The coordinate bond 29

from repulsion between electrons in a metal ion and electrons of ligands in the
hypothetical complex (Figure 2.8).
Now if the rigid shell is allowed instead to concentrate its total force on six
particular spots (for example, the corners of an octahedron), then the sponge is
pressed inward at these positions but bulges outward between them. Compared
with the spherically constricted system, the sponge at the six points of high
pressure is at a higher energy and at the bulges between is at a lower energy.
This corresponds to crystal field splitting with bulges related to t2g orbitals and
points of depression related to eg orbitals.
In the preceding discussion, it was noted that the energy of the d orbitals of a
metal ion increases when ligands approach the ion. This, in itself, suggests that a
complex should be less stable than a free metal ion. However, the fact that
complexes do form indicates that the complex is a lower energy configuration
than the separated metal ion and ligands. The increase in energy of the d orbitals
of the metal ion is more than compensated for by the bonding between metal ion
and ligand. In the case of [TiF6]2+, bonding can be visualized as the result of
electrostatic attraction between negative fluoride ions and the positive titanium
ion.
In an octahedral arrangement of ligands, the t2g and eg sets of d orbitals have
different energies. The energy separation between them is given the symbol ǻo.

Y
It can be proved for an octahedral system that the energy of t2g orbitals is 0.4ǻo

R
less than that of the five hypothetical degenerate d orbitals that result if crystal

T
field splitting is neglected (Figure 2.8). Therefore, the eg orbitals are 0.6ǻo higher

IS be
in energy than the hypothetical degenerate orbitals.
In an octahedral complex that contains one d electron (for example,
[Ti(H2O)6]3+) that electron will reside in the d orbital of lowest energy. Simple

EM cri
electrostatics does not recognize that d orbitals in a complex have different
energies. Therefore, theory predicts that the d electron would have the energy of
the hypothetical degenerate d orbitals. In fact, the d electron goes into a t2g

H
C bs
orbital that has an energy 0.4ǻo less than that of the hypothetical degenerate
orbitals. Thus the complex will be 0.4ǻo more stable than the simple

u
electrostatic model predicts. In simple terms, we can say that the d electron, and

M
hence the whole complex, has a lower energy as a result of the placement of the

A S
electron in a t2g d orbital that is as far from the ligands as possible. The 0.4ǻo is

e
called the crystal field stabilization energy (CFSE) for the complex.

R r
Table 2.1 gives CFSE’s for metal ions in octahedral complexes. The values in

C ha
Table 2.1 are readily calculated by assigning a value of 0.4ǻo for each electron
put in a t2g orbital and a value of –0.6ǻo for each electron put in an eg orbital.

O
Thus, the CFSE for a d5 system is either 3(0.4ǻo) + 2(–0.6ǻo) = 0.0ǻo or 5(0.4

S
ǻo) + 0(–0.6ǻo) = 2.0ǻo, depending on the distribution of the five electrons in the

R e
t2g and eg orbitals.

E
Simple electrostatics treats a metal ion as a spherical electron cloud

Z Li k
surrounding an atomic nucleus. CFT provides a better model, since it admits
that d electrons provide a nonspherical electron cloud in order to avoid positions
in which ligands reside. (They provide a nonspherical electron cloud
 30 Coordination chemistry

Table 2.1 Crystal field stabilization energies for metal ions in octahedral complexes
d electrons
in metal Stabilization, Stabilization,
t
ions 2g eg ǻo t2g eg ǻo
1 0.4
2 0.8
3 1.2
4 0.6 1.6
5 0.0 2.0
6 0.4 2.4
7 0.8 1.8
8 1.2
9 0.6
10 0.0

by residing preferentially in low-energy orbitals that point between ligands.)

Y
Therefore, CFT explains why simple electrostatic calculations consistently

R
underestimate the stability of transition-metal complexes and compounds; the
simple approach neglects the nonspherical electron distribution and the

T
resulting CFSE.

IS be
An early objection to a simple electrostatic treatment of bonding for metal
complexes was that it could not explain the formation of square planar

EM cri
complexes. It was argued that if four negative charges are held to a positive
central ion by electrostatic forces alone, then the negative charges must be at
corners of a tetrahedron. Only in such a structure can negative groups attain

H
C bs
maximum separation and so experience a minimum electrostatic repulsion. This
is correct if the central ion is spherically symmetric. However, such symmetry is
not typical of transition-metal ions, because electrons will reside in low-energy

M u
orbitals that point between ligands and do not have spherical symmetry. In

S
Section 3.1 it is shown that CFT can account for square planar complexes and

A
that it predicts the distortion of certain octahedral complexes.

