On-chip handling of solutions and electrochemiluminescence detection of amino acids

Sensors and Actuators B 122 (2007) 542–548

On-chip handling of solutions and electrochemiluminescence detection of amino acids Hiroki Hosono, Wataru Satoh, Junji Fukuda, Hiroaki Suzuki ∗ Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8573, Japan Received 1 May 2006; received in revised form 23 June 2006; accepted 24 June 2006 Available online 7 August 2006

Abstract On-chip determination of amino acids was conducted by electrochemiluminescence (ECL) following microfluidic transport, mixing of solutions, and generation of ECL. The fabricated system consisted of a polydimethylsilixane (PDMS) substrate with a flow channel structure and threeelectrode systems formed on a glass substrate that resulted in the microfluidic transport and the ECL. An amino acid solution and a reagent solution containing tris(2,2 -bipyridyl)ruthenium(II) (Ru(bpy)3 2+ ) were transported by electrowetting by applying a negative potential to gold working electrodes formed along the flow channels. The two solutions were then mixed in the mixing channel using an additional gold working electrode formed between the electrodes for transport. ECL was generated by applying a potential to a platinum working electrode formed in the mixing channel. The ECL from l-proline, l-lysine, l-leucine, l-valine, and l-histidine was examined. Clear ECL was observed, and its intensity depended on the concentration of the amino acids. By increasing the pH by mixing solutions and generating the oxidized form, Ru(bpy)3 3+ , on the chip, the detection sensitivity could be improved. l-proline showed the strongest ECL, and concentrations of the nM order could be detected. © 2006 Elsevier B.V. All rights reserved. Keywords: Microfluidic transport system; Electrowetting; Electrochemiluminescence; Tris(2,2 -bipyridyl)ruthenium(II); Amino acids

1. Introduction The potential of micro total analysis system (␮TAS) technology has already been demonstrated by the advances that have been made, and innovating devices have been proposed for applications in diverse fields, ranging from basic chemical analyses to biological diagnostics [1–4]. The advantages of the miniaturization of traditional analytical instruments include a low consumption of reagent and sample solutions, a short analysis time, ease of automation, and cost reduction. Although various device concepts may be considered, a common goal will be to achieve the integration of all critical functions, including sample introduction, transport, mixing, reaction, and detection on a single chip. For many of the current microsystems, the procedure relating to the handling of solutions has to be conducted with the help of external pumps, such as microsyringe pumps and peristaltic pumps. Ideally, however, all procedures should be conducted automatically on the chip after filling solutions in injection

∗

Corresponding author. Tel.: +81 29 853 5598; fax: +81 29 853 4490. E-mail address: [email protected] (H. Suzuki).

0925-4005/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2006.06.030

ports. Many approaches have been proposed for microfluidic transport, in particular, those based on silicon micromachining [5–7]. Although excellent performance of the devices has been demonstrated and some of them have been commercialized, their integration in a microchip and widespread use seem to be limited due to the relatively complicated structure and function of the devices along with their high driving voltage and power consumption. As a solution to these problems, devices based on electrowetting have also been developed by mainly focusing on the transport of a droplet [8]. Electrowetting refers to a phenomenon or a technique that changes the wettability of a metal electrode, or the interfacial tension between the electrode/solution interface, by controlling its potential. Although the influence of interfacial tension is negligible in conventional macroscopic systems, it becomes dominant as the dimensions of the flow system decrease. Therefore, the principle is very attractive, considering that further reduction of the flow channel dimensions will be one of the critical issues in the coming decade. We have applied this principle to the transport of a continuous liquid [9]. This simple and versatile microfluidic system has further led us to develop an integrated sensing device for the determination of ammonia, urea, and creatinine [10].

