- Email: [email protected]
A new on-line concentration method of cationic molecules in capillary electrophoresis by a hyphenated micelle to solvent stacking coupling with large amount sample electrokinetic stacking injection
Journal of Chromatography A, 1265 (2012) 176–180
Contents lists available at SciVerse ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
A new on-line concentration method of cationic molecules in capillary electrophoresis by a hyphenated micelle to solvent stacking coupling with large amount sample electrokinetic stacking injection Ya-lei Dong a , Hui-ge Zhang a , Zia Ur Rahman a , Hai-juan Zhang a , Xiao-jiao Chen a , Jing Hu a , Xing-guo Chen a,b,c,∗ a
National Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, China Department of Chemistry, Lanzhou University, Lanzhou 730000, China c Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, Lanzhou 730000, China b
a r t i c l e
i n f o
Article history: Received 26 June 2012 Received in revised form 17 September 2012 Accepted 18 September 2012 Available online 28 September 2012 Keywords: Capillary electrophoresis Modified micelle to solvent stacking Micelle collapse Large amount sample electrokinetic stacking injection Enrichment approach
a b s t r a c t In this paper, we established a new on-line method using micelle to solvent stacking (MSS) technique combining with large amount sample electrokinetic stacking injection (LASEKSI) for the analysis of cationic molecules. In this MSS–LASEKSI, by modulating the integral EOF across the capillary, a equilibrium state was formed and can be maintained for a long time, leading to the continuous stacking of the analytes on the basis of MSS. Thereby, an extremely large amount sample was permitted to be injected into the capillary and then an improved enrichment fold can be achieved comparing with the each case. The variables affecting the performance of MSS–LASEKSI were investigated and discussed. Under the optimized conditions, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights upon normal CZE method (injected at 0.5 psi for 3 s) and number of plates of 2.9 × 106 and 6.5 × 105 were attained for berberine and theophylline, respectively. The developed method described here may provide prospects for exploiting a new concentration technique to achieve higher enrichment factor. © 2012 Elsevier B.V. All rights reserved.
1. Introduction In capillary electrophoresis (CE), a capillary with smaller inside diameter is preferred for rapid and high-resolution separation by applying higher voltages [1]. However, due to the restriction of the short light path and the small injection amount, CE suffers from relatively poor concentration sensitivity for absorbance-based measurements. Considerable interests in improving detection sensitivity of CE have led to the development of numerous techniques to extend light path [2–4] or using more sensitive detectors. Interestingly, ion can be enriched when nanochannels were used according to the previous reports [5,6]. However, in order to avoid the modification of commercial CE instrument or capillary, on-line concentration techniques were proved to be convenient and efficient ways. Over the last 20 years, an array of strategies [7–13] for on-line concentration has been established.
∗ Corresponding author at: National Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, China. Tel.: +86 931 891 2763; fax: +86 931 891 2582. E-mail address: [email protected] (X.-g. Chen). 0021-9673/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2012.09.082
Among them, MSS is a novel concentration strategy relying on the change in direction of the analytes’ effective electrophoretic mobility. This on-line sample concentration approach afforded an order of magnitude improvement in sensitivity according to the previous reports [12,14]. Subsequently, Liu et al. [15] and Zhu et al. [16] demonstrated that the change in electrophoretic mobility can also be caused by micelle collapse. However, approximately 100- to 1000-fold enhancement in sensitivity was obtained and the aforementioned strategies were still restricted by the limited sample amount that can be introduced into the capillary and then confined the further improvement of detection sensitivity. Regarding to the intrinsic weaknesses of the method, a combined concentration approach coupled with large sample injection technique may be an efficient solution to afford higher focusing fold. Therefore, design of hyphenated concentration protocol is available to enable the enhancement of effective sensitivity in CE. Currently, several on-line approaches have been established to achieve large amount sample injection for CE. For instance, Gong et al. [17] proposed a strategy for unlimited volume electrokinetic stacking injection (EKSI), where notly improved the actual injected sample plug lengths of anionic analytes. More recently, Zhang et al. [18] reported a new approach with extremely large volume electrokinetic stacking of alkaloid. The technique
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
modulated the movements of bulk flow inside capillary through a strong acidic buffer, and the stacking frontier of the micelle zone stayed constantly in a certain point along the capillary for a period of time, resulting in that very large amount of cationic analytes can be injected into the capillary and focused. Motivated from the common ground of MSS and Zhang’s EKSI technique, such as the micellar solution being used and the boundary existing in two plugs through the capillary, we assumed that the EKSI may provide a new way to afford large injection amount for MSS. Therefore, development of a modified MSS method coupling with large sample injection amount would be a synergistically effective strategy to maximize enhancement fold. However, to the best of our knowledge, there are no reports regarding the combination of MSS and EKSI up to now. In this paper, we established a new on-line hyphenated MSS method coupling with large amount sample electrokinetic stacking injection (LASEKSI), and then demonstrated that the MSS–LASEKSI approach afforded more enrichment fold than the single method. The principle of focusing, factors for improvement of enhancement fold, and analytical performance were discussed. Alkaloid compounds, berberine and theophylline, were employed as model compounds for evaluating the performance. Berberine is a major isoquinoline alkaloid to counteract toxicity and exhibit antibacterial and anti-inflammatory activity. Theophylline, a drug with various pharmacological functions, is used as a cardiac and respiratory stimulant. Urine was employed as real sample to validate the adaptability of the present method. 2. Experimental 2.1. Apparatus and materials All experiments were performed on a Beckman P/ACE MDQ system (Fullerton, CA) equipped with a diode array UV detector (190–600 nm). Data acquisition and instrument control were carried out using 32 Karat software (Version 7.0). An uncoated fused-silica capillary with dimensions of 60.2 cm length (effective length 50.0 cm) × 50 m I.D. was purchased from Sino Sumtech Optical Conductive Fiber (Handan, China), and was thermostated at 20 ◦ C. A PHS-3C acidity meter (Shanghai REX Instrument Factory, Shanghai, China) was used for the pH measurement. Berberine and theophylline were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Sodium dihydrogen phosphate, phosphoric acid, sodium dodecyl sulfate (SDS), sodium tetraborate, methanol and acetonitrile were products of Tianjin Chemical Reagent Factory (Tianjin, China). All reagents were of analytical grade and were used without further purification. Quartic distilled water was used throughout. 