MnO2@Reduced Graphene Oxide Nanocomposite-Based Electrochemical Sensor for the Simultaneous Determination of Trace Cd(II), Zn(II) and Cu(II) in Water Samples

The rapid detection of trace metals is one of the most important aspect in achieving environmental monitoring and protection. Electrochemical sensors remain a key solution for rapid detection of heavy metals in environmental water matrices. This paper reports the fabrication of an electrochemical sensor obtained by the simultaneous electrodeposition of MnO2 nanoparticles and RGO nanosheets on the surface of a glassy carbon electrode. The successful electrodeposition was confirmed by the enhanced current response on the cyclic voltammograms. The XRD, HR-SEM/EDX, TEM, FTIR, and BET characterization confirmed the successful synthesis of MnO2 nanoparticles, RGO nanosheets, and MnO2@RGO nanocomposite. The electrochemical studies results revealed that MnO2@RGO@GCE nanocomposite considerably improved the current response on the detection of Zn(II), Cd(II) and Cu(II) ions in surface water. These remarkable improvements were due to the interaction between MnO2 nanomaterials and RGO nanosheets. Moreover, the modified sensor electrode portrayed high sensitivity, reproducibility, and stability on the simultaneous determination of Zn(II), Cd(II), and Cu(II) ions. The detection limits of (S/N = 3) ranged from 0.002–0.015 μg L−1 for the simultaneous detection of Zn(II), Cd(II), and Cu(II) ions. The results show that MnO2@RGO nanocomposite can be successfully used for the early detection of heavy metals with higher sensitivity in water sample analysis.


Introduction
The key challenge in environmental pollution remediation is the early detection of the contaminants. Superior remediation outcomes are directly acclimated to the early detection of contaminants. Thus, enabling the application of effective treatment processes that reduces cost amongst other resources [1,2]. Consequently, electrochemical sensors can detect various electroactive compounds by providing specific responses to various analytes by generating Faraday currents at different potentials [3]. Heavy metals such as copper (Cu), cadmium (Cd), and zinc (Zn), amongst others, are electroactive materials that are suitable for electrochemical detection. Evidently, trace amounts below the maximum recommended levels by the world Health Organization (WHO), copper and zinc, amongst others, play an important role in living organisms as they are essential for the metabolism related maintenance [1,4]. However, the excessive levels of these heavy metals beyond the WHO permitted levels can cause serious environmental and human health effects [5]. For instance, water containing Cu beyond the prescribed limit of 1.3 mg L −1 may cause lung

Reagents and Materials
All reagents and chemicals were of analytical reagent grade and doubly distilled water was used in all the experiments. Graphite flakes, potassium permanganate (KMnO 4 ), sulphuric acid (H 2 SO 4 ), phosphoric acid (H 3 PO 4 ), hydrochloric acid (HCl), hydrogen peroxide (H 2 O 2 ), polyvinyl alcohol (PVA), potassium chloride (KCl), hexanol, and absolute ethanol were purchased from Sigma-Aldrich (St. Louis, MO, USA). Spectrascan standard stock solutions of Cd, Cu, and Zn (1000 mg L −1 , Teknolab, Norway) were used to prepare the working standard solutions for the quantification of analyte concentrations in model solutions. Acetate buffer (1.0 mol L −1 , pH 5) was used as a supporting electrolyte.

Instrumentation
Electrochemical measurement such as cyclic voltammetry (CV), differential pulse anodic stripping voltammetry (DPASV), and electrochemical impedance spectroscopy (EIS) were performed using a µ-Autolab TYPE III conventional three-electrode system driven by the NOVA software (Version 1.8). A modified glassy carbon electrode was employed as a working electrode and a platinum wire as the counter electrode and Ag/AgCl/KCl (3 mol L −1 ) reference electrode. Electrochemical measurements were performed at room temperature. All pH measurements were done using H1 9811-5, (HANNA Instruments, Smithfield, RI, USA) pH meter supplied with a combined electrode. The morphology, nano structure and particle size of the material were examined using the transmission electron microscopy (TEM, JEM-2100, JEOL, Tokyo, Japan) and scanning electron microscopy (SEM, TESCAN VEGA 3 XMU, LMH instrument (Czech Republic) coupled with energy dispersive X-ray spectroscopy (EDS) for elemental composition analysis at an accelerating voltage of 20 kV. Infrared spectra were recorded (4000-400 cm −1 ), using a Perkin Elmer spectrum 100 Fourier transform infrared spectrometer (Waltham, MA, USA). The XRD measurements were done on a Rigaku Ultima IV X-ray diffractometer using a Cu-Kα radiation (λ = 0.15406 nm). BET measurements were performed on a Tristar II (Micromeritics, Norcross, GA, USA) after the samples were degassed at 77 • C for 16.5 h. All the BET surface area, pore size, and volume were measured using nitrogen at 77 K.

