화학공학소재연구정보센터
Macromolecules, Vol.45, No.17, 6781-6795, 2012
Experimental and Computational Investigation of the Mechanism of Carbon Dioxide/Cyclohexene Oxide Copolymerization Using a Dizinc Catalyst
A detailed study of the mechanism by which a dizinc catalyst copolymerizes cyclohexene oxide and carbon dioxide is presented. The catalyst, previously published by Williams et al. (Angew. Chem. Int. Ed. 2009, 48, 931), shows high activity under just 1 bar pressure of CO2. This work applies in situ attenuated total reflectance infrared spectroscopy (ATR-FTIR) to study changes to the catalyst structure on reaction with cyclohexene oxide and, subsequently, with carbon dioxide. A computational investigation, using DFT with solvation corrections, is used to calculate the relative free energies for various transition states and intermediates in the cycle for alternating copolymerization catalyzed by this dinuclear complex. Two potentially competing side reactions, sequential epoxide enchainment and sequential carbon dioxide enchainment are also investigated. The two side-reactions are shown to be thermodynamically disfavored, rationalizing the high selectivity exhibited in experimental studies using 1. Furthermore, the DFT calculations show that the rate-determining step is the nucleophilic attack of the coordinated epoxide molecule by the zinc-bound carbonate group in line with previous experimental findings (Delta Delta G(353) = 23.5 kcal/mol; Delta G(353)double dagger = 25.7 kcal/mol). Both in situ spectroscopy and DFT calculations indicate that just one polymer chain is initiated per dizinc catalyst molecule. The catalyst adopts a "bowl" shape conformation, whereby the acetate group coordinated on the concave face is a spectator ligand while that coordinated on the convex face is the initiating group. The spectator carboxylate group plays an important role in the catalytic cycle, counter-balancing chain growth on the opposite face. The DFT was used to predict the activities of two new catalysts, good agreement between experimental turn-over-numbers and DFT predictions were observed.