Construction of a Hydrogen Peroxide Biosensor on Interdigitated Microband Electrodes Fabricated by a Mix-and-Match Process

The reagents used were of analytical grade or the highest commercially available purity and were used as received without further purification. 3-Mercaptopropionic acid (MPA), 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), HRP (261 U mg⁻¹), K₃Fe(CN)₆, and K₄Fe(CN)₆·3H₂O (≥98%) were purchased from Sigma–Aldrich. Na₂HPO₄ (≥99%) and NaH₂PO₄ (≥98%) were purchased from Showa. KCl (99.5%) was obtained from Tedia. N-Hydroxysuccinimide (NHS) was purchased from Alfa Aesar. All samples were prepared with 0.1M phosphate buffer solution at pH 7.0. Deionized water of resistivity not less than 18 MΩ cm was taken from a Milli-Q water purification system (Milli-Q, USA).

A conventional three-electrode system was used, with the Au interdigitated microband electrodes as the working electrode, an Ag/AgCl (3 M KCl) as the reference electrode, and a platinum wire as the counter electrode. Cyclic voltammetry was performed with a CHI 920C Electrochemical Analyzer (CH Instruments, USA). The potential of the generator microband was scanned at 20 mV s⁻¹ while that of the collector microband was held at −0.2 V vs Ag/AgCl. Electrochemical impedance spectroscopy (EIS) measurement was performed with an Autolab PGSTAT30/FRA2 Electrochemical Analyzer (Eco Chemie, Netherlands) at open circuit with a 10 mV amplitude sinusoidal perturbation, applied across the frequency range from 100 kHz to 10 mHz.

The Au interdigitated microband electrodes with 1 μm band width and 1 μm gap width were fabricated using an optical mix-and-match process, which has been described elsewhere.³⁴ The process combines a step-and-repeat process to define the microband electrodes, and a standard lithography to pattern the active areas and contacts through an SU-8 passivation layer. These are regular interdigitated structures composed by 500 μm long bands on a silicon chip over an active area of 500 μm × 500 μm, that is defined by a passivation layer consisting of a 0.5–2 μm thick SU-8 photoresist layer. These chips were attached and wire bonded to printed circuit boards, and encapsulated using Epotek H-77 thermocurable resin. Before activation, the Au interdigitated microband electrodes were washed serially in ethanol, water, and dried under a nitrogen flow. The electrochemical activation for removing traces of the SU-8 epoxy resin consists of a series of potential steps comprising 5 s at −2 V and another 5 s at 0 V in 0.1 M KCl. After the chips were cleaned, the Au interdigitated microband electrodes were immersed in 0.1 M phosphate buffer solution containing 10 mM MPA for 4 h to form a self-assembly layer. The MPA modified interdigitated microband electrodes were immersed in 0.1 M phosphate buffer solution containing 10 mM EDC and 10 mM NHS as a coupling agent for 1 h. After rinsing with 0.1 M phosphate buffer solution, the modified interdigitated microband electrodes were further treated with 0.1 M phosphate buffer solution containing 2 mg ml⁻¹ HRP for 4 h to form HRP/MPA modified interdigitated microband electrodes. Thus, the sensing interdigitated microband electrodes were obtained. The scheme of fabrication steps and probing mechanism for hydrogen peroxide at the HRP/MPA modified interdigitated microband electrodes using generator-collector mode is shown in Figure 1.

