Introduction

Hydrogen peroxide (H2O2) is not only an important intermediate species in food, pharmaceutical and environmental analysis, but also considered as the mediators of the cellular pathology in association with aging and severe human diseases such as cancers and cardiovascular disorders [1]. As a consequence, a sensitive method for reliable determination of H2O2 is necessary. Up to now, several traditional techniques were employed for H2O2 analysis, such as titrimetry, UV-visible spectrophotometry, fluorimetry, chemiluminescence, high performance liquid chromatography and electrochemistry. Among these methods, electrochemical biosensors display better prospect due to their advantages of easy preparation, fast detection, good selectivity, and low-cost [2]. Most previous studies on this subject involved the use of enzymes which can accelerate the electron transfer between the electrodes and H2O2 [3]. The heme protein cytochrome c (Cyt c) is an excellent model for studying the electron transfer of typical enzymes and is widely used to develop biosensors for the determination of H2O2 due to its close similarity to peroxidase [4]. However, the electron transfer at bare electrode surface is interrupted by the three-dimensional structure of enzymes because the redox active centers can be shielded by their insulated protein shells and unfavorable orientation on the electrode surface [5]. Moreover, direct adsorption of enzymes onto the electrode surface may frequently result in their denaturation and the loss of bioactivity. Therefore, a wealth of nanomaterials such as nanoparticles [6], nanotubes [7], nanoporous materials [8] are employed to immobilize the enzymes and at the same time, to connect the active center of redox enzymes to the electrode surface without the protein denaturing.

With the development of nanomaterials, nanoparticles with ordered mesostructures have been studied and have attracted much attention in the fields of both nanotechnology and biomedicine in view of their property of the mesochannels and quantum effects in the nanoscale [9]. Their open-framework structures, large specific surface area, porosity, and nanosize make ordered mesoporous nanoparticles useful in catalysis [10], adsorption [11], controlled drugs release [12], and cellular delivery [13]. Benefiting from the advantages of carbon materials, such as good electrical conductivity, chemical inertness, and nontoxicity, carbon-based mesoporous nanoparticles are ideal carriage for metals, drugs and enzymes, paving a promising platform in biochemistry research [14]. Fang and coworkers demonstrated a novel low-concentration hydrothermal route to synthesize highly ordered mesoporous carbon nanoparticles with spherical morphology and uniform size [15]. The obtained mesoporous carbon nanospheres (MCNs) showed good biocompatibility and high drug capacity.

The layer-by-layer films assembled by alternative adsorption of positively and oppositely charged species from their solutions through electrostatic interactions have been reported as a robust method for facilitating electron transfer [16]. Moreover, layer-by-layer films can provide suitable microenvironments to retain the activity of biomolecules due to their mild conditions [17]. It has been wildly used in the investigation of direct electron transfer between proteins and underlying electrodes. For example, Shi and coworkers studied electrochemistry and electrocatalytic properties of hemoglobin in layer-by-layer films of SiO2 with vapor-surface sol–gel deposition [18]. Xu et al. constructed a layer-by-layer assembled (Hb/CMK-3)n film for the fabrication of an amperometric H2O2 biosensor with high sensitivity and long-term stability [19].

Herein we described a new biosensor toward H2O2 using MCNs as the matrix for the immobilization of Cyt c. The MCNs were synthesized according to the previous report [15] and immobilized on the conducting indium/tin oxide (ITO) electrode via layer-by-layer assembly strategy. The Cyt c was absorbed onto multilayer of MCNs and the direct electrochemistry of Cyt c was investigated. The immobilized Cyt c showed a good electrocatalytic activity to the reduction of H2O2.

Experimental

Reagents and solutions

Horse heart cytochrome c (MW 12384), poly (diallyldimethylammonium chloride) (PDDA, an aqueous solution with a molecular weight of 100,000 – 200,000), and triblock copolymer Pluronic F127 (Mw = 12600, PEO106PPO70PEO106) were purchased from Sigma-Aldrich (http://www.sigmaaldrich.com). H2O2 (wt. 30 %), phenol, formalin aqueous solution (37 wt. %), and neat ethanol were purchased from Shanghai Chemical Plant (http://www.chinanusa.com). Stock solutions of H2O2 were freshly diluted from 30 % solution. The phosphate buffer solutions (PBS), in accordance with the physiological pH value of 7.4, were made of sodium phosphate (Na2HPO4: NaH2PO4 = 81:19, molar ratio) and NaCl dissolved in water at a final concentration of 50 and 10 mM, respectively. All other reagents were of analytical grade, and double distilled water was used throughout. ITO-coated glass plates with a square resistance of ~10 Ω cm−2 were purchased from Shenzhen Nanbo Display Technology Co. Ltd (http://www.csgholding.com/en/index.asp).

