Journal of Chromatography B, 973 (2014) 29–32

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Short Communication

Sensitive determination of four ␤2 -agonists in pig feed by capillary electrophoresis using on-line sample preconcentration with contactless conductivity detection Fan Gao, Minglei Wu, Yi Zhang, Guan Wang, Qingjiang Wang ∗ , Pingang He, Yuzhi Fang Department of Chemistry, East China Normal University, 500 Dongchuan Road, Shanghai 200241, PR China

a r t i c l e

i n f o

Article history: Received 10 June 2014 Accepted 3 October 2014 Available online 12 October 2014 Keywords: Field-enhanced sample injection (FESI) ␤2 -Agonists Capillary electrophoresis Contactless conductivity detection

a b s t r a c t In this work, an on-line preconcentration method of field-enhanced sample injection (FESI) was implemented in the determination of four ␤2 -agoinsts terbutaline (TER), procaterol (PRO), formoterol (FOR) and bambuterol (BAM) by capillary electrophoresis coupled with capacitively coupled contactless conductivity detection (CE-C4 D). Under optimized conditions the background electrolyte (BGE) was 5 mM Tris(hydroxymethyl)aminomethane (Tris) and 10 mM citric acid (Cit) at a pH of 3.2 while the sample dilution solution was obtained by methanol. The detection limits (defined as S/N = 3) of this method were 0.02 mg/L for TER, PRO, FOR, BAM, which were much lower than that of the conventional CEC4 D method without preconcentration procedure, the enhancement factors were greatly improved to be 30–40-fold. The linearity ranges of four ␤2 -agoinsts were 0.1–15 mg/L, with good linear correlation coefficients (r2 > 0.9900). In order to evaluate the application potential of the developed method, real sample from pig feed was analyzed with recoveries of 91.4–106.2%. © 2014 Elsevier B.V. All rights reserved.

1. Introduction ␤2 -Agonists are synthetic derivatives of naturally occurring molecules (catecholamines) [1] (see Fig. S1), and are commonly used in the clinical treatment of pulmonary disorders owing to their bronchodilator activity [2]. However, when these compounds are administered at levels higher than the therapeutic dose, they are active in the repartition metabolism and favor the protein accretion in muscle tissue at the expense of fat. They were therefore employed as growth promoters to improve meat-to-fat ratios. At high concentrations in liver or in meat, the residues of ␤2 -agonists in edible tissues are potentially toxic, since they may cause cardiovascular and other side effects to humans [3]. For this reason, ␤2 -agonists have been banned as growth promoters for the rearing livestock in many European Union (EU) countries and China [4,5]. Consequently, an accurate, reliable and sensitive analytical method to quantification and confirmation of residues for food animal origin, is of considerable importance. ␤2 -Agonists determinations mostly employ HPLC-UV [6,7] and GC-MS [8,9] techniques, because of their good sensitivity.

∗ Corresponding author. Tel.: +86 21 54340015; fax: +86 21 62233508. E-mail address: [email protected] (Q. Wang). http://dx.doi.org/10.1016/j.jchromb.2014.10.004 1570-0232/© 2014 Elsevier B.V. All rights reserved.

However, although the determination of ␤2 -agonists by these methods is satisfactory, extra extraction and purification of analytes are needed prior to analysis, which make these methods either labourious and time-consuming, or require highly specialized equipment to improve sensitivity. On the other hand, capillary electrophoresis (CE) often offers possibilities of simple and rapid analyses. What is more, this method has become a viable technique in separation science on account of its minimum sample and reagent consumption, low cost and high separation efficiency [10–12]. In CE, UV detector is most often used and commercially available. Shi et al. [3] reported the use of CE with UV detection for the determination of ␤2 -agonists in the urine sample and the Chen et al. [13] extended the method to real swine feed samples. Recently, an alternative to UV detection has been developed for CE, namely capacitively coupled contactless conductivity detection (C4 D). In contrast to the optical methods of absorption, conductivity measurements can be more versatile in CE because all ionic species can be directly detected. It is also simpler, less expensive, and has other advantages such as higher robustness as it is not necessary to create an optical window in the protective coating of the separation capillaries [14–16]. Online preconcentration can be regarded as one of the major developments in CE specifically to overcome the sensitivity limitations of electrophoresis. This topic has attracted huge attention in the past 10 years. Among the