2.2. Solutions Standard stock solutions containing 500 g/mL of berberine and theophylline were prepared in methanol/water (10:90, v/v), respectively, and stored at 4 ◦ C. Stock solutions of 400 mM SDS, 500 mM NaH2 PO4 , 100 mM borate and 5 M H3 PO4 were prepared in quartic distilled water. Separation buffer was 20 mM borate (pH 4.0), and prepared by diluting the stock solutions with quartic distilled water. The micellar sample matrix was composed of 20 mM borate and 20 mM SDS (pH 4.0). An acidic buffer solution used to modulate the integral EOF was 20 mM NaH2 PO4 with pH 3.0. Co-solvent buffer (10 mM NaH2 PO4 solution in methanol/acetonitrile/water (60:30:10, v/v/v)) was freshly prepared by diluting stock solution of NaH2 PO4 with pure methanol and acetonitrile. All buffers were adjusted accurately to the desired
177
pH value with 5.0 and 0.1 M H3 PO4 or 1.0 and 0.1 M NaOH, and filtered through a 0.45 m pore size cellulose acetate membrane followed by degassing prior to use. 2.3. Capillary electrophoresis conditions Prior to its first use, the new capillary was conditioned by rinsing sequentially with methanol for 5 min, water for 5 min, 0.5 M HCl for 20 min, water for 5 min, 0.5 M NaOH for 20 min, water for 5 min, and running buffer for 20 min at 20 psi, finally, equilibrated at 20 kV with running buffer for 20 min. To assure a good repeatability, the capillary was rinsed at 20 psi sequentially with water (5 min), 0.5 M HCl (5 min) and water (5 min), at the beginning of each experimental session. Between two runs, a rinse-cycle of water for 2 min, 0.5 M HCl for 3 min, water for 2 min, 0.5 M NaOH for 3 min and running buffer for 5 min at 20 psi was employed. Then the sample solution was introduced into the capillary from the outlet end of capillary for 5 min at 5 psi to avoid to generating bubble inside the capillary, followed by injection of co-solvent buffer at 0.5 psi for 40 s from the inlet end. Sample stacking was performed by applying 10 kV voltage in the normal polarity through the acidic buffer solution (pH 3.0) vial and the sample vial for a period of time. Then a constant voltage of 27.5 kV was applied for the separation between the running buffers. All experiments were carried out with anode at the inlet and cathode at the outlet. 2.4. Sample preparation Urine sample was collected from a female volunteer. 1.0 mL of the human urine sample, spiked with the standard sample at different concentration levels, was centrifuged for 5 min (10,000 rpm). The supernatant was transferred to a glass separator funnel and extracted for three times using the extract (2 mL ethyl acetate for every time), then allowed to stand for phase separation, and the supernatant was combined. The organic extract was evaporated at 37 ◦ C under nitrogen stream until dryness and the leftover was dissolved by 1.0 mL sample matrix buffer solution. 3. Results and discussion 3.1. Schematic representation of MSS–LASEKSI The schematic representation of MSS–LASEKSI was illustrated in Fig. 1. Initially, the whole capillary was filled with the sample matrix (pH 4.0). The co-solvent buffer was injected by applying a pressure of 0.5 psi for a period of time from the inlet end of the capillary. Then a positive voltage of 10 kV was applied across the capillary. As the inner wall of capillary carried negative charges because of the silanol dissociation, electroosmotic flow (EOF) toward the cathode brought the acidic buffer solution (pH 3.0) into the capillary from the inlet end. Along with the injection of strong acidic buffer solution, as shown in Fig. 1a, the dissociation of silanol groups immersed in this section was suppressed, leading to attenuation of the integral EOF across the capillary. When the plug length of the acidic buffer solution (pH 3.0) reached a certain extent (LAB shown in Fig. 1b), the velocity of EOF sped down, leading to the same magnitude of reversely migrating velocity to SDS micelles. Therefore, the apparent migration velocity of EOF fell to nearly zero, and the “balance” was established and could be sustained for a quite long time. This state also can be validated by the dramatic decrease and gradual stabilization of the current (sections a–b in Fig. 2). The negative micelles transported the bound cationic analytes from the sample region to the co-solvent plug, followed by collapsing and releasing as a result of that the CMC of SDS in the organic reagent was higher than that in the water [19–21]. After that, the effective electrophoretic mobility of the released cationic
178
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
from 5 A to approximately 20 A immediately after applying separation voltage, and sustained for some time, due to the existence of the co-solvent and sample matrix in the capillary (sections b–c in Fig. 2). Finally, the current increased to a level when the sample matrix and co-solvent plug were completely pumped out of the capillary by the EOF and the whole capillary was filled with separation buffer (sections c–d in Fig. 2). 3.2. Optimization of the MSS–LASEKSI
Fig. 1. Schematic illustration of sample focusing in MSS–LASEKSI. (a) EOF brought the acidic buffer solution (pH 3.0) into the capillary from the inlet end, leading to attenuation of the integral EOF, (b) the equilibrium state was established because of the comparative velocity of micelles and EOF, and could be sustained for a quite long time, causing the gathering of the sample molecules at the boundary continuously so that a large amount of sample could be electrokineticly injected into the capillary, (c) the separation buffer was adopted and EOF was resumed gradually, leading to the start of the separation procedure. The blue area with symbols of H+ represents the acidic buffer solution (pH 3.0). The white zone is the plug of co-solvent buffer solution. The gray area represents the sample matrix and the separation buffer. LAB is the length of acidic buffer solution. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
analytes reversed from anode to cathode. This caused the gathering of the sample molecules at the boundary between the sample zone and co-solvent plug. With the micelles collapsed continuously, the analytes were focused at the boundary so that a large amount of sample could be electrokineticly injected into the capillary until the equilibrium state was destroyed. After sample injection, the separation buffer was adopted and a separation voltage was applied, leading to the disturbance of the equilibrium state, which can be observed from Fig. 1c. The silanol was dissociated again and EOF was resumed gradually. A separation procedure started and the concentrated analytes were separated by CZE mode. This also can be confirmed from the current graph in Fig. 2. The current increased
Fig. 2. Change of the current in the whole concentration and separation process. “a” is the start of sample injection process under 10 kV; “b” is the start point of the separation after applying 27.5 kV; “c” corresponds to that the matrix and co-solvent were pumped out of the capillary; “d” is the steady current at the time of separation process.
