Scale-up of affinity membrane modules: comparison between lumped and physical models.

Area of Science:

• Biotechnology
• Chemical Engineering
• Separation Science

Background:

• Membrane chromatography is an emerging downstream processing technology in biotechnology.
• Currently used in antibody manufacturing polishing steps, its application in capture steps is under development.
• Reliable simulation tools are needed for predicting performance in large-scale processes.

Purpose of the Study:

• To develop and validate a simulation tool for membrane chromatography.
• To compare the predictive accuracy of a physical model against a lumped model for scale-up.
• To assess the impact of binding kinetics (Langmuir and bi-Langmuir) on model performance.

Main Methods:

• Utilized a physical model to describe protein purification in affinity membrane chromatography.
• Implemented Langmuir and bi-Langmuir binding kinetics.
• Compared model predictions with experimental data from lab-scale and scaled-up systems.

Main Results:

• Both physical and lumped models accurately described lab-scale data.
• Significant differences emerged between models in scaled-up systems, even at early breakthrough stages.
• The physical model demonstrated superior accuracy and safety for scale-up predictions.

Conclusions:

• The physical model is more appropriate and safer for scaling up membrane chromatography processes.
• Model selection is critical for accurate prediction of industrial-scale bioprocessing.
• Further development of simulation tools is essential for broader adoption of membrane chromatography.