Fabrication of Integrated Microelectrodes for Electrochemical

was defined by a manual masking of the unwanted part by an insulator. .... with a CHI 800 electrochemical analyzer (CHI Instruments Co., Austin, T...
0 downloads 0 Views 168KB Size
Anal. Chem. 2003, 75, 5406-5412

Fabrication of Integrated Microelectrodes for Electrochemical Detection on Electrophoresis Microchip by Electroless Deposition and Micromolding in Capillary Technique Jilin Yan,† Yan Du,‡ Jifeng Liu,† Weidong Cao,† Xiuhua Sun,† Weihong Zhou,*,‡ Xiurong Yang,*,† and Erkang Wang*,†

State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, 130022, China, and Department of Material Science and Technology, Jilin University, Nanling District, Changchun, 130021, China

A new method for the fabrication of an integrated microelectrode for electrochemical detection (ECD) on an electrophoresis microchip is described. The pattern of the microelectrode was directly made on the surface of a microscope slide through an electroless deposition procedure. The surface of the slide was first selectively coated with a thin layer of sodium silicate through a micromolding in capillary technique provided by a poly(dimethylsiloxane) (PDMS) microchannel; this left a rough patterned area for the anchoring of catalytic particles. A metal layer was deposited on the pattern guided by these catalytic particles and was used as the working electrode. Factors influencing the fabrication procedure were discussed. The whole chip was built by reversibly sealing the slide to another PDMS layer with electrophoresis microchannels at room temperature. This approach eliminates the need of clean room facilities and expensive apparatus such as for vacuum deposition or sputtering and makes it possible to produce patterned electrodes suitable for ECD on microchip under ordinary chemistry laboratory conditions. Also once the micropattern is ready, it allows the researchers to rebuild the electrode in a short period of time when an electrode failure occurs. Copper and gold microelectrodes were fabricated by this technique. Glucose, dopamine, and catechol as model analytes were tested. During the past decade, considerable interest has been focused on the rapidly growing concept of a micro total analysis system (µTAS) or so-called “lab-on-a-chip” and particular attention has been paid to capillary electrophoresis microchips;1-3 the smallscale devices give a way for fast separation and detection with reduced reagent consumption, production of waste, and use of * Main corresponding author. E-mail: [email protected]. † Changchun Institute of Applied Chemistry. ‡ Jilin University. (1) Manz, A.; Graber, N.; Widmer, H. M. Sens. Actuators, B 1990, 1, 244-248. (2) Reyes, D. R.; Iossifidis, D.; Auroux, P.; Manz A. Anal. Chem. 2002, 74, 2623-2636. (3) Auroux, P.; Iossifidis, D.; Reyes D. R.; Manz, A. Anal. Chem. 2002, 74, 2637-2652.

