Modeling Transport Through Composite Membranes: Impacts of Thin Film and Support Membrane Morphology

Most commercial reverse osmosis (RO) and nanofiltration (NF) membranes are comprised of an ultra-thin polyamide film formed over a microporous “ultrafiltration-like” support membrane (typically polysulfone or polyethersulfone cast over a polyester or polypropylene non-woven fabric). However, transport through these composite structures has not received much attention in the open literature. In particular, to address the basic questions of how the support membrane and the morphology of the typically rugose RO, compared to the smooth NF polyamide films, impact transport through the composite membrane.

In our first theoretical study employing numerical analysis of diffusion through composite membrane structures¹, we found local “hot spots” of high flux (2-9 times the surface area average flux) occurring over support membrane pores; such hot spots could serve to initiate or enhance NF/RO membrane fouling and scaling phenomena. In a second study, we extended the model to consider thin film roughness². Results illustrated how hot spots may be dampened by aligning roughness features directly over support membrane pores, increasing support membrane porosity, and increasing solute permeability through the microporous regions of the support membrane (i.e., the “solid” phase, not the skin layer pores). It was further shown that there is a trade-off between dampening flux hot spots and maintaining high-permeability, suggesting that high flux composite membranes may be inherently fouling-prone.

Here, we build further on our prior efforts and consider heterogeneous films, in which the presence of fluid-filled cavities creates hollow, lobe-like coating films, as have been suggested in the patent and open literature for certain commercial RO membranes ^3,4. The introduction of hollow domains between the coating film and support membrane raises interesting questions with respect to possible transport mechanisms through composite NF/RO membranes. These, and their possible implications for permeability enhancement and fouling-prone hot spots, will be discussed.

1.         G. Z. Ramon, M. C. Y. Wong and E. M. V. Hoek, Journal of Membrane Science, 2012, 415–416, 298-305.

2.         G. Z. Ramon and E. M. V. Hoek, Journal of Membrane Science, 2013, 425–426, 141-148.

3.         M. Kurihara and M. Hanakawa, Desalination, 2013, 308, 131-137.

4.         T. Kato and M. Henmi, Self-Organized Liquid-Crystalline Nanostructured Membranes for Water Treatment: Selective Permeation of Ions, Asilomar Conference Grounds, Pacific Grove, CA, 2013.