- Email: [email protected]
Effect of low concentration sodium dodecyl sulfate on the electromigration of palonosetron hydrochloride stereoisomers in micellar electrokinetic chromatography
Journal of Chromatography A, 1342 (2014) 86–91
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Effect of low concentration sodium dodecyl sulfate on the electromigration of palonosetron hydrochloride stereoisomers in micellar electrokinetic chromatography Shao-Qiang Hu a,∗ , Gui-Xia Wang a , Wen-Bo Guo a , Xu-Ming Guo b , Min Zhao c a
College of Chemistry and Chemical Engineering, Luoyang Normal University, Luoyang 471022, China School of Chemical Engineering & Pharmaceutics, Henan University of Science and Technology, Luoyang 471003, China c College of Chemistry and Molecular Science, Wuhan University, Wuhan 430072, China b
a r t i c l e
i n f o
Article history: Received 30 October 2013 Received in revised form 6 March 2014 Accepted 15 March 2014 Available online 24 March 2014 Keywords: Chiral separation Micellar electrokinetic chromatography Migration order Palonosetron hydrochloride Sodium cholate Sodium dodecyl sulfate
a b s t r a c t The effect of low concentrations of sodium dodecyl sulfate (SDS) on the separation of palonosetron hydrochloride (PALO) stereoisomers by micellar electrokinetic chromatography (MEKC) has been investigated. It was found that the addition of SDS prolongs the migration time and the migration order of four stereoisomers changes regularly with the SDS concentration. Good separations for all the four stereoisomers were achieved at appropriate SDS concentration. The effect of SDS on the electromigration (mobilities) of PALO stereoisomers has been studied, in order to explain its effect on the separation by MEKC. It was found that low concentrations of SDS added into the separation media forms negatively charged complexes with PALO stereoisomers and hence reverses their electromigration direction. Furthermore, the migration order between two enantiomeric pairs is also reversed because the enantiomeric pair with a bigger positive mobility than that of another pair turns to have a bigger negative mobility when bound with SDS. Based on these results, the effect of SDS on the MEKC separation of PALO stereoisomers was elucidated reasonably. The performance of the developed chiral MEKC method was validated by the analysis of a real sample. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Palonosetron hydrochloride (PALO (3aS, 2S)), (3aS)-2-[(S)-1azabi-cyclo[2.2.2]oct-3-yl]-2,3,3a, 4,5,6-hexahydro-1H-benz[de]isoquinolin-1-one hydrochloride, is a highly selective second generation 5-HT3 receptor antagonist used for the prevention of nausea and vomiting associated with chemotherapy [1–4]. It contains two chiral centers in molecular structure (see also Appendix A, Fig. S1) and thus has four stereoisomers belonging respectively to two pairs of enantiomers, i.e. PALO (3aS, 2S), PALO (3aR, 2R), PALO (3aS, 2R) and PALO (3aR, 2S). Among them only PALO (3aS, 2S) possesses pharmacological activity [5,6]. Therefore, development of enantiomeric separation methods for PALO stereoisomers is of great importance for the control of its enantiomeric impurities and avoiding unwanted pharmaceutical and toxicological side effects [5–8].
∗ Corresponding author. Tel.: +86 379 65515113. E-mail address: [email protected] (S.-Q. Hu). http://dx.doi.org/10.1016/j.chroma.2014.03.042 0021-9673/© 2014 Elsevier B.V. All rights reserved.
Micellar electrokinetic chromatography (MEKC), as a combination of chromatographic and electrophoretic mechanisms, is an important separation mode of capillary electrophoresis (CE) [9–12]. The separation based on the former mechanism depends on the selectivity, relative strength of affinity of solutes for the micelles. The separation based on the latter mechanism depends on the mobility difference between two solutes. Separation of the four stereoisomers of PALO by MEKC using sodium cholate (SC) as surfactant and chiral selector has been reported [13,14], which includes the separation in each enantiomeric pair (chiral separation) and the separation between enantiomeric pairs (achiral separation). Our previous research [14] found that the migration orders in each enantiomeric pair are 3aS, 2S < 3aR, 2R and 3aS, 2R < 3aR, 2S (the sequence of migration time), determined by selectivity (chromatographic mechanism). The enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) is eluted before enantiomeric pair (3aS, 2R), (3aR, 2S) due to the mobility difference (electrophoretic mechanism). Because of the offset of the two mechanisms which give opposite migration orders, the peaks of PALO (3aR, 2R) and (3aS, 2R), the second enantiomer of the first pair and the first enantiomer of the second pair, coalesce in pure MEKC at the original
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
87
pH of borate buffer (pH 9.20). Therefore, the separation for all of the four stereoisomers is not achieved, although the resolution in each pair of enantiomers is very good. In order to achieve the separation between PALO (3aR, 2R) and (3aS, 2R), Tian et al. [13] added methanol of high concentration (solvent modified MEKC) and we added butanol of low concentration (cosurfactant modified MEKC) into the micellar solution [14]. The mechanism of separation is tipping the balance between two effects by weakening the chromatographic effect while strengthening the electrophoretic effect, according to the measured selectivity and mobility data. In this work, a commonly used anionic surfactant sodium dodecyl sulfate (SDS) was employed as an alternative to organic solvents in selectivity tuning of MEKC. The separation between PALO (3aR, 2R) and (3aS, 2R) was obtained by adding low concentration of SDS into the separation media. The mechanism of separation is altering the electromigration (mobilities) of PALO stereoisomers through the formation of negatively charged complexes with SDS. 2. Material and methods 2.1. Chemicals and reagents The four enantiomerically pure PALO stereoisomers were purchased from J&K Scientific Ltd. (Beijing, China). Sodium cholate (SC) was purchased from Serva Feinbiochemica (Heidelberg, Germany). PALO injection (PALO (3aS, 2S) content, 50 g mL−1 ), a real sample for the validation of the method, was the product of Chia Tai Tianqing Pharmaceutical Co. Ltd. (Jiangsu, China). Other chemicals used were of analytical regent grade and were used without further purification. 2.2. Preparation of separation media (BGE) and sample solutions Micelle solutions for MEKC were prepared by dissolving appropriate quantities of SC surfactant, SDS if necessary, and sodium tetraborate buffer in distilled water to the desired volume in a flask. The solution was sonicated while covered for 15 min to form a transparent micelle solution. The sample solutions of enantiomerically pure standards, used in the development of the method, were prepared by dissolving appropriate quantities of each enantiomerically pure PALO steroisomer mixedly in appropriate background electrolyte (BGE) to a concentration of 0.1 mg mL−1 . A small amount of methanol was added as an electroosmotic flow (EOF) marker. The spiked sample solutions, for the validation of the developed method, were prepared by adding appropriate quantities of enatiomeric impurity standards mixedly into the PALO injection to concentrations ranging from 0.5 to 5.0 g mL−1 . All solutions were filtered through a 0.45 m filter prior to use. 2.3. Electrophoresis experiments A TH-3100 capillary electrophoresis system equipped with a UV detector (Tianhui Institute of Separation Science, Hebei, China) was employed for all electrophoresis (CE) experiments. The detection wavelength was 214 nm. An uncoated fused silica capillary of id 50 m × od 365 m (Yongnian Photoconductive Fiber Factory, Hebei, China) was used, with a total length (Ltot ) of 60.0 cm and an effective length (Leff ) of 50.0 cm. New capillaries were pretreated by flushing in sequence with distilled water for 5 min, 1.0 M NaOH for 10 min and distilled water for 5 min again at 140 kPa (approx. 20 psi). Between injections, the capillary was rinsed in sequence with distilled water, 1.0 M NaOH, distilled water again and finally BGE for 2 min each at 140 kPa. The capillary cartridge temperature was maintained at 20 ◦ C. Injections were performed hydrodynamically at 10 kPa (approx. 1.5 psi) for 1 s and 5 s, for the solutions
Fig. 1. Effect of SDS concentration on separation of PALO stereoisomers by MEKC. Composition of micellar solution is 30 mM SC in 30 mM sodium tetraborate of pH 9.20. Detection wavelength: 214 nm. Capillary: id 50 m, Ltot 60.0 cm, Leff 50.0 cm. Capillary temperature: 20 ◦ C. Hydrodynamic injection at 10 kPa for 1 s. Applied voltage: 25 kV.
