Communication Anion-modulated switching of retention properties of a zwitterionic stationary phase Jayakumar M. Patil† and Tetsuo Okada* Department of Chemistry, Tokyo Institute of Technology, Meguro-ku, Tokyo 152, Japan. E-mail: tokada@chem.titech.ac.jp Received 2nd November 1998, Accepted 1st December 1998 Stationary phases modified by zwitterionic molecules act both as an anion- and a cation-exchanger. This characteristic property of zwitterionic stationary phases can be modulated only by changing the nature of anions added in mobile phases. Anion-exchange properties emerge when small and well hydrated anions are added in mobile phases, while cation-exchange properties appear when mobile phases contain large and poorly hydrated anions.The nature of cations is much less important than that of anions. These unusual retention properties come from the structure of a zwitterionic molecule used as a stationary phase modifier, which has an inner cationic and an outer anionic group.The surface adsorption of functional molecules has largely enhanced the versatility of reversed-phase chromatography, and allowed the separation of various compounds that are not retained on such hydrophobic stationary phases.1–11 Amphoteric compounds, such as surfactants, have been most extensively used for modifying chromatographic stationary phase surfaces, and are often called ion-pair or ion-interaction reagents.1–9 Ion-pair reagents, for example, make the surfaces of hydrophobic stationary phases ionic.The resulting stationary phases can be used as ion-exchange resins; this mode has significantly contributed to the development of chromatography of ions.12 Unstable retention and poor reproducibility are disadvantages of these so-called ‘dynamically-coated’ stationary phases. However, these have important advantages over chemically bonded ion-exchange resins. The surface densities and the structures of active molecules are two principal factors controlling the retention in this chromatographic mode; these factors can be easily changed in dynamically-coated stationary phases.The former mainly affect the retention ability of the stationary phases, while the latter modify separation selectivity. Recently, Hu, Haraguchi and their coworkers13–19 developed a method named electrostatic chromatography, in which the stationary phases are coated with zwitterionic surfactant and are thus capable of simultaneous retention of cationic and anionic species.The main purpose of their work was the development of a method permitting the use of pure water as a mobile phase for the separation of ions because of the advantage of this highly resistive mobile phase in the conductivity detection of ions. Since the properties of zwitterionic surfaces (as well as those of zwitterionic surfactant micelles) are affected by mobile phase compositions, such as the concentration and the nature of an added salt, detailed studies on salt effects are important for further methodological developments of electrostatic chromatography and for the understanding of its separation mechanism.In the present paper, we demonstrate that the ion uptake properties of stationary phases coated with 3-(N-dodecyl-N,Ndimethyl- ammonio)-propane-1-sulfonate (DDAPS) can be modulated by the nature of the anions added to the mobile phase, i.e., they act as a cation-exchanger under certain conditions, but as an anion-exchanger under different conditions. The characteristic properties of DDAPS-coated stationary phases can be explained by the calculation of electrostatic potential.Experimental section The chromatographic system used was composed of a Tosoh (Tokyo, Japan) computer-controlled pump Model CCPD, a Rheodyne injection valve equipped with a 100 mL sample loop, a JASCO (Tokyo, Japan) UV-Vis detector JASCO Model 875-UV, and a chart recorder. The separation column was immersed in water thermostatted at 25.0 °C.PAR [4-(2-pyridylazo) resorcinol] solution was delivered by another singleplunger pump (Nihonseimitsu Co.) for the postcolumn reaction of transition metal ions. The separation column was a 4.6 3150 mm stainless steel column packed with Wakosil 5C8 (Wako Pure Chemicals, Osaka, Japan) (particle size = 5 mm, specific surface area = 300 m2 g21, mean pore size = 12 nm). The critical micelle concentrations (c.m.c.) of DDAPS were determined by a dye solubilization method with Coomassie Brilliant Blue G-250 under various conditions.20 The adsorption amounts of DDAPS were determined by a breakthrough method; the elution of DDAPS was monitored with a Tosoh refractive index detector Model RI-8010. DDAPS was purchased from Tokyo Kasei (Tokyo, Japan), and recrystallized twice from acetone containing a small amount of ethanol.Other reagents were of analytical grade. Distilled deionized water was used for solution preparation. Results and discussion It is known that anions are partitioned into DDAPS micelles better than cations.This has been confirmed by self-diffusion measurements,21 fluorescence quenching,22 and chromatography. 