In the present concentration strategy, the stable state was formed through the modulated bulk flow velocity. The ionic strength, which affected the mobility of SDS micelles and EOF, was an important factor influencing the enrichment efficiency. The background electrolyte (BGE) was chosen as borate solution at pH 4.0, with concentration ranging from 10 to 40 mM. To avoid the mismatch of conductivity, the separation buffer was a borate solution keeping the same constituent and concentration as the according sample matrix, lacking of SDS micelles. Fig. 3 illustrated that the concentration of borate affecting the migration time and the focusing efficiency remarkably. The “block” in the electropherograms is the co-solvent peak and the negative dip represents the boundary between the co-solvent zone and sample matrix plug [16]. The maximum peak height was achieved when 20 mM of borate was used. Further increasing the concentration of borate did not contribute to the enrichment efficiency. It was because that along with the stepwise increasing in ionic strength, the EOF in the sample matrix plug was suppressed due to the decreased thickness of the electric double layer, as well as the electrophoretic mobility of SDS, resulting from the increased buffer viscosity. Therefore, the migration velocity of SDS micelle was reduced, leading to the decreased injection amount during the same stacking time. The length of the acidic buffer (pH 3.0) zone in the capillary became short and the residual length for separation enhanced, leading to the delayed migration time. And then it will restrict the focusing efficiency with further improving sample matrix concentration. It should be noted that borate buffer (pH 4.0) was chosen because of its superiority
Fig. 3. Effect of the concentration of the background electrolytes (BGE). The sample matrix was consisted of (a) 10 mM, (b) 20 mM, (c) 30 mM, (d) 40 mM of borate (pH 4.0) with 20 mM SDS; the composition of separation buffer was consistent with the sample matrix with the absence of SDS micelles; co-solvent buffer: 10 mM NaH2 PO4 solution in methanol/acetonitrile/water (60:30:10, v/v/v) introduced at 0.5 psi for 40 s; sample was electrokineticly injected using a voltage 10 kV for 10 min and separation voltage is 27.5 kV; B, berberine in sample matrix with concentration of 1.0 g/mL. Detection wavelength: 210 nm; capillary: 50 m I.D. and 60.2 cm total length (50.0 to detector); capillary temperature: 20 ◦ C.
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
Fig. 4. Effect of the SDS concentration on enrichment efficiency. Sample matrix: 20 mM borate with different concentration of SDS (pH 4.0); separation buffer: 20 mM borate (pH 4.0). Other conditions are the same as those in Fig. 3.
upon corresponding H3 BO3 solution or NaH2 PO4 buffer. The buffer solutions consisted of H3 BO3 and NaH2 PO4 (pH 4.0) were tested, which produce proportionable current by CZE mode comparing with the borate solution (20 mM, pH 4.0), indicating the similar electric conductivity. However, the repeatability of EOF was 1.5% using borate buffer solution while 15.5% using H3 BO3 and NaH2 PO4 by methanol as EOF marker. Therefore, 20 mM borate solution (pH 4.0) and 20 mM borate (pH 4.0) containing SDS were chosen as the optimum separation buffer and sample matrix solution for further experiments. The effect of concentration of SDS on the stacking efficiency was investigated in a range of 10–40 mM. As can be observed from Fig. 4, the peak height enhanced as the SDS concentration increasing from 10 to 20 mM. However, further increasing of SDS concentration yielded unobvious improvement of the stacking efficiency, while the resolution between the focused analytes and the co-solvent plug was diminished. This was probably due to the fact that it’s difficult to collapse large concentration micelle once the concentration of co-solvent is not large enough in buffer. In order to collapse the micelle thoroughly, the micelle needs to meet with larger concentration of co-solvent which is in the middle of the co-solvent plug due to its diffusion in the buffer. Accordingly, 20 mM was chosen as an optimal surfactant concentration as sample matrix. In this study, micelle collapse was induced by organic reagents. So a co-solvent buffer solution (10 mM NaH2 PO4 , 60% methanol and 30% acetonitrile) was introduced into the capillary to cause the micelle collapse. Methanol was a preferential reagent according to the previous report [16]. Here, acetonitrile was added to improve the resolution of the focused analytes. The co-solvent zone was also added with 10 mM NaH2 PO4 to match the electric conductivity. Therefore, the length of the introduced organic zone will affect the disintegration capacity of micelle, and then further influence the focusing efficiency. The co-solvent solution was introduced into the capillary by hydrodynamic injection using a pressure of 0.5 psi for a different period. The dependence of the enrichment efficiency on co-solvent buffer amount was investigated by changing the injection time from 20 to 60 s, corresponding to the co-solvent buffer plug length ranging from 0.99 to 2.96 cm. As shown in Fig. 5, the peak height enhanced obviously in the range of 0.99–1.97 cm and reduced with further increased length of co-solvent buffer plug. The co-solvent zone was the inducing factor of micelles collapse, supporting that sufficient amount of organic reagent implied greater capacity of collapsing and releasing. But a partial diffusion of the focused analytes was markedly when the injection length of
179
Fig. 5. Dependence of the co-solvent buffer plug length on enrichment efficiency. The co-solvent buffer was introduced into the capillary at 0.5 psi for different time with length of (a) 0.99 cm, (b) 1.48 cm, (c) 1.97 cm, (d) 2.47 cm, (e) 2.96 cm. Other conditions are the same as those in Fig. 3.
co-solvent buffer solution exceeded 1.97 cm. Thus, 1.97 cm was the best choice for the maximum enrichment efficiency. In the present enrichment method, the cationic analytes carried in negatively charged SDS micelles migrated toward the co-solvent buffer zone continuously from the sample region. When the equilibrium state was established, a large injection amount of sample can be achieved through a long injection time. Therefore, the effect of sampling time on concentration efficiency was studied by electrokinetic injection at a voltage of 10 kV for varying time. It was expected that the focusing efficiency was proportional to the injection time. However, the enrichment fold deteriorated when the injection time was longer than 50 min. It can be deduced that the steady state was destroyed and the boundary between the cosolvent plug and sample matrix was disturbed. Thus, 50 min was chosen as the optimum injection time. The equilibrium state was maintained for less time comparing with the 60 min reported in literature [18], which maybe due to that abundant analytes were enriched in the capillary and then changed the ionic strength of buffer, which disturb the stable state. 3.3. Performance of the MSS–LASEKSI The focusing efficiency and separating power of the method was assessed by comparing its performance with that of normal CZE using berberine and theophylline as model analytes, as illustrated in Fig. 6. The enrichment fold of the proposed method was evaluated by comparing a 50 min stacking with the normal CZE mode injected at 0.5 psi for 3 s. As large amount of sample was injected with MSS–LASEKSI mode, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights for berberine and theophylline were obtained, respectively, higher than the individual mode reported by MSS [12,15,16] and EKSI [18]. It should be noted that berberine exhibited higher focusing efficiency than theophylline, which should be attributed to the stronger affinity of the berberine to SDS micelle than that of theophylline. This phenomenon indicated that the proposed MSS–LASEKSI method may be more effective for the hydrophobic analytes. And, more remarkable, the elution order of berberine and theophylline showed opposite result comparing the MSS–LASEKSI enrichment technique and the normal CZE mode. This was due to the decreased mass-to-charge ratio of theophylline, as a result of the suppressed ionization when basic separation buffer was used in CEZ mode. The linearity, detection limits and the repeatability of the proposed method for detection
180
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180 Table 2 Recovery study of the two test alkaloids at different spiked level in urine sample. Added (g/mL)
Berberine 101
0.3
Recovery (%) RSD (%)a Migration time Peak height Recovery (%) RSD (%) Migration time Peak height
98
0.5
100
1.0
Recovery (%) RSD (%) Migration time Peak height
a
2.4 4.5
3.1 6.0
1.2 7.6
Theophylline 95 3.7 1.0 99 0.9 3.4 112 1.0 5.7
The relative standard derivation was calculated from three recovery data.