of enantiomerically pure standards and the spiked sample solutions, respectively. An applied voltage of 25 kV was used for all experiments. 2.4. UV spectrometry tests The UV spectra of PALO stereoisomers in the presence of varying concentrations of SC or SDS surfactant in 30 mM sodium tetraborate buffer were recorded in the range from 200 to 300 nm with a UV-3200 spectrophotometer (Shanghai Mapada Instruments Co. Ltd., Shanghai, China), using quartz cells with 1 cm path length. The reference cell contained the same concentrations of surfactant and buffer as the sample. 3. Results and discussion 3.1. Effect of SDS on separation of PALO stereoisomers by MEKC At the top of Fig. 1 is the electropherogram of PALO stereoisomers obtained by pure MEKC (no SDS addition) using a 30 mM SC micelle solution. As mentioned earlier, the peaks of PALO (3aR, 2R) and (3aS, 2R) coalesce due to the offset of chromatographic mechanism and electrophoretic mechanism, so the migration order of four stereoisomers is 3aS, 2S < 3aR, 2R = 3aS, 2R < 3aR, 2S (the sequence of migration time). When low concentration of SDS is added into the separation media, the migration time of stereoisomers is prolonged and it increases with the concentration of SDS. Furthermore, the migration order between two enantiomeric pairs is reversed, i.e., the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) move gradually to the rear of those of another pair (3aS, 2R), (3aR, 2S) (Fig. 1). As can be seen in Fig. 1, when 1 mM of SDS is added into the micelle solution, the coalesced peak of PALO (3aR, 2R) and (3aS, 2R) is resolved and the peaks belonging to two enantiomeric pairs begin to interlace, i.e., the first enantiomer of the second pair, PALO (3aS, 2R), is eluted before the second enantiomer of the first
88
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
pair, PALO (3aR, 2R). The migration order becomes 3aS, 2S < 3aS, 2R < 3aR, 2R < 3aR, 2S. When 2 mM of SDS is added, the backward movement of the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) relative to (3aS, 2R), (3aR, 2S) leads to the narrowing of the distance between the peaks of (3aS, 2S) and (3aS, 2R), as well as the peaks of (3aR, 2R) and (3aR, 2S), but the migration order does not change. When 3 mM of SDS is added, the peaks of (3aS, 2S) and (3aS, 2R), as well as the peaks of (3aR, 2R) and (3aR, 2S), coalesce. The migration order becomes 3aS, 2S ≈ 3aS, 2R < 3aR, 2R ≈ 3aR, 2S. When 4 mM of SDS is added, the coalesced peak of (3aS, 2S) plus (3aS, 2R) splits again, but the peak of (3aS, 2R) is in front of that of (3aS, 2S), caused by the further backward movement of the peaks of (3aS, 2S), (3aR, 2R) relative to (3aS, 2R), (3aR, 2S). The migration order becomes 3aS, 2R < 3aS, 2S < 3aR, 2R ≈ 3aR, 2S. When 5 mM of SDS is added, the coalesced peak of (3aR, 2R) and (3aR, 2S) also splits and their order is reversed, too. The migration order becomes 3aS, 2R < 3aS, 2S < 3aR, 2S < 3aR, 2R. When 6 mM of SDS is added, the relative backward movement of the peaks of (3aS, 2S), (3aR, 2R) provides better resolutions, but the migration order does not change. When 7 mM of SDS is added, the relative backward movement of the peaks of (3aS, 2S), (3aR, 2R) deteriorates the resolution between (3aS, 2S) and (3aR, 2S) due to the narrowing of their peak distance, the migration order does not change either. In term of separation for all of four stereoisomers, 1 mM is the optimal SDS concentration, by which baseline resolution is achieved for all the adjacent peaks (Fig. 1). Compared to the reported MEKC methods for PALO stereoisomers [13,14], this method has the advantages of not using organic solvent and low chiral selector (SC) and time consumption. When 6 mM of SDS is used, baseline resolution for all the adjacent peaks is nearly achieved too and different migration order is obtained (Fig. 1). Although the separation time is relatively long, being able to change the migration order will be beneficial to avoiding the interference of some peaks in the chiral separation for real samples. The effects of SDS on the separation of PALO stereoisomers are similar at different SC concentrations, i.e., it always causes the prolongation of the migration time and the reversion of the migration order between two enantiomeric pairs, as shown in Figs. S2 and S3 in Appendix A. However, its effect is weaker at a higher SC concentration. The SDS concentration required to obtain a similar separation increases with the SC concentration. 3.2. Effect of SDS on electromigration of PALO stereoisomers In order to explain the effect of SDS on the separation of PALO stereoisomers in MEKC, its effect on the electromigration (mobilities) of PALO stereoisomers in borate buffer of the same concentration and pH without SC micelles was investigated. Only two diastereoisomers, PALO (3aS, 2S) and (3aS, 2R), were used in this set of experiment, each of them represents a pair of enantiomers in achiral media. As can be seen in Fig. 2, in pure borate buffer, PALO stereoisomers are eluted before the EOF marker methanol due to being positively charged and PALO (3aS, 2S) is eluted before (3aS, 2R), suggesting that it has a bigger electrophoretic mobility than that of PALO (3aS, 2R). When only 1 mM of SDS is added into the buffer, the peaks of PALO stereoisomers move to the rear of methanol peak. And the backward movement of the peaks continues along with the increase of SDS concentration. Referring to the reported fact that ionic species form ion-pair with monomeric surfactant of the opposite charge, resulting in the changes of their mobility in MEKC [15], we attribute this phenomenon to the formation of negatively charged complexes with SDS by electrostatic and/or other interactions: PALO+ + nSDS− PALO · nSDS1−n
Fig. 2. Effect of low concentration SDS on electromigration of PALO stereoisomers in borate buffer. BGE component except for SDS is 30 mM sodium tetraborate, pH 9.20. CE conditions are the same as in Fig. 1.
