1991
Short Communication
Incorporation of Disposable Screen-Printed Electrodes for Use in Capillary Electrophoresis End-Column Amperometric Detection System Dong-Mung Tsai, Hsiu-Wen Tai, Pei-Rong Shih, Ren-Lin Chang, Han Wu, Eric D. Conte, Jyh-Myng Zen* Department of Chemistry, National Chung Hsing University, Taichung 402, Taiwan *e-mail:
[email protected] Received: April 14, 2005 Accepted: June 13, 2005 Abstract The development and performance of an end-column amperometric detection system integrated with disposable screen-printed electrodes for capillary electrophoresis is presented. In this system, the electrode and capillary can be easily replaced and the capillary/electrode alignment procedure is straightforward. The use of easily replaceable screen-printed electrodes offers a tremendous benefit for capillary electrophoresis applications requiring frequent replacement of the working electrode due to fouling. This simple and convenient system is very attractive for routine analyses by capillary electrophoresis with electrochemical detection. The separation and determination of uric acid in human urine is presented. Keywords: Screen-printed electrode, Capillary electrophoresis, Amperometric detection DOI: 10.1002/elan.200503321
Capillary electrophoresis (CE) is a powerful separation technique with low sample consumption and relatively short analysis times. However, the adaptation of detection systems to this separation technique is a great challenge, because minute analyte amounts are separated and thus must be detected. Two main detection schemes: laserinduced fluorescence (LIF) and electrochemical (EC) detection have been implemented to address this challenge. Compared to LIF, EC is relatively simple, inexpensive, applicable to a wide range of analytes, and easily miniaturized. Furthermore, EC detection has the added advantage that concentration detection limits are not compromised by miniaturization [1]. Ewing et al. first reported EC detection for CE using an off-column approach [2]; currently end-column and on-column are the two approaches for coupling CE with EC detection [3]. When end-column detection is used, the negative effect of the applied separation voltage on the working electrode is minimized by using 25-mm capillary columns versus larger 50-mm or 75mm columns [4]. The separation noise current decreases as the working electrode is positioned further from the capillary outlet; however detection sensitivity also decreases as a result of analyte diffusion. To prevent the loss of analyte signal through diffusion, on-column detection can be employed. In this arrangement, the separation potential must be grounded prior to the capillary outlet. Several methods for accomplishing this have been described for conventional CE [5 – 9]. Although excellent results have been obtained on the basis of these decoupler devices, implementation is not an easy task. Depending on the type Electroanalysis 17, 2005, No. 21, 1991 – 1994
of decoupler and operating conditions, complete elimination of effects of the high separation voltage on the detection performance cannot always be ensured and additional problems such as partial loss of cations within the decoupler region and band broadening effects may arise. Two major difficulties for CE-EC involve optimizing alignment between the separation capillary and the working electrode and obtaining reproducibility after working electrode replacement. Wang et al. [10 – 13] reported the use of a powerful and flexible thick-film electrode based EC detector for a CE microchip device. The thick film (screenprinting) microfabrication technology is commonly used for large-scale production of extremely inexpensive and yet highly reproducible electrochemical sensors. In the present study we describe a simple capillary electrophoresis endcolumn amperometric detection system utilizing easily exchangeable screen-printed electrodes (SPEs). The endcolumn amperometric assembled cell [50 30 70 mm3 (L W H)] is depicted in Figure 1. The capillary column (ca. 20 mm I. D. 375 mm O. D., 50 cm in length) (1) was held in a stainless steel tube (ca. 450 mm I. D. 630 mm O. D., 3 cm in length) (2) using epoxy resin. The stainless steel tube served as ground electrode for the separation voltage. When using this amperometric detection system, a complex and time consuming alignment procedure between the capillary column and electrodes is not required. In this system there was no decoupling mechanism necessary to isolate the detection electrode from CE voltage and current. The detection was carried out by potentiostatic control of the electrode potential by means of a three-electrode B 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1992
D.-M. Tsai et al.
Fig. 1. A design of CE-EC cell: 1) capillary column, 2) stainless steel tube, 3) and 4) PMMA plates for holding the screen printed electrode (SPE), 5) SPE, 6) stainless steel plate, and 7) push rod.
system. By taking advantage of versatile SPEs, EC is easily applicable to a broad range of important analytes owing to the variety of interchangeable electrode materials and electrochemical processes that can be exploited for detection. Also, a reproducible surface is ensured by the easy replacement of the working electrode, which proves valuable in applications where electrode fouling is a concern. The coupling of SPEs with CE requires proper attention to the capillary detector spacing. To avoid the serious influence to the working electrode detection current, the working electrode should be positioned further from the capillary outlet. However, the detection sensitivity would also decrease as a result of analyses diffusion. In this study, we vary the distance between the electrode surface and the capillary column outlet simply by increasing the number of insulating layers to result in a greater distance from the capillary outlet and electrode surface. The layer number resulting in a certain SPE thickness is presented in Figure 2A. The spacing of the detector was measured by increasing ca. 10 mm stepwise as the insulating layers increase from one to four. We describe the performance of this SPE/EC detector using uric acid as a Electroanalysis 17, 2005, No. 21, 1991 – 1994
Fig. 2. Influence of the distance between the capillary outlet and SPE upon the electropherograms of uric acid. A) The measured distance of the screen-printed strip from the capillary outlet and B) the measured current response to the number of printed insulating layers. Experimental conditions: 1) 20 mM PBS pH 8.0, 2) [uric acid] ¼ 50 mM, 3) capillary length ¼ 49 cm, I. D. ¼ 20 mm, 4) injection and separation voltage ¼ 10 kV, 5) injection time ¼ 5 s, 6) detection potential ¼ 0.7 V (vs. Ag/AgCl).
