Voltammetric Determination of Rhodamine B Using a ZIF-67/Reduced Graphene Oxide Modified Electrode

University of Sciences, Hue University, 49000, Vietnam TT-Hue Department of Food Safety and Hygiene, 49000, Vietnam University of Education and Science, The University of Da Nang, 50000, Vietnam Ly Thai To High School, 70000, Vietnam Dalat Nuclear Research Institute, 66000, Vietnam Qui nhon University, 55000, Vietnam University of Natural Resource and Environment of Ho Chi Minh City, 70000, Vietnam


Introduction
Rhodamine B (RhB), 9-(2-carboxyphenyl)-6-diethylamino-3-xanthenylidene]-diethylammonium chloride), is one of the xanthene dyes and is highly water-soluble. It is the most common synthetic dyes employed as a colorant in textiles, foodstuffs, plastic, leather, and paper [1]. RhB is thought to be a carcinogen and to exhibit developmental toxicity, neurotoxicity, and chronic toxicity towards humans and animals. It has been illegally used to provide the red color in chili pow-der and sauces as one of the methods of economically motivated adulteration. Thus, due to the health risks arose from RhB consumption, there have been intensive research efforts to develop a simple method for the determination of RhB in different sample matrix. Some methods are available for the determination of RhB such as spectrophotometric determination with micelle-mediated cloud point extraction [2], UV-Vis Spectrophotoscopy following solid phase extraction [3], magnetic solid phase extraction using ionic liquid-coated coreshell magnetic nanoparticles followed by high-performance liquid chromatography [4], and high-pressure liquid chromatography (HPLC) [5,6]. However, the fact that the operational procedure is complicated, and the cost is high limits the application of these approaches.
The voltammetric approaches have been recognized as the potential methods for the detection of organic and inorganic compounds in trace amount as they offer a simple, low cost, highly selective, and in situ operational procedure. The voltammetric method has been employed for the detection of RhB in which it was used to determine RhB in food [7][8][9]. The development of novel electrodes by modifying them with hybrid organic-inorganic materials provides potential tools in trace analysis of not only RhB but other organic and inorganic substances. Several modified electrodes such as exfoliated graphene-modified electrode [7], silica-pillared zirconium phosphate/nafion composite (SPZP/NAF)-modified electrode [8], betacyclodextrin functionalized nanogold/hollow carbon nanosphere nanohybrid (beta-CD-AuNPs/HCNS)modified electrode [9], and core-shell-structured Cu@carbon sphere (Cu@CS) nanohybrid-modified electrode have been developed to determine RhB content in different samples by voltammetric methods. Zeolitic imidazolate frameworks-67 (ZIF-67), which are constructed from the tetrahedrally coordinated divalent cations Co 2+ linked by the imidazolate ligands, are typical ZIFs with high thermal and chemical stability, as well as tunable zeotype topologies. ZIF-67 have been applied in gas storages and adsorption catalyst [10,11] but has been rarely employed in electrochemistry due to its poor electronic conductivity. Reduced graphene oxide (rGO), an oxidized form of graphene, has attracted great interest in the nanoelectronics and electrochemical industry because it has exhibited outstanding properties such as superior electrical conductivity, high surface area, and chemically stable behaviors. The combination of desirable properties of ZIF-67 and rGO expects a versatile electrode modifier. To our knowledge, few works in the literature have studied the voltammetric determination of Rhodamine B using ZIF-67/rGO/GCE.
In the present paper, the ZIF-67/rGO was synthesized by the microwave-assisted method. The ZIF-67/rGO modified GCE was obtained by drop-casting ZIF-67/rGO suspension on the electrode surface. The electrochemical behaviors of ZIF-67/rGO/GCE were investigated by means of cyclic and    (SEM) using SEM JMS-5300LV (Japan) and by transmission electron microscopy (TEM) using JEOL-2100F microscope. Ramma spectra were obtained by XPLORA, HORIBA with laser excitation wavelength of 532 nm. Fourier transformation Infrared (FT-IR) analyses were recorded on a Shimadzu IR Prestige-21 (Japan). X-ray diffraction (XRD) patterns were recorded using D8 Advance-Bruker (Germany) with CuK α radiation (λ = 0:1514 nm). X-ray photoelectron spectra (XPS) were collected by using a Shimadzu Kratos AXISULTRA DLD spectrometer (Japan). Voltammetric experiments were undertaken at room temperature using a CPA-HH5 Computerized Polarography Analyser (Vietnam) electrochemical workstation. A three-electrode cell configuration, comprising a glassy-carbon working electrode (GCE) with a diameter of 2:8 ± 0:1 mm, the GCE or the modified GCE were used for formatting the modified electrode as a working electrode, an Ag/AgCl/3 M KCl reference electrode, and a platinum wire auxiliary electrode was employed in these experiments. The Schimadzu 2030 HPLC system was used to analyze Rhodamine B. The chromatographic conditions were as follows: UV detector (λ = 275 nm), flow rate of 2.0 mL·min -1 with the injection volume of 10 μL, column temperature of 45°C ± 1°C. The quantitative analysis was performed with external standardisation by measuring the peak areas on the chromatograms.      Journal of Nanomaterials oxide. Ascorbic acid (0.15 g) was introduced slowly to the graphene oxide suspension, and the mixture was stirred for 8 hours at 50°C to reduced GO. The reduced graphene oxide (rGO) was separated by centrifugation and washed several times with ethanol and dried at 80°C in a vacuum oven for 5 h.