5406 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

energy.2-7 Also the integration of more sample-handling procedures makes the analytical system more automatic and userfriendly.8,9 Most of the detection methods applied to conventional capillary electrophoresis are adopted onto the microchip format.10 Due to the greatly reduced sample volume detected, sensitive methods such as laser-induced fluorescence are preferred.5,11-12 But the bulk and cost of an off-chip detection system compromises the benefits of miniaturization and portability. Electrochemical detection (ECD) just provides an alternative low-cost method of comparable sensitivity; most importantly, it has an inherent character for miniaturization and is ideally suitable for the microchip format.13-15 These benefits are of great interest to the researchers, and there have been intensive reports on the topic.13-32 Two categories of electrodes may be seen from the (4) Jacobsen, S. C.; Hergenroder, R.; Moore, A. W.; Ramsey, J. M. Anal. Chem. 1994, 66, 4127-4132. (5) Dolnik, V.; Liu, S.; Jovanovich, St. Electrophoresis 2000, 21, 41-54. (6) Kutter, J. P. Trends Anal. Chem. 2000, 19, 352-363. (7) Liu, Y.; Ganser, D.; Schneider, A.; Liu, R.; Grodzinski, P.; Kroutchinina, N. Anal. Chem. 2001, 73, 4196-4201. (8) Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 2000, 72, 5814-5819. (9) Chan, J. H.; Timperman, A. T.; Qin, D.; Aebersold, R. Anal. Chem. 1999, 71, 4437-4444. (10) Schwarz, M. A.; Hauser, P. C. Lab Chip 2001, 1, 1-6. (11) Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 2000, 72, 5814-5819. (12) Hadd, A. G.; Raymond, D. E.; Halliwell, J. W.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 1997, 69, 3407-3412. (13) Wang, J. Talanta 2002, 56, 223-231. (14) Lacher, N. A.; Garrison, K. E.; Martin, R. S.; Lunte, S. M. Electrophoresis 2001, 22, 2526-2536. Vandaveer, W. R., IV; Pasas, S. A.; Martin, R. S.; Lunte, S. M. Electrophoresis 2002, 23, 3667-3677. (15) Wang, J.; Chatrathi, M. P.; Tian, B. Anal. Chim. Acta 2000, 416, 9-14. (16) Wang, J.; Chatrathi, M. P.; Tian, B. Anal. Chem. 2000, 72, 5774-5778. (17) Wang, J.; Chatrathi, M. P.; Tian, B. Anal. Chem. 2001, 73, 1296-1300. (18) Wang, J.; Chatrathi, M. P.; Mulchandani, A.; Chen, W. Anal. Chem. 2001, 73, 1804-1808. (19) Backofen, U.; Matysik, F.; Lunte, C. E. Anal. Chem. 2002, 74, 4054-4059. (20) Fanguy, J. C.; Henry C. S. Electrophoresis 2002, 23, 767-773. (21) Schwarz, M. A.; Galliker, B.; Fluri, K.; Kappes, T.; Hauser, P. C. Analyst 2001, 126, 147-151. (22) Dou, Y. H.; Bao, N.; Xu, J. J.; Chen H. Y. Electrophoresis 2002, 23, 35583566. (23) Woolley, A. T.; Lao, K.; Glazer, A. N.; Mathies, R. A. Anal. Chem. 1998, 70, 684-688. 10.1021/ac034017m CCC: $25.00

© 2003 American Chemical Society Published on Web 09/18/2003

literature: off-chip and on-chip. Wang et al. and other groups used off-chip electrodes similar to that in conventional capillary electrophoresis by placing the working electrode some distance from the very outlet of the electrophoresis channel,15-22 and the electrode can be of various types of material and can be replaced when electrode failure occurs. Most of the other groups took the integrated configuration,23-32 where the integrated electrode allows researchers to build more compact microchip CE-ECD systems and it is well suited for the chip concept; also, the procedure for the alignment of the working electrode to the microchannel is greatly simplified. Most recently, Baldwin et al. have described a fully integrated microchip/electrochemical system in which all the electrodes for electrochemical detection and electrophoresis high voltages were made on one of the glass substrates of the chip,25 and relative palm-size circuits were designed for these purposes. This greatly improved the development of µTAS toward the future pictured by Burns.33 Most of the integrated electrodes were fabricated through photolithography and lift-off procedures developed by the electronics industry. These techniques make it possible to build microelectrode patterns at the level of micrometers. The ordinary operation takes quite a few steps, and the whole procedure is time-consuming and of high cost because of the need for a clean room environment and special apparatus such as for vacuum deposition or sputtering, which often make the technique unavailable to chemists.24,34 Electroless deposition provides a much less expensive way for electrode fabrication; furthermore, the technique is quite compatible with ordinary chemistry laboratories. Hilmi and Luong reported fabrication of a gold working electrode by electroless deposition at the outlet of a glass microchip and applied it to the detection of explosive compounds.28 The pattern of the electrode was defined by a manual masking of the unwanted part by an insulator. There are still no reports on the fabrication of patterned microelectrodes by electroless deposition for microchip CE-ECD. Soft lithography as a newly developed technique provides an easy and useful method for microfabrication almost in the same precision as conventional photolithography.35 It includes a series of techniques such as microcontact printing, micromolding in capillaries (MIMIC), and microtransfer molding. Without the need for any special facilities, micropatterns of the size range of 20100 µm can be made under ambient laboratory conditions, and this has been used in the field of the fabrication of microfluidic (24) Martin, R. S.; Gawron, A. J.; Lunte, S. M.; Henry, C. S. Anal. Chem. 2000, 72, 3196-3202. (25) Baldwin, R. P.; Roussel, T. J.; Crain, M. M.; Bathlagunda, V.; Jackson, D. J.; Gullapalli, J.; Conklin, J. A.; Pai, R.; Naber, J. F.; Walsh, K. M.; Keynton, R. S. Anal. Chem. 2002, 74, 3690-3697. (26) Lapos, J. A.; Manica, D. P.; Ewing, A. G. Anal. Chem. 2002, 74, 33483353. (27) Liu, Y.; Fanguy, J. C.; Bledsoe, J. M.; Henry, C. S. Anal. Chem. 2000, 72, 5939-5944. (28) Hilmi, A.; Luong, J. H. T. Anal. Chem. 2000, 72, 4677-4682. (29) Martin, R. S.; Ratzlaff, K. L.; Huynh, B. H.; Lunte, S. M. Anal. Chem. 2002, 74, 1136-1143. (30) Gawron, A. J.; Martin, R. S.; Lunte, S. M. Electrophoresis 2001, 22, 242248. (31) Manica, D. P.; Ewing A. G. Electrophoresis 2002, 23, 3735-3743. (32) Hebert, N. E.; Kuhr, W. G.; Brazill S. A. Electrophoresis 2002, 23, 37503759. (33) Burns, M. A. Science 2002, 296, 1818-1819. (34) Zoski, C. G. Electroanalysis 2002, 14, 1041-1051. (35) Xia, Y.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37, 550-575.

