• Please log in or register to access this feature.

SEARCH

SEARCH BY CITATION

Keywords:

  • enzymes;
  • carbon nanotubes (CNTs);
  • biofuel cells;
  • glucose oxidase;
  • self-powered devices

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Molecularly ordered composites of polyvinylimidazole-[Os(bipyridine)2Cl] (PVI-[Os(bpy)2Cl]) and glucose oxidase (GOD) are assembled inside a film of aligned carbon nanotubes. The structure of the prepared GOD/PVI-[Os(bpy)2Cl]/CNT composite film is entirely uniform and stable; more than 90% bioelectrocatalytic activity could be maintained even after storage for 6 d. Owing to the ideal positional relationship achieved between enzyme, mediator, and electrode, the prepared film shows a high bioelectrocatalytic activity for glucose oxidation (ca. 15 mA cm−2 at 25 °C) with an extremely high electron-transfer turnover rate (ca. 650 s−1) comparable to the value for GOD solutions, indicating almost every enzyme molecule entrapped within the ensemble (ca. 3 × 1012 enzymes in a 1 mm × 1 mm film) can work to the fullest extent. This free-standing, flexible composite film can be used by winding on a needle device; as an example, a self-powered sugar monitor is demonstrated.


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Controlling the electrical contact of redox enzymes with electrodes is a critical issue for enzymatic biodevices such as biofuel cells and biosensors.1–12 The mutual positioning between enzyme molecules, mediator molecules (not always necessary), and electrode surface determines the efficiency, reproducibility, and stability of the bioelectrocatalysis systems. A conventional engineering for accelerating the electron transfer to the redox enzymes is inclusive immobilization with mediator polymer matrices, in which the successive electron exchange between the neighboring mediator groups connects the enzyme redox center and the electrode surface.10–12 For example, Os-complex-pendant polymers are successful mediating matrices for glucose oxidase (GOD) and can provide glucose oxidation current at a mA cm−2 level.13–16 Barton et al. reported ca. 20 mA cm−2 using GOD, an Os-complex polymer, and a carbon nanotube (CNT)-modified electrode.16 However, because of random mutual positioning in the 3D composite, many enzyme molecules are isolated from the molecular network for continuous bioelectrocatalysis. On the other hand, direct immobilization of an enzyme monolayer on the electrode surface has improved the efficiency of enzyme utilization. A striking example is the reconstituting apo-GOD on a relay-FAD monolayer linked to electrode surfaces.2, 17–20 Since all the enzyme units are oriented in an optimal position with respect to the electrode surface, a high electron-transfer turnover rate comparable to that for bulk GOD reaction (approximately 700 s−1 at 25 °C) has been achieved. However, the drawback of such 2D monolayer engineering is the lower bioelectrocatalytic performance due to the limited amount of immobilized enzymes.

We present herein an enzyme/mediator/electrode ordered ensemble that shows both “high turnover rate” and “large catalytic current”. In order to satisfy both of these requirements, the larger amount of enzymes than monolayer should be immobilized while keeping effective contact with electrodes. We realize such ideal conditions by taking advantage of a film of well-aligned carbon nanotube forest (CNTF)21 consisting of single-walled CNTs arrayed with a pitch of 16 nm. The CNTF was synthesized by water-assisted chemical vapor deposition on a line-patterned Al2O3/Fe catalyst on silicon wafers (see the Experimental Section for details).21 As shown in Figure 1a, the synthesized CNTF film (1.5 mm × 1 mm) was pulled from the substrate and pinched by inverse operating tweezers (electrical lead), to produce an exposed electrode geometric area of ca. 2 mm2 (sum of both faces of a 1 mm × 1 mm sheet). The thickness of the CNTF films (4, 12, or 20μm) was determined by the width of the line-patterns of Al2O3/Fe catalyst. Recently, we reported that the intraspace of the CNTF is useful for immobilization of fructose dehydrogenase and laccase, which are the direct electron transfer (DET)-type enzymes.22 Although there are a few recent reports that also GOD is capable of direct communication with electrodes,23–25 our repeated attempts to prepare a workable GOD/CNTF ensemble electrode without any mediators have failed. Therefore we developed a stepwise process to construct the molecular architecture with polyvinylimidazole-[Os(bipyridine)2Cl] (PVI-[Os(bpy)2Cl]; MW: 15 000) and GOD (EC:1.1.3.4; MW: 186 kDa), as illustrated in Figure 1b and 1c. The PVI-[Os(bpy)2Cl] was synthesized according to a literature method,26 with a molar ratio of imidazole group to [Os(bpy)2Cl] of 5.

