Peroxidase-modified cup-stacked carbon nanofiber networks for electrochemical biosensing with adjustable dynamic range

Abstract

Cup-stacked carbon nanofibers (CSCNFs) were treated with ozone to introduce oxygen-containing groups to the edges of the graphene cups. The O/C atomic ratio at the surface was estimated to be 5.3 × 10⁻². The hydrophilic CSCNFs were used for constructing a three-dimensional network that works both as an electrical nanowire and an enzyme support. Horseradish peroxidase (HRP) molecules were immobilized onto the network for amperometric sensing of H₂O₂ and inhibition-based sensing of cyanide. Both the upper sensing limit of H₂O₂ and the upper and lower sensing limits of cyanide were controlled by about two orders of magnitude by changing the HRP coverage.

---

Introduction

Carbon nanotubes (CNTs) and carbon nanofibers (CNFs) have attracted significant attention as novel electric nanowires and electrode materials for electrochemical biosensors because of their high electronic conductivity and minuscule size.^1–3 They are classified into two groups:^4,5 (i) nanotubes, such as single-walled carbon nanotubes (SWCNTs) and multi-walled carbon nanotubes (MWCNTs), and (ii) nanofibers with stacked graphene, such as platelet carbon nanofibers (PCNFs), herringbone carbon nanofibers (HBCNFs), and cup-stacked carbon nanofibers (CSCNFs). In recent years, it has been reported that the graphene edge of CNTs and CNFs plays an important role in fast electron transfer reactions with redox species, especially redox proteins.

In relation to this, it is well known that heterogeneous electron transfer reactions at edge-oriented pyrolytic graphite (EOPG) electrodes are generally faster than those at basal plane pyrolytic graphite (BPG) electrodes.^6–9 In particular, direct electrochemistry at EOPG of peroxidase,^10,11 hemoglobin,¹² glucose oxidase,¹³ and nitrite reductase¹⁴ have been studied and applied to electrochemical biosensing. The edges of CNTs also allow fast electron exchange with redox proteins; their active centers are reported to be accessible by CNTs.^15–17 However, the walls of CNTs are less active than the edges, similar to BPG.¹⁷ Therefore, CNTs must be oriented vertically to expose their edges. On the other hand, CNFs with stacked graphene provide a high density of graphene edges without control of their orientation. Noda et al. have developed hydrogene peroxide sensors based on CSCNF and horseradish peroxidase (HRP).¹⁸ However, the nature of the CSCNF used and kinetics of the electrochemical reactions have not yet been studied extensively.

Recently, we have successfully synthesized CSCNF (20 and 30 nm in inner and outer diameters, respectively, and several micrometres in length) using a chemical vapor deposition (CVD) method based on the thermal decomposition of poly(ethylene glycol) (PEG) in the presence of Ni.¹⁹ This method requires no explosive or flammable gases and the treatment temperature is 750 °C, whereas conventional catalytic thermal CVD methods require hydrocarbon or hydrogen and thermal treatment at around a 1000 °C.^20–22 Here we treated the CSCNF with ozone to introduce oxygen-containing groups to the graphene edges to make the CSCNF water soluble and adsorptive to proteins. Although strong acids, such as nitric acid, are widely used for the surface treatment of CNTs and CNFs,^23,24 it takes a long time for the washing and filtration processes to eliminate the acids completely. In comparison, the ozone treatment requires no time-consuming post-treatments. We characterized the CSCNF before and after the treatment in detail. CSCNF provides a three dimensional network to which HRP molecules are adsorbed and through which electrons are transported. The HRP/CSCNF-modified electrodes were examined as H₂O₂ sensors, and their upper sensing limit was controlled by changing the amount of HRP molecules. Kinetics of HRP oxidation by H₂O₂ from the Fe(III) to Fe(IV) state and direct electrochemical reduction from Fe(IV) to Fe(III) were studied by amperometry, which is suitable to the sensitive analysis of irreversible reactions. The electrodes were also applied as cyanide sensors based on inhibition of the enzymatic activity. In this case, the dynamic range of the cyanide sensing was controlled by regulating the HRP amount. The three dimensional network allows the control of overall enzymatic activity and regulation of the sensor performances.

