DFT and QSAR Studies of Ethylene Polymerization by Zirconocene Catalysts

Abstract

A computational study of olefin polymerization has been performed on 51 zirconocene catalysts. The catalysts can be categorized into three classes according to the ligand framework: class I, Cp2ZrCl2 (10 catalysts), class II, CpIndZrCl2 (38 catalysts), and class III, Ind2ZrCl2 (3 catalysts), Ind = η5-indenyl. Detailed reaction pathways, including chain propagation and chain termination steps, are modeled for ethylene polymerization using zirconocene catalysts. Optimized structures for reaction coordinates indicated the presence of α-agostic interactions in the transition states (TSs) for both the first and second ethylene insertions, as well as in the ethylene π-complex of the ZrnPr cation. However, β-agostic interactions predominate in the cationic n-propyl and n-pentyl intermediates. The calculated activation energy barrier energies show that the TS for the insertion of ethylene into the Zr–CH3 + bond is the highest point on the computed reaction coordinates. Quantitative structure–activity relationship studies were also performed for 38-mixed zirconocene dichlorides. This study, in concert with the previous work, suggests that the type of ring attached to Zr (Cp vs Ind) affects the reaction kinetics and thermodynamics less significantly than the type of substituents attached to the Cp and indenyl rings and that substituent effects are even greater than those arising from changing the metal (Zr vs Hf).

Computational Discovery of Transition-metal Complexes: From High-throughput Screening to Machine Learning

“…Parveen et al investigated whether through-space geometric descriptors or through-bond electronic effects could better explain differences in 38 sandwich Zr complexes for polymerization . The developed MLR models were predictive of reactivity trends ( R 2 = 0.86) and performed nearly as well as more sophisticated (e.g., ANN) machine learning models .…”

“…Parveen et al investigated whether through-space geometric descriptors or through-bond electronic effects could better explain differences in 38 sandwich Zr complexes for polymerization . The developed MLR models were predictive of reactivity trends ( R 2 = 0.86) and performed nearly as well as more sophisticated (e.g., ANN) machine learning models . Jensen and co-workers employed a combination of cheminformatics-derived topological (see section ) and DFT-calculated descriptors to predict DFT-calculated estimates of activity for 82 ligands in Grubbs Ru olefin metathesis catalysts, where the DFT-calculated activity was also known to correlate to relative experimental activity .…”

Catalyst Activation and Chain Transfer Agents in Metallocene Catalysts for Ethylene Polymerization: A Computational Investigation

“…Nagaoka and coworkers [55] have reported the 1-octene polymerization using HfCat + À MeB(C 6 F 5 ) 3/4 À catalyst system where the monomer insertion rate was found to ΔG � = 17.6 and 17.9 kcal/ mol for the MeB(C 6 F 5 ) 3 À and B(C 6 F 5 ) 4 À cocatalysts. Parveen et al [56] have studied a range of metallocene catalysts for ethylene polymerization. The energy barriers (with respect to πcomplex) for the first ethylene monomer insertion are 12.3 � 1.7 kcal/mol, and for the second monomer, the range observed is 8.3 � 2.8 kcal/mol.…”

Experimental and Theoretical Study of Zirconocene-Catalyzed Oligomerization of 1-Octene

“…Zirconocenes hold a significant position among single-site catalysts of α-olefin polymerization due to high catalytic activity, excellent copolymerization homogeneity, and wide boundaries of regio- and sterecontrol [ 1 , 2 , 3 , 4 , 5 , 6 ]. Even in recent years, many of the studies of the reaction mechanisms [ 7 , 8 , 9 , 10 ] and structure–activity relationships [ 11 , 12 , 13 , 14 , 15 , 16 ] for these catalysts were performed using a generally accepted cationic concept ( Scheme 1 A) based on the fundamental research of Cossee and Arlman [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ].…”