Catalyst Optimization in Olefin Polymerization using an HTE/QSAR workflow: Intuitive 3D-Steric Descriptors to Tackle the Database Size Problem

Catalyst screening in olefin polymerization is an inherently complex problem. Even using high-throughput experimentation (HTE) tools, databases are usually tiny (at least when compared to those in pharma), rarely comprised of more than several dozen catalysts. Complex catalyst synthesis, the need to conduct the polymerization under inert atmosphere, time-consuming polymer analysis and the need to screen the catalysts under various conditions limit the throughput. Extracting reliable information from such relatively small databases and the construction of quantitative structure-activity (QSAR) models for catalyst development face serious hurdles. We have tackled the problem via the introduction of chemically meaningful 3D-steric descriptors – based on Cavallo’s buried volume descriptor – that map the distribution of steric bulk within the active pocket of the catalyst. On the specific example of ansa-zirconocenes for isotactic polypropylene we will be showcasing the development of simple QSAR models that allow for chemical interpretation of the results.1-2 Using these models and additional chemical insight garnered from the screenings, we have developed novel ansa-zirconocenes showing superior performance in all aspects over known literature systems. The new catalysts deliver high-molar mass/high melting point iPP at high polymerization temperatures (120°C).3 In essence, this talk will demonstrate that high-throughput catalyst design could accomplish in 4 years and at a fraction of the cost what the combined effort of industry and academia and billions of dollars could not accomplish in the last 40 years. The possible extension of this approach to cover multiple catalyst classes will also be discussed.