The increase in SDS concentration would enlarge the fraction of complex or the binding number n, causing the increase of the average negative charge carried by PALO stereoisomers and consequently their negative mobility and migration time. The addition of SDS not only prolongs the migration time of PALO stereoisomers, but also reverses their migration order. PALO (3aS, 2S) is eluted after PALO (3aS, 2R) when SDS concentration is above 2 mM (Fig. 2). It suggests that PALO (3aS, 2S) with a bigger positive mobility (eluted first) in pure borate buffer, compared to PALO (3aS, 2R), also has a bigger negative mobility (eluted late) under the existence of SDS, probably because it has stronger interactions with SDS. It is also found in Fig. 2 that the peaks of PALO stereoisomers become very broad and asymmetric when 1 mM to 2 mM of SDS is added. In order to explain this phenomenon, we suppose that PALO stereoisomers may form complexes with different binding numbers and thus different mobilities. If the SDS concentration is very low, the complexes of different stoichiometric ratios can not exchange quickly because of the large distance between PALO stereoisomers and SDS ions. This slow mass transfer may cause the band broadening in CE. The situation is similar to that reported in early literatures about MEKC, the efficiency of MEKC decreases dramatically when surfactant concentrations are close to the CMC because of slow mass transfer caused by large distance between micelles [16]. When SDS concentration is increased from 3 mM to 7 mM, the peaks of PALO stereoisomers become narrow because the increase in the number of SDS ions reduces their average distance to PALO stereoisomers and thus improves the mass transfer. When SDS concentration exceeds the CMC, there is also ion–micelle interaction (partition into micelles) between PALO stereoisomers and SDS; it may provide bridges for the interaction between PALO stereoisomers and SDS monomers, causing the further improvement of the mass transfer. The reported CMC of SDS in pure water is 8.1 mM [11,12], but inorganic salts in the solution may cause the significant decrease of the CMC [17–19]. The CMC value of SDS in 30 mM borate buffer of pH 9.20, the media of this experiment, measured with dye solubilization method [17] in our laboratory is 4.2 mM. Since there are also free surfactant ions of certain concentration (CMC in theory) in micelle solution, the separation media of MEKC, the effect of SDS on the electromigration of PALO stereoisomers in borate buffer containing SC of 12 mM (approx. the CMC) was also investigated. The reported CMC values of SC range from
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
Fig. 3. Effect of low concentration SC and SDS on electromigration of PALO stereoisomers in borate buffer. BGE component except for SC and SDS is 30 mM sodium tetraborate, pH 9.20. CE conditions are the same as in Fig. 1.
10 to 15 mM or even wider, depending on the solution background as well as the measurement method [19–23]. The value in 30 mM borate buffer of pH 9.20 measured with dye solubilization method [17] in our laboratory is 12.1 mM. As shown in Fig. 3, SC at low concentrations has less effect on the electromigration of PALO stereoisomers, compared to SDS. If only SC of different concentrations is added into borate buffer, the peaks of PALO stereoisomers are always in front of that of methanol, showing their positive charge, although they may also interact with negatively charged SC, particularly when the concentration is near the CMC, causing the decrease of their mobility. The migration order is always PALO
89
(3aS, 2S) before (3aS, 2R), suggesting that the relative size of their mobilities is basically not affected. When 1 mM of SDS is added, the peaks of PALO stereoisomers move to the rear of methanol peak. With the increase of SDS concentration, the migration time of PALO stereoisomers is prolonged and their migration order is reversed (Fig. 3). The situation is the same as when only SDS is added (Fig. 2). Our explanation is also the formation of negatively charged complexes with SDS. Obviously, PALO stereoisomers have stronger interactions with the monomers of SDS than SC, although both of them are anionic surfactants. Our explanation for this phenomenon is that besides the electrostatic interaction between PALO cations and SDS anions, there is also hydrophobic interaction between the hydrocarbon chain of SDS and the hydrophobic part of PALO stereoisomers. The flexible long hydrocarbon chain of SDS may be entangled around the hydrophobic part of PALO stereoisomers, reducing their exposure to aqueous solution, thus making them more stable. Monomeric SC should have the same tendency of binding with PALO stereoisomers through hydrophobic interaction, but the rigid structures of both of them (Appendix A, Fig. S1) hamper the close contact between their hydrophobic parts. This explanation is supported by the UV spectrometry experiment. As shown in Fig. 4, the addition of submicellar concentrations of SC up to 10 mM has little effect on the UV spectra of PALO (3aS, 2S) and (3aS, 2R). However, the relative strength of the two bands in the UV spectra is changed evidently by the addition of submicellar concentrations of SDS. The first band (centered at 242 nm for 3aS, 2S and 241 nm for 3aS, 2R) is increased while the second band (centered at 256 nm for 3aS, 2S and 257 nm for 3aS, 2R) is decreased compared to the UV spectra in pure borate buffer, suggesting the change of chemical environment around the chromophores of PALO stereoisomers, perhaps caused by the surrounding of the hydrocarbon chains of SDS.
Fig. 4. Effect of SC and SDS with concentrations below the CMC on the UV spectra of PALO stereoisomers. The concentrations of PALO stereoisomers (0.05 mM, 17 g mL−1 ) and sodium tetraborate buffer (30 mM, pH 9.20) are constant for every spectrum. The reference solutions contain the same concentrations of surfactant and buffer as the sample.
90
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
Table 1 Quantitative parameters of the developed method for the analysis of the PALO injection spiked with enantiomeric impurities.
Calibration range (g mL−1 ) Regression equation
Slope, a Intercept, b Correlation coefficient, R Limit of detection (LOD, g mL−1 ) Limit of quantification (LOQ, g mL−1 ) Recovery (%) a Repeatability (RSD, %) b Intra-day c Inter-day d a b c d
PALO (3aS, 2R)
PALO (3aR, 2R)
PALO (3aR, 2S)
0.5–5.0 y = ax + b y: peak area (mV s); x: concentration of enantiomeric impurities (g mL−1 ) 1.668 −0.001 0.9992 0.08 0.28 100.2
0.5–5.0
0.5–5.0
1.724 −0.083 0.9989 0.08 0.28 104.3
1.600 0.099 0.9982 0.09 0.31 104.2
8.6 9.2
3.8 8.5
7.4 8.9
Values at an enantiomeric impurity concentration of 1.0 g mL−1 . Values at an enantiomeric impurity concentration of 1.0 g mL−1 . RSD of peak areas for five replications. RSD of peak areas for 15 replicates in 3 days, five replicates each day.
3.3. Explanation to effect of SDS on MEKC separation of PALO stereoisomers
concentration but no SC added (Fig. 2), although the EOF are basically equal.
