Coadsorption of Horseradish Peroxidase with Thionine on TiO2

Langmuir 2005, 21, 8409-8413

8409

Coadsorption of Horseradish Peroxidase with Thionine on TiO2 Nanotubes for Biosensing Songqin Liu and Aicheng Chen* Department of Chemistry, Lakehead University, 955 Oliver Road, Thunder Bay, Ontario P7B 5E1, Canada Received April 3, 2005. In Final Form: June 22, 2005 In this study, we investigate the coadsorption of protein with thionine on TiO2 nanotubes for biosensor design. The TiO2 nanotube arrays fabricated by anodic oxidation of titanium substrate possess large surface areas and good uniformity and conformability and are ready for enzyme immobilization. Electrochemical and spectroscopic measurements show that the TiO2 nanotube arrays provide excellent matrixes for the coadsorption of horseradish peroxidase (HRP) and thionine and that the adsorbed HRP on these TiO2 nanotube arrays effectively retains its bioactivity. The immobilized thionine can be electrochemically reduced but cannot be reoxidized in the electrode potential range between -0.7 and 0.0 V. The addition of H2O2 leads to the biocatalytic oxidation of the reduced thionine in the presence of HRP, resulting in developing a novel H2O2 sensor with good stability and reproducibility. The fabricated TiO2 nanotubes offer a stage for further study of immobilization and electrochemistry of proteins. The proposed method opens a way to develop biosensors using nanostructured materials with low electrical conductivity.

1. Introduction The immobilization of proteins on solid substrates for biosensing is a promising direction given the rapid advances in biotechnology. After redox proteins were immobilized on a biocompatible electrode surface, they show enhanced electrochemical activity, allowing the electrochemical measurements of their substrate with higher sensitivity and better selectivity.1-3 However, the direct adsorption of redox proteins on unmodified metal surfaces usually leads to a dramatic change in the protein structures and significant loss of their bioactivity.4-6 Many efforts have been made to improve the interfacial properties by electrode7-11 and protein12 modifications. Among these, nanostructured materials such as calcium phosphate,13 colloidal gold,14-18 montmorillonite,19 clay,20 and NaY molecular sieves21,22 have been identified as very * To whom correspondence should be addressed. E-mail: [email protected]. (1) Rusling, J. F. Acc. Chem. Res. 1998, 31, 363-369. (2) Ciureanu, M.; Goldstein, S.; Mateescu, M. A. J. Electrochem. Soc. 1998, 145, 533-541. (3) Chen, S. M.; Tseng, C. C. Electrochim. Acta 2004, 49, 1903-1914. (4) Holt, R. E.; Cotton, T. M. J. Am. Chem. Soc. 1989, 111, 28152821. (5) Yang, M.; Chung, F. L.; Thompson, M. Anal. Chem. 1993, 65, 3713-3716. (6) Jackson, D. R.; Omanovic, S.; Roscoe, S. G. Langmuir 2000, 16, 5449-5457. (7) Chattopadhyay, K.; Mazumdar, S. Bioelectrochem. Bioenerg. 2001, 53, 17-24. (8) Armstrong, F. A.; Hill, H. A. O.; Walton, N. J. Acc. Chem. Res. 1988, 21, 407-413. (9) Sevilla, J. M.; Pineda, T.; Roma´n, A. J.; Maduenˇo R.; Bla´zquez, M. J. Electroanal. Chem. 1998, 451, 89-93. (10) Xiao, Z.; Lavery, M. J.; Bond, A. M.; Wedd, A. G. Electrochem. Commun. 1999, 1, 309-314. (11) Lisdat, F.; Ge, B.; Scheller, F. W. Electrochem. Commun. 1999, 1, 65-68. (12) Heller, A. Acc. Chem. Res. 1990, 23, 128-134. (13) Cheng, X. L.; Filiaggi, M.; Roscoe, S. G. Biomaterials 2004, 25, 5395-5403. (14) Brown, K. R.; Fox, A. P.; Natan, M. J. J. Am. Chem. Soc. 1996, 118, 1154-1157. (15) Xiao, Y.; Ju, H. X.; Chen, H. Y. Anal. Biochem. 2000, 278, 2228. (16) Shipway, A. N.; Lahav, M.; Willner, I. Adv. Meter. 2000, 12, 993-998. (17) Liu, S. Q.; Ju, H. X. Anal. Biochem. 2002, 307, 110-116. (18) Liu, S. Q.; Ju, H. X. Analyst 2003, 128, 1420-1424. (19) Lei, C.; Deng, J. Anal. Chem. 1996, 68, 3344-3349.