Preparation of Reduced Graphene Oxide (RGO)
Graphene oxide (GO) was initially prepared using the modified Hummers' method [34,35]. Briefly, H 2 SO 4 and H 3 PO 4 were mixed in a 9:1 volume ratio (360:40 mL) and stirred in an ice bath for 30 min. Graphite flakes (3.06 g) and KMnO 4 (18.21 g) were slowly added to the mixing solution under stirring. The mixture was transferred to an oil bath and heated to 85 • C for 30 min. Then, 50 mL of deionized water was added to the mixture, which was maintained under stirring at the same conditions for 60 min. The mixture was placed in an ice bath, and 150 mL of deionized water and 20 mL of H 2 O 2 were slowly added to terminate the reaction. An exothermic reaction occurred, and the solution was then allowed to cool down. The resulting solution was diluted with 1 mol L −1 HCl solution and centrifuged at 6000 rpm for 20 min. The supernatant was removed, and the residuals were washed several times with deionized water until a neutral pH was obtained (pH 7.05). Finally, the prepared samples were dried in an oven at 80 • C for 12 h. The obtained GO was thermally exfoliated by annealing in a muffle furnace at 500 • C at a heating rate of 2 • C min −1 for 2 h and the furnace was allowed to cool to 35 • C. Thus, GO was successfully converted to RGO [35].

Preparation of Manganese Oxide Nanoparticles (MnO 2 )
The synthesis of MnO 2 nanoparticles (NP) followed a previously modified method [36]. Typically, 6 mL of 0.5 mol L −1 KMnO 4 solution was added into 100 mL of the initially prepared ethanol: deionized water (50 mL:50 mL; 1:1) mixture solvent at room temperature. The mixture was stirred for 30 min before a 10% H 2 O 2 (10 mL) and polyvinyl alcohol (PVA) (10 mL) were added, forming a brown precipitate. The supernatant was discarded, and the Membranes 2021, 11, 517 4 of 17 collected brown precipitate was sonicated with a mixture of ethanol (20 mL) and hexanol (20 mL) to remove the PVA. The precipitate was then washed several times with deionized water and a brown powder obtained was dried at 80 • C in an oven.

Preparation of the MnO 2 @RGO/GCE Modified Electrode
The MnO 2 @RGO/GCE modified electrode was fabricated using an electrodeposition method. Initially, the bare GCE was polished successively to a mirror surface using 1.0, 0.3, and 0.05 µm alumina slurries, and then ultrasonically cleaned in deionized water, anhydrous ethanol, and deionized water for 5 min each to remove any organic and inorganic remnants on the electrode surface, prior being air dried at room temperature (24 • C). The electrochemical deposition was achieved by firstly, dispersing 150 mg RGO and 150 mg MnO 2 in 2 mL 0.1 mol L −1 KCl using ultrasonic bath. Then, the pre-cleaned GCE working electrode, Ag/AgCl reference electrode, and a Pt wire counter electrode were immersed in the mixture, and magnetically stirred at a speed of 100 rpm. The MnO 2 @RGO nanocomposite was then electrodeposited onto the GCE surface using cyclic voltammetry (CV) by scanning in the potential range of −0.2 V to +1.2 V with 6 cycles at a scan rate of 25 mV s −1 . The obtained MnO 2 @RGO/GCE electrode was washed with deionized water before being air dried at room temperature.

X-ray Diffraction (XRD)
XRD patterns demonstrating the crystal structure of RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite are displayed in Figure 1. The XRD pattern of MnO 2 nanoparticles demonstrated the existence of MnO 2 as a combination of two phases (α and β phases) at 2-theta angles 18.  [37][38][39]. Evidently, the RGO pattern shows an intense sharp peak corresponding to the (002) plane of carbon at 2-theta angle of 26.3 • [35]. Expectedly, the RGO nanosheets and MnO 2 nanoparticles characteristic peaks were observed on the XRD pattern of the MnO 2 @RGO nanocomposite. Interestingly, the MnO 2 @RGO nanocomposite showed major shifts in lower angles of all the peak positions as shown in Figure 1. This suggests the increase in the lattice parameters confirming the change in the crystal structure due to the formation of a nanocomposite [40,41]. (PVA) (10 mL) were added, forming a brown precipitate. The supernatant was discarded, and the collected brown precipitate was sonicated with a mixture of ethanol (20 mL) and hexanol (20 mL) to remove the PVA. The precipitate was then washed several times with deionized water and a brown powder obtained was dried at 80 °C in an oven.

Preparation of the MnO2@RGO/GCE Modified Electrode
The MnO2@RGO/GCE modified electrode was fabricated using an electrodeposition method. Initially, the bare GCE was polished successively to a mirror surface using 1.0, 0.3, and 0.05 μm alumina slurries, and then ultrasonically cleaned in deionized water, anhydrous ethanol, and deionized water for 5 min each to remove any organic and inorganic remnants on the electrode surface, prior being air dried at room temperature (24 °C). The electrochemical deposition was achieved by firstly, dispersing 150 mg RGO and 150 mg MnO2 in 2 mL 0.1 mol L −1 KCl using ultrasonic bath. Then, the pre-cleaned GCE working electrode, Ag/AgCl reference electrode, and a Pt wire counter electrode were immersed in the mixture, and magnetically stirred at a speed of 100 rpm. The MnO2@RGO nanocomposite was then electrodeposited onto the GCE surface using cyclic voltammetry (CV) by scanning in the potential range of −0.2 V to +1.2 V with 6 cycles at a scan rate of 25 mV s −1 . The obtained MnO2@RGO/GCE electrode was washed with deionized water before being air dried at room temperature.

