Solute Partitioning between the Ionic Liquid 1-n-Butyl-3

Josef Planeta , Lenka Št'avíková , Pavel Karásek , and Michal Roth ... Josef Planeta , Pavel Karásek , Michal Roth ... Michael H. Abraham , Willi...
0 downloads 0 Views 94KB Size
Download PDF
J. Phys. Chem. B 2005, 109, 15165-15171

15165

Solute Partitioning between the Ionic Liquid 1-n-Butyl-3-methylimidazolium Tetrafluoroborate and Supercritical CO2 from Capillary-Column Chromatography Josef Planeta and Michal Roth* Institute of Analytical Chemistry, Academy of Sciences of the Czech Republic, VeVerˇ´ı 97, 61142 Brno, Czech Republic ReceiVed: February 17, 2005; In Final Form: April 25, 2005

Open-tubular capillary-column supercritical fluid chromatography (SFC) with the room-temperature ionic liquid (RTIL) 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF4]) as the stationary liquid and supercritical carbon dioxide (scCO2) as the mobile phase was employed to measure solute retention factors within 313-353 K and 8.1-23.2 MPa. The selection of solutes included 18 compounds of diverse volatilities and chemical functionalities. The retention factors were converted to infinite-dilution solute partition coefficients in the biphasic [bmim][BF4]-scCO2 system. At a constant temperature, an increase in scCO2 density produced distinct shifts in relative retention ()separation factor), thus providing some pressure-tunable selectivity. At a particular temperature and density of CO2, solute partition coefficients can be correlated in terms of linear solvation energy relationships. This important finding indicates a future possibility to estimate the partitioning data in RTIL-scCO2 systems using limited experimental information and the molecular descriptors available for a large variety of prospective solutes. Analysis of the relative retention data by regular solution theory resulted in approximate values of the solubility parameter of CO2-expanded [bmim][BF4].

Introduction Organic salts with melting points near the ambient temperature, or room-temperature ionic liquids (RTILs), enjoy a steady growth in both number and variety of applications.1-3 The most recent efforts include diverse areas such as electrochemistry and solar cells,4 selective organic syntheses,5 catalysis,6,7 membrane separations,8 nanostructure synthesis,9,10 and the use of RTILs in analytical extractions11 and chromatography.12,13 RTILs are also finding their way from academic laboratories to industrial processes.14,15 A significant part of applications benefit from the negligible vapor pressures of most RTILs. The application developments have been backed by detailed characterization of the physical16-19 and thermophysical20,21 properties of pure RTILs by experiments16-21 and molecular dynamics22-25 or Monte Carlo26,27 simulations, though physical data have still been missing for recently synthesized RTILs.28 An important class of RTILs are those containing alkylmethylimidazolium cations. Since these RTILs are largely insoluble in supercritical carbon dioxide (scCO2), Blanchard et al.29,30 suggested the use of scCO2 to extract organic substances from RTILs. Extraction with scCO2 provides for an environment-friendly recovery of nonvolatile or thermally labile nonelectrolytes from a RTIL, thus solving an important issue in the sustainable application of RTILs as reaction media. Besides that, the biphasic RTIL-scCO2 solvent systems have been used in synthetic applications including enzymatic processes,31-34 biphasic catalysis,35-42 and electrochemical reactions.43-45 The growing importance of biphasic RTIL-scCO2 solvent systems has stimulated intense efforts in characterizing the high-pressure phase behavior of the binary systems concerned. Several research groups have been involved in determination of the composition46-53 and volumetric properties46,51 of the imida* Corresponding author. Phone: +420-532-290-171. Fax: +420-541212-113. E-mail: [email protected].

zolium-based RTIL-scCO2 binary systems with some recent attention given to the effects of cosolvents.48,54-56 However, in applications of the biphasic RTIL-scCO2 systems in synthesis and extraction, multicomponent (ternary and higher) phase equilibria involving prospective products and reactants can be even more important than the RTIL-CO2 binary properties. The ternary data will be needed to evaluate material and economic balances of the respective processes as well as to develop predictive thermodynamic models. Despite these emerging needs, very few data sets of this kind have so far been available. Our previous contribution57 employed supercritical fluid chromatography (SFC) to obtain infinite-dilution solute partition coefficients between scCO2 and CO2-expanded 1-n-butyl-3methylimidazolium hexafluorophosphate ([bmim][PF6]). Reasonable agreement has recently been noted58 between the SFC value of the partition coefficient for benzil57 and the value obtained by Sakellarios and Kazarian58 from the combination of transmission-IR and ATR-IR spectroscopy. The present work deals with another weakly coordinating ionic liquid; it is concerned with solute partitioning between scCO2 and CO2expanded 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF4]). As compared with the previous contribution,57 the selection of solutes has been extended here to cover a wider range of solute polarities. This study is a part of our ongoing effort to build up a small database of solute partition coefficients to assist in the development of thermodynamic models for solute-RTIL-scCO2 systems. Experimental Section Materials. Aniline, anisole (methoxybenzene), azulene (bicyclo[5.3.0]decapentaene), benzil (1,2-diphenylethane-1,2dione), benzoic acid, (()-camphor (1,7,7-trimethylbicyclo[2.2.1]heptan-2-one), coumarin (1-benzopyran-2-one), p-cresol (4methylphenol), N,N-dimethylaniline, 1-hexanol, indole (benzo[b]-

10.1021/jp0508251 CCC: $30.25 © 2005 American Chemical Society Published on Web 07/20/2005