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

Concerning the sensing chemistry that may be used in the system, electrochemical or fluorescence-based approaches have predominantly been used. Differently from previous studies, our work focused on testing the detection by electrochemiluminescence (ECL). ECL is a chemiluminescent reaction with species generated electrochemically on an electrode. A convincing advantage of ECL over fluorescence detection is that it does not require a bulky light source and can be generated on an electrode on a chip. This is attractive because the integration of the necessary components is facilitated. Furthermore, differently from the case of fluorescence detection, the background signal is negligible, thereby allowing optical detectors to be used at their maximum sensitivity. For ECL detection, tris(2,2 -bipyridyl)ruthenium(II) (Ru(bpy)3 2+ ) has predominantly been used because of its excellent stability, water solubility, and ability to undergo a series of related reactions at room temperature [11]. Reducing agents, such as aliphatic amines [12], amino acids [13], oxalate [14], codeine [15], and nicotinamide adenine dinucleotide (NADH) [16], have been analyzed for applications such as liquid chromatography [17,18] and flow injection analysis [12,18]. In many bio/chemical analyses, the mixing of solutions plays a key role in taking full advantage of the detection chemistries. The detection by ECL is no exception. The oxidized form, Ru(bpy)3 3+ , is most stable at pH < 6 and becomes unstable in alkaline solutions [13], whereas the sensitivity of detection increases at higher pHs. Therefore, the generation of Ru(bpy)3 3+

543

should take place immediately after the adjustment of the pH. The present study demonstrates that the procedure can be conducted on a chip using amino acids as the target analytes. 2. Experimental 2.1. Materials and reagents The materials and reagents used for the fabrication of the device and the analysis were obtained from the following commercial sources: negative photoresist, OMR-83, from Tokyo Ohka Kogyo (Kawasaki, Japan); thick-film photoresists, SU-8 25 and SU-8 2100, from Micro Chem (Newton, MA, U.S.A.); precursor solution of polydimethylsilixane (PDMS), KE-1300T, from Shin-Etsu Chemical (Tokyo, Japan); tris(2,2 -bipyridyl)dichlororuthenium(II) hexahydrate, from Sigma–Aldrich Japan (Tokyo, Japan); precursor solution of a photosensitive poly(vinyl alcohol), PVA-SbQ, from Toyo Gosei (Chiba, Japan); amino acids, from Wako Pure Chemical Industries (Osaka, Japan). Other reagents used were of analytical reagent grade and obtained from Wako Pure Chemical Industries. All solutions were prepared with distilled deionized water. 2.2. Structure and fabrication of the device The microfluidic system was designed to transport and mix two different solutions on the chip and generate ECL. The basic

Fig. 1. Active microfluidic transport system with integrated components for optical sensing based on ECL: (a) completed chip; (b) planar layout of the entire system; (c) magnified view of the mixing area showing electrodes for the transport, mixing, and generation of the ECL. The dimensions of the chip are 14 mm × 20 mm. W.E., working electrode; R.E., reference electrode; A.E., auxiliary electrode.

544

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

arrangement of the electrodes is shown in Fig. 1. The system was constructed with a glass substrate (20 mm × 14 mm) with driving electrodes and a PDMS substrate (14 mm × 14 mm) with structures to form microflow channels. The electrodes were formed by a thin-film process. The electrode system consisted of elongated gold working electrodes formed along the flow channels, platinum auxiliary electrodes formed in the injection ports, and Ag/AgCl reference electrodes formed in separate compartments. A negative photoresist layer was formed in areas other than the flow channels to confine the solutions in the flow channels more effectively. In order to increase their durability and handle any solutions irrespective of their components, the Ag/AgCl reference electrodes were formed in separate compartments filled with a PVA-SbQ gel containing 0.1 M KCl. The compartments were connected to the injection ports with a liquid junction. A platinum working electrode (200 ␮m × 300 ␮m) and an Ag/AgCl reference electrode (100 ␮m × 300 ␮m) were formed for the generation of ECL at the end of the mixing channel (Fig. 1(c)). PDMS microstructures to form microflow channels were formed by casting the precursor solution of PDMS on template structures formed with the thick-film photoresist, SU-8. The template consists of a double layer of SU-8 to define the height of the flow channel (SU-8 25) and form a protruding structure for the flow channels (SU-8 2100). The patterns of the gold working electrodes on the glass substrate and the protruding structures of PDMS were aligned under a microscope, and the substrates were fixed (Fig. 2). The height of the flow channels was 40 ␮m. The width of the flow channels for the transport of the sample and reagent solutions was 300 ␮m, and that of the flow channel for the mixing of solutions was 500 ␮m. 2.3. Principles of microfluidic transport and detection The procedure for the liquid transport in one of the flow channels is illustrated in Fig. 2. First, a solution is used to fill the injection port. In this status, the solution only wets the edge of the working electrode and is not mobilized because the electrode surface is not sufficiently hydrophilic (Fig. 2(a)). Next, the potential of the working electrode is changed to −0.9 V (versus on-chip Ag/AgCl). Accompanying this change, the surface of the working electrode becomes more hydrophilic, and the solution is transported along the flow channel to the end of the electrode by a capillary action (Fig. 2(b)). While being transported, the solution is confined in a space between the protruding structure and the working electrode by interfacial tension (Fig. 2(c)). The same principle is applicable to the mixing of two solutions. A cross-section of the mixing area is shown in Fig. 2(d). If two flow channels are placed in close proximity and another working electrode is placed between the two flow channels in parallel, solutions that arrive at the mixing area exude through the intervening glass areas and wet the edges of the electrode for mixing. When the potential is applied, the solutions move on the electrode from both sides and mix. In the presence of Ru(bpy)3 2+ , ECL is generated as a result of the energetic electron-transfer reaction between the oxidized

