An Efficient Voltammetric Sensor Based on Graphene Oxide-Decorated Binary Transition Metal Oxides Bi2O3/MnO2 for Trace Determination of Lead Ions

Herein we present a facile synthesis of the graphene oxide-decorated binary transition metal oxides of Bi2O3 and MnO2 nanocomposites (Bi2O3/MnO2/GO) and their applications in the voltammetric detection of lead ions (Pb2+) in water samples. The surface morphologies, crystal structures, electroactive surface area, and charge transferred resistance of the Bi2O3/MnO2/GO nanocomposites were investigated through the scanning electron microscopy (SEM), power X-ray diffraction (XRD), cyclic voltammetry (CV), and electrochemical impedance spectroscopy (EIS) techniques, respectively. The Bi2O3/MnO2/GO nanocomposites were further decorated onto the surface of a glassy carbon electrode (GCE), and Pb2+ was quantitatively analyzed by using square-wave anodic stripping voltammetry (SWASV). We explored the effect of the analytical parameters, including deposition potential, deposition time, and solution pH, on the stripping peak current of Pb2+. The Bi2O3/MnO2/GO nanocomposites enlarged the electroactive surface area and reduced the charge transferred resistance by significant amounts. Moreover, the synergistic enhancement effect of MnO2, Bi2O3 and GO endowed Bi2O3/MnO2/GO/GCE with extraordinary electrocatalytic activity toward Pb2+ stripping. Under optimal conditions, the Bi2O3/MnO2/GO/GCE showed a broad linear detection range (0.01–10 μM) toward Pb2+ detection, with a low limit of detection (LOD, 2.0 nM). The proposed Bi2O3/MnO2/GO/GCE electrode achieved an accurate detection of Pb2+ in water with good recoveries (95.5–105%).


Introduction
As a common heavy metal ion (HMI), Pb 2+ has adverse effects on our health and the environment because of its high toxicity, even at low concentrations. Pb 2+ in the aquatic environment barely degrades and is easily enriched in aquatic food [1,2]. Therefore, Pb 2+ chronically endangers human health via the food chain and gradually induces lifethreatening circumstances. Excessive levels of Pb 2+ in human body can severely destroy our organs and nervous system, which is highly associated with various cancers such as lung, kidney, and brain cancers [3]. Hence, a highly efficient determination of Pb 2+ in water is quite essential to guarantee our health.
Over the last few decades, conventional analytical techniques have been developed to reliably detect Pb 2+ , including X-ray fluorescence spectrometry [4], UV-Vis spectroscopy [5], atomic absorption spectrophotometry [6], inductively coupled plasma mass spectrometry [7], and inductively coupled plasma-atomic emission spectrometry [8]. These analytical techniques are very robust and accurate, even in complex sample matrixes; however, they often require expensive and bulky equipment, cumbersome and time-consuming operation procedures, and highly skilled personnel. Without a doubt, they are not suitable for an on-field analysis. In recent years, stripping voltammetry, especially SWASV, has emerged as a powerful alternative for the trace determination of HMIs due to its advantages of portability, low cost, rapid response, excellent sensitivity, and feasibility for on-site analyses [9,10]. A voltammetric determination of Pb 2+ often involves hanging mercury drop electrodes or mercury film electrodes. Owing to their superior stripping characteristics, these mercurybased electrodes are excellent in their sensitivity and reproducibility [11,12]. However, the toxic mercury contaminates samples and poses health risks to analysts. Alternatively, eco-friendly bismuth film electrodes can provide comparable sensing properties for HMI determination [13,14]. Unlike mercury electrodes, bismuth film electrodes usually suffer from surface passivation, which degrades their stripping signals. Therefore, designing novel materials with extraordinary sensing performance toward Pb 2+ is highly desirable and challenging.
Transition metal oxide nanostructures have been extensively used for the voltammetric detection of HMIs because of their natural abundance, high adsorption capacity, and favorable catalytic activity [15,16]. Among transition metal oxides, MnO 2 has attracted increasing attention due to its earth abundance, low cost, eco-friendliness, favorable electrocatalytic activity, and excellent adsorption capability [17,18]. Nanostructured α-MnO 2 has demonstrated a high affinity for adsorption of Cu 2+ , Pb 2+ , Zn 2+ , Cd 2+ , Hg 2+ , etc. [19,20]. Therefore, MnO 2 nanostructures have recently been used for the voltammetric detection of HMIs [21][22][23]. Owing to its nontoxicity, cost-effectiveness, relatively narrow band gap, high adsorption capacity, and admirable catalytic properties, nanoscale Bi 2 O 3 has also found growing interest in various fields such as photocatalysis [24], electroreduction [25], supercapacitors [26][27][28], and voltammetric sensors [29,30]. It has been reported that nanoscale Bi 2 O 3 displayed a high affinity to HMIs such as Cd 2+ [31,32]. The electrochemical reduction of Bi 2 O 3 can produce a porous Bi layer and further form a "fused alloy" with the heavy metal, which accumulates more HMIs on its surface and eventually enhances the sensitivity [33]. For these reasons, Bi 2 O 3 -based electrodes have emerged as promising alternatives to mercury-based electrodes for HMI determination.
In contrast to single transition metal oxides, binary transition metal oxide electrocatalysts generally show a higher electrocatalytic activity [34,35]. However, binary transition metal oxides have rarely been used to detect HMIs [36][37][38]. Fe 2 O 3 /NiO heterojunctions possess a lower diffusion energy barrier for lead atoms, thus significantly improving the anti-interference ability for detecting Pb 2+ [37]. Bi 2 O 3 /Fe 2 O 3 -decorated graphene oxide (GO) has demonstrated a remarkable electrocatalytic activity toward Cd 2+ determination, having a low LOD of 1.85 ng L −1 [38]. In our recent work, the synergistic interaction between β-Bi 2 O 3 microspheres and shuttle-like α-Fe 2 O 3 nanoparticles enabled the concurrent determination of Cd 2+ and Pb 2+ in environmental and food samples at the nanomolar levels [30].
Binary transition metal oxides such as Bi 2 O 3 /MnO 2 have been successfully used in supercapacitors [39,40] and the voltammetric detection of H 2 O 2 [41]. In addition, the individual Bi 2 O 3 or MnO 2 nanostructures have also been used to detect Pb 2+ . However, to the best of our knowledge, GO-decorated Bi 2 O 3 /MnO x composites have not yet been reported. Herein, we fabricated GO-decorated binary transition metal oxides of Bi 2 O 3 and MnO 2 nanocomposites (Bi 2 O 3 /MnO 2 /GO) and used them as a delicate electrocatalyst for Pb 2+ determination. GO nanoflakes are an electron-rich species that can reduce Pb 2+ into metallic Pb by applying a suitable potential. In addition, abundant oxygen-containing functional groups (OxFGs) such as carboxyl, hydroxy, carbonyl, and epoxide groups in the edge of GO flakes can firmly bind Pb 2+ onto their surface through electrostatic and coordination interactions, which facilitates the adsorption of Pb 2+ [42,43]. Generally, the sensing performance for HMIs mainly relies on the adsorption capacity and electrocatalytic activity of the sensing material that is decorated on the electrode [44], which can be readily tailored using morphology engineering [45,46]. In this regard, we synthesized dandelion-like α-MnO 2 and flower-like β-Bi 2 O 3 nanocomposites to enhance the Pb 2+ adsorption and electrocatalytic activity. With the synergistic interaction of both MnO 2  Dandelion-like α-MnO 2 microspheres were prepared via a facile hydrothermal treatment route [47]. Typically, 1.3522 g of MnSO 4 ·H 2 O, 2.1626 g of K 2 S 2 O 8 and 1.3941 g of K 2 SO 4 were sequentially added to 60 mL of 0.6 M H 2 SO 4 and magnetically stirred for 30 min to completely dissolve. Then, the mixture solution was decanted into a 100 mL Teflon-lined stainless steel container and heated at a temperature of 140 • C for 12 h. The resultant product was repeatedly rinsed with DI water and dried at 60 • C overnight for further use.

