A generalized model for stereoselectivity of C2-symmetric olefin polymerization catalysts

We have recently demonstrated that enantioselectivity models using only a few intuitively meaningful descriptors based on the “buried volume” idea require appropriate dissection of the sphere that is used to calculate said volume.1 For propene polymerization catalysis to isotactic polypropylene (iPP) only a single steric descriptor derived from the comparison of hindrance in differently occupied octants (Δ%VBO) is needed, and electronic effects are unimportant. Moreover, the model (mean absolute deviation, MAD, 0.12 kcal/ mol) works for more than a single catalyst class (metallocenes and Salans), allowing in silico catalyst design. The high accuracy is directly tied to the quality of the underlying high-throughput experimentation (HTE) dataset. However, while the low dimensionality of the descriptor space and their obvious intuitive meaning naturally provide guidelines for further catalyst optimization, some catalyst classes did not fit the correlation. For instance, an individual model was required to describe [OOOO]-catalysts accurately. Here, we describe efforts to overcome these limitations introducing three additional assumptions: (a) enantiomorphic site control has to overcome chain end control, (b) an empty quadrant can be “too empty” and (c) catalyst flexibility has to be correctly accounted for. Indeed, a truly globally applicable model results that is able to describe four different catalyst families (zirconocenes, Salan-, [OOOO]- and Mitsui type catalysts). This research forms part of the research programme of DPI, project #835.