No abstract

The presence of "free" trimethylaluminum (TMA) in methylalumoxane (MAO) solutions can be highly detrimental to the performance of metallocene and "post-metallocene" olefin polymerization catalysts. The most used strategy to remove "free" TMA is to evaporate MAO solutions to dryness, until a free-flowing white powder ("solid MAO") is left. This procedure is tedious and potentially hazardous, because in some cases the distillate is a concentrated hydrocarbon solution of TMA. Moreover, "solid MAO" is poorly soluble in common polymerization media, and once in solution it can regenerate TMA to some extent. This communication reports on a facile alternative, which consists in the controlled addition of a sterically hindered phenol, such as 2,6-di-tert-butylphenol, effectively trapping "free" TMA. We show here that 2,6-di-tert-butylphenol/MAO solutions activate equally well the dichloro-precursors of well-known zirconocene and bis(phenoxyimine)Ti catalysts, and that their use in propene polymerization results in a substantially higher productivity, polymer stereoregularity, and/or average molecular mass compared with activation by MAO alone.

Classical" MgCl 2 -supported Ziegler-Natta catalysts (ZNCs) continue to dominate the industrial production of isotactic polypropylene. There is a growing awareness of the inherent competitive edge of these low-cost systems over single-center (primarily metallocene) catalysts and of the potential for further improvement, particularly if deeper insight into the structure of the catalytic surfaces and the mechanisms of their modification by means of electron donors can be achieved. In the framework of a project ultimately aiming at the implementation of ZNCs with known and controlled surface structures, we are revisiting this whole area by using a combination of advanced computational (periodic DFT) and spectroscopic (high-resolution magicangle-spinning 1 H NMR spectroscopy) tools. In this article, we report on the neat MgCl 2 matrix and on model MgCl 2 /electron-donor adducts. The results indicate that the (104) surface, with five-coordinate Mg cations, is the dominant lateral termination in well-formed large crystals, as well as in highly activated MgCl 2 samples prepared by ball-milling. In the latter case, a minor fraction of surface Mg sites with a higher extent of coordinative unsaturation [e.g., four-coordinate Mg cations on (110) edges and/or at crystal corners or other defective locations] also appear to be present. RMe 2 Si(OMe) (R ) octadecyl) binds to both types of Mg sites, albeit with different strengths resulting in different mobilities. The less-electron-donating RMeSi-(OMe) 2 , in contrast, binds to the more unsaturated Mg sites only. The approach described herein is currently being extended to MgCl 2 /TiCl n systems, as well as to their adducts with internal and external donors of different natures, strengths, and steric demands.