coexisting compounds in urine sample did not influence the CE determination. Fig. 6. The typical electropherograms of a standard mixture solution. (a) The normal CZE mode, injected at a pressure of 0.5 psi for 3 s, using 20 mM of borate as separation buffer without adjustion of pH. The concentration of berberine is 250.0 g/mL and theophylline is 100.0 g/mL. (b) The present MSS–LASEKSI mode, stacking for 10 min under 10 kV; concentration of berberine and theophylline: 1.0 g/mL. (c) The current MSS–LASEKSI method, blank sample. Other conditions are the same as those in Fig. 3. Peak identification: 1, theophylline; 2, berberine. Table 1 Performance of the present MSS–LASEKSI method.
Linear range (ng/mL) Regression equation Correlation coefficient Limit of detection (S/N = 3) (ng/mL) RSD of migration time (n = 3) (%) RSD of peak height (n = 3) (%) a
Berberine
Theophylline
5.0–2000.0 y = 13,147.9 + 38.2xa 0.997 0.4 3.9 8.4
2.0–2000.0 y = 2841.5 + 41.84x 0.999 0.3 0.7 1.6
4. Conclusion In the present work, we established a new on-line enrichment method based on a modified micelle to solvent stacking (MSS) technique coupled with large amount sample electrokinetic stacking injection (LASEKSI). Under the optimum conditions, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights and a high number of plates of 2.9 × 106 and 6.5 × 105 were obtained for the two model analytes, berberine and theophylline, respectively, higher than the single method. The developed MSS–LASEKSI method described here may provide prospects for exploiting novel combined technique to achieve higher concentration fold, therefore advance the CE field. Acknowledgments
y, peak height; x, concentration (ng/mL).
of berberine and theophylline were carried out under the optimum conditions and results were summarized in Table 1. The limit of detection (LOD) of berberine and theophylline were 0.4 and 0.3 ng/mL, respectively, and the calibration curves were linear up to 2000 ng/mL for these compounds with 10 min of sample stacking. In addition, the method produced narrow peaks for berberine and theophylline, which was observed with a high number of plates of 2.9 × 106 and 6.5 × 105 /m, respectively. The RSDs of the migration times and peak height were 3.9 and 8.4%, and 0.7 and 1.6% for berberine and theophylline respectively, indicating good precision. 3.4. Applications In order to evaluate the feasibility of the proposed on-line concentration method for the analysis of berberine and theophylline, urine sample was used as test sample in the study. Under the optimum conditions, recovery experiments were performed by standard addition method. The results were shown in Table 2, indicating that the recovery ranged from 95 to 112%. And the repeatability of the method used in real sample analysis is also satisfactory. The RSD of the migration time ranged from 0.9 to 3.7%, and the peak height ranged from 1.0 to 7.6%. These results suggest that the method proposed here is useful for the detection of berberine and theophylline in urine sample, and other
This work was kindly supported by the National Natural Science Foundation of China (No. 21075056) and the Fundamental Research Funds for the Central Universities (Nos. lzujbky-2012-214 and lzujbky-2011-20). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
J.W. Jorgenson, K.D.A. Lukacs, Anal. Chem. 53 (1981) 1298. S.E. Moring, R.T. Reel, R.E.J. van Soest, Anal. Chem. 65 (1993) 3454. Y. Xue, E.S. Yeung, Anal. Chem. 66 (1994) 3575. Z. Xu, K. Nakamura, A.R. Timerbaev, T. Hirokawa, Anal. Chem. 83 (2011) 398. Q. Pu, J. Yun, H. Temkin, S. Liu, Nano Lett. 4 (2004) 1099. S. Liu, Q. Pu, L. Gao, C. Korzeniewski, C. Matzke, Nano Lett. 5 (2005) 1389. M.R. Mohamadi, N. Kaji, M. Tokeshi, Y. Baba, Anal. Chem. 79 (2007) 3667. P. Britz-McKibbin, K. Otsuka, S. Terabe, Anal. Chem. 74 (2002) 3736. C.X. Cao, Y.Z. He, M. Li, Y.T. Qian, M.F. Gao, L.H. Ge, S.L. Zhou, L. Yang, Q.S. Qu, Anal. Chem. 74 (2002) 4167. J.P. Quirino, S. Terabe, Science 282 (1998) 465. J.P. Quirino, P.R. Haddad, Anal. Chem. 80 (2008) 6824. J.P. Quirino, J. Chromatogr. A 1216 (2009) 294. M.C. Breadmore, J.P. Quirino, Anal. Chem. 80 (2008) 6373. A.M. Guidote Jr., J.P. Quirino, J. Chromatogr. A 1217 (2010) 6290. L. Liu, X. Deng, X. Chen, J. Chromatogr. A 1217 (2010) 175. H. Zhu, C. Ren, S. Hu, X. Zhou, H. Chen, X. Chen, J. Chromatogr. A 1218 (2011) 733. M. Gong, K.R. Wehmeyer, P.A. Limbach, W.R. Heineman, Anal. Chem. 78 (2006) 6035. H. Zhang, J. Zhu, S. Qi, N. Yan, X. Chen, Anal. Chem. 81 (2009) 8886. W. Lu, G.K. Poon, P.L. Carmichael, R.B. Cole, Anal. Chem. 68 (1996) 668. R. Palepu, H. Gharibi, D. Bloor, E. Wyn-Jones, Langmuir 9 (1993) 110. Y. Walbroehl, J.W. Jorgenson, Anal. Chem. 58 (1986) 479.
