In-Situ Construction Molecular Imprinting Electrocatalyst of Au-MoO₃/Graphene for Bisphenol A Determination with Long-Term Stability

Abstract

Molecular imprinting (MI) technology has been used in electrochemical analysis technology because of its unique selectivity and specificity. In this work, an electrochemical sensor based on in-situ inorganic MI-Au-MoO₃/graphene for bisphenol A (BPA) analysis is designed, where MI-MoO₃ is hybridized with graphene nanosheets and Au nanoparticles, and BPA is acted as the temple molecular. Differential pulse voltammetry (DPV) was used to evaluate the sensing performance of the MI-Au-MoO₃/rGO sensor toward BPA determination and it is about 2.0 times that of NI-Au-MoO₃/rGO. The as-constructed sensor presents a wide linear range from 0.01 to 106.04 μM and a low limit of detection of 0.003 μM. It also displays outstanding stability and repeatability up to 20 days, and can be used to analyze the content of BPA in dust leachate and plastic bottle. This sensor offers a promising strategy for environment pollution and food analysis via MI technology.

1. Introduction

Bisphenol A (BPA) is one of the endocrine disrupter chemicals, which can generate many adverse impacts on the biological systems of organisms [1,2,3,4]. Recently, it is inevitable to uptake trace amounts of BPA from daily life on account of the wide application of BPA-based plastic production. It is always used as a raw material to produce food and drink packages, baby bottles and dental sealants, where BPA molecules could be released into food, drinks, and the environment [5,6,7,8,9]. Owing to the serious threat to human health, it is important to develop a simple, high-sensitivity, and selective analytical method for rapid determination of trace BPA in the field of environmental monitoring.

Recently, molecularly imprinted electrochemical sensor (MIES), combined with the merits of MI technology and electrochemical detection, has attracted extensive attention in the electrochemical analysis field [10,11,12]. MI technology, as a technique for the special identification of imprinting molecules, can leave a large number of molecular template sites for the detection to improve the selectivity and the simplicity rapidity of the target molecule [13,14,15]. For example, Lu et al. have developed the MIES of MXene/amino carbon nanotubes for the detection of fisetin [14]; Gao et al. reported an MI-Nb₂O₅ for detecting BPA [16].

Two-dimensional (2D) nanomaterials are exciting for researchers in MIES due to their excellent properties [17,18,19]. The high specific surface area of 2D nanomaterials is beneficial to creating more effective target sites located at the surface or near the surface of the materials by controlling the templates, creating more effective target sites [20]. The frequently used 2D sensing materials include carbon materials [21,22,23,24], transition metal oxides (TMOs) and transition metal chalcogenides (TMCs) [25,26,27,28,29]. Among them, 2D molybdenum trioxide (MoO₃) receives widespread attention in catalysis [30], energy storage [31], and gas and biomedical sensors [32,33] areas, which could be attributed to its unique high stability, conspicuous catalytic activity, excellent electrical and surface charge properties, and easy to fabricate [34]. It was only much more recently that Antoniazzi et al. developed an orthorhombic molybdenum trioxide (α-MoO₃)-modified carbon paste electrode for monitoring BPA leaching from commercial plastic samples to water [35]. These preliminary findings demonstrate the potential of 2D MoO₃ in the electrochemical detection of endocrine disruptors. From this starting point, we can use reduced graphene oxide (rGO) as an electron bridge to further enhance the electron transport capacity of MoO₃ for (bio-) chemical sensing [36], thanks to its high specific surface area and prominent electrical conductivity.

Considering the excellent conductivity and special surface area of rGO and the perfect biocompatibility of Au nanoparticles, decorating MI-MoO₃ with rGO and Au nanoparticles is prepared as a novel electrochemical sensor, marked as “MI-Au-MoO₃/rGO”. The as-constructed sensing platform was used to determine the content of BPA by analyzing the change of current by differential pulse voltammetry technique. The sensor presents a wide linear range and a low limit of detection (LOD). It also shows outstanding stability, and can be employed to analyze BPA content in practical sample including dust leachate, plastic bottle, and tap water with a satisfactory result.

