The addition of a cationic surfactant to the electrophoretic buffer may induce the reversal of the electroosmotic flow (EOF) in the electrophoretic separation. In fact, the EOF was reversed when the concentration of TTAB added in the phosphate buffer (70 mM) at pH 6.0 exceeded 0.2 mM [40]. In this study, on-line concentration of s-triazines was performed under the conditions of reversed EOF, as the concentrations of TTAB employed were much greater than the critical micelle concentration of TTAB which was determined to be 1.6 ± 0.2 mM at pH 6.0 [41].
3.1 Effect of sample matrix on stacking efficiency
In a previous report [39], thirteen s-triazines, including the four chloro-s-triazines and four methylthio-s-triazines selected in this work, were completely separated by MEKC in a phosphate buffer (70 mM) containing TTAB (15 mM) as a cationic surfactant at pH 6.0. For a large volume of sample injection, however, reoptimization of separation parameters, such as phosphate concentration in the sample matrix and micelle concentration in the separation buffer, in particular, is necessary in order to achieve an effective and efficient stacking.
To examine the influences of sample matrix on the stacking efficiency and detection sensitivity of sample analytes, sample analytes were dissolved in an aqueous solution containing varied concentrations of either phosphate buffer in the range 10~70 mM and 4% acetonitrile as well. Fig. 1 shows some typical electropherograms of four
methylthio-s-triazines obtained with sample solutions containing varied concentrations of phosphate buffer (0, 30, 50, 70 mM), while a separation buffer is composed of 40 mM phosphate buffer and 40 mM TTAB at pH 6.0. Sample analytes at a concentration of 10 µg/mL were injected for 30 s. As shown in Fig. 1A, the peaks of the four s-triazines obtained without addition of phosphate buffer in the sample matrix are rather broad and are poorly resolved. Apparently, the analytes are not in the efficient stacking conditions. The resolutions of peaks become even worse with a longer injection time. In contrast, as shown in Fig. 1B-D, with addition of phosphate buffer in the sample matrix, the peaks become sharpened and the peak height of these sample analytes increases with increasing phosphate concentration up to 50 mM. However, the peak height of these s-triazines decreases with further increasing the concentration of phosphate buffer in the sample matrix.
The conductivity values of sample matrices and those of buffer electrolytes at varied concentration were measured. The values of enhancement factor (γ) defined as the ratio of the conductivity of buffer electrolyte to that of sample matrix are indicated in Fig. 1. It should be noted that theγvalue of the most effective stacking of analytes is in the range 1.4-1.2, instead of 1.0. Similar phenomena were observed for chloro-s-triazines as for methylthio-s-triazines and the γ value for most efficient stacking of these analytes was found to be 1.19 [42]. Evidently, the afore-mentioned results reveal that the stacking of analytes is primarily due to sweeping mechanism proposed by Quirino and Terabe [22], although, in this study, it is operated in a normal stacking mode with reversed electrode polarity in the presence of reversed electroosmotic flow. As the γ values of the most effective stacking are not closed to 1.0, the contribution of field-amplified sample stacking to the enhancement of
detection sensitivity may not be completely ignored.
3.2 Stacking of analytes by sweeping
Fig. 2 depicts the schematic stacking mechanism of a neutral analyte dissolved in a sample matrix containing phosphate buffer with a separation buffer containing a cationic surfactant. As illustrated in Fig.
2A, the capillary column is initially filled with a micellar background electrolyte (BGE). A sample zone containing nonmicellar sample matrix or simply water is injected hydrodynamically with pressure for a period much longer than usual. By application of voltage at negative polarity (Fig. 2B), the electroosmotic flow is directed toward the anode (as cationic micelles is adsorbed on the capillary wall) and the micelles migrate toward the cathode. During sweeping, the analytes stacked at the concentration boundary between the region of [mc]s and that of [mc] =0 in the sample zone, where [mc]s denotes the concentration of micelles in the sample zone after sweeping and [mc] =0 indicates that no micelles are present in the original sample zone. The separation is then achieved via MEKC (Fig. 2C).