Synthesis of Flower-like β-Bi 2 O 3 Microspheres
Flower-like β-Bi 2 O 3 microspheres were synthesized by a simple hydrothermal treatment followed by a thermal decomposition at high temperature [30]. In brief, 0.03 mol of Bi(NO 3 ) 3 ·5H 2 O was dissolved into 11 mL of HAc, and 14 mL of anhydrous ethanol was then added to form a white suspension. The resulting suspension was ceaselessly stirred for 45 min, and 28 mL of DMF was then added to yield a clear solution. Subsequently, the mixture solution was poured into a 100 mL Teflon-lined stainless steel autoclave and reacted at a temperature of 120 • C for 40 min. The precursor was centrifuged at 10,000 rpm for 5 min, where it was alternately rinsed with anhydrous alcohol and DI water and allowed to dry at 80 • C overnight. Finally, the resultant β-Bi 2 O 3 precursor was further transferred to a porcelain boat and calcinated at 350 • C for 4 h at a heating rate of 2 • C min −1 in an air atmosphere to yield orange-yellow β-Bi 2 O 3 microspheres.

Preparation of Bi 2 O 3 /MnO 2 /GO Nanocomposites
At first, 5 mg of dandelion-like α-MnO 2 microspheres, flower-like β-Bi 2 O 3 microspheres, and GO nanoflakes were separately dispersed into 10 mL of DI water under ultrasonication to form their respective uniform dispersions at a concentration of 0.5 mg mL −1 . Then, 1 mL of the α-MnO 2 , β-Bi 2 O 3 , and GO dispersions were further mixed and subjected to a 30 min ultrasonication to obtain a uniform Bi 2 O 3 /MnO 2 /GO dispersion. The amount of the three materials were optimized during our preliminary experiments. Bi 2 O 3 /MnO 2 /GO composites containing 5 mg of each of the three materials showed the largest stripping Nanomaterials 2022, 12, 3317 4 of 15 peak current of Pb 2+ . Therefore, we selected the composite with this component content as the sensing material.