chips.36-38 Fortunately, the microelectrodes for ECD on capillary electrophoresis are of the same magnitude. In this work, we combined the MIMIC technique with electroless deposition to fabricate microelectrodes for electrochemical detection on an electrophoresis microchip system, and reproducible results of electrode patterns were obtained. First, a microscope slide was selectively coated with a thin layer of sodium silicate with the help of a poly(dimethylsiloxane) (PDMS) microchannel. This provided a rough silicate area for the anchoring of catalytic particles that were necessary for the electroless deposition. A subsequent electrolesss deposition procedure was taken to make a thin film of metal electrodes. The deposition took place only at the areas where the catalytic particles were, and thus, microelectrodes of the pattern as the PDMS microchannel formed. Most fantastically, the electrode fabricated through this technique could be rebuilt when an electrode failure occured since the silicate layer remained almost uneffected even when the whole metal layer was destroyed, thus providing a robust microchip CE-ECD system EXPERIMENTAL SECTION Chemicals and Materials. A Sylgard 184 PDMS kit was obtained from Dow Corning (Midland, MI), glucose and sodium hydroxide (semiconductor grade) were from Aldrich (Milwaukee, WI), dopamine was from Fluka (Buchs, Switzerland), catechol from Chinese Academy of Military Medical Sciences (Beijing, China), and the solution for gold plating (Na3Au(SO3)2) from Changzhou Institute of Chemical Research (Changzhou, China). Other materials were of analytical grade and all chemicals were used as received without further purification. Fabrication Procedures. PDMS Layers with Microchannels. PDMS microstructures were fabricated through a micromolding procedure with glass masters that were made from a flat sodium lime glass (SG2506, Shaoguang Microelectronics Co., Changsha, China),39 the glass had been coated with a chromium (145 nm) and a positive photoresist (AZ1805, 570 nm) layer by the producer. Patterns of the electrodes and electrophoresis microchannel were first designed in a computer and printed onto a transparent film with a high-resolution laser printer (3000 dpi). These patterns were transferred onto the photoresist layer by a UV light exposure. After the exposed photoresist and corresponding chromium layer were removed, a wet chemical etching was carried out with HF/NH4F buffer (1.7%:2.3%) to make negative relief of the masters. Etching was performed at a rate of ∼0.7 µm per minute, and depth of the microchannels was achieved by controlling the operation time. After that, the left photoresist and chromium layer were removed and high-quality masters for the PDMS molding were obtained. The degassed mixture of the PDMS monomer and curing agent (10:1) was poured onto the masters and cured in an oven at 75 °C for 1.5 h. The PDMS layers were removed from the mold, and patterns of the negative relief microchannels remained (Figure 1A). Reserviors were made by a stainless hole punch at each end of the microchannels. Configuration of the chip was of a simple cross, and microchannels for electrophoresis were typically about 15 µm deep, 48 µm wide (36) McDonald, J. C.; Duffy, D. C.; Anderson, J. R.; Chiu, D. T.; Wu, H.; Schueller, O. J. A.; Whitesides, G. M. Electrophoresis 2000, 21, 27-40. (37) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35, 491-499. (38) Ng, J. M. K.; Gitlin, I.; Stroock, A. D.; Whitesides, G. M. Electrophoresis 2002, 23, 3461-3473.

Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

5407

Figure 1. Procedure for the fabrication of PDMS microstructures (A) and the configuration of the PDMS micropattern for the pretreatment of the slide (B).

at the bottom, and 60 µm wide at the top. The separation channel was of a length of 43 mm. The microchannel for the fabrication of the electrode was of a pattern as illustrated in Figure 1B and of a depth of 1.0 µm. Rinsing of these PDMS layers was performed by sonication in ethanol, and the layers were let to dry under a nitrogen stream. Microelectrodes by MIMIC and Electroless Deposition. A microscope slide was first immersed in Piranha solution overnight to remove any organic impurities, then washed with deionized water, and dried. The PDMS layer with microchannnel was brought into conformal contact with one end of the slide and a reversible sealing formed; this formed microchannels with patterns just like that of our desired electrode (Figure 2). A drop of sodium silicate solution (4.4%) was dripped in one of the reserviors and automatically filled the whole microchannel under capillary action. Excess solution in the reservoirs was drawn away, and the solution in the microchannel was let to dry at room temperature overnight. After that, the PDMS layer was carefully peeled off and a sodium silicate microstructure was obtained. The slide was put into an oven to be heated at 140-150 °C for 20 min to make the silicate sturdy and tightly bonded to the glass substrate. Electroless deposition was then performed. First, the slide was immersed in a sensitizing solution (containing SnCl2 15 g/L, HCl 50 mL/L, and some tin pellets) for 5 min. After a sonication rinse with DI water for 20 s, the sensitized surface was immersed in a solution of 18 mM ammoniacal AgNO3 for another 3 min. Gray particles of the reduced silver appeared on the micropattern. After another 20 s of sonication with DI water, copper plating was carried out in a 1:1 mixture of solutions A (containing 24 mM CuSO4, 10 mM NiCl2, and 0.3 M formaldehyde) and B (containing 0.25 M NaOH and 0.16 M sodium potassium tartrate) at room temperature. Deposition took place over the catalytic silver particles for 45 min to form the micropattern of the electrode. A gold electrode was made by electroless deposition on the copper thin film. The plating was performed by immersion of the copper pattern in a solution containing 6.3 mM Na3Au(SO3)2, 127 mM NaSO3, and 0.5 M formaldehyde for 1 h. 5408 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

Figure 2. Scheme showing the procedure for the fabrication of the copper and gold microelectrodes.

Both electrodes were used without any further purification. Scanning Electronic Microscopy (SEM). SEM images were taken with a JXA840 scanning electronic microscope (JEOL) and were used to give the shape of the electrodes and the morphology of the electrode patterns. Also, images of the sectional part of the

Potentials were applied to the working electrode with a Ag/AgCl (saturated KCl) reference and a platinum wire counter.

Figure 3. Configuration of the microchip with integrated working electrode.

