A Diagnostic Procedure for Identifying Isotherm Models in Liquid Chromatography.

Area of Science:

• Chemical Engineering
• Separation Science
• Computational Chemistry

Background:

• Liquid chromatography is crucial for pharmaceutical purification, relying on mechanistic models like POR and EDM.
• Isotherm model selection significantly impacts chromatography model predictions, especially for complex molecules like peptides.
• Traditional models often fail for complex separations, leading to process-model mismatch.

Purpose of the Study:

• To address process-model mismatch in liquid chromatography by refining isotherm models.
• To develop a model diagnostic procedure that improves the accuracy of chromatography models.
• To provide physically insightful models without resorting to complex data-driven approaches.

Main Methods:

• Adapted a Lagrange multiplier test from kinetic models for isotherm model diagnostics.
• Developed a procedure to refine underfitting isotherm models in chromatography.
• Validated the procedure using three in-silico case studies.

Main Results:

• The diagnostic procedure successfully refined isotherm models, improving predictive accuracy.
• Enhanced models showed better agreement with experimental data compared to traditional approaches.
• The method avoided the need for complex, less interpretable black-box models.

Conclusions:

• The proposed diagnostic procedure effectively addresses process-model mismatch in liquid chromatography.
• This approach allows for the development of accurate and physically insightful chromatography models.
• It offers a valuable tool for optimizing pharmaceutical purification processes.