2070
J. Phys. Chem. B 2007, 111, 2070-2074
Solubility of CO2 in Room Temperature Ionic Liquid [hmim][Tf2N] Mark B. Shiflett*,† and A. Yokozeki‡ DuPont Central Research and DeVelopment, Experimental Station, Wilmington, Delaware 19880, and DuPont Fluoroproducts Laboratory, Chestnut Run Plaza, 711, Wilmington, Delaware 19880 ReceiVed: NoVember 16, 2006; In Final Form: December 29, 2006
Solubility measurements of carbon dioxide in 1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide have been performed with a gravimetric microbalance at temperatures of about 282, 297, 323, and 348 K and pressures up to about 2 MPa. Two different sources for the ionic liquid are examined in this work: an ultrapure sample from NIST (the IUPAC task force sample) and a commercially available sample. Both samples show nearly identical solubility behaviors, being undistinguishable within experimental uncertainties. Solubility (pressure-temperature-composition) data have been well correlated with an equation-of-state (EOS) model used in our previous works. The EOS model calculations are compared with experimental solubility data for the same system in the literature. The present EOS has predicted partial immiscibility at the CO2rich side solutions. To prove this prediction, vapor-liquid-liquid equilibrium experiments have been made, and our predictions have been confirmed.
Introduction Room temperature ionic liquids (RTILs) are molten salts at ambient temperature and are regarded as a new type of solvent.1 Many possible applications with RTILs have been proposed,2 including actual industrial applications.3 Therefore, in the last several years, thermodynamic and thermophysical properties of pure RTILs and their mixtures with various chemicals have been intensively measured by worldwide research groups. However, the quality of such experimental data is often questioned because of large discrepancies among research groups. Some researchers believe that the major cause of such inconsistencies would be the sample purity or chemical stability of RTILs used in the experiment.4 In addition to that, the present authors suspect that there might be other major causes such as systematic measurement errors, T and P calibration errors, inadequate T and P control, unestablished equilibrium conditions, etc. To attain reliable standards for thermodynamic and thermophysical properties of RTILs and their mixtures, the IUPAC task group has been formed.5 Then, as a typical RTIL, 1-hexyl3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([hmim][Tf2N]) was selected for this purpose. Highly purified samples of [hmim][Tf2N] were distributed to several research groups by Drs. J. W. Magee and J. A. Widegren (NIST, Boulder, CO), as members of the task force. After knowing the purpose of the IUPAC group, we decided to join such an international activity, and asked Drs. Magee and Widegren to send us the sample from the same source; they kindly provided us with a small amount of the sample, which we call “IUPAC sample”. In this work, we measure the solubility of CO2 in [hmim][Tf2N] using a gravimetric microbalance6 at temperatures of about 282, 297, 323, and 348 K and pressures up to about 2 MPa. To see any significant differences in the CO2 solubility, two different samples of [hmim][Tf2N] are tested in the present * Address correspondence to this author. E-mail:
[email protected]. † DuPont Central Research and Development, Experimental Station. ‡ DuPont Fluoroproducts Laboratory.
experiments; one source is the IUPAC ultrapure sample, and another is a high-purity commercial sample purchased from EMD Chemicals, Inc. (“EMD sample”). Solubility (pressure-temperature-composition: PTx) data are correlated with our equation of state (EOS) model, which has been successfully used for many mixtures containing RTILs.6-8 The present EOS calculations are compared with solubility data of the present binary system in the literature, and some discussions about inconsistent data will be given. Then, it is shown that this EOS model is not just an empirical fitting function for experimental solubility data but has real predictive power, which any decent theoretical model should possess. The present EOS developed based on vapor-liquidequilibrium (VLE) data alone predicted partial immiscibility (vapor-liquid-liquid equilibria: VLLE) at the CO2-rich side solution. This prediction has been confirmed by performing VLLE experiments similar to our previous works.7,9 Experimental Section and Results Materials. Carbon dioxide (mole fraction purity >0.9999, CAS no. 124-38-9) was purchased from MG Industries (Philadelphia, PA). A molecular sieve trap was installed to remove trace amounts of water from the CO2. The [hmim][Tf2N] (C12H19N3F6O4S2, CAS no. 382150-50-7) was obtained from two sources: National Institute of Standards and Technology (NIST, Boulder, CO) and EMD Chemicals, Inc. (Gibbstown, NJ). NIST supplied 2.0 mL (2.71 g) of [hmim][Tf2N], which was synthesized at the University of Notre Dame (Notre Dame, IN) and purified by NIST as part of IUPAC Project 2002-0051-100. Details about the synthesis and purification can be found in ref 10. A 25 g sample of [hmim][Tf2N] was also purchased from EMD Chemicals, Inc. (mole fraction purity >0.99, Lot EQ500831 632). Both the IUPAC and EMD [hmim][Tf2N] samples were analyzed to verify purity. The initial as-received mass fraction of water was measured by Karl Fischer titration (Aqua-Star C3000, solutions AquaStar Coulomat C and A). The IUPAC sample contained less than 20 ppm of H2O, which was in
10.1021/jp067627+ CCC: $37.00 © 2007 American Chemical Society Published on Web 02/01/2007
Solubility of CO2 in Ionic Liquid [hmim][Tf2N]
J. Phys. Chem. B, Vol. 111, No. 8, 2007 2071
TABLE 1: Chemical Analysis of IUPAC and EMD [hmim][Tf2N] Ionic Liquids IUPAC
EMD
calcd/reported measured calcd/reported measured assay elemental C, % H, % N, % F, % S, % Li, ppm water H2O, ppm halogens F, ppm Cl, ppm Br, ppm
g99.5a
g99.5
g99.0b
g99.0
32.21 4.28 9.39 25.48 14.33 NRd
32.39 4.52 9.42 26.72 15.03 0.273
32.21 4.28 9.39 25.48 14.33 NRd
32.54 4.19 9.50 25.74 14.68 0.203