High Resolution-Scanning Electron Microscopy (HR-SEM)
The HR-SEM images in Figure 2 illustrates the change in morphology during the fabrication of the MnO 2 @RGO sensing electrode. The RGO surface (Figure 2a) shows the thermally exfoliated nanolayers with a thickness ranging between 15 and 25 nm and a lateral size ranging between 55 and 450 nm. The spherical MnO 2 nanoparticles (Figure 2b) appear to be slightly agglomerated, but with a uniform shape. The simultaneous deposition of the MnO 2 and RGO on the surface of the GCE yielded to the MnO 2 @RGO nanocomposite displayed in Figure 2c. Compared to the RGO, the MnO 2 @RGO nanocomposite shows nanolayers with some nanoparticles on their surface, signifying that the RGO nanosheets act as a matrix for connecting the MnO 2 nanoparticles. Furthermore, the nanosheets in the MnO 2 @RGO nanocomposite appear to be thicker (21-36 nm) and larger in size (62-480 nm), due to the MnO 2 nanoparticles amalgamated on the surface of the RGO.
Membranes 2021, 11, x FOR PEER REVIEW 5 The HR-SEM images in Figure 2 illustrates the change in morphology during the rication of the MnO2@RGO sensing electrode. The RGO surface (Figure 2a) shows the mally exfoliated nanolayers with a thickness ranging between 15 and 25 nm and a la size ranging between 55 and 450 nm. The spherical MnO2 nanoparticles (Figure 2b) ap to be slightly agglomerated, but with a uniform shape. The simultaneous depositio the MnO2 and RGO on the surface of the GCE yielded to the MnO2@RGO nanocomp displayed in Figure 2c. Compared to the RGO, the MnO2@RGO nanocomposite sh nanolayers with some nanoparticles on their surface, signifying that the RGO nanosh act as a matrix for connecting the MnO2 nanoparticles. Furthermore, the nanosheets in MnO2@RGO nanocomposite appear to be thicker (21-36 nm) and larger in size (62 nm), due to the MnO2 nanoparticles amalgamated on the surface of the RGO.

Transmission Electron Microscopy (TEM)
To gain further insights on the morphology of the RGO nanosheets, MnO2 nano ticles and MnO2@RGO nanocomposite, TEM was explored ( Figure 3). The R nanosheets (Figure 3a) display the typical wrinkled pattern. The uniformly spheri shaped MnO2 nanoparticles are displayed in Figure 3b. Interestingly, the unifo shaped MnO2 nanoparticles adhere to the surface of the RGO nanosheets after their ultaneous deposition (Figure 3c). This then suggests that there were strong interac between RGO sheets and MnO2 particles, which might then prevent the leaching of M nanoparticles in solution [3]. This observation corroborates with the HR-SEM results pectedly, three amorphous diffraction rings corresponding to the unique (002), (110), facets on the XRD of the RGO were observed in the selected area electron diffrac

Transmission Electron Microscopy (TEM)
To gain further insights on the morphology of the RGO nanosheets, MnO 2 nanoparticles and MnO 2 @RGO nanocomposite, TEM was explored ( Figure 3). The RGO nanosheets ( Figure 3a) display the typical wrinkled pattern. The uniformly spherically shaped MnO 2 nanoparticles are displayed in Figure 3b. Interestingly, the uniformly shaped MnO 2 nanoparticles adhere to the surface of the RGO nanosheets after their simultaneous deposition ( Figure 3c). This then suggests that there were strong interactions between RGO sheets and MnO 2 particles, which might then prevent the leaching of MnO 2 nanoparticles in solution [3]. This observation corroborates with the HR-SEM results. Expectedly, three amorphous diffraction rings corresponding to the unique (002), (110), (211) facets on the XRD of the RGO were observed in the selected area electron diffraction (SAED) pattern ( Figure 3d). Furthermore, multiple polyatomic rings with bright spots were observed on the SAED pattern of the MnO 2 nanoparticles, with rings corresponding to the (110), This further confirms the successful integration of MnO 2 nanoparticles on the surface of RGO nanosheets [42].

