Ester hydrolysis rate constant prediction from quantum topological molecular similarity descriptors

A previously established method [J. Chem. Inf. Comput. Sci. 2001, 41, 764], called quantum topological molecular similarity, is applied to obtain an excellent and statistically validated quantitative structure-activity relationship (QSAR) of base-promoted hydrolysis rate constants for a set of 40 esters. This work is relevant for environmental exposure and risk analysis and proposes a reliable and cheaper altemative to measuring infrared group frequencies for that purpose. Our method draws descriptors from modem ab initio wave functions, which have become affordable by the current abundance of inexpensive computing power. We acquire a 3D geometry-optimized picture of each molecule and characterize its bonds further with four quantities defined by the theory of quantum chemical topology. Without molecular superposition we then construct a variety of models, which all point toward the molecular fragment (O=C)-C-O being most significant to explain the range of hydrolysis rate constants. This highlighted zone is called the active center, and when the model is confined to it, a QSAR of r(2) = 0.930 and q(2) = 0.863 is obtained for all 40 esters.