Journal of Chromatography A, 1326 (2014) 134–138

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage:

Short communication

Penicillin G as a novel chiral selector in capillary electrophoresis Shuchi Dixit ∗ , Jung Hag Park ∗ Department of Chemistry, Yeungnam University, Gyeongsan 712-749, South Korea

a r t i c l e

i n f o

Article history: Received 6 June 2013 Received in revised form 14 November 2013 Accepted 7 December 2013 Available online 15 December 2013 Keywords: Penicillin G Capillary electrophoresis Chiral selector Enantioseparation

a b s t r a c t The penicillin sub-class of ˇ-lactam antibiotics has not been examined for its enantiodiscriminating abilities in capillary electrophoresis (CE) until date. The present work was therefore designed to evaluate penicillin G potassium salt (PenG) as an ion-pair chiral selector (CS) using CE for its several attributes, namely, high solubility in water and lower alcohols, structure allowing multiple interactions with analytes and cost-effectiveness. Systematic experiments were performed to investigate the effect of composition of background electrolyte, applied voltage and capillary temperature on chiral separation. Baseline resolutions of enantiomers of five basic chiral drugs (namely, darifenacin, citalopram, sertraline, propranolol and metoprolol) were attained using a background electrolyte composed of water:methanol (90:10, v/v) and consisting of 10.7 or 16.1 mM CS at 20 ◦ C using an applied voltage of 5 kV. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Capillary electrophoresis (CE) has emerged as an efficient and versatile technique, in recent years, not only for development of enantioresolution methods [1–3] but also to evaluate enantiodiscrimination ability of novel chiral selectors (CSs) [4–7] due to its several advantages, such as, high efficiency, rapidity, simple instrumentation, solution-based chiral discrimination and cost-effective method development. Enantioselective ion-pair CE, where primary ionic interaction is accompanied by additional interactions such as hydrogen bonding, dipole–dipole, charge transfer (␲–␲), hydrophobic and steric interactions, for stereoselective formation of diastereomeric ionpairs between oppositely charged CS and analyte enantiomers, has been proven to be an important tool for chiral analysis in both aqueous and non-aqueous mode [8–15]. Literature reveals that cinchona alkaloids and derivatives, ketopinic and diisoproylideneketogluconic acids, tartrate-boric acid complexes and diaminocyclohexane have been successfully employed as ion-pair CSs [8–14]. Polypeptide, aminoglycoside, macrolide and lincosamide classes of antibiotics have been explored as CSs for their unique structural features and functionalities that allow multiple interactions with analytes having widely different structures [16–22].

Abbreviations: PenG, penicillin G; CS, chiral selector; DAR, darifenacin; CIT, citalopram hydrobromide; SERT, sertraline hydrochloride; METO, metoprolol; PRO, propranolol; MeOH, methanol; MeCN, acetonitrile. ∗ Corresponding authors. Tel.: +82 53 810 2360; fax: +82 53 810 4613. E-mail addresses: [email protected] (S. Dixit), [email protected] (J.H. Park). 0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved.

From ˇ-lactam class of antibiotics, only ceflacor (a member of cephalosporin sub-class) has been employed as CS to prepare chiral stationary phases in liquid chromatography [23,24]; none of the members of penicillin sub-class has been explored as CS in CE. Taking into account the literature mentioned above and the references cited therein, the present work was designed to evaluate penicillin G potassium salt (PenG) as an ion-pair CS using CE. Several attributes of PenG were considered for its implication as a potential ion-pair CS in aqueous mode, namely, (a) high solubility and low viscosity in water and alcohols, (b) prospect of having strong electrostatic interactions with basic analytes, in neutral or basic medium, for existing as an anion (pKa 2.3) in the said range, (c) possibility of additional interactions with analytes due to presence of three stereogenic centers, an aromatic moiety and several functional groups including a carboxylate group and two amide groups in its structure and, (d) easy commercial availability and cost effectiveness. 2. Experimental 2.1. Apparatus An Agilent HP 3D CE System (Palo Alto, USA) equipped with a diode-array UV detector, a high voltage (±30 kV) power supply and an external nitrogen pressure (up to 10 bars) was used to perform electrophoretic experiments. Other equipment used were Corning 135 pH meter (Corning, USA), Agilent Cary 5000 UVspectrophotometer and Elgastat UHQ water purification system (Bucks, UK).

