Enhanced reverse osmosis filtration via chaotic advection induced by patterned membranes: A numerical study

This paper numerically explores the potential of patterned membranes inducing chaotic advection to enhance the efficiency of reverse osmosis (RO) filtration. We compared the filtration performance of four different membrane patterns: flat surface (FS), lateral pattern (LP), herringbone pattern (HP), and staggered herringbone pattern (SHP). The dimensionless pattern depth (dp∗) and the Reynolds number (Re) were chosen as key parameters with influence on the flow and mass transfer characteristics. The numerical scheme was validated by comparing concentration profiles and permeate flux to the experimental data. Poincaré sections were used to represent the dynamical systems in spatially periodic filtration modules. Numerical simulations were conducted for an RO process with an aqueous NaCl solution, varying dp∗ and Re within the ranges of dp∗≤0.3 and 100≤Re≤1000. The HP and SHP significantly enhanced filtration performance compared to the FS and LP, regardless of Re. Chaotic advection induced by the HP and SHP effectively suppressed concentration polarization by transporting solutes away from membrane surfaces and uniformly redistributing them back to the feed stream. Furthermore, this study demonstrates that the patterned membranes inducing chaotic advection can enhance the energy efficiency of RO filtration as well.