Figure 1. Schematic illustration of the stepwise process for constructing bioelectrocatalytic composite inside a CNTF film.

Download figure to PowerPoint

thumbnail image

2. Results and Discusssion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

2.1. Adsorption of PVI-[Os(bpy)2Cl] inside CNTF Films

The CNTF film was first treated with 0.1% Triton X-100 to make it hydrophilic, and then soaked in a stirred phosphate buffer solution (PBS, pH 7.0) containing 1 mg mL−1 PVI-[Os(bpy)2Cl] at 4 °C. As shown in Figure 2a, the cyclic voltammogram (CV) of the treated CNTF showed a symmetric shape typical for adsorbed redox species.27 In fact, the amplitude of peak currents were proportional to the scan rates (Figure 2b). The amount of PVI-[Os(bpy)2Cl] adsorbed within the CNTF films were estimated by integrating the CV currents and is plotted in Figure 2c against the soaking time in the 1 mg mL−1 PVI-[Os(bpy)2Cl] solution. The amount of PVI-[Os(bpy)2Cl] in a CNTF film increased with the soaking time and reached a maximum after 2 h. Importantly, these values are proportional to the CNTF film thickness (7.2 × 10−10 mol for a 12 μm thick film and 12.8 × 10−10 mol for a 20 μm thick film), indicating that the PVI-[Os(bpy)2Cl] molecules can entirely and uniformly adsorbed inside the CNTF films, as illustrated in Figure 1b. A part of the free imidazole groups of the mediator polymer would adsorb on CNT surfaces via ππ interaction.28 The adsorption density of PVI-[Os(bpy)2Cl] calculated using the effective inner surface area of the CNTF films (8.2 cm2 for 20 μm thick film)21 was (1.6 ± 0.1) × 10−10 mol cm−2, which is comparable with the value for a PVI-[Os(bpy)2Cl] film adsorbed on a flat Au surface (3.2 × 10−10 mol cm−2).29

Figure 2. a) Cyclic voltammograms at 10 mV s−1 of the PVI-[Os(bpy)2Cl]-modified CNTF film (20 μm thickness) in PBS (pH 7.0). A CNTF film was soaked in the 1 mg mL−1 PVI-[Os(bpy)2Cl] PBS for 6 h. b) Redox peak currents of the CVs as a function of scan rate. c) The amount of Os(bpy)2 unit inside a CNTF film (film thickness: 12 and 20 μm) as a function of soaking time for 1 mg mL−1 PVI-[Os(bpy)2Cl] PBS solution. The mean values (± standard deviation) of three independent specimens are given.

Download figure to PowerPoint

thumbnail image

2.2. Electrocatalytic Activity of GOD/PVI-[Os(bpy)2Cl]/CNTF Ensemble Films

Subsequent loading of the enzyme GOD was conducted by immersing the PVI-[Os(bpy)2Cl]-adsorbed CNTF films in a stirred PBS solution (pH 7.0) containing 3 mg ml−1 GOD for 1 hour. Figure 3a shows the CVs of GOD/PVI-[Os(bpy)2Cl]/CNTF ensemble films at 10 mV s−1 in a stirred 200 mmd-glucose PBS solution. The catalytic current for glucose oxidation increased in response to the thickness of CNTF films (3.7 mA cm−2 for 4 μm thickness and 14.7 mA cm−2 for 20 μm thickness), indicating that also GOD can entirely penetrate inside the PVI-[Os(bpy)2Cl]-modified CNTF films. For example, the content of GOD incorporated in a 20 μm thick film was measured as ca. 0.86 μg by a C-6667 Protein Quantitation Kit, the value being a little below the case when GOD molecules (6.7 × 6.7 × 21 nm3)30 align to form lines in the interspace of CNTs (1.17 μg). The current density under stirred condition was enhanced to as high as 26.7 mA cm−2 by turning up the buffer temperature to 37.5 °C. This glucose oxidation activity is comparable or superior to those previously reported using GOD.13–16 In a quiescent condition, the current density decreased by half, probably due to the limited mass-transfer inside the film. Importantly, more than 90% of the electrode activity could be maintained even after 6 d storage in an air-saturated PBS solution (Figure 3b), proving the stability of bioelectrocatalytic architecture with the composite of PVI-[Os(bpy)2Cl] polymer and GOD. The anionic GOD molecules could be stably entrapped by electrostatic interaction with cationic Os-complex of the mediator polymer that is anchored on the CNT surface via ππ interaction.28