Synthesis of mesoporous carbon nanospheres

According to the previous report, ordered MCNs were prepared through an aqueous synthesis route containing low concentration reactants and promoted at a high temperature hydrothermal condition [15]. Briefly, commercial available triblock copolymer Pluronic F127 was employed as a template and phenolic resol as a carbon source. Firstly, 0.5 g of phenol, 1.75 mL of formalin aqueous solution (37 wt. %) and 15 mL of NaOH aqueous solution (0.1 M) were mixed and stirred at 70 °C for 0.5 h to obtain low-molecular-weight phernolic resols. Then 0.8 g of F127 dissolved in 15 mL of H2O was added. The mixture was stirred at 66 °C with a stirring speed of 340 ± 40 rpm for 2 h. After that, 50 mL water was added to dilute the solution. The reaction was stopped when the deposit was observed after 16 ~ 18 h. Until the deposit was dissolved again, 17.7 mL of the obtained solution was transferred into an autoclave and diluted with 56 mL of H2O, then followed by heating at 130 °C for 24 h. The final MCNs products were obtained by carbonization at 700 °C in N2 atmosphere for 3 h and the triblcok copolymer templates were removed during this process.

Preparation of the modified electrode

ITO-coated glass plates were thoroughly cleaned by sonicating sequentially for about 15 min in the following solvents: soapy water, neat ethanol, 1 M NaOH, distilled water and dried at room temperature. Then the prepared ITO electrode with original negative surface charge was treated with a polyelectrolyte of cationic PDDA (5 wt. %) for about 60 min to form positive surface charge. Afterwards, the electrode was alternately incubated in negatively charged suspension of MCNs (0.5 mg mL−1) and PDDA for 20 min each. All of the adsorption steps were performed at ambient temperature and after each adsorption step the electrode was rinsed with distilled water. When this cycle procedure was repeated N times, the (MCNs/PDDA)N/ITO electrode was obtained. Cyt c immobilization was achieved by immersing this modified electrode in Cyt c solution (0.5 mg mL−1, dissolved in pH 7.4, 50 mM PBS) at 4 °C refrigerator overnight. Prior to all measurements, the modified electrode was rinsed in 50 mM pH 7.4 PBS to remove non-immobilized protein. Thus, the Cyt c modified electrode was obtained and will be referred as Cyt c/(MCNs/PDDA)N/ITO electrode.

Apparatus

Scanning electron microscopy (SEM) images were recorded on a Philips XL30 microscope operated at 20 kV. Transmission electron microscopy (TEM) images were taken with a JEOL 2011 microscope operated at 200 kV. Cyclic voltammetric (CV) and amperometric measurements were carried out by a computer-controlled CHI 660c electrochemical workstation (CHI Instruments, Shanghai Chenhua Instrument Corp.). A three-electrode system was set up with the modified ITO as the working electrode, a coiled platinum wire as the counter electrode, and a saturated calomel electrode (SCE) as the reference in an electrochemical glass cell containing 10 mL 50 mM PBS at room temperature. A magnetic stirrer provided the convective transport during the amperometric measurement.

Results and discussion

Characterization of MCNs and Cyt c/(MCNs/PDDA)N assembled films

The morphology of MCNs was characterized by SEM and TEM. As shown in Fig. 1a, the typical SEM image displayed uniform spherical morphology of MCNs in large domains. The TEM image of synthesized MCNs showed uniform spheres with diameters of approximately 50 nm (Fig. 1b). Furthermore, it is remarkable that an ordered array of mesopores can be observed, implying an open pore structure on the surface. These characteristic nanospheres with mesostructures could be useful in adsorption, cellular delivery, and immobilization of biomolecules.

Fig. 1
figure 1

(a) SEM image of MCNs. (b) TEM image of MCNs

Full size image

The position of the Soret band of iron heme is usually employed as an indicator of the microenvironment of heme proteins [20]. Figure 2 depicted the UV–vis absorption spectra of 0.5 mg mL−1 Cyt c in solution and the growth process of Cyt c/(MCNs/PDDA)N (N = 1, 3, 5, 7) assembled films. As can be seen, the band for Cyt c adsorbed on the assembly is located at 408 nm, similar as that of native Cyt c in solution (409 nm). This implies that the MCNs matrix has good biocompatibility and is suitable for Cyt c molecules to maintain their native conformation and bioactivity. Moreover, with the increase of (MCNs/PDDA) bilayers, the absorption peak at 409 nm grew gradually, indicating an increase amount of Cyt c immobilized on (MCNs/PDDA)N assembled films.