Fourier Transform Infrared Spectroscopy (FTIR)
The functional groups present in the synthesized RGO nanosheets, MnO2 nanopa cles, and MnO2@RGO nanocomposite were determined using the FTIR ( Figure 4). FTIR spectra of the MnO2 nanoparticles display a broad band around 3404 cm −1 co sponding to the -OH vibration mode of absorbed water [43]. The vibration bands at 1 and 1398 cm −1 can be ascribed to the -OH stretching of the adsorbed water and the meta hydroxide (Mn-OH) formed during the electrodeposition process, respectively [42-The broad band around 647 cm −1 was attributed to the stretching Mn-O bond in M nanoparticles [39,43,44]. The RGO nanosheets display the broad -OH band around 3 cm −1 with a negligible shift in position when compared to the -OH of the MnO2@R nanocomposite [42]. The bands at 2923 and 2862 cm −1 correspond to the -CH and C groups in the RGO structure, respectively. Similarly, these groups showed a neglig change in position on the MnO2@RGO nanocomposite [38,39,43]. Likewise, the ba around 1634 and 1393 cm −1 correspond to the C=C and C-O on the backbone of RGO, w a similar position and negligible change also observed on the MnO2@RGO nanocompo [33,34,42]. The presence of the Mn-O stretching vibration around 633 cm −1 on the na composite suggests the successful integration of the MnO2 with the RGO nanosheets [ 44]. Although the deposition of the RGO nanosheets and MnO2 nanoparticles were d simultaneously to form the MnO2@RGO nanocomposite, the FTIR results obtained s gest that the MnO2@RGO nanocomposite chemical structure resembles that of the R nanosheets, with MnO2 embedded on the structure of the RGO nanosheets. These res agree with the HR-SEM results showing the morphology of RGO as more dominant the MnO2@RGO nanocomposite, suggesting a more conductive material formed [3,42

Fourier Transform Infrared Spectroscopy (FTIR)
The functional groups present in the synthesized RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite were determined using the FTIR (Figure 4). The FTIR spectra of the MnO 2 nanoparticles display a broad band around 3404 cm −1 corresponding to the -OH vibration mode of absorbed water [43]. The vibration bands at 1629 and 1398 cm −1 can be ascribed to the -OH stretching of the adsorbed water and the metallic hydroxide (Mn-OH) formed during the electrodeposition process, respectively [42][43][44]. The broad band around 647 cm −1 was attributed to the stretching Mn-O bond in MnO 2 nanoparticles [39,43,44]. The RGO nanosheets display the broad -OH band around 3140 cm −1 with a negligible shift in position when compared to the -OH of the MnO 2 @RGO nanocomposite [42]. The bands at 2923 and 2862 cm −1 correspond to the -CH and CH 2 groups in the RGO structure, respectively. Similarly, these groups showed a negligible change in position on the MnO 2 @RGO nanocomposite [38,39,43]. Likewise, the bands around 1634 and 1393 cm −1 correspond to the C=C and C-O on the backbone of RGO, with a similar position and negligible change also observed on the MnO 2 @RGO nanocomposite [33,34,42]. The presence of the Mn-O stretching vibration around 633 cm −1 on the nanocomposite suggests the successful integration of the MnO 2 with the RGO nanosheets [42][43][44]. Although the deposition of the RGO nanosheets and MnO 2 nanoparticles were done simultaneously to form the MnO 2 @RGO nanocomposite, the FTIR results obtained suggest that the MnO 2 @RGO nanocomposite chemical structure resembles that of the RGO nanosheets, with MnO 2 embedded on the structure of the RGO nanosheets. These results agree with the HR-SEM results showing the morphology of RGO as more dominant on the MnO 2 @RGO nanocomposite, suggesting a more conductive material formed [3,42].

Nitrogen Absorption-Desorption Studies
The BET properties of RGO nanosheets, MnO2 nanoparticles, and MnO2@RGO nanocomposite were determined from the N2 adsorption-desorption isotherms measured at 77K (Table 1). MnO2@RGO nanocomposite displayed a specific surface area of 148.7 m 2 /g which was by far higher than the 77.5 and 54.1 m 2 /g specific surface area of MnO2 nanoparticles and RGO nanosheets, respectively. The high increase of the surface area can be attributed to the integration of MnO2 nanoparticles with RGO nanosheets. Expectedly, MnO2@RGO nanocomposite displayed higher average pore size and the total pore volume than both RGO nanosheets and MnO2 nanoparticles as summarized in Table 1.

Electrochemical Performance of the Sensing Materials
The electrochemical performance of the functional MnO2@RGO nanocomposite was evaluated on a typical three-electrode setup. The GCE can be defined as the current collector, whereas the RGO nanosheets, MnO2 nanoparticles, and MnO2@RGO nanocomposite are the sensing materials. Prior to trace metal determinations, the bare GCE was immersed in the suspension of each sensing material and cyclic voltammetry (CV) was employed for the electrodeposition of the sensing materials on the polished GCE surface in a potential range of −0.2 to 1.

Nitrogen Absorption-Desorption Studies
The BET properties of RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite were determined from the N 2 adsorption-desorption isotherms measured at 77K (Table 1). MnO 2 @RGO nanocomposite displayed a specific surface area of 148.7 m 2 /g which was by far higher than the 77.5 and 54.1 m 2 /g specific surface area of MnO 2 nanoparticles and RGO nanosheets, respectively. The high increase of the surface area can be attributed to the integration of MnO 2 nanoparticles with RGO nanosheets. Expectedly, MnO 2 @RGO nanocomposite displayed higher average pore size and the total pore volume than both RGO nanosheets and MnO 2 nanoparticles as summarized in Table 1.