Figure 3. a) Cyclic voltammograms of GOD/PVI-Os/CNTF ensemble films at 10 mV s−1 in stirred air-saturated 25 °C (or 37.5 °C) PBS containing 200 mmd-glucose. The thicknesses of CNTF films were 4 or 20 μm. The control voltammograms without GOD and glucose are also shown. We used a total geometric area of the pinched film (2 mm2) for the calculation of the current densities. b) The oxidation current densities at 0.6 V vs. Ag/AgCl for the GOD/PVI-[Os(bby)2Cl]/CNTF film (20 μm thickness) in a stirred 200 mmd-glucose PBS solution, periodically measured during 6 d of storage in PBS solution. c) The current densities at 0.6 V for the 20 μm GOD/PVI-[Os(bby)2Cl]/CNTF film as a function of the glucose concentration, measured in O2-, N2-, and air-saturated stirred PBS (pH 7.0) solutions. The mean values (± standard deviation) of three independent specimens are given.

Download figure to PowerPoint

thumbnail image

The electron-transfer turnover rate for the 20 μm thick film was calculated from the current value at 25 °C (0.29 mA), the Faraday constant (96 500 C mol−1), the molecular weight of GOD (186 000 g mol−1), and the content of GOD molecules in a piece of the ensemble film (ca. 0.86 μg). The derived averaged turnover rate was ca. 650 s−1, being comparable to that of GOD in bulk solution containing the natural electron acceptor O2 (700 s−1) at 25 °C.31 These results indicate that most of ca. 3 × 1012 GOD units within the film could efficiently work to the fullest extent, presumably owing to the molecularly ordered structure of enzyme/mediator/electrode ensemble. Such a high efficiency of the present GOD electrode resulted in a resistance to oxygen inhibition, as shown in Figure 3c. The catalytic performance was almost identical in N2-saturated, air-saturated, and even O2-saturated solutions. In general, glucose oxidation with GOD-modified electrodes is often disturbed by dissolved O2, which is troublesome for glucose sensing.32, 33 However, the ordered Os(bby)2 groups in the present ensemble electrode could effectively accept the electron from GOD in preference to O2, resulted in excellent O2 resistance.

2.3. Application as a Flexible Anode of Biofuel Cells

The present free-standing, bioelectrocatalytic film could be used for miniature biofuel cell devices. We demonstrate here the application of the film to a self-powered sugar indicator designed for inserting into a fruit. For indicating the glucose concentration, the net performance of the biofuel cell system should be controlled by the glucose anode. Because the oxygen in fruits is limited to a lower concentration than glucose, we employ a gas-diffusion biocathode34 for utilizing the abundant oxygen in air outside of the fruits (see the Experimental Section for details). Figure 4a shows the biofuel cell performance measured using 200 mM glucose PBS solution with a couple consisting of a GOD/PVI-[Os(bby)2Cl]/CNTF film anode (20 μm thickness) and a cathode made from bilirubin oxidase (BOD)-modified carbon fabric (1 cm × 1 cm). The open-circuit voltage of the cell was 0.5 V in agreement with the difference between the potentials at which glucose oxidation and oxygen reduction start to occur in cyclic voltammetry (0.1 V in Figure 3a and 0.6 V in Figure S1, see the Supporting Information). The maximum output current (0.27 mA) is almost equivalent to the maximum oxidation current at the composite anode (0.29 mA) that can be calculated by the current density in Figure 3a (14.7 mA cm−2) and the electrode area of 2 mm2. This result indicates that the system is limited by the anode even in 200 mM glucose, a concentration that is markedly higher than that found in raw fruits (a few tens of mM). As shown in Figure 4b, a piece of GOD/PVI-[Os(bby)2Cl]/CNTF film was wound on one lead of a light-emitting-diode (LED) device, whose blinking interval is inversely proportional to the power of the biofuel cell.35, 36 The other lead was connected to the BOD-based gas-diffusion cathode. The blinking interval of the LED upon inserting the device to a grape was coincident with that for the extracted juice (Figure 4c), proving that this device could serve as a sugar indicator by simply being inserted into a grape. We confirmed that there is no corrosion reaction at the LED lead wire during the operation. The present principle of the self-powered sensor could be applied to more important blood sugar monitoring applications; we are planning to develop a GOD/PVI-[Os(bby)2Cl]/CNTF-based device structure suitable for low-invasive insertion into a blood vessel through skins.

Figure 4. a) Performance of a biofuel cell composed of an anode of GOD/PVI-[Os(bby)2Cl]/CNTF film (20 μm thickness) and a cathode of BOD-modified carbon cloth (1 cm × 1 cm) in 200 mM glucose PBS solution, measured by changing the external resistance (28 to 46kΩ). b) Photograph of the LED-based self-powered sugar indicator, at the tip of which the GOD/PVI-[Os(bby)2Cl]/CNTF film was wound. c) The device assembly was inserted in a grape and the LED blinking was measured (inset). The time course of LED emission, taken using an extracted juice, is also shown.

