The XRD plots of the synthesized ZnO nanopowder, GO and GN:ZnO NC composite are given in Fig. 1. XRD analysis of ZnO NPs depicts diffraction peaks at 31.80°, 34.4°, 36.3°, 47.5°, 56.6°, 62.8° and 69.0° in the (Fig. 1A), confirming the hexagonal wurtzite structure of ZnO NPs 41, 52, 53 (JPCDS number: 36-1451) 52. Also, the presence of quite sharp peaks and the absence of any other peak indicate that the synthesized nanopowder does not contain other impurities. The size of ZnO-NPs was also calculated with the help of “Debye-Scherrer formula, 𝑑 = 0.89𝜆/𝛽cos𝜃” 55, where, 0.89 is Scherer’s factor, 𝜆 is wavelength of x-ray used, ‘𝜃’ the “Bragg diffraction angle”, and ‘𝛽’ a constant known as the “full width at half-maximum (FWHM)” of a particular diffraction angle. The mean particle size of the ZnO-NPs corresponding to the FWHM of the most intense peak at 36.3° and corresponding to (101) plane using the “Scherrer formula” 56, 57 was found to be 57.17 nm. The diffraction peak at a 2θ value of 10.810 characteristics of GO (Fig. 1B) disappears in the XRD pattern of as-prepared GN:ZnO NC composite (Fig. 1C). However, a new diffraction peak characteristic of GN at 26.520 with inter-layer spacing (d-spacing) of 0.35 nm was observed in the XRD outline of GN:ZnO NC composite. This confirms that the exfoliated GO has been reduced to GN. The other XRD peaks of the GN:ZnO NC composite can be ascribed to the crystallize ZnO NPs. The ZnO NPs and GN:ZnO NC composites were further analyzed using SEM and EDX techniques. Figure 2A shows the SEM image of as-synthesized ZnO-NPs. As observed, the approximate spherical-shaped ZnO NPs were formed in the range of 40–90 nm size which agrees with the particle size calculated using the “Debye-Scherrer formula” in XRD studies as mentioned above. Figure 2B is the SEM image with distinctive crumpled and wrinkled surface characteristics of GO and Fig. 2C shows the multi-layer structure of graphene sheets stacked together. Figure 2D is the SEM image of GN/ZnO composite which was taken after reduction of GO in presence of ZnO NPs. It is observed that ZnO-NPs are closely spread on the surface of graphene-nanosheets. However, in some regions, few ZnO NPs appear aggregated on the surfaces of GN nanosheets, but most of the ZnONPs are found on the GN sheets and several of which are wrapped by GN sheets.
TEM and HRTEM were used to further characterize the microstructure of synthesized ZnO NPs and GN:ZnO NC. Figure 3A shows the TEM image of ZnO NPs which confirms its formation and the particle size varies from 40–100 nm as analyzed by the particle size analyzer which agrees with the results as obtained in SEM images. HR-TEM image of ZnO NPs from a single nanoparticle is shown in Fig. 3B. The selected area electron-diffraction (SAED) outline so obtained is given in Fig. 3C and the rotational average profile of ZnO NPs is given in Fig. 3D. The lattice spacing values obtained from the SAED outline and the standard “JCPDS” values are given in Table S1. The lattice spacing calculated from the selected area is 2.48 Å which matches perfectly with the value obtained from the standard JCPDS card (2.47590 Å) corresponding to 101 planes of hexagonal ZnO. The values given in Table S1 agree very well within the error limit of ± 2.5% confirming the hexagonal phase of the ZnONPs. Figure 3D represents the rotational average profile of ZnO-NPs. Figure 4A represents the TEM image and Fig. 4B shows the HRTEM image of as-synthesized GN and Fig. 4C shows its corresponding Fast Fourier Transform (FFT) pattern. The hexa-gonal FFT outline shows that the as-synthesized GN is highly crystalline in nature with minimum defects. TEM image of GN:ZnO NC is shown in Fig. 4D which shows that the ZnO-NPs are well dispersed in the graphene sheets.
