Desenvolvimento de processos cromatográficos empregando ensaios de triagem e modelagem fenomenológica

Abstract

Phenomenological models are valuable tools for developing chromatography methods. However, its applicability is impaired by the theorical complexity and accessibility to model parameters. Despite materials are usually select by screening protocols, data obtained during this phase are not used further in the development of the purification process. Here, a joint estimation procedure that integrates data from microplate screening with a single column experiment was proposed to estimate equilibrium and dispersion parameters for modeling gradient elution. Initial estimates of equilibrium isotherm parameters were derived from the screening data, which were then refined by incorporating column chromatogram data to estimate mass transport parameters and enhance isotherm predictions. The proposed method was validated with a reverse phase chromatography using a non-ionic surfactant and functionalized silica. The equilibrium parameters estimated with microplate data predicted a retention volume 8.9% far from experimental results, while the joint estimation prediction was 0.8% in the best-case scenario. In this case, results show that the use of screening data combined with an L-curve method effectively mitigated model overfitting, even when relying on a single gradient elution column assay. The presented protocol should speed up the process development and reduce the consumption of inputs with column experiments.