A Novel Nonenzymatic Hydrogen Peroxide Sensor Based on Magnetic Core-Shell Fe3O4@C/Au Nanoparticle Nanocomposite

Fe3O4@C/Au nanoparticle (AuNP) nanocomposites were prepared through electrostatic adsorption of AuNPs onto PDDA-functionalized core/shell Fe3O4@C magnetic nanospheres, which had been synthesized by a facile solvothermal method. The morphology and composition of the nanocomposites were characterized by transmission electron microscopy (TEM), scanning electron microscopy (SEM), Fourier transform infrared (FT-IR), etc. Moreover, highly electrocatalytic activity to the reduction of hydrogen peroxide (H2O2) was also exhibited on the Fe3O4@C/AuNP-modified indium tin oxide (ITO) electrode. The effect of solution pH and the modification amount of Fe3O4@C/AuNPs on the performance of electrocatalytic H2O2 reduction was investigated. Under the optimal conditions, the catalytic current showed a linear relationship with the increase of H2O2 concentration in the range of 0.007–15 mM and a detection limit of 5 μM. The H2O2 sensor showed high selectivity for H2O2 detection, which could effectively resist the interference of ascorbic acid (AA), uric acid (UA), and citric acid (CA). Finally, the H2O2 sensor was used in the real fetal bovine serum to detect H2O2 and obtained satisfactory results with the recovery values ranging from 95.14 to 103.6%.


Introduction
Magnetic nanoparticles have been popularly applied in anode material [1], degradation [2], electromagnetic wave absorption [3], separation [4], and catalysis [5]. Among these diverse magnetic nanomaterials reported before, Fe 3 O 4 NPs have been studied extensively because of their excellent magnetism, biocompatibility, and especially, highly effective electrocatalytic activity. For example, Kingsley et al. fabricated Fe 3 O 4 nanoparticle-(NP-) modified carbon paste electrode to detect ascorbic acid (AA) and folic acid (FA) at the same time, with the detection limits of 2.51 × 10 -7 and 2.01 × 10 -9 M, respectively [6]. Qu and coworkers established a novel hydrogen peroxide (H 2 O 2 ) sensor using Fe 3 O 4 NPs/chitosan composite-modified electrode, which could electrocatalyze the reduction of H 2 O 2 with a detection limit of 1.53 × 10 -6 M [7]. However, Fe 3 O 4 NPs are unstable and prone to be aggregated and settled in aqueous solution because of the magnetic dipolar-dipolar attraction, resulting in the decrease of their catalytic activity [8]. To overcome this problem, considerable efforts have been devoted to introduce stabilizers on the surface of Fe 3 O 4 NPs.
In recent years, core/shell structured magnetic nanocomposites have been proposed, with Fe 3 O 4 NPs as the core and inorganic materials or polymers as the shell. ese specific core/shell Fe 3 O 4 materials possess enhanced stability and compatibility with different charges, functions, or reactive moieties. Owing to the enlarged effective surface area and the synergetic interaction between core and shell, these nanocomposites usually combine the advantageous properties of both core and shell [9]. For example, Beitollahi and coworkers synthesized a carbon paste electrode modified with magnetic Fe 3 O 4 @SiO 2 /MWCNT nanocomposite and found it exhibiting a very good resolution between the voltammetric peaks of hydrazine and phenol, which made it suitable for the detection of hydrazine in the presence of phenol in real samples [10]. Wang and coworkers showed that the Fe 3 O 4 @ZnO nanoparticles (NPs) have excellent photocatalytic activity and it is convenient to reuse NPs in wastewater treatment owing to the magnetic Fe 3 O 4 core [11]. erefore, magnetic core/shell nanocomposites with multiple properties have been widely used.
Among these shell materials, carbon and gold are the most applied coating materials due to their excellent conductivity and biocompatibility [12,13]. In addition, on the surface of carbon shells, abundant amounts of hydroxyl (-OH) and carboxyl (-COOH) groups can effectively protect the Fe 3 O 4 NPs from aggregation and facilitate further functional modification over the surface. Hence, the applicability of Fe 3 O 4 @C has been extensively promoted. For example, Zhang and coworkers synthesized Fe 3 O 4 @C NPs, which showed dramatic ability to remove methylene blue, with a maximum adsorption of 141.3 mg·g −1 [14]. Zhang and coworkers fabricated another Fe 3 O 4 @C NPs immobilized upon the GCE which was modified with molecular imprinted TiO 2 velamen. It was then made into the uric acid sensor, which delivered a linear range of 0.3-34 μM with a detection limit of 0.02 μM [15]. Meanwhile, functionalized Fe 3 O 4 @Au NPs have also been applied in As (III) [16] and dopamine [17] sensors, respectively.
In this work, Fe 3 O 4 @C core/shell nanocomposite was synthesized by a facile solvothermal method from ferrocene as the raw material. It was not only easily separated by an external magnetic field but also possessed excellent dispersion in water. To further enhance the conductivity and catalytic activity, Au NPs were adsorbed on the surface of the obtained Fe 3 O 4 @C nanospheres by electrostatic attraction. e synthesized Fe 3 O 4 @C/Au NPs were first used to fabricate a nonenzymatic H 2 O 2 sensor and showed high electrochemical sensitivity.  4 and was adjusted to the appropriate pH value. Uric acid (UA), AA, and citric acid (CA) were all purchased from Sigma. Nafion (Nf ) was obtained from DuPont as 5 wt% solution and was used as 0.5 wt% solution after dilution with water. Other reagents were of analytical grade and used without further purification. Double distilled water (DDW) was used throughout the experiments.

