Carbonaceous adsorbents as coatings for ultrafiltration membranes

The feasibility of using adsorptive carbonaceous coatings on ultrafiltration (UF) membranes was evaluated by quantifying atrazine and methylene blue removal capabilities in addition to flux reductions associated with adsorbent applications. Various alternative adsorbents were incorporated in this study, including multi-walled carbon nanotubes (MWCNTs), nano-graphene platelets (NGPs), and superfine powdered activated carbon (S-PAC), for comparison with a more traditional material, powdered activated carbon (PAC). All adsorbents except S-PAC were associated with filtration flux reductions of less than 5% after application as a membrane coating during constant pressure tests, and flux recovery after membrane backwashing was greater than 88% for all materials and masses tested.;For removal of methylene blue in a lab-scale UF system, S-PAC showed fast adsorption kinetics and a steep breakthrough curve. It is likely that the mesopore volume and large external surface area available for adsorption contributed to this phenomenon. MWCNTs were inferior to PAC for overall methylene blue removal, which may be attributable to their lower surface area, extensive aggregation, and lack of micropores. The breakthrough was slower with adsorption occurring in a stirred vessel configuration than through adsorption in a carbon-coated membrane, but overall contaminant removal was roughly equivalent in both cases. With atrazine as the second model contaminant, S-PAC and NGPs were the most efficient for rapid adsorption, while MWCNTs and PAC had slower kinetics. Using a stirred vessel setup rather than the coated membrane resulted in lower retention of atrazine for all adsorbents, which is in contrast with the methylene blue results.;The homogeneous surface diffusion model (HSDM) for packed-column adsorption was applied to the membrane coating results, with significant deviations noted. The deviations likely resulted in part from inaccurate estimation of the surface diffusion coefficient, since adjustment of this parameter in the model yielded well-fitted curves. The linear driving force (LDF) model showed better accuracy, though this model was considered empirical. Considering the minimal flux reduction associated with the application of adsorbents on membranes as well as contaminant removal capabilities, membrane coatings have the potential to be a transformative technology for water treatment.