These results show that there is a good interfacial contact of GN and ZnO-NPs. The existence of ZnO-NPs within the GN layers will help to prevent the agglomeration as well as restacking of plane GN layers and simultaneously its available surface-area gets also increased to improve its electrochemical performance. The GN:ZnO NC was also characterized by EDAX technique. Figure 5 shows the EDAX spectrum of GN:ZnO NC with peaks characteristic of C, O, and Zn elements. The respective weight percentages using EDAX studies of C, O, and Zn are 66.94, 10.85, and 22.21. This reveals a good loading of ZnO-NPs on graphene nanosheets, which agrees with the SEM and TEM interpretations. In this way, a good dispersion of ZnO NPs with GN is achieved.
Raman spectroscopy is a very sensitive technique for electronic structure and is an essential tool to examine well-ordered and disordered crystalline structures and is widely used to characterize carbonaceous materials like graphene58. Raman spectra of the GO, GN, and GN:ZnO-NC are presented in Fig. 6. The two characteristic bands in the Raman spectrum of GN and graphene-based materials are low intensity ‘D-band’ and high intensity ‘G-band’ at 1351 and 1602 cm− 1, corresponding to the disordered carbon and in-plane stretching vibration of sp2 C − C bonds, respectively 44, 59. The ID/IG intensity ratio gives the degree of disorderness and the mean size of the sp2-domains in graphitic materials44, 59, 60. Figures 6(A) and 6(B) shows that ID/IG intensity ratio is increased for GN as compared to GO which indicates that there is a decrease in the in-plane sp2-domain size on graphene-oxide reduction 59. The ID/IG value gets increased from 0.54 in GO to 1.13 in GN:ZnO NC (Fig. 6C). This is because on reduction of GO to GN, oxygen functionalities are removed, which decreases the size of the sp2-domains in graphene and interaction between ZnO NPs and graphene nanosheets 44, 45, 61.
During the reduction of GO to GN, most oxygen based functionality are removed and restores the conductivity of graphene, and also increases the stability of GN:ZnO NC62. However, some residual oxygen-containing groups are still left unreduced on GN [42]. The Zn atoms present in ZnO NPs coordinate with the oxygen atoms of the unreduced functional groups via covalent-interaction which helps in composite formation42. Graphene bearing edges and planes will provide support to spherical ZnO NPs to grow and form GN:ZnO NC.
Electrochemical characterization of GN:ZnO NC/GCE
Electrochemical impedance spectroscopic (EIS) and chrono-coulometric (CC) techniques were applied to characterize the GN:ZnO NC modified GCE. EIS is a suitable technique commonly used to characterize a surface modified GCE45. In the Nyquist diagrams, the semicircular-region obtained at high frequency values correspond to the charge transfer limiting phenomena and from the diameter of the semicircular part, charge transfer values (Rct) are calculated63. EIS analysis shows that the Rct value for GN:ZnO NC/GCE (Fig. 7A) is 3.6 Ω which is lower than that of GN/GCE (6.0 Ω, Fig. 7B). The decrease in Rct value confirms that the ZnO-NPs present in the composite improves the charge transfer characteristics of GN:ZnO NC modified GCE 42. CC was applied to determine the electrochemically active surface area of the electrodes. The slope obtained from the plot of Q against t1/2 using CC and 1 mM K3[Fe(CN)6] standard complex was used to calculate the effective electrochemical surface areas for GN and GN:ZnO NC modified electrodes (Fig.S1A & B). The effective electrochemical surface areas for GN:ZnO NC/GCE and GN/GCE were found to be 0.166 and 0.092 cm2, respectively.
