Determination of the key structural factors affecting permeability and selectivity of PAN and PES polymeric filtration membranes using 3D FIB/SEM

Abstract

Microfiltration (MF) and ultrafiltration (UF) processes are well known in water treatment or separation of valuable biomolecules. They have recently been adapted for microalgae valorization, where filtration employing nanoporous polymer membranes is used to separate and recover lipids and proteins from microalgae extracts. As the design of novel MF and UF membranes with optimized filtration performance (reduced fouling of molecules and increased filtrate fluxes) is leading to increasingly complex pore structures, new characterization methods of filtration membranes are needed. A detailed, nanometer scale, characterization of the three-dimensional pore structure of the membranes and the precise elucidation of the membrane’ structure-performance relationship is thus essential for advancing the development of efficient filtration process operating but also novel MF and UF membranes. In this work, the structural features determining the filtration performances of commercially available polyacrylonitrile (PAN) UF and polyethersulfone (PES) MF membranes are determined using scanning electron microscopy (SEM) coupled with a focused ion beam (FIB) at low electron-doses to produce 3D reconstructions with up to 5 nm resolution. Here, methods to identify key structural parameters of the selective layer or skin of the membranes and to estimate the percentage of blind (dead-end) pores communicating with the membrane surface but not crossing the membrane are presented. Furthermore, the data obtained also indicates that widely used models such as Hagen-Poiseuille equation are insufficient to fully describe asymmetric membranes defined by the presence of a thin selective layer. This work opens up the possibility of providing detailed information, useful not only to illustrate novel filtration membrane designs, but also as input data for more complete nanometer-scale based predictive models.