16–17 This anion-dominated partitioning has been explained in our recent work on chromatographic modeling based on electrostatic theories.23 The polarity of the DDAPS molecule is the most important factor in determining anion-dominated partitioning; i.e., it has an inner cationic and an outer anionic group.According to the developed model, the electrostatic potential at the interface between solution and the DDAPSmodified surface is illustrated in Fig. 1, where 10 mM NaCl or NaClO4 mobile phases are assumed. The transfer free energies of ions are included in the developed model to represent the partition selectivity. The large crystalline ionic radius (thus low charge density) and large polarizability of ClO42 allow its enhanced invasion to the DDAPS layer and the developments of the negative electrostatic potential region as a result.In contrast, the low penetrable nature of Cl2 keeps the potential of the inside DDAPS layer positive, allowing the partition of anions. These calculations predict that DDAPS-modified stationary phase acts as a cation-exchanger in the former case, but as an anion-exchanger in the latter case. The electrostatic potential is † Permanent address: Department of Chemistry, Textile and Engineering Institute, Ichalkaranji-416115, India.Anal. Commun., 1999, 36, 9–11 9not, in contrast, affected by the nature of cations; in most cases, cation effects are negligible. This result agrees well with the experimental facts reported so far.16,17,21,22 The above calculation results were confirmed by chromatographic experiments. Small amounts of DDAPS were added to mobile phases to prevent the desorption of DDAPS from the equilibrated stationary phase surface during a series of measurements. The adsorption equilibrium is usually established for monomer surfactant molecules, suggesting that the c.m.c.of DDAPS is an important factor to optimize the DDAPS concentration in the mobile phases. Salt effects on the c.m.c. of DDAPS were therefore studied. The c.m.c. of DDAPS is 3.0 mM in water (the same as literature values),24 but that the addition of salts reduces the c.m.c. the value of which varies depending upon the nature and the concentration of the salts added (see Table 1).The preliminary experiments showed that 2 mM was high enough to prevent the desorption of DDAPS. Since micellar partition effects are also avoidable for 2 mM DDAPS mobile phases, this concentration was adopted for further chromatographic experiments. It is also predictable that the addition of salts affects the adsorption of DDAPS. However, changes in the adsorption amount were very small; the adsorption amount of DDAPS was 7.1 3 1024 mol per column from 5 mM aqueous solution (without added salts), 6.9 3 1024 mol from 0.1 M NaClO4 solution, and 6.8 3 1024 mol from 0.1 M NaCl solution.Increasing salt concentration affects adsorption in (at least) two ways; (1) enhancing the adsorption by salting-out and (2) lowering the adsorption due to a decrease in the monomer surfactant concentration (due to low c.m.c.). A constant adsorption amount must imply that these opposite effects happen to cancel each other.Thus, we can discuss chromatographic results on the basis of constant surface density of DDAPS molecules and no micellar partitioning. The chromatographic retention of I2 and Cu(ii) on the DDAPS-modified stationary phases was studied as a function of salt concentrations. Fig. 2 clearly indicates the anion-modulated switching of the ionic partition properties of the DDAPS layer. As predicted from the calculation of electrostatic potential (Fig. 1), an anion and a cation show the opposite retention dependence on the nature of anions; the largest retention is seen for Cu(ii) with ClO42 mobile phases, but for I2 with Cl2 mobile phases.It is also an important feature that the plots for anion retention show maxima at particular salt concentrations. This maximum formation was explained by the thickening of the DDAPS layer with increasing salt concentrations.23 The peak broadening was so marked with low concentration mobile phases that maximum appearance was not confirmed for Cu(ii).It can be explained in the following way that DDAPSmodified stationary phase distinguishes anions better than cations. The interaction energy between an ion and the dipolar layer of the DDAPS phase might be substituted by the free energy of transfer from water to a less polar medium of weaker solvation ability [DG°tr (W ?S)] if electrostatic interaction can be ignored. Although electrostatic interaction should be a major source of total interaction energy, it must be negligible if we discuss the selectivity (or relative interaction) of identically charged ions.Although we do not know DG°tr (W ? S) for S = DDAPS (or the dipolar layer of DDAPS), considering typical organic solvents instead of DDAPS is