Hyperthin Organic Membranes for Gas Separations

Abstract Gas separation is practically important in many aspects, e.g., clean energy production and global warming prevention. Compared to other separation technologies like cryogenic distillation and pressure swing adsorption, membrane separation is considered to be more energy efficient. For practical purposes, the ultimate goal is to construct membranes producing high flux and high gas permeation selectivity at the same time. Based on the inverse relationship between flux and membrane thickness, it is clear that fabricating highly selective membranes as thin as possible could increase the flux through the membrane without sacrificing selectivity. But it has proven to be challenging to manufacture selective membranes in the hyperthin (< 100nm) region. [Note: 100 nm is the typical dense layer thickness of commercial membranes to separation gases.] In this dissertation, the focus is on the development of hyperthin selective membranes that were supported by poly(1-trimethylsilyl-1-propyne) (PTMSP), using Langmuir-Blodgett (LB) and Layer-by-Layer (LbL) deposition methods. A "gluing" strategy has been successfully introduced into LB films by our laboratory recently, in which LB monolayers are ionically crosslinked with polyelectrolytes. This success stimulated the pursuance of LB films with improved gas separation properties by: (i) examining calix[n]arene-based surfactants with different sizes (ii) using polymeric surfactants as LB forming materials, and (iii) optimizing the condition of the subphase containing polyelectrolytes. Both a strong polyelectrolyte poly(4-styrene sulfonate) (PSS) and a weak polyelectrolyte poly(acrylic acid) (PAA) were used to create glued LB bilayers. The gas permeation through PSS or PAA-glued LB bilayers made of calix[n]arenes was found to be dominated by solution-diffusion rather than molecular-sieving mechanism. The porous nature of calix[n]arene-based surfactants also turned out to be unnecessary for constructing LB films with high gas selectivities, because a single PAA-glued LB bilayer made of a nonporous polymeric surfactant (ca. 7 nm) was found to exhibit a much higher H2/CO2 selectivity (200), which reached the "upper bound". The key factors that are likely to affect the quality of the glued LB bilayers were: (i) the degree of ionic crosslinking between surfactants and polyelectrolytes, (ii) charge repulsion and (iii) hydrogen bonding among the polyelectrolyte chains. In principle, employing a surfactant bearing a large number of ionic sites (e.g., a polymeric surfactant) that can increase the degree of ionic crosslinking and a weak polyelectrolyte (e.g., PAA) that can be optimized to minimize the charge repulsion and maximize the hydrogen bonding should afford glued LB films with high gas selectivities. The current finding also indicated that the PAA layer filled in between two surfactant monolayers served as the main gas barrier. In a related study, polyelectrolyte multilayers (PEMs) (ca. 14 nm in thickness), which were made from structurally matched (i.e., two oppositely charged polyelectrolytes having identical backbones) and unmatched polyelectrolytes pairs, both showed high CO2/N2 selectivities (ca. 100-150). However, the permeabilities of PEMs derived from structurally matched polyelectrolytes were significantly higher than that from an unmatched pair. The larger solubility of CO2 with respect to N2 in these PEMs should be the reason lead to the high CO2/N2 permeation selectivities.