Contents lists available at SciVerse ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
A new on-line concentration method of cationic molecules in capillary electrophoresis by a hyphenated micelle to solvent stacking coupling with large amount sample electrokinetic stacking injection Ya-lei Dong a , Hui-ge Zhang a , Zia Ur Rahman a , Hai-juan Zhang a , Xiao-jiao Chen a , Jing Hu a , Xing-guo Chen a,b,c,∗ a
National Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, China Department of Chemistry, Lanzhou University, Lanzhou 730000, China c Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, Lanzhou 730000, China b
a r t i c l e
i n f o
Article history: Received 26 June 2012 Received in revised form 17 September 2012 Accepted 18 September 2012 Available online 28 September 2012 Keywords: Capillary electrophoresis Modified micelle to solvent stacking Micelle collapse Large amount sample electrokinetic stacking injection Enrichment approach
a b s t r a c t In this paper, we established a new on-line method using micelle to solvent stacking (MSS) technique combining with large amount sample electrokinetic stacking injection (LASEKSI) for the analysis of cationic molecules. In this MSS–LASEKSI, by modulating the integral EOF across the capillary, a equilibrium state was formed and can be maintained for a long time, leading to the continuous stacking of the analytes on the basis of MSS. Thereby, an extremely large amount sample was permitted to be injected into the capillary and then an improved enrichment fold can be achieved comparing with the each case. The variables affecting the performance of MSS–LASEKSI were investigated and discussed. Under the optimized conditions, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights upon normal CZE method (injected at 0.5 psi for 3 s) and number of plates of 2.9 × 106 and 6.5 × 105 were attained for berberine and theophylline, respectively. The developed method described here may provide prospects for exploiting a new concentration technique to achieve higher enrichment factor. © 2012 Elsevier B.V. All rights reserved.
1. Introduction In capillary electrophoresis (CE), a capillary with smaller inside diameter is preferred for rapid and high-resolution separation by applying higher voltages [1]. However, due to the restriction of the short light path and the small injection amount, CE suffers from relatively poor concentration sensitivity for absorbance-based measurements. Considerable interests in improving detection sensitivity of CE have led to the development of numerous techniques to extend light path [2–4] or using more sensitive detectors. Interestingly, ion can be enriched when nanochannels were used according to the previous reports [5,6]. However, in order to avoid the modification of commercial CE instrument or capillary, on-line concentration techniques were proved to be convenient and efficient ways. Over the last 20 years, an array of strategies [7–13] for on-line concentration has been established.
∗ Corresponding author at: National Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, China. Tel.: +86 931 891 2763; fax: +86 931 891 2582. E-mail address: [email protected] (X.-g. Chen). 0021-9673/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.chroma.2012.09.082
Among them, MSS is a novel concentration strategy relying on the change in direction of the analytes’ effective electrophoretic mobility. This on-line sample concentration approach afforded an order of magnitude improvement in sensitivity according to the previous reports [12,14]. Subsequently, Liu et al. [15] and Zhu et al. [16] demonstrated that the change in electrophoretic mobility can also be caused by micelle collapse. However, approximately 100- to 1000-fold enhancement in sensitivity was obtained and the aforementioned strategies were still restricted by the limited sample amount that can be introduced into the capillary and then confined the further improvement of detection sensitivity. Regarding to the intrinsic weaknesses of the method, a combined concentration approach coupled with large sample injection technique may be an efficient solution to afford higher focusing fold. Therefore, design of hyphenated concentration protocol is available to enable the enhancement of effective sensitivity in CE. Currently, several on-line approaches have been established to achieve large amount sample injection for CE. For instance, Gong et al. [17] proposed a strategy for unlimited volume electrokinetic stacking injection (EKSI), where notly improved the actual injected sample plug lengths of anionic analytes. More recently, Zhang et al. [18] reported a new approach with extremely large volume electrokinetic stacking of alkaloid. The technique
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
modulated the movements of bulk flow inside capillary through a strong acidic buffer, and the stacking frontier of the micelle zone stayed constantly in a certain point along the capillary for a period of time, resulting in that very large amount of cationic analytes can be injected into the capillary and focused. Motivated from the common ground of MSS and Zhang’s EKSI technique, such as the micellar solution being used and the boundary existing in two plugs through the capillary, we assumed that the EKSI may provide a new way to afford large injection amount for MSS. Therefore, development of a modified MSS method coupling with large sample injection amount would be a synergistically effective strategy to maximize enhancement fold. However, to the best of our knowledge, there are no reports regarding the combination of MSS and EKSI up to now. In this paper, we established a new on-line hyphenated MSS method coupling with large amount sample electrokinetic stacking injection (LASEKSI), and then demonstrated that the MSS–LASEKSI approach afforded more enrichment fold than the single method. The principle of focusing, factors for improvement of enhancement fold, and analytical performance were discussed. Alkaloid compounds, berberine and theophylline, were employed as model compounds for evaluating the performance. Berberine is a major isoquinoline alkaloid to counteract toxicity and exhibit antibacterial and anti-inflammatory activity. Theophylline, a drug with various pharmacological functions, is used as a cardiac and respiratory stimulant. Urine was employed as real sample to validate the adaptability of the present method. 2. Experimental 2.1. Apparatus and materials All experiments were performed on a Beckman P/ACE MDQ system (Fullerton, CA) equipped with a diode array UV detector (190–600 nm). Data acquisition and instrument control were carried out using 32 Karat software (Version 7.0). An uncoated fused-silica capillary with dimensions of 60.2 cm length (effective length 50.0 cm) × 50 m I.D. was purchased from Sino Sumtech Optical Conductive Fiber (Handan, China), and was thermostated at 20 ◦ C. A PHS-3C acidity meter (Shanghai REX Instrument Factory, Shanghai, China) was used for the pH measurement. Berberine and theophylline were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Sodium dihydrogen phosphate, phosphoric acid, sodium dodecyl sulfate (SDS), sodium tetraborate, methanol and acetonitrile were products of Tianjin Chemical Reagent Factory (Tianjin, China). All reagents were of analytical grade and were used without further purification. Quartic distilled water was used throughout. 2.2. Solutions Standard stock solutions containing 500 g/mL of berberine and theophylline were prepared in methanol/water (10:90, v/v), respectively, and stored at 4 ◦ C. Stock solutions of 400 mM SDS, 500 mM NaH2 PO4 , 100 mM borate and 5 M H3 PO4 were prepared in quartic distilled water. Separation buffer was 20 mM borate (pH 4.0), and prepared by diluting the stock solutions with quartic distilled water. The micellar sample matrix was composed of 20 mM borate and 20 mM SDS (pH 4.0). An acidic buffer solution used to modulate the integral EOF was 20 mM NaH2 PO4 with pH 3.0. Co-solvent buffer (10 mM NaH2 PO4 solution in methanol/acetonitrile/water (60:30:10, v/v/v)) was freshly prepared by diluting stock solution of NaH2 PO4 with pure methanol and acetonitrile. All buffers were adjusted accurately to the desired
