Abstract
A combined DFT/MM analysis has been carried out on the chain propagation steps and on possible mechanisms for the formation of unsaturations in propene polymerization catalyzed by rac-Me2C(3-t-Bu-1-Ind)(2)ZrCl2/methylalumoxane (1/MAO). The results are compared to the available experimental data on its polymerization performance. The insertion of the si propene enantioface at the R,R site is favored by 4.0-6.0 kcal/mol, in good agreement with the (slightly underestimated) experimental value of 4.6 kcal/mol. Mechanistic aspects related to highly hindered catalysts are also discussed. The large amount of allyl end groups measured in i-PP produced at any [propene] can be rationalized, especially at the highest monomer concentration, by allylic activation of a coordinated propene; allyl end groups formed by this route add to those formed by the unimolecular B-Me transfer reaction. Chain release mechanisms involving a coordinated propene (allylic activation and P-H transfer) kinetically compete. The relatively high rate of chain-end epimerization observed for this catalyst, as well as the presence of internal vinylidene groups, can be rationalized by the relatively high stability of the product of beta -H transfer to the metal due to a H2C-H...Zr agostic interaction. Allylic activation of the growing chain end followed by propene insertion is the source of internal vinylidene unsaturations, but likely not of epimerization. Calculations support the hypothesis that the allylic chain end activation could also occur upon beta -H transfer to a coordinated monomer; hence, the formation of internal vinylidenes could also proceed without developing H-2. Finally, we conclude that within the single-center, two-state catalyst model developed to account for the nonlinear activity/[hl] relationship, the fast propagating site is the Zr-CH2CH(CH3)P, while the Zr-C(CH3)(2)P (tertiary alkyl) formed during the course of chain epimerization, proposed by Busico, represents a suitable model for the catalyst resting state.