Reactions of coordination compounds share some characteristics with reactions of other

molecules, both organic and inorganic, so an understanding of coordination compound

reactions can draw on some familiar concepts. However, the chemistry of coordination
compounds has some additional features because the molecules have more complex
geometries and more possibilities for rearrangement, the metal atoms exhibit more vari-
ability in their reactions, and different factors influence the course of reactions.
Reactions of coordination complexes can be conveniently divided into substitu-
tion reactions at the metal center, oxidation-reduction reactions, and reactions of the
ligands that do not change the attachments to the metal center. Reactions that include
more elaborate rearrangements of ligand structures are more often observed in
organometallic compounds; description of these reactions is given in Chapter 14.

12-1 Synthesis of coordination compounds has always been a major part of chemistry.
HISTORY A N D Although the early chemists did not know the structures of the compounds they
PRlNClPLES worked with, they did learn how to make many of them and described them according
to the style of the time. The synthetic work done by Werner, Jorgensen, and others that
established the current picture of coordination geometry began the systematic
development of reactions for specific purposes. Many years of experimentation and
consideration of possible reaction pathways have led to the ideas described in this
chapter, and even now these ideas must be considered tentative and provisional in
many cases. The unification of reaction theory is still a goal of chemists, whether they
work with organic, inorganic, coordination, organometallic, polymeric, solid-state,
liquid, or gaseous compounds, but the goal is still far in the future. The discovery of
new reactions outruns the explanations, but correlation of these reactions with
theoretical explanations gradually extends our knowledge. Although the ability to
predict products and choose appropriate reaction conditions to obtain the desired
products is still a matter of art as well as science, the list of known reactions is now
long enough to provide considerable guidance.
 12-1 History and Principles 413
The goals of those studying reaction kinetics and mechanisms vary, but a major
underlying reason for such studies is to understand the electronic structure of the com-
pounds and their interactions. The information from these studies also allows more con-
trol of reactions and the design of reaction steps that may be useful for synthesis. A
by-product of synthetic and kinetic studies is the esthetic pleasure of seeing the colors
that are characteristic of many coordination compounds and how they change with
changes in ligands and metal ions.
We will first review some of the background needed to understand reaction mech-
anisms, then consider the major categories of such mechanisms, and finally describe
some of the results of these mechanistic studies.
In general, chemical reactions move from one energy minimum (the reactants)
through a higher energy structure (the transition state) to another energy minimum (the
products). In simple cases, the energies and bond distances can be shown as a three-
dimcnsional surface, with two different bond distances along the base-plane axes and
free energy as the vertical dimension. The reaction MX + Y --+ MY + X begins at
a point representing the short M-X distance of the bond to be broken and the longer
distance between the two reactants MX and Y. As the M-X bond breaks and the
M -Y bond forms, the reaction point moves to represent the short M -Y bond dis-
tance and the longer distance between the two products MY and X. The free energy sur-
face usually has a saddle shape, much like a mountain pass between two valleys. For
more complex reactions, such a visual representation is difficult or impossible, but the
path between the reactants and the products is always the lowest energy pathway and
must be the same regardless of the direction of the reaction. This is the principle of
microscopic reversibility, frequently described by the mountain pass analogy; the lowest
pass going in one direction must also be the lowest pass going in the opposite direction.
If the reaction is such that the conversion from reactants to products takes place
with no hesitation at the transition point as in Figure 12-1 (a), the structure at that state
is called the transition state. If there is a structure that lasts a bit longer as in
Figure 12-l(b), and particularly if it is detectable by some experimental means, it is
called an intermediate. Frequently, the kinetic equations include intermediates, even if
they remain undetected. Their presence allows treatment by a steady-state approxi-
mation, in which the concentration of the intermediate is assumed to be small and es-
sentially unchanging during much of the reaction. Details of this approach are
described later.

Extent of reaction Extent of reaction

(a) (b)

,: FIGURE 12-1 Energy Profiles and Intermediate Formation. (a) No intermediate. The activation energy
is the energy difference between the reactants and the transition state. (b) An intermediate is present at
: '
the small minimum at the top of the curve. The activation energy is measured at the maximum point of
the curve.
, .
414 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

A number of different parameters can be obtained from kinetics experiments.

First, the order of the reaction, indicated by the power of the reactant concentration in
the differential equation that describes it, can be determined, together with the rate
constant that describes the speed of the reaction. By studying a reaction at different
temperatures, the free energy of activation and the enthalpy (or heat) and entropy of
activation can be found. These allow further interpretation of the mechanism and the
energy surface. A somewhat more recent inclusion of pressure dependence provides the
volumebf activation, which offers insight into whether the transition state is larger or
smaller than the reactants.

12-2 12-2-1 INERT AND LABILE COMPOUNDS

SUBSTITUTION
REACTIONS Many synthetic reactions require substitution, or replacing one ligand by another; this is
particularly true when the starting material is in aqueous solution, where the metaI ion
is likely to be in the form [M(H~o),]~'. Some simpler reactions of this type produce
colored products that can be used to identify metal ions:

[ c u ( H ~ o ) ~ ] ~++4 NH3 F='-- [ c u ( N H ~ ) ~ ( H ~ o ) +

~ ]4~H
+20
blue much more intense blue

very pale violet red

These reactions, and others like them, are very fast and form species that can
undergo a variety of reactions that are also very fast. Addition of HNO~(H+),
NaCI(C1-), H ~ P o ~ ( P o ~ ~ KSCN(SCN-),
-), and NaF(F-) successively to a solution of
Fe(N03)3.9 H 2 0 shows this very clearly. The initial solution is yellow because of the
presence of [ F ~ ( H ~ O ) ~ ( O H )and
] ~ ' other "hydrolyzed" species containing both water
and hydroxide ion. Although the exact species formed in this series depend on solution . ,

-
concentrations, the products in the reactions given here are representative:

+ )H+
[F~(H~O)~(OH ]~~ [F~(H~O)~]~*

+]C1-
[F~(H~O)~ ~+
yellow

- colorless (very pale violet)

[ F ~ ( H ~ O ) ~ ( C+
~ )H] ~2 0+

[ F ~ ( H ~ O ) ~ ( C+ -
~ )~] ~0 +4 ~ -
yellow

Fe(H20j5(P04) + C1-
colorless
. .

Fe(H20)5(P04) + SCN- --+ [ F ~ ( H ~ O ) ~ ( S C N+) I~~ +0 4 ~ -

red

+ F-
[F~(H~o)~(scN)]~' -----+ [ F ~ ( H ~ O ) ~ ( Ft
) I SCN-
~+
colorless . .

Compounds such as these that react rapidly are called labile (15'-bil). In many
cases, exchange of one ligand for another can take place in the time of mixing the soh- ::
tions. ~ a u b e 'has suggested a reaction half-life (the time of disappearance of half the ;
initial compound) of one minute or less as the criterion for lability. Compounds that '-'
'H. Taube, Chern. Rev., 1952,50,69.
 12-2 Substitution Reactions 415

react more slowly are called inert or robust (a term used less often). An inert com-
pound is not inert in the usual sense that no reaction can take place; it is simply slower
to react. These kinetic terms must also be distinguished from the thermodynamic terms
stable and unstable. A species such as [ F ~ ( H ~ o ) ~ ( F is
) ]very
~ + stable (has a large equi-
librium constant for formation), but it is also labile. On the other hand, lexa-
aminecobalt(3+) is thermodynamically unstable in acid and can decompose to the
equilibrium mixture on the right

but it reacts very slowly (has a very high activation energy) and is therefore called inert
or robust. The possible confusion of terms is unfortunate, but no other terminology has
gained general acceptance. One possibility is to call the compounds substitutionally or
kinetically labile or inert, but these terms are not in general use at this time.
Werner studied cobalt(III), chromium(III), platinum(II), and platinum(1V) com-
pounds because they are inert and can be more readily characterized than labile com-
pounds. This tendency has continued, and much of the discussion in this chapter is
based on inert compounds because they can be more easily crystallized from solution
and their structures determined. Labile compounds have also been studied extensively,
but their study requires techniques capable of dealing with very short times (stopped
flow or relaxation methods, for example, temperature or pressure jump, nuclear mag-
netic resonance).
Although there are exceptions, general rules can be given for inert and labile elec-
tronic structures. Inert octahedral complexes are generally those with high ligand field
stabilization energies (described in Chapter lo), specifically those with d 3 or low-spin
d4 through d6 electronic structures. Complexes with d 8 configurations generally react
somewhat faster, but slower than the d7, d9, or d l 0 compounds. With strong-field lig-
ands, d 8 atoms form square-planar complexes, many of which are inert. Compounds
with any other electronic structures tend to be labile. Summarizing, we get:
Slow Reactions (Inert) Intermediate Fast Reactions (Labile)
d 3 , low-spin d4, d 5 , and d6 d l , d 2 , high-spin d4, d 5 , and d6
Strong-field d 8 (squareplanar) Weak-field d 8 d7, d9, d i O

12-2-2 MECHANISMS OF SUBSTITUTION

Langford and ~ r have

a described
~ ~ the range of possibilities for substitution reac-
tions, listed in Table [Link] one extreme, the departing ligand lcavcs and a discernible
intermediate with a lower coordination number is formed, a mechanism labeled D for
[Link] the other extreme, the incoming ligand adds to the complex and an in-
termediate with an increased coordination number (discernible either by kinetic or ana-
lytical methods) is formed in a mechanism labeled A for association. Between the two
extremes is interchange, I, in which the incoming ligand is presumed to assist in the re-
action but no detectable intermediates appear. When the degree of assistance is small
and the reaction is primarily dissociative, it is called dissociative interchange, I d .
When the incoming ligand begins forming a bond to the central atom before the depart-
ing ligand bond is weakened appreciably, it is called associative interchange, I,. Many
reactions are described by I, or Id mechanisms rather than by A or D when the kinetic
evidence points to association or dissociation but detection of intermediates is not pos-
sible. Langford and Gray call these categories the stoichiometric mechanisms;the dis-
tinction between activation processes that are associative and dissociative is called the

'c. H . Langford and H . B. Gray, Ligand Substitution Processes, W . A. Benjamin, Ne

416 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

'TABLE 12-1
, ,
Classification of ~ubstitution~echanisrns " . , . ,

Stoichiometric Mechanism
Dissociative Associative
Intimate 5-Coordinate Transition State 7-Coordinate Transition State
Mechanism for Octahedral Reactant for Octahedral Reactant
Dissociatiye activation D Id
Associative activation I,,
Alternative Labels
SN1lim SN2lim
(limiting first-order (limiting second-order
nucleophilic substitution) nucleophilic substitution)

intimate mechanism. The energy profiles for associative and dissociative reactions are
shown in Figure 12-2. The clear separation of these two mechanisms in the figure
should not be taken as an indication that the distinction is easily made. In many cases,
there is no clear-cut evidence to distinguish them, and inferences must be made by
using the available evidence.
Kinetic experiments are frequently carried out with large excess of the incoming
reagent, Y. This simplifies the analysis of the progress of the reaction for each kinetic
run, but requires a number of runs at different concentrations of Y to determine the
order of the reaction with respect to Y.

