Acta Chromatographica, No. 7, 1997 PRINCIPLES AND APPLICATIONS OF ION-EXCLUSION CHROMATOGRAPHY B. K. G³ód Polish Academy of Sciences, Institute - Experimental and Clinical Medical Research Centre, Department of Cellular Signalling, ul. Dworkowa 3, 00-784 Warsaw, Poland SUMMARY Ion-exclusion chromatography (IEC) finds application in the separation of a wide range of small, neutral or partially ionized molecules. In IEC strong and weak electrolytes are eluted unseparated, the former at the beginning of elution and the latter at the end. The retention volumes of the remaining electrolytes were found to be proportional to their dissociation constants. The mechanism of retention of the compounds analysed in ionexclusion chromatography has been described by use of analytical equations and results obtained from computer simulation of column performance (using global thermodynamic and chromatographic equations or the Craig method). The purpose of this paper is to survey the field. It focuses on description of the retention mechanism and on the most important applications. INTRODUCTION Ion-exclusion chromatography (IEC) is a technique widely used to separate ionic compounds from non-ionic compounds and to separate mixtures of acids (or bases), The characteristic feature of IEC is that the sign of the electric charge of the dissociated functional groups on the ionexchange resin is the same as that on the ionic compound analysed. It follows that negatively charged ions, e.g. dissociated acidic compounds, are separated on cation-exchange resins with anionic (usually sulphonic) functional groups. By analogy, positively charged species (bases) are separated on anion-exchange resins containing cationic (usually tetraalkylammonium) functional groups. Although ion-exchange resins are used for this technique, true ion-exchange reactions are not involved, - 72 - although the same columns can be used for both IEC and ion-exchange chromatography. For the specific requirements of ion-exclusion chromatography large ion-exchange capacity is preferential. Column capacity is increased by increasing its dimensions, maximizing the concentration of functional groups on the support, and using a strong ionexchanger (anion- or cation-exchanger). The usual supports are based on macro-porous copolymers of styrene and divinylbenzene in which the degree of cross-linking is characterized by the concentration of divinylbenzene in the reaction mixture. RETENTION MECHANISM When the column is filled with water, which is pumped through as mobile phase, the water molecules build up hydration spheres around the dissociated functional groups of the support. Water contained in the pores of the support and in the hydration spheres is immobilized thus forming the stationary phase and it is possible to regard the resin, its functional group ions and hydrogen ions as being dissolved in that stationary phase. Three parts of the column can, therefore, be considered: the solid resin network; the stationary phase; and the liquid which forms the mobile phase moving between the resin particles. The basis of the mechanism of retention in ion exclusion, recently described in review papers [1-3], is that neutral, uncharged molecules can penetrate the resin, whereas similarly charged co-ions are repelled owing to the presence of dissociated functional groups immobilized in the stationary phase. By analogy with the Donnan membrane equilibrium (Fig. 1) the hydrated resin network behaves like a semi-permeable membrane between the stationary and mobile phases. Except for the covalently bonded functional groups all other species are freely exchanged through such a hypothetical membrane. Because the