Experimental section

Synthesis and characterization of CSCNF

CSCNF was synthesized according to our previous report.¹⁹ The aqueous solution containing 7.5 mM NiCl₂ and 8 M PEG (monomer unit, M_w = 8 000, Wako Pure Chemicals, Japan) was stirred for 30 min at 60 °C and then coated on a silicon wafer (150 mg cm⁻²) uniformly. The silicon wafer was placed on a hot plate for 1 h at 60 °C, followed by cooling at room temperature for 2 h, resulting in formation of a composite PEG–Ni membrane on the silicon wafer. The PEG–Ni membrane-coated silicon wafer was placed in a quartz pipe (7.5 cm in diameter and 85 cm in length) and heated to 750 °C in an electric oven at a heating rate of 15 °C min⁻¹ under an argon atmosphere. The temperature was maintained at 400, 500, and 750 °C for 1 h each to obtain CSCNF.

About 100 mg of the synthesized CSCNF were placed in a glass tube. Air containing 820 ppm ozone generated by an ozone generator POX-10 (Fuji Electric Co. Ltd., Japan) was flushed into the glass tube at a flow rate of 800 mL min⁻¹ for 2 h at room temperature. The surface functional groups of the ozone-treated CSCNF were analyzed by a temperature programmed desorption (TPD) method. A stainless steel tube filled with 100 mg of the ozone-treated CSCNF was heated from 25 to 900 °C at the heating rate of 5 °C min⁻¹ in an electric oven, where argon gas was continuously run into the tube and amounts of CO and CO₂ desorbed from the CSCNF surface were monitored. The surface area of CSCNF was determined by the Brunauer–Emmett–Teller (BET) method based on N₂ adsorption^25,26 onto the CSCNF surface by use of BELSORP 36 apparatus. Graphite structures of the ozone-treated CSCNF were analyzed by Raman spectroscopy (T-640000, Horiba Jobin Yvon Co. Ltd.).

Preparation of HRP/CSCNF-modified electrodes

Ozone-treated CSCNF and HRP (Sigma-Aldrich) were entrapped in a polyion complex membrane formed from poly(L-lysine) (PLL) (Sigma-Aldrich, M_w = 100 000) as a polycation and polystyrene sulfonic acid (PSS) (Sigma-Aldrich, M_w = 1 000 000) as a polyanion and directly immobilized on a transparent indium tin oxide (ITO) electrode. Fig. 1 shows a schematic illustration for the entrapment of CSCNF and HRP. An ITO-coated glass plate (area ∼0.09 cm²) was treated with a 1.0 M sodium hydroxide aqueous solution for 24 h to provide a negatively charged surface, followed by further treatment with 50 mM PLL (monomer unit) aqueous solution for 1 h. After each treatment, the electrode was thoroughly rinsed with distilled water. A 6 μL aliquot of 5 mg mL⁻¹ CSCNF aqueous dispersion was cast onto the PLL-coated electrode surface, followed by further casting of a 6 μL aliquot of 5–500 μM HRP aqueous solution after drying for 1 h at room temperature. HRP molecules are adsorbed onto CSCNF likely due to electrostatic interaction and hydrogen bonding. After rinsing with distilled water to remove excess HRP molecules, a 6 μL aliquot of 50 mM PLL (monomer unit) aqueous solution was cast onto the electrode surface, followed by further casting of a 6 μL aliquot of 50 mM PSS (monomer unit) aqueous solution to form a polyion complex layer, which protects HRP from desorption. The obtained HRP/CSCNF electrode was subjected to morphological observation by a field emission scanning electron microscopy (FE-SEM, S–4500, Hitachi High-Technologies Co. Ltd., Japan). Surface coverage of HRP was determined by spectrophotometric measurements (MCPD-3000 and MC2530, Otsuka Electronics Co. Ltd., Japan) with maximum mole extinction coefficient of ε = 102 mM⁻¹ cm⁻¹.²⁷ All other reagents were at least analytical grade and used without further purification.

Fig. 1 Schematic illustrations of preparation process of the HRP/CSCNF-modified electrode.