Electrochemical Performance of the Sensing Materials
The electrochemical performance of the functional MnO 2 @RGO nanocomposite was evaluated on a typical three-electrode setup. The GCE can be defined as the current collector, whereas the RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite are the sensing materials. Prior to trace metal determinations, the bare GCE was immersed in the suspension of each sensing material and cyclic voltammetry (CV) was employed for the electrodeposition of the sensing materials on the polished GCE surface in a potential range of −0.2 to 1.2 V, at a scan rate of 25 mV s −1 in 0.1 M KCl. The corresponding cyclic voltammograms obtained from the electrodeposition of RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite modified electrodes are displayed in Figure 5a. As seen, the MnO 2 @RGO nanocomposite modified GCE showed persistent enhancement currents when compared to the RGO nanosheets modified GCE and MnO 2 nanoparticles modified GCE. This indicates that the incorporation of MnO 2 nanoparticles on the RGO nanosheets improves the sensing element of the GCE. Evidently, the chargetransfer resistance of the prepared materials was studied, and the EIS measurements for the various electrodes are represented by the Nyquist plots in Figure 5b. Although all the materials display semicircles of different diameters within the high and low frequency regions, the Nyquist plot of the MnO 2 @RGO nanocomposite modified GCE exhibits a very small semicircle, indicating the lowest charge transfer resistance (Rct) (34.71 Ω) than the RGO nanosheets (99.54 Ω) and MnO 2 nanoparticles (124.61 Ω). The low Rct of MnO 2 @RGO nanocomposite modified GCE might be ascribed to the high electrocatalytic activity of the incorporated MnO 2 layer [3] and the increased porous structure of the composite.  Furthermore, the solution resistance (Rs) was determined as 29.39 Ω, 38.23 Ω, a 29.41 Ω for the RGO nanosheets, MnO2 nanoparticles, and MnO2@RGO nanocompos respectively. These results indicate a strong interaction between the electrode-electrol interface region of the MnO2@RGO nanocomposite GCE resulting from the highly poro surface of the MnO2@RGO nanocomposite, as confirmed by the BET results (Table 1). addition to the semicircle observed, the MnO2@RGO nanocomposite modified GCE d play the sharply reduced Warburg impedance (W) initially observed from the low f quencies of the Nyquist plot of the RGO nanosheets. Although MnO2 nanoparticles d play the lowest conductivity, their integration with the conductive RGO nanoshe greatly improves the electron transport capacity and therefore increases the sensitivity shown by the increased current response observed from the cyclic voltammograms (F ure 5a). Following these results, the MnO2@RGO nanocomposite modified GCE was the fore employed to evaluate the response signals of the trace Cd(II), Zn(II), and Cu(II) aqueous solutions.
Prior to the determination of Cu(II), Cd(II), Zn(II) using DPASV, cyclic voltamm grams were recorded at different scan rates Furthermore, the solution resistance (Rs) was determined as 29.39 Ω, 38.23 Ω, and 29.41 Ω for the RGO nanosheets, MnO 2 nanoparticles, and MnO 2 @RGO nanocomposite, respectively. These results indicate a strong interaction between the electrode-electrolyte interface region of the MnO 2 @RGO nanocomposite GCE resulting from the highly porous surface of the MnO 2 @RGO nanocomposite, as confirmed by the BET results (Table 1). In addition to the semicircle observed, the MnO 2 @RGO nanocomposite modified GCE display the sharply reduced Warburg impedance (W) initially observed from the low frequencies of the Nyquist plot of the RGO nanosheets. Although MnO 2 nanoparticles display the lowest conductivity, their integration with the conductive RGO nanosheets greatly improves the electron transport capacity and therefore increases the sensitivity, as shown by the increased current response observed from the cyclic voltammograms (Figure 5a). Following these results, the MnO 2 @RGO nanocomposite modified GCE was therefore employed to evaluate the response signals of the trace Cd(II), Zn(II), and Cu(II) in aqueous solutions.
Prior to the determination of Cu(II), Cd(II), Zn(II) using DPASV, cyclic voltammograms were recorded at different scan rates and v = scan rate (V/s). Using n = 2, C = 5 mM, D 0 = 7.6 × 10 −6 cm 2 s −1 , and the slope (Ip/v 1/2 ), the Aeff values of bare GCE and MnO 2 @RGO nanocomposite modified GCE were calculated as 1.24 and 3.89 cm 2 , respectively. Notably, the Aeff values of the nanocomposite modified GCE is considerably larger than the bare GCE.
the MnO2@RGO nanocomposite modified GCE on the detection of Cu(II), Cd(II), a Zn(II) were presented in Figure 6c. It can be comprehended that the MnO2@RGO nan composite modified GCE produces a large stripping signals compared with the bare GC Furthermore, the influence of the oxygen interference was negligible with favorable s nal-background characteristics. The superior electrocatalytic activity of the MnO2@RG nanocomposite modified GCE achieved can be attributed to the synergistic coupling eff emancipating from the uniformly distributed small MnO2 nanoparticles on the RG nanosheets surface [45]. Although DPASV resulted in satisfactory current responses the detection of Cu(II), Cd(II), and Zn(II) in synthetic samples, the method was optimiz to ensure the accurate determination of the trace metals in real water samples.