15166 J. Phys. Chem. B, Vol. 109, No. 31, 2005 pyrrole), R-ionone [4-(2,6,6-trimethyl-2-cyclohexen-1-yl)-3buten-2-one], N-methylaniline, naphthalene, phenethyl alcohol (2-phenylethanol), phenol, pyrene, and veratrole (1,2-dimethoxybenzene) were purchased from Sigma-Aldrich s.r.o. (Prague, Czech Republic) in the highest purity available and used as received. Carbon dioxide (purity 4.5, mole fraction of residual water 95 mol %), methanol (>99.9 mol %), and methylene chloride (99.9 mol %) were purchased from Sigma-Aldrich. 1-n-Butyl-3methylimidazolium tetrafluoroborate ([bmim][BF4]) (>98 mol %) was supplied by Fluka (Prague, Czech Republic). When attempting to dissolve [bmim][BF4] in methylene chloride to prepare a solution for column coating, we observed the formation of a small amount of very fine white precipitate, probably composed of the products of partial hydrolysis of the tetrafluoroborate anion or of other impurities of ionic nature. Prior to the use for column preparation, therefore, 0.7 mL of the ionic liquid was dissolved in a 5-fold excess (v/v) of methylene chloride, and the fine white precipitate that formed from this operation was separated by centrifugation. From the resultant clear solution, methylene chloride and potentially present volatile organic impurities were driven off by bubbling a gentle stream of CO2 in a temperature-controlled environment, with a final period of 12 h at 130 °C. This purification step produced a clear, nearly colorless [bmim][BF4] that was used for column preparation without any unnecessary delay. Fusedsilica capillary tubing was obtained from CACO s.r.o. (Bratislava, Slovak Republic). Open-Tubular Capillary Column. To improve the wettability of the silica surface with [bmim][BF4], the fused-silica tubing for column preparation was pretreated in the same way as described previously.57 The pretreated tubing was filled with a solution prepared by dissolving 30 µL of purified [bmim][BF4] in 4 mL of methylene chloride and sealed at one end. The other end was exposed to vacuum, and methylene chloride was carefully evaporated, leaving a film of [bmim][BF4] on the inner wall of the capillary tube. The resultant open-tubular capillary column used in retention measurements was 4 m long, had an 85 µm i.d., and contained 9.0 × 10-7 mol of [bmim][BF4]. The equivalent film thickness of pure [bmim][BF4] in the column at 298 K and ambient pressure was 0.16 µm. Apparatus and Procedure. The apparatus used in this study was a modified Varian 3700 gas chromatograph described elsewhere.57 To avoid blending acidic and basic solutes in one injection solution and to secure well resolved peaks at all densities of CO2, the 18 solutes had to be divided among 8 different injection solutions. The solvent was n-hexane, each solution contained naphthalene as a reference, and the concentrations of individual solutes ranged within 1.5-3 mg/mL. The column holdup time was marked by injecting a small amount of methane simultaneously with every injection of solutes.57 The solute retention factors were measured along three isotherms (313, 333, and 353 K) at mean column pressures within 8.123.2 MPa. The pressure drop along the open-tubular column was calculated from the Hagen-Poiseuille equation using the correlation for viscosity of CO2 developed by Vesovic et al.,59 and it did not exceed 0.02 MPa. At a particular isotherm, at least three injections were carried out at each pressure setting, with the retention factors being reproducible to within 1% of the mean value. The measurements were performed in the sequence of increasing pressure. After completing each isotherm, the solute retention factors at the lower pressure limit were measured again to check for any possible loss of [bmim][BF4]

Planeta and Roth from the column during the run. In the same way as in our previous work with [bmim][PF6], we did not observe any loss of [bmim][BF4] from the column in the present measurements. Although [bmim][BF4] has been reported60 to be more soluble in scCO2 than [bmim][PF6], the solubility is still too low to cause a measurable decrease in the amount of [bmim][BF4] in the column during the experimental SFC run. Results and Discussion Below, the solute will be identified by the subscript 1, the principal component of the stationary phase ()[bmim][BF4]) by the subscript 2, and the mobile phase fluid ()CO2) by the subscript 3. The quantities pertaining to the stationary and mobile phases will be denoted by the subscripts s and m, respectively. Partition Coefficients. Retention of a solute in a chromatographic column has customarily been expressed through the solute retention factor

k1 ) (tR - t0)/t0

(1)

where tR is the solute retention time and t0 the column holdup time. Partitioning of the solute between the stationary and mobile phases has usually been described either by the partition coefficient, Kc,

Kc ) c1s/c1m

(2)

where c1s and c1m are the molar concentrations of the solute in the two phases, or by the K-factor

K ) x1m/x1s

(3)