Download figure to PowerPoint

thumbnail image

3. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

An amount of enzyme units larger than that in a monolayer were successfully immobilized while keeping effective electrical contact with the electrode (CNTs), as summarized in Figure 5. In particular, we have succeeded in forming an entirely uniform bioelectrocatalytic architecture with PVI-[Os(bby)2Cl] and GOD inside a CNTF film. The voltammograms of the PVI-[Os(bby)2Cl]-modified CNTF indicated the uniform adsorption of PVI-[Os(bpy)2Cl] on the CNT surface via ππ interaction with the density of ca. (1.6 ± 0.1) × 10−10 mol cm−2. The subsequent GOD seemed to become stably entrapped at the interspaces of PVI-[Os(bby)2Cl]-modified CNTs by the electrostatic interaction. Owing to the ordered positional relationship between GOD, PVI-[Os(bby)2Cl], and CNT, the composite film showed both high activity for glucose oxidation (ca. 15 mA cm−2) and high electron-transfer turnover rate (ca. 650 s−1), indicating almost every enzyme molecules within the film could work to the fullest extent.

Figure 5. Scheme showing correlative characters of: a) an enzyme-film with mediator polymer matrices for larger current, b) an enzyme-monolayer electrode for higher turnover rate, and, c) the present ensemble electrode for both large current and high turnover rate.

Download figure to PowerPoint

thumbnail image

4. Experimental Section

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

CNTF Preparation: CNTF was synthesized in a 1 inch tube furnace by water-assisted chemical vapor deposition at 750 °C with a C2H4 carbon source and an Al2O3 (10 nm)/Fe (1.0 nm) thin-film catalyst grown on silicon wafers.21 We used He with H2 as the carrier gas (total flow 1000 standard cubic centimeters per minute (sccm)) at 1 atm with a controlled amount of water vapor with ethylene (100 sccm) for 10 min.

Quantitative Analysis of the Entrapped Enzymes: The quantitative analysis of GOD was conducted as explained in our previous paper.22 The enzyme-incorporated CNTF film was first washed and immersed in 20 mM sodium phosphate buffer (pH 9.3) containing 0.1 M sodium borate and 1% sodium cholate and dispersed with an ultrasonic homogenizer for 15 min. The GOD in the dispersion was then analyzed using a C-6667 Protein Quantitation Kit (Molecular Probes), using 5 mM (3-(4-carboxybenzoyl)-quinoline-2-carboxaldehyde) (ATTO-TAG CBQCA) and 20 mM KCN to label the enzyme with CBQCA. After 1.5 h of incubation, the fluorescent intensity was measured by a luminescent image analyzer system (Fuji Photo Film, LAS-3000 mini), and the amount of enzyme was determined by referencing a calibration curve.

Preparation of Gas-diffusion Carbon Fabric (CF) Cathodes: The preparation of the cathode basically followed the procedures used for our previous work.34 A 40 μL aliquot of a 10 mg mL−1 multiwalled CNT solution was put on a CF strip and dried in air, followed by thoroughly washing out the surfactant by soaking in an ethanol solution for more than 1 h with stirring. The surface of the CNT-modified CF electrode was further modified with a 0.1 mL solution of 5 mg mL−1 bilirubin oxidase (BOD, EC 1.3.3.5, 2.5 U mg−1, from Myrothecium) in a vacuum oven (0.09 MPa, 35 °C). The strip was additionally coated with the CNT solution to make the surface hydrophobic.

Electrochemical Measurements: The GOD/PVI-[Os(bpy)2Cl]/CNTF ensemble films, anchored at the edge with SUS316L fine tweezers, was analyzed by a three-electrode system (BSA, 730C electrochemical analyzer) in stirred solutions using a Ag/AgCl reference and a platinum counter electrode. The gas-diffusion cathode (BOD-modified CF strip) was put on an air-saturated solution so as to contact the solution by the BOD-modified face during cyclic voltammetry (Figure S1 in the Supporting Information). The performance of a biofuel cell constructed from an GOD-based CNTF anode and an BOD-based CF cathode was evaluated on the basis of the cell voltage upon changing the external resistance between 1 kΩ and 2 MΩ at the time step of 60 s. Unless otherwise indicated, the electrochemical measurements were carried out at room temperature (25 °C).

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Authors express appreciation to Toho Tenax Co. for donation of the carbon fabrics and to Bayer Co. for multiwalled carbon nanotubes.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discusssion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
aenm_201200422_sm_suppl.pdf99Ksuppl

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.