Characterizations of Sensing Materials
The microscopic morphologies of the GO nanoflakes, dandelion-like α-MnO 2 microspheres, flower-like β-Bi 2 O 3 microspheres, and Bi 2 O 3 /MnO 2 /GO nanocomposites were observed using field-emission SEM (Sigma HD, Zeiss, Oberkochen, Germany). Before taking the SEM measurements, a few thin layers of Au were coated onto the surface of the samples. The crystalline structures of these materials were studied using a powder XRD (Rigaku Ultima IV, Tokyo, Japan) with monochromatized Cu Kα radiation (λ = 1.54 A).

Fabrication of Modified Electrodes
Before electrode modification, the GCE was thoroughly polished to a shining mirrorlike surface with 0.05 µm of alumina slurry, and it was ultrasonically cleaned with anhydrous alcohol and DI water for three cycles to remove residual contaminants. Then, the polished GCE was exposed to infrared light to allow the material to adequately dry. The Bi 2 O 3 /MnO 2 /GO/GCE electrode was prepared by using a conventional drop-casting method. Specifically, 5 µL of the Bi 2 O 3 /MnO 2 /GO dispersion was cast on the surface of the freshly polished GCE and dried under the exposure of infrared light to form a firm sensing film. For comparison, the MnO 2 /GO/GCE, Bi 2 O 3 /GO/GCE, and GO/GCE were also fabricated using the same procedure, aside from the dispersion used.

Electrochemical Measurements
All electrochemical measurements were performed on a CHI 660E electrochemical workstation (Chenhua Inc., Shanghai, China) equipped with a classic three-electrode system, which consists of the Bi 2 O 3 /MnO 2 /GO/GCE electrode, a Pt wire, and a saturated calomel electrode (SCE) as the working, counter, and reference electrodes, respectively. A 10 mL electrochemical cell made of glass was used for the electrochemical measurements. Unless otherwise specified, the 0.1 M HAc-NaAc buffer (pH = 5.5) functioned as the supporting electrolyte. To assess the electrochemical performance, the CV curves and Nyquist plots of different modified electrodes were recorded in a solution of 2 mM [Fe(CN) 6 ] 3−/4− and 0.1 M KCl. To improve the stripping responses, a suitable deposition was employed in the Pb 2+ standard solutions. After 30 s of rest, the stripping peak currents of Pb 2+ were recorded between −1.0 V and −0.5 V using the SWASV technique. The frequency, step potential, and pulse amplitude of the SWASV were set at 15 Hz, 4 mV, and 25 mV, respectively. When not in use, the Bi 2 O 3 /MnO 2 /GO/GCE electrode was stored in the air. After each determination, the electrode surface was refreshed by immersing it into a blank solution and applying +0.3 V for 150 s to ensure the complete removal of the residual metals.

Physical Characterization
The microscopic morphologies of the GO nanoflakes, dandelion-like α-MnO 2 microspheres, flower-like β-Bi 2 O 3 microspheres, and Bi 2 O 3 /MnO 2 /GO nanocomposites were observed using the SEM technique, and their SEM images are shown in Figure 1. The GO nanosheets exhibited a typical lamellar structure with obvious wrinkles ( Figure 1A). Dandelion-like nanostructures are found in the image of the α-MnO 2 , consisting of many radially distributed nanorods ( Figure 1B). Typical flower-like β-Bi 2 O 3 microspheres are observed in Figure 1C, which consist of many interconnected thin nanosheets. The unique dandelion-like α-MnO 2 and flower-like Bi 2 O 3 structures enlarge the electroactive surface area, thereby improving the sensing performance. In addition, the interconnected porous microstructures are found in the dandelion-like α-MnO 2 and flower-like Bi 2 O 3 microspheres, which facilitate the electrolyte infiltration and adsorption of HMIs. As illustrated in Figure 1D, typical microspheres are observed in the SEM image of the Bi 2 O 3 /MnO 2 /GO nanocomposites. In addition, the microspheres are partially wrapped by GO nanosheets.
To conform the composition of the microspheres, the energy-dispersive X-ray spectroscopy (EDS) mappings of the Bi 2 O 3 /MnO 2 /GO nanocomposites were also measured ( Figure 1E). The uniformly dispersed C, O, Mn, and Bi distribution suggests the presence of C, O, Mn, and Bi elements. In addition, the distributions of Mn and Bi exhibit obvious microsphere structures, indicating that the microsphere consists of both Bi 2 O 3 and MnO 2 . All of these results indicate the successful synthesis of Bi 2 O 3 /MnO 2 /GO nanocomposites. microstructures are found in the dandelion-like α-MnO2 and flower-like Bi2O3 microspheres, which facilitate the electrolyte infiltration and adsorption of HMIs. As illustrated in Figure 1D, typical microspheres are observed in the SEM image of the Bi2O3/MnO2/GO nanocomposites. In addition, the microspheres are partially wrapped by GO nanosheets. To conform the composition of the microspheres, the energy-dispersive X-ray spectroscopy (EDS) mappings of the Bi2O3/MnO2/GO nanocomposites were also measured ( Figure  1E). The uniformly dispersed C, O, Mn, and Bi distribution suggests the presence of C, O, Mn, and Bi elements. In addition, the distributions of Mn and Bi exhibit obvious microsphere structures, indicating that the microsphere consists of both Bi2O3 and MnO2. All of these results indicate the successful synthesis of Bi2O3/MnO2/GO nanocomposites.     [49]. Additionally, we detected sharp diffraction peaks without any apparent impurity peaks, demonstrating that the as-prepared α-MnO 2 microspheres were of high purity. Broad diffraction peaks were observed for the flower-like β-Bi 2 O 3 at 28.00 • , 32.52 • , 46.32 • , and 55.64 • , which can be indexed to the (201), (220), (222), and (213) crystal facets (JCPDS 651209) [24]. The characteristic diffraction peaks of both the GO α-MnO 2 and β-Bi 2 O 3 can be observed in the XRD pattern of the Bi 2 O 3 /MnO 2 /GO nanocomposites. However, the intensity of the diffraction peaks of the α-MnO 2 and β-Bi 2 O 3 microspheres were significantly weakened, which was mainly due to the presence of a large amount of GO partially masking the diffraction peaks of the α-MnO 2 and β-Bi 2 O 3 . This further confirmed that the Bi 2 O 3 /MnO 2 /GO nanocomposites were successfully synthesized.  [49]. Additionally, we detected sharp diffraction peaks without any apparent impurity peaks, demonstrating that the as-prepared α-MnO2 microspheres were of high purity. Broad diffraction peaks were observed for the flowerlike β-Bi2O3 at 28.00°, 32.52°, 46.32°, and 55.64°, which can be indexed to the (201), (220), (222), and (213) crystal facets (JCPDS 651209) [24]. The characteristic diffraction peaks of both the GO α-MnO2 and β-Bi2O3 can be observed in the XRD pattern of the Bi2O3/MnO2/GO nanocomposites. However, the intensity of the diffraction peaks of the α-MnO2 and β-Bi2O3 microspheres were significantly weakened, which was mainly due to the presence of a large amount of GO partially masking the diffraction peaks of the α-MnO2 and β-Bi2O3. This further confirmed that the Bi2O3/MnO2/GO nanocomposites were successfully synthesized.