Characterization.
e structures and morphologies of the Fe 3 O 4 @C and Fe 3 O 4 @C/AuNPs were characterized by transmission electron microscopy (TEM, JEOL JEM-200CX) and field emission scanning electron microscopy (FESEM, Hitachi S4800). Fourier transform infrared (FT-IR) spectra of KBr powder-pressed pellets were recorded on a Bruker model VECTOR 22 Fourier transform spectrometer. Elemental composition and chemical valence analyses were performed on X-ray diffraction (XRD, Rigaku SmartLab) and X-ray photoelectron spectroscopy (XPS, ermo Scientific K-Alpha), respectively.
Cyclic voltammetry (CV), differential pulse voltammetry (DPV), and electrochemical impedance spectroscopy (EIS) were performed using a CHI 660D electrochemical workstation (Shanghai Chenhua Instruments, China). A threecompartment electrochemical cell contained a saturated calomel reference electrode (SCE), a platinum wire auxiliary electrode, and a modified ITO working electrode with a 3 mm diameter controlled area. e determination of H 2 O 2 was conducted in 0.1 M PBS, and the sensor responses were measured under deductions for background currents from the total ones.

Preparation of Fe 3 O 4 @C/AuNP Magnetic Nanocomposite.
Core/shell structured Fe 3 O 4 @C magnetic nanocomposite was prepared by Wang's method [18]. First, 0.3 g of ferrocene was dissolved in 30 mL of acetone under ultrasonication at room temperature for 30 min. en, 1 mL of H 2 O 2 was added dropwise into the above ferrocene solution. After reacting for 15 min under stirring, the resulting solution was transferred into a Teflon-lined stainless-steel autoclave with a capacity of 50 mL and heated at 210°C for 24 h. After cooling to room temperature, the product was magnetically separated, washed several times with absolute ethanol, and finally dried under vacuum at 50°C for 5 h.
Au colloid solution was prepared by adding 5 mL of 1% (wt. %) sodium citrate into 200 mL of 0.01% (wt. %) HAuCl 4 , which had been heated to boil beforehand. Heating was stopped by the change of solution color from light blue to violet-red and stirring was continued for 15 min. After cooling to room temperature, Au colloid solution was preserved in a refrigerator at 4°C. Generally, an electrostatic attraction method was used for the immobilization of AuNPs onto the surface of the Fe 3 O 4 @C core/shell nanocomposite. 50 mg of Fe 3 O 4 @C nanocomposite was dispersed into 30 mL of 0.2% PDDA that contained 20 mM Tris base and 20 mM NaCl, and the mixture was stirred for 20 min.
e product was collected with the help of a magnet and rinsed with DDW six times to remove the residual PDDA. e positively charged Fe 3 O 4 @C nanocomposite was dispersed in 100 mL of Au colloid solution and stirred for 8 h.
e suspension was separated with the help of a magnet. After rinsed with water and ethanol three times successively, the Fe 3 O 4 @C/AuNPs nanocomposite was dried under vacuum at 50°C for 6 h.