Extent of reaction
(a)

ZI
$
E
FIGURE 12-2 Energy Profiles 0

for Dissociative and Associative 2

Reactions. (a) Dissociative
mechanism. The intermediate has a
lower coordination number than the
starting material. (b) Associative
mechanism. The ihtermediate has
Extent of reaction
a higher coordination number than
the reactant. (b)
 12-3 Kinetic Consequences of Reaction Pathways 417

12-3 Although the kinetic rate law is helpful in determining the mechanism of a reaction, it
KINETIC does not always provide sufficient information. In cases of ambiguity, other evidence
CONSEQUENCES OF must be used to find the mechanism. This chapter will describe a number of examples
REACTION in which the rate law and other experimental evidence have been used to find the
PATHWAYS mechanism of a reaction. Our goal is to provide two related types of information: (1)
the type of information that is used to determine mechanisms, and (2) a selection of
specific reactions for which the mechanisms seem to be fairly completely determined.
The first is the more important, because it enables a chemist to examine data for other
reactions critically and to evaluate the proposed mechanisms. The second is also
helpful, because it provides part of the collection of knowledge that is required for
designing new syntheses. Each of the substitution mechanisms is described with its
required rate law.3

12-3-1 DISSOCIATION (D)

In a dissociative (D) reaction, loss of a ligand to form an intermediate with a lower

coordination number is followed by addition of a new ligand to the intermediate:
kl
MLSX MLS +X
k-1

The stationary-state (or steady-state) hypothesis assumes a very small concentra-

tion of the intermediate, M L S , and requires that the rates of formation and reaction of
the intermediate must be equal. This in turn requires that the rate of change of [ M L S ]be
zero during much of the reaction. Expressed as a rate equation,

Solving for [ M L S ] ,

and substituting into the rate law for formation of the product,

leads to the rate law:

n reactions of this chapter, X will indicate the ligand that is leaving a complex, Y the ligand
3 ~ the
that is entering, and L any ligands that are unchanged during the reaction. In cases of solvent exchange, all
(X, Y, and L) may be chemically the same species, but in the more general case they may all be different.
Charges will be omitted in the general case, but remember that any of the species may be ions. The general
examples will usually be 6-coordinate, hut other coordination numbers cotild he chosen and the discussion
would be similar.
(?'
418 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

One criterion for this mechanism is that the intermediate, ML5, be detectable dur-
ing the reaction. Direct detection at the low concentrations expected is a very difficult
experimental challenge, and there are very few clear-cut dissociative reactions. More
often, the evidence is indirect, but no intermediate has been found. Such reactions are
usually classified as following an interchange mechanism.

INTERCHANGE (I)
12-3-2-
In an interchange (Z) reaction, a rapid equilibrium between the incoming ligand and the
6-coordinate reactant forms an ion pair or Ioosely bonded moIecuIar combination. This
species, which is not described as having an increased coordination number and is not
directly detectable, then reacts to form the product and release the initial ligand.

When k2 << kPl, the reverse reaction of the first step is fast enough that this step is
independent of the second step, and the first step is an equilibrium with K1 = kl/k-l.
Applying the stationary-state hypothesis:

If [Y] is large compared with [ML5X] (a common experimental condition), the

concentration of the unstable transition species may be large enough to significantly
change the concentration of the ML5X, but not that of Y. For this reason, we must solve
for this species in terms of the total initial reactant concentrations of ML5X and Y,
which we will call [MIo and [Y]":

Assuming that the concentration of the final product, [ML5Y], is too small to i,
change the concentration of Y significantly, then

: ,<
From the stationary-state equation, %:
,. 'r

kl([MIo - [ML=jX.Y])[Y]o - k-l[MLSX.Y] - k2[ML5X.Y] = 0 '3

<2

The final rate equation then becomes

where kz/k-l is very small and can be omitted because k2 << k-l is required for the
first step to be an equilibrium.
 12-3 Kinetic Consequences of Reaction Pathways 419

K 1 can be measured experimentally in some cases and estimated theoretically in

others from calculation of the electrostatic energy of the interaction, with fair agree-
ment in cases in which both methods have been used.
Two variations on the interchange mechanism are Id (dissociative interchange)
and I, (associative interchange). The difference between them is in the degree of bond
formation in the first step of the mechanism. If bonding between the incoming ligand
and the metal is more important, it is an I, mechanism. If breaking the bond between
the leaving ligand and the metal is more important, it is an Id mechanism. The distinc-
tion between them is subtle. and careful experimental design is required to determine
which description fits a given reaction.
As can be seen from these equations, both D and I mechanisms have the same
mathematical form for their rate laws. (If both the numerator and the denominator of the
D rate law are divided by k-l/k-l, the equations have the similar forms shown here.)

k[Ml P I k[Mlo[Ylo
Rate = Rate =
[XI + k'[Yl 1 + k1[YIo

At low [Y], the denominator simplifies to [XI for the dissociative and to 1 for the
interchange equation. Both then are second order (first order in M and Y,
rate = k[MIo[Ylo or k[M]o[Y]o/[X]), with the rate of the dissociative reaction slowing
as more free X is formed.
At high [Y], a common condition in kinetic experiments, the second term in
+
the denominator is larger, [XI + kl[Y] = kr[Y] and 1 kf[Y]O = kl[Y]O, and [Y]
cancels, making the reaction first order in complex and zero order in Y
(rate = (k/k1)[MIo).
The change from one rate law to the other depends on the specific values of the
rate constants. The similarity of the rate laws limits their usefulness in determining the
mechanism and requires other means of distinguishing between different mechanisms.

12-3-3ASSOCIATION (A)
In an associative reaction, the first step, forming an intermediate with an increased
coordination number, is the rate-determining step. It is followed by a faster reaction in
which the leaving ligand is lost:

The same stationary-state approach used in the other rate laws results in the rate law

-4
4
4
This is a second-order equation regardless of the concentration of Y.

EXERCISE 12-1
Show that the preceding equation is the result of the stationary-state approach for an
420 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

As with the dissociative mechanism, there are very few clear examples of asso-
ciative mechanisms in which the intermediate is detectable. Most reactions fit better be-
tween the two extremes, following associative or dissociative interchange mechanisms.
The next section summarizes the evidence for the different mechanisms.

12-4 12-4-1 DISSOCIATION

EXPERIMENTAL
EVIDENCE I N Most substitution reactions of octahedral complexes are believed to be dissociative,
OCTAHEDRAL with the complex losing one ligand to become a 5-coordinate square pyramid in the
SUBSTITUTION transition state and the incoming ligand filling the vacant site to form the new octahe-
dral product. Theoretical justification for the inert and labile classifications of Section
12-2-1 comes from ligand field theory, with calculation of the change in LFSE between
the octahedral reactant and the presumed 5-coordinate transition state, either square-
pyramidal or trigonal-bipyramidal in shape. Table 12-2 gives the ligand field activa-
tion energy (LFAE), calculated as the difference between the LFSE of the square-
pyramidal transition state and the LFSE of the octahedral reactant. LFAEs calculated
for trigonal-bipyramidal transition states are generally the same or larger than those for
square-pyramidal transition states. These calculations provide estimates of the energy
necessary to form the transition state. When combined with the general change in en-
thalpies of formation described in Section 10-6, and particularly Figure 10-27, the acti-
vation energies of the square-pyramidal transition state match the experimental facts
(d3 and d 8 complexes are inert in both the strong- and weak-field cases, and d 6 strong-
field complexes are inert). Examination of these numbers shows that the activation en-
ergies of the square-pyramidal transition state match the experimental facts ( d 3 ,
low-spin d 4 through d 6 , and d 8 are inert). Therefore, the calculation of LFAE supports
a square-pyramidal geometry (and a dissociative mechanism) for the transition state.
However, all these numbers assume an idealized geometry not likely to be found in
practice, and the LFAE is only one factor that must be considered in any reaction.

, .
TABLE 12-2
ligand Field Activation 'Energies Calculzited by Angular Overlap ,

Strong Fields (units of e,) Weak Fields (units of e,)

LFSE LFSE LFSE LFSE

System Octahedral Square pyramidal LFAE Octahedral Square pyramidal LFAE
do - 12 -10 2 - 12 -10 2
dl -12 -10 2 - 12 -10 2
d2 - 12 -10 2 -12 -10 2
d3 - 12 -10 2 - 12 -10 2
d4 - 12 - 10 2 -9 -8 1
d5 -12 - 10 2 -6 -5 1
d6 -12 -10 2 -6 -5 1
d7 -9 -8 1 -6 -5 1
d8 -6 -5 1 -6 -5 1
d9 -3 -3 0 -3 -3 0
dl0 0 0 0 0 0 0

: a square-pyramidal transition state, LFAE

N ~ EFor = LFSE (sq. pyr.) - LFSE (oct.), for a
donor only.
 12-4 Experimental Evidence in Octahedral Substitution 421

FIGURE 12-3 Activation Energies

and Reaction Enthalpies. (a), (b),
Large E,, slow reaction. (c) Small
E,, fast reaction. (a), (b), AH < 0,
large equilibrium constant;
(c) AH > 0, small equilibrium
constant. In (b), the intermediate is
potentially detectable. (a)

Even for thermodynamically favorable reactions, a large activation energy means

that the reaction will be slow. For thermodynamically unfavorable reactions, even a fast
reaction (with small activation energy) would be unlikely to occur. The rate of reaction
depends on the activation energy, as in the Arrhenius equation

Some of the possible energy relationships for reactions are shown in Figure 12-3. In (a)
and (b), the reaction is exothermic (AH < 0), and the equilibrium constant is large
(entropy effects could be important, but are ignored for this discussion). In (a), the reac-
tion is spontaneous (AH < 0), but E, is large, so few molecules have enough energy
to get over the barrier and the reaction is slow. In (b), the reaction is spontaneous, with
an intermediate at the dip near the top of the activation energy curve. Intermediates of
this sort are frequently described, but can be detected and identified in only a few cases.
In (c), the reaction can go quickly because of the low activation energy, but has a small
equilibrium constant because the overall enthalpy change is positive.
When s and p orbital influences are added, the results are similar to those of the
thermodynamic case of enthalpy of hydration shown in Figure 10-7, with long half-
lives for d3 and ds and short half-lives for do, d4,d9, and as shown in Figure 12-4.

FIGURE 12-4 Stabilization

Energy and Experimental Half-Lives
for Water Exchange. (Angular
overlap data (solid line) from
J. K. Burdett, J. Chem. Soc. Dalton,
1976, 1725. Half-lives for water
exchange (dashed line) from
F, Basolo and R. G. Pearson,
200 1 I I I I 1 -10
Mechanisms of Inorganic Reactions,
Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn
2nd ed., John Wiley & Sons, 0 2 4 6 8 10
New York, 1967, p. 155.) Number of d electrons
422 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

Other metal ion factors that affect reaction rates of octahedral complexes include
the following (relative rates for ligand exchange are indicated by the inequalities):
1. Oxidation state of the central ion. Central ions with higher oxidation states have
slower ligand exchange rates.

2. Ionic radius. Smaller ions have slower exchange rates.

Both effects can be attributed to a higher electrostatic attraction between the

central atom and the attached ligands. A strong attraction between the two will slow the
reaction, because reaction is presumed to require dissociation of a ligand from the com-
plex. Figure 12-4 shows the half-lives for exchange of water molecules on aquated 2+
transition metal ions. All the ions in the figure are labile, with half-lives for the aqua
complexes shorter than 1 second; measurement of such fast reactions is done by indirect
methods, particularly relaxation methods4 (including temperature jump, pressure jump,
and NMR). The monovalent alkali metal cations have very short half-lives second
or less); of the common 2+ metal ions, only ~ e and ~ v2++ have half-lives as long as
0.01 second. ~ l has~ a half-life
+ approaching 1 second, and cr3+ has a half-life of 40
hours, the only inert aquated transition metal ion.
The evidence for dissociative mechanisms can be grouped as follow^:^, 6' 72

1. The rate of reaction changes only slightly with changes in the incoming ligand. In
many cases, aquation (substitution by water) and anation (substitution by an
anion) rates are comparable. If dissociation is the rate-determining reaction, the
entering group should have no effect at all on the reaction rate. Although there is
no specific criterion for this, changes in rate constant of less than a factor of 10
are generally considered to be insignificant for this purpose.
2. Decreasing negative charge or increasing positive charge on the reactant complex
decreases the rate of substitution. Larger electrostatic attraction between the pos-
itive metal ion and the negative ligand should slow the dissociation.
3. Steric crowding on the reactant complex increases the rate of ligand dissociation.
When ligands on the reactant are crowded, loss of one of the ligands is made eas-
ier. On the other hand, if the reaction has an A or I, mechanism, steric crowding
interferes with the incoming ligand and slows the reaction.

