Furthermore, evolutionary experiments with GAs lead to alternative chemical insight into catalyst performance, as demonstrated hereafter. This pipeline represents a complementary approach to high-throughput screening 13- 17 that becomes comparatively more efficient as the dimensionality of the combinatorial space of catalyst components grows. In the spirit of inverse design, 4- 8 NaviCatGA uses a Genetic Algorithm (GA) 9- 12 to find optimal catalysts (Figure 1). NaviCatGA is a modular part of the broader NaviCat ( Navigating Catalysis) platform for catalyst discovery, which includes other utilities and tools (e. g., database constructors, automatic volcano plot builder, etc.). It manipulates catalyst structures generated in situ from a user-defined library of catalyst components (metal centers, ligands or ligand substituents, scaffolds, etc.) structures can be assembled from the respective components using any molecular representation, including SMILES strings and XYZ coordinates, and evaluated according to any fitness function (e. g., molecular volcano plot descriptors, 1, 2 multivariate linear regression expressions 3).

This work introduces NaviCatGA, a software package capable of optimizing catalysts by exploiting any suitable fitness function that describes their catalytic performance.

Navicat online