Fabrication of Fe 3 O 4 @C/AuNP-Modified ITO Magnetism
Electrode. Before modification, the ITO chips were cleaned with acetone, ethanol, and water, respectively, and then dried under a stream of nitrogen. e active electrode area was confined by bounding a circular opened 3M polyvinyl chloride (PVC) film on the ITO surface, with a 3 mm diameter on the middle of the electrode. On the other side of the ITO, a piece of 8 mm diameter and 5 mm height permanent magnet was adhered by silica rubber adhesive [19]. en, 10 µL of 1 mg·mL −1 Fe 3 O 4 @C/AuNP nanocomposite dispersion (in 5 mL 0.5% Nf ) was dropped onto the surface of the ITO magnetism electrode and dried at room temperature, which was designated as Fe 3 O 4 @C/AuNPs/Nf/ ITO. When the sensor was not in use, it was stored at 4°C in a refrigerator.

Structure and Composition Characterization of Samples.
e structure and morphology of the nanocomposites were investigated by TEM and FESEM. In Figure 1  e characteristic peak appears at 590 cm -1 which is related to the Fe-O stretching vibration [20]. Compared with curve a, there is a new strong band in curve b at around 3446 cm -1 , indicating the presence of -OH in the carbon shell. ese phenomena further proved that the carbon shell is successfully coated on the surface of the Fe 3 O 4 nanosphere. e peak located at 1399 cm −1 is attributed to the COO-symmetric vibration [21]. Actually, the carboxyl group-functionalized Fe 3 O 4 @C nanospheres are negatively charged. is is the reason why we use PDDA to make the surface of carbon shell positively charged for the assembly of negatively charged AuNPs via electrostatic interaction. Curve c shows the FT-IR spectrum of Fe 3 O 4 @C/AuNPs. Because AuNPs do not have absorption in the IR region, the characteristic peak of Fe 3 O 4 @C/AuNPs is almost the same as that of Fe 3 O 4 @C.
To characterize the crystal information of the prepared Fe 3 O 4 @C/AuNPs further, the XRD has been performed. As shown in Figure 2(b), the diffraction peaks appeared at 30°, 35.5°, 43°, 57°, and 62°are originated from Fe 3 O 4 , corresponding to (220), (311), (400), (511), and (440) crystal planes [22]. In addition to those peaks, another peak at 38.2°i s in good accordance with (111) plane of AuNPs. It is difficult to discriminate Fe 3 O 4 phases from c-Fe 2 O 3 by only using XRD patterns due to the similarity of patterns, and XPS is conducted to characterize the electronic structures of Fe element in the nanocomposite, which is known to be sensitive to Fe 2+ and Fe 3+ cations. Figure 2(c) presents two peaks at the binding energies of 724.4 and 711.4 eV, which correspond to the maximum intensity of Fe 2p1/2 and Fe 2p3/2, illustrating the coexistence of Fe 2+ and Fe 3+ cations in Fe 3 O 4 @C/AuNPs [23]. Peaks of Au are also shown in Figure 2(d), corresponding to Au 4f2/7 and Au 4f2/5, respectively. From the above characterization results, the welldefined core/shell structured Fe 3 O 4 @C/AuNPs is confirmed to be constructed.