4 ~ Wilkinson,
. Chemical Kinetics and Reactions Mechanisms, Van Nostrand-Reinhold, New Yurk,
1980, pp. 83-91.
5~ Bas010 and R. G. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., John Wiley & Sons, New
York, 1967, pp. 158-170.
6 ~ G.. Wilkins, The Study of Kinetics and Mechanism qf Reactions of Transition Metal Complexes,
Allyn and Bacon, Boston, 1974, pp. 193-196.
7 ~D.
. Atwood, Inorganic and Organometallic Reaction Mechanisms, Brooks/Cole, Monterey, CA,
1985, pp. 82-83.
'c. H. Langford and T. R. Stengle, Ann. Rev. Phys. Chem., 1968, 19, 193.
 12-4 Experimental Evidence in Octahedral Substitution 423

4. The rate of reaction correlates with the metal-ligand bond strength of the leaving
group, in a linear free energy relationship (LFER, explained in the next section).
5. Activation energies and entropies are consistent with dissociation, although inter-
pretation of these parameters is difficult. Another activation parameter now being
measured by experiments at increased pressure is the volume of activation, the
change in volume on forming the activated complex. Dissociative mechanisms
generally result in positive values for AV,,, because one species splits into two,
and associative mechanisms result in negative AV,,, values because two species
combine into one, with a presumed volume smaller than the total for the reactants.
However, caution is needed in interpreting volume effects because solvation
effects, particularly for highly charged ions, may be larger than the difference
expected for the reaction otherwise.

12-4-2 LINEAR FREE ENERGY

RELATIONSHIPS

Many kinetic effects can be related to thermodynamic effects by a linear free energy
relationship (LFER).~Such effects are seen when, for example, the bond strength of a
metal-ligand bond (a thermodynamic function) plays a major role in determining the
dissociation rate of that ligand (a kinetic function). When this is true, a plot of the loga-
rithm of the rate constants (kinetic) for different leaving ligands versus the logarithm of
the equilibrium constants (thermodynamic) for the same ligands in similar compounds
is linear. The justification for this correlation is found in the Arrhenius equation for tem-
perature dependence of rate constants and the equation for temperature dependence of
equilibrium constants. In logarithmic form, they are

E
L2 -AH0 AS0
l n k = l n A - - RT and l n K = - + -RT
R
kinetic thermodynamic

If the pre-exponential factor, A, and the entropy, AS0, are nearly constant and the
activation energy, E,, depends on the enthalpy of reaction, A H 0 , there will be a linear
correlation between In k and In K. A straight line on such a log-log plot is indirect evi-
dence for a strong influence of the thermodynamic parameter, A H 0 , on the activation en-
ergy of the reaction. In molecular bonding terms, a stronger bond between the metal and
the leaving group results in a larger activation energy, a logical connection for a dissocia-
tive mechanism. Figure 12-5 shows an example from the hydrolysis of [ c o ( N H ~ ) ~ x ] ~ + :

From this evidence, ~ a n ~ f o r dargued

" that the X- group is essentially completely
dissociated and acts as a solvated anion in the transition state of acid hydrolysis and that
water is at most weakly bound in the transition state. Another example from reactions of
square-planar platinum complcxcs is given in Section 12-6-2.

. Moore and R. G. Pearson, Kinetics and Mechanism, 3rd ed., John Wiley & Sons, New York,
9 ~ W.
1981, pp. 357-363.
'OC. H. Langford, Inorg. Chern., 1965,4,265.
424 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

FIGURE 12-5 Linear Free Energy

and [ C O ( N H ~ ) ~ XHydrolysis.
]~+
The log of the rate constant is plotted
against the log of the equilibrium
constant for the acid hydrolysis
reaction of [ C O ( N H ~ ) ~ Xions.
]~*
Measurements were made at 25.0°C.
Points are designated as follows:
1 X = F-. , 2, X- = H2P04-;
3, X- = C1-; 4, X = Br-;
5, X- = I-; 6, X- = NO3-.
(Reproduced with permission from
C. H. Laugford, Inorg. Chem., 1965,
4,265. Data for F- from S. C. Chan,
J. Chem. Soc., 1964,2375,and for I-
from R. G. Yalman, Inorg. Chem.,
1962, 1, [Link] other data from
A. Haim and H. Taube, Inorg.
Chem., 1964,3, 1199.)
Examples of the effect (or lack of effect) of incoming ligand are given in Tables
12-3 and 12-4. In Table 12-3, the data are for the first-order region (large [Y]). The kl
column gives the rate constants for anion exchange; the kl/kl(H20) shows the ratio of
kl to the rate for water exchange. The rate constants are all relatively close to that for
water exchange, as would be expected for a dissociative mechanism. Table 12-4 gives

TABLE p - 3 ' -
LMRing Rate Constants for Anation or water Exchange

-
, ,
. o f ' [ ~ ' a ( ~ H ~ ) ~ ~ i 0 j ~ ~ .+- - a. t ,.- 4 5 " ~ . .
- ,
, , ' , .

+
[ c o ( N H ~ ) ~ H ~ o ] ~ +Ym- [CO(NH~)~Y](~-~')+
H20 +
Ym- ,
k ( I oP6S F ' ) k ~ / ~(Hzo)
k Reference

" W. Schmidt and H. Tauhe, Inorg. Chem., 1963,2,698.

SOURCES:
b ~R. .Hunt and H. Taube, J. Am. Chem. Soc.,1958, 75, 1463.
T. W. Swaddle and G. Guastalla, Inorg. Chem., 1969, 8, 1604.
C. H. Langford and W. R. Muir, J. Am. Chem. Soc., 1967,89,3141

Y k o ~ (id
o M-I s-') K~ ( ~ - l j k o (lo4$ - I )
CH~PO~'-
CH3CO0
NCS-
F-
HF
Hz0
NH3
C5H5N, pyridine
C4H4N2,pyrazine
NH~(cH~)~NM~~'

SOURCE: Adapted with permission from R. G. Mlkins, Ace. Chem. Res., 1970,3,
408; C4H4N2data are from J. M. Malin and R. E. Shepherd, J. horg. Nucl.
Chem., 1972,34,3203.
 ' 1.2-4 Experimental Evidence in Octahedral Substitution 425

data for the second-order region for anation of [ N ~ ( H ~ o ) ~The

] ~ +second-order
. rate
constant, koKo, is the product of the ion pair equilibrium constant, K O , and the rate
constant, ko:

KO is calculated from an electrostatic model that provides good agreement with the few
cases in which experimental evidence is also available. The rate constant, ko, varies by
a factor of 5 or less and is close to the rate constant for the exchange of water. The close
agreement for the wide variety of different ligands shows that the effect of the incoming
ligand on the second step is minor, although the difference in ion pair formation is
significant. Both these reactions are consistent with D or Id mechanisms, with ion pair
formation likely as the first step in the nickel reactions.

12-4-3 ASSOCIATIVE MECHANISMS

Associative reactions are also possiblc in octahcdral substitution, but are much less
common." Table 12-5 gives data for both dissociative and associative interchanges for
similar reactants. In the case of water substitution by several different anions in
[ c ~ ( N H ~ ) ~ ( H ~ othe + , constants are quite similar (within a factor of 6), indicative
) ] ~rate
of an Id mechanism. On the other hand, the same ligands reacting with [ c ~ ( H ~ o ) ~ ] ~ +
show a large variation in rates (more than a 2000-fold difference), indicative of an I,
mechanism. Data for similar Co(II1) complexes are not conclusive, but their reactions
generally seem to have Id mechanisms.
Reactions of Ru(II1) compounds frequently have associative mechanisms and
those of Ru(11) compounds have dissociative mechanisms. The entropies of activation
for substitution reactions of [Ru(III)(EDTA)(H20)]- are negative, indicating associa-
tion as part of the transition state. They also show a very large range of rate constants

. .
TABLE 12-5
Effects of Entering Group and & -,, ~ i,~- ,a n don
s Rates
Rate Constantsfor Anation

SOURCE: Reproduced with permission from J. D. Atwood, hzorganic and

Organometallic Reaction Mechanisms, BooksICole, Monterey, CA, 1985, p. 85;
data from D. Thusius, Inorg. Chem., 1971, 10, 1106; T. Ramasami and
A. G. Sykes, Chettz. Cottzmutz. (Canzbridge), 1978, 378.

" ~ t w o o d ,Inorganic and Organometallic Reaction Mechanisms, p. 85.

426 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

. ,
TABLE 12-6
'Rate Constants for [Ru(lll)(EDTA)(H20)]- [Link]
Ligand k (M-' s-' ) AH* (W 1 n ~ 1 - I ) AS* ( J rnol-I K-I)

Pyrazine 20,000 i 1,000 5.7 i 0.5

Isonicotinamide 8,300 + 600 6.6 * 0.5
Pyridine 6,300 i 500
Imidazole 1,860 + 100
SCN- 270 f 20 8.9 i 0.5
CH3CN 30 i 7 8.3 f 0.5

T. Matsubara and C. Creutz, Inorg. Chenz., 1979, 18, 1956.

SOURCE:

TABLE 12-7
Rate Constants for [RU(II)(EDTA)(H~O)']~-
&,Ibrtituion

Ligand k ,(M-' sp')

Isonicotinamide
CH3CN
SCN-

T. Matsubara and C. Creutz, Inorg. Chem., 1979, 18, 1956.

SOURCE:

depending on the incoming ligand (Table 12-6), as required for an I, mechanism, but
those of Ru(Z1) (Table 12-7) are nearly the same for different ligands, as required for an
Id mechanism. The reasons for this difference are not certain. Both complexes have a
free carboxylate (the EDTA is pentadentate, with the sixth position occupied by a water
molecule). Hydrogen bonding between this free carboxylate and the bound water may
distort the shape sufficiently in the Ru(II1) complex to open a place for entry by the
incoming ligand. Although similar hydrogen bonding may be possible for the Ru(11)
complex, the increased negative charge may reduce the Ru-H20 bond strength
enough to promote dissociation.