concentration of the ions in the stationary phase exceeds that in the mobile phase, osmotic forces tend to drive water into the resin causing it to swell. This swelling is less for more highly cross-linked stationary phases and for mobile phases containing high concentrations of ions. The ratio of the concentrations of ionized to neutral forms of an analysed compound is determined by its dissociation constant and is equivalent to the solute effective charge. Solute retention therefore depends on this constant. Strong acids that are completely dissociated are - 73 - electrostatically repulsed. As a consequence they are eluted unseparated in the column dead volume (VM), which corresponds to the volume of the mobile phase in the column. On the other hand undissociated molecules are able to enter the resin network. They are eluted together with a retention volume equal to the sum of the inner and dead column volumes. The inner column volume means just the volume of its stationary phase. This behaviour makes the determination of the inner and dead column volumes straightforward [4,5]. Only electrolytes of intermediate strength (whose dissociation constants fall in the range 10-7-10-2) can be separated by this technique (Fig. 2). For these electrolytes higher retention volumes are expected for species with higher pKa or pKb values, as was confirmed by Tanaka et al. [4]. The relationship between the distribution coefficient and the pKa value for the analysed acids is analogous to that between the distribution coefficient and the logarithm of molecular weight in sizeexclusion chromatography, The mechanism of retention in IEC has been described in a number of models [3]. In the most general case, assuming that solute retention is influenced by changes of activity coefficient in the stationary phase, changes of stationary phase dielectric constant and buffer and the dissociation constants of stationary phase functional groups, one can derive [6]: 1+ Kd = 1+ 2 K aS γ f K 2f + 4 K f a f γ HS + − K f 2 K aM ⋅Kp (1) K b2 + 4 K b cb − K b where Kd denotes the distribution coefficient, KaS and KaM are the dissociation constants in the stationary and mobile phases, respectively, a is the functional group activity, γ HS the activity coefficient of hydrogen ions in the stationary phase, Kp the neutral molecule partition coefficient, Kf the functional group dissociation constant, cb the buffer concentration and Kb its dissociation constant. If the buffer and the resin functional groups are strong acids and their concentrations are much higher than that of the solute it can be shown that: - 74 - 1+ Kd = K aS a f γ R− S M a K 1+ cb KP (2) It follows from this equation that for buffered mobile phases the retention is independent of solute concentration. In the most simple case, using the crude assumption that pure water is the mobile phase, that the inner column volume equals the dead volume, and that dissociation of the support functional groups is complete, the following simple relationship can be derived for the dependence of solute distribution coefficient on its concentration and its dissociation constant [5]: Kd = 4c max + K a − K a2 + 8 K a c max 4c max − K a + K a2 + 8 K a c max = 2c max + K a − K a2 + 8 K a c max 2c max − 2 K a (3) where cmax is the solute concentration at the peak maximum. From this equation it is apparent that the solute distribution coefficient is a function of only one experimental quantity - the ratio of its concentration to its dissociation constant. Increasing the sample concentration and reducing the sample dissociation constant increase the distribution coefficient. A similar simple relationship can be derived for buffered mobile