Fig. 2
figure 2

UV–vis absorption spectra of bare ITO electrode (curve a), Cyt c/(MCNs/PDDA)N/ITO (N = 1, 3, 5, 7, curve b, c, d, e, respectively), and 0.5 mg mL−1 Cyt c in 50 mM PBS (pH 7.4, curve f)

Full size image

Cyclic voltammetric behavior of Cyt c immobilized on (MCNs/PDDA)N assembled films

Cyclic voltammetry was employed to characterize the electron transfer behavior of Cyt c during the growth of the (MCNs/PDDA)N films. Typical voltammograms were shown in Fig. 3. A pair of well-defined and quasi-reversible redox peaks was observed in each of the cyclic voltammograms at ITO electrodes modified with Cyt c/(MCNs/PDDA)N (N = 1, 3, 5, 7), while only the charge current was obtained at the (MCNs/PDDA)5/ITO surface in 50 mM PBS (pH 7.4). The redox peaks can be attributed to the direct electron transfer of Cyt c on (MCNs/PDDA)N multilayer films. Moreover, both reduction and oxidation peaks grew gradually with the number of bilayers (N) up to 5, and then no longer increased with N. At the scan rate of 0.05 V s−1, the formal potential (E 0  = (E pa   + E pc )/2) of Cyt c obtained at the (MCNs/PDDA)5/ITO electrode is estimated to be 0.141 V (vs. SCE) and a peak-to-peak potential separation (ΔE p ) of the redox peaks is about 0.032 V. Compared with the E 0 of Cyt c in solution with its native state [21], the E 0 of Cyt c at MCNs matrix shifted positively. Previous studies have demonstrated that E 0 is sensitive to the protein conformation and the solvent medium [22]. A small change in the accessibility of the heme group to the protein exterior can modulate the heme reduction potential dramatically. In the present study, the electrostatic interaction between the positively charged Cyt c (pI = 10.5) and negatively charged carbon surface led to a slight conformation change, which accounted for the shift of E 0 in our experiment. The small ΔE p indicated a fast electron transfer rate between Cyt c and (MCNs/PDDA)N/ITO electrode. Former research proved that the negative charged surface can promote the attachment of Cyt c in an orientation favorable for electron transfer [23]. Thus, the mesoporous carbon nanospheres can provide a favorable microenvironment for the direct electron transfer of Cyt c.

Fig. 3
figure 3

Cyclic voltammograms of the bare ITO electrode (curve a), and Cyt c/(MCNs/PDDA)N/ITO (N = 1, 3, 5, 7, curve b, c, d, e, respectively) electrodes in 50 mM PBS (pH 7.4). Scan rate: 0.05 V s−1

Full size image

The average surface concentration (Γ, mol cm−2) of electroactive Cyt c molecules can be estimated from the charge integration of the reduction peak of the cyclic voltammogram of Cyt c/(MCNs/PDDA)N/ITO, according to the Faraday’s law. As shown in Fig. 4, the Γ of Cyt c increased gradually with the number of bilayers (N) up to 5, and then decreased slightly. We speculate that with the increase of N, the distance from proteins in outer layers to the electrode surface was extended accordingly. Cyt c molecules in those layers are not electrochemically addressable at fast scan rates [18]. Thus, a good communication between Cyt c molecules and the electrode surface was maintained when the number of bilayer is as high as 5. In this work, Cyt c molecules immobilized on five layers of (MCNs/PDDA), designated as Cyt c/(MCNs/PDDA)5/ITO, were chosen in the following experiments.

Fig. 4
figure 4

Surface coverage of electrochemical addressable Cyt c immobilized in different number of bilayers of (MCNs/PDDA)N (N = 1, 3, 5, 7)

Full size image

To further investigate the characteristics of Cyt c immobilized on the multilayer films, the effect of scan rates on the voltammetric behavior of Cyt c was studied. In the range from 0.02 to 0.1 V s−1, both the anodic and cathodic peak currents vary linearly with the potential scan rate, indicating a surface-controlled electrochemical process (Electronic Supplementary Material, Fig. S1). According to Laviron theory [24], the charge transfer coefficient (α) was calculated as 0.49. And the heterogeneous electron transfer rate constant (k s ) between the electrode and the surface modification layer was estimated to be 5.9 ± 0.2 s−1. The k s value is comparable to the value for Cyt c immobilized on the WO3 nanoparticles film modified electrode (5.57 ± 0.54 s−1) [25], and higher than those of Cyt c immobilized on nanozeolite-assembled electrode (2.2 s−1) [26], mesoporous materials (2.2 s−1) [27], and multi-walled carbon nanotubes (4.0 ± 0.2 s−1) [28]. The fast electron transfer rate indicates that the host matrix of MCNs offers an ideal microenvironment for facilitating electron transfer between Cyt c and the electrode.