R r e
We have considered the crystal field splitting for octahedral complexes; now

C ha
let us consider complexes of other geometries. It is convenient to start with the
crystal field splitting for an octahedral structure and consider how the splitting

O
is affected by a change in geometry (Figure 2.10). In going from a regular

S
octahedron to a square planar structure, the effect amounts to removal of any

R
two trans ligands from the octahedron. Generally, we speak of the xy plane as

E e
the square plane, which means trans groups are removed from the z axis.

Z Li k
If ligands on the z axis are moved out so that the metal–ligand distance is
only slightly greater than it is for the four ligands in the xy plane, the result is a
tetragonal structure (Figure 2.10). This permits ligands in the xy plane to
 The coordinate bond 31

Figure 2.10 Crystal field splittings of d orbitals of a central ion in complexes

R Y
approach the central ion more closely. Consequently, d orbitals in the xy plane

T
experience a greater repulsion from ligands than they do in an octahedral

IS be
structure, and we find an increase in energy of the dx2–y2 and dxy orbitals (Figure
2.10). At the same time d orbitals along the z axis or in the xz and yz plane
experience a smaller repulsion from ligands, which are now some distance

EM cri
removed along the z axis. This results in a sizable decrease in energy for dz2 orbital
and a slight decrease for dxz and dyz orbitals, relative to the octahedral

H
arrangement.

C bs
The same splitting pattern is found for a square pyramidal structure, in which
there is one ligand on the z axis and the other four ligands plus the central atom

u
are in the xy plane. The complete removal of two ligands on the z axis to give a

M S
square planar configuration is then accompanied by a further increase in energy

A
of the dx2–y2 and dxy orbitals, as well as a further decrease for the dz2, dxz, and dyz

e
orbitals.

R r
The crystal field splitting of d orbitals for a tetrahedral structure is more

C ha
difficult to visualize. We must first try to picture a tetrahedron placed inside a
cube (Figure 2.11). The four corners of the tetrahedron are then located at four of

O S
the corners of the cube. If we now insert x, y, and z axes so they go through the

R
center of the cube and protrude from the centers of its six faces, we can begin to

e
see the position of the four ligands with respect to the d orbitals of the central

E k
atom. The d orbitals along the cartesian axes (dx2–y2 and dz2) are further removed

Z Li
from the four ligands than are orbitals between the axes (dxy, dxz, and dyz).
Therefore, eg orbitals (dx2–y2 and dz2) are the low-energy d orbitals in
 32 Coordination chemistry

Figure 2.11 A tetrahedral complex with its center at the center of a cube.

tetrahedral complexes; t2g orbitals (dxy, dxz, dyz) are of relatively higher energy. It

Y
has been observed that the energy separation between eg and t2g orbitals, the
crystal field splitting ǻt, is only about one-half of ǻo. Hence, crystal field effects

R
favor the formation of octahedral complexes over that of tetrahedral complexes.

T
IS be
Magnetic properties

EM cri
The magnetic properties of transition-metal complexes can readily be
understood in terms of CFT. Transition metal ions have partially filled d orbitals.
If these orbitals are degenerate, Hund’s rule predicts that unpaired electrons will

H
be present. For example, a metal ion containing three d electrons (called a d3

C bs
system) should have three unpaired electrons a d8 metal ion
should have two unpaired electrons and three pairs of electrons

u
). Materials that contain unpaired electrons are attraced to a

M S
magnet and are said to be paramagnetic. (This attraction is much weaker than

A
that exhibited by ferromagnetic materials such as iron.) The magnitude of

R e
attraction of a material to a magnet is a measure of the number of unpaired

r
electrons present.

C ha
Paramagnetism can be measured with a relatively simple device called a Gouy
balance. The sample is placed in a tube suspended from a balance, and the

O S
weight of the sample is measured both in the presence and in the absence of a

R
magnetic field. If the material is paramagnetic, it will weigh more while the

E e
magnetic field is present and attracting it. The increase in weight is a measure of

Z Li k
the number of unpaired electrons in the compound.
A gaseous Co3+ ion, a d6 system, has five degenerate d orbitals and is expected
to have four unpaired electrons. However, some d6 cobalt(III)