Phase-Specific Parameter Estimation in Chiral HPLC Using 1-Site, 2-Site Stochastic Models, and Unified Equation...

On-column enantiomerization remains a challenge for quantitative chiral analysis, but it also provides an opportunity to extract mechanistic information directly from chromatographic data. Here, we present an extended automated workflow in R language for statistical computing for modeling dynamic chromatographic profiles (Batman peaks) of interconverting enantiomers, incorporating a two-site stochastic formulation that distinguishes nonselective and enantioselective interactions on cellulose-based chiral stationary phases. Overload experiments were analyzed using a competitive bi-Langmuir isotherm to estimate site ratios and affinities, which were subsequently integrated into stochastic peak-shape analysis to obtain forward and reverse interconversion rate constants across multiple eluents, temperatures, and flow rates. To decompose these apparent rates into mobile and stationary phase contributions, we developed a mixed-effect model based on capacity factor linearization, followed by empirical Bayesian inference. Mobile phase-specific rate constants exhibited consistent Eyring-Polányi behavior and showed excellent agreement with independently measured off-column circular dichroism kinetics. The results demonstrate that detailed kinetic and thermodynamic characterization of enantiomerization can be obtained directly from routine liquid chromatography measurements. The proposed workflow delivers an integrated, fully automated platform that enhances the analytical utility of dynamic chiral separations and is broadly applicable to systems exhibiting on-column interconversion.