Effects of hydrophilicity of blended submicrogels on the microstructure and performance of thermo-responsive membranes

The hydrophilicity of smart additives thermodynamically and kinetically affects the degree and velocity of the phase separation of casting solution, together with their distribution on the membranes, which ultimately affects the microstructure and properties of smart membrane. However, the reasonable design and successful preparation of additives with different hydrophilicity, as well as the observation of their migration during membrane formation process and final distribution on/in the membranes remain challenges. The three kinds of thermo-responsive submicrogels with similar chemical structures and different hydrophilicity were successfully fabricated and blended into poly(ether sulfone) (PES) as functional gates to prepare smart membranes via vapor-induced phase separation (VIPS). The hydrophilicity of poly(N-n-propylacrylamide) (PNN), poly(N-isopropylacrylamide) (PNI) and poly(N-isopropylmethacrylamide) (PNM) submicrogels were quantitively measured; their diameters and thermo-responsive factors in water were similar. Moreover, the effects of the hydrophilicity of submicrogels on the microstructures and properties of as-prepared smart membranes were systematically investigated and the results were elucidated in principle. The results showed the velocity of phase separation of casting solution, the pore size of membranes and the number of submicrogels distributed on membranes increased whereas the stability of submicrogels on membranes decreased with the enhancement of hydrophilicity of submicrogels. The membrane blended with PNI submicrogels with optimal hydrophilicity, which was provided with both a large number of submicrogels stably distributed on the membrane pores as smart gates and appropriate pore size, exhibited excellent permeability and thermo-responsiveness. The maximum water flux through this membrane was 26271 kg m(-2) h(-1) bar(-1), and its maximum normalized thermo-responsive coefficient N value reached to 28.7. The results will provide valuable guidance for understanding the mechanism of phase separation and further development of smart membranes with satisfactory performances.