This review discusses modifications and extensions of Molecular Mechanics w
hich are designed to model the electronic effects of the valence d and s el
ectrons in transition metal compounds. These effects lead to severe distort
ions away from the ideal geometries predicted by simple VSEPR theory. The s
tereochemical activity of d electrons manifests in a range of structural di
stortions of ionic coordination complexes typified by the Jahn-Teller elong
ations of six-coordinate d(9) Cu2+ species. Modelling these effects require
s an additional term in the strain energy which describes the attendant lig
and field stabilisation energy (LFSE). The LFSE is explicitly incorporated
into the ligand field molecular mechanics (LFMM) method which has been appl
ied to a range of complexes of Cu2+, Ni2+ and Co3+. A single set of LFMM pa
rameters for a given metal-ligand interaction is able to model different co
ordination numbers, spin states and bond lengths. The stereochemical activi
ty of the valence metal s orbital is significant for covalent organometalli
c species such as WMe6 which does not show the regular octahedral geometry
expected for a formally d(0) system. The effect can be treated within Valen
ce Bond theory by modifying the expressions for the angle bending potential
s based on Pauling's strength functions for sd(n) hybrids. There is more th
an one idealised bond angle for sd(3), sd(4) and sd(5) hybrids which correl
ates with the irregular geometries found for hydride, alkyl and aryl compou
nds. The same behaviour can also be obtained within the LFMM scheme by usin
g an extended stabilisation energy term which incorporates the s orbital co
ntributions. The LFMM model also predicts the ligand field contribution to
the activation energy for ligand exchange/substitution and can be used to c
alculate the structures and energies of transition states as illustrated by
model calculations for the reactions of low-spin d(8) complexes. (C) 2001
Elsevier Science B.V. All rights reserved.