12-4-4 THE CONJUGATE BASE MECHANISM

Other cases in which second-order kinetics seemed to require an associative mechanism :

have subsequently been found to have a conjugate base mechanismi2 (called SNICB,
for substitution, nucleophilic, unimolecular, conjugate base in Ingold's notationt3). j
These reactions depend on amine, ammine, or aqua ligands that can lose protons to ;
form amido or hydroxo species that are then more likely to lose one of the other ligands. i
If the structure allows it, the ligand trans to the amido or hydroxo group is frequently f
the one lost. a
[ c o ( N H ~ ) ~ x ] ~++ OH- +
[CO(NH~)~(NH~)X]'H 2 0 (equilibrium) (1) {
[co(NH~)~(NH~)x]+
-+ [ c o ( N H ~ ) ~ ( N H ~ )+
] ~X-
~ ' - (slow) (2) g
+
[ c o ( N H ~ ) ~ ( N H ~ ) ] ~H+2 0 --+ [CO(NH~)~(OH)]~+
(fast) (3) 1
6
Overall, 1
g
+
[ c o ( N H ~ ) ~ x ] ~ + OH- --+ X-
[CO(NH~)~(OH)]~+ + $B
g
e?
" ~ i l k m s , The Study of Kinetics and Mechunlrm of Reactions of Transition Meioi Ci,[Link],
pp. 207-210; Bas010 and Pearson, Mechanisms of Inorganic Reactions, pp. 177-1 93. [
I3C. K. Ingold, Structure and Mechanism in Organic Chemistry, Cornell University Press, Ithaca, NY,
1953, Chapters 5 and 7. [
P
&
 12-4 Experimental Evidence in Octahedral Substitution 427
In the third step, addtion of a ligand other than water is also possible; in basic solution,
the rate constant is koH and the equilibrium constant for the overall reaction is KO".
Additional evidence for the conjugate base mechanism has been provided by
several related studies:
1. Base-catalyzed exchange of hydrogen from the arnine groups takes place under
the same conditions as these reactions.
2. The isotope ratio (180/160) in the product in 180-enriched water is the same as
that in the water regardless of the leaving group (X- = C1-, Br-, NO3-). If an
incoming water molecule had a large influence (an associative mechanism), a
higher concentration of 180should be in the product, because the equilibrium
constant K = 1.040 for the reaction
H2160 + 180H- ~ ~ + 1 '6 0 ~~- 0

3. RNH2 compounds (R = alkyl) react faster than NH3 compounds, possibly be-
cause steric crowding favors the 5-coordinate intermcdiate formed in Step 2.
4. The rate constants and dissociation constants for these compounds form a linear
free energy relationship (LFER), in which a plot of In kOHversus In KOHis linear.
5. When substituted amines are used, and there are no protons on the nitrogens
available for ionization, the reaction is very slow or nonexistent.

Reactions with [ C o ( t r e n ) ( ~ ~ ~ ) Cisomers

l ] ~ + show that the position trans to the
leaving group is the most likely deprotonation site for a conjugate base mechanism.14
The reaction in Figure 12-6(a) is lo4 times faster than that in Figure 12-6(b). In addi-
tion, most of the product in both reactions is best explained by a trigonal-bipyramidal
intermediate or transition state with the deprotonated amine in the trigonal plane. The
reaction in Figure 12-6(a) can form this state immediately; the reaction in Fig-
ure 12-6(b) requires rearrangement of an initial square-pyramidal structure.

3
H,N'
Co

1 'NH,
--+
+

H,N 'dl
N/\I
CoN H 2

'NH,
+
.I- 61
'
,"
H~N'
. Co'

NH3
+

H,N'
CO'

1 O' H
C1 NH,

FIGURE 12-6 Base Hydrolysis of [ ~ o ( t r e n ) ( ~ ~ ~Isomers.

) ~ l ] ~(a)
+ Leaving group (C1-) trans to
\ 'co-
deprotonated nitrogen. (b) Leaving group (Cl-) cis to deprotonated nitrogen. (a) is lo4 faster than
H2N' 1 'NH,
OH
(b), indicating that trans substitution is strongly favored. (Data from D. A. Buckingham,
P. J . Creswell, and A. M. Sargeson, lnorg. Chem., 1975, 14, 1485.) 15%

A. Buckingham,
140. P. J. Cressell, and A. M. Sargeson, Inorg. Chem., 1975, 14, 1485
428 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

Explanations of the promotional effect of the amido group center on its basic 1
strength, either as a a donor or because of ligand to metal rr interaction. The rr interac- :
tion is most effective when the amido group is part of the trigonal plane in a trigonal- ';
bipyramidal geometry, but there is at least one case in which this geometry is not
necessarily achieved.'"

12-4-5 THE KINETIC CHELATE EFFECT

The thermodynamic chelate effect, which causes polydentate complexes to be thermo- - '

dynamically more stable than their monodentate counterpart^,'^ was described in

Section 10-1-1. The difference in the attachment and dissociation of the second (and
third or higher numbered) point of attachment for the ligand is also observed kinetically. '1
Substit~~tion for a chelated ligand is generally a slower reaction than that for a
similar monodentate ligand. Explanations for this effect center on two factors, the in-
creased energy needed to remove the first bound atom and the probability of a reversal
of this first step.17
The reaction must have two dissociation steps for a bidentate ligand, one for each :
bound atom (the addition of water in Steps 2 and 4 is likely to be fast, because of its
high concentration):

The first dissociation (1) is expected to be slower than a similar dissociation of .':
ammonia because the ligand must bend and rotate to move the free amine group away
from the metal. The second dissociation (3) is likely to be slow because the concentra- ;.
tion of the intermediate is low and because the first dissociation can readily reverse. The I

'b. A. Buckingham, P. A. Marzilli, andA. M. Sargeson, Innrg. Chem., 1969.8, 1595.

' 6 ~ a s o l oand Pearson, Mechanisms of Inorganic Reactions, pp. 27, 223; G. Schwarzenbach, Helv. :j
, -4
Chim. Acta, 1952,35,2344.
. Margerum, G. R. Cayley, D. C. Weatherburn, and G. K. Pagenkopf, "Kinetics of Complex
1 7 ~ W.
Formation and Ligmd Exchange," in A. I3 Martell, ed,, Coordination Chemistry. Vol 2, American Chemical
Society Monograph 174, Washington, DC: 1978, pp. 1-220.
 12-5 Stereochemistry of Reactions 429
uncoordinated nitrogen is held near the metal by the rest of the ligand, making reattach-
ment more likely. Overall, this kinetic chelate effect reduces the rates of aquation reac-
tions by factors from 20 to 10'.

A common assumption is that reactions with dissociative mechanisms are more likely
to result in random isomerization or racemization and associative mechanisms are more
likely to result in single-product reactions; however, the evidence is much less clear-
cut. Dissociative mechanisms can lead to single-product reactions with either retention
of configuration or a change of configuration, depending on the circumstances. For
example, base hydrolysis of A - c i s - [ ~ o ( e n ) ~ ~ lin~ ] dilute
' (<0.01 M) hydroxide
yields A-cis-[~o(en)~(OH)~]+, but in more concentrated (>0.25 M) hydroxide it gives
A-cis-[~o(en)~(OH)~]+ (Tables 12-8 and 12-9 and Figure 12-7).18 A conjugate base
mechanism is expected in both cases, with the hydroxide removing a proton from an
ethylenediamine nitrogen, followed by loss of the chloride trans to the deprotonated
nitrogen. In the more concentrated base, the higher concentration of ion pairs
( [ C O ( ~ ~ ) ~ C ~ ~ ] +is
- Oassumed
H-) to result in a water molecule (from the OH- and the
H' removed from ethylenediamine) positioned for easy addition with inversion of the
chiral center.
A similar change in product, this time dependent on temperature, takes place in the
substitution of ammonia for both chlorides in [ ~ o ( e n ) ~ ~ lAt ~ ]low
+ . ~temperatures
~
(-33" C or below, in liquid ammonia), there is inversion of configuration; at higher tem-
peratures (above 25" C in liquid ammonia, alcohol solution, or solid exposed to gaseous
ammonia), there is retention. In both cases, there is also a small fraction of the trans isomer.
Although not a complete explanation of these reactions, all the reported inversion
reactions occur under conditions in which a conjugate base mechanism is possible.20
The orientation of the ligand entcring thc proposed trigonal-bipyratnidal intermediate

.-
(21
I 7;INH2 +OH-
HN
2 1
s- H
LH.,.
,,,, - 1
C1
-----A

H2N,-.. 1
H
,.,N_P
, H
H2N-
c1
1-NH2
--+
%
..-

Co __j Co 'OH- ---$ Co iH

' Co
C l 4 1 \NH2 c141 'N---~*," I
H 2 N J
'N-H H ~\
N, I O' H
' NH2
H2N 3 \ H2N J H
A
H
(b) A
FIGURE 12-7 Mechanisms of Base Hydrolysis of A-cis-[~o(en)2Cl~]+. (a) Retention of configu-
ration in dilute hydroxide. (b) Inversion of configuration in concentrated hydroxide.

"L. J. Boucher, E. Kyuno, and J. C. Bailar, Jr., J. Am. Chem. Soc., 1964,86,3656.
Bailar, Jr., J. H. Haslam, and E. M. Jones, J. Am. Chem. Soc., 1936, 58, 2226; E. Kyuno and
1 9 ~ .C.
J. C. Bailar, Jr., J. Am. Chem. Soc., 1966,88, 1125.
o Pearson, Mechanisms of Inorganic Reactions, p. 272.
2 0 ~ a s o l and
430 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

TABLE 72-8 ,,, -

Stereochemistry of Acid Aquation ' ,

[ C O ( ~ ~ ) ~ L-tX H20
] ~ + --.* [ C O ( ~ ~ ) ~ L H ~ O ]t
( 'X-
+~)+
cis-L X % cis Product trans-L X % cis Product
OH- Cl- 100 OH- C1- 75
OH- BrF 100 OH- Br- 73
Br- * C1- 100 Br- C1- 50
C1- C1- 100 Br- Br- 30
C1- Br - 100 C1- C1- 35
N3- C1- 100 C1- Br- 20
NCS- Cl- 100 NCS- C1- 50-70
NCS- Br- 100 NH? C1- 0
NO;?- Cl- 100 NOz- C1- 0

SOURCE: Data from F, Basolo and R. G. Pearson, Mechanisms of Inorganic Reactions, 2nd ed.. J. Wiley
& Sons, New York, 1967, p. 257.

-
TABLE 12-9 .
i$.~ereochemCstryof' Base .~~[Link] ' ,

[ c o ( ~ ~ ) ~ L x+] ~OH-
+ [CO(~~)~LO+
H ]X-
~'
% cis Product
cis-L X A Racemica A trans-L X % cis Product
OH- C1- 61 36 OH- C1- 94
OH- Br- 96 OH- Br- 90
C1- C1- 21 16 C1- C1- 5
C1- Br- 30 C1- Br - 5
Br- C1- 40 Br - C1- 0
N.1- C1- 51 N3- C1- 13
NCS- C1- 56 24 NCS- C1- 76
NH3 Br- 59 26 NCS Br- 81
NH3 C1- 60 24 NH3 C1- 76
NO2- C1- 46 20 NO2- C1- 6
SOURCE: Data from F, Basolo and R. G. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., J. Wiley
& Sons, New York, 1967, p. 262.
NOTE:The total % cis product is the sum of A and A obtained from the A-cis slarting material.
The optically inactive tram isomer will yield racemic cis; % trans = 100% - % cis.
a Racemic reactant, so the product is also racemic

then dictates the configuration of the product. In some cases, a preferred orientation
of the other ligands may dictate the product. For example, the P form of trim complexes
is more stable than the or form; both are shown in Figure 9-18.

12-5-1 SUBSTITUTION IN TRANS COMPLEXES

Substitution of Y for X in tmns-[M(LL)2BX] (LL = a bidentate ligand such as en) can
proceed by three different pathways. If dissociation of X from the reactant leaves a
square-pyramidal intermediate that then adds the new ligand directly into the vacant
site, the result is retention of configuration and the product, like the reactant, is trans,
shown in Figure 12-8(a). A trigonal-bipyramidal intermediate with B in the trigonal
plane leads to a mixture of trans and cis, as shown in Figure 12-8(b). The incoming
 12-5 Stereochemistry of Reactions 43 1

B trans

FIGURE 12-8 Dissociation

Mechanism and Stereochemical
Changes for trans-[M(LL)2BX].
(a) Square-pyramidal intermediate
(retention of configuration). (b) Trig-
onal-bipyramidal intermediate (three
possible products). (c) Less likely
trigonal-bipyramidal intermediate
(two possible products.) (c)

ligand can enter along any of the three sides of the triangle, resulting in two cis pos-
sibilities and one trans possibility. Dissociation to form a trigonal pyramid with B in
an axial position, Figure 12-8(c), allows two positions for attack by Y, both of which
give cis products (the third side of the triangle is blocked by an LL ring). An inter-
mediate with an axial B is less likely than one with an equatorial B, because an axial
B requires more rearrangement of the ligands (a 90" change by one nitrogen and 30"
changes by two others, in contrast to two 30" changes for the equatorial B) as well as
a larger stretch for the LL ring in the equatorial plane. As a result, the statistical
probability of a change from trans to cis is two thirds for a trigonal-bipyramidal
intermediate.