S. Dixit, J.H. Park / J. Chromatogr. A 1326 (2014) 134–138

2.2. Chemicals and reagents The CS, PenG potassium salt (purity > 98%), was obtained from Sigma–Aldrich (Milwaukee, USA). The analytes, namely, darifenacin (DAR), citalopram hydrobromide (CIT), sertraline hydrochloride (SERT), metoprolol (METO), propranolol (PRO) and all other analytical-grade chemicals were obtained from Sigma–Aldrich (Milwaukee, USA) or TCI (Tokyo, Japan). HPLC-grade methanol (MeOH), acetonitrile (MeCN) and iso-propanol (i PrOH) were obtained from J.T. Baker (Phillipsburg, USA). Double distilled water (18.2 M mL) was used throughout. 2.3. Pretreatment of the capillaries Separations were performed in unmodified fused silica capillaries of 50 ␮m I.D. and 35 cm total length (25 cm effective length) (Polymicro Technologies, Phoenix, USA). Initial activation of the capillaries was done by rinsing them with 1 M HCl (20 min), 1 M NaOH (20 min), water (30 min) and background electrolyte (BGE; 30 min), respectively. The subsequent injections were preceded by a purge with 0.1 M NaOH (5 min), water (5 min) and BGE (15 min). At the end of each day, the capillaries were washed with 1 M HCl (10 min), 1 M NaOH (10 min) and water (20 min).


high background) [21]. The polarity of the recorder was reversed so that positive peaks were obtained. The electropherograms showing baseline resolutions of the enantiomers under optimized conditions have been shown in Fig. 2. Chiral discrimination ability of ion-pair CSs is attributed to stereoselective formation of diastereomeric ion-pairs between oppositely charged CS and analyte enantiomers. PenG possesses three stereogenic centers, an aromatic moiety and, a carboxylate and two amide functional groups. Its key structural feature is a four-membered ˇ-lactam ring fused with a five-membered thiazolidine ring. With one carboxylate group attached to thiazolidine ring (pKa = 2.3), it can exist as an anion in neutral or basic medium. Therefore, under the experimental conditions (apparent pH (pHapp ) = 6–7), the negatively charged carboxylate group of PenG can interact with the protonated amino group in basic drug analytes to form uncharged diastereomeric ion-pairs. The electrophoretic mobilities of free enantiomers are equal whereas the uncharged diastereomeric ion-pairs have no electrophoretic mobility. Thus, enantioseparation of basic drugs can be attributed to difference in (a) mobilities of the free and complexed forms of analyte and, (b) equilibrium constants of ion-pair formation for (R)and (S)-enantiomers of analyte [15,25].

2.4. Separation conditions

3.1. Composition of BGE

The running BGE containing CS, was freshly prepared by dissolving 10.7 mM PenG in the BGE, unless stated otherwise. Stock solutions of the analytes (1 mg/mL) were prepared in water:MeOH (90:10, v/v). All solutions were degassed by sonication under vacuum and then filtered through a 0.45 ␮m filter prior to use. Acetone was used as an electroosmotic flow (EOF) marker. Samples were injected using voltage of 3 kV for 5 s and the applied voltage was 5 kV. Capillary cassette was thermostated at 20 ◦ C, unless stated otherwise. Separations were monitored at 200, 214, 254, 265 and 280 nm. All the measurements were performed five times if not stated otherwise.