Amperometric response toward H2O2 based on Cyt c/(MCNs/PDDA)5/ITO

Figure 5 showed differential pulse voltammogram of Cyt c/(MCNs/PDDA)5/ITO electrode in the absence and presence of H2O2 in 50 mM PBS (pH 7.4). The current responses increased with the H2O2 concentration, indicating the electrocatalytic activity of the immobilized Cyt c for the reduction of H2O2. The typical amperometric responses of the enzyme electrode to successive concentration changes of H2O2 were examined at pH 7.4 and the applied potential of −0.25 V. The corresponding current-time responses are shown in Fig. 6. Well-defined steady state current responses were obtained at Cyt c/(MCNs/PDDA)5/ITO electrode, and the currents increased stepwisely with successive additions of H2O2. The response time t90% (reaching 90 % of the maximum response) was less than 6 s, which indicated a fast process and the immobilized Cyt c could well catalyze the reduction of H2O2. The inset of Fig. 6 corresponds to the calibration plot of the prepared biosensor. The currents had a linear relationship with the concentration of H2O2 in the range of 5 μM to 1.5 mM with a correlation coefficient of 0.9991 (n = 22) and the detection limit was examined to be 1 μM. The analytical performances of the biosensor were summarized in Table 1 and compared with other layer-by-layer assembly-based biosensors towards H2O2 in terms of detection limit and linear response range. It can be seen that Cyt c/(MCNs/PDDA)5/ITO offers a reasonable linear range for H2O2 detection with a lower detection limit.

Fig. 5
figure 5

Differential pulse voltammograms obtained at Cyt c/(MCNs/PDDA)5/ITO electrode in 50 mM PBS (pH 7.4) containing (a) 0 μM, (b) 50 μM, (c) 100 μM, (d) 300 μM H2O2

Full size image
Fig. 6
figure 6

Typical amperometric response of Cyt c/(MCNs/PDDA)5/ITO electrode to successive addition of H2O2 into 10 mL PBS (50 mM, pH 7.4) at the applied potential of −0.25 V. Inset is the calibration curve of the responses of Cyt c/(MCNs/PDDA)5/ITO electrode to H2O2

Full size image
Table 1 Comparison of the performance of H2O2 biosensors based on layer-by-layer assembly strategy
Full size table

The apparent Michaelis-Menten constant (K m ) provides an indication of both the enzymatic affinity to the substrate and the ratio of microscopic kinetic constants. It can be calculated from the Lineweaver-Burk equation [29]:

$$ 1/{I_{{ss}}} = 1/{I_{{\max }}} + {K_m}/{I_{{\max }}}C $$

Where I ss is the steady-state current after the addition of substrate, which can be obtained from amperometric experiments, I max is the maximum current under saturated substrate condition and C is the bulk concentrate of the substrate. The value of K m can be calculated from the slope (K m /I max ) and the intercept (1/I max ) for the plot of the reciprocals of the steady-state current (I ss ) vs. H2O2 concentration (C). The K m value for the Cyt c/(MCNs/PDDA)5/ITO electrode is estimated to be 0.53 mM, indicating that the Cyt c immobilized on the (MCNs/PDDA)5 assembled electrode retains its bioactivity.

In addition, at the applied potential of −0.25 V, the influence of possible interfering species on the current response of the sensor was examined (Electronic Supplementary Material, Fig. S2). The present biosensor was free from the interferences like uric acid (UA), glucose, NO −2 , NO −3 , and SO 2−3 (<2 %). A slight interference (8.5 %) of response currents to 100 μM ascorbic acid (AA) was observed relative to 100 μM H2O2. This interference could be further reduced by using Nafion membranes on the electrode surfaces [30].

Reproducibility and stability of the Cyt c/(MCNs/PDDA)5/ITO electrode

The repeatability of biosensor was evaluated in the current response for successive addition of 100 μM H2O2 at −0.25 V with the same enzyme electrode. The relative standard deviation (R.S.D.) was 4.8 % for nine successive assays. The electrode-to-electrode reproducibility was determined with the same addition of H2O2 using 3 different enzyme electrodes. It shows an acceptable reproducibility with a R.S.D. of 3.1 %. The stability of the biosensor was also examined. No obvious change in cyclic voltammogram was observed when 100 continuous cyclic scans were repeated in 50 mM PBS (pH 7.4) with a scan rate of 0.05 V s-1. When not in use, the electrode was stored in the refrigerator at 4 °C. After 7 days, the amperometric responses of the modified electrode retained 90 % of its initial value, and after 20 days the responses remained 82 %.

Conclusions

A uniform and stable multilayer membrane of (MCNs/PDDA)N were constructed via a simple layer-by-layer assembly strategy and used as enzyme immobilization matrix to fabricate Cyt c/(MCNs/PDDA)N/ITO electrodes. The direct electrochemistry of Cyt c was achieved on the multilayer films modified electrode. Moreover, the resulted Cyt c/(MCNs/PDDA)5/ITO electrode showed a good electrochemical activity to the reduction of H2O2. The presented biosensor exhibited fast amperometric response, wide linear range, low detection limit and good reproducibility.