Electrochemical measurements

Electrochemical measurements were performed in a 67 mM phosphate buffer solution (pH 6.4) with a batch system. An Ag|AgCl|KCl sat. and a coiled platinum wire were used as the reference and counter electrodes, respectively. The catalytic activity of the HRP/CSCNF-modified electrode toward H₂O₂ reduction was evaluated by amperometry with a potentiostat LC-4C (BAS). After the working electrode was polarized at +0.150 V, at which H₂O₂ oxidation and reduction currents were almost negligible at the bare ITO electrode, and a steady-state current was obtained, H₂O₂ solution was added into the electrolyte solution, and the steady-state current was recorded. Note that in our previous work, we applied the same potential to the HRP-modified graphite electrode.²⁸ An inhibition effect of CN⁻ on the HRP/CSCNF-modified electrode was measured as follows. After the working electrode was polarized at +0.150 V, H₂O₂ solution was added into the electrolyte solution (final concentration, 1 μM), followed by the addition of NaCN solution. The inhibition ratio of the H₂O₂ reduction current was recorded.

Results and discussion

Characterization of ozone-treated CSCNF

We initially examined the water dispersibility of ozone-treated CSCNF. Water containing 5 mg mL⁻¹ ozone-treated CSCNF after an ultrasonic treatment for 1 h exhibited a uniform dark color, indicating a good dispersibility. In contrast, untreated CSCNF was poorly dispersed in water even after the ultrasonic treatment. These results suggest that hydrophilic groups are introduced to the CSCNF surface by ozone treatment.

The hydrophilic groups were therefore analyzed by the TPD method. If oxygen-containing groups such as carboxyl, anhydrides, and phenol are present on a surface of a carbon material, CO and CO₂ gases are released by thermal desorption during thermal treatment under an inert gas.²⁹ By monitoring the amounts of released CO and CO₂ and their thermal desorption temperatures, the surface coverage and types of oxygen-containing functional groups on the carbon material can be estimated.²⁹Fig. 2 shows TPD charts of ozone-treated and untreated CSCNF. It is clear that CO and CO₂ are released from the ozone-treated CSCNF surface, whereas desorption from the untreated one is almost negligible. The desorption of CO is attributed to anhydrides (> 400 °C), phenol (> 600 °C), ether, carbonyl, and quinone groups (> 700 °C).²⁹ CO₂ released at 200–400 °C is attributed to carboxyl groups and that released above 400 °C is ascribed to lactone groups and/or anhydrides.²⁹ We therefore conclude that some of those oxygen-containing groups form on the CSCNF surface in the ozone treatment. Since the total amounts of CO and CO₂ released from the ozone-treated CSCNF are 2.10 and 0.35 mmol g⁻¹, respectively, the total amount of oxygen atoms introduced to CSCNF by the ozone treatment is ≥2.80 (= 2.10 + 0.35 × 2) mmol g⁻¹. Since some oxygen atoms such as those of phenol and carbonyl groups could have desorbed as H₂O³⁰ and we cannot accurately separate H₂O molecules desorbed from CSCNF from those released from the substrate and the glass vessel used, it is difficult to determine the total amount of oxygen atoms precisely. On the other hand, the net surface area of the treated CSCNF was determined by the BET method based on N₂ adsorption to be approximately 2.3 × 10⁶ cm² g⁻¹. The surface coverage of oxygen atoms on the ozone-treated CSCNF surface is therefore estimated to be ≥1.2 × 10⁻⁹ mol cm⁻². Considering the typical structure of our CSCNF,¹⁹ the atomic ratio of carbon atoms at the outer surface to the overall carbon atoms is about 2.6 × 10⁻². Therefore, we estimate the atomic ratio of the oxygen atoms to the outer edge carbon atoms to be 5.3 × 10⁻².

Fig. 2 (A) CO₂ and (B) CO TPD spectra of CSCNF (a) before and (b) after the ozone treatment.