glass substrate with electrode pattern were taken to measure the thickness of the electrodes. Electrophoresis Microchip. The PDMS layer with electrophoresis microchannels was rinsed, dried, and then reversibly sealed to the slide with the electrode pattern. Alignment of the electrode was performed with the help of an optical microscope. The configuration was similar to that of Woolley’s in that the working electrode was ∼40 µm to the outlet of the microchannel in the detection reservoir (Figure 3).21 Electrophoresis Procedure and Electrochemical Experiments. All buffers were purified with 0.22-µm membranes (Shanghai Xinya Purification Equipment Co., Shanghai, China). Before the experiment, the microchannel was washed with 1 M sodium hydroxide for 15 min, then rinsed with DI water, and conditioned with running buffer. A 30 mM sodium hydroxide solution was used as the running buffer for the experiments with the copper electrode while 15 mM boric acid (pH ) 8.0) was used for experiments with the gold electrode. A two-channel programmable high-voltage resource (0-2000 V, Xi’an Remax Electronic Co., Xi’an, China) was used for electrokinetic injection and electrophoretic separation. The sample waste reservior was left unused, but buffer was added for balance. The waste reservior was maintained at ground during the whole procedure. Injection and separation were performed by precise control of the high voltages applied to the sample and running buffer reserviors. In the injection period, a high voltage was applied to the sample reservior while a lower one at the value at the cross point was applied to the buffer reservior; this made the solution in the buffer channel steady just as the buffer reservior was floated. A corresponding alternation of the voltages was carried out by applying a higher one at the buffer reservior to make the sample reservior working as floated to perform the separation. All electrochemical experiments were performed with a CHI 800 electrochemical analyzer (CHI Instruments Co., Austin, TX).

RESULT AND DISCUSSION Consideration of the Chip and Electrode Configuration. PDMS as a material for microfluidic chips has been widely accepted by researchers for its flexibility, low cost, and ease of fabrication. It can seal to another PDMS layer, flat silicon, or glass surfaces and even that with slight roughness.32 The sealing can be performed both reversibly and irreversibly.33 In our design, reversible sealing was found to be favorable, as this made it easier to handle the chip when channel blockages occurred. A simple peeling off of the PDMS layer with a subsequent sonication rinse may drive away most of the dirt that caused blockages. Also, this sealing method helped when electrode failures occurred. End-column detection of ECD is the most often used configuration on capillary electrophoresis for its elimination of the preparation of decouplers.13 The electrode pattern adopted here was first introduced by Woolley et al., and similar ones were taken by Baldwin et al. and Lopas et al. for their maximized working areas and minimized exposed parts.23,25,26 Those researches were all taken done on glass microchips. Here we found it much easier to perform on PDMS/glass hybrid chips; the crucial step for hightemperature bonding of glass chip was eliminated and there was almost no failure rate for the fabrication of the microchip CEECD system. Fabrication of the Electrode. The overall reaction for the electroless deposition can be depicted as

HCHO + Cu2+ f Cu + HCOOH (with catalytic surfaces) Catalyst was necessary for the deposition; the only solution for electroless deposition was quite stable under normal conditions without the catalyst.40 In our experiment, freshly generated silver particles were used as the catalyst; it was produced by the reduction of ammoniacal silver by Sn2+ absorbed on the rough area of the silicate pattern.

Sn2+ + Ag(NH3)2+ f Ag + Sn4+ The aim of the surface pretreatment of the glass slide was to produce a rough patterned area for the anchoring of Sn2+ and silver particles. Sodium silicate was used for this purpose; it was often used as an adhesive for glass and has been reported for low-temperature bonding of glass microfluidic chips.41,42 After the solution dried in the PDMS/glass microfluidic channel, the surface of the area coated with silicate was greatly enlarged. Most of the Sn2+ ions absorbed on the flat area of the slide could be driven off during the sonication rinse while these on the patterned area remained. Catalytic silver particles grew on the pattern over the absorbed Sn2+ ions, and deposition only took place over these particles. (39) McCreedy, T. Anal. Chim. Acta 2001, 427, 39-43. (40) Qian, M.; Yao, S.; Zhang, S. Morden Surface Technology; Mechannical Industry Press: Beijing, 1999; pp 80-85. (41) Wang, H. Y. L.; Foote, R. S.; Jacobson, S. C.; Schneibel, J. H.; Ramsey, J. M. Sens. Actuators, B 1997, 45, 199-207. (42) Ito, T.; Soubue, K.; Ohya, S. Sens. Actuators, B 2002, 81, 187-195.

Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

5409

Silicate solution was dropped into the PDMS/glass microchannel and formed an ultrathin liquid layer. Vacuum and slightly heating have been tried to accelerate the drying procedure, but distortion of the pattern was often the result. Room-temperature placement was found to need a longer period of time for the procedure, but it helped to get a better silicate pattern. The configuration of the PDMS microchannel for the pretreatment greatly influenced the formation of the electrode pattern. The greater the depth of the microchannel, the rougher the surface area of the silicate, and the better adhesion of the electrolessly prepared copper layer to the slide. But this introduced a few drawbacks: it made sealing the PDMS layer to the slide harder and difficult to get a flat surface of the electrode pattern; also, the time needed for drying of the solution was greatly prolonged; the solution in a 7-µm-depth microchannel remained partly flowable even after 2 days of room-temperature placement. With less depth, the surface area of the silicate was not enough to provide sufficient sites for anchoring the catalytic particles and the prepared metal thin film was found to give a poor coverage of this patterned area; also, the PDMS microchannel was found more apt to collapse, and this made the complete silicate pattern unavailable. A depth of 1 µm was adopted as a compromise. The electrode pattern at the nonworking area beneath the PDMS layer was used to give electrical connection of the working area to the outer circuits; the greater the width, the lower resistance gotten; but this introduced the high width/depth ratio (the depth was at the level of micrometer to submicrometer), and the same problem as the small channel depth occurred. It was chosen at 300 µm to give good electrical conductivity while remaining as a complete electrode pattern. A gold electrode could also be prepared directly on the glass substrate similar to the copper ones, but this gave only a loose adhesion to the glass substrate and the gold thin film was often destroyed when sealing or peeling off the PDMS layer was carried out. Deposition on the copper film was taken for providing a more robust electrode. Characterization of the Electrodes. Scanning electron microscope images of the electrodes showed good agreement of the electrode pattern with our original design (that of the PDMS microchannel). The gold layer was almost the same as the one for copper as it grew right on the thin film. Full coverage of the electrode pattern was not obtained because the deposition took place where the silver particles were (Figure 4); another image of the sectional part (figure not shown) showed the electrolessly prepared metal thin film was of a thickness of 200 nm. Comparative cyclic voltammetries were performed to study the electrochemical property of the electrolessly prepared metal thin layers (see Figure 5). Both voltammograms of the two electrodes exhibited good agreement with that obtained with conventionally prepared metal disk electrodes, which showed the successful formation the thin films. The electrode was robust enough to stand ordinary operations during the experiment, such as rinsing with buffer, adhering to and peeling off the PDMS layer, applying vacuum at the reservoirs, etc., for the metal grew on the rough surface of the slide and provided a good adhesion. Resistance of the electrode (from one end of the prepared wire for electrical connection to another) was measured with a multimeter, that of the copper electrode was found to be in the range from 30 to 50 5410 Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

Figure 4. SEM image of the electrolessly prepared electrodes. The configuration of the working electrodes: the pattern of a copper electrode (200×) (A) and the morphology of the copper (5000×) (B) and gold (8500×) (C) layers.

Ω and that of the gold electrodes in the range of 8-12 Ω. This low resistance afforded good electrical connection of the electrodes to the electrochemical analyzer. The deposition took place on the catalytic particles and gave good coverage of these particles. In contrast to an electrode made by ordinary thin-film and photolithography techniques, no adhesive layer (often chromium or titanium) was used, thus eliminating the introduction of electrochemical contamination of the working electrode material and the resulting bilayer effect. Rebuilding of the electrode. Electrode failure is one of the problems the researchers on CE-ECD must deal with. When this occurs, a regeneration of the electrode is required, and careful polishing of the electrode is often needed. As with the integrated electrodes, the procedure is quite difficult to perform, for the photolithographically prepared metal layer is rather thin and cannot withstand this process; even the finest alumina polishing powder would damage the layer, and only an electrochemical method was reported to regenerate the electrodes.24 In most cases, discarding the electrode plate or even the whole chip was the result. In our design, the consumption or contamination of the working electrode material had little effect on the silicate micro-

