Novel Membrane Adsorbers with Grafted Zwitterionic Polymers Synthesized by Surface-Initiated ATRP and Their Salt-Modulated Permeability and Protein Binding Properties

A novel zwitterionic polymer functionalized porous membrane adsorber was obtained by grafting poly(N,N-dimethyl-N-methacryloyloxyethyl-N-(3-sulfopropyl) ammonium betaine) (polySPE) to poly(ethylene terephthalate) (PET) track-etched membrane surface via surface-initiated atom transfer radical polymerization (SI-ATRP). The ATRP conditions were optimized, the thus established grafting was well-controlled, and the degree of grafting could be adjusted. Functionalized membranes with a degree of grafting of about 3.5 mu g/cm(2) relative to the specific surface area showed almost zero values of zeta potential estimated from the trans-membrane streaming potential measurements. Typical anti-polyelectrolyte effect was observed for the polySPE grafted membranes. Flux through the membrane was reduced by adding chaotropic chloride and perchlorate salts to the solution which extended the polySPE chains grafted on the membrane pore wall. Perchlorate salt exhibited much stronger effect on polySPE chain conformation than chloride salt and for a membrane with a degree of grafting of 2.7 mu g/cm(2), even 2 mM KClO4 could extend the thickness of the polymer layer to more than two times (similar to 43 nm) of that in pure water (similar to 20 nm). On the contrary, small amounts of kosmotropic ions (10 mM SO42-) further salted out the polySPE chains and led to a slightly increased flux. PolySPE grafted PET membranes with different degree of grafting were then used as membrane adsorber for protein binding. Human IgG was used as model protein and the binding capacity was evaluated under both static (no convective flow through the membrane) and dynamic conditions (flow-through conditions). Static adsorption experiments showed that IgG could be loaded to the membrane at medium salt concentration and 85-95% of bound protein could be eluted at either low (zero) or very high salt concentrations. Dynamic flow-through experiments then revealed the influences of salt concentration and salt type on IgG binding. Effects of two chaotropic salts, NaCl and NaClO4, were evaluated. Slight but not negligible binding of IgG from pure water was suppressed by adding NaCl. IgG binding was then increased in the NaCl concentration range of 100-500 mM and reached a maximum binding capacity value at about 500 mM. Further increase of NaCl concentration led to a decreased binding again. KClO4 showed similar effects onto IgG binding, but this salt functions in a much lower and much narrower concentration range. All results with respect to grafted layer swelling and protein binding followed the empirical Hofmeister series.