Some properties of transition metal compounds can be influenced by static correlation, relativistic effects, and, in some cases, close-lying electronic states could result in difficulties in SCF convergence. All these complications influence geometry optimizations and, hence, the accuracy of the obtained results.
Considering a usual Kohn-Sham DFT calculation of a TM compound with organic ligands:
- What functionals gives accurate equilibrium geometries, compared to high-level calculations or experimental ones?
- What functionals are accurate for electronic and bonding properties?
I'm not asking for a specific functional, but some guidelines on which are better for those purposes.