promising immobilization matrixes because of their high active surface areas for protein binding, regular structures, and good mechanical, thermal, and chemical stability. More recently, Durrant et al.23-25 report that nanoporous TiO2 and ZnO films prepared by screen-printing technologies significantly increase the active surface area by a factor of 150 for adsorption of cytochrome c and hemoglobin. However, their cyclic voltammetric study of the adsorbed cytochrome c and hemoglobin shows an irreversible electrode reaction: a common stumbling block when using these materials, only reduction peaks of the immobilized enzymes are observed, and no oxidative peaks can be seen in the reverse potential scan because of the low electrical conductivity of the nanoporous TiO2 and ZnO films. Although reoxidation of the adsorbed cytochrome c could be achieved by the prolonged application of a positive potential, the electrochemical reoxidation is very slow. This inhibits biosensor applications of nanoporous TiO2 and ZnO2. On the other hand, previous studies show that reduced horseradish peroxidase (HRP) can be reoxidized chemically by hydrogen peroxide (H2O2).26-28 Here, we report on the development of a novel H2O2 electrochemical biosensor involving the coadsorption of HRP and thionine on TiO2 nanotube arrays synthesized by anodic oxidation of Ti substrate. There is growing interest in developing new advanced materials and designing novel devices with control features on a nanometer scale. Nanostructured TiO2-based materials have been receiving significant attention.29-32 Previous studies have shown that nanostructured tita(20) Lei, C.; Lisdat, F.; Wollenberger, U.; Scheller, F. W. Electroanalysis 1999, 11, 274-276. (21) Liu, B.; Hu, R.; Deng, J. Anal. Chem. 1997, 69, 2343-2346. (22) Liu, B.; Yan, F.; Kong, J.; Deng J. Anal. Chim. Acta 1999, 386, 31-39. (23) Topoglidis, E.; Cass, A. E. G.; Gilardi, G.; Sadeghi, S.; Beaumont, N.; Durrant, J. R. Anal. Chem. 1998, 70, 5111-5113. (24) Topoglidis, E.; Campbell, C. J.; Cass, A. E. G.; Durrant, J. R. Langmuir 2001, 17, 7899-7906. (25) Topoglidis, E.; Cass, A. E. G.; O’Regan, B.; Durrant, J. R. J. Electroanal. Chem. 2001, 517, 20-27. (26) Ruan, C.; Yang, F.; Lei, C.; Deng, J. Anal. Chem. 1998, 70, 17211725. (27) Ruan, C.; Yang, R.; Chen, X.; Deng, J. J. Electroanal. Chem. 1998, 455, 121-125. (28) Wollenberger, U.; Drungiliene, A.; Sto¨cklein, W.; Kulys, J. J.; Scheller, F. W. Anal. Chim. Acta 1996, 329, 231-237.

10.1021/la050875x CCC: $30.25 © 2005 American Chemical Society Published on Web 07/29/2005