30

F. Gao et al. / J. Chromatogr. B 973 (2014) 29–32

preconcentration techniques, FESI is most widely used and the simplest method to be performed, based on conductivity difference between the sample solution and background electrolyte (BGE). In this contribution, we developed a simple, cost effective and highly specific method for separation and determination of four kinds of ␤2 -agonists (TER, PRO, FOR, BAR) in pig feed by CE using field-enhanced sample injection with capacitively coupled contactless conductivity detection (FESI-CE-C4 D). Limits of detection, linearity and recoveries were presented for analytes. To the best of our knowledge, this is the first study using FESI-CE-C4 D technique for the simultaneous determination of these four ␤2 -agonists in pig feed. 2. Materials and methods 2.1. Apparatus and chemicals The laboratory-built CE-C4 D system was used for analysis, as shown in Fig. S2. The C4 D (ER125) instrument was purchased from eDAQ (Denistone East, NSW, Australia). The effective length of the capillary tube (75 ␮m i.d. × 375 ␮m o.d, Yongnian Optical Fiber Factory, Hebei, China) was 50 cm to C4 D. The C4 D operates with a sine-wave signal at a frequency of 750 kHz and an effective voltage of 50 V. The new capillary was conditioned by flushing with 0.1 mmol/L hydrochloric acid solution (60 min), followed by deionised water (30 min), 0.1 mmol/L sodium hydroxide solution (60 min) and the running buffer (30 min). Data acquisition and analysis were preformed using e-corder data acquisition system powerchrom 280 (eDAQ, Denistone East, NSW, Australia). The pH was monitored using a PHS-3C Acidometer (Shanghai Jicheng Instrument Factory, China). All experiments were performed at room temperature. 2.2. Materials and chemicals Terbutaline, procaterol, formoterol, bambuterol were purchased from National Institute of Control of Pharmaceutical and Biological Products (Shanghai, China). l-histidine (His), methanol, acetonitrile, ethanol and lactic acid were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Tris(hydroxymethyl)aminomethane (Tris) and citric acid (Cit) from Shanghai Elaboration Reagent Co., Ltd (Shanghai, China). Morpholinoethanesulfonic acid monohydrate (MES) was supplied by Haijing Reagent Co., Ltd (Shanghai, China). All chemicals were of analytical reagent grade and all solutions were stored in a 4 ◦ C refrigerator. Deionised water was prepared using a Milli-Q water purification system (Millipore, Milford, MA). 2.3. Solution preparation The stock solutions of ␤2 -agonists for the CE separation were prepared at a concentration of 100 mg/L. It was necessary to add a small amount of methanol to stock solution of formoterol for it did not dissolve in water. The running buffer was prepared fresh daily by diluting a stock solution of 100 mM of each standard in deionized water. Stock solutions were stored in 4 ◦ C refrigerator. Before the experiments, all solutions were filtered through 0.22 ␮m polypropylene Acrodisc syringe filter (Xinya Purification Instrument Factory, Shanghai, China) and sonicated for 5 min to remove bubbles. Pig feed sample, obtained from a feed producer, was gently pulverized, then dissolved accurate amount (2.00 g) of the powder from sample in deionized water and sonicated with 8 mL methanol for 2 h. Subsequently, the sample solution was filtered through 0.22 ␮m filter twice and stored in the refrigerator at 4 ◦ C.

Fig. 1. Comparison of electropherograms obtained from two kinds of running buffer. Conditions: buffer, (a) 15 mM lactic acid at pH 2.7; (b) 5 mM Tris + 10 mM Cit at pH 3.2; Capillary column, 50 cm × 75 ␮m; separation voltage, +17 kV; standards: 10 mg/L; injection, electrokinetic injection with 17 kV and 5 s. Sample diluted with running buffer; Peak identification, (1) TER; (2) PRO; (3) FOR; (4) BAM.

