Classical molecular dynamics and Monte Carlo simulations are used to calculate the self-diffusivity and solubility of pure and mixed CO(2), H(2), and Ar gases absorbed in the ionic liquid 1-n-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide ([hmim][Tf(2)N]). Overall, the computed absorption isotherms, Henry's law constants, and partial molar enthalpies for pure H(2) agree well with the experimental data obtained by Maurer et al. [J. Chem. Eng. Data 2006, 51, 1364] and the experimental values determined in this work. However, the agreement is poor between the simulations and the experimental data by Noble et al. [Ind. Eng. Chem. Res. 2008, 47, 3453] and Costa Gomes [J. Chem. Eng. Data 2007, 52, 472] at high temperatures. The computed H(2) permeability values are in good agreement with the experimental data at 313 K obtained by Luebke et al. [J. Membr. Sci. 2007, 298, 41; ibid, 2008, 322, 28], but about three times larger than the experimental value at 573 K from the same group. Our computed H(2) solubilities using different H(2) potential models have similar values and solute polarizations were found to have a negligible effect on the predicted gas solubilities for both the H(2) and Ar. The interaction between H(2) and the ionic liquid is weak, about three times smaller than between the ionic liquid and Ar and six times smaller than that of CO(2) with the ionic liquid, results that are consistent with a decreasing solubility from CO(2) to Ar and to H(2). The molar volume of the ionic liquid was found to be the determining factor for the H(2) solubility. For mixed H(2) and Ar gases, the solubilities for both solutes decrease compared to the respective pure gas solubilities. For mixed gases of CO(2) and H(2), the solubility selectivity of CO(2) over H(2) decreases from about 30 at 313 K to about 3 at 573 K. For the permeability, the simulated values for CO(2) in [hmim][Tf(2)N] are about 20-60% different than the experimental data by Luebke et al. [J. Membr. Sci. 2008, 322, 28].

No abstract

No abstract

Polymeric hollow fiber supported ionic liquid membranes (SILMs) were fabricated utilizing Matrimid ® and Torlon ® as the supporting structure and the ionic liquid (IL) 1-hexyl-3methylimidalzolium bis(trifluoromethylsulfonyl)imide ([C 6 mim][Tf 2 N]) as the gas transport media. This IL served as a baseline while the polymer and the fiber pore morphology were varied in order to optimize the support structure. By using sufficiently high fiber porosity, and thus maximizing the IL content of the membrane, it was found that the permeance and selectivity for CO 2 /H 2 separation were comparable for both Matrimid ® and Torlon ® supports. However, the mechanical strength of Matrimid ® supports was low when saturated with IL. Therefore Torlon ® fibers were subsequently investigated because of the higher strength of this material. Molecular modeling was used to investigate the source of the increased strength of Torlon ® , and it was found that the polymer chains in Torlon ® tend to interlock with each other to a greater degree than Matrimid ®. Also, the IL [C 6 mim][Tf 2 N] has less interaction with Torlon ® than with Matrimid ®. In this work the permeance and selectivity for CO 2 /H 2 of these hollow fiber SILMs are reported, as well as the tensile strength, Young's modulus, and glass transition temperature. Threshold image analysis was used to determine the volume fractions of polymer, macro-voids, and micro-voids. X-ray computed tomography scanning was used to non-destructively evaluate the location of IL within the fiber wall.

A constant pressure, mixed gas permeation testing skid was deployed at the National Carbon Capture Center to test membrane performance when continuously exposed to slipstream postcombustion flue gas. Small, laboratory scale membranes can be loaded for testing and the degree of automation allowed the skid to be run unattended for several weeks at a time. In this report, we share our experience in commissioning the skid and quantifying CO2, N2 and O2 permeances of several membranes during the first round of testing. Dense films of polydimethylsiloxane and poly(bistrifluoroethoxyphosphazene) were tested with flue gas for approximately 20 hours each. In addition, we successfully tested four thin film composite hollow fiber membranes made by a dip coating process, consisting of porous Torlon hollow fibers coated with a selective layer of poly(bistrifluoroethoxyphosphazene) or its mixed matrix with a metal organic framework SIFSIX-Cu-2i filler particles. Initial results suggest the polydimethylsiloxane showed comparable results to the literature data, but the coated hollow fiber membranes have lower CO2 permeances relative to N2 or O2 permeances compared to their performance under idealized, dry, contaminant-free mixed gas conditions. While quantification of H2O permeance was performed, we found it was affected by concentration polarization even with small membrane area and a low stage cut.