20 Acta Chim. Slov. 2007, 54, 20–24 Studzi×ska et al.: Chromatographic and Chemometric Methods for Evaluation ... 1. Introduction The interest in properties of room-temperature-ionic liquids is rapidly expanding. In an ionic liquid, the cation is generally large and the anion is either large or small, re- sulting in poorer packing of the larger ions with weaker attraction, so the compound tends to remain liquid. Figure 1 presents the schematic structure of the most commonly used imidazolium ionic liquids. It is this liquid state that is the key to ionic liquids’ useful properties, which include high thermal stability, negligible vapour pressure and good electrochemical stability – making them attractive to any industry using solvents.1–4 Due to their non-volatile feature, the most attractive application of ionic liquids is to replace the traditional volatile organic solvents, which are harmful to the environment. Other features, such as reusability, low viscosity, better reactivity and selectivity, also ensure ionic liquids to be a good candidate for “green chemistry”.1–4 There have been numerous studies concer- ning their preparation, their use as a reaction medium and their physical properties. Therefore, this subject has beco- me especially interesting, because of the possibility of mi- gration of ionic liquids to the environment. The wide ap- plicability of these compounds is the main reason of their use on industrial scale. Figure 1. Schematic structure of imidazolium ionic liquids. As ionic liquids are non-volatile, their presence in air is not probable, in comparison with the possibility of their occurrence in water and soil.5–7 This is the main rea- son why estimation of ionic liquids properties is so impor- tant and is the subject of many works. In other words: the migration of these substances in the environment will de- pend on the interactions between the compounds and wa- ter or soil. It is important to know these interactions, be- cause this will give us a possibility to understand mecha- nisms responsible for ionic liquids activity in the environ- ment. In an attempt to understand the nature of interac- tions of room-temperature-ionic liquids, high performan- ce liquid chromatography may be used. One of the way of Scientific paper Chromatographic and Chemometric Methods for Evaluation of Properties of Ionic Liquids Sylwia Studzi×ska,a Piotr Stepnowski,b Bogusĺaw Buszewskia* a Department of Environmental Chemistry and Bioanalytics, Nicolaus Copernicus University, Faculty of Chemistry, 7 Gagarin Str., PL-87-100 Toruñ, Poland. Tel.: +48(56)6114308, Fax: +48(56)6114837, E-mail: bbusz@chem.uni.torun.pl. b Waste Management Laboratory, Faculty of Chemistry, University of Gda×sk, Sobieskiego 18 Str., PL 80-952 Gda×sk, Poland Received: 19-10-2006 Paper based on a presentation at the 12th International Symposium on Separation Sciences, Lipica, Slovenia, September 27–29, 2006. Abstract The interest in ionic liquids is in constant increase due to their popularity in various applications. The wide applicability of these compounds is the main reason of their use on industrial scale, especially in organic synthesis. This leads to the presence of ionic liquids in water and soils. Therefore, there is a need to study their toxicology and to better understand the basis of it. The main aim of the contribution is to review the existing knowledge about the influence of ionic liquids on living orga- nisms and to provide information about the use of chromatographic and chemometric methods in evaluation of proper- ties of ionic liquids. Keywords: ionic liquids, chromatography, stationary phase, chemometric, QSAR 21Acta Chim. Slov. 2007, 54, 20–24 Studzi×ska et al.: Chromatographic and Chemometric Methods for Evaluation ... changing and optimizing the chromatographic results is to change the column type. This means a change in the stationary phase. Such alterations may be also used to study the interactions between stationary phase ligands and the analyzed compound. Ionic liquids may interact with other compounds in several different ways: van der Waals forces, electrostatic, π…π. However, there is little knowledge about what interaction is the strongest. The main aim of the present contribution was to provide infor- mation about the usefulness of chromatographic and che- mometric methods to estimate the properties of ionic li- quids. 