3.3 Effect of micelle concentration
The stacking efficiency and detection sensitivity of these test analytes are greatly affected by TTAB concentration. The peak height of each individual analyte increases with increasing micelle concentration until reaching the maximum, then it decreases with further increasing micelle concentration. Fig. 3 shows the variations of the peak height of methylthio-s-triazines as a function of TTAB concentration. As can be seen, the optimal TTAB concentrations determined for terbutryn,
prometryn, ametryn and simetryn are about 30, 45, 50 and 70 mM, respectively, at the sample concentration of 10 µg/mL for a 30-s injection.
In the present study, the optimal TTAB concentration for a simultaneous detection of methylthio-s-triazines is 40 mM for a 30-s injection and 60 mM for a 60-s or longer injection time. It should be noted that, with TTAB micelles at a concentration less than 20 mM, a satisfactory stacking of terbutryn for a 30-s injection is difficult.
3.4 Stacking efficiency versus sample plug length
The stacking efficiencies in terms of peak height (SEheight) for a neutral analyte is defined as the peak of the analyte for varied lengths of sample plug (Hstack) divided by the peak height of the corresponding analyte obtained for 1-s injection (H1s). As a certain minimum injection time is required with Beckman A/PCE MDQ system for a particular injection pressure, H1s is defined as the stacking efficient of a minimal injection time (Hmin) divided by the minimal injection time. For example, with an injection pressure of 1 psi, the minimum injection time is 3.5 s.
Then H1s is equivalent to H3.5s divided by 3.5.
For pressure injections, the length of sample plug in a capillary is directly proportional to the product of the injection pressure and injection time. Thus a 30-s injection of sample solution with a pressure of 1 psi corresponds to a sample plug length of 4.23 cm, which is 6.04 % occupancy of the capillary.
Fig. 4 shows the effect of sample plug length on the SEheight using a mixture of four methylthio-s-triazines at a concentration of 1.0 µg/mL.
As illustrated, the stacking efficiency of these test analytes increases
linearly with increasing injection time up to about 70, 60, 35, and 20 s for terbutryn, prometryn, ametryn and simetryn, respectively, then increase gradually with further increasing injection time. Consequently, the maximum detection sensitivity of terbutryn, prometryn, ametryn, and simetryn are obtained with a duration of injection time of 70, 60, 35, and 20 s, respectively.
To demonstrate the stacking efficiency and the enhancement of detection sensitivity, Fig. 5 shows a typical electrophoreogram of the four methylthio-s-triazines obtained for a 30-s sample injection together with an electrophoreogram obtained for a 2.5-s injection under the optimal condition of usual injection time for comparison. Evidently, by the application of sweeping technique, the detection sensitivity of analytes can be greatly enhanced.
3.5 Peak width versus binding constant
It is of interest to note that, under the effective stacking conditions, the peak widths of these s-triazines increase in the order:
simetryn > ametryn > prometryn > terbutryn. As the magnitudes of binding constants of methylthio-s-triazines to TTAB micelles also increase in the same order as for the peak width of the analytes [39], the dependence of the stacking of these analytes on their binding constants is evident. As a matter of fact, it is observed that the stronger the interaction between the analytes and the micelles, the narrower the peak width. As the binding constant of a sample analyte is linearly related to its retention factor, the result is qualitatively consistent with the finding obtained by Quirino and Terabe [22].
3.6 Detection limits and reproducibility
The limits of detection (LOD) at a signal to noise ratio (S/N) is equally to 3, as well as the reproducibility of migration times and peak heights, for these s-triazines were determined. The migration times of these analytes were quite reproducible, with relative standard deviations (RSD) varying in the range 0.8-1.0 % (n=8). The variations of the peak height with RSD less than 10.5 % were obtained. The values of LOD determined for these methylthio-s-triazines with sample concentration in the range of 1000-50 ng/mL for a 30-s injection time are ranging from 9 ng/mL for simetryn to 15 ng/mL for terbutryn. Table 1 gives the data of analysis for these four analytes.