Next, CSCNF was characterized by Raman spectroscopy before and after the ozone treatment. It is known that sp² carbon atoms give rise to Raman peaks at around 1350 or 1580 cm⁻¹.³¹ Carbon atoms consisting the well-ordered graphene structure give a G band signal at 1580 cm⁻¹. On the other hand, carbon atoms adjacent to a defect or a graphene edge give a D band signal at 1350 cm⁻¹. For example, a strong acid treatment of SWCNT and MWCNT disrupts their structures and increases their D/G ratios.^32–34 Also, treatment of SWCNT with 200 ppm ozone for 1 h increases its D/G ratio from 0.12 to 0.74.³⁵ This suggests that some of the C–C bonds in the graphene structure are broken by the ozone treatment. However, in the present case the D/G ratio of CSCNF increased only slightly (Fig. 3A), from 0.63 to 0.66, even under much more severe conditions (800 ppm for 2 h). Since only a few percent of the graphene structure of CSCNF is exposed and the other parts of the cup are covered with another cup, the graphene structure is protected from severe damages.

Fig. 3 (A) Raman spectra of CSCNF (a) before and (b) after the ozone treatment. (B) Visible spectrum of the high HRP/CSCNF-modified electrode. The spectrum of the CSCNF-modified electrode is subtracted. Inset shows a typical SEM image of the surface of the high HRP/CSCNF-modified electrode.

Response of HRP/CSCNF-modified electrode to H₂O₂

We prepared the HRP/CSCNF–modified electrode using the ozone-treated CSCNF. The SEM image of the electrode surface is shown in Fig. 3B (inset). Three dimensional networks of CSCNF with the diameter of 20–50 nm are obvious.

Cathodic current responses of the HRP/CSCNF-modified electrode to H₂O₂ were measured at +0.150 V (vs. Ag|AgCl) in an air saturated 67 mM phosphate buffer solution (pH 6.4). After the back-ground current stabilized, H₂O₂ was injected to the solution and a steady-state cathodic current was observed. This current is likely due to electron transfer from the electrode to HRP via CSCNF. The reaction mechanisms are as follows.^28,36

ferric HRP + H₂O₂ → compound I + H₂O (1)

compound I + e⁻ + H⁺ → compound II (2)

compound II + e⁻ + H⁺ → ferric HRP + H₂O (3)

where compound I and II are oxidized complexes of HRP, in which the Fe(IV) of heme is coordinated to oxygen. Fig. 4A shows the H₂O₂ concentration dependence of the steady-state cathodic current densities for HRP/CSCNF-modified, CSCNF-modified, HRP-modified, and bare ITO electrodes. The cast amount of HRP was 3.3 × 10⁻⁸ mol cm⁻² and that of CSCNF was 0.33 mg cm⁻². The cathodic currents for H₂O₂ at the HRP/CSCNF-modified electrode (plot a) were much larger than those at the CSCNF-modified, HRP-modified, and bare electrodes (plot b, c, and d, respectively). Responses from a PLL/PSS-modified ITO electrode were almost the same as those from the bare ITO. However, the cathodic currents at the CSCNF electrode were two orders of magnitude larger than those at the bare electrode. This is explained in terms of increased electrode area due to the formation of the three-dimensional network of CSCNF. Note that the cathodic currents at the HRP electrode were nearly equal to those at the bare electrode. This result indicates that electrons could not be transferred directly from ITO to HRP molecules. On the other hand, significant enhancement of the currents was observed when the CSCNF electrode was modified with HRP, indicating that electron transfer from CSCNF to HRP was significant. Thus, we conclude that CSCNF can work as an electric nanowire between the electrode and the HRP active site.

Fig. 4 (A) Cathodic current densities of H₂O₂ reduction at (a) the HRP/CSCNF-modified electrode, (b) the CSCNF-modified electrode, (c) the HRP-modified electrode, and (d) bare electrode in an air-saturated 67 mM phosphate buffer solution (pH 6.4) at +0.150 V vs. Ag|AgCl. Cast amount of HRP was 3.3 × 10⁻⁸ mol cm⁻² and that of CSCNF was 0.33 mg cm⁻². (B) Dependence of cathodic current responses to H₂O₂ at (a) the high, (b) middle, and (c) low HRP/CSCNF-modified electrodes in an air-saturated 67 mM phosphate buffer solution (pH 6.4) at +0.150 V vs. Ag|AgCl. Cast amounts of HRP were (a) 3.3 × 10⁻⁸, (b) 3.3 × 10⁻⁹, and (c) 3.3 × 10⁻¹⁰ mol cm⁻² and that of CSCNF was 0.33 mg cm⁻², respectively.