2. Experimental 2.1. Materials and Reagents Standards of 1-n-propyl-3-methyl-imidazolium te- trafluoroborate (PMIM), 1-n-butyl-3-methyl-imidazo- lium tetrafluoroborate (BMIM), 1-n-amyl-3-methyl-imi- dazolium tetrafluoroborate (AMIM), 1-n-hexyl-3- methyl-imidazolium tetrafluoroborate (HMIM) were ob- tained from E. Merck (Darmstadt, Germany) and also supplied by professor Bernd Jastorff (University of Bre- men, Germany). Schematic structures of analyzed ionic liquids and their main properties are given in Table 1. For the preparation of the mobile phase, methanol of “for HPLC” purity (S. Witko, ¬ódź, Poland) was used, as well as deionized water purified using a Milli-Q sys- tem (Millipore, El Passo, TX, USA), KH2PO4 (POCh, Gliwice, Poland), 85% solution of HPLC grade orthop- hosphoric acid (J. T. Baker, Deventer, The Netherlands). In the current studies a series of homemade packing materials with different surface ligands have been utili- zed: cholesterol (SG-CHOL); n-acylamide (SG-CHOL, SG-AP); aminopropyl (SG-CHOL, SG-AP, SG-MIX); cyanopropyl, phenyl, octadecyl, octyl, butyl (SG-MIX, SG-Ph), naphthalene (SG-Ar) and residual silanols locali- Stationary Column Abbreviation Column Silica particle Pore phase type dimensions (mm) size (µm) diameter (Å) octadecyl Gemini 5µ C18 110 Å Gemini 250 × 4.6 5 110 octadecyl, end-capped RP-18e PurospherTM Star Star 250 × 4.6 5 120 octyl Macrosphere 300 C8 5µm SG-C8 250 × 4.6 5 300 butyl Macrosphere 300 C4 5µm SG-C4 250 × 4.6 5 300 alkylamide alkylamide SG-AP 250 × 4.6 5 100 cholesterolic cholesterolic SG-CHOL 250 × 4.6 5 100 mixed mixed SG-MIX 250 × 4.6 7 - phenyl phenyl SG-Ph 125 × 4.6 4.6 - aryl aryl SG-Ar 125 × 4.6 4.6 - Name Abbreviation Structure Molar mass (g mol–1) 1-n-propyl-3-methyl- PMIM 212 imidazolium tetrafluoroborate 1-n-butyl-3-methyl- BMIM 226 imidazolium tetrafluoroborate 1-n-amyl-3-methyl- AMIM 240 imidazolium tetrafluoroborate 1-n-hexyl-3-methyl- HMIM 254 imidazolium tetrafluoroborate Table 1. Structure and basic properties of ionic liquids used. Table 2. Characteristics of HPLC columns used in the study. 22 Acta Chim. Slov. 2007, 54, 20–24 Studzi×ska et al.: Chromatographic and Chemometric Methods for Evaluation ... zed on the silica gel surface of all packings. The reaction mechanism and the conditions for synthesis of alkylami- de, cholesterolic, mixed, phenyl and aryl stationary phases synthesis are given in the literature for: SG-AP,8 SG- CHOL,9 SG-MIX,10 SG-Ph,11 SG-Ar.11 A commercial Macrosphere 300 C4 5 (Alltech, Deerfield, IL, USA; SG- C4); Macrosphere 300 C8 5 (Alltech, Deerfield, IL, USA; SG-C8); Gemini 5µ C18 110 Å (Phenomenex, Torrance, CA, USA) (Gemini); RP-18e PurospherTM Star (E. Merck, Darmstadt, Germany; Purospher) columns have also been applied in the study (Table 2). 2.2. Instrument and Chromatographic Conditions A LC-10Avp (Shimadzu, Kyoto, Japan) high per- formance liquid chromatographic system equipped with a diode-array detector DAD (Shimadzu, Kyoto, Japan), a Rheodyne 7125 manual injection valve (Rheodyne, Ber- keley, CA, USA) with a 20-(micro)L loop were selected for chromatographic measurements. CLASS-VP program was used for the data collection. The elution was carried out with isocratic condi- tions of: 95% v/v 40 mmol L–1 potassium phosphate buf- fer (adjusted with orthophosphoric acid to pH = 4) and 5% v/v methanol in case of all packings except SG-Ar and SG-Ph for which conditions were: 50% v/v 40 mmol L–1 potassium phosphate buffer (adjusted with orthop- hosphoric acid to pH = 4) and 50% v/v methanol. In case of log kW,SG-CHOL determination 40 mmol L –1 potassium phosphate buffer adjusted to pH = 7 was used. The flow rate was 1 ml/min. In the interpretation of the results, the HyperChem v.5.1 package with the ChemPlus extension (HyperCube, Waterloo, Canada) was used. 3. Results and Discussion Chromatographic analysis of ionic liquids on vari- ous column packing types allows for prediction of princi- pal interactions responsible for their retention. Results obtained for 1-butyl-3-methylimidazolium cation are pre- sented in Figure 2 together with a schematic illustration of stationary phase structures and interactions. The rest of results are gathered in Figure 3. The ob- tained data clearly indicate a decrease of the ionic liquid cation retention with a larger alkyl ligand. It was found that the analyzed cations interact stronger with surface of the packing by hydrophobic van der Waals forces. On the other hand, the investigations performed with the use of phenyl and aryl stationary phase showed intense interactions bet- ween ionic liquids cations and chemically bonded aromatic molecules. The retention values of ionic liquids analyzed on these two packings in mobile phase consist of only 5% MeOH v/v were so high, that the analytical conditions had to be changed. Methanol in the mobile phase was increased to 50% v/v. This indicates stronger π…π interaction types in comparison with van der Waals interactions. It can be concluded that main interactions in ionic liquids retention mechanism are: dispersive and π…π type. Figure 2. Dependence 1-butyl-3-methylimidazolium cation reten- tion factor values as a function of different stationary phases. Con- ditions: 95% v/v 40 mM KH2PO4 (pH = 4) and 5% v/v MeOH, ex- cept of SG-Ar and SG-Ph, where conditions are: 50% v/v 40mM KH2PO4 (pH = 4) and 50% v/v MeOH. Figure 3. Dependence of ionic-liquid cations retention as a func- tion of different packing materials. Eluents: 40 mM KH2PO4 (pH = 4) and 5% v/v MeOH, or 50% v/v MeOH. 23Acta Chim. Slov. 2007, 54, 20–24 Studzi×ska et al.: Chromatographic and Chemometric Methods for Evaluation ... We compared the experimentally determined values of logkW obtained for ionic liquid cations on the packing materials with the standard hydrophobicity measure logP. The logP and logkW,IAM values were determined according to Stepnowski et al.,6 who have reported the experimen- tally (with the use of HPLC) measured and theoretically estimated lipophilicity coefficients obtained for represen- tatives of imidazolium ionic liquid cations. Results of the- se studies are presented in Table 3 and Figure 4. Figure 4. Correlation between the theoretically estimated logP va- lues and experimentally determined logkw for IAM and SG- CHOL stationary phases. 6 5 4 3 2 1 0 Figure 5. Correlation between log αW and log αS with logP values for the studied imidazolium-based-ionic liquids. Experimentally measured ionic liquid cations logkW vaule Stationary phase PMIM BMIM AMIM HMIM Imobilozed IAM 0.92 1.32 1.52 1.70 artificial membrane Cholesterolic 1.2 2.63 3.63 5.39 SG-CHOL Theoretically estimated logP values –1.74 –1.44 –1.09 –0.71 Table 3. Comparison of the obtained measure of lipophobicity. An interesting dependence in the case of cholestero- lic stationary phase was found. The correlation coefficient is higher than in the case of IAM packing, which mimics biological membranes and should give better results in this investigation. Such an effect is probably a consequen- ce of possible combination of hydrophobic, donor-accep- tor, π…π, hydrogen bonding interactions on SG-CHOL stationary phase. All of the mentioned interactions are ex- pected to be important in membrane transport, which is probably the reason for such a good correlation between the retention of ionic liquids and hydrophobicity measure in case of cholesterolic packing. Furthermore, the high