Fig. 2. Basic structure and arrangement for the microfluidic transport system: (a and b) cross-section along the flow channel showing the initial status (a) and movement of the solution (b); (c) cross-section of the flow channel area viewed from the direction normal to the flow; (d) cross-section of the mixing area along line X–X in Fig. 1. W.E., working electrode; R.E., reference electrode; A.E., auxiliary electrode.

form of the luminophore (Ru(bpy)3 3+ ) and strong reducing intermediate radical anions [13,17,18]. First, Ru(bpy)3 2+ is oxidized on the platinum working electrode to produce Ru(bpy)3 3+ . Ru(bpy)3 2+ → Ru(bpy)3 3+ + e−

(1)

+

(2)

H3 NCRHCO2 − + OH−  H2 NCRHCO2 − + H2 O

Radical anions are formed in the reaction of the amine sites with Ru(bpy)3 3+ . Following this, deprotonation occurs rapidly. H2 NCRHCO2 − + Ru(bpy)3 3+ ˙ + CRHCO2 − + Ru(bpy)3 2+ → H2 N

(3)

− + ˙ ˙ + CRHCO2 − → H2 NCRCO H2 N 2 +H

(4)

− ˙ is a strong The intermediate radical anion H2 NCRCO 2 reducing agent and produces an excited state, Ru(bpy)3 2+* , in an electron-transfer reaction with Ru(bpy)3 3+ . − 3+ ˙ H2 NCRCO 2 + Ru(bpy)3

→ HNCRCO2 − + Ru(bpy)3 2+∗ + H+

(5)

HNCRCO2 − + H2 O → RCOCO2 − + NH3

(6)

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

545

When the excited state decays to the ground state, the ECL is emitted with a wavelength of 610 nm. Ru(bpy)3 2+∗ → Ru(bpy)3 2+ + hv

(7)