Zoom In Zoom Out Reset image size

Generator-only and generator-collector experiments were carried out at interdigitated microband electrodes in order to investigate the electrode amplification factor and collection efficiency. The amplification factor is the ratio of the anodic peak current using the generator-only mode to the steady state current recorded at the same electrode operated in generator-collector mode. On the other hand, collection efficiency is the ratio of the steady state current at the collector electrode (i_col) which ferricyanide is reduced to ferrocyanide versus the steady state current at the generator electrode (i_gen) which ferrocyanide is oxidized to ferricyanide using the generator-collector mode. Cyclic voltammograms of generator-only and generator-collector in 0.1 M phosphate buffer solution containing 5 mM ferrocyanide at interdigitated microband electrode are shown in Figure 2. It can be seen that a steady state current is obtained with generator-collector mode (Figure 2, curve b) and that the current at the generator electrode is greatly enhanced compared to that of the generator-only mode (Figure 2, curve a). An amplification factor, which is the ratio of the response current obtained with generator-collector mode to that obtained with the generator-only mode, around ∼9.1 was observed, indicating that the interdigitated microband electrode was highly effective in obtaining a high response current using generator-collector mode. This result is attributed to the rapid diffusion of redox species between the two sets of microbands, which are polarized at potentials above and below the formal potential of the ferrocyanide/ferricyanide couple, respectively; resulting in the continuous transformation of species from the reduced to the oxidized forms and vice-versa. This redox cycling makes more redox species available near the generator-collector electrodes and it offers important current enhancement. The collection efficiency estimated from the i_col and i_gen in Figure 2b was 0.97 which is very close to unity. This is attributed to the small width (1 μm) and gap (1 μm) between the interdigitated microband electrodes. Both the obtained collection efficiency and amplification factor imply the reversibility of the redox species, and a fast electron transfer at the interdigitated microband electrodes using generator-collector mode, and are in excellent agreement with previously reported simulations.³⁴

Zoom In Zoom Out Reset image size

EIS was used to monitor the construction of the biosensor on the interdigitated microband electrodes step-by-step. The Nyquist plots obtained after each preparation step in 0.1 M KCl containing 5 mM [Fe(CN)₆]³⁻ and 5 mM [Fe(CN)₆]⁴⁻ are shown in Figure 3. All impedance measurements were performed at open circuit with a 10 mV amplitude sinusoidal perturbation, applied across the frequency range from 100 kHz to 10 mHz. To ensure measurement stability, the EIS experiments for the determination of charge transfer resistance were repeated three times and the relative standard deviation was less than 4%. The Nyquist plot presents a semicircle at high frequencies, consistent with interdigitated microband electrode behavior.²⁹ The Randles equivalent circuit presented in the inset of Figure 3 is appropriate when fitting the experimental data, where R_ct represents the charge transfer resistance. The bare interdigitated microband electrodes (curve a in Figure 3) showed a smaller high-frequency semicircle than the modified interdigitated microband electrode, consistent with faster electron transfer (smaller electron transfer resistance). The R_ct was about 18 ± 0.7% kΩ for bare interdigitated microband electrodes. When MPA was modified onto the surface of interdigitated microband electrode, the diameter of the semicircle increased clearly and the R_ct was about 30 ± 1.3% kΩ (curve b in Figure 3). This is due to the self-assembly of MPA onto the surface of interdigitated microband electrode and effectively excluded the [Fe(CN)₆]^3−/4−. Then, the MPA modified interdigitated microband electrode was immersed in the HRP solution and this further increased R_ct to 57 ± 3.5% kΩ (curve c in Figure 3), suggesting that HRP had been successfully covalently bonded with MPA onto the surface of interdigitated microband electrode and partially blocked its surface. The surface coverage (θ) estimated by θ = (1 - R_ct of bare Au interdigitated microband electrodes/R_ct of HRP/MPA/Au interdigitated microband electrodes)³⁵ was about 0.68 ± 1.3%.

Zoom In Zoom Out Reset image size

In electrochemical biosensors, at least one biological recognition element is attached to an electrode surface. The surface morphology of the HRP/MPA modified interdigitated microband electrode was characterized by AFM to investigate the assembly process step-by-step on the surface of interdigitated microband electrode. The AFM images of bare interdigitated microband electrode, the same electrode modified with MPA, and finally the HRP/MPA modified interdigitated microband electrode are shown in Figure 4. A typical AFM image of HRP/MPA modified interdigitated microband electrode is shown in Figure 4c. This result presents that the HRP was coated onto the surface of interdigitated microband electrode. From the AFM images we also can conclude that the surface of the HRP/MPA modified interdigitated microband electrode is rougher in Figure 4c than those of bare interdigitated microband electrode in Figure 4a and MPA modified interdigitated microband electrode in Figure 4b. The root-mean-square roughness values are 1.9, 2.5, and 8.2 nm for the bare interdigitated microband electrode, MPA modified interdigitated microband electrode, and HRP/MPA modified interdigitated microband electrode, respectively. This is attributed to the fact that the HRP is immobilized on the surface of interdigitated microband electrode.

