Elsevier

Fluid Phase Equilibria

Volume 387, 15 February 2015, Pages 59-72
Fluid Phase Equilibria

Study of interaction between organic compounds and mono or dicationic oxygenated ionic liquids using gas chromatography

https://doi.org/10.1016/j.fluid.2014.11.020Get rights and content

Highlights

  • Measurements of activity coefficients at infinite dilution using GLC.

  • Organic solvents in mono and dicationic based ionic liquids.

  • High selectivity for different problems of separation.

  • LSER correlation was presented.

Abstract

Interactions of selected organic compounds with 16 functionalized based mono or dicationic ionic liquids were studied using inverse gas chromatography. Retention data were used to determine the selectivity and capacity values with these ionic liquids for four separation problems: hexane/benzene, hexane/methanol, hexane/thiophene, and cyclohexane/thiophene. It was found that the choice of the anion, the cation and the functionalized chain play an important role in the efficiency of the separation process. The solvation characteristics of this class of ionic liquids were also evaluated using linear solvation energy relationships (LSERs). Results indicate that mono and dicationic ionic liquids have similar solvation properties.

Introduction

This last decade, ionic liquids (ILs) are being widely promoted as probable substitutes for traditional industrial solvents such as volatile organic compounds in a host of processes. This class of solvent frequently combines the attractive features of excellent chemical stability, high thermal stability, and exceedingly low vapor pressure in a single solvent. ILs are commonly comprised of an asymmetric, bulky organic cation paired with a weakly coordinating anion that may be organic or inorganic in nature. It is now well established that the physico-chemical properties of these solvents strongly depend on the structure of the IL (anion, functionality presented at the cationic or anionic site). Nowadays, it is well established that ILs are among the most intriguing and diverse alternative media available not only for conventional solvent-driven chemical processes like synthesis and catalysis, but also electrolytes, lubricants, and modifiers of mobile and stationary phases within the separation sciences [1], [2], [3]. Numerous works have shown that a large number of ILs exhibit selectivities and capacities better than the solvents typically employed to solve industrial separation problems [4], [5], [6], [7], [8], [9], [10]. For instance, Kekra-Krolik et al. have shown that ILs have a good capacity for deep fuel desulfurization as well as for denitrogenation [11], [12]. Ionic liquids can be also used for the separation of azeotropic mixtures [13], [14]. Marciniak analyzed the influence of the cation and anion structures of the ionic liquid but also the effect of the temperature on the selectivity and the capacity for aliphatics/aromatics and n-hexane/hex-1-ene separation problems [15]. The author showed that the highest values of selectivity is observed with ionic liquids containing small alkyl chains, e.g., based on following cations [MMIM]+, [EMIM]+, [EPY]+, [Et3S]+ coupled to a thiocyanate group in the structure. Unfortunately, when the dialkylimidazolium based ionic liquids reveals high values of the selectivity, the capacity always takes low values. In our previous work, the performance of seven oxygenated trigeminal tricationic ionic liquids (TTILs) was evaluated for different separation problems. The different selectivity and capacity values obtained on these TTILs show that their structure has an important influence on the efficiency of the separation process. It was found that some TTILs have higher selectivity and capacity at infinite dilution than the generally used organic solvents such as NMP or sulfolane, as well as many other ILs in the separation of aromatic, alcohol from aliphatic hydrocarbons. A good capacity is obtained by a moderate lengthening in the alkyl chain grafted to the imidazolium cation.

Different approach may be used to predict the behavior of organic compounds with ionic liquids. Among others, the COSMO-RS model developed by Klamt et al. is a valuable method for predicting the thermodynamic properties of ILs mixtures on the basis of quantum chemical calculations for the individual molecules [16], [17]. Several articles have demonstrated the general suitability of COSMO-RS method to predict the solubilities of several gases in ILs [18], [19]. Moreover, an important feature is that the different intermolecular interactions between the mixture components can be quantified by COSMO-RS. Quantitative structure-activity relationship (QSAR) correlations or quantitative structure–property relationships (QSPR) are proposed to determine physical properties of ILs [20], [21], [22]. Katritzky et al. [21] correlated the melting points of the imidazolium and benzimidazolium bromides by using a QSPR approach. Group contribution methods have been applied to develop QSPR for estimating physicochemical properties of chemicals such as melting point [23], boiling point, vapor pressure, octanol–water partition coefficient, water solubility [24], toxicity [25], viscosity [26], and conductivity [26]. Recently, Gardas and Coutinho [27] proposed a group contribution method based on the Vogel–Tammann–Fulcher (VTF) equation for the prediction of viscosity and conductivity of ILs. The main advantage of using the group contributions approach is its simplicity.

