Automatic procedure for modelling, calibration, and optimization of a three-component chromatographic separation

Area of Science:

• Biopharmaceutical Manufacturing
• Process Engineering
• Digitalization in Drug Development

Background:

• Biopharmaceutical production requires efficient tools for reduced development times and costs.
• Mathematical modeling of process chromatography is crucial for downstream processing.
• Model calibration for multi-component mixtures is complex and time-consuming.

Purpose of the Study:

• To develop an automated model calibration procedure for multi-component mixture purification.
• To implement this procedure using the Orbit software for automated model generation and experimentation.
• To extend the procedure for optimizing operating points via multi-objective optimization.

Main Methods:

• Development of an automated model calibration procedure for linear gradient ion exchange chromatography.
• Utilization of Orbit software for automated model structure generation and experimental data acquisition.
• Application of multi-objective optimization to suggest purification operating points.

Main Results:

• A calibrated mathematical model was generated for a three-component protein mixture using only six assays.
• The model accurately reproduced experimental chromatograms.
• A seventh experiment validated the model's predictive capability under optimized conditions with 95% purity.

Conclusions:

• The automated procedure successfully calibrated a model for complex mixture purification.
• This automation significantly reduces the effort required for model calibration.
• The study advances digitalization in biopharmaceutical development by enabling faster, automated process optimization.