Characterization of Nanocomposite-Modified ITO
Electrodes.
e influence of potential scan rate (v) on the response of the Fe 3 O 4 @C/AuNP-modified ITO electrode in PBS was studied, as shown in Figure 3(a). e reduction and oxidation peak currents (i pa and i pc ) increased linearly with the square root of scan rate (v 1/2 ) accelerating from 40 to 300 mV·s -1 , indicating a surface-controlled electrochemical process.
As shown in Figure 3(b), we also recorded the CVs of various modified electrodes towards 0.1 mM H 2 O 2 in 0.1 M pH 7.4 PBS at a scan rate of 100 mV·s -1 . At the bare ITO, almost no peak of H 2 O 2 is observed (curve a), except for a slight reduction peak. After the electrode was modified with Fe 3 O 4 @C nanocomposites, a small increase in i pa was observed (curve b), indicating the electrocatalytic reduction activity toward H 2 O 2 possessed by Fe 3 O 4 @C nanocomposites. e reduction peak potential was around −0.35 V, similar to another reported nonenzymatic H 2 O 2 sensor [24] and the i pa was 1.7 times higher compared with that in curve a. When the Fe 3 O 4 @C/AuNPs were assembled on the electrode, the peak current significantly increased (curve c) and the i pa is 2.4 times that of curve b, demonstrating that the AuNPs significantly improve the catalytic activity of Fe 3 O 4 @C nanocomposites. e electron transfer rate and the interfacial resistance, as shown in Figures 4(a) and 4(b), respectively, can reflect the electrochemical properties. In Figure 4(a), obvious oxidation and reduction peaks were observed on both Fe 3 O 4 @C-and Fe 3 O 4 @C/AuNP-modified electrodes, belonging to different valences of Fe element. An increased peak current was found after surface modification with AuNPs, indicating that noble metals could promote electron transfer on the electrode. In addition, the electroactive surface area (ESA) was calculated to be 1.628 cm 2 by Randles-Sevcik equation, which is nearly 23 times the geometric area. It means that the modification of Fe 3 O 4 @C/AuNPs increased the roughness of the electrode surface greatly, which could facilitate the absorption, transmission, and catalysis of H 2 O 2 .
Similar results were also confirmed by EIS measurements, as shown in Figure 4

Optimization of Experimental Parameters.
e variations of oxidation-reduction potential of H 2 O 2 in different pH values indicate diverse oxidation ability of H 2 O 2 , resulting in different sensitivities of sensors under different pH circumstances. e effect of solution pH on the electrocatalytic reduction of H 2 O 2 was investigated. Figure 5(a) shows the linear plots of the reduction peak current versus the concentration of H 2 O 2 from 0.1 to 1 mM over the pH range from 6 to 9. e electrocatalytic activity and sensitivity were much higher at the pH value of 7-8 and lower at 6 and 8. e possible reason was that Fe 3 O 4 core was expected to dissolve in acidic circumstances and the catalytic activity faded. Meanwhile, H 2 O 2 is prone to lose oxidizing ability with solution alkalinity increasing. erefore, a nearly neutral condition is suitable for H 2 O 2 detection and the PBS with physiologic pH value of human being which is about 7.4 was chosen as the supporting electrolyte. Figure 5(b) exhibits reduction peak current versus different modification amounts of Fe 3 O 4 @C/AuNPs. e response current towards 1 mM H 2 O 2 rose along with the modification amount of Fe 3 O 4 @C/AuNPs increasing from 0.6 to 1 mg·mL -1 and reached a plateau when the modification amount exceeded 1 mg mL -1 . e response speed was    to be delayed by thickening modification, which brought hidden troubles of its falling off from the electrode surface. It may lead to a decreased stability of the modified electrode and the smaller response current. In this way, 1 mg·mL -1 of Fe 3 O 4 @C/AuNPs was regarded to be the optimal amount of modification for H 2 O 2 detection. Figure 6 displays CVs in the presence of H 2 O 2 at different concentrations under the optimum conditions. e response current shows a good linear relationship with the increase of H 2 O 2 concentration in the range of 0.007-15 mM. e linear regression equations is ∆i p (µA) � 0.10 + 2.01 c (H 2 O 2 ) (mM) (R 2 � 0.9988). e detection limit was estimated at 5 µM, and the theoretical sensitivity of this sensor was calculated to be 28.71 µA mM −1 cm −2 . e response current became saturated when the concentration of H 2 O 2 exceeded 15 mM. Quantitative detection of H 2 O 2 concentration can be realized by the calculation with the linear equation. Furthermore, the performances including electrode modification, linear range, and detection limit of the Fe 3 O 4 @C/ AuNPs/ITO were compared with those of other H 2 O 2 sensors, which are listed in Table 1. It was shown that the proposed H 2 O 2 sensor in this work exhibited a wider linear range and a

