply, J. (1968). Radiat. Res. Rev. 1, 109. Hart, E. J., and Anbar, M. (1970). “The Hydrated Electron.” John Wiley, New York. Hayon, E., Ibata, T., Lichtin, N. N., and Simic, M. (1970).J. Amer. Chem. SOC. 92,3898. Hayon, E., Ibata, T., Lichtin, N. N., and Simic, M. (1972). J. Phys. Chem. 76, 2072. Hayon, E., and Simic, M. (1973a).J. Amer. Chem. SOC.95, 1029. Hayon, E., and Simic, M. (1973b).J. Amer. Chem. SOC.95,2433. Hayon, E., and Simic, M. (1973c).J. Amer. Chem. SOC.95, 6681. Hayon, E., and Simic, M. (1974).Accounts Chem. Res. 7, 114. Henglein, A., Schnabel, W., and Wendenburg, J. (1969). “Einfiihrung in die Strahlenchemie.” Verlag Chemie GMBH, Weinheim/Bergstr. Haysom, H. R., Phillips, J. M., and Scholes, G. (1972). J.C.S. Chem. Comm. 1082. Hine, J. (1962). “Physical Organic Chemistry”, 2nd Ed. McGraw-Hill Book Co. Inc., New York. Janzen, E. G. (1969). Accounts Chem. Res. 2,279.

APPLICATION OF RADIATION TECHNIQUES

295

Jonah, C. D., Hart, E. J., and Matheson, M. S. ( 1 9 7 3 ) . J . Phys. Chem. 7 7 , 1838. Kelm, M., Lilie, J., Henglein, A., and Janata, E. (1974).J. Phys. Chem. 7 8 , 882. Kirino, Y. (1975).J. Phys. Chem. 7 9 , 1296. Kirino, Y., and Schuler, R. H. (1973).J. Amer. Chem. S O C . 9 5 , 6926. Kirino, Y., Southwick, P. L., and Schuler, R. H. (1974).J. Amer. Chem. SOC. 9 6 , 673. Klug-Roth, D., Fridovich, I., and Rabani, J. ( 1 9 7 3 ) . J. Amer. Chem. SOC.9 5 , 2786. Kosower, E. M., Teuerstein, A., and Swallow, A. J. (1973).J. Amer. Chem. S O C . 95,6127. Koster, R., and Asmus, K.-D. ( 1 9 7 1 ) . 2. Naturforsch. 2 6 b , 1108. Koster, R., and Asmus, K.-D. (1973).J. Phys. Chem. 7 7 , 7 4 9 . Land, E. J., and Ebert, M. ( 1 9 6 7 ) . Trans. Faraday SOC.6 3 , 1181. Land, E. J., and Swallow, A. J. (1971).Biochim. Biophys. A c t a 2 3 4 , 3 4 . Laroff, G. P., and Fessenden, R. W. (1971).J. Chem. Phys. 5 5 , 5000. Laroff, G. P., and Fessenden, R. W. ( 1 9 7 3 ) . J . Phys. Chem. 7 7 , 1283. Laroff, G. P., Fessenden, R. W., and Schuler, R. H. (1972).J. Amer. Chem. SOC. 9 4 , 9062. Laroff, G. P., and Fischer, H. ( 1 9 7 3 ) . Helv. Chim. A c t a 5 6 , 2011. Lilie, J. ( 1 9 7 1 ) . 2. Naturforsch. 2 6 b , 197. Lilie, J. (1972).J. Phys. Chem. 7 6 , 1487. Lilie, J., Beck, G., and Henglein, A. ( 1 9 6 8 ) . Ber. Bunsenges. Phys. Chem. 7 2 , 529. Lilie, J., Beck, G., and Henglein, A. ( 1 9 7 1 ) . Ber. Bunsenges. Phys. Chem. 7 5 , 458. Lilie, J., Behar, D., Sujdak, R. J., and Schuler, R. H. (1972).J. Phys. Chem. 7 6 , 2517. Lilie, J., and Fessenden, R. W. ( 1 9 7 3 ) .J. Phys. Chem. 7 7 , 674. Lilie, J., and Henglein, A. ( 1 9 6 9 ) .Ber. Bunsenges. Phys. Chem. 7 3 , 170. Loman, H., and Ebert, M. ( 1 9 7 0 ) . I n t . J. Radiat. Biol. 1 8 , 369. Meisel, D., and Czapski, G. ( 1 9 7 5 ) . J . Phys. Chem. 7 9 , 1503. Michael, B. D., and Hart, E. J. (1970).J. Phys. Chem. 7 4 , 2878. MiCiC, 0. I., and Markovit, V. (1972).Int. J. Radiat. Phys. Chem. 4 , 4 3 . Mittal, J. P., and Hayon, E. ( 1 9 7 2 ) .Nature (London) Phys. Sci. 2 4 0 , 21. Nelson, D. A., and Hayon, E. ( 1 9 7 2 ) . J . Phys. Chem. 7 6 , 3200. Neta, P. (1971).J. Phys. Chem. 7 5 , 2570. Neta,P. (1972a). Chem. Rev. 7 2 , 533. Neta, P. (1972b).J. Phys. Chem. 7 6 , 2399. Neta, P. (1972c).Radiat. Res. 5 2 , 4 7 1 . Neta, P. ( 1 9 7 2 d ) .Radiat. Res. 4 9 , 1. Neta, P., and Dorfman,L. M. ( 1 9 6 8 ) .Adv. Chem. Ser. 81, 222. Neta, P., and Fessenden, R. W. (1970a).J. Phys. Chem. 7 4 , 2263. Neta, P., and Fessenden, R. W. (197Ob).J. Phys. Chem. 7 4 , 3362. Neta, P., and Fessenden, R. W. ( 1 9 7 1 ) . J . Phys. Chem. 7 5 , 738. Neta, P., and Fessenden, R. W. (1972).J. Phys. Chem. 7 6 , 1957. Neta, P., and Fessenden, R. W. (1973).J. Phys. Chem. 7 7 , 620. Neta, P., and Fessenden, R. W. (1974).J. Phys. Chem. 7 8 , 523. Neta, P., Fessenden, R. W., and Schuler, R. H. (1971).J. Phys. Chem. 7 5 , 1654. Neta, P., and Greenstock, C. L. (1973). Radiat. Res. 5 4 , 35. Neta, P., Hoffman, M. Z., and Simic, M. ( 1 9 7 2 ) . J . Phys. Chem. 7 6 , 847. Neta, P., and Patterson, L. K. ( 1 9 7 4 ) .J. Phys. Chem. 7 8 , 2211. Neta, P., and Schuler, R. H. (1973).J. Phys. Chem. 7 7 , 1368. Neta, P., and Schuler, R. Hi ( 1 9 7 5 ) . J. Phys. Chem. 7 9 , 1 ; J . Amer. Chem. S O C . 97, 912.

296

P. NETA

Neta, P., Simic, M., and Hayon, E. (1969).J. Phys. Chem. 73, 4207. Neta, P., Simic, M., and Hayon, E. (1970).J. Phys. Chem. 74, 1214. Nucifora, G., Smaller, B., Remko, R., and Avery, E. C. (1972). Radiat. Res. 49, 96. O’Donnell, J. H., and Sangster, D. F. (1970). “Principles of Radiation Chemistry.” American Elsevier Publishing Co., New York. Packer, J. E., House, D. B., and Rasburn, E. J. (1971).J. Chem. SOC.( B ) 1574. Pagsberg, P., Christensen, H., Rabani, J., Nilsson, G., Fenger, J., and Nielsen, S. 0. (1969).J. Phys. Chem. 73, 1029. Pagsberg, P. B., and Hansen, K. B. (1975). In “Fast Processes in Radiation Chemistry and Biology” (G. E. Adams, E. M. Fielden, and B. D. Michael, eds.). Institute of Physics, London, and Wiley Interscience, London and New York. Patterson, L. K., Bansal, K. M., and Fendler, J. H. (1971). Chem. C o m m . 152. Patterson, L. K., Bansal, K. M., Bogan, G., Infante, G. A., Fendler, E. J., and Fendler, J. H. (1972).J. A m e r . Chem. SOC.94,9028. Patterson, L. K., and Lilie, J. (1974). Znt. J. Radiat. Phys. Chem. 6, 129. Paul, H., and Fischer, H. (1971). Helv. Chim. Acta 54, 485. Peter, F. A., and Neta, P. (1972).J. Phys. Chem. 76, 630. Phillips, G. O., and Cundall, R. B. (eds.). (1968 to present). “Radiation Research Reviews.” Elsevier Publishing Co., Amsterdam (Review series). Pryor, W. A., Lin. T. H., Stanley, J. P., and Henderson, R. W. (1973).J. A m e r . C h e m . SOC.95,6993. Ramler, W., Mavrogenes, G., and Johnson, K. (1975). In “Fast Processes in Radiation Chemistry and Biology” (G. E. Adams, E. M. Fielden, and B. D. Michael, eds.). Institute of Physics, London, and Wiley Intersciznce, London and New York. Rao, P. S., and Hayon, E. (1973). J . Phys. Chem. 77, 2274. Rao, P. S., and Hayon, E. (1974).J. A m e r . Chem. SOC.96, 1287. Samuni, A., and Neta, P. (1973a).J. Phys. Chem. 77, 1629. Samuni, A., and Neta, P. (197313).J. Phys. Chem. 77, 2425. Schmidt, K. H. (1972).Int. J . Radiat. Phys. Chem. 4,439. Schmidt, K. H., and Buck, W. L. (1966). Science 151, 70. Schoneshofer, M. (1972). Z. Naturforsch. 27b, 549. Schuler, M. A., Bhatia, K., and Schuler, R. H. (1974).J. Phys. Chem. 78, 1063. Schuler, M. A., and Fessenden, R. W. (1975). Unpublished results. Schuler, R. H., Laroff, G. P., and Fessenden, R. W. (1973). J . Phys. Chem. 77, 456. Schuler, R. H., and Patterson, L. K. (1974). Chem. Phys. Lett. 27, 369. Schulte-Frohlinde, D., Reutebuch, G., and von Sonntag, C. (1973). Int. J. Radiat. Phys. Chem. 5,331. Schulte-Frohlinde, D., and von Sonntag, C. (1972). Israel J. Chem. 10, 1139. Simic, M., and Hayon, E. (1971). Radiat. Res. 48, 244. Simic, M., and Hayon, E. (1973).J. Phys. Chem. 77,996. Simic, M., and Hoffman, M. Z. (1972).J. Phys. Chem. 76, 1398. Simic, M., Neta, P., and Hayon, E. (1969a).J. Phys. Chem. 73, 3794. Simic, M., Neta, P., and Hayon, E. (1969b).J. Phys. Chem. 73, 4214. Simic, M., Neta, P., and Hayon, E. (1971).Int. J. Radiat. Phys. Chem. 3, 309. Simic, M., Neta, P., and Hayon, E. (1973).J. Phys. Chem. 77, 2662. Smaller, B., Avery, E. C., and Remko, J. R. (1971).J. Chem. Phys. 55, 2414. Sorensen, V. G., Bhale, V. M., McCallum, K. J., and Woods, R. J. (1970). Can. J . C h e m . 48,2542. Soylemez, T., and Schuler, R. H. (1974).J. Phys. Chem. 78, 1052.