Results and Discussion
3.1. Characterization of ZIF-67/rGO Composite. XRD patterns of reduced graphene oxide (rGO), ZIF-67, and ZIF-67/rGO composite were obtained and compared ( Figure 1). Figure 1(a) presents a weak and broad reflection peak at ac. 26°corresponding to the relative short-range order structures in disordered stacked rGO [14], which indicates the successful reduction of graphene oxide. The XRD pattern in Figure 1 The morphology of the obtained materials was observed by TEM and SEM images. The TEM image of rGO shows a stacked and crumpled morphology, revealing deformation because of the exfoliation and restacking process [15] ( Figure 2(a)). The ZIF-67 crystal particles aggregated and were in uniform polyhedral morphology with the size of 140 ± 5:6 nm (counted for 50 particles) (Figure 2(b)). The TEM image of ZIF-67/rGO composite shows that the ZIF-67 particles with an average size of 50-70 nm dispersed well on the rGO sheet ( Figure 2(c)). The formation of composite material in this experimental condition can be explained when considering the possible interactions between the reactants. Upon adding Co 2+ to rGO suspension, there may be an interaction between Co 2+ with oxygencontaining functional group in rGO, although this interaction can be weak. Therefore, as 2-methylimidazole is introduced into the mixture, its bonding to Co 2+ , which is stronger than its competence mentioned above, is formed to generate ZIF-67 as particles of around 10-20 nm dispersing on the rGO matrix as shown in Figure 2(c). The synthesis of ZIF-67 on rGO matrix limits the possibility of ZIF-67 crystals growth. Consequently, its size is significantly smaller compared with that of ZIF-67 synthesized separately as shown in Figure 2(c).
The formation of ZIF-67, rGO, and ZIF-67/rGO composite is also confirmed by the FT-IR spectra ( Figure 3). The position of characteristic absorption bands of ZIF-67 synthesized in this study is in good agreement with those reported previously [16]. The absorption band centered at 3441 cm -1 is assigned to the stretching of OH in water adsorbed in material. The main absorption peaks from 600-1500 cm -1 are assigned to the vibrations of functional groups in 2-methylimidazole, in which the peak at 1578 cm -1 can be attributed to the stretching vibration of C=N bond and the peak at 2924 cm -1 belongs to the symmetric stretching vibration of CH 3 [17]. The absorption band at 1573 cm -1 in the FT-IR spectrum of rGO can be attributed to the skeletal vibration of graphene sheets. The vibration of C-O stretching of epoxy group and alkoxy is observed at broad peak from 1220 cm -1 to 1067 cm -1 due to incomplete reduction of GO [18]. The FT-IR spectrum for ZIF-67/rGO exhibits a combination of infrared peaks of ZIF-67 and rGO indicating that the structures of ZIF-67 and GO are maintained in the composite.
Raman spectroscopy can also be employed to effectively probe the disorder in the carbon structure of the material. In the Raman spectra of rGO and rGO/ZIF-67 (Figure 4), two peaks are clearly visible at 1336 cm -1 and 1602 cm -1 , corresponding to the so-called D and G bands, respectively. It is reported that the D band peak corresponds to the breathing mode of κ-point phonons of A 1g symmetry, and the G band peak represents the first-order scattering of the E 2g phonons [19]. The intensity of the D band is related to the size of the in-plane sp2 domains, and the relative intensity ratio (ID/IG) is a measure of the extent of disorder [19]. An increase in intensity ratio (I D /I G ) and the higher intensity of D band indicate the presence of more isolated graphene domain in ZIF-67/rGO compared to in rGO. It is found that the intensity ratio of I D /I G is 1.02 for rGO and 2.01 for ZIF-67/rGO. The results confirm that the composite consists of rGO and ZIF-67 structures.