Electrochemical performance of GN:ZnO NC modified GCE for p-NP detrmination
Cyclic voltammetric (CV) study was carried out at a scan-rate of 100 mVs− 1 to investigate the electrochemical behavior of 1.96 x10− 6 M p-NP at bare GCE, GN/GCE, and GN:ZnO NC/GCE. Figure 8 (a) is the CV of the blank containing phosphate buffer at pH 6.8 at bare GCE. However, using bare-GCE at the same scan-rate and concentration of 1.96 x10− 6 M, p-NP gives a weak redox couple at 0.20 V ascribed to p-NP (Fig. 8b), while the peak current progressively increased on GN modified GCE (Fig. 8c), and further enhanced with GN:ZnO NC modified GCE (Fig. 8d). Also, the irreversible reduction peak of p-NP at -0.75 V grows progressively at both GN and GN:ZnO NC modified GCE. This shows that both ZnO NPs and GN had an electrocatalytic effect where ZnO NPs act as spacer to enhance the effective-surface area of GN and improve the electrochemical activity while GN decreases the resistance by providing a conducting path. Therefore, the synergic effect of ZnO NPs and GN as cocatalysts is responsible for the electrocatalytic activity of GN:ZnO NC/GCE and its application in the electro-analysis of para-nitrophenol. This is due to the excellent catalytic capability and large effective-surface area of GN:ZnO-NC/GCE which facilitates p-NP to accommodate at the electrode-surface and enhances the electron-transfer process. The current intensity increased progressively on GCE modified with GN and GN:ZnO NC. Figure 8C shows that one large reduction peak appears at -0.75 V on the first cathodic sweep for the p-NP and another redox couple with one anodic-peak at 0.19 V and another cathodic-peak at 0.11 V is obtained. It is observed from the cyclic voltammograms (Fig. 8) that this redox couple grows in peak current at the expense of first irreversible reduction-peak at 0.75 V. This indicates that the irreversible reduction product of p-NP stays on the modified-electrode surface and gets oxidized in the anodic sweep. This is also in agreement with the literature employing the conventional electrodes18, 21, 64, 65. The first irreversible reduction-peak obtained is ascribed to the irreversible reduction of the nitroaromatic moiety to yield the hydroxylamine derivative (NHOH) with the four-electron transfer, according to Scheme 2. The anodic redox couple is due to the redox reaction of hydroxylamine and p-NP involving two electrons in the reaction 18, 30. Figure 9A shows the scan rate effect on peak current using CV for p-nitrophenol (1.96 x10− 6 M) in phosphate-buffer of pH 6.8 at 50, 100, 200, 300, 400, 500, 600, 700, 800 mVs− 1 at GN:ZnO NC/GCE. The plot of scan rate against the cathodic and anodic-peak current (Fig. 9B) of reduction product of p-NP (hydroxylamine) gives straight lines which indicate that both the electrode processes are adsorption controlled. Also, when reduction peak-current of p-nitrophenol was plotted against scan-rate, a straight line was obtained indicating an adsorption-controlled process.
Experimental parameters and their optimization for p-nitrophenol determination
Effect of pH values
Nature as well as the pH of supporting electrolytes play a vital role in the determination of analytes using any chemically modified electrode. Many supporting electrolytes were analyzed. While choosing a specific supporting electrolyte for p-NP determination, peak current as well as peak shape were taken into consideration. 0.1 M each of borate, acetate, phosphate, and Britton Robinson buffer were tested. In citrate buffers of pH 2.1 and KH2PO4–K2HPO4 buffer solution of pH 6.8, high-intensity anodic peak currents were observed. However, in the citrate buffer of pH 2, the anodic peak was unstable. Therefore, for subsequent studies, the optimum buffer solution chosen was phosphate buffer (pH 6.8). Differential pulse voltammetry, square-wave voltammetry, and cyclic voltammetric techniques were applied to optimize the pH values and it was found that square-wave voltammetry responded the best in respect of peak-current and peak-resolution. Square wave adsorptive stripping voltammograms of p-NP (1.66 x10− 6 M) at GN:ZnO NC/GCE and various pH values of 0.1 M potassium phosphate buffer are presented in Fig. 10A. It is observed that the peak current goes on increasing with the increase in pH values up to 6.8 but at 7.4 and 8.0 pH values it decreases. It confirms that the redox reaction of p-NP is favored in the acidic medium. Therefore, for further studies phosphate-buffer of pH 6.8 was selected as the supporting-electrolyte. Also, it is observed from Fig. 10A that the peak-potential is shifting towards a more negative side as the pH increase which confirms the proton involvement in the reaction18. A plot of pH against shift in peak potential (Fig. 10B) gives a straight line with a slope equal to 61.3 mV/pH which confirms that protons and electrons are equally taking part in the electrode reaction as shown in scheme 2.