model analyte. In Figure 2B, the peak current of uric acid at various spacing between the capillary outlet and the electrode surface is presented. At lower thickness (1 and 2 layers), maximum peak current is reached. However, the 2 layers response has better reproducibility than the single layer. A spacing of ca. 25 mm was thus taken as the best spacing between separation and detection systems and was used in all subsequent experiments. Using a 25-mm spacing between the capillary outlet and the electrode surface, the influence of separation voltage upon uric acid response was studied next. The separation voltage affected both the current signal and the baseline response. As expected, increasing the voltage from 6 kV to 18 kV decreases the migration times for uric acid. The largest amperometric signals are observed using the 12 kV
www.electroanalysis.wiley-vch.de
B 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Capillary Electrophoresis End-Column Amperometric Detection System
1993
Fig. 4. Standard addition plot of uric acid in urine. Experimental conditions are the same as in Figure 2 with urine sample diluted by 100 times.
Fig. 3. Effect of detection potential on current response. Experimental conditions are the same as in Figure 2.
separation voltage; however the background reveals a great deal of noise. Overall, the result the amperometric signal at 10 kV is optimum in terms of signal to noise ratio. Figure 3 depicts hydrodynamic voltammograms (HDV) for the oxidation of uric acid. Peak currents were measured stepwise at 100 mV intervals at a 12 kV CE separation potential. The response rises quickly between þ 0.5 and þ 0.7 V, after which current levels off. A dramatic increase in the baseline current, its slope, and the corresponding noise was observed at higher potentials. Subsequent amperometric detection work employed a potential of þ 0.7 V that offered the best signal-to-noise characteristics. Considering the design is aimed at the ability to readily exchange the sensing electrode, the reproducibility of the Electroanalysis 17, 2005, No. 21, 1991 – 1994
system using different screen-printed strips was evaluated. Repeat injections of uric acid were made after exchanging a new SPE between injections and using the same SPE. Both experiments reveal excellent reproducibility. The relative standard deviation for ip measured was 4.65% and 1.67% for 11 different SPEs and 50 continuous injections using the same SPE, respectively. Working electrodes are known to foul during CE runs and replacement does not always lead to reproducible responses. The design of the presented amperometric detector allows for easy replacement of fouled electrodes with accurate alignment that leads to high reproducibility. Calibration plots are obtained for uric acid, in a 20 mM phosphate buffer solution (PBS) at pH 8, from 10 mM to 2 mM. Linearity (r ¼ 0.998) is observed and the resulting calculated limit of detection is 2.75 mM (S/N ¼ 3). Finally, to demonstrate the usefulness of the proposed system, it was applied to the separation and determination of uric acid in human urine. Abnormally high levels of uric acid in urine are indicative of several diseases [14]. Uric acid and Lascorbic acid are present together in biological fluids such as blood and urine and are oxidized at very similar potentials
www.electroanalysis.wiley-vch.de
B 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1994
D.-M. Tsai et al.
