Está en la página 1de 633

COMPREHENSIVE CHEMICAL KINETICS

COMPREHENSIVE
Section 1. THE PRACTICE AND THEORY OF KINETICS
Volume 1 The Practice of Kinetics
Volume 2 The Theory of Kinetics
Volume 3 The Formation and Decay of Excited Species
Section 2. HOMOGENEOUS DECOMPOSITIONAND ISOMERISATIONREACTIONS

Volume 4 Decomposition of Inorganic and Organometallic Compounds


Volume 5 Decomposition and Tsomerisation of Organic Compounds
Section 3.

INORGANIC REACTIONS

Volume 6 Reactions of Non-metallic Inorganic Compounds


Volume 7 Reactions of Metallic Salts and Complexes, and Organometallic Compounds
Section 4. ORGANIC REACTIONS (6 volumes)
Volume 9 Addition and Elimination Reactions of Aliphatic Compounds
Volume I0 Ester Formation and Hydrolysis and Related Reactions
Volume 13 Reactions of Aromatic Compounds
Section 5.

POLYMERISATION REACTIONS (2

Section 6.

OXIDATION AND COMBUSTION REACTIONS

volumes)

(2 volumes)

Section 7. SELECTED ELEMENTARY REACTIONS ( 2 volumes)

Additional Sections
HETEROGENEOUS REACTIONS
SOLID STATE REACTIONS
KINETICS AND TECHNOLOGICAL PROCESSES

CHEMICAL KINETICS
EDITED BY

C. H. BAMFORD
M.A., Ph.D., Sc.D. (Cantab.), F.R.I.C., F.R.S.
Campbell-Brown Professor of Industrial Chemistry,
Unioersity of Liverpool

AND

C. F. H. TIPPER
Ph.D. (Bristol), D.Sc. (Edinburgh)
Senior Lecturer in Physical Chemistry,
Uniuersity of Lioerpool

VOLUME

REACTIONS OF METALLIC SALTS AND COMPLEXES,


AND ORGANOMETALLIC COMPOUNDS

ELSEVIER P U B L I S H I N G C O M P A N Y
AMSTERDAM

- L O N D O N - NEW Y O R K
1972

ELSEVIER PUBLISHING COMPANY

336 I A N V A N G A L E N S T R A A T
P.O. B O X 211. A M S T E R D A M , T H E N E T H E R L A N D S
A M E R I C A N ELSEVIER P U B L I S H I N G C O M P A N Y , I N C .
52 V A N D E R B I L T A V E N U E
N E W Y O R K , N E W Y O R K 10017

L I B R A R Y OF CONGRESS CARD N U M B E R
ISBN 0-444-40861

WITH

COPYRIGHT

@ 1972

76-151731

35 I L L U S T R A T I O N S A N D 228 T A B L E S
BY ELSEVIER PUBLISHING COMPANY, AMSTERDAM

A L L R I G H T S RESERVED

N O P A R T OF T H I S P U B L I C A T I O N M A Y BE R E P R O D U C E D ,
S T O R E D I N A R E T R I E V A L SYSTEM, O R T R A N S M I T T E D I N A N Y F O R M O R BY A N Y M E A N S
ELECTRONIC, MECHANICAL, PHOTOCOPYING, RECORDING, O R OTHERWISE,
W I T H O U T THE PRIOR WRITTEN PERMISSION O F THE PUBLISHER,
E L S E V I E R P U B L I S H I N G C O M P A N Y , J A N V A N G A L E N S T R A A T 335, A M S T E R D A M

PRINTED I N THE NETHERLANDS

C O M P R E H E N S I V E C H E M I C A L KINETICS

ADVISORY BOARD

Professor S. W. BENSON
Professor SIR FREDERICK DAINTON
Professor G. GEE
the late Professor P. GOLDFINGER
Professor G. s. HAMMOND
Professor W. JOST
Professor G. B. KISTIAKOWSKY
Professor v. N. KONDRATIEV
Professor K. J. LAIDLER
Professor M. MAGAT
Professor SIR HARRY MELVILLE
Professor G. NATTA
Professor R. G. W. NORRISH
Professor s. OKAMURA
Professor SIR ERIC RIDEAL
Professor N. N. SEMENOV
Professor z. G. SZABO
Professor 0. WICHTERLE

Contributors to Volume 7

D. BENSON

Department of Chemistry,
Widnes Technical College,
Widnes, Lancs., England

L. J. C S ~ N Y I

Institute of Inorganic and Analytical Chcniistry,


J6szef Attila University,
Szeged, Hungary

T. J. KEMP

School of Molecular Sciences,


University of Warwick,
Coventry, England

C. H. LANGFORD Department of Chemistry,


Carleton University,
Ottawa, Canada

M. PARRIS

Department of Chemistry,
Carleton University,
Ottawa, Canada

P. J. PROLL

Department of Chemistry,
Widnes Technical College,
Widnes, Lancs., England

Preface

The rates of chemical processes and their variation with conditions have been
studied for many years, usually for the purpose of determining reaction mechanisms. Thus, the subject of chemical kinetics is a very extensive and important
part of chemistry as a whole, and has acquired an enormous literature. Despite
the number of books and reviews, in many cases it is by no means easy to find
the required information on specific reactions or types of reaction or on more
general topics in the field. It is the purpose of this series to provide a background
reference work, which will enable such information to be obtained either directly,
or from the original papers or reviews quoted.
The aim is to cover, in a reasonably critical way, the practice and theory of
kinetics and the kinetics of inorganic and organic reactions in gaseous and condensed phases and at interfaces (excluding biochemical and electrochemical kinetics, however, unless very relevant) in more or less detail. The series will be divided
into sections covering a relatively wide field; a section will consist of one or more
volumes, each containing a number of articles written by experts in the various
topics. Mechanisms will be thoroughly discussed and relevant non-kinetic data
will be mentioned in this context. The methods of approach to the various topics
will, of necessity, vary somewhat depending on the subject and the author(s) concerned.
It is obviously impossible to classify chemical reactions in a completely logical
manner, and the editors have in general based their classification on types of chemical element, compound or reaction rather than on mechanisms, since views on
the latter are subject to change. Some duplication is inevitable, but it is felt that
this can be a help rather than a hindrance.
Section 3 deals with reactions in which at least one of the reactants is an inorganic compound. Many of the processes considered also involve organic compounds, but autocatalytic oxidations and flames, polymerisation and reactions of
metals themselves and of certain unstable ionic species, e.g. the solvated electron,
are discussed in later sections. Where appropriate, the effects of low and high energy radiation are considered, as are gas and condensed phase systems but not fully
heterogeneous processes or solid reactions. Rate parameters of individual elementary steps, as well as of overall reactions, are given if available.
In volume 7 reactions of metallic salts, complexes and organometallic compounds are covered. Isomerisation and group transfer reactions of inert metal
complexes and certain organometallics (not involving a change in oxidation state)
are considered first, followed by oxidation-reduction processes (a) between different valency states of the same metallic element (b) between salts of different

VIII

PREFACE

metals and (c) involving covalent inorganic or organic compounds. Finally,


induced reactions are discussed separately.
The Editors desire to record their sincere appreciation of the continuing advice
and support from the members of the Advisory Board.
Liverpool
October, 1971

C. H. BAMFORD
C. F. H. TIPPER

Contents
Preface

................................

VI i

Chapter I (C H . LANGFORD
A N D M PARRIS)

Reactions of inert complexes and metal organic compounds . . . . . . . . .


1. INTRODUCTION

...............................

. STOICHIOMETRIC MECHANISMS . . . . . . . . . . . . . . . . . . . . . . . . .
3 . REACTIONS OF CO(II1) COMPLEXES . . . . . . . . . . . . . . . . . . . . . . .

4 . Cr(ll1). Rh(ll1).

3
7

. . . . . . . . . . . . . .
5 . COMPLEXES OF Pt(1I). Pd(11). AU(II1) AND Rh(1)
. . . . . . . . . . . . . . . . .
6. COMPLEXES WITH B CLASS LIGANDS: THE BINARY CARBONYLS . . . . . . . . . . . .
7. COMPLEXES WITH B CLASS LIGANDS: THE SUBSTITUTED CARBONYLS . . . . . . . . . .
8. THEORETICAL CONSIDERATIONS . . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
RU(II1). I T ( I I 1 ) AND

Pt(1V) COMPLEXES

17

20
25
31
43
52

Chapter 2 (P. J . PROLL)

Reactions in solution between various metal ions of the same element in different oxidation states . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
2 . COPPER.

...............................
SILVER AND GOLD . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1 The exchange reaction between Cu(II) and Cu(1) . . . . . . . . . . . . .
2.2 The exchange reaction between Ag(I1) and Ag(1) . . . . . . . . . . . . .
2.3 The exchange reaction between Au(II1) and Au(1) . . . . . . . . . . . .
2.4 The exchange reaction between Au(II1) and Au(I1) . . . . . . . . . . . .
2.5 The disproportionation of Au(1I) . . . . . . . . . . . . . . . . . . . . .
3. MERCURY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

56

1 INTRODUCTION

.
.
.
.

3.1 The exchange reaction between Hg(l1) and Hg(l); the disproportionation reaction of Hg(I1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 The exchange reaction between Hg(I1) and Hg(1) in non-aqueous media . . . .
4

. THALLIUM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .
5 . T I N AND LEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1 The exchange reaction between Sn(IV) and Sn(I1) in aqueous media . . . . .
4.1 The exchange reaction between TI(III) and TI(1)

5.2 The exchange reaction between Sn(IV) and Sn(I1) in non-aqueous media
5.3 The exchange reaction between Pb(lV) and Pb(I1) in aqueous media . .
5.4 The exchange reaction between Pb(IV) and Pb(1l) in non-aqueous media

56

. . .

. . .
. . .

58
58

58
59
60
60
60
60
62
62
62
68
68
70
70
71

CONTENTS

6.

ARSENIC A N D ANTIMONY

. . . . . . . . . . . . . . . . . . . . . . . . . . .

6.1 The exchange reaction between As(V) and As(II1) . . . . . . . . . . . . .


6.2 The exchange reaction between Sb(V) and Sb(II1) in aqueous media . . . . .
6.3 The exchange reaction between Sb(V) and Sb(II1) in non-aqueous media . . .
6.4 The Sb(II1)-catalysed hydrolysis of Sb(V) . . . . . . . . . . . . . . . . .
7.

. . . . . . . . . . . . . . . . . . . . . .
.
8 . TRANSITION METALS . . . . . . . . . . . . . . . . . . .
8.1 Vanadium and tantalum . . . . . . . . . . . . . .
TELLURIUM

7.1

The exchange reaction between Te(V1) and Te(1V)

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

. . .
. . . .
. . .
. . .

8.1.1 The exchange reaction between V(II1) and V(I1) . . . . . . . . . . .


8.1.2 The exchange reaction between V(1V) and V(11I) . . . . . . . . . .
8.1.3 The exchange reaction between V(V) and V(1V) . . . . . . . . . . .
8.1.4 Reactions between vanadium ions . . . . . . . . . . . . . . . . .
8.1.5 Reactions between tantalum cluster ions . . . . . . . . . . . . . .
8.2 Chromium, molybdenum and tungsten . . . . . . . . . . . . . . . . . .
8.2.1 The exchange reaction between Cr(II1) and Cr(1l) . . . . . . . . . .
8.2.2 The exchange reaction between Cr(V1) and Cr(Il1) . . . . . . . . . .
8.2.3 The reaction between Cr(V1) and Cr(I1) . . . . . . . . . . . . . .
8.2.4 Cr(1l)-catalysed substitution and isomerisation reactions of Cr(I1l) . . .
8.2.5 The exchange reaction between Mo(V) and Mo(1V) . . . . . . . . .
8.2.6 The exchange reaction between W(V) and W(IV) . . . . . . . . . .
8.3 Manganese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.3. I The exchange reaction between Mn(1l) and Mn(1) . . . . . . . . . .
8.3.2 The exchange reaction between Mn(lI1) and Mn(I1)
. . . . . . . .
8.3.3 The exchange reaction between Mn(VI1) and Mn(V1) . . . . . . . .
8.3.4 The exchange reaction between Mn(VI1) and Mn(Il1) . . . . . . . .
8.3.5 The exchange reaction between Mn(VI1) and Mn(1l) . . . . . . . .
8.3.6 The reaction of Mn(VI1) and Mn(I1) . . . . . . . . . . . . . . . .
8.4 Iron. ruthenium and osmium . . . . . . . . . . . . . . . . . . . . . .
8.4. I The exchange reaction between Fe(II1) and Fe(I1) in aqueous media . .
8.4.2 The effect of inorganic ions on the exchange reaction between Fe(Il1)
and Fe(I1) . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.4.3 The effect of organic ligands o n the exchange reaction between Fe(I11)
and Fe(1l) . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.4.4 The exchange reaction between Fe(1Il) and Fe(I1) in non-aqueous and
mixed solvents . . . . . . . . . . . . . . . . . . . . . . . . .
8.4.5 The exchange reaction between hexacyanoferrate(lI1) and hexacyanoferrate(I1) . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.4.6 Reactions of Fe(II1) with Fe(I1) . . . . . . . . . . . . . . . . .
8.4.7 The Fe(I1)-catalysed aquation of Fe(II1) . . . . . . . . . . . . . .
8.4.8 The reaction between Fe(IV) and Fe(I1) . . . . . . . . . . . . . .
8.4.9 The exchange reaction between Ru(VI1) and Ru(VI) . . . . . . . . .
8.4.10 Ru(I1)-catalysed substitution reactions of Ru(II1) . . . . . . . . . .
8.4.1 1 The exchange reaction between Os(1II) and Os(I1) . . . . . . . . . .
8.5 Cobalt, ruthenium and iridium . . . . . . . . . . . . . . . . . . . . . .
8.5.1 The exchange reaction between Co(1lI) and Co(l1) in aqueous media;
the effect of inorganic anions . . . . . . . . . . . . . . . . . . .
8.5.2 Exchange reactions involving complexes of Co(II1) and Co(I1) with ammonia and organic ligands . . . . . . . . . . . . . . . . . . . .
8.5.3 The exchange reaction between Co(1II) and Co(I1) in non-aqueous media
8.5.4 Co(I1)-catalysed substitution reactions of Co(1II) . . . . . . . . . .
8.5.5 The reaction of Co(IV) with Co(I1) . . . . . . . . . . . . . . . .
8.5.6 Rh(1)-catalysed substitution reactions of Rh(II1) . . . . . . . . . .
8.5.7 The exchange reaction between Ir(1V) and Ir(Il1) . . . . . . . . . .
8.5.8 The reaction between complexes of Ir(1V) and Ir(II1) . . . . . . . . .

71
71
71
74
15
75
75
75
75
75
76
17
78
80
80
80
83
84
85
91
92
92
92
92
93
94
94
95
96
96
98
103
105
106
108
109
110
110
110
111

111

111
114
119
119
121
121
122
122

CONTENTS

XI

8.6 Platinum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.6.1 The exchange reaction between Pt(1V) and Pt(1l); Pt(I1)-catalysed substitution reactions of Pt(1V) . . . . . . . . . . . . . . . . . . . .
9.

. . . . . . . . . . . . . . . . . . . . . . . . . . .

122

. . . . . . . . . . . .
9.2 The exchange reaction between Eu(II1) and Eu(l1) . . . . . . . . . . . . .

128
128
130

. . . . . . . . . . . . . . .

130

CERIUM A N D EUROPIUM

9.1 The exchange reaction between Ce(IV) and Ce (111)

10.

122

URANIUM. NEPTUNIUM. PLUTONIUM A N D AMERICIUM

. . . . . . . . . . . . . . . . 130

10.1 Exchange reactions between uranium ions

10.2
10.3
10.4
10.5
10.6
10.7
10.8

Reactions between uranium ions . . . . .


Exchange reactions between neptunium ions .
Reactions between neptunium ions . . . . .
Exchange reactions between plutonium ions
Reactions between plutonium ions . . . . .
Exchange reactions between americium ions
Reactions between americium ions . . . . .

. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

.
.
.
.
REFERENCES . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

. . . . . . . .
. . . . . . . .

. . . . . . . .
. . . . . . . .
. . . . . . . .

132
133
135
138
138
141
141
142

Chupter 3 (D. BENSON)

Oxidation-reduction reactions between complexes of different metals . . . . . 153


1.

INTRODUCTION

2.

OXIDATIONS BY VANADIUM

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .
2.1 Oxidations by vanadium(V) . . . .
2.2 Oxidations by vanadium(1V) . . . .
2.3 Oxidations by vanadium(l1l) . . .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

. . .
. . .
. . .
. . .
. . . . .

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

. . .
. . .
. . .
. . .
. . . .

3 . OXIDATIONS
BY C H R O M I U M AND MOLYBDENUM
3.1 Oxidations by chromium(V1) . . . . . . . . . . . . . . . . . . . . . .
3.2 Oxidations by chromium(ll1) . . . . . . . . . . . . . . . . . . . . . . .
3.3 Oxidations by molybdenum(\/) . . . . . . . . . . . . . . . . . . . . . .

4 . OXIDATIONS BY MANGANESE A N D R H E N I U M . . . . . . . . . . . . . . . . . . .
4.1 Oxidations by manganese(VI1) . . . . . . . . . . . . . . . . . . . . . .
4.2 Oxidations by manganese(lI1) . . . . . . . . . . . . . . . . . . . . . .
4.3 Oxidations by rhenium(VI1) . . . . . . . . . . . . . . . . . . . . . . .
5.

6.

OXIDATIONS BY I R O N ( I I I ) AND R U T H E N I U M ( I I I )

153
154
154
157
159

162
162

167
169
169
169

172
175

. . . . . . . . . . . . . . . . . . 176

5.1 Oxidations by iron(II1) . . . . . . . . . . . . . . . . . . . . . . . . .


5.2 Oxidations by ruthenium(Il1) . . . . . . . . . . . . . . . . . . . . . .

176

. . . . . . . . . . . . . . . . . . . . . . . . . .

188
188

OXIDATIONS BY COBALT(III)

6.1 Inorganic bridging ligands in oxidations by cobalt(II1) complexes . . . . . . .


6.2 Organic bridging ligands in oxidations by cobalt(II1) complexes . . . . . . . .
6.3 Oxidations by aquo complexes of cobalt(II1) . . . . . . . . . . . . . . . .
7. OXIDATIONS BY PLATINUM(IV) . . . . . . . . . . . . . . . . . . . . . . . .
8. OXIDATIONS BY C O P P E R ( I I ) . . . . . . . . . . . . . . . . . . . . . . . . . .
9 . OXIDATIONS BY M E R C U R Y ( I I ) . . . . . . . . . . . . . . . . . . . . . . . . .
10. OXIDATIONS BY T H A L L I U M ( I I I ) . . . . . . . . . . . . . . . . . . . . . . . .
11. OXIDATIONS BY LEAD(IV) . . . . . . . . . . . . . . . . . . . . . . . . . . .
12. OXIDATIONS BY C E R I U M ( I V ) . . . . . . . . . . . . . . . . . . . . . . . . . .
13. OXIDATIONS BY URANIUM , N E P T U N I U MA N D PLUTONIUM . . . . . . . . . . . . . .
13.1 Oxidations by uranium(V1) . . . . . . . . . . . . . . . . . . . . . . .
13.2 Oxidations by neptunium . . . . . . . . . . . . . . . . . . . . . . .
13.3 Oxidations by plutonium . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

188

206
213
227
228
228
230
241
243
253
253
257
261
267

CONTENTS

XI1

Cliupter 4 (T. J . KEMP)

Oxidation-reduction reactions between covalent compounds and metal ions 274


1 . INTRODUCTION . . . . . . . . . . . . . . . . . . . . .
1.1 Scope and pattern of this chapter . . . . . . . . . .
1.2 Categorisation of oxidants as one- or twoequivalent .

..........

..........
. . . . . . . . . . .
2 . OXIDATION BY cr(vr1) AND Mn(vII) . . . . . . . . . . . . . . . . . . . . . .
2.1 General features . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.2 Solution equilibria of oxy-anions of Cr(V1) and Mn(VI1) . . . . . . .
2.2 Oxidation of inorganic covalent species . . . . . . . . . . . . . . . . .
2.2.1 Halide ions . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.2 Cyanide ion . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.3 Oxides of hydrogen . . . . . . . . . . . . . . . . . . . . . . .
2.2.4 Oxy-acids of sulphur . . . . . . . . . . . . . . . . . . . . . .
2.2.5 Nitrite ion . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.6 Trivalent phosphorus compounds . . . . . . . . . . . . . . . .
2.2.7 Arsenious acid . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.8 Carbon monoxide . . . . . . . . . . . . . . . . . . . . . . . .
2.2.9 Molecular hydrogen . . . . . . . . . . . . . . . . . . . . . . .
2.3 Oxidation of monofunctional organic molecules . . . . . . . . . . . . .
2.3.1 Aliphatic hydrocarbons . . . . . . . . . . . . . . . . . . . . . .
2.3.2 Olefins and acetylenes . . . . . . . . . . . . . . . . . . . . . .
2.3.3 Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.4 Aldehydes . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.5 Phenols . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.6 Ketones . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.7 Monocarboxylic acids . . . . . . . . . . . . . . . . . . . . . .
2.3.8 Amines . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.9 Nitroalkanes . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4 Oxidation of polyfunctional organic molecules . . . . . . . . . . . . .
2.4.1 Glycols . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.2 Allylic alcohols . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.3 Ketols, keto-aldehydes and keto-acids . . . . . . . . . . . . . .
2.4.4 Dicarboxylic acids . . . . . . . . . . . . . . . . . . . . . . . .
2.4.5 Hydroxy-acids . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.6 Boronic acids . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.7 Furfurals . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.5 Some metal-ion catalysed reactions of chromic acid . . . . . . . . . . .

3.

OXIDATION BY Pb(Iv), Tl(III),


RU(1II) AND MO(V1)

.
.

3.3

278
278
279
279
279
283
284
285
287
287
288
290
291
292
292
298
300
309
313
313
316
318
319
320
320
322
322
322
324
326
327
327

Hg(II), Hg(I), Bi(V), AU(III),Pt(IV), Pd(II), Rh(III),

.............................

3.1
3.2

274
274
276

General features . . . . . . . . . . . . . . . . . . . . . . . . . . .
Oxidation of inorganic species . . . . . . . . . . . . . . . . . . . . .
3.2.1 Halide and pseudohalide ions . . . . . . . . . . . . . . . . . .
3.2.2 Oxy-acids of sulphur . . . . . . . . . . . . . . . . . . . . . .
3.2.3 Hydrazine . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.4 Nitrite ion . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.5 Hypophosphorous acid . . . . . . . . . . . . . . . . . . . . .
3.2.6 Carbon monoxide . . . . . . . . . . . . . . . . . . . . . . . .
3.2.7 Molecular hydrogen . . . . . . . . . . . . . . . . . . . . . . .
Oxidation of monofunctional organic molecules . . . . . . . . . . . . .
3.3.1 Olefins . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.2 Arylcyclopropanes . . . . . . . . . . . . . . . . . . . . . . . .
3.3.3 Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.4 Hydroperoxides . . . . . . . . . . . . . . . . . . . . . . . . .

329
329
330
331
332
332
333
334
334
335
336
336
342
343
344

XI11

CONTENTS

3.4

3.3.5 Formic acid . . . . . . . . .


3.3.6 Higher carboxylic acids . . . .
3.3.7 Ketones . . . . . . . . . .
3.3.8 Ethers . . . . . . . . . . .
Oxidation of polyfunctional molecules
3.4.1 Glycols . . . . . . . . . . .
3.4.2 a-Hydroxycarboxylic acids . .
3.4.3 Dicarboxylic acids . . . . .

. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

. OXIDATION BY Ag(I1). Ag(1II). CO(III).ce(rv). m ( i I I ) .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .

v(v). rr(1v). NP(VI)AND PU(VI) .


Inorganic chemistry of these oxidation states . . . . . . . . . . . . . . .
4.1.1 Silver species . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.2 Cobalt(II1) . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.3 Cerium(1V) . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.4 Manganese(II1) . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.5 Vanadium(V) . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.6 Iridium(1V) as IrCI6*- . . . . . . . . . . . . . . . . . . . . . .
4.2 Oxidation of inorganic species . . . . . . . . . . . . . . . . . . . . .
4.2.1 Chloride ion . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.2 Bromide ion . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.3 Iodide ion . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.4 Hydrazoic acid . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.5 Bromine . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.6 Chlorine dioxide . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.7 Hydrazine and methylhydrazines . . . . . . . . . . . . . . . . .
4.2.8 Hydroxylamine. 0-methylhydroxylamine and nitrous acid . . . . . . .
4.2.9 Oxides of hydrogen . . . . . . . . . . . . . . . . . . . . . . .
4.2.10 Sulphur compounds . . . . . . . . . . . . . . . . . . . . . . .
4.2.1 1 Hypophosphorous acid . . . . . . . . . . . . . . . . . . . . .
4.2.12 Arsenious acid . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1 3 Antimony(II1) . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 Oxidation of monofunctional organic molecules . . . . . . . . . . . . .
4.3.1 Alkanes . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.2 Activated alkyl groups and polynuclear aromatics . . . . . . . . . .
4.3.3 Olefins . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.4 Benzene . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.5 Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.6 Hydroperoxides . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.1 Aldehydes . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.8 Ketones . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.9 Ethers . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.10 Carboxylic acids . . . . . . . . . . . . . . . . . . . . . . . . .
4.4 Oxidation of polyfunctional organic molecules . . . . . . . . . . . . . .
4.4.1 Unsaturated and benzylic alcohols . . . . . . . . . . . . . . . . .
4.4.2 Glycols . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.3 Unsaturated aldehydes . . . . . . . . . . . . . . . . . . . . . .
4.4.4 Unsaturated carboxylic acids . . . . . . . . . . . . . . . . . . .
4.4.5 Hydroxy ketones . . . . . . . . . . . . . . . . . . . . . . . .
4.4.6 Hydroxy acids . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.7 a-Mercaptocarboxylic acids . . . . . . . . . . . . . . . . . . . .
4.4.8 a-Keto acids . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.9 Oxalic acid . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.10 Malonic acid . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.1 1 Other dicarboxylic acids . . . . . . . . . . . . . . . . . . . . .
4.4.12 Phenols and hydroquinone . . . . . . . . . . . . . . . . . . . .
4.4.13 Aromatic ethers and amines . . . . . . . . . . . . . . . . . . . .
4.4.14 Thioureas . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1

345
346
347
348
349
349
352
352
353
353
354
355
355
355
355
356
356
356
356
358
360
362
362
363
364
365
369
370
371
371
372
373
373
374
375
376
378
378
380
383
384
387
387
388
390
391
391
392
394
395
396
399
402
402
404
406

XIV

CONTENTS

5 . OXIDATION BY Fe(II1). Ag(1).

CU(I1). NP(V) A N D MO(V) . . . . . . . . . . . . .


5.1 Oxidation of inorganic molecules . . . . . . . . . . . . . . . . . . . .
5.1.1 Halide ions . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.2 Pseudohalides; cyanide ion . . . . . . . . . . . . . . . . . . .
5.1.3 Pseudohalides; thiocyanate ion . . . . . . . . . . . . . . . . .
5.1.4 Pseudohalides; azide ion . . . . . . . . . . . . . . . . . . . .
5.1.5 Hydrogen peroxide . . . . . . . . . . . . . . . . . . . . .
.
5.1.6 Thiosulphate ion . . . . . . . . . . . . . . . . . . . . . .
.
5.1.7 Sulphurous acid . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.8 Hypophosphorous acid . . . . . . . . . . . . . . . . . . . .
5.1.9 Persulphate ion . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.10 Phosphorothioic acid . . . . . . . . . . . . . . . . . . . . .
.
5.1.1 1 Hydrazine . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.12 Hydroxylamine . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.13 Carbon monoxide . . . . . . . . . . . . . . . . . . . . . .
.
5.1.14 Molecular hydrogen . . . . . . . . . . . . . . . . . . . . .
.
5.1.15 Antimony(II1) . . . . . . . . . . . . . . . . . . . . . . .
.
5.1.16 Borohydride ion . . . . . . . . . . . . . . . . . . . . . .
.
5.2 Oxidation of organic molecules . . . . . . . . . . . . . . . . . . . .
5.2.1 Thiols . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.2.2 Carbonyl and nitro compounds . . . . . . . . . . . . . . . .
5.2.3 Formic acid . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.2.4 Unsaturated alcohols . . . . . . . . . . . . . . . . . . . . .
.
5.2.5 Hydroxy-ketones (a-ketols. acyloins) . . . . . . . . . . . . . . .
5.2.6 Ascorbic acid . . . . . . . . . . . . . . . . . . . . . . . .
.
5.2.7 Phenols . . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.2.8 Amines . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
5.2.9 Phenylhydrazine and its sulphonic acids . . . . . . . . . . . . .
5.2.10 Ortho-aminoazo compounds . . . . . . . . . . . . . . . . . .
5.2.11 Dichlorophenolindophenol . . . . . . . . . . . . . . . . . . .
5.2.12 Ethylenediaminetetraacetic acid (EDTA) . . . . . . . . . . . . .
5.2.13 Thiourea and thioacetamide . . . . . . . . . . . . . . . . . . .

6.

. 407

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.

.
.
.
.

.
.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .
. .
Simple electron acceptance by inorganic molecules . . . . . . . . . . . . .
6.2.1 Perchlorate ion . . . . . . . . . . . . . . . . . . . . . . .
. .
6.2.2 Chlorate and bromate ions . . . . . . . . . . . . . . . . . . .
6.2.3 Chlorite ion and chlorine dioxide . . . . . . . . . . . . . . . . .
6.2.4 Molecular oxygen . . . . . . . . . . . . . . . . . . . . . .
. .
6.2.5 Water . . . . . . . . . . . . . . . . . . . . . . . . . . .
. .

REDUCTION

6.1
6.2

6.2.6 Sulphur dioxide . . . . . . . . . . . . . . . . . . . . . . .


6.2.7 Xenon trioxide . . . . . . . . . . . . . . . . . . . . . . .
6.3 Simple electron acceptance by organic molecules . . . . . . . . . . . .
6.3.1 Acetylenes . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.2 Quinones . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.3 Nitro compounds . . . . . . . . . . . . . . . . . . . . . .
6.3.4 Carbonyl compounds . . . . . . . . . . . . . . . . . . . .
6.3.5 Unsaturated dicarboxylic acids . . . . . . . . . . . . . . . .
6.4 Electron acceptance followed by cleavage . . . . . . . . . . . . . . .
6.4.1 The Fenton reaction . . . . . . . . . . . . . . . . . . . . .
6.4.2 Hydroperoxides . . . . . . . . . . . . . . . . . . . . . . .
6.4.3 Halogens. cyanogen iodide. hypohalous acids and hydrogen fluoride
6.4.4 Hydroxylamine. hydrazine. hydrazoic acid and azide ion . . . . .
6.4.5 Nitrite . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4.6 Nitrate . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4.7 Peroxodisulphate ion (also called persulphate and peroxydisulphate)

. .
. .
. .
. .
. .
. .
. .

. .
. .
. .
. .
. .
. .
. .
. .

..

408
408
410
411
412
412
414
415
416
416
417
417
419
419
420
422
422
423
423
425
428
428
430
432
433
435
436
436
437
437
438
439
439
440
440
441
442
443
452
452
452
453
453
455
456
456
457
458
458
464
466
470
471
473
475

xv

CONTENTS

6.4.8
6.4.9
6.4.10
6.4.1 1
6.4.12

Peroxomonosulphate ion (Caros acid) . . . . . .


Organic halides . . . . . . . . . . . . . . . . .
p-Substituted alkyl halides . . . . . . . . . . . .
Carbon tetrachloride . . . . . . . . . . . . . .
Aromatic sulphonyl chlorides . . . . . . . . . .

. . . . . . . . . 482

. . .
. . .
. . .
. . . .

482
486
487
488

. . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
ACKNOWLEDGEMENTS
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

488
489
490
491
493
493

7.

. . .
. . .
. . .
. . .

.
.
.
.

.
.
.
.

REDOX REACTIONS BETWEENRADICALS AND METAL IONS

7.1 Stable radicals . . . . . . . . . .


7.2 Growing polymer radicals . . . .
7.3 Transient simple radicals . . . . .

Chapter 5 (L J. C S ~ N Y I )

Induced reactions . . . . . . . . . . . . . . . . . . . . . . . .
I . INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . .
1.1 Definitions . . . . . . . . . . . . . . . . . . . . . . .
1.2 Types of induced reactions . . . . . . . . . . . . . . . .
1.2.1 Coupled reactions . . . . . . . . . . . . . . . . .
1.2.2 Induced chain reactions . . . . . . . . . . . . . .
2.

2.2

2.3

2.4
2.5
2.6
2.7

3.

510

.
.
.
.

510
510
511
512
516

.
.
.
.

.
.
.
.

.
.
.
.

. . .
. . .
. . .
. . .

. . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .

519
Chromium(1V) and chromium(V) species as coupling intermediates . . . . . 519
2.1.1 Reaction between arsenic(II1) and chromium(V1) . . . . . . . . . . 521
2.1.2 Reaction between isopropyl alcohol and chromium(V1) . . . . . . . 525
2.1.3 Oxidation of other alcohols by chromic acid . . . . . . . . . . . . 528
2.1.4 Oxidation of aldehydes and organic acids by chrornium(V1) . . . . . 529
2.1.5 Reaction between iron(1I) and chromium(V1) . . . . . . . . . . . . 532
2.1.6 Oxidation of vanadium(I1) and vanadium(1V) by chromium(V1) . . . . 533
2.1.7 Chromium(V1) as indicator in the induced oxidation of arsenic (111) by
534
molecular oxygen . . . . . . . . . . . . . . . . . . . . . . . .
2.1.8 Properties of the chromium(V) and chromium(1V) intermediates . . . 536
Induced reactions caused by arsenic (IV) intermediates . . . . . . . . . . . 538
2.2.1 Iron(l1)-arsenic(II1)-peroxydisulphate system . . . . . . . . . . . . 538
2.2.2 Iron(1I)-hydrogen peroxide-arsenic(II1) system . . . . . . . . . . . 542
2.2.3 Arsenic(II1)-peroxydisulphate reaction catalyzed by iron(II1) and copper(I1) 543
2.2.4 Polarographic behaviour of the system containing peroxydisulphate,
547
arsenic(II1) and copper(I1) . . . . . . . . . . . . . . . . . . . .
2.2.5 The induced reduction of chlorate by arsenic(II1) . . . . . . . . . . 550
2.2.i Properties of arsenic(1V) intermediate . . . . . . . . . . . . . . . 552
Induced reactions involving H 0 2 and OH radicals . . . . . . . . . . . . . 554
2.3.1 Induced reactions occurring in the H 2 S Z 0 8 - H 2 0 2system . . . . . . . 554
2.3.2 Induced reactions involving other peroxy compounds . . . . . . . . . 563
564
2.3.3 The Fenton reaction . . . . . . . . . . . . . . . . . . . . . . .
Induced reactions involving sulphate radicals . . . . . . . . . . . . . . . 567
Induced reactions involving intermediates produced by partial oxidation of
thiocyanate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
569
Induced reactions effected by reduction of permanganate ions . . . . . . . . 573
Induced reactions involving tin(II1) intermediate . . . . . . . . . . . . . 575
2.7.1 Reaction between iron(II1) and tin(I1) . . . . . . . . . . . . . . . 576
2.7.2 Reaction between tin(I1) and chromate . . . . . . . . . . . . . . 576
2.7.3 Reaction between tin(I1) and perrnanganate . . . . . . . . . . . . 576

EXAMPLES OF INDUCED REACTIONS

2.1

. . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CONCLUSIONS

REFERENCES

577
577
581

This Page Intentionally Left Blank

Chapter I

Reactions of Inert Complexes and Metal


Organic Compounds
C. H. L A N G F O R D A N D M. P A R R I S

1. Introduction

This chapter is concerned with the simplest reactions of inert transition metal
complexes. Fig. 1 shows a typical compound. This is Co(II1) coordinated to six
NH, molecules to form a triply positive cation [ C O ( N H , ) ~ ] ~It+ is
. indicated in
Fig. 1 to be in aqueous solution where water molecules occupy positions in what

Outer sphere

Fig. 1. The C O ( N H ~ ) ion


~~+
in aqueous solution. The inner sphere contains six ammonia
ligands strongly bonded to Co(II1). The outer sphere contains several water molecules.

may be called the outer or second coordination sphere. The reactions in question
are typified by the replacement of one of the NH, molecules by a water molecule
from the outer sphere. Inert will be considered to mean any complex whose
reactions occur slowly enough for conventional experimental kinetic techniques
tG be applied; in general this means half times longer than 10 sec. The inert
complexes have sometimes been called robust but this term seems to be more
suggestive of thermodynamic stability than kinetic non-lability .
This chapter is restricted to treatment of the inert complexes because these
References pp. 52-59

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

provide an extraordinary range of experimental results to examine. The richness


of results arises because a wide variety of closely related structures are synthetically available when ligand substitution does not occur too rapidly. It is also of
theoretical importance to consider reactions of inert complexes because the range
of closely related structures allows careful examination of small structure changes
and correspondingly detailed information about the mechanisms of reactions.
The basic ideas about the mechanistic pathways of ligand substitution that arise
from study of inert complexes also serve as an excellent starting point for analysis
of all ligand substitutions including fast ones. In fact, two cases to be given detailed treatment, Co(II1) and Pt(lI), have served admirably as paradigms for the
general study of ligand substitution. The systems to be considered here include the
octahedral complexes of Co(IIl), Rh(III), Cr(III), Ru(11I) and Ir(IJ1) which
contain principally amine ligands and square planar complexes of Pt(II), Pd(I1)
and Au(II1) which contain similar ligands. For lack of a more preLise characterization of the situation we shall describe these as complexes of the harder ligands
in the sense of Pearsons distinction between hard and soft acids and bases2.
Consideration will also be given to some organometallic systems, principally the
metal carbonyls. These, by contrast with the other inert complexes, essentially
involve ligands which are soft bases in the Pearson classification. A cautious
reading of the evidence suggests separating consideration of these two types of
complex since their reactions occur under quite different circumstances (e.g. type
of solvent), but there is actually no strong evidence that the major generalizations
about mechanism do not extend over both classes of system.
There are essentially two distinct types of experiment which have been utilised
by students of the mechanisms of ligand substitution. One type involves the
detailed analysis of the rate law governing the reaction (including stereochemistry)
and this type can yield (under favorable circumstances) insight into the number
of elementary steps of the reaction and the stoichiometric composition of the
transition state3. However, the experimental rate law for a single reaction cannot
reveal the energetic role played by the various groups in the complex. The value
of the rate coefficient itself reveals the energy difference between two points along
the reaction coordinate. the initial state and the transition state4. The role that
any particular substituent (or component of the environment) plays in determining
that energy difference may only be assessed from a variation of the group (or factor
of the environment). Thus, the energetics of activation must be inferred from a
second type of experiment in which rates of a series of reactions presumed to have
related mechanisms are compared. It is this question of the energetics of activation
which demands our consideration first.
In a ligand substitution reaction, two groups must always receive attention.
There is a bond to the leaving group to be broken and a bond to the entering group
to be formed. The relative importance of these two processes provides a basic
dichotomy for the classification of substitutions. If a reaction rate is sensitive to

STOICHIOMETRIC M E C H A N I S M S

the nature of the entering group, it is clear that the energy of the bond being
formed is important to the activation process; and its influence must be in the
nature of an entering group assistance to activation. If otherwise, then the entering
group could not influence the rate, and the minimum requirement for a substitution reaction is that the bond to the leaving group be broken.
The question of the sensitivity of the rate of substitution to the nature of the
entering ligand provides a basis for the mechanistic dichotomy. A reaction which
is clearly insensitive to the nature of the entering group must reach its transition
state principally by the internal accumulation of the energy to break the bond to
the leaving group within the ground state complex because any significant assistance (in the sense of formation of a new bond) should be selective with respect
to the ligand assisting. Such a reaction, insensitive to the nature of the entering
ligand, will be said to have a dissociative mode of activation. Reactions which are
sensitive to the nature of the entering group will be characterized as having an
associative mode of activation because the assistance of the entering ligand does
play an important part in the determination of the free energy of activation,
although not necessarily to the exclusion of the leaving ligand. As the discussions
of particular cases will suggest, the dichotomy between dissociative and associative
activation does encompass the secondary effects on the energy of activation by
other groups in the complex or by the factors of the environment of the complex
(e.g. solvent effects).
A discussion of ligand exchange reactions of organometallic compounds associated with oxidation-reduction processes leading to free-radical formation
will be found in Volume 14 (Free-radical polymerization).

2. Stoichiometric mechanisms
So far nothing has been said about rate laws. This has been intentional, for there
is no simple relationship between the modes of activation and the concentration
dependencies that determine the rate law. The form of the rate law for a reaction,
that is, the dependence of rate on concentrations over the widest possible variations, tests hypotheses concerning the stoichiometric composition of the transition
state and the number of elementary steps of the reaction. These aspects may be
called the stoichiometric mechanism3. There I S enough variation in activation
mode among reactions of similar stoichiometric mechanism to recommend classifying reactions in two separate ways, first according to stoichiometric mechanism,
then according to activation mode. Moreover, these two distinguishing classifications follow the natural evolution in experiment. Stoichiometric mechanism is
inferred from analysis of rate laws; mode of activation is inferred from consideration of relative rates.
Note that the three transition states, a , b and c , in Fig. 3 all contain both the
References pp. 52-55

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

Reaction coordinate

Fig. 2. Free energy vs. reaction coordinate for the concerted ( I ) and two-step ( A , D ) reactions.

+Y

I
+X
M
\Y

Fig. 3. Possible transition states in the reaction M X + Y

MY+X.

S T O I C H I O M E T R I C MECHANISMS

entering and leaving ligands, Y and X in addition to the remaining metal complex
moiety. All three of these must be reached uia an initial encounter with the entering groups. Since we consider here only slow reactions of non-labile complexes we might write a simple mechanism in all three cases as
MX +Y

K
+
MX, Y

MX, Y

MY + X

(fast equilibrium)
(slow)

(1)

Here MX, Y designates an outer sphere or second sphere complex. There is every
reason to suppose that formation and dissociation of MX, Y occurs at rates approaching the diffusional-control limit so that the slow conversion to MY is a
negligible perturbation on the equilibrium of the first step. There is a similarity
here the Langmuir, the Michaelis-Menten and the Lindemann-Hinshelwood
schemes.
Two common limiting forms of the rate law for mechanism (1) are encountered
experimentally. In the event that the equilibrium constant, K , for outer sphere
complexation is small in relation to the concentration of MX and Y, the rate law
becomes
d[MX1 - k,,,[MX][Y]
dt

where kobs,the second-order rate coefficient, is kK. If K is large, the outer sphere
equilibration may become saturated and the rate law reduce to

dCMX1 - k:,,[MX]
dt

(3)

where kAbs now equals k , the rate coefficient for the slow step. It has not always
beerl realised that these two forms may arise from the same mechanism. Choosing,
as is usual, Y to be in excess, a general expression under mechanism (1) may be
written

Recall that three transition states m:ght be considered as falling within the
pattern (1). Transition state a of Fig. 3 involves strong binding of both X and Y.
In this case, it is quite possible that an intermediate of increased coordination
number is formed during the reaction. Since the initial attack of Y determines
the stereochemical course of any reaction obeying the rate law (4) there is no
References pp, 52-55

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

simple way to identify an intermediate of increased coordination number save


its actual accumulation to the extent that it can be detected. But, should such an
intermediate be detected, the pattern of reaction clearly differs in a qualitative
way from others following the scheme (1). A substitution involving an intermediate
of increased coordination number is a reaction with at least three elementary
steps: ( i ) encounter, ( i i ) addition of entering group, (iii) loss of leaving group.
Recognising that such a path must involve an a transition state it may be usefully
labelled an A , associative, stoichiometric pathway.
The other two transition states that are consistent with mechanism (1) are not
expected to involve an intermediate. They may accomplish an interchange between
the outer and inner coordination spheres in a single elementary step. Such processes
may be felicitously labelled interchange, I , reactions. Note that interchange
reactions may have either a or d transition states. That is, there may be I , or I ,
reactions. It is interesting to note precisely how these cases are distinguished. If
a wide enough concentration range can be studied so that the rate coefficient,
k , for the slow step can be found, it is a straightforward matter. However, interpretation of information from the range covered by equation (2) must normally
be attempted. It must then be determined whether variations in the observed
second-order rate coefficients as Y is varied reflect only variations in the outer
sphere association constant K , or if there are significant variations ink. If variations
in kobsare attributable to K only, k is constant and the reaction employs the d
activation mode. Definite information concerning values of K is unfortunately
scarce.
Simple examples of A , I , and I , reactions are seen to have the same rate law.
There remains another important mechanism that differs. This is the pathway
through an intermediate of reduced coordination number that is possible if a
transition state like d of Fig. 3 occurs. This path may be called dissociative, D .
The mechanism may be represented as
k-,

MX

+ M+X

(slow)

kx

M+Y

5 MY

(fast)

Mechanism (5) gives rise to the rate law

as long as the rate of capture of the intermediate M by Y is large compared to


its rate of recapture by X. If recapture by X becomes important compared to
reaction with Y the expression becomes

STOICHIOMETRIC MECHANISMS

which is not easily distinguished experimentally from expression (4) for mechanism (1). Table 1 summarizes the mechanistic categories described and Fig. 2
indicates their relationships on an energy-reaction progress diagram.
TABLE 1
CLASSIFICATION O F L I G A N D S U B S T I T U T I O N MECHANISMS

Mode of activation

Stoichiornetric niechanisrn

Associative activation
Dissociative activation

Interriiediaie of increased
roordinotion number

One-step process

1,
Id

Interniediute of reduced
coordination number

The categorization just described was proposed recently3. In most of the


literature, substitution reactions have been characterized according to the scheme
introduced by Hughes and Ingold for organic reactions. This scheme has been
critically refined by Basolo and Pearson. Following Basolo and Pearson, the following rough equivalence may be listed
s N 2

sN1

(lim)

s N 2

= 1,

sN1

= 1,

(Iim)

SN denotes substitution, nucleophilic and 1 or 2 the molecularity of the process.


The designation s N 1 seems an objectionable usage since it describes as unimolecular a process which requires the entering group as a stoichiometric
component of the transition state and suggests a too sharp distinction from the
s N 2 or I , process. This could be deemed reasonable if molecularity were defined
as the number of ligands changing covalence but such a definition is probably
no longer an operational one, and it seems unfortunate to diminish the clear
operational significance which molecularity has for gas-phase reactions.
3. Reactions of Co(II1) complexes

The most extensively studied family of non-labile complexes is the cobalt(II1)


ammine series. These are octahedral systems and all those to be considered are
low spin d6 systems. The subtle variations that can be achieved synthetically make
References pp. 52-55

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

them especially attractive for the systematic kineticist and the reactions of Co(II1)
systems have served as the model for development of most of the key concepts
about substitution reactions of octahedral systems.
A good starting place for discussion is the rate law for substitution reactions
in acidic aqueous solutions. One typical reaction is

CO(NH,)~CI'++ H 2 0 + CO(NH,),OH,~++Cl-

(8)

The rate law for this solvolytic process is

which is of course expected in this case from almost any mechanistic picture. When
the solvent is a potential ligand, the perpetual encounter between solvent molecules
and complex severely limits the cases in which the rate equation is directly informative. However, one limit of mechanism is available. The general reaction
(where Y is an arbitrary anionic ligand)

and related ones, in which CI- is replaced by another leaving group, or the NH,
ligands are replaced by others inert to substitution, all follow a rate law similar
to equation (9)'-'. Significantly, the Concentration of Y - does not appear i n the
rate law in any case except when Y - = OH-. Overlooking the exception of OHfor the moment, Y - is not a stoichiometric participant of transition state(s)
leading to its entry and there must be an intermediate in the reactions. There are
two choices of pathway, viz.

I/
-cox

slow

'I

I/

faSt

yco

+Y

I/
-Co'I

X + H,O

I/

-Co-0H2
/I

slow

1,

-/?o

+x

I/

>~o-Y

I/

-CO/I

OH2 + X

I,

+ Y fa51_ 7$o-Y+H20

The first possibility, (1 l), clearly is concerned with a dissociative mode of activation ( d ) . The second, (12), might beassociative led by water attack. But, this interpretation of (12) involves commitment to the proposition that no nucleophile
(possibly excepting OH-) has been discovered which is better than water, i.e. the
associative attack must always involve water in the first instance. The proposition

R E A C T I O N S OF

CO(II1) COMPLEXES

is unattractive and there is an appealing alternative. Pathway (12) might also


represent a case of dissociative activation without a stable intermediate. Since
water will be the predominant component of the second coordination sphere a
dissociative interchange reaction ( I d ) would lead to the aquo complex as the immediate product. The main point at the moment is that either of the attractive
interpretations of the absence of [Y] in rate laws leads to a model involving the
dissociative mode of activation. It is interesting that pathway (12) has been clearly
demonstrated to be common in Co(II1) chemistry6, . In the Co(lI1) case, as in
all others in which reaction with solvent predominates, it is fruitful to adopt the
tentative view that activation is dissociative and seek support for this from other
criteria.
It is pertinent, then, to seek a dependence of substitution rates on (i) leaving
group, (ii) solvent, (iii) steric crowding, (iu) charge, ( v ) nature of non-labile
substituents including stereochemistry, consistent with this picture of the activation mode. If these tests generally support d modes it will be desirable to examine
rate laws closely to attempt a distinction between D and I , stoichiometric pathways.
In a dissociative process the reaction rate is expected to decrease as the strength
of the metal to leaving ligand bond increases. This trend is generally observed in
Co(1II) ammine complexes. As can be seen in Table 2, a partial leaving group
order is

This is to be compared to Yatsimirskiis bond energy order estimated for the


gas phase

Perhaps the closest definition of the role of the leaving group emerges from correlation of rates with equilibria in reactions of the family CO(NH~)~X.In a
dissociative mode the leaving group in the transition state strongly resembles
the leaving group in the product state. If X is an anionic ligand, in the transition
state it should resemble the free anion. The activation free energy should respond
to changes in leaving group in much the same way as the free energy difference
for the overall reaction responds. A linear free energy relationship (see Vol. 2,
Chapter 4) is suggested between the activation energy AG and the free energy of
reaction AGO of the form A(AG) = j?A(AGo)), where A(AG) denotes change in the
free energy quantity with change of X. In the ideal dissociative case, j? would be
unity. This has been realised for the Co(NH,),X2+ family as shown in Fig. 4.
A persuasive reason for preferring the dissociative over the associative interpretation of equation (12) has emerged from recent work on reactions of Co(II1)
References p p . 52-55

10

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

TABLE 2
Co(ll1) COMPLEXES
The labile ligand is italicized.

RATES OF A C I D H Y D R O L Y S I S OF SOME

Complex*
Co(NH3)5 op(ocH3)1
Co(NH3)s NO3
Co(NH3)5 1
Co (N Ha) Br
C O ( N H ~OH2+
)~
Co(NHj)5 CI2+
C O ( N H ~SO4+
)~
CO(N H3) 5 O P 0 3H2
Co(NH3)s NOzZ
Co(NH3)5 N C S 2
Co(NH3)S O H Z +
CO(N H3)63
trans Coen20H CI+
cis CoenzOH Cl+
trans Coen,BrCI+
cis Coen,BrCI+
trans Coen,CI2+
cis CoenzCIz
trans CoenzN3CI+
cis CoenzN3CI+
trans CoenzNCSCI+
cis CoenzNCSCl+
trans Coen,NH3CIZC
cis Coen,NH3C12+
trans Coenz0HZClZ+
cis Coen20H2CI2+
trans Coen,CNCI+
trans CoenZNOzCI+
cis Coen,NO,Cl+
trans Coen2NOzEr+
trans Coen, Br2
trans C O ( N H ~ ) ~ C I ~
cis CO(NH,),CI,
+

AT

k(sec-)

AHX(kca1)

2.5 x 10-4
2.7 x 1 0 - 5
8.3 x I O - ~
6.3 X ~ O - ~
5.8 x I O - ~
1.7 x 1 0 - 6
1.2 x10-6
2.6 X I O - 7
1.15 x
5.0 x I O - O
very slow

25.5
23.5
27
23
19

+6
-9
-24

31
26.2
23.1
24.9
23.5
26.2
21.5
22.5
21.3
30.2
20. I
23.2
24.5

0
-

-lO-*O

1.6 X I O - 3
1.2 x10-2
4.5 X I O - 5
1.4 X I O - 4
3.5 X I O - 5
2.4 x 10-4
2.2 x 10-4
2.0 ~ 1 0 - 4
5 x10-8
1.1 X I O - 5
3.4 X I O - 7
5 XIO-7
2.5 x I O - ~
1.6 x I O - ~
8.2 X I O - 5
9.8 x 10-4
1 . 1 XIO-4
4.0 ~ 1 0 - 3
1.2 x10-6
1.8 X I O - 3
fast

22.5
20.9
21.8
-

250 C

ASt(eu)

ReJ

13

+6

14
14

-4

14
15

+20
+ 10
+2
+ 14
+I4
-5
0
-4
+9
- 14
-11

-6
-

-2

-2
-3
-

14
14

13
16
17
14
18
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
20
20
21, 19
18, 19
18, 19
18, 19
18
23
23

~~-

en represents ethylene diarnine.

complexes in solvents other than water. Work of Tobe, Watts, Langford and their
respective collaborators has demonstrated that these solvents, dimethyl formamide, dimethyl sulphoxide, dimethyl acetamide and methanol also function as
preferred nucleophiles. This reinforces the suggestion that it is the high solvent
concentration and not solvent nucleophilicity that is important. Furthermore, it
is found that in these solvents some direct replacement by an anion may be
observed but that such replacement is always associated with ion pair formation
and that reaction rates show very little sensitivity to the nature of the entering
ion12.

REACTIONS OF

cO(III)

COMPLEXES

11

-3 -

-4-

.-C

X
0
J

-5-

LOP H

eq

Fig. 4. Linear free-energy relationship for the reaction, C O ( N H ~ ) ~ O +HX~- ~*+C O ( N H ~ ) ~ X * +


+HzO. Log k (rate coefficient) us. log K (equilibrium constant).

The interpretation of the next factor, steric crowding, is quite straightforward


if its effects can be isolated. Steric crowding should inhibit an associative reaction
but accelerate a dissociative one. It is frequently difficult to isolate the steric effects
for a reaction in solution since the structure variations that result in crowding
of the reaction site may also modify the surrounding solvent structure in an uncontrollable way or be associated with important electronic effects. There is at
least one clear cut experiment concerned with the steric effect on a Co(II1) complex
that supports the dissociative mode. This is the comparison of the acid hydrolysis
rates for d, 1 and meso Cobn,Cl,+ (bn = 2,3-diaminobutane). It can be seen
that the methyl groups on the chelate rings must be opposed in the meso form and
staggered in the d, 1 form. The meso form hydrolyzes about thirty times faster24.
The other available data are consistent with this suggestion of acceleration by
steric crowding and its implication of dissociation.
The effect of overall charge on the complex is perhaps even more difficult to
isolate than the steric effects. Each change of charge type is accompanied by an
important change in the electronic arrangement about the metal which the discussion (below, p. 12) of electronic effects of non-labile ligands shows to be quite
important. However, an assessment of the rates for the 2' and 3+ charged species
cited in Table 2 suggests that substitution rates at Co(II1) decrease as the overall
charge on the complex is increased. This conforms to expectation for the disReferences pp. 52-55

12

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

sociative model when the leaving group is anionic. For the associative model, one
might expect opposite or at least very small charge effects (see the discussion of
Pt(II), p. 20).
In this catalogue of structure variation experiments to test the hypothesis of
a dissociative activation mode, the last is the role of non-labile ligands. This
question has been examined using the hydrolysis reactions of the family of complexes cis and trans Coen,ACI+ where CI- is the leaving group and A is a variable
non-labile substituent (see rates in Table 2). The first approach taken to the
analysis of the data was to classify the ligands A as electron donors or electron
acceptors on the basis of organic chemical precedent and then to plot the observed
25" rate coefficients as a function of decreasing electron donor-increasing electron
acceptor proper tie^^^. The two branches of the curve (Fig. 5) were given a two-

-1 -

"

-2-

P)
rn

a-- 3 m
J

-4-

-5-

-6-

-'t
NHZ

OH-

N;

CI-

NCS-

NHJ

H,O

CN-

NO,

Fig. 5. Rates of hydrolysis of a series of Coen,ACI" complexes. The abscissa represents the
electron-donating or -accepting power of A .

mechanism interpretation. Good electron donors were supposed to replenish the


depleted electron density at Co(II1) in a dissociative transition state. Electron
acceptors (NOz-, CN-) were supposed to drain away the excess electron density
at Co(1II) in an associative transition state. A difficulty for this attractive hypothesis is that there is no correspondingly simple pattern in the values of AH'.
The extended effort to obtain some evidence for nucleophilic discrimination
in the reactions supposed to involve an associative transition state have been

R E A C T I O N S OF

CO(II1) COMPLEXES

13

reviewed. 26. No direct support has emerged for the postulation of an associative
transition state. Fortunately, there is an alternative account of the situation. The
dissociative transition state is an incipient five coordinate complex. In the extreme
case there are two possible geometries, trigonal bipyramidal and square pyramidal
and different ligands may stabilize geometrically different transition states. Note
(Fig. 6 ) that the square pyramidal form cannot lead to cis-trans isomerization

a-M

trans

cis ; t r a n s
=2.1

cis

Fig. 6 . Stereochemical changes accompanying the dissociative reaction MA4BX + MA4B +


MA4BY.

whereas the trigonal bipyramidal form can. A fairly satisfactory correlation


between stereochemical rearrangement emerges following the suggestion that
the formation of a strongly trigonal form is accompanied by a positive AS. The
two-geometry uniformly dissociative model seems to give the most consistent
account of the effects of non-labile ligands.
The hypothesis of dissociative activation in Co(II1) reactions stands the available tests well. It is therefore profitable to attempt to distinguish the D from the
I d pathways. Fig. 7 summarizes the two pathways consistent with d activation,
and the general methods for establishing the stoichiometric mechanism I d are
illustrated by the example of C O ( N H ~ ) ~ O H , ~ + .
First, a case against the D path may be constructed. A knowledge of the rate
References pp. 52-55

14

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

-M

Fig. 7. The I d us. the D mechanism.

ofwater exchange of CO(NH,),OH,~+ and the rate of hydrolysis of Co(NH3),X2+


under concentration conditions where the reaction goes to completion gives
k--H20and k - x . These may be combined with the overall equilibrium constant
for the reaction to give the ratio k + H 2 0 / k + X
the
, competition ratio for the intermediate CO(NH,),~+.A number of these competition ratios were calculated by
Haim and TaubeZ7including the case X = SCN-. The assumed mechanism may
be tested by generating the intermediate from a different source and checking
the competition ratio by evaluating the immediate product distribution. This was
done by Pearson and Moore6 who generated the intermediate by hydrolysis of
the labile nitrato and bromo pentaamines of Co(II1) in the presence of a large
concentration of SCN- ion. In conflict with the explicit predictions from the D
mechanism27, they found no evidence for capture of the intermediate by
thiocyanate. Unless very small concentrations of Br- or NO3- in solution affect
the reactivity of the intermediate, it is necessary to conclude that it is not there.
Now we can proceed to assemble the positive evidence for the I d path
(I -+ I1 -+ IV, Fig. 7). Once the outer sphere complex, (11), is formed, all replacements of water should occur at the same rate, k-H20.If the ion pairing constant
K, is known, or a limiting rate of anion entry corresponding to saturation of the
association is observable, the rates of conversion of (11) into (IV) may be compared
for various X. All should be equal t o k-HzOif the activation mode is d , but they
will not equal the rate of water exchange which wasidentified with k-H20 on the
D path. The reason is that species (11) has a number of solvent molecules in its

R E A C T I O N S OF

CO(lI1) C O M P L E X E S

15

outer coordination sphere as well as the ligand X. Even if the rate of dissociation
of water is unaffected by the presence of X in the neighborhood, the most probable
result of water loss will be water exchange and not X entry. Thus on the path we
expect all ion pairs tc show closely similar rates of conversion to aniono complexes, but we expect these rates t o lie below the solvent exchange rate by the appropriate statistical factor for the population of the outer sphere. Recently 9 , l 3
the rates of formation of Co(NH,),X from Co(NH,),OH, . . . X have been
reported relative to the water exchange rate for X = SO4, C1-, SCN-, and
H2P04-. The values are 0.24, 0.21, 0.16 and 0.13, respectively. The values span
a range of a factor of two which must be admitted to be a little larger than the
experimental uncertainty and also easily within the differences among the anions
in their probability of occupancy of the crucial outer sphere site adjacent to the
leaving water molecule. All are nearly a factor of five below the water exchange
rate. These results conform neatly to the I d predictions.
Examining the relationship between the probability that a ligand occupies an
outer sphere site, the rate of ligand incorporation, and the solvent exchange rate,
appears to be the most general method for identifying the I,, pathway. It has
been applied to several other systems recently. Rates of anion entry into cis[Coen2NO2(DMS0)I2+in dimethyl sulphoxide (DMSO) have been compared
to the DMSO exchange rate obtained from the deuterium tracer NMR experiments
of Lantzke and Watts29. The pattern is very akin to that for CO(NH,),OH,~
in water. Similarly the rate of sulphate anation in the pair ~ i s - [ C o e n , ( O H ~ ) ~ l ~ +
. . . SO4- has been found to be 0.25 times the water exchange rate of the free
ion 9a.
Several authors have suggested that the Id pathway may prove to be the most
common mechanism in substitution reactions of octahedral complexes generally.
However, the D path can be clearly demonstrated in some cases including at least
two examples from Co(II1) chemistry. The path (I -+ 111 -+ IV, Fig. 7) through the
fivecoordinate intermediate would lead, in the case of rate studies in the presence of
excess anionic ligand, to observed first-order rate constants governed by equation
(13)

This form of [X-] dependence was observed by Haim et


in studies of
the anation of Co(CN),OH,-.
Equation (13) predicts a limiting rate at high
[X-] equal to the solvent exchange rate, and allows substantial variation in reactivity of X- groups. These features are realised in the Co(CN),OH,- system.
The reactivity order toward the intermediate Co(CN),- is: OH- > 1,- > NH,
> SCN- > thiourea > NH, > Br- > S203- > NCO- > H 2 0 , spanning
about four orders of magnitude. A parallel case3 has evolved for the intermediate
~

References p p . 52-55

16

I N E R T C O M P L E X E S A N D METAL O R G A N I C C O M P O U N D S

Co(NH,),SO,+. The order of reactivity toward this five coordinate species is:
OH- > NO,- > CN- > NH, > H,O and spans six orders of magnitude.
These two five coordinated Co(II1) species are, to date, the only ones clearly
established from detailed knowledge of the rate law for substitution (a situation
very clearly subject to change). Strongly suggestive evidence for others has been
accumulated from another approach. Loeliger and Taube3, and Sargeson et al.,,
have examined reactions of a complex where there is a distribution of products
(i.e. 6O and l8O aquo complexes or stereoisomerically different complexes). They
argue that a constant product ratio strictly independent of the nature of the
leaving group implies product formation ufler the leaving group is removed from
the scene of reaction and the existence of an intermediate. The key word here is
strictly. Leaving group effects may often be quite subtle. In I , processes, for
example, they would appear only to the extent that they modified outer sphere
populations. Some persuasive indication of D reaction has been presented for
some of the so-called induced aquations of Co(II1) amine complexes. Induced
aquations include Hg2+-catalyzed halide loss and rapid azide loss catalyzed by
nitrous acid.
Before leaving Co(II1) chemistry we must consider the base hydrolysis reaction
Co(NH,),X+OH-

--+

(14)

Co(NH,),OH+X-

The usual rate law is typified by


-d CCo(N 3)5x = k [Co(N H ,) X] [OH-]
dt
It is important to appreciate that the values of k in equation (15) are often quite
large when compared to the rates summarized in Table 2. Some of these values
appear in Table 3.
The simplest interpretation of equation (15) would assume a nucleophilic attack on Co(II1) by OH-. This, however, would put OH- in an extraordinary
category of nucleophilicity. G a r r i ~ k ,was
~ the first to note that an alternative
explanation for the role of OH- was available. In the alternative, the conjugate
base of the initial complex ammine is presumed to be formed in small amount
and to function as the actual reactive species
Co(NH3),CI2 +OH+

Co(NH,),(NH,)Cl+
C O ( N H ~ ) ~ ( N H ,+H,O
)~+

s Co(NH,),(NH,)CI+ + H,O(fastequilibrium)
+ Co(NH3),(NH2)
+

Co(NH,),OHZf

fC1- (slow)

(16)

(fast)

The mechanism allows for a slow step analogous to the d process observed in

Cr(III), Rh(III), Ru(III), Ir(II1)

AND

Pt(IV)

17

COMPLEXES

TABLE 3
R A T E S O F B A S E H Y D R O L Y S I S OF S O M E

cO(III)

COMPLEXES

The labile ligand is italicized.

koH(l.mole- '.see-')

Cottiplex*

Co(NH3)512+
Co (N H 3) 5Br
Co (N H Cl
Co(NH3)5 N 3 2
CO(NH~)~NO~*+
trans Coen2CI2+
trans Coen,OHCI+
trans Coen2NO2CI+
cis Coen2C12+
cis Coen20HCl+
cis Coen2N02Cl+
trans Co(d,I-bn)2C12+
trans Co(Meso-bn), C12
+

23
7.5
0.85
3 x 10-4
4.2 x
85
0.017
0.080
15.1
0.37
0.03
2100
9800

T("C)

25
25
25
25
25
0
0

0
0

0
0

25
25

E,,(kcal. mole- ')

29
28
29
33
38
23.2
22.8
24.4
24.6
22.4
23. I

AS(eu) Ref.

+42
+40
36
+35
30
-

+
+

35
35
35
35
35
36
36
36
36
36
36
31
37

bn represents 2,3-diaminobutane.

acid hydrolysis, and the extensive evidence of a parallel between acid and base
hydrolytic reactivity has been reviewed3'. It has also been established by measurement of H-D exchange on the ammine l i g a n d ~ ~ ' . that
~ ' the conjugate base
can be formed sufficiently rapidly. The most telling experiments, though, are those
that establish that product formation occurs in a step independent of the initial
hydroxide involvement. Green and T a ~ b have
e ~ shown
~
that the 160/180
isotope
fractionation factors in base hydrolysis are more easily explained assuming in~ ~ shown
corporation of 0 from H 2 0 than from OH-, and Sargeson et u I . have
that other anions (Y-) can effectively compete with OH- in the base catalyzed
pathway to give products C O ( N H , ) ~ Y ~in+ addition to Co(NH3),OH2+.Thus,
there seems little doubt that direct OH- attack on Co(I1I) is excluded.
The reason for high reactivity on the base-catalyzed pathway remains something
of a puzzle. The simplest, but not entirely convincing interpretation, suggests
that the NH2- ligand functions as an electron donor similar to OH-.
has suggested, on spectroscopic and stereochemical grounds, that the conjugate
base species may be labile because it is a high spin d6 complex. This view is rendered
more attractive by Watt and Knifton's recent report45 of a paramagnetic solid
Co(II1) conjugate base species isolated from liquid NH,. Gillard46 has made the
interesting suggestion that OH- may not function as a base but as an electron
donor, to produce a transient OH radical and a labile Co(I1) species.
4. Cr(III), Rh(III), Ru(III), Ir(II1) and Pt(1V) complexes

The related octahedral non-labile complexes which have received some attention will be grouped together simply because there is much less information
References p p . 52-55

18

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

available than there is with respect to Co(II1) complexes. Reactions of Pt(1V)


complexes are very slow and attempts to isolate simple thermal substitutions from
Pt(I1)-catalyzed redox pathways and photochemical reactions have not yet been
very s u c ~ e s s f u 1 ~ Information
~-~~.
on Ir(II1) complexes is very limited to date
and probably not adequate for mechanistic analysis'0*
A reasonable account
of Cr(III), Rh(II1) and Ru(1II) behaviour in acid solutions may be constructed
adhering to the postulate of d activation, although experimental results are less
comprehensive than in the Co(I11) case and the assignment of activation mode is
less secure. Reactions of all three also shed important light on the base hydrolysis
pathway. Some important rate coefficients. for these systems are collected in
Table 4.
Leaving-group orders for acid hydrolyses of both the Cr(NH3),X2+ and
Cr(OH2),X2+systems are parallel to those for Co(III), suggesting the d activation
mode. This point is supported by a correlation between ASf for hydrolysis of
Cr(OH2),X2+ and Sohyd,,
the entropy of hydration of the ions X-, discovered
by Swaddle and King5,. This suggests that X- functions as a solvated anion in
the transition state. The evidence from charge dependence in Cr(III), indicated
in Table 4, is also consistent with a d process. Detailed evidence on steric crowding
of Cr(II1) is lacking but it has been noted66 that Co(CH,NH2),CI2+ hydrolyzes
faster than Co(NH3),CIZ+ whereas Cr(CH3NH2),C12+ hydrolyzes more slowly
than Cr(NH3)5C12+.This seems discordant.
One of the earliest suggestions of what we characterize as the I d mechanism
of reactions of Cr(NH3),Br2+ ion paired
emerged from a study by Jones et
with various organic anions. They found that the initial entering group was water
even though the final thermodynamic product was the complex of the organic
anion. Fairly strong evidence for the stoichiometric I d pathway was presented
by Duffy and Earley6' who established that reactions of Cr(NH3),0H23+ with
SCN-, CI- and H2"0 follow a pattern very similar to the reactions of
Co(NH,),OHZ3+ (seep. 11). It has been shown69that hydrolysis of Cr(OH2),12+
in the presence of CI- produces 10-20 % Cr(OH2),C12+. This could be explained
by the formation of the D intermediate CI-(OH,),~+ which reacts selectively
with C1-. Unfortunately, a much more likely explanation can be given because
of the d e m o n ~ t r a t i o nthat
~ ~ I - has a strong trans-labilizing effect in Cr(OH2)J2+.
The observed product probably arises from rapid hydrolysis of Cr(OH,),CI I + .
Perhaps the most telling piece of information about Rh(II1) reactions in
aqueous solutions is a study7' of the reactions
trans Rhen,CI,+ + 2 X-

trans Rhen,X,+ + 2 CI-

(17)

as a function of the entering group, X. The observations are summarized in Table


5. The small differences very probably reflect secondary factors and there is no
evidence for nucleophilic attack over this quite varied set of nucleophiles. The

Cr(III), Rh(III), Ru(III), Jr(II1)

AND

Pt(IV)

19

COMPLEXES

TABLE 4
25 "C

RATES OF H Y D R O L Y S I S R E A C T I O N S OF SOME O C T A H E D R A L C O M P L E X E S A T

The labile ligand is italicized


A . Acid hydrolysis
Complex

k(sec- l )

E,(kcal.rnole-

Cr(NH3)JZ+
Cr ( N H ) Br
Cr ( N H 3 )sC12
Cr(NH3)63
Cr(OH2)5 F 2+
Cr(OHZ),Cl2+
Cr(OH2)SBr2+
Cr(OH2)5ZZ
Cr(OH2)sN32+
trans Cr(OH2)4C12+
cis Cr(OH2)4CIZ+
trans Cren2CI2+
cis Cren2C12
trans Cren20H2CI2+
cis Cren20H2C12+
Cr(NCS)63Ru(NH3)SCI'

I . O X 10-3
6.8 X
7.3 x 10-6

21
24
24
26
29
25
24
23
33
27

RU

ci63-

trans Ru (OHZ)3C13
Ru(OH2)SCI'
R h (N H 3 ) Br
Ir(NH3)sBr2+
Rh(OH2)CIsZIr(OH2)C152IrC163-

I.OX 10-7

6 . 2 ~
2.8 x 1 0 - 7
3.1 x
8 . 4 1~0 - 5
4.6~
8.3 x 1 0 - 5
4.3 x 1 0 - 5
2.2 x 1 0 - 5
3.3 x 10-4
< 10-6
-3 x
5.3 Y 1 0 - 5
7 . o ~10-7
-1
2.1 x 10-6

Ref.
52
53
53
54
55
55
56
55
55
57
57
58
58
58
58
59
60
61
61
61
62
51
63

23
21
-

23

'.)

26
21

10-8

2x10-'0
3 x 10-4

64

9.4 x 10-6

30

64

Complex

ko,(l.tnole- 'see- ' )

Ea(hcal.ttiole- I )

ReJ

Cr(NH3)5/2+
Cr(N H 3 ) Br2
Cr(NH3)5CIZ+
Ru(NH~)~CI'
Rh(NH3)s/Z+
Rh(NH3)sBr2+
Rh(NH3)5C12+

3.6
6 . 9 ~lo-'
1.7~10-3
4.9
7.3 x lo-'
3.4~10-4
4.1 x 10-4

21
26
27

52
52
52

B. Base hydrolysis

60

33
31
29
-

.-

65
65
65
~

pattern is very similar to the indiscriminate reactivity of Co(l1J) complexes and


very probably means that the initial product of substitution is normally the aquo
compound. This would again recommend provisional assignment of the dissociative mode of activation. However, the evidence available from other effects
on Rh(II1) reactivity does not support the hypothesis as clearly as is the case in
Co(II1) chemistry. Steric crowding does result in faster reactions, but the effect
References p p . 52-55

20

INERT COMPLEXES A N D METAL O R G A N I C COMPOUNDS

is smaller71. Of course, this may be as expected for the larger Rh(I1I) ion. Limited
data7 on the effect of changing charge on the ion is contrary to expectation, for
the more highly-charged species react the faster. It is not here at all clear that only
overall charge effects are being measured.
TABLE 5
R A T E S OF:

trans Rhen *cI2++2x--f trans Rhen zx2++2


c1- A T 8 O 0 C

k x I05(sec-)

X
OH- (0.1 M)

5.1
4.2
4.2
5.2

NOL-(0.1 M)
NOz- (0.05 M )

I - (0.1 M)

I-(0.05 M)
36c1-(0.01 M)
Thiourea (0.1 M)
NH3(5 M)

5.1
4.0

4.9
4.0

A very striking feature of Rh(II1) chemistry is illustrated in Table 5. This is


the reduced importance of the hydroxide-dependent base hydrolysis. Comparison
of data in Table 4 with those in Table 3 reveals that base hydrolysis is also less
important in Cr(II1) than in Co(1II) but that it is a n important feature of Ru(II1)
chemistry. Some clue to this puzzling situation is provided by the observation6
that Ru(NH3)&l2+ undergoes H-D exchange more than 100 times faster than
C O ( N H ~ ) ~ CRh(NH3),C12+
~~+,
or Cr(NH3)&l2+. There could be, it seems,
more of the conjugate base species present in the Ru(II1) system. In fact it has
been estimated that the reactivity for the conjugate base species follows the order:
Cr(II1) > Ru(III)~. There does not as yet appear to be a
Co(II1) >> Rh(II1)
simple explanation for this order of reactivity. One more interesting fact deserves
note. Base hydrolysis is important in reactions of Rh(NH3)5C12+ (see Table 4)
but is undetected in reactions of trans Rhen2pyC12+.An ammine group trans to
the leaving group appears to be required.

5. Complexes of Pt(II), Pd(lI), Au(II1) and Rh(1)


To this point the complexes considered have shared the coordination number
six and approximate octahedral geometry. It has been argued that they also share
the dissociative reaction mode. There are examples of reactions both with and
without intermediates of reduced (that is, 5) coordination, but the insensitivity
to entering ligands is a consistent feature. It will be interesting, shortly, to see if
the dissociative partern persists in more or less organometallic octahedral
systems but first we shall give some attention to the non-labile square planar
systems.
As we turn to complexes with only four groups bound in a planar arrangement
and two potential coordination positions open it should not be too surprising

Pt(II), Pd(II), Au(1II) A N D Rh(1)

21

COMPLEXES

to encounter a change of mechanism. That such a change does occur is revealed


from the rate law which was discovered by Rich and T a ~ b for
e ~ the
~ radiochloride exchange reaction

The rate law for exchange is


rate

(kj + k2[36C1-])[A~C14-]

(19)

and incorporates a first-order term independent of the entering ligand plus a


second-order term dependent upon the entering ligand. This rate law has been found
to be quite general for substitution reactions of square planar systems. The most
detailed experiments have been carried out on Pt(I1) complexes.
The new feature is the k, term in the rate law. Fig..8 shows how this depends
on the nature of the entering ligand for the particular reaction
Pt(dien)Br+ + Y

+ Pt(dien)Y+Br-

(20)

(dien = 1,4,7-triazaheptane,diethylenetriamine;
a tridentate ligand). The figure
shows apparent first-order rate coefficients in the presence of excess Y as functions
of Y, values of k, ranging from "0" and 8.8 x l o w 4I.mole-'.sec-' for OHand C1- to 4.3 x lo-' and 8.3 x lo-' I.mole-'.sec-' for SCN- and thiourea.
The pathway characterized by k , is clearly selective with respect to entering groups
and must be associative. This result raises the interesting question of what comI

"

I
*-

0
X

0
0

NOP,Py.CI

,
0.01

rY1.M

OH-

0.02

0.0 2

Fig. 8. Relative reactivities of different entering groups Y with PtdienBr+ in aqueous solution
at 25 "C. Data from ref. 74.
Referrtires pp. 52-55

22

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

prises a good nucleophile for Pt(I1). Apparently OH-, which is both a strong base
and a good nucleophile for attack at carbon is quite unreactive toward Pt(I1).
Of course, reactivity of nucleophilic reagents is quite sensitively dependent upon
the nature of the substrate. A good deal of effort has been directed toward the
exploration of the correlation of nucleophilic reactivity with easily measurable
characteristics of the nucleophile. A two-parameter approach based on the
basicity (toward H) and polarizability of the nucleophile meets with considerable
success. Nucleophilicity toward Pt(I1) is clearly most dependent upon polarizability, but it should be emphasized that the exact connection between the bulk
polarizability of a species and its nucleophilic reactivity is far from clear.
Nucleophilic reactivity toward Pt(I1) complexes may be conveniently systematized via linear free energy relationship^^^ established between reactions of trans
Ptpy2C1, (py = pyridine) with various nucleophiles and reactions of other Pt(I1)
complexes with the same nucleophiles. First, each nucleophile is characterized by
a nucleophilicity parameter, nopt derived from its reactivity toward the common
substrate, trans PtpyzC1,. Reactivity toward other Pt(I1) substrates is then quite
satisfactorily represented by an equation of the form (21), wherein k, is the value
of k, in the reaction with nucleophile Y

The constant s, characteristic of the substrate complex, reflects its sensitivity to


variation in nucleophilicity as assessed by the Ptpy2C1, reaction. It is called the
nucleophilic discrimination factor (NDF). The intercept log k, turns out to be
related to the value of the k, term in the rate law for the solvent in question. Some
representative ligands involved in attack on Pt(I1) complexes may be listed in
order of decreasing noptas follows75
(C,H,),P

> S=C(NH,), > SeCN- > CN- > SCN- > I - >Br- > Py

> NH, > CI- > CH,OSome values of s (or NDF) fall in the order76
trans Pt[(C,H,),P],CI,

> Ptpy2C12 > Pten2C12 > Pt dieii Br

> Pt dien CI > Pt dien OH2+


The k, term i n the rate law for square planar substitution is very clearly connected with an associative mechanism. The k , term may be also. Consider the
pathway (S = solvent)
L,Pt-X+S

-+

L,Pt-S+X

(slow)

L,Pt-S+Y

--*

L,Pt-Y+S

(fast)

Pt(II), Pd(II), Au(II1)

AND

Rh(1)

COMPLEXES

23

This would accomplish substitution of Y for X by a potentially associative solvent


attack followed by a fast replacement of solvent. This interpretation of the k ,
path is strongly supported by two lines of evidence.
First, the order of k , values in different solvents is quite reasonably interpreted
as a nucleophilicity
(e.g. (CH3),S0 > H,O
CH3N0, > C,H,OH)
and second, the k, rate is (as is the k , ) greatly reduced by steric ~ r o w d i n g ~ ~ . ~ ~ .
Of course, a nucleophilic attack by solvent is a very likely process, apriori. In any
solvent, the solvent itself will be the poorest nucleophile that can be studied since
poorer ones will not effectively compete. Thus the k, term of equation (21) corresponds to the k , value. The aquo intermediate of scheme (22) has been trapped
by using reactions in the presence of OH-, a poor nucleophile but good basea6.
In the reactions of Pt(I1) complexes three ligands, the entering, the leaving and
the trans group, are found to have a large influence upon the rate. As we have
seen, the entering group is quite important and that fact establishes associative
activation. It is probably not surprising that the leaving group is also important.
In fact, leaving group variations span the same wide range of rates as entering
groups. A typical replacement rate series is8'

H,O departs about lo5 times as fast as CN-. A good attacking group is a poor
leaving group with very few exceptions. O H- is notable, for it is a very poor
nucleophile for Pt(I1) but is only very slowly replaced. In the main, lability also
corresponds to thermodynamic instability, for example in the series

< Pt(dien)Br+ < Pt(dien)Cl+


formation constants: Pt(dien)I+ > Pt(dien)Br+ > Pt(dien)Cl+
rates: Pt(dien)I

The third ligand which plays a major role in determining rates of substitution
at Pt(I1) is the ligand which is trans to the leaving group. Its effect is as spectacular
as the effect of the directly involved entering and leaving ligands, and, since first
being recognized and explored by Wernera2 and ChernyaevE3, it has fascinated
students of Pt chemistry. The trans effect has been crucial to the control of synthesis in the square planar series. A fairly constant order of labilizing effect of
tram ligands can be given. If we could study the rate of release of a single leaving
group being replaced by a single entering group with the trans group varied from
CN- to OH-, the rate would decrease by a factor 01 about lo6. A trans effect
ordera4 is given in Table 6. Inspection of the table reveals that a good trans
labilizing ligand corresponds to a good nucleophile and a poor leaving group in
the great majority of cases.
The complementary roles of trans, entering, and leaving ligands when contrasted
with the much smaller role of variable cis ligands suggest an explicit model for
RrJrrenLrs pp. 52-55

24

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

TABLE 6
Trans EFFECT O R D E R I N G O F L I G A N D S
Relative rates
very large

co
CN-

>c-c <

>

large

>

1-

>

moderate

c1-

>

small

>

FH3
OH-

SCNPMe3

> PEt3, H - > CH3- > CbH5-

the transition state. Associative attack at Pt(I1) probably is well represented by


the scheme

In scheme (23) the transition state is concerned, as approximating a trigonal


bipyramidal five-coordinate species, with the leaving, entering and trans ligands
sharing the trigonal plane. It is significant that the model also accounts for the
fact that substitutions at Pt(I1) occur with retention of geometrical isomerismS5
(cis -+ cis and trans -+ trans).
Finally, it is noteworthy that the associative activation model suggested for
reactions of Pt(I1) species is strongly supported by interpretation of the activation
energetics. The characteristic finding86 is a low value of the activation enthalpy
and a highly negative activation entropy. These combine to suggest that formation
of the transition state is accompanied by a net increase in bonding.
If all Pt(I1) complexes are activated associatively, it becomes plausible to suggest that the five-coordinate species in scheme (23) is not merely a transition state
but also an intermediate. Unfortunately, this question of stoichiometric mechanism
cannot be resolved by asking for evidence in the rate law for competitive reactions
with the intermediate. The only way to look for the intermediate directlyis to search
for its accumulation. Such accumulation could be evidenced by departure from
the second-order rate law at high ligand concentration if the step forming the
intermediate becomes rapid compared to its subsequent breakdown. Such kinetic
behaviour has been observed once in a rather unusual Pt(I1) complex containing
sulphur donor atom ligands. Indirect evidence suggesting the intermediate
comes from the preparation of several stable five-coordinate Pt(I1) complexesxx,
such as Pt(SnC13)53-, PtH(SnC13),3- and Pt [(C2H5)3P]2H(SnC13)2-. The
generality of the occurrence of the five-coordinate intermediate must be regarded
as unknown.

25

CLASS LIGANDS: THE B I N A R Y CARBONYLS

This section so far has concentrated on the evidence for a and possibly A reactivity at Pt(I1). Little has been said of the other square planar systems. This is
because much more data are available on Pt(I1) and where data for other complexes exist, they exhibit the pattern made familiar by Pt(I1). The rate coefficients
for reactions of AuC1,- are larger73 than those for PtC1,- suggesting increased
bonding by the entering group in the transition state. It is also true that square
planar Pd(I1) and Ni(I1) complexes react faster than the corresponding Pt(I1)
systems. Probably this reflects easier achievement of five-coordination. Axial
perturbations in solution are more pronounced for these square planar systems.
Some rates of reaction of complexes with pyridine are compared in Table 7. The
Rh(1) and Ir(1) square planar complexes Rh(CO),(p-anisidine)Cl and Ir(CO),
(p-to1uidine)Cl undergo second-order exchange with 14C0 at - 80 C in ethanol,
with k, approx. 2 l.mole-.sec- in both cases.
TABLE 7
R A T E S O F R E A C T I O N S OF A N A L O G U E S OF

[Py] ( M )

Complex

Ptdien SCN
O.OOO10
Pddien SCN
0.00123
Ptdien NO2+
0.00592
Pddien NO2+
0.00124
tr~ns-Pt
[ ( C Z H ~ ) ~ P ] ~ ( O - ~ O I ~0.0062
I)CI
trans-Pd [(C2H5)3P]2(o-to1yl)CI 0.0062
trans-Ni [(C2H5)3P]2(o-tolyl)CI 0.0062
+

Pt(1l)

COMPLEXES W I T H PYRIDINE

Soloent

Temp (C) kOb,(sec-)

H20
HzO
H2O
H20
C2HSOH
C2H5OH
C2H5OH

25
25
25
25
-25
-40
- 65

6.2 X lo-
4 . 2 lo-
~
5 . 0 ~

3.3 x 1 0 - 2
1.7~10-~
5.8 x 10-3
1 . 6 lo-
~

ReJ

81
81
81
81
90
90
90

6. Complexes with B class ligands: the binary carbonyls


The characteristic of these complexes in general is that ligands very often possess
vacant (x-bonding) orbitals. It applies particularly to those unsaturated ligands,
such as CN, CO and hydrocarbons with relatively low energy (and vacant)
antibonding orbitals, less particularly to the ligand atoms with relatively low energy
(and vacant) d orbitals (those which do not belong to the first short period).
We refer to the A , B dichotomy described by Chatt et aL91 and the more general
division into hard and soft acids or bases developed by Pearson. Oxidation
numbers are low, in the most typical cases being zero and very often negative.
In a purely formal sense, this follows from the above, in that the ligand can accommodate some fraction of the formal negative charge on the metal. More
exactly, n* molecular orbitals in the complex will always be lowered in energy by
the presence of such ligands. Coordination numbers are generally low and covalent contributions to the total metal-ligand bond appear to be high.
These four characteristics find expression in a symbiotic principle, that is to
References pp. 52-55

26

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

say, that ligand-metal association tends to form well-matched systems in the sense
that B class ligands combine preferentially with B class metals and even promote
association with other B class ligands. The main virtue of a B class ligand appears
to lie in its minimizing interelectron repulsion, so promoting all of the above
characteristics. There is an independent criterion of this effect, namely the position
of the ligand in the nephelanxetic series94. We will here be concerned primarily
with organometallic rather than coordination compounds, generally, that is,
uncharged complexes with weakly basic ligands in the Brsnsted sense. Thus neither
complexes nor ligands will require a particularly polar solvent environment which
might interfere with kinetic interpretation. Against this advantage must be set,
firstly, the ambiguity which may arise in the oxidation number of the metal and,
secondly, the possibility that ligand reaction or rearrangement may occur without
normal substitution on the metal, and this may dominate the observations. The
stereochemistry of organometallic compounds is more varied9 than that of coordination compounds, and it is evident that more unorthodox substrates are open
to investigation. However, the kinetic information available to date concerns
compounds of conventional, especially octahedral and tetrahedral, stereochemistry. The greatest body of information concerns the carbonyl complexes - in
a sense these may perhaps be characteristic of the whole group of B type compounds, as the ammines have been conveniently taken as characteristic of the A
type compounds.
Some initial assessment of the metal-carbonyl bond may first be made. The
simpler neutral binary carbonyls fall conveniently into three eighteen-electron
classes, uiz. the octahedral chromium group hexacarbonyls, the trigonal bipyramidal iron group pentacarbonyls and tetrahedral nickel tetracarbonyl. In
addition, dimers or polymers with metal-metal bonds are formed by the elements
of groups VII and VIIl while carbonylate anions and some cations are fcrmed
which are structurally akin to the three basic monomeric types. Table 8 indicates
some of these formal relationships between the different carbonyls. Replacement
of CO by other ligands, anionic such as the halides or neutral such as the group V
ligands, hydrocarbons and the like, or a change in the charge on the complex or
a change in its geometry will result in a change in the nature of the M-C bond.
This has been adequately reviewed96: briefly, the total M-C bond owes its peculiar
TABLE 8
SOME B I N A R Y C A R B O N Y L C O M P O U N D S

C L A S S L I G A N D S : THE B I N A R Y C A R B O N Y L S

27

stability to two synergic contributions9, firstly, a a-bond using a predominantly


sp orbital of carbon and, secondly, a n-bond using a 7c*, predominantly antibonding c-0, orbital. In very similar systems the force constant of the c-0 bond
might be expected to be inversely related to that of the M-C bond simply because
a greater contribution of metal to ligand 7c bonding (an increase in M-C bond
order) necessarily results in a smaller net 7c contribution to the C-0 bond. A
priori, there is no good reason why force constant should relate to bond strength
in other than very similar systems, but a like evaluation may be applied cautiously
to the substituted carbonyls and, indeed, is a very useful method of monitoring
directional or other effects of carbonyl substitution. For the moment, the three
series of isoelectronic carbonyls in Table 9 demonstrate fairly adequately that increase in the metal negative charge or decrease in its oxidation number increases
the M-C bond order. Again, there is no reason why a low bond order should
necessarily relate to a high rate of substitution unless the activation step is one of
essentially complete dissociation.
TABLE 9
z~co(cm-*)F O R

SOME CARBONYL COMPOUNDS

Fe(C0)42I786
Mn(C0)51895
I863
v(c0)6I859

Cr(CO)6
1985

cO(c0)41886

Ni(C0)4
2057

Fe(CO)S
2034
2014

Mn(CO)6+
2090

MO(CO),
I990
W(C0)6
1980

* Data taken from refs. 98, 99,

100.

The ligand replacement reactions in the B type complexes superficially resemble


those of the A type, with one important difference, the frequent appearance of a
second-order term. In many cases thermal exchange or substitution in the binary
carbonyls is a first-order reaction, whether it takes place in the gas phase or in
solution. Only octahedral V(CO), , tetrahedral Ni(CO), and closely allied
CO,(CO)~exchange relatively rapidly 03. V(CO),, the only example of a paramagnetic binary carbonyl, must fall into a special category, for, if metal to carbon
7c donation does contribute some part to the overall bond in the carbonyls, then
V(CO), is the only one which in this respect falls short of a full complement of
six cl n electrons. The very rapid exchange of all COs in CO,(CO)~may take place
via a partial one-ended dissociation of a bridging CO together with a rapid
Rejeretrcea pp. 52-55

28

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

randomization of the remaining non-bridged ligands'O'-' 0 2 . This is compatible


with an interpretation of the results for Fe,(C0)12 and the structurally similar
Co4(C0),, (ref. 105) wherein all CO's exchange at the same rate",.
For
Mn2(CO),oand Fe(CO),, containing no such bridging CO's the rates of exchange
are exceedingly slow.
The rapid first-order exchange rate'" for Ni(CO), is tc be contrasted with the
much slower rates', for the carbonylate anions Co(CO),- and Fe(C0)42-,
and this is quite in accordance with a higher M-C bond order in these two latter
compounds, provided that the primary reaction step is a dissociative one. The
group VI hexacarbonyls exchange CO several orders of magnitude more slowly
than Ni(CO), by a first-order process in the gas phase' 0 6 , 07. There is very good
evidence that the photochemical substitutions of these hexacarbonyls proceed
via a dissociative step' 09. The thermal substitution of Mo(CO), with aromatic
and olefinic ligands' follows first-order kinetics, but at much higher concentrations of the more nucleophilic amines and phosphines"
a second-order term
becomes apparent. The slow exchange", of CO in Fe(CO), i s very much accelerated in aqueous acid1I8 and this is considered to be due to formation of
[Fe(CO),H]+. Either the labilizing influence of the H ligand is responsible for
increasing the rate of substitution or else the positive charge on the complex,
reducing the M-C bond order, has this effect (for an alternative interpretation,
see p. 38). In any case, substitution of the CO in Fe(CO), is also an exceedingly
slow reaction"'. Although a second-order dependence in Ni(C0)4 substitution
has not been detected, yet the first-order rate coefficient (triphenyl phosphine as
entering ligand) does show a strong dependence upon the nature of the ~ o l v e n t " ~ .
Generally, saturated solvents lead to slower reactions, higher activation enthalpies
and higher entropies. Table 10 gives some typical values for rate parameters of
exchange and substitution, first-order for the tetrahedral carbonjls, first and secondorder for the octahedral carbonyls.
It has been pointed out that the types of solvents which are used here, are not
generally such as would enter into strong association with the substrate. The
molecularity of the substitution reaction may then stand more chance of being an
operational concept. Amongst the binary carbonyls, the only systems which have
been extensively studied have been nickel tetracarbonyl and the hexacarbonyls
of group VI. For the former, the observation of a first-order rate is at least consistent with a rate-determining dissaiation of one carbonyl ligand followed by
reaction of the intermediate with whichever nucleophile should be available.

'

' 9

Ni(CO),
Ni(CO),

Ni(C0)3 CO

+ L fast_ Ni(CO),L

( 24)

The reaction has been considered'08 as an ideal, or limiting S,l dissociationii6,

29

CLASS LIGANDS: THE BINARY CARBONYLS

T A B L E 10
RATES OF THERMAL SUBSTITUTION I N SOME BINARY CARBONYLS

M(CO),+ L+M(CO).Part 1. Rate

Corrtpoiind

1L-t CO

k,[M(CO).]

C'80
(C6H5)3P
(C6HshP
(C6H 5)3p

14co
14co
14co

1 4 ~ 0

14co

Benzene
Mesitylene
Toluene
Hexamethyl
benzene
Cycloheptatriene
Norbornadiene
(n-B~t)3P
(EtOhP
(C6H 5 )3p

Solvent

Hexane
Hexane
Cyclohexane
Toluene
HzO
Hz0
Gas
Gas
Gas
n-Decanecyclohexane
n-Decanecyclohexane
n-Decanecyclohexane
n-Decanecyclohexane
n-Decanecyclohexane
n-Decanecyclohexane
Decalin
Decalin
Decalin

Temp. ( " C ) k,(sec-')

AH:(kcal. ASI(eri)
mole- I )

0.5
0.5
25
25
60
60
117
116
142
I I2

2.45 x 10-4
24.3
2.45 x 10-4
24.2
6.87 x 10-3
26.6
1 . 9 4 ~ 1 0 - ~20.4
very slow
very slow
2 XIO-5
38.7(E,,)
7.5 X I O - 5
30.8(E,,)
2.6 x I O - ~ 39.8 (E.)
2 XIO-5

1 I2

2.75 x 1 0 - 5

I I2

Ref.

108
108
113
113
130
130
106
107
107
110

110

1 . 7 4 1~0 - 5

110

112

1 . 3 4 10-4
~

110

106

5.6 x 1 0 - 5

30.0(E,)

110

110

112
130.7
112.0
165.7

XIO-4

14
13
20.9
2.0

1.38 x 10-4
2.13 x 10-4
1.15 X

40.2
31.7
39.9

22.6
6.7
13.8

10-5
8.54~

25.5
21.7
29.2
-

- 14.3
- 14.9
-6.9

112
112
112

Part 2. Rate = kZIM(CO).IIL]+kl[M(CO),I


~

Decalin
Decalin
Decalin
Decalin
Decalin
Decalin
Decalin
Decalin
Decalin

130.7
112.0
165.7
130.7
112.0
165.7
130.7
112.0
165.7

2.05 x 10-4
7.1 XIO-5
4.5 X I O - 5
6.96 x 10-4
1.7 X I O - 4
4.31 x 1 0 - 5
1.77 x 10-4
8.88 x

112
112

112
112
112
112
112
112
112

leading to either exchange or substitution, in competition but summing to the


same total rate. There are for example strong correlations between rates and activation parameters for the CO exchange and the substitution. However, it has been
shown113that there is also a strong solvent dependence of the first-order rate coefReferences pp. 52-55

30

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

ficient for triphenylphosphine substitution into nickel tetracarbonyl, from which


considerable contribution of solvent to the transition state may be inferred. Those
solvents (electron-releasing, aromatic) which promotc the reaction are those which
generally form the most stable complexes of, for example, the group VI metals
and which, in Ni(C0)4, decrease the CO bond order114. There are no strong
arguments for either dissociative or associative activation modes. On the one
hand, there is no evidence for the required intermediate Ni(CO), and on the other
it is difficult to envisage a direct attack upon the Ni in Ni(C0)4. There is, however,
an alternative possibility presented below (see p. 31).
In the case of the hexacarbonyls, the rate-expression contains not only the same
type of first-order term but in addition one second-order overall. For good
entering groups (but not CO, for example) the rate expression contains a term
strictly first-order in both the complex and the entering nucleophile. The firstorder rates of CO exchange are practically identical with the rates of substitution
in hydrocarbon solvents, but there is nevertheless some acceleration in ether
(THF, dioxan) solutions. This solvent-dependence is not so well-marked1 as
in the case of nickel tetracarbonyl. The second-order rate of substitution very
strongly depends upon the basicity' ' of the entering nucleophile

'

These observations are, indeed, consistent with an associative activation,


generation of a seven-coordinated intermediate (easier in the case of Mo and W
than for Cr because their larger sizes produce less steric hindrance) by attack
taking place directly upon the metal atom, that is, with an A or a limiting S,2
mechanism116, accompanied by a reaction sequence involving dissociative activation similar to scheme (24) above, oiz.

The experimental results are equally consistent with an initial attack taking
place upon one of the coordinated ligands. A faster nucleophilic attack at the metal
by stronger bases is unexpected in view of the fact that these are electronically
saturated compounds. For this reason base attack upon the coordinated CO should
not be excluded from consideration as a possible mode of activation, and it is
worth noting here that oxygen exchange but not CO exchange occurs' in aqueous
solutions of Re(C0)6f. There will be available empty orbitals, largely located on
the ligands and electron donation into these will promote a loss of the ligands.
In effect, it is likely that the C of the CO ligand bears a net positive charge and as a

''

CLASS LIGANDS: THE SUBSTITUTED CARBONYLS

31

result is subject to nucleophilic attack

In the general case, an incoming nucleophile would be expected to be favoured


by (i) a high basicity consistent with (ii) a high polarizability, and the metal
complex to favour its approach if (iii) it contains electron-acceptive, or B class
ligands. An interpretation of the available data may be essayed on these lines.
The infrared data upon Ni(C0)4 are consistent with a weakening of the C-0
bond114, and it would be of interest to examine the solvent effect upon the Ni-C
bond.

7. Complexes with B class ligands: the substituted carbonyls


There has been considerable success in accounting for the grosser changes in
the carbonyl bond character upon substitution by evaluating only the relative
n-bonding capacity of CO and the substituting ligands. Analyses of IR spectral
changes"' allowed assessment of the contribution of these substituting ligands
to the limited n-electron content in certain of the carbonyls. The methods have
usually ignored the inductive (or a-bonding) dependence of the spectral changes
and to a large measure this appears justified when limited to the tetrahedral
complexes, where n or a interactions between all ligands must be equivalent, to
a first approximation. It is not, however, justified in the case of the octahedral
complexes, for in this geometry there will be a distinct difference between the
mutual n and a interactions of cis and trans ligands. An evaluation of the n
and a contributions to total bond-order In a series of octahedral carbonyl complexes' has shown that these are indeed cooperative rather than complementary
in proportion, and that good n-accepting ligands, such as PF,, NO, CO (the
B class ligands), are ips0 fucto better a-donors. Conversely, the poor n-accepting
ligands, amines, ethers (the A class ligands), are not exceptionally good a-donors
either in these types of compounds. The intermediate ligands may be placed in a
rational sequence of n-acceptance (PX, > P(OR), > PR, > NR,). Since the
ligand-metal bond which is eventually to be broken upon reaction will be extended
in the transition state even of an A reaction it might perhaps be expected that abonding, being less distance-dependent than n-bonding, will contribute relatively
more to a transition state than to a ground state. Accordingly, arguments based
upon bond-orders which have been derived from average force-constants can only
be used with reticence, particularly with regard to any supposed dissociative
reactions.
It may be concluded fairly generally, though, that COYvirtually at the head of

''

References pp. 52-55

32

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

the n-accepting ligand series' 2 2 , should usually gain, thermodynamically, upon


substitution of CO by a ligand lower in the series. In Ni(CO),, say, the tetrahedral
arrangement ensures that a maximum of two d n orbitals may be mixed in with
the n* ligand orbitals. Up to two CO groups are readily replaced; thereafter
further substitution is difficult. In the case of the octahedral carbonyls, up to
three CO's may be so replaced, for there are now three d n orbitals of suitable
symmetry. It has been pointed out, though, that n-acceptance by the CO ligand
is overall possibly less than has been supposed'23 and in any case the argument
is less strong for M(CO), than for M(C0)6 since d-n* overlap is likely to be less
effective for the tetrahedral than for the octahedral corn pound^'^^.
Kinetic experiments have largely been limited to series of compounds based
upon the simple tetrahedral and octahedral structures of nickel tetracarbonyl and
the group VI hexacarbonyls respectively. Replacement of CO by neutral ligands
produces substrates of use in the assessment of directional and non-directional
influences without introducing the complications of altered oxidation number,
charge type, etc. Replacement of CO by NO produces the same net effect as a
reduction to carbonylate anion but without altering the charge, and the resulting
nitrosyl carbonyl may be looked upon as a carbonylate with, attached, the very
electron-acceptive NOf ligand. The replacement of CO by halide has the opposite
effect, tantamount to oxidation of the metal and generation of a carbonyl halide
isoelectronic with its left-neighbouring carbonyl in the Periodic Table. Olefins
behave as neutral, exceptionally polarizable (B type) ligands, and aromatic
compounds (similar ligands) form those exceptionally stable products in which three
CO ligands have been replaced by a single arene. Table 11 shows some of the
relations between groups of carbonyl derivatives.
Exchange and some substitutions in Ni(CO), have been shown' O8 to proceed
by first-order processes with virtually the same activation parameters, consistent
with a rate-determining dissociation. In the series Co(CO),NO through Mn(C0)
T A B L E 11
S O M E RELATED TYPES OF SUBSTITUTED CAKBONYL COMPLEXES
c

Cr(C0)6

Mn(COk+

Mn(C0LNO

Fe(CO)5

Mn(CO)(N0)3 Fe(C0)2(N0)2
and CO(CO)~NONi(CO)b
CO(CO)~(NO)L

ArCr(CO)3

*
Ar

CpMn(CO),

L = neutral monodentate ligand; L-L = neutral bidentate ligand; X


aromatic ligand; Cp = cyclopentadiene.

Ni(CO)3L
Ni(C0)2L2

halide ligand;

33

CLASS LIGANDS: THE SUBSTITUTED CARBONYLS

(NO), the 14C0 exchange is again first order although there is some possibility
of a confusing gas-phase reaction for the former"'. A first-order reaction between
Co(CO),NO and (C6Hs),As apparently occurs, but the remaining tetrahedral
nitrosyl carbonyls (the pseudo-nickel carbonyls) react according to a second-order
rate law. Tables 12 and 13 contain some rate data for these nitrosyl compounds
and show that the rate coefficient generally increases with the basicity of the
entering ligand. Those rates of CO exchange which have been measured are slower
than the second-order substitutions except for the doubtful case of Co(CO),NO,
and it has appeared reasonable to conclude that the substitutions are associative.
There are, indeed, many examples known of pentacoordinated complexes containing B type ligands"'. In the nitrosyl carbonyls, the formal oxidation number
falls: Co( -I), Fe( - 11), Mn( - 111), and it has been argued that this would cause
an increase in the M-C bond order and so reduce the possibility for dissociative
reaction. The decrease in rates in the series (n-But),P > (C,H,),P > (C6H,),As
is consistent with either smaller basicity or, equivalently here, polarizability.
T A B L E I2
(1. mole-'sec-I)

SECOND-ORDER RATE COEFFICIENTS

M(CO)4-,(NO)x+L

M(C0)3-,(NO),L+CO

9 x lo-'
2.6 x 10>200

CO(CO)~(NO)~
Fe(CO)Z(NOLa
Mn(CO)(N0)3b

FOR T H E REACTIONS

10-3
10-3
-54

2.3 x

125
126
127

slow
-5.6

at 25 "C; at 22 "C.
T A B L E 13
SOME DIFFERENCES I N KINETIC BEHAVIOUR BETWEEN

Co(CO),(NO)*
L

(C6H5)3AS (0.1 M )
(C6HshAS (0.1 M )

* Data from ref.


References p p . 52-55

170.

AT

F e ( C O ) 2 ( N O ) zA N D

25 "C. M(CO)4-,(NO),+L -+ M(CO)3-,(NO),L+CO

Solvent

Fe
co
Fe
co
Fe

co

toluene
toluene
THF
THF
toluene
toluene

Fe
co

toluene
toluene

kl(sec-')

0
0

k2(l.rnole-'.sec-')
10-3

10-3

10-3

0
0
0

1.6X
1.3 x 10-3

5.6 x 10-3
4x

kabr

km

1.5 Y l o y 6 sec-'
sec-'

34

I N E R T COMPLEXES A N D M E T A L O R G A N I C C O M P O U N D S

In the case of the nickel compounds, further substitution which is first-order


takes placelZ0
Ni(CO),LL+CO

(27)

+L + Ni(CO),LL + L

(28)

Ni(CO),L+L

-+

or
Ni(CO),L,

While the rate is independent of both the concentration and the nature of L, it
varies with L as: L = Cl,P > (C6H5),P > (n-But),P > (C,H,O),P
>
> (C6H,0),P. The decrease in rate parallels an increase in basicity (a-donation)
and/or a decrease in nephelauxetic effect (n-acceptance) with the exception, that
is, of the phosphite, which appears an anomalously stable compound on both
counts.
Further second-order substitutions can take place also in the cobalt series of
compounds
Co(NO)(CO),L
L

+ L -,Co(NO)(CO)L, + CO

(CH,O),P > (n-But),P

>> (C6H,),P

or
Co(NO)(CO),L

+ L -+

Co(NO)(CO),L

+L

(n-But),P > (C6H5),P > (C6H50),P

L = CH,C(CH,O),P

> (CH,O),P > (C6H5),P

(30)

Generally, the more basic and more polarizable nucleophiles enter into the
complexes more readily, while, as in the case of the nickel compounds, initial
presence of these same ligands aids the reaction. Steric hindrance in entering or
already present ligands markedly reduces the rates. The best entering groups
appear to be small, polarizable bases in which a-donating power appears to be of
more immediate importance than polarizability, while the most readily labilized
complexes are those of the same types of ligands in which lo^ polarizability
becomes more important. These two criteria are no more than general indications
however. It has been previously mentioned that, in a tetrahedral complex, the
presence of the NO ligand could quite readily reduce any dissociative contribution
relative to associative, and this could be explained quite adequately in terms of
an increase in the M-C bond order; this is supported by IR data with regard to the
C-0 bond order. However, it would not be expected to increase the (nucleophiles)
associative contribution by, in effect, reducing the positive charge upon the metal.
An alternative explanation may be that the presence of polarizable non-labile

B CLASS L I G A N D S :

THE S U B S T I T U T E D C A R B O N Y L S

35

ligands is essential in generating the presumably trigonal bipyramidal associative


intermediate. The NO, or any other ligand of good n-accepting plus a-donating
ability, would then be in the position of a trans-labilizing group in square-planar
substitution. It should be ncjted here that second-order kinetics are observed*
for the rapid exchange of tertiary phosphines (L) with Ni(II)L,X, and Co(II)L,X,
(X = C1 > Br > I).
For the octahedral compounds replacement of carbonyl by other ligands has
effects upon substitution kinetics which are somewhat similar t o those for the
tetrahedral compounds but more apparent. Whereas for tetrahedral complexes
increased lability only fairly generally follows upon replacement of CO by the
better n acceptors, in the case of the octahedral complexes a fairly distinct division
may be made between the strongly labilizing substituents (which are A ligands)
and the non-labilizing (which are B ligands). Table 14 shows some of the types of
ligands which fall into these two groups.
Considerable investigation of the octahedral carbonyl complexes has been
carried out. To a certain degree this is because definitive evidence for associative
substitution in the case of type A complexes has been conspicuously lacking
whereas for the type B compounds there seem to be several well-substantiated
examples. A general summary of the main types of octahedral substitutions which
have been kinetically examined is given in Table 15.
T A B L E 14
SOME OCTAHEDRAL COMPOUNDS WHICH UNDERGO SLOW* SUBSTITUTION

M(CO)sX+L -+ M(C0)4LX+CO

M
Mn
Mn
Mn
Cr
Mo
W

Ref.

14co

130
131
132
107
110
107

(C6HS)JP, CsHSN
14co
1 4 ~ 0

R3P
14co

SOME OCTAHEDRAL COMPOUNDS WHICH UNDERGO FAST* SUBSTITUTION

M(CO)5X+L -+ M(CO),LX+CO or M(C0)4X2+2 L, L-L

Mn
Mn
Cr
Mo
W

Br
H
CI, Br, I
cis byridine),
dipyridyl

~~

Relative to the parent reaction, X = L = CO.


References pp. 52-55

L, L-L

--f

M(C0)4L2, M(CO),(L-L)-I2 x
Ref.
133
I34
135
136
137

36

v1

z
3
0

z
0

>.