Optimization of deposition potential and deposition time
Deposition potential and deposition time, both effect the degree of adsorption of the electrochemically reduced p-NP at the electrode surface (NH-OH). Therefore, both these paratmeters need to be optimized. The effect of deposition-potential and deposition-time on peak current was investigated using SW-AdSV. It was found that the deposition-potential largely effects the peak current of para-NP at the modified electrode and the peak current enhanced from − 0.1 to -1.0 V and then diminished. Hence a deposition-potential of -1.0 V was optimized for subsequent studies. Figure S2 (Supplementary material) presents the effect of deposition-time on the anodic peak current of NH-OH for 10.70 x10− 7 M p-NP at GN:ZnO NC/GCE under optimized conditions (deposition potential: -1 V, potassium phosphate-buffer of pH 6.8 as supporting electrolyte) for various deposition times (60, 120, 200, 300, 400 and 500 sec). The peak current increases with the increase in deposition-time from 60–300 sec and then remain almost constant. This is because the modified electrode surface gets saturated due to the accumulation of reduction product (NH-OH). In this way, sensitivity can be improved by increasing the deposition time to determine low concentration levels. A deposition time of 300 s was optimized for subsequent experiments.
Calibration curve
Figure 11A shows the calibration curve for the fabricated sensor using the optimized experimental parameters. The calibration curve was used for the quantitative analysis of p-NP solutions using GN:ZnO NC/GCE employing SW-AdSV using the as-optimized experimental parameters. The stripping-peak current gave a linear relation by changing the concentrations of p-NP from 0.09 x10− 6 M to 21.80 x10− 6. Using the linear plot of the peak current (ipa) versus concentration of p-NP (Fig. 11B) the regression equation obtained is; ipa (µA) = 68.14 C(µM) + 10.67. The limit of detection (LOD) and sensitivity of the developed sensor using this linear regression equation were estimated to be 8.8 x10− 9 M (S/N = 3) and 68.14 µAµM− 1, respectively. This very low detection limit of the sensor shows that the developed method can potentially be applied for the sensitive determination of p-NP. A comparative analysis for the performance of the developed sensor with different electrochemical sensors based on different nanomaterials for the sensitive determination of p-nitrophenol is summarized in Table S2. It is obvious from Table S2 that the sensor based on the present nanocomposite material is superior in respect of its wide linear working range and more important its limit of detection is least compared to those reported by the earlier workers. It can be concluded that the GN:ZnO NC-based electrode is comparatively an first-rate platform for the sensitive determination of p-NP. This improved performance of the electrochemical sensor is attributed to the synergistic effect of GN and ZnONPs based composite material due to their high adsorption ability and excellent electrocatalytic activity.
Effect of potentially interfering substances
Selectivity of the prepared GN:ZnO NC-based sensor for the analysis of p-nitrophenol was evaluated by studying the influence of some important interfering substances of p-NP. The interference study was examined in 0.1 M phosphate-buffer of pH 6.8 in presence of 1.07x10− 6 M p-NP. Most of the phenols like pyrocatechol, o-aminophenol, hydroquinone, hydroxyphenyl,, p-aminophenol, and chlorophenol did not affect the signals of p-NP with deviations less than 4% up to the 100-fold excess. The common nitrophenols that contain the same nitro group as that of p-nitrophenol are o-nitrophenol, m-nitrophenol, and 2, 4, di-nitrophenol. The interference of these nitrophenols was tested on the GN:ZnO NC modified GCE. It was found that o-NP produced a separate reduction peak at -0.10V at the same concentration and suppressed the signal of p-NP at higher concentrations as shown in Fig. 12. Therefore, it can be concluded that both p-nitrophenol and o-nitrophenol can be detected simultaneously at the said modified GCE at a lower concentration. However, the presence of m-NP and 2, 4, DNP did not affect the determination of p-nitrophenol. The results obtained from interference tests indicate that GN:ZnO NC-based sensor will be appropriate for the analytical determination of p-NP in presence of the above-mentioned phenolic compounds and meta and ortho-nitrophenol.