EXERCISE 12-2
h

L
Starting with the structure on the right, follow the example of
Figure 12-8(b) and show that the first two products would be A
rather than A.
X
(Experimentally, the two chiral forms are equally likely, because
how we draw the structures has no effect on the experimental result.)

In fact, experimental results indicate that the statistical distribution is seldom fol-
lowed. With trans reactants, both acid aquation and base substitution reactions result in
i a mixture of isomers; the fractions of cis and trans depend on the retained ligand and
i range from 100% trans to 94% cis, as shown in Tables 12-8 and 12-9. Of course, any

..... ... -. .. . . . . . . . .. .- .. ........

432 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

TABLE 12-10

X cis - trans trans -

Rate Constants for Reactions of [ ~ o ( e n ) ~ ( ~ ~ at
cis
) ~ j "k~
O 25"C,
Racernization
s-I)
H 2 0 Exchange

SOURCE: Adapted with permission from R. G. Wilkins, The Study of Kinetics and Mechanism of
Reactions of Trunaition Metal Comnplaes,Allyn and Bacon, Boston, 1974, p. 344. Data from
M. L. Tobe, in J. H. Ridd, ed., Studies in Structure and Reactivity, Methuen, London, 1966, and
M. N. Hnghes, J. Chem. Soc., A, 1967, 1284.

cis isomer produced from the optically inactive trans reactants will be a racemic mix-
ture of A and A. For both the axial and equatorial types of trigonal-bipyramidal inter-
mediates, the chiral form of the product is determined by which square plane becomes
trigonal in the intermediate; A and A products are equally likely.
Other factors, such as the leaving ligand, X, can strongly influence the mecha-

-
nism and the outcome, making the products deviate further from statistical probabili-
ty. Data on several reactions for [ C O ( ~ ~ ) ~ ( H ~ Oare
) Xshown
] ~ + in Table 12-10. For
X = C1-, SCN-, and H20, racemization and cis tram conversion are nearly
equal in rate, making it likely that they have identical intermediates. Water exchange

-
is faster than the other reactions for all but the hydroxide complex; simple exchange

-
with the solvent requires no rearrangement of the ligands and the high concentration
of water makeb it more likely. For X = NH3, racemization and trans cis
conversion are faster than cis trans conversion. The reasons for this difference
are not clear.

12-5-2 SUBSTITUTION IN CIS COMPLEXES

Substitution in cis complexes can proceed by the same three intermediates as in the
trans complexes. Again, a square-pyramidal intermediate results in retention of config-
uration, providing a cis product in this case. If dissociation of X forms a trigonal bipyra-
mid with B in the trigonal plane, there are three possible locations for the addition of Y,
all in the same trigonal plane. Two of these result in cis products and one in a trans
product. The less likely trigonal bipyramid with an axial B, whether derived from a cis
or a trans reactant, produces two cis products that are half A and half A. These possi-
bilities are all shown in Figure 12-9.
An optically active cis complex can yield products that retain the same configura-
tion, convert to trans geometry, or create a racemic mixture. Statistically, the product of
substitution of a C~S-[M(LL)~BXJ complex through a trigonal-bipyramidal intermediate
should be 116 trans if both intermediates were equally likely and 113 trans if the axial
B form is not formed at all. Experimentally, aquation of C~S-[M(LL)~BX] in acid results
in 100% cis isomer (Table 12-9), indicating a square-pyramidal transition slate. Substi-
tution of optically active cis complexes in base gives products ranging from 95% to
30% cis, with about 2:l retention of chiral configuration (Table 12-9). (Four of the re-
actants listed in Table 12-9 are racemic, so the product is also a racemic cis mixture.)
 12-5 Stereochemistry of Reactions 433

FIGURE 12-9 Dissociation

Mechanism and Stereochemical
Changes for ~ ~ . Y - [ M ( L L ) ~ B X .
(a) Square-pyramidal intermediate
(retention of configuration).
(b) Trigonal-bipyramidal
intermediate (three possible
products). (c) Unlikely trigonal
hipyramidal intermediate (two
possible products).

Among the compounds that retain their optical activity and geometry on hydrolysis are
l~M
[ ~ ( e n ) ~ ~with ] + = Co, Rh, and RU.'~ As a general rule, cis reactants retain their
cis configuration, but trans reactants are more likely to give a mixture of cis and trans
products.

12-5-3 ISOMERIZATION OF CHELATE RINGS

Isomerization has been described previously for a number of complexes with mon-
odentate ligands or with two bidentate ligands. Similar reactions with three bidentate
ligands or with more complex ligands can follow two types of mechanism. In some
cases, one end of a chelate ring dissociates and the resulting 5-coordinate intermedi-
ate rearranges before reattachment of the loose end. This mechanism does not differ
appreciably from the substitution reactions described in Sections 12-5-1 and 12-5-2;
the ligand that dissociates in the first step is the same one that adds in the final step,
after rearrangement.

"s. A. Johnson, F. Basolo, and R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 1741; J. A. Broomhead
and L. Kane-Maguire, Inorg. Chem., 1969, 8,2124.
434 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

FIGURE 12-10 Twist Mechanisms for Isomerization of M(LL)? and [~o(trien)Cl~]+ Complexes.
(a) Trigonal twist. The front triangular face rotates with respect to the back triangular face. (b) Twist
with perpendicular rings. The back ring remains stationary as the front two rings rotate clockwise.
(c) Twist with parallel rings. The back ring remains stationary as the front two rings rotate counter-
clockwise. (d) [Co(trien)~l~]+ a-p isomerization. The connected rings limit this isomerization to a
clockwise trigonal twist of the front triangular face.

Pseudorotation
Other isomerization mechanisms involving compounds containing chelating ligands
are different types of twists. A number of twist mechanisms have been described, with
different movements of the rings; those most commonly considered are shown in
Figure 12- 10.
The trigonal, or Bailar, twist, Figure 12-10(a), requires twisting the two opposite
trigonal faces through a trigonal prismatic transition state to the new structure. In the
tetragonal twists, one chelate ring is held stationary while the other two are twisted to
the new structure. The first one illustrated, Figure 12-10(b), has a transition state with
the stationary ring perpendicular to those being twisted. The second tetragonal twist,
Figure 12-10(c), requires twisting the two rings through a transition state with all three
rings parallel. There have been attempts to determine which of these mechanisms is ap-
plicable, but the complexity of the reactions and the indirect means of measurement
leave them subject to different interpretations. NMR study of tris(trifluoroacety1aceto-
nato) metal(II1) chelates shows that a trigonal twist mechanism is not possible for
M = Al, Ga, In, and fac-Cr, but leaves it a possibility for fac-CO." The multiple-ring
l ~ ] ~ only a trigonal twist in its conversion to the P
structure of c i s - o r - [ ~ o ( t r i e n ) ~allows
isomer, as shown in Figure 12-10(d).

12-6 The products of substitution reactions of square-planar complexes [platinum(II)

SUBSTITUTION complexes are the primary examples] have the same configuration as the reactants,
REACTIONS OF with direct replacement of the departing ligand by the new ligand. The rates vary
SQUARE-PLANAR enormously, and different compounds can be formed, depending both on the entering
COMPLEXES and the leaving ligands. This section and Section 12-7 describe some of these effects.

12-6-1KINETICS AND STEREOCHEMISTRY

OF SQUARE-PLANAR SUBSTITUTIONS

Square-planar substitution reactions frequently show two term rate laws, of the form
Rate = kl[Cplx] t k2[Cplx][Y]

' 2 ~ C.
. Fay and T. S. Piper, Inorg. Chem., 1964,3, 348.
436 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

TABLE 12-11

~ -
Rate Constants and LFER Parameters for Entering Groups
t r ~ n s - P t L ~ C+1 Y trans-PtL2C1Y + C1-
k (10-3 M-' s-I)

SOURCE: Rate constants from U. Belluco, L. Cattalini, F. Basolo, R. G. Pearson,

and A. Turco, J. Am. Chem. Soc., 1965,87,241;PPh3 and q p t data from
R. G. Pearson, H. Sobel, and J. Songstad, J. Am. Chem. Soc., 1968,90,319.
NOTE:s and q are nucleophilic reaction parameters explained in the text.

This mechanism explains naturally the effect of the incoming ligand. A strong
Lewis base is likely to react readily, but the hard-soft nature of the base has an even
larger effect. Pt(I1) is generally a soft acid, so soft ligands react more readily with it. The
order of ligand reactivity depends somewhat on the other ligands on the Pt, but the order
for the reaction

for different Y in methanol was found to be as follows (examples are in Table 12-11):

A similar order, with some shuffling of the center of thc list, is found for reactants
with ligands other than chloride as T. The ratio of the rate constants for the extremes in
the list is very large, with k(PPh3)/k(CH30H) = 9 X lo8. Because T and Y have sim-
ilar positions in the transition state, it is reasonable for them to have similar effects on
the rate, and they do. Discussion of this trans effect is in the next section.
By the same argument, the leaving group X should also have a significant influ-
ence on the rate, and it does (Table 1 2 - 1 The ~ ~order
~ of ligands is nearly the reverse
of that given above, with hard ligands such as C1-, NH3, and NO3- leaving quickly. Soft
ligands with considerable 71. bonding such as CN- and NO2- leave reluctantly; in the
reaction

thc rate increases by a factor of lo5 with H 2 0 as compared with X- = CN- or N O 2 as

the leaving group. The bond-strengthening effect of the metal-to-ligand 71. bonding re-
duces the reactivity of these ligands significantly. In addition, 71. bonding to the leaving
group uses the same orbitals as those bonding to the entering group in the trigonal plane.
These two effects result in the slow displacement of metal-to-ligand 71.-bonding ligands
when compared with ligands with only a bonding or ligand-to-metal IT bonding.

2 4 ~ i l k i n sThe
, Study of Kinetics and Mechanism of Reactions of Transition Metal Complexes, p. 23 1 .
 12-7 The Trans Effect 437

TABLE12-1:2 , ,

',.Ral$
,,. .
Canstants for'teaving Gkups
[pt(dien)x]+ + py -+ [ ~ t ( d i e n ) ~ ~ ]X- +
-
(Rate (kl -I- k2[py])[~t(dien)~]+>
~+

k2 ( M - I S - I )
very fast
5.3 x 10-3
3.5 x 10-3
1.5 X lo-"
1.3 X
4.8 X lo@
3.8 X
2.8 X

SOURCE: Calculated from data in F. Basoto,

H. B. Gray, and R. G. Pearson, J. Am. Chem. Soc.,
1960,82,4200.

Good leaving groups (those that leave easily) show little discrimination between
entering groups. Apparently, the ease of breaking the Pt-X bond takes precedence
over the formation of the Pt-Y bond. On the other hand, for complexes with less re-
active leaving groups, the other ligands have a significant role; the softer PEt3 and
AsEt3 ligands show a large selective effect when compared with the harder dien or en
ligands. The LFER equation25for this comparison is

where ky = rate constant for reaction with Y

ks = rate constant for reaction with solvent
s = nucleophilic discrimination factor (for the complex)
qpt = nucleophilic reactivity constant (for the entering ligand)
The parameter s is defined as 1 for trans-[Pt(~y)~Cl~] and has values from 0.44
for the hard [ ~ t ( d i e n ) ~ ~ to
0 ] 1.43
~ + for the soft t r a n ~ - [ P t ( P E t ~ ) ~ CValues
l ~ ] . of q p t
are found by the equation qpt = log(ky/ks), where ks refers to reaction with
tran~-[Pt(py)~Cl~] in methanol at 30" C. Table 12-11 shows both these factors. For
L = PEt3, the change in rate constant is greater than for L = py because of the larger s
value, and the increase in rate constants parallels the increase in qp,. Each of the para-
meters s and q p , may change by a factor of 3 from fast reactions to slow reactions,
allowing for an overall ratio of lo6 in the rates.

e v ~ ~ the concept of the trans effect in platinum chemistry.