Electrochemical Properties of Different Electrodes
The CV curves for the different electrodes were scanned in a solution of 2.0 mM [Fe(CN)6] 3−/4− and 0.1 M KCl to assess their electrochemical properties. As shown in Figure  3A, a pair of sharp and symmetric redox peaks occur at all electrodes, with an almost identical anodic and cathodic peak current (Ipa and Ipc), indicating that the redox of Fe(III)/Fe(II) is a quasi-reversible process. After the modification of the GO, Bi2O3/GO, MnO2/GO and MnO2/Bi2O3/GO, the redox peak currents were sequentially enhanced. The corresponding effective electroactive areas were also estimated based on the Randles-Sevcik equation: where A is the effective electroactive area, Ageom is the geometric surface area (diameter of 3.0 mm, 7.07 mm 2 ), and the other symbols retain their usual meanings. The effective electroactive area and roughness factor of these electrodes were estimated according to

Electrochemical Properties of Different Electrodes
The CV curves for the different electrodes were scanned in a solution of 2.0 mM [Fe(CN) 6 ] 3−/4− and 0.1 M KCl to assess their electrochemical properties. As shown in Figure 3A, a pair of sharp and symmetric redox peaks occur at all electrodes, with an almost identical anodic and cathodic peak current (I pa and I pc ), indicating that the redox of Fe(III)/Fe(II) is a quasi-reversible process. After the modification of the GO, Bi 2 O 3 /GO, MnO 2 /GO and MnO 2 /Bi 2 O 3 /GO, the redox peak currents were sequentially enhanced. The corresponding effective electroactive areas were also estimated based on the Randles-Sevcik equation: where A is the effective electroactive area, A geom is the geometric surface area (diameter of 3.0 mm, 7.07 mm 2 ), and the other symbols retain their usual meanings. The effective electroactive area and roughness factor of these electrodes were estimated according to Equations (1) and (2)  serving the change in a semicircle diameter [50][51][52]. Figure 3B displays the Nyquist plots of the different electrodes. Typically, a Nyquist diagram includes a semicircle at the higher frequency domain and a straight line at the lower frequency region, which is closely related to the electron-transfer-limited and diffusion-controlled processes, respectively [53,54]. Clearly, the bare GCE showed the largest semicircle (Rct = 4126 Ω), suggesting that the electron transfer was severely retarded in the unmodified bare. When GO was decorated on the GCE, the Rct reduced to 2815 Ω due to the good electrical conductivity of GO. When the β-Bi2O3 and α-MnO2 microspheres were further introduced into the GO/GCE, the respective Rct reduced to 2632 Ω and 1898 Ω, respectively. As anticipated, the smallest semicircle diameter was achieved in the Bi2O3/MnO2/GO/GCE (Rct = 1761 Ω). This indicates that the Bi2O3/MnO2/GO effectively promotes the electron transfer, which ultimately improves the electrochemical sensing performance.   EIS is a useful technique to assess the interfacial properties, mass-transport, and kinetic parameters, in addition to the charge transferred resistance (R ct ) of electrodes by observing the change in a semicircle diameter [50][51][52]. Figure 3B displays the Nyquist plots of the different electrodes. Typically, a Nyquist diagram includes a semicircle at the higher frequency domain and a straight line at the lower frequency region, which is closely related to the electron-transfer-limited and diffusion-controlled processes, respectively [53,54]. Clearly, the bare GCE showed the largest semicircle (R ct = 4126 Ω), suggesting that the electron transfer was severely retarded in the unmodified bare. When GO was decorated on the GCE, the R ct reduced to 2815 Ω due to the good electrical conductivity of GO. When the β-Bi 2 O 3 and α-MnO 2 microspheres were further introduced into the GO/GCE, the respective R ct reduced to 2632 Ω and 1898 Ω, respectively. As anticipated, the smallest semicircle diameter was achieved in the Bi 2 O 3 /MnO 2 /GO/GCE (R ct = 1761 Ω). This indicates that the Bi 2 O 3 /MnO 2 /GO effectively promotes the electron transfer, which ultimately improves the electrochemical sensing performance.