177
pH value with 5.0 and 0.1 M H3 PO4 or 1.0 and 0.1 M NaOH, and filtered through a 0.45 m pore size cellulose acetate membrane followed by degassing prior to use. 2.3. Capillary electrophoresis conditions Prior to its first use, the new capillary was conditioned by rinsing sequentially with methanol for 5 min, water for 5 min, 0.5 M HCl for 20 min, water for 5 min, 0.5 M NaOH for 20 min, water for 5 min, and running buffer for 20 min at 20 psi, finally, equilibrated at 20 kV with running buffer for 20 min. To assure a good repeatability, the capillary was rinsed at 20 psi sequentially with water (5 min), 0.5 M HCl (5 min) and water (5 min), at the beginning of each experimental session. Between two runs, a rinse-cycle of water for 2 min, 0.5 M HCl for 3 min, water for 2 min, 0.5 M NaOH for 3 min and running buffer for 5 min at 20 psi was employed. Then the sample solution was introduced into the capillary from the outlet end of capillary for 5 min at 5 psi to avoid to generating bubble inside the capillary, followed by injection of co-solvent buffer at 0.5 psi for 40 s from the inlet end. Sample stacking was performed by applying 10 kV voltage in the normal polarity through the acidic buffer solution (pH 3.0) vial and the sample vial for a period of time. Then a constant voltage of 27.5 kV was applied for the separation between the running buffers. All experiments were carried out with anode at the inlet and cathode at the outlet. 2.4. Sample preparation Urine sample was collected from a female volunteer. 1.0 mL of the human urine sample, spiked with the standard sample at different concentration levels, was centrifuged for 5 min (10,000 rpm). The supernatant was transferred to a glass separator funnel and extracted for three times using the extract (2 mL ethyl acetate for every time), then allowed to stand for phase separation, and the supernatant was combined. The organic extract was evaporated at 37 ◦ C under nitrogen stream until dryness and the leftover was dissolved by 1.0 mL sample matrix buffer solution. 3. Results and discussion 3.1. Schematic representation of MSS–LASEKSI The schematic representation of MSS–LASEKSI was illustrated in Fig. 1. Initially, the whole capillary was filled with the sample matrix (pH 4.0). The co-solvent buffer was injected by applying a pressure of 0.5 psi for a period of time from the inlet end of the capillary. Then a positive voltage of 10 kV was applied across the capillary. As the inner wall of capillary carried negative charges because of the silanol dissociation, electroosmotic flow (EOF) toward the cathode brought the acidic buffer solution (pH 3.0) into the capillary from the inlet end. Along with the injection of strong acidic buffer solution, as shown in Fig. 1a, the dissociation of silanol groups immersed in this section was suppressed, leading to attenuation of the integral EOF across the capillary. When the plug length of the acidic buffer solution (pH 3.0) reached a certain extent (LAB shown in Fig. 1b), the velocity of EOF sped down, leading to the same magnitude of reversely migrating velocity to SDS micelles. Therefore, the apparent migration velocity of EOF fell to nearly zero, and the “balance” was established and could be sustained for a quite long time. This state also can be validated by the dramatic decrease and gradual stabilization of the current (sections a–b in Fig. 2). The negative micelles transported the bound cationic analytes from the sample region to the co-solvent plug, followed by collapsing and releasing as a result of that the CMC of SDS in the organic reagent was higher than that in the water [19–21]. After that, the effective electrophoretic mobility of the released cationic
178
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
from 5 A to approximately 20 A immediately after applying separation voltage, and sustained for some time, due to the existence of the co-solvent and sample matrix in the capillary (sections b–c in Fig. 2). Finally, the current increased to a level when the sample matrix and co-solvent plug were completely pumped out of the capillary by the EOF and the whole capillary was filled with separation buffer (sections c–d in Fig. 2). 3.2. Optimization of the MSS–LASEKSI
Fig. 1. Schematic illustration of sample focusing in MSS–LASEKSI. (a) EOF brought the acidic buffer solution (pH 3.0) into the capillary from the inlet end, leading to attenuation of the integral EOF, (b) the equilibrium state was established because of the comparative velocity of micelles and EOF, and could be sustained for a quite long time, causing the gathering of the sample molecules at the boundary continuously so that a large amount of sample could be electrokineticly injected into the capillary, (c) the separation buffer was adopted and EOF was resumed gradually, leading to the start of the separation procedure. The blue area with symbols of H+ represents the acidic buffer solution (pH 3.0). The white zone is the plug of co-solvent buffer solution. The gray area represents the sample matrix and the separation buffer. LAB is the length of acidic buffer solution. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
analytes reversed from anode to cathode. This caused the gathering of the sample molecules at the boundary between the sample zone and co-solvent plug. With the micelles collapsed continuously, the analytes were focused at the boundary so that a large amount of sample could be electrokineticly injected into the capillary until the equilibrium state was destroyed. After sample injection, the separation buffer was adopted and a separation voltage was applied, leading to the disturbance of the equilibrium state, which can be observed from Fig. 1c. The silanol was dissociated again and EOF was resumed gradually. A separation procedure started and the concentrated analytes were separated by CZE mode. This also can be confirmed from the current graph in Fig. 2. The current increased
Fig. 2. Change of the current in the whole concentration and separation process. “a” is the start of sample injection process under 10 kV; “b” is the start point of the separation after applying 27.5 kV; “c” corresponds to that the matrix and co-solvent were pumped out of the capillary; “d” is the steady current at the time of separation process.