8410

Langmuir, Vol. 21, No. 18, 2005

nium oxide films can be fabricated by anodizing titanium sheets in a hydrofluoric acid aqueous solution.33-41 The formed TiO2 films are optically transparent, and their surfaces are negatively charged at pH values higher than 6.0.42 Because of nontoxicity, nanostructured TiO2 materials show high biocompatibility and good retention of biological activity for protein binding.23,42 It is also known that dyes are easily adsorbed from the solution to the surface of nanoporous TiO2 films.43 This is attributed to the chelate action between the carboxylate groups of the dye and Ti4+ centers of the titanium oxide surface. In addition, TiO2 nanotubes fabricated by low-cost anodic oxidation of the Ti substrate possess large surface areas and good uniformity and conformability over large areas, desirable for electrochemical biosensor design.33,42 In this work, we first fabricated TiO2 nanotubes by anodizing Ti sheets in a dilute HF solution and then co-immobilized HRP and thionine on the TiO2 nanotube arrays by immersing the Ti/TiO2 electrodes in a mixture of HRP and thionine solution. Electrochemical and spectroscopic measurements show that the TiO2 nanotube arrays provide excellent matrixes for the co-immobilization of thionine and HRP. The immobilized HRP not only retains its bioactivity but also shows a high affinity for H2O2 reduction, resulting in a novel H2O2 sensor. 2. Experimental Procedures 2.1. Chemicals. Horseradish peroxidase (EC 1.11.1.7), 3,3′,5,5′tetramethylbenzidine dihydrochloride, and thionine acetate were purchased from Sigma and used as received. Hydrofluoric acid (49%) came from Fisher Chemical. All other chemicals were of reagent grade and were used as supplied. Water purified by a Nanopure water system was used to prepare all solutions. Measurements were performed in 0.1 M phosphate buffer solution (PB) with different selected pH values obtained with K2HPO4 and KH2PO4. 2.2. Preparation of Biosensors. Grade 1 titanium foil of 0.80 mm thickness was used as the electrode substrate. Prior to anodization, the titanium foils were ultrasonically cleaned in distilled water and then in acetone for 20 min in each. After rinsing thoroughly with pure water, a clean titanium foil electrode (3 × 10 mm, geometrical area ca. 0.6 cm2) was first etched in an 18% HCl solution at 85 °C for 10 min. The etched titanium foil was rinsed thoroughly with pure water and then anodized in an 1:8 acetic acid-water solution containing 1.0 vol % hydrofluoric acid at 20 V for 45 min, forming TiO2 nanotube arrays on the Ti substrate. After rinsing thoroughly with water, the resulting nanotube titanium dioxide electrodes were immersed in 10 µM thionine solution in 5 mM PB at pH 7.0 containing 2 mU mL-1 of HRP for over 5 h to produce the thionine/HRP-modified TiO2 (29) Tian, Z. R. R.; Voigt, J. A.; Liu, J.; Mckenzie, B.; Xu, H. F. J. Am. Chem. Soc. 2003, 125, 12384-12385. (30) Yoo, S.; Akbar, S. A.; Sandhage, K. H. Adv. Mater. 2004, 16, 260-264. (31) Peng, X.; Chen, A. J. Mater. Chem. 2004, 14, 2542-2548. (32) Peng, X.; Chen, A. Appl. Phys. A 2005, 80, 473-476. (33) Gong, D.; Grimes, C. A.; Varghese, O. K.; Hu, W.; Singh, R. S.; Chen, Z.; Dickey, E. C. J. Mater. Res. 2001, 16, 3331-3334. (34) Zwilling, V.; Darque-Ceretti, E.; Boutry-Forveille, A.; David, D.; Perrin, M. Y.; Aucouturier, M. Surf. Interface Anal. 1999, 27, 629-637. (35) Varghese, O. K.; Gong, D.; Paulose, M.; Grimes, C. A.; Dickey, E. C. J. Mater. Res. 2003, 18, 156-165. (36) Mor, G. K.; Varghese, O. K.; Paulose, M.; Mukherjee, N.; Grimes, C. A. J. Mater. Res. 2003, 18, 2588-2593. (37) Chen, Q.; Zhou, W.; Du, G.; Peng, L. Adv. Mater. 2002, 14, 12081211. (38) Zhang, M.; Bando, Y.; Wada, K. J. Mater. Sci. Lett. 2001, 20, 167-170. (39) Kasuga, T.; Hiramatsu, M.; Hoson, A.; Sekino, T.; Niihara, K. Adv. Mater. 1999, 11, 1307-1311. (40) Imai, H.; Takei, Y.; Shimizu, K.; Hirashima, H. J. Mater. Chem. 1999, 9, 2971-2972. (41) Hoyer, P. Langmuir 1996, 12, 1411-1413. (42) Zhang, J. K.; Cass, A. E. G. Anal. Biochem. 2001, 292, 307-310. (43) O’Regan, B.; Grtzel, M. Nature 1991, 353, 737-740.