-1

z
0

a
4
u
a

-I

0
w
E

c
i,

0
w

2
I&

0
I
n

z
L,

0
4

z
3

0
c
c
v1

v1

m
3

y1

.-

s-

Y
4

L UM

ru u
IY

uJ

7
N

d
0-

;I

o\

ci
W

m
W

CLASS LIGANDS: T H E SUBSTITUTED CARBONYLS

z,
0"
z+

2-1
ReJerences ,up. 52-55

v,

37

38

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

The two-term rate law for substitution reactions of the group VI hexacarb ~ n y l s ' ~has
' been previously mentioned (see p. 29) and it will be useful to
summarize the evidence for associative activation in this case.
i. There is reasonably good agreement between the rate of 14C0 exchange in
the gas phase and the jrirst-order rate of substitution in decalin, suggesting that
this term represents a dissociative reaction.
ii. AH', is significantly higher for Cr than for either Mo or W, and this is in
agreement with the generally observed order of stability for those group VI
compounds which are seven-coordinated.
iii. ASs is positive for the first-order rate, negative for the second-order and these
values are expected for, respectively, dissociative and associative activations.
Substitution in the pentacarbonyl manganese and rhenium compounds
(XM(CO),, where X is formally anionic) is invariably more rapid than in the
M(CO),+ cations. The reactions are complicated by the further replacement
steps but, in the first instance, produce cis Mn(CO),XL at a rate independent of
the nature and concentration of the entering ligand (i.e. first-order) and largely
determined by the nature of X','. Usually, the rate of substitution is very much
faster than the exchange in M(CO),+ and when X is a halogen the rates are
d e m ~ n s t r a b l y ' ~ ~ .the same. Ligands which particularly labilize the CO's are
those of type A, of relatively high polarizing power but low polarizability, that is
the halides and NO3-, while no significant acceleration is noted for the type B
ligands. There is one especially interesting exception, for HMn(C0)5 and
HRe(CO), both undergo extremely rapid I4CO exchange. The same effect is
noticed in HCo(CO), and HFe(CO),+ and it has been suggested'54 that the
extreme labilizing ability of H, which is unexpected since H is an extreme type B
ligand', lies in its being able to take part in an insertion (migration) reaction'53
HMn(CO),
(HCO)Mn(CO),+CO

s (HCO)Mn(CO),
e (HCO)Mn(CO),

(31)

For the group VI pentacarbonyl halide anions a first-order substitution of carbon


monoxide by phosphine is also observed'35 to produce cis M(CO),LX-, the rate
of the reaction decreasing (C1 > Br > I) with the non-labile ligand in the same
way as for the Mn, Re carbonyl halides. In this case there is a concurrent secondorder substitution of the halide to produce M(CO)5L. It should be noted that
the order C1 > Br > I is unexpected if a simple D process is the mechanism of the
first-order reaction, for, on the basis of n-bonding capacity (I > Br > CI), the
M-C bond strength should be the greater for CI than for Br, I. Again, n-bonding
arguments rarely give satisfactory predictions concerning reaction rates (see p. 26).
Thus
M(CO),X-

+ A,P

-+

M(CO),(A,P)X-

+ CO

(32)

First-order: M

C L A S S LIGANDS: THE SUBSTITUTED C A R B O N Y L S

Mo > Cr > W, X = Cl > Br > I

M(CO)5X- +A3P + M(CO),(A,P)+XSecond-order: M

39

W, Mo > Cr. X

(33)

Br > CI, I

It is surprising that the second-order rate coefficients decrease as the basicity of


the entering ligand increases; it appears that ligand nucleophilicity is compounded
of basicity and micropolarizability in no simple manner.
The products of this first stage of the reaction, cisMo or Cr(CO),LX- and
cis Mn or Re(CO),LX undergo further substitution. The Mn, Re compounds
behave according to first-order kinetics while the Cry Mo compounds show
second-order kinetics to form trans Cr(CO),L, or cis Mo(CO),L, , rate coefficients increasing this time as expected, with the basicity of the entering ligand L.
Complexity in the manganese and rhenium pentarbonyl halides substitution
arises from the fact that these decompose in inert solvents to form the halogenbridged dimers [M(CO),X], . Both monomers and dimers react with phosphines,
arsines, pyridine, aniline etc. to give the disubstituted compounds M(CO),XL,.
Thus three final products or any mixture of them may be obtained, depending
upon the precise reaction conditions

M(CO),LX

(35)

M(CO),X - co cis M(CO),LX

(36)

(M(CO),X), $ c i s

M(CO),LX

- co

cis M(C0)3L,X

(37)

One factor which promotes all reactions is the electronegativity of X(C1 > Br > I),
although an extensive study of reaction (34) has not been carried out. Release of
CO, reaction (37), is greatly accelerated when L is a basic or a N-donorligand,
(n-But),P rather than (C6HS),P and aniline rather than CNC6Hs, (n-ButO),P.
For a given donor atom, the rate increases with the bulk of the remainder of the
ligand. The reactions
Mn(CO),X

+ cis Mn(CO),LX

(or Mn(CO),LLX)
R e f i ~ t i c epp.
s 52-55

40

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

are first-order1. 3 3 However, the manganese dimers [Mn(CO),X],


cleavage
[Mn(CO),X], +4 L

2 Mn(CO),XL,

undergo

(39)

where L = pyridine or substituted picolines, according to a two-term rate law,


and the corresponding rhenium dimers [Re(CO),X], with the same ligands
(plus (C,H 5)3P)follow second-order kinetics to yield Re(CO),XL.
The simplest overall interpretation of these data is in terms of a rate-determining
dissociation. Entropies of activation are positive and the solvent-dependence for
a better leaving group (CI) is less marked than for a worse one (Bi) in the case
of reaction (38)14. For the dimeric carbonyls, [M(CO),X], , bridge-breaking,
essentially the same dissociation, could result in a rapid pre-equilibrium. If this
were followed by a second dissociative step, then the kinetics could be first-order
(as for Mn), while a rate-determining entry of L could produce second-order
kinetics (as for Re).

There is no real indication of the part which the solvent plays i n these reactions.
The intermediate generated in the fast pre-equilibrium (40) may, for example, be
more or less solvated (as also the species M(CO),X). The effects of varying the
solvent are small, however.
In the case of replacement of CO from the group VI carbonyl compounds there
is additional evidence to the effect that the type A ligands labilize CO whereas the
type B d o not, but rather promote a second-order reaction. For the group VII
octahedral compounds there is no strong evidence in favour of an associative
activation step, except when interpretation is obscured by subsequent or concurrent
reaction (but see ref. 146). There is, however, good reason to believe that such
an associative reaction does occur in certain of the group VI compounds.
The rate of replacement of CO by phosphines and phosphites
M (CO)5L+ L

+ cis

M(CO),LL

+ CO

(43)

does not generally differ greatly from that in the unsubstituted carbonyl when
L = R,P, (RO),P or CNC,H5110. When L = halide or pyridine, that is a type
A ligand, the substitution is very much faster135but it is still first-order, independent of concentration and nature of the entering nucleophile. There ist indeed, a

B CLASS LIGANDS:

THE SUBSTITUTED CARBONYLS

41

second-order halide substitution accompanying CO substitution but the seemingly


important point is that isocyanide, phosphine or phosphite d o not have a labilizing
influence upon the CO whereas pyridine and the halides do so. This is more
convincingly brought out by the reactions which M(CO),B compounds undergo,
where B is a bidentate ligand, uiz. a substituted orrho-phenanthroline, substituted
dipyridyl, 1,2-diamin0-2-methyl-propane,
2,5-dithiahexane or 1,5-cyclooctadiene.
Along this series we see a fall in group basicity and an increase in polarizability.
At one extreme, the o-phenanthrolines are harder ligands than CO, while at
the other, the di-olefin is of comparable softness. This series might perhaps be
extended to the sets of n-(aromatic ligand) MO(CO),~ and 1,3,5-cyclo-heptatriene
M(CO), compounds in which the trend continues to completely second-order
kinetics and replacement, not of CO, but of the hydrocarbon.
When the bidentate B is dipyridyl or a substituted o-phenanthroline and M is
Cr, only a first-order substitution is observed
Cr(CO),(N-N)+L

cis Cr(CO),L(N-N)+CO

(44)

-3.2-

-34-

-3.6-

4.0

50

6.0

P Ka

Fig. 9. Linear free-energy relationship for the reaction


Mo,W).
~ +Cr,
CO
(M
M(C0)4 O - ~ ~ ~ ~ + P ( O C H ~+
) ~M(CO)3
C C H ~O - ~ ~ ~ ~ P ( O C H ~ ) ~ C C H =
Log k , us. pK. o-phen.
References p p . 52-55

42

I N E R T C O M P L E X E S A N D METAL O R G A N I C C O M P O U N D S

Fig. 10. Linear free-energy relationship for the reaction


~ C C H ~= Mo, W).
M(CO)., o-phen+P(OCH2)3CCH3--f M(CO)3 o - ~ ~ ~ ~ P ( O C H , ) +CO(M
Log k2 us. pK,, o-phen.

The same first-order replacements are seen when M is Mo or W, somewhat slower


than in the case of Cr, but still much faster than for the hexacarbonyls. The rate
increases with the pK, of the inert ligand (N-N) and Fig. 9 shows the linear freeenergy relation between log k, and pK,. The relative orders would not have been
expected on the basis of any n-bonding effects since increasing back-donation to
CO would increase the M-C bond order. This increase in M-C bond order is
supported by a decrease in vco with increasing o-phenanthroline basicity. The
same consideralion applies for the pentacarbonyl halide anions where the firstorder rates decrease (C1 > Br > I), unexpectedly as the halide polarizability
increases.
When M is Mo or W, the substitution reaction follows a two-term rate expression and in addition there is some replacement of the dipyridyl or 1,Zdiamino2-methyl propane but not of o-phenanthroline. The proportion of bidentate
replacement to CO replacement is practically independent of the concentration

THEORETICAL CONSIDERATIONS

43

of the entering group41. There is now a decrease in the (second-order) rate


coefficient with the pK, of the inert o-phenanthroline (Fig. 10) and an increase
with what may be judged to be the nucleophilicity of the entering phosphine or
phosphite.
It is apparent that the existence of a second-order term in the rate expression
does not of itself offer any proof of associative or dissociative activation, for there
are two possible alternative mechanisms compatible. These are:
(i) An association of L to give a seven-coordinated intermediate, followed by a
rate-determining loss of either CO or of (N-N), uiz.

(ii) A dissociation of one end of the bidentate ligand, followed by a ratedetermining entry of L, viz.

Alternative (ii) corresponds to the [Re(CO),X], case, equations (41) and (42)
above. However, it was here favoured largely because no second-order term
was observed for the Re(CO),X42 and Re(C0),LX15 substitution. In the case
of Mo(CO),Py, expected to be closely similar to Mo (CO),dipy, a second-order
dependence has been observed.

8. Theoretical considerations
In the final section of this chapter, we shall attempt to give a brief rationalization
of the regularities and peculiarities of the reactions of non-labile complexes which
have been discussed in the previous sections. The theoretical framework in which
the discussion will be conducted is that of molecular orbital theory (MOT).The
MOT is to be preferred to alternative approaches for it allows consideration of
all of the semi-quantitative results of crystal field theory without sacrifice of interest in the bonding system in the complex. In this enterprise we note the apt
remark56: Kinetics is like medicine or linguistics, it is interesting, it is useful,
but it is too early to expect to understand much of it. The electronic theory of
reactivity remains in a fairly primitive state. However, theoretical considerations
may not safely be ignored. They have proved a valuable stimulus to incisive
experiment.
Two remarkably successful generalizations from the preceding pages deserve
attention first of all. Octahedral complexes show a pronounced tendency to react
Re/erciices pp. 52-55

44

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

dissociatively. In contrast, square planar complexes show a very pronounced


tendency to associative reaction. If nothing else were operative, these tendencies
might be anticipated o n steric grounds alone, since the planar complexes have
two open faces for attack, these access routes being closed i n the octahedral
systems. It may be that this steric difference is of prime importance but there is at
least one significant and relevant difference in electronic structure.
Figs. 1 1 and 12 show typical MO diagrams for square planar and octahedral
complexes. Inspection reveals that the metal p z orbital ( z is the axial direction) in
a square planar complex is involved in the TC bonding system and available for D
bonding in the transition state. This is a feature shared by nucleophilic substitution at square planar complexes with the spectacularly associative nucleophilic aromatic substitutions. The octahedral complexes discussed in this chapter

Fig. 1 I . MO diagram for a square planar complex.

THEORETICAL CONSIDERATIONS

45

Fig. 12. MO diagram for an octahedral complex.

stand in sharp contrast. The metal p r orbital is part of the 0 bonding system in
the ground state and, as long as there are three or more d electrons the dn orbitals
are filled and not readily available for 0 bonding in the transition state. Orgel
noted several years ago that an associative mechanism would be quite plausible
if there were fewer than three d electrons. So far, there has been no definitive
evidence in support of this suggestion, although there are some indications in the
results of fast reaction kineticists on the behaviour of V(III)58~ When the
ligands can act as strong n acceptors then the situation may become modified.
n donors, that is, halide ions, amines, oxy ligands can only destabilise the dorbitals
of x symmetry. n accepting ligands on the whole may considerably stabilise these,
to a degree which depends markedly upon the matching (or lack of it) of metal
(dx) and ligand (n*) orbitals. If there is good matching, particularly in the cases
of CO and NO, to a less degree perhaps PF,, the filled orbitals will concentrate
into the internuclear regions leaving in the region of the ligand atom a net positive
References p p . 52-55

46

INERT COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

charge. The orbital system of the octahedral B-type complexes should be qualitatively similar to that of the square-planar d8 complexes.
Moving from consideration of general trends to specific phenomena, perhaps
the most fruitful problem for the theoretician of ligand substitution is the remarkable trans labilizing effect encountered in Pt(I1) chemistry. The molecular
orbital theory of this effect developed by Gray'60 serves admirably as a starting
point for a more general molecular orbital theory of the effects of the non-labile
ligands in a complex. If we appeal to the observation that the leaving group, the
entering group and the group trans to the leaving group play similar roles in square
planar substitutions, we may with reason postulate a transition state approaching
a trigonal bipyramid structure. The theoretical problem becomes comparison of
bonding in the square planar ground state and the trigonal bipyramid activated
state.
First the n bonding may be analyzed. Of the four metal valence orbitals involved
in strong n bonding in a square planar complex, only the p orbitals have trans
directional properties. The trans group and the leaving group must share the same
p orbital and a trans ligand with a strong n interaction with the p orbital must
weaken the bonding to the leaving group. Moreover, when the complex is converted
from the square planar form to the activated trigonal bipyramid an additional
p orbital (thep,) is involved in n bonding to the entering, leaving and trans ligands
(see Fig. 13). It follows that the trans ligand which had approximately a halfshare in a metal p orbital in the ground state may have approximately a twothirds share in a metal p orbital in this transition state. The energy difference
between ground state and transition state should be relatively small for good ligand
(n) to metal ( p ) donors located trans to the leaving group. Calcu1ationsl6' have
shown that ligands such as H - , PR, and CH,- have unusually large overlap
integrals with the Pt 6p orbital. It is probably safe to conclude that the large trans
effect of these ligands is a consequence of the n bonding effect described. Other
trans labilizing ligands must also depend for their effectiveness to a large extent
on this effect.
PZ

Fig. 13. Change in the metal p . orbital structure in square- planar substitution.

THEORETICAL CONSIDERATIONS

47

Some ligands may supplement this CT effect with one operating through the n
bonding system. Three d orbitals have proper symmetry for n interaction, xy, zx
and yz. Assuming coordinates chosen as before, it is clear that the x z and yz
interact only with a pair of trans disposed ligands in the square planar ground
state. For example, the x z orbital may be shared between the leaving and trans
ligands. In the trigonal bipyramid, four d orbitals are of n symmetry (xz, yz,
x2-yz and x y ) and all are shared in n interaction with ligands in the trigonal
plane, viz. the trans, entering and leaving ligands. NOW,in d 8 complexes (all of
the square planar cases we have discussed) all of the n antibonding orbitals derived
from the metal d orbitals are filled. In this case, it is a great advantage for the
trans ligand to have empty, reasonably low-lying ( n ) orbitals which can contribute
to delocalization of the charge in the antibonding n orbitals as a new ligand (the
entering ligand) is attached at the metal. Thus the effect of a good n acceptor
trans ligand is to lower the activation energy for the reaction. Table 16 gives
T A B L E 16
R E L A T I V E U-

A N D Z-trUns E F F F E C T S O F L l G A N D S

(See ref. 3, p. 27)


Estimnted a-effect order
H-

> PR3 > --SCN- > I-, CH3-, CO, CN- > Br- > C1- > NH3 > OH-

Estirnnted n-effect order


H2C = CH2, CO

> CN- >

-NOz-

> -SCN- > I- > Br- > C1- > NH3 > OH-

theoretical estimates of the CT and n trans effect orders as derived from overlap
integral calculations16'. It can be seen from comparison of these orders with the
experimental order of trans effect that both (T and n effects are important. Earlier
emphasis had been placed on n effects but it is important to notice not only that
a very large part of the experimental trans effect order can be discussed using the
CT order alone, but that some ligands (e.g. H-) absolutely require consideration
of the CT effect.
The MOT discussion of the trans effect immediately illuminates the part played
by entering and leaving groups. The ligand which functions effectively in the trans
position should do so also in the other two positions in the trigonal plane; the
entering and the leaving positions. There should be a parallel between the trans
effect series and both the order of effectiveness (nucleophilicity) of entering groups
and the order of inertness of leaving groups, and these correlations are, of course,
well known.
Now, consider leaving group orders in general. The order of inertness at Pt(I1)
(which corresponds to the theoretically comprehensible trans effect series) is
largely opposite to the order of ligands in the well-known spectrochemical seriesL6'.
References pp. 52-55

48

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

Those ligands which are strong field i n the sense of producing a large splitting
between the dn antibonding (xz, y z , x y ) and d a antibonding (x2-y, z) orbitals
of an octahedral complex are frequently ones readily displaced from square
planar complexes of Pt(I1). This circumstance is the opposite of that observed
in the octahedral complexes of Co(1II) and Cr(II1). In these latter two cases the
ligands high in the spectrochemical series are the ones which are least readily
replaced.
The spectrochemical series reflects the splitting of molecular orbitals derived
from the metal d orbitals and a large ligand field splitting is expected in the event
that strong a bonding occurs between metal d and ligand a orbitals. However,
strong a bonding, ligand to metal p orbital, does not make a direct contribution
to the ligand field splitting. In the n bonding framework of the complex the most
important effects are directly reflected in the ligand field splitting. Interaction of
filled metal dn orbitals with lone pair electrons on the ligands in orbitals of n
symmetry makes the dn orbitals antibonding. As this interaction increases in
importance, the ligand field splitting decreases. Conversely, for ligands with lowlying empty TC orbitals (e.g. the n* levels of CO and CN-) the interaction with
the dn orbitals of the metal lowers these levels and increases the ligand field splitting. We can expect, then, that this easily accessible spectroscopic parameter might
provide a general guide to the role of a ligand in d orbital bonding of both a
and n types but little direct information about the important matter of a bonding
involving s a n d p orbitals. More detailed analyses of optical and magnetic resonance spectra promise parameters offering more direct guidance concerning the role
of the s and p metal orbitals. Some recent work is suggestive124*162-164.
Returning to the leaving ligand orders, that order observed for the third transition series metal, Pt(II), could be explained on the hypothesis that changes in a
bonding to metal p orbitals are more important than changes in a bonding to d
orbitals as the leaving group is changed (this hypothesis is implicit in the trans
effect theory). Correlation of lability at Cr(II1) and Co(II1) centres with the
spectrochemical series suggests importance of d orbital a bonding for the lighrer
first transition series metals. Only one later period metal octahedral system has
been studied extensively enough for a definite statement concerning leaving group
orders. In the Rh(1II) complexes the lability order C1- > Br- > I - has been
established6. This is opposite to the Co(II1) order and opposite to that expected
from the ligand field splitting as discussed above. It does correspond, however,
to the lability order for the three ligands at Pt(I1) centres and suggests that this
second transition series metal reflects changes in a binding through metal p orbitals
in its lability order.
As the trans effect theory indicates, there should be some relationship between
lability of a ligand and its role as a labilizing group in another position in a
complex. In an octahedral complex reacting via a dissociative mode of activation,
the transition state has five strongly bound ligands. This state will be stabilized

THEORETICAL CONSIDERATIONS

49

if there should be present within the complex a group which binds strongly with
orbitals formerly utilized by the leaving group. That is, in octahedral as in square
planar complexes, a non-labile group should function to labilize other groups.
In a very rough sense the data given above on the systems Coen,ACl+ where A
is a variable group influencing the rate of loss of CI- could be construed to agree
with this proposition. The labilizing order was

CJ

There is perhaps a significant n bonding effect, yet to be considered, but there


is at least one more immediately obvious point revealed in the results. The leaving
group orders suggested a dominant role for d orbital CJ binding in Co(II1). do
orbitals have four-fold symmetry and there should be no great difference between
the influences of a ligand cis or trans to the leaving group. There should be no
general trans effect in Co(II1) or in Cr(II1) chemistry. However, there should appear a trans effect in octahedral Rh(II1) chemistry if p orbital discrimination
amongst ligands is now of great importance. Recent
166 ha s established
a distinct trans effect in Rh(II1) amine complexes. With several constant leaving
groups, the trans effect order is: I- > Br- > C1-, that is, as for Pt(I1). It has also
been observed7. 1 6 7 that the presence of an amine group trans to the leaving ligand
in a Rh(II1) complex results in hydroxide catalysed (base) hydrolysis but that an
amine in the cis position alone is insufficient. Apparently NH,- labilizes effectively only when trans to the leaving group. It is understandable, then, that extended
search for clear cut trans effects in octahedral complexes is finally bearing fruit
only as studies turn to the second and third transition series metals.
Having some theory of the o effect, we now consider n effects in the dissociative
mode substitution reactions of octahedral complexes. It has already been pointed
out that ligands which have filled orbitals of n symmetry interact with metal drr
orbitals making them function as antibonding levels. This will destabilize ground
states and make such ligands (e.g. C1-, OH,, OH-) rather better leaving groups
than might be anticipated from consideration of a CJ effect alone. Although this
effect should correlate with spectrochemical position it is difficult to distinguish
the role of CJ from n bonding in determining lability. When these ligands with
filled n orbitals function (as non-labile ligands) in modifying the reactivity of
other ligands they play a role opposite to that expected from their positim in
the lability series. This is because the same rr antibonding interaction which weakens
the bond favours the formation of a dissociative transition state. Although this
may be simply phrased in the suggestion that lone pair electrons (of n symmetry)
can be supplied to the increasingly electron deficient metal, a more precise formulation may be the correlation noted68 between the labilizing effect of such ligands
and their position low in the nephelauxetic series. Relatively, stabilization of the
transition state follows from reduction of the interelectron repulsions brought
about by donor ligands.
References pp. 52-55

50

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

Basolo and P e a r ~ o n ' ~were


'
among the first to point out that this particular
type of ligand will function as donor in a stereospecific fashion in the dissociative
transition state. Fig. 14 shows how an electron-filled p orbital on the ligand can
overlap directly with a p orbital being depleted by the departure of a cis ligand
without any rearrangement. On the other hand the same ligands trans to one
another cannot form orbital overlap of 0 orbitals unless there is rearrangement
toward trigonal bipyramid geometry as shown in Fig. 15. These considerations
are, in all probability, the explanation for the greater reactivity of cis CoenzAClf
as compared to trans CoenzAC1+ when A is C1-, Br- or OH-. The suggestion is
supported by the observation that it is the complex of this type that undergoes
isomerization in the course of acid hydrolysis. All cis isomers of the CoenzAClf
A

Fig. 14. The overlap of (1igand)p orbital with avacant cis (meta1)p orbital without rearrangement.

Fig. 15. The overlap of (ligand) p orbital with a vacant orbital upon rearrangement of the rrans
compound to a trigonal bipyramid.

THEORETICAL CONSIDERATIONS

51

system yield cis C O ~ ~ , A O H ,and


~ + of the trans Coen,ACl+ system only those
complexes with A = OH-, C1-, Br- and NCS- lead to any cis product.
Finally, it is important to turn to those ligands which are n acceptors, those with
low-lying empty orbitals that bond with metal dn orbitals. This interaction increases their ligand field splitting, advancing their position in the spectrochemical
series, and at the same time decreases lability by adding additional bonding above
the 0 bonding. It is again true that there is inverse correlation between position
in the spectrochemical series and lability. It is also again true that the n interactions
do not play the same role when the ligand functions to labilize another group.
To a first approximation, n bonding does not stabilize nor seriously destabilize
a dissociative transition state, and one expects such ligands (e.g. NO2-, CN-,
CO) to produce labilizing effects principally within the 0 bonding framework.
These could well be large (see the 0 trans effect series for Pt(1I) in Table 16) but
they will be smaller than anticipated from the ligands' spectroschemical position
or lack of lability. The present suggestion is supported by the fact that NO,- and
CN- as ligands A in Coen,ACl+ are more effective in labilizing C1- from the
trans position. It is not improbable that this small trans effect arises through 0
bonding with mLtal p orbitals. A theoretical assessment of 0 and n contributions
in octahedral systems (such as the Co(II1) complexes) remains to be made, yet
the lability and labilizing orders seem consistent with a theory emphasizing 0
bonding modifiied by some important n effects.
Some assessments of 0 and 71 overlap populations for the simpler carbonyl
systems have been made'24,164. For these, in which the metal is in an abnormally
low oxidation state it is probable that the main variable is within the n system
while the 0 system remains relatively constant for a given geometry. Details differ,
but i t appears that metal d to ligand n* overlap is more effective for the octahedral
than for the tetrahedral systems (for these latter, the metal p orbital is more effective). The most interesting single result is that, in the octahedral systems the
carbon site shows promise of some electrophilic character. The 7~ system is effectively removed from the metal such that at least one cogent argument against
nucleophilic attack directly at the metal is lessened, and it has been pointed
that nucleophilic attack directly at the carbon followed by an internal ligand
migration is not at all unlikely. It is worthy of note that this two-site concept has
recently been convincingly used'70 with regard to some of the anomalies in the
Fe(CO),(NO), substitution kinetics. There is a spectacular solvent effect for, in
THF or methanol, the relative orders of rates for nucleophiles are very different
from those in toluene or dichloromethane (see Table 13). Particularly, (C,H,),As
and CO, verypoor reagents in toluene, are much more effective in THF. In addition, nitrogen bases, quite ineffective for Co(CO),NO, are more effective for
Fe(CO),(NO), than are the phosphines, which are good ligands for Co(CO),NO.
One interpretation of these findings is in terms of initiating solvent or other Atype nucleophilic attack at carbon followed by substitution of CO. This will
References p p . 52-55

52

INERT COMPLEXES A N D METAL O R G A N I C COMPOUNDS

give rise to first-order kinetics if the solvent is a good nucleophile and secondorder kinetics if the entering ligand should be

M-C-O+S

M-C-0
I

M-C-0
I
S

+ B + M-I3 +CO +S

(slow)

(49)

(fast)

(50)

While it is too early to venture a generalization extending to all of the B-type


compounds, specifically why harder ligands appear to exert a labilizing influence in the octahedral compounds, these are some indications that it is not in
order to consider ligands in these B complexes merely as extensions of those in
the A compounds. Reactivity of the coordinated ligand, its rearrangement and
migration may prove to be an important feature.
REFERENCES
1 F. BASOLO
AND R. G . PEARSON,
Mechanisms of Inorganic Reactions, Wiley, New York,1967.
J . A m . C h e m SOC.,85 (1963) 3533.
2 R. G. PEARSON,
A N D H. B. GRAY,
LigandSubstitution Processes, Benjamin, New York, 1966.
3 C. H. LANGFORD
K. S. LAIDLER
AND H. EYRING,
The Theory of Rate Processes, McGraw-Hill,
4 S. GLASSTONE,
New York, 1941.
5 F. BASOLO,
B. D. STONEA N D R. G. PEARSON,
J. Am. Chem. SOC.,76 (1954) 3079.
A N D J. W. MOORE,
Inorg. Chem., 3 (1964) 1334.
6 R. G. PEARSON
AND R. G. WILKINS,
J. Am. Chem. SOC.,76 (1954) 3379.
7 A. W. ADAMSON
J. Chem. SOC.,(1940) 1490.
8 G . W. ETTLEAND C. H. JOHNSON,
9 C. H. LANGFORD
A N D W. R. MUIR,J. A m . Chem. SOC.,87 (1967) 3141.
10 K. B. YATSIMIRSKII
AND L. PANKORA,
Zh. Obshch. Khim., 19 (1949) 611;
K. B. YATSIMIRSKII,
Zh. Obshch. Khim., 20 (1950) 1408.
I I C. H. LANGFORD,
Inorg. Chem., 4 (1965) 265.
12 M. L. TOBE,Rec. Chem. Progr., 27 (1966) 79; Aduan. Chem., 49 (1965) 1 and references
therein.
13 H. TAUBE
A N D W. SCHMIDT,
J. A m . Chem. SOC.,80 (1958) 2642.
14 See the documented table in ref. I , p. 164.
15 H. R. HUNTA N D H. TAUBE,
J . A m . Chem. SOC.,80 (1958) 2642.
J . Chem. SOC.A , (1966) 1.
16 G. C. HALOR,
17 D. L. GAYAND G. C. LALOR,J . Chem. SOC.A , (1966) 1179.
AND M. L. TOBE,
J . Chem. SOC.,
18 M. L. TOBE,Sci. Progr., 48 (1960) 484. See also P. J. STAPLES
(1960) 4803;
M. E. BALDWIN,
S. C. CHANAND
M. L. TOBE,J. Chem. SOC.,(1961)4637;
S. C. CHANA N D M. L. TOBE,
J . Chem. SOC.,(1963) 5200.
19 M. L. TOBE,Inorg. Chem., 7 (1968) 1260.
20 S. C. CHAN,J. Chem. SOC.,(1963) 5137.
21 S. C. CHANA N D M. L. TOBE,J . Chem SOC.,(1963) 514.
AND M. L. TOBE,J . Chem. SOC.,(1963) 506.
22 C. H. LANGFORD
C. R. BOSTONA N D F. BASOLO,
J . Phys. Chem., 59 (1955) 304.
23 R. G. PEARSON,
C. R. BOSTON
AND F. BASOLO,
J. Am. Chem. SOC.,75 (1953) 3089.
24 R. G. PEARSON,
AND M. L. TOBE,Nature, 187 (1960) 477.
25 C. K. INGOLD,R. s. NYHOLM
26 Ref. 1, pp. 79-84.

REFERENCES

53

27 A. HAIMA N D H. TAUBE,
Inorg. Chem., 2 (1963) 1199.
Inorg. Chem., 7 (1968).
28 W. R. MUIRA N D C. H. LANGFORD,
A N D D. W. WATTS,
J. Am. Chem. Soc., 89 (1967) 81 5 .
29 I. R. LANTZKE
29aR. MURRYAND C. BARRACLOUGH,
J. Chem. SOC.,(1965) 7047.
Inorg. Chem., 1 (1962) 573.
30 A. HAIMAND W. K. WILMARTH,
AND W. K. WILMARTH,
Advati. Chem., 49 (1965) 31.
31 A. HAIM,R. J. GRASSI
R. A. PALMER
A N D L. M. BLAKELY,
J. Am. Chem. SOC.,88 (1966) 2877.
32 J. HALPERN,
AND H. TAUBE,
Inorg. Cheni., 5 (1966) 1376.
33 D. A. LOELIGER
I. I. OLSONAND A. M. SARGESON,
Inorg. Chem., 6 (1967) 1807.
34 D. A. BUCKINGHAM,
A ND J. LANG,
J. Chem. SOC.,(1963) 5620; S. C. CHAN,K. Y. HUI,J. MILLERA N D
35 G. C. LALOR
W. S. TANG,J. Chem. SOC.,(1967) 3207.
36 S. C. CHANAND M. L. TOBE,J. Chem. Soc., (1962) 4531; (1963) 514.
R. E. MEEKER
AND F. BASOLO,
J. A m . Chem. SOC.,78 (1956) 709.
37 R. G. PEARSON,
Nufure, 139 (1937) 507.
38 E. J. GARRICK,
39 Ref. 1, p. 177 ff.
40 F. BASOLO,
J. W. PALMER
AND R. G. PEARSON,
J. A m . Chem. SOC.,82 (1960) 1037.
AND F. BASOLO,
J. Phjs. Chem., 64 (1960) 788.
41 J. W. PALMER
42 M. GREENAND H. TAUBE,Inorg. Chem., 2 (1963) 948.
I. I. OLSONA N D A. M. SARGESON,
J. Am. Chem. Soc., 88 (1966) 5443.
43 D. A. BUCKINGHAM,
Advan. Chem., 62 (1967) 542.
44 R. D. ARCHER,
Inorg. Chem., 7 (1968) 1159.
45 G. W. WATTAND J. F. KNIFTON,
J. Chem. SOC.A , (1967) 917.
46 R. D. GILLARD,
S. V. PTITSYN
AND E. D. PASTAKOVA,
Russ. J. Iriorg. Chem., 10 (1965)
47 N. M. NIKOLAEVA,
574.
48 R. C. JOHNSON
AND E. R. BERGER,
Inorg. Chem., 4 (1965) 1262.
AND A. J. POE,J. Chem. SOC.A , (1967) 1633.
49 B. CORAIN
50 A. A. EL-AWADY,
E. J. BOUNSALL
A N D C. S. GARNER,
Inorg. Chern., 6 ( I 967) 79.
51 A. B. LAMBAND L. T. FAIRHALL,
J. Am. Chem. SOC.,45 (1923) 378.
T. P. JONES,W. E. HARRIS
A N D W. J. WALLACE,
J. Am. Chcm. Soc., 83 (1961 )
52 M. A. LEVINE,
2453.
53 A. OGARDAND H. TAUBE,
J. Ant. Chenr. Sor., 80 (1958) 1084.
AND G. C. LAMM,
Acta Chem. Scand., 9 (1955) 216.
54 J. BJERRUM
55 T. F. SWADDLE
A N D E. L. KING,Inorg. Chem., 3 (1964); 4 (1965) 532.
AND E. L. KING,Inorg. Chem., 3 (1964) 916.
56 F. A. GUTHRIE
AND W. C. REYNOLDS,
Inorg. Chem., 2 (1963) 468.
57 H. B. JOHNSON
AND C. S. GARNER,
J. Am. Chem. SOC.,83 (1961) 4152; Inorg. Chern.,
58 D. J. MACDONALD
1 (1962) 20.
59 S. BEHRENDT
AND C. H. LANGFORD,
to be published.
60 J. A. BROOMHEAD,
F. BASOLO
AND R. G. PEARSON,
Inorg. Chem., 3 (1964) 826.
Advances in the Chemistry of Coordination Compounds, Ed., S . KIRSCHER,
61 R. E. CONNICK,
MACMILLAN,
New York, 1961, p. 15.
62 H. B. LAMB,J. A m . Chem. SOC.,61 (1939) 699.
AND R. G. PEARSON,
Inorg. Chem., 3 (1964) 658.
63 J. V. RUND,F. BASOLO
AND
S. GARNER,
J. A m . Chem. Soc., 84 (1962) 2032.
64 J. A. POULSEN
G. C. LALORAND E. A. MOELWYN-HUGHES,
J. Chem. SOC.A , (1966) 719.
65 A. W. BUSHNELL,
66 M. PARRIS,J. Chem. SOC.A , (1967) 583.
A N D W. J. WALLACE,
Can. J. Chem., 39 (1959) 2371.
67 T. P. JONES,W. E. HARRIS
J. A m . Chem. SOC.,89 (1967) 816.
68 N. V. DUFFYAND J. E. EARLEY,
69 M. ARDON,Inorg. Chem., 4 (1965) 372.
AND R. G. PEARSON,
Inorg. Chem., 5 (1966) 223.
70 P. MOORE,F. BASOLO
F. BASOLO
AND R. G. PEARSON,
J. Am. Chem. SOC.,85 (1963) 1781.
71 S. A. JOHNSON,
12 Ref. 3, Chap. 3.
73 R. L. RICHAND H. TAUBE,
J. Phys. Chem., 58 (1954) 1.
74 H. B. GRAY,J. A m . Chem. SOC.,84 (1962) 1548.
L. CATTALINI,
F. BASOLO,
R. G. PEARSON
AND A. TURCO,
J. Am. Chem. Soc.,
75 U. BELLUCO,
87 (1965) 241.

c.

54

I N E R T COMPLEXES A N D METAL O R G A N I C C O M P O U N D S

U. BELLUCO,
Coord. Chem. Reu., 1 (1966) 11 1.
R. G. PEARSON,
H. B. GRAYAND F. BASOLO,
J. Am. Chern. SOC.,82 (1960) 787.
R. S. DRAGO,U. A. MODEAND J. G. KAY,Inorg. Chem., 5 (1966) 2050.
R. WENQUO,Masters thesis, Northwestern University, Euanston, Ill., 1965.
H. B. GRAYAND R. J. OLCOTT,Inorg. Chem., 1 (1962) 481.
F. BASOLO,
H. B. GRAYAND R. G. PEARSON,
J. Am. Chem. SOC.,82 (1960) 4200.
A. WERNER,
Z . Anorg. Allgem. Chem., 3 (1893) 267.
I. I. CHERNYAEV,
Ann. Inst. Platine U.S.S.R., 4 (1926) 261; 120. Sektora Platiny i Drug.
Blagorodn. Metal. Inst. Obsltclt. i Neorgan. Khinr. Akad. Nauk SSSR, 5 (1927) 118.
84 Ref. 2, p. 24.
85 Ref. 2, p. 22.
86 U. BELLUCO,
R. E. HOVE,F. BASOLO,R. G. PEARSON
A N D A. TURCO,
Inorg. Chem., 5 (1966)
76
77
78
79
80
81
82
83

591.
W. C. SEIDEL
A N D J. A. FETCHIN,
J . A m . Chem. SOC.,90 (1968) 2707.
87 J. P. FACKLER,
A. V. LINDSEY,
C. T. PREWITTAND U. G. STOLLBERG,
J. Am. Chem. SOC.,
88 R. D. CRAMER,
87 (1965) 685.
89 A. WOJCICKI
AND H. B. GRAY,Proc. Chem. Soc., (1960) 358.
J. CHATT,H. B. GRAY,R. G. PEARSON
AND B. L. SHAW,
J . Chem. SOC.,(1961)
90 F. BASOLO,
2207.
J. CHATTAND N. R. DAVIES,
Quarf. Reu., 12 (1958) 265.
91 S. AHRLAND,
J. Am. Chem. Sac., 85 (1963) 2922.
92 H. B. GRAYA N D N. A. BEACH,
Inorg. Chem., 8 (1964) 1201.
93 C. K. JORGENSEN,
Absorption Spectra and Cheniical Bonding in Complexes, Addison-Wesley,
94 C. K. JORGENSEN,
Reading, Mass., 1962.
95 M. R. CHURCHILL
A N D R. MASOB,
Aduan. Organornetal. Chem., 5 (1967) 93.
96 J. CHATT,P. L. PAUSONAND L. M. VENANZI,Organometallic Chemistry, Ed. H . Zeiss,
Reinhold, New York, 1960, Chap. 10.
97 L. E. ORGEL,An Introduction to Transition Metal Ion Chemistry: Ligand Field Theory,
Wiley, New York, 1960.
98 F. A. COTTONAND C. S. KRAIHANZEL,
J. A m . Chem. SOC.,84 (1962) 4432.
W. E. WILSONAND R. SUMMITT,
Spectrochim. Acta, 19 (1963) 863.
99 W. F. EDGELL,
100 W. BECKAND R. E. NITZSHMANN,
Z. Naturforsch., 17b (1962) 577.
AND A. WOJCICKI, J . Am. Chem. soc., 83 (1961) 520.
101 F. BASOLO
102 K. NOACK,Helv. Chirn. Acta, 47 (1964) 1064.
AND R. E. JOHNSON,
Inorg. Nucl. Chem., 11 (1959) 33.
103 D. F. KEELEY
0. GAMBINO,
E. SAPPAAND G. A. VAGLIO,Ric. Sci., 37 (1967) 430.
104 G. GETINI,
105 C. H. WEIAND L. F. DAHL,J . Am. Chem. SOC.,88 (1966) 1821.
F. CALDERAZZO
A N D R. ERCOLI,
Gazz. Chirn. ltal., 90 (1960) 1486.
106 G. PAJARO,
Atti Accad. Sci. Torino, 97 (1963) 1197.
107 G. CETINIAND 0. GAMBINO,
R. G. PEARSON,
L. F. KANGAS
A N D P. M. HENRY,
J. Am. Chem. SOC.,
108 J. P. DAY,F. BASOLO,
90 (1968) 1925.
109 I. W. STOLZ,G. R. DOBSON
A N D R. K. SHELINE,
J. Am. Chem. SOC.,85 (1963) 1013.
110 H. WERNER
A N D R. PRINZ,J. Organornetal. Chem., 5 (1966) 79.
11 1 R. J. ANGELICI
AND J. R. GRAHAM,
J . Am. Chem. SOC.,88 (1966) 3658.
AND R. J. ANGELICI,
Inorg. Chem., 6 (1967) 2082.
112 J. R. GRAHAM
I13 R. J. ANGELICI
AND B. E. LEACH,
J . Organornetal. Chem., 11 (1968) 203.
114 G. BOR, Spectrochirn. Acta, 18 (1962) 817.
115 C. A. STREULI,
Anal. Chem., 32 (1960) 985.
116 Ref. 1 , p. 128.
Inorg. Chem., 4 (1965) 1841.
117 E. L. MUERTTERTIES,
A. T. BRAULT
A N D A. J. POE,J . Chem. SOC.,(1964) 676.
118 F. BASOLO,
A ND R. J. ANGELICI,
J. Organomeral. Chern., 8 (1967) 374.
119 E. E. SIEFERT
AND M. L. FIENE,
J . Am. Chem. SOC.,81 (1959) 4200.
120 L. R. MERIWETHER
121 W. A. G. GRAHAM,
Inorg. Chem., l ( 1 9 6 8 ) 315.
122 W. D. HORROCKS
AND R. C. TAYLOR,
Inorg. Chem., 2 (1963) 723.
Inorg. Chem., 4 (1965) 1673.
123 G. R. DOBSON,

REFERENCES

55

124 A. F. SCHREINEV
AND T. L. BROWN,J. Am. Chem. Soc., 90 (1968) 3366.
AND F. BASOLO,
J. Am. Chem. Soc., 88 (1966) 3929.
125 E. M. THORSTEINSON
126 Ref. 1, p. 572.
A N D F. BASOLO,
J. Am. Chern. Soc., 89 (1967) 4626.
127 W. WAWERSIK
AND W. D. HORROCKS,
J. Am. Chem. Soc., 90 (1968) 922.
128 L. H. PIGNOLET
A N D F. BASOLO,
Znorg. Chem., 5 (1966) 1691.
129 E. M. THORSTEINSON
AND K. WOLLMANN,
Chem. Ber., 95 (1962) 1552.
130 W. HIEBER
AND A. WOJCICKI,
J. Am. Chem. Soc., 88 (1966) 844.
131 F. A. HARTMANN
AND F. BASOLO,
J. Am. Chem. Soc., 88 (1966) 2702.
132 S. BREITSCHAFT
I33 R. J. ANGELIC]AND F. BASOLO,Inorg. Chem., 2 (1963) 728.
R. W. HAMILL
A N D H. D. KEASZ,
J. Am. Chem. Doc., 89 (1967) 2851.
134 P. S. BRATERMAN,
Can. J. Chem., 46 (1968) 1655.
135 A. D. ALLENAND P. F. BARRETT,
F. ZINGALES
A N D U. BELLUCO,
Inorg. Chem., 6 (1967) 1582.
136 M. GRAZIANO,
F. BASOLOAND A. J. POE,J. Am. Chem. Soc., 85 (1963) 2215.
137 R. J. ANGELICI,
F. CANZIANI
AND F. BASOLO,
J. Organometal. Chem., 7 (1967) 461.
138 F. ZINGALES,
F. ZINGALES
AND U. BELLUCO,
Inorg. Chem., 6 (1967) 1582.
139 M. GRAZIANI,
AND J. R . GRAHAM,
J. Am. Chem. Soc., 87 (1965) 5586.
140 R. J. ANGELICI
AND R. J. A N G E L I C IAm.
, ~ . Chem. Soc., 87 (1965) 5590.
141 J. R. GRAHAM
A N D J. R. GRAHAM,
Inorg. Chem., 6 (1967) 988.
142 R. J. ANGELICI
AND R. J. ANGELICI,
Inorg. Chem., 6 (1967) 992.
143 J. R. GRAHAM
M. GRAZIANI
A N D U. BELLUCO,
J. A m . Chem. SOC.,89 (1967) 256.
144 F. ZINGALES,
A N D G. R. DOBSON,
Inorg. Chem., 7 (1968) 584.
145 G. C. FABER
M. GRAZIANI,
F. FARAONE
AND U. BELLUCO,
Znorg. Chim. Acta, 1 (1967) 172.
146 F. ZINGALES,
A N D F. BASOLO,
J. Am. Chenz. Soc., 84 (1962) 2495.
147 R. J. ANGELICI
U. SARTORELLI,
F. CANZIANI
AND M. RANEGLIA,
Inorg. Chern., 6 (1967) 154.
148 F. ZINGALES,
A N D U. SARTORELLI,
Inorg. Chem., 6 (1967) 1243.
149 F. ZINGALES
150 A. PIDCOCK,
J. D. SMITHAND B. W. TAYLOR,
J. Chem. Soc. A , (1967) 872.
AND B. W. TAYLOR,
J. Chem. Soc. A , (1967) 877.
151 A. PIDCOCK
AND U. SARTORELLI,
znorg. Chenz., 6 (1967) 1246.
152 F. ZINGALES
153 R. F. HECK,Advan. Organometal. Chem., 4 (1966) 243.
154 Ref. 1, p. 546.
155 E. L. MUERTTERTIES
A N D R. A. SCHUNN,
Quart. Rev. (London), 20 (1966) 245.
Structure and Bonding, 3 (1967) 106.
156 C. K. JORGENSEN,
157 L. E. ORGEL,J. Chern. SOC.,(1952) 4756.
158 B. R. BAKER,N. SUTINAND T. J. WELCH,Inorg. Chem., 6 (1967) 1948.
Inorg. Chem., 7 (1968) 364.
159 W. KRUSEAND D. THUSIUS,
160 Ref. 3, p. 24 ff.
161 Ref. 94, p. 109.
J . Am. Chenz. Soc., 90 (1968)
162 H. B. G R A YA N D J. ALEXANDER,
Chemistry in Britain, (1968) 162.
163 L. M. VENANZI,
A N D R. F. FENSKE,
Znorg. Chem., 7 (1968) 1273.
164 K. G. CAULTON
A N D H. J. POE,J. Chem. SOC.A , (1966) 1275.
165 H. L. BOLT,E. J. BOUNSALL
166 H. L. BOLTA N D H. J. POE,J. Chem. Soc., A , (1967) 205.
quoted in ref. 1, p. 187.
167 U. KLABUNDE,
168 Ref. 3, p. 72.
169 See, for example, ref. 1, p. 173 ff. for a contemporary account.
A N D F. BASOLO,
J. Am. Chern. Soc., 90 (1968) 2536.
170 D. E. MORRIS

Chapter 2

Reactions in Solution Between Various Metal Ions


of the Same Element in Different Oxidation States
P. J. P R O L L

1. Introduction

This chapter attempts to survey the studies which have been made on the various
electron transfer reactions, occurring between metal ions (of the same element)
in homogeneous solution. These reactions include the types known as exchange
reactions
c o 3 ++ c o 2 + + c o 2 ++ c o 3 +
NpO;++NpO:
Mn(CN)i- +Mn(CN):-

+ NpO;+NpO:'
+ Mn(CN):- + Mn(CN)i-

mutual oxidation-reduction reactions


Np(V) + Np(1II)

+ 2 Np(1V)

Cr(VI)+3 Cr(I1)

=4

Cr(II1)

Mn(VII)+4 Mn(I1) = 5 Mn(II1)


catalysed substitutions and isomerisations
Pt(NH3)5C13++C1- Pt(ll) Pt(NH,),Cl;+

+ NH3

Cr(H20),C12++ H 2 0 Cr(ll) Cr(H20);+ +C1CrNC2+ Cr(ll! CrCN"


~

and electron transfer reactions of the type


Fe(II1)X +Fe(II)Y
IrC1:- +IrBr:coci2

+c o 2

+ Fe(II)X+Fe(III)Y
+ IrC1;- + IrBri+ c o 3 + COCI+
+

The latter are often included with the exchange reaction type. Both inner- and

INT RO DUCT I ON

57

outer-sphere reaction mechanisms can occur in these electron transfer reactions.


The isotopic method is the most widely used experimental technique for the
study of exchange reactions. A suitable radioactive tracer is used to label one or
other of the oxidation states of the metal ion; the extent of the exchange reaction
is then followed by the analysis of the amount of radioactivity present in one or
both of the oxidation states after mixing reactants, quenching the reaction and
separating the reactants at various times. Depending on the exchange rate, either
a rapid mixing and quenching technique or the more usual slower technique is
employed. The rates of the exchange reactions are then obtained from plots of
In (1 - F ) versus time (where F is the fraction exchanged at time t ) either from the
slopes of the above plots or from the evaluation of the time of half exchange
( t + ) from these plots and the application of the McKay' equation,
Rate

[X][Y]

CXI+[Yl

.-0.693 - -In
t+

(1 - F ) . -~
[X][Y]
t

[XI + CYI

where X and Y are the two exchanging species. The rates of exchange obtained
for a wide variety of conditions are then used to obtain the rate law, rate coefficient and mechanism in the usual manner. In some systems non-linear McKay
plots have been found and a modified treatment of exchange data has to be
employed. Separation of the two oxidation states of the metal ion can be achieved
using the physical methods of ion migration, ion exchange, diffusion, solvent
extraction, and various chemical methods - usually involving precipitation of one
of the exchanging species. The possibility of separation induced exchange, which
can lead to erroneous results, is usually investigated and corrections applied,
if necessary, to the exchange data. More than one separation method should, if
possible, be used.
Exchange reactions can be sometimes investigated by the techniques of polarimetry, nuclear magnetic resonance and electron spin resonance. The optical
activity method requires polarimetric measurements on the rate of racemization
in mixtures of d-X (or I-X) and I-Y (or d-Y).
The reactions

+ d-Y + I-X
d-X + I-X
I-Y + d-Y

d-X + I-Y

occur. The various kinetic possibilities resulting have been observed and discussed
in papers by Eichler and Wahl' and Im and B u ~ c h ~The
, ~ . nuclear magnetic
resonance method requires measurements on a suitable spectral line (either proton
or metal resonance). From either the position or the broadening, brought about
References pp. 142-152

58

I O N S O F SAME M E T A L I N D I F F E R E N T O X I D A T I O N S T A T E S

by the presence of a paramagnetic exchanging species, of the resonance line of the


diamagnetic exchanging species an estimate of the exchange rate coefficient can
be obtained. An electron spin resonance method has also been used but has met
with little success to date in metal ion-metal ion exchange systems. The theory
underlying these resonance methods has been presented in the literature. Relevant
papers include those by McConnell and Weaver, McConnell and Berger6,
Bruce et al. 7 ,Ward and Weissman, Dietrich and Wahl, and Larsen and Wahl.
In the study of reactions of the types cther than exchange mentioned previously,
the usual technique involves the spectrophotometric examination of reaction
mixtures. The absorbance changes that occur, at a suitable wavelength where only
one species (either reactant or product) absorbs, as the reaction proceeds are
measured (manually or recorded). Treatment of the data uia the Beer-Lambert
law enables rate coefficients and laws to be found in the usual manner. Stopped
flow and temperature jump techniques have been used for very rapid reactions.
In this chapter, the results that have been obtained to date by the various techniques are reviewed. Sections 2 to 7 deal with metal elements in the same order as
their periodic group classification, with the exception of the transition, rare earth
and actinide elements which are dealt with in sections 8, 9 and 10, respectively.
Previously many excellent reviews on these and related topics have appeared
in the literature. Among these are articles by Sutin, Stranks and Wilkins,
HalpernI3, Vlcek14, Amphlett and Zwolinski et al. 16.

2. Copper, silver and gold


2.1

THE EXCHANGE REACTION BETWEEN

CU(I1) A N D CU(1)

McConnell and Weaver, using the NMR line width method, have obtained a
value for the observed rate coefficient, for the reaction in 12 M hydrochloric acid
media, of 5 x lo7 1.mole-.sec-. The width of the 63Cu NMR line from copper(1)
(- 1 M ) was observed in the presence of copper(I1)
to lo- M ) . In this
media the exchanging species are probably C U C ~ , ~and
CuCl,-. Optical
interaction effects have been observed in mixtures of Cu(1) and Cu(I1) in chloride
media.

2.2

THE EXCHANGE REACTION BETWEEN

Ag(I1) A N D Ag(1)

Gordon and Wahl have used the radioisotope OAg as a tracer for a study of
the exchange of silver between Ag(I1) and Ag(1) in acidic media. The precipitation
of Ag(phen)2(C104)2,brought about by the addition of o-phenanthroline, formed
the basis of the separation method. The experimental data were obtained using a

COPPER, S I L V E R A N D

59

GOLD

flow apparatus with reaction times of < 10 sec. Owing to the instability of silver(TI), media 6 M in H + and temperatures in the range - 14.8 to 11.4 "C were
employed. From measurements of the half-times (t,) and the generalised expression

where a and b are orders of the reaction with respect to Ag(1) and Ag(II), a rate
law
rate

~,,,[A~(II)]~

was obtained. The value of kobs (0.2 "C) for media 5.87 M in HC104 was 1020
].mole-'.set-'. An Arrhenius plot led to an activation energy of 12.5 kcal.mole-'

with an entropy of activation of - 1 cal.deg-'.mole-'..


Variation of the hydrogen-ion concentration, over the range 5.87 to 3.94 M
at a constant ionic strength of 5.87 M (C104-), led to the conclusion that the full
rate law was
rate

k'[Ag(II)]2/[H+]4

The mechanism for this exchange was postulated as


Ag2++xHz0

2 Ag(OH),"-")+

+ Ag(OH),"-"'+ + xH+ rapid


+ Ag(III)+Ag(I)

with the value of x possibly equal to 2.


Previously, Bruno and Santoro' had found complete exchange occurred between
the ions Ag(dipy):+ or Ag(phen);+ and Ag+, in nitric acid solution, within the
separation time.

2.3

THE EXCHANGE REACTION BETWEEN

Au(II1)

AND

Au(1)

Under conditions where the dismutation reaction is slow the exchange between
Au(II1) and Au(I) has been shown to proceed at a measurable rate; at 0 "C in
0.09 M HCl, an exchange half-time of about 2 min was observed. The isotopic
method ("*Au) and a separation method based on the precipitation of dipyridine
-chloroaurate(III) was used to obtain data. An acceleration in the exchange rate
was observed as the HCI concentration was increased'.

References pp. 142-152

60

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

2.4

THE EXCHANGE REACTION BETWEEN

Au(II1)

AND

Au(I1)

No direct studies have been made on this exchange system, but Rich and Taube'
have been able to propose a limit of > 1.7 x lo6 I.mole-'.sec-' at 0 "C for the
rate coefficient k i of the reaction

as a result of their studies on the exchange of CI- with AuCI,


Fe(I1).

2.5

THE DISPROPORTIONATION OF

in the presence of

AU(I1)

A limit has been proposed by Rich and Taube' for the rate coefficient k ; of the
process
2A~(II)

A ~ ( I+) A ~ ( I I I )

of > 108 I.mole-'.sec-'

at O "C.

3. Mercury
3.1

T H E EXCHANGE REACTION BETWEEN

Hg(I1)

P R O P O R T I O NATION R E A C T I O N O F

AND

Hg(1);

T H E DIS-

Hg(11)

The earliest reported studies on the exchange, by King', and Haissinsky and
Cottin2, utilised the isotopic method (205* '03Hg) and showed the rapid nature
of the process. Complete exchange was observed within 2 min at room temperature
in acidic nitrate media2. Similar results were obtained with acidic perchlorate
media'. The separation methods used the insolubility of Hg(1) as (u)chloride',',
( b ) chromate' and ( c ) sulphate'.
Wolfgang and Dodson3 in a further study suggested a method of separation
using the picrolonic acid precipitation of Hg(1). In perchlorate media even with
reactant concentrations as low as lo-' M , complete exchange was always observed
by these workers, except in the presence of added CN- (- [Hg(II)]) when the
kinetics could be examined. With reactant concentrations in the ranges Hg(1)
6x
to 7 x lo-' M , and Hg(I1)
1.6 x
M the rate law
to 1.6 x
found was
N

61

MERCURY

The observed rate coefficient (kobs)was found to be dependent on the ratio


[CN-]/[Hg(II)] as was the amount of zero time exchange. At 0 "C, for [CN- J =
[Hg(II)], a value of 5.7 x
I.mole-'.sec-' was calculated, and the overall
activation energy obtained was 14 kcal.mole-' over the range 0 to 30 "C. The rate
of exchange was found to be independent of ionic strength, hydrogen ion concentration and surface area and was not affected by light, oxygen or colloidal
mercury.
Wolfgang and Dodson suggest that the exchange process in the absence of
cyanide is controlled by a rapid dismutation reaction

3 . 2 10'
~ sec-' at 25 "C.
and have calculated a theoretical rate coefficient of
In the presence of cyanide, however, a new rate -determining step, involving a
neutral Hg(I1)-CN complex, possibly
Hg;++HgCN(C104)

is present. King4 and Adamson' have pointed out that the equilibria
Hg2++CN-

+ HgCN'

and
Hg2++ HgCN'

+ HgCN' + Hg2+

must not be rapid for the above mechanisms to be valid, and the observed zerotime exchange should be related to the concentration of free Hg2+.
Wolfgang and Dodson6 have since made a study of the exchange reaction
between H g 2 + and HgCN+ using 203Hg and 10; precipitation of the species
Hg2+.They have proposed a mechanism involving the steps

Hgz+ HgCN'
Hg( CN),

--*

HgCN'

+ Hg2'

ki
+ Hg2 +
HgCN' + HgCN
+

k2

(K)

to account for the experimental observations on the slow exchange. A value for
the equilibrium constant K of 20 at 25 "C has been obtained. With the aid of this
value of K, it was found possible to calculate the zero-time exchange (Z,) mentioned
previously. Good agreement was obtained between the calculated and observed
values of Z , . The rate coefficients k , and k , have values (at 0 "C, p = 0.12 M )
of 1.7 x
and 8.3 x
I.mole-'.set-', respectively, both activation energies
being 17 kcal.mole-'. For the exchange HgCN+-Hg(I), a rate coefficient (0 " C )
References pp. 142-152

62

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

of 7.0 x
l.mole-'.sec-'
and activation entropy of -18 cal.deg-'.mole-'
have been calculated.
Peschanski', during the course of a further study on this exchange in the
presence of cyanide, has obtained results which are in fair agreement with those
above.

3.2

THE EXCHANGE REACTION BETWEEN

Hg(I1)

AND

Hg(1)

IN

N 0 N-A QU E O U S M E D I A

Peschanski', using the isotopic method (z O'Hg), has found complete exchange
(0 "C) in methanol and various other non-aqueous media. The separation methods
used were, (u) paper and column chromatography, ( b ) paper electrophoresis, and
(c) precipitation of Hg(1) with chloride. In the presence of cyanide ions, however,
less than complete exchange could be observed. Zero-time exchange was again
found to vary in the same manner as for aqueous media. Similar effects were
observed in the presence of chloride ions.

4. Thallium

4.1

T H E E X C H A N G E R E A C T I O N B E T W E E N TI(II1) A N D

Tl(1)

Although attempts have been made to study this reaction using ThC as an indicator'P2, the stability of ThC has restricted these investigations. In 1948 the
production of
z06T1enabled more detailed studies to be made by Harbottle
and Dodson3 and by Prestwood and Wah14. In preliminary reports, these authors
presented data for the exchange in perchloric3, 4, hydrochloric3 and nitric4 acid
media, obtained using separation methods involving precipitation of (u ) TI(1)
as chromate3 or bromide4 and (b) Tl(II1) as hydroxide4. A rate law
' 0 4 3

rate

kobs[Tl(IIl)][Tl(I)]

was observed for each of the above three media. In 0.4 M HCIO, kobs (49.5 " c )
has a value of 5.6 x
l.mole-'.sec-' and an associated activation energy3 of
12 kcal.mole-'.
Prestwood and Wahls, in a report of a more detailed study conducted at an
ionic strength of 3.68 M (ClO,-) and temperatures in the range 10 to 50 "C, have
presented results consistent with a more complicated rate expression
rate

( k , k;[H+]-')[TI(III)][TI(I)]

63

THALLIUM

(plots of kobsversus [H+]-' were linear with a positive intercept). This was interpreted in terms of two exchange pathways (rate coefficients k, and k,), viz.
TI3++H,O

+ TIOH'+

+H+

K,

TIOH,+ + ~ i -%
+
with k , > k , and K, having a small value. The values of k, and k;(= k,K,)
at 25 "C are 4.3 x lo-' I.mole-l.sec-' and 2 x 10-6sec-', respectively, with corresponding activation energies of 17.6 and 10.3 kcal.mole-'. The thallium concentration employed in this work was 5 x
to 5 x lo-, M.
Harbottle and Dodson6, have presented data obtained during a study of the
exchange in perchlorate media, under constant high ionic strength conditions
(p = 6.0 M ) . From results obtained with Tl(II1) 1.4 x
to 1.7 x lo-, M ,
TI(1) 1 x
to 2 x
M and H + 3 x lo-' to 6.0 M , it was concluded that
only the step defined by k, was operative and the extent of hydrolysis was large
(plots of k:,: versus [H'] were linear with a positive intercept). On this basis, a
value fork; (25 "C) of 2.6 x 10- I.mole-'.sec-' and an associated activation energy
and entropy of 14.7 kcal.mole-' and - 32 cal.deg-'.mole-' were calculated.
Dodson', in an attempt to resolve the difference in interpretation, carried out
further experiments at various acidities and constant ionic strengths at 50 "C.
He was able to show that the data given by Prestwood and Wahl' and by Harbottle
and Dodson6 were in good agreement with his own, and the additional data he
obtained at p = 2.0, 4.0 and 6.0 M could be interpreted in terms of exchange via
T10H2+ and Tl'. The rate was found to decrease with increasing ionic strength
and a revised activation energy for the process involving k , of 16 kcal.mole-'
was given.
In 1955 Rossotti', utilising the data available at that time on the hydrolysis of
TI(II1)' and Tl(1)' was able to show that the interpretation given by Prestwood
and Wah15 of their data (p = 3.68 M ) was feasible. Estimates were made of k,
and k, (25 "C, p = 3.68 M ) of 4.28 x lo-' and 3.4 x
I.mole-'.sec-', respectively.
In 1961, Roig and Dodson" carried out a further study of the exchange in
perchlorate media under identical conditions (25 "C, p = 3.0 M ) to those in the
Tl(II1) hydrolysis study'. The isotope 'TI was used, with a separation procedure
based on extracting Tl(II1) from reaction mixtures with either methyl isobutyl
ketone or diethyl ether. The exchange was examined in the absence of light,
and a correction procedure to eliminate the catalytic effects of traces of chloride
ions was used since Tl(II1) concentrations of
M were necessary at the
very low acidities employed. Using the known values of the first and second
hydrolysis constants of TI(II1) ( K , and K 3 )

'

References pp. 142-152

.64

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

T10H2

+ H 2 0 + Tl(0H); +H

K3

the values of the rate coefficients k , , k,, and k , for the steps

TI OH^+ +TI+ -%
T~(oH): +TI+ -%
were calculated to be 7.0 x 10- 5, 2.5 x
and < 2.8 x l o p 5 l.mole-'.sec-',
respectively. Activation parameters obtained for the first step ( k , )were E, = 17.4
kcal.mole-' and A S f = -21 cal.deg-'.mole-'.
Gilks and Waind12 have examined the exchange in heavy water and have found
the exchange rate to be about two-thirds of that in ordinary water. A similar
dependence, at constant ionic strength, on D + as on H f was observed and the
rate of exchange was found to decrease as the ionic strength was increased. The
isotope O4TI and chloroplatinate precipitation of TI(1) was employed in this
study. Waind', has discussed the role of water molecules in these exchange reactions.
The accelerating effect of nitrate ions (over the range lo-' to 1.0 M , at p = 3.68
M ) has also been studied5. An additional pathway involving the species TlN032+,
viz.

~ 1 , ++NO,

+ TINO:+

K4

k4

TINO;++TI+ ---+
was thought to occur at [NO;] < 0.4 M . At higher [NO;], pathways involving
'TI(N0,): were also thought likely. The full rate law found by Prestwood and
Wahl' was
rate

( k , k;[H+]-'

+ k~[NO,])[Tl(III)][Tl(I)]

The value of kk( = k4K4) at 25 "C is 4.45 x


12.mole-2.sec-', and the associated activation energy is 16 kcal.mole-'.
The effect of added chloride on the exchange at constant ionic strength and
acidity has been investigated by Harbottle and Dodson6. For Tl(II1)
6.8 x lo-,
to 6.8 x
M , C1- < 3.5 x
Mproduced a decrease
M a n d TI(1) 2 x
M produced an increase. A minimum in the
in the rate whereas C1- > 3.5 x
rate was found to occur at a ratio [Cl~]/[Tl(lll)] 1.5. The exchange pathways

65

THALLIUM

believed to occur are


T1Cl2'+ Ti'
TlCl:

k3

+T1' -%

TICIT + TICiiTlCI,

k7

--+

+TICIi- %-

and it was concluded that k , and k , were less than k , or k , and that k , and k 8
were larger than k , or k,. For the exchange in 0.4 M HCI ( p = 6.0 M ) at temperatures 31.8 to 41.8 "C, an overall activation energy 29.6 kcal.mole-' was found.
Gilks et a l l 4 have since confirmed the above observations and have found a
similar effect to operate in deuterated solvent. At high [Cl-] the rate in D,O is
equal to that in H,O.
Challenger and Masters" have made some observations on the exchange in
to 1.7 x
M
0.4 M H,S04 solution using concentrations of Tl(II1) 8.5 x
and of Tl(1) 1.7 x
to 8.5 x
M , with the tracer ,O4TI and a chloroplatinate precipitation of TI(1). The rate law
rate

= /cobs[TI( I1I)] [T1( I)]

was obeyed and led to a calculated value of kobs(25 " C )1.98 x


l.mole-'.sec-'
with an associated activation energy (0 to 44.3 "C) and entropy of 13.8 kcal.mole-'
and - 22 cal.deg-'.mole-', respectively.
Brubaker and MickelI6 later reported results, obtained in a more detailed study
for sulphate media of ionic strength of 3.68 M with the species Tl(III), TI(I), H +
and SO4'- in the concentration ranges 1.8 x
to 1.08 x lo-, M , 1 . 0 ~
to 2.0 x
to 7 x lo-' M , respectively.
M , 4.5 x lo-' to 2.90 M and 7 x
The chromate precipitation method was used. The experimental data at 24.9 "C
were found to be consistent with a detailed rate equation
rate =

0.0345 + k;[H+]'[SO:-]
[H']'

+ k~o[H']~[SO~-~3[T1(III)][Tl(I)]

+ K,[H'] + K 2 K 3 + K,[H']'[SO:-](l+

KIO[SO:-])

where K2 has been defined previously and K 3 , K 9 , and K , , are given by


K3 =

[TI 0'1 [H' ] ; K , =


T1soa and K , , = [TlSO,]
[TI OH '1
[Ti3'1 [S Oi-1
[Tl'] [ S O:-]

kb and k i 0 are products of exchange rate coefficients and equilibrium constants


References p p . 142-152

66

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

and have numerical values (25 "C) of 0.012 and 0.2, respectively. Measurements
on the exchange at other temperatures were made and the exchange rate was
found to be dependent on the ionic strength.
In a further study, Brubaker et aI.l7 have reported on the effects of the addition
of chloride ion to the sulphate exchange system at virtually constant ionic strength
(3.68 M ) , sulphate and hydrogen-ion concentrations. For the concentration ratio
[CI-]/[Tl(III)] of 9 . 2 ~lo-' to 9.5 at 24.9 "C results analogous to the effect
observed in perchlorate media6 were obtained. The minimum in the rate corresponded to a ratio of
2.5. Results were also presented for the conditions,
constant [CI-] and variable [SO:-] and [H'] ( p = 3.68 M ) . Brubaker et a / . 1 7
have suggested that the exchange paths most likely to occur in sulphate media are

TISO: +T1+

TI(S0,);

+ TISO;

kto

for which activated complexes involving sulphate bridges are likely. Wiles", has
made a further study of this system in the presence of small amounts of added
sulphate and has observed a first order dependence on this anion. Results at higher
concentrations of added sulphate agree with those mentioned previously. A
sulphate bridge activated state was again proposed.
The effect of the addition of cyanide to the exchange reaction in perchlorate
media has been reported by Penna-Franca and DodsonL9.An effect similar to the
addition of chloride was observed with a minimum occurring at a ratio [CN-I/
[Tl(III)] of 3.5 ( p = 0.5 M , 30 "C). It was concluded that the pathways involving
T1+ and TlCN2+ or TI(CN)l are slow whereas the pathways involving T1+ or
cyanide complexes of TI(1) [TlCN, TI(CN);, Tl(CN):-] and TI(CN), or
TI(CN); are more rapid.
Dodson et ~ 1 . ~have
'
also studied in detail the effect of added bromide on the
exchange reaction using precipitation and extraction separations and 04Tl and
202T1as indicators. The variation in the rate of exchange was found to be governed
by the law
rate

k,[T13+][T1+]+ k2[T10H2+][TI+]+k,,[TIBr~]

+ k,,[TIBr,] + k,,[TIBr;][TI+] + k,,[TIBr,][TIBr,]


for added bromide < 0.2 M . The rate coefficients have values (p = 0.5 M , 30 "C):
k , , and k,,, 2.2 x
and 1.25 x
sec-I; k , , and k,,, 1.28 x lo-, and
6.6~
I.mole-'.sec-', respectively. Plots of the observed rate of exchange
versus [Br-] exhibit a maximum followed by a minimum and a further increase
at high bromide ion concentrations. For the step
TIBr; +TlBr;

ki4

-+

THALLIUM

67

the possibility of a symmetrical activated complex was suggested and the steps
having rate coefficients k l i and k , , were proposed as
TlBr:

%T1

TIBr,

%TI

+ Br,
+ Br, + Br-

Brubaker and Andrade have made a study of the effects of acetic and succinic
acids on the exchange; a reduction in the exchange rate was observed.
Some conclusions on the exchange reaction have been made as a result of studies
on the redox reactions between Tl(1) and Ce(TV) and between V(IV) and
Tl(III)23. Gryder and Dorfrnan, have pointed out that TI(II) cannot be invoked
for both the exchange between Tl(II1) and Tl(I) and the reaction of Tl(I) with
Ce(IV) which is slower than the exchange. S y k e ~ has
, ~ shown, on the basis of a
mechanism proposed for the reaction between TI(II1) and V(IV), that Tl(I1)
cannot be involved in the exchange reaction as this would lead to Tl(I) accelerating the above redox reaction. It is interesting to note that as long ago as 1949,
McConnell and DavidsonZ5reported the lack of interaction absorbance phenomenon between TI(II1) and Tl(1). This might be expected if Tl(I1) was absent.
Radiation-induced exchange reactions which are more rapid than the thermal
exchange reactions are thought to involve TI(I1) 5 * 2 6 .
Challenger and Masters have made a detailed study ot the X-ray-induced
exchange reaction in 0.4 M H,S04 solution, over the ranges TI(II1) 2 x lo- to
7 x lo-, M a n d TI(1) 1.4 x
M , with X-ray intensities ( I ) 1.7 x l o i 9
to 7 x
to 4 x 10, eV.l-.min-. The rate law observed was
rate

k,,,[TI(I11)]0510.67

The exchange mechanism proposed involves the steps


TI(I)+OH

+ TI(Il)+OH-

Tl(III)+H

+ TI(II)+HS

TI(I1) + Tl(1)
TI(II)+TI(IZI)
2 TI(I1)

+ exchange
-+

exchange

+ Tlf+Tl3+

Radiation of wavelength 253.7 mp has also been used to induce exchange.


Stranks and Yandel126,on the basis of their observations, have proposed a similar
t The exchange reaction in the presence of CH,CICOOH, CHCI2COOH and CCISCOOH has

recently been examined by McGregor and Wiles.


References pp. 142-152

68

I O N S O F SAME M E T A L I N D I F F E R E N T O X I D A T I O N S T A T E S

mechanism

~ i ~ + ++ T hI ~ + + O H + H +
T 1 0 H Z f + h v + T12++OH
T I + + O H .+ TI'++OHTi3++T1'+
T1'

+T12

exchange

exchange

2 TP+ -+ TI++ ~ 1 3 +
The value of the ratio k,,/k,, (at 25 "C) is 1.2 (see also Stranks and Yandel12*).

5. Tin and lead


5.1

THE EXCHANGE REACTION BETWEEN

Sn(IV)

AND

Sn(I1)

IN AQUEOUS

MEDIA

In aqueous solution this exchange has been studied, in the absence of oxygen,
in chloride'*' and sulphate3 media. The isotope 'I3Sn was used as the indicator
and the separation of Sn(IV) and Sn(I1) was achieved by the formation of the
insoluble salts caesium hexachlorostannate(1V)' - 3 and stannous oxalate3.
Davidson et al.' have investigated the exchange reaction in chloride media,
where the rate law obeyed is
rate

k,,,[Sn(IV)][Sn(II)]

A value for kobsof 9.2 x


I.mole-'.sec-' was calculated from data obtained
at 25.2 "C for 10 A4 HCI solution with Sn(IV) and Sn(I1) in the ranges 6 x
to 3.2 x lo-' M and 9 x
to 5.7 x lo-' M , respectively. The value of kobswas
found to be dependent on the concentration of HCI. An overall activation energy
of 10.8 kcal.mole-' was obtained. These authors, following the observation of
Whitney and Davidson4, have studied i n some detail the interaction observed in
mixtures of Sn(IV) and Sn(I1) and have concluded that an equilibrium

occurs. Exchange could occur via this equilibrium.


Craig and Davidson' have investigated the photochemical exchange, using light
365 mp (a wavelength where the interaction complex absorbs
of wavelengths
N

69

TIN AND LEAD

strongly) from a high intensity source. Uranyl oxalate, which has similar absorption characteristics to the complex (Sn,Cl?;), was used for the actinometer.
Quantum yields for the process were found to be
0.2. It was concluded that
exchange occurs uiu an activated complex which is unsymmetrical.
The exchange in sulphate media, where a similar rate law is obeyed and the
exchange rate is slower than in chloride media, has been studied by Gordon and
Brubaker3. At p = 4.98 M , kobs (25 "C, [ H + ] = 3.99 M and [SO:-] = 0.99 M )
has a value 1.7 x
l.mole-'.sec-'. The effect of variation in the hydrogen ion
(3.30 to 5.0 M ) and sulphate ion (0.70 to 1.40 M )concentrations at constant ionic
strength (LiClO, and Li,SO,) was found to conform to a relationship

suggesting the involvement of two exchange pathways. The possible steps proposed
were
SnOH' -+
SnO(S0,):-

Sn(OH)(SCl,); +Sn(OH)SO,
Sn(OH),SO,+Sn(OH)SO,

-+

-+

The main species of Sn(IV) and Sn(I1) in sulphate media were thought to be
SnOH' and Sn(S0,)2. The overall activation energy (over the range 25-50 "C, in
3.0 M H2S04 with p = 4.98 M ) was calculated as 19 kcal.mole-'.
The addition of chloride ion was found to increase the rate of reaction and, if
present in the range 0-0.5 M , to lead to abnormal orders with respect to Sn(IV)
and Sn(I1). Above 0.5 M C1- a rate law
rate

ko,,[Sn(IV)][Sn(II)][C1-]

was obeyed, with /cobs(25 "C, 1 M [Cl-1) having a value 1.23 x


I2.moleC2.
sec-' with an associated overall activation energy of 22.6 kcal.mole-'. The exchange path

was concluded to occur3. At high temperatures, McKay plots were found to show
deviations from linearity for exchange in presence or absence of chloride, suggesting more than one exchange path is operative.
Interaction absorbance data have also been reported by Gordon and Brubaker3
for mixtures of Sn(IV) and Sn(I1) in sulphate media. Exchange uiu interaction
dimers, Sn(IV)-Sn(II), seems likely.
References p p . 142-152

70

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

5.2 T H E E X C H A N G E

REACTION BETWEEN

Sn(IV)

AND

Sn(II)

I N NON-

AQUEOUS MEDIA

The exchange of I2'Sn between SnCI, and SnCl, in anhydrous methanol and
ethanol in the absence of oxygen has been investigated by Meyer e l ul.'.'. From
the linearity of plots o f t + uersus the inverse of the total tin concentration, which
was varied over the range
4 x lo-* to
3 x lo-' M , a rate law

rate = k,,,[Sn(IV)][Sn(II)]
was suggested. The kinetic parameters found were: (u) methanol solution';
kobs (33 "C) 7.12 x lo-' l.mole-'.sec-', activation energy (over range 33-47 "C)
21 kcal.mole-' and activation entropy - 11.5 cal.deg-'.mole-';
(b) ethanol
solution2; kobs (25 "C) 6.03 x lo-' I.mole-'.sec-', activation energy (over the
range 25-42 "C) 23.7 kcal.mole-' and activation entropy -0.4 cal.deg-'.mole-'.
The separation of Sn(I1) and Sn(IV) was based on the insolubility of the oxalate
of Sn(1I) in alcoholic media.
The mechanism proposed for ethanol media, which.is similar to that for
aqueous media, was
SnCI,

or SnCl:

+ SnCl, + Sn2c16
+SnCl, + Sn2C16

rapid
rapid

Presumably in methanol solution a similar mechanism is operative.

5.3 THE

EXCHANGE REACTION BETWEEN

Pb(1V)

AND

Pb(I1)

I N AQUEOUS

MEDIA

Investigations by Zintl and Ranch', suggested that, i n aqueous alkali, the


oxyanions of lead (plumbate and plumbite) do not exchange at room temperature.
This has been confirmed by Fava2, who detected no exchange in 7 M KOH over
a period of ten days at room temperature, but found measurable exchange at temperatures in the range 57 to 100 "C with reactant concentrations
2 x lo-' M.
The barium plumbate separation method was used with the tracer Ra D.
The corresponding overall activation energy obtained was 33 kcal.mole- '. At
100 "C, the measured half-time for exchange was 13 min.

5.4

71

ARSENIC A N D ANTIMONY

THE EXCHANGE REACTION BETWEEN

Pb(1V)

AND

Pb(I1)

I N NON-

A Q U E O U S MEDIA

Exchange between Pb(W) and Pb(I1) as acetates in acetic acid solution has been
observed' using the isotopic method (Th B). However, Evans et ul. ', using Ra D
to lo-' M ,
as the indicator, with reactant concentrations in the range
concluded that there was no exchange at 80 "C in a period of several hours. The
reaction was carried out in a dry-box system and anhydrous acetic acid was used
as the solvent. Three separation procedures were used: (u) conversion of the lead
species to the plumbate and plumbite anions, followed by precipitation of barium
plumbate (a method used previously for the aqueous exchange system), (6)
precipitation of lead(1V) as the dioxide, and ( c ) rapid cooling and consequent
crystallisation of the lead(1V). Method (b) was reported to bring about induced
exchange.

6. Arsenic and antimony


6.1

THE E X C H A N G E R E A C T I O N

BETWEEN

AS(V)

AND

AS(I1I)

Wilson and Dickenson' observed no exchange, over a period of three hours


at 100 "C,between arsenate and arsenite ions in media ranging from aqueous acid
to aqueous alkali. Martin et ul.' have found similar results for the exchange
between arsenate and thioarsenite ions in aqueous media. However, in liquid
ammonia exchange occurred between ammonium arsenate and arsenic trisulphide'.
The isotopic method was used'.'.

6.2

T H E E X C H A N G E REACTION BETWEEN

Sb(V)

AND

Sb(II1)

I N AQUEOUS

MEDIA

Absorbance measurements on mixtures of Sb(V) and Sb(1II) in hydrochloric


acid media, have led Whitney and Davidson's to propose a dimeric equilibrium

'

SbCl;

+ SbCI, + Sb,Cl:;

Observations on Sb(V) solutions, in the same media, have shown that the absorbance is dependent upon (u) time, and ( b )hydrochloric acid concentration' - 3 .
This has been shown to be due to reactions
SbCI,+x H,O
References p p . 142-152

+ SbC1,-,(OH),+x

H + + x C1-

72

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

Neumann, has measured values of the equilibrium constants ( x = 1 and 2) in


media 6-11 M HCI. Neumann and Ramette4 have obtained data on the rate of
hydrolysis of Sb(V) using spectrophotometry.
Bonner', using the isotopic method (124Sb) and either precipitation of the
oxinate of Sb(II1) or an isopropyl ether extraction of Sb(V), obtained the first
6 M HC1 a complex rate law, viz.
results on this exchange reaction. In media

rate

k'[Sb(111)]0'6[Sb(V)]'.l [HfI4[C1-l9

was found to be consistent with the experimental data. Concentrations were varied
over the ranges: Sb(1II) = Sb(V), 8 x
to 4 x
M ; C1-, 5.4 to 6.1 M;
li3.'.
H', 4.7 to 6.1 M . The value of k' (25 "C) was calculated as 2 . 4 ~
mole 3.7.sec-' with the overall activation energy
27 kcal.mole-'.
Neumann and Brown6, have suggested that the only species of Sb(V) that can
exchange with Sb(II1) is the ion SbCI,. Numerous reasons for this assumption
have been discussed. These authors have shown that the observed half-times of
exchange in media 6-12 M with respect to HCl'-' can be satisfactorily predicted
on the basis of a mechanism

SbCI;

+ SbCI, + SbCI, + SbCl;

(2)

where B represents species of Sb(V) which are present in equilibrium with SbC16-,
in which either reaction (1) or reaction (2) can be rate determining depending on
the conditions. In concentrated HCI the order of the reaction with respect to
[Sb(111)] was demonstrated to be unity. Various transition states, involving
chloride bridges, have been suggested6.
Cheek et al. ', have since published data for HCI media showing ( a )a maximum
exchange rate ( 9 M HCI), (b) non-linear exchange-time plots ( - 7 M HCI),
(c) an order in Sb(l1I) of 1.0 (1 1.8 M HCI) and 0.9 (9.5 M HCI), ( d ) an activation
energy which depends on the acidity, and ( e ) details of rate dependence upon
chloride and hydrogen ion concentrations. The observed rate coefficient (25 "C)
and activation energy are 5.7 x lo-, I.mole-'.sec-' and 17.2 kcal.mole-', respectively, for media 11.8 M in HCI. These authors conclude that exchange occurs
between Sb(1II) and two Sb(V) species which are present in slow equilibrium with
each other.
Kambara et ~ l have
. ~reported data for the exchange in media 0.8 to 4.0 M
with respect to HCI, the highest rate being observed at 2.0 M HCI. Added lithium
chloride and the previous treatment of reactants was also found to affect the rate
of exchange.
Bonner and Goishi" have reviewed the complex exchange kinetics and the

ARSENIC AND ANTIMONY

73

hydrolysis reactions of Sb(V), and have made a further study of the exchange in
the region
7 M HCI. The exchange data were treated according to a reaction
scheme in which Sb(II1) could exchange with two forms of Sb(V) which are
themselves in a slowly established hydrolysis equilibrium. For a medium 7.04 M
in HCl, the values of k , and k, the rate coefficients for exchange between Sb(II1)
and the two forms of Sb(V), are 2.32 x lo-' I.mole-l.sec-' and 2 6.7 x lo-' 1.
mole-'.sec-', respectively, at 24.95 "C. Energy of activation for the step with
coefficient k , is 17.2 kcal.mole-', with an entropy of activation of - 10.3 cal.deg-I.
mole-'. Some evidence was also obtained for a photochemical effect which only
shows itself at low [Sb(III)]. Bonner and Goishi have also pointed out that the
interaction dime? and the Sb(II1)-catalysed hydrolysis of Sb(V)4 must have unsymmetrical activated complexes, which are different from each other and from
the exchange activated complex.
Turco and Faroane" and Turco" have investigated the effect of bromide ions
on the Sb(V)-Sb(II1) exchange reaction in 3.15 M HCI media and have found a
complex rate law"

rate

k,,,[Sb(V)]

1~1[Sb(III)]o~15[H+]4~2[Br-]3~5

which was attributed to the exchange occurring via two pathways, one independent
of Sb(II1) (the main pathway) and one involving both Sb(V) and Sb(II1). A
simplified rate law
rate

k;[Sb(V)]

+ k;[Sb(V)][Sb(III)]

was also found to fit the experimental data. At low [Sb(III)] and [Br-1 (7.5 x
to 1 x lo-' M and 1.0 M, respectively), only the first term in the above rate law
sec-'. The step postulated
was observed. The value of kj (22 "C) is 1.3 x
to account for this exchange was
Sb(V)Br,(X)

+ Sb(III)Bq,,-,)(X)+

Br,

where X represents an hydroxide or chloride ion.


Brubaker and S i n ~ i u s ' l4
~ ,have reported the exchange reaction as not occurring
in sulphate media ([H'], 3 to 12 M ) unless chloride ions are present. McKay
plots deviating from linearity were observed for added chloride ion 0.4 to 6.0 M ,
and complicated dependencies on the ions H', SO:- and Sb(V) were also found.
For mixtures of Sb(II1) and Sb(V), the principle of additivity of absorbance was
obeyed in sulphate media.
TurcoIs has also detected little exchange between Sb(II1) and Sb(V) in alkaline
media (1.8 M KOH) at room temperature.
References p p . 142-152

74
6.3

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

THE EXCHANGE REACTION BETWEEN

Sb(V)

AND

Sb(II1)

I N NoN-

AQUEOUS MEDIA

Barker and Kahn' have made a detailed study of the exchange in carbon
tetrachloride media using the isotope 124Sbto label the Sb(II1) species. The reaction was carried out in sealed ampoules covered with A1 foil in the presence of an
atmosphere of He or Ar gas. The separation method used involved complexing
the Sb(V) with fluoride (brought about by addition of ethanol, HCl and HF)
followed by precipitation of the Sb(II1) with H2S and finally addition of boric acid
and HCI, removal of the CCI,, and treatment with H2S to remove the Sb(V).
Zero-time exchange was
5 %.
Concentration ranges employed in this study were Sb(III), 1.5 x lo-' to
M , the chloride salts being
to 6 . 7 ~
1 . 2 lo-'
~
M , and Sb(V), 5 . 5 ~
used. The experimental data led to a rate law
N

rate = k,[SbCl,]

+ k;[SbC1,]2[SbC1,]

sec-' and 1.8 x lo-,


with values of the coefficientsk , and k ; (at 50.1 "C)1.6 x
12.mole-z.sec-'. Activation energies for the processes represented by k, and k;
were calculated as 19 and 15 kcal.mole-', respectively, from data obtained over
the range 50.1 to 81 "C. The rate of exchange was found to be affected by light, but
not by increase in surface area.
A mechanism involving complex chloride species, uiz.
SbCl,

+ SbCl, + C12
ki

+ SbzCI,,
(rapid)
SbzClIo + SbCl, + SbSbzC113
(rapid)
SbSb2C113 SbCI, + SbSbCllo
2 SbCl,

k2

--t

was proposed to account for these observations. A suggestion was made concerning
the structure of the complex (SbSb,CI,,).
Price and Brubaker'9 have examined the catalytic effect of hydrogen chloride
on the rate of exchange in the same media. Over a range of HCl concentrations,
M , the data obtained was consistent with a rate law
6 x lo-, to 6 x

k , (50 "C) has a value 2.07 x lo-, I.mole-'.sec-'


the exchange pathway
HSbCI,

+ SbCI,

k3
--f

and is the rate coefficient for

75

TRANSITION METALS

It is of interest to note that the 3 2 P exchange between PCl, and PCI, also
proceeds, in CC14 media, via a dissociation step4
k4

PCl, =$ PCl,

+ C12

with a rate law


rate = k4[PC15]

While the phosphorus exchange is


130 times faster than the antimony exchange
for the same basic process the energies of activation (16 and 19 kcal.mole-')
are similar. No light sensitivity was observed in the 32Pexchange.

6.4

THE

Sb(III)-CATALYSED

HYDROLYSIS OF

Sb(V)

Neumann and Ramette' have found the hydrolysis of Sb(V) to be catalysed


by Sb(II1). An activated complex, which must be unsymmetrical, has been
proposed2.
7. Tellurium
7.1

THE EXCHANGE REACTION BETWEEN

Te(V1) A N D Te(1V)

The exchange between the acids H6Te06 and H2Te0, has been investigated
using the radioisotopes 12'Te, '29Te' and I3'Te2. No exchange was detected over
periods of many hours at 95 "C in acidic chloride or perchlorate solution up to
6 M with respect to H+. Separation was achieved using the precipitation of the
tellurite anion which occurs at pH 4.20.

8. Transition metals
8.1

VANADIUM AND TANTALUM

8.1.1 The exchange reaction between V(ZZZ) and V(ZZ)

King and Garner', using the isotopic method ("*V), made an attempt to study
this exchange reaction in both aqueous sulphate and perchlorate media. The
separation methods tried were (a) ion exchange, and ( b ) precipitation of the
vanadium (111) with ammonia, after first complexing the vanadium (11) in
aqueous ethanol as V(dipy):+. Complete exchange was observed within minutes
lo-' M .
at 2"C, in the absence of oxygen, with reactant concentrations of
Evidence was obtained that the exchange of V3+ with V(dipy):+ is slow.

References pp. 142-152

76

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

Krishnamiirty and Wah12 tried a number of separation methods and eventually


used a modified dipyridyl-ammonia separation to obtain kinetic data for this
exchange. The zero-time exchange lay between 25 and 50 o/o, depending on the
conditions; the activity of the V(dipy):+ ion was measured. The rate law obtained
for perchlorate media was
rate = kobs[V(III)][V(II)]
At 25 "C, at an ionic strength of 2.0 M and [H'] of 1 M , kobswas found to have
a value of 1.4 x lo-' I.mole-'.sec-'. The overall activation energy and entropy
were calculated as 13.2 kcal.mole-' and -25 cal.deg-'.mole-', respectively.
The rate of exchange was found to be dependent on the hydrogen-ion concentration (up to 5 x lo-' M ) i n a manner which led to these authors suggesting that
the exchange pathways

exist. Analysis c,f the exchange data obtained at an ionic strength 2.0 M led to
I.mole-'.sec-'
and 3 . 5 ~
sec-',
values of k , and k2K2 of 1 . 0 2 ~
respectively, where K2 refers to
V3++H20

+ VOH2+H+

K2

The estimated value of k , is 1.75 I.mole-'.sec-'.


to 3 x lo-' M , at constant acidity
Addition of chloride in the range 5 x
and ionic strength, was found to increase the rate of exchange. This was interpreted in terms of an exchange pathway involving the VCI2+ ion, viz.

The term k3K3 has a value of 1.4 12.mole-2.sec-' in 2 M perchlorate media of


constant acidity. Other pathways involving chloride are also possible.

8.1.2 The exchange reaction between V ( I V ) and V(II1)

Furman and Garner' have measured the rate of this exchange reaction in
aqueous perchlorate media using the isotope 48V as indicator. Numerous methods

77

T R A N S I T I O N METALS

of separation were tried, kinetic data being obtained using an ion exchange
technique. The rate law obeyed, over the concentrations range V(IV), 1 x lo-'
M , and in the absence of oxygen,
M , and V(III), 8 x
to 5 x
to 5x
was
rate

kobs[V(IV)][V(III)]

The observed rate coefficient kobswas found to be dependent on the hydrogenion concentration (plots of kobsversus [H+]-' were linear passing through the
origin), suggesting that only an acid dependent pathway for exchange occurs.
The pathway suggested was

VOz

+ VOH2+ -%

Using the available data for the equilibrium

V3++Hz0

+ VOH2++Hf

Kl

k , was estimated as 14 I.rnole-'.sec-' at 25 "C and ionic strength of 2.5 M . The


calculated activation energy and entropy obtained for this step were 10.7 kcal.
mole-' and -24 cal.deg-'.mole-'.
The presence of thiocyanate ions appears to catalyse this exchange but no
detailed measurements have been made.

8.1.3 The exchange reaction between V ( V ) arid V ( 1 V )

Tewes et al.', using both precipitation and extraction separation techniques,


observed essentially complete exchange within the separation time (1 min) during
an isotopic study ("'V) of this exchange. The media ranged from 0.3 M perchloric
M.
acid to 7 M hydrochloric acid; reactant concentrations were
More recently Giuliano and McConnell', using an NMR technique based on
line width changes of the 51Vline, have made some measurements of the exchange
rate. The rate law was found to be

rate = kobs[V( V)]'[V( IV)]


in chloride-perchlorate media. At concentrations 3 M CI- and 6.5 M H + , the
value of kobsis 1.5 x lo6 lz.mole-z.sec-'. The rate of exchange was found to be
dependent on the hydrogen and chloride ion concentrations.
References pp. 142-152

78

IONS OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

The mechanism proposed was

2 V(V)

(V(V)),

+( V ) ) ,

+ V(IV) 2(V(V>>,+ V(IV)

with the acid dependence being possibly due to the reaction

2 V0:+2

H+

+V , O ~ + + H , O

Concentrations of V(1V) and V(V) used in this study were


2 x lo- M , respectively.

1.5 x lo- and

8.1.4 Reactions between vanadium ions

A detailed study of the reaction

in perchlorate media, in the absence of air, has been made by Newton and Baker
using a spectrophotometric method, a wavelength of 760 mp (absorption by
V(1V)) being used to obtain rate data. The rate law, obeyed over a range of conditions, was
-d[ V( IV)] /d t

= - d [V( I I)]/d t =

koJ V( IV)] [V( II)]

with no retardation being produced by addition of V(II1). A small but definite


increase in the rate coefficient (kobs)was observed as the hydrogen-ion concentration was increased at constant ionic strength. This effect led to a more detaikd
rate expression
rate

(k + k[H+])[VO+] [V21

and the conclusion was that the main step was

with the possibility of a step

VO2++V++Hf +

or a medium effect operating. For the process represented by k the activation

T R A N S I T I O N METALS

79

enthalpy and entropy were 12.3 kcal.mole-' and - 16.5 cal.deg-'.mole-', and
for that represented by k", 9.8 kcal.mole-' and -31.6 cal.deg-'.mole-'.
During this study, an intermediate absorbing at 425 mp was detected and shown
in a further study2 to be a dimer (VOV4+), with nearly two-thirds of the V(1V)V(I1) reaction proceeding uia this species in an inner-sphere step, the remainder
reacting via an outer-sphere pathway. The mechanism proposed for the reaction
was

V 0 2 ++ V 2 + + 2 H + -P 2 V3++ H 2 0 (k,)

vo2++v2++ vov4+
VOV4+ + H +
VOH2

+ V 3 ++VOH2+

+ H + + V3++ H,O

(k2)
(k3)
(rapid)

From spectrophotometric measurements (at 425 mp) on the appearance and disappearance of dimer, values for the rate coefficients k2 and k , were found to be
and 3.3 x lo-' l.mole-'.sec-',
(at 0 "C and ionic strength 1.0 M ) 6.7 x
respectively. Under the same conditions k , was calculated to be 3.9 x
1.
mole - .sec - I .
Both sulphate and chloride ions were found to accelerate the reaction under
constant ionic strength conditions and, although medium effects may operate,
pathways involving chloride and sulphate ions are possible. For the sulphate ion
addition Newton and Baker' conclude that the step

is a h a y s present. Olver and Ross3, however, during a study of the catalytic


polarographic reduction of V(IIT), were able to obtain data suggesting the dependence of the rate of this reaction on [HSO;] rather than on [SO:-] as above,
and suggested a pathway involving this ion. The rate law proposed was
rate = k'[V02+][V2+]

+ k4[VOHHSO:+][V2+]

Daugherty and Newton4 have examined the reaction between VO: and V3+
and have found their results to be consistent with a rate law

the major part of the reaction proceeding via a pathway associated with the
coefficient k 5 .
References pp. 142-152

80

IONS OF SAME METAL I N DIFFERENT O X I D A T I O N STATES

The reaction
V2++VO:+2Hf

V3++V02++H20

has been investigated by Espenson and Krug', using the stopped-flow technique.
Results indicating a rate law

were obtained. The values of the rate coefficients k,(l.mole-'.set-') and kb(I2.
mole-2.sec-'), at 25 "C and p = 1.01 M , with the associated activation enthalpies
(kcal.mole-') and entropies (cal.deg-'.mole-')
in parentheses are 2.58 x lo3
(1.9, -36.8) and 2.16 x lo3 (1.8 and -37), respectively.

8.1.S Reactions between tantalum cluster ions


Espenson and McCarley' have estimated the rate coefficient for the reaction

from spectrophotometric observations on mixtures of the two reactants to be


greater than lo5 l.mole-'.sec-'.
For the reaction

estimate^^,^ for the rate coefficient of 2 5 x lo7 and 6.8 x lo7 I.mole-'.set-'
have been recently reported.
For the reaction

the value of the rate coefficient proposed2 was 2 10' l.mole-'.sec-'.

8.2

CHROMIUM, MOLYBDENUM A N D T U N G S T E N

8.2.1 The exchange reaction between Cr(III) and Cr(IZ)


The earliest estimate of the rate coefficient for the exchange in perchlorate media
was obtained by Plane and Taube' from data obtained on the catalytic effect of

81

TRANSITION METALS

CriJ on the Cr(H,O):+-H,O exchange reaction. A value of 4.7 x


1.mole-I.
sec-' was proposed.
Anderson and Bonner' made the first detailed kinetic study on the exchange
using the isotopic method ("Cr) and a separation method based on the conversion
of Cr(I1) into Cr(II1) oxalate and an ion-exchange treatment. To prevent oxidation
of Cr(I1) during exchange a hydrogen atmosphere was maintained over the reaction mixture. The rate law found to be obeyed for the concentration ratio range
Cr(III)/Cr(II) of between 3.3 x lo-, and 2.0 in perchlorate media was
rate = kobs[Cr(IIl)][Cr(II)]

From data obtained at p


1.0 M with [H'] varying over the range 0.213 to 1.0
M , kobswas found to be given by

which was considered by Anderson and Bonner to arise from the exchange pathways
cr3+ +cr2+

CrOH2++Cr2+-%The species CrOH2+ being formed by the hydrolysis

+ CrOH2++H+

Cr3++H20

K2

with k ; given by the product K2k2. On this basis k , and k , were evaluated as
2 x lo-' and 7.0 l.mole-'.sec-', respectively (at 24.5 "C p
1.0 M ) . The value
obtained for the overall activation energy was 22 kcal.mole-'. For the second
step (k,) a hydrogen atom transfer mechanism was suggested.
The effect of chloride ion on the exchange was found by these workers to be
very small, whereas Plane and Taube' had estimated a rate coefficient about
five times larger in the presence of lo-' M chloride ion than in perchlorate
solution. Van der Straaten and Aten3 have studied the exchange in media 1 M
with respect to HCl and have estimated a rate coefficient 2 3.0 x lo-, I.mole-'.
sec-l. The isotopic method (51Cr)and a separation procedure based on the precipitation of Cr(I1) as the acetate complex was used.
The exchange reactions

whereXis, F-, C1-, Br-, N;, NCS-, CN-, and H2PO; have beeninvestigated4-*.
ReJerciices pp.

142-IS2

82

IONS OF SAME METAL IN D I F F E R E N T O X I D A T I O N STATES

The radioisotope "Cr was used to tag the C r 2 + species and the separation of
C r 2 + from CrX2+ was achieved using an ion-exchange method, after oxidation
of the C r 2 +to Cr3+ with Fe(II1) or oxygen. The reactions were carried out in the
absence of oxygen in perchlorate media. For the systems involving chloride,
fluoride4, and azide6, King et al. have found a rate law
rate

k3[CrX2+][Cr2+]

to be obeyed over a range of conditions including variation in the [H+1. Values


of the rate coefficient k , (I.mole-'.sec-'), conditions in parentheses, are, for
X = F - , C1-, and N; 2.5 x lo-,, 9.1 (0 "C,p = 1.0 M ) a n d 1.3 (0 "C, p = 0.5 M ) ,
respectively. Assuming a similar rate law when X = Br- and NCS- the calculated
rate coefficients were > 60 (0 "C, p = 1.0 M ) and 1.8 x
(27 "C, p = 1.0
M ) I.mole-'.sec-', respectively. For X = C1- a value of 8.3 I.mole-'sec-' (0 "C,
p
1.0 M ) has been obtained by Taube and King'. Activation parameters
have been obtained for the systems CrF2+-Cr2+ and CrN:+-Cr2 +; these are
AH' = 13.7 and 9.6 kcal.mole-' and AS' = -20 and -22.8 cal.deg-'.mole-',
respectively4,'.
The transition states proposed for these exchange systems involve anion
bridges4, '. For the system CrNCS2+-Cr2+,Ball and King4 suggest that either a
nitrogen bridged transition state or a two stage process

CrNCS2++ C r 2 + + C r 2 ++CrSCN2+
CrSCN" + C r 2 + -+ C r 2 ++CrNCS2+
occurs.
Espenson et aI.'s8 have found for X
dependent rate law
rate

H2PO; and CN- a hydrogen ion

k,b,[~r~2+][~rZ+]

where the observed rate coefficients are given by kobs = kk [H+]-' for X = H,PO;,
(over the range of concentrations, CrH2PO:+ 3.7 x lo-, to 4.0 x
M, Cr2+
1.8 x lo-, to 2.5 x
M , and H + 2.6 x
to 1.0 M ) and by kobs= k , +kk
[ H + ] - ' for X = C N - (over the concentration ranges, CrCN2+ 4.5 x lo-, to
M , and H' 1.2 x l o F 2to 3.5 x lo-'
M , Cr2+ 1 . 0 ~
2.6 x
to 1.9 x

1.

sec-' (25 "C,


Fot the hypophosphito complex, k; has a value 6.11 x
p = 1.0 M ) with the associated activation enthalpy and entropy values of 19.7
kcal.mole-' and -7.2 cal.deg-'.mole-'.
For the exchange involving the monocyanochromium(III), values of k , and
kk are 7.7 x
I.mole-'.sec-' and 4.2 x lo-, sec-' at 25 "C and p = 1.0 M ,

TRANSITION METALS

83

respectively. The activation parameters corresponding to these steps are AH:


9.3 kcal.mole-', AH:. 17.2 kcal.mole-', AS: -32 cal.deg-'.rnole-' and AS:,
- 8 cal.deg-'.mole-'. For the steps associated with kk and k i transition states
having bridges involving the ions X - and OH- have been proposed. For the
pathway associated with k , a mechanism

has been suggested.


Snellgrove and King' have examined the exchange reaction
k7

cis-Cr(N,): +Cr'+ +
and have suggested two azide groups participate in the transition state bridging.
The isotopic method ( 51 Cr) and ion-exchange separation of reactants (after
treatment with HzOz)was used. The observed second-order rate coefficient has
a value 60I.mole-'.sec-' (0 "C, p = 0.5 M ) . The effect of increasing the hydrogen
ion concentration was very small.+
Stranks' has reported the activation energy and entropy for the exchange
system Cr(urea);+-Cr(urea):+, to be 13 kcal.mole-' and -40 cal.deg-'.mole-'.

8.2.2 The exchange reaction between Cr( V l ) and Cr(II1)


The early studies on the exchange system, carried out by Muxart et d.',
Menker
and Garner', and Burgus and Kennedy3, showed the exchange to be slow in
sulphate'. ', perchlorate', nitrate3, and hydroxide media', '. The isotopic method
("Cr) was used with separation of the Cr(II1) and Cr(V1) being achieved by the
precipitation of chromic oxide' and lead chromate3. Some evidence was obtained
for the retardation of the exchange by hydrogen ions.
Altman and King4 made the first detailed study of this system, using Cr(II1)
solutions containing only the monomeric species and Cr(V1) solutions which had
been allowed to age. The isotopic method and lead chromate precipitation separation were used to obtain kinetic data at a temperature of 94.8 "C. Over the range
to 8.4 x lo-' M , Cr(II1) 1.8 x
to
of concentrations, Cr(V1) 2.3 x

+ Data for the exchange of 5'Cr between Cr2+ and the Cr(II1) species Cr(OAc)2+ and cis-

Cr(0x); have also been reported".


References p p . 142-IS2

12.

84

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

8.5 x lo- M , H + 4 x lo- to 8.3 x lo- M , a rate law


rate = (k+k[H+]-2)[Cr3+]43[H2Cr04]23
was obtained, with values of k and k (p = 0.91 M and 94.8 C) of 1 . 4 ~
I.mole-.sec- and 6.6 x 10- mole.l-.sec-, respectively. The exchange was
thought to occur via a slow step, either
Cr(III)+Cr(V)

+ Cr(V)+Cr(III)

Cr(II1) +Cr(V)

+ Cr(IV)+ Cr(IV)

or

with the other steps


Cr(V) + Cr(V1) + Cr(VI) + Cr(V)
Cr(n1) + Cr(V1)

+ Cr(1V) + Cr(V)

also occurring.
8.2.3 The reaction between Cr( V I ) and Cr(IZ)

Hegedus and Haim have examined the reaction of Cr(V1) with Cr(I1)
3 Cr2++Cr(VI) = 2 Cr3++Cr2(0H)42+

using the isotopic method (51Cr)in an attempt to verify the reaction mechanism
proposed by Ardon and Plane, viz.

Cr(V1) Cr2

Cr(V)+Cr+

Cr(V) + Cr3

-+

Cr(IV)+Cr3i

Cr(IV)+ Cr2+ + Cr,(OH);+


Separation of the products was achieved using an ion-exchange method, the reaction taking place in sealed vessels with the Cr(V1) labelled (51Cr). Under various
conditions only 40 % of the Cr appears in the product Cr2(OH)if. This led
these workers to propose a modified mechanism in which the step involving
Cr(V) proceeds in two ways

+
Cr(V) + Cr(1I)

Cr(V) Cr(I1) + Cr(IV) + Cr3+


+

Cr3

+ Cr(1V)

TRANSITION METALS

85

the former step being about four times faster than the latter. No kinetic measurements have been made.

8.2.4 Cr(II)-catalysed substitution and isomerisation reactions of Cr(ZII)


A number of aquation reactions of the type

have been studied and found to be catalysed by Cr(II), usually in the form (3::.
The reactions are usually carried out in a sealed container, in perchlorate media
and in the presence of an inert gas which prevents atmospheric oxidation of the
Cr2 ;the rate data are usually obtained from spectrophotometric measurements.
Studies have been made on the complex ions Cr(II1)X for X = C1- (Taube and
King', Adin and Sykes', and Pennington and Haim3), Br- (Adin et aL4, and
Pennington and Haim,), I- (Adin et uL4, and Pennington and Hairn,),' F- (Adin
et aL4), N; (Doyle et al.'), and NH, (Espenson and Carlyle6). The wavelengths
used in the absorbance studies were 608 mp (Cl-), 622 mp (Br-), 306 or 475 mp
(I-), 408 mp (F-), 433 mp (N;), and 504 to 522 mp (NH,).'+ Since these aquation
reactions can also occur via pathways which are not catalysed by Cr(II), terms
from both catalysed and non-catalysed reactions appear in the rate law, which is
usually of the form
+

d[Cr(III)X]/dt

(k, + k;[H+]-'

+ k'[Cr"])[Cr(III)X]

The rate coefficients k , and k; relate to the non-catalysed process and k' is the
apparent rate coefficient for the catalysed pathways. For the systems where X = C1-,
Br- and I - , k' is given by k ; [H'I-', and a reaction pathway
CrX( OH)+ + Cr2+%
was postulated, with the species Cr(X)(OH)' produced oiu the equilibrium

CrX2+ H,O

+ Cr(X)(OH)+ + H'

K,

and k ; = k,K,. Values of the observed rate coefficient k ; (sec-') are 3.2 x
(Cl-)',
1.74~
(Br-)4, 2.13 x
(I-)4 (at 25 "C, p = 2.0 M ) and
4x
((21-),
2 x lo-, (Br-), and 2 . 2 ~lo-' (I-), (p = 1.0 M , 25 'c),.
Activation parameters (AH* and AS*) of 20.2 kcal.mole-', -7.4 cal.deg-'.mole-'

+ See also Pennington and Haimz3.


t t The reaction for X = CN- has also been investigatedz4.
References pp. 142-152

86

I O N S OF S A M E M E T A L I N D I F F E R E N T O X I D A T I O N S T A T E S

(CI-)', 18.2 kcal.mole-', - 10 cal.deg-'.mole-' (Br-),, and 17 kcal.mole-',


-9.4 cal.deg-'.mole-' (I-)4 were also obtained. For the reaction
C r C I Z + +Cr2++ Cr'++C1-+Cr3+
k3=

- 3a

K3e

the related value of K,, has been determined and an estimate of k,, (at 20 "C,
p = 2.0 M ) of 20 I.mole-'.sec-' has been made'. The observed rate coefficient
k - 3 a has been evaluated as 3 . 4 lo-,
~
1'. mole-'. sec-' ([H'] = 1 . 0 ~
lo-' M
and CI- = 1.0 M a t 40 0C)2.
Sykes et
have proposed a transition state involving a hydroxide bridge
for the processes defined by k , .
For the systems X = F-, N;, and NH,, two catalysed pathways were indicated by the rate law, which took the same form as previously, with
~

k'

= (k;[H+]-'+k,)

Values of k j and k , (at 25 "C and ,u = 2.0 M ) obtained were, 8 . 6 ~


sec-'
sec-' and 5 . 2 ~
and 4 . 4 ~
I.mole-'.sec-' (F- at 25 0C)4, 3 . 4 ~
I.niole-'.sec-' (N; at 40 0C)5and 5.9 x
sec-' and 2.4 x
I.mole-'.sec-'
( N H, at 25 T ) ' , respectively. Activation parameters calculated by Espenson and
Carlyle' for the CrNH:+-Cr2+ system are (AH' and AS') 21.6 kcal.mole-',
-5.6 cal.deg-'.mole-' ( k ; ) and 13.9 kcal.mole-', -33 cal.deg-'.mole-' (k4).'
Transition state complexes proposed for the steps associated with k , are analogous
to those given previously for the chloride, bromide and iodide complex ions. For
the pathway defined by k,, transition states [CrOHCrXH4+]' and [CrH,OCrX"+]'
(X = F- or N;)4-5 and' [CrH20CrNH35f]'have been suggested.
For the reaction between CrN:' and C r 2 + it was found necessary to allow for
the reaction

2 C r 2 + + 2 H + + H N 3 = 2Cr"+NH3+N2
before evaluation of the rate data from the absorbance measure~nents~.
Aquation reactions of some disubstituted aquo ions of Cr(II1) have also been
found to be catalysed by C r 2 + ,viz.

