Electrochemiluminescence of Ru(II) Complexes Immobilized on a

Apr 18, 2007 - Milica Sentic , Milena Milutinovic , Frédéric Kanoufi , Dragan Manojlovic , Stéphane Arbault , Neso Sojic. Chemical Science 2014 5 (...
0 downloads 0 Views 707KB Size
Download PDF
6446

Langmuir 2007, 23, 6446-6452

Electrochemiluminescence of Ru(II) Complexes Immobilized on a Magnetic Microbead Surface: Distribution of Magnetic Microbeads on the Electrode Surface and Effect of Azide Ion Kikuo Komori, Kazutake Takada,† Osamu Hatozaki, and Noboru Oyama* Department of Applied Chemistry, Graduate School of Science and Technology, Tokyo UniVersity of Agriculture and Technology, Naka-cho, Koganei, Tokyo 184-8588, Japan ReceiVed October 25, 2006. In Final Form: February 20, 2007 The electrochemiluminescence (ECL) of magnetic microbeads modified with tris(2,2′-bipyridine)ruthenium(II) ([Ru(bpy)3]2+) was studied in the presence of tri-n-propylamine (TPA) to develop highly sensitive ECL detection system, where the employed microbead has a diameter of 4.5 µm. The ECL signal of the [Ru(bpy)3]2+ derivativemodified magnetic microbeads was found to be affected by the geometrical distribution of the magnetic microbeads on the electrode surface. The ECL peak intensity increased with increasing the number of the beads on the electrode surfaces up to 1.6 × 106 beads cm-2, although above 1.6 × 106 beads cm-2, it decreased. The ECL decrease arises from the physical prevention of the ECL from reaching the photomultiplier tube by the excessive beads. The observed peak ECL signal of the [Ru(bpy)3]2+ derivative-modified magnetic microbeads in the presence of NaN3, which serves as a preservative substance, mainly appeared at a potential of +0.90 V vs Ag/AgCl where [Ru(bpy)3]2+ is hardly oxidized, whereas the ECL signal in the absence of NaN3 appeared at a potential of +1.15 V. The presence of NaN3 on the electrode surface retards formation of an oxide layer on the electrode surfaces and promotes TPA oxidation. The ECL response at +0.90 V was mainly attributed to ECL reaction of excited-state [Ru(bpy)3]2+* formed by oxidation of [Ru(bpy)3]+ with TPA radical cation, where the [Ru(bpy)3]+ was generated by reduction of [Ru(bpy)3]2+ with TPA radical.

Introduction Tris(2,2′-bipyridine)ruthenium(II), [Ru(bpy)3]2+, has been considered a promising photocatalyst1-3 in a solar energy conversion system and a luminophore4-6 in an analytical application. In the latter case, it has been noted with considerable interest that the electrochemiluminescence (ECL) of [Ru(bpy)3]2+ can extensively be applied to sensitive detection of numerous analytes such as oxalate,7,8 NADH,9,10 amines,11,12 and amino acids.13 In recent years, the ECL reaction of [Ru(bpy)3]2+ (or its derivatives) was found to exhibit the highest ECL efficiency in the presence of tri-n-propylamine (TPA),12 and thus this system has been utilized for applications such as immunoassays6,14-18 and DNA analyses.14,19-24 * Corresponding author. E-mail: [email protected]. † Present address: Graduate School of Engineering, Nagoya Institute of Technology, Gokiso, Showa, Nagoya 466-8555, Japan. (1) Oyama, N.; Yamaguchi, S.; Kaneko, M.; Yamada, A. J. Electroanal. Chem. 1982, 139, 215-222. (2) Kaneko, M.; Yamada, A.; Oyama, N.; Yamaguchi, S. Makromol. Chem., Rapid Commun. 1982, 3, 769-772. (3) O’Regan, B.; Gra¨tzel, M. Nature (London) 1991, 353, 737-740. (4) Chang, M.-M.; Saji, T.: Bard, A. J. J. Am. Chem. Soc. 1977, 99, 53995403. (5) Slinker, J.; Bernards, D.; Houston, P. L.; Abrun˜a, H. D.; Bernhard, S.; Malliaras, G. G. Chem. Commun. 2003, 2392-2399. (6) Richter, M. M. Chem. ReV. 2004, 104, 3003-3036 and references therein. (7) Rubinstein, I.; Bard, A. J. J. Am. Chem. Soc. 1981, 103, 512-516. (8) Rubinstein, I.; Martin, C. R.; Bard, A. J. Anal. Chem. 1983, 55, 15801582. (9) Downey, T. M.; Nieman, T. A. Anal. Chem. 1992, 64, 261-268. (10) Martin, A. F.; Nieman, T. A. Anal. Chim. Acta 1993, 281, 475-481. (11) Noffsinger, J. B.; Danielson, N. D. Anal. Chem. 1987, 59, 865-868. (12) Leland, J. K.; Powell, M. J. J. Electrochem. Soc. 1990, 137, 3127-3131. (13) Jackson, W. A.; Bobbitt, D. R. Anal. Chim. Acta 1994, 285, 309-320. (14) Blackburn, G. F.; Shah, H. P.; Kenton, J. H.; Leland, J.; Kamin, R. A.; Link, J.; Peterman, J.; Powell, M. J.; Shah, A.; Talley, D. B.; Tyagi, S. K.; Wilkins, E.; Wu, T.-G.; Massey, R. J. Clin. Chem. 1991, 37, 1534-1539. (15) Deaver, D. R. Nature (London) 1995, 377, 758-760. (16) Yu, H. J. Immun. Methods 1998, 218, 1-8. (17) Namba, Y.; Usami, M.; Suzuki, O. Anal. Sci. 1999, 15, 1087-1093. (18) Namba, Y.; Sawada, T.; Suzuki, O. Anal. Sci. 2000, 16, 757-763.