Liu and Chen

Figure 1. SEM image of TiO2 nanotubes prepared by anodic oxidation of Ti substrate in an acetic acid solution containing 1.0 vol % HF at 20 V for 45 min. nanotube electrode (Th/HRP/TiO2). For comparison, TiO2 nanotube electrodes were immersed in (a) 5 mM PB at pH 7.0, (b) 2 mU mL-1 of HRP in 5 mM PB at pH 7.0, and (c) 10 µM thionine solution in 5 mM PB at pH 7.0 for 5 h to generate the bare, HRP, and thionine-modified TiO2 nanotube electrodes (TiO2, HRP/TiO2, and Th/TiO2), respectively. The resulting electrodes were stored in 5 mM PB at pH 7.0 and 4 °C. 2.3. Electrochemical Measurements. Electrochemical measurements were performed with a Solartron 1287 potentiostat and a Solartron 1252B frequency response analyzer. The threeelectrode system was used for all electrochemical experiments, employing one of the above electrodes, platinum wire, and an Ag/AgCl electrode (3 M KCl) as the working, auxiliary, and reference electrodes, respectively. All experiments were performed at room temperature (22 ( 2 °C) with 0.1 M PB as the background electrolyte. All experimental solutions were deoxygenated by bubbling ultrapure argon for 15 min and maintained under an argon atmosphere during the course of the experiments. The AC impedance experiment was carried out with frequencies ranging from 40 kHz to 40 mHz. Amperometric experiments were performed in a stirred cell with the successive addition of H2O2 into 8.0 mL of supporting solution, while the electrode potential was set at -450 mV. 2.4. Enzyme Loading and Photometric Activity Assay. Photometric measurements were performed with a Cary 5E UVvis-NIR spectrophotometer. For measurement of the HRP loading, a titanium foil with an area of 1 × 1 cm2 was used as the electrode and the reflectance spectra was recorded. To determine the HRP activity,44 the enzyme-loaded electrode was dipped into 1 mL of 0.1 M PB at pH 5.0 containing 0.1 mg/mL TMB and 2 mM H2O2 for 20 min and the absorbance at 647 nm was measured. HRP-free solutions were used as the reference. 2.5. Scanning Electron Microscopic (SEM) Analysis. The morphologies of the TiO2 nanotube electrodes were characterized using SEM analysis (JEOL JSM 5900 LV) at an acceleration voltage of 13 kV. A gold film was applied by argon plasma sputtering for 45 s to the specimens for SEM measurements.

3. Results and Discussion 3.1. Preparation of Titania Nanotubes. The adsorption of dye and protein on nanostructured materials is expected to depend upon the morphology of the substrate. Thus, the morphology of the specific nanomaterials is a vital factor affecting the development and performance of biosensors. Figure 1 shows an SEM image of the titanium oxide nanotubes fabricated by anodic oxidation of a Ti substrate at 20 V for 45 min in an 1:8 acetic acid-water solution containing 1.0 vol % HF. The average pore diameter as estimated from the SEM image (44) Stich, T. M. Anal. Biochem. 1990, 191, 343-346.

Coadsorption on TiO2 Nanotubes for Biosensing

Figure 2. Cyclic voltammograms of (a) bare TiO2, (b) Th/TiO2, and (c) Th/HRP/TiO2 in 0.1 M PB at pH 6.8 and 50 mV/s. (Inset) Plot of reduction peak potential of Th/TiO2 versus pH.