Kinetic studies on the reactions (in the absence of oxygen, in perchlorate media)
have been made'-'' for the ions C r X l where X = CI-, F-, N; and CN-;
the rate data was obtained using spectrophotometric measurements for the
Deutsch and TaubeZ5and Nordrneyer and Taube26 have recently obtained data for the similar
Cr(1II)X systems, X = acetate25and nicotinarnidez6.

87

T R A N S I T I O N METALS

chloride (cis-,245 to 260 mp), (trans-, 448 or 635 m ~ )azide


~ , (cis-,275 m , ~and
) ~
cyanide (cis-,460 mp)" complex ions and the isotopic method (51Cr) with ionexchange separation for the fluoride complex ion'. For these reactions a transition
state, involving the ion X forming a single
has been suggested by the
form of the rate law
-d[CrX:]/dt

= k"[CrX:][CrZ+]

applicable to the catalysed pathway. The non-catalysed aquation reaction was


found to be very much slower than the above reactions.
For the ion CrCI:, Espenson and Slocum7 have obtained values of k" (I.mole-'.
sec-I), at 25 "C, p = 1.0 M , of 401 (trans-) and 288 (cis-) with corresponding
activation parameters (AH', kcal.mole-' and AS', cal.deg-'.mole-') of 4.9 and
-30 (trans-) and 5.7 and -28.2 (cis-), respectively. The value of k" was found
to be independent of [ H + ]in the case of the trans-isomer only. Previously, Taube
and Meyers" had estimated a value of 166 I.mole-'.sec-' at 2 "C (media 1 M
in HC104) and Johnson and ReynoldsI2 had obtained a value of
100 Imole-'.
sec-' for cis-CrC1:). For the reverse reaction

CrC12++ Cr2

+ CI- = CrCI: + Cr2+

Taube and King' have obtained a rate coefficient of (50.2 12.mole-2.sec-' at 0 "C
(media 1 M in HCIO,).
Haim' has found a value of k" of 7.6 I.mole-'sec-' at 25 "C (media 0.2 M in
HCIO,) for the aquation of cis-Cr(N,):. Corresponding activation parameters
of 8.1 kcal.mole-' and -27 cal.deg-'.mole-' werealso obtained. For the aquation ofcis-Cr(CN)l, Birk and Espenson" have reported a value of 4.19 I.mole-'.
sec-' at 25 "C (p = 1.0 M ) .
I.mole-'.sec-' at 25 " c (cisChia and King' have evaluated k" as 1 x
CrF:) and 1.5 x
I.mole-'.sec-' at 25 "C (trans-CrF:). The activation
enthalpq and entropy for the aquation of cis-CrF: are 13 kcal.mole-' and -24
cal.deg-'.mole- ' , respectively.
The species Cr(CN), also undergoes a catalysed aquation
Cr(CN),+Cr2+

C r 2 + + C r C N 2 + + 2CN-

with a second order rate coefficient of 4.56 I.mole-'.sec-' at 25 "C (p = 1.0


M )' '.+
Cr" also catalyses some substitution reactions of the species Cr(NH3)5X2f,
viz.
Cr(NH3)5X2++5H +

C r X 2 + + 5NH:

+ For the similar system involving the species Cr(0x);- see Hutchital".
Relerrnces pp. 142-15.2

88

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

where X is the fluoride, chloride, bromide or iodide group. Ogard and Taube',,
using a spectrophotometric method (detecting the appearance of the product
CrX") have obtained rate data for these reactions, which were carried out in
perchlorate media in the absence of oxygen. The rate law found was
- d [Cr(NH3),X2+]/dr =

k, [Cr(NH,)5X2+]+ k, [Cr(NH,),X'+][Cr2+]

where k , relates to the non-catalysed reactions


Cr(NH,),X2+

Cr(NH,)z+ +X-

and k, (the rate coefficient for the catalysed steps) has values (1.mole-'.sec-'
at 25 "C, p = [ H + ] = 1.0 M ) of 2.7 x l o W 4(F-), 5.1 x
(Cl-), 0.32 (Br-),
and
5.5 (I-). For the species Cr(NH,),Cl2+ k , was found not to be affected
by variation in [H'] but was reduced by change in the solvent from H 2 0 to D 2 0 .
Activation parameters of 13.4, 11.1, 8.5 kcal.mole-' and -30, -23, -33 cal.
deg-'.mole-' were calculated for the species Cr(NH3),F2+, Cr(NH,),Cl'
and
Cr(NH,),Br2+, respectively. The reaction (NH,),CrOHCr(NH,)4C14+ 5H+ =
(NH3),Cr2++5NH: +CrCIZ+ has been found by Hoppenjans et
to be
catalysed by Cr2+.
Cannon14 has made a study of the Cr(I1) catalysed reaction

Cr(NH3),C12++ n OAc-

+ 5 H'

C ~ ( O A C ) ~ ~ -5"NHf
)+

+ C1-

which occurs in acetate buffer. Absorbance measurements at 593 mp were used to


obtain rate data. The rate law observed was
d[Cr(OAc),]/dt

k"'[Cr(NH,)5C12+][Cr(II)]f

from which it was concluded that the mechanism was


Cr2(OAc)iCr(0Ac); +Cr(NH,),C12+
Cr(OAc),CI-

+ 2 Cr(0Ac);
-%Cr(OAc),Cl--+

Cr(OAc),

Cr(OAc),

--+

polymer

+ Cr(0Ac);

Cr( 11) 3 OAc-

K,

+Cr(II)+5 NH;

+ C1-

Using the value of K , , obtained from spectrophotometric measurements, k-, was


calculated as 1.2 l.mole-'.sec-' (25 "C,p = 1.0 M ) with an associated activation

T R A N S I T I O N METALS

89

energy of 12 kcal.mole-'. For the reaction

an estimate of k , 5 0.13 1.mole-'.set-' at 45 "C (p = 1.0 M ) was made.


De Chant and Huntl5'l6 have recently reported some kinetic data on the reactions
cis- or frans-Cr(NH,),Cl2+ + 4 H +

CrCl" +4 NH,f

which are also catalysed by Cr2+.In perchlorate media in the absence of oxygen
a rate law
rate

k , [Cr(NH,),C12+][Cr2+]

was found. In the presence of chloride ions the rate law became
rate

( k , kb[C1-])[Cr(NH3),Cl2+][Cr2+]

Values of the kinetic parameters k,, AH' and AS' are 1.16 I.mole-'.sec-' at
25 "C (p = 1.0 M ) , 9.7 kcal.mole-' and 26cal.deg-'.mole-' for the truns-isomer
and 1.14 x lo-' l.mole-'.seL-', 10.6 kcal.mcle-' and -27.6 cal.deg-'.mole-'
for the cis-species. Both pathways associated with k , and kb are thought to occur
via chloride bridge transition states.
Pennington and Haim" have investigated the similar Cr2+-catalysedaquations
of the cis- and truns-Cr(en),C12+, and the cis- and trans-Cr(en),Cl: species of
Cr(III), in perchlorate media, uiz.
cis- or trans-Cr(en),Cl;

+ Cr2++ 2 H + = Cr2++ 2 enH+ +C1- +CrC12+

cis- or truns-Cr(en),C12+ +CrZ++2 H + = C r 2 + + 2enH+ +CrC12+

In the absence of oxygen the rate law obtained from spectrophotometric measurements at wavelengths 382 mp (trans-isomers) and 510 or 512 mp (cis-isomers)
was found to be of the form
-d[X]/dt

k,,[X][Cr2+] +k;,[X]

where X represents the Cr(II1) species mentioned and k,, and k',, are the rate
coefficients for the catalysed and non-catalysed pathways, respectively. (Only
for the cis-isomers was the term involving k ; , large enough, under experimental
conditions, to appear.) Values of the kinetic parameters k,, (l.mole-'.sec-' at
25 "C,p = 1.0 M ) , AH' (kcal.mole-I) and AS'(cal.deg-'.mole-') are 1.16 x
References pp. 142-1 5 2

90

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N S T A T E S

10.2and -33 for cis-Cr(en),CI2+, 1.95 x


9.0 and -36 for cis-Cr(en),CI:,
4 . 3 4 lo-',
~
8.0 and -33 for frans-Cr(en),C12+ and 5.09 x lo-', 8.7 and -31 for
trans-Cr(en),Cl:, respectively. A chloride bridge transition state was again suggested.
The Cr(I1)-catalysed substitution reaction between Cr(II1) and N-methyliminodiacetic acid H,L
Cr(III)+L2-

CrL+

has been studied by Cannon and EarleyL8.A rate law


d[CrL+]/dt

k"" [Cr3'][Cr2+][HL-] [ H + ] - 2

was found to account for the experimental data. At 25 "C and p 1 .O M , k"" has
sec-'. The mechanism proposed by these workers was
a value 2.2x
H,L ;r HL- + H i
HLCr2 + + L2 Cr3++ H,O

+ L2-+H'
$ CrL
_-

CrOH2++ H t

CrL+CrOH2+ -+ C r L + + C r 2 + + O H which was also concluded to occur in the presence of N-phenyliminodiacetic


acid. A spectrophotometric technique (546 mp) was used to obtain rate data.'
Cr(l1) in the form of the EDTA complex has been found to undergo a very fast
reaction with the Cr(II1) species CrX2+(X = F, C1, or Br). The products are
Cr(II1) EDTA complex ions". However, the kinetic parameters for the reaction
between the Cr(lI1) complex with EDTA and the Cr(I1) complex with frans-1,2diaminocyclohexanetetraacetate have recently been reported by Wilkins and
Yelin", as 3.0x lo3 l.mole-'.sec-', at 25 "C, 5.0 kcal.mole-' and -26 cal.deg-'.
mole-'.
Cr(I1) catalysed isomerisations of the ions CrNC2+ and CrSCN"
CrNC" + C r 2 +

C r 2 ++CrCN2+

CrSCN2++ C r 2 + = Cr2++CrNCS2+
have been studied using spectrophotometric techniques at wavelengths of 520 mp

91

T R A N S I T I O N METALS

and 260 to 350 mp, respectively20-22. For the ion CrSCN"


have been proposed (k,,,k13)

twocatalysed pathways

CrSCN2++Cr2+%CrSCN2++H,0

+ Cr(SCN)(OH)++H+

K,,

Cr(SCN)(OH)+ +Cr2+
the observed rate coefficient being given by

Values of k,, and k,,K,, at 25 "C (p = 1.0 M ) were calculated by Haim and
Sutin2' as 40 I.mole-'.sec-' and 2 sec-', respectively;
For the species CrNC2+,Birk and Espenson22 have found a rate law
-d In [CrNC*+]/df

kob3 = k',,+k;,[Cr2+]

where the terms k i 4 and k',, [Cr"] relate to the non-catalysed and catalysed
processes. The value of k i 5 is 5.7 x lo-' I.mole-'.sec-' at 15 "C ( p = 1.0 M ,
[H'] = 4 x l o - ' M ) . Hutchital" has recently obtained kinetic data for the
Cr2
reaction trans-Cr(Ox), =-cis-Cr(0x);.
+

8.2.5 The exchange reaction between Mo( V ) and M o ( f V )

The exchange of Mo between the anions Mo(CN);- and Mo(CN);- has been
investigated by the isotopic method (99Mo) and the separation methods (a)
precipitation of Mo(CN);- with either ethanol or cadmium ions, and ( 6 ) precipitation of Mo(CN);- with tetraphenylarsonium chloride. Complete exchange
5x
M.
was observed by Wolfgang' even with reactant concentrations
An estimate of the rate coefficient at 2 "C of > lo3 I.rnole-'.sec-' has been sug-

gtsled.
More recently, a value of 3 x lo4 I.mole-'.sec-' has been calculated for the
exchange rate coefficient at 10 "C and zero ionic strength by Campion et a/.'
using the Marcus theory and rate coefficients for the reactions
Mo(CN)i-

+ Os(dipy):+ $
Mo(CN);- + Os(dipy):+
kr

k, and k, for the above reactions were measured by the temperature jump method.
ReJerences pp. 142-IS2

92

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

8.2.6 The exchange reaction between W ( V ) and W ( Z V )


Using the radio-isotope lS5Was the indicator, and a direct injection technique
the exchange between the anions octacyanotungstate(V) and octacyanotungstate
(IV) has been investigated by Goodenow and Garner'. Tetraphenylarsonium
chloride was used to precipitate the W(CN)i- from reaction mixtures. In the
absence of light, in acidic (HC1) or alkaline (KOH) media with reactant concentrations in the range
to
M , complete exchange was observed at tempera1 "C. On this basis a rate coefficient of > 4 x lo4 Lmole-'.set-' has been
tures
proposed.
From a study of the paramagnetic line broadening of aqueous
M)
solutions of W(CN)i- by W(CN):- a limit for k 6 4 x lo8 1.mole-'.sec-' has
been set by Weissman and Garner'.

8.3

MANGANESE

8.3.1 The exchange reaction between Mn(ZZ) and Mn(Z)

Using nuclear magnetic resonance line-broadening measurements (55Mn),


Matteson and Bailey' have studied the exchange between the isonitrile complex
ions of Mn(I1) and Mn(I), (RNC)6Mn2+and (RNC)6Mn+ (where R is either the
ethyl or tertiary butyl group) in the solvents acetonitrile, ethanol and dimethylsulphoxide. In acetonitrile, the evaluated rate coefficients (1.mole- '.sec-l) for the
exchanges at 7 "C, are 6.4 x lo5 (R = ethyl) and 4.0 x lo4 (R = t-butyl). Activation parameters, A H t and A S t , of 1.7 kcal.mole-' and -25 cal.deg-'.mole-'
(R = ethyl) and 4.6 kcal.mole-' and -21 cal.deg-'.mole-' (R = t-butyl) have
also been reported. The exchange in the solvent dimethylsulphoxide was found to
be more rapid than that in ethanol or acetonitrile (see also Matteson and Bailey').
8.3.2 The exchange reaction between Mn(IIZ) and Mn(ZZ)

Polissar' has observed 100 % exchange between Mn(III), as the oxalato ion
Mn(0x); , and manganous sulphate. Adamson', using manganic chloride as the
source of Mn(III), has observed the exchange to be incomplete in a time
15 sec
M. Both workers',' used a separation
with reactant concentrations
technique based on the precipitation of Mn(lV), present via the equilibrium

2 Mn(II1)

+ Mn(II)+Mn(IV)

as the oxide, with perchloric acid solutions as reaction media and the isotope
54Mn as the indicator.

93

TRANSITION METALS

Diebler and Sutin3, using perchlorate salts of Mn(II1) and Mn(1I) and the
isotopic method, have also observed complete exchange, in media 6 M in HClO,,
within a time of one min. Two additional separation techniques were employed:
extraction of the Mn(II1) into benzene as the complex dibutylphosphate and precipitation of the Mn(I1) as the salt MnNH,P04. However, using Marcus theory,
Diebler and Sutin3, have been able to calculate a value for the exchange rate
I.mole-'.sec-', from data obcoefficient at 25 "C (3 M in HC104) of 3 x
tained during a study of the reactions of Co(1II) with Mn(I1) and of Mn(II1) with
numerous Fe(I1) complexes.
Adamson, has reported the exchange between the manganese anions Mn(CN)i and Mn(CN);- to be measurable.
The kinetic parameters for the reaction between the Mn(II1) complex with
EDTA and the Mn(I1) complex with truns-l,2-diaminocyclohexanetetraacetate
have recently been reported by Wilkins and Yelin5 as 1.2l.mole-'.sec-', 7.1
kcal.mole-' and -34 cal.deg-'.mole-'.

8.3.3 The exchange reaction between Mn( VZZ) and Mn( VZ)

This exchange in alkaline media has been studied, using the isotopic method
(54Mn and 56Mn),by many workers. Libby', using the separation afforded by the
insolubility of barium manganate, has found 100% exchange. Hornig et aL2,
attempted to study the exchange using pyridine extraction of the permanganate
but observed 100 % exchange, as did Adamson3, who in addition to the above
separation methods, used a method based on the precipitation of manganese
dioxide. Bonner and Potratz,, using reactant concentrations
lo-, to
M
and extraction of Mn(VI1) as the triphenylsulphonium salt with either ethylenedichloride or chloroform, were able to suggest a limit of > 1500 l.mole-'.sec-'
for the rate coefficient at 0 "C for the exchange in 2 M NaOH.
Sheppard and Wah15 were the first workers to make rate measurements on this
exchange reaction. They used triphenylsulphonium bromide and tetraphenylarsonium chloride as the separating agents; both reagents remove the permanganate
anion. The radio-isotope 54Mn was used to label the manganate ion. In a further
report, Sheppard and Wah16 give details of their special reaction vessel and the
quenching solution. The rate law obeyed is

where kobs has a value of 710 l.mole-'.sec-' at 0.1 "C for 0.16 M NaOH. The
energy and entropy of activation were calculated as 10.5 kcal.mole-' and -9
cal.deg-'.molerespectively.
The addition of various cations (Cs", K", Na' and Li') was found to alter

',

Ref;.rences pp. 142-152

94

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

the value of kobs,the highest value being observed in the presence of Cs'; various
anions were also added but only slight effects were noted. The suggestion of a
cation bridge transition state was made.
A detailed investigation of the catalysis by the caesium cation has been made
by Gjertsen and Wahl', who were able to show from the linear plots of kobstlersus
[ C s + ]that the rate law could be rewritten as
rate

(k' + k"[Cs +I)[Mn O i l [ Mn O:-]

At an ionic strength of 0.16 M the values of k' and k" (0 "C) were 7.1 x lo2 1.
mole- '.sec-' and 1.2 x lo4 12.mole-'.sec-', respectively. Sodium hydroxide was
used to maintain the ionic strength constant.
Meyers and Sheppard' have since made a study of this exchange using the
nuclear magnetic line-broadening method (55Mn). From the results obtained the
rate law
rate

k,,,[MnOi][MnO:-]

was confirmed and, as shown in previous work6, the value of kobswas found to
depend on both the concentration and nature of the cation. I n media 0.57 M with
respect to KOH, kobs(20 "C) has a value 5.6 x lo3 l.mole-'.sec-' ,with an associated activation energy of 8.3 kcal.mole-'.
Britt and Yeng, using the pulsed nuclear resonance technique, have obtained
exchange data comparable to those obtained by the isotopic method. An observed
rate coefficient at 0 "C and p = 1.06 M of 1.23 x lo3 I.mole-'.sec-' has been
calculated for the exchange in the presence of 1.0 M Na+.

8.3.4 The exchange reaction between Mn( VZZ) and Mn(ZZZ)


Polissar' was unable to detect exchange between Mn(VII), as MnO;,
Mn(III), as the complex ion Mn(Ox);, using the isotopic method.

and

8.3.5 The exchange reaction between Mn( V I I ) and M n ( I I )

Using the isotope 56Mnas a means of labelling the oxidation states of manganese,
Polissar' has made a preliminary study of the exchange reaction in perchlorate
media between Mn(VI1) and Mn(I1). He concluded that no measurable exchange
M ) in a time of 15 min.
occurred between these two oxidation states (The reactants were separated by the addition of sodium hydroxide which precipitated manganese dioxide. Adamson2, however, using reactant concentrations

95

TRANSITION METALS

lo-, M , has observed a slow exchange in media 3 M with respect to HCIO,.


The experimental results suggested a complicated rate law of approximately the
form

rate

kobs[MnOJ113[Mn2+]4/3CH+]4'3

with kobs having a value


7.2 x
gested that the mechanism was

l2.mole-'.sec-I

at 25 "C. Adamson sug-

Mn0,+3Mn2++6H+ $Mn02++3Mn3++3H,0
Mn3++H,0

MnO+ +2H+

Mn4++H20 $ Mn02++2H+

K,
K2

K3

M n O + + M n 0 2 +%exchange
Mn02++H,0

-+

MnO,(s)+2 H +

M n 0 2 + + M n Z + + 2 H +$ 2 M n 3 + + H 2 0
Mn4+ + M n Z + $ 2 Mn3+
which leads to a theoretical rate expression
rate

( k : K , Kf/3)f[Mn0,]f[Mn2+]3[H+]

which was considered to be in satisfactory agreement with the observations.


Happe and Martin3, using the radio-isotope 54Mn and a separation method
based on the precipitation of the ion MnO, with tetraphenylarsonium nitrate,
have observed a slow exchange in nitrate media. The experimental results, obtained
over the ranges [HNO,] 1 to 2 M , [MnO,] 2.7 x
to 5.7 x
M and
[Mn"] 8 . 9 ~
to 1 . 3 5 lo-,
~
M , show some agreement with the rate law
found by Adamson2.

8.3.6 The reaction o f M n ( VZZ) and Mn(ZZ)


The reaction of the permanganate and manganous ions can take place under
conditions where either MnO,(s) or Mn(III), which decomposes slowly, are the
main products, viz.
3 M n Z + + 2 MnO, + 2 H,O = 5 Mn02(s)+4 H +
4 MnZf +MnO, + 8 H +

5 Mn3+ + 4 H 2 0

The reaction to produce the hydrated manganese dioxide has been studied by
References pp. 142-IS2

96

I O NS O F SAME METAL I N D I FFE R E N T O X I D A T I O N STATES

many
The reaction in the presence of oxalate4 or periodate' ions has
also been investigated. Oxalato and iodato complexes of Mn(II1) are formed as
intermediates in the reactions.
Rosseinsky and Nico16 have investigated in detail the reaction in highly acidic
perchlorate media which leads to the production of Mn3+. Using [Mn(II)] in the
to 1 x lo-' M with [MnO,]
M the rate law, found from
range 2 x
absorbance measurements at a wavelength of 525 mp, is

-d[MnO;]/dt

k,,,,[MnO,][Mn(II)]*

The variation in the value of the observed rate coefficient kobswith acidity was
found to be of the form

At a constant ionic strength of 3.31 M , values of the rate coefficients k; and k;


at 24.4 "C are 2.04 12. mole-'.sec-'
and 3.05 13.mole-3.sec-', respectively, with
the corresponding activation parameters of -0.6 and f0.2 kcal.mole-' and
- 59 and - 56 cal.deg-'.mole-'.

8.4

IRON, RUTHENIUM AND OSMIUM

8.4.I The exchange reaction between Fe(ZZZ) and Fe(ZZ) in aqueous media
The earliest attempts to measure the rate of exchange between ferrous and ferric
ions in aqueous media utilised the diffusion separation technique. Little agreement was obtained by the different workers' -4. Diffusion separation factors,
found to be
OS',
1.42, 3S3 and
l.24, illustrate the difficulty of the
technique. The isotopes used to label the iron were either "Fe or 59Fe,and exchange was found to be complete in hours2r4 or many days1p3 in perchlorate
media.
Silverman and Dodson' made the first detailed isotopic study of this exchange
system using the separation afforded by the addition of 2,2'-dipyridyl at pH 5,
followed by the precipitation of the ferric iron with either ammonia or 8-hydroxyquinoline. Dodson6, using this separation method, had previously obtained
an overall rate coefficient of 16 1.mole-'.sec-' at 23 "C for 0.4 M perchloric
acid media. The exchange in perchlorate and perchlorate-chloride media was
found to conform to a rate law, first order with respect to both total ferrous and
ferric ion concentrations, with an observed rate constant (kobo)dependent on the
hydrogen-ion concentration, viz.

kobs = k l +k;[H+]-'

TRANSITION METALS

97

This dependence was interpreted in terms of an [H+]-independent pathway


ki

F e 3 + + F e 2 +-+
and an [H+]-dependent pathway
k2

F e O H 2 + + F e 2 ++
the FeOH2+ being produced by the rapid hydrolysis reaction

Fe3+ HzO $ FeOH2+ + H +

K2w

From intercepts and slopes of plots of kobs versus [ H + ] - l , values of k , and


k,Kzw(k;) were evaluated as 3.3 I.mole-'.sec-' and 3.88 sec-' at a temperature
of 21.6 "C.Using known values of KZw(1.43 x lo3 M a t 21.6 "C),k2 was calculated
as 2700 I.mole-'.sec-'. The activation energies obtained for the steps defined by
k , and k, were 9.9 and 7.4 kcal.mole-', respectively, with corresponding entropies
of activation -25 and - 18 cal.deg-'.mole-'. Reasonable agreement with these
results has since been reported by many workers during the course of their studies.
For the step defined by k, a hydrogen-atom transfer mechanism has been suggested by Dodson'. Eimer et a].', using the same technique, observed no effect
on the rate as oxygen present was varied from 4 x lo-' to 2 x lo-' M , thus
eliminating any oxygen dependent mechanism. Hudis and Dodson' have observed
a reduction in both rate coefficients k, and k, when heavy water was used as the
solvent. (At 7.1 "Cin DzO, k, = 0.7 l.mole-'.sec-' and k, = 765 I.mole-'.sec-').
It was concluded that a hydrogen-atom transfer occurred in both pathways.
Fukushima and Reynolds' have reinvestigated the deuterium isotope effect.
Using values of the equilibrium constant K 2 D for

'

obtained by spectrophotometric measurements under the same conditions as


their isotope exchange measurements, these workers obtained values of the rate
coefficients k,, and k,, (25 "C)

for theexchangeinDCI0,mediaatp = 0.5 Mof l.Ol.mole-'.sec-' a n d 3 . 0 4 ~lo3


I.mole-'.sec-', respectively. For the second step (k2,) the activation enthalpy
obtained was 11.5 kcal.mole-l with the entropy -4 cal.deg-'.mole-'.
Refereiices p p . 142-1 52

98

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

Reynolds and Lumry have discussed the role of water in this exchange and
have suggested, for both steps, a mechanism involving water bridges.
Horne has studied the kinetics of exchange in aqueous perchlorate media
at temperatures down to -78 C by the isotopic method (59Fe) and dipyridyl
separation. The same rate law in these ice media as in aqueous solution was
observed, although the acid dependence was small. Horne concluded that the same
exchange mechanism occurs in solid and liquid solvent. Evidence for a Grotthustype mechanism has been ~ummarised~.

8.4.2 The efSect of inorganic ions on the exchange reaction between F e ( l I I ) and
Fe(ll)
The effect of the addition of inorganic ions has been investigated using the
isotopic method (55Fe or 59Fe) and the 2,2-dipyridyl separation. The rate law
in the presence of inorganic anions is given by
rate

kobs[Fe(III)][Fe(I1)]

where kobsis the observed second-order rate coefficient, which is related to the
rate coefficients for the individual reaction pathways and the formation constants
of the species involved in these pathways. The latter constants must be known for
the evaluation of the rate coefficients and constant ionic strength conditions must
be maintained. The reaction pathways
Fe3++ Fe2+

ki
_f

FeOH2++Fe2+

kZ

__f

operative in perchlorate media, have to be taken into account.


The effect of chloride ions was investigated first by Silverman and Dodson.
These authors observed an increase in kobsas the concentration of chloride ion
was increased from 0 to 0.55 M in perchlorate media of constant acidity. The rate
expression found to fit the experimental data was
rate

kl[Fe3+][Fe2+] + k2[FeOH2+][Fe2+]+ k3[FeC12+][Fe2+]


k,[FeCl:][

where k , and k , are the rate coefficients for the reactions


FeC1 + F e 2 +
FeCl:

k3

ks
+ Fe2+-+

+
Fe21

TRANSITION METALS

99

From the estimated values of K3 and K4

+ FeCI2+
FeCI2++C1- + FeCl:
Fe3++C1-

K3

3.01 I.mole-'

K4 = 0.94 I.mole-'

for the exchange conditions, values of k, and k, were calculated as 29 and 51


Lmole-'.set-', respectively, at 20 "C ( p = 0.55 M ) . Entropies and energies of
activation associated with the steps involving chloride (k, and k,) are -24 and
-20 cal.deg-'.mole-' and 8.8 and 9.7 kcal.mole-', respectively.
Sutin et al.' have made a detailed study of the chloride-catalysed paths in
deuterated water media, the exchange occurring more slowly in this solvent than
in water. Accurate values of K,, in both water and heavy water, were obtained
by spectrophotometry and enabled more precise rate parameters to be calculated.
For aqueous media, values are 22.8 1.mole-'.set-' fork, ( p = 0.50 Ma n d 20 "C),
11.5 kcal.mole-' (activation energy), and - 15 cal.deg- '.mole (activation
entropy); for deuterated solvent the corresponding values are 9.1 l.mole-'.sec-',
13.2 kcal.mole-', and - 10.2 cal.deg-'.mole-'.
Sutin et aL3 have recently re-examined the pathways of exchange involving
chloride and have evaluated the rate coefficients as k, = 57.6 and k, = 159 I.
mole-'.sec-' at 25 "C and p = 3.0 M . These workers have pointed out that step
with coefficient k3 can occur in two ways

for which a value k ; = 12.1 I.mole-'.sec-'


obtained3, and

( p = 3.0 M and 25 "C) has been

with k3 being the sum of 2 k j and kj'. Calculations have led to the evaluation of
k;' as 33.4 I.mole-'.sec-' ( p = 3.0 M a n d 25 "C). Home4 has shown that chloride
paths are unimportant at low temperatures in ice media.
Hudis and Wahl' have examined the effect of fluoride ion on the exchange
rate and have found it necessary to include three fluoride exchange paths
FeF2++ Fe2+

+Fe2
FeF, + Fe2+

FeF:

k5
_f

k6
_f

k7
_f

Values obtained for the rate coefficients k , , k6, and k7 (I.mole-'.sec-' at 0 "C
and p = 0.5 M ) were 9.7, 2.5 and 0.5, with activation parameters for the first
References p p . 142-152

100

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N S T A T E S

two steps (k, and k 6 ) of AS' -21 and -22 cal.deg-'.mole-' and E, 9.1 and
9.5 kcal.mole- ', respectively. A hydrogen-atom transfer mechanism was proposed
for these steps.
Menashi et d 6 have examined the exchange when only the FeFZ+ of the
Fe(II1) fluoro-species was present and when the rate of formation of FeFZf and
the rate of the 59Fe exchange are of the same order of magnitude. The results
obtained led these workers to conclude that exchange occurs via an activated
complex [Fe2F4+]*.
Thiocyanate ion also provides additional pathways for exchange, viz.

The rate law found by Laurence', for the exchange i n the presence of thiocyanate
ion in the range 3.2 x lo-, to 8.6 x lo-' M was
rate = k,[Fe3+][Fe2+] +k,[FeOHZ+][Fe2+] +k8[FeSCNZ+][Fe2+]

+ k,[Fe( SCN):][Fe2 '3


At 25 "C and ionic strength 0.5 M, values of the rate coefficients kg and k , obtained were 41.5 and 7.6 l.mole-L.sec-', respectively, with the associated activation energies for these pathways of 7.9 and 8.6 kcal.mole-'. Conocchioli and
Sutin' have recently shown that, since k , = k,,+2 k , , + 2 k,,, where the rate
coefficients k , , , k,, and k , , are for the reactions
FeNCS2++FeZf $ FeNCS++Fe3+
FeNCS" + F e z +

2 FeZ++FeSCNz'

FeNCS"

$ Fe2++FeNCSZf

+Fez+

the value of (k,,+2 k , , ) is


30I.mole-'.sec-'. The value of k , , used was 10.5
l.mole-'.sec-' and k 8 was estimated to be 51.6 I.mole-'.sec-' under the same
conditions ( p = 3.0 M and 25 "C).
Horne and Axelrod, have suggested that this exchange may not be first order
with respect to each iron oxidation state at high thiocyanate concentrations. A
lower value for the rate coefficient k , and a slightly higher activation energy than
that quoted previcusly' have been reported', '.
An azide catalysed path
kl3

FeN:+ +Fez+ +

TRANSITION METALS

101

has also been found by Bunn et a l . ' o ~ l lwho


, have evaluated, from the kinetic
data and the measured formation constant of FeN:', the rate coefficient k , , as
4 . 7 5 ~lo3 l.mole-'.sec-' at 10 "C and p = 0.55 M . The rate coefficient in D 2 0
was found to be lower than i n H,O. Arrhenius plots were linear in the range 0-13
"C,but not above the higher temperature, and led to calculated activation energies
of about 14 kcal.mole-' for both solvents. Positive activation entropies, of 7.0
and 8.5 cal.deg-'.mole-' for H 2 0 and D,O media, were calculated for this pathway. A reaction mechanism of the inner-sphere type was postulated, viz.

3 Fe3+N;Fe2'
Fe3+ N 3Fez 2 Fe2 N 3 Fe3
Fe2+N;Fe3+ 2 Fe2++-N3Fe3+

Fe3+N; + F e 2 +

k-

14

k-IS

k-16

with the relative magnitudes of k-.14 and k,, accounting for the curvature of the
Arrhenius plot.
The presence of phosphoric acid also provides an additional pathway

Sheppard and Brown" have evaluated the rate coefficient k I 7 as 4180 ].mole-I.
sec-' with 0.53 M HCIO, at 20 "C. The overall activation energy and entropy
corresponding to the term K1,k17,where K17is the equilibrium constant of the
react ion
Fe3++HPOi-

+ Fe(HP0,)'

K17

were calculated as 15 kcal.mole-' and 6 cal.deg-'.mole-'.


The presence of bromide ions also leads to alternative pathways for exchange,

uiz.
FeBr2++ Fe2+ -%
FeBr:

+ Fez

Horne13 has obtained a value for k , at 0 "C ( p = 0.55 M ) of 4.9 I.mole-'.sec-'


with a corresponding activation energy of 8.0 kcal.mole-'. The activation energy
found for the second step (k19)was 9.6 kcalmole-'.
Four recent kinetic studies have been made of the exchange in the presence of
the sulphate ion'2,14-'6 , in perchlorate media, some measure of agreement being
observed between the results of the various workers.
'

Reference5 pp. 142-152

102

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

Reynolds and F u k ~ s h i m a 'have


~
made an extensive study of this system and
have interpreted their results, obtained with varying sulphate and hydrogen-ion
concentrations, in terms of the exchange paths

FeSO:

+ Fe2+ %

Fe(S0,); +Fe2+
FeOH2++ FeSO,
At an ionic strength of 0.75 M the calculated values of the rate coefficients k,,,
at 25 " C ) are 5 . 8 6 lo2,
~ 2.07 x lo4 and 1 . 4 lo6,
~
respectively. Although the rate of the step associated with k,, was found to vary
with the ionic strength of the medium, kzl and k Z 2would not appear to depend
on this property. In this study the concentration ranges used were Fe(I1) 6.2 x
to 3.16 x
to 1.1 x
M , SO:- 'up to 8 x loW3M ,
M , Fe(II1) 2 x
and H + 3.8 x lo-, to 1.25 x lo-' M .
Sheppard and Brown', have estimated a value of k,, of 9 . 8 0 ~10, 1.mole-'.
M
sec-' (at 28 "C and p = 1.0 M ) from data obtained with total iron
and sulphate ion up to 8 x lo-, M . Activation energy and entropy for the cornbined term K 2 , k,, , where K , , refers to

k,,, and k 2 , (Lmole-'.sec-'

Fe3++SO:-

+ FeSOl

K2,

were calculated as 13.5 kcal.mole-' and -2 cal.deg-'.mole-'.


Willix15 has estimated from his kinetic data a value of k 2 , of 360 l.mole-'.sec-'
(at 25.7 "C,p = 1.0 M ) and has evaluated the activation energy and entropy as
8.3 kcal.mole-' and - 18.7 cal.deg-'.mole-' for this pathway, using data on the
temperature dependence of the equilibrium K,,.
Bachmann and Lieser16 have estimated rate coefficients (at 25 "C, p = 1.0 M )
of 2.9 x 10, and 1.8 x lo4 1.mole-'.sec-' for the steps defined by k,, and k21.
The corresponding activation energies obtained for these steps were 13.8 and 15
kcal.mole-'. The entropy for the first step (k,,) was evaluated as -1.2 cal.
deg - '.mole-'.
Suggestions that the sulphate catalysed paths may involve a mechanism with
a sulphate-bridged activated complex, as opposed to a hydrogen-atom transfer
mechanism, have been made',.

TRANSITION METALS

103

8.4.3 The effect of organic ligands on the exchange reaction between Fe(III) and
Fe(II)

The effect of some organic acids on the exchange in perchlorate media has been
investigated' -3. Fumaric'. ,, benzoic and o-phthalic acids' have been shown to
cause little or no alteration in the rate, whereas acetic, succinic, carbolic', oxalic'r
and tartaric acids3 have an accelerating influence. The isotopic method (59Fe)
and the 2,2'-dipyridyl separation have been used. The rate law observed was1-,
rate

k,,,[Fe(III)][Fe(II)]

At constant ionic strength and acidity (- 0.55 M ) , with oxalic acid (H20x) up
to 9 x lo-' M , and at temperatures in the range 0.02 "C to 20.57 "C,the rate data
were found by Horne' to be consistent with an expression

rate

R , + k3[FeOx+][Fe2+]

+ k4[Fe(Ox);][Fe2+]

where R, represents the exchange rate at zero oxalate ion concentration. Values
of k , and k4 were calculated as 2140 and 4250 l.mole-'.sec-', respectively, at
20 "C. For the step associated with k , the entropy and energy of activation were
estimated as - 14 cal.deg-'.mole-' and 9.2 kcal.mole-'. Sheppard and Brown2
have reported values of k , (20 "C) of 3300 1.mole-'.set-', and of the entropy of
activation of 28 cal.deg-'.mole-' and the energy of activation of 21 kcal.mole-'
for the overall process associated with k , K , where K3 refers to the equilibrium
Fe3++Ox'-

+ FeOx'

McAuley and Brubaker, have found that for added tartaric acid (H2Tar) up
to 5 x
M , at temperatures in the range 0 to 10 "Cwith constant acidity (0.11
M ) and ionic strength 0.55 M , rate data were consistent with an expression
rate = R,+ k,[FeHTar2+][Fe2+]+k,[Fe(HTar):][Fe2+]
The term defined by k , is relatively unimportant, since plots of kobsversus [H2TarI2
were linear. The pathways for exchange in the presence of tartaric3 and oxalic
acids' are
F e O x+ + F e 2 +
Fe(0x); + F e2 +
FeHTar2++ F e 2 + %
Fe(HTar)l + F e2 + -%References pp. 142-152

I 04

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

in addition to those normally operative in aqueous media. The values of k 5 and


k, could not be calculated owing to lack of data on the formation constants of
Fe(HTar): and FeTar".
Chakrabarty et aL4 have measured the exchange in the presence of 8 - quinolinol
(HQ), using a similar method. The rate law at high acidities (pH < 2) and low
concentrations of Fe3+ was found to be
rate

R,+k,[FeQ2+][Fe2+]

where k, is the rate coefficient for the pathway


k'

F e Q 2 + + F e 2 +--+
At 25 "C and p = 0.5 M the value of k , is 450 l.mole-'.sec-'.
100 % exchange of iron, between the ethylenediaminetetraacetate complexes of
iron(II1) and iron(1I) ( F e y 2 - and Fey-), has been observed within the time
required (15 sec) for either, precipitation, or ion exchange separation of the species
of iron5. Reynolds et d 6have since investigated the exchange of Fe2+ with F e y - ,
which they have found to be measurable and to follow a rate law
rate

k,,,[H

'3

1.7[Fe2+]o.7[FeY -3

These workers have concluded that, in the exchange of F e y 2 - and F e y - , a


dissociation step cannot take part in the mechanism of exchange.
Attempts to measure the exchange between 1,lO-phenanthroline and substituted
1,lO-phenanthroline complexes of iron (11) and iron (111) have led to a lower
limit being placed on the rate coefficients. Eimer and Medalia,, using the isotopic
method (55Fe)and either perchlorate precipitation or extraction using a chloroform solution of camphorsulphonic acid of the iron(I1) complex, observed 100
exchange in 15 sec with the 5,6-dimethylphenanthroline complexes i n 1 M
sulphate media at 0 "C. Eichler and Wahl', using similar separation methods,
but with a direct injection technique, observed complete exchange, between the
phenanthroline complexes at 0 "C, in 3.0 M sulphuric acid, in less than 0.07 sec.
This was calculated to correspond to a rate coefficient in excess of lo5 ].mole-I.
sec-'.
Using the optical activity method, these two workers placed a limit
> 160 I.mole-'.sec-' on the same rate coefficient.
In heavy water 3 A4 with respect to D,S04, a lower limit of k > lo5 Imolz-'.
sec-' at 26 "Chas been proposed by Dietrich and Wah19 and Larsen and Wahl",
as a result of an NMR investigation of the position of the proton absorption bands
of mixtures of the iron (111) complexes and iron (11) complexes. This limit applies to the exchange between iron (111) and iron (11) complexes of 1.lOphe-

'

105

T R A N S I T I O N METALS

nanthroline', 4,7-dimethyl-, 3,4,7,8-tetramethyl-, 3,5,6,8-tetramethyl- and 5,6dimethyl-substituted phenanthroline complexes' '.


For the phenanthroline and 4,7- and 5,6-dimethylphenantholine complexes
a further estimate of k > 3 x lo7 I.mole-'.sec-' has been proposed, on the basis
of NMR line broadening measurements, by Larsen and Wahl".
The isotopic method has been used in conjunction with a flow apparatus by
Stranks' 2, t o measure the exchange between the cyclopentadienyl complexes
of iron (111) and iron (11) in methanol. Separation was based on the insolubility
M
of Fe(C5H5)+in petroleum ether at -80 "C. using Fe(I1) and Fe(II1)
and short reaction times (- msec), a rate coefficient 8.7 x lo5 I.mole-'.sec-' at
-75 "C was obtained. The rate of exchange in the presence of chloride ions and
inert electrolytes was found to be more rapid. Calculations using Marcus Theory
showed reasonable agreement with the experimental observations. In deuterated
acetone, line broadening measurements have led to an estimate9 of this rate
coefficient of > lo5 I.mole-'.sec-' at 26 "C.

8.4.4 The exchange reaction between Fe(ZZZ) and Fe(ZZ) in non-aqueous and mixed
solvents

An isotopic investigation (59Fe) of the exchange of iron (111) and iron (11) as
perchlorates in absolute ethanol, with chromatographic separation on an alumina
column, led to complete exchange being observed within the separation time'.
However, Horne', using the 2,2'-dipyridyl separation method and the isotope
55
Fe has observed a decrease in the exchange rate as water is replaced by acetone,
methanol, ethanol and propanol. The rate tending to zero in the absence of water.
In aqueous ethanol the rate law was found to be
rate = ko,,[Fe(III)][Fe(Il)]
with activation parameters at low water concentrations of about 10 kcal.mole-'
( A H ' ) and -20 cal.deg-'.mole-' (AS'). Increase in the ionic strength led to an
increase in the rate, whereas increase in HC104 concentration brought about a
reduction. The rate of exchange has also been reported to be much slower in
isopropano13 and nitromethane, than in aqeous media. Various reasons for the
slow rate in alcohol media have been put forward2.'.
Menashi et a f . 5have made a detailed isotopic study (59Fe) of the exchange in
dimethylsulphoxide, in the absence of oxygen, using the 2,2'-dipyridyl separation.
to
to 2.8 x lo-, M and 1 x
With Fe(I1) and Fe(II1) in the ranges 1 x
1 x lo-, M , respectively, the rate data was found to conform to a first order
dependence on both oxidation states of iron. The value of the observed rate
coefficient (20 "C), in the presence of 2.5 x
M H,O and 2 x lo-' M HC104,
at p = 0.2 M (NaCIO,), was 18 I.mole-'.sec-'.
Reference5 pp. 142-152

106

I O N S O F SAME METAL IN D I F F E R E N T O X I D A T I O N STATES

The effect of variation in the perchloric acid concentration was found to be


complex. At concentrations lower than 7 x
M , there was a rapid decrease
in the observed rate coefficient as the perchloric acid was increased; above 7 x
M no change occurred in the rate as the [HC104] was altered. Perchlorate ion
concentration had no effect on the rate of exchange. The effect of the addition
of water was found to be dependent on the perchloric acid concentration: either
an increase (HC104 = 2 x
M ) , or a decrease followed by an increase
(HC104 = 2 x lo-' M ) being observed. Ionic strength effects have also been
studied. The overall activation enthalpy and entropy values obtained5 were
9.6 kcal.mole-' and -20 cal.deg-'.mole-', respectively.
It is thought that exchange can occur through the species Fe(DMS0):' and
Fe(DMS0);' possibly via an inner sphere mechanism; the exchange occurs in the
absence of water.
Wada and Reynolds6 have investigated thc effect of chloride ion, over the
range 1 x
to 6 x
M , on this exchange system. A rate law at constant
HC104

was found, which was interpreted i n terms of the reaction mechanism

+ Fe3+ -%Fe3' + F e z +
FeCl' + Fe3 %Fe3 + FeCI'
KI
Fe''+Cl+
FeCl'
(K, = 1 . 4 ~
Fez'

I.mole-')

On this basis, a value of the rate coefficient k , was estimated as 3.2 x lo2 1.mole-'.
sec-' (at 20 "C and p = 0.1 M ) . The energy and entropy of activation corresponding to the term k,K, were calculated as 8.7 kcal.mole-' and -0.5 cal.
deg-'.mole- I , respectively.

8.4.5 The exchange reaction between hexacyanoferrate (ZZZ) and hexacyanoferrate


(11)

The earliest attempts to measure the rate of this exchange reaction met with
little success, 100 % exchange being found to occur during the time necessary for
mixing and separating the two species of
The radioactive indicator used
was usually a mixture of isotopes (59Fe and 55Fe), and the media ranged from
molar aqueous hydrochloric acid to 0.05 M sodium hydroxide. The separation
methods tried included (a) precipitation of the ion Fe(CN);- as the insoluble
salts KCeFe(CN)i, Pb,Fe(CN)i and KFez(CN): and ( b ) physical methods based
on diffusion, ion exchange and electrophoresis4.

T R A N S I T I O N METALS

107

Wahl and Deck' were able to obtain an estimate of an assumed second-order


rate coefficient (- lo3 I.mole-'.sec-'
at 4 "C) using a separation procedure
based on the extraction of Fe(CN);- by a chloroform solution of Ph4AsC1, in
the presence of the ions Co(CN):- and Ru(CN):-, to reduce the exchange
between the iron species in the two liquid phases. A similar estimate was obtained
using a precipitation method in the presence of the carrier Ru(CN):-. A direct
injection technique was used as short reaction times were necessary. Wahl' has
reviewed the large induced exchanges occurring in the chemical separation methods.
The extraction procedure when the carriers Co(CN): - and Ru(CN):- are present
provides the most satisfactory method of separation'.
From the variation in the line width of the single I4N NMR line, observed with
aqueous mixtures of K,Fe(CN), and K,Fe(CN)' of about 0.5 M with respect
to each species of iron, a rate coefficient (9.2 x lo4 l.mole-'.sec-' at 32 "C) has
been calculated7~'. An associated entropy and energy of activation of -24.3
cal.deg-'.mole-' and 4.2 kcal.mole-', respectively, has also been found'. At
temperatures above 40 "C the activation energy may be larger than that mentioned
above.+ Results confirming the positive catalyticeffects of various inorganic cations
(H', Li', Na', K', Rb', Cs', NHf, Mg2+,Ca2+ and Sr2+), reported earlier', wert. also obtained and the possibility of a cation bridge mechanism was
discussed.
Wahl et U I . ~ ,have
' ~ completed the first detailed isotopic study of this exchange
reaction and have shown paths catalysed by cations (H', K', Ca2+, Ba2',
Ph4As+ and various tetraalkylammonium ions' ') occur. The first order with
respect to both Fe(CN):- and Fe(CN)z- was confirmed", viz.

'

rate

ko,,[Fe(CN)~-][Fe(CN):-]

The dependence of the exchange rate on the [cation] is complicated as a result of


complexing, uiz.

+ MFe(CN);M + + Fe( CN); - + MFe( CN$M + +Fe(CN)i-

and the exchange paths involving all species of iron which occur, viz.

+ Fe(CN)i- --+
MFe(CN)i- + Fe(CN):- %
Fe(CN):- + MFe(CN)i- %MFe(CN)i- + MFe(CN)i- --+
Fe(CN);-

ko

k2

See, however, Loewenstein and RonL2.


ReJcrencr> pp. 142-1S2

I08

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

For aqueous potassium hydroxide


M ) the value of the observed rate
coefficient kobsis 226 I.mole-'.sec-' at 0.1 "C. Over the temperature range 0.1
to 20.6 "C an activation energy of 6.0 kcal.mole-' was calculated. At zero cation
concentration k , (0.1 "C), obtained by extrapolation, has a value of 6.0 ].mole-I.
sec-' with an associated activation energy and entropy of 9.1 kcal.mcle-' and
- 24 cal.deg - '.mole- ', respectively.
Wahl et al." have suggested that the rate coefficient obtained by the N M R
method7,*at cation concentrations
5 x lo-' M is that for the step k , in the
above mechanism. There is reasonable agreement between the data obtained by
these two different procedures. A comparison has been made of the experimental
results with those obtained from the Marcus theory' O.
For the exchange reactions between Fe(CN),N:and Fe(CN),N:-,
and Fe(CN),P(C,H,): -,
Fe(CN),NHg - and Fe(CN),NH:-, Fe(CN),P(C,H,):and Fe(dipy)(CN), and Fe(dipy)(CN):-, Stasiw and Wilkins'' have recently
evaluated rate coefficients of 3 x lo3, I x lo', 1 x 10' and 4 x l o 7 I.mole-'.sec-',
respectively, at 25 "C.

8.4.6 Reactions of Fe(I1I) with Fe(II)


A stopped flow technique coupled with spectrophotometric analysis of the iron
(11) complex formed has been used to investigate',' the reactions of some organic
complexes of iron(II1) with the ion Fe2+.The iron(II1) was complexed with 1,lO-

phenanthroline', various substituted ],lo-phenanthrolines (5-methyl-, h i t r o - ,


5-chloro-, 5-phenyl-, 5,6-dimethyl-, 4,7-diphenyl-, and 3,4,7,8-tetramethyl-) and
2,2'-dipyridine, 4,4'-dimethyL2,2'-dipyridine, and 2,2',2"-tripyridine,. The wavelengths used for the analysis lay in the region 500-552 mp.
For the reaction Fe(phen):+-Fe2+ in 0.5 M HClO, a rate coefficient of
3.7 x lo4 l.mole-'.sec-' at 25 "C was calculated by Sutin and Gordon'. The
entropy and energy of activation values obtained were -37 cal.deg-'.mole-' and
0.8 kcal.mole-'. Little acid dependence was observed. The rate coefficients (].mole.
sec-' at 25 "C in 0.5 M HClO,) for the other systems' are : Fe(S-Me-phen):+F e 2 + , 2 x lo4; Fe(5-N02-phen)~+-FeZ', 1.1 x lo6; Fe(5-C1-phen):+-Fe2+,
7.8~
lo3. The rate coefficients for
2.1 x lo'; and Fe(5,6-Me2-phen):+-Fe'+,
these reactions are between 7 and 9 times larger in 0.5 M H2S04. For the systems
Fe(5-ph-phen): +-Fez +,Fe(4,7-phz-phen)i +-Fe2+ and Fe(3,4,7,8-Me4-phen):
Fe2+,the rate coefficients with 0.5 M H zS 0 4 at 25 "C are 3.2 x lo', 3.3 x lo4, and
1.9 x lo3 l.mole-'.sec-', respectively. Ford-Smith and Sutin2 suggest that electron
transfer occurs when the ferrous ion is among the phenanthroline groupings of
the iron(II1) complex.
For the iron(II1) complexes (a) Fe(dipy): ( b ) Fe(4,4'-Me2-dipy):+ and ( c )
Fe(tripy):+ the rate coefficients (l.mole-'.sec-') with perchloric and sulphuric
+-

TRANSITION METALS

109

acids (0.5 M ) are' ( a ) 2 . 7 lo4


~ and 2 . 2 ~
lo5 (6) 6.0 x lo2 and 5.9 x lo3 and
(c) 8.5 x lo4 and 7.4 x 10'. Estimates of second order rate coefficients in excess
of lo8 I.mole-'.sec-', have been made by Gordon et aL3 for the reactions between
Fe( phen): + and Fe(CN);-, Fe(4,4'-Me2-phen): and Fe(3,4,7,8-Me4-phen): +,
and Fe(phen):+ and Fe(4,4'-Me2-dipy):+ at 25 "C in 0.5 M sulphuric or perchloric
acids.
The same technique has been used to measure the rate coefficients for the reactions of ferricyanide ions with ferr~hemoglobin~
and ferrocytochrome c (ref. 5).
At 25 "C, p = 0.1 M and pH 6, the values are 7 x lo4 and 1.6 x lo7 I.mole-'.
sec- respectively.
Stasiw and Wilkins6 have made an investigation of the reactions between
Fe(CN);- and Fe(CN),NH:-,
Fe(CN),H202-, Fe(CN),P(C,H,):and
Fe(dipy)(CN), . The values of the rate coefficients (1.mole- '.sec-') with the
activation enthalpies (kcal.mole- ') and entropies (cal.deg-'.mole- ') and ionic
strength in parentheses, reported by these workers are: 7 x lo3 (3.3, -32;
p = 0.05 M ) , 1 x lo3 (3.2, -33; p = 0.02 M ) , 8 x lo4 (3.3, -25; p = 0.05 M ) ,
and 8 x lo6, respectively. For the reactions between Fe(CN);- and Fe(CN),N:-,
and Fe(dipy)(CN)i- the corresponding values are 8 x lo4 (1.8, -30; p = 0.05 M )
and 2 x lo4, respectively.
The kinetic parameters for the reaction between the Fe(lI1) complex with EDTA
and the Fe(I1) complex with rruns-l,2-diaminocyclohexanetetraacetate have
recently been reported, by Wilkins and Yelin', as 3 x lo4 I.mole-'.sec-', 4.0
kcal.mole-' and -25 cal.deg-'.mole-'.
+

8.4.7 The Fe(II)-catalysed aquation of'Fe(ll1)

Sutin et a / . ' - 3have found iron(l1) to catalyse the aquation of the iron(II1)
ions, FeCI2+ (refs. 1,2) and FeNCS2+ (ref. 3), in perchlorate media. A flow
technique, with spectrophotometric detection at 336 mp (disappearance of FeCl' )
and 460 mp (disappearance of FeNCS2+), was used to obtain rate data. The rate
law
+

'-d [FeX2+]
--____~- kob,([FeX2'1 - [FeX2'],,)
dr
where X is either the chloride or thiocyanate ions, was used to treat the data
obtained. The observed rate coefficient kobsis related to the reactions
FeX2+ + Fe3++XFeX'

+ Fe2++X-

2 F e X + + F e3 +

F e X 2 + + F e2 +

k-

References pp. 142-152

110

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

and in the case of the chlorocomplex, following the study of the effect of hydrogen
ion', the additional reactions,

FeCI2+ H 2 0 $ ClFeOH'
ClFeOH'
ClFeOH++Fe'+

+H +

(K2)

+ FeOH++Cl-

2 FeCl++FeOH2'

k-2

At a constant ionic strength of 3.0 M , Mg(ClO,), , and a temperature of 25 "C,


k, (NCS-) has a value of 10.5 l.mole-'.sec-' (at constant i H + ] = 1.80 M ) and
k, (C1-) of 6.2 l.mole-'.sec-'. The value of the composite term kzK2 (Cl-) is
14.8 sec-' under the same conditions. The step associated with k2 is thought
likely to involve a transition state with a hydroxide bridge.

8.4.8 The reaction between Fe(IV) and Fe(ZZ)


A reaction between iron (IV) and iron (11)

Fe(IV)+Fe2+ + [Fe(III)J,
has been postulated to occur in the oxidation of iron(I1) by various two-equivalent
oxidants'.
8.4.9 The exchange reaction between Ru(VZZ) and Ru(V1)
The exchange between the ruthenium anions RuO, and RuO:- in aqueous
hydroxide media has been found rapid. A limit for the rate coefficient at 0 "C
of > 1.7 x lo3 I.mole-'.sec-' has been proposed by Luoma and Brubaker'.
The isotopic method (' 06Ru), and separation procedures based on the preci pitation of the RuOi- or RuO, species with barium or tetraphenylarsonium ions,
respectively, were used. Attempts to use an ESR technique failed.
8.4.10 Ru(IZ)-catalysed substitution reactions of Ru(IIZ)
Ru(1I) in the form of Ru(NH3)gC catalyses the substitution reactions'
Ru(NH3),0H3" +C1Ru(NH3),C1'+ +H,O

k'

__f

k'i

111

T R A N S I T I O N METALS

At an ionic strength 0.1 M the values of the catalytic rate coefficients k; and k;
(25 "C) were estimated from spectrophotometric measurements to be 4 x lo3
1'.rnole-'.sec-' and 2 x 10' I.mole-'.sec-', respectively.
8.4.1 I The exchange reaction between Os(IZZ) and Os(IZ)

The exchange between I-Os(dipy);+ and d-Os(dipy):+ has been studied by the
optical activity method by Dwyer and Garfas', and Eichler and Wahl'. The former
authors reporting complete exchange in 95 sec at 5 "C for reactants 5 x
M,
while the latter authors report complete exchange in a time < 15 sec at 4 "C under
comparable conditions.
Eichler and Wah12 have attempted an isotopic study (I9lOs and ' 8 5 0 s ) of the
exchange reaction between Os(dipy): + and Os(dipy): + using a direct injection
technique so that reaction times
7x
sec were possible. With total osmium
M in aqueous sulphate media at 0 "C complete exchange was observed.
The separation methods used were, ( a ) perchlorate precipitation (in presence of
iron(I1) carrier) and ( b ) extraction with p-toluenesulphonic acid in nitromethane,
of the osmium(I1) complex. A lower limit of 1 x lo5 1.mole-'.sec-' was placed
on the rate coefficient (0 "C, 3.0 M H'SO,). Dietrich and Wah13 using the line
broadening effect produced by Os(dipy): + on the NMR spectrum of Os(dipy):
have been able to propose a value of > 5 x lo4 I.mole-'.sec-' at 6 "C in D,O
(0.14M [Cl-] and 5 x
M [D']).
Campion et aL4 have calculated a value for the exchange coefficient (10 "C,
zero ionic strength) of 1 x lo7 I.mole-'.sec-' from the observed rate coefficients
for the reactions

Os(dipy):+

+ Mo(CN);- + Os(dipy):+ + Mo(CN);-

obtained by a temperature jump method, and the application of the Marcus theory.

8.5

COBALT, RUTHENIUM A N D IRIDIUM

8.5.1 The exchange reaction between Co(ZIZ) and Co(ZZ) in aqueous media; the

efect of inorganic anions


The earliest investigation of the exchange reaction between the aquated ions of
Co(II1) and Co(I1) was carried out by Hoshowsky et al.', using the isotopic
method (60Co). When sulphate salts (M ) were employed, complete
exchange was observed between the two oxidation states of cobalt, in a time of
less than two min. Two separation methods were employed: ( a ) adsorption on an
alumina column, and ( b ) precipitation of the Co(II1) as the cobaltinitrite.
References p p . 142-152

112

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

The first rate measurements on this system were made by Bonner and Hunt'
who, in a preliminary study, were able to show the exchange in perchlorate media
was overall second-order. The observed rate coefficient (0 "C, 1 M in HCIO4)
was 0.77 I.mole-'.sec-'. The separation method involved complexing with EDTA
at pH 8 before extracting the Co(II), after addition of thiocyanate ions, with
methyl isobutyl ketone. In a further article3 these two authors report in detail
their studies on this system. Overall activation energy and entropy values of 13.2
kcal.mole-' and - 13 cal.deg-'.mole-' are given. From results obtained at
varying acidity and constant ionic strength, Bonner and Hunt3 concluded that
the rate law could be of the form
rate

( k , +k;[H+]-l)[Co(III)][Co(II)]

where k , and k ; have values of 0.83 I.mole-'.sec-' and 0.14 sec-', respectively,
at 3.2 "C and ,LL = 1.0 M . The rate of exchange was found to be influenced by the
addition of nitrate or sulphate ions and ionic strength but not by the presence of
chloride ions, oxygen, light or platinum surfaces. In D 2 0 the rate was found to be
much slower than in H20.
Shankar and De Souza4. ', in a similar study using the isotopic method (60Co),
employed mainly ethylenediamine and ammonia to stop the reaction and a
separation based on the extraction of Co(I1) (as the thiocyanate) with amyl
alcohol-ether mixtures. These two workers were able to prove that the exchange
was accurately first order in both Co(I1) and Co(II1) over a wide range of conditions. Results in general agreement with those mentioned previously3 were
obtained for the effect of ionic strength, heavy water and hydrogen ion concentration on the exchange reaction. The mechanism suggested involved the two exchange pathways

c o 3 ++co'+

ki
-+

CoOH2++ Coz+ -+
k2

with the CoOH2+ species being produced via the equilibrium

Co3+ H,O

$ CoOH2'

+ H+

K2

From a detailed study of the exchange, at various temperatures (in the range
0 to 20 "C) and acidities at a constant ionic strength of p = 1.0 M , the kinetic
parameters were calculated'. k , and k ; ( k ; = k 2 K z ) have values of 0.48 I.mole-'.
sec-' and 0.22 sec-', respectively, at 0 "C. For the exchange pathway associated
with k, , values of the activation enthalpy and entropy of 12.6 kcal.mole-' and
- 14 cal.deg-'.mole- ', respectively, were reported. For the second pathway

113

TRANSITION METALS

( k z ) the authors estimated values of 11.6 kcal.mole-' and -7 cal.deg-'.mole-',


using thermodynamic parameters foundbySutcliffe and Weber6 for the equilibrium.
Comparisons have been made of the activation parameters obtained for this
exchange system with those obtained for similar systems5.
Habib and Hunt7 have continued the study of this reaction, obtaining further
data with special reference to the effects of ionic strength, sulphate and hydrogenion concentrations. From data obtained on the dependence of the rate on the
[H+] at various temperatures, values of the kinetic parameters differing slightly
from those above' have been obtained. Values of A H : and A H ; and AS: and
A S i (at p = 1.0 M ) obtained were 11.8, 5.3 kcal.mole-' and - 17 and -31
cal.deg-'.mole-',
respectively. The value of kz was estimated as
6.7 x lo3 1.
mole-'.sec-' at 18 "C, p = 1.0 M .
The effect of the addition of sulphate and fluoride ions were found by these
workers' to increase the rate of exchange; addition of acetate and trifluoroacetate
ions produced relatively minor changes. For the addition of sulphate ions, a rate
law

rate

( k , + k ; [ H + ] - ' +k;[SO~-])[Co(III)][Co(II)]

was found to fit the experimental observations, suggesting a pathway


coso:

+c o z

k3
+

---f

with k j being given by k3K3,where K3 is the constant for the equilibrium

co3 + so:- + C O S O ~
+

At p = 0.5 M the value of k j is 5.2 x 10' 12.mole-2.sec-' (0.15 "C); the corresponding activation enthalpy and entropy values are 20.5 kcal.mole- and 28.5
cal .deg - '.mole - I.
Shankar and De Souza' have also recently investigated the effect of the additions of various anions to this system in both water and heavy water solvent.
Fluoride was found to have very little influence on the exchange rate while acetate,
nitrate and sulphate ions produced an increase. For the addition of sulphate ions
an estimate of the rate coefficient k , of 20 I.mole-'.sec-' (at 14 "C and p = 2.0
M ) was made. For the addition of nitrate and acetate, values of the coefficients
kk and k; (where kk = k4K4 and k; = k 5 K 5 ) ,viz.

CoNO:++Co2+

kq

ks

CoOAc" +Co2+ -+

+ CONO:+
Co3++OAc- + CoOAc"
c o 3 ++NO;

References pp. 142-152

K4
K5

114

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

at 0 "C are 2.83 x lo4 (p = 1.0 M ) and 0.762 1.2mole-2.sec-' (p = 2.0 M ) ,


respectively.
Shankar and De Souza' also found that the rate of exchange was increased by
the addition of silver ions owing to the occurrence of equilibria such as
Co3++Ag+

+ Co2++Ag2+

Conocchioli et aL9, from spectrophotometric observations on the equilibrium

co3+
+CI-

+COCI~+

have been able to evaluate the rate coefficients for the analogous reactions
coc12++co2+

c03++ C ~ C I +

ki

They found values of 1.0 I.mole-l.sec-' for k, and 26 12.mole-2.sec-' for k , K ,


(at 25 "C,p = 3.0 M ) where K , is the constant for the equilibrium
co2

+ c1 + COCl+

The exchange of 6oCo between the complex anions, 12-tungstocobaltate(III)


( C O O ~ W , ~ O , , ~ and
- ) 12-tungstocobaltate(II) ( C O O ~ W , ~ O ~in~ ~aqueous
-),
chloride media has been studied by Rasmussen and Brubaker". Tetrabutylammonium iodide in acetate media was used to give a separation by precipitation
of the Co(II1). The rate showed a first order dependence on Co(I1) and Co(11I)
concentrations but was independent of the H+ concentration. At p = 1.02 M
(LiCI) the rate coefficient has a value of 1.01 I.mole-'.sec-' at 0 "C. Variation
in the ionic strength and dielectric constant was found to affect the kinetic
parameters. At zero ionic strength, the rate coefficient has a value of 4.5 x lo-,
I.mole-'.sec-' (0 "C). The activation energy of the reaction was reported as
18 kcal.mole-'.
The rate coefficient for the exchange reaction
Co(CN);-

+ Co(CN):-

I.mole-'.sec-'
has been evaluated as < 1 x
using data obtained by Adamson".

at 25 "C by Birk and Halpern"

8.5.2 Exchange reactions involving complexes of Co(ZZZ) and Co(ZZ) with ammonia and organic ligands
The early work' - 4 on exchange reactions involving the complex hexammino
species Co(NH,):+ and Co(NH,);+ led to the conclusion that the 6oCoexchanges

115

T R A N S I T I O N METALS

between Co(NH,);+ and Co(NH,):+, and Co(NH,);+ and Co2+ were extremely
slow. The reactions were studied in the absence of oxygen, which catalyses the
exchange, and in media varying between acid and aqueous ammonia. The separation method used was the extraction of Co(II), as the thiocyanate complex, with
a mixed organic solvent.
Subsequently, Stranks' has investigated a series of cobaltammines with regard
to the electron exchange between cobalt(II1) and cobalt(II), using the isotopic
method 6oCo. The systems studied included the Co(1II) complexes and ion-pairs
Co(NH,):+, Co(NH,):+ OH-, Co(NH,);+ * C1-, Co(NH,),0H2+, trans-Co
and cis-Co(NH,),(OH): with Co(1I) complexes of the type
(NH,),(OH):
Co(NH3); where n has values 3, 4, 5 and 6.t All exchanges refer to perchlorate
13.5 kcal.mole-' and
media in the absence of oxygen. Energies of activation
activation entropies between -28 and -37 cal.deg-'.molehave been found
for these exchanges. The rate coefficients (I.mole-'.set-' at 64.5 "C, U
, = 1.0
M) evaluated from necessary kinetic and ion-pair association data are for the
exchange of Co(NH3):+ with Co(NH,):+.OH- 5.5 x lo-,, with Co(NH,);+.CI7.3 x lo-,, with Co(NH,),0H2+ 9.0 x lo-,, with truns-Co(NH,),(OH):
4.2 x loW3and with cis-Co(NH,),(OH)i 2.5 x
Under the same conditions
is < 1.7 x lo-''
the rate coefficient for the exchange Co(NH,);+-Co(NH,):+
I.mole-'.sec-'. The mechanism proposed by Stranks was

Co(NH3):++H2O

'

+ Co(NH3)2+.OH-+H'

Co(NH,);+.OH-+Co(NH,):+

-+ Co(NH,):++Co(NH,);+

+OH-

involving a hydrogen or hydroxo bridge activated complex. A similar transition


state was thought likely to occur for the Co(NH,):+CI--Co(NH,):+
exchange.
Appleman et uL6 have investigated the exchange of 6oCo between the species
Co(NH,),OH2+ and Co(II), Co(NH,);+ and CO(NH,)(,-~)OH+where n has
values between 0 and 6, in aqueous ammonia. All kinetic data was obtained using
a separation procedure based on the precipitation of the salt C O ( N H , ) ~ H ~ O H ~ C I ~ .
Light and oxygen were excluded from the reaction vessels. A rate law of the form
rate

k [ Co(NH,), OHZ'1 [Co2 '3 [NH,]"

kobs[Co(NH3),0H2+][Co(NH,)f+]

was found from the data obtained over the range of [NH,] 0.092 to 0.167 M,
[Co(II)] in the range 1.5 x
to 9 x l o w 3M with [Co(III)] 4 x lo-, M. The
value of x was found to vary with the ammonia concentration. For the exchange
occurring under conditions when x = 5 the value of kobs(35.5 "C, p = 2.28 M )
is 4.4 x lo-, I.mole-'.sec-' with corresponding activation parameters of 15.6 kcal.

See also Biradar et al.'s.

References pp. 142-152

116

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

mole-' (AHt) and -23.5 cal.deg-'.mole-' (ASS).+ No exchange was detected


between Co(NH3),0HZ+and Co 2 +over a long period of time.
Lewis et u I . ~have studied the exchange between Co(en):+ and Co(NH3);+
and between Co(NH3)if and Co(en):+ occurring in the absence of oxygen. The
observed rate coefficients for the exchanges at p = 0.98 Mare < 1.2 x lo-' 1.mole- '.
sec-' (45 "C) and 2.0 x lo-' l.mole-'.sec-' (25 "C), respectively. The exchange
between Co(en):+ and Co(en):+ was also investigated, using the isotopic method,
by these workers, who found catalytic effects were produced by oxygen and various
added solids. The rate law showed a near first-order dependence on Co(en):+
and an order of between 0.58 and 1.0, with respect to Co(en):+. The activation
energy of the process was found to depend on the ionic strength of the medium
and varied from 12.75 to 15.75 kcal.mole-'. At 25 "C, p = 0.98 M a rate coefficient of 5.2 x
1.mole-'.sec-' was calculated.
Dwyer and Sargeson' have used the optical activity method in their study of the
exchange between the tris-(ethylenediamine) complexes of Co(11I) and Co(I1)

The cobalt(I1) complex, which is optically unstable, was formed in reaction


mixtures by the addition of ethylenediamine (excess) to Co2+. The rate law
obtained from the racemization data was

rate = k , [Co(III)] [Co(II)] k"[Co(III)] [en]


indicating that racemization occurs via two paths only one of which (associated
with k , ) is related to the electron transfer process. At 25 "C and p = 0.98 M , k,
has a value 7.7 x lo-' l.mole-'.sec-', the corresponding values of the activation
energy and entropy are 14.1 kcal.mole-l and -32 cal.deg-'.mole-'. Plots of log
k, versus p* were found to be linear. No specific anion effects were observed.
Strank? has also studied the exchange between tris- and bis-ethylenediamine
complexes of Co(1II) and Co(I1). The isotopic method was used. At low ionic
strengths (CIO,) a rate law
rate = k , [Co(III)] [Co(II)]
was indicated by the kinetic data. The same rates of the exchange were found
between Co(en):+ and Co(en):+ when n had values of 1, 2 or 3. Similar results
were obtained for the exchange between Co(en)i+ and Co(en)i +.Hydroxide and
other anions were found to catalyse the exchange reactions. The rate coefficients
(I.mole-'.sec-') at 50 "C and p = 0.2 M for the various exchange systems, were
+ Recently, Williams and HuntI6 have published further data for this exchange system.

T R A N S I T I O N METALS

117

evaluated by Stranks as : Co(en):+-Co(en)i+, 1.4 x


Co(en):'.OH--Co(en):+Cl--Co(en)f', 5 x
Co(en):+.Br--Co(en)i+
Co(en)i+, 9.2 x W4;
3.0 x
Co(en):+.I--Co(en)f+, 2.0 x
Co(en):+.SO~--Co(en)f', 1.3 x
10- Co(en):+-Co(en)f+, 2.7 x
Co(en),0H2+-Co(en)f+, 9.7 x
and
Co(en),(OH):-Co(en)i+,
2.0 x
For all these processes the activation
13.8 kcal.mole-I with activation entropies in the range -27 to
energies are
-32 cal.deg-'.mole-'. These are similar to the values obtained for the analogous
ammonia complexes5. No exchange has been detected between Co(en):+ and
C o 2 +in aqueous solution over a period of 24 hours'.
The exchange of 6oCo between the EDTA complexes of Co(I1) (Coy2-) and
Co(II1) (COY-), in aqueous perchlorate and nitrate media at pH = 2, has been
investigated by Adamson and Vorres". Using a ion-exchange separation method,
a rate law

rate

k;[CoY -][COY

-1

was found to be obeyed over a range of concenirations, C o y 2 - 1 . 6 ~


to
1.2 x lo-' M and C o y 2 - 1.8 x
to 8.6 x
M. At 85 "C, pH = 2, k; has
a value 2.1 x
I.mole-'.sec-'. The activation parameters are 22 kcal.mole-'
(AH*) and - 17 cal.deg-'.mole-' (AS'). Increase in ionic strength and surface
area and exposure to light produce small increases in the rate of exchange. The
mechanism proposed was a direct electron transfer process: steps involving
dissociation of COY- and C o y 2 - were not considered feasible.
Im and Busch" have made a further study of this reaction. They used the optical
activity method, investigating the rate of racemization of d-COY-, in the presence
of Coy2- which is optically unstable and catalyses the process, viz.
d-CoYd-CoY-+CoY2-

+ I-COY- k
+ CoY2-+I-CoY-

k,

The equation -ln(at/g,,) = kobst was used to treat the experimental data
(kobs = k;[CoY2-]+2 k ) and it was found that k; (the apparent rate coefficient)
had a value of 1.5 x
l.mole-'.sec-' at 85 "C (pH = 2.0). Variation of the
solvent from H 2 0 to D,O did not bring about any change in the rate of racemization, whereas increase in the p H of the medium decreased the rate. This was
explained in terms of the formation of the species Co(H-Y)Co(H-Y)-

+ Coy2- +H +

and the two pathways


d-CoY - + CO(H-Y)d-COY - + COY References p p . 142-152

k4
_f

ks
_f

K4

118

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

The rate coefficients are related by the expression

Values of k4 and k, (I.mole-'.sec-') at 100 "C and the corresponding activation


parameters (AH' and AS' in parentheses) are 8.0 x
(24 kcal.mole-', -9 cal.
deg-'.mole-') and 1.4 x
(20 kcal.mole-' and -21 cal.deg-'.mole-').
Im and Busch", using the optical activity method, have also made a study of
the exchange of the propylenediaminetetraacetate complexes of Co(II1) and
Co(1I) (COPY- and COPY'-). This system can be treated by the McKay equation. Optical rotation data was obtained at temperatures between 80 and 100C
and various pH's between 2.0 and 7.0. This was again found to be consistent with
exchange via two pathways (k, and k,)
~ -H +
L-CO(H-~-PY)-+ ~ - c o ( d - P Y ) +
D-cO(

j-PY)-

D-cO( l-PY)-

+ L-CO( H-d-PY)- 3 D-CO(H-I-PY)- + L-cO(d-PY)+ L-CO(d-PY)'-

k.,D-cO(I-PY)2- +L-CO(d-PY)-

and L, and d and I, refer to the rotational properties of the compound and the
ligand, respectively. Values of the rate coefficientsk, and k , at 100 "Care 7.0 x
and 2.0 x
1.mole-'sec-', respectively.
Baker et aLi3, have obtained some kinetic data on the exchange reactions
between the 1,lO-phenanthroline, 2,2'-dipyridine and 2,2', 2"-tripyridine complexes
of Co(11I) and Co(I1) in both H 2 0 and D 2 0 solution. The isotopic method (60Co)
and separations involving either extraction of Co(I1) with organic solvents (nhexanol-ether or saturated sodium acetate in n-hexanol) or precipitation of Co(II1)
as Co(phen),(I,), were used. For the 1,lO-phenanthroline system, a rate law
D

rate

kobs[Co(11I)] [Co( II)y

was observed, where y has a value of


1 in the presence of perchlorate or nitrate
ions and in the presence of chloride ions. The values of the observed rate coefficients at 0 "C (p = 0.1 M ) , which are dependent on the anion concentration,
are 1.1 l.mole-'.sec-'
(ClO;), 4.0 1.mole-'.sec-' (NO;) and 1 . 2 lo-'
~
l*.
mole-*.set- (Cl-). The mechanism postulated invokes the existence of ionpairs of cobalt(II1) and the anion concerned. The rates of exchange in D 2 0 were
little different from those in H,O. Data have recently been published concerning
this exchange system for media containing polymeric electrolytes' '.
For the exchange between Co(dipy):+-Co(dipy)g+
and Co(tripy):+
-Co(tripy)i+, second-order rate coefficients (0 "C) in perchlorate media of
12

'

T R A N S I T I O N METALS

119

',

respectively, were obtained. Previously, Ellis et a l l 4 had


and 1.6 l.mole-'.secestimated values for the exchange rate coefficients (20 "C) of 18.7 and 4.46 1.
mole-'.sec-' for the 1,lo-phenanthroline and dipridine systems.

8.5.3 The exchange reaction between Co(ZZZ) and Co(IZ) in non-aqueous media
Grossman and Garner' have investigated the exchange reaction between the
species Co(NH3);+ and Co(NH,):+, in the form of their nitrate salts, using the
isotope 6 o Co with anhydrous liquid ammonia as solvent. The reaction was carried out in a sealed system at temperatures of 25" and 45 "C, with a separation
procedure based on ether-pentanol extraction of the thiocyanate complex of
3x
M , observed rate coefficients of
Co(1I). With Co(lI1) and Co(11)
6x
(25 "C) and 7 x
(45 "C) 1.mole-'.sec-' were obtained. The activation energy for the exchange process is 23 kcal.mole-'. Grossman and Garner'
have pointed out that the exchange mechanism could involve the dissociation of
the species Co(NH3)zf, since the rate of ammonia exchange with this species is of
the same order of magnitude.
West2 has observed very slow exchange reactions in pyridine solvent between
Co(II) in the form of Co(OAc), and Co(II1) in the forms tris-(1-nitroso-2napht ho1)cobalt(111) and tr is-(acety1acetone)cobalt (111).
Baker et aL3 have obtained some evidence for a decreasing rate of exchange,
as water is replaced by acetone, with the system Co(phen):+-Co(phen)i+.

8.5.4 Co(ZZ)-catalysed substitution reactions of Co(ZZZ)

Numerous cyanide substitution reactions of cobalt(Il1) complexes, Co"'


(NH3),X and CO"'(CN)~Xhave been investigated (mainly by Halpern et a z . ' ~ ~ , ~ )
and found'-4 to be catalysed by Co(I1) in the form of Co(CN);-. Two reactions
of the pentammine complexes can occur, uiz.
Co"'(NH,),X+S CN-

Co"'(CN),X+5 NH3

Co"'(NH3),X+6 CN-

Co(CN)2-+5 NH3 + X

and

the former by an inner-sphere mechanism with the group X acting as a bridge, uiz.
Co(CN):- +Co1"(NH3),X kl_ [(CN),Co"-X-Co1"(NH3),]'
rate

k [Co"'(NH,) 5X][Co( CN):

References p p . 142-1 52

-3

120

IONS OF SAME METAL I N DIFFERENT OXIDATION STATES

and the latter reaction via an outer-sphere mechanism


Co(CN):-+CN-

+ Co(CN):-

Co(CN):- +Co"'(NH,),X
rate

K2

[(CN),Co"(CN)(X)Co'"(NH,),I:

k , K , [Co"'(NH,),X] [Co(CN):

-1[CN -1

The nature of the group X determines the type of reaction which is the most
important. For X = azide, thiocyanate, hydroxide, chloride', bromide and
iodide' the inner-sphere bath operates while for X = ammonia or oxyanions
(including carboxylates)' the main pathway is the outer-sphere reaction. For X =
fluoride or nitrite the concentration of the cyanide ion present determines which
is the major reaction pathway.
Halpern et a l l , from data recorded using a stopped flow technique with
spectrophotometric detection, have obtained values of k , and k , K2 for each of the
above-mentioned anions.
For the Co(I1I) complex Co(NH,),NO;+, Halpern and Nakamura3 have
obtained spectrophotometric evidence for the inner-sphere reaction occurring via
C O ( C N ) ~ O N O ~which
isomerises to give the product Co(CN),NOl-. The
species Co(NH3),CN2+ also reacts in this manner to give Co(CN)'NC3- and
finally Co(CN):-.
Birk and Halpern4 have studied the Co(II), Co(CN):-, catalysed substitution
reactions

which occur via an inner-sphere mechanism and an activated complex [X(NC),


CO'~'-CN-CO"(CN)~]*to give Co(CN),NC3- and finally Co(CN)i-. Absorbance measurements in the region of I,,, of the species Co"'(CN),X were
used to obtain rate data. The rate law observed was
rate = k,[Co"'(CN),X][Co(CN)~-]
where the rate coefficients k, decrease in the order X = H 2 0 , Br-, C1-, SCN-,
I-, N;, OH- and SO:-. Values of k3 for each of these ions have been determined.
Limits for the rate coefficients of the outer-sphere reactions, involving the ion
Co(CN);-, have also been set by these workers. For the chloride and iodide complexes an additional term, k4 [CO"'(CN)~X],was required in the rate law to
account for a non-catalysed pathway.
Farina and Wilkins' have published data relating to the reactions: Co(tripy)2
with Co(dipy): +,Co(dipy): ,C ~ ( d i p y ),~Co(phen):+, Co(phen)z +,Co(phen)' ,
+

121

T R A N S I T I O N METALS

Co(3,5,6,8-Me4-phen):+, Co3+ and CoOH2+; Co(py),CI:


with Co(tripy)Z +,
Co(dipy):+ and Co(phen):+; Co(phen)2+ with C ~ ( d i p y ) ~ and
+ , Co(dipy):
with Co(tripy):+. Values of the rate coefficients (at 0 "C, I.mole-'.sec-') with
the corresponding activation enthalpies (kcal.mole-I) and entropies (cal.deg mole-') and ionic strengths in parentheses are: 6.4 x 10 (8.4, - 19; p = 0.05 M ) ,
1.3 x lo4 (4.2, -24; p = 0.05 M ) , 6.8 x 10' (10.8, -6; p = 0.05 M ) , 2.8 x 10'
(6.9, -21; p = 0.05 M ) , 3 x lo4 (4.3, -22; p = 0.05 M ) , 1.4 x lo3 (10.9, -4;
p = 0.05 M ) , 6.8 x 10 (10.6, -12; p = 0.02 M ) , 7 . 4 ~
lo4 (3.4, -23; p = 1.0
M ) , 1.7 x lo4 (5.1, -20; p = 0.027 M ) , 3 . 0 ~
lo5 (5.2, - 14; p = 0.00016 M ) ,
1.1 x lo4 (5.2, -20; p = 0.00016 M ) , 9.1 x lo3 (8.4, - 12; p = 0.00016 M ) ,
6 x lo-' (8.7, -30; p = 1.55 M ) , and 2.7 x 10 (8.7, -20; p = 0.05 M ) , respectively.
+

'.

8.5.5 The reaction of Co(IV) with Co(II)

To account for their experimental observations on the reaction of lead(1V)


acetate with cobalt(II) acetate in anhydrous acetic acid solution, Sutcliffe et al.'
have proposed a mechanism involving the Co(1V) steps, viz.
Co(IV)+Co(II)

3 2 Co(II1)

Co(IV)+Pb(II)

$ Co(II)+Pb(IV)

Whilst no direct measurements have been made on these Co(IV) reactions, the
ratio (k2/2 k , ) has been evaluated' as 0.15 at 25 "C.

8.5.6 Rh(I)-catalysed substitution reactions of Rh(ZZI)

Rund et al.' have suggested that the pyridine substitution reaction


RhC1:- + 4 py = trans-Rh py4 Cl';

+ 3 C1-

proceeds via a pathway involving Rh(1). uiz.


Rh(I)+4 py +Rh pyd
Rh py,' +RhCl:(Rh py4-C1-RhC1,)trans-Rh py4CI2++C1-

+(Rh

py4-C1-RhC1,)-

+trans-Rh
_f

slow

py4C12++Rh(I)

trans-Rh py4C1:

Other
have proposed steps involving catalysis by species other than
Rh(1) for some similar reactions.
References p p . 142-IS2

122

IONS O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

8.5.7 The exchange reaction between I r ( I V ) and Ir(III)

The isotopic method ( I 921r)has been used to investigate the exchanges between
the anions IrCli- and IrClz-. The separation methods used by Sloth and Garner'
were: ( a ) extraction of IrCli- with either 2-butanone or acetone-chloroform,
and (6) precipitation of the iridium (IV) as the ammonium salt. These authors
observed complete exchange in less than 42 sec, with reactant concentrations
M , at 1 "C in the absence of light with 1 M HCl as solvent. This corresponds to a rate coefficient in excess of 290 I.mole-'.sec-'.
Recently, Hurwitz and Kustin' have reinvestigated this exchange reaction using
the same isotopic procedure and the 2-butanone separation method, in conjunction
with a stopped flow apparatus. A rate coefficient of 2.3 x lo5 I.mole-l.sec-' was
obtained for the conditions, temperature 25 "C and ionic strength 0.1 M. Application of the Marcus theory to results obtained for the reaction

IrCli-

+ Fe(5,6-Mez-phen); + IrCli- + Fe(5, 6-Mez-phen);'


+

by the temperature jump method3 had led previously tc an estimate for this rate
coefficient of 2.5 x lo4 I.mole-'.sec-'.
8.5.8 The reaction between complexes of Zr(ZV) and Zr(IIZ)
Hurwitz and Kustin' have investigated the reactions
IrCIt-

+ IrBri- + IrCl2- + IrBriki

k2

using the temperature jump method, with spectrophotometric detection (495 mp).
With potassium nitrate media ( p = 0.1 M)values ofk, and k2 (10 "C) are 1.1 x lo7
and 1.5 x lo6 I.mole-'.sec-', respectively, with energies of activation of 5.7 ( k , )
and 7.5 (k,) kcal.mole-'. The rate coefficients were found to depend on the ionic
strength of the medium and possibly on the size of the cation present. Hurwitz
and Kustin' have also obtained a value for the exchange rate coefficient (Marcus
theory) from this data of 1 x lo5 I.mole-'.sec-'.
8.6

PLATINUM

8.6.1 The exchange reaction between Pt(ZV) and Pt(ZI); Pt(ZI)-catalysed substitution reactions of Pt(ZV)
Two types of mechanism have been postulated to explain the experimental
observations on this exchange system.

123

T R A N S I T I O N METALS

Rich and Taube', using the isotopic method and a separation achieved by
precipitation of the ion PtCIi- as Cs,PtC16, have found Pt exchange in lo-, M
HCI media, at 25 "C, to occur in minutes in the presence of uv light and in many
hours under normal conditions. On this basis and using additional information
obtained from experiments on chloride exchange with the ions PtC1:- and
PtCI;-, a mechanism involving a Pt(II1) species, uiz.

+ 2 PtC1:PCCIZ - 5 PtCIi - + c1

PfCI:- +PtClZ-

+ PtC1:-

PtC1:-

+exchange

PtC1:- +PtCIi- --+

exchange

was proposed to account for the experimental observations. A similar chain


mechanism has been postulated by McCarley et a/.,, for the platinum exchange
system, Pt(en)Br,-Pt(en)Br, , uiz.
Pt(en)Br,

5Pt(en)Br, + Br

Pt(en)Br, +Pt(en)Br, --+ exchange


Pt(en)Br, +Pt(en)Br,
Pt(en)Br, +Br

exchange

_ _ f

Pt(en)Br,

__f

in N , N'-dimethylformamide solution in the presence of light. Both of the above


exchange systems show inhibitions by IrCli-.
Basolo et
have found a rate law
~

rate

k'[Pt(IV)][Pt(II)][Cl-J

for the exchange of chloride ions with truns-Pt(en),CI:+ in the presence of


Pt(en):+ in the absence of light; k' has a value 15 12.mole-2.sec-' at 25 "C. The
mechanism postulated was
Pt(en):+ + ~ 1 - + Pt(en),Cl+

+[~1(en),~t-~1-~t(en),~l]~+
[Cl(en)2Pt-CI-Pt(en)2Cl]3 + ~t(en),Cl++ Pt(en),CI:
Pt(en),C~:+ +Pt(en),CI+

which, as Basolo et
have pointed out, provides a pathway for platinum
exchange. This has since been confirmedsp6.
Cox er d 5 using
,
the isotope 195Ptto label the species and tetraphenyl boron
References pp. 142-152

124

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

precipitation of platinum(II), Pt(en):+, as the separation method, have found


to 5 x
M,
the above rate law to be applicable over the ranges [Pt(II)] 8 x
[Pt(IV)] 5 x
to 5 x
M and [Cl-] 1 x
to 1.8 x
M . A value of
k' (25 "C) of 12.1 12.mole-2.sec-' was obtained with the activation entropy and
enthalpy of -20 cal.deg-'.mole-'
and 10.1 kcal.mole- ', respectively. The
activation energy found6 for chloride exchange was 11.5 kcal.mole-'. Using 14C
to label the species Pt(en);', similar confirmation has been obtained for this
exchange mechanism; k' (25 "C) has a value6 of 16 l2.rnoleC2.sec-' for media
9x
M with respect to H'.
Basolo et d 6 ,have found similar platinum(I1)-catalysed chloride exchange
reactions with other Pt(IV) complexes, including cis- and tr~ns-Pt(NH~)~Cl:+,
Pt(NH3),C13+ and trans-Pt(NH,),CI:.
These reactions proceed by the chloride
bridge mechanism above and the apparent rate coefficients (k' 12.mole-'.sec-',
25 "C) for platinum exchange, which was concluded to occur via this pathway,
are: trans-Pt(NH,),Cl:+-Pt(NH,):+,
6.3; c ~ ~ - P ~ ( N H ~ ) ~ C I : + - P ~ ( N H ~ ) : ~ ,
2.5 x
Pt(NH3),CI3+-Pt(NH3):+,
6x
trans-Pt(NH,),CI:Pt(NH,):+,
2.0; and ~~U~~-P~(NH,),CI~-P~(NH~)~C~~,
3.0. No exchange
was detected for the systems trans-Pt(tetrarneen),Cl~+ with either Pt(en):+ or
Pt(tetrameen): where chloride bridges cannot form (en = ethylenediamine and
tetrameen = tetramethyl en).
The systems (pn = propylenediamine)

- -

Wen):+

+ trans-~t(l-pn),~:++ truns-~t(en),~:++Pt(/-pn):+

where X is C1, Br or OH, have been investigated by polarimetry. The platinum


exchange rate increases in the order OH- < C1- < Br-, and the reaction
proceeds via the bridged mechanism6.
Johnson and Basolo' have found for the reactions
Pt(/-pn)g+ + trans-~t(en),clt++ 2 X -

+ trans-Pt(l-pn),X:+
+Pt(en)2, +2 CI+

where X- is Br-, CNS-, CN-, CI-, CNO- and NO;, half times of minutes,
whereas for X- = OH-, SO:-, ClO;, C2H30;, F-, and NO; the half times are
hours. These reactions, which involve exchange and substitution, are also thought
to occur via a bridged dimer mechanism. Polarimetry was again used to obtain
the experimental results.
Ellison et aZ.* have reported that the substitution of one nitrite ion for one
chloride ion in the Pt(IV) species, trans- and ~ i s - P t ( N H ~ ) ~ Cand
l i + trans-Pt(en),
Cl$+,is catalysed by Pt(1I). In the absence of light, a rate law,
rate = k'[Pt(IV)][Pt(II)][NO;]

125

T R A N S I T I O N METALS

and a mechanism similar to that proposed for the chloride exchange reactions
with Pt(1V) was thought to occur. The values of k' (l'.mole-'.sec-')
at 50 "C
and truns-Pt(NH,),CJ,
are: truns-Pt(en)'Clq+, 10.1 ; cis-Pt(NH,),CI:+, 2.1 x
3.25. The species trans-Pt(tetrameen),CI, was found not to undergo substitution.
Other platinum(I1) catalysed substitution reactions which have been reported
are
trans-Pt(NH,),Cl:+ + Br- = truns-Pt(NH,),ClBr'+ + C1(1)
trans-Pt(NH,),ClBr'+ +Brtrans-Pt(NH,),Br$+

+ C1-

trans-Pt(NH3),BrC1'+

+ C1+ NH,

+ C1= trans-Pt(NH,),BrCl'+ + Br-

rruns-Pt(NH,),Br:+

= truns-Pt(NH,),Cl:+

+ Br-

(2)

(3)
(4)

Pt(NH3)5C13++ C1-

Pt(NH3)513++C1-

truns-Pt(NH,),ICl'+

+ NH,

(6)

Pt(NH,),13++Br-

trans-Pt(NH,),IBr'+ +NH,

(7)

Pt(NH3)513f + I -

t r ~ n s - P t ( N H ~ ) ~+NH3
Iq+

(8)

Pt(NH3),Br3+ + Br-

trans-Pt(NH,),Br:+

+NH,

(9)

trans-Pt(NH,),CI~+ +NH,

(10)

truns-Pt(NH,),Cl:+

Pt(NH3),CI3+ C1-

(5)

for which rate data has been obtained using spectrophotometric methods, by
Rettew and Johnson' for reactions (l), (2), (3) and (4), Johnson and Berger"
for reaction (5) and Mason and Johnson" for reactions (6), (7), (8), (9) and (10).
Values of the apparent rate coefficients (I'.mole-'.sec-')
at 25 "C for reactions
(1) to (lo), with the corresponding activation enthalpies (kcal.mole-') and
entropies (cal.deg- '.mole-') and ionic strength in parentheses, are : 108(8,- 24;
U
, = 0.2 M ) , 1 . 9 lo4
~ (3, -30; p = 0.2 M ) , 6.3 (11, -20; ,U = 0.2 M ) , 4 . 2 lo3
~
(6, -22; ,U = 0.2 M ) , 1.21 (6, -37; p = 0.375 M ) , 5.6 x 10' (11, - 10; p = 0.016
M ) , 1 . 2 lo4
~ (8, -15; ,U = 0.016 M ) , 3 . 9 10'
~ (6, -29; p = 0.016 M ) , 12
(10, -19; ,U = 3.2 M ) and 1.2 x lo-, (18, - 13; p = 3.2 M ) , respectively. In
all the above reactions a halide-bridged species participates. Johnson and Berger'
have made some rate measurements on the similar catalysed system
trans-Pt(NH,),pyC13+

+ C1- + trans-Pt(NH,),CIZf

+py

In dimethylformamide solution, platinum exchange between Pt(en)Br, and


Pt(en)Br, has been found to be light-sensitive and catalysed by bromide ions. In
the absence of light, a value < 1.33 x lo-' I.mole-'.sec-' was obtained for the
observed rate coefficient (25 "C) from data obtained by the isotopic method
('"Pt) and separation using an ether precipitation of Pt(en)Br,. The bridge
mechanism was thought to occur'.
References p p . 142-152

126

IONS OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

In methanol solution reactions of the type (diars = o-phenylenebisdimethylarsine)


trans-Pt(PEt,),X,

+Pt(diars):+

+ Y - = Pt(diars),XY'+

+ trans-Pt(PEt,),X,

+ Xu

where X- and Y - are either chloride or bromide ions, also occur via a
halide bridge mechanism. Peloso and Ettore", who have made a study of these
reactions using a spectrophotometric method, have found a rate law
rate = k; K[Pt(diars):+][Pt(PEt,),X,][Y-]/(l+

K[Y-1)

where k; is an apparent rate coefficient and K the constant of the equilibrium


Pt(diars)$+ + Y -

+ Pt(diars),Y'

to be applicable. Values of k; at 30 "C (p = 0.0385 M ) and the associated activation parameters (AH* and AS') obtained by these workers are: 4 . 1 4 ~lo-' I.
mole-'.sec-', 6.4 kcal.mole-' and -39 cal.deg-'.mole-'(X- = C1-, Y- = Cl-)
7.65 1.mole-'.sec-', 5.3 kcal.mole-' and -41 cal.deg-'.mole-'
(X- = Cl-,
Y- = Br-); 2.65 x lo3 1.mole-'.set-', 3.4 kcal.mole-' and -32cal.deg- '.mole-'
(X-= Br-, Y - = Cl-); and 3.9 x lo3 l.mole-'.sec-', 3.8 kcal.mole-' and -30
cal.deg-'.mole-' (X- = Br-, Y- = Br-). The mechanism proposed was
trans-Pt(PEt,),X,

+Pt(diars),Y+ + X,(PEt,),Pt-X-Pt(diars),Y+

X,(PEt,),Pt-X-Pt(diars),Y

+ +

trans-Pt(PEt,),X, +Pt(diars),XY2

+ X-

The species trans-Pt(PEt,),X, and Pt(diars),XY'+ can undergo non-catalysed


substitution by the ion Y-. Reactions with Y- = I- have also been studied;
the final products are Pt(PEt,),I, and I;.
Reports have recently appeared in the literature on studies of the catalysed
reactions
truns-Pt(dien)NH,Cl;+
trans-Pt(CN),Br:-

+ 2Br-

= trans-Pt(dien)NH,Br:

+CI- = trans-Pt(CN),C1Br2-

trans-Pt(en)(NO,),Cl,
trans-Pt(NO,),Br~ -

+ 2Br+ CI-

+ 2C1-

+ Br-

= trans-Pt(en)(NO,),Br,

trans-Pt(NO,),C1Br2-

(1 1)
(12)

-I-2C1-

+ Br-

(13)
(14)

by Mason',, reactions (12) and (14), and Syamal and Johnson14, reactions (11)

T R A N S I T I O N METALS

and (13); and the reactions

+ Pt(dien)Br+ +2Br-

trans-Pt(dien)BrCl;

trans-Pt(dien)Br: +Pt(dien)Br+ + 2C1-

+Pt(dien)Br+ +2Br= trans-Pt(dien)Br: + Pt(dien)NO:


+ + Pt(dien)Br+ +2Br-

trans-Pt(dien)NO,Cl;

trans-Pt(NH,),Cl;

trans-Pt(NH,),CI:

+ 2C1-

+ 2Br-

trans-Pt(dien)NH,Bri

+ Pt(NH,): + 2Cl+

+ +Pt(dien)Br+ +2Br-

trans-Pt(dien)NH,Cl:

rate

trans-Pt(dien)NO,Br: +Pt(NH3):+

+Pt(dien)NH:+
=

by Bailey and Johnson".

+2C1-

+ Pt(dien)NO: + 2Br-

trans-Pt(NH,),Cl:+

trans-Pt(dien)Br+ +Pt(NH,):+

+2CI-

trans-Pt(dien)Br:

+ Pt(dien)NH:+ + 2C1-

For reactions (13) to (20) a rate law

k'[Pt(IV)][Pt(II)][X]

where X is either the chloride or bromide ion, has been ~ b t a i n e d ' ~ . 'For
~ . reactions (11) and (12) the rate laws obtained were
rate

(k'[Br] +k"[Br]')~t(IV)][Pt(II)]

rate

( k , +k'[Cl-])[Pt(IV)][Pt(II)]

and

respectively.
For reactions (1 1)-(20) the values of the rate coefficients ( k ' ) (12.mole-'.sec-')
with the associated activation enthalpy (kcal.mole- ') and entropy (cal.deg. mole-') and ionic strength in parentheses are: 3.0 x 10' (4.2, -33; p = 0.1 M)I4,
7.2~
10 (2.5, -42; p = 1.01 M)',, 1 . 6 4 ~10' (5.8, -29; p = 0.1 M)',, 7 . 6 ~
lo-' (-1.6, -64; p = 1.01 M ) ' , at 25 "C, and 2 . 0 10'
~ (6.2, -27; p = 0.2 M ) ,
~ (7.6, -23; p = 0.2 M ) , 1 . 2 lo2
~
9 . 3 10'
~ (5.0, -28; p = 0.2M), 1 . 3 10'
(8.4, -21; p = 0.2 M ) , 4 . 0 lo2
~ (5.9, -27; p = 0.2 M ) , and 1.7 x 10' (8.0,
-21; p = 0.2 M ) 1 5 at 24.2 "C, respectively. The corresponding values associated
with the rate coefficients k , and k" (at 25 "C) are 7.8 l.mole-'.sec-' (-0.5,

'

References pp. 142-152

128

I O N S O F SAME M E T A L I N D I F F E R E N T O X I D A T I O N S T A T E S

-56; p = 1.01 M)13 and 1 . 7 ~


lo3 13.mole-3.sec-' (5.9, -24; p = 0.1 M)',,
respectively.
Data has also been reported concerning some Pt(I1) catalysed reactions for
aqueous methanol and aqueous dioxane media".

9. Cerium and europium


9.1

THE E X C H A N G E REACTION BETWEEN

Ce(IV)

AND

Ce(II1)

Ion migration' and diffusion separation2 techniques, in conjunction with the


isotopic method (I4'Ce and '44Ce) have led to complete exchange being observed
in both sulphate and perchlorate media. However, in 6 M nitric acid solution with
a separation involving an ether extraction of Ce(IV) an observed rate coefficient
(25 "C) of 0.27 I.mole-'.sec-' and an activation energy of 13.4 kcal.mole-' have
been reported2.
As a result of a more detailed study, Gryder and Dodson3 have found, for
media
6 M in (u ) nitrate and (6) perchlorate, an order with respect to Ce(1II)
(1.7 x
to 1.9 x
M ) ofunity. The order withrespect to Ce(1V) (4.25 x
to 4.25 x
M ) , lay between zero and one in perchlorate media, and was about
0.90 in nitrate media. For perchlorate solution the rate law obtained was

rate = k'[Ce(III)]

+ ki[Ce(IV)][Ce(III)]

For media 6.18 M [HCIO,] at 0 "C, k' and k; have values 8.3 x
sec-' and
6.7 x
l.mole-'.sec-' with activation energies for these steps (k' and k ; ) of
19.4 and 16.8 kcal.mole- ', respectively. The mechanism proposed involved the
exchange steps
Ce(IV) + Ce(I1I)

-+

Ce(lV)+Ce(III)*

-+

and an excitation process, which depends on the environment

+ + Ce(III)* + S

Ce(II1) S

From the dependence of the rate on [H'] it was concluded that hydrolysed species
are involved in the excitation step. For nitrate media where the step associated
with k ; predominates under the experimental conditions, the effect of variation in
the [H'] (1-6 M ) at ,u = 6.18 M was studied and results were interpreted on the
basis of a rate law
rate = k ; [Ce(Iv)][Ce( III)]

CERIUM A N D E U R O P I U M

129

The variation in the rate coefficient k ; , viz.


k ; = k , +k;"H+]-'

was consistent with two exchange pathways, one involving fully aquated ions and
one hydrolysed species. At 0 "C k, and k;' have values of 2.5 x lo-' 1.mole-'.
sec-' and 0.3 mole.1-'.sec-' with activation energies of 7.7 and 24 kcal.mole-',
respectively.
Using a separation method based on the extraction of Ce(IV) by butyl phosphate
Parchen and Duke4 have made a further study of this exchange reaction. In
perchlorate media (- 6 M ) , at 0 "C, with Ce(1V) and Ce(II1) in the ranges 8 x
to
A4 and 8 x
to 8 x lop3M , respectively, the exchange data indicated
a rate law
rate

k ; [Ce(III)][Ce(IV)]

+ k;[Ce(III)][Ce(IV)]'

Variation in the hydrogen-ion concentration, 5.85 to 2.04 M at ,u = 5.9 M


(NaCIO,), gave data in accordance with the participation of hydrolysed species of
Ce(IV), the rate at low concentrations of H + being much faster than at high [H'].
Two pathways were suggested
Ce(0H):

+ Ce(II1)

ki

k2

CeOCeOH5++Ce(III)+

It is interesting to note that a dimer of Ce(rV) has also been invoked to account
for observations on this exchange system at a platinum surface5. The rate of exchange between the fully aquated ions of Ce(1V) and Ce(II1) was concluded to
be relatively slow4.
Observations, made by Sigler and Masters6, on the exchange in 0.4 M H2S04
solution, have led to a rate law
rate

kobs[Ce(IV)] [Ce(III)]

M and Ce(III), 5 x
to
which was obeyed over the ranges Ce(IV),
to 1.1 x lo-' M . kobsat 0 "C has a value
4 . 2 ~lo-' 1.mole-'.set-'. These
authors have also reported their results for the exchange reaction induced by
hydrogen peroxide. A mechanism

HzOz Ce(1V) + HOz +Ce(III)

H 0 2 +Ce(IV)
References pp. 142-IS2

+ H"

-+ Ce(III)+O,+H+

130

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

was proposed. The isotopic method ('44Ce) and n-butylphosphate separations


were used.
Challenger and Masters7 have found an X-ray induced cerium exchange reaction, which is faster than the normal exchange reaction, occurs in nitric and
sulphuric acid media.
In nitrate media (- 6 M ) , fluoride ion has a catalytic effect on the exchange
reaction between Ce(IV) and Ce(II1). Hornig and Libby* have made a detailed
study of this effect, over the range of added KF, 0 to 8 . 4 ~
M , and have
concluded that a pathway involving a monofluoro complex occurs, possibly involving a fluoride-bridged activated complex.
9.2

THE EXCHANGE REACTION BETWEEN

Eu(II1)

AND

Eu(I1)

The isotope ls2Eu has been used as a tracer in the study of this exchange
system'*2. In view of the instability of the Eu(1I) species the absence of both light
and oxygen are necessary. Separation of the two species, Eu(II1) and Eu(II),
was achieved using a hydroxide precipitation of the trivalent europium, brought
about by the addition of aqueous ammonia. Observations on the exchange in
perchlorate media were found to be obscured by the ClO; oxidation of Eu(1I).
However, in the presence of chloride ions the exchange rate could be measured in
perchlorate-containing media. A rate law for constant ionic strength (2.0 M )
rate

kob,[Eu(III)][Eu(I1)][Cl-]

was found to be obeyed. The concentration ranges employed were Eu(1II)


3 . 5 lo-'
~
to 1 . 2 lo-'
~
M , Eu(I1) 2.6 x lo-' to 6.8 x
M and C1- 9 x lo-'
to 1.86 M . The value of kobs(v = 2.0 M and 39.4 "C) is 1.8 x lo w3I2.mole-'.
sec- almost independent of hydrogen-ion concentration. The associated activation energy obtained over the range 32 to 50 "C was 20.8 kcal.mole-'. Meier and
Garner' have proposed the exchange pathway

',

EuC12

+ EuZ+

which is much more rapid than any involving the fully aquated or hydrolysed ions
of Eu(II1) and Eu(II), to occur.

10. Uranium, neptunium, plutonium and americium


10.1 E X C H A N G E

REACTIONS BETWEEN URANIUM IONS

The exchange reaction between U(V1) and U(IV) in aqueous sulphate',


chloride' and perchlorate3 media and in various mixed aqueous-organic solvents

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

131

containing hydrochloric acid4-' has been studied. The isotopic method (with
233Uas the tracer) was used with separations of the two oxidation states achieved
by ( a ) precipitation of the U(IV) as either the cupferron ~ o m p l e x or
~ , as
~ the
fluoride UF2 4 , 5 p and (b) aqueous extraction of U(IV) after the addition of
thenoyltrifluoroacetone in benzene3.
Betts' obtained evidence for the exchange in sulphate media being catalysed
by light of wavelengths
340 mp, the region of absorption of U(V1). In the presence of light a rate law, approximately

rate = k'[ U(V1)lt [U( IV)]t[H']-f

'

and activation energy of 8.5 kcal.mole- were found.


Rona2 has obtained results on the exchange in chloride media over the ranges
[U(VI)J 2 . 4 ~
to 1 . 5 lo-'
~
M , [U(IV)] 2 . 5 ~
to 6 . 4 ~
M and
[H'] 2.5 x
to 1.4 x lo-' M . which are consistent with a rate law
rate = k" [u(vr)][L.J(IV)]~
[H ]+

The rate was found to be unaffected by light and either [CI-] or [CIO,]. The
overall activation energy obtained was 33.4 kcal.mole- I. The mechanism of exchange suggested by Rona' was

+UOH3++H+
UOH3++ UO; + 2 H 2 0 +
U4++HZ0

(or,

OH

O=U-0-U-OH

O=U-0-U-OH

(K)

+ UOH3

OH d H
+

+2 H'

exchange

Masters and Schwartz3 have made a study of the exchange in perchlorate media
and have found two exchange pathways operate. One pathway involves exchange
via the process

U(V1) U(1V)

+ 2U(V)

with an activation energy of 38 kcal.mole-', the other via the pathway observed
for hydrochloric acid media with an activation energy of
28 kcal.mole-'. For
the former pathway a rate law
rate = kOb,[U(VI)][U(IV)]
where kobs = k"'([H+]3+K[H+]2)-1, was found to be obeyed over the range of
References p p . 142-152

132

I O N S O F SAME METAL IN D I F F E R E N T O X I D A T I O N STATES

conditions [U(Vi)] 2 x
to 4 x
M , [U(iV)] 1 x
to 5 x
M,
[H'] 6 x
to 2 x lo-' M , at temperatures of 25.1 to 47.1 "C. At 25.1 "Cand
p = 2.0 M , k"' hasavalue of 2 . 1 3 ~
mole2.1-2.sec-'. Values of the activation parameters of 37.5 kcal.mole-' and - 36 cal.deg-'.mole-' were calculated.
Ultraviolet radiation was found to increase the exchange rate in perchlorate media.
~ , found with various mixtures of aqueous hydrochloric acid
Amis et u I . ~ - have
with e t h a n ~ l ~acetone6,
.~,
or ethylene glycol7, exchange dependences with respect
to U(VI), U(IV) and H + which vary with the organic solvent content. Values of
2.70, 0 and - 1.26 (100 % ethanol) and 0.85, 0.66 and -0.53 (100 % ethylene
glycol) have been reported for the orders with respect to U(VI), U(IV) and H + .
With aqueous ethanol the activation energy was also found to depend upon the
solvent content; light was found to have no effect on the exchange rate'. Changes
in the mechanism are thought to O C C U ~ ~ * ~ * ' .
The rate coefficient for the exchange between U(Vi) and U(V) has been estimated as 52 I.mole-'.set-', at 25 "C in 1.0 M HC104, by Gordon and Taube',
from observations of the U(V)-catalysed water exchange reaction of aqueous
U(Vi).

