| Description |
This dissertation describes the development of nanoporous membranes formed via the assembly of silica nanoparticles modified with polymer brushes. This dissertation explores the use of such "hairy" nanoparticles (HNPs) for the purpose of membrane separations with size and charge selectivity. These HNP membranes have the advantages of pore size tunability (varied by changing the silica diameter) as well as cost effectiveness, and potential reusability. The latter is based on the ability to disassemble the membrane building blocks for cleaning and then to reassemble the membrane without performance loss. The first set of HNP membranes described display the reversibility and size selectivity. Silica nanoparticles were modified with polymer brushes containing monomers with acid/base groups or a neutral monomer. The acid/base HNP membranes were stable in organic solvents and the neutral HNP membranes were stable in water. Both types of membranes could be deposited, disassembled (in water for acid/base HNP membranes and organic solvents for neutral HNP membranes), and redeposited. Cut-off experiments were also performed, confirming that the ultrafiltration membranes could be used for size selective separations. To further explore the applications of HNP membranes, charged polymer brushes were grown from the surface of silica to produce nanofiltration membranes capable of charge rejection. Copolymers of differing lengths and sulfonic group content were grown from the surface of silica to study the effect of membrane charge density on the rejection of salts. Increasing the percentage of sulfonic groups in the copolymers increased the rejection of salts in water. It was observed that the introduction of salt solutions to the membranes affected the polymer conformation. Size-based separation experiments were also performed, confirming that both size and charge selectivity can be achieved. The arrangement of polymer-coated silica spheres in the HNP membranes forms a classic hard-sphere/soft-shell system. Typical geometrical calculations of particle cutoff size are inaccurate for randomly positioned spheres, and tests indicated that pore size is also not well predicted by the Kozeny-Carman equation or geometric sphere model. Hence, a percolation model is presented for accurate estimate of the pore size cut-off for any random soft-shell hard-core sphere system. The model is created assuming a discrete (pixelated) geometry for the sphere system and can be easily modified for any number of percolation or transport studies. |
