Chemistry Reference
In-Depth Information
4
Shape
4.1
Getting in Shape
Coordination complexes adopt a limited number of basic shapes. We have developed in
Chapter 3 a predictive set of molecular shapes evolving from an electrostatic model of the
distributions predicted for from two to six point charges dispersed on a spherical surface.
These shapes, evolving from a modification of the valence shell electron pair repulsion
(VSEPR) model that was itself developed initially for main group compounds, are satisfac-
tory as models for many of the basic shapes met experimentally for complexes throughout
the Periodic Table. The original VSEPR model and its electron counting rules have limited
predictive value for shape in complexes of d-block elements compared with its application
for p-block elements. This relates to the defined directional properties of lone pairs in
p-block elements, whereas in transition elements nonbonding electrons play a much re-
duced role in defining shape. Rather, it is simply the number of donor groups bound about
the metal that is the key to shape in transition metal complexes. This is recognized in the
Kepert model, which is a variation of the VSEPR concept developed for transition elements
that ignores nonbonding electrons and considers only the set of donor groups represented
as point charges on a surface. This essentially electrostatic model has limitations, as shape
is influenced by other factors such as inherent ligand shape and steric interactions between
ligands, as well as the size and valence electron set of the central metal ion.
The actual shape of complexes is now able to be determined readily in many cases.
The advent of X-ray crystallography at a level where highly automated instruments allow
rapid determination of accurate and absolute three-dimensional structures of coordination
complexes in the crystalline form has been a boon for the chemist. Provided a complex
can be crystallized, its structure in the solid state can be accurately defined. Although it is
important to recognize that, for a coordination complex, solid state structure and structure
in solution can differ, it is nevertheless true that they often are essentially the same, so we
have at our fingertips an exceptionally fine method for structural characterization. This is a
technique that can define angles with an error approaching 0.01
◦
and bond distances (which
are typically between 100 and 300 pm) with an error as low as 0.1 pm. It is least successful
at detecting the very light hydrogen atoms, although the closely related neutron diffraction
method provides greater resolution of these. An example structure is shown in Figure 4.1,
in which all atom locations were accurately defined except for hydrogen atoms, which have
been placed at calculated positions.
What crystallography has now shown clearly is that the predicted shapes we developed
earlier are often observed but usually not achieved ideally - we notice that bond angles are
often not exactly those anticipated, and, at times, geometries occur that are clearly better