10.2

R E A C T I O N S BETWEEN U R A N I U M IONS

Numerous studies have been made on the forward reaction of the equilibrium,
2 U(V)

+ U(Vi)+ U(iV)

using polarographic'

potenti~metric'.~
and spectrophotometric techniques'

-7,

O-

'

A majority of the investigations led to the rate law


- d [U( V)]/d t =

k',[H3' [U( V)]

and a mechanism
UO: + H +
UO: + U O O H 2 +
UOOH'

+ UOOH2+

--f

UO;++UOOH+
U(IV)

The value of the rate coefficient k', (12.mole-2.sec-') has been calculated as 130
(p = 0.4 M ) ' , 1800 (p = 3.8 M ) 2 , 156 ( p = 0.5 M ) 4 , 260 ( p = 0.5 M , D20)4,
143 (p = 0.4
436 (p = 2.1 M ) 6 , 192 (p = 1.0 M ) 7 , 417 (p = 2.0 M ) 9 ,
150 (p = 0.5 M)" and 135 (p = 1.0 M)" at 25 "C in perchlorate media. Imai6
has also reported that addition of the anions C1-, Br-, I- and NO; accelerates

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

I33

this reaction. Activation parameters AH* (kcal.mole-') and AS* (cal.deg-'.


mole-') for p = 2.1 M of 10.2 and -17 (perchlorate), 12.7 and -2.7 (0.1 M
added chloride) and 14.8 and +0.4 (1 M added bromide) have been calculated.
For chloride media, corresponding values of k ; (12.mole-2.sec-') of 175 ( p = 0.5
M ) , and
100 (p = 0.15 M ) 3 at 25 "C have been obtained. With 0.5 A4 sulphate
media a value of 1200 12.mole-2.sec-' at 30.4 "C has been found', with the corresponding activation energy 9.5 kcal.mole- '.
Newton and Baker", using wavelengths of 737 mp or 648 mp (U(1V) absorption), have obtained similar overall rate coefficients to those obtained by polarograph^^.^. However, they have found U(V1) to retard the disproportionation
reaction due to the presence of an additional process

U(VI)+ U(V) +U,(XI)


U,(XI)+ U(V)

K2

2 U(VI)+U(IV)

The dimer U,(XI), U20:+, was detected by its absorption at 737 mp. Analysis
of spectrophotometric data using the rate law

led to values of k ; (12.mole-2.sec-'), k , (Lmole-'.set-') and K2 (I.mole-') at


25.1 "C and p = 2.0 M with [ H + ] 5 x lo-' to 2 x lo-' M of
500,
14 and
16.7, respectively. The values of these coefficients show some dependence on
[H']; various reasons for this effect, including medium effects, have been fully
discussed by these workers. Activation parameters for the step associated with
k ; of 11 kcal.mole-' (AH*)and - 11 cal.deg-'.mole-' (AS*) have been evaluated.
The reverse reaction of the disproportionation equilibrium has been investigated
by Masters and S ~ h w a r t z ' ~It. provides a pathway for the exchange of U(V1)
and U(1V) as has been previously mentioned.

- -

10.3

EXCHANGE REACTIONS BETWEEN NEPTUNIUM IONS

The exchange reaction between Np(V1) and Np(V) has been investigated using
the isotopic method (239Np) and an extraction separation (Np(V1) with tributylphosphate or thenoyltrifluoroacetone in toluene)'-3. Cohen et al.' have
found, for the exchange in perchlorate media, a rate law

to be obeyed over the ranges [Np(VI)] 1.8 x


References p p . 142-152

to 9 . 2 7 ~

M , [Np(V)]

134

IONS O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

2.6~
M , and [H'] 0.32 to 0.99 M . At p = 1.0 M and a
to 1 . 0 4 ~
temperature of 0 "C, kobshas a value of
30 I.mole-'.set-', with an associated
energy and entropy G f activation of 8.3 kcal.mole-' and -24 cal.deg-'.mole-',
respectively.
Cohen et aL2 and Sullivan et aL3, in further articles, report some data obtained
under higher ionic strength conditions and have suggested that kobs depends on
[H+] in the manner3

Values obtained for the coefficients k , and k ; at 4.5 "C ( p = 3.0 M ) were 73.5
1.mole- '.sec-' and 15 12.mole-2.sec-', respectively, with corresponding activation energies of 12 and 15 kcal.mole-'.
In deuterated solvent the rate of exchange was found to be lower than in aqueous
media; ths ratio kob,(H20)/kob,(D20)was found to be dependent on the acid
concentration. Sullivan et aL3 have suggested a hydrogen atom transfer process
and a water bridging process
NpO:

+H + + N p 0 2 H 2 +

N p 0 2 H 2 ++NpO:+
NpO:+ +NpO:

K2

5exchange

+ H 2 0 +exchange
ki

to account for the two exchange pathways.


Cohen et aL2 have studied the effect of the addition of both chloride and
nitrate ions (up to 3 M a t p = 3.0 M ) on this exchange. Whereas nitrate ion was
found to have vertually no effect on the exchange rate, chloride ion had an accelerating influence. For exchange in the presence of chloride ions rate data was
found to be fitted by an expression
rate

with values of the rate coefficients k b , k 3 and k, at 0 "C ( p = 3.0 M , [H'] = 3.0
M ) of 88.8, 215, and 86 I.mole-'.sec-', respectively. The energies and entropies
of activation were found to have values of 10.6 kcal.mole-', - 12.6 cal.deg-'.
mole-' (kb), 15.4 kcal.mole-', 6.7 cal.deg-'.mole-' (k3) and 15.3 kcal.mole-',
5 cal.deg- '.mole- ' (k,). A chlorine-atom transfer mechanism has been discussed.
Cohen et ~ l . have
~ , also made some observations on the exchange in watersucrose and water-ethylene glycol mixed solvents containing perchloric acid
(0.106 M ) . Over a range of dielectric constant 68 to 88, no alteration in the exchange rate was observed.

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

135

Sullivan et aL5, have studied the perchlorate media exchange reaction between
Np(V) and Np(1V) using the isotopic method (239Nptracer) and a separation
based on the extraction of Np(IV) with thenoyltrifluoacetone in benzene. Data
to 4.9 x lo-' M , [Np(IV)] 3.9 x
obtained over the ranges [Np(V)] 5 x
to 3.2 x
to
4.5
x
lo-'
M were found to fit a rate expresM a n d [H'] 4 x
sion
rate

'

k [Np( V)] [H '3

+ k h[Np( V)] '[ N p( IV)] '[H '1 -

The rate coefficients k ; and k i at 25 "C (p = 1.2 M) have values and corresponding activation parameters of 1.10 x
12.mole-2.sec-', 17.6 kcal.
mole-', -22 cal.deg-'.mole-' and 6.45 x lo-' mole.l-l.szc-', 36.8 kcal.mole-',
32 cal.deg-'mole- ', respectively. Some surface catalysis was found; the above
results relate to teflon reaction vessels. The processes suggested by these authors
were, for the path associated with k ; ,
2NpO:+H+

+ NpO:'+NpOzH+

+ Np( OH):'

exchange

NpO,H+

and for the path associated with kk


Np4++NpO:
Np(OH):+ +NpO:+

+ NpO:++Np3'
+exchange

with the species Np(OH)i+ produced via the equilibria

+ NpOH3++ H +
N p O H 3 + + H 2 0+ Np(OH):++H'
Np4+ H,O

The above processes are related to the disproportionation reaction of Np(V)


which has also been investigated. However exchange rate constants and activation
parameters are different from those predicted from disproportionation data5.

10.4

REACTIONS BETWEEN NEPTUNIUM IONS

The reaction between Np(V) and Np(II1)


NpO: + N p 3 +

2 2 Np4'

in the absence of oxygen in perchlorate media has been studied by Hindeman


References pp. 142-152

136

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

al.', using a spectrophotometric method (723 mp, Np(IV) absorption). The


rate law governing the reaction

et

-d[NpO:]/dt

= k;[NpO:][Np3+][H+]

to 1 . 0 ~
was found to be obeyed over the ranges [NpOl] 2.5 x
M,
[Np3'] 2.98 x
M and [H'] 1.08 x lo-' to 1.98 M . At 25 "C
to 1.04 x
and p = 2.0 M , k ; has a value 43.2 12.mole-2.sec-' (HzO solution) and 34.3 1'.
mole-'.sec-' (D'O solution). The activation energies obtained in this study were
6.52 (H'O) and 5.62 (D'O) kcal.mole-l. The mechanism suggested by these
workers involves reaction of the species Np(O)(OH)'+, which is produced by the
equilibrium
NpO:

+ H + + Np(O)(OH)'+

Kl

and Np3+. The equilibrium reaction between Np(V1) and Np(IV)


Np(VI)+Np(IV)

2 2 Np(V)
k4

in aqueous perchlorate, ethylene glycol-water perchlorate and sulphate media


has also been investigated by Hindeman et al.2-5.The rate law controlling the
forward reaction ( k 3 ) was found from spectrophotometric data [724 mp for
Np(1V) or 983 mp for Np(V)] for perchlorate media to be
- d[Np(IV)]/dt or +d[Np(V)]/dt

k,,,[NpO:+][Np4+]"

where kobs= k;[H+]-', and a has a value of 3 in ethylene glycol-perchlcrate


media' and 1 in aqueous perchlorate media'. The rate was also found to be
dependent on the ethylene glycol concentration. In media 12.1 M ethylene glycol
k ; has a value of 1.58 x lo-' mo1et.l-.*sec-' (25 "C); in aqueous perchlorate
the corresponding value is 4 . 5 lo-'
~
mole.l-'.sec-' (at 25 "C, p = 2.0 M ) ,
with activation parameters (AH' and AS*) of 24.6 kcal.mole-' and 17.8 cal.deg-'.
mole-'. Using the data above and that available on the equilibrium, Hindeman
et al.', obtained values of kk (at 25 "C, p = 1.0 M ) and associated activation
13.mole-3.sec-', 17.1 kcal.mole-'and -38.1 cal.deg-'.
parameters of 9.57 x
mole-' for the reverse reaction, the rate law suggested being
- d [N p( V)]/d t =

'

kk[Np( V)] [H '3'

The mechanism proposed by these authors was

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

137

+ N p 0 2 ++ 2 H + or
+ Np(OH)z++2 H +

Np4+ H 2 0
N p 4 + + 2 H,O

N p O f + + N p 0 2 ++N p O i + N p 0 3 +
Np03++H20
NpOZf+Np(OH)ff
Np(O)(OH),+

+ NpO:

--f

+ 2 H + or

2 Np(O)(OH),'

+ NpOl +H+

A reinvestigation of the rate dependence on [H']


was concluded that the relation

has since been made3 and it

kobs = k;[H+]-2+k;'[H+]-3
was more accurate. Values of k; and k;' at 24.9 "C ( p = 2.2 M ) obtained were
4.27 x lo-, mole.l-'.sec-' and 5.04 x
mole2.1-2.sec- respectively.
For sulphate media much more complex rate laws for both the forward and
reverse steps of the reaction were found by Sullivan et a ~ the~ observed
,
rate
coefficient (kobs)being dependent on both [H'] and [HSO;]. For the forward
reaction

',

The values of the coefficients k;, k , , k6, k, and k, at 25 "C ( p = 2.2 M ) are
4 . 4 8 ~lo-, mole.l-l.sec-', 7 . 1 2 ~lo-' I.mole-'.sec-', 0.119 1.mole-'.sec-',
3.6 mole2.1-, and 1.7 mole3.1-j, respectively. The value of the overall activation
energy for the reaction in sulphate media is 23 kcal.mole-'.
For the reverse process the rate law obtained (sulphate media) was
-d[Np(V)]/dt

k'[Np(V)]2[H']o'09[HSO~]

(k,[HSO,I

+ k10[HS0,l2)(kI 1 + kl2[H+I)"P(V)l2

Values of k,, k,,, k,, and k,, at 25 "C ( p = 2.2 M ) are 5 x


12.mole-2.sec-',
1.46~
13.mole-3.sec-1, 7.39 x lo-' and 1.35 x lo-' mole.1-', respectively.
Possible transition states have been discussed4.
The reactions have also been studied in deuterated media3, the corresponding
values of the terms k; and k;' are 8 . 5 6 ~
mole.1-'.sec-' and 1 . 0 2 ~
lo-'
mole2.1-2.sec-' (at 24.9 "C, p = 2.2 A4 perchlorate media); an overall activation
References pp. 142-152

138

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N S T A T E S

energy for the reaction of 26.85 kcal.mole-' was also calculated from the experimental data. For the forward reaction in either perchlorate or sulphate media,
the reaction rate ratio (H,O: D20) lies between 4.33 and 5.65 depending on the
temperature. For the back reaction the same ratio has a value of 0.4, and the values
of the terms k , and k , , at 25 "C (p = 2.0 M ) for deuterated sulphate media are
1.17 x
12.mole-2.sec-' and 5.8 x
13.mole-3.sec-', respectively3.

10.5

EXCHANGE REACTIONS BETWEEN PLUTONIUM IONS

Keenan'v2 has made an investigation of the exchange reaction between Pu(IV)


and Pu(II1) in perchlorate media. The isotopic method was used with an a energy
analyser to separate the tracer activity (238Pu)from that normally present from
the major constituent (239Pu).Tributylphosphate extraction of the Pu(1V) formed
the basis of the separation method. It was shown that the rate law has the approximate form
rate

~,,,[Pu(IV)][PU(III)]

For 0.5 M perchloric acid kobs has a value 1.8 x lo2 I.mole-'.sec-' at 0 "C'.
From results obtained at p = 2.0 M , [H'] = 0.40 to 2.0 M , Keenan concluded
that the pathways operative were

ki

PU4+ +pu3+ +
kz

PuOH~++PU~+

the hydrolysed species PuOH3+, being produced via the equilibrium

Pu4+ + H,O

+ PuOH3++ H +

K,

Using known values of the constant K 2 , the values of k , and k 2 were calculated
as 1.8 x lo2 and 1.3 x lo4 I.mole-'.sec-',
respectively, at 0 "C and p = 2.0 M.
The activation energies and entropies obtained for the k , step were 7.7 kcal.
mole-' and -31 cal.deg-'.mole-'; for the k 2 step values of 2.8 kcal.mole-'
and -32 cal.deg-'.mole-'were found.

10.6

REACTIONS BETWEEN PLUTONIUM IONS

Rabideau et al. 1 , 2 have used a spectrophotometric method [wavelengths,


600 mp for (Pu(II1)' and 830 mp for ( P U ( V I ) ~to
] investigate the reactions

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

puo;+ + pu3+

3.puo:
7

139

+pu4+

k2

The rate law


-d[PuO;+]/dt

-d[Pu3+]/dt

k l ( l - K/K*)[PuO;+][Pu3+]

(where K and K * are the values of [Pu3+][Pu0;+]/[Pu0:][Pu4+]at equilibrium


and at any time t , respectively) was found to be obeyed in both aqueous and
deuterated perchlorate media. The calculated values of k , at 25 "C, ( p = 1.0 M ) ,
which were found not to depend on [H'], are 2.66 1.mole-'.sec-' (H,O) and
1.49 I.mole-'.sec-' (D,O). The values of k z under the same conditions after
correction for hydrolysis effects were calculated as 37.4 I.mole- '.sec-' (H20)
and 56.5 l.rnole-'.sec-' (D20). Activation parameters (AH' and AS') were:
4.82 kcal.mole-' and -40.4 cal.deg-'.mole-' (kl), 13.6 kcal.mole-' and -5.7
cal.deg-'.mole-' (k,) for H,O solution; 4.5 kcal.mole-' and -42.6 cal.deg-'.
mole-' (kl), 11.8 kcal.mole-' and - 11 cal.deg-'.mole-' ( k , ) for D 2 0 solution.
Neither of the rate coefficients (k, , k , ) was affected by the addition of chloride
ions.
In acetic acid media in the presence of 0.5 M HC104, a modified reaction
P ~ ( v I ) +pu(rrr)
~

3 P~(IV)

occurs, owing to the instability of Pu(V) in this media. Alexi et u I . ~have used a
spectrophotometric method to make a brief investigation and have suggested that
the steps

are amongst the processes occurring. The reaction appears to depend on the concentration of the Pu present.
From spectrophotometric studies on 0.5 A4 HCI solutions of plutonium containing Pu(V), Connick4 attempted to investigate the disproportionation reaction
of this species
k'a

2 Pu(V) $ Pu(IV)+Pu(VI)
k's

From his results, Connick concluded that the reactions also occurring were
Pu(V)+Pu(III)

22Pup)
k.6

Pu(V)+Pu(IV)
References pp. 142-152

2 Pu(VI)+Pu(III)

k'

140

I O N S OF SAME METAL IN D I F F E R E N T O X I D A T I O N S T A T E S

and he was able to obtain values of the rate coefficients k; and k ; at 25 "C of
5.8 x lo-' and < 2 x
Lmole-'.set-', respectively; in conjunction with
equilibrium data, estimates of kk and kk for the same conditions, of 3 . 5 ~
and < 1.2 x
I.mole-'.sec-', respectively, were made.
Rabideaus, using EMF measurements to obtain rate data, has since reinvestigated
the disproportionation reaction of Pu(V), using solutions containing very small
amounts of Pu(IV) and Pu(II1). He was able to conclude from the dependence
of the rate coefficient (kj) on the first power of [H'] that the steps and equilibrium
present were
PuO:

+ H + + Pu(O)OH'+

PuO: +Pu(O)OH'+ +PuOf++Pu(O)OH+


Pu(O)OH+ +Pu(IV)
The values of k ; obtained from this work are 4.0 x
I.mole-'.sec-' (at 25 "C,
= 1.0 M , in HzO media) and 5.25 x
l.mole-'.sec-' (for
DzO media under the same conditions). For aqueous media, activation parameters
of A H f = 19 kcal.mole-' and A S f = -5.8 cal.deg-'.mole-' were calculated.
Rabideau6 has also investigated, using EMF measurements, the overall reaction

p = 1.0 M , [H']

3Pup)

+ Pu(VI)+2 Pu(rI1)

which is governed by the rate law


-d[Pu(IV)]/dt

3 k~(l-K*/K)[Pu(IV)IZ

(where K and K" are the values of [Pu(III)Iz[Pu(VI)]/[PU(IV)]~at equilibrium


and at time t , respectively) and involves the reactions defined by kk , k; , k ; , and
k ; . From results obtained with perchlorate media of various acidities (kk dependent
on [H'I3) it was concluded that the equilibria and pathways operating were
P u 4 + + H z 0 ;t P u O H 3 + + H f
PuOH3+ + H,O
PuOH3++Pu(0H):'

+ Pu(OH):+ + H +
--+Pu3++Pu0:

+H30+

puo: +pu4+ + PU3++ P u O f +


The value of kk at 25 "C ( p = 1.0 M ) obtained was 2 . 7 ~lo-' I.mole-'.sec-'.
Previously, Connick and McVey7, using data availableat that time for this system
had calculated a value of kd of 2.5 x
I.mole-'.sec-' for approximately the
same conditions.

10

U R A N I U M , N E P T U N I U M , P L U T O N I U M A N D AMERICIUM

141

Rabideau and Cowan' have made a similar investigation of this reaction in


hydrochloric acid media and have found identical features. At 25 "C (media
I.mole-'.sec-' with associated
1 M i n HCI, p = 1 .O M ) kk has the value 1.5 x
activation parameters (AH* and ASt) of 39 kcal.mole-' and 53 cal.deg-'.moleThese results show good agreement with those calculated previously by Connick
McVey' (kk = 9.8 x
I.mole-'.sec-', AH* = 40 kcal.mole-' and AS* = 60
cal.deg- '.mole- ') for similar conditions.

'.

10.7

E X C H A N G E REACTIONS BETWEEN A M E R I C I U M IONS

Keenan et a[.' have carried out a brief survey on the exchange reactions
A m O l + Am3+ -+
AmO:'

+AmOz

-+

In perchlorate media the latter reaction was found to be rapid (100 % exchange
in < 60 sec) and the former reaction slow (half times of > 200 hours at 100 "C).
The isotopic method was used (242Am).

10.8

REACTIONS BETWEEN AMERICIUM IONS

Am(V), AmOz, undergoes disproportionation and redox reactions in aqueous


acidic media, which can be followed by spectrophotometry at wavelengths of
812 mp for Am(III),
715 mp for Am(V) and
992 mp for Am(V1). The
earlier work'-5 on solutions of Am(V) in hydrochloric', nitric5, sulphuric5, and
perchloric acid' - media, which were hindered by a-radiation reactions (241Am),
led to the conclusion that the rate law

-d[AmOl]/dt

k,,,[Am0~]'[H+]4

was ~ b e y e d ' . ~ - ~ .
In 1962, Coleman6, using the relatively stable 243Am, was able to study the
reaction of Am(V) in various media in more detail and found a rate law
-d[AmO:]/dt

k;[AmOl]2[H+]'

+ k',"Am0:]2[H+]3

In perchlorate media, values of k ; and k;' at 75.7 "C (p = 2.0 M ) are 7 x


13.mole-3.sec-' and 4.6 x
14.mo1e-4.sec-1, respectively. The rate increased
in the solvent order HC104 < HNO, < HCI < H,S04. No rate measurements
in dibutylphosphoric acid media were possible.
References pp. 142-152

142

I O N S OF S A M E M E T A L I N D I F F E R E N T O X I D A T I O N S T A T E S

The stoichiometric equation found for the overall reaction was


3 Am(V) = 2 Am(VI)+Am(lII)

with the probable steps


2 Am(V)

+ Am(V1) + Am(IV)

Am(IV)+ Am(V)

+ Am(III)+Am(VI)

The reaction between Am(V1) and Am(1II) has been detected6 but no kinetic
data has been reported.
Americium(1V) has been found to undergo similar reactions, viz.

+ Am(V)+ Am(II1)
Am(IV) + Am( V) + Am(111) + Am( VI)
2 Am(1V)

Penneman et ~ l . have
~ , studied these reactions in acidic sulphate, nitrate and
perchlorate media, and have found that only in sulphate media is the latter step
relatively important. The disproportionation reaction rate coefficient has been
I.mole-'.sec-' at 0 "C (media5 x lo-' M i n HNO,).
estimated as > 1.03 x

REFERENCES
1. Introduction

1 H. A. C. MCKAY,Nature, 42 (1938) 997.


2 E. EICHLER
AND A. C. WAHL,J. Am. Chem. Soc., 80 (1958) 4145.
3 Y.A. IM A N D D. H. BUSCH,J. Am. Chem. Soc., 83 (1961) 3357.
4 Y. A. IM AND D. H. BUSCH,J. Am. Chem. Soc., 83 (1961) 3362.
5 H. M. MCCONNELL
AND H. E. WEAVER,
J. Chem. Phys., 25 (1956) 307.
AND S. B. BERGER,
J. Chem. Phys., 27 (1957) 230.
6 H. M. MCCONNELL
AND S. I. WEISSMAN,
J. Chem. Phys., 24 (1956) 473.
7 C. R. BRUCE,R. E. NORBERG
8 R. L. WARDAND S. I. WEISSMAN,
J. Am. Chem. Soc., 79 (1957) 2086.
AND A. C. WAHL,J. Chem. Phys., 38 (1963) 1591.
9 M. W. DIETRICH
10 D. W. LARSENAND A. C. WAHL,J. Chem. Phys., 41 (1964) 908.
I 1 N. SUTIN,Ann. Rev. Nucl. Sci., 12 (1962) 285.
AND R. G. WILKINS,Chem. Rev., 57 (1957) 743.
12 D. R. STRANKS
Quart. Reo. (London), 15 (1961) 207.
13 J. HALPERN,
14 A. A. VLCEK,Chemie, 9 (1957) 305.
15 C. B. AMPHLETT,
Quart. Reo. (London), 8 (1954) 219.
R. J. MARCUSAND H. EYRING,
Chem. Reo., 55 (1955) 157.
16 B. J. ZWOLINSKI,
2.1 The exchange reaction between Cu(l1) and Cu(1)
1 H. M. MCCONNELL
AND H. E. WEAVER,
J. Chem. Phys., 25 (1956) 307.
A N D N. DAVIDSON,
J . Am. Chem. SOC.,72 (1950) 3168
2 H. M. MCCONNELL

REFERENCES

143

2.2 The exchange reaction between Ag(I1) and A g ( I )


1 B. M. GORDON
A N D A. C. WAHL,J . Am. Chem. Soc., 80 (1958) 273.
Ric. Sci., 26 (1956) 3072.
2 M. BRUNOA N D V. SANTORO,

2.3 The exchange reaction between Au(II1) and Au(1)


1 A. TURCO
AND G. SORDILLO,
Gazz. Chim. Ital., 85 (1955) 977.

2.4 The exchange reaction between A u ( I I I ) and A u ( I I )


1 R. L. RICHAND H. TAUBE,
J. Phys. Chem., 58 (1954) 6.
2.5 The disproportionation of Au(1I)
1 R. L. RICHA N D H. TAUBE,
J. Phys. Chem., 58 (1954) 6.
3.1 The exchange reaction between H g ( I I ) and H g ( I ) ; the disproportionation reaction of Hg(I1)
1
2
3
4
5
6
7

E. L. KI NG ,J . Am. Chem. Soc., 71 (1949) 3553.


M. HAISSINSKY
A N D M. COTTIN,
J. Chim. Phys., 46 (1949) 476.
R. L. WOLFGANG
AND R. W. DODSON,
J. Phys. Chern., 56 (1952) 872.
E. L. KING,J. Phys. Chem., 56 (1952) 876.
J. Phys. Chem., 56 (1952) 876.
A. W. ADAMSON,
R. L. WOLFGANG
AND R. W. DODSON,
J. Am. Chem. Soc., 16 (1954) 2004.
D. PESCHANSKI,
J. Chini. Phys., 50 (1953) 640.

3.2 The exchange reaction between H g ( I I ) and H g ( I ) in non-aqueous media


1 D. PESCHANSKI,
J. Chim. Phys., 50 (1953) 634.
4..1 The exchange reaction between TI(III)and T I ( I )
1 J. ZIRKLER,
Z. Physik., 97 (1934) 410; 98 (1935) 75; 99 (1936) 669; Z. Pliysik. Chern.,

A187 (1940) 103.


V. MAJER,Z. Physik. Chern., A179 (1937) 51.
G. HARBOTTLE
A N D R. W. DODSON,
J . Am. Chem. Soc., 70 (1948) 880.
R. J. PRESTWOOD
AND A. C. WAHL,J. Am. Chem. Soc., 70 (1948) 880.
R. J. PRESTWOOD
AND A. C. WAHL,J. Am. Chern. Soc., 71 (1949) 3137.
G. HARBOTTLE
AND R. W. DODSON,
J. Am. Chem. Soc., 73 (1951) 2442.
R. W. DODSON,J . Am. Chem. Soc., 75 (1953) 1795.
F. J. C. ROSSOTTI,
J. Inorg. Nucl. Chem., 1 (1955) 159.
G. BIEDERMANN,
Arkiv Kemi, 5 (1953) 441.
R. P. BELLAND J. H. B. GEORGE,Trans. Faraday Soc., 49 (1953) 619.
E. ROIGAND R. W. DODSON,
J. Phys. Chem., 65 (1961) 2175.
s. GILKSA ND G . M. WAIND,Discussions Faraday Soc., 29 (1960) 102.
G. M. WAIND,Discussions Furaduy Soc., 29 (1960) 135.
S. GILKS,T. RODGERS
AND G. M. WAIND,Trans. Faraday SOC.,57 (1961) 1371.
G. E. CHALLENGER
AND B. J. MASTERS,
J . Am. Chem. SOC.,78 (1956) 3012.
C. H. BRUBAKER
AND J. P. MICKEL,
J. Inorg. Nucl. Chem., 4 (1957) 55.
C. H. BRUBAKER,
K. 0.GROVES,
J. P. MICKELAND C. P. KNOP,J. Am. Chem. Soc., 79
(1957) 4641.
18 D. R. WILES,Can. J. Chem., 36 (1958) 167.
AND R. W. DODSON,
J . Am. Chem. Soc., 77 (1955) 2651.
19 E. PENNA-FRANCA
M. H. FORD-SMITH,
R. p. BELLAND R. W. DODSON,Discussions
20 L. G. CARPENTER,
Faruday SOC.,29 (1960) 92.

2
3
4
5
6
7
8
9
10
I1
12
13
14
15
16
17

144

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES


21 C. H. BRUBAKER
AND C. ANDRADE,
J. Am. Chem. SOC.,81 (1959) 5282.
A N D M. C. DORFMAN,
J. Am. Chem. SOC.,83 (1961) 1254.
22 J. W. GRYDER
D. R. ROSSEINSKY,
J. B. STEADAND A. G . SYKES,Discussions
23 W. C. E. HIGGINSON,
Faruday SOC.,29 (1960) 49.
24 A. G . SYKES,
J. Chem. SOC.,(1961) 5549.
25 H. MCCONNELL
AND N. DAVIDSON,
J. Am. Chem. SOC.,71 (1949) 3845.
AND J. R. YANDELL,
Exchange Reactions, I.A.E.A., Vienna, 1965, p. 83.
26 D. R. STRANKS
AND D. R. WILES,J . Chem. SOC.A, (1970) 323.
27 R. G. MCGREGOR
AND J. K. YANDELL,
J. Phys. Chem., 73 (1969) 840.
28 D. R. STRANKS

5.1 The exchange reaction between Sn(1V) and Sn(l1) in aqueous media
1
2
3
4

C. I. BROWNE,
R. P. CRAIGA N D N. DAVIDSON,
J. Am. Chem. Sac., 73 (1951) 1946.
R. P. CRAIGAND N. DAVIDSON,
J. Am. Chem. SOC.,73 (1951) 1951.
G. GORDON
A N D C. H. BRUBAKER,
J. Am. Chem. SOC.,82 (1960) 4448.
J. E. WHITNEY
AND N. DAVIDSON,
J. Am. Chem. SOC.,71 (1949) 3809.

5.2 The exchange reaction between Sn(1V) and Sn(l1) in non-aqueous media
1 E. G. MEYERAND
M. A. MELNICK,
J. Phys. Chem., 61 (1957) 367.
A N D M. KAHN,J. Am. Chem. SOC.,73 (1951) 4950.
2 E. G. MEYER
5.3

The exchange reaction between Pb(1V) and Pb(IZ) in aqueous media


1 E. ZINTLAND A. RANCH,
Ber., 57B (1924) 1743.
2 A. FAVA,J . Chim. Phys., 50 (1953) 403.

5.4 The exchange reaction between Pb(1V) and Pb(l1) in non-aqueous media
1 G. V. HEVESY
A N D L. ZECHMEISTER,
Ber., 53B (1920) 415.
AND T. H. NORRIS,
J. Am. Chem. SOC.,74 (1952) 4986.
2 E. A. EVANS,J. L. HUSTON
6.1 The exchange reaction between As( V ) and As(ll1)

1 J. N. WILSONAND R. G. DICKENSON,
J. Am. Chem. SOC.,59 (1937) 1358.
P. DAUDEL,
R. DAUDEL
AND P. MAGNIER,
Compt. Rend., 224 (1947) 195.
2 M. MARTIN,
6.2 The exchange reaction between Sb( V ) and Sb(ll1) in aqueous media
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

J. E. WHITNEY
A N D N. DAVIDSON,
J . Am. Chem. Soc., 69 (1947) 2076.
J. E. WHITNEY
AND N. DAVIDSON,
J. Am. Chem. SOC.,71 (1949) 3809.
H. M. NEUMANN,
J. Am. Chem. SOC.,76 (1954) 2611.
H. M. NEUMANN
AND R. W. RAMETTE,
J. Am. Chem. SOC.,78 (1956) 1848.
N. A. BONNER,
J. Am. Chem. SOC.,71 (1949) 3909.
H. M. NEUMANN
AND H. BROWN,
J. Am. Chem. Soc., 78 (1956) 1843.
C. H. CHEEK,Ph. D. Thesis, Washington University, 1953.
C. H. CHEEK,
N. A. BONNER
AND A. C. WAHL,J. Am. Chem. SOC.,83 (1961) 80.
T. KAMBARA,
K. YAMAGUCHI
AND S. YASUBA,
Exchange Reactions, I.A.E.A., Vienna,
1965, p. 101.
N. A. BONNER
AND W. GOISHI,
J. Am. Chem. SOC.,83 (1961) 85.
A. TURCOA N D G. FAROANE,
Ric. Sci., 25 (1955) 2887.
A. TURCO,Gazz. Chim. Ital., 88 (1958) 365.
C. H. BRUBAKER
AND J. A. SINCIUS,
Abstr. Paper 139th Meeting Am. Chem. SOC.,Missouri, 1961.
C. H. BRUBAKER
AND J. A. SINCIUS,
J. Phys. Chem., 65 (1961) 867.
A. TURCO,Gazz. Chim. ltal., 2 8 A (1953) 231.

REFERENCES

145

6.3 The exchange reaction between Sb( V ) and Sb(III) in non-aqueous rnediu

F. B. BARKER
AND M. KAHN,J. A m . Chent. Soc., 78 (1956) 1317.
K. R. PRICEA N D C. H. BRUBAKER,
Exchange Reactions, I.A.E.A., Vienna, 1965, p. 113.
K. R. PRICEAND C. H. BRUBAKER,
Inorg. Chenr., 4 (1965) 1351.
W. E. BECKER
A N D R. E. JOHNSON,
J. Am. Chern. SOC.,79 (1957) 5157.

I
2
3
4

6.4 The Sb(1II)-catalysed hydrolysis of Sb( V )


I H. M. NEUMANN
AND R. W. RAMETTE,
J. A m . Chem. Soc., 78 (1956) 1848.
2 N. A. BONNER
AND W. GOISHI,
J. Am. Chenr. SOC.,83 (1961) 85.

7.1 The exchange reaction between Te( V I ) and T e ( I V )


I M. HAISSINSKY
AND M. COTTIN,Anal. Chim. Acta, 3 (1949) 226.
2 M. W. HANSON
AND T. C. HOERING,
J. Phys. Chem., 61 (1957) 699.

8.1.1 The exchange reaction between V ( I I I ) and V ( I I )


1 W. R. KINGA N D C. S. GARNER,
J. Am. Chem. SOC.,74 (1952) 3709.
2 K. V. KRISHNAMURTY
AND A. C. WAHL,J. Am. Chenr. Soc., 80 (1958) 5921.

The exchange reaction between V ( I V )and V ( I 1 I )

8.1.2

1 S . G. FURMAN
AND C. S. GARNER,
J. Am. Chern. Soc., 14 (1952) 2333.

The exchange reuction between V( V ) and V ( I V )

5.1.3

I H. A. TEWES,J. B. RAMSEY
A N D C. S. GARNER,
J. Ant. Chent. Soc., 72 (1950) 2422.
2 C. R. GIULIANO
AND H. M. MCCONNELL,
J. Inorg. Nucl. Chem., 9 (1959) 171.

8.1.4 Reactions between vanadium ions

I
2
3
4
5
8. I .5

T. W. NEWTON
A N D F. B. BAKER,
J. Phys. Chern., 68 (1964) 228.
T. W. NEWTON
A N D F. B. BAKER,
Inorg. Chem., 3 (1964) 569.
J. W. OLVERAND J. W. Ross, J. Phys. Chem., 66 (1962) 1699.
N . A. DAUGHERTY
AND T. W. NEWTON,
J. Phys. Chem., 68 (1964) 612.
J. H. ESPENSON
A N D L. A. KRUG,Inorg. Chern., 8 (1969) 2633.
Reactions between tantalum cluster ions

1 J. H. ESPENSON
A N D R. E. MCCARLEY,
J. Am. Chem. Soc., 88 (1966) 1063.
2 J. H. ESPENSON
A N D D. J. BOONE,Inorg. Chem., 7 (1968) 636.
3 N. WINOGRAD
A N D T. KUWANA,
J . Am. Chem. Soc., 92 (1970) 224.

8.2.1

The exchange reaction between Cr(II1) and Cr(1I)


1 R. A. PLANE
A N D H. TAUBE,
J. Phys. Chern., 56 (1952) 33.
A. ANDERSON
A N D N. A. BONNER,
J. Am. Chem. Soc., 76 (1954) 3826.

2
3
4
5
6
7
8
9

H. VAN DER STRAATEN


AND A. H. W. ATEN,Rec. Trav. Chirn., 73 (1954) 89.
D. L. BALLAND E. L. KING,J. An?. Chem. Soc., 80 (1958) 1091.
H. TAUBE
A N D E. L. KING,J. Am. Chem. SOC.,76 (1954) 4053.
R. SNELLGROVE
A N D E. L. KING,Inorg. Chem., 3 (1964) 288.
K. A. SCHRODER
A N D J. H. ESPENSON,
J. Am. Chem. Soc., 89 (1967) 2548.
J. P. BIRKA N D J. H. ESPENSON,
J. Am. Chem. SOC.,90 (1968) 2266.
R. SNELLGROVE
AND E. L. KING,J. Am. Chem. SOC.,84 (1962) 4609.

146

I O N S OF SAME METAL I N D I F F E R E N T O X I D A T I O N STATES

10 D. R. STRANKS,
Discussions Faraday Soc., 29 (1960) 79.
11 E. DEUTSCH
AND H. TAUBE,
Inorg. Chem., 7 (1968) 1532.
Inorg. Chem., 9 (1970) 486.
12 D. H. HUTCHITAL,

8.2.2 The exchange reaction between Cr( V I ) and Cr(II1)


1 R. MUXART,
P. DAUDEL,
R. DAUDEL
AND M. HAISSINSKY,
Nature, 159 (1947) 538.
2 H. E. MENKER
AND C. S. GARNER,
J. Am. Chem. SOC.,71 (1949) 371.
AND J. W. KENNEDY,
J . Chem. Phys., 18 (1950) 97.
3 W. H. BURGUS
4 C. ALTMAN
AND E. L. KING,J. Am. Chem. Soc., 83 (1961) 2825.

8.2.3 The reaction between Cr( V I ) and Cr(II)


1 L. S . HEGEDUS
AND A. HAIM,Inorg. Chem., 6 (1967) 664.
2 M. ARDONA N D R. A. PLANE,J. Am. Chem. SOC.,81 (1959) 3197.
8.2.4 Cr(l1)-catalysedsubstitution and isomerisation reactions of Cr(l1l)
1 H. TAUBE
AND E. L. KING,J. Am. Chem. SOC.,76 (1954) 4053.
J. Chem. Soc., A (1966) 1518.
2 A. ADINAND A. G. SYKES,
AND A. HAIM,J. Am. Chem. Soc., 88 (1966) 3450.
3 D. E. PENNINGTON
AND A. G. SYKES,
J. Chem. SOC.A , (1967) 1504.
4 A. ADIN,J. DOYLE
5 J. DOYLE,A. G. SYKESAND A. ADIN,J . Chem. SOC.A , (1968) 1314.
6 J. H. ESPENSON
AND D. W. CARLYLE,
Inorg. Chem., 5 (1966) 586.
AND S. G. SLOCUM,
Inorg. Chem., 6 (1967) 906.
7 J. H. ESPENSON
8 Y. CHIAAND E. L.KING,Discussions Faraday Soc., 29 (1960) 109.
9 A. HAIM,J. Am. Chem. Soc., 88 (1966) 2325.
J. Am. Chem. SOC.,90 (1968) 2266.
10 J. P. BIRKAND J. H. ESPENSON,
A N D H. MEYERS,
J. Am. Chem. SOC.,76 (1954) 2103.
11 H. TAUBE
AND w. L. REYNOLDS,
Inorg. Chem., 2 (1963) 468.
12 H. B. JOHNSON
J. Am. Chem. SOC.,80 (1958) 1084.
13 A. E. OGARDAND H. TAUBE,
14 R. D. CANNON,
J. Chem. SOC.A , (1968) 1098.
1 5 M. J. DE CHANT
AND J. B. HUNT,J. Am. Chem. SOC.,89 (1967) 5988.
16 M. J. DE CHANTAND J. B. HUNT,J. Am. Chem. SOC.,90 (1968) 3695.
AND A. HAIM,Inorg. Chem., 5 (1966) 1887.
17 D. E. PENNINGTON
AND J. E. EARLEY,
J . Chem. Soc. A , (1968) 1102.
18 R. D. CANNON
19 R. F. N. THORNELEY,
B. KIPLINGAND A. G. SYKES,
J. Chem. SOC.A, (1968) 2847.
20 A. HAIMAND N. SUTIN,J. Am. Chem. SOC.,87 (1965) 4210.
21 A. HAIMAND N. SUTIN,J. Am. Chem. SOC.,88 (1966) 434.
J. Am. Chem. SOC.,90 (1968) 1 1 53.
22 J. P. BIRKAND J. H. ESPENSON,
AND A. HAIM,Inorg. Chem., 6 (1967) 2138.
23 D. E. PENNINGTON
Inorg. Chem., 7 (1968) 991.
24 J. P. BIRKAND J. H. ESPENSON,
25 E. DEUTSCH
AND H. TAUBE,
Inorg. Chem., 7 (1968) 1532.
AND H. TAUBE,
J. Am. Chem. SOC.,90 (1968) 1162.
26 F. NORDMEYER
Inorg. Chem., 9 (1970) 486.
27 D. H. HUTCHITAL,
J. B. HUNTAND L. PENZHORN,
Inorg. Chem., 7 (1968) 1467.
28 D. W. HOPPENJANS,
AND R. E. YELIN,Inorg. Chem., 7 (1968) 2667.
29 R. G. WILKINS

8.2.5 The exchange reaction between Mo( V ) and M o ( I V )


1 R. L. WOLFGANG,
J. Am. Chem. SOC.,74 (1952) 6144.
2 R. CAMPION,
N. PURDIE
AND N. SUTIN,
J. Am. Chem, Soc., 85 (1963) 3528.

8.2.6 The exchange reacrion between W ( V ) and W ( I V )

c. s. GARNER,
J. Am. Chem. Soc., 77 (1955) 5272.
s. I. WEISSMAN AND c. s. GARNER,J. Am. Chem. Soc., 78 (1956) 1072.

1 E. L.GOODENOWAND

REFERENCES

147

8.3.1 The exchange reaction between M n ( I I ) and M n ( I )

I D. S. MATTESON
AND R. A. BAILEY,
J. Am. C h e m Soc., 89 (1967) 6389.
AND R. A. BAILEY,
J. Am. Chem. Soc., 91 (1969) 1975.
2 D. S. MATTESON
8.3.2 The exchange reaction between Mn(III) and M n ( I I )
1 M. POLISSAR,
J . Am. Chem. SOC.,58 (1936) 1372.
J . Phys. Coll. Chem., 5 5 (1951) 293.
2 A. W. ADAMSON,
AND N. SUTIN,J . Phys. Chem., 68 (1964) 174.
3 H. DIEBLER
J . Phys. Chem., 56 (1952) 858.
4 A. W. ADAMSON,
5 R. G. WILKINS
AND R. E. YELIN,Inorg. Chem., 7 (1968) 2667.

8.3.3 The exchange reaction between Mn( V I I ) and Mn( V I )


1
2
3
4
5

6
7
8
9

W. F. LIBBY,J . Am. Chem. SOC.,62 (1940) 1930.


H. C. HORNIG,G. L. ZIMMERMAN
AND W. F. LIBBY,
J. Am. Chem. Soc., 72 (1950) 3808.
A. W. ADAMSON,
J. Phys. Chem., 55 (1951) 293.
N. A. BONNER
A N D H. A. POTRATZ,
J . Am. Chem. SOC.,73 (1951) 1845.
J. C. SHEPPARD
AND A. C. WAHL,J. Am. Chem. SOC.,7 5 (1953) 5133.
J. C. SHEPPARD
AND A. C. WAHL,J. Am. Chem. SOC.,79 (1957) 1020.
L. GJERTSEN
AND A. C. WAHL,J. Am. Chem. SOC.,81 (1959) 1572.
0. E. MEYERS
A N D J. C. SHEPPARD,
J. Am. Chem. SOC.,83 (1961) 4730.
A. D. BRITTAND W. M. YEN,J. Am. Chem. SOC.,83 (1961) 4516.

8.3.4 The exchange reaction between Mn( V I I ) and M n ( I I I )


1 M. POLISSAR,
J. Am. Chem. Soc., 58 (1936) 1372.
8.3.5 The exchange reaction between Mn(VI1) and M n ( I I )
1 M. POLISSAR,
J. Am. Chem. Soc., 58 (1936) 1372.
J. Phys. Chem., 5 5 (1951) 293.
2 A. W. ADAMSON,
J. Am. Chem. SOC.,77 (1955) 4212.
3 J. A. HAPPEAND D. S. MARTIN,
8.3.6 The reaction between Mn( V I I ) and Mn(II)
1 F. C. TOMPKINS,
Trans. Faraday SOC..38 (1942) 128.

2
3
4
5
6

M. J. POLISSAR,
J. Phys. Chem., 39 (1935) 1057.
M. A. GUYARD,
Bull. Soc. Chim., 1 (1864) 89.
H. F. LAUNER
AND D. M. YOST,J. Am. Chem. SOC.,
56 (1934) 2571.
G. R. WATERBURY,
A. M. HAYESAND D. S. MARTIN,J. Am. Chem. Soc., 74 (1952) 15.
D. R. ROSSEINSKY
AND M. J. NICOL,Trans. Faraday SOC.,61 (1965) 2718.

8.4.1 The exchange reaction between Fe(III) and Fe(II) in aqueous media

I
2
3
4
5
6
7
8
9
10

L. VAN ALTENAND C. N. RICE,J. Am. Chem. SOC.,70 (1948) 883.


V. J. LINNENBOM
AND A. C. WAHL,J. Am. Chem. SOC.,71 (1949) 2589.
H. A. KIERSTEAD,
J . Chem. Phys., 18 (1950) 856.
R. H. BETTS,H. S. A. GILMOUR
AND R. K. LEIGH,J. Am. Chem. SOC.,72 (1950) 4978.
J. SILVERMAN
AND R. W. DODSON,
J. Phys. Chem., 56 (1952) 846.
R. W. DODSON,J . Am. Chem. SOC.,72 (1950) 3315.
R. W. DODSON,
J . Phys. Chem., 56 (1952) 852.
L. EIMER,
A. I. MEDALIA
AND R. W. DODSON,
J. Chem. Phys., 20 (1952) 743.
J. HUDISAND R. W. DODSON,
J. Am. Chem. SOC.,78 (1956) 912.
S. FUKUSHIMA
AND W. L. REYNOLDS,
Talanta, 1 I (1964) 283.

148

I O N S O F SAME METAL I N D I F F E R E N T O X I D A T I O N STATES


11 H. L. REYNOLDS
A N D R. W. LUMRY,
J. Chem. Phys., 23 (1955) 2460.
12 R. A. HORNE,
J . Inorg. Nucl. Chem., 25 (1963) 1139.
13 R. A. HORNE
A N D E. H. AXELROD,
J. Chem. Phys., 40 (1964) 1518.

8.4.2 The effect of inorganic ions on the exchunge between Fe(III) and Fe(II)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

J. SILVERMAN
AND R. W. DODSON,
J. Phys. Chern., 56 (1952) 846.
N. SUTIN,J. K. ROWLEY
AND R. W. DODSON,
J. Phys. Chem., 65 (1961) 1248.
R. J. CAMPION,
J. J. CONOCCHIOLI
AND N. SUTIN,
J. Am. Cheni. Soc., 86 (1964) 4591.
R. A. HORNE,
J . Inorg. Nucl. Chem., 25 ( I 963) 1 139.
J. HUDISA N D A. C. WAHL,J. A m . Chetn. Soc., 75 (1953) 4153.
J. MENASHI,
S. FUKUSHIMA,
C. Foxx AND W. L. REYNOLDS,
Inorg. Chem., 3 (1964) 1242.
G. S. LAURENCE,
Trans. Faraday Soc., 53 (1957) 1326.
T. J. CONOCCHIOLI
AND N. SUTIN,
J. Am. Chem. Soc., 89 (1967) 282.
R. A. HORNE
A N D E. H. AXELROD,
J . Cheni. Phys., 40 (1964) 1518.
D. BUNN,F. S. DAINTON
A N D S. DUCKWORTH,
Trans. Faraday Soc., 57 (1961) 1131.
D. BUNN,F. S. DAINTON
AND S. DUCKWORTH,
Trans. Faraday Soc., 55 (1959) 1267.
J. C. SHEPPARD
A N D L. C. BROWN,
J. Phys. Chern., 67 (1963) 1025.
R. A. HORNE,
Ph. D . Thesis, Columbia (1955).
W. L. REYNOLDS
AND S. FUKUSHIMA,
Inorg. Chern., 2 (1963) 176.
R. L. S. WILLIX,Trans. Faraday Soc., 59 (1963) 1315.
K. BACHMANN
AND K. H. LIESER,
Z. Physik. Chem. N.F., 36 (1963) 3.

8.4.3 The effect of organic ligands on the exchange between Fe(III) und Fe(II)

I R. A. HORNE,J. Phys. Cheni., 64 (1960) 1512.


J. C. SHEPPARD
A N D L. C. BROWN,
J. Phys. Chem., 67 (1963) 1025.
A. MCAULEY
A N D C. H. BRUBAKER,
Inorg. Chem., 3 (1964) 273.
M. R. CHAKRABARTY,
J. F. STEPHENS
A N D E. S. HANRAHAN,
Inorg. Chenz., 5 (1966) 1617.
A. W. ADAMSON
A N D K. S. VORRES,
J. Inorg. Nucl. Chetn., 3 (1956) 206.
W. L. REYNOLDS,
N. LIUAND J. MICKUS,
J. Am. Chem. SOC.,83 (1961) 1078.
L. E I M E ARN D A. I. MEDALIA,
J. Am. Chern. Soc., 74 (1952) 1592.
E. EICHLER
A N D A. C. WAHL,J. A m . Chern. SOC.,80 (1958) 4145.
M. W. DIETRICH
A N D A. C. WAHL,
J. Chem. Phys., 38 (1963) 1591.
D. W. LARSEN
A N D A. C. WAHL,J. Chern. Phys., 41 (1964) 908.
D. W. LARSEN
A N D A. C. WAHL,J. Chetn. Phys., 43 (1964) 3765.
D. R. STRANKS,
Discussions Faraday Soc., 29 (1960) 73.

2
3
4
5
6
7
8
9
10
II
12

8.4.4 The exchange reaction between Fe(II1) and Fe(II) in non-aqueous and mixed soluents

I D. PESCHANSKI,
J. Chim. Phys., 50 (1953) 634.
R. A. HORNE,Exchange Reuctions, I.A.E.A., Vienna, 1965, p. 67.
N. SUTIN,J. Phys. Chem., 64 (1960) 1766.
A. G. MADDOCK,
Trans. Faraday Soc., 55 (1959) 1267.
J. MENASHI,
W. R. REYNOLDS
A N D G. VAN AUKEN,
Inorg. Chern., 4 (1965) 299.
G. WADAAND W. L. REYNOLDS,
Exchange Reactions, I.A.E.A., Vienna, 1965, p. 59.

2
3
4
5
6

8.4.5 The exchange reaction between hexacyanoferrafe ( I I I ) and hexmyanoferrate ( I I )

I
2
3
4
5
6
7

A. C. THOMSON,
J. Am. Chem. Soc., 70 (1948) 1045.
C. HAENNY
A N D E. WICKLER,
Helu. Chim. Acta, 32 (1949) 2444.
C. HAENNY
A N D G. ROCHAT,
Helu. Chim. Acfa, 32 (1949) 2441.
J. W. COPPLEA N D A. W. ADAMSON,
J . Am. Cheni. Soc., 72 (1950) 2276.
A. C. WAHLA N D C. F. DECK,J. Am. Chem. Soc., 76 (1954) 4054.
A. C. WAHL,Z. Elecfrochenr.. 64 (1964) 90.
A. LOEWENSTEIN,
M. SHPORER
A N D G. NAVON,
J . An]. Cheni. SOC.,85 (1963) 2855.

REFERENCES

149

8 M. SHPORER,
G. RON,A. LOEWENSTEIN
A N D G. NAVON,
Inorg. Chem., 4 (1965) 361.
9 P. KING,C. F. DECKA N D A. C. WAHL,Ahstr. Paper 139th Meeting AD?.Chem. Soc.,
Missouri, 1961.
10 R. J. CAMPION,
C. F. DECK,P. KINGA N D A. C. WAHL,Inorg. Chein., 6 (1967) 672.
1 1 R. STASIW
A N D R. G. WILKINS,
Inorg. Chem., 8 (1969) 156.
12 A. LOEWENSTEIN
AND G. RON,Inorg. Chein., 6 (1967) 1604.
8.4.6 Reactions of Fe(Il1) with Fe(I1)

N. SUTINA N D B. GORDON,
J. Am. Chem. Soc., 83 (1961) 70.
M. H. FORD-SMITH
A N D N. SUTIN,
J. Am. Chem. Sor., 83 (1961) 1830.
B. M. GORDON,
L. L. WILLIAMS
A N D N. SUTIN,
J. Am. Chenr. Sue., 83 (1961) 2061.
N. SUTIN,Nufirre, 190 (1961) 438.
5 N. SUTINA N D D. R. CHRISTMAN,
J. Am. Chenr. Soc., 83 (1961) 1773.
6 R. STASIWA N D R. G. WILKINS,
Inorg. Chon., 8 (1969) 156.
A N D R. E. YELIN,
Inorg. Ch~~irr.,
7 (1968) 2667.
7 R. G. WILKINS

1
2
3
4

8.4.7 The Fe(II)-rcrtalysed qiiotion o/Fe(111J

1 R. J. CAMPION,
T. J. CONOCCHIOLI
A N D N . SUTIN,J. Am. Chem. Sue., 86 (1964) 4591.
AND N. SUTIN,
J. Am. Chern. Soe., 89 (1967) 282.
2 T. J. CONOCCHIOLI
3 E. G. MOORHEAD
A N D N. SUTIN,Inorg. Chenr., 6 (1967) 428.
8.4.8 The reaction Detween Fe(IV) and Fe(1I)

1 T. J. CONOCCHIOLI,
E. J. HAMILTON
AND N. SUTIN,
J. A / I IChenr.
.
Sue., 87 (1965) 926.
8.4.9 The exchange reaction of Ru( V I I ) and Ru( V I )
1 E. V. LUOMA
AND C. H. BRUBAKER,
Inorg. Chern., 5 (1966) 1618.
8.4.10 Ru(II)-catn/ysedsubstitution reactions of Ru(II1)
1 J. F. ENDICOTT
A N D H. TAUBE,
J. Am. Chem. Soc., 84 (1962) 4985.

8.4.1 1 The exchange reaction between Os(1II) and Os(1I)


1
2
3
4

F. P. DWYER
AND E. c. GARFAS,
Nature, 166 ( I 950) 48 I .
E. EICHLER
AND A. C. WAHL,
J. A m . Chem. Sor., 80 (1958) 4145.
M. W. DIETRICH
A N D A. C. WAHL,J. Chein. Phys., 38 (1963) 1591.
R. CAMPION,
N. PURDIEA N D N. SUTIN,J. Am. Chenr. Soc., 85 (1963) 3528.

8.5.1 The exchange reaction between C o ( I I I ) and C o ( l 1 ) in aqueous niedia


1
2
3
4
5
6
7
8
9
10
11
12

S. A. HOSHOWSKY,
0. G., HOLMES
AND K. J. MCCALLUM,
Can. J. Res. B , 27 (1949)258.
N. A. BONNER
AND J. P. HUNT,J. Am. Chem. Soc., 74 (1952) 1866.
N. A. BONNER
A N D 5. P. HUNT,J . A m . Chenr. Soc., 82 (1960) 3826.
J. SHANKAR
A N D B. C. DE SOUZA,
J. Inorg. Nucl. Chein., 24 (1962) 187.
J. SHANKAR
AND 3. C. DE SOUZA,
J. Inorg. Nucl. Chem., 24 (1962) 693.
L. H. SUTCLIFFE
AND J. R. WEBER,
Trans. Faraday Soc., 52 (1956) 1225.
H. S. HABIB
AND J. P. HUNT,J. Am. Chern. Soe., 88 (1966) 1668.
J. SHANKAR
A N D B. C. DE SOUZA,
J. Inorg. Nrtcl. Chem., 29 (1967) 1983.
T. J. CONOCCHIOLI,
G. H. NANCOLLAS
AND N. SUTIN,Inorg. Chenr., 5 (1966) 1.
P. G. RASMUSSEN
AND C. H. BRUBAKER,Inorg. Chem., 3 (1964) 977.
J. P. BIRKAND J. HALPERN,
J . Am. Chern. Soc., 90 (1968) 305.
A. W. ADAMSON,
J. Am. Chem. Soc., 73 (1951) 5710.

150

IONS O F SAME METAL I N DIFFERENT OXIDATION STATES

8.5.2 Exchange reactions involving complexes of Co(II1) and Co(II) with ammonia and organic
ligands
1 J. H. FLAGG,
J. Am. Chem. SOC.,63 (1941) 557.

2
3
4
5
6
7
8

K. J. MCCALLUM
AND S. A. HOSHOWSKY,
J. Chern. Phys., 16 (1948) 254.
S. A. HOSHOWSKY,
0. G. HOLMES
AND K. J. MCCALLUM,
Can. J. Res. B, 27 (1949) 258.
w. B. LEWIS,c. D. CORYELL
AND J. w. IRVINE, J. Chem. SOC.S, (1949) 386.
D. R. STRANKS,
Discussions Furuday SOC.,29 (1960) 73, 131.
E. APPLEMAN,
M. ANBARAND H. TAIJBE,J. Phys. Chem., 63 (1959) 126.
F. P. DWYER
AND A. M. SARGESON,
J. Phys. Chem., 65 (1961) 1892.
D. R. STRANKS,
Advances in Chemistry of Coordination Compounds, Macmillan, New

York, 1961.
9 B. WEST,J.Chem. Soc., (1952) 3115.
10 A. W. ADAMSON
AND K. S. VORRES,
J. Inorg. Nucl. Chem., 3 (1956) 206.
11 Y . A. 1 M AND D. H. Busca, J. Am. Chem. Soc., 83 (1961) 3357.
12 Y.A. 1M A N D D. H. BUSCH,J. Am. Chern. SOC.,83 (1961) 3362.
F. BASOLO
AND H. M. NEUMANN,
J. Phys. Chem., 63 (1959) 371.
13 B. R. BAKER,
AND M. J. G. WILLIAMS,
J. Chem. SOC.,(1957) 4456.
14 P. ELLIS,R. G. WILKINS
15 N. S. BIRADAR,
D. R. STRANKS
AND M. S. VAIDYA,
Trans. Faraday SOC.,58 (1962) 2421.
A N D J. P. HUNT,J. Am. Chem. Soc., 90 (1968) 7210.
16 T. J. WILLIAMS
17 S. BRUCKNER,
V. CRESCENZI
AND F. QUADRIFOGLIO,
J. Cheni. SOC.,A (1970) 1168.
8.5.3 The exchange reaction between Co(III) and Co(I1) in non-aqueous media
1 J. J. GROSSMAN
A N D C. S. GARNER,
J. Chem. Phys., 28 (1958) 268.
2 B. WEST,J. Chem. SOC.,
(1952) 3115.
3 B. R. BAKER,
F. BASOLO
AND H. M. NEUMANN,
J. Phys. Chem., 63 (1959) 371.

8.5.4 Co(I1)-catalysedsubstitution reactions of Co(II1)

HALPERN
AND S. NAKAMURA,
J. Am. Chern. SOC.,
85 (1963) 2517.
A. W. ADAMSON,
J. Am. Chern. SOC.,78 (1956) 4260.
J. HALPERN
AND S. NAKAMURA,
J . Am. Chern. SOC.,87 (1965) 3002.
J. P. BIRKAND J. HALPERN,
J. Am. Chern. SOC.,90 (1968) 305.
R. FARINA
A N D R. G. WILKINS,
Inorg. Chem., 7 (1968) 514.

1 J. P. CANDLIN,
J.

2
3
4
5

8.5.5 The reaction of C o ( I V ) with C o ( I I )


1 D. BENSON,P. J. PROLL,J. WALKLEY
A N D L. H. SUTCLIFFE,
Discussions Faraday SOC.,

29 (1960) 60.
2 P. J. PROLL,Ph. D. Thesis, Liverpool University (1962).
8.5.6 Rh(I)-catalysed substitution reactions of Rh(III)
1 J. V. RUND,F. BASOLO
AND R. G. PEARSON,
Inorg. Chem., 3 (1964) 659.
J. A. OSBORNEAND
G. WILKINSON,
J. Chem. SOC.,(1965) 1951.
2 R. D. GILLARD,
3 R. D. GILLARD,
J. A. OSBORNE
A N D G. WILKINSON,
J. Chem. Soc., (1965) 4107.

8.5.7 The exchange reaction between I r ( l V ) and Ir(III)


1 E. N. SLOTHA N D C. S. GARNER,
J. Am. Chem. SOC.,77 (1955) 1440.
2 P. HURWITZAND K. KUSTIN,Trans. Faraday SOC.,62 (1966) 427.
3 P. HIJRWITZ
A N D K. KUSTIN,
Inorg. Chem., 3 (1963) 823.
8.5.8 The reaction between complexes of Ir(f V ) and I r ( I I f )
1 P. HURWITZ
AND K. KUSTIN,Trans. Faraday Soc., 62 (1966) 427.

REFERENCES

151

8.6. I The exchnnge reaction between P t ( I V ) a n d P t ( I I ) ;Pt(II)-catalysed substitution reactions of

Pt(IV)
1 R. L. RICHAND H. TAUBE,
J. Am, Chem. Soc., 76 (1954) 2608.
D. S. MARTINAND L. T. Cox, J. Inorg. Nucl. Cheni., 7 (1958) 113.
2 R. E. MCCARLEY,
P. H. WILKS,R. G. PEARSON
AND R. G. WILKINS,
J. Inorg. Nucl. Chem.,
3 F. BASOLO,
6 (1958) 161.
A. F. MESSING,P. H. WILKS,R. G. WILKINS
A N D R. G. PEARSON,
J. Inorg.
4 F. BASOLO,
Nucl. Chern., 8 (1958) 201.
5 L. T. Cox, S. B. COLLINS
AND D. S. MARTIN,
J. Inorg. Nucl. Chem., 17 (1961) 383.
6 F. BASOLO,
M. L. MORRISA N D R. G. PEARSON,
Discussions Faraduy Soc., 29 (1960) 80.
AND F. BASOLO,
J. Inorg. Nucl. Chem., 13 (1960) 36.
7 R. C. JOHNSON
8 H . R. ELLISON,
F. BASOLO
AND R. G. PEARSON,
J. Am. Chem. Soc., 83 (1961) 3943.
9 R. R. RETTEW
AND R. C. JOHNSON,
Inorg. Chem., 4 (1965) 1565.
10 R. C. JOHNSON
AND E. R. BERGER,
Inorg. Chem., 4 (1965) 1262.
Inorg. Chem., 4 (1965) 1258.
11 W. R. MASONA N D R. C. JOHNSON,
AND R. ETTORE,
J. Chem. Soc. A, (1968) 2253.
12 A. PELOSO
13 W. R. MASON,Inorg. Chem., 8 (1969) 1756.
14 A. SYAMAL
AND R. C. JOHNSON,
Inorg. Chem., 9 (1970) 265.
AND R. C. JOHNSON,
Inorg. Chem., 8 (1969) 2596.
15 S. G. BAILEY
AND E. R. BERGER,
Inorg. Chem., 7 (1968) 1656.
16 R. C. JOHNSON

9.1 The exchange reaction between C e ( I V ) and Ce(III)


1
2
3
4
5
6
7
8

V. J. LINNENBOM
AND A. C. WAHL,J. Am. Chem. Soc., 71 (1949) 2589.
J. W. GRYDER
AND R. W. DODSON,
J. Am. Chem. Soc., 71 (1949) 1894.
J. W. GRYDER
AND R. W. DODSON,
J. Am. Cheni. Soc., 73 (1951) 2890.
F. R. PARCHEN
AND F. R. DUKE,
J. Am. Chem. Soc., 78 (1956) 1540.
S. FRONAEUS
A N D C. 0. OSTMAN,
Acta Chem. Scond., 10 (1956) 769.
P. B. SIGLER
AND B. J. MASTERS,
J. Am. Chem. Soc., 79 (1957) 6353.
G. E. CHALLENGER
A N D B. J. MASTERS,
J. Am. Chem. Soc., 77 (1955) 1063.
H. C. HORNIGAND W. F. LIBBY,J. Phys. Chem., 56 (1952) 869.

9.2 The exchange reaction between Eir(III) and Eu(II)


1 D. J. MEIERAND C. S. GARNER,
J. Am. Chem. Soc., 73 (1951) 1894.
2 D. J. MEIERA N D C. S. GARNER,
J. Phys. Chem., 56 (1952) 853.

10. I Exchange reactions between uranium ions

I R. H. BETTS,Con. J. Res., 26B (1948) 702.


E. RONA,J. Am. Chem. Soc., 72 (1950) 4339.
B. J. MASTERSA N D L. L. SCHWARTZ,
J. Am. Chem. SOC.,83 (1961) 2620.
D. M. MATHEWS,
J. D. HEFLEY
AND E. S. AMIS,J. Phys. Chem., 63 (1959) 1236.
A. ~ N D E L L AND
I
E. S. AMIS,J. Am. Chem. Soc., 81 (1959) 4180.
S. L. MILTON,A. ~ N D E L L IAND E. S. AMIS,J. Inorg. Nucl. Chem., 17 (1961) 325.
S. L. MILTON,J. 0. WEARA N D E. S. AMIS,J. Inorg. Nucl. Chem., 17 (1961) 317.
G. GORDON
AND H. TAUBE,
J. Inorg. Nucl. Chem., 16 (1961) 272.

2
3
4
5
6
7
8

10.2 Reactions between uranium ions


1 D. M. H. KERNAND E. F. ORLEMAN,
J. Am. Chem. Soc., 71 (1949) 2102.

2
3
4
5

D. M. H. KERNAND E. F. ORLEMAN,
J. Am. Chem. Soc., 75 (1953) 3059.
K. A. KRAUS,F. NELSON
A N D G. L. JOHNSON,
J. Am. Chem. Soc., 71 (1949) 2510.
F. R. DUKEAND R. C. PINKERTON,
J. A m . Chem. Soc., 73 (1951) 2361.
J. KORYTA
AND J. KOUTECKY,
CON.Czech. Chem. Cornmun., 20 (1955) 423.

152

I O N S O F S A M E METAL IN D I F F E R E N T O X I D A T I O N STATES
6
7
8
9
10
11
12
13

H. IMAI,Biill. Chem. SOC.Japan, 30 (1957) 873.


D. T. PENCEA N D G. L. BOOMAN,
Anal. Chem., 38 (1966) 11 12.
H. G. HEALAND J. G. N . THOMAS,
Trans. Faraday SOC.,45 (1949) 11.
0. FISHER
A N D 0. DRAKA,
CON.Czech. Chem. Commun., 24 (1959) 3046.
L. J. HEIDTA N D K. A. MOON,J. Am. Cliem. Soc., 75 (1953) 5803.
L. J. HEIDT,J. Am. Chern. SOC.,76 (1954) 5962.
T. W. NEWTON
A N D F. B. BAKER,
Inorg. Chenz., 4 (1965) 1166.
B. J. MASTERS
AND L. L. SCHWARTZ,
J. Am. Chem. Soc., 83 (1961) 2620

10.3 Exchange reartions between neptuniuni ions

1 D. COHEN,
J. C. SULLIVAN
A N D J. C. HINDMAN,
J. Am. Chem. Soe., 76 (1954) 352.
J. C. SULLIVAN
A N D J. C. HINDMAN,
J. Am. Chem. SOC.,77 (1955) 4964.
2 D. COHEN,
3 J . C. SULLIVAN,
D. COHEN
AND J. C. HINDMAN,
J. Am. Chem. Soe., 79 (1957) 3672.
4 D. COHEN,
J. C. SULLIVAN,
E. S. AMISAND J. C. HINDMAN,
J. A m . Chem. SOC.,78 (1956)
1543.
5 J. C. SULLIVAN,
D. COHENA N D J. C. HINDMAN,
J. Am. Cheni. Soc., 76 (1954) 4275.
10.4 Reuctions between neptunium ions
1
2
3
4
5

J. C. HINDMAN,
J. C . SULLIVAN
A N D D. COHEN,
J. Am. Chem. Soc., 80 (1958) 1812.
J. C. HINDMAN,
J. C . SULLIVAN
AND D. COHEN,
J. Am. Chem. SOC.,76 (1954) 3278.
J. C. HINDMAN,
J. C . SULLIVAN
A N D D. COHEN,
J. Am. Chem. Soc., 81 (1959) 2316.
J. C. SULLIVAN,
D. COHEN
AND J. C. HINDMAN,
J. Am. Chem. Soc., 79 (1957) 402Y.
D. COHEN, E. S. AMIS,J. C. SULLIVAN
A N D J. C. HINDMAN,
J. Phjs. chem., 60 (1956) 701.

10.5 Exchtrnye reactions between plutonium ions

1 T. K. KEENAN,
J . Am. Chem. Soc., 78 (1956) 2339.
2 T. K. KEENAN,J. Phys. Chern., 61 (1957) 1 117.
10.6 Recictions between plirtonirim ions
1 0. E. OCARDA N D S. W. RABIDEAU,
J. Pliys. Chem., 60 (1956) 812.
2 s. W. RABIDEAU
A N D R. J. KLINE,J . Phys. Cheni., 62 (1958) 617.
3 M. ALEXI,Q. C. JOHNSON,
H. D. COWANA N D J. F. LEMONS,
J. Inorg. Niicl. Chem., 29
(1967) 2327.
J. Am. Chem. Soe., 71 (1949) 1528.
4 R. E. CONNICK,
5 S. W. RABIDEAU,
J. Am. Chem. SOC.,79 (1957) 6350.
6 S. W. RABIDEAU,
J. Am. Chem. Soc., 75 (1953) 798.
7 R. E. CONNICK
A N D W. H. MCVEY,
J. Am. Chem. Soc., 75 (1952) 474.
8 S. W. RABIDEAU
A N D H. D. COWAN,
J. Am. Chem. SOC.,77 (1955) 6145.
10.7 Exchiinye reactions bet ween americiiim ions
1 T. K. KEENAN,
R. A. PENNEMAN
A N D J. F.

SU1 TLE,

J . Phys. C ~ ~ I I59I .(1955)


,
381.

10.8 Reactions between americium ions


1
2
3
4

G. R. HALLAND P. D. HERNIMAN,
J. Chem. Soe., (1954) 2214.
G. R. HALLA N D T. L. MARKIN,
J . Inorg. Nircl. Chem., 4 (1957) 296.
S. R. G U N NA N D B. B. CUNNINGHAM,
J. Am. Chem. Soc., 79 (1957) 1563.
R. A. PENNEMAN
A N D L. B. ASPREY,
Intern. Con5 Peaceful Uses Atomic Energ), Vol. 7,

1955, p. 355.
5 A. A. ZAITSEV,
V. N. KOSYAKOV,
A. G . RYKOV,Yu. P. SOBELEV
A N D G . N. YAKOVLEV,
Radioklziniiyci, 2 (1960) 339.
6 J. S. COLEMAN,
Inorg. Chem., 2 (1963) 53.
7 R. A. PENNEMAN,
J. S. COLEMAN
AND T. K. KFENAN.J . Inorg. Nircl. Chem., 17 (1961 ) 138.

Chapter 3

Oxidation-Reduction Reactions
Between Complexes of Different Metals
D. B E N S O N

1. Introduction
The last two decades have seen a growing interest in the mechanism of inorganic reactions in solution. Nowhere is this activity more evident than in the
topic covered by this review: the oxidation-reduction processes of metal complexes. This subject has been reviewed a number of times previously, notably by
Taube' (1959), Halpern' (1961), Sutin3 (1966), and Sykes4 (1967). Other articles
and books concerned, wholly or partly, with the topic include those by Stranks',
Fraser6, Strehlow7, Reynolds and Lumry', Basolo and Pearson', and Candlin
et af.'Ot. Important recent articles on the theoretical aspects are those by Marcus"
and Ruff 12. Elementary accounts of redox reactions are included in the books by
Edwards13, SykesI4 and Benson". The object of the present review is to provide a
more detailed survey of the experimental work than has hitherto been available.
The material included is organised according to the periodic table as follows.
Classification, in the first place, is on the basis of oxidants, arranged in order of
their position across the periodic table, i.e., from vanadium to lead. The reactions
of the lanthanide cerium(1V) and the actinides are treated last. Within each section
the order for each oxidant is of decreasing oxidation number. In general, reductants
within each section and sub-section are arranged, again according to the periodic
table, but in order of increasing oxidation number. However, in sections (6.1)
and (6.2), dealing with oxidations by cobalt(III), the subject matter is such that
no classification on the basis of reductant has been attempted.
In terms of gross features of mechanism, a redox reaction between transition
metal complexes, having adjacent stable oxidation states, generally takes place in
a simple one-equivalent change. For the post-transition and actinide elements,
where there is usually a difference of two between the stable oxidation states, both
single two-equivalent and consecutive one-equivalent changes are possible.
AS regards intimate mechanism, electron transfer reactions of metal complexes
are of two basic types. These have become known as outer-sphere and innersphere (see Chapter 4, Volume 2). In principle, an outer-sphere process occurs
with substitution-inert reactants whose coordination shells remain intact in
t See, also, S ~ t i n ~ ' ~ .
References pp. 267-273

154

C O M P L E X E S O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

forming the activated complex. In other words, electron transfer proceeds more
rapidly than substitution in the coordination shell of one of the reactants. Conversely, in inner-sphere processes substitution takes place prior to electron
transfer. The direct and indirect criteria used in distinguishing between inner- and
outer-sphere mechanisms have been discussed in detail by Sutin3.

2. Oxidations by vanadium
2.1

OXIDATIONS BY VANADIUM(V)

The oxidation by vanadium(V) of iron(II), a reaction in which a metal-oxygen


bond is broken, takes place according to the stoichiometric equation

V 0 2 + Fe( 11) + 2 H +

V 0 2 + + Fe(III)+ H 2 0

Oxidation potentials lead to a value of 7 . 9 ~


lo3 for the equilibrium constant.
Kinetic data for the reaction (from 0 to 55.6 " C ) in acid perchlorate solutions
(over the range 0.047-1.0 M) have been obtained spectrophotometrically by
following the disappearance of V(V) (which absorbs strongly between 305 and
350 mp) as a function of timeI6. The second-order nature of the rate law
-d[V(V)]/dt

-d[Fe(II)]/dt

k'[V(V)][Fe(II)]

is rigidly adhered to over a wide range of reactant concentrations. The observed


rate coefficient, k', shows a dependence on [H'], viz.

k' = u / [ H+ ] + b + c [ H + ]
The three terms of this expression indicate three competitive activation processes,
respectively
V 0 2 + + F e 2 + +H 2 0 +
V 0 2 + +Fe2+

_f

(V02FeOH2+)'+H+
(V02Fe3+)*

V 0 2 ++ Fe2++ H + +(HV02Fe4+)*
assuming inner-sphere structures for convenience. The term in a is barely significant, and the c term ytrongly predominates over the b term. At 25 "C and
p = 1 M the value of c [the rate coefficient for the most important path (2.3)]
is 3400 12.mole-2.sec-'. Under the same conditions a value of 60 I.mole-'.sec-'
can be assigned tentatively to b. The activation parameters corresponding to

155

OXIDATIONS BY V A N A D I U M

step (2.3), AH' and AS', are 1.52 kcal.mole-' and -37.3 cal.deg-'.mole-',
respectively. It is noteworthy that the following reduction reactions of the analogous MOz+ actinide ions show a similar first-order hydrogen-ion dependence:
( N p 0 2 ++Fe2+)17, ( N p 0 2 + + N p 3 + ) ' * , ( P U O , + + P U O ~ ~ )and
' ~ , ( UO z + +
U 0 2 + ) 2 0 .In Table 1 the activation parameters of the V(V)+Fe(II) reaction are
TABLE 1
ACTIVATION PARAMETERS FOR REACTIONS H A V I N G

AH '(kcal.

Net activation process

mole-')

+4

A C T I V A T E D COMPLEXES"

AS' (cal.deg rnole- I )

st

Ref.

corpku
-~

VO,+ + F e 2 + + H + + (HV02Fe4+)I
NpO,+ + F e z + + H + + (HNpO2Fe4+)'
Fez --Fe3
H zO (FeOH Fe4 )' 4-H
Cr2+T C ~ +~H+z O + (CrOHCr4+)'+H+
Fez+ L C o 3 + + H z 0 + (FeOHCo4+)X+H+
V 3 + + V O Z + + H z 0+ ( V O H V 0 4 + ) ' + H +
v3 'voZ + ( v o 2 v 4 +):
+

1.52
8.6
19.4
22
18.8
20. I
16.6

-37.3&0.6
-38
9.9*1.4
- 2 +5
+I6 ~ t 4
- 3 k5
5 1 6

-7Ofl
-69

-70+1
-79f5
-66*4
-77+5
-6516

16
17
22
23
24
25
26

From Daugherty and NewtonL6.


S i complex is the formal ionic entropy of the activated complex: SfCOmplCi
=AS$+Z1Sorcactantrz1.
Some values used for SoreacLan,s
are estimated ones.
a

contrasted with those of other redox reactions for which the net charge on the
activated complex is +4. The V(V) + Fe(I1) reaction is remarkable in having a
very low enthalpy of activation, considerably smaller than AH' for the Np(V)+
Fe(I1) system (which also involves the breaking of a metal-oxygen bond), and
rather surprisingly, lower than AH' for the exchange between Fe(CN)63- and
Fe(CN),4-. The AS' values in Table 1 are seen to vary widely from -38 to 16
cal.deg- '.mole-'. However, the (formal) ionic entropies (S'complex)
of the + 4
activated complexes are similar (-65 to -79 cal.deg-'.mole-') and comparable
to those for reaction between two actinide ions". Thus it is apparent that the
prime factor in deciding the entropy of these types of activated complexes is the
charge and not the size of the complex.
The reduction of V(V) to V(1V) by Fe(I1) has been also studied by Nicol and
Rosseinskyz7.z8. These authors used a polarographic method employing a
rotating platinum electrode-calomel electrode pair. Under certain conditions of
applied voltage, the diffusion current depends only upon the concentration of
Fe(1I). Nicol and Rosseinsky conclude, after studying the reaction over wider
ranges of conditions than those of Daugherty and Newton16, that the complex
dependence on hydrogen-ion concentration is better described by

k' = ~ + c [ H + ] + ~ [ H + ] ~

They attribute this to the existence of three reaction paths, given by eqns. (2.2)
References pp. 266273

156

COMPLEXES OF DIFFERENT METALS: OXIDATION-REDUCTION

and (2.3) together with


V O 2 + + F e 2 + + 2H + --+
(H2V02FeS+)*
At 25 "C and p = 3.0 M , b, c and d values are quoted of 203 I.mole-'.sec-',
6730 12.mole-Z.sec-', and 1830 13.mole-3.sec-', respectively.
When an excess of vanadium(V) is reacted with tin(l1) in dilute hydrochloric
acid media, V(1V) is produced in amounts equivalent to Sn(1I) oxidised, and no
V(II1) is obtained29t. When vanadium(V) is reduced by an excess of tin(1I) in
the same media, both V(IV) and V(1II) are formed as products: the ratio of V(V)
reduced to V(II1) formed is constant except when the ratio [V(V)],/[Sn(II)], is
close to unity. V(1V) and V(II1) were determined spectrophotometrically at
755 mp and 400 mp, respectively. There is considerable evidence against the scheme
V(IV)+Sn(IIl)

(2.4)

V(V)+Sn(IIl) -+ V(IV)+Sn(IV)

(2.5)

V(V)+Sn(II)

V(IV)+Sn(III)
V(V)+V(III)

-+

--f

V(III)+Sn(W)

(2.6)

2 V(N)

(2.7)

in which the reactant species undergo one-equivalent changes. Tnstead, an alternative scheme, made up of a primary two-equivalent change
V(V)+ Sn(1I)

V(111) Sn(IV)

(2.8)

coupled with steps (2.4), (2.5) and (2.7), is more in accord with the experimental
results on stoichiometry. From a detailed analysis of the kinetic implications of
this reaction sequencc it is concluded that about 90 % of effective encounters
between V(V) and Sn(1l) result in the direct formation of V(I1I)and Sn(IV), whilst
the remaining 10 7; yield V(1V) and Sn(II1). It is conceivable that the reaction
involves an association complex represented as (V+ Sn)"" which, in addition
to dccaying to V([Il)+Sn(lV), or V(IV)+Sn(III) (steps (2.8) and (2.4), respectively), can persist for a sufficient time in solution to react further with V(V) to
generate V(1V) and Sn(IV), viz.

Evidence has been cited for a similar 1 : 1 complex formed between Sn(I1) and
U(V1) in hydrochloric acid3'. Path (2.9) is expected to become morc important
at relatively high initial concentrations of V(V). For conditions in which the
t Although the rate was found to be too great to measure in hydrochloric acid media, the

V(V) i-Sn(l1) reaction has been investigated more recently in perchloric acid2".

O X I D A T I O N S BY V A N A D I U M

I57

initial concentration of V(V) is in excess of Sn(II), the principal mechanism for


the generation of V(IV) is step (2.8) in combination with step (2.7).

2.2

OXIDATIONS BY

VANADIUM(IV)

A V(III)-Cr(III) dimer has been detected as an intermediate in the V(IV)+


Cr(I1) reaction3 and its formation and subsequent decomposition studied32.
Since V(IV) was arranged to be in excess of Cr2+, no complications ensued from
the latter ion bringing about further reduction of V(1II) to V(I1). Under these
conditions the overall reaction corresponds accurately to

'

Kinetic data were obtained over wavelengths ranging from 260 to 760 mp. The
presence of an intermediate is readily apparent: on mixing pale-blue solutions
of V(1V) and Cr(I1) a bright green colour develops immediately, and then fades
slowly to the blue-purple colour characteristic of V3+ and Cr3+.From the response
of the rate to variations in acidity, it is deduced that the reaction proceeds mainly
via a rapid "direct" route and that only a small portion occurs via the intenselycoloured intermediate at a measurable rate. It is clear that the direct path cannot
be a two-equivalent change, involving V(11) and Cr(IV) as intermediates, since the
product Cr(II1) species is the aquo ion, and not the dimer which would be produced
from the interaction of Cr(IV) and Cr(II)33. Also, as the dimeric species VOV4+
(formed from V(IV) +V(II)34) is not detectable at 425 mp, appreciable concentrations of V(I1) cannot be present. The rate of disappearance of the intermediate
{which follows first-order kinetics) increases as the hydrogen-ion concentration
increases although the dependence is complex: over ranges of 0.010 to 0.99 M
HC104 and 5" to 25 "C at p = 1.00 M the following expression applies
k' = ( A K + B[H+])/(K+ [H'])

(2. lo)

where A and B are constants, and K is the equilibrium constant for equilibrium
between acid and base forms of the intermediate. Formulating the intermediate
as VO(OH),,Cr("-")+ and the equilibrium as

then a comparable expression to (2.10), viz.


k'

(koK+kl[H+])/(K+ [I+])

References pp. 267-273

158

COMPLEXES O F DIFFERENT METALS: OXIDATION-REDUCTION

can be obtained by assuming that the intermediate decays by the two competitive
processes

VO(OH),+ ,Cd3-")+

products

VO(OH),Cr(4-")+ (and/or VO(OH),+ ' C T ( ~ - " )++H + ) --%products


On this basis k, = 0.0170 sec-', k, = 0.645 sec-l, and K = 0.739 mole.l-' at
25 "C. The corresponding activation parameters were determined also by Esp e n ~ o n By
~ ~ a. method involving extrapolation of the first-order rate plots at
various wavelengths to zero time, the absorption spectrum of the intermediate
was revealed (Fig. 1). Furthermore, the value of K obtained from the kinetics
was compatible with that derived from measurements on the acid dependence of
the spectrum of the intermediate. Rate data for a number of binuclear inter~ ~ there to be a correlation
mediates are collected in Table 2. E ~ p e n s o nshows
between the rate of decomposition of the dimer and the substitution lability of the
more labile metal ion component. The latter is assessed in terms of the rate of
substitution of SCN- in the hydration sphere of the more labile hydrated metal
ion.
The one-equivalent oxidation of Eu(I1) by V(IV) takes place rapidly: at 0 "C,
0.2 M [H'] and
M reactant concentrations, the second-order rate coeffilo3 I.mole-'.sec-'
(2.64 x lo3 l.mole-'.sec-l at 25 "C and p =
cientj' is
1.0 M 2 7 6 ) .

- -

Wavelength (mp)

Fig. 1 . Absorption spectrum of V(II1)-Cr(II1) dimer formed as an intermediate in the


V(1V) fCr(I1) reaction (the ordinate is directly proportional to the absorbance); 0.200 M HC104;
temp., 15.0 "C. (From E ~ p e n s o nby
~ ~courtesy
,
of The American Chemical Society.)

159

OXIDATIONS B Y V A N A D I U M

TABLE 2
DECOMPOSITION RATES OF BINUCLEAR METAL COMPLEXES'

Mode of formation

Binuclear species

Specific decomposition
rate at 25.0 "C (sec-')

log k for formation


of S C N - complex

Ref.

Cr(IV) +Cr(ll)
and
Cr(III)+OH-

CrOHCrS

1.5 x I O - ~

Cr(IlI),

-5.7

33b, 36

Np(V1) +Cr(lII)
and
N p (V) Cr (11)

Np02Cr4+

2.3 x

Cr(III),

-5.7

37

V(IV)+V(II)
and
V(II1) + O H -

vov4+

0.03 +1.54[H+]

V(III),

t1.8

34

VO(OH).Cr(4-n)+

0.017+0.645[Ht]

V(III),

+1.8

32

0.35 + 3 . 5 [ H + ]

Fe(III),

+2.1

38

V(1V) +Cr(II)
Fe(IV)+Fe(II)
and
Fe(III)+OHa

FeOFe4+

From Espenson3*.

From ref. 35 (k in I.mole-'.sec-').

2.3

OXIDATIONS BY VANADIUM(III)

The kinetics of the oxidation of chromium(I1) by vanadium(II1) in acid


perchlorate media have been studied spectrophotometrically between 0.2" and
35.0 "C over a range of 0.027-0.500 A4 HC10440. The oxygen-sensitivity of both
reactants meant that the air had to be excluded in all kinetic runs. Also, since
V(TI1) slowly reduces perchlorate ion, fresh solutions of V(II1) were required for
each experiment. In terms of stoichiometry the reaction conforms accurately to

and the rate law is given by

- d[V3+]/dt = k'[V3+ ] [ C r Z f ]
over wide ranges of wavelengths and reactant concentrations. No direct evidence
could be adduced for the existence of intermediates. The rate response to variations
in acidity demonstrates that the observed rate coefficient, k', can be considered in
terms of two empirical parameters q and r

k' = q / ( r + [H'])
References p p . 266-273

(2.11)

160

COMPLEXES O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

To account for the inverse dependence of k' on [H'], Espenson4' has proposed
the following general mechanism
V 3 + + C r 2 + + nH 2 0

+ V(OH),Cr(5-")++n H +
ki

(2.12)

k-i

V(OH),Cr(5-")++(n-lj H + 3 (e.g., V 2 + + C r O H 2 + )
(CrOH2++Hf

(2.13)

C r 3 + + H 2 0 ,rapid equilibrium)

(2.14)

It will be noted that conversion of the intermediate V(OH),Cr(S-")+ to products


involves a different number of H + ions than its conversion back to reactants.
It is considered likely that the binuclear intermediate has an inner-sphere structure. On applying the steady-state approximation to the concentration of this
intermediate, it follows that

Comparison of equations (2.11) and (2.15) reveals q and r to be k,k2/k-1 and


k 2 / k - , , respectively. This enables k, to be calculated from q/r. In its simplest
forms the structure of the reactive intermediate can be viewed as V(OH)Cr4+
(when II is 1 ) or as VOCr3+ (when n is 2). Similar species which have been characterized or implied kinetically are CrOCr4+ (ref. 33), Np02Cr4+ (ref. 37), U02Cr 4+
(ref. 31), VOV4+ (ref. 34), U O P U O ~ ~(ref.
' 41), Pu02Fe4+(ref. 42) and FeOFe4+
(ref. 38). Predictions on the rate of the V(III)+Cr(lI) system, based upon Marcus
theory43. have been made by Dulz and S ~ t i on
n ~the~ assumption that an outersphere process applies. The value arrived at by these authors is
60 times lower
than the experimcntal value.
Hai11-1~~
has commented on the ambiguities inherent in the interpretation of
rate laws, using the V(III)+Cr(lI) reaction as an example. The experimental rate
law found by Espenson4'

rate

q[v3+][cr2+]
-.
r

+ [H']

has the limiting forms


rate = q[V3+][Cr2+]/rat low [ H + ]
and
rate

q[V3+][Cr2+]/[H+]at high [H',

It follows that the reaction proceeds via two consecutive activated complexes:

OXIDATIONS BY VANADIUM

161

(VCr")'
and (V(OH)Cr4+)'. In Espenson's mechanism, equations (2.12) to
(2.14), (VCr")' is formed directly from the interaction of V 3 + + C r 2 + ,an event
which is followed by the spontaneous aquation of (V(OH)Cr4+)t. However, as
Haim points out, (V(OH)Cr4+)', formed from VOH" +Cr2+, may precede
(VCr5+)*,formed from V(OH)Cr4++ H + , as shown by

+ VOH2++ H + rapid equilibrium, K


VOH2++Cr2+ 2 V(OH)Cr4+
V3++ H 2 0

(2.16)
(2.17)

k-3

V(OH)Cr4+ + H + -%= VZ++Cr3++H,O

(2.18)

On this mechanism
rate

k , k4 K[V3+][Cr2+]
k-3 +k4[H+]

whence k 3 = 3.12 x lo2 I.mole-'.sec-' and k - 3 / k 4 = 0.108 mole.1-' (at 25 "C),


as compared with k , = 5.76 I.mole-'.sec-' and k 2 / k - , = 0.108 mole.1-'. The
experimental rate Iaw, whilst defining the compositions of the activated complexes,
is incapable of defining the sequence in which they are formed.
Adin and S y k e ~have
~ ~ re-examined the hydrogen-ion dependence of the
V(III)+Cr(lI) system over a broad range of H + concentrations from 0.45 M
down to 0.016 M. They confirm the type of acid dependence quoted by Espenson
and support the interpretation given by Haim, equations (2.16) to (2.18)+. At
25 "C and p = 0.5 M , the experimental parameters q and r are 0.50 and 0.10,
respectively, so that k , = 3 . 5 7 ~lo2 I.mole-'.sec-', and k - , / k 4 = 0.1 rnole.1-'.
Adin and S y k e find
~ ~ no
~ evidence for the existence of the more complex hydrogenion dependence originally suggested by Sykes4' in a re-analysis of Espenson's
data4'.
The reduction of V(II1) by Eu(I1)
V(III)+ Eu(I1) = V(II)+ Eu(II1)
has been examined in perchlorate media4'. Contrary to an earlier
Eu(I1) is not oxidised by perchlorate ions. A spectrophotometric method was
employed in the kinetic work, the formation of V(I1) being monitored at its peak
of 850 mp. Oxygen was rigorously excluded in all kinetic runs. The rate law is
In a later paper, Espenson and Parker277, reporting on an extensive study of the V(III)+
Cr(I1) system in chloride and perchlorate media, favour Haim's mechanism on grounds of
reactivity patterns.
References p p . 266-273

162

COMPLEXES OF DIFFERENT

M E T A L SO :X I D A T I O N - R E D U C T I O N

which is indicative of the steps


V3'+Eu2+

''
ks

V O H ~ ++ E ~ -f
~ +

(2.19)
products
(2.20)

At 25 "C, k , is 9 . 0 l~o w 3I.mole-'.sec-' and k , = k'/K = 2.0 I.mole-'.sec-'


( K is the hydrolysis constant of V3+). Steps (2.19) and (2.20) have associated
A H f values of 11.4 and 6.2 kcal.mole-', and AS' values of -30.1 k 5 and
-35t-5 cal.deg-'.mole-', respectively. The much slower rate of the V(IIl)+
Eu(l1) reaction, as compared to the V(III)+Cr([I) system, is ascribed by Adin
and Sykes to the relative difficulty in transferring electrons from f-orbitals than
from d-orbitals. The sensitivity of the Eu(I1) reaction to chloride ions is noted
but not reported in detail.

3. Oxidations by chromium and molybdenum


3.1

OXIDATIONS BY

CHROMIUM(VI)

The stoichiometric equation for oxidation of vanadium(1V) by chromium(V1)


i n acid perchlorate solutions is essentially
HCr04-+3 VO2++Hf

Cr3'+3V02++H20

That HCr04-, V 0 2 + and V 0 2 + are the predominant species under the conditions
of Espenson's kinetic study5' originates from evidence cited by Tong and King5',
No binuclear species are detectable
and Rossotti and R o ~ s o t t i respectively.
~~,
in the Cr(lI1) product. Vanadium(V) retards the reaction and the full form of the
rate law is

lo4 13.mole-3.sec-' at 25 "C


where k' = 0,563 I.mole-'.sec-' and k" = 5 . 4 ~
and p = 1 M . The first term of the rate law predominates at low concentrations
of Cr(V1) ( 5 4 x lo-' M ) and at low acidities ( 5 0.03 M ) . At higher concentraM2)
tions of Cr(V1) and/or H + (such that [HCrO,-][H+] exceeds 2 x
the second term becomes important. Restricting attention to the first term, the
form suggests a transition state of (HCr04V2++n H20)' with an average oxida-

OXIDATIONS BY

Cr

AND

163

Mo

tion number for Cr and V of 4.5, i.e., 0.5 (2x4+6-5). Such a transition state
could stem from a V(II1) + Cr(V1) or V(IV) Cr(V) combination. In terms of
oxidation numbers the possible mechanisms involve either disproportionation
of V(1V) followed by a two-equivalent oxidation of V(II1) ( A ) , or a sequence of
three one-equivalent steps ( B ) ;viz.

2 V(IV)

2 V(V)+V(III)

rapid, K

k-i

( A ) ~(111) Cr(VI) 3 V(V> Cr(IV)

slow

V(IV)+Cr(IV)

+ V(V)+Cr(III)

rapid

V(IV)+Cr(VI)

+ V(V)+Cr(V)

rapid, unfavourable

(3.2)

slow

(3.3)

rapid

(3.4)

or
( B ) V ( I V ) + C r ( V ) d V(V)+Cr(IV)

V(IV)+Cr(IV)

+ V(V)+Cr(III)

If mechanism ( A ) applied the Cr(VI)+V(IV) system would be anomalous when


compared with the Cr(VT)+ Fe(1I) and Ce(IV)+ Cr(I1T) reactions which have
similar rate laws and Cr(V) + Cr(1V) transformations as rate-controlling steps.
Apart from this there are other good reasons for rcjecting mechanism ( A ) . At
25 "C, K is lo-'' and k' is 0.56 I.mole-'.sec-', allowing k , to be calculated as
0 . 5 6 ~10" I.mole-'.sec-' (since k' = k,K). Furthermore, V(V) can only retard
the reaction if the intermediate V(Il1) reacts with V(V) in preference to Cr(VI),
i.e., the conditions k - , [V(V)] >> k,[Cr(VT)] must apply. A value of k - l =
1.4 x lo4 I.mole-'.sec-' at 25 "C has been obtained by Daugherty and Newton53
from a study of the V(III)+V(V) system. Using this value would mean that the
ratio [V(V)]/[Cr(VT)] should be very much larger than 4 x lO5(k,/k-,). Mechanism
( A ) can be discarded on the grounds that this result is much higher than that
3 x 10').
achieved experimentally (the highest value of [V(V)]/[Cr(VI)] is
Mechanism ( B ) can be made more explicit by making use of the result that the
rate is independent of [H']. Thus, steps (3.2) and (3.3) can be rewritten as

VOz+ HCr0,-

+ H,O + VOz+ + H3Cr04

V 0 2 ++H3Cr0,

(H3Cr0,VO2+)'

rapid
slow

on the assumption that Cr(V) exhibits a coordination number of 4. The second


term of the rate law (3.1), in which there is a second-order dependence on
[HCrO,-1, signifies the participation of Cr,O,'- as a reactantt, and a transition
t The species HCr04- and Cr,O,'-

are in equilibrium by

2 HCr04- $ C r , 0 7 2 - + H , 0
as 98 ].mole-' at 25 "C
The equilibrium constant has been measured spectroph~tometrically~'
and ,u = 1 M .
References pp. 267-273

164

C O M P L E X E S 0 F D I FF E R E N T M E T A L S: 0 X I D AT I 0 N-R E D U C TI 0N

state, consisting of two Cr atoms and one V atom, of composition (HCr207V2+f


rn H20)*. A reasonable mechanism, similar t o (B), would have Cr207'- as a
reactant in the first step and a dimer of Cr(V) and Cr(V1) as the intermediate
in the slow stage.
The kinetics of the oxidation of iron@) by chromium(V1)
HCrO,-+3 F e Z + + 7H +

Cr3++3Fe3++4H 2 0

were studied first by Bensons4 and subsequently by Gortner" and by Wagner


and Preisss6. The early literature is discussed in detail by Westheimers7. More
recently the reaction has been the subject of a comprehensive investigation by
Espenson and KingSs under conditions where the dominant species of Cr(V1)
and Fe(II1) are HCr04-, FeCrO,+, and Fe(H20)63+.There is no evidence to
suggest that the product Cr(1II) is other than the simple hexaaquo ion; the formation of a dimeric species of Cr(II1) is considered unlikely. The extent of complex
formation between Fe3+ and HCrO,-, uiz.
Fe3++ HCr0,-

+ FeCrO," + H +

K,

received detailed study and a value of Kl = 1.4 was assigned at 0 "C and ,u =
0.0839 M t . Effectively, this result means that under these conditions 8 to 40 %
of Cr(V1) is present as Fe(II1)-chromate complex. Rate measurements on the
Cr(V1) Fe(I1) reaction were made both by a direct spectrophotometric approach,
and by the titrimetric method first used by B e n ~ o n(utilising
~~
the induced oxidation of iodide ion). Iron(II1) retards the reaction and the corresponding rate law
takes the form

- -

with k' = 6.2 x 10' 14.mole-4.sec-' and k"


0 "C. The gross features of the mechanism are

Cr(V1) Fe(I1)
Cr(V)+ Fe(l1)
Cr(IV)+Fe(II)

2.2 x

1OI2

ls.mole-s.sec-'

at

+ Cr(V) + Fe(III)
4

Cr(IV)+Fe(lII)

rate-determining

Cr(lII)+ Fe(1II)

rapid

The first term of (3.5) implies a corraponding net activation process" represented
+ See, also, ref. 278.

,.

OXIDATIONSBY

Cr A N D Mo

H C r 0 4 - + 2 F e Z + + 3H + + ( n - 2 ) H,O

-+

165

[FeOCrO(OH,),3+]t+Fe3+

That is, the activated complex contains one Cr(V) atom and one Fe(I1) atom.
Espensonsobhas shown, from a consideration of the induced oxidation of iodide
ion, that the reaction between Cr(V1) and Fe(I1) requires one added proton.
Consequently, the [H+I3-dependence of the rate can be viewed as the addition
of two protons in a pre-equilibrium followed by the addition of a further proton
in the slow step, viz.
HCr04-+FeZ++2H +
H3Cr04+Fe2++ H +

+ H3Cr04+Fe3+

rapid

(H4Cr04Fe3f)t

slow

-+

The second term of (3.5) is indicative of an activated complex containing two


Cr atoms with an average oxidation number of + 5 . 5 together with one Fe(I1)
atom. This result suggests that the reactive entity in the rate-controlling step may
well involve one Cr(V) atom and one Cr(V1) atom, e.g., Cr207H,("-3)+. ROS, ~while finding no evidence for this pathway, have detected
seinsky and N i c 0 1 '~
an additional term first-order in Fe(I1).
The kinetics of the reduction of Cr(V1) by tris(1,lO-phenanthroline)iron(II),
viz.

HCrO,-+3 Fe(phen),'++7 H+

C r 3 + + 3 Fe(phen),"+4

H,O

have been studied by Espenson and King5* over the temperature range 0-40 "C
by a spectrophotometric method. No unequivocal conclusions can be drawn
about the mechanism from the rate data except the lack of effect of F e ( ~ h e n ) , ~ +
concentration on the rate suggests that reduction of a Cr(V) species is not ratedetermining. In addition, the rate of the Cr(V1) + Fe(phen),'+ reaction is much
smaller than the Cr(VI)+FeZ+ reaction. This may be due to different thermodynamic tendencies in the primary and rate-determining step
Cr(VI)+Fe(II)

-+

Cr(V)+Fe(ITI)

as reflected by differences in equilibrium constants of


lo6. A further factor
contributing to the greater rate of the Cr(Vl)+ Fe(I1) reaction may be the availability of an inner-sphere transition state with an oxygen-bridged structure,
Fe-0-Cr. In contrast, reaction of Fe(phen)32+ must needs occur by an outersphere path'. These interpretations are supported by the results of BirkZE0on
reactions of Cr(V1) with other Fe(I1) complexes.
References pp. 267-273

166

COMPLEXES OF DIFFERENT METALS: OXIDATION-REDUCTION

In the multi-equivalent oxidation of Sn(I1) by Cr(V1)


2 Cr(VI) + 3 Sn(I1)

2 Cr(Ill)

+ 3 Sn(IV)

there is some evidence for the presence of Sn(II1) or a related reducing species59.
In the Mo(V) + Sn(I1) system, however, formation of intermediates seems insignificantz8'.
Mason and Kowalak6O have examined the oxidation of As(I1I) by Cr(V1)
i n 0.2 M acetic acid-0.2 M potassium acetate buffers. When As(II1) is in large
excess over Cr(V1) the rate law is
-d[Cr(VI)]
dt
~

- k , K,[Cr(VI)][As(III)]
-~

+ K,[As(III)]

i n accord with thc scheme

As(III)+ HCr04As(III).HCrO,-

+ As(III).HCrO,k3

--+

products

pre-equilibrium, K ,
rate-determining

At 25 "C, and with the ionic strength adjusted to 1.5 M with potassium nitrate,
sec-' and 22.4 l.mole-', respectively.
the values of k 3 and K , are 3 . 7 6 ~
At higher concentrations of Cr(VI) an analogous mechanism applies, HCr04being replaced by Cr,072-. No evidence is cited for the formation of As(1V);
it seems likely that As(V) and Cr(IV) are formed in the slow step.
Kinetic data for the oxidation of neptunium(V) by chromium(VI), uiz.
Cr(V1)+3 Np(V)

Cr(III)+3 Np(V1)

have been obtained by spectrophotometric means at wavelengths of 350 mp


[where the absorbance is largely due to Cr(VI)] and 980 mp [where the absorbance
is entirely due to Np(V)I6'. Formally, the process may be described as the simple
transfer of one of the two available Sfelectrons from Np(V) to the oxidant since
both Np(V) and Np(V1) have the same linear 0-Np-0 structure. Fresh solutions
of neptunium were preparcd directly before each kinetic experiment to reduce
the possibility of peroxide formation by radiolysis of the solutions by a-particles
from the decay of 237Np.The empirical rate law is not in line with the rate laws
of the Cr(VI)+Fe(II) (p. 164) and Cr(VI)+V(IV) (p. 162) reactions and was
derived by a computer programme designed to approximate, from a set of experimental data, the first derivative of an unknown function. At constant acidity
the derived law takes the form

O X I D A T I O N S BY

Cr

AND

Mo

167

and has the following indefinite integral

i n which the subscripts refer to initial concentrations of reactants, and the parameters A and B are defined in terms of initial concentrations and rate coefficients
k' and k". Values for A , B and C , computed from rate data (concentration of
Np(V) z'ersus time) using equation (3.7), were used to evaluate k' and k". Agreement was attained between these results and those calculated directly from equation (3.6). The rate law was checked by observing the constancy of k' and k"
(at constant perchloric acid concentration and ionic strength) over a range of
initial concentrations of Cr(VI), Np(V) and Np(V1). The average values are
k' = 8.26 I.mole-'.sec-'
and k" = 0.42 at [H'] = 1.51 M , ,D = 2.00 M and
25 "C. As a mechanism, Sullivan6' has proposed a sequence of one-equivalent
steps

Cr(VI)+NpO,+

+ Cr(V)+NpOZ2+
k4

k-4

Cr(V)+NpO,+
Cr(lV)+NpO,+

3 Cr(IV)+NpO;+
+

Cr(III)+Np022+

rapid

On this basis Cr(V), not Cr(IV), is the kinetically important intermediate such
that k' = 3 k , and k" = k-,/k,. The hydrogen-ion dependence of the reaction
rate has been discussed. Furthermore, comparisons are drawn with the rate of
the Cr(VI)+ F e ( ~ h e n ) , ~reaction",
+
and Sullivan has speculated on the intimate
nature of both mechanisms in the light of Marcus theory4j.

3.2

OXIDATIONSBY CHROMIUM(III)

Oxidation of hexaaquovanadium(l1) by Cr(H,0),SCN2+


stagesh2.2'2
CrSCN2++ V 2 + -% Cr2++VNCS2+

occurs in two

(3.8)

VNCS2+ kz_ V 3 + + SCNA value for k , of 9.8 I.mole-'.sec-', at [H'] = 1.0 M , p = 1.0 M and 25 "C,
was obtained by following the rate of disappearance of CrSCN" at 262 mp2".
By contrast, the rate coefficient for the reaction of CrNCS2+ with V 2 + is 1.7 x lo-,
l.mole-'.sec-' under the same conditions62. The formation and subsequent disReferences pp -767-273

168

C 0 M P L E X E S OF D I F F E R E N T MET A L S: 0 X I D A TI 0 N-R E D U CT I O N

appearance of VNCS" was followed at 350mp and 400mp: k, is quoted as


0.99 sec-' at 25 "C and p = 1.0 M . Reaction of CrSCN" with Cr", U ~ Z . 6~4 ~

is negligible in comparison to reaction (3.8). Baker et a1.62suggest that the reaction of CrSCN" with V2+ takes place by a thiocyanato-bridged transition state,
in which the anion is bonded to both metal atoms.
The equilibrium constant of the system

2 Cr(II)+ Eu(II1)

Cr(III)+ Eu(I1)

k-3

is e ~ t i m a t e d ~(from
~ , ~ ' redox potentials) to be 2.2. Spectrophotometric measurements (at 575 mp, the absorption maximum for C r 3 + )reveal both forward and
backward reactions to be extremely slow (5 % reaction in one day). Approximate
1 . 7 lo-'
~
I.mole-'.sec-'
and k-,
1 . 4 lo-'
~
values ( f 2 0 % ) are k,
I.mole-'.sec-' at 25 "C and 0.5 M HCIO,. The addition of chloride ion, whilst
having no influence on k,, catalyses the back reaction appreciably, and

d[Cr(Ztl)]/dt

k - , [Cr(II)][Eu(IIl)] [CI - ]

k - , being 9.8 x lo-, 12.mole-2.sec-' under the same conditions. The substitutioninert CrC12+ species is formed as a product. With CrC12+ as the Cr(II1) species,
the reduction

I.mole-'.sec-' at 25 "C and activation


has a rate coefficient k , of 2.23 x
parameters A H t and ASt of 17.1 kcal.mole-' and 14.0f 5 cal.deg-'.mole-',
respectively. The marked increase in reactivity of CrCI2+ as compared with Cr3+
indicates that an inner-sphere activated complex is involved. Although (EuCICr4+)'
seems the likely structure, the other possibilities, ( E u H , O C I C ~ ~ +and
) ~ (EuH,
OCrCI4+)', cannot be excluded. It is of interest that, in the reactions of Eu(I1)
with Cr(IZI), V(nl), V(1V) and Fe(lll), there is an approximately linear relationship between the logarithms of the rate coefficients and the standard free-energy
changes.
Adin and Sykesb5 have reported on the reduction of monohalogenochromium(II1) complexes by Eu(l1)

in perchloric acid solutions at p

1.0 M . Corrections were applied for the

OXIDATIONS BY

Mn

AND

Re

169

simultaneous uncatalysed and Cr(I1)-catalysed aquations of CrX",


CrX2+

uiz.