To develop highly sensitive detection systems based on ECL, magnetic microbeads have been employed.6,14-18,24 The magnetic microbeads modified with the [Ru(bpy)3]2+ complex can readily be collected on the electrode surface by using a magnet, resulting in a construction of a highly condensed [Ru(bpy)3]2+ complex domain. The use in the area of electroanalytical chemistry seems very promising, since an electrode can be reused by simply washing out the beads from the surface. Although there have been numerous reports on applications of the ECL detection systems, such as immunoassays and DNA analyses, to the best of our knowledge, none of them has studied structural specificity of the ECL reactions for [Ru(bpy)3]2+ immobilized on the magnetic microbeads in detail.25 Therefore, one of the objectives of the present research is to explore the influence of the microbeads distribution at the electrode surface on the ECL responses. The ECL reaction mechanism of dissolved [Ru(bpy)3]2+ species with TPA has mainly been investigated by the Leland and Bard groups.12,26-28 To generate an ECL signal at high efficiency, oxidation of TPA plays an important role in the ECL mechanism. Schemes 1-3 show the proposed mechanisms, where TPA+• ) (CH3CH2CH2)3N+•, TPAH+ ) Pr3NH+, TPA• ) Pr2NC•HCH2(19) Kenten, J. H.; Casadei, J.; Link, J.; Lupoid, S.; Willey, J.; Powell, M.; Rees, A.; Massey, R. Clin. Chem. 1991, 37, 1626-1632. (20) Gudibande, S. R.; Kenten, J. H.; Link, J.; Friedman, K.; Massey, R. J. Mol. Cell. Probes 1992, 6, 495-503. (21) Xu, X.-H.; Bard, A. J. J. Am. Chem. Soc. 1995, 117, 2627-2631. (22) Hsueh, Y. T.; Smith, R. L.; Northrup, M. A. Sens. Actuators, B 1996, 33, 110-114. (23) Boom, R.; Sol, C.; Weel, J.; Gerrits, Y.; de Boer, M.; Wertheim-van, Dillen, P. J. Clin. Microbiol. 1999, 37, 1489-1497. (24) Miao, W.; Bard, A. J. Anal. Chem. 2004, 76, 5379-5386. (25) Oyama, N.; Komori, K.; Hatozaki, O. Stud. Surf. Sci. Catal. 2001, 132, 427-430. (26) Zu, Y.; Bard, A. J. Anal. Chem. 2000, 72, 3223-3232. (27) Kanoufi, F.; Zu, Y.; Bard, A. J. J. Phys. Chem. B 2001, 105, 210-216. (28) Miao, W.; Choi, J.-P.; Bard, A. J. J. Am. Chem. Soc. 2002, 124, 1447814485.

