Chromatography is a fundamental laboratory technique used for separation mixture into its components. In biopharmacy, it is used to separate target proteins from other impurities in biologically active compounds.
This master's thesis focuses on mechanistic modeling of multimodal anion-exchange chromatography (MAC), which is an important technique in the production of biopharmaceuticals due to its high selectivity and its adaptability to process conditions. As experimental optimisation of chromatographic processes is time and resource intensive, the models serve to support our development and deepen our understanding of the process. We have experimentally calibrated a chromatographic model consisting of a Transport Dispersive column model (TDM) and different adsorption models. We have shown that simple models such as the linear and Langmuir isotherms are not suitable for modelling MAC chromatography under selected process conditions, while various upgrades of the Steric Mass Action (SMA) model provide better results. The latter were evaluated with various industrially relevant metrics such as purity and process yield prediction. Finally, we validated the model predictions on an independent dataset.
|