Cr3++ X -

Cr 2 + + C r X 2 ++ c r 3 + + C r 2 + + X At 25 "C the rate coefficients @mole-'.set-') for the reductions are given as
> CrBr"
( 3 . 4 ~lo-,) > CrCI2+ ( 1 . 4 ~
> CrF2+
Cr12+ (4.1 x
(6.0 x
The reactivity sequence follows the "normal" order of bridging
efficiency (see p. 194) andisthe reverse to that found in the reactions of Eu(I1)
and Fe(1I) with C O ( N H ~ ) ~ X '(p.
+ 190).
3.3

OXIDATIONS BY MOLYBDENUM(V)

The rate of oxidation of tris(2,2'-bipyridine)osmium(II)


lybdate(V)
Mo(CN),~- + Os(bipy),'+

by octacyanomo-

4 Mo(CN)$- + O ~ ( b i p y ) , ~ +
kr

has been measured by the temperature-jump method66. In practice, application


of this method to a rapid bimolecular electron-transfer can only be successful
if the equilibrium constant of the system is close to unity. Consequently, for the
perturbation to effect a measurable shift in the equilibrium position the system
must show a considerable AS" value, as occurs, for example, in reactions between
oppositely-charged species. The change in equilibrium between Mo(CN),~and O ~ ( b i p y ) , ~followed
+
by recording the absorbance of the latter at 480 mp
10" had been induced.
as a function of time after a temperature difference of
Calculation revealed k, and k, to be 2 . 0 lo9
~ I.mole-'.sec-'
and 4 . 0 lo9
~
I.mole-'.sec-', respectively, at 10 "C and p = 0.50 M . These values are close to
the diffusion-controlled limits as given by the Debye equation67. Application of
Marcus theory43 (see p. 247) allows rate coefficients to be estimated for the
O ~ ( b i p y ) , ~++O ~ ( b i p y ) , ~ +
and M o ( C N ) , ~ - + Mo(CN),,- exchange reactions:
the calculated values are 1 x l o 7 I.mole-'.sec-' and 3 x lo4 I.mole-'.sec-' whereas those found experimentally6' are > 1 x lo5 l.mole-'.sec-' and
3 x lo4
l.mole-'.sec-', respectively, at 10 "C and p = 0.

4. Oxidations by manganese and rhenium


4.1

OXIDATIONS BY MANGANESE(VII)

The kinetics of the oxidation of Fe(CN),4- by permanganate have been


examined in phosphate buffers over the pH range69 1.6-6.3. The stoichiometry
References p p . 267-273

170

c oM P L E x E s oF

D I F F E R E N T M E T A L s : O X I D A T I o N-R E Du c T I o N

of the reaction was confirmed by measurements on the product Fe(CN)63- at


420 mp. The technique employed to obtain kinetic data incorporated a rapidmixing device; the rate of disappearance of Mn0,-(Fe(CN)64- in excess) was
followed at 520 mp using photographic recording of oscilloscope traces". The reaction is first-order in both reactants; second-order rate coefficients were deduced
from the equation

E'CI

log -L
[At:

5ck't
2.303

=-.-

+ constant

where A , is the absorbance of MnO,-, c = ([Fe(CN)64-]o/5)- [MnO,-], and


is the molar absorptivity of Mn0,-. At 15.1 "C in phosphate buffcr of pH
6.30 and p = 0.0933 M , the rate coefficient k' is 2 . 6 4 ~lo4 I.mole-'.sec-'. This
value was verified by monitoring the rate of formation of Fe(CN)63- at 420 mp,
by which method k' = 2.56 x lo4 I.mole-'.sec-'. Increase in acid concentration
increases the rate of reaction; activation parameters are quoted at low and high
acidities. In the range pH 5-6 the rate-controlling step may involve an ion-pair
formed with K' (or Na') from the buffer

E'

K++Fe(CN),,MnO,-+KFe(CN)?-

+ KFe(CN):4 MnO,'-

rapid,

K,

+KFe(CN)62-

350 I.mole-*

slow

As the p H is reduced to pH 3 the rate increases, and eventually achieves a limiting


value at
pH 2.5. The following scheme, with protonated species, is suggested
as likely at p H 1-2.

H'
Fe(CN),"H'
HFe(CN)63-

Mn0,-

+ HFe(CN)63+ H2Fe(CN),'-I

rapid equilibria

+ H2Fe(CN),'- -% MnO,'- + H,Fe(CN),-

slow

The reaction between MnO,- and HFe(CN)63- (rate coefficient, k , ) is important


in the intermediate pH region. Detailed treatment o f the data yields a value of
1 7 . 4 ~lo4 I.molc-'.sec-' for k , . The resemblance of k , to k, ( 5 2 . 0 ~lo4 1.
mole-'.sec-') suggests that, in terms of intimate mechanism, a single proton is
necessary as a bridge between the reactants, electrostatic effects being unimportant. Comparison of the rate parameters with those calculated on the basis of
Marcus theory43 suggests that an outer-sphere process obtains. Replacement of
K + by Na' ion in the buffer solution is shown not to produce specific cation effects.
Permanganate oxidises formatopentaamminecobalt(II1) in a complicated

OXIDATIONS BY

Mn

AND

171

Re

fashion7', the over-all reaction being compounded of a mixture of


MnO,- + 3 Co(NH,),.OCHO'+

-+

3 CO, + 3 Co2+ MnO,

and

2 Mn0,- + 3 Co(NH,),.OCHO*+ -+ 3 C 0 , + 3 (NH,),CO.OH,~+


+ 2 MnO,.
The ratio of unreduced to reduced cobalt is variable and is dependent on the concentration of Mn0,-. uiz.
[(NH,),CO.OH,~+]/[CO~+]
= 3 x 10, [MnO,-]
By way of contrast, oxidation of the organic ligand in oxalatopentaamminecobaltand p-aldehydobenzoatopentaaniminecobalt(IlI)73 is accompanied by
reduction of the cobalt(II1) centre in the case of one-equivalent oxidants, e.g.
Ce(IV), but not in the case of two-equivalent oxidants (e.g. CI,). The rate law
is simple
-d[Co(NH3),.OCHO2+]/df

k , [CO(NH,),.OCHO~~][M~O,-]

k,, independent of [H'], is


2.5 x lo-, I.mole-'.sec-' at 0.1 "C, and 0.21
I.mole-'.sec-'
at 25 "C, [HCIO,] = 0.1 M and p = 1.0 M . Replacement of
hydrogen by deuterium in the formato group, whilst bringing about no change in
stoichiometry, leads to a
10-fold reduction in rate. Candlin and Halpern7'
consider that the first stage in the reaction is the formation of an intermediate

Mn0,-

+ (NH,),CO"'(OCHO-)~+ +HMn0,- + (NH3)sC~11r(C02-)2t

i n a one-equivalent step by abstraction of a hydrogen atom. The fate of thc


intermediate is to produce either Co2++ CO, or (NH3)5C01"*OH23+ CO,.
It seems likely also that the oxidation of free HCOO-, in the oxidation of formic
acid by MnO,-, occurs by hydrogen-atom and not hydrogen-ion t r a n ~ f e r ~ , , ~ ~ .
The oxidation of (NH,),Co.OCHOZf by Co3+ and S,O,'- (catalysed by Ag')
generates (NH,),CO.OH,~+ as the major product, whereas the oxalato and p aldehydobenzoato complexes yield almost exclusively Co2+.
In the multi-equivalent reduction of permanganate by Sn(l1) there is evidence
to suggest the presence of Sn(I1I) in the reacting ~ y s t e m , ~In. this respect the
react ion

Mn(V)+ Sn(I1) --j Mn(IV) Sn(II1)


is considered the likely source of Sn(I1I).
References pp. 267-273

co M P LEX E s OF

1 72

4.2

D I FF E R E N T M E T A L s : ox I DA T I ON-RED

u C T IO N

OXIDATIONS BY MANGANESE(III)

There is a paucity of information on reactions of the strongly-oxidising


manganese(Il1) with inorganic substrates. The main reason for this neglect lies
in the tendency of Mn(lI1) to disproportionate, uiz.
2 Mn(II1)

+ Mn(II)+ Mn(IV)(as MnO,)

This probably takes place through the self-condensation of hydrolysed species


and thus disproportionation is reduced at high acidities. This tendency can be
suppressed also by the use of strong sulphuric acid ( 2 5 M ) as the medium,
or by employing the oxalato or pyrophosphato complex. In both these cases
strong complexing occurs. Perchlorate possesses no such stabilising ability but
R o ~ s e i n s k yreports
~~
that reasonably stable solutions of Mn(II1) perchlorate
( 5 lop3 M in 4 M acid) can be prepared from acid permanganate and
Mn(1I) solutions, a large excess (- 25-fold) of the latter preventing disproportionation of Mn(1II).
Rosseinsky and N i c 0 1 ~have
~
made a kinetic study of the Mn(Ill)+V(IV)
reaction
N

Mn(III)+V(IV) = Mn(II)+V(V)
by following the rate of appearance of V(V) at 325 mp
A simple second-order rate law

(E =

284 l.mole-'.cm").

obtains over the range 0.8 to 3.0 M perchloric acid at temperatures from 5 "C
to 20 "C. With
M reactant concentrations, half-lives were of the order
of 10-20 sec. At 20 "C and 3 M HC104, kobsis 1.53 x lo2 l.mole-'.sec-'. The
rate increase with decrease in hydrogen-ion concentration is i n keeping with a
mechanism in which a hydrolysed species of Mn(II1) takes part, uiz.

Mn3++V02++H20

M n 2 + + V 0 2 + + 2H +

Mn3+ H 2 0 + MnOH2++ H +

rapid equilibrium, K h

M n O H 2 + + V O Z +3 M n 2 + + V 0 2 + + H t
The rate data were fitted to the relationship

OXIDATIONS BY

Mn

AND

173

Re

Although the activation parameters corresponding to steps k,, and k , are given,
there is some doubt over the correct value for Kh (and the accompanying AH,,)
under the conditions employed. There seems to be no special kinetic feature
about the oxygen ligand in V 0 2 + , since the rates of reaction of the latter oxocation
withvarious oxidants follow the same sequence (V(V)" > Mn(III)79 > C0(111)~~
> Fe(lX)28 > T1(III)26) as the corresponding reactions of Fe(lI)22,24,28~79*81.
An important study of Mn(II1) reactions is that by Diebler and Sutins2.
Mn(lI1) was prepared by anodic oxidation of up to 0.1 M Mn(II) in 1 to 6 M
HCIO,. The nature of Mn(I1I) in perchloric acid is discusseds3. The oxidation
of Fe(I1) by M n ( m ) is kinetically second-order with corresponding rate coefficients
of 1 . 4 6 lo4
~ and 1 . 6 7 lo4
~ I.mole-l.sec-' in 3.0 and 1.0 M perchloric acid,
respectively, at ,u = 3.1 M and 25 "C. The latter rate coefficient is consistent with
the value of 6 . 9 2 lo3
~ 1.mole-'.sec-' at 15 "C (1.04 M HClO,, p = 3.04 M )
recorded by Rosseinskys4. The oxidation of various substituted 1,lO-phenanthroline complexes of Fe(I1) was investigated also; a linear relationship exists
between log k l , 2 and log K l , 2 (Fig. 2). The gradient of the plot is 0.45, a value
close to the theoretical value of 0.5 given by the Marcus equation [eqn. (12.6), p.
2471. Rate coefficients for the oxidation of Fe(I1)-phenanthroline complexes in
pyrophosphate-sulphate media are presented. A linear relationship is observed
between log k l , 2 and the formal oxidation potentials of the Fe(I1) complexes
(Fig. 3). A rapid-mixing and flow apparatus44 was employed for all the above
reactions. In a recent paper, Rosseinsky and Nico17' confirm that the Mn(IlI)+
Fe(I1) reaction is second order from measurements on the changes in the diffusion current of iron(I1) at a rotating platinum electrode. In the ranges 0.54-3 M
perchloric acid and 0.3-15 "C the observed rate coefficient has an acid dependence

0 in 3M HCIO4

I in

4.0

6.0

8.0

1M HCQ

10.0

log Kip

Fig. 2. Relationship between the logarithms of the rate coefficients (k,,2) and the logarithms of the
equilibrium constants ( K , , 2) for the oxidation of various substituted Fe(I1)-phenanthroline
complexes by Mn(II1) in 1 M and 3 M HC104 at 25.0 "C. (From Diebler and Sutins2,by courtesy
of The American Chemical Society.)
References pp. 267-273

174

COMPLEXES O F D I F F E R E N T METALS: OXIDATION-REDUCTION

085

090

095

100

105

110

E: (volt)

Fig. 3. Relationship between the logarithms of the rate coefficients ( I ? , . ~ ) for the oxidation of
various substituted Fe(I1)-phenanthroline complexes by Mn(1II) pyrophosphate (pH I .O,
ionic strength
0.5 M ) at 25.0 ' C and the formal oxidation potentials ( E ; ) of the Fe(11) c o m plexes. (From Diebler oiid S i i f i i i s Z ,by coiirti'sy uf The Aiiiericuti Chcriiicul Sociery.)

similar to that encountered in the Mn(III)+V(IV) reaction"


cordingly, Rosseinsky and Nicol suggest the analogous scheme

(p. 172). Ac-

M n 3 + + F e 2 + -+ Mn2++Fe 3+

Mn3+ H 2 0

+ MnOH2++ H +

rapid equilibrium, K,,

M n O H 2 + + F e 2 +-+ M n 2 + + F e 3 + + O H As noted above (p. 173), a lack of certainty over the correct value of Kh under
the prevailing conditions gives rise to ambiguities in the values of individual
rate coefficients.
Ro~seinsky'~has examined the oxidation of Hg(1) and Mn(II1) in aqueous
perchloric acid at 50 "C by a method involving titration of Mn(1II) with Fe(I1).
At constant acidity ([H'] = 3.0 M ) his rate data can be expressed as

where Hg(I), represents Hg22++ Hg2C104+. Thus the reaction is retarded by


both Hg(ll) and Mn(I1). This result suggests two rapid pre-equilibria

OXIDATIONS BY

Mn

AND

Re

175

and
2 Mn(1II)

+ Mn(IV)+Mn(II)

K2

lo-,

which are followed by two slow and rate-controlling steps


Mn(III)+Hg(O)

--%Mn(II)+Hg(I)

slow

Mn(IV)+Hg(I),

3 Mn(II)+2Hg(II)

slow

together with
Mn(lTI)+ Hg(1) .--,Mn(II)+Hg(II)

rapid

In this scheme Hg(1) is presenta5 as Hg+ ions (possibly ion-paired with CIO,-),
and Mn(1V) is likely to be M n 4 + + M n O Z + .Step (4.2) may proceed either as
written (two-equivalent change) or via two consecutive one-equivalent stages
with HgZ3+,or Hg(l), +Hg(II), as intermediates. On this basis k' and k" are
identified as k , K , and k, K 2 , respectively. Very approximate values of k 2 and k,
were estimated (on the basis of a number of rather drastic assumptions) as 5 x lo3
I.rnole-'.sec-' and 2 50 I.mole-'.sec-', respectively, at 50 "C. Other possible
mechanisms tested and found inadequate in explaining retardation of the reaction
by products include (a) reactions Mn(II1) Hg(I), and Mn(IV)+ Hg(O), and
( b ) a back reaction in step (4.1) and omission of (4.2). It is of interest to note that
Mn(1II) is capable of oxidising both Hg(0) and Hg(1) whereas Tl(I11) oxidises
only the formera6 and Co(Il1) the latters0. Also, of the three oxidants, only
Co(1II) is capable of oxidising HgZ2+,a result which has led R o ~ s e i n s k yto
~~
speculate on the ability of Co(IJ1) to react via a high-spin species which can tZsoverlap with HgZ2+.

4.3

OXIDATIONS BY RHENIUM(VII)

Rhenium(VII), as the perrhenate ion ReO,-, is reduced rapidly to rhenium(\/)


by tin(I1) in hydrochloric acid solutions. Further reduction to rhenium(1V) takes
place slowlya7. Banerjea and Mohans8 have followed the rate of formation of
Re(IV) at 550 mp. At 30 "C, with [Re(V)] = 3 x lo-, M and [Sn(II)] = 3 x lo-'
M , the observed rate coefficient is 6 . 2 ~
lo-, sec-' in 3 M hydrochloric acid
solution. The following mechanism is considered appropriate
ReOC1,'-

+ H2SnCI,

---f

Re(OH)C152- +HSnCI,

ReOC1,2- +HSnCI, +Re(OH)CI,'References pp. 26 7-273

+SnCI,

176

co M P L E x ES OF D I F F E R E N T M E T A L S : ox I DA T I o N-RE D u c T I o N
5. Oxidations by iron(I1I) and ruthenium(IT1)
5.1

OXIDATIONS BY IRON(III)

The reaction between complexes of the type Fe(H,O),X+ and V(H20)62f


has been investigated by a flow method44 by measurements at the absorption
maximum of the former species6. Rate coefficients are recorded in Table 3 when
TABLE 3
RATE COEFFICIENTS FOR THE REACTION BETWEEN
A N D ,u =

Fe(H20)5XZ+A N D V ( H Z O ) Q ~A+T 2 5 C

1.0 M (ref. 6 2 )

X
HzO
OHCI NCSN3-

k(l.mole-.sec- * )
1.8 x lo4
<4 xi05
4.6 x 1 0 5

6.6 x lo5
5.2 x 105

X = H,O, OH-, C1-, NCS-, and N 3 - ; in all these cases the rates of reduction
are rapid compared with the rates of substitution of X- in Fe(H,0)63+. No
evidence is evinced for formation of VX+ complexes. It is noteworthy that
OH- exerts little effect on the rate of this reaction although it has a pronounced
influence on the rate of other reactions of Fe(IlI)89. This result, together with the
observation that the other anions have similar effects on the rate, would seem to
indicate that the FeX2++V2+ reactions proceed via an outer sphere path. The
suggestion is made, however, that, although the Fe3++ V 2 + reaction is undoubtedly of the outer-sphere type (since the rate of water replacement in V2+ is much
slower than the electron-transfer), the FeX2++V+ reactions occur through
inner-sphere, water-bridged activated complexes resulting from the loss of a
water molecule coordinated to Fe(lI1).
Fe(II1) oxidises V(111) to V(1V). Superficially the reaction is unexceptional:
the stoichiometry is simple
Fe(III)+V(III) = Fe(lI)+V(IV)

(5.1)

and the process is first-order in each reactant (Fig. 4;plot A). However, a number
of interesting features have been revealed by Higginson and Sykes in the course
of a detailed kinetic studyg0. Measurements were made at 755 mp (where V 0 2 +
absorbs strongly). Precautions were taken against atmospheric oxidation. Reduction of perchlorate to chloride by V(III) was slight under the conditions prevailing. On the assumption that the reaction proceeded as a single-stage process
V(IV) was added (in concentrations similar to those of the reactants) in an effort

OXIDATIONS BY

Fe(II1)

AND

Ru(II1)

177

Fig. 4. Effect of initial presence of products upon the rate of the Fe(III)+V(III) reaction. Initial
M (A, B, C),
concentrations: [Fe(III)] = 0.0226 M ; [V(III)] = 0.022 M, [Fe(II)]
0.07-0.12 M (D); [V(IV)]
M (A, D), 0.057 M (B), 0.094 M ( C ) ; [ H +] = 1.00 M,
ionic strength, 3.0 M, temp., 20.0 "C. (From Higginson and Sykesg0, by courtesy of The Chemical
Society.)
N

to disclose a back-reaction. Instead V(IV) has a pronounced accelerating and not


a retarding influence (Fig. 4; plots B and C ) . On the other hand, the presence of
Fe(I1) leads to a decrease in rate which then achieves a limiting value after the
addition of a certain amount of Fe(l1) (Fig. 4; plot D). A quantitative analysis
of these and similar results allows the empirical rate law to be deduced as

-d [V(III)]/dt =
= k, [Fe(II1)][V(lII)]

+k'[Fe(IIl)] [V(TII)][V(IV)]/[Fe(II)]

(5.2)

The first term in (5.2) corresponds to the simple path represented by (5.1), rate
coefficient k, . The form of the second term can be understood on the basis of the
sequence
Fe(III)+V(IV)

$ Fe(II)+V(V)

k-z

(5.3)

V(V)+V(III) 5 . 2 V(IV)

(5.4)

Although k2/kd2 is very small ([V(V)] 0.1 % of [V(III)] and [Fe(III)]), the
forward reaction in (5.3) is sufficiently rapid for (5.3) and (5.4) to make a significant contribution in the absence of large amounts of Fe(I1). The presence of
added Fe(1I) so reduces the amount of V(V) that (5.3) and (5.4) are no longer
References pp. 267-273

178

COMPLEXES OF D I F F E R E N T METALS: OXIDATION-REDUCTION

effective. Applying the steady-state approximation to V(V) in (5.3) and (5.4),


it follows that

Equation (5.5) reduces to the form of (5.2) if k - , [Fe(II)] >> k , [V(III)], whereupon k' = k 3 k , / k - , . Equation (5.5) can be written as
CFe(W1[V(IV)I
- k-,[Fe(II)]
( - d [V( I I I )]/d t ) - k [Fe( III)] [V( I1I)]
kz k3[V( 111)]

+ -1

k2

Plots of the left-hand side of this equation uersus corresponding ratios of [Fe(II)]/
[V(IlI)] are linear. Such plots appear to pass through the origin so that k , must
be large. The gradients of the plots ( k - , / k , k , ) yield a value for k3 since k - , / k ,
was determined separately by measurements on the EMF of a cell composed of
Fe(III)+Fe(II)andV(V)+V(IV) half-cells. At 25 "C, /A = 3.0 M a n d [H'] = 1.0
M , k , is calculated to be 260 l.mole-'.sec-'. Step (5.1) was found to be acid
dependent

kl

b+(~/[H+l)+(d/[H+]~)

At 25 "C, approximate values of the constants b, c and d a r e 2.2 x lo-, Lmole-'.


sec-', 3.0 x lo-, sec-' and 3.4 x lo-, mole.1-'.sec-',
respectively. Three
parallel routes (with activation complexes differing in the number of bound
protons) are indicated. Activation parameters corresponding to k , and k, are
reported.
Besides being catalysed by V(IV), the reaction is also subject to catalysis by
Cu(I1) due to
Cu(II)+V(III) -%Cu(I)+V(IV)
Fe(III)+ Cu(1) -+ Fe(lI)+ Cu(I1)

(5.6)
rapid

(5.7)

With Cu(I1) present in similar concentrations to Fe(II1) the contribution of the


uncatalysed path is small. The rate coefficient k , varies with acidity according
to k , = g+h/[H+]+. From Arrhenius plots of g and h, A H f and AS* were calculated and compared with those corresponding to path b. At 25 "C in 1.0 M
perchloric acid, k , = 0.30 I.mole-'.sec-', AH: = 21.3 kcal.mole-' and AS: =
5 f 5 cal.deg-'.mole-'. Under the same conditions k, = 0.025 1.mole-'.sec-',
t See, also, ref. 283.

OXIDATIONS BY

Fe(II1)

AND

Ru(II1)

179

AH: = 17.3k4.1 kcal.mole-' and AS: = - 15+ 13 cal.deg-'.mole-'. Thus the


efficiency of Cu(I1) as a catalyst is due to a more favourable entropy of activation.
In fact the enthalpy of activation for the catalysed reaction is greater than that
for the uncatalysed reaction2,. Recently, Espenson et al. 9 2 , 2 8 4 have measured
the rate of the reverse of reaction (5.6) as
d [V3+]/dt = e[V02 '1 [Cu' 1[H+]
where e = 3.84 x lo2 12.mole-2.sec-' at 25 "C and p = 3.0 M2',. These authors
question the reality of the term involving g in the expression for k,, as included
by Higginson and Sykes". Consequently, since h = 0.384 sec-' then the equilibrium constant for reaction (5.6) is calculated as 1.1 x lo3 I2.moleC2 (c/h) at
25 OCZs4.Espenson et al.92,285have been able to measure also the rate coefficient
of reaction (5.7) as 1.6 x lo5 I.mole-'.sec-' at 25 "C in 1 M HC10,2s5.
Fe(II1) oxidises the tantalum cluster-ion Ta,C1,22+ by way of two consecutive
one-electron transfer steps93 (see, also, refs. 286 and 287).

At 15 "C and 0.020 M perchloric acid k , = 620 I.mole-'.sec-', k, = 3.72 I.


mole-'.sec-', k-, = 201 I.mole-'.sec-' and K = k,/k-, = 0.018 (cf. spectrophotometric value of 0.016 at 22 "C). As might be expected for redox reactions
in which little molecular rearrangement is required, these steps are rapid.
The oxidation of Cr(I1) by Fe(II1) in perchloric acid is markedly catalysed
by chloride ion. Taube and Myers9, found that Cr(H20)5C12+is formed along
with C I - ( H ~ O ) , ~as
+ products of the oxidation, the relative proportions of the
two species depending on concentrations of H+and CI-. The suggestion was made
that Cr(H20)5C12+is produced by reaction of C I - ( H ~ O ) , ~with
+ a chloro complex of Fe(III), viz.

and that the mechanism can be described in terms of an inner-sphere activated


complex of composition [(H20)5FeC1Cr(H20),4+]f.Ardon et d9'investigated
the catalysed oxidation in 5.27 Mperchloric acid at - 50 "C. Under these conditions
the process was described as instantaneous. The green product solution was
analysed by ion-exchange, revealing Cr(H2O),CI2+ to be the chief product
of reaction. Since at -50 "C the formation of Fe(H2O),CI2+ is relatively slow,
it is deduced that Fe3+ may be incorporated into the activated complex without
requiring the prior formation of Fe(H2O),CI2+. Dulz and Sutin9, have published
References pp. 267-273

180

COMPLEXES OF D I F F E R E N T METALS: OXIDATION-REDUCTION

a detailed kinetic study of the Cl--catalysed oxidation (at 25 "C) based upon the
use of flow technique^^^'. The reaction is first-order in both Fe(ll1) and Cr(II),
the second-order rate coefficient being 7.7 x lo3 l.mole-'.sec-'
for 1.00 M
HClO,. The acid dependence of the rate can be expressed as
rate = k, [Fe3+][Cr"] +k,fFeOH2'][Cr2+]
Since Kh as defined by
Fe3'+H20

+ FeOH2++H'

is much less than [H'] then


k' = rate/[Fe(III)][Cr2'] = k7+k8Kh/[H+]

A plot of k' versus l / [ H f ] is linear. Using a value9, for Kh of 1.69 x 1O-j enables
k, and k, to be calculated from the slope and intercept: k, = 2.3 x lo3 l.mole-'.
sec-' and k, = 3.3 x lo6 I.mole-'.sec-', at 25 "C and p = 1.00 M. The effect
of chloride was examined systematically as followstt. The formation and dissociation of Fe(H20)5C12' by
Fe(H20)63++c1-

ks
+
Fe(H2O),CI2+

k-9

has rate coefficients k, and k-9 of 19.4 I.mole-'.sec-' and 6.7 sec-', respectively,
at 25 "C and p = 1.00 M9,. With Cr(I1) present, the rate of disappearance of
Fe(H20),CI2+is given by
-d[FeCI2 +]/dt = k-, [FeCI'

'1 - kg[Fe3'][CI-]

+ klo[FeCl2+][Cr2'1

(5.8)

where k,, is defined by


Fe(HzO),CI2++Cr2+

kio
__f

products

It was shown that k , , is much larger than both k, and k-9 whereupon equation
(5.8) becomes
-d [FeCI2+]/dt = k,,[FeCIZ+] [Cr2']
Fe(H20)J12+ was followed spectrophotometrically at 336 mp and k,, was
t The comparable reaction between FeBr*+ and CrZ+ has been examined by Carlyle and

EspensonzsS.
t t Catalysis by chloride has been examined for the comparable Fe(II1) fNp(I1I) reactionzs9.

O X I D A T I O N SBY

181

Fe(II1) A N D Ru(II1)

obtained as 2 x lo7 I.mole-'.sec-' at 25 "C and ,u = 1.00 M . On the assumption


that there exists a CI--catalysed path for the reaction, not requiring the participation of Fe(H20)5C12+as a reactant, then the expression
-d [Fe(III)]/dt

k , [Fe3+][Cr"] + k 6 [FeOH2+][Cr2+]

+k9[Fe3+][CI- ] k , , [Fe3+][Cl- ] [Cr2+]


holds for experiments in which C1- is added only to the Cr(I1) solution. For
conditions where [Cr(II)] >> [Fe(III)] and [Cl-] >> [Fe(III)], the apparent firstorder rate coefficient is given by

Plots of k' versus [Cr"] at fixed [Cl-] are linear and allow k , , to be calculated
from the slopes as 2 . 2 ~lo4 12.mole-2.sec-', at 25 "C and p = 1.00 M . Product
yields of Cr(H,0)5C12+ and Cr(H,0),3+ obtained experimentally were in excellent agreement with those calculated on the basis of the kinetic scheme. Dulz
and S ~ t i conclude
n ~ ~ that two routes exist for the chloride-catalysed oxidation,
Cr(H20),Cl2+ being formed in both paths, viz. inner-sphere process

Fe(Hz0)5C12+ Cr(H,O),'+

+[(HzO)5Fe-CI-Cr(H20)54+1'

Fe(H20),'+ +Cr(Hz0)5C12+

and outer-sphere process, either

Fe( H20)63 Cr( H, O),Cl+


+

[( H, O),FeH, 0-Cl-Cr( H, 0)54+]*

(or [(H20)5FeH20-CrCI(H20)$+]*)
+Fe( HZO),,

+ Cr(H,0),C12

or
(H20),Fe3+C1-

+ Cr(H20)62 +[(H,0),FeH20-Cl-Cr(H20)54+]*
+

+Fe(H,O),,

+Cr( H20),C12

The rate of the path involving Fe(Hz0)5C12f as a reactant is


3000 times
faster than the alternative (outer-sphere) path. Table 4 summarises results on this
reaction and the corresponding electron exchange reactions in terms of relative
rate coefficients.
When a solution of Fe(I1I) and thiocyanate is mixed with a solution containing
excess Cr2+ in a flow apparatus44 the absorbance change can be seen to involve
three distinct stages, corresponding to (a) the very rapid reaction between
FeNCS" and Cr2+ ( k 2 2 x lo7 1.mole-'.set-'); ( b ) the reactions Fe3++Cr2+,
References pp. 267-273

182

COMPLEXES O F DIFFERENT METALS: OXIDATION-REDUCTION

TABLE 4
K L L A T I V E R A T E C O E F F I C I E N T S FOR T H E F e ( I I I ) + C r ( I l )

H2O

OH -

CI -

From rcl'. 2 3 .

I
>8 <IO4U
> 2 Y 106b

From ref. 22.

From rcf. 99.

A N D RELATED REACTIONSg6

I
1.4 4 103

1
1.1

6.2d

104

\
,

I O X L

From ref. 100.

FeOH2++Cr'+, and F e 3 + + C r 2 + + S C N - (k = 2.3 x lo3 I.mole-'.sec-',


3.3 x 10, I.mole-'.sec-', and
2 x lo5 12.mole-2.sec-', respectively); (c) the
Cr(t1)-catalysed isomerisation of CrSCN" produced in ( u ) ( k = 42 1.molesec-1)63. Rate coefficients pertain to 1 M HCIO, solutions at 25 "C. Thus an
inner-sphere mechanism is demonstrated. The S-bonded thiocyanato complex,
CrSCN", is not produced when a solution of Cr2++SCN- is oxidised by Fe(tI1).
CrSCN2+ can be prepared by the gradual addition of a 5 x
M Cr2+ solution
to an equal volume of a well-stirred solution of 5.5 x
M Fe([II) and 4.5 x
M SCN-. The product solution is green whereas CrNCS" solutions are purple.
Oxidation of the pentacyano complex of Co(II), C O ( C N ) , ~ - ,by Fe(CN)63results in thc oxidant being retained in the coordination sphere of cobalt, and the
inert binuclear ion (NC),F"CN CO"'(CN),~- is formed"'. This constitutes a
piece of direct evidence for a bridged mechanism. Similarly, reaction of C O ( C N ) ~ ~ with oxygen results in the production of (NC),Co"'OO CO"'(CN),~-.
The reduction of Fe(CN),3 - by cobalt(I1) ethylenediaminetetraacetate,
Co(EDTA)'-, yields Fe(CN),,- and Co(EDTA)- indirectly in two stagesIo2

Co(EDTA)'-

'.

+ Fe(CN)63- + CO(EDTA).F~(CN),~rapid equilibrium, K,

2Co(EDTA)- + Fe(CN),,-

C O ( E D T A).F~(CN),~-

rate-determining

Under conditions of a large excess of Co(EDTA)'- the back-reaction is negligible


and the rate is given by

where [Blolal] represents the sum of the concentrations of Fe(CN)63- and


Co(EDTA).Fe(CN),'-, and kobsis the observed (first-order) rate coefficient. In
accord with the rate law, plots of log ( A , - A , ) versus time are linear, where A

OXIDATIONS B Y

Fe(II1)

AND

Ru(III)

183

and A , are the absorbance values at 420 mp, the absorption maximum of
Fe(CN)63-, at time I and after complete reaction. Furthermore, the form of the
rate law is verified from the linearity of plots of l/kobsversus l/[Co(EDTA)'-].
Slope and intercept values of such plots yield values for k,, and K , at 25 "C of
6.2 x
sec- and 670 1.mole-', respectively. An independent value of K , (710 1.
mole- ') was obtained spectrophotometrically. The overall activation energy and
entropy are estimated to be 26 kcal.mole-' and 16 cal.deg- '.mole-', respectively.
Adamson and Gonick"' are of the opinion that the reaction sequence is more
correctly a three-stage process involving the formation of the species (EDTA)Co"NC-Fe111(CN),5- which first undergoes charge-transfer to (EDTA)Co"'-NCFe"(CN),'and then breaks down to Co"'(EDTA)- and Fe11(CN),4-. The
second of these binuclear species is taken to be the intermediate, since magnetic
susceptibility measurements reveal the intermediate to be diamagnetic. Structurally
the intermediate resembles (NC) ,CO"'-NC-F~"(CN),~- formed between
C O ( C N ) , ~ -and Fe(CN),'-'O'.
The same system has been investigated in greater
detail by Huchital and Wilkin~"~,using rapid reaction techniques to characterise
the intermediates formed. The kinetics of formation and decomposition of the
bridged cyanide intermediate, (EDTA)CO"'-NC-F~"(CN),~-

'

Co(EDTA)'-

+ Fe(CN);-

k-

(EDTA)CO"'-NC-F~"(CN),~K,

(5.9)

I3

have been studied with the aid of temperature-jump and stopped-flow equipment.
It is suggested that the intermediate is formed by replacement of water from the
Co(l1) complex, which may react as Co(EDTA)(H,O)'-.
In addition, evidence
from stopped-flow measurements is presented for the inclusion of a second intermediate, viz.

(EDTA)CO"'-NC-F~"(CN)~~+Fe(CN),3-

2 (EDTA)Co"'-NC-Fe"'
k-14

(CN),4-+Fe(CN),4-

K,

(5.10)

TABLE 5
R A T E P A R A M E TE R S F 0 R

(5.9)

1.3 105
0.9 1 0 5
/

Temperature jump method.


Refirctrces p p . 267-273

Fe(CN)63-

+Co(EDTA)'-

8 6"
b

Stopped flow method.

RE ACT I 0 N I o 3

1 . 5 103
~ =
1.6 x 103 b
1.6 x lo3

Direct spectrophotometry.

184

C 0M P L EX E S 0F D 1FFE R E N T M E T A L S : 0XI D A T I 0N-R E D U C T I 0 N

Rate parameters, at 25 "C, for reactions (5.9) and (5.10) are collected in Table 5.
Reaction (5.10) is analogous to"'
(CN),Co1"-NC-Fe"(CN),6-

+Fe(CN);-

$(CN),Co"'-NC-Fe"'

k-15

(CN),'- +Fe(CN)64- K 3
Huchital and W i l k i n ~ " ~report k15 = 1.1 x lo3 1.mole-'.sec-', k - 1 5 = 1.6 x lo4
I.mole-'.sec-l and K3 = 0.07, at p = 0.1 M and 25 "C. These results are close
to those for reaction (5.10).
The reaction between Fe(U1) and Sn(I1) in dilute perchloric acid in the presence
of chloride ions is first-order in Fe(lI1) concentrationlo4. The order is maintained
when bromide or iodide is present. The kinetic data seem to point to a fourthorder dependence on chloride ion. A minimum of three C1- ions in the activated
complex seems necessary for the reaction to proceed at a measurable rate. Bromide
and iodide show third-order dependences. The reaction is retarded by Sn(I1)
(first-order dependence) due to removal of halide ions from solution by complex
formation. Estimates are given for the formation constants of the monochloro
and monobromo Sn(I1) complexes. In terms of catalytic power I- > Br- > C1and this is also the order of decreasing ease of oxidation of the halide ion by
Fe(1lI). However, the state of complexing of Sn(1I)and Fe(I1I)is given by C1- >
Br- > I-. Apparently, electrostatic effects are not effective in deciding the rate.
For the case of chloride ions, the chief activated complex is likely to have the
composition (FeSnCl,+)*. The kinetic data cannot resolve the way in which the
C1- ions are distributed between Fe(I11) and Sn(I1).
The problem has been partially resolved in a later note by Peterson and Duke'"
describing their investigation of the reaction between Sn(I1) and the ferricinium
ion. Ferricinium perchlorate was prepared by oxidation of ferrocene with AgCIO,
in aqueous perchloric acid; from the nature of the ferricinium structure, Fe(II1)
is unlikely to complex with more than one chloride ion. The reaction, followed by
absorbance measurements on the ferricinium ion at 615 mp, is first-order in both
reactants. The chloride-ion dependence indicates a total of five C1- ions in the
activated complex, four of which are deduced to be associated with Sn(1I) as
SnCI,-.
Wetton and Higginson5' have briefly investigated the reaction between Fe(II1)
and Sn(I1). Spectrophotometric data for the rate of disappearance of Fe(II1)
were obtained at 335 mp. In the absence of substantial amounts of Fe(I1) and
Sn(IV) the kinetics in 1.0 M hydrochloric acid are of simple second order, indicating that Sn(I1) is present as the monomeric species. The addition of Fe(l1)
to the reacting system produces a pronounced retardation, the second-order plots
showing curvature after 75 "/, reaction. The appropriate mechanism is

OXIDATIONS BY

Fe(II1)

AND

Ru(II1)

185

Fe(III)+Sn(II) 3 Fe(II)+Sn(III)
k-16

Fe(II1) + Sn(II1)

Fe(I1) + Sn(1V)

This mechanism gives a quantitative fit with the observed kinetics: at 25 "C,
k , , / k - 1 6 is 1280 and k,, is 4.4 l.mole-'.sec-'. Furthermore, experiments in which
the Co(II1) complex, Co(YOH)H20, is included in the reacting system provide
evidence for the presence of Sn(II1) intermediates (H4Y = EDTA).
By means of a stopped-flow technique, Carlyle and E ~ p e n ~have
~ subn ~ ~ ~ ~ *
jected the reaction between europium(I1) and iron(II1) to a detailed examination.
In perchloric acid solution two processes are discerned, viz.

and
Fe(H,O):++Eu:l

+ H 2 0 -+-[FeEu(H,O),OH4+]*+Ht

In the presence of chloride ions there exists both an anion-catalysed reaction


F e ( H , O ) ~ + + E u+C1~~

_f

Fe(H,O)z++Eui;

+Cl-

and an inner-sphere reaction (see, also, ref. 288)

U(1V) is oxidised quantitatively to U(V1) by Fe(II1) in dilute perchloric acid


solutions, viz.
2 Fe(lII)+U(lV)

2 Fe(II)+U(VI)

The reaction, as studied by Bettslo6, was followed by measuring the amount of


Fe(I1) formed as a function of time, aliquots of the reaction mixture being
quenched by a solution of o-phenanthroline at pH 4. At constant acidity and
ionic strength the reaction is first-order in both Fe(II1) and U(IV), viz.
d[Fe(ll)]/dt = -d[Fe(IIl)]/dt = -2 d[U(IV)]/dt = 2 k'[Fe(III)][U(IV)]
when [HClO,] = 1.02 M and p = 1.02 M .
where k' is 12.4 l.mole-'.sec-'
Increase in acidity has a retarding influence: plots of log [H'] versus log k' indicate a - 1.8-order dependence on hydrogen-ion concentration. Increase of ionic
strength (by addition of NaClO,) is shown to produce a small increase in rate.
References pp. 267-273

186

COMPLEXES OF DIFFERENT METALS: OXIDATION-REDUCTION

At constant acidity both of following schemes are plausible


Fe(lll)+ U(IV) --+

Fe(ll)+ U(V)

slow

(A)

Fe(III)+ U(V) --+Fe(lI)+U(VI)

rapid

or
Fe(III)+ U(1V) -+ Fe(lI)+ U(V)

slow

(B)

2 U(V) --+ U(VI)+ U(tV)

rapid

Scheme ( B ) is considered unlikely on the grounds that, at the low concentrations


of U(V) involved, the latter would disappear by oxidation with Fe(ZI1) rather
by dismutation. The hydrogen-ion dependence suggests that the rate-controlling
step between Fe(1ll) and U(1V) can be visualised in terms of a series of competitive
reactions of hydrolysed species of both reactants viz.

* FeOH + H +
U 4 + + H 2 0 + UOH3++H+

KR

F e 3 + + 2 H,O ;= Fe(OH),++2 H +

K,

.+u ( o H ) ~+ ~2 +H +

K,

Fe3+ H 2 0

u4++ 2

H,O

K4

F e 3 + + U O H 3 +%Fe++U(V)
FeOH

+ U4+ 2Fe+ + U(V)

F e O H 2 + + U O H 3 +% Fe 2++U(V)
k21

Fe(OH),+ + U 4 + --+

Fe3+ U(OH)+

k22

Fe++U(V)

(5.11)

+ U(V)

(5.12)

+Fez+

On the assumption that the concentration of U(OH)+ and Fe(OH)+ are


negligible when compared with the total concentration of Fe(I1I) and U(IV), the
derived rate law is

so that

(5.13)

OXIDATIONS BY

Fe(II1)

AND

Ru(III)

187

where K' = k , , K 4 + k l , K h and K" = k , , K h K 4 + k 2 , K , + k , , K , . Writing D for


the denominator of the right-hand side of (5.13)

k'D

K'[H+]+K"

and plots of k'D (calculated from known values of Kh and K 4 ) against [H'] are
linear at temperature between 3.1 "C and 24.8 "C. The constants K' and K",
evaluated from the slopes and intercepts of such plots, are 2.98 sec-' and 20.6
mole.1- l.sec-', respectively, at 24.8 "C; the corresponding apparent activation
energies are 22.5 and 24.5 kcal.mole-'. If Fe(OH),+ and U(OH),'+, known to
be present in extremely low concentrations, are insignificant kinetically ( k Z 1=
k,, = 0) then
AH"

E2,+AH,,+AH4

where AHh and AH4 are the enthalpies of hydrolysis of Fe(1II) and U(IV),
respectively, and E,, is the true activation energy for FeOH2++UOH3+. On
this basis E,, = 24.2- 12.3- 10.6 = 1.3 kcal.mole-', a most unrealistic figure
for electron-transfer processes. It is concluded, therefore, that the sequence involving two OH groups is, in fact, made up of (5.11) and (5.12). Maximum rates
and activation parameters of the paths involving one OH group are given in
Table 6. The Fe(III)+ U(1V) reaction is greatly accelerated by HS04- ions"'.
This is ascribed to the participation of sulphate complexes of both reactants, e.g.
U4++ HS04-

+ US04'+ + H +

A - 1 .2-order dependence on hydrogen-ion concentration is noted. In sulphuric


acid media the possibility that U(V) is consumed by disproportionation cannot
be ruled out.
The rate law for the reduction of Fe(1II) by Np(IV), uiz.
Fe(lII)+Np(lV)

+ Fe(II)+Np(V)
TABLE 6

MAXIMUM VALUES OF

Condition

References pp. 267-273

kls

AND

k I 9 I N Fe(IIl)i-U(lV)

REACT1ONlob

Rate coeficient(l.inole- I.
sec-', 25" C )

E(kcaLmole- I )

ASt(cal.degmole- I )

107

11.9

-11.2

1860

10.2

- 11.3

I.

188

C O M P L E X E S 0 F D I F F E R E N T M ETA L S: 0 XI D A T 1 0 N-R E D U C T I O N

is reported to be
d [Np(IV)]/dt = k"[Np(V)] [Fe(II)][H' 1-k' [Np(IV)][Fe( 1II)]/[H'l3
The value. given by Huizenga and Magnusson" for k' and k" are 3.4 mole2.1-'.
min-' and 4.7 12.mole-2.min-', respectively, at 25 "C in 1.0 M perchlorate
solutions. The apparent activation energy corresponding to k' is 35 kcal.mole-'.
Nitrate complexing of Np(1V) reduces the forward rate.

5.2

OXIDATIONS BY RUTHENIUM(III)

Ru(NH,),~+, Ru(NH3),CI2+, Ru(H2O),CI2+, and C~~-RU(NH,),CL,~


are
reduced by Cr2+ in perchlorate media'08,291-293or p-toluenesulphonic acid
media293. Chloride ion strongly catalyses the former reaction"' according to
d[Ru(II)]/dt

( k +k'[Cl-])[Ru(

] [Cr2+1

At p = 0.022 My k = 28 l.mole-'.sec-',
k' = 6 x 10' 12.mole-'.sec-'.
exchange experiments indicate that CrCI2+ is the major product species.

Ion-

6. Oxidations by cobalt(II1)
A great deal of attention has been given t o the oxidation-reduction reactions

of cobalt(Ir1). For convenience this section is subdivided into three parts: reactions involving inorganic bridging ligands (some outer-sphere systems are discussed also for comparison), reactions involving organic bridging ligands, and reactions of aquo complexes.

6.1

INORGANIC BRIDGING

LIGANDS I N OXIDATIONS BY

COBALT(III)

COMPLEXES

The oxidation of Cr2+ with substitution-inert complexes of the type


lo9.
CO(NH,)~X'+ has been the subject of a pioneering study by Taube et
The range of complexes studied.are those for which X = C1-, Br-, I-. F-,
H 2 0 and NH,. The most significant fact to emerge from such reactions
is that they are accompanied by quantitative transfer of X to the reductant.
When X is CI-, no exchange occurs with chloride ion in solution during the
course of reaction and it is surmised that an inner-sphere activated complex
containing a Cr-C1-Co bridge is involved, i.e., [(NH3)5Co-X-Cr4+]*. Earley

OXIDATIONS BY

CO(II1)

189

and Gorbitz'", confirming the earlier results of Taube"', have shown that
the principal path of reduction of Co(NH3),ClZ+ by Cr2+ in the presence of
pyrophosphate yields product Cr(I1I) incorporating both chloride and pyrophosphate.
The rates of oxidation of V2+ by complexes of the type CO(NH,)~X,where
X = HzO, NH, and C1-, have been examined by Zwickel and Taube1I2 in HzO
and in D,O solution. These workers have compared the results on such systems
with data on the CrZ++Co(NH,)63+reaction in HzO and D,O. The product
V 3 + , unlike Cr3+, is substitution-labile. Consequently, a different approach is
necessary for V2+ oxidations than is customary for CrZ+ oxidations. The specific
rate coefficients for Co(NH,),,
V2 and CO(NH,)~,++CrZ+ vary with CIconcentration according to
+

kobs = k k"C1-l
The rate coefficients, k and k', and the corresponding activation parameters are
given in Tables 7 and 8. Cr(H2O),Cl2+ is the primary product in the chloridedependent path for the co(NH3),,+ Cr2+ reaction. The specific rate coefficient
for the Co(NH3),CIzf +V2+ reaction was determined as 342 I.mole-'.min-' at
p = 1.00 M and 25 "C. In general, reactions of V2+ contrast sharply with those
of Cr2+ (except for CO(NH,),~+ CrZ+)and Cr(bipy),'+ by exhibiting chloride

TABLE 7
R A T E D A T A F O R C O ( N H J ) ~ ~ R+E A C T I O N S 1 1 2

Temp.("C)

Reductant

Cr2

k'(12.mole-2.min-')

0.0053
0.014
0.01 1"
0.22
0.41
0.24"

0.74
1.70
1.35"
1.27
3.27
1.95"

25.0
37.0
37.0
25.0
37.0
37.0

V2+

Ionic strength 0.40 M .

k(l.mole-'.min-')

In 100 % DzO.
TABLE 8

A C T I V A T I O N P A R A M E T E R S F O R C O ( N H J ) ~ ~R+E A C T I O N S L L Z

Reductant

Path

A H t (kcalmole- l )

Cr2
Cr2
V2+
V2+

k
k'
k

14.7
12.4
9.1
14.1

References pp. 26 7-273

k'

AS: (caLdeg - '.mole- ')


- 30
-25
-40

-20

TABLE 9
_

S E C O N D - O R D E R R A T E C O E F F I C I E N T S (I.mole-'.sec-'
~

C r z +a

Oxidant

8 . 9 1~0 - 5
CO(NH~)~'+
0.5
Co(NH3)50Hz3+
1.5 x lo6
CO(NH~)~OH'+
9 xi05
Co(NH 3 ) F2
2.6 x lo6
Co(NH &I2+
> 2 XI06
Co(NH3),Br2+
> 2 x106
Co(NH3)512+
C O ( N H ~ ) ~ N ~ ~ + 3 xi05
19
Co(NH,),NCS'+
18
CO(NH~)~SO~+
0.18'
Co (N H 3 ) OAc2
90
CO(NH~)~NO~''
Co(NH3)5S203+
Co(NH 3 ) ,(maleate)
4.8 x 109
Co(NH3) d'04
8.3 x 103
Co(NH3)5P04H
C O ( N H ~ ) ~ P O ~ H ~ ~0.3
+
C O ( N H ~ ) ~ P O ~ H ~ ~0.3
+
-2 ~ 1 0 - 5
C~(en)~~

3 . 7 10-3
~

d,=

0.5'
-

0.3
7.8'

0.43'

2 x10-2
0.15
2.6 x 104
3.9 x 102
2.5 x lo2
1.2x102
1.9 x 10'
0.7
1 . 4 ~10'

0.18'
-1

x 10'

I .4 x 107
1 . 6 lo2
~
2.3
4.5
d

COMPLEXES

cr(bipu)32'

'RU(NH3)S2+ '

Fe2+ a

m
X
m
CA

13

Co(l11)

zP
P

2.6
-5
25
1.2 x 102

THE REDUCTION OF VARIOUS

Eli2+ a

y2+ a

at 25 " C ) F O R

-2

XIO-4

6.9 x loz
5 xi04

1.1 x 10-2
3.0
4 x10-2

1.8 x 103
2.6 x lo2
I .6 x 103
6.7 x 103
1.2

8 x105
5 XI06
-

4.1 x 104
1 . 0 104
~
4.5 x 104
1 . 2 103
~
8 xi04
L O X 103

6.6~
1.3 x 10-3
7.3 x 10-4
-

8.8 x
<3

XIO-6

<5

XIO-5

tJ

5 XIO'
6
-3
-5

XIO-3

1.8 x 103

tJ
C

Data from refs. 113, 125 and 126, unless otherwise specified.
0.2 M.
,u = 0.1 M, C104- medium. ,LL
1.0 M , C104- medium, unless otherwise noted.
from ref. 114. 8 from ref. 127.
from ref. 128.
from ref. 122. j from ref. 129. Ir from ref. 130.

a /A =

n
/I =

0.4 M.

from refs. 112 and 121.

=!
0
z

191

CO(1II)

OXIDATIONS BY

catalysis'. The rate data on the CO(NH,),(H,O)~++ V 2 + reaction are marred by


pronounced scatter of results. Previous arguments' used to explain the differences
in the C O ( N H ~ ) , ( H ~ O+)C
~ r+2 + and Fe3++Cr'+ reactions, based upon variations in the lability of Co(II1) and Fe(III), are considered doubtful. Zwickel and
Taube"' conclude that oxidations of V2+ by Co(111) complexes are more likely
to proceed by outer-sphere routest?. These authors speculate on the intimate nature
of the processes involved in the transfer of an electron from the reductant to the
Co(II1) centre.
Candlin et ~ 1 . ' 'have
~
reported the results of kinetic investigations on the reduction of various pentaamminecobalt(II1) complexes by C r 2 + ,V 2 + , Eu*+ and
C r ( b i ~ y ) ~ ' +In. most cases a stopped-flow apparatus44 was used, the rate of disappearance of the Co(II1) complex being measured at
500 mp. Data on
C r ( b i ~ y ) , ~reduction
+
were obtained by following its decrease in absorbance at
562 mp. Oxygen was eliminated during all kinetic experiments. The flow technique
sec to 30 sec. Second-order
could measure half-lives within the range 5 x
rate coefficients and activation parameters are given in Tables 9 and 10. Rate
coefficients for reduction of phosphatopentaamminecobalt(I1I) complexes are
recorded also; the order of reactivity with C r 2 + ,V 2 + a nd EuZ+is Co(NH3),P04 >
Co(NH3),P04Hf > Co(NH3),PO4HZ2+ CO(NH,),PO,H~~+. The innersphere character of Cr2+ reductions is recognisable from the substantial effects
caused by variation of the nature of ligand X in the oxidant. All the oxidants
studied, excluding CO(NH,),~+ and C ~ ( e n ) ~ , which
+
react by outer-sphere
processes, react with Cr2+ through the intervention of a bridging group. Reductions by Cr(bipy)," are relatively insensitive to the nature of X and are design-

