Investigating transport through sub-nanometer zeolites pores

Abstract

Membrane-based reverse osmosis (RO), which accounts for over 40% of the current worldwide desalination capacity, is limited by the solution-diffusion mode of water transport through a tortuous polymeric active layer. Alternatively, simulations suggest that introducing rigid sub-nanometer pores into the active layer could substantially increase the water permeability and achieve perfect salt rejection. In this thesis, we synthesized MFI (Mobil Five) zeolites that have uniform, well-defined pores of ~~ 5.5 Å and experimentally investigated the transport across these sub-nanometer pores. This porous structure serves as a model framework to experimentally investigate water and salt transport and can be used to suggest the potential performance of such microporous active layers in RO-based desalination. We first developed an experimental methodology that combined vapor sorption analysis with high-pressure infiltration to probe both the role of crystal size and internal surface chemistry on the transport properties. For purely siliceous MFI zeolites, upwards of 100 MPa was required to saturate the porous network to the framework capacity of 35 water molecules per unit cell. However, by introducing hydrophilic (i.e, acidic) defects within the structure, this infiltration pressure was reduced by five orders of magnitude (to ~~ 1 kPa). While increasing the defect density increased the amount of water within the pores at typical RO pressures (~~ 5 - 6 MPa), the diffusivity of this infiltrated water within the more defective zeolites was 1 - 2 orders of magnitude lower than that of the water within the purely siliceous MFI zeolite. This decreased diffusivity, which was attributed to the strong attraction of water to the hydrophilic defect sites, resulted in ~~ 10x decrease in the estimated permeability. We subsequently microfabricated sub-micron thick zeolite-based membranes to investigate transport limited to the zeolite crystals. By quantifying the water flux generated by a concentration gradient (i.e., forward osmosis), the flux across the more hydrophobic MFI zeolites was ~~ 8 - 10x higher than that across the more hydrophilic MFI zeolites. Additionally, although a small amount of meso/macro-sized defects existed in the membranes, no salt transport (within experimental uncertainty) was detected across the zeolite pores, which demonstrated that the pores of MFI zeolites were capable of selectively transporting water and rejecting hydrated salt ions. Collectively, this work presents an improved fundamental understanding of the transport of water that is confined within the sub-nanometer pore structure of zeolites. The insights gained demonstrate the potential of size-selective membranes in water desalination and offers approaches toward improving the performance of future RO membranes.