where x1s and x1m are the mole fractions of the solute in the two phases. The relationships used here to convert k1 to Kc and K were the same as those employed previously.57 To ensure consistency in the k1-to-Kc and k1-to-K conversions, we used a single source51 of composition and volumetric data on the liquid phase of the [bmim][BF4]-scCO2 binary system, although there was only a very little overlap in the pressure range of the [bmim][BF4]-scCO2 phase equilibrium data51 and our chromatographic measurements. The density of pure CO2 was calculated from the equation of state (EOS) of Span and Wagner.61 Table 1 shows the relative values of solute partition coefficients, RAB, at 313 K with respect to naphthalene under the particular conditions. The relative values of solute partition coefficients at 333 and 353 K are given in Tables 2 and 3, respectively. The retention factors, partition coefficients, and K-factors of naphthalene are listed in Table 4. In Table 4, the data labeled “RTIL w/ CO2” were calculated employing the composition and volumetric data interpolated from the results of Aki et al.51 The data labeled “RTIL w/o CO2” contain no correction for the dissolution of CO2 in [bmim][BF4]. They were calculated assuming insolubility of CO2 in [bmim][BF4] and incompressibility of pure [bmim][BF4] and were only corrected for the effect of temperature on the geometric volumes of both phases in the column using the temperature dependence of the density of pure [bmim][BF4].19,21 In the k1-to-K conversion, the effect of correction for CO2 dissolution in [bmim][BF4] is more important than that in the k1-to-Kc conversion for the same reasons as discussed elsewhere with [bmim][PF6].57 At a particular temperature and pressure, the uncorrected Kc values listed in Table 4 are somewhat lower than the corresponding data for [bmim][PF6].57 On the basis of preliminary experiments, the relative values of partition coefficients in Tables 1-3 can

Solute Partitioning in the [bmim][BF4]-scCO2 System

J. Phys. Chem. B, Vol. 109, No. 31, 2005 15167

TABLE 1: Relative Values of Solute Partition Coefficients with Respect to Naphthalene at 313 K

TABLE 3: Relative Values of Solute Partition Coefficients with Respect to Naphthalene at 353 K RABa

RABa solute aniline anisole azulene benzil benzoic acid camphor coumarin p-cresol N,N-dimethylaniline 1-hexanol indole R-ionone N-methylaniline phenethyl alcohol phenol pyrene veratrole

8.5, 8.8, 9.2, 10.5, 13.2, P/MPa: 8.1, Fm/kg‚m-3: 289.8 353.9 429.1 532.0 660.1 747.4 6.24 0.239 2.52

6.47 6.78 6.78 0.266 0.242 2.43 2.25 2.12

0.185

0.182

26.8

0.410 0.233 0.305 2.33

1.49

13.3 11.4 18.2 17.9 0.444 0.471 0.510 0.253 0.302 40.7 0.266 0.242 2.40 2.42 2.45 4.48 3.95 3.67 22.8 23.2 1.43

1.31

6.61

6.24

1.94 1.62 5.96 3.63 23.5 19.0 9.00 6.61 16.7 15.2 0.563 0.605 47.9

40.5

2.40 2.31 3.34 2.95 22.5 21.1 16.9 10.5 1.26 1.17

11.8, 13.3, 15.3, 17.6, 23.2, P/MPa: 10.0, Fm/kg‚m-3: 221.6 288.6 352.4 439.9 526.3 658.3

solute aniline anisole azulene benzil benzoic acid camphor coumarin p-cresol N,N-dimethylaniline 1-hexanol indole R-ionone N-methylaniline phenethyl alcohol phenol pyrene veratrole

4.70 0.248 2.46

0.243

0.452 0.228 0.407 2.01

1.41

4.95 5.13 0.290 0.308 2.34 2.22 12.2 22.3 0.243 0.253 14.8 13.8 0.471 0.496 0.254 0.292 46.2 43.3 0.333 0.308 2.07 2.12 4.18 16.1 29.1 1.34 1.27

5.14

4.98

2.00 7.69 5.50 2.92 19.4 16.7 12.3 11.3 12.7 0.533 0.354 38.0

8.46 5.50 11.5 9.2 0.587 32.2

23.6

2.10 2.05 3.47 2.93 2.44 15.8 15.0 12.4 18.8 13.6 6.94 1.17

For solutes A and B, RAB ) kA/kB ) KcA/KcB ) KB/KA (B ) naphthalene).

For solutes A and B, RAB ) kA/kB ) KcA/KcB ) KB/KA (B ) naphthalene).

TABLE 2: Relative Values of Solute Partition Coefficients with Respect to Naphthalene at 333 K

TABLE 4: Retention Factors, Partition Coefficients, and K-Factors of Naphthalene

a

a

RABa solute aniline anisole azulene benzil benzoic acid camphor coumarin p-cresol N,N-dimethylaniline 1-hexanol indole R-ionone N-methylaniline phenethyl alcohol phenol pyrene veratrole

P/MPa: 8.7, 10, 10.9, 12.1, 13.3, 16.7, 21.6, Fm/kg‚m-3: 221.3 290.0 350.4 442.0 523.8 657.0 747.1 5.18 0.223 2.59

0.218

0.434 0.212 0.416 2.14

1.54

5.58 5.76 5.99 5.97 0.260 0.282 0.252 2.44 2.34 2.19 2.02 9.59 6.70 3.83 2.69 24.7 22.2 16.9 13.3 0.212 0.211 0.222 13.0 10.4 6.99 5.13 15.6 14.8 12.8 10.7 0.450 0.462 0.495 0.541 0.235 0.261 0.310 0.363 48.9 43.8 33.6 26.7 0.321 0.282 0.252 2.20 2.23 2.29 2.27 4.40 4.18 3.86 3.47 2.96 2.51 19.2 19.1 17.4 14.9 24.6 17.7 10.1 6.80 1.43 1.36 1.28 1.20

For solutes A and B, RAB ) kA/kB ) KcA/KcB ) KB/KA (B ) naphthalene). a

be estimated to be reproducible within 3% of the mean value, whereas the reproducibility of the absolute partition coefficients in Table 4 can be estimated to be (10% of the mean value. The primary factor determining the accuracy of the data in Table 4 is the uncertainty in the amount of [bmim][BF4] in the column. Phase equilibria in binary RTIL-scCO2 systems have been known46,51 to be affected by the presence of a small amount of water in the system. In imidazolium-based RTILs, both water62 and CO263 have been shown to interact primarily with the anion of the RTIL. The effect of water on RTIL-CO2 equilibrium would certainly transfer into the k1-to-Kc and k1-to-K conversions, if not to the mechanism of chromatographic retention itself. In the SFC experiment, the steady-state concentration of water in the RTIL is determined by the concentration of water in the incoming stream of CO2 and by the partition coefficient of water between the RTIL and scCO2. Since the mole fraction of residual water in the particular grade of CO2 employed in this study was as low as 5 × 10-6 or less, we conclude that the results listed in Tables 1-4 are not likely to be influenced by the presence of moisture.