In the present concentration strategy, the stable state was formed through the modulated bulk flow velocity. The ionic strength, which affected the mobility of SDS micelles and EOF, was an important factor influencing the enrichment efficiency. The background electrolyte (BGE) was chosen as borate solution at pH 4.0, with concentration ranging from 10 to 40 mM. To avoid the mismatch of conductivity, the separation buffer was a borate solution keeping the same constituent and concentration as the according sample matrix, lacking of SDS micelles. Fig. 3 illustrated that the concentration of borate affecting the migration time and the focusing efficiency remarkably. The “block” in the electropherograms is the co-solvent peak and the negative dip represents the boundary between the co-solvent zone and sample matrix plug [16]. The maximum peak height was achieved when 20 mM of borate was used. Further increasing the concentration of borate did not contribute to the enrichment efficiency. It was because that along with the stepwise increasing in ionic strength, the EOF in the sample matrix plug was suppressed due to the decreased thickness of the electric double layer, as well as the electrophoretic mobility of SDS, resulting from the increased buffer viscosity. Therefore, the migration velocity of SDS micelle was reduced, leading to the decreased injection amount during the same stacking time. The length of the acidic buffer (pH 3.0) zone in the capillary became short and the residual length for separation enhanced, leading to the delayed migration time. And then it will restrict the focusing efficiency with further improving sample matrix concentration. It should be noted that borate buffer (pH 4.0) was chosen because of its superiority
Fig. 3. Effect of the concentration of the background electrolytes (BGE). The sample matrix was consisted of (a) 10 mM, (b) 20 mM, (c) 30 mM, (d) 40 mM of borate (pH 4.0) with 20 mM SDS; the composition of separation buffer was consistent with the sample matrix with the absence of SDS micelles; co-solvent buffer: 10 mM NaH2 PO4 solution in methanol/acetonitrile/water (60:30:10, v/v/v) introduced at 0.5 psi for 40 s; sample was electrokineticly injected using a voltage 10 kV for 10 min and separation voltage is 27.5 kV; B, berberine in sample matrix with concentration of 1.0 g/mL. Detection wavelength: 210 nm; capillary: 50 m I.D. and 60.2 cm total length (50.0 to detector); capillary temperature: 20 ◦ C.
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180
Fig. 4. Effect of the SDS concentration on enrichment efficiency. Sample matrix: 20 mM borate with different concentration of SDS (pH 4.0); separation buffer: 20 mM borate (pH 4.0). Other conditions are the same as those in Fig. 3.
upon corresponding H3 BO3 solution or NaH2 PO4 buffer. The buffer solutions consisted of H3 BO3 and NaH2 PO4 (pH 4.0) were tested, which produce proportionable current by CZE mode comparing with the borate solution (20 mM, pH 4.0), indicating the similar electric conductivity. However, the repeatability of EOF was 1.5% using borate buffer solution while 15.5% using H3 BO3 and NaH2 PO4 by methanol as EOF marker. Therefore, 20 mM borate solution (pH 4.0) and 20 mM borate (pH 4.0) containing SDS were chosen as the optimum separation buffer and sample matrix solution for further experiments. The effect of concentration of SDS on the stacking efficiency was investigated in a range of 10–40 mM. As can be observed from Fig. 4, the peak height enhanced as the SDS concentration increasing from 10 to 20 mM. However, further increasing of SDS concentration yielded unobvious improvement of the stacking efficiency, while the resolution between the focused analytes and the co-solvent plug was diminished. This was probably due to the fact that it’s difficult to collapse large concentration micelle once the concentration of co-solvent is not large enough in buffer. In order to collapse the micelle thoroughly, the micelle needs to meet with larger concentration of co-solvent which is in the middle of the co-solvent plug due to its diffusion in the buffer. Accordingly, 20 mM was chosen as an optimal surfactant concentration as sample matrix. In this study, micelle collapse was induced by organic reagents. So a co-solvent buffer solution (10 mM NaH2 PO4 , 60% methanol and 30% acetonitrile) was introduced into the capillary to cause the micelle collapse. Methanol was a preferential reagent according to the previous report [16]. Here, acetonitrile was added to improve the resolution of the focused analytes. The co-solvent zone was also added with 10 mM NaH2 PO4 to match the electric conductivity. Therefore, the length of the introduced organic zone will affect the disintegration capacity of micelle, and then further influence the focusing efficiency. The co-solvent solution was introduced into the capillary by hydrodynamic injection using a pressure of 0.5 psi for a different period. The dependence of the enrichment efficiency on co-solvent buffer amount was investigated by changing the injection time from 20 to 60 s, corresponding to the co-solvent buffer plug length ranging from 0.99 to 2.96 cm. As shown in Fig. 5, the peak height enhanced obviously in the range of 0.99–1.97 cm and reduced with further increased length of co-solvent buffer plug. The co-solvent zone was the inducing factor of micelles collapse, supporting that sufficient amount of organic reagent implied greater capacity of collapsing and releasing. But a partial diffusion of the focused analytes was markedly when the injection length of
179
Fig. 5. Dependence of the co-solvent buffer plug length on enrichment efficiency. The co-solvent buffer was introduced into the capillary at 0.5 psi for different time with length of (a) 0.99 cm, (b) 1.48 cm, (c) 1.97 cm, (d) 2.47 cm, (e) 2.96 cm. Other conditions are the same as those in Fig. 3.
co-solvent buffer solution exceeded 1.97 cm. Thus, 1.97 cm was the best choice for the maximum enrichment efficiency. In the present enrichment method, the cationic analytes carried in negatively charged SDS micelles migrated toward the co-solvent buffer zone continuously from the sample region. When the equilibrium state was established, a large injection amount of sample can be achieved through a long injection time. Therefore, the effect of sampling time on concentration efficiency was studied by electrokinetic injection at a voltage of 10 kV for varying time. It was expected that the focusing efficiency was proportional to the injection time. However, the enrichment fold deteriorated when the injection time was longer than 50 min. It can be deduced that the steady state was destroyed and the boundary between the cosolvent plug and sample matrix was disturbed. Thus, 50 min was chosen as the optimum injection time. The equilibrium state was maintained for less time comparing with the 60 min reported in literature [18], which maybe due to that abundant analytes were enriched in the capillary and then changed the ionic strength of buffer, which disturb the stable state. 3.3. Performance of the MSS–LASEKSI The focusing efficiency and separating power of the method was assessed by comparing its performance with that of normal CZE using berberine and theophylline as model analytes, as illustrated in Fig. 6. The enrichment fold of the proposed method was evaluated by comparing a 50 min stacking with the normal CZE mode injected at 0.5 psi for 3 s. As large amount of sample was injected with MSS–LASEKSI mode, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights for berberine and theophylline were obtained, respectively, higher than the individual mode reported by MSS [12,15,16] and EKSI [18]. It should be noted that berberine exhibited higher focusing efficiency than theophylline, which should be attributed to the stronger affinity of the berberine to SDS micelle than that of theophylline. This phenomenon indicated that the proposed MSS–LASEKSI method may be more effective for the hydrophobic analytes. And, more remarkable, the elution order of berberine and theophylline showed opposite result comparing the MSS–LASEKSI enrichment technique and the normal CZE mode. This was due to the decreased mass-to-charge ratio of theophylline, as a result of the suppressed ionization when basic separation buffer was used in CEZ mode. The linearity, detection limits and the repeatability of the proposed method for detection
180
Y.-l. Dong et al. / J. Chromatogr. A 1265 (2012) 176–180 Table 2 Recovery study of the two test alkaloids at different spiked level in urine sample. Added (g/mL)
Berberine 101
0.3
Recovery (%) RSD (%)a Migration time Peak height Recovery (%) RSD (%) Migration time Peak height
98
0.5
100
1.0
Recovery (%) RSD (%) Migration time Peak height
a
2.4 4.5
3.1 6.0
1.2 7.6
Theophylline 95 3.7 1.0 99 0.9 3.4 112 1.0 5.7
The relative standard derivation was calculated from three recovery data.