phases if it is assumed that: the buffer concentration is much higher than that of the analysed solute; a strong acid or base is used as a buffer; the concentration of the resin functional group is much higher than that of the solute; and the resin functional groups are completely dissociated. The distribution coefficient, Kd, is the described by [7]: K a K HVA + c f VS + 1 Kd = Ka KaK2 + c b cb2 + 1 (3) - 75 - where cf denotes functional group concentration, K2 the second dissociation constant, KH the linear Henry adsorption isotherm constant and VA the adsorption layer volume. It should be remarked that ion exclusion can seldom be considered as the sole retention mechanism even on an ion-exclusion resin. Also relevant are hydrophobic adsorption on the resin network (as in reversedphase chromatography), size exclusion, effect of functional group screening in the analysed sample, normal phase retention, and van der Waals and polar interactions of the sample compound with the support (Fig. 2) [3,7]. APPLICATIONS IEC finds an increasing range of applications in the determination of weak and moderately strong acids, both organic and inorganic, amines and even sugars. Its major advantage lies in its ready applicability to samples with very complex compositions. Typical applications are the analysis of sea and mineral waters, biological fluids, foodstuffs, wine and other drinks, detergents, etc. (Table I). IEC is used most frequently for the analysis of carboxylic acids. Aromatic acids can be detected by UV-detection in the range 250-280 nm. Aliphatic and aromatic acids can be detected by UV detection at 200-220 nm. The detection limit obtained is below the ppm level [11,50]. Determination of benzoic acid in mustard, in which it is used as a preservative [5], is shown in Fig. 3. All inorganic compounds were eluted in the column dead volume whereas undissociated compounds were not detected by the electrokinetic detector. The determination of aliphatic fatty acids in wine [8] is shown in Fig. 4. It is worth emphasizing that except for filtration to remove insoluble impurities no special sample pretreatment was involved in the analysis of either sample. Despite this modest pretreatment the samples have passed through the column without damaging it and do not influence the long-term effectiveness of the column. It is an essential advantage of IEC that this technique is applicable to samples of various and complicated composition, even though it is conventional to regard the technique as having the disadvantage of limited retention. This is especially important in biological samples with complex matrices. Analysis of acids in urine [50,56] or pharmaceutical solutions [36] is performed simply by filtration and eventually dilution. In blood or plasma samples proteins should be - 76 - precipitated by acidification or removed by addition of hydroxide and acetonitrile (ACN) and then centrifugation [52]. For some compounds the observed retention is greater than that implied by the ion-exclusion mechanism [3,6]; this is especially true for aromatic compounds owing to the strong π-electron interaction with the aromatic rings of the resin network. This can be of practical use in the separation of aliphatic compounds from aromatic on an ion exclusion column (Fig. 5) [47]. Hydrophobic adsorption can be controlled (reduced) by addition of an organic modifier to the mobile phase. It has recently [47] been shown that addition of cyclodextrin to the mobile phase can also reduce retention. Such changes of retention are very selective, so that even enantiomers of methylphenobarbital are effectively separated. The second group of compounds frequently analysed by IEC