The effect of SDS on MEKC separation of PALO stereoisomers can be elucidated reasonably based on our investigation about its effect on the electromigration of PALO stereoisomers. According to the basic theory of MEKC, the observed velocity of a solute in MEKC is the weighted average for its different forms, including all the forms distributed in water phase and micelles [9–11]. For the separation without SDS addition, the positively charged PALO stereoisomers in water phase swim downstream to the direction of capillary outlet while the PALO stereoisomers distributed in negatively charged micelles swim upstream to the direction of capillary inlet. As the upstream migration is partially offset by the downstream migration, the observed (average) velocities are relatively high and the migration times are relatively short. If the PALO stereoisomers in water phase form negatively charged complexes with SDS, they also swim upstream, thus the observed velocities decrease and the migration times are prolonged. When the SDS concentration is increased, the average negative charge carried by PALO stereoisomers increases, due to the increase of the fraction of complexes or the binding number. Therefore, their negative mobility increases, leading to the further decrease of observed velocities and finally the prolongation of migration times. Since PALO (3aS, 2S) with a bigger positive mobility (eluted first) than that of PALO (3aS, 2R) in pure borate buffer turns to have a bigger negative mobility (eluted late) when binding with SDS (each of them represents a pair of enantiomers), the migration order between two enantiomeric pairs is reversed. Shown in Fig. 1, the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) move to the rear of those of another pair (3aS, 2R), (3aR, 2S) gradually when the SDS concentration is increased. Different from in pure borate buffer (Fig. 2), the addition of SDS of even very low concentration into SC micelle in MEKC does not broaden the peaks (Fig. 1), probably because high concentration of SC micelles provide bridges for the interaction between PALO stereoisomers and SDS, thus improving the mass transfer. In our opinion, the weaker effect of SDS on the separation of PALO stereoisomers at a higher SC concentration may be ascribed, on one hand, to the decrease in the fraction of PALO–SDS complexes, caused by the competition effect from the partition of PALO stereoisomers into SC micelles. On the other hand, part of SDS may combine with SC at high concentration, reducing its interaction with PALO stereoisomers. This can be demonstrated by the fact that the migration times of PALO stereoisomers when 12 mM of SC is added (Fig. 3) are shorter than those at the same SDS
3.4. Validation of the method The developed method was used for the analysis of a real sample, PALO injection, in order to check the applicability. Its nominal content of effective ingredient, PALO (3aS, 2S), is 50 g mL−1 . As the enantiomeric impurities, i.e. PALO (3aS, 2R), PALO (3aR, 2R) and PALO (3aR, 2S), were not detected (Fig. 5), the standards of them were spiked into the PALO injection to concentrations ranging from 0.5 to 5.0 g mL−1 . The spiked samples were injected directly. In order to improve the sensitivity and precision, the sample injection time was increased from 1 s to 5 s. Other conditions optimized above remained unchanged. The peak widths of the enantiomeric impurities are basically not changed with the increase of injection time, whereas the peak width of the main ingredient, PALO (3aS, 2S), increases from 1.3 s to 1.7 s (W1/2 ) as the injection time is changed from 1 s to 5 s. However, a baseline resolution between PALO (3aS, 2S) and PALO (3aS, 2R) can still be obtained
Fig. 5. Electropherograms of PALO injection spiked with different concentrations of enantiomeric impurities. BGE composition is 30 mM SC and 1 mM SDS in 30 mM sodium tetraborate of pH 9.20. CE conditions are the same as in Fig. 1, except that the injection time is 5 s.
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
(Fig. 5). Further increase of injection time influences the separation between PALO (3aS, 2S) and PALO (3aS, 2R), but baseline resolutions among PALO (3aS, 2R), PALO (3aR, 2R) and PALO (3aR, 2S) can be obtained till an injection time of 10 s (data not given). Good precision, recovery, and linear relationship between the peak areas and the impurity concentrations were obtained (Table 1). The limit of detection (LOD, S/N = 3) and limit of quantification (LOQ, S/N = 10) for the enantiomeric impurities are approximately 0.1 g mL−1 and 0.3 g mL−1 , respectively. 4. Conclusions SDS can be an alternative to organic solvents in selectivity tuning, in the separation of PALO stereoisomers by MEKC using SC as chiral selector. Based on the analysis of electropherograms, low concentrations of SDS added into the separation media forms negatively charged complexes with PALO stereoisomers and hence reverses their electromigration direction. Consequently, the separation results of PALO stereoisomers are changed dramatically. The peaks originally coalesced are resolved and the migration order of stereoisomers is changed regularly with the concentration of SDS. Good separations for all the four stereoisomers can be achieved at appropriate SDS concentration. The optimized method has good precision, accuracy, and linear relationship in the analysis of enantiomeric impurities in the PALO injection, although the sensitivity is slightly low. Acknowledgment The financial support (Grant no. 12A150017) as Key Project of Science and Technology Research by Department of Education of Henan Province, PR China, is gratefully acknowledged.
91
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.chroma.2014.03.042. References [1] R. de Wit, M. Aapro, P. Blower, Cancer Chemother. Pharmacol. 56 (2005) 231. [2] M. Aapro, P. Blower, Cancer 104 (2005) 1. [3] J.T. Hickok, J.A. Roscoe, G.R. Morrow, C.W. Bole, H. Zhao, K.L. Hoelzer, S.R. Dakhil, T. Moore, T.R. Fitch, Lancet Oncol. 6 (2005) 765. [4] P. Eisenberg, J. Figueroa-Vadillo, R. Zamora, V. Charu, J. Hajdenberg, A. Cartmell, A. Macciocchi, S. Grunberg, Cancer 98 (2003) 2473. [5] M. Wang, X.-P. Ding, H.-L. Chen, X.-G. Chen, Anal. Sci. 25 (2009) 1217. [6] M.V. Murthy, K. Jyothirmayi, K. Srinivas, K. Mukkanti, R. Kumar, G. Samanta, Am. J. Anal. Chem. 2 (2011) 437. [7] X.-R. Yu, M. Song, T.-J. Hang, Chin. J. New Drugs 17 (2008) 870. [8] P. Radhakrishnanand, D.V. Subba Rao, V. Himabindu, Chromatographia 69 (2009) 369. [9] S. Terabe, K. Otsuka, K. Ichikawa, A. Tsuchiya, T. Ando, Anal. Chem. 56 (1984) 111. [10] S. Terabe, K. Otsuka, T. Ando, Anal. Chem. 57 (1985) 834. [11] K.R. Nielsen, J.P. Foley, in: P. Camilleri (Ed.), Capillary Electrophesis—Theory and Practice, CRC Press, Boca Raton, FL, 1998, p. 135. [12] S.E. Deeb, M.A. Iriban, R. Gust, Electrophoresis 32 (2011) 166. [13] K. Tian, H.-L. Chen, J.-H. Tang, X.-G. Chen, Z.-D. Hu, J. Chromatogr. A 1132 (2006) 333. [14] S.-Q. Hu, H.-B. Yang, H.-J. Shi, Y.-H. Zhang, Z. Yang, Electrophoresis 34 (2013) 3086. [15] T. Kaneta, S. Tanaka, M. Taga, H. Yoshida, Anal. Chem. 64 (1992) 798. [16] M.J. Sepaniak, R.O. Cole, Anal. Chem. 59 (1987) 472. [17] H. Nakamura, A. Sano, K. Matsuura, Anal. Sci. 14 (1998) 379. [18] G.-X. He, L. Zhang, G. Li, Chin. Chem. Res. 22 (2011) 68. [19] E. Fuguet, C. Ràfols, M. Rosés, E. Bosch, Anal. Chim. Acta 548 (2005) 95. [20] C.M. Hebling, L.E. Thompson, K.W. Eckenroad, G.A. Manley, R.A. Fry, K.T. Mueller, T.G. Strein, D. Rovnyak, Langmuir 24 (2008) 13866. [21] S. Reis, C.G. Moutinho, C. Matos, B. de Castro, P. Gameiro, J.L.F.C. Lima, Anal. Biochem. 334 (2004) 117. [22] P. Garidel, A. Hildebrand, R. Neubert, A. Blume, Langmuir 16 (2000) 5267. [23] A. Hildebrand, P. Garidel, R. Neubert, A. Blume, Langmuir 20 (2003) 320.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Effect of low concentration sodium dodecyl sulfate on the electromigration of palonosetron hydrochloride stereoisomers in micellar electrokinetic chromatography Shao-Qiang Hu a,∗ , Gui-Xia Wang a , Wen-Bo Guo a , Xu-Ming Guo b , Min Zhao c a
