Predictive modeling of electrocatalyst structure based on structure-to-property correlations of x-ray photoelectron spectroscopic and electrochemical measurements.

Chemical structure and catalytic activity of nonplatinum porphyrin-based electrocatalyst for oxygen reduction is characterized by combination of X-ray photoelectron spectroscopy (XPS) and rotating disk electrode. The goal of the study is to show how modifications in the molecular structure affect catalytic characteristics and how to use these structural modifications in a purposeful manner to increase catalytic activity. Initial correlation of structure to electrochemical performance is achieved through the application of principal component analysis (PCA) to curve-fits of high-resolution XPS spectra combined with results of electrochemical measurements. Furthermore, a predictive model that describes this correlation is build using the combination of genetic algorithm (GA) and multiple linear regression (MLR). Based on structure-to-property correlations, two types of active sites responsible for the catalytic activity, i.e., Co associated with pyropolymer and Co particles covered by oxide layer, are determined, and a dual-site for oxygen reduction on cobalt porphyrins is hypothesized, allowing for designing a catalyst structure with optimal performance characteristics.