The produced Ru(bpy)3 2+ can be oxidized again on the platinum working electrode as in Eq. (1), and the same sequence of reactions is repeated. 2.4. Procedures for the detection of amino acids A standard solution containing an amino acid and a reagent solution containing necessary reagents to generate the ECL were used. When in use, the latter may be stored in the system, and only a sample solution may be injected into the injection port. The reagent solution contained 50 mM Ru(bpy)3 2+ , 10 mM KCl, and 2.0 wt.% ethanol. Ethanol was added to promote the dissolution of Ru(bpy)3 2+ . Amino acids were dissolved in buffer solutions containing 10 mM KCl. The following buffer solutions were used: 0.1 M KH2 PO4 –NaOH (pH 7.0 and 8.0), 0.1 M H3 BO3 –NaOH (pH 9.0 and 10.0), and 0.1 M K2 HPO4 –NaOH (pH 11.0 and 12.0). The standard solution and the reagent solution were transported through the two flow channels as mentioned earlier. After the solutions reached the end of the respective flow channels and wetted the longitudinal edges of the working electrode in the mixing channel, the potentials that had been applied to the working electrodes for transport were switched off. The same potential was immediately applied to the working electrode in the mixing channel to mix the solutions and transport the mixed solution to the platinum working electrode for the ECL. After the potential was switched off, the ECL was generated by applying +1.2 V (versus on-chip Ag/AgCl) to the platinum working electrode immediately by a Hokuto Denko HA-151 potentiostat/galvanostat, and the image was recorded by using a high-sensitivity cooled CCD color camera VB-7010 (Keyence, Osaka, Japan). The exposure time for the camera was adjusted depending on the intensity of the ECL. The ECL intensity was obtained from the image by using image-measurement and image-analyzing software, VH-H1A5 (Keyence, Osaka, Japan). Calibration plots were obtained with a fixed exposure time. 3. Results and discussion 3.1. Microfluidic transport and mixing of solutions The basic performance of microfluidic transport based on this principle was characterized in an earlier study [9]. The flow velocity of the transported solution depends on the potential of the working electrode and the geometry of the flow channel. The flow velocity increases as the working electrode is polarized more to the negative side and as the flow channel becomes narrower. Fig. 3 presents a series of pictures showing the movement of solutions and the generation of ECL. First, the reagent solution containing Ru(bpy)3 2+ and a sample solution containing an

Fig. 3. Transport of fluorescein solutions and generation of ECL: (a and b) two solutions are transported to the end of the electrode; (c) the solutions exude through the glass area to the mixing electrode; (d) the two solutions are mixed on the mixing electrode; (e) the mixed solution is transported to the working electrode for the ECL; (f) the ECL is generated on the working electrode when a potential is applied. W.E.: working electrode.

amino acid were filled in the injection ports. The two solutions were then transported to the mixing area by applying the potential to the respective working electrodes for transport (Fig. 3(a and b)). The solutions slightly exuded through the hydrophilic glass gap area and wetted the longitudinal edges of the mixing electrode (Fig. 3(c)). Since the reagent solution contained ethanol, it spread more on the mixing electrode. When the potential was applied to the mixing electrode, the two solutions mixed

546

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

(Fig. 3(d)). To mobilize the solutions, the mixing electrode must be connected electrically with the reference and auxiliary electrodes in the respective flow channels. In our system, either or both of the solutions on the mixing electrode can be mobilized by connecting the electrodes to one or two potentiostats. Here, only the electrodes for the sample solution side were used, considering the spreading of the reagent solution. As a result, the mixing ratio of the two solutions became approximately 1:1. After mixing, the mixed solution was transported to the working electrode for the ECL (Fig. 3(e)). As seen in Fig. 3(f), red luminescence was observed on the platinum working electrode when the electrode was switched on. The reproducibility of the obtained data was fairly good, as seen in the experimental results shown later. 3.2. pH dependence of the ECL intensity The species that participate in the reactions of the ECL are anions produced from the amino acids. The pH dependence of the ECL of the Ru(bpy)3 3+ –amino acid system was investigated, and the strongest ECL was reported in a pH region between 10 and 11 [13]. As the pH of the solutions increases, the production of the anions is enhanced. As a result, the reaction of the ECL is promoted. The pH dependence of the ECL intensity obtained with l-leucine is shown in Fig. 4. Here, the ECL intensity was measured at pHs between 7 and 12. There were no significant differences in the ECL intensity among the different buffers. The ECL intensity increased significantly at pH > 8. However, the background signal also increased at pH > 9. Hydroxide ions are known to produce ECL with Ru(bpy)3 3+ [13]. The net signal of the ECL originating from the reaction of amino acids was calculated by subtracting the background signal from the total signal. The corrected plot shows that the ECL intensity leveled off at pH > 10. In the following experiments, the pH of the standard solution was set at pH 10.

Fig. 5. Time courses of the ECL intensity of the amino acids. The abscissa indicates the time at which the potential was applied to the working electrode for ECL. The ECL was recorded for 1 s after the indicated time. Three runs were made for the respective amino acids. Averages and standard deviations are shown (most of the bars are behind the symbols). () l-proline; (䊉) l-lysine; () l-leucine; () l-valine; () l-histidine.