2. Results and Discussion

2.1. Morphological and Structure Characterization

The morphologies of as-synthesized MoO₃ and Au-MoO₃/rGO were characterized via SEM. As shown in Figure 1a, the as-synthesized MoO₃ was shown a sheet-like structure in general. Further introduced Au nanoparticles and rGO into MoO₃, the sheet-like structure was not changed (Figure 1b), only some tiny particles appeared on the surface of both sheet-like structures. XRD patterns were used to explore the crystal structures and crystalline phases of all samples in Figure S1a. It can be seen that all characteristic peaks of two MoO₃ based samples were matched up well with the orthorhombic α-MoO₃ (JCPDS No. 05-0508) [37], indicating that introduction of rGO and Au did not change the structure of MoO₃. There is no distinct diffraction peak of rGO in the XRD pattern of Au-MoO₃/rGO owning to the trace content of rGO [38]. For Au-MoO₃/rGO, three diffraction peaks at 38.2°, 44.4°, and 64.6° were detected apart from the peaks of MoO₃, which could be severally ascribed to the (111), (200), and (220) planes of Au [38], indicating the successful formation of Au nanoparticles in the composites of Au-MoO₃/rGO.

Moreover, Fourier transforming infrared spectra (FT-IR) was employed to explore the organic composition of MoO₃ in the fabrication process. As shown in Figure 1c, For BPA and MI-MoS₂, the FT-IR spectrum displayed three typical absorption peaks. The peaks at 1508 cm⁻¹ and 1603 cm⁻¹ assign to aromatic rings framework vibration while the peak at 2969 cm⁻¹ assigns to C-H vibration. However, those peaks were not appeared in NI-MoS₂, which indicates BPA molecules were successfully in-situ grown on the surface of MoS₂. Furthermore, that the peaks stood on the 3376 cm⁻¹ and 1603 cm⁻¹ could be ascribed to the stretching and bending mode of O-H active groups, which was beneficial to bind specifically BPA molecules through hydrogen bonding interaction. Additionally, these characteristic peaks were totally disappeared after the annealing process in MI-MoO₃, which indicates that BPA molecules were removed and specific recognition sites could be left on the surface of MoO₃ [16,39].

X-ray photoelectron spectroscopy (XPS) was used to detect the elements of MoO₃ and Au-MoO₃/rGO. In addition to the signals of the O 1s and Mo 3d, a new peak can be indexed to Au 4f in Au-MoO₃/rGO (Figure S1b), indicating Au nanoparticles were successfully introduced on the surface of MoO₃/rGO. The Mo 3d spectrum (Figure 1d) of pure MoO₃ exhibits two major peaks at 231.8 eV and 234.9 eV, which was attributed to Mo 3d_5/2 and Mo 3d_3/2, respectively, corresponding to Mo⁶⁺. Interestingly, the Mo 3d peaks in Au-MoO₃/rGO shifted to higher binding energies compared to pure MoO₃. The deconvolution of the O 1s spectrum (Figure 1e) shows two peaks: the lattice oxygen in MoO₃ at 529.5 eV and the interstitial oxygen of MoO₃ at 531.4 eV [40]. Same with Mo 3d, the peak signals of O 1s in Au-MoO₃/rGO also shifted to a higher position at 529.9 and 531.7 eV, respectively. This phenomenon could be ascribed to the interaction of among Au, MoO₃ and rGO, causing the surface electrons of MoO₃ to transfer to Au [41]. Furthermore, the peak signals at 83.6 and 87.3 eV were severally attributed to Au 4f_7/2 and Au 4f_5/2 (Figure 1f) [42], indicating that the Au nanoparticles were triumphantly loaded on the MoO₃/rGO sample.