APPLICATION OF RADIATION TECHNIQUES

29 7

Stevens, G. C., Clarke, R. M., and Hart, E. J. (1972).J. Phys. Chem. 76, 3863. Stockhausen, K., and Henglein, A. (1971). Ber. Bunsenges. Phys. Chem. 75, 833. Swallow, A. J. (1973). “Introduction to Radiation Chemistry.” J o h n Wiley, New York. Thomas, J. K. (1967).J. Phys. Chem. 71,1919. Verma, N. C., and Fessenden, R. W. (1973).J. Chem. Phys. 58, 2501. von Sonntag, C., Ansorge, G., Sugimori, A., Omori, T., Koltzenburg, G., and Schulte-Frohlinde, D. (1972). 2. Naturforsch. 27b, 471. von Sonntag, C., and Thorns, E. (1970). 2. Naturforsch. 25b, 1405. Wallace, S. C., and Thomas, J. K. (1973). Radiat. Res. 54,49. Wander, R., Neta, P., and Dorfman, L. M. (1968).J. Phys. Chem. 72, 2946. Wertz, J. E., and Bolton, J. R. (1972). “Electron Spin Resonance. Elementary Theory and Practical Applications.” McGraw-Hill, New York. Wiberg, K. B. (1964). “Physical Organic Chemistry.” John Wiley, New York. Wigger, A., Henglein, A., and Asmus, K.-D. (1967). Ber. Bunsenges. Phys. Chem. 71,513. Willson, R. L. (1971a). Chem. Comm. 1249. Willson, R. 1,. (1971b). Trans. Faraday SOC.67, 3020. Zemel, H., and Neta, P. (1973).Radiat. Res. 55, 393. Zimbrick, J. D., Ward, J. F., and Myers, L. S., Jr. (1969).Znt. J. Radiat. Biol. 16, 505.

This Page Intentionally Left Blank

Author Index Numbers in italics refer t o the pages on which references are listed at the end of each article.

A

Aalbersberg, W. I., 76,122 Aaron, J. J., 177, 182, 183, 184, 192, 193, 210,215, 216 Abate, K., 182, 215 Abu-Elgheit, M., 195, 218 Adams, G. E., 236, 244, 254, 259, 261, 268, 269, 272, 278, 279, 282, 284, 291, 291, 293 Aldrich, J. E., 226,291 Adams, R. N., 2, 7, 8, 9, 43, 51, 77, 78, 79, 116, 124, 125 Adamson, A. W., 147,216 Adita, S., 177, 215 Aikens, D. A., 57,118 Akermark, B., 29,129 Albert, A., 136, 215 Allen, A. O., 290,294 Allen, M. J., 27, 36, 46, 56, 116 Allendoerfer, R. D., 106,116 Allinger, N. L., 111, 129 Amelotti, C. W., 14, 119 Anbar, M., 66, 122, 224, 233, 234, 282, 283,292, 294 Anderson, J. D., 2, 83, 84,116, 126 Anderson, L. B., 75, 116, 128 Anderson, J., 105,116 Andreades, S., 11, 43, 56, 57,116 Andrulis, P. J., 3, 116 Andrusev, M. M., 91,119 Aniskova, L. V., 57, 116 Ansorge, G., 287,296 Aoki, K., 172,220 Arapakos, P. G., 12,116 Argauer, R., 137, 138, 214, 215 Arzak, A. A., 56,127 Asahara, T., 13, 67, 69,116 Asmus, K-D., 228, 238, 244, 258, 263, 266, 268, 272, 275, 278, 279,281, 286,292, 295, 297 Aten, A. C., 76, 107,116 Au-Young, Y. K., 17,57,117 299

Avaca, A., 13, 69,116 Avery, E. C., 229, 253, 278, 292, 296 Avigal, I., 144, 164, 215 Aylward, G. H., 107,116

B Baggaley, A. J., 13, 56,116 Bailey, D. N., 213,215 Baizer, M. M., 2, 9, 12, 15, 17, 34, 75, 82, 83, 84, 86, 95, 111, 116, 126, 129 Balkas, T. I., 231, 287, 292 Ballard, R E., 138,177, 182,183, 215 Bambenek, M., 233, 292 Bank, S., 111, 116 Bansal, K. M., 228,238,258,276, 283,284, 286,288, 290,292, 293, 294, 296 Bard, A. J., 3, 10, 71, 76, 78, 81, 85, 116, 118,126, 127 Barker, C. C., 192, 215 Barnes, K. K., 2,43,99, 110,124 Barradas, R. G., 23,91,116, 118 Bartocci, G., 213, 215 Bartok, W., 161, 164, 192,215 Barton, F. E., 108, 121 Bascombe, K. N., 211,216 Basolo, F., 3, 116 Batrakov, V. V., 6, 22, 119 Bauer, D., 43, 48,49,116 Baumann, H.,111,125 Bazoin, M., 43, 117 Bechgaard, K., 9, 10, 16, 57,116,127,128 Beck, F., 2,4, 33, 34, 86,95,117, 122 Beck,G., 228,258, 263, 266, 278, 279, 281, 292, 295 Beck, J. P., 43, 49, 116 Beck, P., 19,122 Behar, D., 248, 249, 258, 280, 281, 287, 292, 295 Bell, R. P., 184, 203, 211, 216

AUTHOR INDEX

Belleau, B., 17, 56, 57, 117, 129 Bellus, D., 187, 218 Benkeser, R. A., 13, 67, 69, 117 Benson, S. W., 59,128 Berezin, I. V., 203, 218 Berg, R A., 143, 220 Berger, P. A., 111, 116 Bernal, I., 12, 126 Bernath, T., 3, 60,123 Bernhardsson, E., 13, 117 Bertozzi, R. J., 43,118 Bertrin, J., 205, 216 Bertram, J., 43, 51,117 Bessiere, J., 43,126 Beveridge, D. L., 192, 207,218 Bewick, A., 2, 13, 69, 98,116, 117 Bhale, V. M., 283, 296 Bhatia, K., 226, 254, 275, 278, 280, 286, 288,292, 296 Bickley, H. T., 59, 126 Bielski, B. H. J., 264, 291, 292, 293 Bilevich, K. A., 3, 111 , l1 7 Billon, J. P., 12,43, 117 Birch, A. J., 13, 67,117 Birks, J. B., 132, 143, 146, 216 Bishop, N. I., 3,117 Blackburn, G. M., 57, 11 7 Blaunstein, R. R., 133,217 Blomgren, E., 23, 91, 117 Blount, H. N., 56, 7 9 , l 1 7 Boag, J. W., 236,291 Bobbitt, J. M., 16,98, 117 Bockris, J. O’M., 22, 23, 46, 87, 91, 106, 117, 122, 124 Boekelheide, V., 111,121 Bogan, G., 290,293, 296 Bolton, J. R, 247, 297 Bonneau, R., 191, 202,216, 21 7 Bonner, W. A., 13, 56,57,117, 124 Bordi, S., 91, 117 Bortolus, P., 213, 215 Bose, E., 58, 11 7 Bouwhuis, E., 177, 198, 216 Bowie, L. J., 215, 216 Boyd, R. H., 21 1,216 Brago, I. N., 2, 117 Brand, L., 146, 147, 170, 196, 201, 202, 214, 215, 216, 219, 220 Brandt, W. W., 144,170,177,219 Bratzel, M. P., 183, 192, 210, 216 Brauman, J. I., 204, 216 Braun, W., 59,117 Breant, M., 43, 117 Bremphis, R V., 195, 220 Brennan, M. P. J., 113, 1 1 4 , l l 7 Breslow, R., 111, 117 Brett, C. L., 235, 238, 292 Brettle, R., 13,56, 113, 114, 116, 117

