Evaluation of PM3(tm) as a geometry generator in theoretical studies of transition-metal-based catalysts for polymerizing olefins

The PM3(tm) method has been applied to several systems of relevance to catalytic polymerization of olefins, for catalysts containing Ti, Zr or Cr. With some exceptions, PM3(tm) calculations reproduce experimental geometries of stable, closed-shell, precursors well. For stationary points along the path of monomer insertion into a metal-alkyl bond, the comparison is made to structures obtained by optimization using various first-principle methods. Large errors are uncovered for the transient structures, in particular pertaining to metal-ethylene coordination and agostic interactions. The energy profiles for four insertion reactions are computed by gradient-corrected density functional (DFTG) methods, using molecular structures taken from PM3(tm) and first-principle geometry optimizations, respectively. The chromium case is promising, giving values for the barrier to monomer insertion of 11 and 9 kcal/mol based on PM3(tm) and DFTG geometries, respectively. The Ti- and Zr-based systems are predicted to proceed downhill based on PM3(tm) structures, whereas small barriers are found when using first-principle structures. A hybrid PM3(tm)-DFTG procedure is suggested for geometry optimization, which facilitates an accurate estimate of the barrier when applied to one of the zirconium systems.