Zoom In Zoom Out Reset image size

To evaluate the potential application of the prepared HRP/MPA modified interdigitated microband electrodes for the determination of hydrogen peroxide in aqueous solution, different concentrations of hydrogen peroxide were added into 0.1 M phosphate buffer solution (pH 7). The typical cyclic voltammetric responses of the proposed hydrogen peroxide biosensor to successive additions of hydrogen peroxide in 0.1 M phosphate buffer solution (pH 7) containing 5 mM ferrocyanide using generator-collector mode are shown in Figure 5. It can be seen that both the oxidation (generator) and reduction (collector) currents responded sensitively to the addition of hydrogen peroxide and reached steady state at about 0.35 V vs. Ag/AgCl. The corresponding calibration curves for the generator and collector electrodes shown in Figure 5 (see the inset) indicates that the prepared hydrogen peroxide exhibits a linear range up to 500 μM (i (μA) = 0.0717 × c (μM) + 0.4860; r² = 0.9996), a sensitivity of 114.75 A M⁻¹ cm⁻², and a detection limit of 9.96 μM. The detection limit is determined according with the 3 S_b/m criterion, where m is the slope of the linear calibration plot and S_b is estimated as the standard deviation. The sensitivity (114.75 A M⁻¹ cm⁻²) of the present hydrogen peroxide biosensor is better than previous reported hydrogen peroxide biosensors based on HRP such as adamantane-modified HRP/polythiolated-β-cyclodextrin/Au (109 μA M⁻¹ cm⁻²)³⁶ and HRP/nano-Au/chitosan-entrapped carbon paste electrode (0.013 A M⁻¹ cm⁻²).³⁷ The reason for that is that the interdigitated microband electrode performing with generator-collector mode provides access to enhanced rates of mass transport that improve the sensitivity of the biosensor.

Zoom In Zoom Out Reset image size

Co-electrooxidation of potential interferences such as ascorbic acid (AA) and uric acid (UA) is a key factor for the evaluation of hydrogen peroxide biosensors. The interference tests were carried out in 0.1 M phosphate buffer solution (pH 7) containing 10 μM hydrogen peroxide in the presence of the same concentration of AA and UA. The responses at the HRP/MPA modified interdigitated microband electrode to the successive additions of 10 μM hydrogen peroxide, 10 μM AA, and 10 μM UA in 0.1 M phosphate buffer solution (pH 7) containing 5 mM ferrocyanide using generator-collector mode are shown in Figure 6. No significant responses can be observed after the additions of 10 μM AA and 10 μM UA to 10 μM hydrogen peroxide in 0.1 M phosphate buffer solution (pH 7) at the HRP/MPA modified interdigitated microband electrode. The results demonstrate that these species coexisting in the sample matrix did not affect the determination of hydrogen peroxide. The reason for this is that the current amplification resulting from operating in generator-collector mode relies on the reversibility of the redox couple involved. On the other hand, UA and AA are irreversibly oxidized, which means that their oxidation products are not reduced back at the collector electrode. This means that their signal is not amplified at all or nearly as much as that of the ferrocyanide/ferricyanide interacting with the HRP, and the result is a selective biosensor with a high selectivity for hydrogen peroxide determination.

Zoom In Zoom Out Reset image size

Long-term biosensor stability is crucial for practical application of the proposed hydrogen peroxide biosensor. To evaluate the storage stability of the prepared hydrogen peroxide biosensor, the HRP/MPA modified interdigitated microband electrodes were kept in 0.1 M phosphate buffer solution (pH 7) at 4°C in order to remain its activity when not in use. The stability was examined by periodic measurement of the biosensor responses to 10 μM hydrogen peroxide in 0.1 M phosphate buffer solution (pH 7) containing 5 mM ferrocyanide at the HRP/MPA modified interdigitated microband electrode using generator-collector mode. The change of the generator current at the HRP/MPA modified interdigitated microband electrode for 13 days is shown in Figure 7. It decreased to about 96% of its initial response current on day 3 and about 83% on day 13. The loss of the response current may result from the decrease of the enzyme activity and the materials losses by diffusion into solution during storage.

Zoom In Zoom Out Reset image size