4. CONCLUSION
On-line concentration of neutral species of s-triazine herbicides in MEKC using a cationic surfactant is demonstrated. The stacking efficiency of analytes can be greatly enhanced by sweeping with addition of buffer electrolyte in the sample matrix and with an appropriate micelle concentration in the separation buffer. Reoptimization of separation parameters is necessary for a large-volume sample injection. For analytes with considerably different binding constants to the micelles, the optimal micelle concentration for an efficient stacking may be different from one analyte to the other.
ACKNOWLEDGMENT
We thank the National Science Council of ROC on Taiwan for financial support.
REFERENCES
[1] L. Hernandez, N. Joshi, P. Verdeguer, N.A. Guzman, in Capillary Electrophoresis Technology, N.A. Guzman (Ed), Marcel Dekker, New York, NY, 1993, pp. 605-614.
[2] D.B. Craig, C.Y. Wong, N.J. Dovichi, Anal. Chem., 68 (1996) 69.
[3] R.L. Chien, D.S. Burgi, J. Chromatogr., 559 (1991) 141.
[4] R.L. Chien, D.S. Burgi, J. Chromatogr., 559 (1991) 153.
[5] D.S. Burgi, R.L. Chien, Anal. Chem., 63 (1991) 2042.
[6] R.L. Chien, D.S. Burgi, Anal. Chem., 64 (1992) 489A.
[7] D.S. Burgi, R.L. Chien, J. Microcolumn Sep., 3 (1991) 199.
[8] D.S. Burgi, Anal. Chem., 65 (1993) 3726.
[9] Z. Liu, P. Sam, S.R. Sirimanne, P.C. McClure, J. Grainger, D.G.
Patterson, J. Chromatogr., 673 (1994) 125.
[10] K.R. Nielsen, J.P. Foley, J. Chromatogr. A, 686 (1994) 283.
[11] J.P. Quirino, S. Terabe, J. Chromatogr. A, 781 (1997) 119.
[12] J.P. Quirino, S. Terabe, J. Chromatogr. A, 791 (1997) 255.
[13] J.P. Quirino, S. Terabe, Anal. Chem., 70 (1998) 149.
[14] J.P. Quirino, S. Terabe, Anal. Chem., 70 (1998) 1893.
[15] J.P. Quirino, K. Otsuka, S. Terabe, J. Chromatogr. B, 714 (1998) 29.
[16] J.P. Quirino, S. Terabe, J. Chromatogr. A, 798 (1998) 251.
[17] P. Gebauer, W. Thormann, P. Bocek, J. Chromatogr., 608 (1992) 47.
[18] C. Schwer, F.J. Lottspeich, J. Chromatogr., 623 (1992) 345.
[19] T.J. Thompson, F. Foret, P. Vouros, B.L. Karger, Anal. Chem., 65 (1993) 900.
[20] C.X. Zhang, W. Thormann, Anal. Chem., 68 (1998) 2523.
[21] C.X. Zhang, W. Thormann, Anal. Chem., 70 (1998) 540.
[22] J.P. Quirino, S. Terabe, Science, 282 (1998) 465.
[23] J.P. Quirino, S. Terabe, K.Otsuka, J.B. Vincent, G. Vigh, J.
Chromatogr. A, 838 (1999) 3.
[24] J.P. Quirino, S. Terabe, J. Chromatogr. A, 850 (1999) 339.
[25] J.P. Quirino, S. Terabe, Anal. Chem., 71 (1999) 1638.
[26] J.P. Quirino, S. Terabe, J. Chromatogr. A, 856 (1999) 465.
[27] J.P. Quirino, S. Terabe, Electrophoresis, 21 (2000) 355.
[28] J.P. Quirino, S. Terabe, Anal. Chem., 72 (2000) 1023.
[29] J.P. Quirino, S. Terabe, P. Bocek, Anal. Chem., 72 (2000) 1934.