We also examined the dependence of the cathodic current responses to H₂O₂ on the surface coverage of HRP (Fig. 4B). The cast amounts of HRP were 3.3 × 10⁻¹⁰, 3.3 × 10⁻⁹, and 3.3 × 10⁻⁸ mol cm⁻² (low, middle, and high HRP/CSCNF-modified electrodes, respectively) and that of CSCNF was 0.33 mg cm⁻². As a result, approximately the same sensitivity (current/H₂O₂ concentration) was obtained in the low H₂O₂ concentration region (≤3 × 10⁻⁶ M). These results indicate that the rate of H₂O₂ reduction by HRP (eqn 1) is faster than that of H₂O₂ diffusion from the bulk solution to HRP; the system is diffusion-controlled. If the rate of H₂O₂ reduction is slower than that of the H₂O₂ diffusion, the current response should increase as the HRP coverage on the electrode increases. This is also supported by the transient response to H₂O₂; the current decreased gradually after addition of H₂O₂. In contrast, saturation currents in the higher H₂O₂ concentration region increased as the cast amount of HRP increased. These results indicate that the electron transfer reactions (eqn 2 and 3) are not limited by diffusion of H⁺; the system is kinetically controlled. This is supported by the transient response; a current decrease, which is observed in a diffusion-controlled system, was not observed and the current was almost constant after immediate increase of the current in response to addition of H₂O₂. We can therefore estimate the kinetics of the electron transfer reactions if we know the amount of HRP.

We therefore subjected the low, middle, and high HRP/CSCNF-modified electrodes to spectrophotometric measurements. As a result, only the high HRP/CSCNF-modified electrode exhibited a spectrum attributed to HRP (Fig. 3B, the spectrum of CSCNF-modified electrode is subtracted). On the basis of the spectrum, we determined the minimum surface coverage of HRP. Assuming that the molar extinction coefficient ε of the immobilized HRP at the peak wavelength is equal to that of dissolved HRP (ε = 102 mM⁻¹ cm⁻¹),²⁷ the minimum surface coverage of the high HRP/CSCNF-modified electrode is estimated to be 2.6 ± 0.7 × 10⁻⁹ mol cm⁻² (mean ± SE, n = 4). This value is one order of magnitude lower than the cast amount (3.3 × 10⁻⁸ mol cm⁻²). This is likely due to the shielding of light by CSCNF and/or detachment of cast HRP molecules before overcoating HRP and CSCNF with PLL and PSS. Since the surface area of CSCNF is about 2.3 × 10⁶ cm² g⁻¹, the surface area of CSCNF immobilized on the electrode surface was roughly calculated to be about 7.5 × 10² cm². Hence, the minimum surface coverage of HRP on the CSCNF surface is estimated to be 3.4 × 10⁻¹² mol cm⁻². This value is close to the spectroscopically determined surface coverage of a HRP monolayer on a fluorine-doped tin oxide (FTO) electrode, 7.0 × 10⁻¹² mol cm⁻²,³⁷ suggesting that HRP molecules adsorb onto CSCNF as a monolayer.

Kinetic analysis of the HRP reactions

Kinetics of the reactions were analyzed on the basis of the observed currents (Fig. 4B, plot a) and the HRP coverage (ca. 2.6 × 10⁻⁹ mol cm⁻²). As mentioned above, the current response in the low H₂O₂ concentration range (≤1.0 × 10⁻⁵ M) is determined by the rate of reaction 1. If reaction 1 is not diffusion-controlled but kinetically controlled, the cathodic current i₁ is given by eqn 4.^28,38,39

i ₁ = 2Fk₁ΓC_S (4)

where F is the Faraday constant, k₁ is the rate constant of eqn 1, Γ is the surface coverage of HRP, and C_S is the H₂O₂ concentration in the bulk solution. On the other hand, the current response in the high H₂O₂ concentration range (≥ 1.0 × 10⁻³ M) is determined by the rate of reactions 2 and 3. If the reactions are not controlled by diffusion of protons but by the enzymatic reactions including electron transfer from CSCNF to compound I and II, the cathodic current i₂₃ is given by eqn 5.^28,38,39

i ₂₃ = 2FΓC_Hk₂k₃/(k₂ + k₃) (5)

where C_H is the proton concentration in the bulk solution and k₂ and k₃ are the rate constants of eqn 2 and 3, respectively. Thus, the rate constants k₁ and k₂k₃/(k₂ + k₃) can be estimated on the basis of eqn 4 and 5, respectively.