3.3. Determination of amino acids based on the ECL detection The time courses of the ECL intensity of the examined amino acids are shown in Fig. 5. The exposure time was 1 s. Immediately after the potential was applied, Ru(bpy)3 3+ was produced on the working electrode and reacted with the amino acids that existed in the vicinity of the electrode. The ECL intensity increased rapidly and reached a maximum. The ECL intensity then decreased and settled at a steady state about 4 s later as a result of the depletion of the amino acids. The relative ECL peak intensities for the examined amino acids are summarized in Table 1. The intensities were normalized to that of l-proline. Of the amino acids examined, only l-proline is a secondary amine amino acid, and its ECL intensity was stronger than that of the other amino acids. Generally, the reducing ability of the secondary amine is stronger than that of the primary amine. The same tendency has been observed in reactions of primary and secondary amines with Ru(bpy)3 3+ [12,17,18]. The ECL intensity of the primary amines is affected by the side chains. An alkyl chain has electron-donating ability, and the ECL intensity became stronger as the alkyl chain length increased [17]. Table 1 Relative ECL intensities of amino acids

Fig. 4. Dependence of the ECL intensity of l-leucine on pH. Five runs were made for the respective pHs. Averages and standard deviations are shown (most of the bars are behind the symbols). (䊉) Measured ECL; () background ECL; () output originating from the ECL of l-leucine.

Amino acida

Relative ECL intensityb

l-proline l-lysine l-leucine l-valine l-histidine

1.00 0.47 0.34 0.18 0.05

a b

± ± ± ± ±

0.02 0.01 0.01 0.02 0.01

Concentration: 100 mM. Intensities normalized to that of l-proline. Exposure time: 1 s.

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

547

ties from the amino acids could be distinguished down to 0.5 nM for l-proline, 5 ␮M for l-leucine and l-lysine, and 10 ␮M for lvaline. However, there was a limitation because the background intensity increased with the increase in the exposure time. The standard deviations of the measured ECL intensity increased at concentrations lower than 0.1 mM. The detection limit should be defined by considering the scattering of the measured values. Lowering the detection limits was not our concern in this study because the detection limits are determined by the instrument used rather than by the method. Although the detection limit was on the order of nM in the best case using our present instrument, a detection limit of the fM order has been reported in an analysis based on ECL [19]. Even with our microsystem, higher sensitivities and lower detection limits will be achieved by using a better detection system. Fig. 6. Dependence of the ECL intensity on the concentration of amino acids. Five runs were made for the respective concentrations. Averages and standard deviations are shown (most of the bars are behind the symbols). Exposure time: 20 s. () l-proline; (䊉) l-lysine; () l-leucine; () l-valine.

The dependence of the ECL intensity on the concentration of the amino acids was examined by increasing the exposure time to 20 s (Fig. 6). The ECL had started to be recorded when the potential was applied to the working electrode. An increase in the ECL intensity was observed with the increase in the concentration of the amino acids. The ECL intensities of the primary amine amino acids (l-leucine, l-valine, and l-lysine) were weak and could not be detected at concentrations lower than 50 ␮M. To examine the calibration plot for lower concentrations of l-proline, the exposure time was increased to 40 s (Fig. 7). A clear tendency was again observed over the range between 10 nM and 100 ␮M. Furthermore, to examine the detection limit with our instrument and method, the exposure time was increased to 60 s, which is the maximum exposure time of the CCD camera. The ECL intensi-

3.4. Future directions In this system, the required volumes of the sample solution and the reagent solution in the flow channels were approximately 50 nl. The measurement by the ECL was accomplished using a very simple structure, and the total time required for the analysis was within 5 min by the manual sequential procedure used in this study. The analysis time will be shortened by using an automated, more sensitive detection system that does not require the exposure procedure. One challenge with our system is to distinguish one amino acid from the others when sample solutions contain several amino acids or other ECL-generating compounds, because the ECL reaction with Ru(bpy)3 3+ is not specific to a particular compound and is common in amino acids. For this purpose, a separating or filtering function may be necessary for our system to be successfully used. The ECL system may also be used to detect specific DNAs or proteins [19–22]. Our next step is to expand the applicable fields in chemical and biological diagnostics. 4. Conclusions

Fig. 7. Dependence of the ECL intensity on the concentration of l-proline in a lower concentration range. Five runs were made for the respective concentrations. Averages and standard deviations are shown. Exposure time: 40 s.