Kc T/K P/MPa Fm/kg‚m-3 313

333

353

8.1 8.5 8.8 9.2 10.5 13.2 8.7 10 10.9 12.1 13.3 16.7 21.6 10 11.8 13.3 15.3 17.6 23.2

289.8 353.9 429.1 532.0 660.1 747.4 221.3 290.0 350.4 442.0 523.8 657.0 747.1 221.6 288.6 352.4 439.9 526.3 658.3

k1 1.88 1.17 0.417 0.201 0.109 0.0788 2.70 1.16 0.692 0.312 0.175 0.0875 0.0685 1.48 0.743 0.410 0.204 0.122 0.0772

K

RTIL RTIL RTIL RTIL w/ CO2 w/o CO2 w/ CO2 w/o CO2 204 126

302

248 0.00660 0.00323 154 0.00889 0.00425 55.1 0.00983 26.5 0.0165 14.4 0.0245 10.4 0.0298 352 0.00533 0.00294 151 0.00522 90.3 0.00725 40.8 0.0127 22.8 0.0192 11.4 0.0306 8.94 0.0343 190 0.00537 95.7 0.00821 52.8 0.0122 26.3 0.0196 15.7 0.0275 9.95 0.0346

Solute Retention versus Solute Properties. The properties of a system containing an organic nonelectrolyte, a RTIL, and scCO2 obviously result from the very large variety of intermolecular interactions involved. A rigorous thermodynamic modeling of even the binary RTIL-supercritical fluid systems has still been in an infant stage, although important progress has recently been made.64,65 To gain some insight into the partitioning data from SFC, we have therefore taken a more empirical approach based on linear solvation energy relationships (LSERs).66-68 Within the general framework of LSERs, the relative partition coefficient of a solute A with respect to naphthalene ()B) at a particular temperature and pressure can be expressed by

log(KcA/KcB) ) rRA + sπHA + a

∑RHA + b∑βHA + VVA

(4)

where the molecular descriptors of the solute A include the excess molar refraction, RA, the solute dipolarity/polarizability, πHA, the solute “overall” hydrogen-bond acidity, ∑RHA, the solute overall hydrogen-bond basicity, ∑βHA, and McGowan’s

15168 J. Phys. Chem. B, Vol. 109, No. 31, 2005

Planeta and Roth

Figure 1. LSER fit of relative partition coefficients (eq 5).

characteristic volume69 of the solute, VA. The coefficients r, s, a, b, and V are then obtained by regression of experimental values of log(KcA/KcB) on tabulated values70 of the solute descriptors. At 333 K and 12.1 MPa, for example, one obtains

log(KcA/KcB) ) 1.510RA + 0.2929πHA + 2.658 0.9453

∑RHA +

∑ βHA - 2.276VA

(5)

Figure 1 shows a comparison between the experimental data and the values calculated from eq 5. In eq 5, the coefficient at ∑RHA has the largest magnitude, suggesting that, under the particular conditions, the solute retention is rather sensitive to the hydrogen-bond acidity of the solute. However, in regard to the relative weight of individual terms on the right-hand side of eq 5 in the calculated values of log(KcA/KcB), the most important are the terms containing VA and RA. The large magnitude and negative sign of the coefficient at VA indicate that the energy needed to form a cavity for the solute molecule in [bmim][BF4] works against the solute retention. These findings are not limited to the particular conditions of eq 5; at 313 K and 10.5 MPa, with a slightly different set of solutes, the resultant fit is

log(KcA/KcB) ) 1.307RA + 0.08867πHA + 2.699 0.7017

Figure 2. (KcA/KcB)(PAsat/PBsat) as a function of CO2 density at 333 K (B ) naphthalene): O, aniline; b, anisole; 0, azulene; 9, benzoic acid; 4, 1-hexanol; 2, indole; 3, N,N-dimethylaniline; 4, N-methylaniline; ], p-cresol; [, phenethyl alcohol; x, phenol.

∑RHA +

∑ βHA - 1.769VA

(6)

and all the comments given for eq 5 apply to eq 6 as well. The results shown in Figure 1 and eqs 5 and 6 justify a moderate optimism in regard to the future possibility to estimate the RTIL-scCO2 partitioning characteristics using LSER-based correlations. SFC retention measurements with a limited selection of solutes can be used to establish a correlation such as eq 5, and the correlation can be applied to estimate the data for other solutes of interest employing the molecular descriptors available for a broad range of solutes.70 It should be noted that the base of experimental data used to obtain the solute descriptors70 involved neither systems containing RTILs nor systems containing scCO2. The strong effect of RTIL-solute interactions is most readily apparent when considering the range of solute retention factors against the range of solute vapor pressures. At 313 K, for example, the solute retention factors range well within 3 orders of magnitude (see Tables 1 and 4), whereas the solute vapor pressures71,72 span across more than 5 orders of magnitude (anisole 1148 Pa, pyrene 3.6 × 10-3 Pa). To separate the effect of solute volatility from the effects of solute-RTIL and solutescCO2 interactions, it is expedient to consider the quantity (KcA/

Figure 3. KcA/KcB as a function of CO2 density at 333 K (B ) naphthalene): O, aniline; b, anisole; 0, azulene; 9, benzoic acid; 4, 1-hexanol; 2, indole; 3, N,N-dimethylaniline; 4, N-methylaniline; ], p-cresol; [, phenethyl alcohol; x, phenol.

