Development of ab-initio structure determination program for primary particles of Ziegier-Natta catalysts using machine learning

摘要

The smallest structure unit in MgCl2-supported Ziegler-Natta catalysts is called primary particles. They are regarded as MgCl2 nanoparticles or nanoplates whose lateral surfaces are coordinatively terminated by TiCl4 and donors. Behind the expected significance on the primary structure distribution of resultant polymers, the determination of the structure of the primary particles has not yet been achieved. The disordered nature of the primary particles encompasses an experimental difficulty, while huge compositional and configurational spaces have hindered computational modeling based on first-principle calculations.