are weak inorganic acids and their salts (Table 1). Separation of borate involves the use of a polyalcohol (mannitol [57] and D-sorbitol [58]) in the mobile phase. They form polyborate complexes which are more highly ionized and better detected by conductiometric detection. Very important environmentally is the analysis of NO2- on a column filled with mordenite [47]. Alcohols are separated on an ion-exclusion column by a hydrophobic adsorption retention mechanism [25]. They have been monitored using conductiometric or refractive index detectors. Dimethylsulphoxide (DMSO) has been determined in sea water by use of a UV detector [15]. Sugars can be separated on ion exclusion columns but their retention is affected by both size exclusion and hydrophobic adsorption. Oligomers elute in order of decreasing molecular weight. They separate better on a resin in the Ca2+ form [25]. Usually the pulsed amperometric or refractive index detectors are used. It is also possible to separate basic compounds, such as amines [59] and alkanolamines [42], on a column with tetraalkylammonium functional groups. Corradini et al. [60] separated proteins on a silica-based column and found that at very low salt concentrations proteins were separated by an electrostatic retention mechanism (ion exchange or ion exclusion); separation was by size exclusion at higher salt concentrations and by hydrophobic adsorption at very high salt concentrations. Alcohols can be determined by IEC with water as mobile phase. Stevens et al. [17] reversed this situation and analysed water using alcohol as the mobile phase. Sulphuric acid was added to the mobile phase and indirect conductiometric detection was used. - 77 - Fig. 1 Schematic illustration of ion exclusion. - 78 - Fig. 2 Effect of acid pKa on its retention volume, VR [47], on a 300 × 7.8 mm i.d. × 9 µm particle diameter Bio-Rad Aminex HPX-87H ion-exclusion organic acid analysis column (8% cross-linked cation exchanger in the hydrogen form). The mobile phase was 0.5 mM H2SO4. Solute acids: „, separated by pure ion-exclusion (perchloric, nitric, sulphosalicylic, chloroacetic, formic and acetic acids and methanol); ∆, affected by the screening effect (oxalic, maleic, tartaric, fumaric, citric, lactic and ascorbic acids); š, long chain aliphatic acids (propionic, butyric and valeric); and {, aromatic acids (phthalic, isophthalic, terephthalic, o-, mand p-nitrobenzoic, benzoic, salicylic, m-nitrophenol, o-chlorophenol, o-, m- and p-toluic, anisic, p-chlorobenzoic and tannic). - 79 - Fig. 3 Separation of benzoic acid (2) from other ionic components (1) in mustard [5]. Electrokinetic detection using streaming current (SCD) and streaming potential (SPD) detectors (insensitive to non-ionic compounds). - 80 - Fig. 4 Chromatogram obtained from a white wine [8]. The acids identified were: 1, tartaric; 2, malic; 3, lactic; 4, succinic; 5, unknown. Detection: potentiometric using a copper wire and UV at 210 nm. - 81 - Fig. 5 Ion exclusion chromatogram obtained from Fanta [47]. The acids identified were: 1, citric; 2, ascorbic; 3, benzoic. Detection by UV at 210 nm. Chromatographic conditions as for Fig. 2. - 82 - Table I Examples of practical application of ion-exclusion chromatography Sample Solutes Eluenta Column Detectionb Ref. Mustard Benzoic acid LiChrosorb KAT Water Electrokin. 