Bride, M. H., 192,215 Bridges, J. W., 161, 177, 216 Briegleb, G., 109, 117 Bnnen, J. S., 177, 216 Brinn, I., 167, 170, 201, 202, 207, 219 Britt, W. J., 10, 81, 121 Britton, W. E., 105, 121 Brixius, D. W., 3, 129 Broman, R., 79, 118 Bronskill, M. J., 227, 292 Brookes, C. J., 17, 118 Brouwer, D. M., 52, 118 Brown, 0. R., 115, 127, 118 Brown, R, 43,128 Buck, W. I., 228,296 Buhler, R E., 278, 293 Buisson, J., 43, 11 7 Bullock, G., 278,292 Bulska, It, 177, 216 Bunt, J. C., 60, 123 Bunton, C. A., 21 1, 216 Bunyan, P. J., 57, 118 Burchill, C. E., 278, 279, 280, 283, 292 Burdon, J., 56, 118 Burgert, B. E., 56, 57,125 Burr, J. G., 191,213,216 Burton, M., 224, 293 Buvet, R., 43, 126 Buxton, G. V.,256,258, 270,293 Byrd, L., 9, 118

C

Caldin, E. F., 203,216 Callingham, M., 43,122 Campbell, C. B., 57,124 Capomacchia, A. C., 167. 170. 171. 177. 182, 196, 198, 200, 2i5,216, 218,219; 220,221 Carroll, F. A., 157,216 Cksky, P., 268, 293 Carter, J. G., 133,217 Casanova, J., 2, 12, 15, 97, 105,118, 119 Caullet, C., 9, 13, 56, 123 Cauquis, G., 7, 9, 36, 43, 56, 64, 118 Cedheim, L., 17, 37, 97, 118 Cercek, B., 244, 274, 275, 278, 281, 292, 293 Cetina, R, 207,Zl 6 Cetorelli, J. T., 138, 167, 177, 221 Chalvet, O., 205, 207, 216 Chambers, J. Q,, 50, 91, 118, 124 Chambers, K. W., 278,293

301

AUTHOR INDEX

D Chan, A., 191,213,216 Chan, A. W. K., 212,217 Dahms, H., 2 3 , 9 1 , 1 1 9 Chan, P. C., 264, 291,292, 293 Chaney, A., 62,118 Dainton, S., Chang, J., 2, 56,118 Dalle-Molle, E., 114, 1 2 0 Chawla, 0. P., 287, 288, 293 Damaskin, B. B., 6, 22, 91, 119, 1 2 2 Chen, R. F., 214, 216 Daniels, F., 3, 119 Chibisov, A. K., 141,Zl 6 Danly, D., 1 7 , 1 1 9 Childs, R F., 21 3, 218 Dannerberg, J. .J., 205, 216 Childs, W. V., 85, 118 Dauben, H. J., 1 5 , 1 2 9 Chkir, M., 17, 118 Daudel, R., 204, 205, 207, 216 Chodkowska, A., 177,216 Davies, C. W., 150, 216 Christensen, H., 236, 244, 247, 272, 284, Davies, D. S., 177, 216 293, 296 Davies, G. R., 36, 43,129 Christensen, H. C., 244, 272, 284, 293 Davis, R E., 113, 1 2 3 Christophorou, L. G., 133,21 7 De Boer, E., 76, 122 Chruma,J. L., 12, 15, 111, 116 Debye, P., 154, 21 7 Chrysochoos, J., 284, 293 Degner, D., 93, 94, 1 2 2 Chu, W., 111,117 DeKlein, W. J., 56, 119 Chung, L. L., 111,121 DeLucca, M., 214, 215, 216 Chutny, B., 247, 281, 288, 293 Demas, J. N., 147, 216 Cimino, G. M., 211, 219 Depatie, C. B., 59, 119 Clapper, G. L., 98,126 Derkacheva, L. D., 170, 188, 191, 212, 216 Clark, D.B., 12,48, 57, 65, 118, 1 2 3 Desmond, M. M., 108,125 Clarke, R M., 244,272, 278,297 Dessau, R M., 3, 119 Cleghorn, H. P., 98, 117 DeTar, D. F., 59, 119 Clusius, K., 33,118 Dewar, M. J. S., 3,116, 205,216 Coe, P. L., 17, 118 Deyrup, A. J., 209, 21 7 Coetzee, J. F., 43, 57, 118, 1 2 3 Dhingra, R. C., 133, 216 Cohen, H., 280,293 Dickinson, T., 33,119 Cohen, L A., 18, 118, 122 Dienes, A., 21 2 , 2 2 0 Cohen, M. J., 3 , 1 1 8 Dietz, R, 3, 14, 28, 65, 82, 107, 116, 1 1 9 Cohen, S. G., 275,293 Dirlam, J. P., 13, 56, 96,119, 1 2 6 Coleman, A. E., 5 7 , 1 1 8 Dolphin, D., 43, 44, 1 2 7 Colemann, J. P., 16, 51, 93, 117, 118, Dorfman, L. M., 3, 71, 119, 230, 236, 237, 127 244, 272, 273, 278, 290,293, 295, 297 Coleman, L. B., 3,118 Doty, J. C., 177, 1 7 9 , 2 1 7 Collinson, E., 278, 293 Doughty, A., 43,119 Colombi, M., 57, 60, 62, 1 2 1 Doyle, M. P., 37, 119 Colpa, J. P., 203, 207, 216 Draganic, I. G., 224, 234, 293 Conway, B. E., 2, 13, 16, 23, 91, 98, 104, Draganic, 2. D., 224, 234,293 112,116,118, 1 1 9 Draimaix, R., 172, 177, 220 Cook, C. D., 59,119 Drewer, R. J., 191, 219 Cooke, B. A., 2 , 1 1 9 Dubinin, A. G., 5 7 , 1 1 6 Cooper, R, 278,292 Dubois, J-E., 43,119 Corey, E. J., 15,119 Dunlop, I., 238, 249, 251, 256, 263, 281, Cottrell, P. T., 64, 119 286,294 Coulson, C. A., 207, 216 Dupin, M., 43, 117 Courbis, P., 11,119 Dutton, D. R., 205, 220 Courtot-Coupez, J., 43, 4 4 , 1 1 9 Duty, R. C., 12, 26, 28, 129 Covitz, F. H., 2, 15, 119, 1 2 7 Cowan, D. O., 2,120 Cowgill, R. W., 191, 216 Cramer, J., 3, 1 2 9 Cu, A., 213,216 E Cundall, R. B., 224, 296 Curphey, T. C., 1 4 , l 19 Eads. D. K.. 111.128 Czapski, G., 256, 279, 282, 293, 294, 295 Easterly, C.’E., 133,217

F.

278,203

302

AUTHOR INDEX

Eberson, L., 2,4, 10, 11, 12, 13, 16, 17, 18, 25, 28, 32, 33, 35, 36, 37, 43, 44, 56, 57, 58, 59, 60, 63, 64, 71, 72, 73, 75, 77, 81, 92, 93, 96, 97, 98, 99, 104, 105, 108, 113, 116, 117, 118, 119, 120, 125, 126 Ebert, M., 224, 247, 272, 275, 281, 284, 292, 293, 295 Edward, J. T., 21 1,217 Edwards, J. O., 111,120 Edwards, J. W., 138, 177, 182, 183, 215 Eiben, K., 229, 230, 238, 247, 248, 250, 264, 269,286,293 Eigen, M., 252, 269, 293 Eirich, F. R., 59,117 El-Bayoumi, M. A., 195,215,218 Ellerhorst, R H., 159, 161, 166, 192, 217 Ellis, D. W., 165, 21 7 Elving, P. J., 2, 43, 120, 129 Epling, G. A., 15,129 Epstein, B. D., 114, 120 Erickson, R. E., 18,120 Evans, D. F., 137,Zl 7 Evans, D. H., 98,126

Fioshin, M. Ya., 2, 14, 17, 36, 44, 56, 57, 87,116,120,128 Firoi, G., 57, 60, 62,121 Fischer, C. M., 18,120 Fischer, R, 256, 289, 295 Fleischmann, M., 2, 4, 12, 41, 43, 44, 48, 50, 51, 56, 57, 64, 65,87, 99, 105, 113, 117,118, 119, 120, 121, 123 Flurry, R L., 203, 207, 21 7 Fojtik, A., 256, 263, 268, 289, 293, 294 Foldvary, P., 226,291 Formaro, L., 57, 60, 62,121 Forster, T., 132, 134, 144, 145, 155, 157, 159,170,197,198,217 Foucault, A., 48, 49, 116 Fraenkel, G. K., 12,126 Fratev, F., 207, 220 Fridovich, I., 291,295 Fritsch, J. M., 10, 43, 53, 81, 121, 124 Frumkin, A. N., 22,119 Fry, A. J., 2, 29, 95, 97, 105, 108, 111, 121, 127 Fuchs, H., 19,122 Fueno, T., 56,129

F

G

Faita, G., 56, 120 Falck, J. R., 16,120 Farber, L., 18,118 Fa.rhataziz, 234, 236, 292, 293 Famngton, G. C., 43,44,127 Faure, J., 191,217 Favaro, G., 143, 177, 179, 192, 203, 217 Fayet, M., 191, 202,217 Feitelson, J., 140, 144, 146, 147, 156, 164, 170, 201,202, 214,215, 218, 219 Feldberg, S. W., 77, 120 Fendler, E. J., 224, 290, 292, 293, 296 Fendler, J. R,224, 231, 238, 287, 290,292, 293, 296 Feng, E., 43,122 Fenger, J., 236,296 Fernando, Q., 182, 183, 207,Zl 7 Ferraris, J., 2, 120 Fessenden, R. W., 228, 229, 230, 235, 243, 247, 248, 249, 250, 251, 252, 253, 255, 257, 258, 263, 264, 265, 266, 269, 270, 272, 277, 280, 281, 285, 286, 287, 288, 292, 293, 294, 295, 296, 297 Fichter, F., 2, 26, 36,120 Ficquelmont, A. M., 43,119 Fielden, E. M., 290, 294 Fink, D. W., 191, 21 7 Finkelstein, M., 11, 13, 43, 56, 57, 60, 61, 64,113,114,115,120, 127