KcB)(PAsat/PBsat), where PAsat and PBsat are the vapor pressure of solute A and the vapor pressure of naphthalene at the particular temperature, respectively. If the liquid phase obeyed the Raoult law and the gas phase were an ideal gas, the quantity (KcA/KcB)(PAsat/PBsat) would equal unity, as the solute retention would be inversely proportional to the solute vapor pressure. The actual situation is illustrated by Figure 2, showing (KcA/ KcB)(PAsat/PBsat) for selected solutes at 333 K as a function of CO2 density. At a particular density, the position of a solute along the vertical axis gives a relative measure of the strength of solute-stationary phase interactions. Figure 2 confirms some expectable relationships regarding the interaction strength, for example, aniline > N-methylaniline > N,N-dimethylaniline or phenol > p-cresol. The somewhat surprising position of benzoic acid in the lower part of the plot may be due to low vapor pressure because of the known tendency of benzoic acid to dimerize73 in the vapor phase. The retention of benzoic acid is comparable to that of phenol, as readily seen from Figure 3, showing the relative retention KcA/KcB at 333 K as a function of the density of CO2. Figure 3 also indicates some shifts in relative retention among the various solutes with changing density of CO2, although the shifts are not large enough to enable an efficient control of selectivity by density changes in process applications of RTIL-scCO2 systems. In 1-hexanol and anilines, the relative retention with respect to naphthalene increases with raising density of CO2, while a decrease is observed in phenols, benzoic acid, phenethyl alcohol, and indole.

Solute Partitioning in the [bmim][BF4]-scCO2 System

J. Phys. Chem. B, Vol. 109, No. 31, 2005 15169

Figure 4. (KcA/KcB)(PAsat/PBsat) at 333 K and 12.1 MPa as a function of the sum of solute hydrogen-bonding descriptors. In “doubly” represented solutes that are solids at 333 K, the lower symbol pertains to the solid and the upper symbol to subcooled liquid.

The quantity (KcA/KcB)(PAsat/PBsat) can also be used to illustrate the relative importance of hydrogen bonding in the solute-[bmim][BF4] interactions. In Figure 4, (KcA/KcB)(PAsat/ PBsat) at 333 K and 12.1 MPa is plotted as a function of the sum of solute hydrogen-bonding descriptors. A major part of the solutes capable of hydrogen bonding are enclosed within the tilted ellipse, indicating a loose correlation between (KcA/ KcB)(PAsat/PBsat) and ∑RHA + ∑βHA. Azulene, benzoic acid, and pyrene are solids at 333 K. In the plot, therefore, these solutes are represented by the vapor pressure of the solid as well as by the vapor pressure of the pure subcooled liquid solute at 333 K. The vapor pressures of subcooled liquids were obtained from the thermodynamic cycle described by Prausnitz et al.74 The fact that naphthalene (the reference solute) is also a solid at 333 K does not bear on the configuration of data points in Figures 2 and 4, as the replacement of the vapor pressure of solid naphthalene with the vapor pressure of subcooled liquid would only shift both plots along the y-axis. Variation of Retention Factors with CO2 Density. At a constant temperature, the retention factor of a solute decreases with increasing density, Fm, of CO2, and the relationship between ln k1 and ln Fm is often nearly linear. Mean numerical values of the slopes (∂ ln k1/∂ ln Fm)T in all solutes and retention isotherms are included in the Supporting Information. In naphthalene, the slopes are slightly less negative than the corresponding data in [bmim][PF6],57 but no clear-cut trend between the two RTILs can be identified in the other solutes. The change of the solute retention factor with the density of CO2 is given by

( ) ∂ ln k1 ∂ ln Fm

T

)

[

( ) ( )]

∂µ∞1s 1 Vj∞1m - Vj∞1s RTβmT ∂x3s

T,P,n2s

∂x3s ∂P

-

T,σ

Vs βsTσ (7) Vm βmT

where R is the molar gas constant, T is the temperature, P is the pressure, Vs and Vm are the geometric volumes of the stationary and mobile phases in the column, respectively, x3s is the mole fraction of CO2 in the CO2-expanded RTIL, µ∞1s is the infinite-dilution chemical potential of the solute in the CO2expanded RTIL, βmT is the isothermal compressibility of CO2, βsTσ is the isothermal compressibility of the CO2-expanded RTIL at saturation with CO2, and Vj∞1m and Vj∞1s are the infinite-dilution partial molar volumes of the solute in CO2 and in the CO2expanded RTIL, respectively. For a particular solute at a

