Thursday, November 12, 2015: 12:51 PM
155B (Salt Palace Convention Center)
Prior research on the performance of RTIL-membranes theorized that membrane performance correlated with the RTIL properties of viscosity, molar volume, or the ability to chemically interact with the faster fluxing chemical species. Because RTIL-membranes are dominated by transport through the liquid phase, the lower the RTIL’s viscosity the faster the transport by diffusivity. The other theories on membrane performance (molar volume and chemical interactions) relate to the RTIL-membrane selectivities being solubility dominated, instead of diffusion dominated. The literature reports on numerous attempts to increase the solubility of the faster fluxing species in order to increase the RTIL-membrane selectivities. However, it could also be fruitful to look at hindering the transport of the second species in the separation pair, for example nitrogen, air, or methane. Unfortunately, except for Carvalho and Coutinho’s evaluation of only four ILs in 2011, this approach is the “road less traveled” in RTIL gas separations. In brief, Carvalho and Coutinho suggested that increasing the ionic liquid’s hydrogen-bond acceptor ability would lead to lower methane solubility and, therefore, higher selectivities. To the best of our knowledge, the literature still does not contain any membrane research based on their proposal. Therefore, we tested of the hypothesis that rejection of methane will dominate in determining the water/methane selectivity of RTIL-membranes. We did find that RTILs that we fabricated with high hydrogen bond acceptance potential had significantly better performance (2700 GPU-water and water/methane selectivity = 100,000) compared to previously tested RTIL-membranes. However, their behavior and membrane fabrication is complicated. One hypothesis is that increasing the hydrogen-bond acceptance ability of the RTIL increases the ability of the RTIL to interact with and solubilize the polymers used in the liquid stabilizing porous matrix of the tested RTIL-membranes. We report a strong correlation (R2 = 0.98) for an exponential fit between the water/methane selectivity and the Kamlet-Taft Parameter for hydrogen-bond accepting.