3. Results and discussion 3.1. Optimization of CE conditions 3.1.1. Effect of running buffer component, pH value and concentration In the selection of BGE, the main consideration is the ionization characteristic of the analytes. According to the research previously [17], the pKa value of this species are close to 9.0. At high pHvalue, it will take long time for each analysis. The buffer used in the separation thus has a pH-value which is lower than 9.0 for the purpose of achieving protonation of ␤2 -agoinsts as cation, which is a prerequisite for its electrophoretic separation and detection by conductivity measurement. A solution of lactic acid at 15 mM (pH 2.7), which had been employed successfully in previous work for the determination of salbutamol sulfate in medicaments [18], was thus adopted as the running buffer. However, as shown in electropherogram (a) of Fig. 1, a serious peak overlap between the procaterol and formoterol was observed. Then, different solutions including MES, His, Cit, Tris, acetate and borate were evaluated. Among this different tested BGEs, the combination of Tris and His presented the best option for performing the complete separation of ␤2 -agoinsts by comparing the signals of the analytes in the presence of other different solutions (Fig. 1(b)).

F. Gao et al. / J. Chromatogr. B 973 (2014) 29–32

31

Fig. 3. The effect of different sample dilution solvents on peak area in the FESIprocedure. 1. TER; 2. PRO; 3. FOR; 4. BAM; standards: 5 mg/L; running buffer, 5 mM Tris + 10 Mm Cit (pH 3.2); Sample injection was performed at 17 kV and 5 s; other conditions as for Fig. 2.

were injected to compare the enrichment effect. As shown in Fig. 3, it is obviously that the strongest signal was obtained in the case of methanol as the sample dilution solution, followed by ethanol, water and diluted buffer solution. This behavior may due to the biggest conductivity difference between methanol and the running buffer. Thus, the methanol was selected as optimum solvent for samples for subsequent experiments. Fig. 2. The effect of different concentrations of running buffer on separation of analytes. (a) 5 mM Tris + 5 Mm Cit; (b) 5 mM Tris + 10 Mm Cit; (c) 10 mM Tris + 5 Mm Cit; (d) 10 mM Tris + 10 Mm Cit; (e) 10 mM Tris + 15 Mm Cit; Peaks identification (from left), TER, PRO, FOR, BAM; other conditions as for Fig. 1.

The influence of the Tris–His concentration was also evaluated in the interval from 5 to 15 mM, and the results were illustrated in Fig. 2. As the Fig. 2 shown, with increasing background electrolyte concentration, an improvement in the signal-to-noise ratio and resolution was observed. However, high concentration of buffer made the peak broader and migrate time longer due to Joule heating inside the capillary. As a consequence, 5 mM Tris–15 mM His (pH 3.2) (Fig. 1(b)) was utilized as separation BGE. 3.1.2. Effect of the separation voltage Effect of the separation voltage was investigated ranging from 10 to 20 kV. Migration time of the analytes significantly shortened and corresponding peak currents increased gradually with the increase of separation voltage up to 17 kV. However, when voltage larger than 17 kV more Joule heating effect was produced, resulting in peak broadening and in turn affected separation efficiency. Therefore, a voltage of 17 kV was selected as the best separation voltage. 3.2. Optimization of FESI conditions 3.2.1. The influence of sample matrix In FESI, the sample solution is of lower conductivity than the running buffer. Theoretically, the amount of stacking is proportional to the conductivity difference between the running buffer and the sample solution. Adding organic solvents to the sample solution can result in an increase in sensitivity due to the lower conductivity of the sample. Therefore, samples diluted in methanol, acetonitrile, ethanol, diluted buffer and deionized water solution