Galiano, F. R., 14,123 Gandel’sman, L. Z., 192, 221 Garito, A. F., 3, 118 Garnett, J. L., 107, 116 Garrison, W. M., 280,294 Garst, J. F., 108, 121 Gasowski, G. L., 238,293 Gassman, P. G., 92, 121 Gavrilova, A. E., 48, 121 Gebicki, J. M., 290, 294 Geiger, W. E., 81, 121 Geldof, P. A,133,Zl 7 Gerdil, R, 15,121 Gerovich, V. M., 91, 119 Gershon, H., 182, 183, 192, 193, 210,215, 216, 21 9 Geske, D. H., 10, 50, 81,121, 123 Gilde, R-G., 2, 14, 17, 56,121, 128 Gileadi, E., 2, 4, 9, 22, 91, 93, 94,121, 122, 123, 126 Gilles, J.-M., 111, 125 Ginns, I. S., 279, 292 Given, P. H., 76,121 Gladkikh, I. P., 91,119 Glass, G. E., 64, 121 Glass, R. S., 10, 81, 121 Glasstone, S., 26, 121 Gleicher, G. J., 108, 121 Gleicher, M. K., 108, 121

AUTHOR INDEX Godfrey, T. S., 212,217 Gohlke,J. R., 146, 147, 201, 202, 214, 215, 219 Gold, V., 235, 238, 292 Goldschmidt, C. R., 151, 154, 218 Goldschmidt, S., 31, 32,121 Goluber, A. I., 35,122 Gonikberg, M. G., 48,121 Gonzalez, E. R., 115,118 Goodridge, F., 4,121 Gordon, L., 191, 218 Gordon, S., 227, 294 Gourley, R. N., 93,94,121 Govindjee, 3, 126 Grabowska,A., 137,177,205,207,216, 217 Grabowski, Z. R., 171, 207,217, 219 GrZnse, S., 32, 33, 120 Gratzel, M., 228, 256,284, 290,292, 294 Green, M., 23, 91,117, 119 Greenstock, C. L., 238, 249, 251, 256, 263, 269, 279, 281, 286, 290, 291, 294, 295 Griffin, G. W., 15, 129 Griffiths, J., 192, 21 7 Griller, D., 59,121 Grimshaw, J., 93, 94,121 Grisdale, P. J., 177, 179, 217 Grunbein, W.,263, 268, 289, 294 Gruver, G. A., 9,121 Guillemonat, A., 11, 119 Gutmann, F., 2, 121 Gutsche, C. D., 212, 213,217

H Haas, H., 31, 32,121 Habersbergerova, A., 245, 294 Hagen, R., 172,217 Hakozaki, S., 11 1,128 Hale, J. M., 107,121 Hallas, G., 192, 215 Hammerich, O., 9, 10, 12, 16, 52, 7 7 , 79, 81, 82, 86, 116, 121, 122, 127 Hammett, L. P., 209, 21 7 Hammond, G. S., 157,216 Hancock,C.K., 161, 165, 192, 218 Hansen, K. B., 227,296 Hamman, A., 157,217 Harris, L. B., 35, 122 Harrison, J. A., 127,118 Hart, E. J., 66, 122, 224, 227, 233, 234, 244, 272, 278, 284, 288, 293, 294, 295, 29 7 Hartman, R. B., 161, 164, 192, 215 Haufe, J., 2, 33, 122 Hawkes, G. E., 97,122 Hayes, J. W., 146, 147, 201, 202, 214, 215, 21 9

303

Haylock, J. C., 139, 140, 150, 182, 199, 210,217 Hayon, E., 237, 244, 245, 253, 256, 258, 259, 260, 261, 262, 264, 265, 267, 272, 273, 282, 289,294,295,296 Haysom, H. R., 276, 280, 289, 294 Headridge, J. B., 43,122 Heeger, A. J., 3, 118 Heiba, E. I., 3, 119 Heiland, W., 9 1, 122 Heilbronner, E., 172, 21 7 Heineman, W. R., 79,118 Heinzelmann, W., 141,218 Helgee, B., 4, 108,120 Heller, A., 214, 221 Henderson, R. W., 235, 296 Henglein, A., 224, 228. 244, 247, 256, 258, 259, 263, 266, 268, 272, 275, 276, 278, 279, 280, 281, 284, 289,292, 294. 295, 297 Henry, B. R., 133, 134, 158, 217 Henson, R. M. C., 170, 205, 21 7 Herbert, M. A., 191, 21 7 Hercules, D. M., 2, 10, 122, 165, 171, 201, 213,215, 21 7, 221 Hert, R. C., 177, 216 Herz, J., 57, 123 Hey, D. H., 57,118 Hickling, A., 26, 121 Hickling, G. G., 280,292 Hiebaum, G., 207,220 Himel, C. M., 165,219 Hine, J., 267,294 Hine, K. E., 213, 218 Hirota, K., 194,218 Hochstrasser, R. M., 206, 218 Hoffman, M. Z., 143, 220, 237, 245, 251, 256, 259, 260, 266, 269,272, 285, 288, 295, 296 Hoffmann, A. K., 12,124 Hoffmann, H., 19,122 Hogeveen, H., 52,118 Hogfeldt, E., 211, 218 Hoggett, J. G., 36,122 Hoijtink, G. J., 2, 76, 107, 108, 116, 122, 133, 21 7 Holy, N. L., 3, 122 Hopkins, T. A., 191, 214, 219 Hopkinson, A. C., 161, 170, 171, 172, 188, 210,211,218 Horner, L., 12, 14, 18, 19, 93,94,97,120, 122 House, D. B., 278, 283, 296 House, H. O., 43,122 Howard, A., 224, 293 Hrdlovic, P., 187, 218 Huang, S. K., 211,216 Hudnall, P. M., 11 1, 126

304

AUTHOR INDEX

Huebert, B. J., 76, 107, 122 Humffray, A. A., 2, 122 Hunt, J. W., 226, 227, 238, 251, 256, 263, 269,290,291, 292, 294 Hunt, R. L., 3, 116 Qurley, R, 213,218 Hussain, S. K., 158,218

I Ibata, T., 244, 259, 260, 272, 294 Idoux, J. P., 161, 165, 192, 218 Ikeda,A., 11, 13, 111, 127, 1 2 8 Ikenoya, S., 57,122 Infante, G. A., 290, 296 Ingram, K. C., 195, 215, 218 Inoue, T., 13,122 Ireland, J. F., 136, 138, 141, 142, 157, 159, 185,187,188,193, 205,213,218 Irwin, R., 215, 216 Isaks, M., 192, 218 Isoe, S., 18,122 Isserlis, G., 2, 122 Iversen, P., 28, 41, 43,122, 1 2 4 Iwakura, C., 12, 57,124

J Jackman, D., 182, 215, 221 Jackson, G., 136, 141, 143, 154, 170, 177, 182, 205,208,218 Jacobs, J., 207,216 Jaffd, H. H., 157, 159, 161, 166, 167, 192, 207,217, 218 Jagur-Grodzinski, J., 85, 128 Jain, D. V. S., 207,216 Janata, E., 228, 258, 280, 284, 292, 295 Janata, J., 212, 21 8 Janovsky, I., 245, 294 Jansen, G., 21 2,218 Janssen, M. J., 177, 198, 216 Janzen, E. G., 251,294 Jason, M., 15, 129 Jenevein, R. M., 98,128 Jesch, C., 23, 117 Johns, H. E., 191, 21 7 Johnson, K., 227,296 Jonah, C. D., 227,295 Jones, H. L., 157, 167, 192, 207, 218 Jones, J. R., 203, 218 Jordan, J. W., 213,221 Jordan, R. W., 14,123 J o s h , T., 65, 123 Joussot-Dubien, J., 133, 154, 191, 217, 218

Jubault, M., 94,122 Juneau, R. J., 191, 200, 214,221

K Kadyrov, M. Kh., 35,122 Kaganovich, R. I., 9 1,119, 122 Kaiser, E. M., 13, 67, 69, 117 Kaisheva, M. K., 91,122 Kamm, K. S., 143,147, 213,221 Kamneva, A-L., 14, 17, 56,120 Kaneko, H., 13,67,69,116 Kardos, A. M., 12,129 Kariv, E., 93, 94, 123 Kasche, V., 143, 219 Kasha, M., 132,218 Kato, S., 191, 213, 218 Katz, M., 17, 57, 123 Keating, J. T., 35,123 Kelm, M., 280,295 Kern, D. H., 98,128 Keszthelyi, C. R., 85,118 Kim, K. S., 113,123 Kimura, K., 12, 57,124 Kirino, Y., 257, 258, 285, 286, 295 Waning, U. K., 151, 154,218 Klen, R., 15,126 Klug-Roth, D., 291,295 Knight, A. E. W., 147,218 Kobayashi, K., 99,123 Koch, V. R., 12, 65,123 Kochi, J. K., 3,60,123,127 Koehl, W. J., 35, 57, 113,123 Koehler, W. R., 191, 21 7 Koizumi,M., 144,191,213,218, 219 Kokubun, H., 188,Zl 8 Kollmar, H., 52, 123 Kolthoff, 1. M., 57, 123 Koltzenburg, G., 287,296 Kongshaug, M., 278, 281,293 Konschin, H., 164, 218 Korinek, K., 2,124 Kornienko, A. G., 14,17,56,120 Kornprobst, J.-M., 12, 92,123 Kosaka, T., 11, 1 2 8 Koshechkina, L. I., 35,122 Kosower, E. M., 277,295 Koster, R., 238, 286, 295 Kourim, P., 245,294 Kovacic, P., 48,123 Kovi, P. J., 161, 170, 171, 182, 185, 188, 193, 194, 195, 198, 200, 215, 216, 218, 219,220 Koyama, K., 11, 13, 27, 56, 57, 122, 123, 129 Koziol, J., 196,220 Koziolowa, A., 196, 220