Run buffers consisting of various concentrations of CS and organic modifiers were evaluated to achieve successful enantioseparation of analytes. The eo and pHapp values, for these different run buffers, have been compiled in Tables 1 and 2 (supplementary data). Literature reveals that run buffers without addition of supporting or additional electrolytes have been successfully employed for separation of complex mixtures using CE [26]. Supporting electrolytes are added in the run buffer to provide a conductive medium that can produce currents in the microampere range. In this work, potassium salt of PenG in BGE played roles of both the chiral selector and the electrolyte. As stable and sufficient currents were generated throughout the runs to carry the separations in reasonable times, no additional electrolyte was added in run buffer. Various concentrations of PenG [i.e., 5.3 mM (0.2%, w/v, pHapp 6.60), 10.7 mM (0.4%, w/v, pHapp 6.41), 16.1 mM (0.6%, w/v, pHapp 6.36) and 21.5 mM (0.8%, w/v, pHapp 6.10)] were used to study the effect of concentration of CS on enantioseparation using water:MeOH (90:10, v/v). The effect of concentration of CS on eff and eo , for DAR and METO, has been shown in Fig. 3A. In addition, effect on effective mobility differences (eff ) and resolution (Rs ) values, for DAR and METO, has been shown in Fig. 3B. An increase in the concentration of CS from 5.3 to 10.7 mM resulted in increased resolution factors for all the analytes. The improved resolution can be attributed to a greater extent of interaction/complexation between the CS and the analytes. According to Bjørnsdottir et al. [15], the principle of ion-pair separation is analogous to the principle of chiral separation with cyclodextrins, described by Wren and Rowe [25]. That is, at an optimum concentration of CS, the maximum difference in mobilities of the two enantiomers (max ) is obtained and the maximum separation of the two enantiomers occurs. The best resolutions of DAR, CIT, SERT and PRO were attained at 10.7 mM concentration of CS and in case of METO the best resolution was observed at 16.1 mM concentration. The maximum resolution values at these optimized concentrations can be attributed to the maximum value of eff (Fig. 3B). A further increase from the optimized CS concentration caused lowering of eff values and hence a lowering of resolution factor. An increase in the concentration of CS caused an increment in the migration time of the analytes (Table 1, supplementary

2.5. Calculation of separations parameters The separation factors (˛eff ) were calculated according to ˛eff = eff1 /eff2 , where eff = app − eo = Ltot ·Leff /Vt − Ltot ·Leff /Vto (where eff , effective mobility; app , apparent mobility; eo , electroosmotic mobility; Ltot , total length of capillary; Leff , the effective length of capillary; V, applied voltage; t, migration time of the enantiomer and; to , migration time of the EOF marker). Resolution values (Rs ) were calculated according to Rs = 2(t2 − t1 )/(w1 + w2 ) (where t1 and t2 are the migration times, and w1 and w2 are the peak widths at the base of the first and second eluting enantiomers, respectively). 3. Results and discussion The present work describes enantioseparation of a set of five basic chiral drugs (namely, DAR, CIT, SERT, METO and PRO) in BGE composed of water/MeOH and containing PenG potassium salt as a CS. The structures of CS and analytes are shown in Fig. 1. Systematic experiments were performed to study the effect of various factors influencing enantioresolution, namely, composition of BGE, applied voltage and capillary temperature by taking DAR and METO as representative analytes. The separation data including migration times, effective mobilities, separation factors and resolution values of the resolved enantiomers have been compiled in Tables 1 and 2 (supplementary data). The analytes were monitored using indirect spectrophotometric detection in account of UV absorption of PenG (maxima at approximately 264 nm). This type of detection usually results in negative peaks (due to a diminished absorbance from a


S. Dixit, J.H. Park / J. Chromatogr. A 1326 (2014) 134–138

Fig. 1. Structures of the chiral selector and analytes.

data). A similar pattern has been reported in the literature and can be explained by the slowdown of EOF [22]. No enantioseparation was observed without addition of organic solvent in BGE. Literature reveals that addition of an organic modifier can alter or enhance enantioseparation by influencing viscosity of BGE, solubility and effective charge of analytes and CS, and interactions involved in chiral recognition mechanism [11,16–19]. Therefore, different ratios of water/MeOH [i.e., 95:5, 90:10 and 85:15, v/v] were investigated for two representative analytes, DAR and METO, to study the effect of concentration of organic modifier on enantioresolution. While an addition of 5% MeOH led to insignificant changes in resolution and separation selectivity; an increment up to 10% MeOH resulted in baseline resolution of all the analytes. An increase in the concentration of organic modifier

resulted in increased migration times of analytes due to decreased EOF. Table 2 indicates that exchange of MeOH with MeCN or i PrOH did not provide any improvement in enantioresolution for any of the analytes. While MeCN provided only partial separation, no enantioseparation was observed with i PrOH therefore it can be concluded that the replacement of MeOH with MeCN/i PrOH led to suppression of the interactions between the CS and the analytes. The longest migration times were observed with i PrOH followed by MeOH and MeCN, this trend can be attributed to the order of viscosity of solvents (i.e., i PrOH > MeOH > MeCN). The higher viscosity of constituting organic modifier causes higher viscosity of BGE and results in decreased EOF causing increased migration times of the analytes. In view of the highest resolution values,

