Tunable retention and recovery via pore-to-particle interactions in amine functionalized micro- and ultrafiltration...

Membrane-based technology is a cornerstone of bioprocessing and nanoparticle separation, yet the relationship between membrane pore size and particle retention or transmission remains poorly standardized across applications. This study systematically investigates the role of the pore-to-particle size ratio ( $\lambda =d_m/d_p$ ) in determining membrane performance, integrating literature data with controlled filtration experiments using model nanoparticles, proteins, and virus-like particles. We demonstrate that high retention (≥95 %) occurs consistently at $\lambda <0.5$ across both microfiltration and ultrafiltration regimes; however, flux stability varied widely with $\lambda \ll 1$ , offering minimal flux decline. As $\lambda$ increases above 1, transmission dominates, but selectivity can be dramatically enhanced through membrane surface modification. Amine functionalization of the membrane pore with a hydrogel exhibited >70 % retention of 20 nm negatively charged particles even when $\lambda >40$ and 80 % retention of DNA and protein when $\lambda >160$ , highlighting the significant role of electrostatic interactions. Furthermore, modeling the prediction of flux decline and pore size reduction further validated the increased interactions via functionalization. Notably, these findings have implications for adeno-associated virus (AAV) purification: while conventional microfiltration membranes transmit AAV with minimal selectivity, functionalized membranes offer a potential route to selectively retain impurities without compromising vector recovery. Together, this work establishes a unified framework based on $\lambda$ and demonstrates how targeted functionalization can expand the functional range of porous membranes. These results provide a mechanistic foundation for improved membrane selection and design, especially in applications requiring high-resolution particle separation, such as gene therapy manufacturing.