305

AUTHOR INDEX Kramer, H. E. A., 207, 220 Krone, L. H., 111 , 1 2 8 Krupicka, J., 15, 111, 1 2 9 Kujawa, E. P., 57, 124 Kunugi, A., 5 7 , 1 2 3 Kunzel, H., 14, 17, 56,127 Kuwana, T., 3,9, 43, 48, 79,117, 118, 121, 123, 1 2 5 Kuwata, K., 10, 8 1 , 1 2 3 Kuz’min, M. G., 203, 218 Kysel, O., 187, 207, 214, 218

L Labhart, H., 141,218 Lacaze, P-C., 43, 119 Lahiri, S. C., 177, 215 Lam, F. L., 275, 293 Lambert, F. L., 99,123 Lambert, R. F., 13, 67, 69, 1 1 7 Lamm, B., 12, 123 Lamy, E., 82,123 Land, E. J., 243, 244, 247, 284, 295 Larcombe, B. E., 82,119 Large, R. F., 2, 56,118 Laroff, G. P., 248, 249, 250, 251, 252, 255, 257, 258, 269, 270, 285, 286, 289,293, 296 Lasser, N., 140, 214, 218 Laurent, A., 12,92,123 Laurent, E., 12, 1 2 3 Laurent-Dieuzeide, E., 12, 92, 1 2 3 Layloff, T. P., 14,119 Lazar, M., 187,218 Le Dtmtzet, M., 43, 44, 1 1 9 Ledger, M. B., 206,218 Ledwith, A., 3, 71, 78,116, 1 2 3 Lehmkuhl, H., 2, 19,123 Leicher, W., 31, 32, 121 Lelandais, D., 17,118 Leonhardt, H., 191, 218 Lesclaux, R., 133, 154, 218 Letsinger, R L., 63,123 Leung, M., 57,123 Leute, R., 205, 216 Levy, M., 30,123 Libert, M., 9, 13, 56, 123 Lichtin, N. N., 244, 259, 260, 272, 294 Liedke, P., 159, 161, 193, 196, 197, 218 Lietaer, D., 159, 170, 171, 172, 203, 204, 220 Liler, M., 210,219 Lilie, J., 226, 228, 244, 247, 258, 259, 264, 267, 280, 281, 284, 286,287,294, 295, 296

tin, T. R, 235,296 Lindqvist, L., 143, 191, 219 Lindsey, R. V., 14, 17, 56, 1 2 3 Linstead, R. P., 60,123 Lipkin, D., 1 4 , 1 2 3 Lippert, E., 173, 219 Little, W. A., 3, 123 Liu, M. Y., 1 0 4 , 1 2 6 Livingston, R., 191, 218 Loken, M., 21 5,216 Loken,M. R., 146, 147, 201, 202, 214, 215, 219 Loman, R, 281,295 Longchamp, S., 9, 13, 5 6 , 1 2 3 Longworth,J. W., 184, 219 Loveland, W., 13, 119 Lucchesi, P. J., 161, 164, 192, 215 Ludman, C. J., 56,123 Lumry, R. W., 3, 101,126 Lund, H., 2, 11, 12, 18, 28, 29, 41, 43, 56, 65,71, 108,119, 122, 124 Lunnala, R., 164, 218 Lutz, H., 206, 218 Lykos, P. G., 203,207,217 Lyons, E., 2, 121

M

McCain, J. H., 63, 1 2 3 McCall, M. T., 157,216 McCallum, K.J., 283,296 McCapra, F., 2, 10,124 McCarrou, E. M., 56, 123 McCarthy, W. J., 214, 219 McCartney, R. L., 14,119 McClelland, R. A., 211, 212,221 McDowell, C. S., 12, 124 McKillop, T. F. W., 2, 124 Mackor, E. L., 203, 207, 216 Maclean, C., 203, 207, 216 McVeigh, H., 188, 220 Magee, J. L., 224, 293 Magnusson, C., 11, 57, 60, 124 Mairanovskii, S. G., 2, 87, 1 2 8 Maloy, J. T., 85,118 Mango, F. D., 13, 56, 57, 117, 124 Mann, C. K.,2, 41, 43, 57, 64, 99, 110,119. 124, 126 Manning, G., 77, 78,124 Manousek, O., 12,124 Mansfield, J. R., 113,121 Marcoux, L. S., 43, 77, 79, 124 Mark, H. B., 108,124 Markby, R. E., 13, 67, 69, 70, 1 2 8 Markova, A. V., 99,128

306

AUTHOR INDEX

Markovi;, V., 234, 295 Martin, J., 227, 294 Marziano, N. C., 211, 219 Masetti, F., 143, 177, 179, 192, 203, 217 Mason, S. F., 139, 140, 150, 161, 170, 172, 177, 182, 198, 204, 207, 210, 217, 219 Masui, M., 57, 122, 124 Matheson, M. S., 227, 295 Matsuda, Y., 12, 57,124 Matsuoka, T., 13, 122 Matthews, D. B., 106,124 Mattson, J. S., 114, 120 Maugh, T., 43,128 Maurey, M., 9,118 Mavrogenes, G., 227, 296 Mayeda, E. A., 38, 39, 40, 43, 109, 124 Mayer, R. T., 165,219 Mazur, Z., 111, 117 Mazzucato, U., 143, 177, 179, 192, 203, 213,215,217 Meier, E. P., 91, 124 Meier, W., 172, 21 7 Meisel, D., 282, 295 Mels, S. J., 13, 117 Mentrup, A., 19,122 Meyerstein, D., 280, 293 Michael, B. D., 236, 244, 254, 261, 272, 278, 279, 282, 284, 288,291, 295 Michel, M.-A., 29, 108,124 Midid, 0. I., 234, 295 Millar, P. G., 93, 94, 121 Miller, C. L., 161, 193, 218, 220 Miller, L. L., 3, 9, 12, 16, 38, 39, 40, 43, 57,65,109,118, 120, 123, 124 Miller, W. N., 10, 81, 121 Millington, J. P., 57, 124 Mills, N. S., 21 3, 221 Mirkind, L. A., 14, 17, 44, 56, 57, 116, 120 Mishutushkina, I. P., 91,119 Misono, A., 13, 69,124 Mittal, J. P., 244, 295 Mockel, H., 292 Moe, N. S., 9, 116, 122 Mohilner, D. M., 13, 67, 69, 70, 128 Moodie, R. B., 36,122 Morita, M., 191, 213,218 Morita, T., 172, 220 Morkved, E. H., 59,126 Morris, M. D., 104, 126 Morton, R. A., 191, 214,219 Muck, D. L., 12, 92, 105,124 Mulliken, S. B., 57,126 Munro, I. H., 143,216 Murata, Y., 77, 79, 125, 127 Murrell, J. N., 189, 205, 219 Myer, J. A., 212, 220 Myers, L. S., Jr., 283, 297

N

Nadjo, L., 82, 123 Nagase, S., 2, 56, 125 Nagaura, S., 57, 123 Nakamaru, K., 144,219 Nasielski, J., 159, 170, 171, 172, 177, 203, 204,220 Neikam, W. C., 13, 108,119, 1 2 5 Nelson, D. A., 260, 261, 262, 264, 295 Nelson, R. F., 43, 125 Ness, N. M., 2, 119 Neta, P., 143, 219, 230, 234, 235, 237, 238, 244, 245, 246, 248, 249, 250, 251, 252, 256, 258, 259, 260, 263, 264, 265, 266, 269, 272, 273, 274, 276, 277, 281, 282, 283, 285, 286, 287, 288, 289, 292, 294, 295, 296, 297 Neumann, H., 12,122 Nicholson, R. S., 8, 1 2 3 Nielsen, S. O., 236, 296 Nikolic, A., 234, 293 Nilsson, A., 57, 125 Nilsson, G., 236, 296 Nilsson,S., 11, 13, 57, 63, 113,120, 125 Noguchi, I., 98, 117 Nordblom, C. D., 109,124 Norman, R. 0. C., 3,125 Norris, R. K., 3,127 Noyd, D. A., 111,116 Noyes, R. M., 154, 219 Nucifora, G., 229, 253, 278,296 Niirnberg, H. W., 91, 129 Nuzuma, S., 144,219 Nyberg, K., 2, 4, 10, 11, 12, 13, 16, 33, 34, 36, 38, 39, 43, 56, 57, 58, 60, 61, 64, 65, 81, 98, 117, 120, 124, 125, 126

0 O’Connor, J. J.. 75,125 O’Donnell, J. F., 43,125, 224, 296 Ofran, M., 146, 147,156,170,201,202,219 Ohmori, H., 57, 122, 124 Ohnesorge, W. E., 10,128 Ohno, A., 43,128 Okhlobystin, 0. Yu, 3, 111, 117 Olofsson, B., 11, 12, 32, 33, 36, 44, 57, 60, 64,120, 124 O’Malley, R. F., 56,123, 125 Omori, T., 287, 296 Ort, M. R., 82, 84,126 Osa,T., 13, 43, 48, 69, 124, 1 2 5 Oth, J. F. M., 111,125 Ottolenghi, M., 132, 144, 151, 154, 164, 215, 218, 219

307

AUTHOR INDEX Oude-Alink, B. A. M., 212, 213,217 Ovchinnikova, T. M., 35, 125 Owen, D. M., 17,118