S. Dixit, J.H. Park / J. Chromatogr. A 1326 (2014) 134–138


Fig. 2. Electropherograms showing baseline resolution of five basic chiral analytes using PenG as a CS. Conditions: capillary, 35 cm (25 cm effective length) × 50 ␮m I.D.; BGE, water:methanol (90:10, v/v) containing 10.7 mM (for CIT, DAR, PRO and SERT) and 16.1 mM (for METO) CS; applied voltage, 5 kV; injection, 3 kV, 6 s; capillary temperature, 20 ◦ C; detection, 214 nm.

reasonable migration times and excellent compatibility with UV detection, MeOH was found to be the best organic modifier. Buffer pH plays an important role in chiral discrimination by affecting extent of ion-pair interactions between oppositely charged CS and analytes [11]. The BGE having apparent pH value 6.41 and composed of 10.7 mM CS and water:methanol (90:10, v/v) provided the highest resolution values for enantioseparation of DAR, CIT, SERT and PRO whereas the BGE having apparent pH value 6.36 and composed of 16.1 mM CS and water:methanol (90:10,

v/v) provided the highest resolution values for enantioseparation of METO. The apparent pH values and separation data for various run buffers, compiled in Tables 1 and 2 (supplementary data), reveals that higher or lower pH values than the optimized pH resulted in lowering of resolution values. It indicates that the strongest ionpair interaction between negatively charged carboxylate group of PenG and protonated amino group in basic drug analytes occurred at optimized buffer compositions in the intermediate pH range (ca. 6.4).

Fig. 3. Effect of concentration of CS on eff and eo values (A) and, on eff and Rs values (B) for DAR and METO. Experimental conditions are the same as in Fig. 2.


S. Dixit, J.H. Park / J. Chromatogr. A 1326 (2014) 134–138

3.2. Applied voltage Applied voltage can affect EOF and Joule heating [5,20,21]. Effect of applied voltage was investigated over a range of 2–10 kV. An increment in applied voltage resulted in faster migration due to increased EOF. Resolution values increased upon increase of the voltage from 2 to 5 kV. This is likely due to decreased longitudinal diffusion of the analyte molecules as less time is available for the process with increasing EOF. Further increment in the voltage resulted in lowered resolution values for the analytes (Table 2, supplementary data). The lower resolution values can be explained by peak broadening caused by increased Joule heating at higher voltages. In account of satisfactory peak shapes, resolution values and migration times, 5 kV was considered to be the optimized applied voltage.

excellent technique not only to develop enantioseparation methods but also to evaluate chiral discriminating ability of novel CSs for its simplicity, rapidity and cost effectivity. Acknowledgement This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. 2012R1A1A2002635). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at 2013.12.018.

3.3. Temperature References Temperature can affect migration time and resolution of enantiomers by influencing the mobilities of analytes and CSs, and binding stability of two enantiomers to the CS [19]. Effect of temperature was investigated over a range of 15–25 ◦ C. On account of the highest resolution and reasonable migration time, 20 ◦ C was found to be the optimized temperature. In the present work with PenG as a CS, the repeatabilities of migration times of separated enantiomers were found to be less than 2.5% intraday and less than 3.5% interday. The repeatabilities of the enantioresolution values were lower than 3.0% for all studied analytes. Repeatability of areas was lower than 3.5% and 4.0% interday and intraday, respectively. Adsorption of basic analytes on inner walls of fused silica capillaries may lead to poor repeatability of migration time and peak area. However, the reproducibility data and symmetry of peaks reveals that the basic analytes were not significantly adsorbed on the inner wall of fused silica capillaries in the present study. 4. Conclusions Pen G, has been successfully explored for the first time for its chiral discriminating abilities in CE. This study reveals PenG as an effective CS for the CE enantioseparation of basic analytes. A CS concentration of 10.7 mM (0.4%, w/v) was found to give the best resolution for enantiomers of darifenacin, citalopram, sertraline and propranolol; while 16.1 mM (0.6%, w/v) was found to be the best for metoprolol. The best chiral resolutions were obtained in BGE composed of water:methanol (10:90, v/v) at 20 ◦ C with applied voltage of 5 kV. The present study indicates that the other members of ˇ-lactam class of antibiotics can also be examined for their enantiodiscriminating abilities. Also, CE has once again been proved an