correlation values indicate that logP for ionic liquid corre- lates well with their partitioning into a given stationary phase. Another way of estimating the logP values is the use of chemometric approach. The HyperChem software al- lows for estimation of a variety of molecular descriptors commonly used in Quantitative Structure Activity Rela- tionships (QSAR) studies, e.g. van der Waals surface area, solvent accessible area, which can be used to draw conc- lusions regarding the lipophilicity of analysed molecules. The imidazolium-based ionic liquid cations used in this study are built on the same skeleton containing an imida- zolium ring with one methyl group bonded to both nitro- gen atoms. The main QSAR properties were determined for all compounds. Then, the šselectivity’ parameter was determined, as follows: αS = ionic liquid solvent accessible area / skeleton solvent accessible area, αW = ionic liquid van der Waals surface area / skele- ton van der Waals surface area. The obtained values were compared with logP va- lues and are presented in Figure 5. As it can be seen, the QSAR parameters correlate well with the logP value. In the case of ionic liquid cation the structure differs only in the number and position of methyl groups and the main interactions in which they can take part are van der Waals (dispersion) force, similar as in the case of interactions ta- king part when logP determined. These investigations show that with the use of chemometric approach it is pos- sible to predict hydrophobicity of ionic liquids cations. 4. Conclusions The studies of ionic liquids interactions are still scarce and in this contribution, liquid chromatography provided interesting results. HPLC allowed for the des- criptions of the main interactions, which were found to be π…π type and van der Waals dispersive forces. The use of different chemometric approaches allows for furt- her conclusions regarding the interactions. The present work proved the usefulness of HPLC and chemometric methods in the preliminary studies of ionic liquids pro- perties. 24 Acta Chim. Slov. 2007, 54, 20–24 Studzi×ska et al.: Chromatographic and Chemometric Methods for Evaluation ... 5. Acknowledgements Financial support provided by the Polish Ministry of Science and Higher Education (Warsaw, Poland) under grant 2P04G08329 is acknowledged. The authors grate- fully acknowledge S. Witko (Ĺódź, Poland) for HPLC sol- vents and silica support (SG) donation. 6. References 1. H. Olivier-Bourbigou, L. Magna, J. Mol. Cat. A: Chemical 2002, 3484, 1–19. 2. J. D. Holbrey, K. R. Seddon, Clean Prod. Proc. 1999, 1, 223–236. 3. K. N. Marsh, J. A. Boxall, R. Lichtenthaler, Fluid P. Equil. 2004, 219, 93–98. 4. M. J. Earle, J. R. Seddon, Pure Appl. Chem. 2000, 72, 1391–1398. 5. P. Stepnowski, Aust. J. Chem. 2005, 58, 170–173. 6. P. Stepnowski, P. Storoniak, Environ. Sci. Poll. Res. 2005, 12(4), 199–204. 7. A. Lata ĺa, P. Stepnowski, M. N¥dzi, W. Mrozik, Aquatic To- xicol. 2005, 73, 91–98. 8. R. Gadza ĺa-Kopciuch, B. Buszewski, J. Sep. Sci. 2003, 26, 1273–1283. 9. B. Buszewski, M. Jezierska-Świta ĺa, R. Kaliszan, A. Wojtc- zak, K. Albert, S. Bochmann, M. T. Matyska, J. J. Pesek, Chromatographia 2001, 53, 204–212. 10. B. Buszewski, R. M. Gadza ĺ a-Kopciuch, R. Kaliszan, M. Markuszewski, M. T. Matyska, J. J. Pesek, Chromatographia 1998, 48, 615–622. 11. R. Gadza ĺa-Kopciuch, M. Kluska, M. Weĺniak, B. Buszew- ski, Mat. Chem. Phys. 2005, 89, 228–237. Povzetek Zanimanje za ionske teko~ine nara{~a zaradi {iroke uporabnosti v industrijskih procesih, {e posebej v organski sintezi, kar je vodilo do njihove prisotnosti v okolju, zlasti v vodi in prsti. Zato so se pokazale potrebe, da bolje razumemo nji- hove toksi~ne lastnosti. Glavni cilj tega prispevka je, da poda pregled obstoje~e literature o vplivu ionskih teko~in na `ive organizme in da oce- ni uporabnost kromatografskih in kemometrijskih metod za oceno lastnosti ionskih teko~in.