Stripping Voltammetric Responses of Pb 2+ on Different Electrodes
The voltammetric behavior of 1.0 µM of Pb 2+ on the different modified electrodes were studied using the SWASV technique ( Figure 4). As a control, we also recorded the SWASV curves of the different electrodes in the absence of Pb 2+ . In the absence of Pb 2+ , no noticeable response peaks were found in any of the electrodes ( Figure S1). In unmodified GCE, a weak stripping peak was observed at −0.656 V with an anodic stripping peak current (I pa ) of 4.118 µA, indicating that a sluggish oxidation process occurred in unmodified GCE. When GO nanoflakes were decorated on the GCE surface, the I pa (Pb 2+ ) increased to 6.277 µA because GO, with its large surface area and abundant OXFGs, facilitates the adsorption of Pb 2+ . When the flower-like β-Bi 2 O 3 and dandelion-like α-MnO 2 microspheres were introduced into the GO/GCE, their I pa (Pb 2+ ) significantly increased to 9.541 µA and 10.95 µA, respectively, while their respective anodic stripping peak potentials (E pa ) also decreased. This suggests that the decoration of flower-like β-Bi 2 O 3 and dandelion-like α-MnO 2 microspheres promotes an efficient electron transfer, which is closely related to high affinity capacity and extraordinary electrocatalytic activity toward Pb 2+ . As expected, the GO-coated binary transition metal oxides of Bi 2 O 3 /MnO 2 remarkably improved the stripping voltammetric response of Pb 2+ , with the highest I pa of 58.07 µA and the lowest E pa (−0.667 V). Notably, the stripping peak current for the GO-coated binary transition metal oxide was about five times higher than that of the GO-coated single metal oxides, suggesting that the synergistic effect between the flower-like β-Bi 2 O 3 and dandelion-like α-MnO 2 microspheres is attributed to the enhanced I pa and reduction in overpotential.
The voltammetric behavior of 1.0 μM of Pb 2+ on the different modified electrod were studied using the SWASV technique ( Figure 4). As a control, we also recorded t SWASV curves of the different electrodes in the absence of Pb 2+ . In the absence of Pb 2+ , noticeable response peaks were found in any of the electrodes ( Figure S1). In unmodifi GCE, a weak stripping peak was observed at −0.656 V with an anodic stripping peak cu rent (Ipa) of 4.118 μA, indicating that a sluggish oxidation process occurred in unmodifi GCE. When GO nanoflakes were decorated on the GCE surface, the Ipa (Pb 2+ ) increased 6.277 μA because GO, with its large surface area and abundant OXFGs, facilitates the a sorption of Pb 2+ . When the flower-like β-Bi2O3 and dandelion-like α-MnO2 microsphe were introduced into the GO/GCE, their Ipa (Pb 2+ ) significantly increased to 9.541 μA a 10.95 μA, respectively, while their respective anodic stripping peak potentials (Epa) a decreased. This suggests that the decoration of flower-like β-Bi2O3 and dandelion-like MnO2 microspheres promotes an efficient electron transfer, which is closely related high affinity capacity and extraordinary electrocatalytic activity toward Pb 2+ . As expecte the GO-coated binary transition metal oxides of Bi2O3/MnO2 remarkably improved t stripping voltammetric response of Pb 2+ , with the highest Ipa of 58.07 μA and the low Epa (−0.667 V). Notably, the stripping peak current for the GO-coated binary transiti metal oxide was about five times higher than that of the GO-coated single metal oxid suggesting that the synergistic effect between the flower-like β-Bi2O3 and dandelion-l α-MnO2 microspheres is attributed to the enhanced Ipa and reduction in overpotential.

Effect of Deposition Parameters
Deposition parameters have a prominent effect on the voltammetric behavior of Pb As illustrated in Figure 5A, the Ipa (Pb 2+ ) gradually increased when the deposition pote tial shifted from −1.3 V to −1.0 V, then sharply declined as the deposition potentials shift further. At an excessively negative deposition potential, hydrogen bubbles would be ge erated on the surface of the Bi2O3/MnO2/GO/GCE, resulting in the exfoliation of the d posited Pb 2+ . When the deposition potential was higher than -1.0 V, the electrochemi energy was not sufficient to reduce the deposited Pb 2+ . Therefore, the optimal depositi potential was set at −1.0 V. Generally, prolonging the deposition time can enhance t adsorption amount of HMIs on the electrode surface, thereby increasing the strippi peak current. As presented in Figure 5B, the Ipa (Pb 2+ ) steadily increased with depositi time until reaching a plateau at 300 s. This was mainly because the surface adsorption si of the Bi2O3/MnO2/GO/GCE were saturated at 300 s. Thus, the optimum deposition tim was set at 300 s.