In 1926, ~ h e r n ~ a introduced
12-7
THE TRANS EFFECT In reactions of square-planar Pt(I1) compounds, ligands trans to chloride are more
easily replaced than those trans to ligands such as ammonia; chloride is said to have a
stronger trans effect than ammonia. When coupled with the fact that chloride itself is
more easily replaced than ammonia, this trans effect allows the formation of isomeric

2 s ~ t ~ o oInorganic
d, and OrganometallicReaction Mechanisms, pp. 60-63.
2 6 ~ I.
. Chernyaev, Ann. Inst. Platine USSR., 1926,4,261.
438 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

NH3
I
(a) NH, -Pt -NH,
I
NH3

NH, C1

FIGURE 12-12 Stereochemistry

and the trans Effect in Pt(1I)
Reactions. Charges have been
omitted for clarity. In (a) tl~rough(f),
the first substitution can be at any C1
position, with the second controlled I
by the trans effect. In (g) and (h), (h) C1-Pt -NH,
both substitutions are controlled by I
the lability of chloride. NH3

Pt compounds, as shown in the reactions of Figure 12-12. In reaction (a), after the first
ammonia is replaced, the second replacement is trans to the first CI-. In reaction (b),
the second replacement is trans to C1- (replacement of ammonia in the second
reaction is possible, but then the reactant and product are identical). The first steps in
reactions (c) through (f) are the possible replacements, with nearly equal probabilities
for replacement of ammonia or pyridine in any posilion. The second steps of (c) '
through (f) depend on the trans effect of C1-. Both steps of (g) and (h) depend on the ':

greater lability of chloride. By using reactions such as these, it is possible to prepare :

specific isomers with different ligands. Chernyaev and coworkers did much of this, i
 12-7 The Trans Effect 439

preparing a wide variety of compounds and establishing the order of trans effect
ligands:
CN- - CO - -
C2H4 > pH3 SH2 > NO2- > I- > Br- > C1- > NH3 py > OH- > H 2 0 -
I EXERCISE 12-3

I Predict the products of these reactions (there may be more than one product when there are
conflicting preferences).

[ptc1412- + NO2- -- (a) (a) + NH3 - (b)

[PtC13NH3]- + NO2-

[Ptcl(~H,),]+
- -
+ NOz-
(c)

(e)
(c) + NO2-

(el + N02- -
-----)

--
(d)

(0
[ ~ t c l , ] ~+
[PtIJ + C1-
- I-
- (g)
(i)
(g) + 1-
(i) + Cl-
(h)
(J)

12-7-1 EXPLANATIONS OF THE TRANS EFFECT27

Sigma-bonding effects
Two factors dominate the explanations of the trans effect, weakening of the Pt-X
bond and stabilization of the presumed 5-coordinate transition state. The energy rela-
tionships are given in Figurc 12-13, with the activation energy the difference between
the reactant ground state and the first transition state.

FIGURE 12-13 Activation Energy

and the trans Effect. The depth of the
energy curve for the intermediate and
the relative heights of the two
maxima will vary with the specific
reaclion. (a) Poor trans effect, low
ground state, high transition state.
(b) a-Bonding effect, higher ground
state (trans influence).
(c) n-Bonding effect, lower
transition state, (tmns effect).

The Pt-X bond is influenced by the Pt-T bond, because both use the Pt p,
and d X ~ - y 2
orbitals. When the Pt-T o bond is strong, it uses a larger part of these
orbitals and leaves less for the Pt -X bond (Figure 12-14).As a result, the Pt-X bond
is weaker and its ground state (sigma-bonding orbital) is higher in energy, leading to a
smaller activation energy for the breaking of this bond [Figure 12-13(b)]. This ground
state effect is sometimes called the trans influence and applies primarily to the leaving
group. It is a thermodynamic effect, contributing to the overall kinetic result by chang-
ing the reactant ground state. This part of the explanation predicts the order for the trans
FIGURE 12-14Sigma-Bonding effect based on the relative o-donor properties of the ligands:
Effect A strong u bond between Pt
and T weakens the Pt-X bond

The order given here is not quite correct for the trans effect, particularly for CO
and CN-, which have strong trans effects.

27Atwood, Inorganic and Orgunometallic Reactions Mechanisms, p. 54; Basolo and Pearson,
Mechanisms of Znorgaizic Reactions, p. 355.
440 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

Pi-bonding effects
The additional factor needed is .rr bonding in the Pt-T bond. When the T ligand forms
a strong n-acceptor bond with Pt, charge is removed from Pt and the entrance of anoth-
er ligand to form a 5-coordinate species is more likely. In addition to the charge effect,
the dX2-y2 orbital, which is involved in o bonding in the square-planar geometry, and
both the d,, and dyz orbitals can contribute to n bonding in the trigonal-bipyramidal
transition state. Here, the effect on the ground state of the reactant is small, but the
energy of the transition state is lowered, again reducing the activation energy
[Figure 12-13(c)]. The order of n-acceptor ability of the ligands is

C2H4 - CO > CN- > NO2- > SCN- > I- > Br- > C1- > NH3 > OH-
The expanded overall trans effect list is then the result of the combination of the
two effects:

NO2- - SCN- - I- > Br- > C1- > py, NH3 - OH- - H 2 0
Ligands highest in the series are strong r r acceptors, followed by strong u donors.
Ligands at the low end of the series have neither strong u-donor nor n-acceptor
abilities. The trans effect can be very large; rates may differ as much as lo6 between
complexes with strong trans effect ligands and those with weak trans effect ligands.

EXERCISE 12-4
It is possible to prepare different isomers of Pt(I1) complexes with four different ligands. Predict
the products expected if 1 mole of [ ~ t ~ l ~is ]reacted
" successively with the following reagents
(e.g., the product of reaction a is used in reaction b):

a. 2 moles of ammonia
b. 2 moles of pyridine [see Reactions (g) and (h) in Figure 12-12]
c. 2 moles of chloride
d. 1 mole of nitrite, NOz-

- - - Oxidation-reduction reactions of transition metal complexes, like all redox reactions,

12-8
OXIDATION- involve the transfer of an electron from one species to another-in this case, from one
REDUCTION complex to another. The two molecules may be connected by a common ligand through
REACTIONS which the electron is transferred. in which case the reaction is called a bridging or
inner-spherc reaction, or the exchange may occur between two separate coordination
spheres in a nonbridging or outer-sphere reaction.
The rates have been studied by many different methods, including chemical
analysis of the products, stopped-flow spectrophotometry, and the use of radioactive
and stable isotope tracers. Taube's research group has been responsible for a large
amount of the data, and their reviews cover the field.28
The rate of reaction for electron transfer depends on many factors, including the
rate of substitution in the coordination sphere of the reactants, the match of energy lev-
els of the two reactants, solvation of the two reactants, and the nature of the ligands.

2 8 ~ .. Meyer and H. Taube, "Electron Transfer Reactions," in G. Wilkinson, R. D. Gillard, and

I.
J. A. McCleverty, eds., Pergamon, Comprehensive Coordination Chemistry, Vol. 1 , London, 1987,
pp. 33 1-384; H. Taube, Electron Transfer Reactions of Complex Ions in Sohttion, Academic Press, New York,
1970; Chem. Rev., 1952,50,69;J. Chem. Educ., 1968,45,452.
 12-8 Oxidation-Reduction Reactions 441

12-8-1 INNER- AND OUTER-SPHERE

REACTIONS

When the ligands of both reactants are tightly held and there is no change in the coordi-
nation sphere on reaction, the reaction proceeds by outer-sphere electron transfer.
Examples of these reactions are given in Table 12-13 with their rate constants.

TABLE 12-13
Rate Constants for Outer-Sphere Electron Transfer Reactionsa
Reductants

[co(NH,),(NH,)I~+ 6.9 x 10" 1.1 x 10:~

[co(NH~)~F)I~+ 1.8 X l o 3
[cwH~)~(oH)I~+ 3 x 10" 4 x lo-2
~CO(NH,)~NO~)I~~ 3.4 x 10'
[WNHMH~QI~' 5 x lo4 3.0
LCO(NHMJ)I~+ 8 X 1@ 2.6 X lo2
ICO(NH~)~(B~)I~+ 5 x 106 1.6 x 103
LCO(NHM)I~+ 6.7 x lo3
SOURCE:[ ~ r ( b i ~ ~data ] ~ ' J.~ P. Candlin, J. Halpern, and D. L. Trimrn, J. Am. Clzem. Soc., 1964,
) ~ from
] ~ ' from J. F. Endicott and H. Taube, J. Am. Chem. Soc., 1964, 86, 1686.
86, 1019. [ R U ( N H ~ ) ~ data
NOTE:"econd-order rate constants in M-' s-' at 2S°C.

The rates show very large differences, depending on the details of the reactions.
Characteristically, the rates depend on the ability of the electrons to tunnel through the
ligands. This is a quantum mechanical property whereby electrons can pass through po-
tential barriers that are too high to permit ordinary transfer. Ligands with T or p elec-
trons or orbitals that can be used in bonding (as described in Chapter 10 for T-donor
and T-acceptor ligands) provide good pathways for tunneling; those like NH3, with no
extra lone pairs and no low-lying antibonding orbitals, do not.
In outer-sphere reactions, where the ligands in the coordination sphere do not
change, the primary change on electron transfer is a change in bond distance. A higher
oxidation state on the metal leads to shorter a bonds, with the extent of change de-
pending on the electronic structure. The changes in bond distance are larger when eg
electrons are involved, as in the change from high-spin Co(I1) (tZg5eg2) to low-spin
Co(II1) (t2g6).Because the eg orbitals are antibonding, removal of electrons from these
orbitals results in a more stable compound and shorter bond distances. A larger ligand
field stabilization energy makes oxidation easier. Comparing water and ammonia as
ligands, we can see that the stronger field of ammonia makes oxidation of Co(I1) rela-
tively easy. [ c o ( N H ~ ) ~ ]is~ +
a very weak oxidizing agent. The aqueous Co(II1) ion, on
the other hand, has a large enough potential to oxidize water:

Inner-sphere reactions also use the tunneling phenomenon, but in this case a sin-
gle ligand is the conduit. The reactions proceed in three steps: (1) a substitution reaction
that leaves the oxidant and reductant linked by the bridging ligand, (2) the actual trans-
fer of the electron (frequently accompanied by transfer of the ligand), and (3) separation
of the products:29

. Candlin
2 9 ~ P. and J. Halpern, Imrg. Chenz., 1965,4, 766
442 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

[ C O ( N H ~ ) ~ ( C ~+) ][~c+~ ( H ~ o ) ~-]-~++ [ ( N H ~ ) ~ C O ( C ~ ) C ~ ( H+

~OH)2~0] ~ + (1)
Co(II1) oxidant Cr(I1) reductant Co(1II) Cr(I1)

In this case, these are followed by a reaction made possible by the labile nature of
Co(I1):

The transfer of chloride to the chromium in these reactions is easy to follow ex-
perimentally because Cr(II1) is substitutionally inert and the products can be separated
by ion exchange techniques and their composition can be determined. When this i s
done, all the Cr(II1) appears as Crcl2+. The [c~(H~o)~]~+-[c~(H~o)~cI]~+ exchange
reaction (which results in no net change) has also been studied, using radioactive 'lcr
as a tracer.30All the chloride in the product came from the reactant, with none entering
from excess C1- in the solution. The rate of the reaction could also be determined by
following the amount of radioactivity found in the crc12+ at different times during the
reaction.
In many cases, the choice between inner- and outer-sphere mechanisms is difti-
cult. In the examples of Table 12-13, the outer-sphere mechanism is required by the re-
ducing agent. [ R U ( N H ~ ) ~is] an
~ +inert species and does not allow formation of bridging
species fast enough for the rate constants observed. Although [ ~ r ( b i ~ ~is)labile,
~ ] ~ the
+
parallels in the rate constants of the two species strongly suggest that its redox reactions
are also outer-sphere. In other cases, the oxidant may dictate an outer-sphere mecha-
nism. In Table 12-14, [ c o ( N H ~ ) ~ ]and
~ + [ ~ o ( e n ) ~have
] ~ +outer-sphere mechanisms be-
cause their ligands have no lone pairs with which to form bonds to the reductant. The
other reactions are less certain, although cr2+(aq) is usually assumed to react by inner-
sphere mechanisms in all cases in which bridging is possible.