College of Chemistry and Chemical Engineering, Luoyang Normal University, Luoyang 471022, China School of Chemical Engineering & Pharmaceutics, Henan University of Science and Technology, Luoyang 471003, China c College of Chemistry and Molecular Science, Wuhan University, Wuhan 430072, China b
a r t i c l e
i n f o
Article history: Received 30 October 2013 Received in revised form 6 March 2014 Accepted 15 March 2014 Available online 24 March 2014 Keywords: Chiral separation Micellar electrokinetic chromatography Migration order Palonosetron hydrochloride Sodium cholate Sodium dodecyl sulfate
a b s t r a c t The effect of low concentrations of sodium dodecyl sulfate (SDS) on the separation of palonosetron hydrochloride (PALO) stereoisomers by micellar electrokinetic chromatography (MEKC) has been investigated. It was found that the addition of SDS prolongs the migration time and the migration order of four stereoisomers changes regularly with the SDS concentration. Good separations for all the four stereoisomers were achieved at appropriate SDS concentration. The effect of SDS on the electromigration (mobilities) of PALO stereoisomers has been studied, in order to explain its effect on the separation by MEKC. It was found that low concentrations of SDS added into the separation media forms negatively charged complexes with PALO stereoisomers and hence reverses their electromigration direction. Furthermore, the migration order between two enantiomeric pairs is also reversed because the enantiomeric pair with a bigger positive mobility than that of another pair turns to have a bigger negative mobility when bound with SDS. Based on these results, the effect of SDS on the MEKC separation of PALO stereoisomers was elucidated reasonably. The performance of the developed chiral MEKC method was validated by the analysis of a real sample. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Palonosetron hydrochloride (PALO (3aS, 2S)), (3aS)-2-[(S)-1azabi-cyclo[2.2.2]oct-3-yl]-2,3,3a, 4,5,6-hexahydro-1H-benz[de]isoquinolin-1-one hydrochloride, is a highly selective second generation 5-HT3 receptor antagonist used for the prevention of nausea and vomiting associated with chemotherapy [1–4]. It contains two chiral centers in molecular structure (see also Appendix A, Fig. S1) and thus has four stereoisomers belonging respectively to two pairs of enantiomers, i.e. PALO (3aS, 2S), PALO (3aR, 2R), PALO (3aS, 2R) and PALO (3aR, 2S). Among them only PALO (3aS, 2S) possesses pharmacological activity [5,6]. Therefore, development of enantiomeric separation methods for PALO stereoisomers is of great importance for the control of its enantiomeric impurities and avoiding unwanted pharmaceutical and toxicological side effects [5–8].
∗ Corresponding author. Tel.: +86 379 65515113. E-mail address: [email protected] (S.-Q. Hu). http://dx.doi.org/10.1016/j.chroma.2014.03.042 0021-9673/© 2014 Elsevier B.V. All rights reserved.
Micellar electrokinetic chromatography (MEKC), as a combination of chromatographic and electrophoretic mechanisms, is an important separation mode of capillary electrophoresis (CE) [9–12]. The separation based on the former mechanism depends on the selectivity, relative strength of affinity of solutes for the micelles. The separation based on the latter mechanism depends on the mobility difference between two solutes. Separation of the four stereoisomers of PALO by MEKC using sodium cholate (SC) as surfactant and chiral selector has been reported [13,14], which includes the separation in each enantiomeric pair (chiral separation) and the separation between enantiomeric pairs (achiral separation). Our previous research [14] found that the migration orders in each enantiomeric pair are 3aS, 2S < 3aR, 2R and 3aS, 2R < 3aR, 2S (the sequence of migration time), determined by selectivity (chromatographic mechanism). The enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) is eluted before enantiomeric pair (3aS, 2R), (3aR, 2S) due to the mobility difference (electrophoretic mechanism). Because of the offset of the two mechanisms which give opposite migration orders, the peaks of PALO (3aR, 2R) and (3aS, 2R), the second enantiomer of the first pair and the first enantiomer of the second pair, coalesce in pure MEKC at the original
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
87
pH of borate buffer (pH 9.20). Therefore, the separation for all of the four stereoisomers is not achieved, although the resolution in each pair of enantiomers is very good. In order to achieve the separation between PALO (3aR, 2R) and (3aS, 2R), Tian et al. [13] added methanol of high concentration (solvent modified MEKC) and we added butanol of low concentration (cosurfactant modified MEKC) into the micellar solution [14]. The mechanism of separation is tipping the balance between two effects by weakening the chromatographic effect while strengthening the electrophoretic effect, according to the measured selectivity and mobility data. In this work, a commonly used anionic surfactant sodium dodecyl sulfate (SDS) was employed as an alternative to organic solvents in selectivity tuning of MEKC. The separation between PALO (3aR, 2R) and (3aS, 2R) was obtained by adding low concentration of SDS into the separation media. The mechanism of separation is altering the electromigration (mobilities) of PALO stereoisomers through the formation of negatively charged complexes with SDS. 2. Material and methods 2.1. Chemicals and reagents The four enantiomerically pure PALO stereoisomers were purchased from J&K Scientific Ltd. (Beijing, China). Sodium cholate (SC) was purchased from Serva Feinbiochemica (Heidelberg, Germany). PALO injection (PALO (3aS, 2S) content, 50 g mL−1 ), a real sample for the validation of the method, was the product of Chia Tai Tianqing Pharmaceutical Co. Ltd. (Jiangsu, China). Other chemicals used were of analytical regent grade and were used without further purification. 2.2. Preparation of separation media (BGE) and sample solutions Micelle solutions for MEKC were prepared by dissolving appropriate quantities of SC surfactant, SDS if necessary, and sodium tetraborate buffer in distilled water to the desired volume in a flask. The solution was sonicated while covered for 15 min to form a transparent micelle solution. The sample solutions of enantiomerically pure standards, used in the development of the method, were prepared by dissolving appropriate quantities of each enantiomerically pure PALO steroisomer mixedly in appropriate background electrolyte (BGE) to a concentration of 0.1 mg mL−1 . A small amount of methanol was added as an electroosmotic flow (EOF) marker. The spiked sample solutions, for the validation of the developed method, were prepared by adding appropriate quantities of enatiomeric impurity standards mixedly into the PALO injection to concentrations ranging from 0.5 to 5.0 g mL−1 . All solutions were filtered through a 0.45 m filter prior to use. 2.3. Electrophoresis experiments A TH-3100 capillary electrophoresis system equipped with a UV detector (Tianhui Institute of Separation Science, Hebei, China) was employed for all electrophoresis (CE) experiments. The detection wavelength was 214 nm. An uncoated fused silica capillary of id 50 m × od 365 m (Yongnian Photoconductive Fiber Factory, Hebei, China) was used, with a total length (Ltot ) of 60.0 cm and an effective length (Leff ) of 50.0 cm. New capillaries were pretreated by flushing in sequence with distilled water for 5 min, 1.0 M NaOH for 10 min and distilled water for 5 min again at 140 kPa (approx. 20 psi). Between injections, the capillary was rinsed in sequence with distilled water, 1.0 M NaOH, distilled water again and finally BGE for 2 min each at 140 kPa. The capillary cartridge temperature was maintained at 20 ◦ C. Injections were performed hydrodynamically at 10 kPa (approx. 1.5 psi) for 1 s and 5 s, for the solutions
Fig. 1. Effect of SDS concentration on separation of PALO stereoisomers by MEKC. Composition of micellar solution is 30 mM SC in 30 mM sodium tetraborate of pH 9.20. Detection wavelength: 214 nm. Capillary: id 50 m, Ltot 60.0 cm, Leff 50.0 cm. Capillary temperature: 20 ◦ C. Hydrodynamic injection at 10 kPa for 1 s. Applied voltage: 25 kV.