2.2. Electrochemical Behaviors

The electrochemical activity of different electrodes was studied by cyclic voltammetry (CVs) and electrochemical impedance spectroscopy (EIS). MI-Au-MoO₃/rGO electrode possesses a bigger oxidation peak current density compared to NI-Au-MoO₃/rGO electrode in Figure 2a. Moreover, a pair of redox peaks with peak-to-peak separation (ΔE_p) of 103 mV was observed at MI-Au-MoO₃/rGO, which was smaller than the ΔEp of NI-Au-MoO₃/rGO (160 mV). For comparison, pure samples such as GCE, Au/GCE, and rGO/GCE were also probed, and they showed only a weak signal. This phenomenon could be ascribed to the excellent conductivity and the increased active sites of MI-Au-MoO₃/rGO after removing the template. The Nyquist plots were used to explore the electron transfer processes of different sensors in Figure 2b. Generally, the smaller diameter of the arc (DIA) means a faster charge transfer rate, and these results also demonstrate a similar trend with CVs. The simulated circuit diagram was used to analyze the EIS data (Figure 2c), and R_s, R_ct, Q, and Z_w represent the resistance of the solution, the charge transfer resistance value between the electrode and solution, the electrode double-layer capacitance and Warburg impedance, respectively [43]. The results show that the R_ct values of GCE, Au/GCE, rGO/GCE, NI-Au-MoO₃/rGO/GCE and MI-Au-MoO₃/rGO/GCE can be estimated to 5.31 × 10⁴ Ω, 3.01 × 10⁴ Ω, 9.97 × 10³ Ω, 4.95 × 10³ Ω and 9.38 × 10² Ω, respectively, indicating that MI-Au-MoO₃/rGO sensor has good electronic transmission performance [41].

The electrochemical behavior of the oMI-Au-MoO₃/rGO electrode toward BPA was explored. As shown in Figure 2d, for NI-Au-MoO₃/rGO, the electro-oxidation potential of BPA appeared at 0.48 V and the relative oxidation peak current was 11.54 μA. However, for MI-Au-MoO₃/rGO, the enhanced oxidation peak occurred at a lower potential (0.475 V), meanwhile, the corresponding oxidation peak current reached 23.05 μA, which improved 2.0 times compared with NI-Au-MoO₃/rGO. These indicate that the MI-Au-MoO₃/rGO electrode displays a better electrochemical response for BPA. This may be due to some special activate sites left on the surface of MI-Au-MoO₃/rGO after annealing, which can selectively adsorb BPA molecules to enhance the electrochemical response ability towards BPA [16].

2.3. Optimizing Conditions of Detection

The accumulation is an important factor for the electrochemical sensor. The oxidation peak current was increased dramatically with increasing accumulation time from 0 to 180 s (Figure 3a). The surface amount of BPA was accumulated with increasing time and thus enhance the oxidation signal of BPA [44]. However, the current signal significantly decreases when exceeded 180 s owing to that BPA at the electrode rapidly reaching saturation. Meantime, the optimization for accumulation potential was also investigated in the range of −0.4 V to 0.3 V (Figure 3b). The highest oxidation peak current density is achieved at −0.1 V. From the above results, 180 s and −0.1 V were selected as the optimum accumulation time and potential, respectively. Finally, the amount of MI-Au-MoO₃/rGO loading was also explored. In the experiment, the concentration of MI-Au-MoO₃/rGO was immobilized at 2 mg mL⁻¹, and the suspension volume on the modified electrode was optimized. It could be obviously concluded from Figure 3c that the oxidation peak current of BPA increases with increasing the amount of MI-Au-MoO₃/rGO concentration from 1 to 3 μL. When further increasing the amount of MI-Au-MoO₃/rGO, the oxidation peak current of BPA dramatically decreases. This phenomenon can be ascribed to the limited electron transport ability in the oxidation of BPA in a thicker film, which led to the diminution of oxidation peak current. Hence, 3 μL of 2 mg mL⁻¹ of MI-Au-MoO₃/rGO suspensions was chosen to modify the electrode throughout the experiment.

The effects of pH were displayed by DPV from 5.2 to 8.7 (Figure 4a). As can be seen in Figure 4b (red line), the oxidation peak current steadily increased with the pH value raised from 5.2 to 7.0, while the oxidation current decreased on the contrary when the solution pH exceeded 7.0. Moreover, the pKa of BPA (pKa = 9.73) was distinctly higher than the optimized pH value, indicating that the absorbing capacity of on MI-Au-MoO₃/rGO electrode surface towards undissociated BPA was better than the dissociated molecules in neutral solution [45]. In consideration of achieving higher sensitivity, the phosphate buffer with pH 7.0 was chosen as the optimal supporting electrolyte for determining BPA. It can be seen that a good linear correlation of oxidation peak potential and pH value (Figure 4b, black line), the peak potential of BPA was negatively shifted with the increase was accompanied by pH value. As well, the linear regression equation was E_pa (V) = −0.0595 pH + 0.8797 (R² = 0.9942). The obtained slope of 59.5 mV pH⁻¹ was approximately to the theoretical value (59 mV pH⁻¹), revealing that there had equal numbers of transferred electrons and protons on the oxidation process of BPA on the MI-Au-MoO₃/rGO electrode [46]. Combined with the previous conclusion that there were two transferred electrons during BPA electro-oxidation process [43], the oxidation of BPA on MI-Au-MoO₃/rGO could be a two-electron and two-proton process.