The fabricated microfluidic system can be a basic device for biochemical analyses, which require on-chip microfluidic transport, mixing, and detection. Necessary solutions can be transported on the elongated electrodes in the flow channels and mixed in the mixing area by just applying a negative potential to the working electrode. One of the representative, promising applications is the ECL generated by Ru(bpy)3 3+ and amino acids. ECL can be generated on the chip by applying a potential to the working electrode. ECL can also be enhanced by increasing the pH of the solution using the mixing mechanism. Even with the present device and instrumentation, a detection limit of a nM order can be achieved. Since this device requires solutions only in the order of nl, the consumption of the sample and reagent solutions can be significantly reduced. Although we focused our attention on the combination of the active microfluidic transport and the sensing chemistry, integrated systems based on the ECL detection, which incorporated both a photodetector and the ECL-generating electrodes, have been proposed [23,24]. In addition, another flow channel or mechanism may be incorpo-

548

H. Hosono et al. / Sensors and Actuators B 122 (2007) 542–548

rated to adjust the sample solution pH to an appropriate value on the chip. By combining these technologies, more sophisticated systems with higher functions should be constructed. Acknowledgements This study was partially supported by Grants-in-Aid for Scientific Research on Priority Areas and by the 21st Century COE Program, both of which are under the Ministry of Education, Culture, Sports, Science, and Technology, and Grants-in-Aid for Scientific Research (B), which is under the Japan Society for the Promotion of Science. References [1] D. Figeys, D. Pinto, Lab-on-a-chip: a revolution in biological and medical sciences, Anal. Chem. 72 (2000) 330A–335A. [2] D. Erickson, D. Li, Integrated microfluidic devices, Anal. Chim. Acta 507 (2004) 11–26. [3] T. Vilkner, D. Janasek, A. Manz, Micro total analysis systems. Recent developments, Anal. Chem. 76 (2004) 3373–3386. [4] A. Bange, H.B. Halsall, W.R. Heineman, Microfluidic immunosensor systems, Biosens. Bioelectron. 20 (2005) 2488–2503. [5] M. Elwenspoek, T.S.J. Lammerink, R. Miyake, J.H.J. Fluitman, Towards integrated microliquid handling systems, J. Micromech. Microeng. 4 (1994) 227–245. [6] P. Gravesen, J. Branebjerg, O.S. Jensen, Microfluidics—a review, J. Micromech. Microeng. 3 (1993) 168–182. [7] S. Shoji, M. Esashi, Microflow devices and systems, J. Micromech. Microeng. 4 (1994) 157–171. [8] F. Mugele, J.-C. Baret, Electrowetting: from basics to applications, J. Phys.: Condens. Matter 17 (2005) R705–R774. [9] W. Satoh, M. Loughran, H. Suzuki, Microfluidic transport based on direct electrowetting, J. Appl. Phys. 96 (2004) 835–841. [10] W. Satoh, H. Hosono, H. Suzuki, On-chip microfluidic transport and mixing using electrowetting and incorporation of sensing functions, Anal. Chem. 77 (2005) 6857–6863. [11] R.D. Gerardi, N.W. Barnett, S.W. Lewis, Analytical applications of tris(2,2 -bipyridyl)ruthenium(III) as a chemiluminescent reagent, Anal. Chim. Acta 378 (1999) 1–41. [12] J.B. Noffsinger, N.D. Danielson, Generation of chemiluminescence upon reaction of aliphatic amines with tris(2,2 -bipyridine)ruthenium(III), Anal. Chem. 59 (1987) 865–868. [13] S.N. Brune, D.R. Bobbitt, Effect of pH on the reaction of tris(2,2 bipyridyl)ruthenium(III) with amino-acids: implications for their detection, Talanta 38 (1991) 419–424. [14] I. Rubinstein, C.R. Martin, A.J. Bard, Electrogenerated chemiluminescent determination of oxalate, Anal. Chem. 55 (1983) 1580–1582. [15] G.M. Greenway, L.J. Nelstrop, S.N. Port, Tris(2,2-bipyridyl)ruthenium (II) chemiluminescence in a microflow injection system for codeine determination, Anal. Chim. Acta 405 (2000) 43–50. [16] T.M. Downey, T.A. Nieman, Chemiluminescence detection using regenerable tris(2,2 -bipyridyl)ruthenium(II) immobilized in Nafion, Anal. Chem. 64 (1992) 261–268.