Effect of Deposition Parameters
Deposition parameters have a prominent effect on the voltammetric behavior of Pb 2+ . As illustrated in Figure 5A, the I pa (Pb 2+ ) gradually increased when the deposition potential shifted from −1.3 V to −1.0 V, then sharply declined as the deposition potentials shifted further. At an excessively negative deposition potential, hydrogen bubbles would be generated on the surface of the Bi 2 O 3 /MnO 2 /GO/GCE, resulting in the exfoliation of the deposited Pb 2+ . When the deposition potential was higher than -1.0 V, the electrochemical energy was not sufficient to reduce the deposited Pb 2+ . Therefore, the optimal deposition potential was set at −1.0 V. Generally, prolonging the deposition time can enhance the adsorption amount of HMIs on the electrode surface, thereby increasing the stripping peak current. As presented in Figure 5B, the I pa (Pb 2+ ) steadily increased with deposition time until reaching a plateau at 300 s. This was mainly because the surface adsorption sites of the Bi 2 O 3 /MnO 2 /GO/GCE were saturated at 300 s. Thus, the optimum deposition time was set at 300 s.

Effect of Solution pH
It is well-known that a solution's pH has a significant impact on the Ipa (Pb 2+ ). Therefore, the influence of the solution's pH was also explored. As illustrated in Figure 6, the Ipa (Pb 2+ ) slowly increased as the pH increased from 3.0 to 4.5, and then sharply increased until the pH of 5.5. Afterwards, the Ipa (Pb 2+ ) dramatically decreased when the pH exceeded 5.5. Therefore, pH = 5.5 was chosen as the optimal solution pH. This phenomenon can be interpreted as follows. At lower pH values, the H + adsorption on the electrode surface neutralizes the negative charge on the electrode surface, which reduces the adsorption of Pb 2+ , resulting in a decrease in the Ipa (Pb 2+ ). Pb 2+ tends to be hydrolyzed in a solution with a higher pH so that the concentration of free Pb 2+ in the solution decreases and the Ipa (Pb 2+ ) decreases.

Effect of Solution pH
It is well-known that a solution's pH has a significant impact on the I pa (Pb 2+ ). Therefore, the influence of the solution's pH was also explored. As illustrated in Figure 6, the I pa (Pb 2+ ) slowly increased as the pH increased from 3.0 to 4.5, and then sharply increased until the pH of 5.5. Afterwards, the I pa (Pb 2+ ) dramatically decreased when the pH exceeded 5.5. Therefore, pH = 5.5 was chosen as the optimal solution pH. This phenomenon can be interpreted as follows. At lower pH values, the H + adsorption on the electrode surface neutralizes the negative charge on the electrode surface, which reduces the adsorption of Pb 2+ , resulting in a decrease in the I pa (Pb 2+ ). Pb 2+ tends to be hydrolyzed in a solution with a higher pH so that the concentration of free Pb 2+ in the solution decreases and the I pa (Pb 2+ ) decreases.

Effect of Solution pH
It is well-known that a solution's pH has a significant impact on the Ipa (Pb 2+ ). Therefore, the influence of the solution's pH was also explored. As illustrated in Figure 6, the Ipa (Pb 2+ ) slowly increased as the pH increased from 3.0 to 4.5, and then sharply increased until the pH of 5.5. Afterwards, the Ipa (Pb 2+ ) dramatically decreased when the pH exceeded 5.5. Therefore, pH = 5.5 was chosen as the optimal solution pH. This phenomenon can be interpreted as follows. At lower pH values, the H + adsorption on the electrode surface neutralizes the negative charge on the electrode surface, which reduces the adsorption of Pb 2+ , resulting in a decrease in the Ipa (Pb 2+ ). Pb 2+ tends to be hydrolyzed in a solution with a higher pH so that the concentration of free Pb 2+ in the solution decreases and the Ipa (Pb 2+ ) decreases.

Stripping Kinetics of Pb 2+ on the Bi 2 O 3 /MnO 2 /GO/GCE
In order to study the stripping kinetics of Pb 2+ , the cyclic voltammograms of 1.0 µM of Pb 2+ were measured by the Bi 2 O 3 /MnO 2 /GO/GCE at different scanning rates (0.05-0.40 V s −1 ). Figure 7A shows the CV curves of 1.0 µM of Pb 2+ at various scanning rates. A pair of well-shaped redox peaks occurred at all scanning rates with almost identical I pa and I pc (I pa /I pc ≈ 1), suggesting that Pb 2+ stripping is a quasi-reversible process. As the scanning rate increased, the I pa and I pc gradually increased. In addition, the anodic peaks shift to more positive potential while the cathodic peaks shift to more negative potential. As illustrated in Figure 7B, both the I pa and I pc are linearly correlated to the square root of scanning rates (v 1/2 ), demonstrating that Pb 2+ stripping was primarily controlled by the diffusion.