5 Wine Aliphatic acids Aminex HPX-87H 1 mM CSA LTV, P-Cu 8 Acid rain Carboxylic acids Dionex HPICE-AS2 2 mM HCl C Antarctic ice Carboxylic acids Aminex HPX-87H 5 mM MSA UV-200 10 Urine Carboxylic acids Aminex HPX-87H 25 mM H2SO4 UV-200 11 Blood Carboxylic acids Dionex HPICE-AS1 10 mM HCl C 12 Milk Fatty acids Aminex HPX-87H 10 mM H2SO4 UV-210 13 Mineral water As(V), As(III) Aminex HPX-87H 10 mM H3PO4 A 14 Sea water DMSO Aminex HPX-87H 5 mM H3PO4 UV 15 Detergents Phosphates Hitachi 2613 Dioxane - water K 16 Organic solvents Water Aminex 50W-X4 MeOH, H2O, HCl C 17 Sea water Inorganic anions Dionex B-1 2.4 mM CO32- C 18 Air Aliphatic acids Wescan ion exclusion 0.25 mM H2SO4 C 19 20 9 Sewage waters Acetic acid, carbonate YEW SCX-252 2 mM H2SO4 C Latex Aliphatic acids Dionex AS-1 1 mM HCl C 21 Caprolactam Acetic, propionic acids Dionex ion exclusion 0.1 mM HCl C 22 Polyurethanes Aliphatic acids Dionex ion exclusion 0.1 mM HCl C 23 Frozen food Sulphites Wescan ion exclusion 5 mM H2SO4 A (Pt) 24 Beer, sake Aliphatic acids TSK SCX 1 mM H2SO4 C 25 Beer Glycerol, ethanol Dionex AS-1 0.1 mM HClO4 A(pulsat.) 26 Ham Nitrite Polypore H 20 mM mM H2SO4 A(Pt) 27 Milk Lactose, glucose TSK-gel SCX 0.1 mM H3BO3 C, UV-200 28 Milk Iodides Dionex AS-1 0.01 mM NaNO3 A(Ag) 29 Coffee Phosphates, fatty acids Dionex ICE 10 mM HCl C 30 Wescan 269-051 5 mM H2SO4 A 31 Potatoes, paprika Sulphites Maple syrup Aliphatic acids Dionex ion exclusion 0.05 mM HCl C 32 Blood, serum Org., inorg. acids, salts Dionex ICE Aqueous buffers C 33 Liver, lung Esters Dionex AS-1 1 mM HCl C 34 Pancreas juice Acids Dionex AS-1 Water C 35 Medicines Acids, alcohols, sugars Dionex AS-1 25 mM H2SO4 RI 36 Toothpaste Fluorides Wescan ion exclusion 3 mM H2SO4 C 37 Paracetamol Acetate Dionex ion exclusion 0.1 mM HCl C 38 Plating bath Citric, lactic acids Home made 1 mM HCl C 39 Tomatoes Fatty acids AS5 FBA, pH 2.8 C 40 Tech. sugar Fatty acids HPICE-AS1 Acidic solutions C 41 Cosmetics Alkanolamines PS/DVB, OH- form 2% Glycerine UV 200 42 Forensic Carboxylic acids Waters Ion Exclusion 1 mM OSA C 43 Tap water Silicate Ion exclusion dil. HClO4 CL 44 - 83 - Silage effluents IonPac - AS5 0.9 mM PFBA Plasma and urine TCEOH Aliphatic acids Aminex A-5, K+ form 10 mM K2S04/KOH RI 46 Fanta Aminex HPX-87H 1 mM CSA 47 Carboxylic acids Csuppr. UV-210 45 Soil Borate Wescan ion exclusion 0.3 M D-sorbitol C 48 Air Nitrogen dioxide Wescan HS 5 mM H2SO4 A (Pt) 49 10 mM H2SO4 Urine Carboxylic acids Interaction ORH-801 Blood, serum Fluoride Wescan anion exclusion 2 mM H2SO4 UV-254 50 C 51 Plasma Valproic acid Dionex ICE 0.5 mM H2CO3 C 52 Urine Uric, oxalic acids Vydac SCX Buffer pH 2.8 A 53 Brain tissue N-acetylaspartic acid Interaction ION 300 0.5 mM H2SO4 UV-215 54 Murine tumours Lactate IonPac ICE AS 1 0.5 mM OSA C 55 a b CSA, camphorsulphonic acid; MSA, methanesulphonic acid; OSA, octanesulphonic acid; FBA, fluorobutyric acid; MeOH, methanol; PFBA, perfluorobutyric acid; TCEOH, 2,2,2-trichloroethanol. P, potentiometric; C, conductiometric; A, amperometric; K, coulometric; RI, refractive index; CL, chemiluminescence. - 84 - REFERENCES [1] R.E. Smith, ‘Ion Chromatography Applications’, CRC Press, Boca Raton, 1988 [2] B.K. G³ód, ‘Ion-Exclusion Chromatography’, Wyd. UMCS, Lublin, in press [3] B.K. G³ód, Chem. Anal. (Warsaw), 39, 1 (1994) [4] K. Tanaka, T. Ishizuka and H. Sunahara, J. Chromatogr., 174, 153 (1979) [5] B.K. G³ód and W. Kemula, J. Chromatogr., 366, 39 (1986) [6] B.K. G³ód, Acta Chromatogr., 6, 101 (1996) [7] B.K. G³ód and J. Stafiej, J. Chromatogr., 654, 197 (1993) [8] B.K. G³ód, P.R. Haddad and P.W. Alexander, J. Chromatogr., 589, 209 (1992) [9] S.R. Bachman and M.E. Peden, Water Air Soil Pollut., 33, 129 (1987) [10] P.R. Haddad and P.E. Jackson, J. Chromatogr., 447, 155 (1988) [11] D.N. Buchanan and J.G. Thoene, Anal. Biochem., 124, 108 (1982) [12] W.E. Rich and E.L. Johnson, Eur. Pat. Appl., EP. 38720 (1981) [13] T. Jupille, M. Gray, B. Black and M. Gould, Am. Lab., 13, 80 (1981) [14] E.C.V. Butler, J. Chromatogr., 450, 353 (1988) [15] J.P. Ivey and