Gas chromatography is a good tool to understand the behaviour of the solutes and the stationary phase through the measurements of partition coefficients or activity coefficients at infinite dilution [1], [6], [10]. Crowhurst et al. [28] proposed to characterize the physico-chemical properties of ionic liquids using the Kamlet and Taft’s approach [29], [30], [31], [32]. Parameters such as hydrogen bond acidity, hydrogen bond basicity and dipolarity/polarizability are determined using UV–vis technique. Similar approaches based on gas chromatography technique were proposed to quantify various intermolecular solute-IL interactions. Among others, Abraham et al. have developed the linear solvation energy relationship model (LSER) allowing to correlate thermodynamic properties of phase transfer processes [33], [34], [35], [36]. Abraham solvation parameter model for both the gas-to-solvent partition coefficient, KL, and the water-to-solvent partition coefficient, P have the following expression:Log KL=c+e×E+s×S+a×A+b×B+l×LLogP=c+e×E+s×S+a×A+b×B+v×VThe dependent variables in Eqs. (1) and (2) are solute descriptors as follows: E and S refer to the excess molar refraction in units of (cm3 mol−1)/10 and a dipolarity/polarizability description of the solute, respectively, A and B are measures of the solute hydrogen-bond acidity and basicity, V is the McGowan volume in units of (cm3 mol−1)/100, and L is the logarithm of the gas-to-hexadecane partition coefficient at 298 K. The coefficients c, e, s, a, b and l (or v) are not simply fitting coefficients, but they reflect complementary properties of the solvent phase.

The system constants are identified as the opposing contributions of cavity formation and dispersion interactions, l, the contribution from interactions with lone pair electrons, e, the contribution from dipole-type interactions, s, the contribution from the hydrogen-bond basicity of the stationary phase (because a basic phase will interact with an acid solute), a and b the contribution from the hydrogen-bond acidity of the stationary phase. The system constants are determined by multiple linear regression analysis of experimental log SP (log KL in this work) values for a group of solutes of sufficient number and variety to establish the statistical and chemical validity of the model.

Acree and co-workers reported mathematical correlations based on Abraham’s solvation model for the gas-to-ionic liquid, KL, and water-to-ionic liquid, P, partition coefficients [37], [38], [39]. Sprunger et al. [40], [41], [42], [43] modified Abraham’s solvation model by rewriting each of the six solvent equation coefficients as a summation of their respective cation and anion contribution. Recently, our team proposed to estimate LSER coefficients using a group contribution method [44]. The group contribution model coupled to LSER (GC-LSER) enables one to predict with good accuracy log KL and log P at 298 K of not only alkyl based ionic liquids but also task-specific ionic liquids.

In this work, gas–liquid chromatography was used to quantify the interaction between organic compounds and 16 ILs composed of a polar alkyl chain grafted on the cation and of tetrafluoroborate [BF4] or dicyanamide [N(CN)2] or bis(trifluoromethylsulfonyl)imide [N(Tf)2] as anion. These ILs are 1-methyl-3-propoxymethylimidazolium bis(trifluoromethyl-sulfonyl)imide [IL 1], 1-methyl-3-propoxymethylimidazolium tetrafluoroborate [IL 2], 1-methyl-3-propoxymethylimidazolium dicyanamide [IL 3], 2-methyl-1-octyl-3-propoxymethylimidazolium bis(trifluoromethylsulfonyl)imide [IL 4], 2-methyl-1-octyl-3-propoxymethylimidazolium dicyanamide [IL 5], 1-benzyl-3-propoxymethylimidazolium bis(trifluoromethyl-sulfonyl)imide [IL 6], 1-benzyl-3-propoxymethylimidazolium tetrafluoroborate [IL 7], 1-benzyl-3-propoxymethylimidazolium dicyanamide [IL 8], 3,3′-[1,7-(2,6-dioxaheptane)]bis(1-methylimidazolium)bis(trifluoromethylsulfonyl)imide [IL 9], 3,3′-[1,7-(2,6-dioxaheptane)]bis(1-methylimidazolium)dicyanamide [IL 10], 3,3′-[1,7-(2,6-dioxaheptane)]bis(2-methy-1-octylimidazolium)bis(trifluoromethylsulfonyl)imide [IL 11], 3,3′-[1,7-(2,6-dioxaheptane)]bis(2-methy-1-octylimidazolium)tetrafluoroborate [IL 12], 3,3′-[1,7-(2,6-dioxaheptane)]bis(2-methy-1-octylimidazolium)dicyanamide [IL 13], 3,3′-[1,7-(2,6-dioxaheptane)]bis(1-benzylimidazolium)dicyanamide [IL 14], 4-dimethylamino-1-propoxymethyl pyridinium bis(trifluoromethylsulfonyl)imide [IL 15], 1,1′-[1,7-(2,6-dioxaheptane)]bis(4-dimethylaminopyridinium)bis(trifluoromethylsulfonyl) imide [IL 16]. Their molecular structures are given in Table 1.

To our knowledge, this is the first time that activity coefficients have been measured for solutes dissolved in these mono or dicationic oxygenated ILs. The results of these activity coefficient measurements are used to calculate log KL and log P values for solutes dissolved in anhydrous ILs and to derive predictive Abraham model for describing solute transfer into these families of ILs from both the gas phase.

Section snippets

Synthesis of tested ILs

Ionic liquids studied in this work were synthesized as previously described [45], [46], [47]. The nomenclature and structure of ILs are shown in Table 1. All obtained ILs are air-stable under ambient conditions and may be handled under normal laboratory conditions, a purity of 99.5%.