is 80 nm with a standard deviation of 13 nm, and the wall thickness is 29 nm with a standard deviation of 10 nm. These results are in good agreement with Grimes’s work.33,35,36 When the applied anodization potential is lower than 10 V (for instance, between 5 and 10 V) or the HF concentration is lower than 0.5% (for instance between 0.2 and 0.5%), our SEM study shows that the formed TiO2 thin film has a homogeneous structure with a narrow particle-size distribution. Without the etching procedure or in the absence of HF in the electrolyte, an inhomogeneous, uneven nanoporous TiO2 film was produced. Further study showed that the electrochemical response of adsorbed thionine on the inhomogeneous TiO2 film is poor. In contrast, a well-defined reduction peak of the adsorbed thionine on both the uniform nanotubes and the homogeneous narrow particle TiO2 surface was observed. Thus, only the results using the TiO2 nanotube electrodes are presented in the following sections. 3.2. Electrochemical Characteristics of Adsorbed Thionine. Figure 2 shows the cyclic voltammograms of different electrodes in 0.1 M PB at pH 6.8 in the absence of dissolved thionine at 50 mV/s. A well-defined reduction peak appears at -0.477 V for the Th/TiO2 electrode during the first cycle when scanning the electrode potential from 0 to -0.7 V (curve b in Figure 2), while no peak is observed for both the bare Ti foil and TiO2 nanotube electrode (curve a in Figure 2) in the same potential range. The response of the Th/TiO2 electrode is therefore attributed to the reduction of the adsorbed thionine on the electrode surface. No oxidative peaks are seen in the reverse scans, and no peaks in the following scans can be clearly resolved for Th/TiO2, indicating that the immobilized thionine on the TiO2 nanotube electrode is changed completely to its reductive form during the first scan. In addition, the peak potential is more negative than the reduction peak of free thionine in solution (E0′ ) -0.12 V versus SCE at pH ) 7.0) on a gold electrode.26 Control experiments were also performed in 0.1 M PB at pH 6.5 containing 0.1 mM thionine using the TiO2 nanotube electrode and a Pt electrode. When cycling the electrode potential in the range between -0.7 and 0.0 V versus Ag/AgCl, a reduction peak is observed at -0.536 V, but no oxidation peak is seen during the reversed scan. In contrast, a pair of large and well-defined oxidation/reduction peaks centered at -0.153/-0.235 V appear in the CV curve of the Pt electrode. The lack of the oxidation peaks and the negative shift of the reduction potentials for the adsorbed thionine are thus attributed to the low electrical conductivity of

Langmuir, Vol. 21, No. 18, 2005 8411

Figure 3. Electrochemical impedance spectra of (a) bare Ti, (b) TiO2 nanotube, (c) Th/TiO2, and (d) Th/HRP/TiO2 recorded at the open circuit potential in 0.1 M PB at pH 6.6 containing 30 mM K3Fe(CN)6 and 10 mM K4Fe(CN)6.

the TiO2 nanotube electrodes. This is consistent with the electrochemical and spectroelectrochemical studies of immobilized cytochrome c on nanoporous TiO2 films by Topoglidis et al.24,25 Our further study shows that the electrode potential of the reduction peak strongly depends on the pH of the electrolytes. The inset of Figure 2 presents a plot of the electrode potential of the reduction peak versus pH. Increasing the pH of the electrolyte (from 5.0 to 9.0) results in a linear negative shift of the reduction peak potential with a slope of -(51.1 ( 2.7) mV pH-1. This value was close to the expected value (-58.0 mV pH-1) for an electrochemical reduction involving a single proton transfer and one electron transfer.45 The reduction of Th can be described as follows

3.3. Coadsorption of HRP and Thionine on the TiO2 Nanotube Arrays. Similar to the Th/TiO2, a reduction peak is observed for the Th/HRP/TiO2 electrode during the first cycle in 0.1 M PB at pH 6.8 (curve c in Figure 2). In contrast, no peak is seen at the HRP/TiO2 electrode in the same potential range. The response of Th/HRP/TiO2 is therefore attributed to the reduction of the adsorbed thionine. However, the peak current is smaller and the peak potential has a positive shift (∼30 mV) compared to the reduction peak of Th/TiO2. These results indicate that HRP and thionine are co-immobilized on the TiO2 nanotube arrays and that the adsorption of HRP decreases the thionine load on the electrode surface. The coadsorption of HRP and thionine on the TiO2 nanotube arrays is demonstrated by the AC impedance measurements. Figure 3 shows the AC impedance spectra for (a) titanium foil etched in HCl, (b) TiO2, (c) Th/TiO2, and (d) Th/HRP/TiO2, recorded at the open circuit potential in 0.1 M PB at pH 6.6 containing 30 mM potassium ferricyanide(III) and 10 mM potassium hexacyanoferrate(II) trihydrate. The redox couple Fe(CN)63-/4- is widely used as an electrochemical probe in the electrochemical impedance study, especially for the characterization of (45) Clavilier, J.; Svetlicic, V.; Zutic, V. J. Electroanal. Chem. 1995, 386, 157-163.

