Abstract
The identification of key materials’ parameters that correlate with the catalytic performance can accelerate the development of improved heterogeneous catalysts and unveil the relevant underlying physical processes. However, the analysis of correlations is often hindered by inconsistent data. Besides, nontrivial, yet unknown relationships may be important, and the intricacy of the various processes may be significant. Here, we tackle these challenges for the CO oxidation reaction catalyzed by perovskites via a combination of rigorous experiments and artificial intelligence (AI). A series of 13 ABO3 (A = La, Pr, Nd, Sm; B = Cr, Mn, Fe, Co) perovskites was synthesized, characterized, and tested in catalysis. To the resulting dataset, we applied the symbolic-regression SISSO approach. We identified an analytical expression as a descriptor for the activity that contains, as key materials’ parameters, the normalized unit cell volume, the Pauling electronegativity of the elements A and B, and the ionization energy of the element B. Therefore, the activity is described by crystallographic distortions and by the chemical nature of A and B elements. The generalizability of the identified descriptor is confirmed by the good overall quality of the predictions for the activity of 3 additional ABO3 and of 16 chemically more complex AMn(1-x)B’xO3 (A = La, Pr, Nd; B’ = Fe, Co Ni Cu, Zn) perovskites. These AMn(1-x)B’xO3 materials contain substitutions of Mn at the B sites as well as chemical elements that were not part of the training set (Ni, Cu, Zn).