P.R. Haddad, J. Chromatogr., 391, 309 (1987) [16] K. Tanaka and T. Ishizuka, J. Chromatogr., 190, 77 (1980) [17] T.S. Stevens, K.M. Chritz and H. Small, Anal. Chem., 59, 1716 (1987) [18] W. Rich, F. Smith, L. McNeil and T. Sidebottom, in E. Sawicki and J.D. Mulik (eds), ‘Ion Chromatographic Analysis of Environmental Pollutants’, vol. 11, Ann Arbor Science, Ann Arbor, NE, 1979, p. 17 [19] C. Rosenberg, W. Winiwarter, M. Gregori, G. Pech, V. Casensky and H. Puxbaum, Fresenius’ Z. Anal. Chem., 331, 1 (1988) [20] T. Murayama, T. Kubota, Y. Hanaoka, S. Rokushika, K. Kihara and H. Hatano, J. Chromatogr., 435, 417 (1988) [21] R.C. Crafts, Polymer Testing, 5, 193 (1985) [22] H. Muller, W. Nielinger and A. Horbach, Angew. Makromol. Chem., 108, 1 (1982) [23] J. Von Unterrichter-Worthman and F. Quella, Kunstoffe, 74, 682 (1984) [24] H.-J. Kim and Y.-K. Kim, J. Food Sci., 51, 1360 (1986) [25] K. Tanaka and J.S. Fritz, J. Chromatogr., 409, 271 (1987) - 85 - [26] R.D. Rocklin, LC, 1, 504 (1983) [27] H.-J. Kim and Y.-K. Kim, in P. Jandik and R.M. Cassidy (eds), ‘Advances in Ion Chromatography’, vol. 1, Century International, Franklin, MA, 1989, p. 391 [28] J.F. Lawrence, Chromatographia, 24, 45 (1987) [29] Dionex Application Note 38 [30] G.M. Bauer, Lebensm.-Biotechnol., 1, 18 (1985) [31] J.H. Nguyen, H.-J. Kim and D.T. Gjerde, Am. Lab., May 1988, 12 [32] B.J. Shevenell and W.C. Shortle, Phytopathology, 76, 132 (1986) [33] W.E. Rich, E. Johnson, L. Lois, B.E. Stafford, P.M. Kabra and L.J. Marton, in L.J. Kabra and L.J. Marton (eds), ‘Liquid Chromatography in Clinical Analysis’, Humana, Clifton, NJ, 1981, p. 393 [34] A.R. Dahl, S.C. Miller and J. Petridou-Fischer, Toxicol. Lett., 36, 129 (1987) [35] J.R. Kreling and J. DeZwaan, Anal. Chem., 58, 3028 (1986) [36] G. Iwinski and D.R. Jenke, J. Chromatogr., 392 397 (1987), [37] J. Behnert, P. Behrend and A. Kipplinger, Labor Praxis, 9, 38 (1985) [38] B. Kreilgard and F.M. Anderson, Arch. Pharm. Chem., Sci. Ed., 12, 85 (1984) [39] C. Pohland, S. Afr. J. Sci., 80, 208 (1984) [40] B. Daubert, M. Guerere and J. Estienne, Ann. Falsif. Expert. Chim. Toxicol., 83, 401 (1990) [41] C. A. Accorsi and G. Blo, J. Chromatogr., 555, 65 (1991) [42] M. Fukui, H. Koniski, K. Ohta and K. Tanaka, Bunseki Kagaku, 41, T27 (1992) [43] L.A. Kaine, J.B. Crowe and K.A. Wolnik, J. Chromatogr., 602, 141 (1992) [44] H. Sakai, T. Fujiwara and T. Kumamaru, Bull. Chem. Soc. Jpn., 66, 3401 (1993) [45] K. Fisher, C. Corsten, P. Leidman, D. Bieniek and A. Kettrup, Chromatographia, 38, 43 (1994) [46] H. Itoh, S. Ikeda and N. Ichinose, Analyst, 119, 409 (1994) [47] B.K. G³ód and G. Perez, in preparation [48] H.C. Mehra, K.D. Huysmans and W.T. Frankenberger, J. Chromatogr., 508, 265 (1990) [49] H.-J. Kim and Y.K. Kim, Anal. Chem., 61, 1485 (1989) [50] D.J. Woo and J. R. Benson, Am. Clin. Prod. Rev., January 1984, 20 [51] J. Behnert, P. Behrend and A. Kipplinger, Labor Praxis, 9, 39 (1985) - 86 - [52] H. Itoh, Y. Shinbori and N. Tamura, Bull. Chem. Soc. Jpn., 59, 997 (1986) [53] W.J. Mayer, J.P. Mcarthy and M.S. Greenberg, J. Chromatogr. Sci., 17, 656 (1979) [54] D. S. Dunlop, D.M. McHale and A. Lajtha, Brain Res., 580, 44 (1992) [55] M.R. Stratford, C.S. Parkins, S.A. Everett, M.P. Dennis, M. Stubbs and S.A. Hill, J. Chromatogr. A, 706, 459 (1995) [56] V.T. Turkelson and M. Richards, Anal. Chem., 50, 1420 (1978) [57] T. Okada and T. Kuwamoto, Anal. Chem., 58, 1375 (1986) [58] H.C. Mehra, K.D. Huysmans and W.T. Frankenberg, J. Chromatogr., 508, 265 (1990) [59] P.R. Haddad, F. Hao and B.K. G³ód, J. Chromatogr. A, 671, 3 (1994) [60] D. Corradini, R. Filippetti and C. Corradini, J. Liquid Chromatogr., 16, 3393 (1993) - 87 - - 88 - View publication stats