Stability and Repeatability.
e application of chemical sensors put forward higher demands on stability and catalytic signal repeatability. DPV and EIS measurements for  Figure 7, it was found that the DPV peak remained nearly unchanged after 10 cycles of scanning. On the other hand, the Nyquist plots nearly overlapped for 10 times of detection, and the R ct value maintained around 152 Ω. It demonstrated that the immobilized Fe 3 O 4 @C/ AuNP modification layer could keep a good level of stability in a continuous scanning system.
In addition, a repeated measurement towards 1 mM H 2 O 2 in 0.1 M pH 7.4 PBS was conducted to verify the repeatability of the sensor. e response signal recorded on 5 parallel electrodes is displayed in Figure 7(c), and the relative standard deviation (RSD) value between them was calculated as 1.8%. e good stability and repeatability of the fabricated H 2 O 2 sensor are attributed to the simple modification method, stable property, and good catalytic ability of Fe 3 O 4 @C/AuNPs.

Interference Study.
To evaluate the selectivity of Fe 3 O 4 @C/AuNPs, an interference study was carried out. As it is known, AA, UA, and CA are the most common interfering electroactive species during the detection of H 2 O 2 . e current responses with the addition of H 2 O 2 ,  AA, UA, and CA at the potential of −0.35 V are shown in Figure 8, among which the responses caused by AA, UA, and CA interfering species were no more than 10% response of H 2 O 2 . e result implied that the Fe 3 O 4 @C/ AuNPs possessed high selectivity for H 2 O 2 detection at the potential of −0.35 V.

Determination of H 2 O 2 in Fetal Bovine Serum (FBS).
e determination of H 2 O 2 in FBS was also investigated. e FBS was diluted with 0.1 M of pH 7.4 PBS via the volume ratio of 1 : 9. We measured the recovery by adding different concentrations of H 2 O 2 into the FBS dilution. e recovery values were calculated ranging from 95.14 to 103.6%, which are presented in Table 2. e relative standard deviations (RSDs) are less than 5%, proving a good application prospect.

Conclusion
In summary, a novel core-shell structured Fe 3 O 4 @C/AuNPs nanocomposite is successfully synthesized and utilized for fabricating a nonenzymatic H 2 O 2 sensor, exhibiting a wider linear range from 0.007 to 15 mM and a lower detection limit of 5 μM with a sensitivity of 28.71 µA mM −1 cm −2 . e advantages of the proposed sensor over the previous works are concluded as four points: (1) the superparamagnetic property of Fe 3 O 4 @C/AuNPs nanocomposite contributes to the construction of a magnetic electrode modification,  (3) e C/AuNPs bilayer possesses hydrophilicity, biocompatibility, electroactivity, and conductivity, facilitating the substance transmission and electron transfer between bulk solution and electrode surface. (4) e detection of H 2 O 2 in FBS illustrates high selectivity, displaying good potential for real applications. Of course, the fabricated Fe 3 O 4 @C/AuNPs nanocomposite could also be used as substrate or nanolabel in other sensing systems.
Data Availability e data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
e authors declare no conflicts of interest.