TABLE 12-14
Rate Constants for Aquated Redudantsa
cr2+ EU~+ v 2+
j
[Co(en)d3+ -2 x 10F -5 x lo-" -2 x
ICO(NH~)~~+ 8.9 x lop5 2 x 3.7x lo-* P1
1
ICO(NH~)S(H,~)I~+ 5 x 10-' 1.5 x 10-I -5 x lo-'
1
,
[CO(NH~)~(NO~)I~+ -9 x 10' -1 x 102 1
g
ICO(NH~)S(C~)I~+ 6 x lo5 3.9 x lo2 -5
B
1.4 x 2.5 x 2.5 x 10'
IC~(NH~)S(B~)I~'
[CO(NH,)~(~)I~+ 3 x
lo6
lo6 1.2 x
lo2
lo2 1.2 x lo2
HjI
SOURCE: Data from J. P. Candlin, .I
Halpern,
. and D. L. Trirnm, J. Am. Chem. Soc., 1964,86, 1019; data
for cr2+ reactions with halide complexes from J. P. Candlin and J. Halpem, Inorg. Chem., 1965,4,756; '
data for [c~(NH&]~+reactions with cr2+ and v2+from A. Zwickel and H. Taube, .I. Am. Chem. Soc.,
1961,83,793.
a Rate constants in M-' sf'.

'OD. L. Ball and E. L. King, J. Am. Chem. Soc., 1958,80, 1091.

 12-8 Oxidation-Reduction Reactions 443

v2+(aq) reactions appear to be similar to those of cr2+(aq),although the range of

rate constants is smaller than that for cr2+. This seems to indicate that the ligands are
less important and makes an outer-sphere mechanism more likely. This is reinforced by
comparison of the rate constants for the reactions of [ ~ r ( b i ~ ~(outer-sphere,
)~]~' Table
12-13) and v2+(Table 12-14) with the same oxidants. v2+may have different mecha-
nisms for different oxidants, just as c r 2 + does.
~ u ~ + ( aisq an
) unusual case. The rate constants do not parallel those of either the
more common inner- or outer-sphere reactants, and the halide data are in reverse order
from any others. The explanation offered for these rate constants is that the thermody-
namic stability of the EUX+ species helps drive the reaction faster for F-, with slower
rates and stabilities as we go down the series. Because of the smaller range of rate con-
stants, E U ~ +reactions are usually classed as outer-sphere reactions.
When [co(cN)~]~-reacts with Co(111) oxidants ( [ c o ( N H ~ ) ~ x ] ~that+ ) have po-
tentially bridging ligands, the product is [co(cN)~x]~+,evidence for an inner-sphere
mechanism. Rate constants for a number of these reactions are given in Table 12-15.
The reaction with hexamminecobalt(Il1) must be outer-sphere, but has a rate constant
similar to the others. The reactions with thiocyanate or nitrite as bridging groups also
show interesting behavior. With N-bonded [(NH~)~coNcs]", it reacts by bonding to
the free S end of the ligand, because the cyanides soften the normally hard c o 2 + ion.
With S-bonded [(NH~)~coscN]~+, it reacts initially by bonding to the free N cnd of the
ligand and then rearranges rapidly to the more stable S-bonded form. In a similar fash-
ion, a transient 0-bonded intermediate is detected in reactions of [ ( N H ~ ) ~ C O ( N O ~ ) ] ~ +
with [ c o ( c N ) ~ ] ~ ~ . ~ ~

TABLE 12-15
Rate Constants for Reactions
with [co(cN)~]~-
Oxidant k (M-' s-')

SOURCE: Data from J. P. Candlin, J. Halpern, and

S. Nakamura, J. Am. Chem. Soc., 1963,85.2517.
NOTE:aOuter-sphere mechanism caused by the
oxidant. Complcxcs with other potential bridging
groups (pod3-, sod2-,C O ~ ~and - , several
carboxylic acids) also react by an outer-sphere
mechanism, with constants ranging hum 5 X 10'
to 4 x 104.

Other reactions that follow an inner-sphere mechanism have been studied to de-
termine which ligands bridge best. The overall rate of reaction usually depends on the
first two steps (substitution and transfer of electron), and in some cases it is possible to
draw conclusions about the rates of the individual steps. For example, ligands that are
reducible provide better pathways, and their complexes are more quickly r e d ~ c e d . ' ~
Benzoic acid is difficult to reduce, but 4-carboxy-N-methylpyridine is relatively easy to

".I. Halpern and S. Nakamura, J. Am. Clzern. Soc., 1965, 87, 3002, J. L. Burmeister, Inorg. Chern.,
1964,3,919.
32~aube,Electron Transfer Reactions of Complex Ions in Solution, pp. 64-66; E. S. Gould and
H. Taube, J. Am. Chem. Soc., 1964,86, 13 18.
444 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

TABLE 12-16
Ligand Reducibility and Electron Transfer

[ ( N H ~ ) ~ c o L ]+
L
~ +[cI(H~o)~]*+ -
Rate constants for the reaction
co2+ + 5 NH3 + [ C I ( H ~ O ) ~ L ]+~ +
k ? (M-' sp')
H20
Comments

II
C6HsC -0
0.15 Benzoate is difficult to reduce
0
II
CH,C -0
0.34 Acetic acid is difficult to reduce
0
II
CH3NC,H4C -0
1.3 N-methyl-4-carboxypyridine is more reducible

3.1 Glyoxylate is easy to reduce

0
II
HOCH,C -0
7 x lo3 Glycolate is very easy to reduce

SOURCE: H. Taube, Electron Transfer Reactions of Complex Ions in Solution, Academic Press, New York,
1970, pp. 64-66.

reduce. The rate constants for the reaction of the corresponding pentammine Co(II1)
complexes of these two ligands with Cr(I1) differ by a factor of 10, although both have
similar structures and transition states (Table 12-16). For both ligands, the mechanism
is inner-sphere, with transfer of the ligand to chromium, indicating that coordination to
the Cr(I1) is through the carbonyl oxygen. The substitution reactions should have simi-
lar rates, so the difference in overall rates is a result of the transfer of electrons through
the ligand. The data of Table 12-16 show these effects and extend the data to glyoxylate
and glycolate, which are still more easily reduced. The transfer of an electron through
such ligands is very fast when compared with similar reactions with ligands that are not
reducible.
Remote attack on ligands with two potentially bonding groups is also found. Iso-
nicotinamide bonded through the pyridine nitrogen can react with cr2+ through the car-
bony1 oxygen on the other end of the molecule, transferring the ligand to the chromium
and an electron through the ligand from the chromium to the other metal. The rate con-
stants for different metals are shown in Table 12-17. The rate constants for the cobalt
pentammine and the chromium pentaaqua complexes are much closer than usual. The
rate for Co compounds with other bridging ligands is frequently as much as 10' larger
than the rate for corresponding Cr compounds, primarily because of the greater oxidiz- I,
ing power of Co(II1). With isonicotinamide compounds, the rate seems to depend more
on the rate of electron transfer from cr2' to the bridging ligand, and the readily re- ,

ducible isonicotinamide makes the two reactions more nearly equal in rate. The much ;
faster rate found for the ruthenium pentammine has been explained as the result of the ;
transfer of an electron through the .rr system of the ligand into the t2, levels of Ru(II1) :'1!
(low-spin Ru(II1) has a vacancy in the tZg level). A similar electron transfer to Co(1II) or
Cr(II1) places the incoming electron in the eg levels, which have o symmetry.33 '1
:j
1
3 3 ~Tauhe
. and E,. S. Gonld, Acc. Chem. Res., 1969,2, 321.
 12-8 Oxidation-Reduction Reactions 445
, ,
TABLEl2-17 - ,
, ,

Rate Copstants for deifuction of .tsClnicotit%[Link](4-Pyridine

Carboxylic'[Link]*)[Link] [ z ~ ( H ~ Q * ) ~ ~-,+
Oxidant k2 (M-' s-')
0

SOURCE:H. Taube, Electron Transfer Reactions of Complex Ions in Solution,

Academic Press, New York, 1970, pp. 66-68.

12-8-2 CONDITIONS FOR HIGH AND LOW

OXIDATION NUMBERS
The overall stability of complexes with different charges on the metal ion depends on
many factors, including LFSE, bonding energy of ligands, and redox properties of the
ligands. When other factors are more or less equal, the hard and soft character of the lig-
ands also has an effect. For example, all the very high oxidation numbers for the transi-
tion metals are found in combination with hard ligands, such as fluoride and oxide.
Examples include Mn04-, Cro4'-, and ~ e 0 4 with ~ - oxide, and AgF2, RuFS, PtF6, and
OsF6 with fluoride. At the other extreme, the lowest oxidation states are found with soft
ligands, with carbon monoxide being one of the most common. Zero is a common for-
mal oxidation state for carbonyls; V(CO)6, Cr(C0)6, Fe(CO)5, C O ~ ( C O )and ~,
Ni(C0)4 are examples. All these are stable enough for characterization in air, but some
react slowly with air or decompose easily to the metal and CO. Their structures and
reactions are explained further in Chapters 13 and 14.
Reactions of copper complexes show these ligand effects. Table 12-18 lists some
of these reactions and their electrode potentials. If the reactions of the aquated Cu(I1)

TABLE 12-18 ' . '

Electr~dePotentials of cab& and Cqppet $ekes . . '

Cu(II)-Cu(I)Reactions cFO( V )
cuZi(aq) + 2 C N + e- F== [ c ~ ( C N ) ~ ] - ( a q ) +1.103
cn2+(aq)+ I- + e- =CuI(s) I 0.86
cu2+(aq) C1- + + e ---' CuCl(s) +0.538
+
cu2+(aq) e +cu+(aq) t0.153
[ c u ( N H ~ ) ~ ]+
~ 'e- *
[CU(NH&]++ 2 NH3 -0.01
Cu(II)-Cu(0)Reactions 6"(V)
cu2+(aq)+ 2 e =Cu(s) +0.337
[ c u ( N H ~ ) ~ ] ~+' 2 e +Cu(s) + 4 NH3 -0.05
Co(III)-Co(I1)Reactions 8"(V)

SOURCE: T. Moeller, Inorganic Chemistry, Wiley-Interscience, New York,

1982, p. 742.
446 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

and Cu(1) are taken as the basis for comparison, it can be seen that complexing Cu(11)
with the hard ligand ammonia reduces the potential, stabilizing the higher oxidation
state as compared with either Cu(1) or Cu(0). On the other hand, the soft ligand cyanide
favors Cu(I), as do the halides (increased potentials). The halide cases are complicated
by precipitation, but still show the effects and also show that the soft iodide ligand
makes Cu(1) more stable than the harder chloride.
In other cases, almost any ligand can serve to stabilize a particular species, and
competing effects will have different results. Perhaps the most obvious example is the
Co(1II)-Co(I1) couple, mentioned earlier in Section 12-8-1. As the hydrated ion (or aqua
complex), Co(111) is a very strong oxidizing agent, reacting readily with water to form
oxygen and Co(11). However, when coordinated with any ligand other than water or flu-
oride, Co(1U) is kinetically stable, and almost stable in the thermodynamic sense as well.
Part of the explanation is that A, is quite large with any ligand, leading to an easy change
from the high-spin Co(lI) configuration tZg5eg2 to the low-spin Co(1II) configuration
[Link] means that the reverse reduction is much less favorable, and the complex ions
have little tendency to oxidize other species. The reduction potentials (Table 12-18) for
Co(111)-Co(11) with different ligands are in the order H 2 0 > NH3 > CN-, the order of
increasing A, and decreasing hardness. The increasing LFSE change is strong enough to
overcome the usual effect of softer ligands stabilizing lower oxidation states.