Optimization of the DPASV Experimental Parameters
The influential parameters (deposition time (DT), deposition potential (DP), pu amplitude (PA), and step potential (SP)) affecting the determination of Cu(II), Cd( Zn(II) using MnO2@RGO nanocomposite modified GCE were optimized using the sma central composite design (SCCD). The design of the experiments (DOE) is shown in Ta S1, alongside the current response of each element when subjected to the influential p rameters (Table S2). Using the results obtained in Table S2, the significance of each fac was examined using analysis of variance (ANOVA). The ANOVA results were rep sented in the form of pareto charts (Figures 7, S4 and S5). Considering the fact th ANOVA is a linear model, only linear factors (factors with the index L) were consider while quadratic factors (factors with the index Q) were not considered on the pareto ch analysis [46]. Evidently, it can be clearly seen in Figure 7 that the three interactions (1. The differential pulse voltammograms showing the responses of the bare GCE and the MnO 2 @RGO nanocomposite modified GCE on the detection of Cu(II), Cd(II), and Zn(II) were presented in Figure 6c. It can be comprehended that the MnO 2 @RGO nanocomposite modified GCE produces a large stripping signals compared with the bare GCE. Furthermore, the influence of the oxygen interference was negligible with favorable signalbackground characteristics. The superior electrocatalytic activity of the MnO 2 @RGO nanocomposite modified GCE achieved can be attributed to the synergistic coupling effect emancipating from the uniformly distributed small MnO 2 nanoparticles on the RGO nanosheets surface [45]. Although DPASV resulted in satisfactory current responses on the detection of Cu(II), Cd(II), and Zn(II) in synthetic samples, the method was optimized to ensure the accurate determination of the trace metals in real water samples.