8412

Langmuir, Vol. 21, No. 18, 2005

Figure 4. UV-vis reflection spectra of (a) Th/TiO2 and (b) Th/HRP/TiO2. (Inset) UV-vis spectra of (c) 9 µM thionine in 0.1 M PB at pH 6.5, (d) 12 mU mL-1 HRP in 0.1 M PB at pH 6.5, and (e) 6.5 µM thionine, 2.6 mU mL-1 HRP, and 0.1 M PB at pH 6.5.

biomaterial-modified electrodes.46,47 An equivalent circuit Rs(RctCPE) was used to model the impedance data, thus enabling the extraction of electrical parameters, such as resistance, from the impedance spectra.48 In this circuit, Rs represents the uncompensated solution resistance, while the parallel combination of the charge-transfer resistance (Rct) and the constant phase element (CPE) leads to a depressed semicircle in the corresponding Nyquist impedance plot. The CPE is defined by CPE-T and CPE-P. If CPE-P equals 1, then the CPE is considered a capacitor Cdl. The charge-transfer resistance (Rct) changes in the following order: Ti (6251 Ω) , TiO2 (52556 Ω) < Th/TiO2 (82823 Ω) < Th/HRP/TiO2 (92629 Ω). These results show that (i) the conductivity of the TiO2 nanotubes is much lower than that of the etched Ti foil and (ii) adsorption of thionine and coadsorption of HRP and thionine further increase the charge-transfer resistance of the TiO2 nanotubes. The increase in charge-transfer resistance of Th/HRP/TiO2 compared to Th/TiO2 demonstrates the coadsorption of HRP and thionine on the TiO2 surface. In addition, the adsorption of thionine and coadsorption of HRP and thionine decrease the capacitance (CPE-T) of the TiO2 nanotubes. The coadsorption of HRP and thionine on the TiO2 nanotube electrode is further supported by the spectroscopic measurements. Figure 4 shows the absorption spectra of both the Th/TiO2 (curve a) and Th/HRP/TiO2 (curve b) electrodes. An absorbance at 599 nm is observed for the Th/TiO2 electrode, which is due to the absorbance of thionine. This is very close to 601 nm for the absorbance of thionine in solution (curve c in the inset of Figure 4). The absorbance spectrum of the Th-HRP/TiO2 electrode surface shows, in addition to the absorbance observed for Th/TiO2, a new absorbance at 403 nm. This additional adsorption is assigned to the Soret band of the heme of HRP. The absorbance of the immobilized HRP (403 nm) is very close to the absorbance of the free HRP in solution (405 nm) (curve d in the inset of Figure 4), indicating that (46) Zayats, M.; Katz, E.; Willner, I. J. Am. Chem. Soc. 2002, 124, 14724-14735. (47) Katz, E.; Alfonta, L.; Willner, I. Sens. Actuators, B 2001, 76, 134-141. (48) Chen, A.; La Russa, D. J.; Miller, B. Langmuir 2004, 20, 96959702.

Liu and Chen

Figure 5. Cyclic voltammograms of Th/HRP/TiO2 and Th/ TiO2 (inset) in 0.1 M PB at pH 6.8 (a and c) before and (b and d) after the addition of 0.2 mM H2O2.

Figure 6. (A) Amperometric response of (a) Th/HRP/TiO2 and (b) Th/TiO2 at -450 mV upon successive additions of 0.134 mM H2O2 into 0.1 M PB at pH 6.8. (B) Plot of the reduction current versus the H2O2 concentration. (Inset) Linear calibration curve.

the immobilized HRP retains its native structure.49 This is desirable for electrochemical biosensor applications. In addition, the activity of immobilized HRP on the TiO2 nanotube arrays was determined with a photometric activity assay using tetramethyl benzidine as the sub(49) Nassar, A.-E. F.; Willis, W. S.; Rusling, J. F. Anal. Chem. 1995, 67, 2386-2392.