12-9 The reactions described to this point are either substitution reactions or oxidation-
REACTIONS OF reduction reactions. Other reactions are primarily those of the ligands; in these
COORDINATED reactions, coordination to the metal changes the ligand properties sufficiently to
change the rate of a reaction or to make possible a reaction that would otherwise not
LIGANDS
take place. Such reactions are important for many different types of compounds and
many different conditions. Chapter 14 describes such reactions for organometallic
compounds and Chapter 16 describes some reactions important in biochemistry. In this
chapter, we describe only a few examples of these reactions; the interested reader can
find many more examples in the references cited.
Organic chemists have long used inorganic compounds as reagents. For example,
Lewis acids such as AlC13, FeC13, SnC14, ZnC12, and SbCIS are used in Friedel-Crafts
electrophilic substitutions. The labile complexes formed by acyl or alkyl halides and
these Lewis acids create positively charged carbon atoms that can react readily with
aromatic compounds. The reactions are generally the same as without the metal salts,
but their use speeds the reaclions and makes them much more useful.
As usual, it is easier to study reactions of inert compounds, such as those of
Co(III), Cr(III), Pt(II), and Pt(IV), in which the products remain complexed to the metal
and can be isolated for more complete study. However, useful catalysis requires that the
products be easily separated from the catalyst, so relatively rapid dissociation from the
metal is a desirable feature. Although many of the reactions described here do not have
this capability, those with biological significance do, and chemists studying ligand
reactions for synthetic purposes try to incorporate it into their reactions.

12-9-1 HYDROLYSIS OF ESTERS, AMIDES,

A N D PEPTIDES

Amino acid esters, amides, and peptides can be hydrolyzed in basic solution, and the
addition of many different metal ions speeds the reactions. Labile complexes of Cu(II),
Co(II), Ni(U), Mn(II), Ca(II), and Mg(U), as well as other metal ions, promote the reac-
tions. Whether the mechanism is through bidentate coordination of the a-amino group
and the carbonyl, or only through the amine, is uncertain, but seems to depend on the
 12-9 Reactions of Coordinated Ligands 447
relative concentrations. Because the reactions depend on complex formation and hy-
drolysis as separate steps, their temperature dependence is complex, and interpretation
of all the effects is d i f i i c ~ l t . ~ ~
Co(111) complexes promote similar reactions. When four of the six octahedral po-
sitions are occupied by amine ligands and two cis positions are available for further re-
actions, it is possible to study not only the hydrolysis itself, but the steric preferences of
the complexes. In general, these compounds catalyze the hydrolysis of N-terminal
amino acids from peptides, and the amino acid that is removcd rcmains as part of the
complex. The reactions apparently proceed by coordination of the free amine to cobalt,
followed either by coordination of the carbonyl to cobalt and subsequent reaction with
OH- or H 2 0 from the solution (path A in Figure 12-15) or reaction of the carbonyl
carbon with coordinated hydroxide (path B ) . AS ~ ~a result, the N-terminal amino acid is
removed from the peptide and left as part of the cobalt complex in which the a-amino
nitrogen and the carbonyl oxygen are bonded to the cobalt. Esters and amides are also
hydrolyzed by the same mechanism, with the relative importance of the two pathways
dependent on the specific compounds used.
Other compounds such as phosphate esters, pyrophosphates, and amides of phos-
phoric acid, are hydrolyzed in similar reactions. Coordination may be through only one
oxygen of these phosphate compounds, but the overall effect is similar.

H
N- C-C-

H H O
I I I1
+ N-C-C-

FIGURE 12-15 Peptide Hydrolysis by [ ~ o ( t r i e n ) ( ~ ~ 0 ) ( 0 ~(Data . D. A. Buckingbam,

) ] ~ +from
J. P. Collman, D. A. R. Hopper, and L. G. Marzilli, J. Am. Chem. Soc., 1967, 89, 1082).

' 4 ~ . M. Jones, Ligand Reactivity and Catalysis, Academic Press, New York, 1968. Chapter 111
summarizes the arguments and mechanisms.
3 5 ~ .P. Collman and D. A. Buckingham, J: Am. Chern. Soc., 1963, 85, 3039; D. A. Buckingham,
J. P. Collman, D. A. R. Hopper, and L. G. Marzelli, J. Am. Chem. Soc., 1967,89. 1082.
448 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

12-9-2 TEMPLATE REACTIONS

Template reactions are those in which formation of a complex places the ligands in the
correct geometry for reaction. One of the earliest was for the formation of phthalocya-
nines, shown in Figure 12-16. Although the compounds were known earlier, their study
really began in 1928 after discovery of a dark blue impurity in phthalimide prepared by
reaction of phthalic anhydride with ammonia in an enameled vessel. This impurity was
later discovered to be the iron phthalocyanine complex, created from iron released into
the mixture by a break in the enamel surface. A similar reaction takes place with copper,
which forms more useful pigments. The intermediates shown in Figure 12-16 have been
isolated. Phthalic acid and ammonia first form phthalimide, then 1-keto-3-iminoisoin-
doline, and then 1-amino-3-iminoisoindolenine. The cyclization reaction then occurs,
probably with the assistance of the metal ion, which holds the chelated reactants in po-
sition. This is confirmed by the lack of cyclization in the absence of the metals.36Other
reagents can be used for this synthesis, but the essential feature of all these reactions is
the formation of the cyclic compound by coordination to a metal ion.
More recently, similar reactions have been used extensively in the formation of
macrocyclic compounds. Imine or Schiff base complexes (R N =CHR2) have been ex-
tensively studied. In this case, the compounds can be formed without complexation, but
the reaction is much faster in the presence of metal ions. An example is shown in Figure
12-17. In the absence of copper, benzothiazoline is favored in the final step rather than
the imine; very little of the Schiff base is present at equilibrium.
A major feature of template reactions is geometric; formation of the complex
brings the reactants into close proximity with the proper orientation for reaction. In ad-
dition, complexation may change the electronic structure sufficiently to promote the re-
action. Both are common to all coordinated ligand reactions, but the geometric factor is

Phthalic Phthalimide 1-Keto-3-imino- I-Amino-3-iminoiso-

anhydride isoindoline indolenine

FIGURE 12-16 Phthalocyanine

Synthesis. Cu(I1) phthalocyanine 1
:F

3 6 ~ Price,
. "Dyes and Pigments," in G. Wilkinson, R. D. Giliard, and J. A. McCleverty, eds.,
Comprehensive Coordination Chemistry, Vol. 6, Pergamon Press, Oxford, 1987, pp. 88-89.
 12-9 Reactions of Coordinated Ligands 449

Qk CHO
____j

Schiff base

FIGURE 12-17 Schiff Base Template Reaction. (a) The Ni(I1)-o-aminothiophenolcomplex reacts
with pyridine-l -carboxaldehyde to form the Schiff base complex. (b) In the absence of the metal ion,
the product is benzthiazoline; very little of the Schiff base is formed. (From L. F. Lindoy and
S. E. Livingstone, Inorg. Chetn., 1968, 7, 1149.)

more obvious in these; the final product has a structure determined by the coordination
geometry. Template reactions have been reviewed and a large number of reactions and
products d e ~ c r i b e d . ~ ~

12-9-3 ELECTROPHILIC SUBSTITUTION

Acetylacetone complexes are known to undergo a wide variety of reactions that are at
least superficially similar to aromatic electrophilic substitutions. Brornination, nitra-
tion, and similar reactions have been studied.38In all cases, coordination forces the lig-
and into an en01 form and promotes reaction at the center carbon by preventing reaction
at the oxygens and concentrating negative charge on carbon 3. Figure 12-18 shows the
reactions and a possible mechanism.

FIGURE 12-18 Electrophilic

Substitution on Acetylacetone
Complexes. X = C1, Br, SCN, SAr,
SCI, NO2, CH2C1, CH2N(CH3j2,
COR, CHO.

3 7 ~ St.
. C. Black, "Stoichiometric Reactions of Coordinated Ligands," in Wilkinson, Gillard, and
McCleverty, Comprehensive Coordination Chen~istry,pp. 155-226.
. Collman, A n g m Chem., Int. Ed., 1965, 4, 132.
3 8 ~P.
450 Chapter 12 Coordination Chemistry IV: Reactions and Mechanisms

C E N ERAL The general principles of kinetics and mechanisms have been described by J. W. Moore
REFERENCES and R. G. Pearson, Kinetics and Mechanism, 3rd ed., Wiley-Interscience, New York,
1981, and F. Wilkinson, Chemical Kinetics and Reaction Mechanisms, Van Nostrand-
Reinhold, New York, 1980. The classic for coordination compounds is F. Basolo and R.
G. Pearson, Mechanisms of Inorganic Reactions, 2nd ed., John Wiley & Sons, New
York, 1967. More recent books are by J. D. Atwood, Inorganic and Organometallic
Reaction Mechanisms, Brooks/Cole, Monterey, CA, 1985, and D. Katakis and G. Gordon,
Mechanisms of Inorganic Reactions, Wiley-Interscience, New York, 1987. The reviews
in G. Wilkinson, R. D. Gillard, and J. A. McCleverty, eds., Comprehensive Coordina-
tion Chemistry, Pergamon Press, Elmsford, New York, 1987, provide a more compre-
hensive collection and discussion of the data. Volume 1, Theory and Background,
covers substitution and redox reactions, and Volume 6, Applications, is particularly rich
in data on ligand reactions.

PROBLEMS 12-1 The high-spin d4 complex [c~(H~o)~]*+ is labile, but the low-spin d4 complex ion
[ c ~ ( c N ) ~ ] ~is- inert. Explain.

Why is the existence of a series of entering groups with different rate constants
evidence for an associative mechanism (A or I,)?

Predict whether these complexes would be labile or inert and explain your choices. The
magnetic moment is given in Bohr magnetons ( p ~after) each complex.

Ammonium oxopentachlorochromate(V) 1.82

Potassium hexaiodomanganate(1V) 3.82
Potassium hexacyanoferrate(II1) 2.40
Hexammineiron(I1) chloride 5.45

Consider the half-lives of substitution reactions of the pairs of complexes:

Half-Lives Shorter than 1Minute Half-Lives Longer than 1Day

IC~(CN),I~- [cr(cN),13-
[F~(HZO)~~+ [WCN)~I~-
IC~(H~O)~I~+ [ C O ( N H ~ ) ~ ( H ~ O( )H] 2~0+exchange)

Interpret the differences in half-lives in terms of the electronic structures.

The general rate law for substitution in square-planar Pt(1I) complexes is vadid for
the reaction

Design the experiments needed to verify this and to determine the rate constants. What
experimental data are needed, and how are the data to be treated?