CCL: conformational search for the realistic-size transition metal species

Dear Computational Chemistry List Members,

Many of you study the reaction mechanisms involving quite large (100-200 atoms) transition metal complexes (or organometallic species) with high degree of conformational freedom. Undoubtedly, the realistic Gibbs free energy profiles require knowledge of the most stable conformers for all the species involved (or to be even more strict – one would need to know the Boltzmann distribution of the conformers).

Could you share your experience/expertise in the conformational search of the realistic-size transition metal complexes with large number (10-20) of rotatable bonds? What software do you use? Is it free/commercial? Are you satisfied with it? Do you trust the results? Another similar question – do you find it useful to carry out the conformational search on the transition states?

My two cents:

Usually, when possible I use the X-Ray structures of the species I am interested in. Even though the most stable gas phase conformation might be different from the one in the solid state, still I believe the conformation from the X-Ray is more reliable than the one from the quick conformational search, especially when the large number of rotatable bonds is the case.

When the X-Ray structure is not accessible, and the number of reasonable conformers is not that large (5-10) – I try to optimize them at DFT level (since with the Density fitting algorithms it is often affordable).

I have tried to use HyperChem, OpenBabel, Tinker, PCModel for the systematic conformational search at molecular mechanics level, but I do not know how reliable the results of such searches are for the transition metal species.

Wrt to the conformational search on the transition state – I try to avoid it because if the conformation of the transition state is different from the conformation of the reactant, then another transition state associated with conformational change has to be found, and sometimes the transition state associated with the conformational change might be higher in energy than the original TS we were looking at.

Any comments, suggestions, constructive discussions are appreciated and deeply encouraged.

With kind regards,