The Active Site of Nitrogenase FeMo-Cofactor

The X-ray structure of the resting state of FeMoco exhibits a central iron prism with three p-S atoms on top of the three vertical edges of the prism. On top of the two triangular faces of the Fe6 prism we find another Fe and a Mo atom so that two tetrahedral structures arise. The triangular faces of these tetrahedra are each capped with three additional sulfur atoms (Fig. 1). 

The FeMoco can be extracted intact (20) from the acid-denatured MoFe protein, which has the important consequence that modeling of its function may start with studies on the structure of the isolated cluster. The protein environment (and also solvent molecules) may be neglected in the first modeling step though they are essential for the reduction mechanism. This simplification appears to be mandatory since the electronic structure of the isolated cluster is extremely complicated. 

To arrive at significant insight for the biological reduction process, all theoretical studies should be taken into account since none of them is free of model-inherent approximations. We emphasize that none of the recent studies, which utilized DFT methods, discussed the theoretical complications with standard density functionals arising from different spin states, which are discussed in the Appendix. The authors thus have assumed that the dependence of the results on the chosen density functional is of little importance in 

Pioneering studies with semi-empirical methods were carried out shortly after the discovery of the FeMoco structure via X-ray diffraction experiments (21,22). These extended-Hiickel calculations have been carried out for structurally fixed models of FeMoco (23,24) and of FeMoco, FeVco, and FeFeco (25), in which the three amino acid ligands have been replaced by hydride ligands. The main results of these studies are deduced based on the assumption that the character of the frontier orbitals, i.e., their composition in terms of atomic orbitals, 

In view of the many approximations in these semi-empirical calculations their results can only be regarded as first hints for more elaborate methods. Such are DFT calculations on FeMoco-type clusters and model complexes. As it is discussed, for instance, in Ref. (31), N2 usually binds very weakly to transition metal complexes. It is thus not certain whether end-on coordination to one or two iron centers can lead to sufficiently strong coordination of dinitrogen. 

---