P Pace, I., 167, 171, 201, 219 Packer, J. E., 278, 283, 296 Page, C. L., 113,127 Pagsberg, P., 227, 236, 296 Pakula, B., 137, 177, 205, 207, 217 Palluel, A. L. L., 2, 119 Papeschi, G., 91,117 Paquette, L. A., 75, 116 Park, E. H., 191, 213, 21 6 Pirkhyi, C., 108,125, 129 Parker, C. A., 132, 137, 138, 143, 219 Parker, V. D., 7, 9, 10, 12, 13, 16, 18, 25, 37, 50, 52, 56, 57, 71, 72, 73, 75, 76, 77, 78, 79, 81, 82, 86, 116, 118, 120, 121,122,124,125,126,127,128 Parsons, I. W., 56, 118 Partridge, L. K., 91,126 Passerini, R. C., 21 1, 219 Patterson, L. K, 226, 238, 244, 245, 256, 260, 263, 272, 273, 274, 283, 286, 288, 290,292, 293, 294, 295, 296 Paul, R,256,296 Paul, W. L., 161, 167, 171, 185, 194, 200, 21 9, 220 Payne, R., 87, 88,126 Pearl, J. A., 75,125 Pearson, R. G., 3,116 Pedler, A. E., 17, 118 Peet, N. P., 43,122 Peltier, D., 94, 122 Penton, J. R., 36,122 Peover, M. E.,. 2,. 7, 14, 47. 76, 107, 119, 12i, 126 Peradeiordi. F.. 207.21 6 Peridon, J.,.43,.126' Perlstein, J. H., 2,120 Pemn, C. L., 2,129 Perron, K. M., 283,292 Peter, F. A., 234, 238,296 Peter, H. H., 191,219 Petersen, R C., 11, 13, 43, 57, 60, 61, 114, 115,120,127 Peterson, M. L., 14, 17, 56, 123 Petit, G., 43, 126 Petkov, D., 207,220 Petkovit, Lj., 234, 293 Petrii, 0. A,, 6, 22,119 Petrovich, J. P., 2, 17, 82, 84, 116, 126 Petukhov, V. A, 188, 191, 212,216 Phelps, J., 10, 81, 116

Phillips, G. O., 224, 296 Phillips, J. M., 276, 280, 289, 290, 294 Philp, J., 182, 199, 207, 219 Pierre, G., 56, 118 Piersma, B. J., 2, 4, 9,126 Pistorius, R., 14, 17, 18, 56,127 Pletcher, D., 2, 9, 12, 41, 43, 44, 48, 50, 51, 56, 57, 64, 65, 99, 105, 117, 118, 119, 120, 121, 122, 123 Pointreau, R,76, 126 Ponce, C., 205, 216 Poole, J. A., 133, 216 Popp, F. D., 2,126 Popp, G., 2, 12, 56, 118, 126 Porter, A. S., 91, 126 Porter, G., 136, 141, 143, 154, 159, 161, 170, 177, 182, 200, 201, 205, 206, 208, 210,212,217, 218, 219, 220 Postle, M. J., 21 1, 21 9 Powell, J. S., 107, 126 W11, E. J., 83, 116 Pryor, W. A., 59,126,235,296 Puglisi, V. J., 85, 98, 126 Pullman, B., 2, 120 Pysh, E. S., 108, 109,126

R Rabani, J., 236, 291, 295, 296 Rabinowitch, E., 3, 126 Race, G. M., 48, 105,121 Radlick, P., 15, 126 Rahn, R. O., 184, 219 Rajbenbach, L., 59,117 Ralph, B., 113,127 Ramachandran, V., 12,124 Ramler, W., 227, 296 Rao, P. S., 261, 267, 282, 296 Rao, R. R., 57,126 Raoult, E., 94,122 Rapp, A., 19,122 Rasburn, E. J., 278, 283,296 Rawlinson, D. J., 60, 126 Rayner, D. M., 142, 146, 153, 187, 206, 211,219 Rebattu, J.-M., 43, 117 Reddy,A.K.N., 22,46, 87, 106,117 Reddy, T. B., 43, 56,129 Reed, R G., 29, 95, 108,121,127 Reichenbacher, P. H., 35, 104, 126, 128 Reid, C., 215, 21 9 Reid, D. H., 188, 219 Reilley, C. N., 9, 129 Reinmuth, W. H., 12,126 Reitz, N. C., 12, 126 Remko, R., 229, 253,278,292,296

308

AUTHOR INDEX

Rettschnick, R. P. H., 133, 21 7 Reutebuch, G., 238, 296 Reynolds, W. L., 3, 101,126 Rich, W. E., 111,127 Richtol, H. H., 57, 118 Ridgway, T. H., 111,127 Rieger, P. H., 12, 106,116, 126 Rieke, R. D., 111,126, 127 Rietta, M. S., 177, 220 Rietz, B., 13, 63, 117, 120 Rifi, M. R., 2, 15, 105,127 Roberts, B. P., 59,121 Roberts, T. D., 213, 221 Robertson, P. M., 2,127 Robinson, S. L., 57, 126 Rochester, C. H,, 21 0,219 Rockett, B. W., 157, 21 7 Roder, H., 14, 97, 122 Roe, D. K., 43,128, 213,215 Rogers, H. R,15,118 Rogers, L. B., 137, 161, 164, 165, 167, 170, 192,198,217, 221 Ronlin, A., 9, 10, 16, 57, 79, 81, 86, 116, 125, 127, 128 Rosebrook,D. D., 144, 170, 177, 216, 219 Rosenberg, J. L., 167, 170, 201, 202, 207, 219 Ross, A. B., 233, 234, 236, 292, 293 Ross, S. D., 11, 13, 43, 56, 57, 60, 61, 64, 113,114,115,120,127 Rotinyan, A. L., 35,125 Rotkiewicz, K., 171, 207,217, 219 Rousseau, K., 43, 44, 1 2 7 Rudd, E. J., 57, 60, 64, 127 Russell, G. A., 3, 127 Russell, P. J., 3, 123 Rutledge, J. M., 191, 220 Ryde-Petterson, G., 105, 120

S Sabol, M. A., 13,117 Sadlej, A. J., 171, 207, 21 7 Saeki, T., 56,129 Sainsbury, M., 75,127 Salmin, L. A., 14, 17, 56, 120 Salzberg, H. W., 57,123 Samuelsson, B., 12, 123 Samuni, A., 249, 251, 259, 280, 285, 292, 296 Sanders, L. B., 159, 161, 164, 182, 192, 21 3,220,221 Sandman, D. J., 3,118 Sandorfy, C., 207,219 Sangster, D. F., 224, 296 Santhanam, K. S. V., 76,127

Sargent, E. P., 136, 215 Sato, N., 113, 127 Saveant, J. M., 82,123 Saygin, O., 17, 57,123 Sayo, H., 57,122 Schafer, H., 2, 13, 14, 16, 17, 18, 28, 40, 56, 57,120,127 Schanzer, W., 33,118 Scheffler, M., 228,294 Schmidt, K., 227, 294 Schmidt, K. H., 228,296 Schnabel, W., 224, 294 Schofield, K., 36, 122 Scholes, G., 276, 280, 289, 290, 294 Schoneshofer, M., 243, 296 Schooten, J., van, 76,122 Schroder, G., 111,125 Schubert, C. C., 113,127 Schuler, M. A., 226, 270, 275, 286, 288, 296 Schuler, R. H., 226, 229, 231, 235, 237, 238, 244, 246, 247, 248, 249, 250, 251, 252, 254, 258, 272, 275, 278, 280, 283, 285, 286, 287, 288, 292, 293, 294, 295, 296, 297 Schulman, S., 182,183, 207,217 Schulman, S. G., 132, 138, 157, 159, 161, 164, 167, 170, 171, 177, 182, 183, 185, 191, 192, 194, 195, 196, 197, 198, 200, 201, 207, 210, 213, 214,215, 216, 218, 219, 220, 221 Schulte-Frohlinde, D., 238, 283, 286, 287, 293, 296 Schultz, H. P., 2, 126 Schwartz, J., 204,216 Scott, G. W., 206, 218 Scott, M. K., 12, 116 Sease, J. W., 29, 108,127 Sehested, K., 244, 272, 284,293 Seiler, P., 172, 21 7 Sekine, T., 113,127 Seliger, H. It, 191, 214, 219 Selinger, B. E., 147,218 Seliskar, C. J., 170, 196, 220 Sellers, R. M., 256, 258, 293 Semenov, N. N., 71,127 Seno, M., 13, 67, 69, 116 Serve, D., 36, 43, 64, 118 Shain, I., 8, 125 Shank, C. V., 212,220 Sharp, J. H., 107,116 Shekhvatov, M. S., 35,122 Sheldon, R. A., 3, 60,127 Shephard, B. R., 60,123 Shih, S., 3, 119 Shimuzi, T., 57,123 Shine, H. J., 3, 71, 77, 78, 79, 116, 125, 127