Figure 5. Solubility parameter of CO2-expanded [bmim][BF4] calculated from relative retention factors of anisole (An), R-ionone (Io), and benzil (Be) with respect to naphthalene. The values at zero density of CO2 are square roots of cohesive energy densities of pure [bmim][BF4] obtained from molecular dynamics (MD) simulations25 and from solvent effects on reaction rates (RR).75

particular isotherm, the cause for the different slopes (∂ ln k1/∂ ln Fm)T in columns with [bmim][BF4] and [bmim][PF6] is most likely associated with the terms containing x3s. Recent data on the high-pressure phase behavior of CO2 with imidazoliumbased RTILs51 suggest a difference in the pressure coefficient of CO2 solubility, (∂x3s/∂P)T,σ, between [bmim][BF4] and [bmim][PF6]. The composition derivative of solute chemical potential in eq 7, however, does not lend itself to a direct experimental measurement, and it can only be evaluated from a theoretical model or from a molecular simulation. Solubility Parameter of CO2-Expanded [bmim][BF4]. The relative partition coefficient ()relative retention factor), RAB, of a solute A with respect to a reference solute B ()naphthalene) can be employed to estimate the solubility parameter, δs, of CO2-expanded [bmim][BF4] from57

(V0BL - V0AL)δ2s + 2(V0ALδA - V0BLδB)δs + (V0AL - V0BL)δ2m + 2(V0BLδB - V0ALδA)δm - RT ln RAB ) 0 (8) Apart from RAB, all other quantities needed to estimate δs from eq 8 are pure-component properties, namely, the molar volumes of pure liquid (or subcooled liquid) solutes, V0AL and V0BL, the solubility parameters of pure liquid (or subcooled liquid) solutes, δA and δB, and the solubility parameter, δm, of pure CO2 at the particular temperature and pressure. The derivation of eq 8 and the data sources for the individual properties have been described elsewhere.57 To retain consistency with the previous values of δs for CO2-expanded [bmim][PF6], we have used the same solutes in the present work. The results are shown in Figure 5, and the respective numerical data are included in the Supporting Information. Considering a limited validity of the regular solution theory in systems containing RTILs, a satisfactory degree of consistency was obtained between the SFC-derived values of δs and the square roots of cohesive energy densities of pure [bmim][BF4] at 298 K obtained from molecular dynamics simulations25 and from solvent effects on reaction rates.75 Conclusion In our previous contribution,57 we used SFC retention measurements to obtain [bmim][PF6]-scCO2 partition coefficients for several solutes covering a wide range of volatility. In the present study of partitioning between [bmim][BF4] and scCO2, the selection of model solutes has been extended to cover

15170 J. Phys. Chem. B, Vol. 109, No. 31, 2005 a wider range of solute polarities and to include some hydrogenbond donors. We have found that, at a fixed temperature and pressure, infinite-dilution solute partition coefficients in the [bmim][BF4]-scCO2 system can reasonably be correlated within the linear solvation energy relationships. This finding is the most important result of this work, as it opens a way to estimate the RTIL-scCO2 partition coefficients from the molecular descriptors available for many prospective solutes of diverse structural types.70 Such estimations can be valuable in the engineering design of real-world processes employing biphasic RTIL-scCO2 solvent systems. There is also another reason that the feasibility of SFC measurements in an open-tubular column with [bmim][BF4] is important. At a particular temperature within 10-90 °C, the viscosity of [bmim][BF4] has been reported to be lower than the viscosity of [bmim][PF6] by a factor ranging within 1.63.2.19,21 Therefore, the results of this work give some hope that the SFC studies of solute partitioning in biphasic RTIL-scCO2 systems may be feasible even with still less viscous RTILs. This feature has some potential importance, as future process applications of RTIL-scCO2 solvent systems will probably be focused on low-viscosity RTILs. Acknowledgment. We thank Dr. Joan F. Brennecke (Department of Chemical Engineering, University of Notre Dame, Notre Dame, IN) for a preprint of ref 51. We acknowledge the support of this work by the Grant Agency of the Academy of Sciences of the Czech Republic (Project No. B400310504). Supporting Information Available: Table showing mean values of the slopes (∂ ln k1/∂ ln Fm)T in all solutes and estimated values of the solubility parameter of CO2-expanded [bmim][BF4]. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Welton, T. Chem. ReV. 1999, 99, 2071-2083. (2) Wasserscheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39, 3772-3789. (3) Ionic Liquids As Green SolVents. Progress and Prospects; Rogers, R. D., Seddon, K. D., Eds.; ACS Symposium Series 856; American Chemical Society: Washington, DC, 2003. (4) Buzzeo, M. C.; Evans, R. G.; Compton, R. G. ChemPhysChem 2004, 5, 1106-1120. (5) Earle, M. J.; Katdare, S. P.; Seddon, K. R. Org. Lett. 2004, 6, 707710. (6) Wilkes, J. S. J. Mol. Catal. A 2004, 214, 11-17. (7) Welton, T. Coord. Chem. ReV. 2004, 248, 2459-2477. (8) Scovazzo, P.; Kieft, J.; Finan, D. A.; Koval, C.; DuBois, D.; Noble, R. J. Membr. Sci. 2004, 238, 57-63. (9) Zhou, Y.; Schattka, J. H.; Antonietti, M. Nano Lett. 2004, 4, 477481. (10) Antonietti, M.; Kuang, D. B.; Smarsly, B.; Zhou, Y. Angew. Chem., Int. Ed. 2004, 43, 4988-4992. (11) Luo, H. M.; Dai, S.; Bonnesen, P. V.; Buchanan, A. C.; Holbrey, J. D.; Bridges, N. J.; Rogers, R. D. Anal. Chem. 2004, 76, 3078-3083. (12) Ding, J.; Welton, T.; Armstrong, D. W. Anal. Chem. 2004, 76, 6819-6822. (13) Andre, M.; Loidl, J.; Laus, G.; Schottenberger, H.; Bentivoglio, G.; Wurst, K.; Ongania, K.-H. Anal. Chem. 2005, 77, 702-705. (14) Weyershausen, B.; Lehmann, K. Green Chem. 2005, 7, 15-19. (15) Weyershausen, B.; Hell, K.; Hesse, U. Green Chem. 2005, 7, 283287. (16) Huddleston, J. G.; Visser, A. E.; Reichert, W. M.; Willauer, H. D.; Broker, G. A.; Rogers, R. D. Green Chem. 2001, 3, 156-164. (17) Dzyuba, S. V.; Bartsch, R. A. ChemPhysChem 2002, 3, 161-166. (18) Gu, Z. Y.; Brennecke, J. F. J. Chem. Eng. Data 2002, 47, 339345. (19) Seddon, K. R.; Stark, A.; Torres, M.-J. Viscosity and Density of 1-Alkyl-3-methylimidazolium Ionic Liquids. In Clean SolVentssAlternatiVe Media for Chemical Reactions and Processing; Abraham, M. A., Moens, L., Eds.; ACS Symposium Series 819; American Chemical Society: Washington, DC, 2002; Chapter 4, pp 34-49.

