A COMSOL-Based Modeling Framework for Predicting Buffer-Dependent Selectivity in Single-Pass Nanodiafiltration

Membrane selectivity in single-pass nano-diafiltration systems is strongly influenced by buffer type and concentration arising from boundary-layer formation due to molecular interactions between solutes, buffer components, and the membrane. In this study, a COMSOL Multiphysics model was created to quantitatively describe the relationship between apparent solute rejection and the properties of the buffer system, to create an engineering-oriented design and prediction tool. The model parameters were calibrated using selected experimental datasets, and predictive performance was subsequently evaluated against an independent dataset at an intermediate buffer concentration. The calibrated model was then applied to assess how diavolume and flow conditions shape the separation performance of raffinose and uridine as a model mixture representing an oligosaccharide product and a low-molecular-weight nucleoside by-product of enzymatic glycan synthesis in the flow modul used. Simulation allowed for precise predictions of the relation between recovery rate and product purity as well as their dependency on buffer concentration. The good agreement between simulated and measured values indicates that the model successfully captures the major experimentally observed trends governing buffer-dependent selectivity. ... mehrThese insights form the basis for more rational process design, and rapid computer-aided optimization may substantially reduce experimental effort in the development of nano-diafiltration processes.