3.2.2. The influence of injection parameters Injection time (t) and voltage (V) are the most crucial factors that affect the sensitivity enhancement in FESI. By varying the injection time from 2 to 7 s, the peak heights of these four ␤2 -agonists increased. But the band broadening of signals and low separation efficiency were also observed. When the injection time exceeded 6 s, the peak broadening problem in conjunction with peak distortion became more obvious (Fig. S3). Owing to the facts above, one considered that 6 s injection time can be optimal as the compromise between the signal amplification and separation efficiency. Optimization of injection voltages was carried out by injecting the sample electrokinetically in the range of 10–20 kV at 6 s. On one hand, low voltage was not sufficient for effective stacking and high voltage could cause shorter analysis time and obtain sharper peak shape. On the other hand, the excessive voltage was not suitable in consideration of Joule heating and bubble formation. The greatest signal enhancement was achieved after the utilization of 17 kV during the stacking steps. Electropherograms under the above FESI and conventional injection of samples are shown in Fig. 4 for comparison. The stacking efficiency (peak area enhancement factors) could be calculated by multiplying the peak area ratios with the concentration dilution factors [19]. As a result, the stacking efficiencies of four ␤2 -agoinsts with preconcentration method were improved 31, 31, 34, 36 compared with the conventional injection. 3.3. Method evaluation The linearity of the method was determined by constructing a calibration curve with different concentrations of the four ␤2 agoinsts. The concentration of series standard mixture solutions of ␤2 -agoinsts was range from 0.10 to 15.00 mg/L. All the validation data gained were shown in Table 1.

32

F. Gao et al. / J. Chromatogr. B 973 (2014) 29–32

Table 1 The results of linearity, limits of detection (LOD), precision and enrichment factor. Compound

Linear regressiona

TS PH FF BH

y = 10.247x + 0.934 y = 10.068x + 1.747 y = 10.245x + 0.791 y = 10.296x + 0.975

a b

Linearity (mg/L) 0.1–15 0.1–15 0.1–15 0.1–15

Correlation (R2 , n = 6)

LODb (mg/L)

RSD% (time, n = 5)

RSD% (area, n = 5)

Enrichment factor

0.9965 0.9946 0.9969 0.9983

0.02 0.02 0.02 0.02

2.1 1.3 1.0 1.7

4.0 4.9 2.1 4.8

31 31 34 36

Linear regression based on peak area (mV s) vs. concentration (mg/L). Estimated on the basis of S/N = 3.

4. Conclusions A method for the analysis of four ␤2 -agonists by using FESI-CEC4 D was demonstrated for the first time. The experimental results showed that the developed method was simple, robust and sensitive. Compared with the conventional LC–UV and methods, the FESI-CE-C4 D approach not only provided a comparable detection limit of 0.02 mg/L, but also possessed several potential advantages including the fast and highly efficient separation with small injection amount. The applicability of the developed methods to the real pig feed sample analysis demonstrated satisfactory recoveries. The current results provide a great incentive to further investigate of the applicability FESI-CE-C4 D method for achieving detection limits lower than legislated maximum concentration limits. Acknowledgements

Fig. 4. Comparison electropherograms between without concentration (a) and after concentration (b) and (c). Conditions: Capillary column, 50 cm × 75 ␮m; separation voltage, 17 kV; buffer, 5 mM Tris + 10 mM Cit (pH 3.2); (a) the concentration of four ␤2 -agonists are 10 mg/L, sample dissolved in running buffer; (b) the concentration of four ␤2 -agonists are 0.1 mg/L, sample dissolved in methanol; (c) the concentration of four ␤2 -agonists are 5 mg/L, sample dissolved in methanol; injection, electrokinetic injection 17 kV for 6 s; Peaks identification (from left), TER, PRO, FOR, BAM.

