Polyolefins, primarily polyethene (PE) and polypropene (PP), are among the most important end-products made from petrochemicals. PE can be made by radical polymerization or metal-catalyzed polymerization; for PP, only the metal-catalyzed route works.
The "classical" catalyst for both is the heterogeneous Ziegler-Natta catalyst (TiCl3, usually on a support, activated with Al alkyls). However, dramatic improvement of our understanding of the catalysis came only after the discovery of homogeneous "metallocene" catalysts (activated by methylalumoxane, MAO).
We now know that chain growth follows a migratory insertion path originally proposed by Cossee and Arlman (with later modification by Green and Rooney), that the dominant chain termination mechanism is usually chain transfer to monomer, and that stereocontrol (one of the key issues for PP) generally follows the "ligand orients chain, chain orients monomer" sequence first suggested by Corradini.
Even for homogeneous catalysts, there are still many issues that are not yet (fully) understood:
For the classical Ziegler-Natta catalysts, which to date remain dominant in industry, our understanding is much less detailed and even the precise nature of the active site(s) remains uncertain.
We use computational methods to address questions about details of polymerization catalysis, mainly in collaboration with the group of Vincenzo Busico. For relevant papers, see numbers 167-165, 163-161, 159 and 157 of my publication list.