LAUSR

Abstract Based on the quantum chemical descriptors and the activities for isoprene polymerization of 12 neodymium carboxylates obtained by authors' earlier work, statistical analysis of data was made and a QSAR model correlating the quantum chemical descriptors and the activity was built with the partial least square (PLS) approach. The model is: A = 51,602 eHOMO + 6 σ + 12,546, which indicates that the catalytic activity A is positively correlated with the HOMO energy eHOMO and the ligand polarizability σ, with the contribution of σ being larger, eHOMO smaller. The model's coefficient of determination r2 = 0.96 and that of cross validation q2 = 0.94, both being close to 1, which means that its quality is well and its predictive power is strong. Analysis of the modeling process and the resulted QSAR model, together with the interpretation of the model's mechanism, also shows that the model obtained from this study is valid and reliable. According to the QSAR model, the mechanism of catalyst activity can be interpreted as that neodymium carboxylates with higher HOMO energy and larger ligand polarizability are easier to react with the co-catalysts so as to produce more active and stable centers of catalyst, resulting in a higher activity.