Optimization of the DPASV Experimental Parameters
The influential parameters (deposition time (DT), deposition potential (DP), pulse amplitude (PA), and step potential (SP)) affecting the determination of Cu(II), Cd(II), Zn(II) using MnO 2 @RGO nanocomposite modified GCE were optimized using the small-central composite design (SCCD). The design of the experiments (DOE) is shown in Table S1, alongside the current response of each element when subjected to the influential parameters (Table S2). Using the results obtained in Table S2, the significance of each factor was examined using analysis of variance (ANOVA). The ANOVA results were represented in the form of pareto charts (Figure 7, Figures S4 and S5). Considering the fact that ANOVA is a linear model, only linear factors (factors with the index L) were considered, while quadratic factors (factors with the index Q) were not considered on the pareto chart analysis [46]. Evidently, it can be clearly seen in Figure 7 that the three interactions (1. DT and DP; 2. DP and PA; 3. DT and PA), alongside the independent PA factor were significant at 95% confidence level for Zn. However, Cd ( Figure S4) only shows one interaction between DP and DT as significant at 95% confidence level, alongside PA, which is the only independent factor that was significant. The pareto chart for Cu ( Figure S5) shows that none of the factors were significant at 95% confidence level. However, PA and SP were found to be important and influential on analytical response. It can therefore be clearly seen that the PA has the overall most significant effect on the analytical response (peak current).
Membranes 2021, 11, x FOR PEER REVIEW 10 of 18 and DP; 2. DP and PA; 3. DT and PA), alongside the independent PA factor were significant at 95% confidence level for Zn. However, Cd ( Figure S4) only shows one interaction between DP and DT as significant at 95% confidence level, alongside PA, which is the only independent factor that was significant. The pareto chart for Cu ( Figure S5) shows that none of the factors were significant at 95% confidence level. However, PA and SP were found to be important and influential on analytical response. It can therefore be clearly seen that the PA has the overall most significant effect on the analytical response (peak current). The response surface methodology (RSM) was used to investigate the interaction and the quadratic effects of the main parameters DP, DT, and SP and PA, using the data generated from the experimental SCCD, resulting in 3D surface plots for the Zn(II) optimization ( Figure 8). Clearly, from Figure 8a, increasing DT to any value greater than 250 results in higher current response, even at low DT values. Similarly, the DP at higher values (0 to −0.7) lead to higher current response at low DT values, signifying the significance of the interaction of DP and DT. Furthermore, the interaction of SP with DP in Figure 8b clearly shows that any given SP value (from −4 to 1.4/higher) interacting with low DP values (from −1.6 to −2.0/lower) result to higher current responses. Figure 8c shows that increasing the PA to higher values (>90) leads to a higher current response at lower or higher DP values, indicating the significance of PA, alongside the insignificance of DP. Figure 8d shows that a higher current response can be achieved at any SP value (−4 to 14/higher) interaction with higher DT (>350) or lower DT (<0), indicating the significance of the interaction of SP and DT at given values. Furthermore, low PA and low DT (both below 0) give high response as shown in Figure 8e. Finally, the interaction of PA and SP shows that higher PA (>110) correspond to higher response at any given SP, indicating the significance of PA and the insignificance of SP at given values (Figure 8f). The results obtained in these RSM plots, noticing the significant interactions, correspond to the pareto chart analysis.
The RSM plots for Cd(II) are shown in Figure S6. In Figure S6a, it can be seen that at higher DT values and higher DP values, there is a higher current response. Figure S6b shows that a DP between −1.6 to −0.4 and any SP between the given DP range leads to a higher current response. Furthermore, higher DP (>−1.8) and a higher PA (>120) interaction leads to a higher current response as shown in Figure S6c. DT between 0 and 350 The response surface methodology (RSM) was used to investigate the interaction and the quadratic effects of the main parameters DP, DT, and SP and PA, using the data generated from the experimental SCCD, resulting in 3D surface plots for the Zn(II) optimization ( Figure 8). Clearly, from Figure 8a, increasing DT to any value greater than 250 results in higher current response, even at low DT values. Similarly, the DP at higher values (0 to −0.7) lead to higher current response at low DT values, signifying the significance of the interaction of DP and DT. Furthermore, the interaction of SP with DP in Figure 8b clearly shows that any given SP value (from −4 to 1.4/higher) interacting with low DP values (from −1.6 to −2.0/lower) result to higher current responses. Figure 8c shows that increasing the PA to higher values (>90) leads to a higher current response at lower or higher DP values, indicating the significance of PA, alongside the insignificance of DP. Figure 8d shows that a higher current response can be achieved at any SP value (−4 to 14/higher) interaction with higher DT (>350) or lower DT (<0), indicating the significance of the interaction of SP and DT at given values. Furthermore, low PA and low DT (both below 0) give high response as shown in Figure 8e. Finally, the interaction of PA and SP shows that higher PA (>110) correspond to higher response at any given SP, indicating the significance of PA and the insignificance of SP at given values (Figure 8f). The results obtained in these RSM plots, noticing the significant interactions, correspond to the pareto chart analysis.
( Figure S6d). Any DT value interacting with PA (<0) gives a maximum response (Figu S6e). Lastly, any SP value interacting with PA (<0) gives a maximum response (Figu S6f). These results are in agreement with the pareto chart analysis, as SP and DT we clearly insignificant. The RSM plots of Cu(II) shown in Figure S7 show a similar patter with the RSM plots of Cd(II). Subsequently, DP, DT and SP were clearly insignificant o the current response, thus agreeing with the pareto chart analysis.  The RSM plots for Cd(II) are shown in Figure S6. In Figure S6a, it can be seen that at higher DT values and higher DP values, there is a higher current response. Figure S6b shows that a DP between −1.6 to −0.4 and any SP between the given DP range leads to a higher current response. Furthermore, higher DP (>−1.8) and a higher PA (>120) interaction leads to a higher current response as shown in Figure S6c. DT between 0 and 350 interacting with any SP between the DP values given result to a higher current response ( Figure S6d). Any DT value interacting with PA (<0) gives a maximum response ( Figure S6e). Lastly, any SP value interacting with PA (<0) gives a maximum response ( Figure S6f). These results are in agreement with the pareto chart analysis, as SP and DT were clearly insignificant. The RSM plots of Cu(II) shown in Figure S7 show a similar pattern with the RSM plots of Cd(II). Subsequently, DP, DT and SP were clearly insignificant on the current response, thus agreeing with the pareto chart analysis.
The simultaneous estimations of the optimum conditions leading to the maximum current response were evaluated using the desirability function (DF) for all the parameters and metals (Figure 9, Figures S8 and S9

Analytical Figures of Merit
Under optimum conditions, calibration curves from the differential pulse volta grams using MnO2@RGO/GCE were studied for different concentrations of each me (Cu, Zn and Cd). The limits of detection (LOD = 3Sb/m) and the limits of quantifi (LOQ = 10Sd/m) are presented in Table 2, where Sd and m are standard deviation blank signal (n = 20) and the slope of the calibration curve, respectively. The intrad peatability) precision of the MnO2@RGO/GCE was investigated by performing 15 urements of the standard solution of Cd, Cu and Zn ions at a concentration of 25 μ Additionally, the interday (reproducibility) precision was evaluated by preparin MnO2@RGO/GCE from the same batch in five consecutive working days MnO2@RGO/GCE were used for the determination of 25 μg L −1 of metal ions. The sions were expressed in terms of relative standard deviation (RSD) and the resu summarized in Table 2. Consequently, the intraday and interday precision MnO2@RGO/GCE were acceptable, as they were less than 5%. Moreove MnO2@RGO/GCE was found to be reusable for four weeks and after the response MnO2@RGO/GCE decreased.