10.1021/la063120e CCC: $37.00 © 2007 American Chemical Society Published on Web 04/18/2007

Electrochemiluminescence of Ru(II) Complexes

Langmuir, Vol. 23, No. 11, 2007 6447

Scheme 1

Scheme 2 Figure 1. (A) SEM image of magnetic microbeads modified with [Ru(bpy)3]2+-NHS-labeled mouse monoclonal antibody. (B) Schematic illustration of the labeled antibody immobilized on magnetic microbead surface. (C) Structure of the [Ru(bpy)3]2+-NHS.

Scheme 3

previously been available. Furthermore, the effect of the microbead distribution on the electrode surface upon the ECL response has remained unknown. Therefore, in this report, ECL behaviors with the attached complexes are described. In order to perform the clinical usage of ECL detection system, we study the correlation of the ECL intensity with distribution of luminomagnetic microbeads collected and reconstructed on the electrode surfaces by a magnet in batch and flow injection systems. Detailed analyses of the ECL responses and their mechanisms on luminomagnetic microbeads are performed in a phosphate buffer solution containing TPA. In addition, we report that an azide added as a stabilizer to serums or an oxidant for organic compounds exhibits a marked positive influence on the ECL intensity. Experimental Section

CH3, P1 ) Pr2N+CHdCH2CH3, and P2 ) Pr2NH + CH3CH2CHO.6,12,28 To date, it has been believed that ECL is produced upon concomitant electrooxidation of [Ru(bpy)3]2+ and TPA (Schemes 1 and 2).12,26,27 Upon oxidation, the short-lived TPA radical cation (TPA•+) is believed to lose a proton from the R-carbon to form a strong reducing species TPA•.6,29,30 This radical can then reduce [Ru(bpy)3]3+ to [Ru(bpy)3]2+* (Scheme 1). TPA• can also reduce [Ru(bpy)3]2+ to [Ru(bpy)3]+, followed by the annihilation reaction between [Ru(bpy)3]3+ and [Ru(bpy)3]+ (Scheme 2).12,26,27 It should be noted that, although TPA can also be oxidized by [Ru(bpy)3]3+, this oxidation reaction occurs only when [Ru(bpy)3]2+ concentration is high enough. In other words, oxidation of TPA by [Ru(bpy)3]3+ is negligible when the concentration of [Ru(bpy)3]3+ is low.26 Recently, Bard’s group proposed that TPA•+ oxidizes [Ru(bpy)3]+ to [Ru(bpy)3]2+* at a more negative-side potential than that of the oxidation of [Ru(bpy)3]2+, without electrochemically oxidizing [Ru(bpy)3]2+ to [Ru(bpy)3]3+ (Scheme 3).28 As described above, the ECL mechanism of dissolved [Ru(bpy)3]2+ species with TPA has been well-researched, but that of the [Ru(bpy)3]2+ derivative immobilized on a magnetic microbead (luminomagnetic microbead) has not been described nor elucidated in detail so far. Data that would allow a direct quantitative comparison with the ECL behaviors at a [Ru(bpy)3]2+ complex in both attached and unattached states have not (29) Smith, P. J.; Mann, C. K. J. Org. Chem. 1969, 34, 1821-1826. (30) Portis, L. C.; Bhat, V. V.; Mann, C. K. J. Org. Chem. 1970, 35, 21752178.