[15 – 19]. Therefore it is important to separate uric acid in the presence of L-ascorbic acid for accurate quantification by amperometry. Uric acid was spiked into a real urine sample from 40 mM to 160 mM and again a linear response was obtained (r ¼ 0.999). Results of this sample spike are presented in Figure 4. Referred to the literature [20], UV spectra of uric acid were also applied for the real sample analysis to verify the accuracy of the present approach. Two characteristic absorption peaks appear at 290 and 235 nm for uric acid was observed; and the wavelength of 290 nm was selected for the detection of uric acid. Using standard addition method, a value of 3762 mM in human urine by the proposed method compared favorably to result of 3840 mM obtained with the UV absorption spectra method. The relatively higher value detected by the spectra is because of interference of other compounds present in biological fluids. The results demonstrate that the combination of SPEs with conventional CE systems results in a versatile analytical device. The ability to readily exchange the sensing electrode should be extremely useful for many practical applications that result in surface electrode poisoning. Such flexibility and low cost are united with low detection limits and good precision. Experimental All of the other compounds used in this work were ACScertified reagent grade. Distilled, deionized water (E-pure water purification system, Barnstead, Taiwan) was used for preparing the standard solutions. A 200-ppm Cu(II) solution in 0.1 M HNO3 was used for the platting experiments. Unless otherwise mentioned, all the base electrolyte solutions were pH 8 PBS (I ¼ 0.1 M). Phenols standard in pH 8 PBS was prepared daily and its aliquots were used for electrochemical experiments. Urine sample was diluted by 100-fold and acidified to pH 2.24 with phosphorous acid to precipitate protein. A 2 mL aliquot of this mixture was passed through a 0.22-mm filter for UVand CE-EC analysis. Electrophoresis was driven by a high voltage dc (0 – 30 kV) power supply (Spellman, Plainview, NY, USA). Cyclic voltammetric measurements and chronoamperometric (i-t) experiments were carried out with a CH 900 electrochemical workstation (Austin, TX, USA). The disposable SPEs (5) were purchased from Zensor R&D (Taichung, Taiwan) and had the dimensions of 50 15.5 0.8 mm3 (L W H). A stencil was used in printing the conducting carbon on a flexible polypropylene film (50 70 mm). A silver layer (not in the working portion) was first printed before coating the carbon ink (Acheron, Japan) to make the SPE effectively conductive. Then, the unit was cured in an oven at 100 8C for 30 min. After drying, an insulating layer was finally printed over the SPE leaving a suitable working area. The measured average resistance was 200 2.10 W/cm. Specialized machinery equipment was used to cut a grove (15.5 25 0.5 mm3 (L W H)) in component (4) to hold the SPE in place. Openings were drilled that matched openings on a stainless holder (6) [50 Electroanalysis 17, 2005, No. 21, 1991 – 1994
30 5 mm3 (L W H)] and PMMA plate (4). Bolts (35 mm 2.9 mm diameter) were placed in these openings to secure the entire assembly. The top of component 3 has two openings (dia. ¼ 0.7 mm), one for the stainless steel tube and two for the waste solution. The stainless steel tube with adhered capillary column was placed through the middle position opening of component 3. The capillary extended 1 mm longer than the stainless steel tube and the stainless steel tube extended about 1 mm above the PMMA surface. Component 4 was used to align the capillary column and working electrode. The adjustable push rod (45 mm 5.8 mm, component 7) allows easy replacement of the SPE. Before each run, the capillary was flushed with the sequence of 0.1 M HCl, double distilled water, 0.1 M NaOH, and double distilled water. During the experiments the separation voltage was applied across the capillary and the detection potential was applied at the working electrode. After the electroosmotic current reached a constant value, electromigration injection was carried out and the electropherogram was recorded. All potentials were measured versus Ag/AgCl reference electrode. Acknowledgements The authors gratefully acknowledge financial support from the National Science Council of the Republic of China. Sabbatical leave support was provided for E. D. C. through a Fulbright Fellowship and Western Kentucky University. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
[13]
[14] [15] [16] [17] [18] [19] [20]
F. M. Matysik, Electroanalysis 2000, 12, 1349. R. A. Wallingford, A. G. Ewing Anal. Chem. 1987, 59, 1762. L. A. Holland, A. M. Leigh, Electrophoresis 2002, 23, 3649. T. Kappes, P. C. Hauser, Electroanalysis 2000, 12, 165. W. T. Kok, Y. Sahin, Anal. Chem. 1993, 65, 2497. S. Park, S. M. Lunte, C. E. Lunte, Anal. Chem. 1995, 67, 911. D. M. Osbourn, C. E. Lunte, Anal. Chem. 2001, 73, 5961. D. M. Osbourn, C. E. Lunte, Anal. Chem. 2003, 75, 2710. S. S. Zhang, H. X. Liu, Y. J. Wu, C. L. Yu, Analyst 2001, 126, 441. J. Wang, B. Tian, E. Sahlin, Anal. Chem. 1999, 71, 5436. J. Wang, M. P. Chatrathi, B. Tian, R. Polsky, Electroanalysis 2000, 12, 691. J. Wang, M. Pumera, M. P. Chatrathi, A. Rodriguez, S. Spillman, R. S. Martin, S. M. Lunte Electroanalysis 2002, 14, 1251. J. Wang, M. Pumera, M. P. Chatrathi, A. Escarpa, R. Konrad, A. Griebel, W. Dorner, H. Lowe, Electrophoresis 2002, 23, 596. V. S. E. Dutt, H. A. Mottola, Anal. Chem. 1974, 46, 1777. H. Ernst, M. Knoll, Anal. Chim. Acta 2001, 449, 129. J.-M. Zen, P.-J. Chen, Anal. Chem. 1997, 69, 5087. J.-M. Zen, A. Senthil Kumar, D.-M. Tsai, Electroanalysis 2003, 15, 1073. J.-M. Zen, G. Ilangovan, J.-J. Jou, Anal. Chem. 1999, 71, 2797. J.-M. Zen, C.-T. Hsu, Y.-L. Hsu, J.-W. Sue, E. D. Conte, Anal. Chem. 2004, 76, 4251. J.-F. Jen, S.-L. Hsiao, K.-H. Liu, Talanta 2002, 58, 711.
www.electroanalysis.wiley-vch.de
B 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