AUTHOR INDEX Sliizuka, X., 172, 220 Shono,T., 11, 13, 111,127, 1 2 8 Shragge, P. C., 238, 251, 256, 263, 269, 290, 294 Shukla, P. R., 195, 220 Shukla, S. N., 33,128 Shulman, R. G., 184, 21 9 Sicher, J., 15, 111,129 Siebrand, W., 133, 134, 158, 21 7 Simic, M., 143, 220, 237, 244, 245, 251, 253, 256, 258, 259, 260, 265, 266, 269, 272, 273, 285, 288, 289, 294, 295, 296 Simon, J. M., 43, 118 Simonet, J., 12, 29, 108,123, 1 2 4 Sims, J. J., 15, 126 Sioda, R. E., 57, 79, 128 Skaletz, D., 93, 94,122 Skell, P. S., 35, 104, 123, 126, 1 2 8 Smaller, B., 229, 253, 278, 292, 296 Smirnov, V. A., 2, 8 7 , 9 9 , 1 2 8 Smith, B. E., 139, 140, 150, 161, 170, 172, 177, 182, 199, 204, 207, 210, 217, 219 Smith, D. E., 76, 107,122 Smith, H. O., 52,123 Smith, W. B., 14, 17, 56,128 Sokol’skii, D. V., 65, 1 2 8 Solly, R. K., 59,128 Sommer, U., 207,220 Song, P A . , 196, 207, 220 Sorensen, V. G., 283, 296 Sosnovsky, G., 60,126 Southwick, P. L., 286,295 Southwick, E. W., 2, 1 2 9 Soylemez, T., 244, 272, 275, 297 Spurlock, S., 15, 126 Stamp, A, 192, 215 Stanley, J. P., 235, 296 Staples, T. L., 85,128 Steckhan, E., 16, 17, 57,127 Stefani, A. P., 31, 128 Steinberg, M., 30, 1 2 3 Stenberg, V. I., 205, 220 Stermitz, F. R., 16, 120 Stern, G., 151, 154, 218 Sternberg, H. W., 13, 67, 69, 7 0 , 1 2 8 Sternerup, H., 4, 33, 57, 96, 97, 119, 120, 128 Stevens, G. C., 244, 272, 278, 290, 294, 297 Stevens, R. D. S., 191, 202, 216 Stickler, S. J., 143, 220 Stocker, J. H., 98,128 Stockhausen, K., 244, 258, 272, 278, 297 Stryer, L,170, 198, 201, 202, 220 Stuart, J. D., 10,128 Suarez, C., 239, 286, 293 Subba Rao, G., 13, 67,117

309

Sugimori, A., 287, 296 Sugino, K., 113, 1 2 7 Sujdak, R. J., 287,295 Sullivan, P. J., 195, 220 Sun, M., 1 9 6 , 2 2 0 Sundheim, B. R., 201, 202, 220 Sundholm, F., 164,218 Suppan, P., 21 2, 21 7 Susuki, T., 11, 27, 56, 57, 123 Sutton, J. R., 13, 117 Svanholm, U., 9, 10, 16, 56, 57, 78, 79, 81, 116, 1 2 8 Svensmark Jensen, B., 9, 76, 1 2 8 Swain, C. G., 43, 1 2 8 Swallow, A. J., 224, 243, 277, 281, 288, 293, 295, 297 Swann, S., 2, 111, 1 2 8 Swinkels, D. A. J., 2 3 , 9 1, 117 Szwarc, M., 3, 30, 66, 71, 74, 85, 123, 1 2 8

T Taft, R. W., 192, 211, 220 Tamura, H., 1 2 , 5 7 , 1 2 4 Tang, R. T., 3, 60,123 Tansley, A. C., 91, 126 Taran, L. A., 35, 1 2 5 Tardivel, R., 12, 123 Tatlow, J. C., 17, 56, 118 Taub, I. A., 244, 278,292, 293 Taylor, W. B., 226, 227, 291, 292 Teply, J., 245, 294 Terni, H. A., 93, 94, 1 2 3 Testa, A. C., 177, 213, 216, 218, 2 2 4 221 Teuerstein, A., 277, 295 Texier, P., 43, 49,116 Thielen, D. R., 75, 1 2 8 Thiry, P., 159, 172, 220 Thomas, C. B., 3 , 1 2 5 Thomas, H. G., 1 2 , 9 2 , 1 2 8 Thomas, J. K., 278, 280, 290, 294, 297 Thompson, G. F., 278, 292 Thorn, E., 283, 296 Tichy, M., 188, 219 Tidwell, P. T., 138, 167, 177, 221 Tincher, C. A., 13, 117 Titus, J. A., 44,128 Tomilov, A. P., 2, 36, 56, 87, 1 2 8 Torosian, G., 188, 19 1, 220 Trahanovsky, W. S., 3, 1 2 9 Tramer, A., 161, 220 Trieff, N. M., 201, 202,220 Trinajstic, N., 205, 216 Trojanek, A., 57, 1 2 5 Trotter, W., 213, 220

310

AUTHOR INDEX

Trozzolo, A. M., 212, 220 Tsutsumi, K., 172,220 Tsutsumi, S., 11, 13, 27, 56, 57, 122, 123, 129 Turner, W. R., 43,129 Tussey, B., 182, 220 Tyutyulkov, N., 207,220

U

Udenfriend, S., 138, 214, 220 Uebel, J. J., 13, 56,120, 127 Umberger, J. Q., 157, 158, 220 Urban, W., 170, 172, 210, 221 Utley, J. H. P., 2, 18,93,97,118, 122, 129 Uzhinov, B. M., 203, 218

V Valenta, P., 91,129 Vance, C. J., 99,121 VanderDonckt,E., 132, 136, 159, 161, 167, 170, 171, 172, 177, 192, 195, 200, 201, 203,204, 205,210,220 Van Duyne, R. P., 9 , 1 2 9 van Hemmen, J. J., 279,291 van Tarnelen, E. E., 15, 29, 126, 129, 204, 216 Vellturo, A. F., 15, 129 Verma, N. C., 229, 252, 297 Vielstich, W., 2, 4, 129 Vijh, A. K., 2, 16, 98, 104, 112,118 Vogels, C., 172, 177, 220 Volke, J., 12, 129 von Sonntag, C., 238, 283, 287,296

W

Wagenknecht, J. H., 12, 15, 28,129 Wahl, P., 191, 202, 21 7 Wai, H., 21 1, 221 Walatka, V., 2,120 Walker, 0. J., 33, 128 Wallace, S. C., 290, 297 Wander, R., 230, 272, 273,297 Ward, J. F., 283, 297 Ward, G. A., 56,129 Ward, P. J., 3, 125 Ware, W. R., 195,220 Watanabe, H., 1 3 , l 1 7 Watkins, A. R., 140, 151, 157, 159, 170, 198, 202,220, 221 Wawzonek, S., 2, 12, 14, 26, 28,129 Wearring, D., 14, 129

Weber, K., 132, 221 Weedon, B. C. L., 16, 26, 60, 93,118, 123, 129 Wehry, E. L., 137, 161, 164, 167, 170, 192, 193,198,221 Weinberg, H. R., 2, 27,129 Weinberg, N. L., 2, 27, 43, 56, 57, 129 Weingarten, H., 10, 53, 81, 121 Weisstuch, A., 177, 221 Welch, G., 21 1, 221 Weller, A., 132, 137, 144, 145, 146, 148, 149, 154, 155, 157, 158, 159, 170, 172, 177, 184, 193, 201, 202, 208, 210,221 Wemer, T. C., 171,201,221 Wendenburg, J., 224,294 Wender, I., 13, 67, 69, 70, 128 Wendt, H., 17, 30, 57,123, 129 Wertz, J. E., 247, 297 West, M. A., 141, 219 West, R., 64, 121 Westberg, M. H., 15,129 Wethermann, D. P., 143, 147, 213,221 White, B. S., 7, 126 White, C. E., 137, 138, 214, 215 Whitesides, T., 15,126 Wiberg, K. B., 15, 129, 267, 297 Wierenga, W., 37, 119 Wiesgraber, K. H., 98, 117 Wigger, A., 244, 258, 263, 266, 268, 272, 275, 278, 281,292, 297 Wilkinson, R G., 99, 120 Will, J. P., 57, 117 Williams, D. L., 214, 221 Williams, G. H., 58, 60, 64,129 Williams, J. H., 14, 119 Williams, J. L. R., 177, 179, 21 7 Williams, K. T., 177, 216 Williams, R. T., 161, 21 6 Willson,J. S., 3,125 Willson, R. L., 244, 254, 259, 261, 268, 269, 278,279,282,290,291,294,297 Wilshire, J. F. K., 57,129 Wilson, A.M., 111,129 Wilson, E. R, 12, 92, 105, 124 Wilson, G. S., 10, 81, 121 Wilson, J. D., 10, 53, 81,121 Wilson, R. K., 203,217 Winefordner, J. D., 132, 138, 159, 161, 164, 167, 177, 182, 183, 184, 192, 193, 210, 213,214,215, 216, 219, 220, 221 Winkler, H., 19, 122 Winograd, N., 113,123 Wolfrom, M. L., 62,118 Wolff, R. K., 226, 227, 291, 292 Wollner, G. P., 278, 280, 292 Woodhall, B. J., 36,43,129 Woods, R. J., 283,296 Wooldridge, J., 21 3, 221

31 1

AUTHOR INDEX Wright, C. M., 56,129 Wubbels, G. G., 213, 221 Wudl, F., 2,129 Wyatt, P. A. H., 136, 138, 141, 142, 146, 153, 157, 158, 159, 161, 170, 171, 172, 187, 188, 193, 205, 206,210, 211, 213, 21 7, 21 8, 219, 221 Wynne-Jones, W. F. K., 33, 113,119, 121

Y Yagi, H., 98, 117 Yagupol’skii, L. M., 192, 221 Yakatan, G. J., 159,170, 191,193,200,214, 221 Yamagishi, F. G., 3,118 Yamagishi, T., 13, 69,124 Yang, N. C., 108, 109,126 Yates, G. B., 18, 97, 122, 129

Yates, Ic, 211, 212, 221 Yildiz, A., 43, 48, 125 Yoshida, K., 56, 129 Young, J. F., 161, 171, 191, 194, 200,220, 221

L

Zahnow, E. W., 11,43, 56, 57,116 Zahradnik, R., 108, 125, 129, 188, 219, 268,293 Zalar, F. V., 92,121 Zalis, B., 182, 221 Zavada, J., 15, 111, 129 Zemel, H., 276, 297 Zhurinov, M. Zh., 44,120 Zimbrick, J. D., 283, 297 Zimmerman, H. E., 143, 147, 213, 221 Zorn, H., 238, 286,293 Zuman, P., 2, 12, 50, 106,124, 129