2.4. Effect of Scan Rates

The effect of scan rates was investigated by CV (Figure 5a). The oxidation peak currents gradually increase with increasing the scan rate. A good linear relationship of oxidation peak current (I_pa) versus square root of scan rates was obtained in Figure 5b, and the linear regression equation was fitted as follows: I_pa (μA) = 5.03 v^1/2 (mV s⁻¹) − 21.5 (R² = 0.995), which indicates that the electrochemical oxidation of BPA at MI-Au-MoO₃/rGO electrode was a typical diffusion-controlled process. Figure 5c displays the dependence of peak potential (E_pa) with ln v, and the regression equation is described as E_pa (V) = 0.0272 ln v (mV s⁻¹) + 0.4224 (R² = 0.9934). With regard to a diffusion-controlled and completely irreversible electrode process, the Laviron equation is usually used to define the E_pa, and the equation is exhibited as follows:

$$E_{pa}=E^{\theta }+(\raisebox{1ex}{$RT$}\!\left/ \!\raisebox{-1ex}{$\alpha nF$}\right.)\ln (\raisebox{1ex}{$RT{k}^{\theta }$}\!\left/ \!\raisebox{-1ex}{$\alpha nF$}\right.)+(\raisebox{1ex}{$RT$}\!\left/ \!\raisebox{-1ex}{$\alpha nF$}\right.)\ln v$$

(1)

where R, T, and F have usually meanings (R = 8.314 J mol⁻¹ K⁻¹, T = 298 K, F = 96,480 C mol⁻¹), n is the number of transfer electrons, E^θ, k^θ and α represent formal redox potential, standard rate constant in the reaction and transfer coefficient, respectively. As for α in the completely irreversible electrode process, in general, speaking, is assumed to be 0.5. According to Equation (1), the slope is equivalent to RT/αnF [45]. Hence, the electron transfer number (n) is determined to be 1.89. These indicate BPA electro-oxidation is a two-electron transfer process.

2.5. Determination of Detection Limit, Stability and Real Samples Analysis

Under the optimal experimental conditions, the MI-Au-MoO₃/rGO sensor was employed to detect different concentrations of BPA (Figure 6a). The peak current increases with the continuous addition of BPA. Moreover, the linear relation of oxidation peak current vs. the logarithm of the BPA concentration from 0.01 μM to 116.04 μM was shown in Figure 6b. The corresponding linear regression equation is I_pa (μA) = 9.335 + 1.689 log C (μM) (R² = 0.9468) and I_pa (μA) = 6.057 + 8.083 log C (μM) (R² = 0.9938), respectively. The LOD was estimated to be 0.003 μM at a ratio of signal to noise of 3 (S/N = 3). The sensor shows lower LOD, compare with previous reports (Table S1). The lower LOD of the MI-Au-MoO₃/rGO sensor could be explained: i, BPA template molecules were in-situ grew in the hydrothermal process of MoS₂ nanomaterials, after annealing treatment, the surface of MoO₃ nanosheets could generate a mass of peculiar recognition active sites with special shapes, sizes, and functional groups, ii. The peculiar recognition active sites would improve the adsorption of BPA, thereby enhancing the electrochemical response of BPA.

The stability was investigated by DPV (Figure 6c). MI-Au-MoO₃/rGO was kept in the refrigerator at 4 ℃ during the test period. When used per 5 days, the current response was reduced by ca. 19%, indicating this sensor has excellent stability for up to 20 days. The reproducibility was measured by the DPV technique through 5 different sensors in Figure S2 and all sensors showed a relative standard deviation (R.S.D.) of 2.69%. This result shows a good reproducibility for BPA detection at the prepared sensor. The sensor was also used to detect actual samples including dust leachate, drinking water bottles and tap water by using the standard addition technique via DPV (Table S2), and the recovery ranged from 99.9% to 101.5% as well as the R.S.D. was less than 2.8%, suggesting the sensor can be applied for detecting real samples.

