Modeling, evaluation, and optimization of a self-bearing spinfilter for bioseparations

Rotating filters promise many advantages for long-term separation processes of complex media such as cell cultures, but have never been commercialized on a large scale due to practical challenges and limited scalability. A bearingless spinfilter circumvents many of the drawbacks, yet additional considerations must be taken into account. Thus, an existing particle-force based model for cross-flow filtration, dependent on the membrane shear rate and the transmembrane pressure, is adapted to the spinfilter and predicts the behavior of retained particles at the membrane. Computational fluid dynamics simulations allow to calculate the shear rate at the membrane for the model. Additionally, with the simulated distribution of the shear rate and the energy dissipation rate within the spinfilter, a prediction of the hydrodynamic stress on cell cultures can be made. The model is verified with a spinfilter prototype on a yeast cell culture and is in good agreement with experimental results. Additional separation experiments with bovine blood serve to indicate the extent of cell damage induced by hydrodynamic stress within the spinfilter. The results demonstrate that operating the spinfilter at the optimum operation point leads to a faster separation process at a lower level of cell damage per volume of filtrate in comparison to a passive tangential flow filtration in hollow fibers.