Preparation of Luminomagnetic Microbeads. First, mouse monoclonal antibody was labeled with [Ru(bpy)3]2+ by adding [Ru(bpy)3]2+-NHS (succinimide) (Igen) dissolved in anhydrous dimethyl sulfoxide (Aldrich) to a 0.15 M phosphate buffer solution (pH 7.8) containing the antibody and 0.15 M NaCl, and then the solution was stirred for 30 min in the dark at room temperature. The [Ru(bpy)3]2+labeled antibody was then collected by gel filtration (Sephadex G-25 column). Polystyrene-coated magnetic microbeads with diameter of 4.5 µm (Dynal, Dynabeads D-450) were modified with the [Ru(bpy)3]2+-labeled antibody by immersing the beads in a 0.15 M phosphate buffer solution containing the antibody (Figure 1). The mixture was stirred with a rotating cultivator for at least 12 h at room temperature. The beads were then washed with deionized water several times to remove free antibody. Since the hydrophobic part of the antibody has been known to adsorb on the surface of the beads, the hydrophilic part of the antibody, which is the immunoreactive site, faces outside as shown in Figure 1B.31 The averaged numbers of the antibody labeled with [Ru(bpy)3]2+ on one bead and the [Ru(bpy)3]2+ incorporated that the antibody were determined to be ca. 8000 and 15 molecules, respectively, by spectrophotometric measurements and BCA protein assays. The prepared luminomagnetic microbeads are stored in 0.2 M phosphate buffer solution containing NaN3 (Wako), Triton-X 100 (Aldrich), and Tween 20 (Aldrich). Electrochemical Measurements. Electrochemical measurements for both batch and flow systems were performed in 0.2 M phosphate buffer solution containing TPA (Acros Organics), NaN3, Triton-X 100, and Tween 20. The pH of the electrolyte solution was adjusted with NaOH. A Pt electrode (0.28 cm2), a Ag/AgCl (KCl satd), and a Pt plate were used as working, reference, and counter electrodes, respectively. The areas of the counter electrodes were 1.0 and 0.25 cm-2 for batch and flow systems, respectively. (31) Kawaguchi, H.; Sakamoto, K.; Ohtsuka, Y.; Ohtake, T.; Sekiguchi, H.; Iri, H. Biomaterials 1989, 10, 225-229.

6448 Langmuir, Vol. 23, No. 11, 2007

Figure 2. Schematic depiction of (A) batch and (B) flow cell systems for the ECL detection. The experimental setup is schematically illustrated in Figure 2. First, a magnet (4000 G, 0.5 cm diameter) was set under the working electrode to collect and hold luminomagnetic microbeads. Then, a phosphate buffer containing the microbeads was filled out on the working electrode surfaces in the batch system case or injected to the carrier solution in the flow system case. A peristaltic pump was used for the flow system. Potential was applied to the working electrode with a potentiostat (Polarization Unit PS-02, Toho Technical Research) after the magnet was removed. The ECL signal, measured with a photomultiplier tube (PMT, Hamamatsu H6780-02) to which a voltage of 800 V was applied, was recorded simultaneously with cyclic voltammetry (CV). A CCD camera was used to monitor the geometrical distribution of the luminomagnetic microbeads on the electrode surfaces.

Results and Discussion Distribution of Luminomagnetic Microbeads on the Electrode Surface. In situ monitoring of the luminomagnetic microbeads (7.8 × 104 to 5.8 × 106 beads cm-2) on the electrode surface as shown in Figure 2A was carried out using a CCD video camera. The luminomagnetic microbeads were collected on the electrode surface using a magnet. Figure 3A,B shows the CCD images of luminomagnetic microbeads before and after, respectively, a magnet was removed. Before a magnet was removed, i.e., was set beneath the electrode, we recognized that the luminomagnetic microbeads formed many towerlike structures perpendicularly on the electrode surface along the line of the magnetic force. When luminomagentic microbeads of 1.6 × 106 beads cm-2 in a test cell were gathered on the surface, the average number of beads in each tower built up on the electrode surface was determined to be ca. 30 beads. By removing the magnet behind the electrode, the towers fell and laid down on the electrode surface (Figure 3BII), which indicated that the beads align themselves along the line of the magnetic force. The linked beads

Komori et al.

Figure 3. CCD images of the luminomagnetic microbeads collected on the electrode surface (A) when the magnet was positioned at the backside of the electrode and (B) after the magnet was removed.