Since the current in the low H₂O₂ concentration range is limited by diffusion, we can only estimated the minimum k₁ value from eqn 4. The value was estimated to be 88 M⁻¹ s⁻¹ from the i₁/C_S value (4.4 × 10⁻² A cm⁻² M⁻¹ at 3.0 × 10⁻⁶ to 1.0 × 10⁻⁵ M H₂O₂ for the high HRP/CSCNF-modified electrode, Fig. 4B) and the Γ value (2.6 × 10⁻⁹ mol cm⁻²). The minimum k₁ value is actually lower than those obtained for HRP-adsorbed graphite electrodes by Ruzgas et al. (3.9 × 10⁴ M⁻¹ s⁻¹)⁴⁰ and Tatsuma et al. (3.1 × 10⁴ M⁻¹ s⁻¹).²⁸

We also estimated the k₂k₃/(k₂ + k₃) value on the basis of eqn 5 from the i₂₃ value of 1.7 × 10⁻⁵ A cm⁻² at 1.0 × 10⁻³ M H₂O₂ for the high HRP/CSCNF-modified electrode, the Γ value, and the C_H value of 4.0 × 10⁻⁷ M to be about 8.5 × 10⁴ M⁻¹ s⁻¹. This value is about two orders of magnitude lower than that at pH 7.0 of the HRP-adsorbed graphite electrode reported by Ruzgas et al. (6.6 × 10⁶ M⁻¹ s⁻¹).⁴⁰ This difference may be explained in terms of the coexising polymers (i.e., PLL and PSS), which may interfere with the direct electrochemistry of HRP.²⁸ Local pH around HRP molecules may also be high (i.e., the C_H value may be low) in the present system because of PLL.

Application to inhibitor sensing

The reduction of H₂O₂ catalyzed by HRP (eqn 1–3) is suppressed when an inhibitor such as cyanide is added to the system, because the inhibitor coordinates to the sixth coordination site of ferric HRP, which is the reaction site of H₂O₂. The concentration of the inhibitor can be evaluated from the current decrease.^36,41 From the current response to H₂O₂ before and after addition of an inhibitor, a %inhibition value R is determined.

R = 100 × (i₀ − i)/i₀ (6)

where i₀ and i are the current responses to H₂O₂ before and after addition of the inhibitor, respectively. When eqn 1 is the rate-determining step in eqn 1–3, eqn 6 can be rewritten as follows.^36,42

R = 100/(1/K_appC_i + 1) (7)

K _app = K/(1 + k₁Γd/D) (8)

where C_i is the inhibitor concentration, K is the formation constant of the HRP-inhibitor complex (K = [HRP–inhibitor]/[HRP][inhibitor]), K_app is the apparent formation constant, d is the diffusion layer thickness, and D is the diffusion coefficient of H₂O₂. The inhibitor concentration C_i can be determined from the R value on the basis of eqn 7. It is obvious from eqn 7 that R ∼ 100% when K_appC_i >> 1 and R ∼ 0% when K_appC_i ≪ 1. In addition, R = 50% when C_i = 1/K_app. As eqn 8 shows, K ∼ K_app if the Γ value is sufficiently small and the catalytic reaction of HRP for H₂O₂ is kinetically controlled. If we increase the Γ value so that the diffusion of H₂O₂ controls the reaction, the K_app value becomes smaller than K; the dynamic range of the inhibitor sensing shifts to a higher concentration range. Thus we can regulate the dynamic range if we control the amount of HRP on CSCNF.

