Abstract
A series of membrane filtration experiments was carried out in a crossflow flat slit geometry to study permeate flux dependence on particle size and operational mode. Dilute monodisperse suspensions of polystyrene particles 20, 50, 100, and 680 nm in diameter were filtered in both constant permeate flux (CF) and constant transmembrane pressure (CP) crossflow regimes using laboratory membrane cell with a known flow configuration. In addition to particle size, other controlled parameters included temperature, crossflow velocity, bulk suspension concentration, transmembrane pressure, membrane hydraulic resistance, and surface chemistry of particles and the membrane. Obtained experimental specific permeate flux profiles were used to evaluate the transient permeate flux model by Sethi and Wiesner. The extended model was found to significantly overpredict permeate flux for medium-size particles (50–100 nm) while providing reasonable agreement with experimental data for small (20 nm) and big (680 nm) particles. Despite these discrepancies, experimentally observed differences between performance in CP and CF modes were in good qualitative agreement with model predictions. The CP operation mode yielded higher specific permeate flux, with the difference decreasing as particle size increased. Further, the extended model was modified to allow for parametric studies of permeate flux using descriptors of the cake morphology as fitting parameters. It was shown that changes in cake morphology alone could not account for the discrepancies observed for medium-sized particles. These results suggest additional back-transport mechanism(s) not currently included in the model. Preferential resuspension and scouring of medium-sized particles were hypothesized as such mechanisms.
Get full access to this article
View all access options for this article.