were crumbled and widespread on the surface by repeatedly moving the magnet on and off. This phenomenon was observed regardless of the number of beads applied onto the electrode surface. ECL Responses from Luminomagnetic Microbeads. Figure 4 shows the CVs and the simultaneously obtained ECL potential curves for a Pt electrode covered with luminomagnetic microbeads (7.8 × 105 beads cm-2) in a 0.2 M phosphate buffer solution containing 100 mM TPA, 0.05% Triton-X 100, and 0.05% Tween 20. The Triton-X 100 and Tween 20, which were surfactants, were used to stabilize adsorption of the luminomagnetic microbeads on the electrode surfaces. The observed ECL intensity increased in the presence of the surfactants. This may be because their adsorption on the electrode surface would retard the formation of the oxide layer, which prevents generation of the highly reducing radical, TPA•.32-34 As shown in Figure 4 (curve I), an anodic current was observed at a potential higher than +0.60 V, and the ECL peak with the intensity of 350 mV appeared at +1.15 V. This ECL peak should be attributed to the ECL reactions between [Ru(bpy)3]2+-NHS and TPA. A similar result has been reported for the Pt electrode modified with [Ru(bpy)3]2+, which exhibits an ECL peak at ca. +1.20 V (vs Ag/AgCl) in 0.19 M phosphate buffer solution (pH 7) containing 0.13 M TPA.35 The anodic current shown in Figure 4 can mainly be attributed to the oxidation of TPA. In fact, the CV of the luminomagnetic microbead-covered (7.8 × 105 beads cm-2) electrode in the absence of TPA did not exhibit a redox peak corresponding to the [Ru(bpy)3]2+/3+-NHS couple due to the low concentration of (32) Workman, S.; Richter, M. M.; Anal. Chem. 2000, 72, 5556-5561. (33) Zu, Y.; Bard, A. J. Anal. Chem. 2001, 73, 3960-3964. (34) Factor, B.; Muegge, B.; Workman, S.; Bolton, E.; Bos, J.; Richter, M. M. Anal. Chem. 2001, 73, 4621-4624. (35) Xu, X.-H.; Bard, A. J. Langmuir 1994, 10, 2409-2414.

Electrochemiluminescence of Ru(II) Complexes

Langmuir, Vol. 23, No. 11, 2007 6449

Figure 5. Dependences of the ECL intensity on the number of the luminomagnetic microbeads (a) collected by the magnet positioned at the backside of the electrode and (b) after the magnet was removed in 0.2 M phosphate buffer containing 100 mM TPA, 7.7 mM NaN3, 0.05% Triton-X 100, and 0.05% Tween 20 at 100 mV s-1.

Figure 4. (A) Cyclic voltammograms and (B) light emission curves of a Pt electrode covered with luminomagnetic microbeads (7.8 × 105 beads cm-2) in 0.2 M phosphate buffer solution (pH 7.5) containing 100 mM TPA, 0.05% Triton-X 100, 0.05% Tween 20, and NaN3 ((I) 0 and (II) 7.7 mM). Scan rates are 100 mV s-1.

[Ru(bpy)3]2+-NHS. Under the assumption that the luminomagnetic microbeads (4.5 µm diameter) on the electrode surface form a monolayer which has a hexagonal close-packed structure, the concentration of [Ru(bpy)3]2+-NHS was estimated to be ca. 0.36 µM (see Experimental Section). However, the effective concentration of [Ru(bpy)3]2+-NHS participating in the electrontransfer reaction should be lower, because only [Ru(bpy)3]2+NHS species on the microbeads in contact directly with the electrode surface can be electrochemically oxidized. Thus, the current resulting from the oxidation of the [Ru(bpy)3]2+-NHS is negligibly small. In order to develop a practical detection system, NaN3, which is one of the preservatives commonly used in the biological chemistry field, was added to the electrolyte solution. It is surprising that, in the presence of 7.7 mM NaN3, the ECL peaks appeared at +0.90 and +1.40 V with intensities of ca. 1800 and 320 mV, respectively (Figure 4, curve II). It should be noted that in the presence of NaN3 the first ECL peak appeared at a negative potential where the [Ru(bpy)3]2+-NHS species are hardly oxidized, and the intensity was significantly higher than that in the absence of NaN3. In addition, the anodic peak current in the presence of NaN3 was larger than that in its absence. The influence of N3- on the ECL response will be described again in the following section. Influence of ECL Intensity on Luminomagnetic Microbead Distribution. Figure 5 demonstrates the change of the observed ECL peak intensities against the number of the luminomagnetic microbeads in the test cell. The ECL intensity of the beads was enhanced by removing the magnet. This is due to the observed towerlike structure of the microbeads, which makes the number of the luminomagnetic microbeads in direct contact with the electrode surface smaller. However, when the magnet was removed, the microbeads were widespread on the electrode surface and the number of the luminomagnetic microbeads in direct contact with the surface increased. To ascertain whether the magnetic force of a magnet has influence on the photon responses generated with the ECL reaction, the ECL measurement was