Table 1
A: Precision and accuracy analysis for p-nitrophenol in pre-analyzed samples by the developed method (SW-AdSV)
Added
(10− 8M)
|
Found
(10− 8M)
|
(%R)
|
Precision
(% R.S.D, n = 5)
|
Accuracy
(% Bias)
|
Intraday
|
|
|
|
|
19.60
|
19.34
|
98.63
|
1.21
|
-1.32
|
38.4
|
38.35
|
99.8
|
1.78
|
-0.13
|
74.0
|
75.20
|
98.37
|
1.90
|
1.62
|
Interday
|
|
|
|
|
19.6
|
19.80
|
98.97
|
1.89
|
1.02
|
38.4
|
38.76
|
100.39
|
2.07
|
0.93
|
74.0
|
73.01
|
98.66
|
1.88
|
-1.33
|
Table 1B: Percentage recovery assay for p-nitrophenol in real water samples by the developed procedure (SW-AdSV)
|
Sample
|
Added(10− 8M)
|
Founda(10− 8 M) (n = 5)
|
% Recovery
|
Panchganga river water
|
35.61
|
34.78
|
97.66
|
71.22
|
70.01
|
98.30
|
106.83
|
103.99
|
97.34
|
Ground water
|
35.61
|
34.59
|
97.13
|
71.22
|
69.20
|
97.16
|
|
106.83
|
104.31
|
97.64
|
Average of five measurementsa |
The fabricated electrode was also evaluated for its reproducibility and stability. The reproducibility for the detection of 1.66 x10− 6 M p-NP using 0.1 M phosphate-buffer of pH 6.8 employing five electrodes was found to have a relative-standard-deviation (RSD) value of 4.7 %, revealing a good reproducibility of the sensor. The stability of the sensor was also examined in 0.1 M phosphate-buffer of pH 6.8 for 1.66 x10− 6 M p-nitrophenol and the current response obtained was periodically monitored. An initial peak current response of about 93.4 % of the sensor was retained for 15 days which proved good stability of the sensor. The repeatability of the as developed sensor was also carried out for 10 successive measurements with an RSD value of 4.0 % indicating good repeatability.
Accuracy and precision of the method
The accuracy of the method as developed was determined by spiking an exactly weighed amount (pre-analyzed amount) of p-NP. The accuracy is represented as a mean relative error (Table 1A). The mean percentage recovery was found to be 98.93 and 99.34 for intra-day and inter-day assays, respectively. The recovery values are useful to provide good confidence in the accuracy of the developed method.
The precision of the developed method was also determined. The concentration of p-nitrophenol was analyzed in pre-analyzed sample solutions five times in intra-day assay and successively for five days in an inter-day assay using the SW-AdSV techniques. The precision and the percentage recovery values of the developed method obtained as the mean of five separate measurements are given in Table 1A. The average values of variation-coefficients for intra-day and inter-day assays based on five measurements were found to be 1.63 and 1.94%, respectively. Thus, the results obtained confirm a good reproducibility of the fabricated electrode and a high precision of the developed method.
Practical application of the developed senor in the determination of para-NP in real samples
The GN:ZnONPs composite based sensor was used to analyze p-NP in various real water samples. Groundwater sample was collected from Andheri (East) and river water sample from Panchganga River, Mumbai. 0.1 M potassium phosphate-buffer (supporting electrolyte) was directly prepared in 100 ml of each sample which was used as an analyte. 25 cm− 3 of sample volume was taken in an electrochemical cell for detection of p-NP using the standard addition method. The developed method was applied under the optimized experimental parameters (deposition potential: -1.0 V and deposition time : 300 sec) and the measurements were performed three times. It was found that neither p-NP nor o-NP was detected in both these water samples. Therefore, the desired amount of standard p-nitrophenol was spiked into the collected water samples for performing the recovery tests. The percentage recovery of both the samples is given in Table (1B). A recovery of 97.66 and 97.13% was obtained in river and groundwater samples respectively. These observations validate the suitability of the developed method for the determination of p-nitrophenol in the natural water samples.