Data-Integrated Elucidation of Structure-Activity Relationships toward the Rational Design of...

Understanding and predicting catalyst performance from structural and electronic information remains a central challenge in organocatalysis. Here, we present a data-integrated framework that quantitatively combines experimental, computational, and machine learning (ML) approaches to elucidate the structure-activity relationships of halogen-bond (XB) donor catalysts based on perfluoroiodoarene cores. Single-crystal X-ray diffraction and chloride-binding analyses revealed that linker-containing two-point donors exhibit significantly stronger binding and higher catalytic activity than one-point donors. Through density functional theory calculations and an ML regression analysis integrating crystallographic and electronic descriptors, the Gibbs free energy change (ΔG) and binding constant (K) were identified as primary determinants of activity. Meanwhile, a Shapley additive explanation analysis highlighted the σ/π-hole potentials and nucleophilicity (N value) as additional electronic factors. This integrated experimental-computational-ML approach enables the quantitative extraction of key electronic factors governing XB donor catalysis and provides a physically interpretable framework for extending noncovalent-interaction-driven organocatalysis beyond XB formation, encompassing hydrogen- and chalcogen-bond donor systems.