[17] S.N. Brune, D.R. Bobbitt, Role of electron-donating/withdrawing character, pH, and stoichiometry on the chemiluminescenct reaction of tris(2,2 -bipyridyl)ruthenium(III) with amino acids, Anal. Chem. 64 (1992) 166–170. [18] W.A. Jackson, D.R. Bobbitt, Chemiluminescent detection of amino acids using in situ generated Ru(bpy)3 3+ , Anal. Chim. Acta 285 (1994) 309–320. [19] G.F. Blackburn, H.P. Shah, J.H. Kenten, J. Leland, R.A. Kamin, J. Link, J. Peterman, M.J. Powell, A. Shah, D.B. Talley, S.K. Tyagi, E. Wilkins, T.-G. Wu, R.J. Massey, Electrochemiluminescence detection for development of immunoassays and DNA probe assays for clinical diagnostics, Clin. Chem. 37 (1991) 1534–1539. [20] M. Xue, T. Haruyama, E. Kobatake, M. Aizawa, Electrochemical luminescence immunosensor for ␣-fetoprotein, Sens. Actuators B 35–36 (1996) 458–462. [21] H. Yu, J.W. Raymonda, T.M. McMahon, A.A. Campagnari, Detection of biological threat agents by immunomagnetic microsphere-based solid phase fluorogenic- and electro-chemiluminescence, Biosens. Bioelectron. 14 (2000) 829–840. [22] L.J. Obenauer-Kutner, S.J. Jacobs, K. Kolz, L.M. Tobias, R.W. Bordens, A highly sensitive electrochemiluminescence immunoassay for interferon alfa-2b in human serum, J. Immunol. Methods 206 (1997) 25–33. [23] G.C. Fiaccabrino, N.F. de Rooij, M. Koudelka-Hep, On-chip generation and detection of electrochemiluminescence, Anal. Chim. Acta 359 (1998) 263–267. [24] E. L’ Hostis, P.E. Michel, G.C. Fiaccabrino, D.J. Strike, N.F. de Rooij, M. Koudelka-Hep, Mircoreactor and electrochemical detectors fabricated using Si and EPON SU-8, Sens. Actuators B 64 (2000) 156–162.

Biographies Hiroki Hosono received his BE degree in materials science from University of Tsukuba, Japan, in 2005. At present, he is working on a master degree at the Graduate School of Pure and Applied Sciences, University of Tsukuba. His current research interests include ␮TAS and microfluidic transport systems. Wataru Satoh received his BE and ME degrees in materials science from University of Tsukuba, Japan, in 2003 and 2005. In 2005, he became a Research Fellow of the Japan Society for the Promotion of Science. At present, he is working on a doctoral degree at the Graduate School of Pure and Applied Sciences, University of Tsukuba. His current research interests include the development of micro pumps based on electrochemical principles, micro bio/chemical sensors, and ␮TAS. Junji Fukuda is an assistant professor of the Graduate School of Pure and Applied Sciences, University of Tsukuba, Japan. He obtained his PhD and master degrees in chemical engineering from Kyushu University, Japan. His research area has been in the synthesis, processing, and evaluation of new biomaterials for tissue engineering and bioartificial organs. Hiroaki Suzuki received his BE and ME degrees in applied physics and his PhD degree in bioelectronics and biotechnology from the University of Tokyo in 1981, 1983, and 1993, respectively. In 1983, he joined Fujitsu Laboratories, Ltd., Japan. In 1996, he moved to the Institute of Materials Science, University of Tsukuba, Japan, where he became an associate professor of materials science. In 2004, he was promoted to full professor of the Graduate School of Pure and Applied Sciences, University of Tsukuba. His current research interests include micromachining, micro bio/chemical sensors, and ␮TAS.