Planeta and Roth (20) Fredlake, C. P.; Crosthwaite, J. M.; Hert, D. G.; Aki, S. N. V. K.; Brennecke, J. F. J. Chem. Eng. Data 2004, 49, 954-964. (21) Tokuda, H.; Hayamizu, K.; Ishii, K.; Abu Bin Hasan Susan, M.; Watanabe, M. J. Phys. Chem. B 2004, 108, 16593-16600. (22) Hanke, C. G.; Price, S. L.; Lynden-Bell, R. M. Mol. Phys. 2001, 99, 801-809. (23) Morrow, T. I.; Maginn, E. J. J. Phys. Chem. B 2002, 106, 1280712813. (24) de Andrade, J.; Bo¨es, E. S.; Stassen, H. J. Phys. Chem. B 2002, 106, 13344-13351. (25) Liu, Z. P.; Huang, S. P.; Wang, W. C. J. Phys. Chem. B 2004, 108, 12978-12989. (26) Shah, J. K.; Brennecke, J. F.; Maginn, E. J. Green Chem. 2002, 4, 112-118. (27) Shah, J. K.; Maginn, E. J. Fluid Phase Equilib. 2004, 222-223, 195-203. (28) Marsh, K. N.; Boxall, J. A.; Lichtenthaler, R. Fluid Phase Equilib. 2004, 219, 93-98. (29) Blanchard, L. A.; Hancu, D.; Beckman, E. J.; Brennecke, J. F. Nature 1999, 399, 28-29. (30) Blanchard, L. A.; Brennecke, J. F. Ind. Eng. Chem. Res. 2001, 40, 287-292. (31) Lozano, P.; de Diego, T.; Carrie´, D.; Vaultier, M.; Iborra, J. L. Chem. Commun. 2002, 692-693. (32) Reetz, M. T.; Wiesenho¨fer, W.; Francio, G.; Leitner, W. Chem. Commun. 2002, 992-993. (33) Lozano, P.; de Diego, T.; Gmouh, S.; Vaultier, M.; Iborra, J. L. Biotechnol. Prog. 2004, 20, 661-669. (34) Reetz, M. T.; Wiesenho¨fer, W.; Francio, G.; Leitner, W. AdV. Synth. Catal. 2003, 345, 1221-1228. (35) Liu, F.; Abrams, M. B.; Baker, R. T.; Tumas, W. Chem. Commun. 2001, 433-434. (36) Brown, R. A.; Pollet, P.; McKoon, E.; Eckert, C. A.; Liotta, C. L.; Jessop, P. G. J. Am. Chem. Soc. 2001, 123, 1254-1255. (37) Sellin, M. F.; Webb, P. B.; Cole-Hamilton, D. J. Chem. Commun. 2001, 781-782. (38) Jessop, P. G.; Stanley, R. R.; Brown, R. A.; Eckert, C. A.; Liotta, C. L.; Ngo, T. T.; Pollet, P. Green Chem. 2003, 5, 123-128. (39) Ballivet-Tkatchenko, D.; Picquet, M.; Solinas, M.; Francio, G.; Wasserscheid, P.; Leitner, W. Green Chem. 2003, 5, 232-235. (40) Webb, P. B.; Sellin, M. F.; Kunene, T. E.; Williamson, S.; Slawin, A. M. Z.; Cole-Hamilton, D. J. J. Am. Chem. Soc. 2003, 125, 1557715588. (41) Solinas, M.; Pfaltz, A.; Cozzi, P. G.; Leitner, W. J. Am. Chem. Soc. 2004, 126, 16142-16147. (42) Gao, L. A.; Tao, J. A.; Zhao, G. Y.; Mu, T. C.; Wu, W. Z.; Hou, Z. S.; Han, B. X. J. Supercrit. Fluids 2004, 29, 107-111. (43) Zhang, J. M.; Yang, C. H.; Hou, Z. S.; Han, B. X.; Jiang, T.; Li, X. H.; Zhao, G. Y.; Li, Y. F.; Liu, Z. M.; Zhao, D. B.; Kou, Y. New J. Chem. 2003, 27, 333-336. (44) Zhao, G. Y.; Jiang, T.; Wu, W. Z.; Han, B. X.; Liu, Z. M.; Gao, H. X. J. Phys. Chem. B 2004, 108, 13052-13057. (45) Zhao, G. Y.; Jiang, T.; Han, B. X.; Li, Z. H.; Zhang, J. M.; Liu, Z. M.; He, J.; Wu, W. Z. J. Supercrit. Fluids 2004, 32, 287-291. (46) Blanchard, L. A.; Gu, Z.; Brennecke, J. F. J. Phys. Chem. B 2001, 105, 2437-2444. (47) Kamps, A. P.-S.; Tuma, D.; Xia, J. Z.; Maurer, G. J. Chem. Eng. Data 2003, 48, 746-749. (48) Liu, Z. M.; Wu, W. Z.; Han, B. X.; Dong, Z. X.; Zhao, G. Y.; Wang, J. Q.; Jiang, T.; Yang, G. Y. Chem.sEur. J. 2003, 9, 3897-3903. (49) Shariati, A.; Peters, C. J. J. Supercrit. Fluids 2004, 29, 43-48. (50) Shariati, A.; Peters, C. J. J. Supercrit. Fluids 2004, 30, 139-144. (51) Aki, S. N. V. K.; Mellein, B. R.; Saurer, E. M.; Brennecke, J. F. J. Phys. Chem. B 2004, 108, 20355-20365. (52) Costantini, M.; Toussaint, V. A.; Shariati, A.; Peters, C. J.; Kikic, I. J. Chem. Eng. Data 2005, 50, 52-55. (53) Kroon, M. C.; Shariati, A.; Costantini, M.; van Spronsen, J.; Witkamp, G.-J.; Sheldon, R. A.; Peters, C. J. J. Chem. Eng. Data 2005, 50, 173-176. (54) Scurto, A. M.; Aki, S. N. V. K.; Brennecke, J. F. J. Am. Chem. Soc. 2002, 124, 10276-10277. (55) Zhang, Z. F.; Wu, W. Z.; Liu, Z. M.; Han, B. X.; Gao, H. X.; Jiang, T. Phys. Chem. Chem. Phys. 2004, 6, 2352-2357. (56) Zhang, Z. F.; Wu, W. Z.; Gao, H. X.; Han, B. X.; Wang, B.; Huang, Y. Phys. Chem. Chem. Phys. 2004, 6, 5051-5055. (57) Planeta, J.; Roth, M. J. Phys. Chem. B 2004, 108, 11244-11249. (58) Sakellarios, N. I.; Kazarian, S. G. J. Chem. Thermodyn. 2005, 37, 621-626. (59) Vesovic, V.; Wakeham, W. A.; Olchowy, G. A.; Sengers, J. V.; Watson, J. T. R.; Millat, J. J. Phys. Chem. Ref. Data 1990, 19, 763-808. (60) Wu, W. Z.; Li, W. J.; Han, B. X.; Jiang, T.; Shen, D.; Zhang, Z. F.; Sun, D. H.; Wang, B. J. Chem. Eng. Data 2004, 49, 1597-1601.