[1] P. Jac, G.K. Scriba, J. Sep. Sci. 36 (2013) 52. [2] D.A. Tsioupi, S.R. Stefan-van, C.P. Kapnissi-Christodoulou, Electrophoresis 34 (2013) 178. [3] L. Sanchez-Hernandez, M. Castro-Puyana, M.L. Marina, A.L. Crego, Electrophoresis 33 (2012) 228. [4] A.P. Kumar, J.H. Park, J. Chromatogr. A 1218 (2011) 1314. [5] T. Yu, Y. Du, B. Chen, Electrophoresis 32 (2011) 1898. [6] B. Chen, Y. Du, P. Li, Electrophoresis 30 (2009) 2747. [7] A.F. Prokhorova, E.N. Shapovalova, A.V. Shpak, S.M. Staroverov, O.A. Shpigun, J. Chromatogr. A 1216 (2009) 3674. [8] L.J. Wang, J. Yang, G.L. Yang, X.G. Chen, J. Chromatogr. A 1248 (2012) 182. [9] L.J. Wang, S.Q. Hu, Q.L. Guo, G.L. Yang, X.G. Chen, J. Chromatogr. A 1218 (2011) 1300. [10] Y. Hedeland, J. Lehtinen, C. Pettersson, J. Chromatogr. A 1141 (2007) 287. [11] S. Kodama, A. Yamamoto, A. Matsunaga, K. Hayakawa, Electrophoresis 24 (2003) 2711. [12] E. Zarbl, M. Lämmerhofer, P. Franco, M. Petracs, W. Lindner, Electrophoresis 22 (2001) 3297. [13] M. Lämmerhofer, E. Zarbl, V. Piette, J. Crommen, W. Lindner, J. Sep. Sci. 24 (2001) 706. [14] V. Piette, M. Fillet, W. Lindner, J. Crommen, J. Chromatogr. A 875 (2000) 353. [15] I. Bjørnsdottir, S.H. Hansen, S. Terabe, J. Chromatogr. A 745 (1996) 37. [16] S. Dixit, J.H. Park, Biomed. Chromatogr. (2013), 10.1002/bmc.2950. [17] A.F. Prokhorova, E.N. Shapovalova, O.A. Shpigun, J. Pharma. Biomed. Anal. 53 (2010) 1170. [18] T.J. Ward, A.B. Farris, J. Chromatogr. A 906 (2001) 73. [19] C. Desiderio, S. Fanali, J. Chromatogr. A 807 (1998) 37. [20] B. Chen, Y. Du, J. Chromatogr. A 1217 (2010) 1806. [21] B. Chen, Y. Du, H. Wang, Electrophoresis 31 (2010) 371. [22] D.W. Armstrong, K.L. Rundlett, G.L. Reid, Anal. Chem. 66 (1994) 1690. [23] H.J. Choi, H.J. Ha, M.S. Shin, J.S. Jin, M.H. Hyun, J. Liq. Chromatogr. Relat. Technol. 32 (2009) 1879. [24] H.S. Cho, H.J. Ha, H.J. Choi, J.S. Jin, M.H. Hyun, Bull. Korean Chem. Soc. 29 (2008) 2135. [25] S.A.C. Wren, R.C. Rowe, J. Chromatogr. 603 (1992) 235. [26] P.B. Wright, A.S. Lister, J.G. Dorsey, Anal. Chem. 69 (1997) 3251.

Penicillin G as a novel chiral selector in capillary electrophoresis.

The penicillin sub-class of β-lactam antibiotics has not been examined for its enantiodiscriminating abilities in capillary electrophoresis (CE) until...
914KB Sizes 0 Downloads 0 Views