Abstract: Metalloporphyrin complexes have attracted much attention because of their excellent catalytic performance and good selectivity in oxygen reduction reactions (ORR). However, massive investigations on substituent effects with conventional experimental synthetic approaches as well as high accuracy quantum chemistry computations would be difficult. Herein, quantitative structure-activity relationship (QSAR) methods were utilized. Based on collected physicochemical parameters, calculated topological parameters and quantum-chemical parameters of cobalt porphyrins with various substituents by Chem3D and Gaussian, the correlation analysis was performed by SPSS program, which enables quick screening of feature descriptors on dioxygen activation mediated by cobalt porphyrins. Subsequently, a stepwise regression analysis was performed, which outputs a multi-variant regression equation with acceptable goodness-of-fit and generalization ability. Our work systematically demonstrated the calculation and collection of representative descriptors in QSAR, and the detailed process of correlation analysis and regression analysis were shown. This work can assist in self-learning the relevant functions of Chem3D, Gaussian, and SPSS software, enhancing practical abilities in molecular modeling, computational chemistry and statistical software, as well as data analysis and processing skills.

Key words: Cobalt porphyrin, Oxygen binding energy, QSAR, Correlation analysis, Stepwise regression analysis