Molecular modelling for transition metal complexes: Dealing with d-electron effects

The development and application of molecular mechanics methods which include an explicit treatment of d-electron effects are reviewed. The origins of the authors' ligand field molecular mechanics (LFMM) method are traced from Hitchman's simple expression for predicting the M-L bond length change accompanying a d-d transition through to the general implementation of LFMM within the molecular operating environment, including analytical energy gradients, explicit M-L pi-bonding and d-s mixing effects, plus a term for handling different spin states. The LFMM applications cover simple coordination complexes of Cu(II) where a single parameter set treats multiple coordination numbers; the treatment of the complete Jahn-Teller effect in six-coordinate d(9) Cu(II) complexes including elongated and compressed structures and the intervening barrier height: the parameterisation of Co(III) and Ni(II) species to enable both high- and low-spin complexes to be treated with a single parameter set: the unusual and highly distorted Type I centre in 'blue copper' enzymes; and dinuclear species containing the [Cu2O2](2+) moiety found in Type 3 copper enzymes and the important Cu-O pi-bonding effects. The LFMM has also spawned two approaches, the ligand field extension of the sum of interactions between fragments ab initio (SIBFA) model and an extension to the general utility lattice program (GULP). The former has been applied to some simple Cu(II) species while the latter has been applied to d(4) Mn(III)-oxides. Both are limited to sigma effects only. In addition, the related effective crystal field model of Tschougreeff is discussed and compared with the LFMM approach. Providing one is prepared to develop the necessary parameters, the LFMM and related approaches are capable of delivering DFT-quality results but up to four orders of magnitude faster. (C) 2008 Elsevier B.V. All rights reserved.