[30] F.A. Gunther (Editor), Residue Reviews, Springer-Verlag, Heidelberg, Germany, 1970,Vol. 32.
[31] D. Barcelo, J. Chromatogr. A, 643 (1993) 117.
[32] J.R. Dean, G. Wade, I.J. Barnabas, J. Chromatogr. A, 733 (1996) 295.
[33] V. Pacakova, K. Stulik, J. Jiskra, J. Chromatogr. A, 754 (1996) 17.
[34] J.T. Smith, W. Nashabeh, Z. El Rassi, Anal. Chem., 66 (1994) 1119.
[35] C. Desiderio, S. Fanali, Electrophoresis, 13 (1992) 698.
[36] R.C. Martinez, E.R. Gonzalo, A.I.M. Dominguez, J.D. Alvarez, J.H.
Mendez, J. Chromatogr. A, 733 (1996) 349.
[37] J. Cai, Z. El Rassi, J. Chromatogr., 608 (1992) 31.
[38] W.M. Nelson, C.S. Lee, Anal. Chem., 68 (1996) 3265.
[39] C.E. Lin, C.C. Hsueh, T.Z. Wang, T.C. Chiu, Y.C. Chen, (1999) J.
[42] C.E. Lin, T.Y. Yang, Y.C. Liu, T.Z. Wang, (in preparation)
Table 1 Limits of detection (LOD, S/N=3) and reproducibility of methylthio-s-triazines for 60-s injection time.a
Methylthio-s-triazines Simetryn Ametryn Prometryn Terbytryn Equation of line y=2.2324x
–0.0159
y=2.3910x –0.0232
y=2.0220x –0.0294
y=1.9795x +0.0488 Coefficient of variation (R2) 0.9998 0.9999 0.9991 0.9987
LOD (ng / mL) 13 9 14 15
RSD (%)
Migration time 1.01 1.00 0.80 0.83
Peak height 8.7 9.1 9.4 10.5
a Sample matrix: 50 mM phosphate buffer containing less than 5% CH3CN.
Buffer electrolyte: 40 mM phosphate buffer containing 40 mM TTAB at pH 6.0
Figur e Captions
Fig. 1 Effect of sample matrix on the stacking efficiency and detection sensitivity of methylthio-s-triazines. Sample matrix：(A), water; (B), 30 mM phosphate buffer; (C), 50 mM phosphate buffer; (D), 70 mM phosphate buffer.
Separation buffer, 40 mM TTAB in 40 mM phosphate buffer at pH 6.0;
injection pressure, 1 psi; injection time, 30 s; capillary, 70 cm × 50 µm I.D;
applied voltage, -20 kV, detection wavelength, 222 nm; temperature, 25℃;
sample concentration, 10 µg/ml, sample dissolved in a sample matrix containing 4 % acetonitrile solution. Peak identification, 1 = simetryn, 2 = ametryn, 3 = prometryn, 4 = terbutryn.
Fig. 2 Schematic diagram of a stacking mechanism by sweeping using a cationic surfactant (with sample matrix containing phosphate buffer).
Fig. 3 The variation of peak heights of sample analytes as a function of TTAB concentration. The electrophoretic conditions are the same as for Fig. 1C.
Fig. 4 Plots of SEheight versus injection time with a mixture of four methylthio-s-triazines at sample concentration of 1.0 µg/mL.
Electrophoretic conditions are the same as for Fig. 1C, except sample concentration.
Fig. 5 Detection sensitivity of analytes measured under two different separation conditions: (A) without sample stacking (2.5-s injection with an injection pressure of 0.4 psi; sample concentration, 1.0 µg/mL); (B) with
sweeping-stacking (30-s injection with an injection pressure of 1 psi;
sample concentration, 1.0 µg/mL) Other operating conditions are the same as for Fig. 1C.
0.0 4.0 8.0 12.0 16.0 20.0 4
1 3 2
25 mAU
γ = 1.06 γ = 1.40 γ = 2.21 γ = 1.10x103