T A B L E 10
A C T I V A T I O N P A R A M E T E R S FOR R E D U C T I O N O F V A R I O U S
.

~~~~

Cr2+

AH:

Oxidant
+

14.7"
2.9b
-

ASZ

References p p . 267-273

AH:

- 30"

9.1"

- 52b
-

6.2
6.9

- 32

9.1
(4.6)
11.7
6.1'

-29

A H : in kcal.mole-'; AS$ in cal.deg-'.mole-'.


122.
+ See, also, ref. 294.
t t See, however, ref. 295.

Eu2

AS:

~~

V2+

~-

CO(NH~)~NH~~
CO(NH~)~OH~~+
Co(NH3)SCI2+
CO(NH~)~B~'+
Co(NH3)512
CO(NH~)~N~~+
CO(NH3) 5 S 0 4
Co (N H 3 ) NCS2

CO(II1) C O M P L E X E S 1 1 3

From ref. 112.

-40"
-

-22
(- 32)
- 14
-34'

5.0
4.7

- 30

-32
-

5.5
6.1

-30
-26

From ref. 121.

From ref.

192

COMPLEXES OF D I F F E R E N TMETALS: OXIDATION-REDUCTION

o -Cr(biPY)j2*
3

O-V2'

-g

mercury electrode

:I

-2

- 3 t C i ,

i , i ,

-4
C o ( e n ) p ROAc2+ RNCS2* RS04*
RNHJ3*

RF2'

RCIZi

R O H P RN?'

R12*
RBr2+

Fig. 5. Relative rate coefficients (kRx/kROH,j+)


at 25 "C, referred to the rate for ROHz3+ [R =
Co(NH,),], for the reduction of various Co(II1) complexes arranged in approximate order of
increasing reactivity. Ionic strengths: Cr(bipy),* +,0.1 M ; V 2 + , 1.0 M, dropping mercury
electrode (from data of VIEek1l6), 0.4 M . (From Cundlin et uL113,by courtesy of The American
Chemical Society.)

ated as outer-sphere reactions (as originally suggested by Zwickel and Taube114).


Similarly, the kinetic data of Table 9 indicates that reactions of V2+ are probably
outer-sphere. The close resemblance between V2+ and Cr(bipy),'+, as regards
patterns of reactivity, is clear from Fig. 5 and both show a correlation with the
rates of electrochemical reduction of the Co(ll1) complexes as measured by
VlEek using a dropping mercury electrode"'. Vl15ek"~>'~~
has drawn attention
to the fact that the order of increasing reactivity of the Co(II1) complexes is very
nearly the same as the order of decreasing excitation energy, t2,6 --t f2g5eg.The
latter is evident from the shift in A,, values with changes in field strength of ligand
X. Reductions by Eu(I1) are difficult to categorise: the pattern of reactivity towards halogenocomplexes is Co(NH3)'F2+ > Co(NH,),Cl2+ > Co(NH,)'Br2+
> C O ( N H , ) ~ I ~the
+ , opposite of that shown byV2+,C r ( b i ~ y ) , ~and
+ CO(CN)~,-.
Candlin et al.' l 3 conclude that Eu(1I) reductions proceed predominantly by an
inner-sphere route. These authors comment also on the dangers inherent in assuming that reactions of even closely-related oxidants with a common reductant
proceed by a common mechanism, particularly in the light of evidence to show
that reduction of Co(NH,),X complexes by Co(CN),,- may occur through both
inner- and outer-sphere
and that reduction by V2+ of complexes

OXIDATIONS BY

193

cO(III)

containing certain conjugated organic ligands takes place by remote attack on the
ligand"9, I2O
Diebler and Taubel2, have quoted rate parameters for the reduction of
halopentaamminecobalt(I1I) complexes by Fe(II) and V(II). Rate coeffients for
V 2 + reductions are in reasonable agreement with those reported by Candlin
et ~ 1 . ' '(except
~
for C O ( N H , ) ~ B ~ ~So
+ )also
. are Fe2+ reductions, taking into account differences in conditions. At 25.5 "C and p = 1.7 M , the second-order rate
coefficients for Fez+ reduction of Co(NH3),F2+, Co(NH3),CI2+ and
C O ( N H , ) ~ B ~ are
~ + 7.6 x lo-,, 1.6 x lo-,, and 0.92 x
I.mole-'.sec-',
respectively (compare results in Table 9). The respective values of AH* are 13.4,
14.5 and 15.6 kcal.mole-', and of AS* are -23, -23 and -20 cal.deg-'.mole-'.
The observed reactivity pattern F- > CI- > Br- is the reverse of the pattern
encountered in the oxidation of Cr2+ with CrX2+ (ref. 99) and Cr(NH,),X2+
(ref. 124) and in the reduction of CO(NH,)~X'+ by C r ( b i ~ y ) , ~ +
(ref. 114),
R u ( N H ~ ) (ref.
~ ~ +125) and V2+ (ref. 113)+.However, the pattern F- > C1- > Brapplies also in the Co(NH3),X2++Eu2+ ~ y ste m "~ . In the case of the
Co(NH,),X2+ + Fe2+ systems there seems to be a correlation between rate and
the thermodynamic stability of the complexes of Fe3+ with X-: the relevant
stability constants are 1.5 x lo5, 4 and < 0.1 for FeF", FeCI2+and FeBr2+,
respectively. The reactions are assisted by free halide ions, as shown by

rate

k'[FeZf][Co(NH3),X2+][XJ

When CO(NH,)~F'+ is the oxidant, values of k' are 3 . 5 ~


2.1 x lo-'
and 13 12.mole-2.sec-' for X- = Br-, C1- and F - , respectively. With
Co(NH,),Br2+, values of k' are 5 1 x lo-, and412.mole-2.sec-' for X- = C1and F - , respectively. These results apply for p = 1.7 M and 25.8 "C except
for X- = F- where the temperature is 25.1 "C and p = 2.0 M . Since the ratio
of k' for F- compared to k' for CI- is considerably different for CO(NH,)~F'+
and Co(NH3),Br2+ (6 x lo2 as opposed to > 4 x lo3) the stabilization order of
Fe3+ by the halide cannot be invoked t o explain the greater effect of F- as
compared to C1-.
Espenson'26 has made a study of reductions of acidopentaamminecobaIt(I1I)
ions by Fe2+, viz.
CO(NH,)~X'++ FeZf + 5 H +

Co2++ F e 3 + + X - + 5 NH,'

It proved necessary to correct the rate data in the case of X = Cl- and Brfor the simultaneous aquation of the complex. Rate coefficients are essentially
in agreement with those given in Table 9. Activation parameters are quoted when
X = F - , CI-, Br-, and also for CO(C,O,),~-.
t Parker and E ~ p e n s o n 'find
~ ~ the pattern BrReferences p p . 26 7-273

> CI- > F -

in similar reductions by Cu'

194

COMPLEXES 0 F D 1FFE R E N T MET A L S: OX I D A T I ON-R E D U CT I 0 N

Candlin and Halpern comment that the sequence of rapid rates observed
for Cr2+ as a reductant (i.e. C O ( N H ~ ) ~ I >+ CO(NH,)~B~+
> CO(NH,)~CI+
> Co(NH,),F+) is contrary to that found for the slow reactions of Fez+
(ref. 126) and Eu (ref. 113). All three reductants would appear to favour innersphere mechanisms, but in the case of Fez and Eu the order of reactivity seems
to be connected with the stability of the product halide complex (FeX or
EuX2+) which increases in the order X = I- to X = F-. Or in other words, as
pointed out by Halpern and Rabani131, in the generalised inner-sphere reaction
Co-X

+red +Co +X-red

it is reasonable to suppose that the order of reactivity, as X is varied, will depend


on variations in the strength of the bond being broken (Co-X) and the bond being
formed (X-red). The strength of both bonds increases in the order I- < Br- <
C1- < F-. The order of reactivity of various reductants (when X = C1-) is
H > CO(CN),~-> CrZ+ > Eu2+ > Fez+. In the case of mild reductants
(for example, EuZ+ and Fe) the formation of a bond to the reductant in the
transition state is considered more important than bond breaking. Thus the
reactivity order will be C O ( N H ~ ) ~ F >+ C O ( N H ~ ) ~ C ~> +
Co(NH3),BrZ+ >
Co(NH3),12+ (see, however, Adin and S y k e ~on~CrX2++EuZ+,
~
p. 169). For
more reactive reductants (e.g. hydrogen atoms, Cr2+ and CO(CN),~-)bondbreaking is more important in the transition state than bond-making so that the
reactivity order is decided by the order of bond strengths, i.e., CO(NH,)~I+ >
Co(NH,),Br+ > CO(NH~)~CI+
> CO(NH~)~F+.
The oxygen-(I) and sulphur-bonded(I1) isomers of thiosulphatopentaamminecobalt(1 11)

0
I
(NH3) ~CO-0-S-S
I

0
I

(NH3)sCo-S-S-0

0
I1

are of comparable ~tability~.However, isomer I is reduced by Cr+ 70 times


more rapidly than isomer 11. The rate of reduction of I resembles the rates of
reduction of the ~ u l p h a t o and
~ sulphito complexes (Table 11).
Miller el al.133have examined in detail the relative efficiencies of oxoanions
as bridging groups by means of a study of the rates of reaction of Cr(II), V(II),
Eu(I1) and Ti(II1) with oxoanion complexes of pentaammine- and tetraamminecobalt(lI1). The complexes used include metaborato, carbonato, nitro, nitrito,
nitrato, sulphito, sulphato, aquosulphato, thiosulphato, selinito, selanato and
phosphato. The rates of reduction correlate with the position in the periodic
table of the central atom of the oxoanion group: the rate increases from Group

195

CO(II1)

OXIDATIONS BY

TABLE 11
REDUCTION OF

C O ( N H ~ ) ~CX
OMPLEXES

BY

k(l.rnole-'.sec-', at 25 " C ) AHX(kcal.mole-')

so42so32-

0-bonded S 2 0 3 2 S-bonded SZOo2-

Cr2+ (from ref. 132)

AS:(cal.deg- ' m o l e - ' )

ReJ

-25
-26
-39
3

113
132
132
132

8.3
8.3
4.2
24.6

18
18.6
13.3
0.18

I11 to a maximum value at Group V, and then decreases. Furthermore, for a given
ligand, the rate increases as the atomic weight of the central atom increases.
The reduction of cyanopentaamminecobaIt(I1I) by Cr2+ in acidic perchlorate
297
media proceeds in two distinct
Co(NH3),CN2++Cr2++5H +

_f

Coz++CrNC2++5NH4+

CrNC2++CrCN2+

(6.1)
(6.2)

The oxidation-reduction stage (6.1) has a second-order rate coefficient of 22.8 1.


mole-'.sec-' (p = 0.15 M ) at 15 "C. The linkage isomerisation stage (6.2) is
much slower with a first-order rate coefficient of 1.0 x lo-, sec-' in the absence
of CrZ+at 15 "C, [H'] = 0.4 M , p = 1.0 M Z g 7 .Since the rate of this stage is
dependent on the concentration of Cr2+2 9 7 , the isomerisation occurs also by the
step
CrNC2++CrZ+

_f

Cr2++CrCN2+

(6.3)

The species CrCNz+ was characterised by means of its behaviour to ion-exchange,


its absorption spectrum, and analysis of the CN-/Cr ratio in the separated complex. The existence of the species CrNCZ+was less clearly established from its
absorption spectra.
The rates of reduction of cis- and trans-Co(NH,),(N,),+ and CO(NH,),N,~'
by Fez+ in aqueous perchloric acid have been studied by Haimlz8. The cis
isomer of the former complex reacts at an acid-independent rate whereas the
rate for the trans isomer is acid-dependent. The respective rate laws at 25 "C are
rate = 11.1 [Fez+][cis-Co(NH,),(N,),

*]

and
rate

(4.4+ 82[H+])[FeZ+]
[trans-Co(NH,),(N,),+ ]

where time is expressed in min. CO(NH,),N,~+ reacts at an acid-independent


rate given by
rate = 0.52 [Fe2+][Co(NH3),N32+]
References pp. 26 7-2 73

196

C O M P L E X E S O F D I F F E R E N TM E T A L S : O X I D A T I O N - R E D U C T I O N

The order of reactivity towards Fe2+ is given by cis-Co(NH,),(N,),+


>
trans-Co(NH,),(N,),+ >> C O ( N H , ) ~ N , ~ ' . In view of the lability of Fe(II1)
to substitution it is not possible to decide whether these reactions proceed by
inner- or outer-sphere routes. But the findings can be explained readily on the
basis of an inner-sphere mechanism. The suggestion is made'35 that cis-Co(NH,),(N3)2+ reacts with Fe2+ through a double-bridged activated complex analogous
to that formed between Cr(N3)2+ and Cr2+. Interpretation of the relatively high
reactivity of tran~-Co(NH,)~(N,),+in terms of a trans effect 13' receives support
from the observed acid catalysis which probably arises from the fact that removal
of N,- rrans to the bridging N,- is aided by attachment of a proton (see also
ref. 137).
With a view to determining the equilibrium constant for the isomerisation, the
rates of reduction of an equilibrium mixture of cis-and trans-C0(NH,),(0H,)N~~
with Fe2+ have been measured by Haim'38. At Fe2+ concentrations above
1.5 x lo-, M the reaction with Fe2+ is too rapid for equilibrium to be established
between cis and trans isomers, and two rates are observed. For Fe2+ concentrations below 1 x
M , however, equilibrium between cis and trans forms is
maintainzd and only one rate is observed. Detailed analysis of the rate data
yields the individual rate coefficients for the reduction of the tram and cis isomers
by Fe2+ (24 1.mole-'.set-' and 0.355 1.mole-'.sec-') as well as the rate coefficient and equilibrium constant for the cis to trans isomerisation (1.42 x lop3sec-'
and 0.22, respectively). All these results apply at perchlorate concentrations of
0.50 M and at 25 "C. Rate coefficients for the reduction of various azidoamminecobalt(lI1) complexes are collected in Table 12. Haim138 discusses the implications of these results on the basis that all these systems make use of azide bridges*.
The effect of substitution in Co(1lI) by a non-bridging ligand is remarkable
in terms of reactivity towards Fez+. The order of reactivity, transCO(NH,),(OH,)N,~' > t r a n ~ - C o ( N H , ) ~ ( N ,>
) ~CO(NH,),N,~+,
~
is at va+

T A B L E 12
RATE

COEFFICIENTS FOR

THE

REDUCTION OF

Co(ll1)

AZIDE

COMPLEXES BY

Fez+

(25 0C)138
k (Lrnole- .set- ')

Complex

Co(NH3)5N3"
rr~ns-Co(NH~)~(N~)~
cis-Co(N H 3)4(N3)2 +
cis-Co(N H3)4(0HZ)N32
~~U~S-CO(NH~)~(N~H)N~'
rr~ns-Co(NH3)4(0HZ)N3'
+

0.0057
0.0733
0.185
0.355
> 1.37
24

* A direct test of this mechanism is awaited. Haim'38 notes that the primary Fe(II1) product
of reduction of t r u n s - C 0 ( N H ~ ) ~ ( 0 H ~ )should
N ~ ~ + be detectable using a suitable flow apparatus.

OXIDATIONS BY

197

CO(lI1)

riance with Orgel's ~ u g g e s t i o n 'that


~ ~ the smaller the ligand field strength of the
group trans to the bridging ligand (N3- < H,O < NH, is accepted as the order
of field strength) the higher the rate. Instead, this behaviour can be rationalised
on the grounde that electron transfer necessitates the movement of both the
bridging ligand and the group trans to it away from the Co(1II) centre (see also
is more reactive than transref. 128). Thus ~~u~s-CO(NH,),(OH,)N,~+
Co(NH,),(N,),+
because H,O is more easily removed from Co(1II) than is
N3-.
Reductions of various Co(1II) complexes by Fe(I1) have been studied under
high pressures' '. The motivation for performing such experiments resides in the
possibility that the volume of activation (AV'), like the entropy of activation,
might be a criterion for distinguishing between inner- and outer-sphere reactions.
For reactions of the type
( N H 3 ) , C o X Z f + F e Z + + 5H +

C o 2 + + F e 3 + + X - + 5 NH,'

the formation of the activated complex on an inner-sphere mechanism, unlike


that for an outer-sphere route, is accompanied by the release of a water molecule
from the coordination shell of the reductant to the rolvent, viz.
(NH3),CoX2++ Fe(H,O),'+

71 [(NH,),CO-X-F~(H,O),~+]'

H,O
inner-sphere
li [(NH,),CoX H,O Fe(H,0),4+]*
outer-sphere

On this basis AV' should be more positive for an inner-sphere than for an outersphere reaction since a water molecule occupies a greater volume in the liquid
phase than if it is coordinated. Second-order rate coefficients were determined
at various pressures in the range 0.001 to 3.5 kbars, the rate decreasing with increase in pressure. The apparatus used was a modification of that first described
by Osborn and Whalley14'. Values of AV' were calculated from the slopes of
plots of log k versus pressure, since
-d log, kldP

AVf/RT

The results obtained by Candlin and H a l ~ e r n are


' ~ ~given in Table 13; in all
cases, it is seen that AV' is positive. These results strongly suggest that the Fe(I1)
reductions proceed by inner-sphere routes. However, to be convincing the method
requires calibration by reactions of known mechanism.
Endicott and Taube", consider that there is cause for doubt over the generallyheld views that Cr(bipy),'+ is oxidised by an outer-sphere rnechani~m"~.They
suggest that, since the complex is very labile to substitution, coordination sites
References

pp.

267-273

198

COMPLEXES O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

T A B L E 13
VOLUMES OF ACTIVATION

(Avt) F O R

cO(III)

REDUCTION OF VARIOUS

COMPLEXES BY

~e(11)13
9

A V~(cm3.rnole-')

Complex

+I1
+8
+8

+ 14
++2.2
14
+2.8
+3
HC104 media; temp., 25 "C except when noted. a At 35 "C.
HC104.
At 20 "C; Y4- = ethylenediaminetetraacetate.

0.001 M HC104.

0.02 M

could well be exposed by the opening-up of a chelate ring. The complex


Ru(NH,)~'+ is a better choice for an outer-sphere reductant since it is substitution-inert and the oxidation product is R u ( N H ~ ) ~ Endicott
~+.
and Taubelz5
have surveyed the reaction of this reagent with a variety of pentaamminecobalt(111) complexest. The data given in Table 9 (p. 190) displays the parallelism
between Cr(bipy),'+ and R u ( N H , ) ~ ~reductions
+
as regards general patterns of
reactivity, e . g . C O ( N H , ) ~ O H , ~is+ more reactive than Cr(NH3)50H2+towards
both reagents. These results imply that the two reagents react by a common
mechanism. Furthermore, there is a resemblance in the rates of reduction of the
Co(lI1) complexes at the dropping mercury ele~trode"~and by R u ( N H , ) ~ ~ + .
Outer-sphere routes may operate for reduction of C O ( N H ~ ) ~ N H , ~ + ,
C O ( N H , ) ~ O H , ~and
+ CO(NH,)~CI'+ by V'+. The pattern of reactivity for
Cr" is quite different than that of the other reductants: with the exception of
Co(NH3)63+,CrZ+ reacts via an inner-sphere mechanism. It is significant that
C O ( N H , ) ~ O H ~is+reduced much more rapidly than C O ( N H , ) ~ O H ~From
~+. a
restricted study of Cu+ reductions it appears likely that this reductant favours
an outer-sphere process. Taube et
and Pate1 and E n d i ~ o t discuss
t ~ ~ ~ the
results from the standpoint of the Marcus theory of electron-transfer reactions43.
Endicott and Taube' O8 have investigated salt effects in the reactions of
halogenopentaamminecobalt(lI1) complexes with R u ( N H , ) ~ ~ i.e.
+,
~

C O ( N H , ) , B ~ ~ + + R U ( N H , +) ~5 ~H+
+ =
C O ~ + + R U ( N H , ) ~NH,++Br~++~

(k')

C ~ ( ~ ~ ) , C ~ , ~ + R U ( N H ,H+
) ~ '=+ + ~
C O ~ + + R U ( N H , )+~2~enH+
+
+ 2 cl-

(k")

t Meyer and TaubeZg8havesimilarlyinvestigated reduction of Fe(l1I)complexes by Ru(NH3)6'+.

OXIDATIONS BY

cO(III)

199

At ionic strengths less than 0.015 M


log k' = 4.4 p'+log

kb

and
log k"

2.5 +'p

log k y

In the case of the reduction of the complex C O ( N H ~ ) ~ I 'by


+ Ru(NH,)~'+ the
kinetic concentration of Ru(NH,),12+ is higher than the estimated equilibrium
concentration, and thus the reaction would appear to occur by direct group
transfer.
The reduction of a variety of cis- and trans-chlorobis(ethylenediamine)cobalt(HI)complexes (of the type C O ( ~ ~ ) ~ X C Iby
" +Fe2+
)
have been examined by
Benson and HaimI4lt. These authors summarise the available information concerning the nature of the bridging ligand for complexes of this type. When
X = NH,, C1- must be the bridging ligand. When X = SCN-, since
Co(NH3),C12+ reacts very much faster'26 than Co(NH3),NCS2+, again C1must act as the bridge. When X = H,O, C1- bridging must occur because
Co(NH,),Cl'+ is much more reactive than C O ( N H , ) ~ O H ~For
~ + the
. case where
X = N3-, it appears that N,- functions as the bridging ligand for the following
reasons: (u) Co(NH3),N,'+
is reduced by Fez+ more rapidly than is
Co(NH,),Cl'+, (refs. 126, 128, 138) i.e., N3- is more efficient than C1- when
NH, is in a trans position to the bridging group; ( b ) replacement of NH3 by N3in the trans position of CO(NH,),N,~" produces a much less marked effect in
reactivity than does the replacement of NH, by CI- in the trans position of
Co(en),NH,Cl'+, and therefore if C1- and N3- function similarly as trans
ligands in reduction of trans-Co(en),N,Cl+ then the presence of a trans C1should help attack at N3-. When X = Br-, it is likely that both C1- and Brare involved in bridging. The nature of the bridging group may be different for
reduction of cis and trans isomers particularly when X and C1- have similar
abilities for bridging. It is clear, however, that for those cases where X = H20,
NH3 or SCN- the bridge is formed by Cl- irrespective of whether X and C1are cis or trans to one another. Benson and HaimI4l interpret their results (Table
14) for reductions of those Co(en),XCI"+ complexes, which make use of the same
bridge, on a model which assigns relative reactivities to two factors: (a) the ligand
field strength of the group trans to the bridge'36, and (6) the energy required to
stretch the metal-ligand bond along the z axis'. The latter factor is of importance
since such stretching will bring about a lowering in the energy of the dz2orbital
of the Co(II1) centre, and therefore increase the availability of the orbital to an
incoming electron. The general lack of information on force constants makes
+ Compare ref. 300.
Refereirres pp. 26 7-273

200

COMPLEXES OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

T A B L E 14
COMPARISON

OF

THE

R A T E OF

REDUCTION

O F SOME

CHROMIUM(I1) A N D IRON(I1) A T

CO(en)2XY"+
2 5 0c64

COMPLEXES

RY

Relative rate coeficients


Oxidant

Reductarit

Co(en)zNH3(NCS)Z
+
Co(en)z(NCS)2
Co(en)20H2(NCS)Z+
Co(en)2NH3C12+
Co(en),(NCS)CI+ a
Co(en)zOHZCIZ+
+

Cr2
Cr2
Cr2
Fez
Fez
Fez

trans-Xlcis-X
1.2
9 xlo-1
3 '<I01
3.7
8 x10-1
5.2 x 10'

cis-Xlcis-NH3

1 .o
4.8
1.5 x 101
1 .o
10
2.6 x 10'

trans-Xltraris-N H 3
1.o

3.7
3.1 x 10'
1 .o
2

3.6 x lo3

' Data from ref. 141.


direct comparisons difficult. It should be noted that measurements on nitrogen
isotopic-fractionation factors for the reduction of various Co(1II) ammine
complexes with Cr2+ have not revealed stretching of Co-N bonds'42. These
results are to be expected in the light of the small difference in the bond distances
in Co(l1) and Co(l1l) am mine^'^^. However, De Chant and Hunt'44 have shown
that substantial distortion of the bond to the trans ligand occurs in the reaction
of trans-Cr(NH,),(OH2)C12+ with Cr2+. For the trans series of Co(en),XCI"+
the order of reactivity towards Fe2+ is X = H 2 0 > Br- > C1- > SCN- > NH,
(overall change in reactivity of 3600). With the exception of X = H 2 0 , this
coincides with the order expected in the light of relative ligand field strengths and
ease of removal of the trans groups away from the Co(II1) centre. For the cis series
of complexes the order of reactivity is C1- > H 2 0 > SCN- > NH, (overall
change in reactivity of 90). Since, in all these cases, the group trans to the bridging
ligand is ethylenediamine, then reactivity should be little affected by modification
of the cis group. Any slight differences must arise from other factors than those
considered to operate for trans isomers.
The rates of reaction of Cr(I1) witha number of cis-and trans-Co(en),(NCS)X"+
complexes (where X = H 2 0 , NH,, CI- and SCN-) have been examined by Haim
and S ~ t i nusing
~ ~ a flow apparatus44. Quantitative transfer of thiocyanate
from cobalt to chromium is observed for all complexes except cis- and transCo(en),(NCS)CI+. These chloro complexes react via a chloride-bridged activated
complex, the other complexes by thiocyanate-bridged activated complexes. The
reduction of tran~-Co(en),OH,(NCS)~+by excess Cr2+ proceeds in two stages.
Firstly, reduction of Co(Il1) takes place to give CrSCN", and this is followed
by the chromium(l1)-catalysed isomerisation of the latter to give CrNCS"
( k = 42 I.mole-'.sec-' at 25 "C in 1 M HC104: see also ref. 63). The overall
rate increases with decreasing acidity, the observed rate coefficient being given by

OXIDATIONS BY

20 1

cO(llI)

where k , and k, are the rate coefficients for rrans-C~(en),OH,(NCS)~+


+Cr2+
and trans-Co(en),(NCS)OH+ Cr2+,respectively, and Kh is the equilibrium constant for

tr~ns-Co(en),OH,(NCS)~~ trans-Co(en),(NCS)OH+

+H+

From a linear plot of k' versus I/[H+], k , = 1 . 4 lo3


~ I.rnole-'.sec-' and
k 2 K h = 6.6 sec-'. Using a value of
5x
mole.l-' for Kh, k, is estimated
as
1 x lo7 I.mole-'.sec-' at 25 "C. A decision as to whether SCN- or OH-

acts as a bridging group is possible by considering the influence of acidity on


thiocyanate transfer. On the assumption that the aquo and the hydroxo complexes
use thiocyanate and hydroxide bridges, respectively, viz.
t r ~ n s - C o ( e n ) , O H , ( N C S )+~C~r 2 + -% (Co-NCS-Cr)'

+CrSCN2+

+ CrNCS2++ Co( 11)


H
I

trans-Co(en),(NCS)OH+ + C r 2 + 3 (Co-0-Cr)* -+ Cr3++Co(II)


then the ratio ([CrSCN2+]+[CrNCS2+])/[Cr3+]is equal to k, [H+]/k,K, =
2.1 x 10, [H']. This result is compatible with the value 1 . 4 10,
~ [H'] derived
experimentally by analysis of the products. CrSCN", formed in the reduction
of truns-C~(en),OH,(NCS)~+,is believed to be formed from the attack of Cr2+
on the S atom of the thiocyanate group. On the other hand, CrNCSZf is probably
formed by adjacent attack on the thiocyanate ligand, i.e. attack by Cr2+ on the
N atom. In Table 14 a comparison is made between the rates of reduction of
Co(en),(NCS)X"+ by Cr(I1) and Co(en),CIX"+ by Fe(l1). The following points
are noteworthy: (a) when X = NH, or SCN- there are only slight differences
between the rates for reduction of cis and trans isomers, (6) substitution of SCNfor NH, in either cis or trans positions produces only slight increases in rate, (c)
substitution of H,O for NH, produces a substantial rate increase, particularly for
trans isomers. The overall similarities in the reactivity patterns for Fe(I1) and
Cr(1I) are additional evidence for supposing that Fe(I1) reductions use chloride
bridges. Table 15 shows, for a number of related reactions, acomparison between
the rate coefficients observed and those calculated from the Marcus equation
(eqn. (12.6) p. 247) modified to

1, where, for example, k,,, and K1,, are the rate coefficient
assumingf,,,lf,,,
and equilibrium constant for the C O ( N H , ) ~ C I ~++Cr2+ reaction, k,, and K1,3
are the corresponding constants for the Co(NH,),CI2+ + Fe 2 + reaction, andf,,,
References pp. 267-273

202

C 0 M P L E X E S O F D I F F E R E N T MET A LS: 0 X I D A T I 0 N-R E D U C T I 0 N

T A B L E 15
RATE

COEFFICIENTS

(1.mole-'sec-'.)

FOR

SOME REACTIONS OF C H R O M I U M ( I I ) AND


25.0"C AND^ = 1.0 M 8

I R O N ( I I ) W I T H C O B A L T ( I I I ) C O M P L E X E S AT

C O ( N H ~ ) ~ C+Fez+
P+
C O ( N H ~ ) ~ C+Cr2+
I~+
C O ( N H ~ ) ~+Fe2
F~
C O ( N H ~ ) ~+Cr2+
F~+
C O ( N H ~ ) ~+Fez+
N~~
C O ( N H ~ ) ~ +Crz+
N~~+
cis-Co(en)z(NCS)C12++Fez
~is-Co(en)~(NCS)Cl'+Cr2
rruns-Co(en)z(NCS)C12 +Fez+
fruns-Co(en)z(NCS)C12++Cr2 +
+

From Haim and

and refs. cited therein.

-2.9
6.4
-2.2
5.9
-2.0
5.5
-3.8
6.3
- 3.9
6.4

6.6
5.5
5.7
5.1
-

5.6

Calculated from eqn. (6.4).

andf,,, are as defined by Marcus". Although equation (6.4) was first derived
for outer-sphere processes, evidently it can be applied as well to inner-sphere
reactions since observed and calculated rate coefficients are in good agreement
(particularly among the pentaammine series).
Cannon and E a r l e ~ ' ~have
' measured the rates of reduction of cis- and transX = NH, or H,O) by Cr2+. From the observed inC O ( ~ ~ ) , X ( H , O )(where
~+
verse dependence on hydrogen-ion concentration it is concluded that the ratedetermining step is the attack of the reductant on the conjugate base of the aquo
ion. A bridged mechanism is supported by the work of Kruse and T a ~ b e ' ~ ~
who showed that the Cr2+ reduction of cis-Co(en),(H20)OH2+ results in the
transfer of a single oxygen atom. Table 16 allows results of Cr2+ and Fe2+
reductions141 to be compared; the data on the FeZf reactions refer to the corresponding chloro complexes. There is a vast difference in rates between these
two systems: the rates of the Co(en),(NH,)OH2+ +Cr2+ reactions are about
10" times greater than the rates of the Co(en),(NH,)CI2+ +Fez+ reactions.
Assuming that this disparity in rate arises entirely from free energy differences
in the Cr(III)-Cr(lI) and Fe(III)-Fe(rI) couples (A(AG) = 27.6 kcal.mole-I),
it is interesting that the Marcus treatment43 gives

although the equation was intended originally to describe the behaviour of outersphere reactions. For both Cr2+ and Fe2+ reductions the most reactive complex
is the trans-aquo one. However, the effect of the trans ligand is less pronounced
when Cr2+ is the reductant. Outward motion of the trans ligand in the formation
of the activated complex is present in both Cr2+ and Fe2+ reductions but is much

OXIDATIONS B Y

203

CO(111)

TABLE 16
Co(en)2XC1"+ A N D C O ( ~ ~ ) ~ X O H
C O"M
+P L E X E S
Fez+ A N D Cr2+

RATE COEFFICIENTS FOR REDUCTION OF


BY

Fez+ reductant a
X

1 04k ( I . mole-

.set- I )

Cr2+ reductant
10 - 6k ( h o l e - '.set - )

trans-isomers
NH3
NCS
C1

Br
N3
H2O

cis-isomers
NH3
NCS
H20
CI
Chloro complexes141:[C104-]

0.66
1.3
320
360
620
2400

0.22
-

0.18
1.7
4.5
16

0.20

1.0 M , 25 "C.

2.6

0.79

Hydroxo complexes14s: 25.5 "C.

more important in the latter case. This is consistent with the low isotopic fractionation factor found for trans nitrogen in Cr2+ reduction^'^^.
E ~ p e n s o n ' ~has
' shown that the reaction of ~ i s - C o ( e n ) ~ ( N ~
with
) ~ +V2+ takes
place by an inner-sphere mechanism. This Co(II1) complex was selected for investigation because it is particularly reactive towards V2+,and also the dissociation of monoazido vanadium(lI1) is relatively slow. At low V2+ concentrations
(2-20 x
M ) the second-order rate coefficient is 32.9 I.mole-'.sec-' at 25 "C,
[H'] = 0.10 M and p = 1.0 M . At higher VZf concentrations (- 0.1 M ) ,
using a stopped-flow apparatus, the kinetics are apparently first order at 520 mp,
a wavelength where VN32f shows negligible absorbance. The rate coefficient under
these conditions agrees with that obtained at low V 2 + concentrations. However,
the data obtained by monitoring the reaction at 350 mp, an absorption maximum
for VN32+,deviates from first-order behaviour. Subtraction of the absorbances
of stable reactants and products from the net absorption at 350 mp gives the
absorbance due to the VN3'+ entity. Fig. 6 shows the formation and decay of
the latter in a typical experiment. The variation of VN32+ concentration with
ime was calculated on the basis of an inner-sphere sequence
+V2+
ci~-Co(en),(N,)~+

-% VN32+

V3+ + N 3 -

where k,, and k,, are the rate constants for the redox and aquation reactions,
respectively. As shown in Fig. 6, the calculated and observed values agree nicely.
That the VN," intermediate does not originate from interaction of the Co(I1I)
Rejercnres pp. 267-273

204

COMPLEXES O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

30

02

06

04
Time b e c )

Fig. The formation and decay (at 350 mp) of the VN32 intermediate tmed in the cis-Co(en)2(N3)2+- 1 - V2 reaction. The points are the observed concentrations and the line is calculated from
the known rate parameters. [ C O ( I I I ) ]=
~ 1.5 x
M , [ H + ] = 0.10 M , [ V 2 + ] 0.1 M . The
+

intermediate attains a maximum concentration of 2.7 x l o w 5M a t 0.17 sec; calculated, 2.74 'i
M at 0.113 sec. (From E~penson'~', by courtesy of The American Chemical Society.)

complex with VN,' was proved by the observation that added HN, has no effect
on the rate.
It is of incidental interest that a little work has been done on dicobalt systems.
Doyle and Sykes14* have made a study of the reduction of decammine-p-amidodiby V(I1). Since the rate is independent
cobalt(IIl), (NH3),Co~NHz~Co(NH,)ss+,
of hydrogen-ion concentration the mechanism cannot involve an amide bridge
and must be outer-sphere, as it is in the case of the reduction of CO(NH,),~+
by V(11)112~
149. Both the binuclear complex and C O ( N H , ) ~ , +
are inert to substitution but the former is capable of functioning as a two-equivalent oxidant.
Thus the two likely mechanisms are
n+

( N H 3 ) , C ~ ~ N H 2 ~ C ~ ( N H+3V) zs+s +--+

CO'++CO(NH,),~+
+V3+
+ 5 N H 4 + (6.5)

H+

CO(NH,),~+f V 2 + -+ C o 2 + + V 3 + + 6NH,'

(6.6 1

and
(NH3)5C~.NH2Co(NH,),+
5+
V 2 + -f 2 Coz++V(IV)+ 1 1 NH4+ (6.7)
v(Il)+v(rv~

_f

2 V(II1)

On the basis of the rate law alone


rate

k [V(II)] [(NH,) ,Co.NH,-Co(NH,)

55

(6.8)

cO(III)

OXIDATIONS BY

205

the mechanisms are indistinguishable. For experiments in which V(I1) was


present in large excess over the complex, the formation and decay of an absorption
peak at 425 mp was noted during the initial stages of the reaction. Step (6.8)
proceeds in part by means of a binuclear intermediate VOV4+, characterised by
6800 I.mole-'.cm-l) at the same ~ a v e le n g th , ~ .
its intense absorption ( E
Although this appears to be evidence for the two-equivalent mechanism, in fact
VOV4+ is produced by a side reaction between V(I1) and H f ions or water?. It is
concluded that the reaction between V(I1) and (NH3),Co~NH2Co(NH,),5 f
takes place predominantly through the one-equivalent route, (6.5) and (6.6), with
the intermediate formation of Co(NH,),,+. Activation parameters for step (6.5)
are AH' = 9.5 kcal.mole-' and A S f = -31 f 3 cal.deg-'.mole-' at p = 0.4 M .
Activation parameters for step (6.6) are AH' = 9.1 kcal.mole-' and AS' =
-40 cal.deg-'.mole-' at p = 1 M (ref. 149). Step (6.5) is faster than (6.6) by a
factor of
30 at 25 "C; the differences in rate must arise from differences in activation entropies. Table 17 gives direct and relative kinetic data on the effect
of anions on the reactions of V2+ with (NH3),Co~NH2~Co(NH3)ss+
and
Co(NH,),,+. The reactivity order is the same in both cases, i.e. F- > SO,'- >
CI-, and the relative rates are similar. The suggestion is made that the anion is
brought into the activated complex by V2+ and is held there at the side of V(I1)
remote from the cobalt centre. Sykes'50-'52 has examined other dicobalt systems:
the reductions by Fe(I1) of the peroxo complexes, (NH,),Co.p(NH,, O2)*Co
(NH3)44f and ( N H , ) , C O ~ O , C O ( N H ~ )(see,
~ ~ +also, refs. 302 and 303, in which
reduction of the latter complex has been studied using Cr(II), V(I1) and Eu(I1)).
The reductions by Fe(l1) of chloro(ethy1enediaminetriacetatoacetate)cobaltate
(HI), Co(Y)CI2-, and its conjugate acid, Co(HY)Cl-, have been investigated
by Pidcock and Higgin~on"~.At hydrogen-ion concentrations > 5 x
M

T A B L E 17
ANION EFFECTS I N THE REDUCTION OF
BY

(NH3)5Co.NH2.Co(NH3)5s+A N D C O ( N H & ~ +

vz+A T 25 'CL4*

(NH,) 5 CO.NHZ.Co(NH3) 5'


Anion

Uncatalysed
+CI-

+sod2+F-

k'(12.mole-z.sec-')

k'lk

0.149 a
2.1

14

1000
8600

Ionic strength, 0.4 M. kobr = k - t k " X - 1 .

6700

58000

CO(NH3)6

k'(12.niole-2.sec-')

0.00441

0.035
8.50
91.7

k'lk
-

8
1930
2 1000

k values in I.mole-'.sec-L.

The side reaction can be obviated by setting aside the V(I1) before commencing a reaction,
thus removing traces of oxygen301.
References pp. 267-273

206

c o M P L E x ES o F

D I F F E R E N T METALS:

ox I DA T I o N-RE D u C T Io N

the chief reaction is


Co(HY)Cl-

+ Fe2+ = Co2++ Fe(lI1)

It should be noted that the Co(I1)-EDTA complex is unstable and dissociates


to give the simple Co2+ ion. The product Fe(II1) is largely Fe3+ along with some
EDTA complex. The rate law obeyed is
-d[Co(HY)CI-]/dt

= k,[Co(HY)Cl-][FeZ+]

where log k, = log ko+AIH+], and k, and A are constants. The complexes
Co(Y)C12- and Co(HY)CI- react with Fez+ at comparable rates. Comparisons
are instructive between this redox reaction and other cation-catalysed chlorideabstraction reactions154 which occur without simultaneous oxidation-reduction.
The intimate mechanisms are likely to be

Co"'(HY)CI- +Fez+ + [Co"'(HY)CI


Co"( HY)-

+Fe"'C1'

- - Fe"] +[Co(HY) - - C1

* *

Fe]*

and

+ [Co"'(HY)CI MI']
C1 - MI1]*
Co"'(Y) + H + + M"CI+

Co"'(HY)CI- + M 2 +
+[Co"'(HY)

* *

* * *

where M2+ = Mn2+, Co2+, NiZ+, Cd2+, Pb2+ and Hgz+. The fact that the
reduction by Fe2+ has a lower activation energy may be a consequence of the
weakening of the Co-Cl bond in the activated complex on reduction of the charge
on the Co centre.

6.2

ORGANIC

B R I D G I N G LIGANDS

IN

OXIDATIONS

BY

COBALT(III)

COMPLEXES

A partial knowledge of the function of organic ligands as bridging groups has


arisen out of an examination of the rates of reduction of over a hundred carboxylatopentaamminecobalt(l1I) complexes with Cr(ll), V(lI), Fe(II), and other
reductants. It seems clear that the role of the organic ligand is to act as a mediator
for electron transfer. For this to occur the ligand must contain groups capable
of associating strongly with the reductant. Alternatively, it must contain a conjugated system of bonds. Three classes of interaction are recognised: adjacent
attack, adjacent attack with chelation, and remote attack. The topic has been the
subject of an extensive review by Taube' 5 5 (see, also, Taube and Gould304).
Table 18 is a collection of rate parameters (taken, mainly, from Fraser13')

T A B L E 18

h,

u
h

TE P

R A M E T E R S ~F O R R D U C T I O N O F C

Reductant
Ligand

kb(l.mole- '. AHt(kca1.


sec-')
mole-')

From Fraser13'.

0.35'
0.10
0.11
0.074
0.15'
0.074
0.082
-

0.075'
0.093'
0.21
-

0.13
0.18
7d (7.2')
0.052"

8.2'
7.9
4.0
2.5
9.0'
6.0
2.8
5.1'
2.6'
10.0
9.6
7.5
8.3'
-

For ,u = 1.0 M, 25 "C.

Co(II1) C O M P L E X E S B Y Cr2+,V + ,

VZ+

CfZ+

Acetato
Chloroacetato
Cyanoacetato
Dichloroacetato
Benzoato
o-Chlorobenzoato
o-Iodobenzoato
Salicylato
o-Phthalato
m-Phthalato
p-Chlorobenzoato
p-Iodobenzoato
p-Hydroxybenzoato
p-Cyanobenzoato
Formato
Trifluoroacetato
a

RBOXYLATOPENTAAMMINE

ASt(cal. kb(l.mole-'.
deg-'.rnole-') sec-')
-33'
-37
-49
-55
-33'
-43
- 54

0.43
1.25
1.13
1.03
0.52
0.57
0.90

-45'
- 56'
-28

1.01
0.60
0.60
0.37
0.53
0.88

- 30

-37
-27'
From ref. 156.

AHt(kca1.
mole-')

5.8
9.4
9.4
9.6
6.7
10.5
15.8
9.3
10.2
9.0
8.0
9.0
9.3
10.2
-

From ref. 155.

Euz

ASt(ca1.
deg-'.mole-')
-41
- 27

-27
-26
-31
-24
- 5
-27
-24
-29

kb(l.mole-'
sec-')
0.18
3.16
2.08
0.24
0.39
0.28
2.16
0.64

AHz(kca1.
mole-')
4.4
9.3
-

6.9
6.2
3.2
3.3

-51
-

- 30
-33
-

2.1

-24
-

From ref. 304.

-47
-25
-35
-40
-50

8.2
7.9

0.27
-

From ref. 157.

ASt(ca1.
deg-'.rnole-')

-33

Eu2+

-30

-28

ND

208

C O M P L E X E S OF D I F F E R E N TMETALS: O X I D A T I O N - R E D U C T I O N

for the reduction by C r 2 + ,V 2 + , and Eu2+ of pentaamminecobalt(II1) complexes


containing those carboxylato groups which do not favour reaction by remote
attack and have no tendency for chelation with the reductant (except for the
salicylato and phthalato complexes). For these ligands, attack by the reducing
agent occurs at the carboxyl group adjacent to the Co(I1I) centre ( A )

and the path for electron transfer is Co-0-C-0-red. The two oxygen atoms in
the carboxylato complexes are not equivalent'5s (a result in conflict with a previous
report'59). In a systematic study, FraserI3' has investigated the effect of varying
the substituent R. The rates are, in general, independent of hydrogen-ion concentration. viz.
-d[Co(NH3),L2+]/dt

k[Co(NH3),L2+][red]

where red = C r 2 + ,V 2 + , or E u 2 + ,and the net reaction is, for example,


( N H 3 ) 5 C o L 2 + + C r 2 + + 5H +

C o 2 + + C r L 2 + + 5NH4+

The results show that:


( a ) variations in the nature of R have only a small effect on the rate coefficients
although there is a 104-105-fold variation in the dissociation constants of the
corresponding acids;
( 6 ) there are, however, considerable variations in AH' and AS'; furthermore,
all AS' values are strongly negative. A good linear relationship exists between
the activation parameters, viz.
TASI

LYAH'+P

For C r 2 + , LY = 0.94 and j = 18.0 kcal.mole-'. For V 2 + , a = 0.97 and P = 17.5


kcal.mole( c ) the rate decreases in the sequence CH3COO- > C1CH2C00- >
C12CHCOO-, as might be expected since the electron-withdrawing power of the
R group increases.
An explanation of the overall
lO-fold variation in the rates for the various
complexes has been given by invoking steric factors16'. Steric effects may account
also for the fact that, although HCOO- is less basic than CH3COO-, the formato
complex reacts much more rapidly than the acetato

O X I D A T I O N S BY

cO(III)

209

The rate of reaction is enhanced if the ligand is able to chelate with the reductant. Examples of chelating ligands are u-hydroxy acids, e.g., glycolate ( B ) and
lactate'61; and those ligands containing a carbonyl group ( C ) or a hydroxy group
( D ) in an ortho position to the co-ordinated carboxyl
157*161-164
(Table 19). o-Nitrobenzoato ( E ) (but not o-aminobenzoato) pentaammine complexes provide further examples. Chelation with Cr(I1) can take place via a
sulphur atom15" 164 , as in the S-benzylthioglycolato complex (F). The efficiency
of such ligands in promoting electron transfer from reductant to cobalt is imperfectly understood but is probably related to the provision of a firm route
along which an electron may move. Good evidence for the existence of chelation
is reported by Butler and Taube16' in the case of reduction of glycolatopentaamminecobalt(II1) by Cr2+. The first product of reaction is a metastable Cr(1II)
species which has a higher absorptivity than the stable glycolatochromium(II1)
ion, and reverts to the latter with a half-life of
22 h. Chelation of Cr2+ by the
u-hydroxy group seems to explain the rate sequence glycolate < lactate < methyl-

T A B L E 19
R A T E P A R A M E T E R S F O R R E D U C T I O N OF S E L E C T E D C A R B O X Y L A T O P E N T A A M M I N E
COMPLEXES B Y
S H O W I N G EFFECT O F C H E L A T I O N ' 5 5

cr2+,

Ligand
Glycolato
Methoxyacetato
Lactato
Methyllactato
a-Malato
p-Malato
Malonato
Salicylato
Phthalato ion
Salicylato ion

k (I.mole - .set- l )

9.0
9.3

3.1
0.42
6.7
11.8
2.7
0.36
0.29
0.15 a
2.7
2x108 =

9. I

ASf(cal.deg- '.mole- I )
-26
-23

-24
-

Ionic strength, 1.0 M; temp., 25 "C. a For p


From ref. 181.
RejPrences p p . 267-273

AH f (kcal.mole- )

3.0 M, ref. 157.

cO(III)

From ref. 129.

210

COMPLEXES OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

lactate since replacement of H by CH3 on the ct carbon should increase the basicity
of the O H group. Chelation is an important factor in the reduction by C r 2 + of
a variety of heterocyclic complexes (derived from pyridine, pyrazole and pyrazine)
as is evidenced by marked spectral effects'63. Huchital and T a ~ b e 'have
~~
examined the rate of ring closure of malonatopentaamminecobalt(II1). They
find that a chelate ring is formed before oxidation of C r Z +occurs. It is considered
that the hydrogen-ion dependent path is not to be conceived in terms of a remote
attack mechanism but that the activated complex is probably of the type ( G ) , uiz.

Ester hydrolysis accompanies electron transfer during the reduction by Cr(l1)


of pentaammine complexes containing half-esters of conjugated dibasic acids as
ligands, for example*, methyl fumarate" 9*166a, phenyl f ~ m a r a t e " ~ " ~methyl
~,
maleate120, methyl terephthalate, and phenyl terephthalate'68. Complete hydrolysis of the ester occurs in the reduction by V2+ and Eu2+ of the half-ester complex containing the methyl succinato
Huchital and Taube17' have
investigated the products formed in the reaction of methyl- and ethylmalonato
pentaamminecobalt(1II) complexes with Cr(1I). In the case of the methyl complex, spectrophotometry and ion-exchange show that about 50 % of the ligand
appears in the chelated form, Cr(OOC)2CHzf, and for the ethyl complex, about
67 %. The corresponding amount of alcohol is found free in solution. The rest
of the ligand is found as the monodentate ester malonato complex of Cr(II1).
V(I1) and Eu(I1) are ineffective in inducing hydrolysis. Contrary to an earlier
report'69, no ester hydrolysis occurs with the succinato half-ester complex on
reaction with Cr(II), V(rI), or Eu(I1). This latter observation can be rationalised
on the basis that the chelate ring is stable for the succinato complex than it is
for the malonato, the former complex reacting via simple adjacent attackL7'.
T a ~ b e suggests
' ~ ~ that in reductions by Cr(I1) an intermediate of the type ( H )

is formed which can decompose by rupture at the Cr-0 bond (to form the half-

* On reinvestigation'66b, the reduction of (methylfumarato)pentaamminecobalt(llI) by


Cr(lI), and by V(I1) and Eu(II), has been found to produce little ester hydrolysis.

OXIDATIONS BY

cO(III)

21 1

ester complex), or at the C - 0 bond (leading to hydrolysis and preservation of


the chelate ring). However, this does not explain reductions by V(I1) and Eu(I1)
where M 3 + - 0 bond cleavage would predominate since V(IT1) and Eu(II1) form
complexes which are highly labile. Neither does it account for the oxygen-tracer
observation that about 40 % alkyl-oxygen fission occurs in such systems'69. The
effects of chelation by non-bridging ligands have been covered by Fraser'" who
has shown that the rate of reduction of amminecobalt(1II) acetato complexes by
Cr(I1) and V(1I) is not altered when four NH, groups are replaced by two
ethylenediamines, or when five ammonias are replaced by a tetraethylenepentamine group. However, increasing chelation brings about a reduction in the
rate for sulphato complexes.
0~ ,O-CdNH&
C

H/C+cCH
I
o'/c\OH
(I)

When a dibasic ligand, containing a conjugated eystem, is attached to Co(III),


e.g., fumarate(Z) or terephthalate, the CrZ+ reductant may react at either of the
two carboxyl groups, i.e., by both adjacent and remote attack. In these cases the
rate law is compounded of two terms: a hydrogen-ion independent term and a
hydrogen-ion dependent term" 9,129,168. The precise effects of conjugation on
the rates of reduction of Co(1II) complexes has given rise to much interest and
speculation. Attempts have been
to relate the rate of remote attack
to the mobile bond order'74 between the terminal atoms of the bridging ligand.
Another theoretical treatment is that given by Libby'75. The remote attack
mechanism has been suggested for a large number of other ligands, for example,
oxalate, maleate"', methyl monoesters of fumaric and maleic acid"'~'66~'67~'76,
p-aldehyd~benzoatel'~,4-~arboxylatopyridine'~~,
ni~otinarnide'~', substituted
pyridines and p y r a ~ o l e ' ~ ~ .
Fraser'" has investigated the rates of reduction by Cr2+ of various cobalt(ll1)
ammine complexes containing ligands coordinated by nitrogen, e.g., urethane,
methyl glycinate, benzocaine, ethyl nicotinate and isonicotinate and ethyl-4aminobutyrate. These complexes react very much more rapidly than complexes
with similar ligands attached to cobalt through oxygen. It seems likely that the
first stage in the reduction of Co(IIJ) complexes is the formation of a radical ion
as a result of electron transfer t o the organic ligand, the electron being subsequently
transmitted to the Co(II1) centre. Furthermore, it is possible that remote attack
takes place only if the organic ligand is reducible. There is strong evidence to
suggest the formation of an analogous radical-ion intermediate in the oxidation
.
in principle
of formatopentaamminecobalt(II1) by ~ e r m an g a n a te ~ 'Although
ESR spectroscopy is capable of detecting such intermediates, there is at present no
References pp. 267-273

212

COMPLEXES OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

direct evidence of this kind. The subject of remote attack is discussed at length
by Taube','.
Halide effects have been reported for the Cr2+ reduction of Co(NH,),L"+
complexes (where L = NH,, acetato or fumarato, FuH-)lE2. Chloride exerts a
stronger catalytic effect than bromide; in all cases the second-order rate coefficient takes the form kobs= k+k,[X-] where k is pH-independent for L = NH,
or OAc-, but not for L = FuH-. When L = NH, or FuH-, halide is captured
into the inner coordination sphere of chromium, and products of the type XCr2+
and XCrFuH' result. CO(NH,),NH,~+ is known to react via an outer-sphere
mechanism'12, whereas CO(NH~)~OAC'+
and Co(NH,),FuH2+ use bridge
mechanism^"^*'^^ . It is surprising that C1- and Br- have similar influences on
both outer- and inner-sphere processes (Table 20). This similarity suggests that a
Cr(I1) halide complex, CrX', is formed rapidly and that this species takes part
in the formation of the activated complex.
T A B L E 20
KINETIC

L
NH3

FuHOAc-

PARAMETERS

IN

THE

p(M)

k(l.mole-'.
sec-')

2.60
3.4
1.65

0.0072

HALIDE-CATALYSED
B Y CrZ+(AT 25 oC)'82

REDUCTION

kc,(12.mole-2. kB,(12.mole-2.
sec-I)
sec- l)
0.60
0.69
0.20

1 .oo

0.34

0.3 1
0.30
0.064

kc,/kB,
1.9
2.3
3.1

OF

CO(NH3)sL"'

kc,/k(l. kBr/k(l.
mole-') mole-')
83
0.69
0.59

43
0.30
0.10

T A B L E 21
RATE PARAMETERS FOR

T H E OXIDATION OF

CO(1II)

Complex
co(Y)Co(HY)H,O
Co(Y)HZOCo(Y0H)HZO
Co(Y0H)OHCo(HY)CI-

A H i (kcal.mole- ')
11.8
11.1
22.1
14.2
13.7
11.3

Fe(I1)

B Y VARIOUS

EDTA

AND

HEDTA

COMPLEXES'84

ASi(cal.deg- .mole - ')


-36 1 5
-33 1 3
+27 4 ~ 6
-23.511.4
+6 f 5
-21.9f1.5

k (25 "C)(/.mole- '.min - ')


0.0362
0.510
53.9
0.277
1.45 x lo6
81.7

The kinetics for the reduction of glyoxalatopentaamminecobalt(I11) by Cr2+


indicate that the carbonyl form is far more reactive (k > 7 x lo3 I.mole-'.sec-'
at 25 "C) than the hydrate (k 1 I.mole-'.sec-' at 25 OC)IE3.The hydrate form
predominates and its rate of dehydration has been studied.
Wood and Higgin~on"~have made a detailed study of the kinetics of oxidation
of Fe(1l) by a number of complexes of Co(1II) with ethylenediaminetetraacetic
acid (H4Y = EDTA) and hydroxyethylethylenediaminetriaceticacid (H3YOH =
HEDTA). Rate data and activation parameters are quoted (Table 21) for the

OXIDATIONS BY

cO(III)

213

reduction of Co(Y)-, Co(HY)H,O, Co(Y)H,O-, Co(YOH)H,O, Co(Y0H)OH-,


and Co(HY)CI- in perchloric acid media. The last two complexes are believed to
react with Fe(l1) through inner-sphere activated complexes involving hydroxo
and chloride bridges, respectively. On the other hand, the reductions of Co(Y)and Co(YOH)H,O are best interpreted as proceeding via carboxylato bridges.
When the Co(I1I)-EDTA complex is reduced by Cr(II), three carboxylato groups
are transferred which suggests that a triply-bridged transition state is involved in
this rea~tion"~.A number of redox reactions involving metal-EDTA and similar
complexes have been studied briefly by Wilkins and Yelin305.

6.3

OXIDATIONS BY AQUO COMPLEXES OF COBALT(III)

The course of the oxidation of V(II1) by Co(II1) has been the subject of an investigation by Rosseinsky and Higginson". Fig. 7 illustrates the dependence of
the concentrations of reactants and products upon time when Co(II1) is in excess
over V(II1). Under the conditions applying in Fig. 7, V(Il1) is oxidised to V(1V)
until 90% of V(II1) has been used up (i.e. at
28 min). Then V(V) begins to
be produced, and after V(II1) has been totally consumed (- 37 min) the reaction
corresponds to the oxidation of V(1V) by Co(II1). Co(II1) reacts with V(II1) and
V(1V) at a similar rate under the same conditions*. It is clear that Co(lI1) is
reduced by V(IV) as well as by V(II1) during the reaction, and that reduction by

5*0h

4.0

3 .O

Co (ID)

c\

20

40

60

00

Time (rnin)

Fig. 7. Dependence of concentrations of reactants and products upon time in the Co(lII)+V(IlI)
reaction. Initial concentrations: [Co(III) J = 4.99 x
M ; [V(III)] = 2.93 X
M , [V(lV)]=
3 x lo-' M ; [H+]
= 2.96 M, ionic strength = 3.0 M ; temp., 5.0 "C. (From Rosseinsky and
Higginsonso, b y courtesy of the Cheniical Society.)

* The redox potentials of the V 0 2 + - V 0 2 + and V02+-V3+ couples are 1.00 and 0.36 V,
respectively.
References pp. 267-273

214

COMPLEXES OF DIFFERENT METALS: OXIDATION-REDUCTION

V(1V) becomes progressively more important as the reaction proceeds. The fact
that V(V), the product of oxidation, is not detectable until all the V(II1) is consumed is ascribed to the rapid reaction V(III)+ V(V) + 2 V(IV) with a rate coef5 x lo3 I.mole-'.min-' at 5 "C, [ H f ] = 3 M and p = 3 M. When the
ficient of
initial concentration of V(I1I) is greater than that of Co(III), the reaction between
V(II1) and V(V) is unimportant and

Values of k, were derived from plots of this equation using known values of kb (see
below). At 0 "C, [H'] = 1 M and ,u = 3 M , k, is equal to 0.192 I.mole-'.sec-'.
k, varies inversely with hydrogen-ion concentration, the precise form of the
function being uncertain but of the type k, = k, +k'/[H+].
The reaction between Co(II1) and V(IV)
Co(III)+V(IV)

co(Il)+v(v)

in perchloric acid media (followed by monitoring Co(I1I) at 400 mp) is second


order and the rate is little affected by the presence of a large excess of V(V)".
However, high concentrations of Co(I1) (- 0.1 M ) increase the rate; also chloride
ions are found in the reaction mixture. Apparently, Co(I1) catalyses the reduction
of perchloric acid by V(1V). The second-order rate coefficient, k,, can be expressed
as kb = k,+k"/[H+] where k, and k" are 0.245 1.mole-'.sec-' and 2.08 sec-',
respectively, at 20 "C and p = 3.0 M. The corresponding activation energies are
21.8 and 15.9 kcal.mole-', and the entropies of activation are 12+9 and - 5 + 6
cal.deg-'.mole-', respectively. At 0 "C, [ H + ] = 1 M and p = 3 M , k, is 0.260 I.
mole- '.set- I.Anion effects were not investigated.
Although the oxidation of Cr(lI1) by Ce(IV) is a rapid reaction in perchlorate
solutions, the oxidation of Cr(II1) by Co(II1) in the same media occurs at a rate
similar to or slower than that of the thermal decomposition of Co(Il1) in perchloric
acid (3 M ) . However, the Co(ITI)+Cr(III) reaction is subject to catalysis by Ag(J)
ion and Kirwin et af.'86have made a kinetic study of this system, viz.
3 Co(lII)+Cr(lll) = 3 Co(II)+Cr(VI)
At?([)

Kinetic data were obtained by following the rate of appearance of Cr(V1) at


475 mp, a wavelength where Cr(III), Co(T1) and Co(1II) absorb only slightly.
Ag(1I) is said to exhibit an absorption maximum at 475 mp but does not interfere
with the observations since it is present in low concentrations. The concentration
of Cr(1II) was maintained in excess over Co(II1) and Ag(1). Plots of log (A,-A,)
versus time are linear where A , and A , are the absorbancies due to Cr(V1) after
complete reaction, and after time t. At constant Cr(III), Co(1I) and Co(1II)

OXIDATIONS BY

cO(III)

215

concentrations the observed rate coefficient (kobs)is proportional to Ag(1) concentrarion. In addition, plots of l/kobsversus l/[Cr(III)] and 1/k& versus [cO(II)]
are linear. Variations in acidity and perchlorate concentration have no effect on
the rate. The kinetic results are discussed on the basis of the following mechanism
Co(III)+Ag(I)

2 Co(II)+Ag(II)

k-3

Cr(III)+Ag(II) 3 Cr(IV)+Ag(I)
Cr(IV)+Ag(II)

--+

Cr(V)+ Ag(1)

Cr(V)+ Ag(I1) +Cr(VI)+ Ag(1)


Application of the steady-state treatment to the concentrations of Ag(II), Cr(1V)
and Cr(V) leads to the rate law
- d[Co( III)]

- d[Cr( VI)]

3 dt

dt

III)]
+ 3 k,[Cr(III)]

- k3 k4[Cr( III)][Ag( I)][Co(

k-,[Co(II)]

Since
dt
d[Cr(VI)]

-3 dt
d[Co(III)]

3
kob,[Co(III)]

inversion of the rate law gives

in keeping with the linear relationships referred to above. Evaluation of the


slopes and intercepts of these plots (along with the value of k 3 / k - 3 previously
determined) allows the individual rate coefficients, k3, k-, and k,, to be calculated.
Table 22 contains these values together with the associated activation parameters.
The rate coefficient for the oxidation of Mn(n) by Co(I1I) has been determined
T A B L E 22
RATE

PARAMETERS FOR

THE Ag(l)-CATALYSED

Rate coeficient (/.mole- '.see- ')


3717
5017
1611
a From Kirwin et
error by a factor of 3.

References p p . 267-273

CO(I1I)+Cr(III)

AHt(kcal.mole-

I)

REACTION

AT

25

"c"

'

ASt(cal.deg- '.mole- )
7jI 6
-32+10
- 14120

As noted by Sutin et aI.lg1,the original table has rate coefficients in

2 16

COMPLEXES 0 F DIFFERENT MET A L S : O X I D AT I ON-R E D U C T I 0 N

by Diebler and Sutina2 as 1 . 0 0 ~


lo2 I.mole-'.sec-' in 3.0 M perchloric acid
at 25 "C.
The reduction of Co(ll1) by Fe(I1) in perchloric acid solution proceeds at
a rate which is just accessible to conventional spectrophotometric measurement^^^.
At 2 "C in 1 M acid with [Co(IlI)] = [Fe(II)] 5 x
M the half-life is of the
order of 4 sec. Kinetic data were obtained by sampling the reactant solution for
unreacted Fe(I1) at various times. To achieve this, aliquots of the reaction mixture
were run into a quenching solution made up of ammoniacal 2,2'-bipyridine, and
the absorbance of the F e ( b i ~ y ) , ~ complex
+
measured at 522 mp. Absorbancies
of Fe(I1I) and Co(ll1) hydroxides and Co(bipy),'+ are negligiblz at this wavelength. With the reactant concentrations equal, plots of l/[Fe(II)] versus time
are accurately linear (over a sixty-fold range of concentrations), showing the
reaction to be second order, uiz.
N

-d[Fe(II)]/dt

-d[Co(III)]/dt = k'[Co(III)][Fe(II)]

No indication is given of the reaction of Co(1II) polymers although these are


present in the reaction solution^'^^. It is noteworthy that the intercepts of the
above plots do not coincide with the values obtained from the initial Fe(1I) concentration. The "zero-time oxidation" is believed to arise from a finite quenching
time together with a rapid reaction of hydrolysed species of the reactants. The rate
of reaction is inversely proportional to the concentration of hydrogen ions. This
result is taken as implying competitive reactions between CoOHZf + F e z + and
C o 3 + + F e 2 + ,as described by the rate law

rate = k , [Co3 ][Fez ] k6 [CoOH2 ][Fez


+

Thus the observed rate coefficient, k', is given by


k' = k,+k,K,/[H+]
where Kl is the hydrolysis constant"" for Co3+. Slope and intercept values
of plots of k' Versus 1/[H+J enable k , and k, to be calculated. At 0 'C and
6500 I.mole-'.sec-', respectively. The
p = 1.0 M , k , and k, are 10 and
corresponding activation energies and activation entropies are 9.1 and 7.9 kcal.
mole-' and -23 and - 14 cal.deg-'.mole-', respectively. It is apparent that the
preferred path is that involving CoOH2+. It is a common observation in metalion redox systems that the greater part of reaction proceeds via hydrolysed species
as, for example, in the reactions Fe(II1) + Fe(II)", Co(III)+ C0(11)'"~, and
Fe(II1) Cr(II)94. Sulphate, unlike fluoride, increases the rate of reaction. Assuming this is due to the introduction of the step

CoS04++ F e z + k.,Co2++ FeS04+

OXIDATIONS B Y CO(II1)

217

then k, has a value of


4900 1.mole-'.sec-' at 0 "C. Anion effects were not
investigated in detail.
Some key experiments have been performed by Sutin et ~ 1 . ' ~on' the reaction
between Co(II1) and Fe(I1) in the presence of chloride ions, and have established
the process

CoC12+ Fe2+_+ (CoC1Fe4+)' +Co2++ FeC12+


to be of the inner-sphere variety. A flow technique44 was necessary since both
reactants and both products are substitution-labile. When a solution containing
M Co(III), 2.9 x
M Co(II), 4.0 x lo-, M Cl- in 3 M perchloric
9.6 x
M Fe(1I) in 2.7 M perchloric acid, the
acid is mixed with a solution of 9.2 x
reaction of CoC12+ with Fe2+ is complete within 2 or 3 msec (k 2 5 x lo3 1.
mole-'.sec-'), and the FeCI2+ produced decays with a half-life of 280 msec.
FeCl" has a A,,
of 336 mp. When a solution of 9 . 6 ~
M Co(1II) and
5.0 x 1 O - j M Co(I1) in 3 M HC104 is mixed with one containing 9.2 x lo-' M
Fe(l1) and 4 x lo-, M C1- in 2.7 M HC104, FeC12+ is not detected at 336 mp.
However, at larger Co(II1) concentrations (4.8 x lo-, M ) the slow formation
of FeCI2+ is observed due to

Fe3+ C1-

*FeC12+

The rate of oxidation of Fe(I1) by Co(I1I) has been studied in the presence
of Ag(I) by means of a stopped-flow apparatus'". Under conditions of excess
Ag(1) and Fe(II), Ag(I1) competes with Co(1II) for Fe(II), and the reactions are

Co(II1) Ag(1)
k-3

Co(11) Ag(I1)

A Ag(I)+Fe(III)
Co(III)+Fe(II) A Co(II)+Fe(III)
Ag(II)+Fe(II)

At sufficiently high concentrations of Fe(1I) the observed second-order rate


coefficient for the disappearance of Co(II1) can be expressed as

thus allowing the determination of k, and k9 from a plot of kobsversus [Ag(I)]/


[Fe(II)]: k, = 41 + 3 l.mole-'.sec-' and k, = 330+7 1.mole-'.sec-' at 25 "C.
The oxidation of Co(I1) by Ag(I1) was also examined (at 470 mp, the absorption
maximum for Ag(lI)) in the presence of excess Co(I1) and Ag(I), when the firstReferences pp. 267-273

218

C O M P L E X E S OF DIFFERENT METALS: OXIDATION-REDUCTION

order rate coefficient describing the approach to equilibrium is given by

Values of kobsvary linearly with [Co(Il)], at constant [Ag(L)], and k , and k-,
were obtained as 37+4 I.mole-lsec-' and 1.75 x lo3 I.mole-'.sec-', respectively
(cf. Table 22). The two independently-determined values of k , are seen to be
in close agreement. The equilibrium constant, k , / k _ , , of the Co(III)+ Ag(1)
system is, thus, 2.1 x lo-* (using average k , ) in 4 M perchloric acid at 25 "C. A
value of 4 + 2 x lo-' is arrived at by considering the oxidation potentials of the
Ag(I)-Ag(LI) and Co(11)-Co(II1) couples.
Despite the lability of Fe(II1) complexes, Haim and Sutin192 have been able
to study the rapid reductions of some Co(I1I) complexes by Fe(I1) and identify
the primary Fe(II1) products. A flow apparatus, described by Dulz and S ~ t i n ~ ~ ,
enabled the rate of formation and decay of the spectra of the Fe(II1) products

Fig. 8. Transmittance uersus time curves for the C O ( C ~ O ~+)F~e z~+- reaction. Upper curve
shows disappearance of
(wavelength, 600 mp; abscissa scale, 500 msec per major
division). Lower curve shows formation and decay of the intermediate FeC2O4+ (wavelength,
310 mp; abscissa scale, 2 sec per major division). [ C O ( C ~ O ~ )=~ ~
1 -. ]0 ~
M ; [Fez+] =
2.5 X lo-' M ; [HC104] = 0.92 M ; ionic strength = 1.0 M ; temp., 25 "C. (Froin Haim and
Sutin' 92, by courtesy of The American Chemical Society.)

OXIDATIONS BY

cO(III)

219

to be observed and recorded (Fig. 8). A two-stage mechanism holds for the
C O ( C ~ O , ) , ~ -Fe2+ reaction, uiz.

CO(C,O,),~- + F e 2 +
F e C 2 0 4 + + 2H +

_f

FeC,O,+ +Co2+ + 2 C,O?-

+ Fe3++H2Cz04

The disappearance of CO(C,O,),~-, followed at 600 mp, gives a rate coefficient


of 33 I.mole-'.sec-' at 25 "C and p = 1.0 A4 (cf. the value of 1.15 x lo3 I.mole-'.
sec-' as determined by Barrett and BaxendaleI9, for 20 "C and p = 0). The rate
of formation and disappearance of FeCzO,+ was followed at 310 mp. Under the
conditions described in Fig. 8, the half-lives for the redox process and the decay
of FeC,O,+ are 0.85 sec and 2.4 sec, respectively. The latter value agrees well with
the half-life as calculated from the data of Moorhead and S ~ t i n ' ~The
~ . time
required for FeC204+ to achieve its maximum concentration is 2.1 sec (cf. 2.0
sec calculated on the basis of the kinetic scheme) and the ratio [FeC,O,+],,,/
[Co(IIl)], is 0.51 (cf. 0.56 from the kinetic scheme). The results indicate that an
oxalate-bridging mechanism applies. Similar conclusions are drawn for the
Fe(1I)-reduction ofCo(H.EDTA)CI-, Co(en),(OH,)Cl2 +,Co(NH3),(OH2),C12+,
C O ( N H ~ ) ~ ( O H ~ ) and
~ N ~C~O+( N
, H , ) , ( O H , ) C ~ O ~ In
~ . all cases a bridged
transition state applies, and observations were made on the growth and decay of
the corresponding Fe(lI1) complex.
The kinetics of the Co(l11) oxidation of a number of substituted tris(1,lOphenanthroline) complexes of Fe(I1) have been studied by Campion et a[.' 9 5 ,
using a flow te~hnique,~.
The second-order rate coefficients obtained for 5-methyl1,lo-phenanthroline, 1,lO-phenanthroline, 5-chloro-1,lo-phenanthroline and 5nitro-1,lO-phenanthrolinein 3 M HCIO, at 25 'Care given in Table 23 along with
the formal oxidation potentials of the complexes. Table 24 gives observed and
calculated values of slopes and intercepts of plots of (AG;,, 1.15 RT logf)
uersiis AGO,,, (see eqn. (12.8), p. 247) for various redox reactions involving
substituted Fe(I1) phenanthroline complexes. The slopes, particularly, agree
well with the calculated values. However, the intercepts for all the reactions are
smaller than those calculated from the Marcus treatment. According to Campion

T A B L E 23

S E C 0 N D 0 R D E R R A T E C O E F F I C I E N T S F O R T H E 0 XI D A T I 0 N 0 F

COMPLEXES B Y

CO(1IL)

IN

Ligand
5-Methyl-l , 10-phenanthroline

I ,lo-Phenanthroline
5-Chloro-I, 10-phenanthroline
5-Nitro- 1,lO-phenanthroline
References p p . 267-273

3 M

Fe(l1)- P H E N A N T H R 0 L1 NE
25.0 0c'95

PERCHLORIC ACID AT

E"( b')
I .02
1.06
1.12
1.25

10- 3k12(l.mo/e'.set- l )

15.0
14.0
5.02
1.49

220

C O M P L E X E S OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

T A B L E 24
FOR OXIDATIONSLOPES A N D INTERCEPTS O F PLOTS O F ( A G t , , , + l . l 5
1ogf)vs.
R E D U C T I O N RE A C T I O N S I N V O L V I N O S U B S T I T U T E D Fe(Il)-PH E N A N T H R O L I N E C O M P L E X E S
AT

25.0 "C

Slope
Reaction

Intercepr(kcal.mole- I )

Observed Calculated

F e ( ~ h e n )+Fez
~~
a
Ce(IV)+Fe(phen)32+
Mn(IIl)+Fe(phen)3Z+
Co(III)+Fe(phen),Z+
+

0.56
0.48
0.49
0.51

From ref. 196.

Observed
14.8
14.8
17.6
19.5

0.50

0.50
0.50
0.50

From ref. 44.

From ref. 82.

Calculated

Medium
0.5 kfH2SO4
0.5 M HzS0.t
1 and 3 M HC104

13.0

13.0
15.8
12.9

3 M HCIO,

From ref. 195.

et u Z . ' ~ ' , this deviation may stem from non-cancellation of the non-electrostatic
contributions to the work required to bring together the various pairs of reactants.
The reduction of Co(II1) by Ag(1) in perchlorate solutions has been studied
by Sutcliffe el
Since the initial product of reaction is the very reactive
Ag(1l) species, all solutions were subject to preliminary ozonolysis to remove
traces of reducible impurities. The final products of reaction are Co(I1) and Ag(1).
Kinetic data were obtained spectrophotometrically by following the disappearance
of Co(II1) at 605 mp, a small correction being applied for the absorbance of
Co(l1). With Ag(1) in excess, the disappearance of Co(II1) is second order, i.e.,
plots of the reciprocal of the corrected absorbance uersus time are linear. The rate
is directly proportional to the concentration of Ag(I), and inversely proportional
to the square of the concentration of Co(II). These results can be understood in
terms of the mechanism

Co(III)+Ag(I)

Co(II)+Ag(II)

K2

k-3

Ag(l1)

+ products

Transient concentrations of Ag(I1) were detected spectrophotometrically, and


by electron spin resonance. The thermal decomposition of Ag(I1) perchlorate, the
subject of a separate studylg8,takes place by

+ Ag' +Ag3+
A g 3 + + H 2 0+ A g 0 + + 2 H +
2 Ag2'

Ago+ %Agf

+to,

rapid equilibrium, K3
rapid equilibrium, K4
slow

O X I D A T I O N S BY

CO(II1)

22 1

and the rate law is

Reverting to the Co(IIl)+Ag(l) reaction, the concentration of Ag(l1) is fixed


by K2 as

and the derived rate law is


-d[Co(III)]
dt

- -d[Ag2+]
dt

under conditions where [Ag(I)] > [Co(III)]. In keeping with this relationship,
the rate is dependent on [H+]-' at constant ionic strength; the observed rate
3.3 M perchloric acid. The
constant goes through a pronounced minimum at
overall A H f values for the Ag(I1) d e c o m p o s i t i ~ n ' and
~ ~ the Co(III)+ Ag(1)
r e a ~ t i o n ' ~ are
'
1 1 + 2 and 34+4 kcal.mole-', respectively. Thus K2 has an associated overall AH of 12 kcal.mole-'. The value of 0.76 obtained for K2 is in
poor agreement with the value of 2.1 x l o v 2 given in ref. 191 (see p. 218).
The slow reduction of cobalt(I1I) by mercury(1)

2 Co(II1) Hg(l), = 2 Co(I1) 2 Hg(l1)


is stoichiometric under conditions of equivalent concentrations of the two reactants, or when mercury(1) is present in excessso. However, when Co(I1I) is in
great excess the slow oxidation of water by Co(II1) becomes important, and the
reaction is then non-stoichiometric. The second-order nature of the rate law, uiz.
- 2 d[Hg(I)Jdt

= -d[Co(III)]/dt

= 2 k,,~s[C~(III)][Hg(I)~]

is maintained in the presence of a large excess of Co(I1) and Hg(II), the products
of reaction. Kinetic data were obtained in most cases by monitoring the disappearance of Co(111) at 603 mp. Variation in perchlorate-ion concentration does
not affect the rate. This last observation contrasts with the behaviour shown by
the TI(III)+ H B ( I ) ~system, where the rate varies inversely with C104- concentrationS6. It follows from the observed acidity dependence that the second-order rate
coefficient kobs can be written as kobs= k+k'/[H+] where k = 0.02 I.mole-'.
sec-' and k' = 0.36 sec-' at 19.9 "C and p = 3.0 M. The corresponding activation
References pp. 267-273

222

C O M P L E X E S O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

energies and activation entropies are 22.1 and 29.2 kcal.mole-', and 9 + 6 and
37 + 4 cal.deg-'.mole-', respectively. Either of the following schemes is likely
Co(III)+ Hg(I),

-+

Co(II)+Hg(II)+ Hg(1)

Co(IIl)+Hg(I) -+ Co(II)+Hg(ll)
or
Co(IlI)+Hg(I),

slow
rapid

C0(11)+Hg,~+

slow

C0(111)+Hg,~+ -+ Co(II)+2 Hg(I1)

rapid

Since the rate does not display an inverse dependence on Hg(1I) concentration,
the oxidation of Hg atoms, in equilibrium with mercury(1) and mercury(II), can
be discounted, although Hg atoms are kinetically important in the reduction of
thallium(II1) by mercury(I)86. It seems likely that the acid-dependent path ( k ' )
involves CoOH2+.Anion effects were not investigated.
Co(1II) oxidises TI(1)'l according to
2 Co(III)+TI(I)

2 Co(II)+TI(IIl)

in a manner resembling the oxidation of Fe(I1) by Tl(III)199. Owing to the instability of Co(II1) in aqueous perchloric acid solutions, the kinetics of the reaction were examined under conditions where TI@) was in excess over Co(III), the
disappearance of the latter being followed by titration. By employing low initial
concentrations of Co(II1) relative to TI(I), and by avoiding low hydrogen-ion
concentrations, the rate of decomposition was reduced to less than 10 % of the
overall rate of disappearance. The presence initially of high concentrations of
Co(II1) causes a marked reduction in rate; on the other hand, TI(II1) has little
effect. The simplest scheme is

where TI(I1) is competed for by Co(1I) and Co(II1). At low Co(1I) concentrations
-d[Co(III)]/dt

2 k,,bs[C~(III)][Tl(I)]

and the pH-dependence of kobs was investigated over the range 0.25-2.50 M
HCIO, between 0 "C and 25 "C at p = 2.70 M . First-order plots (of log [Co(III)]
versus time) deviated slightly, but significantly, from linearity at the beginning of
reaction. This is thought to come about from the reaction of dimeric species

OXIDATIONS BY

cO(III)

223

of Co(I11). The variation of kobswith [H'] can be expressed as

Plots of kobsversus l/[H+] allow k,, to be evaluated, but k,, cannot be defined
exactly due to doubt over the value of K , , the hydrolysis constant of Co3+, and
I.mole-'.sec-' at 15 "C, and 2.5 x
also to scatter in the data. k , is 6.8 x
l.mole-'.sec-'
at 25 "C. Using a value of 5 x
mole.l-' for K , (ref. 200),
k,,
1 . 7 lo-*
~
I.mole-'.sec-'
at 15 "C. Unsuccessful attempts were made to
correct for the reduction of Co(II1) by water, and thus errors of up to
10%
may be present in the quoted values of k,,. The latter has an associated activation
energy and entropy of 26.4 kcal.mole-' and 2 2 5 7 cal.deg-'.mole-l, respectively.
The rate of the Co(m)+Tl(I) reaction is increased by the addition of sulphate.
It is deduced that the SO4,- ion, and not the HS04- ion, is responsible for the
catalysis, although the latter is present in much larger amounts. The back-reaction
of Co(I1) with TI(I1) is much less important in the presence of sulphate. Ashurst
and Higginson" make a tentative suggestion that perchlorate ions participate
in the Co(III)+Tl(I) system, i.e., the activated complex contains one or more
CI04- ions.
The binuclear complex of Co(III), bi-p-hydroxobis(bioxaIatocobaltate(II1)
[(C204)2Co"'(OH)2C01'1(Cz04)24-],
is reduced by Sn(1I) in dilute hydrochloric
acid solutions2". The products of reaction are the mononuclear complexes,
probably cis-bioxalatodiaquocobaltate(II1). If trioxalatocobaltate(II1) is included
in the reaction mixture, then a considerable fraction of it is consumed during the
course of the reaction of Sn(I1) with the binuclear species. Since trioxalatocobaltate(111) does not react with Sn(1I) under the prevailing conditions, this result indicates
that the reactive reducing species is Sn(III), generated by a one-equivalent process.
bring
~ + about oxidaThe Co(I1I) complexes CO(NH,),~+and C O ( N H ~ ) ~ O H
tion of stannate(I1) ion in strongly basic solution202.The rates were found to
be independent of the concentration of the Co(II1) complex. It is proposed that
stannate(I1) exists as a dimer, and that the monomer is the reactive species, the
rate being close to half-order in stannate(I1). Cyanide and thiosulphate catalyse
the reaction but CO(CN),~-is immune to attack by stannate(II) ion. The experimental difficulties encountered in this study preclude a full analysis as regards
mechanism.
The reaction between Co(I1I) and Ce(II1) has been the subject of a detailed
study in a series of papers by Sutcliffe and Weber's8~20s~206.
Of particular value
is the thorough investigation of the influence of anions on the rate in perchlorate
media. The reaction was followed by meaeuring the disappearance of Co(II1) at
its absorption maximum of 650 mp, a wavelength where both oxidation states of
cerium are transparent and Co(T1) absorbs only slightly. Changes in temperature
and ionic strength affect the spectrum of Ce(II1) at the 296 mp maximum, but the

References p p . 267-273

224

C O M P L E X E S OF D I F F E R E N T ME r A L S : O X I D A T I O N - R E D U C T I O N

spectrum is unaltered by variations in perchloric acid concentration"'. From


these results it is concluded that a complex, presumed to be CeC1042f, exists
between Ce(II1) and perchlorate, and analysis of the spectral data allowed the
equilibrium constant ( K = 1.4 Lmole-' at ,u = 1.14 M and 25 "C) and associated
thermodynamic parameters to be calculated. Using a different approach, Heidt
and Berestecki203have obtained a value for K of 0.86 Lmole-' at ,u = 4.50 M
and 25 "C. These authors suggest that the perchlorate complex is an extended
(outer-sphere) one, namely, Ce(H20)63fC104-. The oxidation of Ce(II1) by
Co(lI1) is stoichiometric, and first-order in both reactants. The back-reaction is
>
not detectable in the presence of a large excess of Co(I1) ([CO(II)]~/[CO(III)]~
200). The dependence on HC104 concentration is described by

k'

= a/[HCIO,]

where k' is the observed second-order rate coefficient and a is a constant. At


acidities below 0.2 M a colloidal Ce(IV) compound is formed, leading to interference in optical measurements. At constant acidity the rate increases with increase
in ClO,- concentration (to show this effect the ionic strength was maintained
constant, and lanthanum perchlorate substituted for sodium perchlorate). This
dependence on C104- concentration can be expressed as

l/k' = b+c/[CIO,-]
where b and c are constants. Since previous work had precluded the possibility
of a reactive hydrolysed species of cerium(III), the hydrogen-ion dependence was
taken as evidence for the existence of CoOH2+.The equilibrium constant for
Co3

+ H 2 0 +-CoOHZ++ H +

Kl

was determined successfully (Kl = 1.75 x


mole.l-' at 25 "C), despite
experimental difficulties arising from the reduction of Co3+ by water at low
acidities. It is deduced that the reaction

COOH~ C~CIO,~
+

co(11) Ce(IV)

is the rate-controlling step, as a result of the observed rate response to variations


in acid and perchlorate concentrations. On the assumption that the total Co(II1)
[CeC1042+],
and Ce(II1) concentrations are [Co3+]+ [CoOH2'] and [Ce"]
respectively, it follows that

OXIDATIONS BY

CO(II1)

225

since K , is less than [H']. The empirical parameters. a, b and c, are then evaluated
as follows

TABLE 25
RATE PARAMETERS F O R T H E

CO(III)+Ce(III)

REACTION188*20s*206

k(l.mole- '.see-') at 25 "C AHf(kcal.mole-I )

CoOH2++CeC1042+
CoOH2++CeNOJ2+
CoOHZC+CeFZf
CoS04+ +CeS04+
CoS04+ +CeC104*+
coSO4+ +Ce3+
c o 3 ++CeS04+

95
93
8500
12000
250
5 350
I
200

ASf(cal.deg- '.mole- I )

14&7
-5*l

19&2
14*2

The true rate coefficient, derived from these relationships, is given in Table 25,
along with the corresponding AH' and AS' terms. A later spectrophotometric
study'*' of Co(II1) perchlorate solution has cast some doubt over the value of
the formation constant (Kl) for CoOH". It is suggested that CoOH" undergoes
slow dimerisation to C0-O-Co4+ and/or Co-0-CoOH3 , although further
hydrolysis and/or polymerisation is possible. Furthermore, reaction between
hydrolysed and dimeric species has been proposed as the rate-determining step
in the reduction of Co(1II) by waterZo0.A kinetic study of the formation of
the monochloro complex of Co(II1) has given a considerably higher value for Kl
ot 0.22 mole.1- at 25 OCZo4.Nitrate ion increases slightly the rate of the Co(II1)
Ce(l1I) reaction in terms of a first-order dependenceZo5.The effect is attributed
to the introduction of a step involving CoOH" and CeN032+.The formation
constant of the latter ion received a spectrophotometric study. In contrast to the
slight effect of NO,-, the addition of fluoride ion increases the rate markedly:
M F- is sufficient to double the rate coefficient. The linear
the presence of
dependence of rate on F- concentration is discussed on the basis of reaction
between CoOH" and CeF2+. Rate parameters are collected in Table 25: experimental difficulties arising from the insolubility of cerium(I11) fluoride did not
allow calculation of the AH' and ASf values for the fluoride-catalysedpath. That
the product Ce(IV) is strongly complexed is shown by the observation that spent
reaction mixtures were colourless instead of yellow. Sulphate complexes of both
reactants participate in the Co(III)+ Ce(lI1) reaction in bisulphate media; a
linear relationship is found between the rate and the anion concentrationZo6.
+

References pp. 267-273

226

COMPLEXES OF D I F F E R E N TMETALS: OXIDATION-REDUCTION

The predominant reactive species of Co(lI1) are C0S04+ at 20 "C, and Co(SO,),at higher temperatures than 30 "C. Rate data (in the form of maximum rate
coefficients) are given in Table 25 for the various mechanistic steps.
The oxidation of Np(V) by Co(II1) has the stoichiometry
Co(III)+Np(V) = Co(lI)+Np(VI)
and a rate law
-d[Np(V)]/dt

~'[N~(V)][CO(III)][H+]-~"'

in perchlorate media at 25 "CZo7.Kinetic data were obtained by recording the


absorbance of Np(V) (A,,, = 980 mp) as a function of time. No catalysis was
observed on the addition of SO4'-, CI-, HP04'- or F- ions. The form of the
rate law indicates that the activated complex is composed of one Co(II1) and one
Np(V) ion (along with an undetermined number of water molecules). The small
hydrogen-ion dependence is attributed to ionic strength effects arising from
changes in solution composition. Application of the Marcus equation (p. 247)
gives rise to a value of 2.8 x lo6 I.mole-'.sec-' for the second-order rate coefficient at 25 "C. The observed value is 3.35 x lo2 I.mole-'.sec-'. Similar discrepancies are noted for other Co(II1) systems.
Np(V), present as a weak complex with Cr(IIl)t, is oxidised by Co(II1) according to
Co(lIl)+ Np(V)Cr(IlI)

Co(II)+Np(VI)+ Cr(II1)

Kinetic data were obtained by following the rate of disappearance of the complex
at 993 mp, and the Cr(II1) product was identified"' as Cr(H20)63+.The rate is
described by
-d[Np(V)Cr(III)]/dt

k"[Co(III)][Np(V)Cr(III)][H+]

at 25 "C. This result suggests a hydrolysis of Co(III) as a pre-equilibrium, followed


by attack of the hydrolysed species at the Np(V) site in the complex as the ratedetermining step. By this means the net formal charge of the activated complex is
reduced. However, the stoichiometry and kinetics are consistent also with a scheme
in which Co(ll1) attacks the complex at the Cr(II1) centre, viz.
Co(Ill)--Cr(lll)~Np(V)

Co(II)+ Cr(IV)*Np(V)

t At 25 "C, the Np(V).Cr(llI) complex has an equilibrium constant of 2.62, and a rate coeffisec-' (p. 259). The dissociation of this species during the
cient for decomposition of 2.32 x
course of the redox reaction is negligiblezo8.

Pt(IV)

OXlDATlONS BY

227

whereupon Cr(IV) is destroyed either by direct electron exchange with Np(V),


or by dissociation of the Cr(IV).Np(V) complex, with subsequent reduction of
Cr(IV) and oxidation of Np(V).

7. Oxidations by platinum(1V)
Beattie and B a s o 1 0 ~ ' ~have investigated the reactions of the substitutioninert octahedral complexes of Pt(1V) with tris(bipyridine)chromium(II). A rapidmixing, stopped-flow apparatus"' was made use of in the majority of experiments.
Kinetic data were obtained by following the disappearance of Cr(bipy),''
at
562 mp, usually with Pt(IV) in excess. The stoichiometry corresponds to
Pt(IV)+2 Cr(1I) = products
and the rate law is simple second-order. The rate coefficients obtained are given
in Table 26. The complex Pt(en),,+ is reduced 20 times more rapidly than its
conjugate base, Pt(en),(en-H),
(where en-H represents ethylenediamine minus
a hydrogen atom). A comparison of the rates of reduction of Pt(1V) and C0(111)"~
(see Table 9, p. 190) by the common reductant Cr(bipy),'+ reveals: (a) the
order of halogenopentaamminecobalt(III) complexes is F- < C1- < Br-,
whereas that for the analogous Pt(1V) complexes is C1- < I - < Br-, ( b )
Pt(en),,+ is reduced ten times faster than Pt(NH3)64+ whereas Co(en),,+ reacts
four times slower than CO(NH,)~,+.The presence of a hydroxo ligand imparts
inertness to Pt(IV), truns-Pt(NH,),CIOH'+ reacting nearly lo4 times slower
than truns-Pt(NH,),ClZZ+. There is a general correlation between the rates of
reduction of Pt(1V) complexes and their polarographic half-wave potentials.
Beattie and BasoloZo9suggest that the rate-controlling step in the reductions is
the formation of a Pt(II1) intermediate by a one-electron, outer-sphere process,
cf. the oxidation of Pt(I1) complexes by hexachloroiridate (IV),06.
+

T A B L E 26
RATES OF REDUCTION OF

Pt(1V)

COMPLEXES BY

AT

25 "C AND^

k (I.mole-'.sec-')

Complex
Pt(NH3)5C13
Pt(NH3)5Br3+
Pt(NH3)513+
+

(2.9*0.3) x lo5
(8 &4 )x106
(3 f l )x106
(7.7 I .o) x 103
( 4+1 )x102
(7.7*1.0)x10~

Pt(en),(er~-H)~
Pt(NH3)64+
trans-Pt(N H3)4CIZZ
rrans-Pt(NH3),C10H2 +
+

References pp. 267-273

Cr(bipy):+

107

(2 *I

)xi03

= 0.1 MZo9

228

c o M P L E x E s OF D I F F E R E N T M E T A L s: O X I D A T I o N-R

E D u cT I oN

8. Oxidations by copper(I1)
The rate of reduction of Cu(I1) by Cr(I1) has been measured in aqueous
perchloric
3 0 7 . With Cu(I1) in excess and in the absence of oxygen, the
reaction corresponds to
Cu2++Cr'+ = C u + + C r 3 +
and is first-order in each reactant, the rate showing an acid dependence given by
d [Cu' ]/dt

= (a

+b/[H+])[Cu2+][Cr"

At 24.6 "C and 1.00 M C104- concentration, a is 0.17 I.mole-'.sec-' (AH' =


12.5 kcal.mole-', AS' = -20.2 cal.deg-'.mole-') and b is 0.587 sec-' (AH' =
17.1 kcal.mole-', AS' = -2.1 c a l . d e g - ' . r n ~ l e - ' ) ~ ~ ~The
.
stoichiometry was
checked by noting the decrease in absorbance of Cu2+ at 750mp. When CrZf
is present initially in concentrations greater than C u 2 + , then metallic copper is
produced (as a supersaturated solution) according to the step
C u + + C r Z ++C u o + C r 3 +

Cu' is generated also in the reduction of C u 2 + by V2+ in perchloric acid


media92.308. Unlike the C u 2 ++Cr'+ system, the rate is pH-independent and
the second-order rate coefficient is 26.6 l.mole-'.sec-' at 25 "C in the range
0.04-1.0 M hydrogen-ion concentration. AH' and AS' are 11.4 kcal.mole-' and
- 13.8 cal.deg-'.mole-', respectively. By analogy with Vzf reductions the rapid
rate and insensitivity to acid are taken as indicating an outer-sphere mechanism6'.
The greater reactivity of V2+ as compared to Cr2+ is interesting since Cr" is,
in general, a much more powerful reducing agent. C u + cannot be prepared
satisfactorily using Eu2+ as reductant since Cu metal is formed: the second-order
rate coefficient for the C u 2 + + E u 2 +reaction is estimated as
3 l.mole-'.sec-'
at 25 0C92.

9. Oxidations by mercury(I1)
The rate law for the oxidation of V(I11) by Hg(1I)
2 Hg(1I) + 2 V(I1I) = Hg(I),
is complexz6,viz.

+2 V(IV)

OXIDATIONS BY

Hg(I1)

229

The second term contributes only 10-20 % of the overall rate and its precise form
is uncertain. The kinetic parameters have been evaluated for the reaction at 15 "C.
The first term is consistent with the sequence of steps

Hg(I1) V(I11)
Hg(I)+V(III)

+ Hg(l)+ V(1V)
Hg(O)+V(IV)

-+

Hg(I1) Hg(0)

-+

Hg(I),

rapid

The second term suggests the existence of an alternative path-way, possibly


2 V(II1)
Hg(II)+V(II)

+ V(IV)+V(II)
Hg(O)+V(IV)

-+

Hg(II)+ Hg(0)

-,Hg(I),

rapid

Mercury(I1) oxidises Cr(I1) in aqueous perchloric acid solutions according


to the stoichiometric equation

2 Hg(II)+2 Cr(I1) = Hg(I),+2 Cr(II1)


Spectrophotometric measurements on the rate of appearance of Cr(II1) were
made at 408 mp over the temperature range 5-20 "C at p = 2.0 M (ref. 21 1). With
Hg(I1) in excess the rate law is
d[ Cr(III)]/d t = 2 kobs[Hg(II)][Cr( II)]
where kobs= k, +k,'/[H+]. The hydrogen-ion dependent term predominates:
at 20.0 "C, k, = 0.040 I.mole-'.sec-' and k,' = 0.895 sec-'. The k, term corresponds to the reaction between Cr2+ and Hg2+, and the k2' term to Cr2+
HgOH' (k,' = kzKA where k, is the specific rate coefficient and KA is the acid
mole.1-l). At 25 "C, k, is 5.25 x lo3
dissociation constant of HgZ+,2.8 x
I.mole-'.sec-'. Added Hg(I), or Cr(lI1) have no influence on the rate of the
Hg(lI)+Cr(ll) system. However, when Cr(l1) is added to Hg(I), in the absence
of Hg(II), then mercury metal precipitates since Cr(I1) reduces the Hg(I1) formed
in the disproportionation

The following mechanism is likely


Hg(I1) + Cr(I1)
References p p . 267-273

-+

Hg(1) + Cr(II1)

rate-determining

230

COMPLEXES OF D I F F E R E N T METALS: OXIDATION-REDUCTION

Hg(1) Cr( 11)

-+

Hg(O)+ Cr(II1)

rapid

(9.2)

In this respect the final Cr(I1I) product is the monomeric species and not the green
dimer. Furthermore, if Hg(1I) is not i n excess then reaction (9.3) cannot occur,
and mercury precipitates. The reaction is appreciably catalysed by chloride ions.
At 10 "C, the specific rate coefficients for the reaction of C r 2 + with HgCl',
I.mole-'.sec-', respecHgOH', and Hg2' are 1.5, 2.3 x lo3, and 1.75 x
tively. The species HgCI, is much less reactive than HgCI'.
A slow reaction takes place between Hg(I1) and Fe(1I) in perchloric acid solution
viz.

2 Hg(I1) + 2 Fe(I1)

Hg(I),

+ 2 Fe(C1I)

Preliminary work2I2records values for second-order rate coefficients of 2.03 x


I.mole-'.sec-' at 80 "C and 3.75 x
I.mole-'.sec-' at 90 "C. The activation
energy is quoted as 15.5 kcal.mole-', and ASt as -42 cal.deg-'.mole-'. The step
Hg2++Fez'

Hg'

-+

+ Fe3+

slow

is rate-determining and is followed either by


2 Hg+

+ Hg2'+

rapid

or by the two steps


Hg+ + Fe2

-+

Hg + Fe3

Hg+Hg2+ -+ Hg2"

rapid
rapid

In dilute perchloric acid solution the reaction between Hg(l1) and Sn(I1) is
M . After a detailed
complete within 1 min for reactant concentrations of
consideration, Wetton and Higginson' conclude that a two-equivalent primary
reaction is likely.

10. Oxidations by thallium(II1)


The development of fast reaction techniques has allowed a detailed kinetic
study of the TI(III)+V(III) system. Daugherty213followed the course of the reaction by monitoring the appearance of V(IV) at 760 mp. 70-90 % completion of
reaction corresponded to 25-30 sec. Spectrophotometric observations revealed

10

OXIDATIONS B Y

23 1

Tl(III)

no oxidation of V(IV) to V(V)*. The simple rate law


- d [TI( I1I)] /d t = k,,,[TI( 111)] [V( I1I)]

is uncomplicated by V(IV) or TI([) dependences: at reactant concentrations of


M the products of reaction do not affect the rate, even when present at
M . Variations in perchloric acid concentrations over the
concentrations of
range 0.30 to 1.90 M ( p = 2.0 M ) at temperatures 0.5, 14.0 and 25.1 "C show a
- 1.22-order dependence of the apparent second-order rate coefficient on H +
concentration. This result strongly suggests that the mechanism can be visualised
in terms of two parallel steps, of which the most important is TIOH2++V3+.
Activation parameters for this major path are given in Table 27, which includes
T A B L E 27
ACTIVATION PARAMETERS FOR VARIOUS RELATED REACTIONS213

AHT.(kcal. AST(cal.deg- I .
mole- ')
mole- l )

Ner ncrirntion process


TI3+--V3++S-H2O
+ (TIOHV5+)1+H+
F e 3 + - V 3 + i - H 2 0 + (FeOHV5+)Z+H+
Ti3+ - F e 3 + ? H 2 O + ( T i O H F e 5 + ) $ + H +
a

S:,,,,,,.

-lS~+ZSo,,,,,,,,,zl.

13.9
17.6
13.4

4.1i-2.4
-10.2i-5.3
-8.3

St complex

-86
- 128
- 124

ReJ
213
90
b

Higginson, unpublished results quoted in ref. 213.

also comparable information for the Fe3++V3+ and Fe3++Ti3+ reactions.


Replacement of HC104 by HCI results in a pronounced decrease in rate, and the
rate law deviates from pure second order; this chloride effect is attributed to the
reduced reactivity of chloro complexes of TI(IIC1). The rate coefficient of the
V(V)+V(III) reactions3 is too small (by a factor of
14) to account for the
observed rate of formation of V(1V) in the mechanism

TI(III)+V(tII)

+ TI(I)+V(V)

rate-determining
(A)

V(V)+V(III)

2 V(1V)

rapid

and the alternative one-electron process is considered more appropriate, oiz.

Tl(III)+ V(11r) .+ TI(II)+V(IV)

rate-determining

(B)
-rI(rr)+v(rrr)

Ti(r)+v(rv)

'

Higginson ef a1.26 have observed the formation of V(V) in dilute sulphuric acid media in
the presence of an excess ofTl(II1). More V(V) is formed if V(1V) is present initially. In this media
they suggest that mechanism ( B ) operates together with
TI(II)+V(IV) + TI(I)+V(V)
V(V)+V(11l) + 2 V(1V)
ReJcrenws p p . 267-273

232

c o M P L E x E s o F D I F F E R E N T M E T A L S:

O X I DA T I o N-R E D u c T I o N

This proposal is supported by the observation that the normally slow reaction
between TI(II1) and Fe(1I) can be induced by V(II1) (p, 233). Consequently,
Fe(1I) must react with an active intermediate, which is likely to be TI(I1).
The rate of oxidation of V(1V) by Tl(II1) is unaffected by the presence of TI(1)
but is considerably decreased by the addition of V(V)". The rate law is

-d[ V( IV)] k' [TI( 111)] [V( IV)]'


dt

~"CV(V)l+L-V(~V)l

where k'
2.2 x lo-' I.mole-'.sec-'
and k"
42 at 80 "C, [ H f ] = 1.8 M
and p = 3 M . Consequently, the mechanism is similar to the TI(III)+Fe(Il)
i.e.

+
+ T1(11) +V(V)
TI (11) + V(IV) .+ Tl(1) + V(V)

TI(II1) V( IV)

This scheme has a bearing2I4 on the TI(II1) + Tl(1) exchange reaction where,
on the basis of the observed rate law only (rate a[TI(III)][TI(I)]), it is impossible
to discriminate between a single-stage and a two-stage two-equivalent process.
Since the rate of the exchange reaction is comparable to the rate of oxidation of
V(IV) by TI(III), appreciable exchange occurs during the course of the latter
reaction. It is argued2I4 that the addition of TI(1) should increase the rate of the
TI(III)+V(IV) reaction if Tl(I1) were an intermediate in the exchange. However,
as noted above, TI(1) has no such effect, and it is concluded that the exchange takes
place by a single two-equivalent step.
The rate of reduction of TI(II1) by Fe(I1) was studied titrimetrically by Johnbetween 25 "C and 45 "C in aqueous perchloric acid (0.5 M to 2.0 M ) at
p = 3.00 M . At constant acidity the rate data in the initial stages of reaction conform to a second-order equation, the rate coefficient of which is not dependent
on whether TI(I1I) or Fe(I1) is in excess. The second-order character of the reaction confirms early work on this system216. A non-linearity in the second-order
plots in the last 30 % of reaction was noted, and proved to be particularly significant. Ashurst and Higginson' 9 9 observed that Fe(II1) retards the oxidation,
thereby accounting for the curvature of the rate plots in the last stages of reaction.
On the other hand, the addition of TI(1) has no significant effect. On this basis,
they proposed the scheme
TI(II1) + Fe(II)% Tl(I1) + Fe(II1)
k-

TI(II)+Fe(II)

Tl(I)+Fe(III)

10

OXIDATIONS BY

TI(II1)

233

in which Fe(II1) and Fe(I1) compete for the TI(1I) intermediate. The derived rate
law, applicable to the whole course of the reaction, is then
-d[Fe(II)]
dt

k, k2[Fe(II)]'[TI(III)]
k,[Fe(II)]
k-,[Fe(III)]

-2
-

assuming a stationary-state concentration for TI(I1). The reaction was followed


by the disappearance of Fe(I1). From a detailed analysis of the kinetic data, values
for k'lk-, and k , were obtained of 30.8 and 2.41 x lo-' I.mole-'.sec-', respectively, at [H'] = 1.00 M , p = 3.00 M and 25 "C. Contrary to earlier observat i o n ~ " ~ oxygen
,
has no detectable influence on the rate. Strict conformity of the
data to the rate equation rules out the possibility that the alternative route
involving Fe(1V) can make a significant contribution. Increase in acidity decreases
the rate of the reaction, and the results are treated'I5 in terms of a mechanism
involving two pathways (with activated complexes made up of one or two OHions) such that
Kl
k1 = ( K , +[H+])

k;'K2

(k" m)

Plots of k , ( K , + [H+]/K,) uersus l / [ H f ] are linear. The values of k', (intercept)


and k;'KZ (slope) are dependent on the value assigned to K,, the first hydrolysis
constant of TI3+.The latter quantity was estimated from previous data218. K 2
is the second hydrolysis constant of TI3+. Rate parameters are collected in Table
28; k; has an associated activation energy'" of 18.4 kcal.mole-' and a AS'
value215 of -7.3 cal.deg-'.mole-'. The results of Ashurst and H i g g i n ~ o n ' ~ ~
at 25 "C are in excellent agreement with those of Johnson'15. Sulphate ion is
reported to catalyse the TI(1II) + Fe(I1) rea~ tio n " ~On
. the other hand, chloride
ion inhibits the reaction in an unusual manner"7~2'9. The rate exhibits a pronounced minimum at 0.1 to 0.2 M C1- concentration, and attains a limiting value
at
2 M C1- concentration. The conclusion is drawn that TI(1I) must complex
with chloride.
The slow reaction between TI(1II) and Fe(I1) is induced by V(III)'20. The
results indicate that an intermediate, probably TI(II), is formed by the reaction

T A B L E 28
RATE

COEFFICIENTS A N D

25.0
35.0
45.0
References p p . 267-273

EQUILIBRIUM

6.4
9.4
12.7

CONSTANTS

I .41
3.91
9.96

OF

TI(III)+Fe(ll)

I .24
4.44
12.8

234

COMPLEXES O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

of TI(IrI) and V(III), and that this intermediate is capable of reacting with Fe(I1).
The TI(III)+U(IV) system is also subject to induction by V(III), although TI(I1)
is not as reactive towards U ( W ) as it is to Fe(II). Induction experiments on the
effect of V(I1) on the TI(III)+U(IV) system reveal that TI(I1) reacts much more
rapidly with V(I1) than with V(1II). Alternatively, they imply that Tl(II) is not
formed in the TI(III)+V(II) reaction. Spectrophotometric data for the rate of
oxidation of V(I1) by TI(II1) were obtained from measurements at 760 mp, a
wavelength where V(IV) absorbs more strongly than V(lT), V(III), and TI(lI1).
As the reaction proceeds the observed second-order rate coefficient increases. This
effect is attributable to the catalytic influence of V(II1). The following sequence
of reactions is an appropriate mechanistic framework

Tl(III)+V(Il)

-+

TI(I)+V(IV)

(10.1)

+
TI (11) + V(II1)
TI (111) + V( I I I ) -+ TI (11) + V( IV)
TI(I1I) V( I I)

TI(II)+V(II)
TI(II)+V(I11)
T1(111)+2 V(I1)
2 Tl(I1)
v(rr)+v(iv)

--+

(10.2)
(10.3)

Tl(T)+V(III)

(10.4)

TI(I)+V(IV)

(10.5)

T1(1)+2 V(1II)

(10.6)

TI(I)+TI(III)

(10.7)

2 vpn)

(10.8)

Of these steps, the last three can be discounted: (10.6) on the grounds that there
is no significant V(I1) dependence, (10.7) is considered unimportant since TI(I1)
is present only in minute concentrations, (10.8) is slow by comparison with the
other steps in the set (k 0.13 I.mole-l.sec- in 1 M HCIO, at 0 OC)l. Both
rate and stoichiometric data infer that the reaction between TI(IT1) and V(II)
occurs essentially by a two-electron oxidation (step (10.1)). In the presence of
chloride, less V(IV) is produced. It is interesting to note that oxidation of V(II)
by molecular oxygen or hydrogen peroxide generates V(IV). However, the
oxidation of V(1II) by Tl(II1) does not occur as a two-electron step (see p. 231).
The kinetics of the oxidation of tris(bipyridyl)osmium(II) by T1(III)223

TI(III)+ 2 Os(bipy),+ = Tl(1) 2 Os(bipy),,

show deviations from second-order behaviour similar to those encountered in the


TI(III)+ Fe(I1) reaction99. It is probable that an analogous mechanism operates,
uiz.

TI (111) + Os(bipy ),
TI (11) + Os(bipy ),

r TI(11) + Os(bipy ),
-+-TI( I)

+Os(bipy ),

+
+

10

OXIDATIONS BY

TI(III)

235

The reaction was followed by means of the strong absorption of the Os(1I)
complex at 480 mp. Unlike the TI(III)fFe(II) system, there is a slight increase
in rate as the hydrogen-ion concentration is increased. The kinetic data were
interpreted on the basis that both TI3+ and T10H2+ react with Os(bipy),2+
(with rate coefficients k, and k,, respectively). At 24.5 "C and p = 2.99 M ,
k, = 36.0 I.mole-'.sec-' and k, = 14.7 I.mole-'.sec-'; corresponding activation
energies are 6.90 and 11.5 kcal.mole-l. The latter values are considerably smaller
than those for the Tl(III)+T1(1) exchange,,, and for the TI(III)+Fe(II) react i o 1 1 l ~ On
~ . the other hand, all three reactions are subject to retardation by C1ions.
Ce(IV), although a strong oxidant, reacts only very slowly with Hg(I), whereas
TI(III), a two-equivalent oxidant, reacts relatively rapidlyE6,2 2 5 according to
Tl(III)+Hg(I),

TI(I)+2 Hg(1I)

Rate data were obtained by following the rate of disappearance of Hg(I),


spectrophotometrically at 236 mpS6. The rate, unaltered by the addition of TI(1)
but reduced by Hg(II), is represented by

which becomes, on integration

where the subscript o refers to initial concentrations. Plots of the left-hand side
of this equation versus time are linear over a wide range of concentrations of reactants and products. The observed rate coefficient k', obtainable from the slopes
of such plots, is found to be inversely proportional to HCIO, concentration
(2.4 to 6.0 M at p = 6.0 M ) , and also to ClO,- concentration. The following
scheme accounts very adequately for the observed kinetics
Hg,,+ +CIO,-

+ Hg,CIO,+
K3

HgZ2+$ H g 2 + + H g

(10.9)

rapid equilibria

2 TIOH,+ + H +
TIOH2++ Hg 2T I + + H g 2 + + O H - rate-determining
TP+ +H,O

References p p . 267-273

(10.10)
(10.11 )
(10.12)

236

c o M P L E x E s o F D I F F E R E N T M E T A L S:

O X I DA T I o N-R E D u c T I o N

Accordingly, the derived rate law is


-d[Hg(I)']dt

= k , [T10H2'][Hg]

or, alternatively

Thus k' is given by

mole.l-' and K , = 0.073


Using values of K3 = 0.91 I.mole-', K4 = 5.5 x
mole.l-' in conjunction with a k' value of 4.2 x lo-' sec-l (at [H'] = 3 M and
[C104-] = 3 M ) , allows k, to be calculated as 1 x lo6 I.mole-'.sec-' at 25 "C.
Other metal ions known to complex with C104- ion include Fe3+ (ref. 226) and
Ce3+ (ref. 188). Perchlorate complexing of Hgi' has been reported independently'".
Equation (10.13) gives the apparent activation energy E' in terms of the
true activation energy E , [AH* of step (10.12)] and the enthalpies of reaction,
A H , , AH4 and A H , , corresponding to steps (10.9), (10.10) and (10.11) as
E'

E,+AH,+AH,-AH,

(10.13)

Estimated values of A H 3 , AH4 and A H 1 are -0, 10.6 and -0.5 kcal.mole-',
respectively, and E' is 24.4 kcal.mole-' and therefore E , is given as -14+3
kcal.mole-'. The corresponding A S f term is 13 f10 cal.deg-'.mole-'.
It is
noteworthy that the kinetics demonstrate the direct reaction

to be negligible, although the concentration of HgZ2+ionscan be lo5 times


greater than that of Hg atoms. The slow step (10.12) bears a formal resemblance
to the rate-determining step in the TI(I[I)+TI(I) exchange"*, viz.
TIOH'++TI'

+ TI++Tl3'+OH-

particularly as the TI' and Hg species are isoelectronic. However, this step is
lo9 to 10" times slower than (10.12) even though the activation energies are
similar. The profound difference in rate strongly suggests that the principal factor
responsible is the charge of the reactant ions: this is reflected in disparate AS'

10

O X I D A T I O N S BY

TI(II1)

237

terms. Chloride and bromide ions catalyse the Tl(III) + Hg(I)2 reaction (showing
non-linear dependences) by complexing with HgZ+,thus increasing the concentration of Hg atoms by displacing the dismutation step (10.10).
Wetton and HigginsonSg report that the reaction between Tl(JI1) and Sn(I1)
is complete in less than 10 sec at -6 "C in
1.3 M HCI for
lo-' M reactant
concentrations. It appears that it takes place as a single-stage two-equivalent
process without the intervention of Sn(III), since added CO(NH~)~(H,O)CI~+
is not consumed during the course of the reaction.
The two-equivalent oxidation of U(IV) by Tl(III), viz.

TI (111) U(IV)

T1(I) U( VI)

obeys a simple second-order rate law, addition of products having no effect on


the rate2". Rate measurements were made at 650 mp, the absorption maximum
of U(V1). Increase of HC104 concentration leads to a decrease in rate: rate
cr[H+]-" where n ranges from 1.48 at 16 "C to 1.39 at 25 "C. This result is interpreted as arising from the availability of two simultaneous pathways involving
hydrolysed species (UOH3+ and/or TIOH"). The corresponding rate law
-d[U(IV)]/dt

= [T13+][U4+](k;[H+]-'

+ k;[H']-2)

can be written in terms of K , and Kl , the hydrolysis constants of U4+ and TI3+,
viz.

-d[U(IV)]/dt

= [Tl(III)][U(IV)](k;[H+]

+k;)/([H+] +K,)([H+l+ K I )

Thus kobs,the observed (second-order) rate coefficient, is given by

Use was made of known values of K 5 = 0.021 mole.l-' and Kl = 0.073 mole.1-'
at 25 "C and p
3 M , along with estimated values of these constants at other
temperatures (assuming, in each case, an enthalpy of hydrolysis of 11.0 kcal.
mole-'). Plots of kobs([H']+K,)([H']+Kl) uersus [H'] are linear, and k',
and k ; are obtained from the slopes and intercepts. At 25 "C and p = 2.9 M , the
rate parameters for the first path are: k; = 2.11 x
sec-', A H f = 24.6 kcal.
For the second path, the cormole-' and A S f = 16f7 cal.deg-'.mole-'.
responding values are: k; = 2.13 x
mole.l-'.sec-', A H f = 21.7 kcal.mole-'
and ASf = 7+7 cal.deg-'.mole-'. Paths 1 and 2 involve the activated complexes
(UOHTI6+)' and (UOTI5+)', respectively. The formation of U-0 bonds is a
necessary requirement since changes in coordination occur from U(IV) to U(V)
( U 0 2 + )to U(V1) (UOZ2+).Addition of NaC104 increases the rate of the reac-

References p p . 26 7-273

238

c oM P L E X E s oF

D I F F E R E N T M E T A L s:

ox I D A T I o N-R

E Du c T I o N

tion. The effect of various anions and cations on the reaction is summarised as
follows. C1- inhibits the reaction whereas SO4- enhances the rate, results ascribed to the formation of complexes of Tl(111): parallel behaviour is noted in
the T1(TII)+TI(I) exchangezz9, and in the TI(III)+Fe(ll) r e a ~ t i o n ~ Ad.
dition of small amounts of Cu(lI), Ag(I) and Hg(I1) leave the rate unaffected.
These ions have a marked effect on the rate of oxidation of U(IV) by oxygen, a
reaction known to proceed by a chain mechanism230. It is concluded that the
general features of the Tl(III)+ U(1V) reaction are in accord more with a single
two-electron step than with successive one-electron steps for the following reasons:
(a) The reaction is considerably faster than the TI(III)+ Fe(I1) system, generally
accepted to occur via one-electron steps;
( b ) The reaction shows no abnormality in AHt and AS which seems to be
characteristic of processes, e.g. TI(II1) Fe(ll), involving TI(I1).
( c ) If the primary process generated TI(II), then a chain process of the sort

U(IV)+TI(II)

--+

U(V)+T1(1)

U(V) + TI(II1) + U(V1) + TI(I1)


would result. However, there is no such indication, particularly since the rate is
not sensitive to metallic ions like Cu(I1).
The reaction between TI(II1) and U(IV) is one of the few redox reactions
which have been studied in a mixed solventz31.Solutions were kept under nitrogen.
There are striking differences between the rate in aqueous perchloric acid and
methanol-aqueous perchloric acid solutions. In the latter media the order with
respect to TI(III), U(IV), and Hf alters as the solvent composition is changed
(Table 29). For 25 % methanol-75 % water solvent the kinetic orders of 1.0,
1.5 and - 1.33 with respect to U(lV), TI(III), and H + , respectively, are consistent
with the existence of two competing path? whose net activation processes are

TP+ + u4++ H,O

5( T I . H O . U ~ ++) H~ +

2 T 1 3 + + U 4 + + 2H z O ~ ( T I . H 0 . U . 0 H . T l s + ) : + 2H f
T A B L E 29
TI(III)+U(1V)

REACTION: KINETIC ORDERS I N AQUEOUS METHANOLZ3

Order with respect to


Methanol (%)

W V )

TI(III)

Hf

1.oo
1 .oo

1 .oo
1.50
1.25
0.67

-1.39 to -1.48
- 1.33
-0.25
-0.67

25
50
75

0.50
0.33

10

OXIDATIONS BY

239

TI(III)

The corresponding rate law is


- d [U(IV)]/dt = k' [T13 ] [U4
+

3 [H'I-

+k" [Ti3

1' [U4+] [H

]-

'

and

Using values of K , and K , , identical with those made use of by Harkness and
Halpernz2*, the rate coefficients k' and k" were evaluated from the slopes and
intercepts of linear plots of the left-hand side of equation (10.14) versus ([H']

T A B L E 30
Tl(III)+U(IV)

REACTION: KINETIC DATA I N

2 5 % METHANOL-75 % W A T E R

MEDIAz3'

Path

Rare coefficienr(sec-') at 25 "C

AHZ(kcal.mole-')

ASt(cal.deg-'.mole-')

k'
k"

0.0041
0.98

3.11
4.47

-59.0
-43.6

Ionic strength, 2.9 M .

Kl)/[Tl(III)]. Table 30 contains these results together with AH' and A S f values.
The reaction in 25 % methanol is faster than in water, although ASs is strongly
negative. To some extent this may be rationalised in terms of the structure of the
activated complex (TI.HO.U-OH.TI*')*, since the formation of U0,'+ would
appear to be facilitated by the positions of the two oxygen atoms relative to the
U atom. The role of the solvent in governing the progress of the reaction is discussed by Jones and A m i ~ ' ~ ' In
. contrast to the reaction in aqueous solution,
chloride is ineffective as an inhibitor and sulphate is a poor catalyst in 75 %
methanol media. Moreover, Cu(1I) and Hg(I1) exert a marked catalytic effect
whereas Ag(I) strongly retards the reaction. These latter observations recall those
recorded for the U(IV)+02 system230.A step-wise process, involving U(V) and
TI(II), is advanced, uiz.

Tl(1II) U(1V) + Tl(I1) U(V)


U(V)+Cu(ll)

+ U(VI)+Cu(I)

TI(II)L Cu(1) + TI(1) Cu(1I)


Ag(1) is believed to act as an inhibitor, as is likely in the U(IV)+02 reaction,
by bringing about a chain-breaking step with U(V): the well-defined induction
period is proportional to the initial Ag(1) concentration. During this time, a
References pp. 267-273

240

C 0M P L E X E S 0 F D I F F E R E N T M ETA L S : OX I D A TI ON-R E D U CTI 0 N

colloidal suspension of metallic Ag is formed, and after this is complete the reaction proceeds at its normal rate.
Wearz3' has subjected the TI(III)+U(IV) reaction to a close scrutiny, the
kinetics being studied over a 2000-fold range of reactant concentrations. Solutions
were kept under nitrogen. The range of concentrations examined by Harkness and
Halpernzz8 were U(1V) = 3.5 x
to 11.0 x
M , TI(rI1) = 5 x
to
21 x
M. Wear finds that the orders with respect to the reactant vary with
T A B L E 31
TI(III)+U(IV)
Region

I
I1
111
IV

REACTION:

V A R I A T I O N OF K I N E T I C O R D E R S W I T H C O N C E N T R A T I O N

1 0 3 [ U ( W ) ]( M ) 103[Tr(III)]( M ) [ H + ] ( M ) U ( I V ) order TI(II1) order H+order


0.5-50
0.03-0.5
6.3
0.07-0.22

1-50
5.0
0.02-1.0
0.12-0.36

0.25-2.75
0.90
0.90
0.50-1.90

0.88
1.2
0.88
1.6

0.90
0.90
0.67
0.38

-1.4
-1.4
-1.4
-1.5

IoNc strength, 2.9 M ; temp., 25 "C. From Wearf32 (by courtesy of Sandia Laboratories and

U.S. Atomic Energy Commission).

concentrationz32.His results are summarised in Table 3 I*. Qualitatively, the data


are equally consistent with a chain mechanism

Tl(II1) + U(1V) --+ TI(II) U(V)


Tl(ll)+U(IV)

TI(Il)+U(V)
2 U(V)

+ TI(I)+U(V)

-+

TI(I)+U(VI)

U(IV)+U(VI)

2 TI(I1) + Tl(T)+Tl(ITI)
Tl(III)+U(V)

+ TI(II)+U(VI)

Love et al.z33have examined the oxidation of U(1V) by Tl(II1) in the presence of


tartaric acid. The kinetics are complex: tartaric acid increases the initial rate but,
in its presence, reaction ceases before all the TI(II1) is consumed. The disagreement between the rate coefficients of Love et al. and those of Harkness and
Halpernzz8 is ascribed to the presence of Fe(II1) impurities in Tl(II1) stock
solutions. However, as pointed out by Newton and Bakerz5O, this explanation
cannot be valid since the results of Harkness and Halpern are not dependent on
the initial reactant concentrations.

The rate laws reported by Wear2j2 have been criticised by Newton and Bakerzs0.

11

OXIDATIONS BY

24 1

Pb(IV)

11. Oxidations by lead(1V)


The reduction of Pb(1V) acetate by Co(I1) acetate in acetic acid234 exhibits
more complex kinetics than the Pb(1V) + Ce(II1) system (p. 242). The expected
stoichiometry of 1 : 2, corresponding to
Pb(IV)+2 Co(11) = Pb(11)+2 Co(II1)
is found in the anhydrous solvent but the presence of water causes a change in
stoichiometry (Fig. 9). Alcohol also brings about a similar effect (the addition of
10 % by volume of methanol causes the stoichiometry to be reduced to 1 : 1.1).
The fact that methanol and water d o not effect the decomposition of Pb(IV) or
of Co(rI1) acetates, under these conditions, strongly suggests that these additives
must be reacting with a transient intermediate, thus causing the final concentration of Co(II1) to decrease. Kinetic data, for the most part obtained for the appearance of Co(I1I) at 400 mp, show that the order with respect to Pb(1V) is
unity under all conditions, but that the order with respect to Co(11) is non-integral
(1.5). At 25.0 "C the observed rate coefficient is 1.70 I*.mole-*.sec-'; the apparent activation energy and entropy are 17.5 kcal.mole-' and -8 + 4 cal.deg-'.
mole- respectively. Furthermore, the presence of Pb(1I) causes the order in
Co(I1) to increase. Of the two reaction products, only Pb(I1) in large excess causes
a retardation. The kinetic observations are interpreted in terms of a scheme in-

',

20

I
40

60

[H20] ( M I

Fig. 9. Effect of water o n the stoichiometry of the Pb(LV)+Co(lI) reaction in acetic acid at
and 37 "C ( 0 ) .(From Benson et
by courtesy of The Faraday
temperatures of 23 "C (0)
Society.)
References pp. 267-273

242

co M P L E x ES o F D I F F E R E N T METALS: ox I DA T I o N-R E D u c T I o N

volving a Co(l1) dimer, and Pb(Il1) and Co(IV) as intermediate species, ciz.

CO(II)+C~(II) [co(rr)12
Pb(IV)

+ [Co(II)], -,Pb(I1) +2 Co(II1)


Pb(IIr)+co(rIr)+co(rI)

Pb(rv)+ [c0(11)],

-+

Pb(IV)+ [Co(II)],

-+ Pb(Il)+Co(IV)+

Pb(IV)+ Co(1I)
Pb(IV)+Co(II)

Co(1I)

+ Pb(II)+ Co(IV)
-+

Pb(III)+Co(III)

co(rv)+co(r~) 2 c O ( q
Pb(III)+Co(II)

-+

Pb(II)+ Co(1II)

The postulation of the + 4 oxidation state of cobalt is necessary to account for


the retarding influence of Pb(II). The existence of a dimeric species of Co(I1)
acetate is required by the rate law and is confirmed by spectrophotometric and
solubility
The existence of ionic species of the reactants is
inferred by the rate increase on addition of sodium acetate, an observation which
cannot be attributed to a salt effect because sodium perchlorate produces a rate
decrease. On this scheme an explanation of the effect of water on the stoichiometry
is that the step
co(Iv)+co(II)

-+

2 Co("1)

is eliminated, Co(1V) being reduced directly by water, possibly by


Co(IV)+H,O = C0(11)+2 H + + &0,
The even greater influence of methanol may be due to its reaction with Pb(1II).
Benson and S ~ t c l i f f e 'have
~ ~ made a kinetic study of the reaction between
lead(lV) acetate and cerium(I1I) acetate in anhydrous acetic acid, viz.

Pb(1V) + 2 Ce(lI1) = Pb(II) 2 Ce(l3')


The reaction was followed by observing the appearance of the yellow colour of
Ce(1V) at 400 mp, with Pb(IV) present in excess concentration. Pb(1V) was varied
in the region 8.6 x
M to 4.4 x
M while Ce(II1) was kept at
4x
M. A practical difficulty encountered was the photochemical instability of Ce(1V)
acetate. Under the above conditions and in the temperature range 30-47 "C, the
reaction is strictly first-order in each reactant. The observed rate coefficient at
30.0 "C is 1.48 x lo-' I.mole-'.sec-' and the apparent activation energy and

12

OXIDATIONS BY

Ce(IV)

243

entropy are 19.5 kcal.mole-l and -2.5f 1.5 cal.deg-.mole-, respectively.


The rate is unaltered by the presence of Ce(1V) at similar concentrations to that
of the initial Ce(II1) concentration. Similarly, Pb(I1) has no influence on the rate,
unless present in very large excess when a slight increase is observed. The presence
of sodium perchlorate produces a marked retardation and the results were treated
in terms of an ionic strength effect. On the other hand, added sodium acetate
increases the rate, there being an initial first-order dependence on the concentration of this salt. The addition of ethanol or benzene leaves the order of reaction
unaffected, but both these additives produce an increase in rate. In the case of
ethanol there is a first-order dependence, whereas for benzene a linear relationship
exists between log kobsand the reciprocal of the bulk dielectric constant, suggesting
a solvent effect. Since both Pb(II) and Ce(1V) fail to retard the reaction, the
mechanism is probably
Pb(IV)+Ce(III)

--t

Pb(lII)+Ce(IV)

slow

Pb(IlI)+Ce(III)

Pb(II)+Ce(IV)

rapid

although no direct evidence was adduced for the participation of Pb(I1I). In


the presence of sodium acetate, ionic species take part in the reactionz3, e.g.,
Pb(OAc),-, Pb(OAc)5- and Ce(OAc)4-. Confirmatory evidence for these
comes from ion-migration and spectrophotometric studies. The effect of ethanol
is attributed to the formation of a reactive complex between the alcohol and
Pb(IV). In this respect, it is interesting that ethanol also increases the rate
of oxidation of t-butyl hydroperoxide by Pb(IV)238.In the latter system a complex
exists between the organic substrate and Pb(IV), the rate-controlling step is
Pb(IV).ROOH + ROOH

-+

products

and the reaction is second-order in peroxide. Ethanol supplants the peroxide


molecule from the complex so that the step becomes
Pb(IV).EtOH + ROOH

-+

products

and the reaction order with respect to peroxide approaches unity.

12. Oxidations by cerium(1V)

In acidic sulphate media, the multi-equivalent oxidation of Cr(II1) by Ce(1V)


3 Ce(IV)+ Cr(rr1)

3 Ce(rrr)+ Cr(V1)

takes place at a convenient rate for spectrophotometric measurements239. HowReferences pp. 267-273

244

co M P L E x ES o F D I F F E R E N T M E T A L s: O X I DA T I o N-RE D u c T 10N

ever, the reaction was found to be inconveniently fast in perchloric acid solutions.
Kinetic results were obtained at 492 mp and 500 mp, wavelengths where Ce(II1)
is transparent but Ce(IV), Cr(II1) and Cr(V1) absorb to varying degrees. No
evidence was obtained for the presence of dimeric species of Ce(1V); this reactant
is presumed to be highly complexed with sulphate (or bisulphate) ions240t. Although dimers of Cr(V1) are possible, monomer and dimer have similar absorptivity values at the chosen wavelengths''. The reaction demonstrates an inverse
dependence on the concentration of Ce(III), and a square dependence on Ce(IV)
concentration, viz.

Consequently, the activated complex (X*) of the rate-determining step is composed


of one Ce atom and one Cr atom, the average oxidation state of each atom being
A sequence of one-equivalent steps
+ 4 [x' = 2 c~(Iv)+c~(III)-c~(III)].
are in accord with the rate law, viz.

The concentration of Cr(1V) is maintained by equilibrium (12.1) as

and since the rate is coatrolled by (12.2)


rate

k,[Ce(IV)][Cr(IV)]

k ,k [ce( IV)] [Cr (III)]

(12.4)

k-lCWWI
Thus the observed rate coefficient kobsis equivalent to k, k z / k - , . Because of the
limitations of the kinetic procedure, there is some small doubt over the inclusion
of a term in Ce(IV) concentration in the denominator of (12.4). However, the
relative insignificance of this term means that k - , [Ce(III)] > k2[Ce(IV)]. Also,
the inequality k3[Ce(IV)] >> k - , [Ce(III)] is a consequence of the observed
kinetics. A? in the Cr(VI)+Fe(II) system5*, the slow stage involves the inter+ See, also, ref. 309.

12

OXIDATIONS BY

Ce(1V)

245

conversion of Cr(IV) and Cr(V). There is strong evidence that such a transformation requires a change in coordination number of the metal ion. The observed
rate coefficient shows a n approximate inverse dependence on the square of HSO,concentration.
The cerium(1V) oxidation of Mn(I1) has been briefly reported on by Aspray
et ~ 2 1 . ~ The
~ ' ~ reaction
.
was followed from the rate of appearance of Mn(II1) at
505 mp, using solutions in 4.5 M sulphuric acid. The results indicate the occurrence
of an equilibrium
Ce(IV)+ Mn(I1)

+ Ce(III)+ Mn(lT1)

At 20.8 "C, [Mn(II)] = 8.21 x lo-' M , [Ce(IV)] = 1.63 x


M , the forward
rate coefficient is 0.213 I.mole-'.sec-'; the corresponding equilibrium constant
is 0.015, in good agreement with redox potential data. Rechnitz et aLZ4lbhave
examined the system in greater detail. They find the values for the apparent rate
coefficients of the forward and backward steps to be 0.435 l.mole-'.sec-' and
5.9 I.mole-'.sec-', respectively, in 3.0 M sulphuric acid at 25 "C. Rates were
measured by monitoring cerium(1V) at 400 mp. The forward path has A H f and
ASt values of 13.0 kcal.mole-' and - 17 cal.deg-'.mole-'; the corresponding
values for the Ce(III)+ Mn(II1) reaction are 15.9 kcal.mole-' and - 1.7 cal.
deg-'.mole-'. From the observed dependences on reaction media the kinetically
important Ce(1V) and Mn(1II) species are Ce(S04)2 and MnOH".
The kinetics of the oxidation of Fe(1I) by Ce(IV) in aqueous perchloric acid
have been studied, using reactant concentrations in the range lo-' to
M
(ref. 242). A quenching method was utilised to monitor the disappearance of
Fe(I1). The reaction conforms to a 1 : 1 stoichiometry and is of simple second
order, viz.

- d[Fe(II)]/dt

k' [Ce(IV)] [Fe(II])

However, the intercepts of log [Fe(II)]/[Ce(IV)] versus time plots deviate from
the values expected for the initial concentrations of the reactants. This "apparent zero-time oxidation", which is reproducible, is believed to result from a
finite quenching time, and the reaction of Fe(l1) with a very reactive Ce(IV)
species. Added amounts of Ce(1II) and Fe(II1) leave the rate unaffected. At
constant ionic strength, k' varies inversely with hydrogen-ion concentration in
the range 0.05 to 1.00 M ; for [H'] > 1.0 M , k' increases with increasing [H'].
In general

k'

= u [ H + ] + b + c/ [ H + ]

which signifies reaction by three simultaneous routes. These are


References p p . 267-273

246

coM P L E x E s oF

D I F F E R E N T M E T A L S: O X I D A T I o N-RED

u cT I O N

Ce(111) + Fe( 111)

Ce4+ Fe( 11)

CeOH3++Fe(I1) 3 Ce(ll1) + Fe(II1)


Ce(OH)22++ Fe(I1) A Ce(II1) + Fe(II1)
Therefore
(12.5)
where Kl and K , are the hydrolysis constants of Ce4+ and CeOH3+,respectively.
When [H'] < 1 M , Kl >> [H'] and equation (12.5) reduces to

In accordance with this relationship, plots of k' versus K 2 / ( K 2+ [H']) are fairly
linear, using a value of 0.08 for K 2 . Values of k , and k6 are obtained from the
slopes ( = k,-k,) and intercepts ( = k,) of such plots. At 0.3 "C and p = 2.0 M ,
k, = 5865+ 1500 1.mole-'.sec-',
k , = 1000+200 I.mole-'.sec-',
and k , =
4830+500 I.mole-'.sec-'. At unit ionic strength, k' has an associated AH'
of 9.4 kcal.mole-', and a AS' of -6+3 cal.deg-'.mole-'.
Dainton et uLZ4',
suggest that the similarity between k , and k , may arise from common transition
states of the type (CelV-OH3+ . . * Fe") and (OHCe1"-OH2+ * * Fe"). Electron
transfer may take place by hydrogen-atom transfer, or electron conduction through
a hydrogen-bonded intermediate. The possible mechanism of the direct aciddependent path is discussed in terms of
( a ) charge transfer through symmetrical bridges of protonated water molecules,
i.e.
H

(H,O),,CeiVO . . H-0'-H
I

H
*

OFe"(H,O),
I

and
( 6 ) electron tunnelling through extended acid chains of the type -(HClO,-),
which may exist in cold, concentrated acid solutions.
The Ce(lV)+Fe(II) system is catalysed by HS04- and F - , but not by CIions. Since there is a linear relationship between k' and [HSO,-], the pathway
CeS0,"

+Fez+

5Ce(III)+Fe(III)

12

OXIDATIONS BY

Ce(1V)

247

is likely. Assuming a value of 3500 for the equilibrium constant of

Ce4+ HS04-

+ CeS04'+ + H+

then k , is
5 x lo3 I.mole-'.sec-' at 0 "C and p = 0.23 M. This value agrees
quite well with that calculated from the data of Dulz and S ~ t i for
n ~the~ reaction
in sulphuric acid solution.
Dulz and S ~ t i have
n ~ ~measured the rates of oxidation of Fe(I1) and tris(1,lOphenanthroline) Fe(I1) complexes by Ce(IV) with a view to test the linear free
energy relationships' predicted by M a r c u 244
~ .~ Th
~ ese
~ ~rapid reactions were
studied in sulphuric acid media using a stopped-flow apparatus. In 0.50 M
H2S04 at 25 "C the Ce(IV)+Fe(II) and Ce(IV)+Fe(phen),'+ systems have
~ I.mole-'.sec-'
and 1 . 4 2 ~10' 1.
(second-order) rate coefficients of 1 . 3 lo6
mole- '.set- respectively. The corresponding activation energies are 9.5 kcal.
mole-' and 6.5 kcal.mole-'. According to the Marcus theory of electron transfer
reactions43, if k l , z and Kl,2 are the rate and equilibrium constant, respectively,
for the oxidation-reduction reaction, and kl, and k z , z are the rate coefficients of
the exchange reactions, then

',

where f is defined by
(12.7)

Z is the collision frequency of two uncharged molecules in solution, and has a


value of 10" I.mole-'.sec-'. In terms of free energies, equation (12.6) can be
written as
AGi, 2 = 0.50 AGi. +0.50 AG;, z+0.50 AGO,, 2 - 1.15RTlogf

(12.8)

where AG' represents the free energies of activation of the various processes,
and AGO,, is the standard free-energy change of the redox reaction. Consequently,
a linear relationship between (AG;,,-0.5
AG;, 1 ) and AG?, is predicted.
Fig. 10 contains such a plot for the Ce(1V) oxidation of a number of Fe(I1)
phenanthroline complexes. Similar data are included for the oxidation of Fez+
by various Fe(II1) phenanthroline complexes in 0.50 M HC104 at 25 0C196.Rate
coefficients and (formal) oxidation potentials are set out in Table 32. A straight
* Other linear free-energy relationships are those for the oxidation of a series of Fe(I1)
phenanthroline complexes by Co(II1)' 9s and Mn(III)S2.
References pp. 267-273

248

C 0M P L EX E S 0F D I F F E R E N T M ETA L S : 0X I D A TI 0N-R E D U C T I 0N

54-

3-

21-

-2

-4

-6

-8

I
-10

I
-12

I
-14

A G$(kcal mole-')

Fig. 10. Relationship between (AG:,,,-0.5 AG',.,) and the standard free energy change (AGOlz)
of the redox reactions at 25 "C. Open circles, Ce(IV)+Fe(phen)32+ reactions in 0.50 M H 2 S 0 4 .
Closed circles, Fez++ F e ( ~ h e n ) , ~ reactions
+
in 0.50 M HC104. Numbers refer to complexes in
~ ~courtesy
,
of The American Chemical Society.)
Table 32. (From Dulz and S u ~ i nby
T A B L E 32
Fe(l1)
HZS04A T 25.0 0C44

R A T E COEFFICIENTS FOR T H E OXIDATION OF

Complex
Iron (11)
Tris(3,4,7,8-tetramethyl-l ,lo-phenanthroline)iron(lI)
Tris(5,6-dimethyl-l ,lo-phenanthroline)iron(II)
Tris(5-methyl-1,1O-phenanthroline)iron(II)
Tris(1,IO-phenanthroline)iron(lI)
Tris(5-phenyl-l,lO-phenanthroline)iron(lI)
Tris(5-chloro-I, 10-phenanthroline)iron(ll)
Tris(5-nitro-1 ,lo-phenanthroline)iron(lI)

COMPLEXES BY

E o of coniplex (V)'

(1)
(2)
(3 )

0.68
0.83
0.99

(5)

1.04
1.08
1.10

(6)
(7)

1.14
1.26

(4)

Ce(IV)

IN

0.50 M

k(l.mole-'.sec- ')
1.3 \
1.6 x
4.3
2.2 x
1.42
1.2 x
2.5 Y
3.9

10'

lob
105
105

105
105

104
103

The formal oxidation potentials of the complexes.

line of slope 0.50 has been drawn through the data for reactions which have
AGO,, close to zero, when equation (12.8) simplifies to
A G S2 = 0.50 AG:, +0.50 AGis 2+o.50 AGO,,

(12.9)

The intercept, 6.5 kcal.mole-', leads to a value of 13.0 kcal.mole-' for the
(average) AG' of the various phenanthroline Fe(I1)-Fe(lI1) exchanges. Such a
AG:, value corresponds to an average rate coefficient, k2, 2 , of 2 x lo3 1.mole-'.

12

OXIDATIONS BY

Ce(1V)

249

T A B L E 33
COMPARISON OF OBSERVED A N D CALCULATED RATE COEFFICIENTS AT

Reaction

k , . , (/.molesec- I )

k2,2(l.mole-I .
sec-I)

4.0
5 2 XIO-5
1.ox 1 0 - 2

II XIO-4
5 2 ~10-

4.4
4.0
4.0
4.0

-.

-5
-5

1.0 Y 1 0 - 2
s 2 XIO-5
4.0

k l , 2 calculated
(Lmole- '.see- I )

-.

Fe(I1) +Ce(IV)
Cr(II)+Fe(III)
V(II)+Fe(III)
Eu(II)+Fe(III)
Cr(I1) +Co(III)
V(II)+Co(III)
Cr2 +V3
FeZ++Co3+

k , ,2 observed
(/.mole-'.set- I )

25.0 V4

1.3 x lo6

8 X103

> 105 a
> 105 a
> 3 x 102a
> 3x102"

I . O v 10-2

42 a

-5

6 x lo5
1 6x

lo5

9 X lo5
2 2 x 106

51 x 10'0
-2X1010
2-2 x 10-2
-6x1O6

From ref. 24.

sec-'. The curve (AGO,, more negative than 5 kcal.mole-') was calculated from
equation (12.8) using this value for k z , in equation (12.7). The general conclusion
is that the data adequately comply with the predictions of the Marcus theory.
Table 33 is a collection of observed and calculated rate coefficients for a variety
of redox reactions. Adamson et ~ 1 . ~have
~ ' criticised the value of kl,z for the
Ce(IV)+Fe(II) system, as calculated by Dulz and S ~ t i non
~ ~the, grounds that
the rate of the Fe(JII)+Fe(II) exchange (kl, 1) and the corresponding oxidation
potential relate to HC104 media, whereas the rate ( k 2 , 2 )and oxidation potentia
of the Ce(IV)+Ce(III) system are for H2S04 media. Adamson e l ~ 2 1 . ~arrive
~'
at
a calculated value of 1.3 x lo6 l.mole-'.sec-' for the rate coefficient (k1,2)of the
Ce(IV)+Fe(II) reaction in 0.5 M HC104 at 0 "C. Since this value is very much at
variance with the observed value (700 1.mole- '.sec-'), they conclude that this
oxidation takes place by an atom-transfer mechanism, to which the theoretical
treatment of Marcus is not appropriate.
Rate coefficients (and oxidation potentials) are given in Table 34 for the Ce(IV)
T A B L E 34
SECOND-ORDER

RATE COEFFICIENTS

Ce(1V)

IN

Fe(phen),2+
Fe(phen)2(CN)2
Fe(phen)(CN)+'Fe(bipy),2+
Fe(bipy)2(CN)Z
Fe(bipy)(CN)42Fe(CN)64a

From ref. 44.

References pp. 267-273

0.5 M

FOR

1.07
0.79
0.65
I .05
0.81
0.67
0.69

O F Fe(l1)
25.0C195

THE OXIDATION

SULPHURIC ACID AT

0.142 a
7.1 1
8.88
0.196
8.40
12.5
I .90

COMPLEXES

BY

250

C O M P L E X E S OF D I F F E R E N T METALS:

ox I DA T I O N-RE

DU CTION

TABLE 35
C O M P A R I S O N OF O B S E R V E D A N D C A L C U L A T E D R A T E C O E F F I C I E N T S ' 9 5

kl.' obserued(l.rnole-'.sec-')

Reaction

> 108
1.9~10~
1 . 4 107
~
6.1 X lo7
3.8 x 105
1.9 x lo6
5 . 0 lo6
~
3.0 x 104
4.3 x 104

Ce(lV) W(CN)84Ce(1V) +Fe(CN)64Ce(IV)+ M O ( C N ) ~ ~ IrC16'- +W(CN)84IrCI6'- +Fe(CN)641 ~ 1 ~ +2 M


- ~ ( c N ) ~ ~ MO(CN)8'- +W(CN)S4Mo(CN)s3- +Fe(CN)64Fe(CN)63- W(CN)84-

k l B zcalculated(l.rnole- '.set-')
6.1 x lo8

6.0 x
1.3 x
8.1 x
5.7 x

lo6
107
107

105

1 .o x 106

1.7 x 107
2.1 x 104
5.1 x 104

oxidation of Fe(bipy),'+, Fe(bipy),(CN), , Fe(bipy)(CN),'-,


Fe(CN),,-, and
corresponding phenanthroline complexes' 9 5 . Mixed-ligand complexes react more
rapidly than complexes containing identical ligands.
Campion et
have compared the rates of oxidation of W(CN),,-,
Mo(CN),~-, and Fe(CN),,- by Ce(IV) with those calculated from the Marcus
theory,,. The results on these systems are given in Table 35 together with data
on a number of related reactions. Agreement between observed and calculated
rate coefficients is good.
In conforming to an expected linear free energy relationship, the Ce(1V) oxidation of various 1,lO-phenanthroline and bipyridyl complexes of Ru(I1) in 0.5 M
sulphuric acid are consistent with the requirements of the Marcus treatment','.
The results for the oxidation of the 3- and 5-sulphonic-substituted ferroin complexes by Ce(IV) suggest that the ligand does not function as an electron mediator,
and that the mechanism is outer-sphere in type. Second-order rate coefficients for
the oxidation of Ru(phen),'+, Ru(bipy),'+, and Ru(terpy),'+ are 5.8 x lo3,
8.8 x lo3, and 7.0 x lo3 l.mole-l.sec-', respectively, in 0.5 M H2S04 at 25 "C;
a rapid-mixing device was employed.
The rate of oxidation of mercury(1) by Ce(1V) is slow in any medium but
N 3.6 times faster in 2 M perchloric acid than in 1 M sulphuric acid, achieving
a maximum in the former medium at 4 M , and then decrea~ing'~,.Sulphate ion
retards the reaction: the rate increase observed in HC104 solutions is ascribed to
the formation of less complexed, more reactive species of Ce(1V). The kinetics
of the reaction between Hg(I) perchlorate and Ce(1V) sulphate have been examined
in 2.0 M perchloric acid at 50.0 "C, under which conditions the rate law

- d[Ce(IV)]/dt

k'[Ce(IV)][Hg(I),]

obtains, Hg(I1) and Ce(II1) having no effect on the rate. Mechanistically the reac-

12

OXIDATIONS BY

25 1

Ce(1V)

tion is best envisaged as


Ce(IV)+ (Hg-Hg),'
Ce(IV)+Hg(I)

-+

Ce(III)+Hg(I)+Hg(II)

slow

Ce(III)+Hg(II)

rapid

a scheme in which the Hg-Hg bondis broken simultaneously with electron transfer.
At 50 "C in 2.0 A4 HClO,, the rate coefficient of the slow step is given as 0.14
I.mole- '.set- ';the corresponding apparent activation energy is 14.4 kcal.mole-'.
Silver(]) is effective as a catalyst in the oxidation of mercury(1) by Ce(IV)26.
In dilute HCIO, solutions no direct reaction takes place between Ce(1V) and
mercury(I), or between Ce(IV) and Ag(1). The catalysed reaction obeys, in the
presence of excess Hg(l),, the expression
-d[Ce(IV)]/dt
=

= -2 d[Hg(I),]/dt

(12.10)

2 ki[ce(Iv)] = 2 k,[Ce(IV)][Ag(I)],

where [Ag(I)], represents the initial concentration of Ag(1). When Hg(l), is


in small excess only and Ce(lI1) is present initially, first-order plots for the rate
of disappearance of Ce(1V) show curvature, the observed rate coefficient (kk)
decreasing as Ce(II1) increases. This result suggests a back-reaction involving
Ce(ll1). The scheme proposed is
Ce(IV) +Ag( I)

2 Ce(II1) +Ag(I1)

k-8

Ag(II)+Hg(I),

2Ag(l)+Hg(I)+Hg(II)

Ce(IV)+Hg(l)

-+

Ce(IlI)+Hg(ll)

rapid

Assuming the steady-state hypothesis to apply to Ag(l1)

If Hg(I), is in large excess, equation (12.11) reduces to (12.10) since k - ,

-= k9

(k-,/k9 = 0.198 at 1.5 M [H'], p = 3.0 M and 20 "c).

In mixed solutions, 2.0 M i n HC104 and 0.1 M i n H2S0,, the rates of oxidation
of Hg(I), by Ce(IV), as catalysed by Ag(1) and Mn(Il), are 247
- d[Ce(IV)]/dt = 0.304[Ce(IV)] [Ag(l)] +O.O218[Ce(IV)][Ag(l)]/[Hg(I),]

and

- d [Ce(lV)/dt

1.28[Ce(IV)] [Mn(lI)] 0.143 [Ce(lV)] [Mn(ll)]/[Hg(I),]

respectively, at a temperature of 50.0 "C (time expressed in min); corresponding


References pp. 26 7-273

252

co M P L E x E s o F D I FF E R E N T M E T A L S :

O X I D A T I ON-R E D u c TI O N

activation energies are 13.5 and 12.1 kcal.mole-'. In the latter case the scheme
suggested is
Ce(IV)+ Mn(I1)
Mn(III)+ Hg(I),

Ce( IV) Hg( I )

+ Ce(IlI)+Mn(III)
-+

Hg(I)+Hg(Il)+ Mn(I1)

-+

Ce(II1) Hg(I1)

rapid

Ce(IV)+ Mn(II1) -+ Ce(IIl)+Mn(lV)


Mn(lV)+Hg(I),

-+

Mn(U)+2 Hg(I1)

although this does not explain the inverse dependence on [Hg(I),] at low concentrations. Spectrophotometric evidence was adduced for the presence of
Mn(1II) as an intermediate. The Ce(IV)+ Hg(I)2 reaction, doubly catalysed by
Ag(I)+ Mn(II), was examined in both HC104 and H2S04.In the former medium,
the kinetics are complex, and the rate shows an enhanced catalytic effect; in the
latter medium, the kinetics are analytically soluble, and the catalytic effect is
additive. Tentative mechanisms are proposed.
The mechanism proposed by Dorfman and G r ~ d e r , ~to
' account for the reduction of Ce(IV) by Tl(I), in 6.18 A4 nitric acid at 54 "C, includes a dimeric species
of Ce(IV) and a Ce(1V)-Ce(II1) binuclear species, viz.

+ [Ce(IV)],
Ce(III)+ Ce(1V) + [Ce(Ill)Ce(lV)]
2 Ce(IV)

klo
+
k-

Ce(lV)+ OH-

K3

K4

Ce(III)+ OH

10

T1( I) + OH

kr
+
T1( 11)+ OHk-it
I

k12

Ce(IV)+TI(I)

f
k-12

Ce( IV) +Tl(II)

Ce(III)+Tl(II)

2 Ce(Ill) +TI(III)

Using concentrations in mole.1-' and time in sec, Dorfman and Gryder find
K , = 18, K4 = 2, k , , = 1 . 3 3 lo-',
~
k 1 2 = 3.81 x
k-.lo/kll = 0.021,
k - l o k - , , / k l 1k , , = 5.4 x lo-', k - , , / k , , = 2.6 x lO-,,and k - , 2 / k , , = 1.52 x lo-,
The Ag(1)-catalysed oxidation of Tl(1) by Ce(IV) can be explained26 by a
series of reactions
Ce(IV)+Ag(I)

ks
+

Ce(III)+Ag(II)

k-8

Ag( 11) +T1( I)


Ce(IV)+TI(II)

2 Ag( I) +T1(11)
-+

Ce(III)+TI(III)

rapid

13

OXIDATIONS BY

u , Np,

PU

253

analogous to those for the Ag(1)-catalysed reaction of Ce(IV) and mercury(I)26.


From these the derived rate law is

However, the term k-,[Ce(III)] cannot be neglected since k-8/k14 is 35.7. Neither
can k, be evaluated from first-order plots for the disappearance of Ce(1V). The
integrated form of equation (12.12) allows k-,/k,, and k, to be obtained. The
ratio k,lk,,, obtained from k-,/k9 and k-,/k14, is 180 at 1.5 M [H'] and
20 "C. This corresponds to the ratio of the rate coefficients for the oxidation of
Hg(I), and of Tl(1) by Ag(I1). Higginson et dZ6
quote a value of 185 for the ratio
of rate coefficients for oxidation of these two species by Co(II1). The dependence
of k, on hydrogen-ion concentration is described. Schenk and Bazzelle31 have
studied the Ce(1V) +TI([) system, both uncatalysed and catalysed by Ag(1) and
Mn(III), in sulphuric acid media.
Cerium(1V) oxidises tin(I1) in aqueous sulphuric acid probably by a two-step
path involving Sn(III)249. At low Sn(1V) concentrations and low sulphate concentration the reaction is second order, and the suggestion is made that the reactant species are Ce(SO,),'- and SnSO,. In mixed chloride-sulphate media the
Ce(1V) Sn(I1) reaction, in the presence of trioxalatocobaltate(lII), produces
an intermediate which consumes the Co(II1) complex5'. This result is interpreted
as being evidence for the presence of Sn(II1) in the reacting system.

13. Oxidations by uranium, neptunium and plutonium


The redox reactions of the actinide elements have been the subject of a recent
and authoritative review by Newton and BakerZS0.The net activation process
concept is used to interpret the experimental data. Empirical correlations shown
to exist include those between the entropies of the activated complexes and their
charges, and, for a set of similar reactions, between AGt and AGO, and A H f and
AH'. The present state of the evidence for binuclear species is discussed.

13.1

OXIDATIONS BY URANIUM(VI)

The oxidation of vanadium(l.1) by u r a n i ~ m ( V l ) ~viz.


~',
U(VI)+V(II)

U(V)+V(lII)

U(V) V(I1) + U(1V) + V(Il1)


Referenres p p . 267-273

(13.1)
(13.2)

254

C O M P L E X E S O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

is complicated by the presence of the additional reactions

2 U(V)

( 1 3.3)

U(IV)+ U(V1)

and
U(V)+V(III)

U(IV)+V(IV)

( 13.4)

However, in the presence of excess V(1V) the U ( V ) formed in reaction (13.1) can
be reoxidised to U(V1) by
U(V)+V(lV)

U(VC)+V(III)

(I 3.5)

Since reaction (13.5) is more rapid than (13.2), (13.3), or (13.4), reactions (13.1)
and (13.5) are predominant under these conditions, and the overall process corresponds to the U(V1)-catalysed reaction of V(1I) and V(IV). In the absence of
U(V1) this reaction is relatively slow221. Kinetic data were obtained spectrophotometrically at 760 mp (where V(IV) is the principal absorbing species) in
0.05 to 2.0 M perchloric acid between 0.6" and 36.8 "C. The rate law is

- d IV( IV)]/dt

= - d[V(II)]/dt =

k,[V(Il)][V( Iv)] + k , I u ( v l ) ] [ v( 1I)]

where k , and k , refer to the uncatalysed and catalysed reactions, respectively.


At 25 "C, [H'] = 1 M , p = 2.0 M , [V(II)] = 2 x
M , [V(IV)] = 2.55 x
M , k , and k , have values of 1.6 and
71 I.mole-'.sec-', respectively. No direct
evidence was adduced for the presence of U(IV). The response of the rate to
variations in hydrogen-ion concentration is only very Flight. Accordingly, the
principal net activation process for the catalysed reaction is

UO,'+

+ v 2 ++ (uo2v4+):

and the corresponding AS' and AH' values are quoted as -26.1 f 0 . 4 cal.deg-'.
mole-' and 7.1 kcal.mole-' at p = 2.0 M . It is not known whether the reaction
is of the inner-sphere type or not, as no evidence was obtained for a binuclear
intermediate. The reaction is catalysed by chloride and sulphate ions; the former
effect receives a detailed discussion by Newton and Baker2", as does the ionic
strength dependence.
U(V1) oxidises V(111) slowly in acid perchlorate solutions, the first step being
U(V1) + V( 111)

3 U(V) +V( IV)

(13.6)

k-2

which is followed by disproportionation of U(V). Newton and Baker252have


made use of the rapid reaction of V(V) with Fe(IIl)106

U(V) + Fe( 111)

2 U(VI) + Fe( 11)

(13.7)

13

OXIDATIONS BY

u, Np, PU

255

to obtain kinetic information on reaction (13.6). Reactions (13.6) and (13.7),


taken in combination, represent the U(V1)-catalysed oxidation of V(II1) by
Fe(II1). The uncatalysed reaction has been described by Higginson and Sykesgo(p.
176), uiz.

2V(IV)+Fe(II)
V(IV) + Fe(II1)
V(V) + Fe(I1)
k-s
V(III)+V(V) 2.
2 V(IV)

V(III)+Fe(III)

(13.8)
(13.9)
(13.10)

The catalysed reaction was followed by measuring the rate of appearance of V(IV)
at 760 mp. Comprehensively, on the basis of reactions (13.6) to (13.10), the
rate law is

In equation (13.1 l), the first term corresponds to the catalysed part of the reaction
and the remaining terms, which make a relatively small contribution, apply to the
uncatalysed part. Kinetic data at constant acidity were in good agreement with the
integrated form of the calculated rate expression. The rate coefficients k,, k,, k, ,
and the ratio k - s/k6 were evaluated. Almost linear plots of log k z versus log [H']
were obtained at four temperatures with slopes close to - 1.8. This result suggests
that the dominant activated complex is that formed by loss of two H f ions, uiz.
U O Z z f + V 3 + + H Z O+ (VO*UOz3+)'+2H +

(13.12)

although the alternative path, involving the loss of one H + ion, viz.
UOzz+ + V 3 + + H Z O+ (VOH-UOZ4+)'+H+

(13.13)

must be present also. The rate data are shown to be more in accord with consecutive reactions and an inner-sphere binuclear intermediate, rather than with
parallel reactions. A number of mechanisms are possible, e.g.

v3++H,O + VOH,+ + H +

rapid equilibrium

+ VO.UOz3++ H +
V O ~ U O ~+~voz+
+
+uo,+

rate-determining

UOzz++VOHZf

References p p . 267-273

rate-determining

256

C O M P L E X E S O F D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

T A B L E 36
ACTIVATION

PARAMETERS2"

FOR

AH!(kcal.rnole- ' )

17.710.3
13 1 2
16 + I

NP
PU

M02''+V3++H2O

A S (cal.deg -9
-5

--f

(VOHM024+)s+H*

.mole- )

3.8h0.9
j=6
+3

ReJ
252
a

254

Recalculated from data in ref. 253.

This contrasts with the Pu(V1) + V(II1) reaction which occurs by parallel paths.
The activation parameters corresponding to the net activation processes (13.12)
and (13.13) are A H f = 22.1 and 17.7 kcal.mole-', and ASs = 12.9k0.5 and
3.8 k 0 . 9 cal.deg-'.mole-'. respectively. Table 36 contains similar data for the
analogous reactions of NpOz2+ (ref. 253) and P u O ~ (ref.
~ + 254) with V3+. The
positive AS* value for the UOzz++V3+ reaction has led Newton and BakerZS2
to suppose that the U(VI) reaction is inner-sphere, whereas the Np(V1) and Pu(V1)
reactions are outer-sphere.
On mixing acid solutions of U(V1) and Cr(I1) at 0 "C, a rapid reaction takes
place and a bright green solution is produced. A slower reaction then occurs to
yield a solution having the darker green colour characteristic of U(W) and
Cr(lI1). The appearance of the absorption spectrum of the principal intermediate
leads to the conclusion3' that it is a complex formed between U(V) and Cr(lll),
viz.

U(Vl)+Cr(lI)

U(V).Cr(III)

From the lack of pH dependence it is likely that the intermediate species is


CrOU04+. Reduction of this complex by Cr(I1)
U(V).Cr(lIl)+Cr(II)

--t

2 Cr(III)+U(IV)

proceeds rapidly with half-lives between 4 and 8 min at 0 "C. The intermediate
species is reactive towards TI(II1) and V(IV), displaying rates which are proportional to the concentration of the intermediate, but independent of the concentration of oxidant. Efficient transfer of oxygen from U 0 2 2 +to Cr(H20)63f
takes place2s5.Evidence was sought for an analogous intermediate in the Pu(VI)+
Cr(I1) system but the results were negative.
B a e ~ has
~'~
briefly investigated the reduction of U(V1) by Fe(l1) in the strongly
complexing medium of phosphoric acid.
Moore" has made a brief study of the slow reduction of U(V1) by Sn(I1) in
hydrochloric acid media. Chloride ion has a pronounced effect on the rate.
Spectrophotometric evidence is cited for the complexing interaction of U(V1)
and Sn(I1).

13

OXIDATIONS B Y

u , Np, PU

13.2 OXIDATIONS B Y

257

NEPTUNIUM

The reaction between Np(V1) and V(1II) in perchlorate media proceeds via
consecutive steps of comparable ratez53,viz.
Np(VI)+V(III) -% Np(V)+V(IV)
Np(VI)+V(IV) -% Np(V)+V(V)
At 25 "C and in 2.0 M HC104, k, is estimated as 17.5k1.6 1.mole-'.sec-' and
k, as 30f: 15 1.mole-'.sec-'. The hydrogen-ion dependence of the reaction is
rate

(k+ k'/[H+])[Np(VI)][V(III)]

where k and k' have values of 6.3 Lmole-'.sec-' and 20.3 sec-', respectively.
Likely steps are between NpOZ2++V3+, together with NpOzz++VOH2+ and/or
NpO,OH+ +V3+. The activation energies and entropies of the k and k' routes
are, respectively, 32 kcal.mole-' and 52+ 16 cal.deg-'.mole-',
and 13 kcal.
mole-' and -9+6 cal.deg-'.mole-' (as recalculated by Newton and Baker "').
Np(V1) oxidises U(1V) in a two-equivalent process

2 Np(V1) + U(1V) = 2 Np(V) U(V1)


In aqueous perchloric acid the rate law found by Sullivan
d[NpO,+]/dt

el

al.257is

2 k'[NpOzZ+][U4+]/[H+]

which suggests the scheme

+ UOH3++ H + Kl
NpOzZ++ UOH3+
N p 0 2 ++ U(V)
rate-determining
rapid
NpOzz++ U(V) NpOz++ U02'+
U4++ HzO

Assuming a steady-state concentration for U(V), k' is identified as K,k,. The


parameters describing the net activation process
N p O z 2 f + U 4 f + H z 0 + (UOHNp 0 2 5 f)* + H+
At 25 "C
are A H f = 18.2 kcal.mole-' and ASf = 7.4k0.8 cal.deg-'.mole-'.
and p = 2 M , k' is 10.8 sec-'. Spectrophotometric, potentiometric and proton
relaxation evidence has been cited for the specific interaction of Np(V) and U(V1)
References p p . 267-273

coM P L E x E s oF

258

D I F F E R E N T M E T A L s : o x I D A T I o N-R E D u c T I o N

in acid media'". A value of 0.690 I.mole-' has been derived for the equilibrium
constant of the complex N p 0 2 + * U 0 2 2 +
at 25 "C.
Np(V) oxidises V(II1) to V(IV) in perchloric acid according to
Np(V)+V(III) = Np(IV)+V(IV)
that is
Np02++V3++2H+

Np4++V02++H20

However, at acid concentrations less than 0.5 M , Np3+ is formed by


N p 4 + + V 3 + + H 2 0+ N p 3 + + V 0 2 + + 2 H +

where K2 6 x
the main reaction is
- d"PO2

dt

at 25 "C and [ClO,-]

+I = ( k , + k ' X 1 )
[V02 +I

K2

3.0 M (ref. 259). The rate law of

[Np02+][V3+]

Plots of (-d[Np02+]/dt)/[NpOz+][V3+]
versus [Np4']/[V02+] are linear with
intercepts k, and slopes k'. The following mechanism accounts fairly adequately
for the kinetics
N p 0 2 + + V 3 +-% N p 0 2 + + V 0 2 +
Np0'++2 H +

-+N p 4 + + H 2 0

rapid equilibrium

+' N p 3 + + V 0 2 + + 2H +

rapid equilibrium, K 2

Np4++V3++H20

N ~ ( v+
) N~(III+

2~ p p )

Thus

whereupon k' = ksK2/[H+]'. At 25 "C, [C104-] = 3.0 M , and [H'] = 0.13 M ,


k , is 35 I.mole-'.sec-', and k' is calculated as 1.3 I.mole-'.sec-' (the observed
value is 3.6 1.mole-'.set-'). The postulated mechanism has points in commcn
with that for the Fe(III)+V(III) systemg0.
Np(V) and Cr(II1) interact to form a binuclear complex, oiz.
0-Np-@+ + C T ( H ~ O ) ~=~O-Np-OCr(H20),4+
+

+H 2 0

13

OXIDATIONS BY

u, Np, PU

259

&

...,....-'----..I.

400

500

,
-\

600

700

800

Wavelength ( m p )

Wavelength ( m p )

Fig. 1 1 . Absorption spectrum of Np(V)-Cr(1II) dimer (--),


Cr(II1) (- - -), and Np(V)
(- -); 1.0 M HC104; temp., 25 "C. (From
by courtesy of The American Chemical
Society.)

The complex has been separated by ion exchange and characterised by direct
analysisz0*. The complex has a distinctive absorption spectrum (Fig. 1l), quite
unlike that of Np(V) and Cr(II1). The rate coefficient for the first-order decomposec-' at 25 "C in 1.0 M HCIO,. Sullivan37
sition of the complex is 2.32 x
has obtained a value for the equilibrium constant of the complex, K = [Np(V)
-Cr(III)]/[Np(V)] [Cr(III) 1, of 2.62 & 0.48 at 25 "C by spectrophotometric
experiments. The associated thermodynamic functions are: AH = - 3.3 kcal.
mole-' and A S = -9.0 cal.deg-'.mole-'. The rates of decay and aquation of
the complex, measured at 992 mp, were investigated in detail. The same complex
is formed when Np(V1) is reduced by Cr(II), and it is concluded that the latter
reaction proceeds through both inner- and outer-sphere paths. It is noteworthy
that the substitution-inert Rh(lrI), like Cr(III), forms a complex with NP(V)'~'.
This bright-yellow Np(V).Rh(III) dimer has been separated by ion-exchange
References pp. 267-273

260

COMPLEXES OF DIFFERENT METALS: OXIDATION-REDUCTION

and its absorption spectrum recorded. Using spectrophotometric means, the


equilibrium constant, K = [Np(V).Rh(III)]/[Np(V)][Rh(III)], has been determined as 3.31 1.mole-' at 25 "C. Associated values of A H and A S are -3.6
kcal.mole-' and -10+3 cal.deg-'.mole-'. It may be significant that K for the
Np(V).Cr(IIl) analogue has a similar value at 25 "C (2.62 l.mole-1)37. The rate
coefficient for the dissociation of the complex into Np(V) and Rh(II1) is 4.38 x
sec-' at 50 "C ([H'] = 1.00 M , p = 1.00 M ) ; the activation energy is 27.3 kcal.
mole-' and ASf is 8.6k0.9 cal.deg-'.mole-'. The rate of approach to equilibrium
has also been measured. The dissociations of Np(V)-Rh(III) and Np(V)Cr(III)
are catalysed by HF in a first-order manner.
In 1 M perchloric acid solution, excess Np(V1) is reduced by Cr(1I) to Np(1V)
along with smaller amounts of Np(V) and Np(V)-Cr(Il1) complex37. Excess
Np(V) is reduced by Cr(I1) according to
Np(V)+Cr(Il) = Np(IV)+Cr(III)
and a Np(I11) intermediate has been detected at [H'] < 0.2 M . Thompson and
Sullivanz6' discuss their kinetic results on this reaction in the light of the following scheme
Np(V) + Cr(I1)

Np(1V) Cr(II1)

(1 3.14)

Np(IV)+Cr(II) ~ N p ( I I I ) + C r ( I I I )

(13.15)

Np(V)+Np(III) -% 2 Np(IV)

(13.16)

The alternative scheme

+
+Cr(1V)
Cr(II) + Cr(IV) --+ 2 Cr(II1)
Np(V) Cr(1I) + Np(lI1)

Np(V)+Np(IZI)-

2 Np(rV)

is rejected on the grounds that (u ) no dimeric Cr(II1) species was detected, and
( b ) for agreement with the empirical rate data, the Np(III)+Np(V) reaction
would require to have a rate coefficient
40 times greater than the observed
value. The rate law for the one-equivalent reduction of Np(V) by Cr(II), reaction
(13.14), is

The term k6[Hf]0.78 is re-expressed as k0[Hf]$["+' according to the Harned

13

OXIDATIONS B Y

u , Np,

PU

261

treatment with k, = 1146 12.mole-2.sec-' and j3 = - 1.142 at 25 "C. Activation


parameters corresponding to the k, term are AH' = 1.85 kcal.mole-' and
ASf = - 38.4 cal.deg-'.mole-'.
Oxygen-18 tracer experiments strongly suggest
the participation of an inner-sphere activated complex. Kinetic data on the
reaction between Np(V) and Np(III), reaction (13.16), conform to the expression
-d[Np(V)]/dt

~~[N~(V)][NP(III)][H+]~.~~

where k 5 = 5.40 (mole, 1 and sec units) at 25 "C and p = 0.2 M (see also refs.
18 and 259). The effect of C1- and HSO,- ions on the rate of the Np(V)+Np(III)
and Np(V)+Cr(II) reactions are similar in that C1- is ineffective as a catalyst,
whereas both reactions are sensitive to HS04-.
Shastri et aLZ6' have examined the kinetics of the reduction of Np(V) by U(1V).
The rate of the reaction, zero order in Np(V) and first order in U(IV), shows
an inverse dependence on the square of the hydrogen-ion concentration. The
gross features of the mechanism are believed to be
Np(V)+U(IV)
Np(IV)+U(IV)
Np(V)+Np(III)
Np(V)+U(V)

Np(1V) U(V)

-+

-+

-+

-+

-+

Np(lV)+U(V)
Np(III)+U(V)
2 Np(IV)
Np(lV)+U(VI)

Np(1II) U(VI)

At [H'] = 0.1 M , U(IV) reacts about six times more rapidly with Np(IV) than
with Np(V). Unlike nitrate and sulphate, chloride ion accelerates the reaction.
The reaction between Np(1V) and Cr(II), reaction (13.15), has been the subject
of a separate investigation by Thompson and Sullivan263.The rate law is
-d[Np(IV)]/dt = k'[Np(lV)][Cr(II)][H+]-'.27
in perchloric acid solutions at 25 "Ct. At the same temperature, p = 1.00 M
and [HCIO,] = 1.00 M , k' is 4.29 I.mole-'.sec-'. In the presence of 1.0 x
M
C1- and 1 . 0 ~
M HS04-, the k' values are 4.33 and 6.26 I.mole-'.sec-',
respectively.
13.3

OXIDATIONSBY PLUTONIUM

In contrast to the reactions of Pu(VI) and Pu(1V) with Ti(III), the reaction
between Pu(V) and Ti(II1) was found by Rabideau and Kline264 to be immeaThe Np(IV)+V(II) reaction shows a quite different hydrogen-ion dependence and is thought
to be outer-sphere in type.
References p p . 267-273

262

C O M P L E X E S OF D I F F E R E N T M E T A L S : O X I D A T I O N - R E D U C T I O N

surably fast, and therefore they consider the overall reduction of Pu(V1) to Pu(1II)
as being composed of two consecutive second-order reactions, viz.
Pu(VI)+2 Ti(1II) = Pu(lV)+2 Ti(1V)

(1 3.17)

Pu(IV)+Ti(III) = Pu(III)+Ti(IV)

( 13.18)

The interaction of Pu(V1) and Pu(II1) is insignificant. From absorbance-time


data for PuOzz+ at 830 mp

- d[Pu(VI)]/dt

= k; [Pu(VI)] [Ti(III)]/[H+ ]

This form of rate law suggests

as the rate-controlling step. Using a value of 65.5 sec-' for k ; , the observed rate
coefficient of (13.18), the kinetic data were given an iterative treatment by a
computer method. At 25 "C in 1 M perchloric acid, the average value of k; was
shown to be 108 sec-'. For the net process

PuOz2++Ti3+ HzO -+ ( P u O , T ~ O H ~+
+H
) ~+

A H f and ASs are 10.3 kcal.mole-' and - 14.7+ 1.3 cal.deg-'.mole-'. A practical
difficulty is that Ti(II1) slowly reduces perchlorate ions to chloridez65. However,
added chloride ion was demonstrated to have no effect on the rate of the Pu(VI)+
Ti(lI1) system. It remains unclear why the reduction of Pu02'+ by Ti(II1) should
be more difficult to bring about than is the reduction of P u 0 2 + . It is interesting
that, with V(II1) as reductant, the reduction of Pu02+ takes place slowly254.
Pu(V1) oxidises V3+ by
P u O z Z + + V 3 + + H 2 0= P U O ~ + + V O ~ H
+ ++ ~
The progress of the reaction was followed at 830 mp254.Kinetically, the oxidation
is second order and it proceeds via parallel paths showing a dependence on both
the inverse first and inverse second powers of [H'], the respective rate coefficients
being 2.12 sec-' and 0.228 1.mole-'.set-' at 25 "C and p = 2 M. The route in
which a single hydrogen ion is liberated predominates at 25 "C, and has AH' =
15.5 kcal.mole-' and ASf = -5+2 cal.deg-'.mole-'.
R a b i d e a ~suggests
~~~
the following scheme as likely

v~++H+
~ oV O H ~ + + H +

rapid

(13.19)

slow

(13.20)