Solute Partitioning in the [bmim][BF4]-scCO2 System (61) Span, R.; Wagner, W. J. Phys. Chem. Ref. Data 1996, 25, 15091596. (62) Cammarata, L.; Kazarian, S. G.; Salter, P. A.; Welton T. Phys. Chem. Chem. Phys. 2001, 3, 5192-5200. (63) Cadena, C.; Anthony, J. L.; Shah, J. K.; Morrow, T. I.; Brennecke, J. F.; Maginn, E. J. J. Am. Chem. Soc. 2004, 126, 5300-5308. (64) Shariati, A.; Peters, C. J. J. Supercrit. Fluids 2003, 25, 109-117. (65) Ally, M. R.; Braunstein, J.; Baltus, R. E.; Dai, S.; DePaoli, D. W.; Simonson, J. M. Ind. Eng. Chem. Res. 2004, 43, 1296-1301. (66) Kamlet, M. J.; Taft, R. W. J. Am. Chem. Soc. 1976, 98, 377-383. (67) Taft, R. W.; Kamlet, M. J. J. Am. Chem. Soc. 1976, 98, 28862894. (68) Kamlet, M. J.; Abboud, J. L.; Taft, R. W. J. Am. Chem. Soc. 1977, 99, 6027-6038. (69) Abraham, M. H.; McGowan, J. C. Chromatographia 1987, 23, 243-246.

J. Phys. Chem. B, Vol. 109, No. 31, 2005 15171 (70) Abraham, M. H.; Chadha, H. S.; Whiting, G. S.; Mitchell, R. C. J. Pharm. Sci. 1994, 83, 1085-1100. (71) NIST Chemistry WebBook. NIST Standard Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.; National Institute of Standards and Technology: Gaithersburg, MD, 2003 (http:// webbook.nist.gov/chemistry). (72) Dykyj, J.; Repa´sˇ, M. Tlak nasy´ tenej pary organicky´ ch zlu´ cˇ enı´n (Saturation Vapor Pressure of Organic Compounds); Veda: Bratislava, Slovak Republic, 1979. (73) Murata, S.; Sakiyama, M.; Seki, S. J. Chem. Thermodyn. 1982, 14, 723-731. (74) Prausnitz, J. M.; Lichtenthaler, R. N.; Gomes de Azevedo, E. Molecular Thermodynamics of Fluid-Phase Equilibria; Prentice Hall: Upper Saddle River, NJ, 1999; pp 638-641. (75) Swiderski, K.; McLean, A.; Gordon, C. M.; Vaughan, D. H. Chem. Commun. 2004, 2178-2179.