Multi-stage Aqueous Two-Phase Extraction

ATPE is an alternative, cheap and continuous protein purification technique which can handle large and varying material types as well as achieve high purities and yields. It could be used to shift protein manufacturing from a batch to a continuous process. The implementation of continuous processing would be beneficial to the biopharmaceutical industry as it increases the product quality, decreases costs, and increases throughput. However, ATPE is associated with a low resolution and reliability which has led to a lack of wide acceptance within an industrial setting. This thesis uses ATPE and the McCabe Thiele method to better understand ATPE and improve resolution.

This work experimentally evaluated MSE using model pigmented proteins. The yield and purities demonstrated there were differences between the separation methods tested. The gravity settled systems (to a distinct horizontal interface formation) had a lower resolution than centrifuge settled systems despite literature suggesting they were comparable. To understand the differences, settling of ATPE was evaluated using traditional dispersion height measurements, turbidity measurements, microscopic observation, and changes in protein concentration in each of the phases. It was found protein precipitated and settled from the bulk liquid phases into the horizontal interface after its formation. Single-stage protein partitioning data was experimentally determined and used in a modified McCabe Thiele method to predict multi-stage behaviour; this was compared to MSE experimental systems. Previous literature only considered two bulk liquid phases in the partitioning data. This work found that horizontal interface partitioning data was required to accurately describe multi-stage ATPE. Deming regression was used to show uncertainty in horizontal interface partitioning data did not impact the stage-wise modelling. This thesis has shown considering partitioning into the horizontal interface of a system is required to model multi-stage ATPE. Accurate understanding and modelling of multi-stage ATPE will aid in more reliable and successful process design.