P U O ~ ~ + + V ( O H+) ~P+u O z + + V 0 2 + + H z 0 slow

(13.21)

Pu02'+

+ VOH2+ + P u 0 2 ++ VOz++ H +

13

OXIDATIONS BY

u , Np, PU

263

The kinetics are equally in accord with the replacement of reaction (13.20)
by one between P u 0 2 0 H f +V3+, and reaction (13.21) by PuO,OH+ +VOH2+.
M ) Pu(V1) is reduced by Fe(I1) to Pu(V), viz.
At low concentrations ( c

(13.22)

Pu(VI)+ Fe(I1) = Pu(V)+ Fe(II1)


At higher concentrations Pu(V) is reduced further
Pu(V)+Fe(II) = Pu(IV)+Fe(III)

(13.23)

The inverse hydrogen-ion dependence of the rate of reaction (13.22) is exceptionally complex and is summarised by
-d[Pu(VI)]/dt

[Pu(VI)][Fe(II)]{A

+ (B+C[H+I)-'}

from experiments in the range 0.05 to 2.0 M HC104 at p = 2.0 M (ref. 42). Rate
data were obtained by spectrophotometric measurements of Pu(V1) at 830 mp
between 0 "C and 25 "C. The existence of a binuclear intermediate, of the type
Pu(V).Fe(III), is inferred from the form of the rate law. Thermodynamic quantities of activation are reported for the three activated complexes through which the
reaction proceeds. Ionic strength effects were investigated, as also was the influence of chloride ions on the rate. The Pu(V).Fe(lIl) dimer is analogous to
the complexes formed between Np(V) and Fe(III)208,and U(V) and Cr(III)31.
Likewise, chloride ion has a marked accelerating effect on the two-equivalent
reduction of Pu(V1) by Sn(1I). On the assumption that the rate-determining step is
Pu(Vl)+Sn(II)

Pu(V)+Sn(IlI)

two mechanisms are possible266: ( a ) a second Pu(V1) is reduced by Sn(II1);


the Pu(V) formed then disproportionates, viz.
Pu(VI)+Sn(III)
2 Pu(V)

Pu(V)+Sn(IV)

+ Pu(IV)+Pu(VI)

(6) the reduction of a second Pu(V1) takes place, along with reduction of Pu(V)
by Sn(Il), viz.
Pu(VI)+Sn(III)
Pu(V)+Sn(II)

-+

Pu(V)+Sn(IV)

-+

Pu(IV)+Sn(III)

Pu(V)+Sn(III) + Pu(IV)+Sn(IV)
References pp. 267-273

264

COMPLEXES OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

Neither of these possibilities are likely: the disproportionation of Pu(V), and the
Pu(V)+Sn(II) reaction are too slow to account for the rapidity and kinetics of
the overall reaction. That the Pu(VI)+Sn(II) reaction is very much faster than
the Pu(V)+Sn(II) reduction is taken as evidence for the occurrence of a single
two-equivalent process, viz.
Pu(VI)+Sn(II)

--f

Pu(IV)+Sn(IV)

The chloride-ion dependence indicates the importance of two activated complexes,


(PuO,SnCl,+)f and (PuO,SnCI,)*, with A H f values of 14.0 and 14.6 kcal.mole-',
and ASf values of 4 . 4 f 7 and 8.0f5.5 cal.deg-'.mole-', respectively (as recalculated by Newton and Baker'"). In terms of the apparent rate coefficient, k'

k'

a[CI-I3 +b[C1-I4

where a = 433 14.mole-4.sec-' and b = 768 15.mole-5.sec-' at 2.4 "C. As in the


Pu(IV)+ Sn(I1) system, the rate is insensitive to variations in acidity.
When 2 x
A4 solutions of Pu(V1) and U(1V) are mixed in 1 M HCIO,
solution, the main plutonium product is Pu(V) along with smaller amounts of
PU(ITI)~'.When lo-, A4 concentrations are used, the stoichiometry does not
deviate significantly from

Newton41 has shown that no complications ensue from the reaction of the intermediate U(V) with oxygen, since the latter has no effect on the rate. A simple
second-order rate equation applies, the disappearance of Pu(V1) being followed
at 830 mp, and the probable mechanism is
Pu(V1) + U(IV)

Pu(V)+ U(V)

rate-determining

( 13.24)

rapid

(13.25)

followed by

Pu(V1) U(V)

-P

Pu(V)

+ U(V1)

or

2 U(V)

-,U(IV)+U(VI)

rapid

( I 3.26)

To be consistent with the observed first-order dependences on Pu(V1) and U(IV),


it is necessary that steps (13.24) and (13.25) do not occur simultaneously. The
rate of reaction decreases with increasing hydrogen-ion concentration. It appears
that two activated complexes (and a binuclear intermediate) are involved in the

13

OXIDATIONS BY

265

u , Np, PU

reaction, as represented by the net activation processes


P u O , ~ + + U ~ +H,O
+ ~ + (H20UOHPu02S+)*+H+

(13.27)

+ ~( H O U O H P U O ~ ~ + )H
' ++~
P u O ~ ~ + + U H~ 2+0 +

(13.28)

and

The respective A H f and AS' values of (13.27) and (13.28) are 17.6 kcal.mole-'
and 3.4+ 1.5 cal.deg-'.mole-', and 21.4 kcal.mole-' and 18.1k 1.0 cal.deg-'.
mole- '. Newton4' discusses the apparent similarities between the U(V1) U(IV)267
NP(VI)+NP(IV)~~*,
P u ( V I ) + P U ( I V ) ~and
~ ~ , Pu(VI)+U(IV) systems.
The reduction of Pu(1V) by Ti(IT1) to blue Pu(JI1)

Pu(IV) + Ti(II1) = Pu(lI1) +Ti(IV)


has been investigated by Rabideau and Kline269, usi.ng a spectrophotometric
method to follow the disappearance of Pu(IV) at 469 mp. Rate data can be expressed in terms of the principal species as
-d[Pu(IV)]/dt =

k;[Pu4+][Ti3+]/[H+]

At 2.4 "C and p = 2.02 M , k; is 12.7 sec-'; at 25 "C and p = 1.02 M , k; is 65.5
sec-'. In terms of the net activation process
P u 4 + + T i 3 + + H 2 0+ (PuTiOH6+)'+Hf
AH' and AS' are 16.7 kcal.mole-' and 5.9$.2 cal.deg-'.mole-', respectively.
Chloride ion has a slight accelerating influence on the reaction.
Pu(IV) oxidises V(II1) stoichiometrically in perchloric acid solution, uiz.

and the rate law is given by

- d [Pu(IV)]/dt

270

= k'[Pu4+][V3+]/[H+]
=

+~ " [ P u ~ [V3+]/
']
[H']'

kobs[ P ~ )[v3
I +1

Spectrophotometric values for k' and k" at 2.4 "C ( p = 2 M ) are 1.70 sec-' and
1.71 mole.l-'.sec-', respectively. The corresponding activation parameters, AHS
and AS*, are 17.1 kcal.mole-' and 4.8k1.7 cal.deg-'.mole-' for the k' path,
and 21.5 kcal.mole-' and 20.8f 1.4 cal.deg-'.mole-' for the k" path. The rate of
reaction is not susceptible to the addition of chloride ions. This result is of importance because C104- is reduced to C1- by V(II1).
References pp. 26 7-273

266

C O M P L E X E S OF D I F F E R E N T M E T A L S : O X I D A T I O N - R E D U C T I O N

The Pu(IV) Fe(I1) reaction proceeds stoichiometrically in perchloric acid


media271, viz.
Pu(IV)+ Fe(I1) = Pu(III)+ Fe(lI1)
The process is first order in each of the reactants and the rate is unaffected by the
presence of Fe(lI1). However, there is some uncertainty about the inclusion of a
term in Pu(II1) in the rate law. From the observed hydrogen-ion dependence,
Newton and C ~ w a n conclude
~~'
that the principal reaction path has an activated
complex formed from Pu4+, Fe2+ and water with the prior loss of one hydrogen
ion. The probable form of rate law is

-d [Pu(l V)]/dt

k , [Pu4+][Fe2+]+k' [Pu4+][Fe2+]/[H+]

where k3 = 0.177 1.mole-'.sec-' and k' = 3.12 sec-' at 2.5 "C. The activated
complex (PuOHFeS+)' has AH' = 19.1 kcal.mole-', and AS' = 13.3 cal.deg-'.
mole-'. The pronounced increase of rate on the addition of chloride ion is attributed to the provision of a new reaction path involving (PuClFe")', for which
AH' = 14.4 kcal.mole-' and AS' = 0.6 cal.deg-'.mole-l. Sulphate ion also
increases the rate.
Sn(I1) reduces Pu(IV) by
2 Pu(IV)+Sn(II) = 2 Pu(III)+Sn(IV)
In perchloric acid media the reaction is extremely slow and is complicated by the
formation of polymeric species of tin, and by heterogeneity. R a b i d e a ~has
~~~
examined the kinetics in mixed perchlorate-chloride solutions, in which media
no turbidity is apparent. The rate expression is complex, viz.

- d [Sn(II)l)/dt

c[Pu4+][Sn''] [C1-l4 d[Pu4+][Sn2+][CI-]'

but shows a lack of dependence on hydrogen-ion concentration. At 25 "C, c and


d are 720 1s.mole-5.sec-' and 1636 16.mole-6.sec-', respectively. The net activation process

pu4+ Sn2

+4 ~ 1 -,
- ( P U C I , S +~ )'~

has AH' and AS' equal to 26.9 kcal.mole-' and 44.7 cal.deg-'.mole-',
the process
pu4+ +Sn2++ 5 CI-

-+

whereas

(PuClsSn+)'

has a A H f of 24.1 kcal.mole-', and a AS' of 37.0 cal.deg-'.mole-'.

A possible

REFERENCES

267

mechanism is

although there is no direct evidence for the participation of Sn(I1l).


The reduction of Pu(IV) by U(IV)

obeys the rate


-d[Pu(IV)]/dt

~[Pu~+][U~+]/[H+]~

in accordance with the net activation process

Pu4+ U4++ H 2 0 + (PuOU6+)' + 2 Ht


The corresponding activation parameters are AH' = 24.3 kcal.mole-'
ASf = 30.1 1.9 cal.deg-'.mole-'. Sulphate ion catalyses the reaction.

and

REFERENCES
1 H. TAUBE,
in Advances in Inorganic Chemistry and Radiochemistry, ed. H. J. E M E L ~ AND
US
A. G. SHARPE,
Vol. 1, Academic Press, New York, 1959, p. 1.
2 J. HALPERN,
Quart. Reu., 15 (1961) 207.
3 N. SUTIN,Ann. Rev. Phys. Chem., 17 (1966) 119.
4 A. G. SYKES,
in Advances in Inorganic Chemistry and Radiochemisfry,ed. H. J. E M E L ~ AUNSD
A. G. SHARPE,
Vol. 10, Academic Press, New York, 1967, p. 153.
5 D. R. STRANKS,
in Modern Coordination Chemistry, ed. J. LEWISA N D R. G. WILKINS,Interscience, 1960, p. 78.
6 R. T. M. FRASER,
Reu. Pure Appl. Chem., 11 (1961) 64.
Ann. Rev. Phys. Chem., 16 (1965) 167.
7 H. STREHLOW,
8 W. L. REYNOLDS
AND R. W. LUMRY,
Mechanisms of Electron Transfer, Ronald Press, New
York, 1966.
9 F. BASOLO
A N D R. G. PEARSON,
Mechanism of Inorganic Reactions, 2nd Ed., Wiley, New
York, 1967.
10 J. P. CANDLIN,
K. A. TAYLOR
AND D. T. THOMPSON,
Reactions of Transition Metal Complexes,
Elsevier, Amsterdam, 1968.
11 R. A. MARCUS,Ann. Rev. Phys. Chem., 15 (1964) 155.
12 1. RUFF,Quart. Rev., 22 (1968) 199.
13 J. 0. EDWARDS,
Inorganic Reaction Mechanisms, Benjamin, New York, 1964.
14 A. G. SYKES,
Kinetics of Inorganic Reactions, Pergamon, London, 1966.
15 D. BENSON,Mechanisms of Inorganic Reactions in Solution, McGraw-Hill, London, 1968.

268

COMPLEXES O F D I F F E R E N T METALS: OX1 D AT ION-RE D U CTION

N. A. DAUGHERTY
AND T. W. NEWTON,
J . Phys. Chem., 67 (1963) 1090.
J. R. HUIZENGA
AND L. B. MAGNUSSON,
J. Am. Chem. SOC.,73 (1951) 3202.
J. C. HINDMAN,
J. C. SULLIVAN
AND D. COHEN,
J. Am. Chem. SOC.,80 (1958) 1812.
S. W. RABIDEAU,
J . Am. Chem. SOC.,79 (1957) 6350.
H. IMAI,Bull. Chem. SOC.Japan, 30 (1957) 873.
T. W. NEWTON
AND S. W. RABIDEAU,
J. Phys. Chem., 63 (1959) 365.
J. SILVERMAN
AND R. W. DODSON,
J . Phys. Chem., 56 (1952) 846.
A. ANDERSON
AND N. A. BONNER,
J. Am. Chem. SOC.,76 (1954) 3826.
L. E. BENNETT
AND J. C. SHEPPARD,
J. Phys. Chem., 66 (1962) 1275.
S. C. FURMAN
AND C. S. GARNER,
J . Am. Chem. SOC.,74 (1952) 2333.
W. C. E. HIGGINSON,
D. R. ROSSEINSKY,
J. B. STEADAND A. G. SYKES,Discussions Faraday
SOC.,29 (1960) 49.
27 M. J. NICOLAND D. R. ROSSEINSKY,
Proc. Chem. SOC.,(1963) 16.
AND M. J. NICOL,Electrochinz. Acta, 11 (1966) 1069.
28 D. R. ROSSEINSKY
AND P. KNOWLES,
J . Chem. SOC.,(1962) 1137.
29 D. J. DRYE,W. C. E. HIGGINSON
30 R. L. MOORE,J . Am. Chem. SOC.,77 (1955) 1504.
31 T. W. NEWTON
AND F. B. BAKER,
Inorg. Chem., 1 (1962) 368.
Inorg. Chem., 4 (1965) 1533.
32 J. H. ESPENSON,
33aM. ARDONAND R. A. PLANE,J . Am. Chem. Soc., 81 (1959) 3197.
148th National Meeting of American Chemical
33bR. E. CONNICKAND M. G. THOMPSON,
Society, Inorg. Division, Papers 23 and 24, September, 1964.
A N D F. B. BAKER,
Inorg. Chem., 3 (1964) 569.
34 T. W. NEWTON
35 M. EICENA N D R. G. WILKINS,Advan. Chem. Ser., ed. R. F. GOULD,No. 49, American
Chemical Society, 1965, p. 5 5 .
AND R. A. PLANE,
Inorg. Chem., 3 (1964) 322.
36 R. W. KOLACZKOWSKI
Inorg. Chem., 3 (1964) 315.
37 J. C. SULLIVAN,
38 T. J. CONOCCHIOLI,
E. J. HAMILTON
A N D N. SUTIN,
J . Am. Chem. SOC.,87 (1965) 926.
Nature, 209 (1966) 804.
39 A. ADINAND A. G. SYKES,
Inorg. Chem., 4 (1965) 1025.
40 J. H. ESPENSON,
41 T. W. NEWTON,
J . Phys. Chem., 62 (1958) 943.
J. Phys. Chetn., 67 (1963) 1425.
42 T. W. NEWTONA N D F. B. BAKER,
J . Phys. Chem., 67 (1963) 853.
43 R. A. MARCUS,
44 G. DULZAND N. SUTIN,Inorg. Chem., 2 (1963) 917.
45 A. HAIM,Inorg. Chem., 5 (1966) 2081.
46 A. ADINAND A. G. SYKES,J. Chem. SOC.A, (1968) 351.
47 A. G. SYKES,Chem. Commun., (1965) 442.
J. Chem. SOC.A, (1966), 1230.
48 A. ADINAND A. G. SYKES,
J . Phys. Chem., 56 (1952) 853.
49 D. J. MEIERAND C. S. GARNER,
J. Am. Chem. SOC.,86 (1964) 1883;
5OaJ. H. ESPENSON,
J. Am. Chem. SOC.,86 (1964) 5101.
50bJ. H. ESPENSON,
5 1 J. Y. P. TONGAND E. L. KING,J. Am. Chem. SOC.,75 (1953) 6180.
AND H. S. ROSSOTTI,
Acta Chem. Scnnd., 9 (1955) 1177; 10 (1956) 957.
52 F. J. C. ROSSOTTI
53 N. A. DAUGHERTY
A N D T. W. NEWTON,
J. Phys. Chem., 68 (1964) 612.
J. Phys. Chem., 7 (1903) 1 , 356.
54 C. BENSON,
J . Phys. Chem., 12 (1908) 632.
55 R. A. GORTNER,
AND W. PREISS,
2. Anorg. Chem., 168 (1928) 265.
56 C. WAGNER
Chem. Rev., 45 (1949) 419.
57 F. H. WESTHEIMER,
AND E. L. KING,J. Am. Chem. SOC.,85 (1963) 3328.
58 J. H. ESPENSON
J . Chem. SOC.,(1965) 5890.
59 E. A. M. WETTONA N D W. C. E. HIGCINSON,
Inorg. Chem., 3 (1964) 1248.
60 J. G. MASONAND A. D. KOWALAK,
J . Am. Chem. SOC.,87 (1965) 1495.
61 J. C. SULLIVAN,
62 B. R. BAKER,
M. ORHANOVIC
AND N. SUTIN,J . Am. Chetn. SOC.,89 (1967) 722.
63 A. HAIMAND N. SUTIN,J . Am. Chem. SOC.,87 (1965) 4210.
64 A. HAIMAND N. SUTIN,J. Am. Chem. SOC.,88 (1966) 434.
65 A. ADINAND A. G. SYKES,J. Chem. SOC.A, (1968) 354.
AND M. EIGEN,Z. Elektrochem., 63 (1959) 652; G. G. HAMMES
AND J. 1.
66 C. CZERLINSKI
16
17
18
19
20
21
22
23
24
25
26

REFERENCES

269

STEINFELD,
J. Am. Chem. SOC.,84 (1962) 4639.
67 P. DEBYE,Trans. Electrochem. SOC.,82 (1942) 265.
68 E. EICHLER
AND A. C. WAHL,J. Am. Chem. SOC.,80 (1958) 4145; M. W. DIETRICH
AND A. C.
WAHL,J. Chem. Phys., 38 (1963) 1591.
69 M. A. RAWOOF
AND J. R. SUTTER,
J. Phys. Chem., 71 (1967) 2767.
70 L. J. KIRSCHENBAUM
AND J. R. SUTTER,
J. Phys. Chem., 70 (1966) 3863.
A N D J. HALPERN,
J. Am. Chem. Soc., 85 (1963) 2518.
71 J. P. CANDLIN
72 P. SAFFIRAND H. TAUBE,
J. Am. Chern. Soc., 82 (1960) 13.
73 R. T. M. FRASER
AND H. TAUBE,
J. Am. Chem. SOC.,82 (1960) 4152.
74 S. M. TAYLOR
AND J. HALPERN,
J. Am. Chem. SOC.,81 (1959) 2933.
75 K. B. WIBERGAND R. STEWART,
J. Am. Chern. SOC.,78 (1956) 1214.
76 D. R. ROSSEINSKY,
J. Chem. SOC.,(1963) 1181.
AND M. J. NICOL,J. Chem. SOC.A , (1968) 1022.
77 D. R. ROSSEINSKY
78 C. R. GIULIANO
AND H. M. MCCONNELL,
J. Inorg. Nucl. Chem., 9 (1959) 171.
79 D. R. ROSSEINSKY
AND M. J. NICOL,Trans. Faraday SOC.,64 (1968) 2410.
80 D. R. ROSSEINSKY
AND W. C. E. HIGGINSON,
J. Chem. SOC.,(1960) 31.
81 K. G. ASHURST
AND W. C. E. HIGGINSON,
J. Chem. SOC.,(1956) 343.
AND N. SUTIN,J. Phys. Chem., 68 (1964) 174.
82 H. DIEBLER
83 see also, G. DAVIES,L. J. KIRSCHENBAUM
AND K. KUSTIN,Znorg. Chem., 7 (1968) 146.
Chem. Znd. London, (1963) 1166.
84 M. J. NICOLA N D D. R. ROSSEINSKY,
85 A. W. ADAMSON,
J. Phys. Colloid Chem., 55 (1951) 293.
86 A. M. ARMSTRONG
AND J. HALPERN,
Can. J. Chem., 35 (1957) 1020.
J. Am. Chem. SOC.,72 (1950) 2254.
87 E. K. MAUNAND N. DAVIDSON,
88 D. BANERJEA AND M. S. MOHAN,J. Indian Chem. soc., 40 (1963) 188.
AND N. SUTIN,Inorg. Chem.. 6 (1967) 428.
89 E. G. MOORHEAD
90 W. C. E. HIGGINSON
AND A. G. SYKES,
J. Chem. SOC.,(1962) 2841. See also ref. 45.
91 E. L. KINGAND C. S. GARNER,
J. Phys. Chem., 58 (1954) 29.
92 J. H. ESPENSON,
K. SHAWAND 0. J. PARKER,
J. Am. Chem. SOC.,89 (1967) 5730.
93 J. H. ESPENSON
AND R. E. MCCARLEY,
J. Am. Chem. SOC.,88 (1966) 1063.
AND H. MYERS,J. Am. Chem. SOC.,76 (1954) 2103.
94 H. TAUBE
AND H. TAUBE,
J. Am. Chem. SOC.,84 (1962) 872.
95 A. ARDON,J. LEVITAN
96 G. DULZA N D N. SUTIN,J. Am. Chem. SJC.,86 (1964) 829.
97 R. M. MILBURN
AND W. C. VOSBURGH,
J. Am. Chem. SOC.,77 (1955) 1352.
98 From results of R. E. CONNICK
A N D C. P. COPPEL,
J. Am. Chem. SOC.,81 (1959) 6389, and
M. J. M. WOODS,P. K. GALLAGHER
AND E. L. KING,Inorg. Chem., 1 (1962) 55.
99 D. L. BALLAND E. L. KING,J. Am. Chem. SOC.,80 (1958) 1091.
100 N. SUTIN,J. K. ROWLEY
AND R. W. DODSON,
J. Phys. Chem., 65 (1961) 1248.
101 A. H A I M A NW.
D K. WILMARTH,
J. Am. Chem. SOC.,83 (1961) 509.
102 A. W. ADAMSON
AND E. GONICK,Inorg. Chem., 2 (1963) 129.
103 D. H. HUCHITAL
AND R. G. WILKINS,
Inorg. Chem., 6 (1967) 1022.
104 F. R. DUKEAND R. C. PINKERTON,
J. Am. Chem. SOC.,73 (1951) 3045.
105 N. C. PETERSON
A N D F. R. DUKE,J. Phys. Chem., 67 (1963) 531.
105aD. W. CARLYLE
AND J. H. ESPENSON,
J. Am. Chem. SOC.,90 (1968) 2272.
106 R. H. BETTS,Can. J. Chem., 33 (1955) 1780.
107 S. MINC,J. SOBKOWSKI
AND M. STOK,Nukleonika, 10 (1965) 747.
108 J. F. ENDICOTT
AND H. TAUBE,
Inorg. Chem., 4 (1965) 437.
H. MYERSAND R. L. RICH,J. Am. Chem. SOC.,75 (1953) 4118.
109 H. TAUBE,
110 J. E. EARLEY
AND J. H. GORBITZ,
J. Znorg. Nucl. Chem., 25 (1963) 306.
111 H. TAUBE,J. A m Chem. SOC.,77 (1955) 4481.
112 A. ZWICKEL
AND H. TAUBE,
J. Am. Chem. SOC.,83 (1961) 793.
113 J. P. CANDLIN,
J. HALPERN
AND D. L. TRIMM,
J. Am. Chem. SOC.,86 (1964) 1019.
114 A. ZWICKEL
AND H. TAUBE,
Discussions Faraday SOC.,29 (1960) 42.
I 15 A. A. VLEEK,Advances in the Chemistry of the Coordination Compounds, ed. S . KIRSCHNER,
MacMillan, New York, 1961, p. 590.
116 A. A. VLEEK,Discussions Faraday SOC.,26 (1958) 164.
117 J. P. CANDLIN,
J. HALPERN
AND S. NAKAMURA,
J. Am. Chem. SOC.,85 (1963) 2517.

270

c o M P L E x E s OF

D I F F E R E N T M E T A L S:

ox I D A T I o N-R

E D u cT I oN

118 .I.
HALPERN
AND S. NAKAMURA,
1. Am. Chem. SOC.,87 (1965) 3002.
AND H. TAUBE,
J. Am. Chem. SOC.,83 (1961) 2239.
I19 R. T. M. FRASER
AND H. TAUBE,
J. Am. Chem. SOC.,83 (1961) 2242.
120 R. T. M. FRASER
121 A. ZWICKEL
AND H. TAUBE,
J. Am. Chem. SOC.,81 (1959) 1288.
Inorg. Chem. 2 (1963) 954.
122 R. T. M. FRASER,
A N D H. TAUBE,
Inorg. Chem., 4 (1965) 1029.
123 H. DIEBLER
J. Am. Chem. SOC.,80 (1958) 1084.
124 A. E. OGARDAND H. TAUBE,
AND H. TAUBE,
J. Am. Chem. Soc., 86 (1964) 1686.
125 J. F. ENDICOTT
126 J. H. ESPENSON,
Inorg. Chem., 4 (1965) 121.
AND J. HALPERN,
Inorg. Chem., 4 (1965) 766.
127 J. P. CANDLIN
128 A. HAIM,J. Am. Chem. SOC.,85 (1963) 1016.
AND H. TAUBE,
J. Am. Chem. SOC.,83 (1961) 1785.
129 D. K. SEBERA
Advances in the Chemistry of Coordination Compounds, ed. S . KIRSCHNER,
130 R. T. M. FRASER,
MacMillan, New York, 1961, p. 287.
131 J. HALPERN
AND J. RABANI,
J. Am. Chem. SOC.,88 (1966) 699.
AND R. T. M. FRASER,
J. Am. Chem. SOC.,87 (1965) 2758.
132 D. E. PETERS
A N D R. T. M. FRASER,
Proc. Symp. Exchange Reactions, Brook133 R. G. MILLER,D. E. PETERS
haven National Laboratory, Intern. Atomic Energy Agency, Vienna, 1965, p. 203.
134 J. H. ESPENSON
AND J. P. BIRK,J. Am. Chem. SOC.,87 (1965) 3280.
135 R. SNELLGROVE
AND E. L. KING,J. Am, Chem. SOC.,84 (1962) 4610.
136 L. ORGEL,Rept. 10th Solvay Conference, Brussels, 1956, p. 289.
137 A. HAIMAND W. K. WILMARTH,
Inorg. Chem., 1 (1962) 583.
138 A. HAIM,J. Am. Chem. SOC.,86 (1964) 2352.
AND J. HALPERN,
Inorg. Chem., 4 (1965) 1086.
139 J. P. CANDLIN
AND E. WHALLEY,
Can. J. Chem., 39 (1961) 1094.
140 A. R. OSBORN
AND A. HAIM,J. Am. Chem. Soc., 87 (1965) 3826.
141 P. BENSON
Inorg. Chem., 4 (1965) 1184.
142 M. GREEN,K. SCHUGAND H. TAUBE,
B. M. CRAVEN,
H. C. FREEMAN,
N, E. KIMEAND J. A. IBERS,Chem. Corrimun.,
143 M. T. BARNET,
(1966) 307.
144 J. M. DE CHANTAND J. B. HUNT,J. Am. Chem. SOC.,89 (1967) 5988.
145 R. D. CANNON
AND J. E. EARLEY,
J. Am. Chem. SOC.,87 (1965) 5264.
J. Am. Chem. Soc., 82 (1960) 526.
146 w. KRUSEAND H. TAUBE,
J. Am. Chem. SOC.,89 (1967) 1276.
147 J. H. ESPENSON,
AND A. G. SYKES,
J. Chem. SOC.A , (1967) 795.
148 J. DOYLE
Z. Phys. Chem., 44 (1965) 92.
149 P. H. DODELAND H. TAUBE,
150 A. G. SYKES,Trans. Faraday SOC.,58 (1962) 543.
151 A. G. SYKES,Trans. Faraday SOC.,59 (1963) 1325.
152 A. G. SYKES,Trans. Faraday SOC.,59 (1963) 1334.
J. Chem. SOC.,(1963) 2798.
153 A. PIDCOCKAND W. C. E. HIGGINSON,
J. Chem. SOC.,(1963) 2788.
154 R. D Y K EAND W. C. E. HIGGINSON,
155 H. TAUBE,
in Mechanisms of Inorganic Reactions, Advan. Chem. Ser., ed. R. F. COULD,
NO. 49, Am. Chem. SOC., 1965, pp. 107,117.
Can. J. Chem., 37 (1959) 129.
156 H. TAUBE,
AND H. TAUBE,
J. Am. Chem. SOC.,86 (1964) 1318.
157 E. S. COULD
158 E. B. FLEISCHER
AND R. FROST,J. Am. Chem. SOC.,87 (1965) 3998.
Nature, 202 (1964) 691.
159 R. T. M. FRASER,
160 R. T. M. FRASER,Nature, 205 (1965) 1207.
AND H. TAUBE,
J. Am. Chem. SOC.,87 (1965) 5597.
161 R. D. BUTLER
J. Am, Chem. SOC.,85 (1963) 1747.
162 R. T. M. FRASER,
163 E. S. COULD,J. Am. Chem. SOC.,87 (1965) 4730.
164 E. S. COULD,J. Am. Chem. SOC.,88 (1966) 2983.
AND H. TAUBE,
Inorg. Chem., 4 (1965) 1660.
165 D. H. HUCHITAL
166aR. T. M. FRASER,
D. K. SEBERA
AND H. TAUBE,
J. Am. Chem. SOC.,81 (1959) 2906.
J. Am. Chem. SOC.,90 (1968) 1178.
166bJ. K. HURSTAND H. TAUBE,
AND H. TAUBE,
J. Am. Chem. SOC.,81 (1959) 5000.
167 R. T. M. FRASER
168 R. T. M. FRASER,
J. Am. Chem. SOC.,83 (1961) 564.

REFERENCES

27 1

169 R. T. M. FRASER,
J. Art?. Chem. SOC.,84 (1962) 3436.
AND H. TAUBE,
J. Am. Chem. SOC.,87 (1965) 5371.
170 D. H. HUCHITAL
AND L. E. ORGEL,
Discussions Faraday SOC.,29 (1960) 32.
171 J. HALPERN
J. Am. Chem. SOC.,83 (1961) 4920.
172 R. T. M. FRASER,
R. C. JARNAGIN
AND M. SILVER,
J. Phys. Chem., 68 (1964) 265.
173 P. V. MANNING,
A N D H. C. LONGUET-HIGGINS,
Proc. Roy. SOC.London, Ser. A , 191 (1947) 39.
174 C. A. COULSON
175 W. F. LIBBY,J. Chem. Phys., 38 (1963) 420.
176 R. T. M. FRASER
AND H. TAUBE,
J. Am. Chem. SOC.,81 (1959) 5514.
A N D H. TAUBE,
J. Am. Chem. SOC.,85 (1963) 3706.
177 E. S. COULD
AND H. TAUBE,
J. Am. Chem. SOC.,88 (1966) 4295; 90 (1968) 1162.
178 F. R. NORDMEYER
179 E. S. COULD,
J. Am. Chem. SOC.,89 (1967) 5792.
Inorg. Chem., 3 (1964) 1561.
180 R. T. M. FRASER,
AND H. TAUBE,
J. Am. Chem. SOC.,83 (1961) 4172.
181 G. SVATOS
AND R. C. JARNAGIN,
J. Phys. Chem., 67 (1963) 2884.
182 P. V. MANNING
J. Am. Chem. SOC.,89 (1967) 269.
183 H. J. PRICEAND H. TAUBE,
J. Chem. SOC.,(1965) 21 16.
184 P. B. WOODAND W. C. E. HIGGINSON,
Proc. Chem. SOC.,(1964) 109.
185 P. 9. WOODAND W. C. E. HIGGINSON,
Trans. Faraday SOC.,60 (1964) 119.
186 J. B. KIRWIN,P. J. PROLLAND L. H. SUTCLIFFE,
AND J. R. WEBER,
J. Inorg. Nucl. Chem., 12 (1960) 281.
187 L. H. SUTCLIFFE
AND J. R. WEBER,Trans. Faraday SOC.,52 (1956) 1225.
188 L. H. SUTCLIFFE
AND J. P. HUNT,J. Am. Chem. SOC.,82 (1960) 3826.
189 N. A. BONNER
G. H. NANCOLLAS
AND N. SUTIN,J. Am. Chem. SOC.,86 (1964) 1453.
190 T. J. CONOCCHIOLI,
N. SUTINAND B. WARNQVIST,
Inorg. Chem., 6 (1967) 838.
191 D. H. HUCHITAL,
192 A. HAIMAND N. SUTIN,J. Am. Chem. SOC.,88 (1966) 5343.
AND J. H. BAXENDALE,
Trans. Faraday SOC.,52 (1956) 210.
193 J. BARRETT
AND N. SUTIN,Inorg. Chem., 5 (1966) 1866.
194 E. G. MOORHEAD
N. PURDIEAND N. SUTIN,Inorg. Chem., 3 (1964) 1091.
195 R. J. CAMPION,
AND N. SUTIN,J. Am. Chem., 83 (1961) 1830.
196 M. H. FORD-SMITH
J. Phys. Chem., 67 (1963) 2288.
197 J. B. KIRWIN,F. D. PEAT,P. J. PROLLAND L. H. SUTCLIFFE,
J. Phys. Chem., 67 (1963) 1617.
198 J. B. KIRWIN,F. D. PEAT,P. J. PROLLAND L. H. SUTCLIFFE,
AND W. C. E. HIGGINSON,
J. Chem. SOC.,(1953) 3044.
199 K. G. ASHURST
A N D C. F. WELLS,Trans. Faraday SOC.,53 (1957) 800.
200 J. H. BAXENDALE
R. T. LEIGHAND R. NIGHTINGALE,
J. Chem. SOC.,(1962) 435.
201 W. C. E. HIGGINSON,
AND F. R. DUKE,J. Am. Chem. Soc., 74 (1952) 1535.
202 R.C. PINKERTON
J. Am. Chem. SOC.,77 (1955) 2049.
203 L. J. HEIDTAND J. BERESTECKI,
G. H. NANCOLLAS
AND N. SUTIN,Inorg. Chem., 5 (1966) 1.
204 T. J. CONOCCHIOLI,
205 L. H. SUTCLIFFE
AND J. R. WEBER,Trans. Faraday SOC.,55 (1959) 1892.
AND J. R. WEBER,
Trans. Faraday Soc., 57 (1961) 91.
206 L. H. SUTCLIFFE
AND R. C. THOMPSON,
Inorg. Chem., 6 (1967) 1795.
207 J. C. SULLIVAN
J. Am. Chem. SOC.,84 (1962) 4256.
208 J. C. SULLIVAN,
AND F. BASOLO,
Inorg. Chem., 6 (1967) 2069.
209 J. K. BEATTIE
AND J. W. MOORE,
Inorg. Chem., 5 (1966) 1523.
210 R. G. PEARSON
J. Chem. SOC.A , (1968) 215.
211 J. DOYLEAND A. G. SYKES,
Discussions Faraday SOC.,29 (1960) 125.
212 A. W. ADAMSON,
J. Am. Chem. SOC.,87 (1965) 5026.
213 N. A. DAUGHERTY,
214 A. G. SYKES,J. Chem. SOC.,(1961) 5549.
J. Am. Chem. SOC.,14 (1952) 959.
215 C. E. JOHNSON,
J. Chem. SOC.,121 (1922) 394.
216 A. J. BERRY,
R.P. EPPLE,J. Am. Chem. SOC.,74 (1952) 5772.
217 0. L. FORCHHEIMERAND
AND A. C. WAHL,J. Am. Chern. Soc., 71 (1949) 3137.
218 R. J. PRESTWOOD
J. Phys. Chem., 60 (1956) 1015.
219 F. R. DUKEAND B. BORNONG,
220 F. 9. BAKER,
W. D. BREWER
AND T. W. NEWTON,
Inorg. Chem., 5 (1966) 1294. See also ref.
213.
221 T. W. NEWTON
AND F. 9. BAKER,
J. Phys. Chem., 68 (1964) 228.
Inorg. Chem., 4 (1965) 1069.
222 J. H. SWINEHART,
J. Chem. SOC.,(1957) 1841.
223 D. H. IRVINE,

272

C O M P L E X E S OF D I F F E R E N T METALS: O X I D A T I O N - R E D U C T I O N

224 G. HARBOTTLE
A N D R. W. DODSON,
J. Am. Chem. SOC.,73 (1951) 2442.
J. HALPERN
AND W. C. E. HIGGINSON,
J. Phys. Chem., 60 (1956) 1661.
225 A. M. ARMSTRONG,
in Kinetics and Mechanism of Inorganic Reactions in Solution, Chemical Society,
226 K. W. SYKES,
London, 1954, p. 64.
227 S. HIETANEN
AND L. G. SILLEN,
Arkiu Kemi, 10 (1956) 103.
AND J. HALPERN,
J. Am. Chem. SOC.,81 (1959) 3526.
228 A. C. HARKNESS
229 See, for example, C. H. BRUBAKER
AND J. P. MICKEL,
J. Inorg. Nucl. Chem., 4 (1957) 55.
230 J. HALPERN
AND J. G. SMITH,
Can. J. Chem., 34 (1956) 1419.
231 F. A. JONESAND E. S. AMIS,J. Inorg. Nucl. Chem., 26 (1964) 1045.
232 J. 0. WEAR,J. Chem. SOC.,(1965) 5596.
AND C. H. BRUBAKER,
J . Inorg. Nucl. Chem., 27 (1965) 2183.
233 C. M. LOVE,L. P. QUINN
P. J. PROLL,L.H. SUTCLIFFE AND J. wALKLEY,DiSCUSSiOnS Faraday soc., 29 (1960)
234 D. BENSON,
60.
AND A. W. DAVIDSON,
Inorg. Chem., 2 (1963) 823.
235 W. P. TAPPMEYER
A N D L. H. SUTCLIFFE,
Trans. Faraday SOC.,56 (1960) 246.
236 D. BENSON
L. H. SUTCLIFFE
AND J. WALKLEY,
J. Am. Chem. SOC.,81 (1959) 4488.
237 See also, D. BENSON,
Trans. Faraday SOC.,55 (1959) 2107.
238 D. BENSONAND L. H. SUTCLIFFE,
AND E. L. KING,J. Am. Chem. SOC.,82 (1960) 3805.
239 J. Y. P. TONG
240 See, for example, T. J. HARDWICKE
AND E. ROBERTSON,
Can. J. Chem., 29 (1951) 828.
D. R. ROSSEINSKY
AND G. B. SHAW,Chem. Ind. London, (1963) 911.
241a M. J. ASPRAY,
G. N. RAOAND G. P. RAO,Anal. Chem., 38 (1966) 1900.
241 b G. A. RECHNITZ,
F. S. DAINTON
AND P. GLENTWORTH,
Trans Faraday SOC.,61 (1965) 689.
242 M. G. ADAMSON,
Discussions Faraday SOC.,29 (1960) 21.
243 R. A. MARCUS,
244 R. A. MARCUS,Can. J. Chem., 37 (1959) 155.
AND R. H. PRINCE,
J. Chem. SOC.,(1965) 5749; J. Chem. SOC.A, (1966) 1370.
245 J. D. MILLER
AND G. G. GUILBAULT,
J . Phys. Chem., 64 (1960) 1825.
246 W. H. MCCURDY
AND W. H. MCCURDY,
J. Phys. Chem., 70 (1966) 656.
247 G. G. GUILBAULT
AND J. W. GRYDER,
Inorg. Chem., 1 (1962) 799; J. W. GRYDER
AND M. K.
248 M. K. DORFMAN
DORFMAN,
J. Am. Chem. SOC.,83 (1961) 1254.
249 C. H. BRUBAKER
AND A. J. COURT,
J. Am. Chem. SOC.,78 (1956) 5530.
AND F. B. BAKER,
in Advan. Chem. Ser., No. 71, Am. Chem. SOC.,1967,
250 T. W. NEWTON
p. 268.
251 T. W. NEWTON
AND F. B. BAKER,
J. Phys. Chern., 69 (1965) 176.
AND F. B. BAKER,
J. Phys. Chem., 70 (1966) 1943.
252 T. W. NEWTON
J. Phys. Chem., 68 (1964) 1190.
253 J. C. SHEPPARD,
J. Phys. Chem., 62 (1958) 414.
254 S. W. RABIDEAU,
Inorg. Chem., 2 (1963) 1277.
255 G. GORDON,
256 C. F. BAES,J. Phys. Chem., 60 (1956) 805.
A. J. ZIELEN
AND J. C. HINDMAN,
J. Am. Chem. SOC.,82 (1960) 5288.
257 J. C. SULLIVAN,
J. C. HINDMAN
A N D A. J. ZIELEN,
J. Am. Chem. SOC.,83 (1961) 3373.
258 J. C. SULLIVAN,
AND J. C. SULLIVAN,
J. Phys. Chem., 66 (1962) 442.
259 E. H. APPELMAN
AND J. C. SULLIVAN,
Inorg. Chem., 6 (1967) 892.
260 R. K. MURMANN
AND J. C. SULLIVAN,
J. Am. Chem. SOC.,89 (1967) 1098.
261 R. C. THOMPSON
E. S. AMISAND J. 0. WEAR,J. Inorg. Nucl. Chem., 27 (1965) 2413.
262 N. K. SHASTRI,
263 R. C. THOMPSON
AND J. C. SULLIVAN,
J. Am. Chem. SOC.,89 (1967) 1096.
AND R. J. KLINE,
J. Phys. Chem., 63 (1959) 1502.
264 S. W. RABIDEAU
265 F. R. DUKEAND P. R. QUINNEY,
J. Am. Chem. SOC.,76 (1954) 3800.
AND B. J. MASTERS,
J. Phys. Chem., 65 (1961) 1256.
266 S. W. RABIDEAU
J. Am. Chem. SOC.,79 (1957) 6350.
267 See for example, S. W. RABIDEAU,
J. C. SULLIVAN
AND D. COHEN,
J. Am. Chem. SOC.,76
268 See for example, J. C. HINDMAN,
(1954) 3278.
AND R. J. KLINE,
J. Phys. Chem., 64 (1960) 193.
269 S. W. RABIDEAU
AND R. J. KLINE,J. Inorg. Nucl. Chem., 14 (1960) 91.
270 S. W. RABIDEAU
AND H. D. COWAN,
J. Phys. Chem., 64 (1960) 244.
271 T. W. NEWTON
272 S. W. RABIDEAU,
J. Phys. Chem., 64 (1960) 1491.
J. Phys. Chem., 63 (1959) 1493.
273 T. W. NEWTON,

REFERENCES

273

SUPPLEMENTARY REFERENCES
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311

N . SUTIN,Accounts Chem. Res., 1 (1968) 225.


N. A. DAUGHERTY
A N D B. SCHIEFELBEIN,
J . Am. Chem. SOC.,91 (1969) 4328.
J. H. ESPENSON
A N D R. J. CHRISTENSEN,
J. Am. Chem. SOC.,91 (1969) 7311.
J. H. ESPENSON
A N D 0. J. PARKER,
J. Am. Chem. SOC.,
90 (1968) 3689.
J. H. ESPENSON
AND S. R. HELZER,
Inorg. Chem., 8 (1969) 1051.
D. R. ROSSEINSKY
A N D M. J. NICOL,J. Chem. SOC.A, (1969) 2887.
J. P. BIRK,J. Am. Chem. SOC.,91 (1969) 3189.
A. A. BERGHA N D G. P. HAIGHT,Inorg. Chem., 8 (1969) 189.
M. O R H A N O VH.
I ~ ,N. Po AND N. SUTIN,J. Am. Chem. SOC.,90 (1968) 7224.
0.J. PARKER
AND J. H. ESPENSON,
J. Am. Chem. soc., 91 (1969) 1313.
K. SHAWAND J. H. ESPENSON,
J. Am. Chem. Soc., 90 (1968) 6622.
0. J. PARKER
A N D J. H. ESPENSON,
Inorg. Chem., 8 (1969) 1523.
J. H. ESPENSON,
Inorg. Chem., 7 (1968) 631.
J. H. ESPENSON
A N D D. J. BOONE,
Inorg. Chem., 7 (1968) 636.
D. W. CARLYLE
A N D J. H. ESPENSON,
Inorg. Chem., 8 (1969) 575.
T. W. NEWTON,
G. E. MCCRARY
A N D W. G. CLARK,
J. Phys. Chern., 72 (1968) 4333.
D. W. CARLYLE
AND J. H. ESPENSON,
J. Am. Chem. Soc., 91 (1969) 599.
J. A. STRITAR
AND H. TAUBE,
Inorg. Chem., 8 (1969) 2281.
D. SEEWALD,
N. SUTINAND K. D. WATKINS,
J. Am. Chem. SOC.,91 (1969) 7301.
W. G . Movlus AND R. G . LINCK,J. Am. Chem. Soc., 91 (1969) 5394.
D. E. PENNINGTON
AND A. HAIM,Inorg. Chem., 7 (1968) 1659.
K. M. DAVIESAND J. H. ESPENSON,
Chem. Commun., (1969) 111; J. Am. Chem. SOC..91
(1969) 3093.
D. J. PARKER
AND J. H. ESPENSON,
J. Am. Chern. Sac., 91 (1969) 1968.
J. P. BIRKAND J. H. ESPENSON,
J. Am. Chem. SOC.,90 (1968) 1153.
T. J. MEYERAND H. TAUBE,Inorg. Chern., 7 (1968) 2369.
R. C. PATELAND J. F. ENDICOTT,
J. Am. Chem. Soc., 90 (1968) 6364.
R. G. LINCK,Inorg. Chem., 7 (1968) 2394.
J. DOYLEAND A. G. SYKES,J . Chem. SOC.A, (1968) 2836.
R. D A V ~ EAND
S A. G . SYKES,J. Chem. Soc. A, (1968) 2831.
A. B. HOFFMAN
AND H. TAUBE,
Inorg. Chem., 7 (1968) 1971.
H. TAUBE
AND E. S. COULD,Accounts Chem. Res., 2 (1969) 321.
R. G. WILKINS
A N D R. E. YELIN,Inorg. Chem., 7 (1968) 2667.
J. HALPERN
AND M. P R I B A N IJ.
~ ,Am. Chem. Soc., 90 (1968) 5942.
K. SHAWAND J. H. ESPENSON,
Inorg. Chem., 7 (1968) 1619.
0. J. PARKER
AND J. H. ESPENSON,
Inorg. Chem., 8 (1969) 185.
M. W. Hsu, H. G. KRUSZYNA
AND R. M. MILBURN,
Inorg. Chem., 8 (1969) 2201.
G . H. SCHENK
A N D W. E. BAZZELLE,
Anal. Chem., 40 (1968) 162.
M. J. BURKHART
A N D T. W. NEWTON,
J. Phys. Chem., 73 (1969) 1741.

Chapter 4

Oxidation-Reduction Reactions between Covalent


Compounds and Metal Ions
T.J. K E M P

1. Introduction
1.1

SCOPE A N D PATTERN OF THIS CHAPTER

The vast number of thermodynamically possible reactions obtained by permuting


oxidants and reductants within the scope of this review present major problems
of classification and selection. To only a limited extent is the modernity or detail
of a paper indicative of its relevance, some of the definitive papers having been
published before 1950. Discussion has been concentrated, therefore, at points
where a kinetic investigation of a reaction has resulted in a real advance in our
understanding both of its mechanism and of those of related reactions, and work
which has been more of a confirmatory nature will not receive comparable
consideration. Detailed reference to products, spectra, etc. will be made only
when the kinetics produce real ambiguities.
Presentation of existing data can be made in several ways. Wiberg and Stewart
have taken a series of oxidising metal ions singly or in very small groups and have
examined the reactions of each member with a range of reductants. An alternative
approach, adopted by Waters3, is t o compare the reactions of a given reductant,
e.g. iodide ion, hydrogen peroxide, aliphatic hydrocarbon, ketone, alcohol, olefin
and acetylene, with a range of oxidising metal ions, noting the roles of complexformation, hydrolysis, inner and outer-sphere processes and other mechanistic
features.
Classification exclusively in terms of a few basic mechanisms is the ideal approach, but in a comprehensive review of this kind, one is presented with all
reactions, and not merely the well-documented (and well-behaved) ones which
are readily denoted as inner- or outer-sphere electron transfer, hydrogen atom
transfer from coordinated solvent, ligand transfer, concerted electron transfer,
etc. Such an approach has been made on a more limited scale4. Turney has
considered reactions in terms of the charges and complexing of oxidant and
reductant but this approach leaves a large number to be coped with under further
categories.
As regards oxidation by metal ions, we have chosen to select groups of from
two to eleven metal ions, the members of which are known to display similar

275

INTRODUCTION

mechanisms in their reaction with several substrates. This basic similarity of


mechanism implies certain properties common to a group, such as equal increments between stable valency states, comparable redox potentials, ability to form
n-complexes, etc. This means that different oxidation states of a given transition
metal may be represented in different groups. These groups consist of
Cr(V1) and Mn(Vll)
Pb(IV), Tl(III), Hg(II), Hg(I), Bi(V), Au(III), Pt(IV), Pd(II), Rh(III),
Ru(1II) and Mo(V1)
Ag(II), Ag(III), Co(III), Ce(IV), Mn(III), V(V), Ir(IV), Np(V1) and Pu(V1)
Fe(I1I) (including ferricyanide), Ag(I), Cu(II), Cu(I), Np(V) and Mo(V)
Reductions by metal ions are covered in Section 6 in terms of ( i ) electronacceptance and (ii)electron-acceptance concerted with homolytic fission. One group
of reactions, which includes oxidations and reductions by metal ions, is that
between a metal ion and a neutral free radical. These form a self-contained class
which is treated separately in Section 7.
This procedure emphasises the overall division of oxidants into "one-equivalent"
and "two-equivalent'' types defined6-' in terms of how many equivalents of
reducing species, e.g. electrons or hydrogen atoms, are taken up by the oxidant
in the primary act, as deduced from the behaviour of the oxidant towards either
hydrazine6, sulphite ion* or captive ligand'. l o according to the schemes (u)-(c)
respectively.
One-equivalent oxidant

N4H6 -+ N 2 + 2 NH,

2 30;
(c)

-+

(3)

S 2 0 i - (dithionate)

(5)

M"++[(NH,),CO(III)-OCOCO~H]~+M("-')+
+[(NH,),CO(III)-OCOCO~-]~+
+H+
-+

[(NH,),CO(III)-OCOCO,-]~+

-+

[(NH,),CO(II)OH~]~+
+ 2 C 0 2 (very fast)

References p p . 493-SO9

c o v A L E N T c o M P o u N D S,

276

M E T A L I o N S : O X I D A T I o N-R E D u c T I o N

Two-equivalent oxidant

M"++ N2H4

(a)

-+

N,H,

+ M("-2)++ 2 H +

2 N2Hz -+ N4H4
NZ

N4H4

(c)

+ N2H4

NH3+HN3

M"++[(NH,),CO(III)-OCOCO~H]~+
+ M("-')'
+[(NH,),CO(III)OH~]~'+~C 0 2 + H +

It should be noted that although an oxidant may be categorised as two-equivalent


from these reactions, it may, on occasion, function as a one-equivalent reagent.
Three-equivalent changes are extremely rare.
The sub-classification of the oxidising metal ions derives from overall reactivity,
which is only crudely related to redox potential and is gauged largely with hindsight.
The coverage of reducing substrates is in terms of increasing molecular complexity. The inorganic substrates are dealt with in terms of periodic group although
azide and cyanide appear with halide, and sulphurous, phosphorous and arsenious
acids are taken together, as are also the gases carbon monoxide and hydrogen.
The organic compounds are considered in the order aliphatic and aromatic
hydrocarbons, alcohols, aldehydes, carboxylic acids, ethers and amines followed
by polyfunctional compounds (unless one of the functional groups is not in any
way involved in the reaction).
The cases of oxidation and reduction of "bound" ligands are included in the
section appropriate to the ligand involved.

1.2

C A T E G O R I S A T I O N O F O X I D A N T S AS O N E - OR T W O - E Q U I V A L E N T

While it is very convenient to consider separately (i) those oxidants which


undergo changes of oxidation state by only one unit altogether or by a series of
changes of which the first involves a single unit and (ii) those which undergo
changes of oxidation state by two units, this scheme does depend on an initial
correct classification. This is obvious in certain instances, thus cerium exists only
in two oxidation states, + 4 and +3, and can participate only in one-equivalent
changes. In many cases, however, the stages of reduction are far from obvious
and recourse to the modes of oxidation of sulphite, hydrazine and captive ligand
must be made (vide supra).

INTRODUCTION

277

The use of sulphite has, however, been called into question recently. According
to the scheme all one-equivalent oxidants should produce at least some dithionate
(eqns. 4 and 5). However, Vepkek-SiHka et a1.l2-I4 have found that while substitution-labile one-equivalent oxidising metal complexes give sulphate and
dithionate, substitution-inert one-equivalent oxidising complexes give sulphate
only.
The following mechanisms for generation of both products were suggested

The species (MSO3)+IVrepresents an inner-sphere complex between sulphite and


the oxidant formed by ligand-displacement. (MS03)+Vis formed by abstracting
an electron from (M S O ~ )+I v .
Reactions ( 6 ) , (7) and (9) occur with substitution-labile oxidants and (6),
(7) and (8) with substitution-inert oxidants. It is postulated that the species
(MS03)+IV involves metal-oxygen coordination for labile M, and metalsulphur coordination for inert M. The latter mode is thought to prevent sulphursulphur bond formation.
Brown and Higginson have proposed an alternative interpretation of the
data. The stationary-state concentration of (MSO3)+IVis likely to be very low
for substitution into an inert oxidising complex and hence step (9), which depends
on the square of this concentration, is rather unlikely, particularly as it involves
rupture of two inert metal-sulphur bonds. They have also discussed the parallel
case of hydrazine, for which a similar distinction between substitution-labile and
-inert one-equivalent oxidising complexes is apparent. Substitution-inert ferricyanide ion16 produces only nitrogen, whilst substitution-labile Mn(II1) pyrophosphate gives some nitrogen and ammonia via N4H, (the analogue of
S2062-)and some nitrogen uia N2H2 (the analogue of S042-). (The distinction
between the separate origins of the nitrogen is based on isotopic studies).
Coordination isomerism of the type proposed for sulphite is not feasible for N2H4
and a more general explanation of the dichotomy is preferable. Brown and
Higginson suggest that whereas substitution-labile oxidants may release the
sulphite radical-ion, which then dimerises or is oxidised to sulphate, substitutioninert oxidants retain the radical-anion until a second oxidation (7) occurs.
References p p . 493-SO9

278

cov A L E N T C O M P O U N D S ,

METAL IONS:

0 x 1D AT ION-RED U C T I O N

2. Oxidation by Cr(V1) and Mn(VI1)


2.1

G E N E R A L FEATURES

Reactions of these oxidants share a variety of features. Both function as either


one- or two-equivalent reagents, depending on the substrate involved. Complete
reaction involves several stages of reduction of the metal ions and considerable
evidence for the transient existence of unusual valency states has been obtained.
In the form of chromic acid and acidic, neutral or alkaline permanganate, Cr(V1)
and Mn(VI1) have been very widely used in synthetic organic chemistry and in
analytical chemistry and a correspondingly large body of non-kinetic information
relevant to the mechanisms has accumulated. It is probable that more kinetic
studies have been performed using Cr(V1) and Mn(VI1) than any other metal-ion
oxidant. One contrasting feature is that the oxidising power of Cr(V1) is normally
confined to acidic solutions.
Some relevant oxidation potentials are

+ 14 H + + 6 e- + 2 C r 3 + + 7 H,O
M n 0 4 - + 8 H + + 5 e - + M n 2 + + 4 H,O
MnO,-+e- + MnO2-

Cr,O,r

M n 0 4 - + 4 H + + 3 e-

+ Mn OZ+2 H,O

+1.33 V

+1.51 V

+ 0.564 V

+ 1.695 V

2.1.1 Oxidation states involved in reduction of Cr( V I ) and Mn( V I I )


Cr(V1) is normally reduced to Cr(II1) since Cr(V) and Cr(1V) are very unstable
under ordinary reaction conditions. Westheimer has discussed critically the
roles of Cr(V) and Cr(1V) as reactive intermediates, (vide infra), and Wiberg
has summarised the inorganic chemistry of compounds containing Cr(V) and
Cr(1V).
The situation with Mn(VI1) is more complicated. The final stage of reduction
depends both on the pH of the medium and upon the presence of some ion
capable of stabilising an intermediate valency state by forming an insoluble
precipitate or a complex ion. Oxidations by permanganate in alkaline solution
normally produce a precipitate of MnO,, but addition of barium ionsz2 results
in the precipitation of barium manganate, BaMnO,, containing Mn(V1). This
does not imply a one-equivalent reduction of Mn(VI1) however, for the reaction
Mn(V)+ Mn(VI1) + 2 Mn(V1)
is rapidz3. Mn(V) exists only in strong alkali and its role in the reduction of
Mn(VI1) is essentially that of a reactive intermediate.

OXIDATIONBY

Cr(V1) A N D Mn(VI1)

279

In acidic solution MnO, is usually the end product, although particularly


vigorous reductants, e.g. iodide and oxalate ions, convert permanganate to
manganous ions. Mn(II1) is stable only in acidic solution or in the form of a
complex, e.g. with pyrophosphate ion, and it has seldom been reported as the end
product of a permanganate oxidation, e.g. for that of Mn(n) in a phosphate
buffer and for those of alcohols and ethers in the presence of fluoride ion2,.
2.1.2 Solution equilibria of oxy-anions of Cr( VZ) and Mn( VZZ)
Equilibria relevant only to kinetic studies of oxidations will be mentioned.
For Cr(V1) in aqueous solution these are at 25 OC2', 2 6
H2Cr0,
HCrO,

+ H + + HCrO,
+ Hf+CrO:-

2HCrO; + C r 2 0 , 2 - + H 2 0

mole.1-'

K,

= 4.1

K,

1.3 x

K3

155

mole.l-'

Addition of ions such as chloride27, sulphate and phosphate (denoted X-),',


produces a new equilibrium
0
II

H+ HO-Cr-0- + X -

29

0
II

+ X-Cr-0- + H 2 0

II

II

Addition of acetic acid27or trifluoroacetic acid3' also influences the acid chromate
ion (oide infra).
It transpires2' that Cr20,'- plays a minor part in oxidation mechanisms and
that H2Cr0, is the reactive form of Cr(V1) and accordingly most reactions are
acid-catal ysed.
Permanganic acid has a pK of -2.25 in perchloric acid3' and one of -4.6 in
sulphuric acid3, when the Hammett acidity function, H,, is used. Accordingly
HMnO, is present to a significant extent only in strongly acidic solutions and
comparatively few of the reactions which have been examined involve anything
other than MnO;.
2.2

O X I D A T I O N O F I N O R G A N I C C O V A L E N T SPECIES

2.2.1 Halide ions


The oxidation of iodide ion by aqueous chromic acid at low acidity is very
slow but is subject to marked enhancement in rate on addition of ferrous ion33.
References pp. 493-SO9

280

cov A L E N T C O M P O U N D S , M E T A L I O N S : 0x1D A T I ON-RED u C T I O N

Iodide ion is very slowly oxidised by Fe(II1) and the iodide must therefore be
oxidised by some species other than Cr(VI) or Fe(II1). For a large excess of iodide,
the stoichiometry has been reported to be3,
Cr(VI)+2 I-+Fe(II) = Cr(III)+I,+Fe(III)
from which it is clear that the Fe(I1) is not acting as a true catalyst, but that it is
probably involved in a scheme
Cr(VI)

+Fe(I1) + Cr (V) + Fe(111)

Cr(V)+I10-+I-+2Hf

(10)

Cr(III)+IO-

(1)

-+

12+H20

(12)

This is an example of an induced reaction, which is the subject of the next chapter
of this book. That step (1 1 ) is involved rather than
Cr(V)

+ I-

Cr (IV) 1.

is argued by Westheimer from the relative effects of arsenious acid upon the
chromic acrd oxidations of manganous and iodide ions. The latter has an induction
factor, defined as the number of equivalents of substrate oxidised to the number
of equivalents of inductor oxidised, of 2 but the former has one of 0.5. The reactions involved are:
Manganous ion
H,AsO,

+ Cr(V1) + H,AsO, + Cr(IV)

Cr(N)+ Mn(II)
H,AsO,+Cr(VI)+Mn(II)

-+

Cr(III)+Mn(III)

= H3A~O4+Cr(III)+Mn(III)

Iodide ion
H,AsO,

+ Cr(VI)

H3As04 Cr(rV)

-+

2 Cr(V)

2 [Cr(V)+I-

-+

Cr(III)+IO-]

2 [10-+1-+2H+

-+

12+H20]

H3As0,+2 Cr(VI)+4I-

H,As0,+2 Cr(III)+2 I,

Cr(N)+Cr(VI)

It is assumed that Cr(IV) reacts quickly with Mn(II) but slowly with iodide ion
for otherwise the induction factors would not be as found.

OXIDATION BY

28 1

Cr(V1) A N D Mn(VI1)

Westheimer' has reviewed other inductions of the chromic acid oxidation of


iodide, indicating how these reactions afford insight into the mechanism of the
simple oxidation.
An early
of the kinetics of the simple iodide-chromic acid reaction
yielded the rate law

-d[Cr(VI)]/dt = k[Cr(VI)]{ [H+I[I-] k'[H+lz [I-]')

A more recent e ~ a m i n a t i o nconfirmed


~~
the existence of an unusual ionic-strength
dependence of the reaction rate3', which features a minimum at p
0.83, but
it was noted that the initial kinetics differ from those occurring later in the reaction. Consideration of the equilibria prevailing in aqueous solutions of Cr(V1)
produced a simplified rate law
-d[I-]/dt

= k[HCrOJ[I-]2[H30+]'

where k at 25 "C ( p = 0.1) is (1.2550.08)~lo3 14.mole-4.sec-1 and E is 5.6


kcal.mole- '. The autoretardation is connected with the production of molecular
iodine since addition of Iz reduces the initial reaction rate. An ion-pair or a
complex between Cr(VI) and I- is viewed as the active oxidant giving Cr(1V) and
iodine which can, however, back-react, viz.
Complex + I -

k 2

k-13

Cr(IV) + Cr(VI)
Cr(V)+I2H++IO-+I-

1'

+ Cr(IV)

(13)

3 2 Cr(V)

(14)

k-14

IO-+Cr(III)

+ H20+12

Further analysis of the kinetics over the course of reaction afforded values for
k14/k-13, e.g. 0.10 at 20 "C (p = 0.1).
Measurements over a wider range of reactant concentration^^^^ favour a more
complex rate law
-d[HCrOJ/dt

[HCrO,-](k,

[H30+][I-]

+ k2[H30+]2[J-]2
+ k 3 W 3 0+ 1"1- 1)

where, at 20.3 "C (p = 0.130 M NaClO,), k , = 0.206+0.009 I'.mole-'.sec-',


k2 = 111+7 14.mole-4.sec-', k , = 154f3 14.mole-4.sec-', E, = 10.351.5
kcal.mole-', AS: = - 16+6 eu, E, = 6.9k1.0 kcal.mole-' and AS: = -27k3
eu. The second order term in [I-] is considered to correspond to attack of
ICr03- upon I-.
References p p . 493-SO9

282

COVALENT COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

The permanganate oxidation of iodide has been the subject of a recent detailed
study by a rapid mixing technique3'. At 35 "C and at an ionic strength of 0.9 M
over the p H range 3-6 the rate expression is
-d[MnO,]/dt

[Mn0,][I-]{k,+k3a,+}

In the region of pH 5, k, = 51.2 1.mole-'.sec-' and k, = 1 . 7 0 lo7


~ 1.2mole-2.
sec-'. Activation parameters for the second-order path are E = 1.90 kcal.mole-'
and ASt = -45.8 eu while those for the third order path are 4.37 kcal.mole-'
and - 14.4 eu, respectively. The authors accommodate both the low energies of
activation and the observed orders in the mechanisms:
pH-independent path
MnO, +I-

+ (03MnOI)2-

(O,MIIOI)~-+ H 2 0 -, HOI+HMnOz-

(rapid)
(slow)

pH-dependent path

( 03P/lnOI)2- + H,O+

-, HOI

+ H2Mn0,

(slow)

The HOI would be rapidly reduced by iodide and the Mn(V) species would be
expected either to disproportionate or to oxidise further iodide. This reaction
scheme has features in common with the analogous reaction with cyanide ion
discussed below.
The oxidation of bromide ion by Cr(V1) has been examined at length by
Bobtelsky et a1.38-40.The rate equation is
d[Br2]/dt

k[Cr(VI)l[Br-If"H2S041

The apparent first-order rate coefficient obtained using excess oxidant increased
exponentially with increase in acidity in the range 5 N < [ H 3 0 + ] < 12 N . The
reaction is first-order with respect to added manganous ions (k increasing sharply),
but the activation energy (1 1.0 kcal.mole-') remains unchanged. At appreciable
catalyst concentrations the reaction becomes almost zero-order with respect to
bromide ion. The mechanism appears to be a slow oxidation of Mn(I1) to Mn(II1)
followed by a rapid reduction of the latter by bromide. This reaction is considered
further in the section on Mn(1I)-catalysis of chromic acid oxidations (p. 327).

OXIDATION B Y

283

Cr(V1) A N D Mn(VI1)

2.2.2 Cyanide ion

Only the permanganate oxidation of this ion has been studied kinetically. At
pH 12-14.6 the reaction has the stoichiometry
2MnO;+CN-+2

OH- = 2MnO:-+OCN-+H,O

but at pH 6-1 2, the reaction is complex and non-stoichiometric, yielding cyanate,


cyanide and carbon dioxide, whilst cyanogen is formed between p H 6 and 9.
The reaction is very slow in acid solution, and has a maximum rate near pH 9.
It is quite rapid, however, in the high p H region in which Stewart and Van der
Linden4 found evidence for two reaction paths. At pH > 13 and at low reactant
concentrations. the kinetics are

- d[MnOJ/df

k,[MnO,][CN-]

but at lower basicity and at higher reactant concentrations a more complex and
pH-dependent path dominates, the rate law being

- d[MnOJ/d

f =

k , [MnOJ[ CN-][ OH-] -

k ; [Mn 0, ][CN - ] [H 1

k;MnO,][CN-][HCN]

Freund4, has reported similar kinetics for the simple reaction and gives k , as
6.4 x lo7 exp( -9.0 x 103/RT)I.mole-.sec-.
Stewart and Van der Linden4 also examined the incorporation of 0 into
the cyclnate from labelled permanganate. The percentage of transfer varied with
alkalinity and the authors believe that significant oxygen-transfer occurs in the
second-order reaction, but not in the complex reaction. Accordingly the mechanism
for the second-order reaction is proposed to be
MnO, + C N - -+ [O,Mn

. . . 0 . . . CN12- + MnO;

+ OCNMnO,

+ MnO,

+ 2 OH-

-+

2 MnOi-

+ H,O

(slow)

(fast)

Conversely, a mechanism involving oxygen-transfer cannot be written for the


complex reaction, and the authors propose a sequence
HCN+CNMnO;

+ H(CN),

(CN),+2 OH3 Mn0:-+4


References p p . 493-504

+ H(CN);
+ (CN),
+

+ H + + Mn0:-

CN-+OCN-+HZO

H,O = 2 M n 0 2 + M n 0 i + 8 OH-

(fast)
(slow)
(fast)
(fast)

284

C O V A L E N T COMPOUNDS, M E T A L IONS: O X I D A T I O N - R E D U C T I O N

The slow step of this reaction corresponds to removal of hydride from an anion
and finds several counterparts in oxidations of organic compounds by MnO; .
The anion may have the structure
H-C-NI

C=N

2.2.3 Oxides of hydrogen


Vepfek-SiSka et a1.43-45have recently shown that the oxidation of hydroxide
ion by Mn(VI1) is very much slower than has been suggested, being profoundly
influenced by trace quantities of transition metal cations.
The reaction between chromic acid and hydrogen peroxide gives CrO, and
involves no immediate change in the oxidation state of the metal atom. However,
CrO, is decomposed by acid to Cr(H,O):+ although there is some evidence for
the intermediacy of other species, and a brief reference to the overall reaction
merits inclusion in this review.
of the formation of CrO, yielded the following rate
A stopped-flow
law for acidities
of 0.01 to 0.05 M
c
d[CrO,]/dt = k,[H+][Hz02][HCr0.J
where k , = (4.4+2.0)x lo7 exp[-(4.5f0.2)~ 103/RT], 2 . 0 ~lo4 12.mole-2.
sec-' at 25 "C. At higher acidities (up to 6 M HNO,) and at 4 "C, kobs was
reasonably well r e ~ r e s e n t e dby
~~
kobs = 5 . 0 ~
lo3 L-H2OZI[H+I
(l+O.l[H+])
These data are considered to favour the mechanism
HCrOh

+ H + + H2Cr04

H2Cr04+H,0,

-, H 2 C r 0 , + H 2 0

H 2 C r 0 5 + H 2 0 2+ CrO,+2 H20

(slow)
(fast)

The disappearance of CrO, in perchloric acid follows the law


-d[CrO,]/dt
with k;

= k;[Cr0,][H+]2

6.8 x lo9 (exp

- 12.8 x 103/RT)I2.mole-'.sec-',

2.7 at 25 "C.

OXIDATIONBY

Cr(V1) A N D Mn(VI1)

285

Permanganate does not form a stable peroxy compound with hydrogen peroxide
but oxidises it rapidly to oxygen and water. The stoichiometric equation depends
on pH, with Mn(II), Mn(1V) and Mn(V1) being formed in acidic, neutral and
alkaline solutions respectively. Chang4* found DzOz to react with MnO; in D,O
with a rate only 15 % of that for H 2 0 2 in H 2 0 , the reaction being auto-catalysed
by Mn2+ ion. At 0.1 Macidity k , equals 2 . 9 ~lo3 I.mole-'.sec-' (18 0C)49.

2.2.4 Oxy-acids of sulphur


Haight et aL5' have published a detailed account of the kinetics and stoichiometry of the oxidation of buffered bisulphite ion by chromic acid. The reaction
is fast and its study requires a rapid mixing technique. The stoichiometry varies
from a Cr(VI)/S(IV) molar ratio of 1 : 2 to 2 : 3 as theinitialconcentrationsare
changed in the range 0.12 < [Cr(VI)]/[S(IV)] < 1.4 and this was explained in
terms of competition between two overall reactions
HS0;+6 H +

2 C r 3 + + 2 SO:-+S,0,2-+6 H 2 0

(15)

2 HCrO, + 3 HSO; + 5 H +

2 C r 3 + + 3S 0 : - + 5

(16)

2 HCrO;+4

HzO

The observed rate law, which applies only to (15) in 0.5 M sodium acetate
buffer in the pH range 4.18-5.05 is

K , , which has a value of 36 1.mole-' at 25 "C, is interpreted as being the equilibrium constant for the reaction

kobs is 1.37 X lo8 13.mole-3.sec-' at 25.0 " c and the activation energy and entropy
are 4.5 kcal.mole-' and - 13 eu, respectively. A further point is that the sulphate
formed during the reaction is bound to Cr(II1) at its conclusion.
The presence of the denominator term in the rate equation (17) suggests that
the equilibrium (18) precedes the oxidation step. Two sequences of reactions are
proposed (see below), depending on whether the sulphite radical ion dimerises
(20) or attacks further acid chromate ion (21). It should be noted that of the
species prevalent in dilute aqueous chromic acid, namely CrO:-, Cr,O:-,
HCrOg and H2Cr04, only the last is regarded as possessing oxidising powers.
This fact, noted by Westheimer", is tacitly assumed in all recent discussion of
References pp. 493-509

286

C O V A L E N T COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

mechanisms of oxidations by Cr(V1)


HSO;+H+
SO,+CrSO,Z-

+S02+H20 K
+ {[O~SOC~O,OSO,]~-}'
(slow)

{[02SOCr020S0,]2-}*+4 H , 0 + 2 H + -+ [S04Cr(H20),]+ +-SO, (19)

s20;-

(20)

-+

SOz-+Cr(V)

(21)

-+

O,CrOCrO,OSO:-

2.s0,

-+

-SO, +HCrO,
H + + HCrO, +CrSO;-

+ H,O
(22)

O,C~OC~O,OSO:Cr(V)+S(tV)

-+

-+

2 Cr(V)+SO:-

(23)

Cr(III)+S(VI)

(24)

The entity marked with a double dagger is regarded by the authors as an


activated complex. Its breakdown (19) may well consist of a sequence of rapid
steps rather than the single step implied, which involves a three-electron transfer
and double protonation of a transition state subsequent to its formation. Steps
(22)-(24> were invoked to explain the complete oxidation of S(1V) to S(V1) at
higher Cr(V1) concentrations.
The authors contrast this oxidation with those of As(II1) and of alcohols, both
of which involve analogous pre-equilibria (vide infru).
At high ratios of reductant to oxidant, conditions which favour tetrathionate
formation at the expense of sulphate, the Cr(V1) oxidation of thiosulphate follows
kinetics6 9 5

and at 20 "C k , = (1.42f0.21) x lo2 I.mole-'.sec-', k , = (1.55k0.22)~lo4


12.mole-2.sec-'. E , = 8.850.7 kcal.mole-', AS: = -20.8k2.5 eu, E , = 6.8k
2.4 kcal.mole-' and AS: = - 17.7k8.0 eu. K , the formation constant of
c~s,o~-,
HCrO,

+ HS,O; + H,O +CrS,O;-

is unusually high ( 1 . 2 4 ~lo4 I.mole-') and essentially all Cr(V1) is in the form
of the complex under the reaction conditions of high [S,Oi-]. The activated
complex contains one Cr(V1) and two S,OZ - species and two-electron reduction
of Cr(V1) occurs with concomitant one-electron oxidation of both S,Oi - species

O X I D A T I O N BY

Cr(V1) A N D Mn(VI1)

287

to yield S402-. No evidence for Cr(V) or S , 0 3 . was found from attempts to


initiate induction and polymerisation reactions.
A similar investigation715 produced a simpler rate law
-d[Cr(VI)]/dt

k [Cr(VI)] [S,0:-]2(H30f]

with k = 5.5 x lo4 13.mole-3.sec-' (25 "C, pH 4, 11 = 0.23 M ) , E = 7.4 kcal.


mole- and kDZ0/kHz0
= 1.4. Spectrophotometric evidence for CrS,O,
was
obtained, yielding K = 2.2 x lo5 Lmole-' (18.4 "C).
The chromic acid oxidation of dithionic acid is independent of oxidant concentration and its rate is equal to that of the acid-catalysed hydrolysis to sulphite
and sulphate, which must therefore constitute the rate-determining process5'.

'

2.2.5 Nitrite ion


The rate law for the MnO, oxidation of NO; to nitrate is reported to be51a
-d[MnO,with E

]/dt

k[Mn04- I [NO,-

11.5 kcal.mole-' and AS'

I( 1 + k ' /[ H 3 0 +I)

-9.9 eu.

2.2.6 Trivalent phosphorus compounds


Haight et a1.52-52b
have made a comparative study of the Cr(V1) oxidations
of phosphorous and hypophosphorous acids and several organophosphorus
analogues. Neither H,PO,- nor H,PO,- ions are oxidised at an appreciable
rate; the former ion yields an anhydride, HO2POCrO3'-, with a formation
constant of 7 + 1 Lmole-' (24 "C, acetic acid-sodium acetate buffer at uH 4)
but H 2 P 0 2 - does not form an anhydride. In acidic media both neutral acids
are oxidised according to the rate law5'* 52a

where, at 25 "C, for H3P03 k , = 2 . 0 ~


k, = 6 . 0 ~
(both Iz.mole-2
sec-') and K = 16 I.mole-' and for H 3 P 0 2 k , = 6 . 5 lou3,
~
k2 = 1 3 . 5 ~
and K = 11 (same units, respectively). The slower oxidation of DP(OH),=O
indicates a primary kinetic isotope effect of ca. 4 and chloride ion inhibits reaction.
The form of the rate law suggests significant complex formation, cf. H,P03-,
References pp. 493-509

288

c o v A L E N T c o M P O u N D S,

MET A L I o NS : O X I D A T I o N-R E D u c T I o N

and the isotope effect demonstrates slow P-H cleavage, viz.


HCr0,-

+ HP(OH),-O + -OCr02-O-PH(OH)-O+
-Hf

H20

I1 + H +

HOCr0,-0-PH(0H)-0

II

HO-Cr0,-0-P-OH

-+

HO-CrO, - 0-P-OH

+ BH

H:B
2 Cr(1V) -+ Cr(V)+Cr(III)
P(III)+ Cr(V) -+ P(V)+ Cr(TZ1)
(B is a base, e.g. H,O or H,P03-)
This is entirely analogous to the Westheimer mechanism for isopropanol
oxidation (Section 2.2.3).
Clearly for P(II1) to be oxidised, the P atom must be bonded simultaneously
to (i) a hydrogen atom, (ii) an oxygen atom linked to the oxidant and (iii) a free,
unioniyd hydroxyl group [for otherwise H2P03- would reduce Cr(VI)]52.In
conformity with these conclusions replacement of the hydroxylic protons in H3P03
by ethyl groups to give diethyl phosphite results in a reduction of Cr(V1) characterised by rate-determining ester-hydrolysis to ethanol and monoethyl phosphite,
both of which are oxidised normally52b.Also DP(OC,H5),-0 is oxidised at
virtually the same rate as HP(OC,H,),=O but oxidation of C,H5P(OC2H,),-0
proceeds very slowly.
A r e i n v e ~ t i g a t i o nof~ ~the
~ hypophosphite oxidation produced rather slower
rates and a differing rate law including a term k o [HCrO; ][H,PO,] corresponding
to a complex H,PCrO; for which optical measurements provided support.

2.2.7 Arsenious acid


The chromic acid oxidation of As(1II) was first studied by De L u ~ y whose
~~,
results have been recalculated by Westheimer to afford a rate law
-d[Cr(VI)]/dt = ~[HC~O;][H,ASO,][H]~
Westheimer has also reviewed the induced oxidations by the Cr(V1)-As(II1)
couple of iodide, bromide and manganous ions (vide supra). The induction factor
of 0.5 for Mn(I1) implies an intermediate tetravalent chromium species; however,
the factor of 2 for iodide points to a pentavalent chromium intermediate. Both

OXIDATION BY

Cr(VI)

AND

Mn(V1I)

289

observations are best accommodated in the following scheme, which does not
involve the unstable As(1V)
HCr04-+As(III)

Cr(IV)+AsO,

Cr(tV)+Cr(VI)

-+

2 Cr(V)

Cr(V)+As(III)

-+

Cr(IIt)+AsO,

Edwards5 has also re-examined De Lurys data and finds evidence for an extra
term in the rate law, which becomes

+ k[HCrO,][H,AsO,][H+]

- d[Cr(VI)]/dt = k[HCrO,][H,AsO,][H+]

The reaction in 0.2 M acetic acid-0.2 M sodium acetate buffers at an ionic


strength of 1.5 M (KNO,) and a temperature of 25 C, has been reinvestigated
recently by Mason and Kowalakj6. They report the following rate law for a high
[As(In)]/ [Cr(VI)] ratio
2 kK[As(III)][HCrO,]
1+ K[As(III)]

- d[Cr(VI)]/dt =

For the conditions quoted, k = 3 . 7 6 ~


law is consistent with the mechanism
HCrO,+As(lII)
As(III).HCrO,

sec- and K

+ As(III).HCrO,

+ products

(slow)

22.4 1.mole-I. The

A second series of experiments was conducted with [Cr(VI)] > [As(III)] and
a modified rate law was obtained, viz.
-d[Cr(VI)]/dt

1.68 x lo-2[As(IIIj][HCr0,]
1 +22.4 [HCrOJ
+2.47 x 10-2[Cr20~-][As(III)]

The authors believe that a similar mechanism to the above operates for the
dichromate-ion dependent path, but that the equilibrium constant K is much
smaller. This oxidation presents two novel features, namely the lack of an acidity
dependence of the rate and the participation of a term involving dichromatic ion.
Kolthoff and Fineman have examined the reaction in alkaline media in the
presence and absence of oxygen, obtaining the rate law
-d[Cr(VI)]/dt
References pp. 493-509

= k,[CrO~-][As(III)]

290

cov A L E N T COMPOUNDS,
METAL

I O N S:

0x1D A T I ON-RED u C T I o N

with k , = 1.61 x
I.mole-'.sec-' at 30 "C, pH 9.1 and an ionic strength of
1.75. k , was independent of acidity above this pH, but the reaction was acid
catalysed at lower pH. This conclusion was based on a single measurement,
however, and the disagreement with the results of Mason and K ~ w a l a kis ~not
~
serious. The cr(vr)-As(rrr) reaction was found to induce the reduction of
molecular oxygen, thereby confirming the early work of Kessler5*. For further
details, the reader is referred to the chapter on induced reactions.
Wiberg5' has reviewed this oxidation and suggests that the complex of Mason
and Kowalak is a mixed anhydride, cf. the oxidations of sulphur dioxide (p. 285)
and formic acid (p. 316). The anhydride cannot be of neutral charge, however, as
no acidity dependence is observed and a possibility is

H3As03 HCrO;

+ H,O

+ HO-As-0-Cr0,-0I

OH
HO-As-O-CrO,-O-

-+ As(V)+Cr(IV)

(slow)

OH
Further complication is apparent when the reaction is investigated in
H,PW-HPO:while the rate shows a first-order dependence on
total [Cr(VI)] and [As(III)], a complex dependence on the buffer composition
is found, indicating two activated complexes of composition H,PO, .HCrO,.
As(II1) and HPO:-*HCrO,*As(III), which correspond to attack of HCrP0:upon neutral As(II1) and As(OH),O- respectively.

2.2.8 Carbon monoxide

The oxidation by permanganate proceeds readily in acid and neutral solutions


(to give MnO,) and in basic solution (to give MnOi-). The rate law is60*61
-d[CO]/dt

k[CO][MnOJ

with E = 13.6 kcal.mole-' and AS' = 17 eu over the p H range 1 to 13. Catalysis
by Agf and Hg2+ ions was observed (cf. oxidation of hydrogen by MnO;)
following a general rate law6'
-d[CO]/dt

k[CO][MnO;][cat]

For cat = Ag', k(0 "C) = 1 . 1 0 ~lo5 I2.mo1e-".sec-', E = 1.80 kcal.mole-'


and AS' = -31 eu; for cat = Hg2+, k(0 "C) = 1 . 0 6 lo3
~ 12.mole-2.sec-',
E = 7.0 kcal.mole-' and ASf = 21 eu. It was suggested that intermediates of

OXIDATION BY

29 1

Cr(V1) A N D Mn(VI1)

the type
[-Hg-CO-OMnO,]
are involved. It is pertinent6 that CO is readily oxidised by solid AgMnO,.

2.2.9 Molecular hydrogen

Unlike Mn(VI1) chromic acid oxidises hydrogen only in the presence of a


catalyst, namely, argentous ion63. The stoichiometry is
Cr,03-+3 H,+8 H +

2 C r 3 + + 7 H,O

and no change in the concentration of Ag(1) was found. No evidence of complex


formation between Cr(V1) and Ag(1) was obtained, and the rate law at 0.50 A4
ionic strength is
-d[H,]/dt

k[AgI2[H2]

with k = 4 x lo8 exp( - 15.8 x 103/RT)1Z.mole-2.sec-1.


When ceric ions were substituted for chromic acid, the reaction was still zeroorder with respect to metal ion, the rate of reduction of which was unchanged.
The mechanism favoured by the authors depends on formation of a complex of
silver ions and hydrogen, viz.
Agf+H2
AgH: +Agf
(Ag,

+ AgH:

+ (Ag, . . . H,)

. . . H2)++3 Cr(V1) -+

2 A g f + 3 C r 3 + + 2 Hf

(fast)
(slow)
(fast)

Other possible pre-equilibria were also considered.


Webster and H a l ~ e r found
n ~ ~ the stoichiometry of the permanganate oxidation
in acidic solution to be
2 M n 0 4 + 3 H 2 + 2 H + = 2MnO2+4HzO
and in alkaline solution (0.3-0.6 M sodium hydroxide) to be
2MnO;+H2+2OH-

= 2Mn0:-+2Hz0

In acidic solution the kinetics are


-d[H,]/dt
References pp. 493-SO9

k,[H,][MnOh]

c o v A L E N T c o M P o u N D S,

292

M E T A L I O N S: O X I DA T I o N-R E D u c T I O N

with k2 = 4 . 2 lo9
~ exp[-(14.7&0.5) x 103/RT]I.mole-'.sec-'
(p = 0.3 M ) .
The reaction is subject to slight acid catalysis. In neutral solution the rate expression has the same f ~ r m ~ ~but
. ~ k' ,is lower by 15 %. In basic solution also no
change in the law was observed and no significant salt effect was found at any pH.
The reaction has the same velocity in deuterium oxide66.
The authors favour a two-equivalent oxidation in order to avoid the necessity
to postulate free hydrogen atoms as intermediates in the reaction, viz.
MnO,+H,

Mn0;-+2 H + or HMnO:-+H+

or
MnO,+H,

MnO;+H,O

The reaction was also found to be catalysed by silver ion6,, being of the third
order, with k , = 7.5 x lo7 exp[( -9.3f.0.5) x 103/RT]I2.mole-'.sec-' (p = 0.3
M ) . Ag+ is thought to form a complex with MnO, (a process for which there is
spectroscopic support), which attacks the hydrogen molecule, viz.
Ag'
*
*

+ MnO, + AgMnO,

+ H 2 + AgH+ + Mn0:- + H +
AgH+ + MnO,
Ag+ + H + Mn0:AgMnO,

--f

2.3

(fast)
(slow)
(fast)

O X I D A T I O N O F MONOFUNCTIONAL O R G A N I C MOLECULES

It is not within the scope of this review to deal with products of oxidation
of organic molecules in cases other than those for which a kinetic analysis has also
been attempted. However, much insight into the modes of reactivity of Cr(V1)
and Mn(VI1) is to be gained from such information and the reader is referred to
the recent excellent reviews of Wiberg67 on Cr(V1) and Stewart6' on Mn(VII1).
One major difference between the published accounts of oxidation of inorganic
and organic species is the more prevalent use of non-aqueous solvents for the latter,
which gives rise to several complicating features.
2.3.1 Aliphatic hydrocarbons

Work has been concentrated on hydrocarbons containing activating groups,


such as phenyl or tert-butyl, in order to achieve convenient rates of reaction. The
relative rates of oxidation of two series of hydrocarbons with chromic acid in
95 % acetic acid are6's7'
(a)

cyclohexane methylcyclohexane toluene ethylbenzene diphenylmethane


0.01
0.08
0.16
0.50
1.oo

(b)

OXIDATION BY

toluene ethylbenzene
1
7.2

Cr(V1)

AND

isopropylbenzene
71.1

Mn(VI1)

293

tert-butylbenzene
0.019

By comparing a number of aliphatic hydrocarbons of different structure, Mareg


and RoCek7' compute relative rates of oxidation of methyl, methylene and methine
groups to be 1 : 114 : 7000-18,000.
The simplest compounds which have been examined are the series
CH,(CH,),CH371. The observed rate law is7'
- d[Cr(VI)]/dt = k[Cr(VI)] [alkanelh,

k was found to increase regularly with n and at a temperature of 50 "C, an acidity


of 0.2 M (H2S04) and a solvent composition of 99 % acetic acid, k is in good
agreement with values calculated from

nkCHz

where kCH2= 5.73 x lo-' I.mole-'.sec-'. The same rate law was found for
c y c l ~ h e x a n e and
~ ~ the oxidations of both cyclohexane and n-heptane were
greatly retarded on addition of 4 % of water to the acetic acid medium. This effect has been found subsequently for many oxidations of organic substances by
chromic acid in acetic acid (vide infra).
RoCek et ~ 1 have
. also
~ measured
~
rate coefficients for a series of cyclo-alkanes,
(CH,),(n = 4 to 14), and find the analogue of kCH2
in equation (25) to fluctuate
with ring size in a manner corresponding exactly to the enthalpy of combustion
of the cycloalkane concerned per methylene group, provided n is greater than five,
i.e. there exists a direct correlation between reactivity and thermochemical strain.
For straight chain and cycloalkanes, RoCek et ~ 1 . ' prefer
~
a mechanism involving hydride ion abstraction to give a partly-developed carbonium ion which
suffers further reaction with the Cr(rV) portion before it can become free to give
acetate or olefin

The oxidation by Cr(V1) of aliphatic hydrocarbons containing a tertiary carbon


atom has been studied by several groups of workers. Sager and Bradley74 showed
that oxidation of triethylmethane yields triethylcarbinol as the primary product
with a primary kinetic isotope effect of about 1.6 (later corrected by Wiberg and
Foster7' to 3.1) for deuterium substitution at the tertiary C-H bond. Oxidations
References pp. 493-SO9

294

COVALENT COMPOUNDS, METAL I O N S :O X I D A T I O N - R E D U C T I O N

of norbornane and bornane proceed readily74, but the production of P-camphor


from the latter indicates the mefhylene group to be oxidised in preference to the
bridgehead tertiary C-H bond. Mare3 and RoCek7' sought evidence for steric
hindrance in their examination of the relative rates of oxidation by Cr(V1) of the
following series
(CH,),CH
(relative) k ,
1

(CH3),CH(neo-C5Hii) CH,CH(neo-C,H,,),
0.674
0.104

They also compared compounds which might be expected76 to give strong


neighbouring group effects should a carbonium ion intermediatc be formed, viz.

k, (relative)

cyclohexane
1

bornane
2.0

isobornane
8.0

These effects are very much smaller than those found for the first-order solvolyses
of bornyl and isobornyl chlorides, which differ in relative rate by several orders of
magnitude. However, the authors argue that this does not necessarily disprove
participation by a carbonium ion in the mechanism which they proposed for the
oxidation of straight-chain alkanes (see above).
Wiberg and Foster75 oxidised (+)-3-methylheptane in 91 % acetic acid with
chromic acid and obtained (+)-3-methyI-3-heptanol with 70-85 % retention of
Configuration. This implies that should the initial product of oxidation be either
MeEtPrC' or MeEtPrC., then neither diffuses away from the solvent cage before
reacting further. Slow initial formation of [R3c.Cr(V)] would be expected to be
followed by rapid transfer of a second electron to give [R,C+*Cr(IV)]. The latter
could be an ester of Cr(1V) which suffers Cr-0 fission to give R,COH with
retention of configuration.
Closely related to this work is that of Wiberg and Evans69 on the Cr(VI)
oxidation of diphenylmethane in 95 % acetic acid. This has the rate law

- d [Cr(VI)]/df

= k [substrate] [ C r 0 3 ] h ,

and activation parameters; E = 15.8 kcal.mole-', AS' = 21.2 eu. A primary


kinetic isotope effect of 6.4 was obtained at 30 "C and electron-releasing groups
enhanced the reaction rate (p' = - 1.17). The appearance of the total concentration
of hexavalent chromium in the rate equation is significant. In many oxidations,
of both organic and inorganic substrates, reactivity is confined to H,Cr04 even
in an aqueous acetic acid medium, e.g. aromatic aldehydes (p. 310). Evidently
dichromate ion has comparable reactivity in this particular oxidation.
Writing the oxidant as H2Cr04 while noting the participation of HCr,O;,
Wiberg and Evans69 present five mechanisms which accommodate both the

OXIDATION BY

Cr(V1)

AND

Mn(VI1)

295

kinetics and the C-H cleavage implied by the isotope effect, viz.
A.

B.
C. Ph,C

-H

H\

O=Cr03H,

0.

OH

E.

0 = Cr0,H2
PhfCH-Ph
I
H

>

Ph2C--H

HO8H

+ Cr(lV)

Ph,C-H

5'-

OH

bH

0 -Cr(OH),O'
I
Ph-CH-Ph

The lack of steric effects i n oxidations of hydrocarbons by Cr(V1) renders D


and E unacceptable. The activated complex of scheme C is non-linear and hence
does not comply with the magnitude of the observed isotope effect. Two pieces
of evidence are quoted which indicate A t o be the more probable of the remaining
two. Firstly, the p + constant of - 1.17 is more in agreement with that obtained
for bromine atom abstraction from toluenes (-1.369 to -1.806)77 than those
found for solvolyses involving electron-deficient carbon (- 2.57 to - 4.67)78.
Secondly, the correlation between the relative rates of oxidation of the series

and the corresponding relative rates of hydrogen abstraction by CCI,. is much


superior to that between oxidation rates and relative rates of solvolysis of the
corresponding chlorides. It should be noted, however, that this type of argument
has received criticism79.
Wiberg prefers mechanism A to the carbonium-ion mechanism with the proviso
that the radical is oxidised before inversion occurs. The carbonium ion formed
must rapidly acquire an oxygen atom to prevent inversion and the two processes
may be synchronous. The minor role which free carbonium ions may play in the
reaction has been discussed".
A quantity of earlier work exists on chromic acid oxidation of hydrocarbons.
It was noted that diphenylmethane and other hydrocarbons in glacial acetic acid
solution are oxidised rapidly during the initial stages but that reaction is autoretarded". The autoretardation is eliminated on adding 2.5 % of sulphuric acid.
The orders of the reaction with respect to diphenylmethane and Cr(V1) are one
and two respectively*', the latter differing from that found by Wiberg and
References p p . 493-509

296

COVALENT COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

Evans69. The more recent result is to be preferred in view of the wider range of
Cr(V1) concentration employed.
Electron-releasing groups facilitate the oxidation of a series of para-substituted
toluenes, although determination of the rate coefficients again depended on a
second-order decay plot of the concentration of Cr(VI)83.
Chromyl chloride combines with arylalkanes in inert solvents to deposit an
Etard complex of the general formula ArH.2 CrO,CI,. On hydrolysis this yields
the expected oxidation product. Study of the reaction has been from three standpoints, namely the formation and the spectroscopy of the complex and the products
of its decomposition. The latter aspect has been covered by Wiberg67, who gives
details of the reaction of o-nitrotoluene to give an apparently normal complex
which, on hydrolysis, reverts mainly to the starting compound84. This implies
that oxidation occurs only during hydrolysis. However, both the magnetic
susceptibility85*8 5 a (3.2+_0.06 Bohr magnetons per Cr atom) and the ESR spectrum86 of the complex with toluene, indicate the presence of Cr(IV)67.
The oxidations of toluene, diphenylmethane and triphenylmethane are second
order, relative rates at 22 C being
PhCH,
k,(l.mole-.sec-)

1.28 x

Ph,CH2

Ph,CH

1 . 6 7 ~lo-

0.16

, it was concluded that the


The effect of solvent upon k , has been r e p ~ r t e d ~and
activated complex is not sufficiently polar to be called ionic. The oxidations
of toluene and triphenylmethane exhibit primary kinetic deuterium isotope
effects of 2.4 and ca. 4 respectively. No isotopic mixing occurred during formation
of the Etard complex from a mixture of normal and deuterated o-nitr~toluene~.
The chromyl chloride oxidation of a series of substituted diphenylmethanes
revealed that electron-withdrawing substituents slow reaction while electronreleasing groups have the opposite effect, the values of p and p + being - 2.28 & 0.08
and -2.20k0.07 r e ~ p e c t i v e l y ~ ~ .
Stairs93 comments that this p value is strongly dependent on the temperature
but his data have been criticised by Duffin and Tucker94, who prefer their method
of observing the rate of formation of the adduct to that of estimating total residual
oxidising power employed by Stairs, and they find p + (25) to be -2.32k0.10 as
compared to a value of -2.20k0.07 at 40 C. These values are considerably more
negative than those found for chromic acid oxidation of diphenylmethanes
(- 1.17) and toluenes ( - 1.12).
The relative rates of oxidation of phenylmethanes cover too small a range to
be compatible with carbonium ion formation (cJthe discussion on chromic acid
oxidation of diphenylmethane, p. 295), and an initial reaction to give a radical
plus Cr(V) followed by rapid transfer of a second electron to form Cr(1V) is more

OXIDATION BY

Cr(V1) A N D Mn(VI1)

297

likely67*", viz.
Ar3CH+ CrO2CI2 + [Ar3c.0Cr(OH)C121
[Ar3~.0Cr(OH)CI,]+ Ar,C-O-Cr(OH)CI,
Ar,CH, would react further to give Ar2C[OCr(OH)CI,]2, hydrolysis of which
would give the observed products.
NecSoiu et al.79.9 s dispute these conclusions and prefer an S,2 process, viz.
\
\

-C-H

o:c/rcl,

:
'

-C
/ ' ,

H..-O

CI

L-o-&-cl
I
OH

The transition state does not involve a large degree of charge separation and hence
the relative rates of oxidation of toluene, diphenylmethane and triphenylmethane
may be explained in terms of an ionic mechanism.
The aqueous permanganate oxidation of hydrocarbons has been relatively
neglected for reasons of low solubility, and discussion of the oxidations of C-H
groups has derived largely from studies of certain carboxylic acids or alcohols
when the functional group is not directly involved. The oxidations of 4-methylhexanoic acid, 5-methylheptane sulphonic acid and p-sec-butylbenzoic acids are
of a total kinetic order of
at pH 13. Oxidation of 4-methylhexanoic acid
proceeds with 35 % retention of configuration at pH 13, but that of p-sec-butylbenzoic acid displays only 6 % retention. About 25 % of l80was transferred to
the product, 4-hydroxy-4-methylhexanoate lactone, from labelled MnO,, but
only 0.6 to a-hydroxy-p-sec-butylbenzoic acid. As in the case of chromic acid,
three modes of electron-transfer deserve consideration
R3CH

MnO,

R3C+ + M n ( V )

The factor of seven variation between k, for the ordinary and benzylic tertiary
hydrogen is too small to be associated with carbonium ion formation. The observed degrees of retention and l8O transfer imply a caged radical rapidly reducing
Mn(V1) to Mn(V) by accepting oxygen
[R3C.MnO4H-] + R,C-O-Mn03H
Solvolysis follows, with carboxyl participation where this is feasible, to yield
R3COH, although several modes are involved. Other stereochemical work has
been summarised by Stewart6'.
References pp. 493-SO4

298

COVALENT COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

A subsequent detailed analysis665of the permanganate oxidation of the tertiary


hydrogen atom of 4-phenylvaleric acid in 2.5 M potassium hydroxide solution
supports the caged radical mechanism. The reaction order is two overall, k,/k,
is ca. 11.5, ring substitution has little effect on the rate ( p 0) and the oxidation
proceeds with a net 30-40
retention of optical configuration.
In an earlier series of experiments, Cullis and Ladbury9' examined the kinetics
of the permanganate oxidation of hydrocarbons in acetic acid solution. Initial
attack on toluene occurs at the methyl group and a total order of two was found.
Electron-withdrawing agents reduced the rate of oxidation. However, the effects
of added salts were complex and the authors believe that lower oxidation states of
manganese are responsible for the oxidation. The oxidation of ethylbenzene
produced acetophenone, the process being second-order with

k,

1.9 x lO"exp(-

14.8 x 103/RT) I.mole-'.set-'.

However, added pyrosphosphate reduced the rate by 40 %.

2.3.2 Olefns and acetylenes


The oxidation of olefins by Cr(VT) follows kinetics97a

- d [Cr(VI)]/df

k , [Cr(Vr)][olefin][H,O+]

The reaction is insensitive to steric effects, but k , increases with the number of
alkyl substituents although it is not influenced by position of the alkyl groups
(unlike oxidation by TI(T11))97b.In these respects Cr(V1) oxidation resembles
bromination, chlorination and epoxidation and a symmetrical transition state of
the type depicted is favoured.

Chromyl chloride oxidation of alkenes proceeds 2:iu the formation of adducts


at a rate necessitating stopped-flow techniques. At 15 "C the formation of 1 : 1
adduct from styrene and oxidant in CCI, solution707is simple second-order with
k2 = 37.0 I.mole- '.set-'. Measurements with substituted styrenes yielded p + =
- 1.99. E = 9.0 kcal.mole-' and AS* = -23.8 eu for styrene itself. Hydrolysis
of the styrene adduct yields mostly phenylacetaldehyde (76.5 %)and benzaldehyde
(21.1 %). Essentially similar results were obtained with a set of 15 a l k e n e and
~~~~

OXIDATION B Y

Cr(V1)

AND

Mn(VI1)

299

some c y c l ~ a l k e n e swith
~ ~ ~E and A S f values in the ranges 6.0-7.6 kcal.mole'-'
and 27-41 eu, respectively. Formation of rearranged products
R3
I

RIR,C=CR,R,

-+

Rl-C-COR,
1

R,
suggests alkyl migration on hydrolysis of the adduct.
The oxidation by permanganate has been investigated both at pH 6.5 and
13 using salts of unsaturated carboxylic acids to facilitate d i s s ~ l u t i o n ~At~ .low
substrate concentrations (M ) and pH 13, the reaction is first-order with
respect to both oxidant and substrate but at 2-3 x
M substrate concentration
significant deviation is apparent, especially at 0 "C. The initial rate is independent
of base concentration but the rate at later stages is strongly affected. This suggests
that the reduction of a second permanganate ion is influenced by hydroxide ion
as in the following scheme
MnO, +crotonate ion

5B

(26)

Thus d[MnO,]/dt = - k , [MnO,][crotonate]-k,[MnO,][B]


and d[B]/dt =
k , [MnO, ] [crotonate] - k , [MnO, ] [B]. Elimination of time as a variable followed by integration and further manipulation leads to an expression for k,.
At 25 "C and p H 13, k , = 220 I.mole-'.sec-'; at pH 6.5, k , = 232 Lmole-'.
sec-'. The reaction at pH 6.5 showed no deviation from simple second-order
behaviour even at 0" C and the individual rate plots were straight. Evidently
hydroxide ion does not affect the initial stage of reaction, which has a low activation energy (5.4k0.7 kcal.mole-') and an activation entropy of - 3 2 + 2 eu.
At pH 13 the final products are manganate and cis-diol. The observation of a
large degree of transfer of l8O from labelled MnO, into oleate ion99 at pH 12
suggests formation of a cyclic ester, uiz.
'C'

II

/c\

+ MnO;

C'-O,

c
,-,0

M
, "'

0"0

The oxidation of a series of olefins reveals the reaction to be very insensitive


to electronic effectsg8. Phenyl and methyl substitution of the olefin mildly accelerate reaction. In all cases k , is pH-independent. Data are collected in Table 1.
The rate of permanganate oxidation of acetylenedicarboxylic acid to CO, at
pH 0.25-5.0 requires the use of the stopped-flow method, which yielded simple
References p p . 493-509

300

c o v A L E N T c o M P o u N D S,

M E T A L I o N s : o x I D A TIo N-R E D u c T I o N

TABLE 1
R A T E P A R A M E T E R S FOR F I R S T S T A G E O F T H E O X I D A T I O N O F O L E F I N S B Y P E R M A N GANATE

Temperature
Olefin
Ally1 alcohol
Acrylate ion
Crotonate ion
Vinylacetate ion
4-Bromocrotonate ion
2-Pentenoate ion
3-Pentenoate ion
4-Pentenoate ion
Cinnarnate ion
p-Methylcinnamate ion
p-Methoxycinnamate ion
p-Chlorocinnarnate ion
Oleate ion

= 25

k (I.mole- '.set- l )

C; pH 13.
E

137f7
330f 17
220f 11
111*6
325j=16
242k 12
455f23
167f8
503525
278k 14
264% 14
302+ 15
70

(kcal.mole-)

ASt(eu)

8.1 f0.7
5.9f0.7
5.4f0.7
7.0k0.7

-24f2
-3Of2
-32f2
-28k2

7.2&0.7
3.9f0.7

-27f2
-36f2

__

second-order kinetics7 lo. Although the removal of MnO, (A,,, 540 nm) under
conditions of excess substrate showed clean first-order behaviour, a transient
built up and then decayed at ca. 250 nm, the region of absorption by Mn(II1).
Quenching the reaction at this point in time enabled identification of oxalic acid
as another intermediate. At pH > 3, when the dianion of the substrate is the only
significant species, E = 6.7 kcal.mole- and ASt = -32 eu.