performed in 0.2 M phosphate buffer solution containing free 0.1 µM [Ru(bpy)3]2+, 100 mM TPA, 7.7 mM NaN3, 0.05% Triton-X 100, and 0.05% Tween 20 by setting and removing a magnet repeatedly. It was found that the ECL peak intensities were unaffected by the magnetic force. Thus, the difference in the ECL peak intensities of the luminomagnetic microbeads can be attributed to the difference in the numbers of the beads attached directly on the electrode surface. As shown in Figure 5, plot b, the ECL peak intensity increased with increasing number of beads on the electrode surface up to 1.6 × 106 beads cm-2. However, when it became larger than 1.6 × 106 beads cm-2, the intensity appeared to decrease. These results indicate that the accumulated beads physically block the ECL reaching the PMT. Since the half-life of TPA+• has been known to be ca. 0.2 ms with the assumption that the TPA+• deprotonation reaction is a first-order process,28 only the [Ru(bpy)3]2+-NHS species located adjacent to the electrode should produce the ECL. When the luminomagnetic microbeads (7.8 × 105 beads cm-2) were gathered to the concave area on the electrode, where the area is prepared specially to collect the beads, without using a magnet on the electrode, the beads disseminated. The ECL peak intensity observed was 15% higher than that of luminomagnetic microbeads collected by the magnet. This result suggests that the ECL intensity increases when luminomagnetic microbeads are widespread on the electrode surface with dissemination. On the contrary, the aggregation of the luminomagnetic microbeads induced the decrease in the ECL intensity. ECL for [Ru(bpy)3]2+ Species Dissolved in the Solution Containing N3-. In order to elucidate the effect of N3- on the ECL reaction, the ECL measurement was carried out in 0.2 M phosphate buffer solution (pH 7.5) containing free 0.1 µM [Ru(bpy)3]2+, 100 mM TPA, and 7.7 mM NaN3 without the surfactants (Figure 6, curve I). Since the concentration of [Ru(bpy)3]2+ in the present study was far lower than that of TPA, its oxidation current was negligibly small, as in the case of the luminomagnetic microbead system. The anodic current was found to increase from +0.60 V during the anodic potential scan, and the ECL signals appeared to concurrently increase as the oxidation of TPA occurred. The curve of ECL emission vs potential displayed two peaks at +0.90 and +1.15 V with intensities of ca. 800 and 600 mV, respectively. This result indicates that ECL reaction between [Ru(bpy)3]2+ and TPA may occur via at least two parallel reaction routes. In the absence of NaN3, the anodic current started to increase from +0.65 V, and a single ECL peak appeared at +1.05 V with intensity of 380 mV (Figure 6, curve

6450 Langmuir, Vol. 23, No. 11, 2007

Figure 6. (A) Cyclic voltammograms and (B) light emission curves of a Pt electrode in 0.2 M phosphate buffer solution (pH 7.5) containing (I) 0.1 µM [Ru(bpy)3]2+, 7.7 mM NaN3, and 100 mM TPA; (II) 0.1 µM [Ru(bpy)3]2+ and 100 mM TPA; and (III) 0.1 µM [Ru(bpy)3]2+ and 7.7 mM NaN3. Scan rates are 100 mV s-1.

II). Since the ECL intensity in the presence of NaN3 was higher than that in the absence of NaN3, the reaction of [Ru(bpy)3]2+ with TPA was certainly accelerated by N3-, which was electrochemically oxidized to N3• at ca. +1.12 V.36,37 However, since no ECL peak was observed in 0.2 M phosphate buffer containing [Ru(bpy)3]2+ and NaN3 in the absence of TPA (Figure 6, curve III), it can be stated that N3- and/or its oxidized species are not directly involved in the ECL reaction. The anodic current above +0.60 V in the presence of NaN3 was found to be larger than that in its absence. N3- has been known to adsorb on a Pt electrode and to retard the production of an oxide layer,38 resulting in promotion of the electrochemical oxidation of TPA. In addition, the oxidized species of N3- (N3•) is probably protonated. Thus, we suppose that the proton dissociated from TPA•+ on the electrode surface is accepted by N3 -, resulting in promotion of the deprotonation of TPA•+ preferable to the ECL reaction. Dependence of the ECL intensity at +0.90 V on NaN3 concentration was examined in a 0.2 M phosphate buffer solution containing 0.1 µM [Ru(bpy)3]2+ and 100 mM TPA. The ECL intensity increased with increasing NaN3 concentration to 7.7 mM, and then followed by decrease. At NaN3 concentration higher than 7.7 mM, [Ru(bpy)3]2+* and TPA•+ might be reduced by N3-. When experiments were also performed at various TPA concentrations (0.01-0.1 M), there were optimal NaN3 concentrations giving rise to the highest ECL intensity at each TPA concentration. Bard’s group has reported that the TPA oxidation is promoted when halide ions adsorbed on an electrode surface, and the optimal concentration of halide ion exists at each TPA concentration.26 Thus, N3- in a solution may exhibit a similar effect to halide ions. Note that the ECL peaks in the presence (36) Ward, G. A.; Wright, C. M. J. Electroanal. Chem. 1964, 8, 302-309. (37) Alfassi, Z. B.; Harriman, A.; Huie, R. E.; Mosseri, S.; Neta, P. J. Phys. Chem. 1987, 91, 2120-2122. (38) Roscoe, S. G.; Conway, B. E. J. Electroanal. Chem. 1988, 249, 217-239.