Stripping Kinetics of Pb 2+ on the Bi2O3/MnO2/GO/GCE
In order to study the stripping kinetics of Pb 2+ , the cyclic voltammograms of 1.0 μM of Pb 2+ were measured by the Bi2O3/MnO2/GO/GCE at different scanning rates (0.05-0.40 V s − 1 ). Figure 7A shows the CV curves of 1.0 μM of Pb 2+ at various scanning rates. A pair of well-shaped redox peaks occurred at all scanning rates with almost identical Ipa and Ipc (Ipa/Ipc ≈ 1), suggesting that Pb 2+ stripping is a quasi-reversible process. As the scanning rate increased, the Ipa and Ipc gradually increased. In addition, the anodic peaks shift to more positive potential while the cathodic peaks shift to more negative potential. As illustrated in Figure 7B, both the Ipa and Ipc are linearly correlated to the square root of scanning rates (v 1/2 ), demonstrating that Pb 2+ stripping was primarily controlled by the diffusion.

Calibration Plot, LDR, and LOD
Under optimal determination conditions, the Ipa (Pb 2+ ) at various concentrations were measured on the Bi2O3/MnO2/GO/GCE via the SWASV technique. As illustrated in Figure  8A, well-defined stripping peaks of Pb 2+ occurred at about −0.65 V with a slight positive shift at higher concentrations. As shown in the inset of Figure 8A, the stripping peaks of low concentrations of Pb 2+ slightly shifted to more negative biases, probably due to the electrode surfaces not being exactly the same. However, the obvious positive shift in the peak potential at higher concentrations is probably due to the occurrence of concentration polarization. Moreover, the Ipa (Pb 2+ ) gradually increased with Pb 2+ concentration. The Ipa (Pb 2+ ) are in good proportion to Pb 2+ concentration from 0.01 to 10 μM ( Figure 8B). The corresponding linear regression equation was expressed as Ipa(μA) = 53.45C (μM) + 0.578, with a good correlation coefficient (R 2 ) of 0.998. The LOD was calculated as 2.0 nM (0.41 μg L − 1 ) based on 3σ/s (where σ is the standard deviation in blank solution and s is the slope of the calibration plot). A comparison of the analytical properties for Pb 2+ was also made between the Bi2O3/MnO2/GO/GCE composite and previously reported ones. As shown in Table 2, the analytical properties of the Bi2O3/MnO2/GO/GCE composite, including the LDR, LOD, and sensitivity, well matches or even exceeds the previously reported electrodes.

Calibration Plot, LDR, and LOD
Under optimal determination conditions, the I pa (Pb 2+ ) at various concentrations were measured on the Bi 2 O 3 /MnO 2 /GO/GCE via the SWASV technique. As illustrated in Figure 8A, well-defined stripping peaks of Pb 2+ occurred at about −0.65 V with a slight positive shift at higher concentrations. As shown in the inset of Figure 8A, the stripping peaks of low concentrations of Pb 2+ slightly shifted to more negative biases, probably due to the electrode surfaces not being exactly the same. However, the obvious positive shift in the peak potential at higher concentrations is probably due to the occurrence of concentration polarization. Moreover, the I pa (Pb 2+ ) gradually increased with Pb 2+ concentration. The I pa (Pb 2+ ) are in good proportion to Pb 2+ concentration from 0.01 to 10 µM ( Figure 8B). The corresponding linear regression equation was expressed as I pa (µA) = 53.45C (µM) + 0.578, with a good correlation coefficient (R 2 ) of 0.998. The LOD was calculated as 2.0 nM (0.41 µg L −1 ) based on 3σ/s (where σ is the standard deviation in blank solution and s is the slope of the calibration plot). A comparison of the analytical properties for Pb 2+ was also made between the Bi 2 O 3 /MnO 2 /GO/GCE composite and previously reported ones. As shown in Table 2, the analytical properties of the Bi 2 O 3 /MnO 2 /GO/GCE composite, including the LDR, LOD, and sensitivity, well matches or even exceeds the previously reported electrodes. Refs.

Anti-Interference Ability
Excellent selectivity is essential for trace determination of HMIs in complex sample matrix. Therefore, the anti-interfering ability of the Bi2O3/MnO2/GO/GCE was also studied. To explore the anti-interfering ability, the SWASV responses of 1.0 μM Pb 2+ were recorded on the Bi2O3/MnO2/GO/GCE in presence of 100-fold concentration interfering species, such as common cations (i.e., Na + , K + , Ca 2+ , Mg 2+ , Zn 2+ , Fe 2+ , Co 2+ , Cu 2+ , Cd 2+ , Al 3+ ) and anions (i.e., Cl − , NO3 − , SO4 2− , PO4 3− ). The relative errors are less than 5% in presence of these potential interfering species (Table S1), indicating that the Bi2O3/MnO2/GO/GCE possesses excellent selectivity. The extraordinary selectivity of the Bi2O3/MnO2/GO/GCE may due to the higher affinity of Bi2O3/MnO2/GO for Pb 2+ .Interestingly, a sharp stripping peak of Cd 2+ was also observed at −0.865 V on the Bi2O3/MnO2/GO/GCE. In addition, the stripping peaks of Pb 2+ and Cd 2+ did not overlap each other with a broad potential separation of