of enantiomerically pure standards and the spiked sample solutions, respectively. An applied voltage of 25 kV was used for all experiments. 2.4. UV spectrometry tests The UV spectra of PALO stereoisomers in the presence of varying concentrations of SC or SDS surfactant in 30 mM sodium tetraborate buffer were recorded in the range from 200 to 300 nm with a UV-3200 spectrophotometer (Shanghai Mapada Instruments Co. Ltd., Shanghai, China), using quartz cells with 1 cm path length. The reference cell contained the same concentrations of surfactant and buffer as the sample. 3. Results and discussion 3.1. Effect of SDS on separation of PALO stereoisomers by MEKC At the top of Fig. 1 is the electropherogram of PALO stereoisomers obtained by pure MEKC (no SDS addition) using a 30 mM SC micelle solution. As mentioned earlier, the peaks of PALO (3aR, 2R) and (3aS, 2R) coalesce due to the offset of chromatographic mechanism and electrophoretic mechanism, so the migration order of four stereoisomers is 3aS, 2S < 3aR, 2R = 3aS, 2R < 3aR, 2S (the sequence of migration time). When low concentration of SDS is added into the separation media, the migration time of stereoisomers is prolonged and it increases with the concentration of SDS. Furthermore, the migration order between two enantiomeric pairs is reversed, i.e., the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) move gradually to the rear of those of another pair (3aS, 2R), (3aR, 2S) (Fig. 1). As can be seen in Fig. 1, when 1 mM of SDS is added into the micelle solution, the coalesced peak of PALO (3aR, 2R) and (3aS, 2R) is resolved and the peaks belonging to two enantiomeric pairs begin to interlace, i.e., the first enantiomer of the second pair, PALO (3aS, 2R), is eluted before the second enantiomer of the first
88
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
pair, PALO (3aR, 2R). The migration order becomes 3aS, 2S < 3aS, 2R < 3aR, 2R < 3aR, 2S. When 2 mM of SDS is added, the backward movement of the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) relative to (3aS, 2R), (3aR, 2S) leads to the narrowing of the distance between the peaks of (3aS, 2S) and (3aS, 2R), as well as the peaks of (3aR, 2R) and (3aR, 2S), but the migration order does not change. When 3 mM of SDS is added, the peaks of (3aS, 2S) and (3aS, 2R), as well as the peaks of (3aR, 2R) and (3aR, 2S), coalesce. The migration order becomes 3aS, 2S ≈ 3aS, 2R < 3aR, 2R ≈ 3aR, 2S. When 4 mM of SDS is added, the coalesced peak of (3aS, 2S) plus (3aS, 2R) splits again, but the peak of (3aS, 2R) is in front of that of (3aS, 2S), caused by the further backward movement of the peaks of (3aS, 2S), (3aR, 2R) relative to (3aS, 2R), (3aR, 2S). The migration order becomes 3aS, 2R < 3aS, 2S < 3aR, 2R ≈ 3aR, 2S. When 5 mM of SDS is added, the coalesced peak of (3aR, 2R) and (3aR, 2S) also splits and their order is reversed, too. The migration order becomes 3aS, 2R < 3aS, 2S < 3aR, 2S < 3aR, 2R. When 6 mM of SDS is added, the relative backward movement of the peaks of (3aS, 2S), (3aR, 2R) provides better resolutions, but the migration order does not change. When 7 mM of SDS is added, the relative backward movement of the peaks of (3aS, 2S), (3aR, 2R) deteriorates the resolution between (3aS, 2S) and (3aR, 2S) due to the narrowing of their peak distance, the migration order does not change either. In term of separation for all of four stereoisomers, 1 mM is the optimal SDS concentration, by which baseline resolution is achieved for all the adjacent peaks (Fig. 1). Compared to the reported MEKC methods for PALO stereoisomers [13,14], this method has the advantages of not using organic solvent and low chiral selector (SC) and time consumption. When 6 mM of SDS is used, baseline resolution for all the adjacent peaks is nearly achieved too and different migration order is obtained (Fig. 1). Although the separation time is relatively long, being able to change the migration order will be beneficial to avoiding the interference of some peaks in the chiral separation for real samples. The effects of SDS on the separation of PALO stereoisomers are similar at different SC concentrations, i.e., it always causes the prolongation of the migration time and the reversion of the migration order between two enantiomeric pairs, as shown in Figs. S2 and S3 in Appendix A. However, its effect is weaker at a higher SC concentration. The SDS concentration required to obtain a similar separation increases with the SC concentration. 3.2. Effect of SDS on electromigration of PALO stereoisomers In order to explain the effect of SDS on the separation of PALO stereoisomers in MEKC, its effect on the electromigration (mobilities) of PALO stereoisomers in borate buffer of the same concentration and pH without SC micelles was investigated. Only two diastereoisomers, PALO (3aS, 2S) and (3aS, 2R), were used in this set of experiment, each of them represents a pair of enantiomers in achiral media. As can be seen in Fig. 2, in pure borate buffer, PALO stereoisomers are eluted before the EOF marker methanol due to being positively charged and PALO (3aS, 2S) is eluted before (3aS, 2R), suggesting that it has a bigger electrophoretic mobility than that of PALO (3aS, 2R). When only 1 mM of SDS is added into the buffer, the peaks of PALO stereoisomers move to the rear of methanol peak. And the backward movement of the peaks continues along with the increase of SDS concentration. Referring to the reported fact that ionic species form ion-pair with monomeric surfactant of the opposite charge, resulting in the changes of their mobility in MEKC [15], we attribute this phenomenon to the formation of negatively charged complexes with SDS by electrostatic and/or other interactions: PALO+ + nSDS− PALO · nSDS1−n
Fig. 2. Effect of low concentration SDS on electromigration of PALO stereoisomers in borate buffer. BGE component except for SDS is 30 mM sodium tetraborate, pH 9.20. CE conditions are the same as in Fig. 1.