This Page Intentionally Left Blank

Cumulative Index to Authors

Anbar, M., 7, 115 Bell, R. P., 4, 1 Bennett, J. E., 8 , 1 Bentley, T. W., 8 , 151 Bethell, D., 7, 153; 10, 53 Brand, J. C. D., 1, 365 Brinkman, M. R., 10, 53 Brown, H. C., 1, 35 Cabell-Whiting, P. W., 10, 129 Cacace, F., 8 , 79 Carter, R. E., 10, 1 Collins, C. J., 2, 1 Cornelisse, J., 11, 225 Crampton, M. R., 7, 211 de Gunst, G. P., 11, 225 Eberson, L., 12, 1 Farnum, D. G., 11, 123 Fendler, E. J., 8 , 271 Fendler, J. H., 8 , 271 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, 5 3 Gillespie, R. J., 9, 1 Gold, V., 7,259 Greenwood, H. H., 4, 73 Havinga, E., 11, 225 Hogeveen, H., 10, 29, 129 Ireland, J. F., 12, 131 Johnson, S. L., 5, 237 Johnstone, R. A. W., 8 , 151 Kohnstam, G., 5 , 121 Kramer, G. M., 11, 177 Kreevoy, M. M., 6, 63 Liler, M., 11, 267 Long, F. A., 1, 1 Maccoll, A., 3 , 9 1 McWeeny, R., 4 , 7 3 Melander, L., 10, 1 Mile, B., 8 , 1

314 Miller, S. I., 6, 185 Modena, G., 9, 185 More O’Ferrall, R. A., 5, 331 Neta, P., 12, 223 Norman, R. 0. C., 5 , 5 3 Nyberg, K., 12, 1 Olah, G. A., 4 , 305 Parker, A. J., 5, 173 Peel, T. E., 9, 1 Perkampus, H. H., 4, 195 Pittmann, C. U., Jr., 4, 305 Pletcher, D., 10, 155 Ramirez, F., 9, 25 Rappoport, Z., 7 , l Reeves, L. W., 3, 187 Robertson, J. M., 1, 203 Samuel, D., 3, 123 Schaleger, L. L., 1, 1 Scheraga, H. A., 6, 103 Shatenshtein, A. I., 1, 156 Silver, B. L., 3, 123 Simonyi, M., 9, 127 Stock, L. M., 1, 35 Symons, M. C. R., 1, 284 Thomas, A., 8, 1 Tonellato, U., 9, 185 Tiidos, F., 9, 127 Turner, D. W., 4, 31 Ugi, I., 9, 25 Ward, B., 8, 1 Whalley, E., 2, 9 3 Williams, J. M., Jr., 6, 63 Williamson, D. G., 1, 365 Wolf, A. P., 2, 201 Wyatt, P. A. H., 12, 131 Zollinger, H., 2, 163 Zuman, P., 5, 1

INDEX

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 properties of electronically excited states of organic molecules, 12, 131 Acids, reactions of aliphatic diazo compounds with, 5, 331 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 Aliphatic diazo compounds, reactions with acids, 5, 331 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 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 Arynes, mechanisms of formation and reactions at high temperatures, 6, 1 A - S E ~reactions, 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 Carbene chemistry, structure and mechanism in, 7, 163 Carbon atoms, energetic, reactions with organic compounds, 3, 201 Carbon monoxide, reactivity of carbonium ions towards, 10, 29 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 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, 27 1 Cations, vinyl, 9, 185 Charge density-N.M.R. chemical shift correlations in organic ions, 11, 125 Chemically induced dynamic nuclear spin polarization and its applications, 10, 53 CIDNP and its applications, 10, 53 Conformations of polypeptides, calculations of, 6, 103 Conjugated molecules, reactivity indices in, 4, 73 Diazo compounds, aliphatic reactions with acids, 5, 331

316

CUMULATIVE INDEX

Dipolar aprotic and protic solvents, rates of bimolecular substitution reactions in, 5, 173 Directive effects in aromatic substitution, a quantitative treatment of, 1, 35 Electrochemistry, organic, structure and mechanism in, 12, 1 Electrode processes, physical parameters for the control of, 10, 155 Electron spin resonance, identification of organic free radicals by, 1, 284 Electron spin resonance studies of short-lived organic radicals, 5, 23 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 Entropies of activation and mechanisms of reactions in solution, 1, 1 Enzymatic catalysis, physical organic model systems and the problem of, 11, 1 Equilibrium constants, N.M.R. measurements of, as a function of temperatures, 3, 187 Ester hydrolysis, general base and nucleophilic catalysis, 5, 237 Exchange reactions, hydrogen isotope, of organic compounds in liquid ammonia, 1, 156 Exchange reactions, oxygen isotope, of organic compounds, 3, 123 Excited molecules, structure of electronically, 1, 365 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 pyrolysis of small-ring hydrocarbons, 4, 147 General base and nucleophilic catalysis of ester hydrolysis and related reactions, 5 , 237 H20-D20 Mixtures, protolytic processes in, 7, 259 Heat capacities of activation and their uses in mechanistic studies, 5, 121 Heterolysis, gas-phase, 3,91 Hydrated electrons, reactions of, with organic compounds, 7, 115 Hydration, reversible, of carbonyl compounds, 4, 1 Hydrocarbons, small-ring, gas-phase pyrolysis of, 4, 147 Hydrogen atom abstraction from 0-H bonds, 9, 127 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 Ionization potentials, 4, 3 1 Ions, organic, charge density-N.M.R. chemical shift correlations, 11, 125 Isomerization, permutational, of pentavalent phosphorus compounds, 9,25 Isotope effects, steric, experiments on the nature of, 10, 1 Isotope effects, hydrogen, in aromatic substitution reactions, 2, 163 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, reaction, polarography and, 5 , 1 Mass spectrometry, mechanism and structure in: a comparison with other chemical processes, 8, 152 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, organic reaction, isotopes and, 2, 1 Mechanisms, reaction, use of volumes of activation for determining, 2 , 9 3 Mechanisms of formation and reactions of arynes at high temperatures, 6, 1 Mechanisms of reactions in solution, entropies of activation and, 1, 1 Mechanistic studies, heat capacities of activation and their use in, 5, 121

CUMULATIVE INDEX

317

Meisenheimer complexes, 7, 21 1 Micellar catalysis in organic reactions: kinetic and mechanistic implications, 8, 27 1 N.M.R. chemical shift-charge density correlations, 11, 125 N.M.R. measurements of reaction velocities and equilibrium constants as a function of temperature, 3, 187 Non-planar and planar aromatic systems, 1, 203 Norbomyl cation: reappraisal of structure, 11, 179 Nuclear magnetic resonance, see N.M.R. Nucleophilic aromatic photosubstitution, 11, 225 Nucleophilic catalysis of hydrolysis and related reactions, 4, 237 Nucleophilic vinylic substitution, 7, 1 0-H bonds, hydrogen atom abstraction from, 9, 127 Oxygen isotope exchange reactions of organic compounds, 3, 123 Permutational isomerization of pentavalent phosphorus compounds, 9 , 25 Phosphorus compounds, pentavalent, turnstile rearrangement and pseudorotation in permutational isomerization, 9 , 25 Photochemistry of carbonium ions, 10, 129 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 Protic and dipolar aprotic solvents, rates of bimolecular substitution reactions in, 5, 173 F’rotolytic processes in H,O-D,O mixtures, 7, 259 Protonation sites in ambident conjugated systems, 11, 267 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 Radicals, organic, application of radiation techniques, 12, 223 Radicals, organic free, identification by electron spin resonance, 1, 284 Radicals, short-lived organic, electron spin resonance studies of, 5, 53 Reaction kinetics, polarography and, 5, 1 Reactions mechanisms, use of volumes of activation for determining, 2, 9 3 Reaction mechanisms in solution, entropies of activation and, 1, 1 Reaction velocities and equilibrium constants, N.M.R. measurements of, as a function of temperature, 3, 187 Reactions of hydrated electrons with organic compounds, 7, 1 1 5 Reactivity indices in conjugated molecules, 4, 73 Refractivity, molecular, and polarizability, 3, 1 Resonance, electron-spin, identification of organic free radicals by, 1, 284 Resonance, electron-spin, studies of short-lived organic radicals, 5, 63 Short-lived organic radicals, electron-spin resonance studies of, 5, 5 3 Small-ring hydrocarbons, gas-phase pyrolysis of, 4, 147 Solution, reactions in, entropies of activation and mechanisms, 1, 1 Solvents, protic and dipolar aprotic, rates of bimolecular substitution reactions in, 5, 173 Spectroscopic observation of alkylcarbonium ions in strong acid solutions, 4, 305 Steric isotope effects, experiments on the nature of, 10, 1 Stereoselection in elementary steps of organic reactions, 6, 185 Structure and mechanism in carbene chemistry, 7, 153 Structure and mechanism in organic electrochemistry, 12, 1 Structure of electronically excited molecules, 1, 365 Substitution, aromatic, a quantitative treatment of directive effects in, 1, 35 Substitution reactions, bimolecular, in protic and dipolar aprotic solvents, 5, 173

318

CUMULATIVE INDEX

Substitution reactions, aromatic, hydrogen isotope effects in, 2, 163 Superacid systems, 9, 1 Temperature, N.M.R. measurements of reaction velocities and equilibrium constants as a function of, 3, 187 Tritiated molecules, gaseous carbonium ions from the decay of, 8, 79 Tritium atoms, energetic, reactions with organic compounds, 2, 20 1 Turnstile rearrangement in isomerization of pentavalent phosphorus compounds, 9, 25 Unsaturated compounds, basicity of, 4, 195 Vinyl cations, 9, 185 Volumes of activation, use of, for determining reaction mechanisms, 2, 93