3. Experiments

The Materials and characterizations involved in this work are discussed in the Supplementary Material (Text S1 and S2).

3.1. In-Situ Synthesis of MI-MoO₃ Nanosheets

Firstly, 0.25 mmol of (NH₄)₆Mo₇O₂₄·4H₂O, 0.75 mmol of CH₄N₂S, and a certain amount of BPA molecules dissolved in ethanol were ultrasonically dispersed into 20 mL of distilled water. The resulting solution was transported into 25 mL Teflon autoclave and heated at 210 ℃ for 24 h. In the process of hydrothermal reaction, BPA molecules could be adsorbed on the surface of MoS₂. After cooling down to room temperature, the black precipitates were subsequently obtained by centrifugal as well as washed with water and ethanol several times, respectively. The powders were dried in an oven at 60 ℃ overnight. Then, in order to remove the template molecules of BPA, the sediments were annealed at 500 ℃ for 2 h (Scheme 1), obtaining the target product of MI-MoO₃. For comparison, non-imprinted MoO₃ (NI-MoO₃) was prepared except for adding BPA molecules in the same synthesis process.

3.2. Synthesis of MI-Au-MoO₃/rGO

The MI-Au-MoO₃/rGO was synthesized as follows: 20 mg of as-prepared MI-MoO₃ nanosheets and 0.21 mL of HAuCl₄ (0.0486 M) were dispersed into 15 mL of ethylene glycol solution containing 2.0 mg GO under ultrasonic processing for 30 min. After adjusting the pH value of the above solution to 10.5 via 0.1 M NaOH solution, the homogeneous solution was transferred to a Teflon-lined stainless-steel autoclave (25 mL) and heated at 140 ℃ for 4 h. After completion of the reaction, the black precipitates were centrifuged and washed with double distilled water to neutral and then dried at 60 ℃ overnight, obtaining MI-Au-MoO₃/rGO (Scheme 1). For comparison, NI-Au-MoO₃/rGO was synthesized in the same method, only replacing MI-MoO₃ with NI-MoO₃.

3.3. Preparation of Sensor

For fabricating the sensor, the as-obtained MI-Au-MoO₃/rGO (2 mg) powder and Nafion (5%, 10 μL) were dispersed into 1 mL ethanol/water mixed solution (V_ethanol:V_water = 1) by ultrasonic agitation. Meanwhile, a glassy carbon electrode (GCE) (Diameter = 3 mm, geometric area = 0.07 cm²) was mechanically polished with 0.05 um α-Al₂O₃ powder to be a bright mirror-like surface and thoroughly washed with water as well as ethanol. Then, 3 μL MI-Au-MoO₃/rGO dispersion solution (2 mg/mL) was coated on the GCE surface and dried in the air to evaporate the solvent, resulting in a sensor of MI-Au-MoO₃/rGO electrode (Scheme 1). For comparison, NI-Au-MoO₃/rGO electrode was constructed by a similar method.

3.4. Electrochemical Measurement

Electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) were carried out in 0.1 M KCl solution containing 2.5 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆] mixture. Unless otherwise stated, 0.1 M phosphate buffer (pH 7.0) was used as the determining medium for BPA. CVs with different scan rates were recorded from 0.2 to 0.8 V. Differential pulse voltammetry (DPV) measurement was operated from 0.2 to 0.8 V with a pulse time of 0.05 s, pulse amplitude of 0.05 V, pulse period of 0.5 s and incremental E of 0.004 V. All the electrochemical experiments were measured at room temperature.

4. Conclusions

A new inorganic MI-Au-MoO₃/rGO electrochemical sensor was successfully prepared for detecting BPA. In the preparation process, BPA molecules can be successfully combined and removed from the Au-MoO₃/rGO nanocomposites, which offer lots of recognition sites on the surface of MI-Au-MoO₃/rGO for BPA. The sensor presents a wide linear range from 0.01 to 106.04 μM and a low LOD of 0.003 μM. It also shows excellent long-term stability up to 20 days, and can be applied to analyze BPA in real samples. This simple-to-operate method offers a new window for great potential application in environmental pollution and food health monitoring.