A novel probe design strategy based on multiple regression analysis: Emerging computational methods and experimental validation for efficient heavy water detection

Heavy water (D₂O) plays a vital role in various fields, including nuclear fission moderation, metabolic flow tracing, and deuterated drug synthesis. Precise detection of D₂O purity is essential due to its susceptibility to H₂O contamination. This study explores novel computational strategies utilizing Electrostatic potential (ESP), Bonding energy (BE), and Laplacian bond order (LBO) to predict the performance of optical probes designed for D₂O detection. Through multiple regression analysis, a quantitative evaluation formula was developed and validated experimentally using modified coumarin-based probes. Probes such as 4-MU-CN and 2OH-CN demonstrated superior detection capabilities in UV-Vis and fluorescence spectroscopy, exhibiting high linearity and low limits of detection (LOD). Molecular properties were further elucidated via NMR and density functional theory (DFT), confirming the mechanism of D₂O recognition and highlighting the role of intramolecular charge transfer (ICT) in enhancing probe sensitivity. Additionally, the study explores the applicability of the computational approach to different probe structures, screening of an amino-based optical probe AMC-CN-Ac with good detection properties. These findings underscore the efficacy of computational tools in optimizing probe design for precise D₂O detection.