Fig. 5 shows dependence of the %inhibition value for the low and high HRP/CSCNF-modified electrodes on the cyanide concentration. Here the H₂O₂ concentration was 1.0 μM, where the eqn 1 is the rate-determining step. The inhibition curve for the high HRP/CSCNF-modified electrode apparently shifts towards the higher concentration region compared to that for the low HRP/CSCNF-modified electrode. From the inhibition curves, the K_app values were determined to be 1.5 ± 0.5 × 10⁵ M⁻¹ (n = 3) and 3.1 ± 1.2 × 10³ M⁻¹ (n = 4) for the low and high HRP/CSCNF-modified electrodes, respectively. In a previous report regarding a similar cyanide sensing by a pyrolytic graphite electrode with adsorbed HRP (HRP/PG), the K_app value was regulated between 1.3 × 10⁵ M⁻¹ and 1.7 × 10⁴ M⁻¹.³⁶ The control of the dynamic range was over about one order of magnitude for the previous HRP/PG electrode with a two dimensional surface geometry. In the present work, the three dimensional network of CSCNF allowed control of the HRP amount and the dynamic range over about two orders of magnitude. In addition to cyanide, it is expected to tune the dynamic range for other inhibitors, such as azide, thiourea, and dichromate.^43,44

Fig. 5 Dependence of the inhibition ratio of (a) the low and (b) high HRP/CSCNF–modified electrodes on the cyanide concentration in an air-saturated 67 mM phosphate buffer solution (pH 6.4) containing 1.0 μM H₂O₂ at +0.150 V vs. Ag|AgCl.

Conclusion

CSCNF was treated with ozone to introduce oxygen-containing groups onto the edges of the graphene cups. The hydrophilic CSCNF was used to construct a three-dimensional network that works both as an electrical nanowire and an enzyme support. By taking advantage of the network, we developed biosensors for H₂O₂ and cyanide, and regulated their sensing limits. Future improvement in the contact between the graphene edges and the enzymes would facilitate the direct electrochemistry of enzymes.

Acknowledgements

We are grateful to Dr Y. Kato for help with the Raman spectroscopy measurement. This work was supported by an International cooperative program of the Japan Science and Technology Agency (JST, Japan).

