Mechanistic process models have gained increasing interest in digitalizing the optimization and characterization preparative chromatography, in particular of ion exchange (IEX) processes. For industrial application, a robust and reliable calibration workflow is key to adoption and management buy-in. Several model-based workflows have been presented in the past trying to streamline the application of mechanistic models in process development. For models relying on the most common steric mass action (SMA) isotherm, model calibration is thereby performed predominantly in the linear adsorption range using well-established approaches like the Yamamoto method. The model is afterwards extended to the non-linear adsorption range by considering a limited number of chromatography runs at high-load density. While this workflow allows for a fast and cost-efficient calibration and application of mechanistic models, it requires that the entire process range can be adequately reflected by the SMA model which can only be confirmed with the high-load density runs.<br/>In this study, we present lessons learned from using the aforementioned workflow for an intensified industrial antibody IEX process. Using the Yamamoto method, a SMA model was thereby pre-calibrated based and extended to the non-linear range. Observed deviations between the calibrated model and experimental data at high protein load densities required an extension of the traditional SMA model that is incompatible with the initial Yamamoto approach. The planned workflow could therefore not be adhered to.<br/>Based on the lessons learned, we present an alternative model-based workflow using a non-stoichiometric chromatography model. In addition to cost-efficient model calibration strategies, the presented workflow describes the application of the mechanistic model to identify optimal process parameter ranges and optimize the Design-of-Experiments studies for the characterization and validation of the IEX process.

Lessons learned from establishing a model-based workflow for efficient process characterization

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.