Multivariate Linear Regression Models to Predict Monomer Poisoning Effect in Ethylene/Polar Monomer Copolymerization Catalyzed by Late Transition Metals

Abstract

This study combined density functional theory (DFT) calculations and multivariate linear regression (MLR) to analyze the monomer poisoning effect in ethylene/polar monomer copolymerization catalyzed by the Brookhart-type catalysts. The calculation results showed that the poisoning effect of polar monomers with relatively electron-deficient functional groups is weaker, such as ethers, and halogens. On the contrary, polar monomers with electron-rich functional groups (carbonyl, carboxyl, and acyl groups) exert a stronger poisoning effect. In addition, three descriptors that significantly affect the poisoning effect have been proposed on the basis of the multiple linear regression model, viz., the chemical shift of the vinyl carbon atom and heteroatom of polar monomer as well as the metal-X distance in the σ-coordination structure. It is expected that these models could guide the development of efficient catalytic copolymerization system in this field.