References

1. J. Wang, Electroanalysis, 2005, 17, 7–14.
2. B. Mahar, C. Laslau, R. Yip and Y. Sun, IEEE Sens. J., 2007, 7, 266–284.
3. J. Wang and Y. Lin, TrAC, Trends Anal. Chem., 2008, 27, 619–626.
4. J. J. Gooding, Electrochim. Acta, 2005, 50, 3049–3060.
5. P. Serp, M. Corrias and P. Kalck, Appl. Catal., A, 2003, 253, 337–358.
6. R. M. Wightman, M. R. Deakin, P. M. Kovach, W. G. Kuhr and K. J. Stutts, J. Electrochem. Soc., 1984, 131, 1578–1583.
7. R. J. Bowling, R. T. Packard and R. L. McCreery, J. Am. Chem. Soc., 1989, 111, 1217–1223.
8. C. E. Banks, R. R. Moore, T. J. Davies and R. G. Compton, Chem. Commun., 2004, 1804–1805.
9. M. Pacios, M. del Valle, J. Bartroli and M. J. Esplandiu, J. Electroanal. Chem., 2008, 619–620, 117–124.
10. G. Jönsson and L. Gorton, Electroanalysis, 1989, 1, 465–468.
11. Y. Xu, W. Peng, X. Liu and G. Li, Biosens. Bioelectron., 2004, 20, 533–537.
12. J. Pang, C. Fan, X. Liu, T. Chen and G. Li, Biosens. Bioelectron., 2003, 19, 441–445.
13. G. Wang, N. M. Thai and S.-T. Yau, Electrochem. Commun., 2006, 8, 987–992.
14. C. M. Silveira, S. P. Gomes, A. N. Araújo, M. C. B. S. M. Montenegro, S. Todorovic, A. S. Viana, R. J. C. Silva, J. J. G. Moura and M. G. Almeida, Biosens. Bioelectron., 2010, 25, 2026–2032.
15. J. J. Gooding, R. Wibowo, J. Liu, W. Yang, D. Losic, S. Orbons, F. J. Mearns, J. G. Shapter and D. B. Hibbert, J. Am. Chem. Soc., 2003, 125, 9006–9007.
16. F. Patolsky, Y. Weizmann and I. Willner, Angew. Chem., Int. Ed., 2004, 43, 2113–2117.
17. J. Liu, A. Chou, R. Rahmat, M. N. Paddon-Row and J. J. Gooding, Electroanalysis, 2005, 17, 38–46.
18. T. Noda, T. Ukai and T. Yao, Anal. Sci., 2010, 26, 675–679.
19. Y. Takahashi, H. Fujita, W.-H. Lin, Y.-Y. Li, T. Fujii and A. Sakoda, Adsorption, 2010, 16, 57–68.
20. J. Kong, A. M. Cassel and H. Dai, Chem. Phys. Lett., 1998, 292, 567–574.
21. G. Che, B. B. Lakshmi, C. R. Martin, E. R. Fisher and R. S. Ruoff, Chem. Mater., 1998, 10, 260–267.
22. C. L. Cheung, A. Kurtz, H. Park and C. M. Lieber, J. Phys. Chem. B, 2002, 106, 2429–2433.
23. S. Nagasawa, M. Yudasaka, K. Hirahara, T. Ichihashi and S. Iijima, Chem. Phys. Lett., 2000, 328, 374–380.
24. H. Hu, B. Zhao, M. E. Itkis and R. C. Haddon, J. Phys. Chem. B, 2003, 107, 13838–13842.
25. M. Eswaramoorthy, R. Sen and C. N. R. Rao, Chem. Phys. Lett., 1999, 304, 207–210.
26. H. P. Boehm, Carbon, 1994, 32, 759–769.
27. P. J. Ohlsson, Acta Chem. Scand., Ser. B, 1976, 30b, 373–375.
28. T. Tatsuma, K. Ariyama and N. Oyama, J. Electroanal. Chem., 1998, 446, 205–209.
29. J. L. Figueiredo, M. F. R. Pereira, M. M. A. Freitas and J. J. M. Órfão, Carbon, 1999, 37, 1379–1389.
30. S. Haydar, C. Moreno-Castilla, M. A. Ferro-García, F. Carrasco-Marín, J. Rivera-Utrilla, A. Perrard and J. P. Joly, Carbon, 2000, 38, 1297–1308.
31. M. S. Dresselhaus, A. Jorio, M. Hofmann, G. Dresselhaus and R. Saito, Nano Lett., 2010, 10, 751–758.
32. K. C. Park, T. Hayashi, H. Tomiyasu, M. Endo and M. S. Dresselhaus, J. Mater. Chem., 2005, 15, 407–411.
33. M. T. Martínez, M. A. Callejas, A. M. Benito, M. Cochet, T. Seeger, A. Ansón, J. Schreiber, C. Gordon, C. Marhic, O. Chauvet, J. L. G. Fierro and W. K. Maser, Carbon, 2003, 41, 2247–2256.
34. S. Costa and E. B. Palen, Acta Phys. Pol. A, 2009, 116, 32–35.
35. W. Sun, U. Khaled, H. Tomita, Z. Li, K. Imasaka and J. Suehiro, Jpn. J. Appl. Phys., 2010, 49, 055002.
36. T. Tatsuma and N. Oyama, Anal. Chem., 1996, 68, 1612–1615.
37. T. Tatsuma, Y. Okawa and T. Watanabe, Anal. Chem., 1989, 61, 2352–2355.
38. T. Tatsuma, M. Gondaira and T. Watanabe, Anal. Chem., 1992, 64, 1183–1187.
39. T. Tatsuma and T. Watanabe, Anal. Chem., 1992, 64, 625–630.
40. T. Ruzgas, L. Gorton, J. Emnéus and G. Marko-Varga, J. Electroanal. Chem., 1995, 391, 41–49.
41. M. H. Smit and A. E. G. Cass, Anal. Chem., 1990, 62, 2429–2436.
42. K. Komori, K. Takada and T. Tatsuma, J. Electroanal. Chem., 2005, 585, 89–96.
43. H. Zollner, Handbook of enzyme inhibitors part A, 2nd Ed., VCH Verlagagesellschaft GmbH, Germany, 1993, 367–368.
44. J. Zhao, R. W. Henkens and A. L. Crumbliss, Biotechnol. Prog., 1996, 12, 703–708.

---