Testing DFT ability to predict the stereoselectivity of group 4 metallocenes in propylene polymerization

In this study we have tested the ability of a standard DFT computational protocol to reproduce the experimentally obtained stereoselectivity of 26 different C2-symmetric zirconocene catalysts active in propylene polymerization. The catalysts were chosen for their relevance in metallocene catalyzed polymerization of propylene. To this end, primary insertion of both si- and re-propylene enantiofaces into the Zr-CH2-CH(CH3)2 bond was considered to simulate the growing chains step. The energy difference between these two transition states, ΔEre-si, was taken as a measure of the stereoselectivity (pentad: mmmm%) of different catalysts. The results clearly indicated that there was a good agreement between ΔEre-si and the mmmm% values, so that greater ΔEre-si could correspond to higher mmmm%. A model was fitted to the experimentally obtained mmmm% against theoretical ΔEre-si. The coefficient of determination (R2) of the resultant plot was 0.9793, which indicated a good accuracy of the model. Finally, to quantify the steric role of the studied ligands in the observed stereoselectivity, the analysis of the buried volume (VBur) and of the steric maps was performed for two representative complexes. The images revealed that a greater asymmetric localization of the %VBur around the metal center led to a higher mmmm% in the resultant polymer.