The calibration curves exhibited a good linearity with correlation coefficients (R) in the range of 0.9946–0.9983. The LODs, determined form the peak height as the average concentration corresponding to the signal/noise ratio equal to 3, can reach 0.02 mg/L for terbutaline, procaterol, formoterol and bambuterol, indicating that the analytical assay was sensitive to determine ␤2 -agoinsts. The reproducibility of the CE peak areas was studied for five replicate experiments for a sample containing 5 mg/L of each four ␤2 -agoinsts. Consequently, practical repeatability was obtained for all analytes with RSD values (n = 5) were 4.9% or better for peak areas and 2.1% or better for peak time, which indicating good precision of this method. 3.4. Real sample analysis Pig feed sample with four ␤2 -agoinsts was treated with the procedure depicted in Section 2.3 prior to FESI-CE-C4 D analysis, and the results were shown in Table S1. According to the results, none of the ␤2 -agoinsts was detected in pig feed sample by using FESI-CE-C4 D method, which suggests that the ␤2 -agoinsts content was below the detection limits. Under the optimum conditions, recovery experiments were performed by standard addition method. Consequently, the recoveries for ␤2 -agoinsts were ranged from 91.4 to 106.2%. This suggests that the proposed FESI-CE-C4 D approach could be used to determine the concentration of four ␤2 -agoinsts in real pig feed samples.

This work was financially supported by the Program for New Century Excellent Talents in University (NCET-08-0191) and the National Program on Development of Scientific Instruments and Equipment (2011YQ150072). 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.jchromb.2014.10.004. References [1] O.R. Joassard, A.-C. Durieux, D.G. Freyssenet, Int. J. Biochem. Cell B 45 (2013) 2309–2321. [2] S. Collins, M. O’Keeffe, M.R. Smyth, Analyst 119 (1994) 2671–2674. [3] O. Anurukvorakun, W. Buchberger, M. Himmelsbach, L. Suntornsuk, Biomed. Chromatogr. 24 (2010) 588–599. [4] W. Xiu-Juan, Z. Feng, D. Fei, L. Wei-Qing, C. Qing-Yu, C. Xiao-Gang, X. Cheng-Bao, J. Chromatogr. A 1278 (2013) 82–88. [5] J.C. Domínguez-Romero, J.F. García-Reyes, R. Martínez-Romero, J. Chromatogr. B 923–924 (2013) 128–135. [6] A. Blomgren, C. Berggren, A. Holmberg, F. Larsson, J. Chromatogr. A 975 (2002) 157–164. [7] A. Koole, J. Bosman, J.P. Franke, R.A. de Zeeuw, J. Chromatogr. Biomed. Appl. 726 (1999) 149–156. [8] M. Hernández-Carrasquilla, Anal. Chim. Acta 408 (2000) 285–290. [9] L. He, Y. Su, Z. Zeng, Y. Liu, X. Huang, Anim. Feed Sci. Technol. 132 (2007) 316–323. ˇ ˚ [10] P. Tuma, K. Málková, E. Samcová, K. Stulík, Anal. Chim. Acta 698 (2011) 1–5. [11] F.S. Felix, L. Ferreira, d.O.P. Rossini, Talanta 101 (2012) 220–225. [12] J.L. Costa, A.R. Morrone, R.R. Resende, J. Chromatogr. B 945–946 (2014) 84–91. [13] Q. Chen, L.-Y. Fan, W. Zhang, C.-X. Cao, Talanta 76 (2008) 282–287. ˇ P.C. Hauser, Anal. Chim. Acta 607 (2008) 15–29. [14] P. Kubán, [15] A.J. Zemann, Electrophoresis 24 (2003) 2125–2137. [16] T.T.T. Pham, H.H. See, R. Morand, S. Krähenbühl, P.C. Hauser, J. Chromatogr. B 907 (2012) 74–78. [17] F.S. Felix, M.S. Quintino, A.Z. Carvalho, L.H. Coelho, J. Pharm. Biomed. Anal. 40 (2006) 1288–1292. [18] C. Chen, H. Li, Y. Fan, Chin. J. Chromatogr. 29 (2011) 137–140. [19] Q. Weng, G. Xu, K. Yuan, P. Tang, J. Chromatogr. B 835 (2006) 55–61.

Sensitive determination of four β2-agonists in pig feed by capillary electrophoresis using on-line sample preconcentration with contactless conductivity detection.

In this work, an on-line preconcentration method of field-enhanced sample injection (FESI) was implemented in the determination of four β2-agoinsts te...
508KB Sizes 0 Downloads 9 Views

Recommend Documents