2.3.3 Alcohols

The oxidation of alcohols to carbonyl compounds by hexavalent chromium


is a reaction of the greatest synthetic importance. Wibergs recent review67 covers
the range of compounds to which the method has been applied together with the
resulting yields and products. Very many examples of the reaction have been
studied from the mechanistic point of view, but the work of Westheimer and his
group has contributed most to our understanding of the reaction, providing, indeed, a model for many subsequent investigations. In the late nineteen-fifties
RoCekoo, 1 0 1 proposed a mechanism alternative to Westheimers so-called ester
mechanism20 but more recently the two proponents have achieved a common
viewpoint02. The general reaction has been reviewed several
and it is
sufficient to enumerate the main points, which relate to isopropanol, unless stated
to the contrary.

( a ) Products
In general, the substrate RlR2CHOH, where R, or R, and R, may be H, is

OXIDATION BY

Cr(V1) A N D Mn(VI1)

301

oxidised to R,R,C-0 and the oxidant is reduced to chromic ion. Complications


may occur when R, is not a simple alkyl group (see later).
(6) Kinetics
In aqueous solutions of moderate acidity ( < 1 M ) the rate expression i ~ ' ' ~ , ' ' ~

- d[ Cr(VI)]/d t

= k , [HCr OJ

[R,CHOH] [H30 '3

+ k,[HCr0,][R2CHOH][H30+]z
The dependence on [HCrO,] is a feature of many other oxidations by chromic
acid and is characterised by a falling-off of the apparent first-order rate coefficient for disappearance of Cr(V1) at higher concentrations of oxidant. In more
strongly acidic solutions the kinetics are1"*

- d [Cr(VI)/dt

k[Cr(VI)] [R,CHOH]h,

~ exp(-13.1 x 103/RT)I.mole-'.sec-'. At
At an ho value of 1, k = 1 . 4 lo7
acidities Ho = -3 to -9, the rates of oxidation of 2-propanol, l,l,l-trifluoro2-propanol and 1,1,1,3,3,3-hexafluoro-2-propanol
reach maxima' O 5 at different
values of h,.
Primary isotope eflect
At low acidity the relative rates of oxidation of isopropanol and (CH,),CDOH
(contaminated with "light" isopropanol) were found to be 6.4+ 1.9 at 40 "C after
correction for incomplete deuteration' 0 6 . k,/k, has also been determined and the
results are in agreement'O7. Under extremely acid conditions k,/k, decreases to
1.3'05.
(c)

(d) Solvent isotope eflect

Results on the relative rates of oxidation in light and heavy water were considered' O8 in terms of the two components of the rate expression. k,(D,O/H,O) is
2.44 and k,(D,O/H,O) is 6.26.
( e ) Replacement of hydroxylic hydrogen by an alkyl group
This replacement (by an isopropyl group to give di-isopropyl ether) reduces
the oxidation rate'" by a factor of about 1500.

(f)Effects of substituents in the group R,


The oxidation of a series of meta- and para-substituted cr-phenylethanols shows
that electron-donating substituents facilitate reaction ( p = - 1.01)'09. A similar
study"' of primary aliphatic alcohols confirmed this trend (p* = - 1.06f0.06).
References p p . 493-SO9

302

c o v A L E N T c o M P o u N D S,

M E T A L I o NS: OX I D A T I o N-RED

ucT I o N

( 9 ) Base catalysis

An o b s e r ~ a t i o n that
' ~ ~ the oxidation is subject to marked catalysis by pyridine
was later refuted'0'*'''.

( h ) Induced oxidation
The oxidation induces the oxidation of manganous ion to MnOz with an induction factor (see p. 280) of 0.5'12. The rate of reduction of Cr(V1) is reduced
simultaneously to a maximum extent of 50 %lo3.

(i) Steric eflect


The alcohol 3p, 28-diacetoxy-6p-hydroxy-l8/312-oleanene was oxidised in an
aqueous acetic acid medium"'. Deuteration at the six position had no effect on
the rate in solvents of high (> 80 %) acetic acid content, but the isotope effect
reached 2 in 60 % acetic acid. Increasing the acetic acid content of the medium
produced a much larger effect on the oxidation rate of cyclohexanol than of the
polycyclic alcohol.
( j ) Stoppedjlow investigation
The resolution of the overall reaction into steps implied by the steric effect
(above) has been achieved"3."3b for the oxidation of isopropanol. In 97 %
aqueous acetic acid a rapid reaction, k, M 1.25 x lo4 1.mole-'.sec-' (15 "C,
p = 0.183 A4 NaClO,), which is unaffected by deuteration, precedes the oxidation.
Evidence for a n intermediate has been reported for the oxidation of l,l,l-trifluoro-2-propanol at very high acidities' 0 5 .
Numerous further investigations of the reaction have been reported, but those
outlined above bear most directly on the problem.
Westheimer's "ester mechanism" is
H C r O i + H30'

=H2Cr04

H20

(28)cfast)

R2C-0-Cr -OH
I
11
H
O

(29) (fast)

0
II

R2CHOH

H2Cr04

0
I1
R2C-OpCr-OH

IJ

H20)

R,C=O

0
&-OH

H30t

0-

(30)(slow)

The importance of Cr(V) as an intermediate has been confirmed by a stopped-

OXIDATION BY

Cr(V1) A N D Mn(VI1)

303

flow i n v e ~ t i g a t i o n ' ~ ~ *A" ~new


~ . species is formed absorbing at 510 nm and
displaying a narrow ESR signal centred on g = 1.9805. This species is reduced by
the alcohol with a primary kinetic isotope effect of 9.1 *0.81'3b. The term in
[H30+I2in the rate expression corresponds to a protonated ester.
Oxidation of cyclobutanol by a Cr(V1)-V(IV) couple appears to involve attack of Cr(1V)upon the substrate to yield a free radical, (Section 2.5). This implies
the following possible variation in the Westheimer scheme for a secondary
alcoholz0
Cr(IV)+R,CHOH

C r ( I I I ) + R 2 c O H+ HS

+
+
Cr(V) + R2CHOH + R,C = 0 + Cr(II1)

R 2 c O H Cr(V1) -+ RzC = 0 Cr(V) H

The rival mechanism advanced by RoCek and KrupiCka'

has, as the first step

The two major differences are: (i) the absence of any complex or ester in the
second mechanism, and (ii) the direction of the movement of the electron pair
of the C-H bond, the hydrogen becoming proton-like in the transition state of
the Westheimer scheme but hydride-like in that of the RoCek mechanism.
Both schemes accommodate the kinetics, the primary isotope effect and the
induction factor, which indicates that Cr(IV) is the initial stage of reduction of
the oxidant. RoCek's mechanism does not accord with the solvent isotope effect
of 2.5 per proton, which has just the value to be expected for acid-catalysis, for
the 0-H bond cleavage should be subject to a primary isotope effect of about
7. The ester mechanism is not open to this criticim.
The main objections to the ester mechanism are based on the substituent effects
and, to a lesser degree, the kinetics in strongly acidic solutions. The departure of
the hydrogen as a proton in the slow step should be facilitated by electron-withdrawing groups, but the converse is found. Substituents are, of course, bound
to affect both the esterification equilibrium (29) and the oxidation step (30) and
the observed p is the sum of the individual values. Spectroscopic evidence has been
~ b t a i n e d " ~ . ~, ho
" wever, for the existence of 1 : 1 complexes in solution between
alcohols and acid chromate ion, e.g. KzsOfor isopropanol is 0.083 I.mole-'
( p = 0.001 M ) . These esters can be extracted into benzene'04 and are rapidly
decomposed therein on addition of pyridine to give the oxidation product, although
hydrolysis is a competing reaction' 16. The pyridine-catalysed oxidative breakdown in benzene has a primary kinetic isotope effect of about 5 (ref. 116).
References pp. 493-509

304

c o v A L E N T co M PO u N D s, M E T A L I o N S : o X I D A T I o N-RE D u c T 10N

These results do not prove that the ester is an essential intermediate in aqueous
solution even though it is present, but the result with the hindred triterpene is
convincinglo2. In this case the esterification step, which is normally fast, has
become rate-determining and the disappearance of the isotope effect must mean
that C-H cleavage occurs after the formation of the ester and not independently of
it. The generality of this result is apparent from the stopped-flow investigation of
isopropanol oxidation ' 1 3 .
The h, dependence of the oxidation in strongly acidic media was taken by RoCek
as implying that no water molecule is involved in the transition state and that
consequently the ester mechanism as portrayed above cannot hold. However,
the Zucker-Hammett hypothesis upon which this argument is based, i.e. that a
reaction forming a transition state containing a water molecule will follow a
[H,O+] dependence, but that otherwise an h, dependence will be followed, may
not be valid, and in any case the ester can be depicted as breaking down as foll o W s 1 l 7 , 118

The maximum rate reported at very high acidity accords with protonation of the
ester' 0 5 . To summarise, the ester mechanism has gained general acceptance,
although the substituent effects have yet to be explained wholly satisfactorily and
the exact nature of the transition state, i.e. whether it is of considerable or only
slight carbonyl character, remains a contentious issue67, '.
Many interesting observations not bearing directly on the general mechanism
have been reported. Certain secondary alcohols undergo cleavage' 2 0 , e.g.

''

H
I

C6H ,-C-C(CH,),
I
OH

Cr(V1)
--f

C,H ,C-C(CH,),,

C 6 H 5 C H0and (CH,),COH

II

which is up to 70 % of the total reaction in the case of phenyl-tevt-butylcarbinol.


Addition of manganous or cerous ions, which react with Cr(IV), greatly reduces
the degree of cleavage. C,H,CD(OH)C(CH,), is oxidised at only one-twelfth
of the rate of the ''light'' alcohol at 0 "C. Cleavage does not break the C-H(D)
bond and hence the isotope effect implies that cleavage does not occur by attack
of Cr(V1) upon the alcohol, for otherwise the oxidation route would be completely
usurped for the "heavy" alcohol. Presumably, therefore, cleavage is entirely due
to attack of Cr(IV) or Cr(V).
The following mechanism for cleavage was proposed'2'

OXIDATION BY

Cr(VI) A N D Mn(VI1)

C6H5CH(OH)C(CH,)3 Cr(V)

305

+ C6H5-CH-C(CH,),
I
I

0-Cr (V)
C6H5-CH-C(CH,), + C6H5CHO+Cr(lII)+ (CH3),C+
I

0-Cr(V)
(CH3),Cf H 2 0 4 (CH,),COH

+H +

This is supported by the observation'20 that 180-labelled alcohol is cleaved to


unlabelled t-butanol. Wiberg67 has given other examples of this type of cleavage,
and has dealt with other complications such as further oxidation of products,
formation of hemiacetals, etc.
The reactivity of Cr(IV) towards cyclobutanol to give 4-hydroxybutyraldehyde
(ref. 1 8 6 b ) (Section 2.5) suggests the possibility of participation of this species in
cleavage processes, and Nave and Trahanovsky6 have demonstrated radical
processes in the Cr(V1)oxidation of 2-aryl-I-phenylethanols by trapping C6H5CH2*
with molecular oxygen (to give benzyl alcohol). Ring-substitution of the 2-aryl
group produced good Hammett plots both for ketone formation ( p = -0.1Of
0.02) and for cleavage ( p = 1.06f0.04). The Westheimer cleavage mechanism is
accordingly modified to

+
Cr(IV) + RCHOHR' -,Cr(II1) + RCHO + R',
Cr(V) + R'OH
Cr(V1) + R'.
2Cr(V) + Cr(V1) + Cr(IV)

Cr(V1) + RCHOHR'

Cr(1V) RCOR'

Interest has been shown by several groups on the effect of solvent and of added
anions upon the oxidation of alcohols. The oxidation of isopropanol proceeds
2500 times faster in 86.5 % acetic acid than in water at the same hydrogen ion
c ~ n c e n t r a t i o n ~The
~ . kinetics and primary kinetic isotope effect are essentially
the same as in water. Addition of chloride ion strongly inhibits the oxidation and
the spectrum of chromic acid is modified. The effect of chloride ion was ratioin terms of the equilibrium,
H++Cl-+HCrO;

+ CICrO;+H,O

where K== lo5 12.mole-2, and the chlorochromate ion is presumed to be of


negligible oxidising power. This type of equilibrium has been detected spectrophotometrically for phosphate2', sulphate'22, 1 2 3 and other inorganic anions2'
and its influence on the rate of oxidation of isopropanol has been examined
systematically2'. The indications from optical measurements are that the anion
A- does not enter HCrO; to any great extent but that it does enter the neutral
References p p . 493-SO9

306

c o v A L E N T c OM P o u N D s, M E T A L I o N s: O X I D A TI o N-R

E D u CTI o N

acid, viz.
H,CrO,

+ HA + H,O + HCr0,A

where A = OP03H,, -Cl, -OS03H, -0CI0, and -ONO,. A correlation existsz9


(with the exception of HNO,) between the pK values of chromic acid in aqueous
solution of the mineral acids and pK values of the mineral acids themselves.
These conclusions receive support from the kinetic dataz9.The rate of oxidation
by the acid chromate ion is independent of the nature of the mineral acid but at
higher acidity, when the oxidation by chromic acid becomes dominant, the rate
depends not only upon the acidity but also upon the acid concerned, the oxidising
ability of the species, HCr03A, increasing in the order
H,CrPO, < HCrCIO, < H,CrSO, < HCrClO, < HCrNO,
for a given value of h,. From the mechanistic point of view, however, this merely
means the replacement of an OH group on the chromium atoms in reaction (30)
by A.
The effect of adding large quantities of acetic acid to the medium is more
complicated. The acceleration of the oxidation rate of isopropanol was ascribed
initially to a shift of the esterification equilibrium to the right (reaction 29).
However, RoCek found that acceleration by acetic acid occurs for oxidations which
cannot involve a pre-equilibrium esterification, e.g. those of aliphatic7' and
a l i c y c l i ~ ~ ~hydrocarbons.
*
The obvious alternative, i.e. that acetic acid combines
with chromic acid, viz.
H,Cr04 + CH,CO,H

+ CH3COOCr0,0H + H,O

has been proposed'24*


and is supported by a very detailed examination of the
oxidation of isopropanol in trifluoroacetic acid3'. A plot of log k, for the latter
medium against Ho revealed a kink at H, = 0.30. A similar kink appears on graphs
of the acidity dependences of the oxidation in various aqueous mineral acids (vide
supra) and corresponds to a changeover of the oxidant from HCrO; to HCr0,A.
The reaction is independent of H, for H, < - 1.9. The thermodynamic parameters
and primary kinetic isotope effect for the reaction in CF,CO,H are approximately
the same as in water. At high acidities chloride ion exhibited a very strong specific
retarding effect ascribed to the process
HOCrO20COCF,

+ C1- + HOCr0,CI- +CF,CO;

Of relevance also is the effect on chromic acid oxidation of aqueous isopropanol


of systematically replacing water by acetonelz6. Most parameters, including the

OXIDATION BY

307

Cr(V1) A N D Mn(VI1)

primary kinetic isotope effect, the Hammett p value and salt effects are largely
unaltered, but the rate is enhanced by a factor of 7 x 10 in 93.3 % acetone compared with that in water and depends on the h, function. These observations cannot
be accommodated by invoking substitution of an anion into HCrO, and the enhancement may originate in a shift of the equilibrium (29) to the right.
A number of effects of conformation have been reported118. Crowding of
the hydroxyl group results in faster oxidation and alcohols with axial hydroxyl
groups are oxidised more readily than the equatorial isomers, e.g. cis-4-tertbutylcyclohexanol is oxidised about three times fasterlZ7than the trans isomer at
25 C.
TABLE 2
OXIDATION OF SUBSTITUTED PHENYLTRIFLUOROMETHYLCARBINOLS BY CHROMIC
A C I D AT

Substituent

25C

k2 (acreraged) for light alcohol (I.mole- '.set- )


0.544
0.256
0.1 1 1
0.0494
0.01 17

CH3
H
m-Br
m-NO2
3,5-Dinitro

kl,/kD
7.4040.20
8.53&0.15
9.803=0.20
12.20*0.20
I2.93f0.78

One further striking kinetic feature is the magnitude of the primary kinetic
isotope effect for the oxidation of certain aryltrifluoromethylcarbinols18 (Table
2 ) . It is clear that a correlation exists between the isotope effect and the oxidation
rate, slow oxidations exhibiting large isotopes effects. The magnitude of the
isotope effects may be due to one of several effects including loss of bending
frequencies in the transition state or quantum-mechanical t~nnelling~.That
large isotope effects are often associated with neighbouring groups of fluorine
atoms is apparent from the analogous results with permanganate (p. 308).
The oxidation of tertiary alcohols by chromic acid is comparatively slow and
shows a zero-order dependence of the rate upon oxidant concentration130*1 3 1 . For
l-methylcyclohexanol31 the kinetics are
-d[Cr(VI)]/dt = k[alcohol]h,
Clearly the rate-determining step is the elimination of water from the protonated
alcohol to form a carbonium ion, which loses a proton to give the readily oxidised
olefin.
The Cr(V1) oxidation of di- and triarylcarbinols in acetic acid-sulphuric acid
mixtures shows normal kinetics (first-order each in alcohol and Cr(V1) and an
h, dependence up to H,
- l)131a. Electron-donating substituents appear in the
phenolic component and the rate-determining step is believed to involve a 1,2-

References p p . 493-SO9

308

cov A L E N T co M P O U N D s, M E T A L I O N S : 0x1D AT ION-RED u C T IO N

aryl shift, uiz.

a+

Ar3C-O-Cr03H,+

--+

Ar&-O..
;a+;

. Cr03H,

-+

Ar,C-0-Ar

Is';

Ar,C-0-Ar -I-HzO --+ Ar,C=O t ArOH H +


The reaction parameters for tri- and diarylcarbinol oxidation respectively are
p + = -0.879 and p = -0.54. For migration in triarylcarbinols p + = - 1.44.
For triphenylcarbinol at an HR of -3.95, E = 12.4 kcal.mole-' and ASt =
-25.1 eu whilst for benzhydrol at HR = -2.62, E = 6.5 kcal.mole-' and AS' =
-37.1 eu.
The oxidation of alcohols to carbonyl compounds by permanganate proceeds
most rapidly in basic solution and it is with this medium that the majority of
kinetic studies have been performed.
As with chromic acid, tertiary alcohols are oxidised only very slowly with
degradation. The rate expression for oxidation of secondary alcohols is'32, 1 3 3
-d[MnOJ/dr

k[R,CHOH][MnO,][OH-]

and in basic solution the stoichiometry for benzhydrol oxidation is'34


(C6H,),CHOH+2 MnO; + 2 OH- = (C6H,),C0+2 Mn0;-+2 H,O
The kinetic parameters are E = 6.3 kcal.mole-' and AS' = -38.4 eu, and at
25 "C the reaction exhibits a primary kinetic isotope effect of 6.6. When "0labelled MnO; was employed, no labelled oxygen appeared in the benzophenone.
The r n e ~ h a n i s r n 'involves
~~
abstraction of hydrogen, either as a hydride ion or a
hydrogen atom, from the anion of the alcohol
RzCHOH
R,CHO-

+ R,CHO- + H +

HMnOz+MnO; < R,C-O+


R2C-O- + HMnO,

A final decision between these alternatives has yet to be made.


The oxidation by basic permanganate of phenyltrifluoromethyl carbinols' 34* 134a,
which has the same rate law as benzydrol, is characterised by very large primary
kinetic isotope effects (cf. Cr(V1) p. 307).
Substituent
knlb
kHIkT

H
16.0
57.1 & 2.4

p-CH3
16.1

rn-Br
16.2

OXIDATION BY

Cr(V1) A N D Mn(VI1)

309

That reaction rates are barely affected by the nuclear substitution ill-accords with
the hydride abstraction mechanism. However, nuclear substitution affects both
the ionisation of the alcohol as well as the oxidation step. The value of kH/kT
agrees well with that of 55.5 calculated from k,/k, by means of the Swain rel a t i ~ n s h i p , 'i.e.
~ ~ ~( k ~ / k , ) " ~ ~=' kH/kT.
A few results have been reported on the oxidation of cyclohexanol by acidic
permanganateZ4*13'. In the absence of added fluoride ions the reaction is firstorder in both alcohol and oxidant',', the apparent first-order rate coefficient
(for excess alcohol) at 25 "C following an acidity dependence k = 3.5+ 16.0
[H,O+]sec-'. k,/k, depends on acidity (3.2 in dilute acid, 2.4 in 1 M acid) and
kDzO/kH20
is 1.74. Addition of fluoridez4 permitted observation of the reaction
for longer periods (before precipitation) and under these conditions methanol is
attacked at about the same rates as di-isopropyl ether, although dioxan is oxidised
over twenty times more slowly. The lack of specificity and the isotope effect indicates
that a hydride-ion abstraction mechanism operates under these conditions. (The
reactivity of di-isopropyl ether towards two-equivalent oxidants is illustrated by
its reaction with Hg(II).) Similar results were obtained with buffered permanganate.
An analogous study668of the oxidation of benzyl alcohol by permanganate in
aqueous perchloric acid yielded a rate law
-d[MnO,-]/dt

k, [MnO4-][C6H5CHZOH][H3O+
]

with k, = 0.32 12.mole-'.sec-' at 35 "C (p = 1.92 M ) . Mnzf and F- ions are


without effect on the rate and E = 11.4 kcal.mole-', AS* = -27.2 eu. Hydrideion transfer to HMn04 is proposed as the mechanism of reaction.
The MnOi oxidation of di- and triarylcarbinols in moderately strong sulphuric
acid (H, < -0.50) is first-order in alcohol and zero-order in oxidant; plots of
log k versus Ho give good straight lines of slope approximately - 1 for both series of
substrate^'^^'. k,/k, for oxidation of (C6H5),CDOH is only 1.08 and the Hammett p values are - 1.02 for secondary alcohols and - 1.39 for tertiary alcohols.
These results indicate rate-determining carbonium-ion formation, viz.
Ar'CHOH

+ H+ + ArzCH++ H,O
2.3.4 Aldehydes

Aldehydes exist in aqueous solution in one or both of the forms

RCH=O HZO

+ RCH(OH),

(31)

and in any example the solute may act either as a gem-diol or a carbonyl compound.
References pp. 493-SO9

310

COVALENT COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

The chromic acid oxidation of benzaldehyde to benzoic acid in aqueous solution


was examined by Graham and Westheimer"', and in acetic acid solution by
Wiberg and
With water as solvent the kinetics are
-d[Cr(Vl)]/dt

k[RCHO] [HCrO, I{ [ H 3 0 + ]+k'[H30+ I*}

and in acetic acid solution


-d[Cr(VI)]/dt

k[RCHO][HCrO,]h,

For the latter reaction E = 13.2 kcal.mole-' and ASt = -28 eu.
Both groups reported a sharp decrease (40-65%) in the rate on addition of
Mn(I1) ions. Wiberg and
reported a deuterium isotope effect of 4.3
(30 "C) and that electron-withdrawing groups facilitate reaction ( p = 1.02),
the latter result confirming the data of L ~ c c h i ' ~ ' , ( p = + 1.06). The opposite
effect of electron-withdrawing groups for aliphatic aldehydes has been reported
by ROC^^'^^ (p* = - 1.2), who explains the difference in terms of the hydration
pre-equilibrium (3 1 ) which favours the gem-diol form for aliphatic aldehydes and
the carbonyl form for aromatic aldehydes. Otherwise the oxidation of aliphatic
aldehydes strictly parallels that of benzaldehyde as regards kinetics'40* 14'
(formaldehyde), Arrhenius parameters (f~rmaldehyde'~', acetaldehyde, nbutyraldehyde, i~obutyraldehyde'~~),
primary kinetic and solvent isotope effects
(f~rmaldehyde'~',a ~ e t a l d e h y d e ' ~ ~and
) , retardation by Mn(I1) with induced
oxidation (f~rmaldehyde'~').
Accordingly, the most reasonable mechanism is the exact analogue of that
proposed for alcohol oxidation

''*

RCHO + H'

HCr 04-

?H
R-C-OCr03H
I

Although the fate of Cr(1V) is uncertain, (cf. the alcohol oxidation), some characteristics of the intermediate chromium species have been obtained by Wiberg and
Richards01-1'~~
from a study of competitions between benzaldehyde and each of
several substituted benzaldehydes. The competition between the two aldehydes
for Cr(V1) is measured simply by their separate reactivities; that for the Cr(V) or
Cr(IV) is obtained from estimation of residual aldehyde by a ''C-labelling
technique. If Cr(V) is involved thenp values for oxidation by Cr(V1) and Cr(V) are
0.77 and 0.45, respectively. An isotope effect of 4.1 for oxidation of benzaldehyde
by Cr(V) was obtained likewise.
Wiberg and L e p ~ e have
' ~ ~ examined the oxidation of benzaldehyde by chromyl

OXIDATIONBY

31 1

Cr(V1) A N D Mn(VI1)

acetate in an acetic anhydride medium. The orange colour of the oxidant changes
to dark brown on addition of the aldehyde. Analysis showed that each chromium
atom retains one equivalent of oxidising power and the spectrum rules out the
possibility of a Cr(V1): Cr(1II) complex. Isotope dilution analysis indicated that
almost 25 % of the oxidant was consumed by the solvent under some conditions
and this effect was minimised by employing excess aldehyde. The oxidation
proceeds in two stages, both steps displaying a total kinetic order of two. Some
substituted benzaldehydes show detailed differences of order in one step or the
other. The p values for the two steps were estimated to be -0.2 and -0.9
respectively, and both display a moderate primary kinetic isotope effect. "0 is
transferred from labelled oxidant to the aldehyde and an ESR spectrum obtained
from the reaction mixture was assigned to Cr(V). The authors prefer a oneequivalent oxidation mechanism to give an acyl radical (corresponding to the
first step); the resulting Cr(V) effects a similar one-equivalent oxidation (corresponding to the second step), viz.
RCHO +(CH,CO),CrO,

-+

Rc-O+ Cr(V)

+
RCHO + Cr(V) + R e 0 + Cr(IV)

R e 0 +(CH,CO),CrO,

RcO+Cr(V)

-+

R e 0 Cr(V)

+ R6O+Cr(IV)

The permanganate oxidation of aromatic aldehydes has been studied by several


The production of benzoic acid from benzaldehyde necessitates
groups'46*.'41
the use of buffers in all save the most basic solution (pH 12). Wiberg and Stewart146 confined their measurements to a p H range of 5 to 13; at pH < 5 an
autocatalytic reaction appeared. Under these conditions the end product is Mn(IV)
or Mn(V1) (depending on pH) and the reaction is first-order in both aldehyde and
oxidant. The acidity dependence is complex; the reaction showing a small negative
p value (-0.248) in neutral solution and a large positive value ( 1.83) in an
alkaline medium. It is clear that two paths exist, one predominating at about pH 7
and the other at p H 1 1 . The primary kinetic isotope effect is much larger at pH
6.5 than at pH 11-12. Again, 75 % oxygen transfer (determined with ' * O )occurs
from labelled MnO; to aldehyde at pH 5.5 but only about 26 % at pH 12.4.
Rate parameters were obtained at p H 6.5 ( E = 10.3 kcal.mole-' and A S f =
- 26.2 eu for benzaldehyde). Detailed examination of the acidity dependence for
the two paths indicated that in acidic solution general acid catalysis is operative,
but that in alkaline solution catalysis is specific to hydroxide ion, the order with
respect to OH- being one-halft. A comparison with the manganate-ion oxidation
of benzaldehydes was made and it was found that Mn(V1) attacks a series of

t This result has been questioned recently706.


References p p . 493-509

3 12

co v A L E N T c OM P ou N D S,

M E T A L I o N s : ox I D A T I ON-R E D u CTIo N

substituted benzaldehydes at similar rates as alkaline MnO; and an identical p


value is obtained.
Two mechanisms are proposed by the authors. For the acidic-neutral solutions

+ MnO, + H 2 0 +

ArCHO

OH 0
I
I1
Ar-C-0-Mn-0

OH
I

II

+OH-

OH
I

Ar-C-0-MnO,
I

+ B: + Ar-C=O+

Mn(V)O;

HB

H
3MnO;+H20

2Mn02+MnO;+20H-

H,O may be replaced by any acid, HA, and a cyclic mechanism for the breakdown of the ester is quite feasible. For oxidation in alkali the fractional order in
hydroxide ion, the low k,lk, and low degree of oxygen-transfer from oxidant are
taken as symptomatic of a free-radical chain reaction of the type
MnO;

+ OH-

MnOt-

+ OH.

OH-+RCHO --+ R-CH(0H)-0.


R-CH(OH)-O.+MnO;

-+

RCO,H+MnO;

+ OH*

with termination by dimerisation of OH.. This particular scheme does not fit the
observed kinetics and it should operate for any organic substrate readily oxidised
by hydroxyl radical, of which there are a large number. The kinetics are, however,
unique to aromatic aldehydes.
Investigation of the oxidation of aliphatic aldehydes32 has been confined to
gem-diols which behave as secondary alcohols, being most easily oxidised in the
anionic forms, e.g.
CF3CH(OH) and CF3CH-OI
I

0-

0-

although in very strong acid the permanganic acid itself effects rapid oxidation.
All these routes for fluoral hydrate display a substantial primary kinetic isotope
effect and Arrhenius parameters have been obtained (Table 3).
As has frequently been the case, the evidence at present available does not
permit discrimination between the possibilities of hydride ion- or hydrogen atomabstraction by Mn(VI1).

O X I D A T I O NBY

Cr(V1) A N D Mn(VI1)

313

TABLE 3
K I N E T I C ISOTOPE EFFECTS AND

A R R H E N I UPSA R A M E T E R S F O R T H E
Mn(VI1) A N D Mn(V1)
~E (kcal. mole-')
kH/kD

OXIDATION OF

FLUORAL HYDRATE BY
~~

Reaction

~~

MnOa2--dianion
MnO,--dianion
Mn0,--monoanion

H Mn04-neutral
aldehyde

pH 10.2
DH 7.6

ASt(eu)

_____~

4.8
5.1
10.1
13.6

1.2
13.0

-31.4
-18.4

6.3

13.5

- 17.8

2.3.5 Phenols

In a perchlorate medium the oxidation of hydroquinone to p-benzoquinone


by chromic acid obeys the rate e x p r e ~ s i o n ' ~ ~
-d[Cr(VI)]/dt = k[Cr(VI)][hydr~quinone][H,O+]'.~~ [quinoneIo
however, a power series of the type

a+b[H,O+]

+ c[H3OfI2

also represents the acidity dependence adequately. At pH 2 the second-order


M ) . The activation energies
coefficient is 53.4+_1.2 I.mole-'.sec-' (p = 5 x
for the terms involving b and c are 4.6 1.O and 5.2 f0.9 kcal.mole- respectively.
Use of a stopped-flow apparatus and high reactant concentrations in an effort to
identify an ester intermediate was unsuccessful.
The permanganate oxidation of phenols is complicated by the intervention of
lower oxidation states of manganese, (cf. the oxidation of toluene, p. 298). For
example, the oxidation of 2,6-dinitrophen01'~~~
5 0 in weakly acidic solution
displays an induction period, following second-order kinetics thereafter. However,
addition of potassium fluoride inhibits reaction almost completely, but manganous
ions strongly accelerate it.
Free-radical intermediates in the oxidation of a series of quinol phosphates by
permanganate at pH 11.7 have been characterised by an ESR rapid-mixing technique"'. Clearly a one-equivalent oxidation step to give a semiquinone phosphate
radical operates in this case.

',

'

2.3.6 Ketones
The chromic acid oxidation of cyclohexanone to adipic acid via 2-hydroxycyclohexanone and cyclohe~ane-l,2-dione~~~
is third order' 53, viz.
-d[Cr(V1)l
dt
References pp. 493-509

k[ketone][HCrO,][H,O+]

3 14

c o v A L E N T c o M P o u N D S,

M E T A L I O N S: O X I D A T I o N-R E D u c T I o N

where k = 1.05 x
12.mole-2.sec-' at 50 "C (p = 0.646 M ) . The rate of
oxidation is considerably lower than that of enolisation at the same temperature
and the question arises at to whether oxidation proceeds via attack of Cr(V1)
upon the ketone or the enol, and the observation of a primary kinetic isotope
effect of 4.0 (50 "C) on changing the substrate to 2,2,6,6-tetradeuterocyclohexanone does not clarify matters, for the enolisation itself is subject to an isotope
effect of the same order of magnitudelS4. However, the reaction proceeds four
times faster in heavy than in light water and this increase, which is considerably
larger than the factor of 2.0 to 2.5 for proton-catalysed reactions, was taken
as evidence by the authors that enolisation must be taking place as a necessary
pre-requisite for reaction, for this equilibrium is favoured by a factor of 2.0-2.5
on changing to heavy water. The mechanism therefore becomes
(last)

+H'+HCrO~

(last)

O - croz- OH

-rf r0, -% H

+ CrO,

H20

+ OH-

(slow)

(fast)

RoCek and Rieh1lS5 have pointed out that the enolisation scheme leads to
kinetics

k [ ketone] [Cr( VI)]


k , k[Cr(VI)]

- d[Cr( VI)] - k ,
-

dt

This forecasts that should k[Cr(VI)] considerably exceed k , then a change of


reaction order with respect to Cr(V1) from one to zero is to be expected. This
they observed for isobutyrophenone in 99 % acetic acid and 2-chlorocyclohexanone in water at moderate concentrations of oxidant.
Permanganate attacks ketones at both high and low pH. Alkaline permanganate
oxidises acetonelS6,the rate law being
-d[MnO,]/dt

k,,,[acetone][OH-][MnOJ

where kobs= 107 12.mole-2.sec-' at 25 "C. These kinetics suggest the mechanism
acetone + OHenolate + MnO;

+ enolate + H,O
ki

k-

4 product + Mn(V)

,2 Mn( VI)
Mn(V) + Mn(VI1) "

OXIDATION B Y

Cr(VI)

315

Mn(VI1)

AND

A steady-state approximation for enolate ion yields


- d[MnO,]

-.

- 2 k , k,[MnO,][acetone][OH-]
-

k - *[H20] k[MnOi]

dt

Inclusion of an oxidation of neutral enol is incompatible with the observed rate


law.
Oxidation of [2-'4C]acetone produced the following yields of products identified
by isotope dilution analysis
4%

acetol

pyruvic acid 27 %

pyruvaldehyde 0 %

lactic acid

6%

19 %

acetic acid

47 %

oxalic acid

These figures suggest an oxidation sequence


,CHJC02H
CH3C CH,-CH3CCH20H-CH,C

CHO - C H 3 C 0

CO,H

'(CO2H),

/OHCH3C H (0H)CO;

Rate laws and coefficients were determined for the oxidation of all intermediate
compounds by MnO, and were compatible with the scheme as presented, i.e.
including a route for direct oxidation to pyruvaldehyde. An estimate of k2
(- 5 x lo7 I.mole-'.sec-') suggests an initial electron-transfer to give CH3
COCH,. . This can then be rapidly oxidised in two ways

0
II

CH,CCH,.

0
II

+ Mn(V1) -+

+ CH,CCH:

CH,COCH,OH

CH3C-CH-OMn( VI)O;
I

H
OH-

0
11

CH,C-CHO

+ H 2 0+ Mn(1V)

A few kinetic measurements on the acid permanganate oxidation of cyclohexanone


have been reported by LittlerIJ7.The reaction is first order in ketone but the order
in oxidant was not clearly established.
References pp. 493-SO9

316

COVALENT

C O M P O U N DMSE, T A L

IONS: O X I D A T I O N - R E D U C T I O N

2.3.7 Monocarboxylic acids

Formic acid is rather atypical and should be considered separately. The chromic
acid oxidation has been examined in some detail by Kemp and water^'^', following the earlier work of St~ethlage'~'.At acidities of up to 3.6 M (HC104) the
reaction follows kinetics
-d[Cr(V1)l
dt

- k[formic a~ i d l [ H C r O J[ H , 0 + ] ~

with E = 11.6 kcal.mole-' and AS* = -38 eu. However, at an ionic strength
of 6.0 M (CIO;), the [ H 3 0 + l 2term reverts t o h,. Several effects associated also
with the Cr(V1) oxidation of alcohols were noted, including a primary kinetic
isotope effect of 7.15 at 25 "C, a solvent isotope effect of 5.74 and a retardation
of rate by Mn(lI), the latter in a sulphate-ion medium only. It is probable, therefore
that a Westheimer-type mechanism operates through a n intermediate formylchromic anhydride, HCOOCrO,H, which undergoes protonation before breaking
down to C 0 2 and Cr(IV), cf. reactions (28)-(30).
The permanganate oxidation of formic acid has attracted much attention. The
reaction is pH-independent above pH 5 and involves formate ion. At lower pH's
the rate is much lower until permanganic acids begins to be formed at very low
pHls9.
Above pH 5 the reaction presumably involves two anions and, indeed, a positive
salt effect is found'60*16' and the rate expression is

164,
Alargeprimary kineticisotopeeffect has been reported by several
the most comprehensive study having been made by Bell and Onwood160in an
attempt to assess the role of quantum-mechanical tunnelling. This was found to
be unimportant and the rate coefficients at infinite dilution were expressed as

k,
k,

=
=

(4.72k0.60) x 10' exp(- 11,849~75/R~)l.mole-'.sec-~


(3.94k0.06) x 10' exp( - 13,056+ 13/RT)l.mole-'.sec-'.

A small solvent isotope effect was found by Bell and OnwoodL6' (kH20/kD20=
1.08) in contradiction to that of only 0.38 reported by Taylor and H a 1 ~ e t - n ' ~ ~ .
Over one-third of the oxygen present in the carbonate originated from the oxidant
when "0-labelled permanganate was used'63. The reaction is subject to pronounced catalysis by ferric ions'63.
The available data do not permit a conclusive mechanism to be formulated.
C-H cleavage is clearly rate-determining but no choice between a one- or two-

OXIDATIONBY

Cr(V1) A N D Mn(VI1)

317

equivalent process can yet be made. A simple hydride-ion or hydrogen-atom


transfer is not compatible with the observed '*O exchange.
The reaction between permangante ion and neutral formic acid follows similar
bimolecular kinetics'63 with k, = 1.1 x lo9 exp(- 16.4 x 1O3/RT)l.rnole-'.sec-'.
No primary kinetic isotope effect was found for this path either in light or heavy
water. However, Mocek and Stewart'" have reported that in very strong sulphuric
acid the oxidations of neutral substrate by both HMnO, and MnO: display
substantial isotope effects.
The oxidation of formate bound to cobalt(II1) ("captive formate") has been
examined by Candlin and Halpern". The kinetics are
-d[complex]jdt

k[(NH,),Co.02CH2+][Mn0,]

and the stoichiometry follows two paths

++MnO;

(NH,),Co*O2CHZ+

= CO,

+ Coz+++MnO,

(NH3),Co.0zCH2+++MnO; = CO, +(NH,),Co.OH:+

++MnO,

C-deuteration of the captive formate reduces the rate by a factor of 10.5 without
affecting the stoichiometry. The ratio of unreduced to reduced cobalt was found
to depend on oxidant concentration

The authors propose that an initial abstraction of a hydrogen atom produces


a complex containing bound carbon dioxide radical-anion. This complex could
break down in two ways

[ ( N H , ) , C O . ( - O ~ C H ) ] ~ + + M ~O,, [(NH,),CO(III)(CO;)]~++HM~O~

''

co2

[(NH3),C~(III)(CO;)]2+
MnOT

+ co2

[(NH,),CO(~~I)~H~]~++C~~

The ten-fold reduction in rate on deuteration without changing the stoichiometry


provides strong evidence that both sets of products originate in a common step.
The Arrhenius parameters for this reaction, ( E = 13.3 kcal.mole-', AS* = - 17
eu) invite comparison with those for the oxidation of free formate ion ( E = 12.4
kcal.mole-', AS* = -15 eu).
The oxidation of other monocarboxylic acids by both Cr(V1) and Mn(VI1) is
slow. Mare5 and R ~ t e k ' ~examined
'
the effect of CO,H groups on the oxidation
rates of methine and methylene groups. With a series of dicarboxylic acids
References p p . 493-SO9

3 18

c o vA LEN T

COM P O U N D S. M E T A L I O N S : O X I DA T I ON-R E D u C T I o N

H02C(CH2),C02H it was found that beyond a certain point the increment in the
rate coefficient for an increases of one i n n is constant at 5.2 x l o p 3 I.mole- '.set- ',
which compares with a value of 5.73 x
I.mole-'.sec-' for n-paraffin oxidation
(p. 293). It is clear that carboxylic acids behave as paraffins except that a slight
retardation due to the inductive effect of -C02H is apparent. Permanganate
behaves in much the same way and some examples of carboxylic acid oxidation
have been cited in the section on hydrocarbons.
2.3.8 Amities
A comprehensive account of the oxidation of benzylamine by both neutral and
alkaline permanganate has been published recently'66. The neutral amine is the
active reductant at all pH values and at pH 9.9 good second-order kinetics were
obtained. Under these conditions the stoichiometry is

3 C6H5CH2NH2+2MnO,

=:

3 C6H5CH=NH+2M n 0 2 + 2 OH-

+ 2 H2O
The pH-dependence of the rate is complicated. The rate increases very rapidly
between pH values of 8 and 10 and the authors correlate this with the literature
value of pKBH+ of benzylamine of 9.34. The rate levels off at pH 11-12, but at
pH 13-14 a reaction the rate of which is linearly dependent on hydroxyl ion concentration occurs.
In the pH range 8-10 the reaction shows a deuterium kinetic isotope effect of
7.0 for oxidation of C6H5CD2NH2.C6H5CH2ND2(in D,O) is oxidised at the
same rate as the light compound. A series of ring-substituted analogues were
examined and the data afforded a p + value of -0.28. Activation parameters were
determined for a series of benzylamines and all fell in a range of 11.0 to 12.3
kcal.mole-' (E) and -21.6 to -24.8 eu (AS'). These large negative values are
of interest as they resemble those found for permanganate oxidation of both
neutral molecules (H2, HC02H) and of anions (alkoxides). For this path the
authors propose a transition state formed thus

C6H5CH2NH2 + Mn0,-

----C

This could decompose in one or both of two ways

OXIDATION BY

Cr(VI) A N D Mn(VI1)

319

The present information does not allow discrimination between these two routes.
The high-pH path probably involves concerted attack of OH- and MnO, upon
neutral amine because the possibility of attack of oxidant upon C6H5CH2NHor C6H,CHNH2 is too remote in view of known acidities of amines.
In the same paper the authors list the responses of a number of amines and
amides towards permanganate and record the effect of freezing the medium upon
the oxidation of benzylamine.
TABLE 4
R A T E D A T A FOR T H E P E R M A N G A N A T E O X I D A T I O N O F SOME A L K Y L A M I N E S

Temperature unspecified, probably about 20 "C.


Amine

k,(l.mole- '.see- ')

P Ka

3.08
0.944
0.0828

10.65
10.98

10.63
9.92
10.155

3.36
1.82

Rosenblatt et aZ.167have examined the effect of structure and isotopic substitution


upon the permanganate oxidation of some alkylamines (Table 4). The isotope
effect of 1.84 is considered to be sufficiently low to be compatible with aminium
radical-cation formation, and it is felt that, while C-H cleavage is significant for
oxidation of primary amines, the dominant mode of oxidation of tertiary amines
is electron-transfer, e.g.

MnO, +N(CH,),
+
"(CH,),

H2C . . . N(CH,),

+ MnO,

+ -N(CH,), + MnOi+ H++H2C


- N(CH,),
* *

+ H2C=N(CH3I2+

(slow)
(fast)
(fast)

H,C-N(CH,),+ + H 2 0 + H2CO+HN(CH3),

2.3.9 Nitroalkanes
A stopped-flow examinati~n'~'' of the alkaline permanganate oxidation of
phenylnitromethane
3 C,H5CH-NO;+2

MnO; +H,O = 2 C6H5CH0+2 MnO,


+ 3 N 0 , + 2 OH-

indicates a rate law


-d[MnO,]/dt
References pp. 493-509

= k2[Mn04][C6H5CH=NO;]

320

co v A L E N T c o M PO u N D s,

M E T A L I o N s : ox I D A T I o N-RE D u c T I o N

which suggests a transition state

The kinetics and mechanisms of the MnO; oxidations of nitrocyclohexane and


nitrocyclopentane are entirely similar'67b. The combined rate data for solutions
at 0.5 M ionic strength are
kZ(274"K)(l.mole-'.set- I )

Substrate
Phenylnitromethane
Nitrocyclo hexane
Nitrocyclopentane

2.4

E(kcal.mole8.1 5 0 . 2
8.05
11.1

180&20
310*20
52&10

I)

ASi(eu)
-20*1
-27.8
-20.0

O X I D A T I O N O F P O L Y F U N C T I O N A L O R G A N I C MOLECULES

2.4.1 Glycols

Oxidation of glycols can proceed by two routes: (i) formation of an a-hydroxycarbonyl compound and (ii) carbon-carbon fission. In the case of chromic acid,
successive C-methylation of ethylene glycol increases the degree of cleavage16'
until, with pinacol, cleavage is at least 70 % q ~ a n t i t a t i v e ' ~Chatterji
~.
and Mukherjee169 - 1 7 1 have examined four glycols obtaining rate laws and Arrhenius
parameters. In all cases the oxidation rate depends on the first powers of the
glycol and acid chromate ion concentrations. The remaining data are summarised
in Table 5. Manganous ion strongly retards the oxidation of all the glycols except
pinacol. Despite its much larger activation energy, the oxidation of pinacol is
cu. 400 times faster than that of ethylene glyc01'~'. It is clear that an ordinary
TABLE 5
ACIDITY-DEPENDENCES

A N D RATE PARAMETERS FOR THE CHROMIC ACID OXIDATION

O F C(-GLYCOLS

Glycol
Ethylene glycol"
Propylene glycol
2.3-Butylene glycol
Pinacol

Dependence on [ H 3 0 + ] A (13mole-3.sec-1) E (kcal.mole-')


[H30+l+k[H30+lz
[H30+12

IH30+12
[H30+1

2.2 x i 0 5
8.1 x i 0 5
4.5 x 107
b1.07 x l o L 2

Arrhenius data refer to [ H 3 0 + l 2 dependence term.


Units 12.mole-2.sec-1.

10.7
10.4
12.2
17.2

Ref.
170
169, 171
169, 171
169, 171

OXIDATION BY

Cr(VI)

AND

Mn(VI1)

321

alcohol type oxidation mechanism operates almost exclusively for ethylene glycol,
but that increasing C-methylation favours the second (cleavage) mechanism. This
view is supported: (i) by the fitting of the result for ethylene glycol to the Taft
plot for oxidation of primary alcohols and ( i i ) by its proximity to the point for
2-methoxyethanol on this plot" '.
The oxidation of pinacol was studied in further depth by Chang and Westheinier'72. The total kinetic order of three was confirmed and the reaction was
found to proceed 2.7 times faster in heavy than light water. In contrast with
ethylene glycol, 0-methylation of one group reduces the oxidation rate by a factor
of the order of 1500, producing complex kinetics containing an essentially zeroorder term in chromic acid. The reaction is not sensitive to oxygen, but induces
the oxidation of Mn(I1) ion to MnO, with a factor just under 0.3. The oxidation
rate is reduced by Mn(I1) by a factor of about 2.5 at chromic acid concentrations
of 6 x
M . At high oxidant concentrations (0.145 M ) no retardation is found.
This probably accounts for the failure of Chatterji and Mukherjee to observe
retardation' '.
The solvent isotope effect suggests that no 0-H cleavage is involved in the slow
step and the effect of 0-methylation indicates that a cyclic complex is involved.
The induction factor is probably obscured by the reaction of Mn(I1I) and MnO,
with pinacol itself. The typical glycol-cleavage mechanism advocated for oxidations
by Pb(IV) and I(VI1) (p. 349) may well operate, uiz.
H 2 C r 0 4 + RC
,

R2C

R2C
R,

-OH

- OH

- 0,

C -' 0

R2C-0,
R2C

- 0'

Cr

+ 2H,O

(fast)

CrO,

(slow)

\O
R2C = 0

\O

R2C=0

Strong support for the cyclic ester intermediate comes from the measurement
of the relative rates of oxidation by chromic acid of cis- and truns-1,2-dimethyl1,2-~yclopentanediol~~~.
In water and in 90 % acetic acid kcis/k,,,,,is 17,000 and
800, respectively. Both oxidations are first order both in glycol and HCrO;.
The rate with the cis-isomer involves a hydrogen ion concentration dependence;
the reaction with the trans-isomer was examined in more acidic solutions and the
rate followed an h, dependence. The solvent isotope effect of kD20/kH20
of 2.0 is
also normal. Oxidation of some diastereoisomeric 1,2-diols has revealed lower
k(dl)/k(meso)ratios than expected on the basis of decomposition of a cyclic ester.
This has been i n t e r ~ r e t e d " in
~ terms of rate-determining ester formation (cf. the
hindered triterpene study of Westheimer et al.lo2,p. 302). A similar proposal
has been made by Kwart and B r e t ~ g e r " to
~ explain the relative rates of oxidation
of cyclohexyl and cyclopentyl pinacols.
References pp. 493-509

322

COVALENT COMPOUNDS, METAL IONS: OXIDATION-REDUCTION

2.4.2 Allylic alcohols


The equatorial allylic alcohol 3/l-hydroxyandrost-4-ene is oxidised by Cr(V1)
3 10 times faster than the saturated 3/l-hydroxy-5a-androstan-17-one,
5.7 times

faster than the axial 3a-hydroxyandrost-4-ene and 6.9 times faster than its 3deuterated a n a l ~ g u e " ~The
.
greater speed of oxidation of the equatorial isomer
is in contrast to the pattern observed for saturated alcohols and probably arises
from resonance between the double bond and the incipient carbonyl group.
Ally1 alcohols behave essentially as olefins towards MnO; (p. 300).

2.4.3 Ketols, keto-aldehydes and keto-acids


The rapid oxidations of certain of these polyfunctional compounds by alkaline
permanganate were examined' 56 as a supplement to the study of acetone oxidation
(p. 314). The oxidations of acetol and pyruvaldehyde show identical rate laws of
the form
-d[MnO,]/dt

k,[substrate][MnO;]

+ k,[substrate][MnO;][OH-]

with the following values of the rate coefficients at 25 "C

k,(I.moIe- '.set- ')


k,(12.mole-2.sec-')

acetol

pyruvaldehyde

50
1 . 2 103
~

5.6 x 10'
9.3 x 10,

Pyruvate ion is oxidised according to the kinetics


-d[MnO;]/dt

k,[substrate][MnO;][OH-]

+ k,[substrate][OH-]

with k, = 4.0 x 10, 12.mole-2.sec-' and k, = 0.9 I.mole-'.sec-'. These various


results were regarded by the authors'56 as of too preliminary a nature to justify any
mechanistic proposals.

2.4.4 Dicarboxylic acids


The examples of oxalic and malonic acids are atypical and are considered
separately. Other dicarboxylic acids react essentially as alkanes (p. 297).

OXIDATIONBY

Cr(V1) A N D Mn(VI1)

323

Early work of Dhar'77 established that oxidation of oxalic acid by chromic acid
occurs readily, but some of his kinetic data are unreliable as the substrate itself
acted as the source of hydrogen ions. The reaction is first-order in oxidant and is
subject to strong manganous ion catalysis (as opposed to the customary retardation), the catalysed reaction being zero-order in chromic acid. This observation
is related to those found in the manganous-ion catalysed oxidations of several
organic compounds discussed at the end of this section.
' ~ ~also Bakore and Jain178adetermined the rate law for
Chandra et ~ 1 . and
the uncatalysed reaction to be
-d[Cr(VI)]/dt = k,[HCrO; ][oxalic a ~ i d ] ' [ H , O + ] ~
with k, = 3.7 x lo7 e x p ( l 2 . 0 ~103/RT) I'.mole-'.sec-'.
tain a different rate law

Rao and A ~ y a r ob'~~

-d[Cr(VI)]/dt = k[Cr(VI)] [oxalic acid]"'


A third set of results, those of

-d[Cr(VI)]/dt

indicates two paths, viz.

( k , [ C , O 2 - ] + k2[C204'-]2)[H30+]3[HCrOq]

Of the various results, these appear to the reviewer to be the most meticulously
obtained and they do not fit the other two rate laws suggested. At 25 "C ( p = 1 .O M
perchlorate), k, = 5 . 1 0 ~
lo5 14.mole-4.sec-' and k, = 4.78 x 10" I ' . m ~ l e - ~ .
sec- The corresponding reaction intermediates are a neutral chelate monooxalato complex and a (non-chelate) bis-oxalato complex of Cr(V1).
The permanganate oxidation of oxalic acid has been studied exhaustively and
has been reviewed by Ladbury and C u l l i ~ ' ~It~is. characterised by an induction
period and a sigmoid dependence of rate upon time. Addition of manganous
ions eliminates the induction period and produces first-order decay kinetics"'. l E 2 .
Addition of fluoride ions, however, practically eliminates reaction"'.
It is clear that some slow reduction of permanganate occurs during the induction
period to give lower valency states of manganese which form oxalato complexes of
varying stability. Oxidative break-down of these gives manganous ions which
react rapidly with permanganate unless removed by complexing, e.g. with fluoride.
Oxidation of malonic acid by Cr(V1) has been noted briefly by SnethlagelE3
and later by Kemp and WaterslS4. The kinetics are simple second-order but the
acidity dependence is complex. Heckner et ~ 1 . ' ~find
~ ' the alkaline permanganate
oxidation of malonic acid (and also of 0- andp-toluic acids and of p-toluenesulphonic acid) to be retarded by added Mn(VI), uiz.

'.

-d[MnO;]
dt
References p p . 493-SO9

- k[MnO~][OH-][sub~trate]~.~-~.~
- ____

[M n 0,'

-1

324

c o v A L E N T c o M P o u N D S,

M E T A L I o N s : O X I D A T I o N-R E D u CT I O N

These authors invoke the pre-equilibrium


OH-

+ MnO,- + Mn04,-

+OH.

the validity of which rests on the value of no for the couple OH*/OH-. Latimer's
value'' of +2.0 V, compared with that of +0.56 V for the Mn(VII)/Mn(VI)
couple, would seem to exclude it but Stein's value of +1.25 V (pH 14)'34d increases its plausibility. However, the extreme sensitivity of alkaline permanganate
to trace metal
renders Heckner's conclusions suspect as no special precautions to remove these ions appear to have been taken.

2.4.5 Hydroxy-acids

Bakore and Narain"' obtained the following kinetics for the oxidations by
chromic acid of lactic, malic and mandelic acids
- d[Cr(VI)]/dt = k[HCr O i l [ hydroxy-acid] [H,

0'1

The rate of oxidation is reduced by one half on addition of manganous ions and
the following Arrhenius parameters were recorded

Lactic acid
Malic acid
Mandelic acid

A(l2.mole-*.sec-')

E(kcal.mole-

5.2 x 104
6.6 x lo4
2.2 x 104

9.04
8.96
7.90

Kemp and Waters'86 found a primary kinetic isotope effect of 8.7 for oxidation
of C-deuterated mandelic acid and noted a large difference in rate between the
oxidations of mandelic acid ( k , at 24.4 "C = 1.7 I.mole-'.sec-') and u-hydroxyisobutyric acid ( k , at 24.4 "C = 5 . 6 ~
I.mole-'.set-') - a difference not
reproduced for the oxidation of these compounds by the one-equivalent reagent,
manganic sulphate. The various data are fully in accord with a Westheimer-type
mechanism, viz.
RtH(OH)CO2H

HCrOi

+ H,O+

H,C,R
H02C'

R
C'
I,,,

r /

\\o

II
'0-Cr-OH

II
0

OH
I
Cr-OH
II

Decarboxylation of the keto-acid would then ensue.

2H20

(slow)

OXIDATION BY

Cr(V1) A N D Mn(VI1)

325

The oxidation of lactate ion by alkaline permanganate (0.1 M base) was


examined as part of the general study of the oxidation of acetone and its possible
oxidation product^"^. The reaction is first-order in lactate ion, and, if it is assumed
that it is also first-order in oxidant, then the second-order rate coefficient is 2
I.mole-.sec- at 25 C.
No published work exists on the kinetics of acid permanganate oxidation of
hydroxy-acids although Pink and Stewart have noted that benzilic acid is
oxidised by acid permanganate to benzophenone in an autocatalytic process
with k1,20/kDz0of unity.
In unpublished work88 Kemp has found that although the sigmoid character
of the [MnO,] uersus time plots obtained with acidic solutions makes kinetic
analysis difficult, several features are apparent from comparison of collected plots
(Figs. 1 and 2). Firstly, the slow or induction period is longer for mandelic than
for C-deuterated mandelic acid and even the fast stage involving lower valency
states of Mn is more prolonged for the deuterated acid. Secondly, although the
induction periods for oxidation of several a-hydroxyacids follow the sequence,
(CH,),C(OH)C02H > CH,(OH)CO,H > CH,CH(OH)CO,H > C6H,CH
(OH)CO,H the rates of the fast stages follow a different sequence from the
reverse of the above, viz. C6H,CH(OH)CO2H > (CH,),C(OH)C02H >
CH3CH(OH)CO2H > CH,(OH)CO,H.
Experience with both two- and oneequivalent reagents with respect to this series suggests that the induction stage

Fig. 1. Isotope effect for acid permanganate oxidation of mandelic acid. Temperature = 16.2 C;
~
M ; [H2S04] = 1.69 M .
[Mn04-] = 1 . 4 lo-

lo3 [C6H5CH(OH)C02H]
lo3 [C6H5CD(OH)C02H]
References pp. 493-SO9

10.4
9.9

1.3
6.9

5.2
4.9

326

C O V A L E N T COMPOUNDS, M E T A L IONS: OXIDATION-REDUCTION

Fig. 2. Relative rates of oxidation of some a-hydroxy acids by acid permangante. Temperature =
25.6 "C; [MnO,-] = 1 . 4 ~ 1 0 -M~; [HZS04]= 1.69 M . 0,C6H5(OH)C02H (0.0693 M ) ;
0,(CH3)2C(OH)COzH (0.0630 M ) ; X , CH3CH(OH)COzH (0.0704 M ) ; A, CHz(0H)COZH
(0.142 M ) .

probably involves two-equivalent oxidation of the hydrogen atom bound to the


hydroxylated carbon atom (to give the primary kinetic isotope effect and the
sluggish oxidation of a-hydroxyisobutyric acid) and that the second stage involves
one-equivalent attack by Mn(II1) to give a rate sequence identical with that for
direct attack by manganic sulphate'86 (Fig. 4, p. 394).

2.4.6 Boronic acids

The oxidation of tert-butylboronic acid by chromic acid has the stoichiometry


CrO,

+ t-C,H,B(OH),

"Ao t-C,H,OH+

B(OH),

+ Cr3+

and follows simple second-order kinetic^'^

- d[Cr(VI)]/dt

= k[HCrO~][t-C,H,B(OH)z]

The acidity dependence is complex and indicates that no extra proton to give
H,CrO, is required, but that H,CrOf is an active oxidant in this reaction. The
rate is very sensitive to the nature of the alkyl group, uiz.
R

CH3

CZH5

t-C,H,

lo7 k2(l.mole-'.sec-') at 30 "C in 0.114 M HCIO,

2.4

6600

750,000

OXIDATION BY

327

Cr(V1) A N D Mn(VI1)

The authors propose a mechanism

H C r O i t RB(OH)2 F

B (OH),

+O
; H + C r (P)

2.4.7 Furfurals
Examination of the permanganate oxidation of furfural anc 5-substituted
furfurals, at pH 11.5-13.3 in a stopped-flow apparatus, viz.

indicates a rate law7

- d [MnO,-]/dt

k , [furfural] [MnO,-]

+ k , [furfural][MnO,-

][OH-]

The second term dominates and is characterised by p = +1.30 (using a-meta


constants), E = 10.8 kcal.mole-, AS* = -22.8 eu and k,/kD > 1.8. 0 experiments show the major source of oxygen in the acid produced is the solvent, which
suggests the hydrate anion is the reactive form, viz.

OH

OH

2.5

SOME METAL-ION CATALYSED REACTIONS O F C H R O M I C A C I D

D h a ~ - noted
~
that the oxidation of oxalic acid by chromic acid is markedly
accelerated on adding manganous ions, the reaction order in Cr(V1) changing
from one to zero. Bobtelsky and Glasner4 found the oxidation of bromide ions
by chromic acid in aqueous sulphuric acid to follow kinetics
-d[Cr(VI)]/dt
References p p . 493-509

k[Mn(II)][Cr(VI)]f [H,O+]

328

C 0 V A L E N T C 0 M PO U N D S, M E T A L I O N S: OX1 D A TI ON-R E D U C T I 0 N

i.e. to be independent of the concentration of bromide (p. 282). Their results were
confirmed by Kemp and Watersxs4using aqueous sulphuric acid (2.4 M ) and are
readily rationalised in terms of the scheme

+ Cr(V)+ Mn(II1)
Cr(V)+ Mn(I1) + Cr(IV)+Mn(III)
Cr(IV)+ Mn(I1) + Cr(III)+ Mn(1II)
Cr(VI)+3 Mn(I1) + Cr(II1)+3 Mn(1II)
Cr(V1) + Mn(I1)

Mn(III)+Br2 Br.

--$

Mn(II)+Br.

Br2

(35)

The slow step is (32), which excludes any participation by bromide in determining the overall rate.
Kemp and WaterslS4 also found the oxidations of cyclohexanone and of
mandelic, malonic and a-hydroxyisobutyric acids by Cr(V1) to be Mn(I1)catalysed. In these cases, as with oxalic acid, the [Cr(VI)] versus time plots are
almost linear and the reaction becomes first order in substrate (or involves
Michaelis-Menten kinetics), and, except at lowest catalyst concentrations, approximately first order in [Mn(II)]. Detailed examination of the initial rate of
oxidation of a-hydroxyrobutyric acid as a function of oxidant concentration
revealed, however, that the dependence is
-d[Cr(VI)]/dt

= k[substrate][Mn(II)][HCrO~]*

This can be explained in terms of the pre-equilibria (32)-(34) followed by a slow


oxidation of the substrate by Mn(II1). It is probable that the substrates are chelated
to Mn(I1) and Mn(II1) throughout the process. The rate of oxidation of the substrate is given by
rate = k[Mn(III)] [substrate]
However, [Mn(IrI)] is related to [Cr(VI)] by equilibrium (35), and
[Mn(IIl)] = Kf [HCrO; ]* [Mn(II)]/ [Cr(III)]*
hence

- d[Cr(VI)]/dt

kKf [substrate] [Mn(II)][HCrO,]*/[Cr(III)]*

In confirmation of the suggestion that the active reagent of the Cr(VL)-Mn(I1)


couple is Mn(1II) it was foundls6 that, while the rates of oxidation of mandelic

Pb, T1, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

SPECIES

329

and a-hydroxyisobutyric acids by Cr(V1) are in the ratio 300 : 1, the ratios for
oxidation by the couple and by Mn(II1) are, respectively, 2.9 and 2.3. Furthermore,
the values of k,/k, for oxidation of C-deuteromandelic acid by Cr(VI), Mn(Il1)
and Cr(V1)-Mn(II) respectively are 8.7, 1.1 and approximately 1 .O.
More detailed results on the less strongly-catalysed oxidation of mandelic
acid186a shows that here the catalytic effect depends on the initial [Mn(II)]/
[Cr(VI)] ratio; at ratios less than lo2 Mn(I1) retards reaction. At high ratios the
catalysed reaction is not retarded by a hundred-fold excess of Cr(II1) ions, which is
evidence against the multiple equilibria (32)-(34), and for this substrate the reaction order in substrate is fractional. Step (33) is discounted in a modified reaction
scheme which attributes part of the oxidation to attack by Cr(IV) and Mn(II1)
upon the substrate, the Cr(1V) being formed during initial attack on substrate by
Cr(V1). At low [Mn(II)], the more strongly oxidising Cr(IV) is replaced by
Mn(II1) via reaction (34); at high [Mn(II)], Cr(V1) is reduced by Mn(1I) as well
as by mandelic acid.
Preliminary results have been reported'86b of oxidation of cyclobutanol by
the Cr(V1)-V(IV) couple to 4-hydroxybutyraldehyde. This proceeds at the same
rate as the oxidation of V(IV) by Cr(VI)la6" and cannot involve attack of Cr(V)
upon the alcohol, for this oxidation state is formed in a rapid pre-equilibrium:
but rather attack by Cr(IV), uiz.
Cr(VI)+V(N) f: Cr(V)+V(V)

Cr (V) V( IV)
Cr(IV)+C,H,OH
R.+V(V)

(fast)

+ Cr (IV) +V(V)

(slow)

Cr(III)+R.

(fast)

HOCH,CH,CH,CHO+V(IV)

(fast)

+
--f

The oxidation by chromic acid alone leads to a mixture of cyclobutanone and 4hydroxybutyraldehyde; the existence of an isotope effect for the oxidation of
1-deuteriocyclohexanol suggests that Cr(V1) produces the ketone and lower
oxidation states of chromium produce the cleavage product.

3. Oxidation by Pb(IV), TI(III), Hg(II), Hg(I), Bi(V), Au(III), Pt(IV), Pd(II),


Rh(ZII), Ru(II1) and Mo(V1)
3.1

GENERAL FEATURES

The first three members of this series appear at the bottom of the B subgroups
of the periodic groups 4, 3 and 2. They exhibit the so-called "inert-pair'' effect and
normally assume oxidation states of +4, + 2; + 3, + 1 and + 2,O respectively, i.e.
differing by two units. The species Hg', TI2+ and Pb3+ are of high energy and
References pp. 493-SO9

330

co v A L E N T C O M P O U N D S ,

M E T A L I O N S : O X I DA T I O N-RED U C T I O N

appear only as short-lived intermediates, e.g. in the oxidation of TI' by OH.


during pulse radiolysis' 9 0 . The other oxidants are also mainly two-equivalent in
their action.
All these oxidants form 7r-complexesby accepting electrons from olefinic bonds,
a property which has been widely discussed' 91. Oxidations by these species are
not, however, restricted to olefinic compounds and there is considerable evidence
that they are not totally restricted to two-equivalent action. Kinetic data on their
oxidations, once rare, has become profuse in the last decade both for aqueous and
non-aqueous media.
The redox potentials are given in Table 6 .
TABLE 6
OXIDATION POTENTIALS OF SOME TWO-EQUIVALENT REAGENTSI9

Reaction

Pb4+ $ 2
TI3++ 2
2 Hg2++ 2
Hgz2++ 2
Au3++ 2
PtC162-+2
PdZC+ 2

PbZ+

(Yolr)

=2

ca. +1.7
iI.25
+0.920
+0.789

cu. +1.41

e- =

eeeee-

no

= TI+
= HgzZ+

Hg
Auf
= PtC142-+2 CI-

e- = Pd

+0.68
$0.987

Pb(1V) is most usually employed as the tetraacetate and the action of this
compound is complex in that it can function either as a two-equivalent oxidant
giving Pb(I1) or as a source of acetoxy radicals, uiz.
Pb(O.COCH,),

Pb(O.COCH,),*+ CH,C02.

It can also act as a source of CH2C0,H radicals (vide infra).


A recent authoritative review on the numerous reactions which have been
studied without determination of kinetics is available' 9 2 and discussion here will
be restricted to the relatively few reactions which have been examined kinetically.

3.2

O X I D A T I O N OF INORGANIC SPECIES

Although Pb(IV) is sufficiently strong an oxidant to oxidise halides, no kinetic


data are available. Complexes of Pt(IV) and Au(II1) oxidise iodide and thiocyanate
ions but the other oxidants are weaker and form stable halo-complexes. However,
some simple molecules such as hypophosphorous acid, carbon monoxide and
molecular hydrogen are oxidised by the weaker members.

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

33 1

SPECIES

3.2.1 Halide and pseudohalide ions


The Pt(1V) oxidation of iodide ion was studied193 using bis-triethylarsine and
bis-triethylphosphine complexes to prevent interference from substitution reactions. The observation of isosbestic point: with trans- [Pt(AsEt,),CI,] and trans[Pt(AsEt,),Br,] indicates a single step reaction and quantitative production of
1; per mole of oxidant was recorded. Simple second-order kinetics were obtained
with these oxidants and also with trans-[Pt(PEt,),Cl,] and rate data are summarised in Table 7. The stoichiometry is
trans-[Pt(AsEt,),C1,]

+5 I-

trans-[Pt(AsEt,)I,] + 4 C1-

+ i;

TABLE I
RATE D A T A FOR T H E O X I D A T I O N O F IODIDE ION BY COMPLEXES O F

Complex

k2

(I.niole-'.sec-')

trans- [ P ~ ( A S E ~ ~ ) ~ C I ~
(0.5
] 10.03) x
trans- [Pt(AsEt&Br4]
1.56k0.04
trans- [Pt(PEt&CI4]
(3.5 &0.15) x

Temperature
("C)

E
(kcalmole- ')

35
25
35

20.1 f 0 . 6
10.6k0.4
16.2k0.4

pt(lV)'93
A.Si(eu)
- 1 1 k3
-25 + 2
- 19.5+2

The authors acknowledge that the present data are insufficient to allow a detailed
mechanism to be given but they reject the possibility of slow S,2 substitution
followed by a rapid redox process.
The oxidation of potassium thiocyanate by AuBr, is first-order in oxidant and
the pH dependence indicates that it is also first-order in thiocyanate ion, which is
oxidised much faster than HCNS19,. The activation parameters are E = 6.4k0.4
kcal.mole-' and AS' = 26+2 eu.
Bi(V) in aqueous perchloric acid is very strongly oxidising but kinetic studies
have been confined to a few stopped-flow measurements on oxidation of iodide,
bromide and chloride ions6 '. The appearance of Bi(II1)-halide complexes was
first-order with respect to Bi(II1) and in all cases the first-order rate coefficient, k,,
was the same, i.e. 161+8 sec-l at 25 "C ([H,Of] = 0.5 M , p = 2 . 0 M ) , irrespective of the nature or concentration of the halide. A preliminary attack on
solvent is compatible with these interesting results, viz.
Bi(V) + H,O
Bi(IV)+XOH*+X-

$ Bi(IV)

+H + +OH.

Bi(III)+X.

OH-+X.

2x. + x,
These authors also cite unpublished work on Sb(V) oxidation of I - which follows
a rate law involving the term [Sb(V)][1-l2.
References pp. 493-509

332

c o v A L E N T c o M P o u N D S,

M E T A L I o N s : o x I D A T I ON-R E D u c T I o N

3.2.2 Oxy-acids of sulphur

The work with iodide (preceding sub-section) was extended to t h i ~ s u l p h a t e ' ~ ~ ,


and isosbestic points and second-order kinetics were again obtained with the
various Pt(IV) complexes (Table 8). Two SZO,,- ions are consumed per mole
of Pt(IV) reduced, suggesting tetrathionate to be the product of oxidation, viz.
trans-[Pt(AsEt,),CI,]

+ 2 S,O:-

+ S402- + 2 C1-

trans-[Pt(AsEt,),CI,]

TABLE 8
R A T E D A T A F O R T H E O X I D A T I O N O F T H I O S U L P H A T E ION B Y C O M P L E X E S O F

Complex

k2 (l.mole-'.sec-')

Temperature ("C) E (kcalmole- ')

r r ~ n s - [ P t ( A s E t ~ ) ~ C l ~ ] 0.86&0.03
(2.6 f 0 . 1 ) ~ 103
tr~ns-[Pt(AsEt~)~Br,]
2.06&0.08
trans- [Pt(PEt3)2C14]

35
25
35

15.5+0.4
9.8+0.5
14.5+0.4

The rate of oxidation of sulphite ion by tetrabromoaurate'


and analysis of the kinetic data leads to the rate law

- d[AuBr,]/dt

94

pt(IV)'93
ASt(eir)
--10.5&2
-12 *3
-12 +2

is pH-dependent

k,[HSO;][AuBr,]

i.e. SO, and H,S03 are unreactive. At 16 "C k, is ( 4 . 0 f . 0 . 4 ) ~lo4 ].mole-'.


sec-' a n d E i s 19.4+ 1 kcal.mole-'.

3.2.3 Hydrazine

Oxidation by Mo(V1) as Moo4'- at low pH (1.2 to 3.2) gives N, stoichiometrically, viz.


4 Mo(VI)+N2H5+ = N 2 + 2 [Mo(V)I2+5 H +
The kinetics are194a

k; is fractionally acid-inverse, which probably reflects equilibria involving the


oxidising ion. E = 14.3 kcal.mole-' and ASt = -8.4 eu at pH 1.60, p = 0.22 M .

Di-imide was detected as an intermediate both mass spectrometrically and by


trapping with cis-1 ,2-cyclohexane dicarboxylic acid. The products indicate purely

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

333

SPECIES

two-equivalent behaviour, uiz. the steps

+ Mo(V1) + N,H, + 3 H + + Mo(1V)


H + + 2 N,H, -,N 2 + N , H 5 +

N,H,+

Mo(VI)+Mo(IV) + 2 Mo(V)

+ [Mo(V)],

(slow)
(fast)
(fast)

The catalysis by Mo(V1) of the oxidation of N,H,+ to Nz by methylene blue


depends on the steps given above, Mo(V1) being regenerated by methylene blue
oxidation of the Mo(V)
The latter reaction was studied independently
and
-d[methylene blue]/dt

k[Mo(V),]

~
sec-' and E = 23.4k0.2 kcal.mole-'. These
where k (25 "C) = ( 8 . 6 k 0 . 4 ) lo-,
figures are very similar to those reported for oxidation of Mo(V) dimer by iodine
and oxygen (pp. 468 and 450) and suggest a rate-determining first-order dissociation of the dimer to yield a reactive, monomeric Mo(V).

3.2.4 Nitrite ion

propose that the substitution of nitrite ion into chloroamEllison et


mineplatinum(1V) complexes involves oxidation of some nitrite by Pt(1V) to give
Pt(I1) which then functions as a catalyst. Prior addition of Pt(1I) removes an induction period and produces a new rate law
rate = k,[Pt(IV)][Pt(II)][NO;]
Substitution into tr~ns-[Pt(en),Cl,]~+catalysed by [Pt(en),]"
diamine) would proceed as follows
Pt(en):+ +NO;
Pt(en),Cl:++Pt(en),NO:

P Pt(en),NO:
lllow
fast

[CIPt(en),
slow

ethylene-

(fast)

. . . C1.. . Pt(en),NOz13+

11 fast

ClPt(en):
Pt(en),Cl+

(en

+ CIPt(en),NO:+

+ Pt(en)2,++CI-

(fast)

At 50 "C ( p = 0.22 M ) with these reactants k3 equals 10.1 12.mole-2.sec-'.


Redox behaviour of this type is considered to influence many substitution
reactions of Pt(IV)'94.
References p p . 493-SO9

334

COVALENT C O M P O U N D S , METAL I O N S : OXIDATION-REDUCTION

3.2.5 Hypophosphorous acid

Mitchell' 96 has shown that mercuric chloride oxidises this substrate to phosphorous acid

2 HgClz H3P02 + H,O = HgzClz+ 2 HCl+ H3P03


The reaction shows a first-order dependence on substrate concentration but,
except at very low concentration, is zero-order with respect to oxidant; moreover,
the zero-order rate coefficient is the same as that observed with oxidations by
iodine, cupric chloride and silver nitrate. The reaction is acid-catalysed. The
oxidation is completely analogous to the halogenation of ketones and involves a
slow tautomeric equilibrium followed by rapid oxidation, viz.

H
I

slow

O=P-OH

+ HO-P-OH %
fast

P(OH),

+ M("-')+

In contrast, thallic perchlorate oxidation follows a rate law

suggesting breakdown of a complex formed in a pre-equilibrium step669.Addition


of chloride ion increased the rate by a factor of up to ten, presumably as a result
of complexing. For the pre-equilibrium, AH = 7.8 kcal.mole-', AS = 31 eu and
for the redox breakdown of the complex, E = 13.7 kcal.mole-', A S f = - 14.2 eu.

3.2.6 Carbon monoxide


Tl(II1) and Hg(1) are inert to this substrate but Hg(I1) oxidises it to CO? at
moderate temperatures' "
2 H g 2 + + C O + H z 0 = Hg:'+CO2+2

H+

The rate law is


-d[CO]/dt = kz[CO][Hg2+]
with E

15.2 kcal.mole-' and AS' = - 13 eu.

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

335

SPECIES

Methanolic mercuric acetate absorbs carbon monoxide to give a stable derivative.


A C O . H ~ C O . O C H ,i.e.
~ ~ CO
~ , has inserted into the Hg-OCH3 bond. This implies
the possibility of a similar insertion between Hg and a water ligand, uiz.
[H20-Hg-OHz]+

+ CO -+

[H,O-Hg-CO-OH]+
Hg+Hg+

[HzO-Hg-CO-OH]+

+H +

(Slow)

-+

Hg+ COZ H +

(fast)

-+

Hg;+

(fast)

Pd(II) in the form of K,PdBr, oxidises CO to CO, in aqueous solution199


PdBr:-+CO+H,O

Pd+COz+2Hf+4Br-

The appearance of COz in the gas phase before precipitation of metal implies the
production of a soluble Pd(0) complex such as [Pd(CO)zBr,]2-. This receives
support from the observation that three times as much CO is consumed as CO,
liberated before deposition of Pd. The following reactions may be involved
PdBri[PdBr,CO]-

+ CO + [PdBr,CO]- + Br+ CO + [PdBr,(CO)2]0 + Br-

[Pd(II)Brz(CO)z]o+2HzO 5 Pd(0) complex + Pd(0)+2 C 0 2 + 4 H +


+ 2 BrThe authors believe the Pd(0) complex to be dimeric, containing Pd(0) and Pd(1I).
At 5 C, k3 is of the order of 1.0 I.mole-.sec-, but it is influenced in a curious
manner by addition of bromide ion, being increased initially and then sharply
decreased. The effect itself is very temperature-dependent.
A few data have been reported on the anaIogous oxidation by PdCli- which
proceeds by a similar mechanismZo0.

3.2.7 Molecular hydrogen


Along with Cu, MnO, and Ag+, the two oxidation states of mercury are
reduced by molecular hydrogen, O . H alpernZo1*
2 0 3 considers that hydrogen
is oxidised by two general mechanisms corresponding to rate equations of the
type
Class I -d[oxidant]/dt = kz [oxidant][H,]
Class I1 -d[oxidant]/dt = k3 [oxidant][H,]
References pp. 493-509

336

c o v A L E N T c o M P o u N D S,

M E T A L I o N s : O X I D A T I o N-R E D u c T I o N

(Hg(1) should fall into Class I on this basis but it is regarded as a source of two
oxidising ions per molecule.)
The first kinetic class probably corresponds to heterolysis of the H-H bond,
the second class to homolysis. Hg(I1) falls, with Cu(II), Ag(1) and MnO,, into
Class I and Hg(I), with Ag(1) and the Ag(1)-MnO, couple, into Class 11. The
activation energies for Hg(1) and Hg(I1) are 20.4f0.6 and 18.1 f0.6 kcal.mole-l,
respectively, and the activation entropies are, respectively, - 10.2k2 and
- 12.2+2 eu. The rate-determining steps are believed to be
H g Z + + H 2-+ H g H + + H +

Hg;+

+ H,

2 HgH+

The configuration of the transition state for Class I reactions is seen aszo3

(L = any other ligand). Strong complexing of the mercury atom, e.g. by CI- or
ethylenediamine, reduces the rate by 2 or 3 powers of tenzo4, but this is in any
case a general feature of Hg(I1) oxidations.
PdCIi-, Rh(II1) and Ru(II1) act as homogeneous catalysts for reduction of
The kinetics of all three activation reactions
FeCI, by molecular
fall into Class I. The Arrhenius parameters are
A(I.mole-'.sec-

PdQ2Rh(II1)
Ru(II1)

l)

E(kcal.mole- ')

6.6 x 10"
2.3 x 1015
4.0 x 1014

20
25.2
23.8

(36) represents the mechanism. Ru(II1) is singular in not oxidising hydrogen in


the absence of FeC1,207. Oxidation by Ru(1V) proceeds via Ru(II1) catalysiszo7.

3.3

O X I D A T I O N O F M O N O F U N C T I O N A L O R G A N I C MOLECULES

3.3.1 Olefns
These ions, with the exception of Pb(IV), form complexes with olefins and this
process is a preliminary to oxidation when this occurs. A recent reviewZo8of the
action of Pd(I1) covers aspects of structure and bonding as well as kinetics, and
a similar but older, review exists for Hg(II)Z09.The oxidation of olefins by thallic

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

337

SPECIES

species was discovered only recently and no previous review has included discussion of it.
Acetoxylation at an allylic position2' is the typical reaction of lead tetraacetate
but no kinetic data are available. Product studies favour a heterolytic mechanism2' .
Henry has examined in detail the oxidation of several olefins both by thallic
perchlorate in an aqueous perchloric acid
2 1 3 and by thallic acetate in
aqueous acetic acid214. The reaction displays mixed stoichiometry

'

CR1R,=CR3R4+TI(III)+ H 2 0 = TI(I)+ CRlR2R3COR4+2 H +


CRiR,=CR,R, +Tl(III)+2 HZO = T1(1)+ HOCR1RzCR3R40H + 2 H +
the relative importance of the paths depending on the nature of the alkyl groups
R, etc. The kinetics in aqueous mineral acid are

-d[Tl(III)]/dt = k 2 [TI(III)][olefin]
k 2 increases with increase of acidity but this is a salt effect and the dependence is
really an inverse one with respect to the activity of water. The relative rates in
the two media (but not between the media) are
Ethylene

Propene

1 -Butene

cis-2-Butene

1
1

I67
153

162
157

58
60

Perchloric acid
Acetic acid

trans-2-Butene Isobutene
13.6
35

-200,000
-230,000

The product distribution is also largely unaffected by change of medium. The basic
kinetics in acetic acid are also unchanged but the effect of addition of acetate ions
indicates a linear relationship between k2 and [TI(OAc); ]/[TI(III)]. TI(0Ac); is
regarded as the most significant oxidising entity over a wide range of acetate
concentration although TI(OAc)'+ and TI3+ become important at low acetate
concentrations. The mechanism, which is basically the same for both sets of reaction conditions, is

H20

k2

2*TI-CRlR2-CR3R40H

+ H'

k-2

H20 TI' + CR+Xt4-CRJR40H + H *


' + T I -CR1R2-CR3R40H
TL* + CR, R2R3COR4

H'

k , is very large and (37) is unlikely to be rate-determining. (39) or (40) cannot be


References p p . 493-SO9

33 8

c o v A L E NT c o M P o u N D S,

M E T A L I o N s : O X I D A T I o N-R E D u c T I o N

the slow step because the acid-retardation implied by step (38) is not found. Reaction (38) is probably, therefore, the slow step. Several schemes for the breakdown
of the oxythallation adduct are considered to account for the shift of the group
R,, and it is probable that more than one mechanism operate^"^.
The reaction between Hg(1I) and olefins has been examined from several angles
and work prior to 1950 has been summarised by ChattZo9.Several types of complex
and product are formed, depending on the olefin, which involve no change in the
oxidation state of the mercury atom. Propenyl ethers have long been known to
produce the corresponding glycol plus metallic mercury but no kinetics are
available215.
It has been shown recently216that the 1 : 1 complexes between Hg(II) and a
number of olefins are themselves subject to allylic oxidation by Hg(I1) at 80 "C.
Typical stoichiometry is
CH3CH(OH)CHZ.Hg+ + 3 Hg2+ = CH,=CH.CHO-t2 Hgi+ + 3 H t
The kinetics, which involve examination of two species containing Hg(1I) are of
the form

+ d[acraldehyde]

- -

!d14

dt

k,[adduct][Hg2+]

dt

where Hgz+ refers to mercuric trifluoroacetate and k, = 4.1 x loL4exp(-28.8 x


103/RT)I.mole-'.set-'. Further alkylation of the olefin results in large increases
in rate; a -CH,R group is oxidized to -COR in preference to a methyl group;
the double bond always remains intact.
Oxidation of propene labelled with 13C(CH3-CH=*CH2) yields acraldehyde
(88 %), acetone (10 %) and propanal (2 % ) 2 1 7 . The labelled acraldehyde consists

of equal amounts of CH;=CH-CHO and CH,=CH-CHO. The scrambling of


the I3C is seen as a result of the formation of a symmetrical intermediate

This could be oxidised by a second Hg(I1) species in the slow step, which must involve two molecules containing Hg(I1).
The stoichiometry of the oxidation of ethylene by palladous chloride is
C2H,+PdC12+HzO

CH,CHO+Pd+2 HCI

Products from very many other olefins have been detailed by Smidt et aLZ1'.
All monoolefins with at least one hydrogen atom on each carbon atom of the

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

339

SPECIES

double bond yield the corresponding ketone, with the carbonyl group appearing
at the carbon atom initially subject to the higher inductive effect. Reaction in
acetic acid produces complex mixtures of compounds including glycol acetates.
Considerable kinetic and other data are available, and are summarised below
for the reaction in aqueous solution.
(i) The kinetics for the oxidation of ethylene, propene and the three butenes
are of the form219-223

where K , is the equilibrium constant for the reaction


PdCI,'-

+ RCH=CHR +

RCH
_f

PdCl,

1-

+ CI-

k' varies according to an Arrhenius expression with E = 20.4 kcal.mole-I and


A S f = -8.7 eu. K , is 17.4k0.4 at 25 "C with p = 1 M (CIO,-) and is pHindependent.
(ii) The absorption of olefin by the solution is initially very rapid and to an
extent greater than required for simple saturation. The degree of absorption
diminishes as the chloride ion concentration is increased but is unaffected by
change of pH219*220.
(iii) Oxidation of C2D4 produces only a secondary isotope effect219.
(iu) Oxidation of C2H4 in D 2 0 proceeds four times more slowly224than in
H 2 0 and no introduction of deuterium into the acetaldehyde occurs225.
( u ) The dependence of rate upon ionic strength (added NaCIO,) involves a
maximum rate at ,u = 0.4 M 2 ' 9 , 2 2 ' .
(ui) A dimeric species [PdC12C2H4],can be prepared in benzene andisolated226.
It is decomposed by water to acetaldehyde, palladium metal and hydrochloric
acid.
A mechanism accommodating these data has been proposed2"; the preequilibrium (41) is rapidly attained and is followed by

[PdCI,C2H4]- + H 2 0 + [PdC12(H20)C2H,] +Clk2

k-2

(K2 = k2/k-2)

[PdC12(H20)C2H4] H 2 0

.+ [PdC12(OH)C2H4]- + H 3 0 +

k-3

(K3

[PdC12(OH)C2H4]-

(42)

k3

+ CH20H-CH2-PdCl + C12 C H 3 C H0+ Pd(0) + HCI


k4

k,/k-,)
(slow)

(43)
(44)

k-4

CH20H-CH2-PdCl
References pp. 493-509

(45)

340

C O V A L E N T C O M P O U N D S , METAL IONS: O X I D A T I O N - R E D U C T I O N

This gives the rate expression

in agreement with experiment. The slow step (44) corresponds to the transformation of a n-complex into a a-complex, viz.
CI,
CL/

CH\
Pd

OH

H 0

CH2

\ Pd /

CI'

H2\
HO'

CH2

OH2

The breakdown of the a-complex is envisaged by Henryzi9as


H
H
I
I
H-C-C-0

ClPd

H
I
CH,-C=O

OH2

CIPd-

H30+

Cu(I1) compounds are frequently used in conjunction with Pd(I1) in the


oxidation of olefins in the Wacker process. Their role has been viewed as that of
catalyst for autoxidation of Pd metal back to Pd(I1). Dozono and ShibazZ7
report the rate of oxidation of ethylene by a PdCIz-CuCI, couple to be given by

with E = 10.4 kcal.mole-'. The agreement with the uncatalysed reactions and
the non-involvement of Cu(1I) concentration supports the catalytic role. The
situation may, however, be more complicated (vide infra).
The oxidation of ethylene by palladous acetate in acetic acid has been examined
by Moiseev et a l.z2 8 -2 3.0This reaction shows mixed stoichiometryZz8

C2H4 PdClz

+ 2 CH3CO;

= CH3COzCH-CHZ +Pd(O)

+CH3C02H
+2

C,H4+PdClz+2 CH3CO;

c1-

(CH3COz)zCHCH3+Pd(O)+2 C1-

The kinetics arezz9


rate

k,[Pd(II)][C,H4]

with E = 17.2 kcal.mole-' and AS* = - 10.7 eu. No incorporation of deuterium


into the ethylidene diacetate occurs when the reaction is carried out in
CH3COzDZ '.

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

341

SPECIES

One interesting difference between Pd(I1) in aqueous and acetic acid solutions
is that whilst oxidation of C2D4 by aqueous Pd(I1) displays no primary kinetic
isotope effect, the oxidation of CH,-CD-CH, by PdCl, in acetic acid-isooctane
affords aZ3' k,/k, value of 2.8. The products are a mixture of propenyl and
isopropenyl acetates in a 64 : 36 ratio unaffected by deuteration and with isotopic
retention. A hydride-ion shift in the slow step is proposed to account for the
isotope effect. This would be expected to result in a lower yield for labelled
isopropenyl acetate compared with propenyl acetate. Agui16208has commented
that Pd(I1)-catalysed isomerisation of the olefin could account for the isotopic
pattern of the products.
Moiseev et al., who proposed initially232that ethylidene diacetate was produced
from addition of acetic acid to vinyl acetate, later showed this to be impossible
from the result of reaction in CH3C02D,preferring the following mechanism230
+

CH,CO,H

[C~3Pd-CHz-CHzOzCCH3]2-+

H*

1 .

3 C I - + Pdo+CH3-CH-OzCCH3
C
, Hz=CH

-O,CCH,+

H+

(46)

C H3CH02CCH3
\CH,COZH
'(CH3C0z)z

CHCH, + H '

(47)

Under certain conditions a combination of Pd(I1) and Cu(I1) in acetic acid


oxidises olefins to saturated products which neither reagent produces alone233.
Although Cu(I1) continues to catalyse the production of vinyl acetate through
step (46) by a redox mechanism, the following new reaction can be effected
CH3C02CH2CH2CI
CH3C02CH2CH202CCH3+CuCI
CH3C02H+Hz0 CH,C0,CH2CH20H

PdCI,

C,H4+CuCl,+CH3CO;

Palladium metal is not produced in the new reaction and the substitution of a
twenty-fold excess of lithium chloride for cupric chloride prevented reaction;
kinetic data revealed first-order dependences upon both Pd(1I) and Cu(1I). The
distribution of products varied in an unpredictable way with reactant concentrations. The following mechanism was proposed by Henry233(X = C1- or CH,CO;)

C2H4 PdX2 CH3CO;

+ XPdCH2CH202CCH3 +X-

XPdCHZCH202CCH3 + HPdX
XPdCH2CH202CCH3 + 2 CUX,

+ CH2=CHO,CCH3

PdX, +XCH2CH,02CCH3 + 2 CUX

The nature of this final step is ill-defined.


Rhodium trichloride oxidises ethylene in dimethylacetamide solution to a
References pp. 493-509

342

cov A L E N T co M P O u N DS,

M E T A L I o N S: O X I D A T I o N-R E D u c T I O N

mixture of acetaldehyde, but-1-ene and but-2-ene. In general, 1 mole of C H 3 C H 0


is produced per mole of Rh(IZ1). Initial rate measurements indicate the
-d[C2H4]/dt

k[Rh(III)][C
H 4]
-2

'""

[a-I

The following scheme is consistent with the available results


Rh(III)CI,

ki

+ Rh(III)Cl,,k-

+Cl-

Rh(lII)CI,-,(C2H4)+H,0
Rh(I)CI,-

-, Rh(I)CI,- I

+CH3CHO+2 H +

+ C2H4 -, Rh(1) complex

(fast)
(fast)

The H 2 0 is considered to come from coordinated water in Rh(ZII)C1,.3 H 2 0 .


Assuming a steady state for Rh(ZI1) CI,-l yields
_
-d[C2H,I
_

dt

kl k2[Rh(II1)CInICC2H4I
k- I [cl-]fk2[C2H4]

The observed results fit this equation reasonably well, and at 80 "C k , z
sec-', k-,[Cl-]/k, z 0.025 mole.l-' with (for k , ) E = 10.1 kcal.mole-I and
AS' = -46 eu.
3.3.2 Arylcyclopropanes

Thallium(I[I) triacetate oxidatively cleaves phenylcyclopropane and its ringsubstituted analogues in anhydrous acetic
to yield mainly (> 90 %) the
corresponding I -aryl-1 , 3-diacetoxypropane together with a little of the cinnamyl
acetate. The reactions are of the first order in each component provided account
is taken of the double salt formation between TI(I1I) and TI(1) which renders an
additional TI(II1) molecule inactive following the reduction of each TI(II1) in a
two-equivalent process. For phenylcyclopropane E = 13.0k0.2 kcal.mole-' and
AS' = -29.2k0.6 eu. Electron-releasing groups facilitate reaction ( p = -4.3).
By analogy with Hg(I1) acetate which cleaves the ring to yield an organomercurial
compound, the proposed mechanism is
C 6 H 5 u

TI(OAC)~

C6H5

&

Tl(OAc),

fast

C 6 H 5 5 OAc + C 6 H 5 v O A C

+ T lOlc

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

343

SPECIES

using Hg(OAc), showed that for this reagent


An entirely analogous
reaction ceases at the organometallic compound unless the temperature is insec-').
creased to 135 "C (kdec = 8 . 6 ~
Lead tetraacetate reacts more slowly with arylcyclopropanes then either
thallic or mercuric acetates although the distribution of products is similar234c.
The reactions are of the first order in each reactant; for phenylcyclopropane E =
20.2k0.3 kcal.mole-' and AS' = - 17.8f 1.0 eu and a good correlation between
rate of cleavage and '
0 is obtained, giving p+ = - 1.75. Addition of perchloric
acids catalyses reaction through equilibrium formation of the more highly electrophilic Pb(OAc),+ C104- (K2980K = 23.7 I.mole-'); under these conditions p+
is - 1.3 which argues against an ion-pair description of the reactive species in the
tetraacetate oxidation.
3.3.3 Alcohols

Kinetic studies have been reported only for the oxidations by Hg(I1) and
TI(III), although the oxidation by Pb(IV) is receiving mechanistic study at the
present time.
1-Propanol reacts with lead tetraacetate in boiling benzene solution to give a
complex mixture of products including 1-propylacetate (35 %) and 1,l-dipropoxypropane (10 %)*". The intermediate RCH2CH2-0-Pb(OAc), is considered to
decompose both homolytically and heterolytically on the basis of observed
products, e.g. 2-propanol gives 2 % of isopropyl phenyl ether in benzene solution
and the reaction is subject to catalysis by pyridine. The production of substituted
tetrahydrofurans from long-chain alcohols236 has been reviewed2j7. Triphenyl
methanol gives high yields of hemiketal ester238*2 3 9 ; in general the course of
reaction is

e.g.

X = NO2

X=NOz,

Y=n

X=H.

Y = NO2

The migratory aptitude for p-nitrophenyl relative to phenyl is 4 . 4 k 0 . 3 which was


interpreted as indicating an exclusively homolytic mechanism. The following
chain scheme was proposed
Ar,COH+Pb(OAc),
Ar3C-0-Pb(OAc),
Ar,C-O.
Ar2cOAr Ar3CO-Pb(OAc),
Ar,COPb(OAc),
Ar2cOAr (AcO),Pb.

References pp. 493-SO9

+ Ar,C-0-Pb(OAc), + HOAc
+

Ar,C-O. +Pb(OAc),.

+ Ar,t]OAr
+ product

+ Ar3COPb(OAc),

-+ Pb(OAc),

+ product

(slow)

+Ar,CO*

+ Pb(OAc),

344

C O V A L E N T C O M P O U N D S , METAL I O N S : O X I D A TI O N - R ED U C TI O N

A brief study of the oxidations of cyclohexanol and cyclohexanol-1-d by


TI(II1) indicated the rate expression to be'35
-d [TI(III)]/dt = k [alcohol] [Tl(III)]f [ H 3 0 +]
The values of k,/kD and kD20/kH20are, respectively, 5.5 and 1.7. The acidity
dependence was explained in terms of the partial hydrolysis240
TI3++ 2 H 2 0 + TI(OH)2++ H 3 0 +
where TI3+ is the only significant oxidant. This does not, however, explain the
solvent isotope effect, for K (hydrolysis) is the same in H 2 0 / D 2 0mixtures as in
H 2 0 , and its origin may lie in differential solvation of the transition states.
Two studies have been performed by Littler24*1 3 5 on the oxidation of cyclohexanol by Hg(II), the second leading to more detailed and reliable data. The
reaction is first-order in both oxidant and substrate but the rate is independent
of acidity. E is 24.8 kcal.mole-', AS' is 1 eu, k,/k, is 3.0 and kD20/kH20is 1.30.
At 50 "C di-isopropyl ether is attacked at about one-half the rate of isopropanol,
which implies that hydride ion abstraction is occurring in both cases. This is
supported in the case of cyclohexanol by the isotope effects.
The oxidation of alcohols by Pd(II) and Rh(III) has been noted but no kinetic
data are yet a ~ a i I a b l e ~ ~ ' - ~ ~ ~ .

3.3.4 Hydroperoxides
Only tert-hydroperoxides have been examined kinetically and discussion will
be restricted to these. The stoichiometry of the lead tetraacetate oxidation is not
straightforward, but the main product is the corresponding t e r t - a l ~ o h oand
l ~ ~one
~
mole of tert-butyl hydroperoxide consumes just over two moles of oxidant245.
The kinetics with this substrate are245
-d[Pb(IV)]/dt

k3 [ P ~ ( I V ) ] [ ~ - B U O O H ] ~

'.

with an activation energy of 20.0k0.5 kcal.mole- Pb(I1) fails to retard reaction,


but added sodium acetate accelerates reaction to an extent directly proportional
to its concentration. Sodium acetate is very weakly dissociated in acetic acid and
this observation implies a second-order dependence on acetate ion, as in the autodecomposition of lead tetraacetate (p. 346). Addition of benzene and also of
ethanol increases the rate, the latter to a degree proportional to its concentration.
In pure methanol the reaction order in hydroperoxide changes to unity.

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo S P E C I E S

345

The kinetics are in keeping with the scheme


Pb( OAc),
Pb(0Ac);

+ OAc- + Pb( OAc),


+ OAc- + Pb(0Ac);-

Pb(0Ac);- +ROOH

$ Pb(OAc)i-.ROOH

Pb(0Ac); +ROOH

+ [Pb(OAc),.ROOH]

[Pb(OAc),.ROOH]-

+ROOH

-+

R0,.

Pb(OAc)~-.ROOH+ROOH -+ R 0 , .
Added alcohol can replace ROOH in its purely coordinative capacity.
A non-kinetic study of the oxidation of cumyl h y d r ~ p e r o x i d eby
~ ~Pb(1V)
~
to acetophenone and dimethylphenylcarbinol gives useful complementary data.

3.3.5 Formic acid


Although formic acid is oxidised to CO, by Pb(IV), kinetic data exist only for
the oxidations with Hg(II), Hg(I) and TI(II1) in aqueous mineral
and with Pd(I1) in acetic
Rate equations are of the form

Kinetics are obtained directly for Hg(II), Hg(1) and TI(III). With Pd(I1) a catalytic
system was
depending on the continuous reoxidation of Pd(0) by a
Cu(II)-02 couple. The inverse acidity dependence could result from a hydrolysis
of the type
TI3++H,O

+ TIOH'' + H +

where only T10H2+is strongly oxidising, but is more probably due to ionisation
of the substrate to the readily oxidised HCO; ;thus

- d[HCO,H]/dt

k,[HCO;][Ox]

The data are summarised in Table 9. At high concentrations of formic acid the
reaction becomes less than first-order in substrate248;this indicates the possibility
of complex-formation, but a medium effect may also be influential in the vicinity
of 1 M formic acid. Complex-formation affects the kinetics of the TI(IrI) oxidation
at all but the lowest reactant concentration^^^^.
References pp. 493-SO9

c o v A L E N T c o M P ou N D S,

346

M E T A L I o N S: OX I D A T I o N-R E D u CTI o N

TABLE 9
SUMMARY OF DATA FOR OXIDATION OF FORMATE 1 0 ~ 2 4 7 . 2 4 9 3

Oxidant

Temp. range ( " C )

E (kcal.mole-

l)

AS:(eu)

~~~~

Hg2
Hg2
~

1 +3

Pd(I1)

36-61
60-80
65-85
100-1 25

20.6
21.6
26.6
22.7

kHlkD
~

3
0
21
-6

3.4 (46.5 "C)


3.9 (65.1 "C)
3.4 (75.2 "C)
2.0(112 "C)

A two-equivalent oxidation mechanism is favoured for all oxidants247s249,


viz.

HCO;

K
+ M"+ +
HCO,.M'"-')+

H C O , . M ( ~ - ~ ) + H + + co , + M ( " - ~ ) +

(fast)
(slow)

In the case of TI(III) the overall rate coefficient has been resolved into a product
k K for the two steps247*249. The large positive AS' is due almost entirely to the
initial association, which was also studied spectroscopically. An alternative rate
determining step in the Pd(II) oxidation is hydride ion transfer to Pd(II)249a.
Oxidation of formic acid by mercuric chloride is the subject of several early
kinetic studies. D h a ~ - "showed
~
the reaction to be first-order in oxidant and substrate and to be subject to strong retardation by added chloride ions in agreement
with earlier work. The reaction is also subject to retardation by added acid and
presumably involves formate ion as the principal reactant.

3.3.6 Higher carboxylic acids

The auto-decomposition of lead tetraacetate in acetic acid, which normally


occurs at reflux temperature2", can be studied at 50 "C in the presence of sodium
acetate25 2 5 l a . The principal products of both the uncatalysed and catalysed decompositions are acetoxyacetic acid and carbon dioxide. The kinetic order of the
"normal" decay of Pb(IV) is complex and evidence was obtained that oxidation
of products is significant after the earliest stages. The evidence indicates that slow,
simple homolytic breakage of lead tetraacetate to give Pb(OAc),. and AcO.
does not occur but that the solvent plays an integral part, e . g .
' 9

Pb(OAc),+ HOAC + (AcO),P~.+HOAC+.CH~CO,H


The radical C H 2 C 0 2 His regarded as originating from the solvent. This primary
step is followed by a multitude of secondary reactions. An entirely analogous
scheme, producing (CH,),CH-O-~(CH,), , was proposed for decomposition in
~di-i~opropylether~~~.

Pb, TI, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

SPECIES

347

In the presence of acetate ions (1.0 M ) the rate of decomposition decreases as


the oxidant concentration is increased2* l a , although the reaction tends towards first-order behaviour both in Pb(1V) and CH3C02- as the oxidant concentration is lowered towards 2 x
MZ5la
. Under these simplyfying conditionsE = 12.2+0.2 kcal.mole- andAS* = -38+_0.7eu. N o r m a n a n d P o u ~ t i e ~ ~ ~
favour a non-radical mechanism, oiz.

AcO-Pb(OAC)p
\

6 7 ,/

ACO.CH,CO;

+ BH+

pb(oAc)z

CHz-CO

which is supported by the abilities of other bases to catalyse the decomposition.


Benson et ~ 1 . ~prefer a second-order dependence on acetate ion and produce
evidence that the complexity of the kinetic decay of Pb(IV) at higher concentrations stems from product-retardation. They write a radical mechanism with
Pb(OAc),- as the active oxidant. Norman and P o u ~ t i e ~propose
~ ~ a base-catalysed dimerisation in competition with the oxidation to account for the kinetic
deviation, and were able to isolate a new lead compound from reaction mixtures
exhibiting 4 equivalents of oxidising power per mole.

3.3.7 Ketones

Pb(IV), Tl(II1) and Hg(1I) attack ketones at rates approximating to the corresponding rate of enolisation
OH

11

-C-CHR-

+ -C=CR-

Lead tetraacetate attacks several ketones, the rates being those of e n o l i s a t i ~ n ~ ~ ~ ,


uiz.
rate

k[ketone][Pb(IV)]O

Clearly a two-equivalent oxidation of the enol is taking place. Product work has
been summarised by Criegee 53 and in general a-acetoxylation occurs.
Oxidation of cyclohexanone by thallic per~hlorate~has similar kinetics but
includes an acidity dependence of the form (at 25 C, p = 1.3 M )
k

k ( 4 . 8 ~I O - + l l . l

10-[H30+])

The rate of oxidation is, however, considerably slower than the rate of enolisation
as measured by iodination.
References pp. 493-SO9

348

co v A L E N T C O M P O U N D S , M E T A L I O N S:

O X I D A TI ON-RED

u C T IO N

Mercuric perchlorate has been shown to attack cyclohexanone, the reaction


being zero-order in the salt and the rate being that of enolisation with a primary
kinetic isotope effect of the same magnitude, to give a mercurated ketonelS4

-CH=C-

-CH-C-

Hg+

I1
0

distinguished by its ultraviolet spectrum (A,,


2260 A, E = 8.53 x lo3). Similar
data were reported for acetonelS4. These are not oxidation processes but are
necessary preliminaries.
The ultimate product of Hg(I1) oxidation is cyclohexane-1,2-dione which is
probably produced from 2-hydroxycyclohexanone. The oxidation by thallic
perchlorate produces cyclopentanecarboxylic acid in 75 % yield2s4;this contrasts
with the production of 2-acetoxycyclohexanone by thallic acetate2". The
mechanism favoured for the production of the carboxylic acid is

2-hydroxycyclohexanone does not produce the acid on treatment with thallic


perchlorate and the reaction with 2,2',6,6'-tetradeutero-~yclohexanoneproduces
a tri-deuterated acid. Other possibilities were eliminated.

3.3.8 Ethers
The ready oxidation of di-isopropyl ether by Hg(l1) p e r ~ h l o r a t eis~ a~ good
indication that this oxidant can function as a hydride-ion acceptor, uiz.
CH3

Hg2+ H-C-O-CH(CH,),

CH

HgH+ +

--f

Hg22++ H +

CH3
HgH'

+ Hg2+

3'C=(;-CH(CH3)2
CH3/

CH3\
+
,C-O-CH(CH,),
CH3/

+ H 2 0 --*

CH3\
'C=O+ (CH3)zCHOH
CH3/

(slow)

(fast)
(fast)

The kinetics are simple second-order. At 40 "C ( p = [H'] = 0.188 M ) k , is


5.1 x
I.mole-'.sec-' and E is found to be 16.6 kcal.mole-'. These data are,
of course, relevant to the problem of Hg(1I) oxidation of secondary alcohols.

Pb, T1, Hg, Bi, Au, Pt, Pd, Rh, Ru, Mo

3.4

SPECIES

349

O X I D A T I O N OF P O L Y F U N C T I O N A L MOLECULES

3.4.1 Glycols
Lead tetraacetate is an important reagent for glycol cleavage, which has the
stoichiometry
RZC-OH
1

RZC-OH

+ Pb(OAc),

2 RzC-O+ P b ( 0 A ~ ) z 2 HOAC

It can be used both in polar and non-polar solvents and complements the Malaprade oxidation with periodate, with which it has been compared and contrasted256.
Much has been published on the mechanism of its action, particularly on the
stereochemical aspects of its reactivity, and only a summary of the main features
can be presented.
The rate equation for the oxidation of ethylene glycol in acetic acid isz5'
-d[Pb(lV)]/dt

k,[Pb(IV)][glycol]

with k, = 1.95 x 1014 exp(-20.9 x 103/RT) I.mole-'sec-'. This has been confirmed for a series of glycols by Cordner and PausackerZ5*,who also showed that
electron-releasing substituents accelerate oxidation of benzpinacols while electronwithdrawing groups retard it.
Acidification with trichloracetic acid catalyses oxidationz59 , the fractional increase in the rate coefficient per mole of acid added, uiz. (Ak/k,)/[acid], being
of the order of two. Strong catalysis by alkali metal acetates has been observed for
several oxidations, e.g. of cis-cyclohexane-1,2-diolz60.formic acidz6', methyl
mannoside and galactoside261 and several a-hydroxycarboxylic acidsz6'.
Retardation by acetic acid itself, which reduces the oxidation rate of transcyclohexane-l:2-diol by three orders of magnitudez6', suggests the existence of
a pre-equilibrium
RZC-OH
I
+Pb(OAc),
RZC-OH

RzC-O-Pb(OAc),
I

RZC-OH

+ HOAc

Changing the reaction medium from acetic acid to water does not reduce the
efficiency of the oxidant263and, indeed, gradual dilution of an acetic acid medium
with water, methanol or benzene increases the rates of oxidation of several glycols
of factors of 500 to 1000 (ref. 260). This effect raises the question of whether the
catalysis by trichloracetic acid (uide supra) is solel