In addition to the treatment mentioned above, each IL was further purified by subjecting the liquid to a very low pressure of about 5 Pa at 343 K for approximately 24 h. Next, packed columns were conditioned for a 12 h

Theoretical basis

Activity coefficients at infinite dilution for solute 1 in IL 2, γ1,2, were calculated with the following expression [48]:lnγ1,2=ln(n2×R×TVN×P10)P10×B11V10RT+2×B13V1RT×J×P0where n2 is the number of moles of stationary phase component within the column, R is the gas constant, T is the oven temperature, VN is the net retention volume, B11 is the second virial coefficient of the solute in the gaseous state at temperature T, B13 is the mutual virial coefficient between solute 1 and the

Activity coefficients at infinite dilution of organic compounds in ionic liquids

The uncertainty of γi values may be obtained from the law of propagation of errors. The following measured parameters exhibit uncertainties which must be taken into account in the error calculations with their corresponding standard deviations: the adjusted retention time tR, ±0.01 min; the flow rate of the carrier gas, ±0.1 cm3/min; mass of the stationary phase, ±2%; the inlet and outlet pressures, ±0.002 bar; the temperature of the oven, ±0.1 K. The main source of uncertainty in the calculation

Conclusion

In this work, the performance of 16 new mono or dicationic ionic liquids was evaluated for different separation problems. The different selectivity and capacity values obtained on these mono and dicationic ILs show that their structure has an important influence on the efficiency of the separation process. It was found that the best selectivity and capacity at infinite dilution values are obtained with monocationic ionic liquids. The performance of these ILs are similar than the generally used

References (53)

  • F. Mutelet et al.

    J. Chromatogr. A

    (2006)
  • U. Domanska et al.

    J. Chem. Thermodyn.

    (2013)
  • A. Marciniak

    Fluid Phase Equilib.

    (2010)
  • V. Dohnal et al.

    Chem. Eng. J.

    (2014)
  • A. Marciniak

    Fluid Phase Equilib.

    (2010)
  • M.M. Diedenhofen et al.

    Fluid Phase Equilib.

    (2010)
  • J. Palomar et al.

    J. Sep. Purif. Technol.

    (2011)
  • N. Sun et al.

    Fluid Phase Equilib.

    (2006)
  • P. Luis et al.

    Ecotoxicol. Environ. Saf.

    (2007)
  • H. Matsuda et al.

    Fluid Phase Equilib.

    (2007)
  • M.H. Abraham et al.

    J. Chromatogr.

    (1991)
  • L.M. Sprunger et al.

    Fluid Phase Equilib.

    (2008)
  • L.M. Sprunger et al.

    Fluid Phase Equilib.

    (2009)
  • P. Jensen et al.

    J. Solid State Chem.

    (1999)
  • D.W. Armstrong et al.

    Anal. Chem.

    (1999)
  • G.W. Meindersma et al.

    Monatsh. Chem.

    (2007)
  • X. Han et al.

    Acc. Chem. Res.

    (2007)
  • A.L. Revelli et al.

    J. Chem. Eng. Data

    (2009)
  • A.L. Revelli et al.

    J. Chem. Eng. Data

    (2008)
  • F. Mutelet et al.

    J. Chem. Eng. Data

    (2012)
  • F. Mutelet et al.

    J. Chem. Eng. Data

    (2006)
  • K. Kędra-Krolik et al.

    Ind. Eng. Chem. Res.

    (2011)
  • K. Kędra-Krolik et al.

    Energy Fuels

    (2011)
  • M. Seiler et al.

    AIChE J.

    (2004)
  • A. Klamt et al.

    Annu. Rev. Chem. Biomed. Eng.

    (2010)
  • M. Gonzalez-Miquel et al.

    Ind. Eng. Chem. Res.

    (2011)
  • Cited by (16)

    • Determination of physicochemical properties of ionic liquids by gas chromatography

      2021, Journal of Chromatography A
      Citation Excerpt :

      The experimental procedure employed by these authors has changed little in contemporary studies. Infinite dilution activity coefficients for ionic liquids have been determined by gas chromatography at temperatures within the range 25°C < T < 105°C, support loadings typically from 30 to 60 % (w/w), and on columns generally ≤ 1 m, Table 3 [27,35-38,47,50,56,67-71,78-236]. Columns of different length and phase loading are used to accommodate the wide range of retention times and to evaluate the repeatability of the measurements.

    • Characterization of the solubilizing ability of tetraalkylammonium ionic liquids containing a pendant alkyl chain bearing a basic N,N-dimethylamino or N,N-dimethylaminoethoxy functionality

      2019, Journal of Molecular Liquids
      Citation Excerpt :

      The estimated uncertainty for activity coefficients was 3%, taking into account the possible uncertainties in determining the retention data, the solute thermodynamic properties, and the column loading. It should be pointed out that the solute gas phase partition coefficients into water (Kw) required in Eq. (3) are available from our earlier publications [11–25]. It should also be noted that water-to-IL partition coefficients calculated through Eq. (3) pertain to a hypothetical partitioning process involving solute transfer from water to the anhydrous IL.

    View all citing articles on Scopus
    View full text