The increase in SDS concentration would enlarge the fraction of complex or the binding number n, causing the increase of the average negative charge carried by PALO stereoisomers and consequently their negative mobility and migration time. The addition of SDS not only prolongs the migration time of PALO stereoisomers, but also reverses their migration order. PALO (3aS, 2S) is eluted after PALO (3aS, 2R) when SDS concentration is above 2 mM (Fig. 2). It suggests that PALO (3aS, 2S) with a bigger positive mobility (eluted first) in pure borate buffer, compared to PALO (3aS, 2R), also has a bigger negative mobility (eluted late) under the existence of SDS, probably because it has stronger interactions with SDS. It is also found in Fig. 2 that the peaks of PALO stereoisomers become very broad and asymmetric when 1 mM to 2 mM of SDS is added. In order to explain this phenomenon, we suppose that PALO stereoisomers may form complexes with different binding numbers and thus different mobilities. If the SDS concentration is very low, the complexes of different stoichiometric ratios can not exchange quickly because of the large distance between PALO stereoisomers and SDS ions. This slow mass transfer may cause the band broadening in CE. The situation is similar to that reported in early literatures about MEKC, the efficiency of MEKC decreases dramatically when surfactant concentrations are close to the CMC because of slow mass transfer caused by large distance between micelles [16]. When SDS concentration is increased from 3 mM to 7 mM, the peaks of PALO stereoisomers become narrow because the increase in the number of SDS ions reduces their average distance to PALO stereoisomers and thus improves the mass transfer. When SDS concentration exceeds the CMC, there is also ion–micelle interaction (partition into micelles) between PALO stereoisomers and SDS; it may provide bridges for the interaction between PALO stereoisomers and SDS monomers, causing the further improvement of the mass transfer. The reported CMC of SDS in pure water is 8.1 mM [11,12], but inorganic salts in the solution may cause the significant decrease of the CMC [17–19]. The CMC value of SDS in 30 mM borate buffer of pH 9.20, the media of this experiment, measured with dye solubilization method [17] in our laboratory is 4.2 mM. Since there are also free surfactant ions of certain concentration (CMC in theory) in micelle solution, the separation media of MEKC, the effect of SDS on the electromigration of PALO stereoisomers in borate buffer containing SC of 12 mM (approx. the CMC) was also investigated. The reported CMC values of SC range from
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
Fig. 3. Effect of low concentration SC and SDS on electromigration of PALO stereoisomers in borate buffer. BGE component except for SC and SDS is 30 mM sodium tetraborate, pH 9.20. CE conditions are the same as in Fig. 1.
10 to 15 mM or even wider, depending on the solution background as well as the measurement method [19–23]. The value in 30 mM borate buffer of pH 9.20 measured with dye solubilization method [17] in our laboratory is 12.1 mM. As shown in Fig. 3, SC at low concentrations has less effect on the electromigration of PALO stereoisomers, compared to SDS. If only SC of different concentrations is added into borate buffer, the peaks of PALO stereoisomers are always in front of that of methanol, showing their positive charge, although they may also interact with negatively charged SC, particularly when the concentration is near the CMC, causing the decrease of their mobility. The migration order is always PALO
89
(3aS, 2S) before (3aS, 2R), suggesting that the relative size of their mobilities is basically not affected. When 1 mM of SDS is added, the peaks of PALO stereoisomers move to the rear of methanol peak. With the increase of SDS concentration, the migration time of PALO stereoisomers is prolonged and their migration order is reversed (Fig. 3). The situation is the same as when only SDS is added (Fig. 2). Our explanation is also the formation of negatively charged complexes with SDS. Obviously, PALO stereoisomers have stronger interactions with the monomers of SDS than SC, although both of them are anionic surfactants. Our explanation for this phenomenon is that besides the electrostatic interaction between PALO cations and SDS anions, there is also hydrophobic interaction between the hydrocarbon chain of SDS and the hydrophobic part of PALO stereoisomers. The flexible long hydrocarbon chain of SDS may be entangled around the hydrophobic part of PALO stereoisomers, reducing their exposure to aqueous solution, thus making them more stable. Monomeric SC should have the same tendency of binding with PALO stereoisomers through hydrophobic interaction, but the rigid structures of both of them (Appendix A, Fig. S1) hamper the close contact between their hydrophobic parts. This explanation is supported by the UV spectrometry experiment. As shown in Fig. 4, the addition of submicellar concentrations of SC up to 10 mM has little effect on the UV spectra of PALO (3aS, 2S) and (3aS, 2R). However, the relative strength of the two bands in the UV spectra is changed evidently by the addition of submicellar concentrations of SDS. The first band (centered at 242 nm for 3aS, 2S and 241 nm for 3aS, 2R) is increased while the second band (centered at 256 nm for 3aS, 2S and 257 nm for 3aS, 2R) is decreased compared to the UV spectra in pure borate buffer, suggesting the change of chemical environment around the chromophores of PALO stereoisomers, perhaps caused by the surrounding of the hydrocarbon chains of SDS.
Fig. 4. Effect of SC and SDS with concentrations below the CMC on the UV spectra of PALO stereoisomers. The concentrations of PALO stereoisomers (0.05 mM, 17 g mL−1 ) and sodium tetraborate buffer (30 mM, pH 9.20) are constant for every spectrum. The reference solutions contain the same concentrations of surfactant and buffer as the sample.
90
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
Table 1 Quantitative parameters of the developed method for the analysis of the PALO injection spiked with enantiomeric impurities.
Calibration range (g mL−1 ) Regression equation
Slope, a Intercept, b Correlation coefficient, R Limit of detection (LOD, g mL−1 ) Limit of quantification (LOQ, g mL−1 ) Recovery (%) a Repeatability (RSD, %) b Intra-day c Inter-day d a b c d
PALO (3aS, 2R)
PALO (3aR, 2R)
PALO (3aR, 2S)
0.5–5.0 y = ax + b y: peak area (mV s); x: concentration of enantiomeric impurities (g mL−1 ) 1.668 −0.001 0.9992 0.08 0.28 100.2
0.5–5.0
0.5–5.0
1.724 −0.083 0.9989 0.08 0.28 104.3
1.600 0.099 0.9982 0.09 0.31 104.2
8.6 9.2
3.8 8.5
7.4 8.9
Values at an enantiomeric impurity concentration of 1.0 g mL−1 . Values at an enantiomeric impurity concentration of 1.0 g mL−1 . RSD of peak areas for five replications. RSD of peak areas for 15 replicates in 3 days, five replicates each day.
3.3. Explanation to effect of SDS on MEKC separation of PALO stereoisomers
concentration but no SC added (Fig. 2), although the EOF are basically equal.