Coadsorption on TiO2 Nanotubes for Biosensing

Langmuir, Vol. 21, No. 18, 2005 8413

strate. This measurement reveals that 0.09 nU cm-2 of HRP was adsorbed on the electrode surface. All of these results show that HRP can be effectively immobilized on TiO2 nanotube arrays by a coadsorption procedure and that the immobilized HRP on the TiO2 nanotubes effectively retains its bioactivity. The coadsorption of HRP and thionine is further demonstrated by their biocatalytic activity. 3.4. Th/HRP/TiO2 as a H2O2 Sensor. Figure 5 presents the CVs of Th/HRP/TiO2 recorded in 0.1 M PB at pH 7.0 in the absence of H2O2 (a) and in the presence of 0.2 mM H2O2 (b) at 50 mV/s. Th/HRP was immobilized on the electrode surface, and there is no Th and HRP in the presence of the solutions. The CV curves presented in Figure 5 thus represent the response of the Th/HRP layer on the surface. A dramatic increase of the reduction current is observed in curve b compared to curve a. In contrast, the reduction current of both Th/TiO2 and HRP/TiO2 electrodes (curves c and d in the inset of Figure 5, respectively) only slightly increases after injecting 0.2 mM H2O2 into the electrochemical cell. Thus, the large increase of reduction current for the Th/HRP/TiO2 electrode shows a high electrocatalytic activity of the immobilized HRP for thionine oxidation in the presence of H2O2. Both immobilized HRP and thionine are necessary to maintain the cycle of the electrocatalytic reactions. The response mechanism of the Th/HRP/TiO2 electrode to H2O2 is outlined as follow:26-28

HRP(red) + H2O2 f compound I + H2O

(2)

compound I + thionine(red) f compound II + thionine(ox)* (3) compound II + thionine(ox)* + 2H+ f HRP(red) + thionine(ox) + H2O (4) thionine(ox) + 2e- + 2H+ f thionine(red)

(5)

where the thionine(ox)* represents the free radical formed during the reaction. First, the immobilized HRP(red) chemically reduces the H2O2 to form water and oxidized HRP denoted as HRP(ox). Then, the HRP(ox) chemically oxidizes the reduced Th(red) to form HRP(red) and Th(ox). Finally, the Th(ox) is again electrochemically reduced to form Th(red), resulting in the reduction current in the CV curves. In addition, the reduction current of Th/HRP/TiO2 increases with an increasing H2O2 concentration. Figure

6A presents the chronoamperometric response of the Th/ HRP/TiO2 (a) and Th/TiO2 (b) electrode with successive additions of H2O2 into 0.1 M PB at pH 6.8, recorded under a stirred system at the potential of -0.45 V. The sensor shows an increasing amperometric response to H2O2 from 1.1 × 10-5 to 2 × 10-3 M (Figure 6B). The linear response range of the sensor to the H2O2 concentration is from 1.1 × 10-5 to 1.1 × 10-3 M as seen in the inset of Figure 6B. The linear regression equation is log I (µA) ) 3.679 + 0.736 log c (M), with a correlation coefficient of 0.9987. The detection limit is estimated as 1.2 × 10-6 M at 3σ based on the slope. When the Th/HRP/TiO2 electrode was not in use, it was stored in PB in a refrigerator at 4 °C. It retained 86% of its initial current response after 2 weeks of storage, showing good storage stability. The relative standard deviation is 8.4% for six successive determinations at an H2O2 concentration of 0.2 mM. The fabrication of six electrodes, made independently with the Ti foils, shows an acceptable reproducibility with a relative standard deviation of 12.2% for the current determined at 0.2 mM H2O2. 4. Summary In this study, we have demonstrated that HRP can be effectively immobilized on TiO2 nanotube arrays by a coadsorption procedure. The TiO2 nanotube arrays fabricated by anodic oxidation of titanium substrate possess large surface areas and good uniformity and conformability over large areas and are ready for enzyme immobilization. Our electrochemical and spectroscopic measurements show that the TiO2 nanotube arrays provide excellent matrixes for the co-immobilization of thionine and HRP and that the immobilized HRP on the TiO2 nanotubes effectively retains its bioactivity. The immobilized thionine can be electrochemically reduced but cannot be reoxidized in the electrode potential between -0.7 and 0.0 V in the absence of HRP and/or H2O2. The addition of H2O2 results in biocatalytic oxidation of the reduced thionine in the presence of HRP. This allowed us to develop a novel H2O2 sensor with a calibration ranging from 1.1 × 10-5 to 2 × 10-3 M and a detection limit of 1.2 × 10-6 M at 3σ. Acknowledgment. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (NSERC). We thank Dr. Xinsheng Peng for valuable discussion. A. C. acknowledges the CFI New Opportunities Award. LA050875X