Machine learning approaches towards connecting catalyst performance, reactor conditions and polymer properties

N. Lambic
United States

Keywords: machine learning, polymer properties, polyolefin


The presentation will highlight combined application of high-throughput experimentation and machine learning (ML) across multiple platforms, including catalyst development, process condition optimization and polyolefin product performance. The combination of modelling tools across such broad spectrum (catalyst, product and process) highlights the data driven approach towards design and implementation of improved systems. In order to illustrate this approach, examples of polypropylene catalyst development, as well as optimized preparation and performance properties of high-melt strength polypropylene will be provided.