The effect of SDS on MEKC separation of PALO stereoisomers can be elucidated reasonably based on our investigation about its effect on the electromigration of PALO stereoisomers. According to the basic theory of MEKC, the observed velocity of a solute in MEKC is the weighted average for its different forms, including all the forms distributed in water phase and micelles [9–11]. For the separation without SDS addition, the positively charged PALO stereoisomers in water phase swim downstream to the direction of capillary outlet while the PALO stereoisomers distributed in negatively charged micelles swim upstream to the direction of capillary inlet. As the upstream migration is partially offset by the downstream migration, the observed (average) velocities are relatively high and the migration times are relatively short. If the PALO stereoisomers in water phase form negatively charged complexes with SDS, they also swim upstream, thus the observed velocities decrease and the migration times are prolonged. When the SDS concentration is increased, the average negative charge carried by PALO stereoisomers increases, due to the increase of the fraction of complexes or the binding number. Therefore, their negative mobility increases, leading to the further decrease of observed velocities and finally the prolongation of migration times. Since PALO (3aS, 2S) with a bigger positive mobility (eluted first) than that of PALO (3aS, 2R) in pure borate buffer turns to have a bigger negative mobility (eluted late) when binding with SDS (each of them represents a pair of enantiomers), the migration order between two enantiomeric pairs is reversed. Shown in Fig. 1, the peaks of enantiomeric pair (3aS, 2S), (3aR, 2R) (as a whole) move to the rear of those of another pair (3aS, 2R), (3aR, 2S) gradually when the SDS concentration is increased. Different from in pure borate buffer (Fig. 2), the addition of SDS of even very low concentration into SC micelle in MEKC does not broaden the peaks (Fig. 1), probably because high concentration of SC micelles provide bridges for the interaction between PALO stereoisomers and SDS, thus improving the mass transfer. In our opinion, the weaker effect of SDS on the separation of PALO stereoisomers at a higher SC concentration may be ascribed, on one hand, to the decrease in the fraction of PALO–SDS complexes, caused by the competition effect from the partition of PALO stereoisomers into SC micelles. On the other hand, part of SDS may combine with SC at high concentration, reducing its interaction with PALO stereoisomers. This can be demonstrated by the fact that the migration times of PALO stereoisomers when 12 mM of SC is added (Fig. 3) are shorter than those at the same SDS
3.4. Validation of the method The developed method was used for the analysis of a real sample, PALO injection, in order to check the applicability. Its nominal content of effective ingredient, PALO (3aS, 2S), is 50 g mL−1 . As the enantiomeric impurities, i.e. PALO (3aS, 2R), PALO (3aR, 2R) and PALO (3aR, 2S), were not detected (Fig. 5), the standards of them were spiked into the PALO injection to concentrations ranging from 0.5 to 5.0 g mL−1 . The spiked samples were injected directly. In order to improve the sensitivity and precision, the sample injection time was increased from 1 s to 5 s. Other conditions optimized above remained unchanged. The peak widths of the enantiomeric impurities are basically not changed with the increase of injection time, whereas the peak width of the main ingredient, PALO (3aS, 2S), increases from 1.3 s to 1.7 s (W1/2 ) as the injection time is changed from 1 s to 5 s. However, a baseline resolution between PALO (3aS, 2S) and PALO (3aS, 2R) can still be obtained
Fig. 5. Electropherograms of PALO injection spiked with different concentrations of enantiomeric impurities. BGE composition is 30 mM SC and 1 mM SDS in 30 mM sodium tetraborate of pH 9.20. CE conditions are the same as in Fig. 1, except that the injection time is 5 s.
S.-Q. Hu et al. / J. Chromatogr. A 1342 (2014) 86–91
(Fig. 5). Further increase of injection time influences the separation between PALO (3aS, 2S) and PALO (3aS, 2R), but baseline resolutions among PALO (3aS, 2R), PALO (3aR, 2R) and PALO (3aR, 2S) can be obtained till an injection time of 10 s (data not given). Good precision, recovery, and linear relationship between the peak areas and the impurity concentrations were obtained (Table 1). The limit of detection (LOD, S/N = 3) and limit of quantification (LOQ, S/N = 10) for the enantiomeric impurities are approximately 0.1 g mL−1 and 0.3 g mL−1 , respectively. 4. Conclusions SDS can be an alternative to organic solvents in selectivity tuning, in the separation of PALO stereoisomers by MEKC using SC as chiral selector. Based on the analysis of electropherograms, low concentrations of SDS added into the separation media forms negatively charged complexes with PALO stereoisomers and hence reverses their electromigration direction. Consequently, the separation results of PALO stereoisomers are changed dramatically. The peaks originally coalesced are resolved and the migration order of stereoisomers is changed regularly with the concentration of SDS. Good separations for all the four stereoisomers can be achieved at appropriate SDS concentration. The optimized method has good precision, accuracy, and linear relationship in the analysis of enantiomeric impurities in the PALO injection, although the sensitivity is slightly low. Acknowledgment The financial support (Grant no. 12A150017) as Key Project of Science and Technology Research by Department of Education of Henan Province, PR China, is gratefully acknowledged.
91
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.chroma.2014.03.042. References [1] R. de Wit, M. Aapro, P. Blower, Cancer Chemother. Pharmacol. 56 (2005) 231. [2] M. Aapro, P. Blower, Cancer 104 (2005) 1. [3] J.T. Hickok, J.A. Roscoe, G.R. Morrow, C.W. Bole, H. Zhao, K.L. Hoelzer, S.R. Dakhil, T. Moore, T.R. Fitch, Lancet Oncol. 6 (2005) 765. [4] P. Eisenberg, J. Figueroa-Vadillo, R. Zamora, V. Charu, J. Hajdenberg, A. Cartmell, A. Macciocchi, S. Grunberg, Cancer 98 (2003) 2473. [5] M. Wang, X.-P. Ding, H.-L. Chen, X.-G. Chen, Anal. Sci. 25 (2009) 1217. [6] M.V. Murthy, K. Jyothirmayi, K. Srinivas, K. Mukkanti, R. Kumar, G. Samanta, Am. J. Anal. Chem. 2 (2011) 437. [7] X.-R. Yu, M. Song, T.-J. Hang, Chin. J. New Drugs 17 (2008) 870. [8] P. Radhakrishnanand, D.V. Subba Rao, V. Himabindu, Chromatographia 69 (2009) 369. [9] S. Terabe, K. Otsuka, K. Ichikawa, A. Tsuchiya, T. Ando, Anal. Chem. 56 (1984) 111. [10] S. Terabe, K. Otsuka, T. Ando, Anal. Chem. 57 (1985) 834. [11] K.R. Nielsen, J.P. Foley, in: P. Camilleri (Ed.), Capillary Electrophesis—Theory and Practice, CRC Press, Boca Raton, FL, 1998, p. 135. [12] S.E. Deeb, M.A. Iriban, R. Gust, Electrophoresis 32 (2011) 166. [13] K. Tian, H.-L. Chen, J.-H. Tang, X.-G. Chen, Z.-D. Hu, J. Chromatogr. A 1132 (2006) 333. [14] S.-Q. Hu, H.-B. Yang, H.-J. Shi, Y.-H. Zhang, Z. Yang, Electrophoresis 34 (2013) 3086. [15] T. Kaneta, S. Tanaka, M. Taga, H. Yoshida, Anal. Chem. 64 (1992) 798. [16] M.J. Sepaniak, R.O. Cole, Anal. Chem. 59 (1987) 472. [17] H. Nakamura, A. Sano, K. Matsuura, Anal. Sci. 14 (1998) 379. [18] G.-X. He, L. Zhang, G. Li, Chin. Chem. Res. 22 (2011) 68. [19] E. Fuguet, C. Ràfols, M. Rosés, E. Bosch, Anal. Chim. Acta 548 (2005) 95. [20] C.M. Hebling, L.E. Thompson, K.W. Eckenroad, G.A. Manley, R.A. Fry, K.T. Mueller, T.G. Strein, D. Rovnyak, Langmuir 24 (2008) 13866. [21] S. Reis, C.G. Moutinho, C. Matos, B. de Castro, P. Gameiro, J.L.F.C. Lima, Anal. Biochem. 334 (2004) 117. [22] P. Garidel, A. Hildebrand, R. Neubert, A. Blume, Langmuir 16 (2000) 5267. [23] A. Hildebrand, P. Garidel, R. Neubert, A. Blume, Langmuir 20 (2003) 320.