significant to infer the origin of the anion selectivity of the DDAPS-modified phase. DG°tr (W ? S) values of cations do not vary as much as those of anions; e.g., differences in DG°tr (W ? S) between Na+ and Cs+ are 1.3 kJ mol21 for S = methanol, 0.4 kJ mol21 for S = DMF, 28.8 kJ mol21 for S = acetonitrile, and 2.1 kJ mol21 for S = DMSO, while those between Cl2 and I2 are 25.9 kJ mol21, 227.3 kJ mol21, 223.5 kJ mol21, and 229.4 kJ mol21.25 This might be explained by the fact that anions are predominantly solvated through hydrogen bonds in water but mainly by dispersion energy in other solvents.Though we do not know the proper circumstances in the dipolar layer of the DDAPS phase, ions should be solvated in a different way from solvation in bulk water.Since the solvation of anions in the DDAPS layer must be weaker than that in water (similar to that in organic solvents), it can be reasonably understood that the Fig. 1 Profiles of electrostatic potential at the interface of solution and the DDAPS-modified surface. Table 1 The c.m.c. of DDAPS in various solutions Solution c.m.c./mM Water 3.0 0.01 M NaClO4 2.4 0.05 M NaClO4 2.0 0.1 M NaClO4 2.0 0.1 M NaCl 2.6 0.2 M NaCl 2.2 0.4 M NaCl 2.0 1.0 M NaCl 2.0 Fig. 2 Changes in the retention of (a) I2 and (b) Cu(ii) on the DDAPSmodified stationary phase with the salt concentration. 10 Anal. Commun., 1999, 36, 9–11DDAPS phase shows anion-selectivity rather than cationselectivity. The above results clearly suggest that the properties of the DDAPS-modified stationary phases can be adjusted to a particular separation in a very simple way, i.e., by changing electrolyte compositions. Fig. 3A shows the separation of some UV-absorbing anions with Cl2 eluents. It should be noted that the selectivity is identical with that in usual anion-exchange chromatography. In contrast, a less solvated anion should be added in mobile phases for the separation of cations. Fig. 3B is an example of separation of transition metal ions with the mixed mobile phase of tartaric acid and NaClO4; the former is a complexing agent, while the latter is necessary to make the stationary surface negative.Thus, switching the nature of the stationary phase in ionic separation is possible only by varying anions added in mobile phases. This is an important characteristics of zwitterionic stationary phases. Thus, although uses of zwitterionic micelles or zwitterionic surfaces in separation have not been common in developing analytical methods, we believe that this approach is of potential importance and versatile applicability. Acknowledgement This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Culture and Sports, Japan (Monbusho).J.M.P. thanks the UNESCO and the Monbusho for providing a fellowship. References 1 H. Miwa and M. Yamamoto, J. Chromatogr. A, 1996, 721, 261. 2 M. Piotte, F. Boss�anyi, F. Perreault and C. Jolicoeur, J. Chromatogr. A, 1995, 704, 377. 3 S. Fichtner, F. Th. Lange, W. Schmidt and H.-J. Brauch, Fresenius’ J. Anal. Chem., 1995, 353, 57. 4 J.-F. Jen and C.-S.Chen, Anal. Chim. Acta, 1992, 270, 55. 5 Y. Michigami, K. Fujii and K. Ueda, J. Chromatogr. A, 1994, 664, 117. 6 M. Adachi, K. Oguma and R. Kuroda, Chromatographia, 1990, 29, 579. 7 C. Sarzanini, M. C. Bruzzoniti, G. Sacchero and E. Mentasti, Anal. Chem., 1996, 68, 4494. 8 Y. Inoue, K. Kawabata and Y. Suzuki, J. Anal. At. Spectrom., 1995, 10, 363. 9 S. Zappoli and C. Bottura, Anal. Chem., 1994, 66, 3492. 10 J. D. Lamb and R. G. Smith, J. Chromatogr., 1993, 640, 33. 11 S.H. Hansen and J. Tjørnelund, J. Chromatogr., 1991, 556, 353. 12 M. C. Gennaro, Adv. Chromatogr., 1995, 35, 343. 13 W. Hu, T. Takeuchi and H. Haraguchi, Anal. Chim. Acta, 1992, 267, 141. 14 W. Hu and H. Haraguchi, Anal. Chim. Acta, 1994, 289, 231. 15 W. Hu and H. Haraguchi, Anal. Chim. Acta, 1994, 285, 335. 16 W. Hu, T. Takeuchi and H. Haraguchi, Anal. Chem., 1993, 65, 2204. 17 W. Hu, H. Tao and H. Haraguchi, Anal. Chem., 1994, 66, 2514. 18 W. Hu, A. Miyazaki, H. Tao, A. Itoh, T. Umemura and H. Haraguchi, Anal. Chem., 1995, 67, 3713. 19 W. Hu and H. Haraguchi, Bull. Chem. Soc. Jpn., 1993, 66, 1420. 20 C. Samsonoff, J. Daily, R. Almog and D. S. Berns, J. Colloid Interface Sci., 1986, 109, 325. 21 N. Kamenka, M. Chorro, Y. Chevalier, H. Levy and R. Zana, Langmuir, 1995, 11, 4243. 22 S. Brochsztain, P. B. Filho, V. G. Toscano, H. Chaimovich and M. J. Politi, J. Phys. Chem., 1990, 94, 6781. 23 T. Okada and J. M. Patil, Langmuir, 1998, 14, 6241. 24 J. P. Berry and S. G. Weber, J. Chromatogr. Sci., 1987, 25, 307. 25 I. Sakamoto and S. Okazaki, Yobai in Ion (Solvents and Ions), Taniguchi Insatsu, Matsue, 1990. Paper 8/08477G Fig. 3 Typical separation of selected anions (A) and transition metal ions (B). Mobile phase: (A) 2 mM DDAPPS + 10 mM NaCl; (B) 2 mM DDAPS + 20 mM tartaric acid + 0.1 M NaClO4 (pH 4.0). Detection: (A) UV at 220 nm; (B) PAR postcolumn reaction, detection at 540 nm. Anal. Commun., 1999, 36, 9–11