Komori et al.

of N3- appeared at the same potentials as in the case of iodide.26 Meanwhile, since no ECL peak at +0.90 V appeared in the presence of bromide, the ECL peak in the presence of N3- is different from the case when bromide adsorbs on the electrode surface. This difference could be ascribed to the degree of the adsorption strength on the electrode surfaces. Dependence of the ECL responses upon pH was also examined. The ECL intensity at +0.90 V was measured at various pHs in 0.2 M phosphate buffer solution containing 0.1 µM [Ru(bpy)3]2+, 10 mM TPA, and 7.7 mM NaN3. The pH was adjusted by adding NaOH or H3PO4 to the electrolyte solution. The ECL signal increased over the pH range from 6 to 10 and then leveled off at pH above 10. This result may be explained on the basis of pKa of TPA (10.4), because neutral TPA molecules are readily oxidized in comparison to the protonated ones.34 In addition, since TPA became insoluble as pH increased, the ECL signal remained constant. This denotes the same tendency as a reaction system of [Ru(bpy)3]2+ and TPA dissolved in the solution containing no NaN3.12 In addition, a similar pH dependence of the ECL responses was also observed for the luminomagnetic microbead system. Mechanism under Luminomagnetic Microbeads. As mentioned above, the ECL curve of the luminomagnetic microbeads in the presence of NaN3 was found to be different from that in the absence of NaN3. On the basis of the results shown in the previous sections, the ECL mechanism of luminomagnetic microbeads in the presence of NaN3 is proposed as follows. Since the ECL peak of luminomagnetic microbeads appeared at ca. +0.90 V in the presence of NaN3 (Figure 4, curve II), Scheme 3 is anticipated mainly in the ECL mechanism.28 Since at higher concentration of [Ru(bpy)3]2+ (g1 mM) a reversible CV wave ascribed to the [Ru(bpy)3]2+/3+ redox reaction was observed with E0 ≈ +1.10 V, [Ru(bpy)3]3+ is scarcely generated at +0.90 V, as described above. If the ECL reactions of Schemes 1 and/or 212,26 predominately occur, the ECL peak should appear over the potential of +1.10 V. The second ECL peak of the luminomagnetic microbeads appeared at ca. +1.40 V, where the ECL mechanism is expected to be mainly the reaction shown by Scheme 1. Note that the ECL intensity of the second peak was lower than that of [Ru(bpy)3]2+ dissolved in the solution, where the second ECL peak at +1.15 V in the presence of NaN3 (Figure 6, curve I) is mainly related to the annihilation reaction ([Ru(bpy)3]3+ + [Ru(bpy)3]+ f [Ru(bpy)3]2+* + [Ru(bpy)3]2+). Considering that the [Ru(bpy)3]2+-NHS species are immobilized on the magnetic microbeads, the annihilation reaction between [Ru(bpy)3]3+-NHS and [Ru(bpy)3]+-NHS should hardly occur. However, it has previously been reported that in the case of the [Ru(bpy)3]2+ derivative immobilized directly to an electrode surface,35 the ECL peak appeared at a potential of ca. +1.20 V, indicating the annihilation reaction of vicinal [Ru(bpy)3]2+ derivatives. Therefore, in the present study, [Ru(bpy)3]2+-NHS complexes on the microbead surface are likely immobilized apart from each other. Thus, the dominant mechanism of the ECL likely changes in the order of Scheme 3, Scheme 2, and Scheme 1, when the potential is scanned to positive. In the absence of NaN3, TPA is not sufficiently oxidized. Thus, the ECL reactions of Schemes 2 and/or 3 would be the dominant reaction. The decrease of the ECL intensity at a potential higher than +1.1 V could be attributed to the formation of an oxide layer on the Pt electrode surface. Responses in a Flow Injection System. Flow injection systems using magnetic microbeads have been developed and applied for clinical analyses.6,14,15,17,18 However, effects of microbead distribution at the electrode surface on the ECL intensities have