Analytical Figures of Merit
Under optimum conditions, calibration curves from the differential pulse voltammograms using MnO 2 @RGO/GCE were studied for different concentrations of each metal ion (Cu, Zn and Cd). The limits of detection (LOD = 3Sb/m) and the limits of quantification (LOQ = 10Sd/m) are presented in Table 2, where Sd and m are standard deviation of the blank signal (n = 20) and the slope of the calibration curve, respectively. The intraday (repeatability) precision of the MnO 2 @RGO/GCE was investigated by performing 15 measurements of the standard solution of Cd, Cu and Zn ions at a concentration of 25 µg L −1 . Additionally, the interday (reproducibility) precision was evaluated by preparing five MnO 2 @RGO/GCE from the same batch in five consecutive working days. The MnO 2 @RGO/GCE were used for the determination of 25 µg L −1 of metal ions. The precisions were expressed in terms of relative standard deviation (RSD) and the results are summarized in Table 2. Consequently, the intraday and interday precisions for MnO 2 @RGO/GCE were acceptable, as they were less than 5%. Moreover, the MnO 2 @RGO/GCE was found to be reusable for four weeks and after the response of the MnO 2 @RGO/GCE decreased. Subsequently, the comparison of MnO 2 @RGO/GCE with other reported modified electrodes in the determination of Cd, Cu and Zn ions is presented in Table 3. It can be seen from the table that the LOD and sensitivity towards determination of Zn improved.
Additionally, it can be perceived that the present method exhibited a wider dynamic range and better or similar precision compared those reported in the literature.

Surface Water (SW) Sample Analysis
The applicability and validity of the proposed MnO 2 @RGO/GCE was investigated by the application on the determination of Zn 2+ , Cu 2+, and Cd 2+ ions in surface water samples. Initially, the accuracy of the developed method was assessed by analyzing three water samples spiked with 2.00, 5.00, and 10.0 µg L −1 concentrations of Cu(II), Zn(II), and Cd(II). The concentration of the target elements was achieved by comparing the values obtained by using the calibration curve method. The results and percentage recovery values are shown in Table 4. According to the results obtained, a good agreement between the added and an obtained analyte concentration was achieved. The percentage recoveries ranged from 97-99.9%. These results reveal that that the developed electrochemical sensor method has a relatively good accuracy in real water sample matrix. The applicability of the developed electrochemical sensor for the simultaneous determination of Cd, Cu, and Zn in real water samples was investigated. The analytical results obtained are presented in Tables 4 and 5. It can be seen in Table 5 that the results obtained by the developed electrochemical method show a good agreement with those obtained by the ICP-OES analysis. Therefore, these findings confirmed the applicability of the proposed method for precise and accurate simultaneous determination of Cd, Cu and Zn in real environmental water samples. In addition, these results revealed that the current electrochemical sensor has capabilities to determine the trace metal ions the real water samples without significant interferences.

Stability of the Electrochemical Sensor
The stability of the modified GCE was investigated by repetitively measuring 25 µg L −1 of Cd, Cu, and Zn using the same modified electrode. These experiments were conducted under optimal experimental conditions. After every use, the MnO 2 @RGO nanocomposite modified GCE was immersed in 2 M nitric acid to dissolve any adsorbed metal ions and stored in the refrigerator at 4 • C. The experimental results for the long-term stability of the MnO 2 @RGO nanocomposite modified GCE are displayed in Figure 10. Evidently, the developed sensor was found to be stable as it retained about 95% of the initial current response in the determination of Cu(II), Cd(II), and Zn(II). This implies that the sensor could be used continuously for at least 7 days without any significant loss in performance.

Conclusions
This work reported on the determination of Cu(II), Cd(II), and Zn(II) using the fabricated MnO2@RGO/GCE modified electrode using DPASV. The electrodeposited MnO2@RGO/GCE exhibited high electrocatalytic activity owing to its increased surface area, low charge-transfer resistance at the electrolyte-electrode interface, and high selectivity to Cu(II), Cd(II), and Zn(II). Experimental parameters such as deposition potential, deposition time, pulse amplitude, and step potential were optimized using the small central composite design and the response surface methodology for Cu(II), Cd(II), and Zn(II). The experimental electrochemical results using the optimum conditions suggest that the MnO2@RGO nanocomposite enhances the current response on the detection of Cu(II), Cd(II), and Zn(II) in surface water. Additionally, the MnO2@RGO modified GCE exhibited several advantages, such as the simple preparation, reproducibility, and good stability.

Conclusions
This work reported on the determination of Cu(II), Cd(II), and Zn(II) using the fabricated MnO 2 @RGO/GCE modified electrode using DPASV. The electrodeposited MnO 2 @RGO/GCE exhibited high electrocatalytic activity owing to its increased surface area, low charge-transfer resistance at the electrolyte-electrode interface, and high selectivity to Cu(II), Cd(II), and Zn(II). Experimental parameters such as deposition potential, deposition time, pulse amplitude, and step potential were optimized using the small central composite design and the response surface methodology for Cu(II), Cd(II), and Zn(II).
The experimental electrochemical results using the optimum conditions suggest that the MnO 2 @RGO nanocomposite enhances the current response on the detection of Cu(II), Cd(II), and Zn(II) in surface water. Additionally, the MnO 2 @RGO modified GCE exhibited several advantages, such as the simple preparation, reproducibility, and good stability. Furthermore, the fabricated electrochemical sensor proved to be a better alternative method that can be used for the determination of heavy metals in environmental matrices without the sample preparation step and can be used in routine analysis.