"Oscillating" metallocene catalysts: what stops the oscillation?

The 150 MHz C-13 NMR microstructural analysis of polypropylene samples produced with two representative "oscillating" metallocene catalysts of largely different steric hindrance, namely [(2-(3,5-di-tert-butyl-4-methoxyphenyl)indenyl)(2)ZrP](+) and [(2-phenylindenyl)(2)ZrP](+) (P = polymeryl), and the implications on the origin of the stereocontrol are presented and discussed in detail. The original mechanistic proposal of an "oscillation" between a rac-like (isotactic-selective) and a meso-like (nonstereoselective) conformation cannot explain the observed polymer configuration. The isotactic-stereoblock nature of the polymers obtained with the former catalyst proves unambiguously that the active cation "oscillates" between the two enantiomorphous rac-like conformations at an average frequency that, even at high propene, concentration, is only slightly lower than that of monomer insertion. The less-hindered [(2-phenylindenyl)(2)ZrP](+) gives instead a largely stereoirregular polypropylene, which is the logical consequence of a faster ligand rotation; however, depending on the use conditions (in particular, on the nature of the cocatalyst and the polarity of the solvent), the polymerization products may also contain appreciable amounts of a fairly isotactic fraction. The peculiar microstructure of this fraction, with isotactic blocks of the same relative configuration spanned by very short atactic ones, rules out the possibility that the latter are due to an active species in meso-like conformation and points rather to a conformationally "locked" rac-like species with restricted ring mobility. The hypothesis of a stereorigidity induced by the proximity to a counteranion, which would play the role of the interannular bridge in the rac-bis(indenyl) ansa-metallocenes, was tested by computer modeling on a [rac-(2-phenylindenyl)(2)ZrMe(C3H6)][B(C6F5)(4)] ion couple and found viable.