Electrochemiluminescence of Ru(II) Complexes

Langmuir, Vol. 23, No. 11, 2007 6451

Figure 7. (A) Photographs of Pt electrode surfaces attaching the luminomagnetic microbeads (7.8 × 105 beads cm-2) gathered at flow rates of (b) 2.4, (c) 6.3, and (d) 10.3 mL min-1. Photograph (a) is a Pt electrode surface without luminomagnetic microbeads. (B) CCD images of luminomagnetic microbeads on a Pt electrode surface. The numbering of images corresponds to those in (A).

not sufficiently been studied. Therefore, the relation between the distribution of magnetic microbeads and the ECL intensities should be examined and compared with the batch system. In the present study, the distribution of luminomagnetic microbeads and ECL intensity were studied by gathering the beads at various flow rates using a flow injection system illustrated in Figure 2B. Figure 7 represents CCD images of the luminomagnetic microbeads gathered on the electrode surface using a magnet at various flow rates. At a lower flow rate, the luminomagnetic microbeads accumulated on a part of the electrode close to the inlet (Figure 7Ab), whereas at a higher flow rate, the microbeads comparatively disseminated over the entire surfaces of the

electrode (Figure 7Ad). In order to examine the influence of the magnetic force on the distribution of the luminomagnetic microbeads, the luminomagnetic microbeads on the electrode were monitored with and without the magnet on the backside of the electrode after the flow was stopped (Figure 7B). The images appeared to be virtually the same, indicating no magnet effect upon the distribution. This result suggests that, since the solution flows at right angle with the magnetic force, construction of the towerlike structure for microbeads on the electrode surface might be difficult. In addition, the ECL measurements with CVs were performed in a 0.2 M phosphate buffer solution containing 100 mM TPA, 7.7 mM NaN3, 0.05% Triton X-100, and 0.05% Tween

6452 Langmuir, Vol. 23, No. 11, 2007

Komori et al.

to 4.5 mL min-1, whereas at flow rates higher than 8.4 mL min-1, the ECL intensity was virtually constant. At a lower flow rate, the ECL is likely blocked by aggregation and accumulation of the microbeads as described above, whereas at a higher flow rate, it would be enhanced due to disseminating of the beads, as is likely in the case of the batch system.

Conclusions

Figure 8. Flow rate dependences upon ECL intensity of luminomagnetic microbead-modified Pt electrode (7.8 × 105 beads cm-2) with a magnet (a) and without a magnet (b) at the backside of the electrode in 0.2 M phosphate buffer solution (pH 7.5) containing 100 mM TPA, 7.7 mM NaN3, 0.05% Triton-X 100, and 0.05% Tween 20.

20 after the flow was stopped. The curves of ECL emission and current vs potential showed a similar tendency to those of the batch system. However, the ECL intensities of luminomagnetic microbeads upon switching the magnet on and off on the backside of the electrode remained virtually constant (Figure 8). Thus, unlike the batch system, the ECL measurement can be performed at the condition that the magnet is placed at backside of the electrode in the flow injection system. Plots of the ECL intensity vs flow rates were found to be linear over the range from 0.80

The ECL intensity of luminomagnetic microbeads covering the electrode surface was found to be affected by geometrical distribution of the luminomagnetic microbeads on the surfaces of an electrode for both batch and flow injection systems. The ECL intensity increased with disseminating the luminomagnetic microbeads. The ECL peak signal in the present detection system appeared at +0.90 V vs Ag/AgCl, where [Ru(bpy)3]2+ is hardly oxidized, in a phosphate buffer solution containing NaN3. It was found that NaN3, which is used as a preservative for a test solution, enhances ECL intensity. Acknowledgment. The authors are grateful to Dr. J. Premkumar for useful discussion and S. Ishioka for his help with construction of the ECL measurement system and preparation of luminomagnetic microbeads. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan nos. 11555228 (N.O.), 13022214 (N.O.), and 13555235 (O.H.). LA063120E