coexisting compounds in urine sample did not influence the CE determination. Fig. 6. The typical electropherograms of a standard mixture solution. (a) The normal CZE mode, injected at a pressure of 0.5 psi for 3 s, using 20 mM of borate as separation buffer without adjustion of pH. The concentration of berberine is 250.0 g/mL and theophylline is 100.0 g/mL. (b) The present MSS–LASEKSI mode, stacking for 10 min under 10 kV; concentration of berberine and theophylline: 1.0 g/mL. (c) The current MSS–LASEKSI method, blank sample. Other conditions are the same as those in Fig. 3. Peak identification: 1, theophylline; 2, berberine. Table 1 Performance of the present MSS–LASEKSI method.
Linear range (ng/mL) Regression equation Correlation coefficient Limit of detection (S/N = 3) (ng/mL) RSD of migration time (n = 3) (%) RSD of peak height (n = 3) (%) a
Berberine
Theophylline
5.0–2000.0 y = 13,147.9 + 38.2xa 0.997 0.4 3.9 8.4
2.0–2000.0 y = 2841.5 + 41.84x 0.999 0.3 0.7 1.6
4. Conclusion In the present work, we established a new on-line enrichment method based on a modified micelle to solvent stacking (MSS) technique coupled with large amount sample electrokinetic stacking injection (LASEKSI). Under the optimum conditions, 6.3 × 103 - and 6.4 × 102 -fold enrichment in peak heights and a high number of plates of 2.9 × 106 and 6.5 × 105 were obtained for the two model analytes, berberine and theophylline, respectively, higher than the single method. The developed MSS–LASEKSI method described here may provide prospects for exploiting novel combined technique to achieve higher concentration fold, therefore advance the CE field. Acknowledgments
y, peak height; x, concentration (ng/mL).
of berberine and theophylline were carried out under the optimum conditions and results were summarized in Table 1. The limit of detection (LOD) of berberine and theophylline were 0.4 and 0.3 ng/mL, respectively, and the calibration curves were linear up to 2000 ng/mL for these compounds with 10 min of sample stacking. In addition, the method produced narrow peaks for berberine and theophylline, which was observed with a high number of plates of 2.9 × 106 and 6.5 × 105 /m, respectively. The RSDs of the migration times and peak height were 3.9 and 8.4%, and 0.7 and 1.6% for berberine and theophylline respectively, indicating good precision. 3.4. Applications In order to evaluate the feasibility of the proposed on-line concentration method for the analysis of berberine and theophylline, urine sample was used as test sample in the study. Under the optimum conditions, recovery experiments were performed by standard addition method. The results were shown in Table 2, indicating that the recovery ranged from 95 to 112%. And the repeatability of the method used in real sample analysis is also satisfactory. The RSD of the migration time ranged from 0.9 to 3.7%, and the peak height ranged from 1.0 to 7.6%. These results suggest that the method proposed here is useful for the detection of berberine and theophylline in urine sample, and other
This work was kindly supported by the National Natural Science Foundation of China (No. 21075056) and the Fundamental Research Funds for the Central Universities (Nos. lzujbky-2012-214 and lzujbky-2011-20). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
J.W. Jorgenson, K.D.A. Lukacs, Anal. Chem. 53 (1981) 1298. S.E. Moring, R.T. Reel, R.E.J. van Soest, Anal. Chem. 65 (1993) 3454. Y. Xue, E.S. Yeung, Anal. Chem. 66 (1994) 3575. Z. Xu, K. Nakamura, A.R. Timerbaev, T. Hirokawa, Anal. Chem. 83 (2011) 398. Q. Pu, J. Yun, H. Temkin, S. Liu, Nano Lett. 4 (2004) 1099. S. Liu, Q. Pu, L. Gao, C. Korzeniewski, C. Matzke, Nano Lett. 5 (2005) 1389. M.R. Mohamadi, N. Kaji, M. Tokeshi, Y. Baba, Anal. Chem. 79 (2007) 3667. P. Britz-McKibbin, K. Otsuka, S. Terabe, Anal. Chem. 74 (2002) 3736. C.X. Cao, Y.Z. He, M. Li, Y.T. Qian, M.F. Gao, L.H. Ge, S.L. Zhou, L. Yang, Q.S. Qu, Anal. Chem. 74 (2002) 4167. J.P. Quirino, S. Terabe, Science 282 (1998) 465. J.P. Quirino, P.R. Haddad, Anal. Chem. 80 (2008) 6824. J.P. Quirino, J. Chromatogr. A 1216 (2009) 294. M.C. Breadmore, J.P. Quirino, Anal. Chem. 80 (2008) 6373. A.M. Guidote Jr., J.P. Quirino, J. Chromatogr. A 1217 (2010) 6290. L. Liu, X. Deng, X. Chen, J. Chromatogr. A 1217 (2010) 175. H. Zhu, C. Ren, S. Hu, X. Zhou, H. Chen, X. Chen, J. Chromatogr. A 1218 (2011) 733. M. Gong, K.R. Wehmeyer, P.A. Limbach, W.R. Heineman, Anal. Chem. 78 (2006) 6035. H. Zhang, J. Zhu, S. Qi, N. Yan, X. Chen, Anal. Chem. 81 (2009) 8886. W. Lu, G.K. Poon, P.L. Carmichael, R.B. Cole, Anal. Chem. 68 (1996) 668. R. Palepu, H. Gharibi, D. Bloor, E. Wyn-Jones, Langmuir 9 (1993) 110. Y. Walbroehl, J.W. Jorgenson, Anal. Chem. 58 (1986) 479.