A computational approach to evaluation of nanoporous zeolitic membranes for reverse osmosis desalination

Abstract

Water scarcity has become one of the major global concerns. Research has shown that more than one-third of the population throughout the world resides in water-stressed regions now. Reverse osmosis (RO) desalination has recently been highlighted as a clean technology that is less energy intensive using membranes rather than directly consuming fossil fuels. Advances in nanomaterials have opened new possibilities for RO membrane materials. Zeolites are one of the candidate materials due to their crystalline structures and durability. However, the performance of zeolite membranes has not been successful for practical use due to poor water flux. Thus, we used molecular dynamics simulations to investigate, at the molecular level, a series of potential zeolite types which have been rarely studied for application in desalination. We determine diffusion coefficients and the structure of water when it passes through pores of these zeolites. In addition, we examine the potential of mean force for water or ions as they move into and through a zeolite membrane in order to evaluate how favourable the passage of the molecule of interest through each membrane is. This provided us with some criteria (i.e. water permeability and ion selectivity) to judge if the membrane is likely to have suitable desalination performance. Without this molecular-level understanding, selection of zeolite materials for desalination membranes is difficult. Some widely-studied types and potential types of zeolites were selected for our study. The common types have all 3-dimensional (3-D) pore structures, while the potential zeolites have 1-dimensional (1-D) cylindrical pores. To investigate the water dynamics and structure in these different pores, we employed molecular (MD) dynamics simulations. The MD results showed the water self-diffusivity, which indicates a molecular mobility, in the 1-D pores is up to 18-time higher than that of the 3-D pores. As it was found that water molecules formed clusters, water droplets, and moved collectively at low water density in the zeolite pores, the water collective diffusivity, which is directly related to water flux, was also measured for the all the case of zeolites. This collective diffusivity through the 1-D pore zeolites was around one order of magnitude higher than that of the 3-D pore zeolites, suggesting our 1-D zeolites selected are promising as high flux membranes. To evaluate the thermodynamic stability for water or ion of interest when they pass across the zeolite membranes, the potential of mean force calculations were carried out using MD simulations. We selected one of the 3-D pore zeolites (LTA) and another of the 1-D pore zeolites (VET) for the membrane. In general, these membranes had a moderate energy barrier to water transport, but a very high barrier to sodium or chloride ion transport except the VET membrane. The chloride ion had a minimum in the potential of mean force at the pore entrance and a preference for the chloride ion to enter the pore than to enter the bulk solution for the system considered. However, VET has a lower energy barrier to water passage and a comparable high energy barrier to sodium ion transport compared to the LTA membrane. The VET membrane may be feasible as a desalination membrane if there is some chemical modification done on the pore to effectively reject the chloride ion. This issue will be more studied and addressed in our future work.