Table 1. Analytical Performance of the Electrolessly Made Copper and Gold Electrodes on the Model Analytes electrode material

copper

analyte linear range correlation coefficient limit of detection RSD of peak current (n ) 6) RSD of migration time (n ) 6)

glucose 20 µM-1 mM 0.9932 2.7 µM 4.32% 2.12%

gold dopamine 10-500 µM 0.9950 1.2 µM 3.52% 2.35%

catechol 10-500 µM 0.9913 1.8 µM 4.66% 3.26%

Figure 6. Electropherograms obtained with integrated electrodes. Copper electrode (A): six consecutive runnings of a 100 µM glucose sample (the below) and detection of a 10 µM sample; sample injection at 1200 V for 5 s; running buffer 30 mM sodium hydroxide; detection potential: 700 mV. Gold electrode (B): separation of a mixture of dopamine and catechol (100 µM each): injection at 1200 V for 5 s; running bufferm 15 mM borate acid (pH ) 8.0); detection potential, 800 mV.

Figure 5. Comparative cyclic voltammograms between the electrolessly prepared electrodes (solid line) and conventional prepared microdisk electrodes (dash line): (A) copper electrodes in 30 mM NaOH and (B)gold electrodes in 10 mM K3Fe(CN)6 and 1 M NaCl, both with a scan rate of 100 mV/s.

structure beneath it and the sites for the anchoring catalytic particles remained. This provided the possibility to rebuild the electrode with the same pattern. The left metal could be easily dissolved in a lift-off solution (HCl/H2O2/H2O (1:1:2) was found effective for removing the remaining copper while aqua regia was used for gold.). A subsequent electroless deposition procedure was used to refabricate the electrode. Often, with the same procedure used the same electrode pattern was gotten, because the etching-off procedure had little effect on the silicate microstructure during for the continuing deposition process. The whole chip CE-ECD system could be rebuilt in 1 or 2 h.

Performance of the Integrated Electrode on the Electrophoresis Microchip. The analytical performance of the CE-ECD system with a copper electrode was characterized by the detection of glucose, which was reported to give good response at that electrode material(Figure 6).43 Both injection and separation were performed under a field strength of 250 V/cm. Hydrodynamic voltammetry was carried out to optimize the detection potential, and 700 mV was adopted to give an appropriate S/N ratio. Copper oxidized and dissolved in the high-pH NaOH buffer; the phenomenon was evident while a positive potential was applied. When one part of the electrode was broken, the working exposed area in the detection cell reduced and thus resulted in the drastic fall of the electrochemical current and disappearance or great reduction of electrochemical response of the analyte. The electrode had a lifetime ∼1 h, but during this period, reproducible analytical results were obtained. When electrode break occurred, the PDMS layer was peeled off and the destroyed electrode and the whole microchip could be refabricated through the procedure mentioned above. The detection limit was estimated to be 2.7 µM from an electropherogramm of a 10 µM solution of the analyte (see Table 1). The separation efficiency obtained in our procedure was lower (8600 plates/m), It was the reasonable result of the larger sample plug introduced by the injection method and the difference of electroosmotic flow mobility between the glass substrate and the PDMS layer. (43) Baldwin, R. P. J. Pharm. Biomed. Anal. 1999, 19, 69-81.

Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

5411

Dopamine and catechol were chosen for the characterization of the microchip system with the integrated gold electrode. The gold electrode was more stable than the copper one for it was a more noble metal and the near-neutral running buffer used had less effect on it. It could stand for a longer period of use (a few hours). Also, a hydrodynamic voltammetry was performed and 800 mV was chosen for the detection. Dopamine and catechol were separated and detected in less than 100 s. Detection limits at the micromolar lever were obtained. CONCLUSION A simple method for the fabrication of an integrated electrode for electrochemical detection of an electrophoresis microchip has been developed. The combination of micromolding in capillary technique with electroless deposition gives a convenient way for

5412

Analytical Chemistry, Vol. 75, No. 20, October 15, 2003

the fabrication of microelectrodes, almost with the same precision as a conventional photolithography technique, and the method is compatible with ordinary chemistry laboratories. The PDMS/glass hybridmicrochip CE-ECD system built through this method was found to be robust and suitable for various applications. ACKNOWLEDGMENT This work is supported by the National Natural Science Foundation of China through Grant 20299030 and National Key Basic Research Program 2002CB513110.

Received for review January 8, 2003. Accepted August 8, 2003. AC034017M