Anti-Interference Ability
Excellent selectivity is essential for trace determination of HMIs in complex sample matrix. Therefore, the anti-interfering ability of the Bi 2 O 3 /MnO 2 /GO/GCE was also studied. To explore the anti-interfering ability, the SWASV responses of 1.0 µM Pb 2+ were recorded on the Bi 2 O 3 /MnO 2 /GO/GCE in presence of 100-fold concentration interfering species, such as common cations (i.e., Na + , K + , Ca 2+ , Mg 2+ , Zn 2+ , Fe 2+ , Co 2+ , Cu 2+ , Cd 2+ , Al 3+ ) and anions (i.e., Cl − , NO 3 − , SO 4 2− , PO 4 3− ). The relative errors are less than 5% in presence of these potential interfering species (Table S1), indicating that the Bi 2 O 3 /MnO 2 /GO/GCE possesses excellent selectivity. The extraordinary selectivity of the Bi 2 O 3 /MnO 2 /GO/GCE may due to the higher affinity of Bi 2 O 3 /MnO 2 /GO for Pb 2+ . Interestingly, a sharp stripping peak of Cd 2+ was also observed at −0.865 V on the Bi 2 O 3 /MnO 2 /GO/GCE. In addition, the stripping peaks of Pb 2+ and Cd 2+ did not overlap each other with a broad potential separation of 0.215 V, suggesting the feasibility of simultaneous detection of Pb 2+ and Cd 2+ .

Reproducibility, Repeatability and Stability
To assess the practicability of the Bi 2 O 3 /MnO 2 /GO/GCE composite, we also studied their reproducibility, repeatability, and stability. The relative standard deviation (RSD) for parallel detections of 10 µM of Pb 2+ using five independent Bi 2 O 3 /MnO 2 /GO/GCEs was 4.59% ( Figure S2), indicating satisfactory reproducibility. The RSD for five consecutive detections of 10 µM of Pb 2+ was 5.38% ( Figure S3), suggesting admirable repeatability. Moreover, the I pa of 10 µM of Pb 2+ retained 92.05% of its initial values after one week ( Figure S4), indicating excellent storage stability.

Determination of Trace Pb 2+ in Water Samples
Under the optimal determination conditions, the Pb 2+ concentrations in the water samples were quantitatively determined by the SWASV technique using the Bi 2 O 3 /MnO 2 /GO/ GCE composite. As summarized in Table 3, the Pb 2+ concentration from a local lake was determined to be 0.121 µM, while no Pb 2+ was determined in the tap water. To further confirm the accuracy and precision, a series of Pb 2+ standard solutions of known concentrations were separately spiked into the water samples, and recovery assays were then performed. The Bi 2 O 3 /MnO 2 /GO/GCE exhibited acceptable RSD values (3.83-5.89%) and satisfactory recoveries (95.5-105%). The Bi 2 O 3 /MnO 2 /GO/GCE has demonstrated tremendous prospects in the sensitive determination of Pb 2+ from complex matrixes.

Conclusions
In this study, GO-coated binary transition metal oxides of Bi 2 O 3 /MnO 2 nanocomposites were used to fabricate a sensitive voltammetric sensor for the trace detection of lead ions in water samples. The Bi 2 O 3 /MnO 2 /GO nanocomposites boosted the electroactive surface area and significantly reduced the charge transferred resistance. Moreover, the synergistic enhancement effect from the GO nanoflakes, dandelion-like α-MnO 2 microspheres, and flower-like β-Bi 2 O 3 microspheres endowed Bi 2 O 3 /MnO 2 /GO/GCE with extraordinary electrocatalytic activity toward the stripping voltammetric behavior of Pb 2+ . Under optimal detection conditions, the Bi 2 O 3 /MnO 2 /GO/GCE exhibited a relatively wide LDR (0.01-10 µM), low LOD (2.0 nM) and high sensitivity (53.43 µA µM −1 ). Moreover, the Bi 2 O 3 /MnO 2 /GO/GCE exhibited an anti-interference ability even in presence of a 100-fold concentration of common cations and anions, as well as outstanding reproducibility, repeatability, and stability. The MnO 2 /Bi 2 O 3 /GO/GCE composite realized the sensitive detection of trace Pb 2+ in water samples with satisfactory recovery. Together with portable and smart electrochemical devices, the proposed Bi 2 O 3 /MnO 2 /GO nanocomposites demonstrate promising prospects in the in situ detection of HMIs.
Supplementary Materials: The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/nano12193317/s1: Figure S1: The SWASV curves for the Bi 2 O 3 / MnO 2 /GO/GCE recorded in the absence of Pb 2+ (0.1 M HAc-NaAc buffer, pH = 5.5); Figure S2: The stripping peak current for parallel detections of 10 µM of Pb 2+ using five independent Bi 2 O 3 /MnO 2 / GO/GCEs; Figure S3: The stripping peak current for five consecutive detections of 10 µM of Pb 2+ using the same Bi 2 O 3 /MnO 2 /GO/GCE; Figure S4: The change in the stripping peak current of 10 µM of Pb 2+ within a week; and Table S1 Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data are available upon reasonable request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.