Descriptor–property relationships in heterogeneous catalysis: exploiting synergies between statistics and fundamental kinetic modelling

Abstract

Kinetics-driven design of heterogeneous catalysts is a promising methodology to optimize material performances for a reaction of interest. However, most of the fundamental kinetic models suffer from a limited applicability due to ill-defined relationships between catalyst features which impact the kinetics, i.e. catalyst descriptors, and properties determined from material characterization. In order to overcome this limitation, a comprehensive methodology is proposed, combining kinetic simulations with a selection of statistical tools. The aim is to identify similarities between experimental data and simulated performances, and to assess the significance of the descriptor–property relationships that can be established. Oxidative coupling of methane (OCM) was selected as a case study to demonstrate this methodology. The obtained qualitative relationships indicated that the electronic properties of surface oxygen species are key in the optimization of OCM catalysts. This proof of concept highlights the proposed methodology as a tool for catalyst design for a broad variety of reactions, provided that a kinetic model and (a limited amount of) experimental data are available.