The oxidation states of elements were examined by XPS ( Figure 5). The XPS survey spectrum presents the existence of Co, O, N, and C in ZIF-67/rGO composite ( Figure 5(a)). The C1s spectrum of ZIF-67/rGO ( Figure 5(b)) is able to be fitted by three components at 284.59, 285.87, and 288. 16 eV, corresponding to carbon in aromatic rings, epoxy C-O-C, and carboxyl COOH groups, respectively [15,20]. These components indicate that the reduction of graphene  Journal of Nanomaterials oxide is not complete. The N1s XPS spectrum at 400 eV represents a dominant contribution from nitrogen atom in the pyrrolic-N and a shoulder at 398.8 eV, which possibly due to the interaction of N-imidazolate with graphene sheets (Figure 5(c)) [21]. The Co2p spectrum exhibits two main peaks coupled with two satellites, 805.01 eV coupled with 797.7 eV for Co2p1/2 and 785.7 eV coupled with 780.5 eV for Co2p3/2 [22] (Figure 5(d)). The estimation of the oxidation states of Co cations can be done by analyzing the energy gap between the Co2p main peak and the satellite peak. Typically, this energy gap of Co(II) cation is ca. 6.0 eV, while that of Co(III) cation is approximately 9-10 eV [23,24]. In the present case, the energy gap of cobalt cation is 6.27 eV for Co2p1/2 and 5.2 for Co2p3/2. Therefore, Co(II) is the main form in the as-prepared ZIF-67 material. The textural properties of ZIF-67, rGO, and ZIF-67/rGO were examined via nitrogen adsorption-desorption isotherms ( Figure 6). The ZIF-67 material exhibits a typical type-I adsorption isotherm with a high amount of N 2 adsorption indicating high porosity. An increase in adsorption volume at a low relative pressure indicates the existence of micropores in all ZIF-67 samples. The isotherm curves of rGO and ZIF-67/rGO belong to typical type IV according to IUPAC classification. The presence of the hysteresis loop at high relative pressure region indicates the existence of mesopore which can be attributed to the void among the primary particles [25,26]. The specific surface areas of ZIF-67, rGO, and ZIF-67/rGO derived from the BET model are found to be 1330 m 2 .g -1 , 200 m 2 .g -1 , and 498 m 2 .g -1 , respectively. The surface area ZIF-67/rGO sample in this study is higher than that of rGO based materials reported previously [27,28].  (Figures 7(b) and 7(d)), the value of A is found to be 0.070 cm 2 for bare GCE and 0.117 cm 2 for ZIF-67/rGO-GCE indicating that the effective electroactive surface area of the modified electrodes is improved significantly.

Cyclic Voltammetry Behavior of RhB at ZIF-67/rGO/GCE
(1) Cyclic Voltammetry Behavior. Figure 8(a) presents the CVs of RhB at bare GCE, ZIF-67/GCE, rGO/GCE, and ZIF-67/rGO-GCE. It is clear that there is no obvious redox peak on the CVs with the use of bare GCE and ZIF-67/GCE. When rGO/GCE was used, on the scan toward positive potential, the rise of oxidation current at 0.92 V indicating an oxidation process of RhB at the electrode surface. No reduction process can be observed on the reversed scan suggesting that this is an irreversible process. However, the CV behavior on ZIF-67/rGO-GCE assembles that on rGO/GCE, except the fact that this electrode significantly enhances not only the shape of the oxidation peak but also the peak current at 0.92 V. It is found that the peak current magnitude is 2.2 times higher compared with rGO/GCE (Figure 8(b)). The ZIF-67/rGO-GCE exhibits good electroactivity toward the oxidation of RhB probably due to the synergistic effect of high electrical conductivity of rGO [30] and excellent textural properties of ZIF-67, including large surface area, abundant number of adsorptive sites, and large volume of porosity.
(2) Effect of pH. Figure 9 presents the effect of pH on electrochemical responses of CV. At solution pH lower than 5, the oxidation peak cannot be observed (Figure 9(a)). However, the peak becomes more apparent at pH> 6 and the peak current magnitude reaches a maximum value at pH 7 (Figures 9(b) and 9(c)). Hence, pH = 7 was selected for further experiments. Furthermore, upon increasing pH from 6 to 10, the peak potential shifts to more negative value from 0.96 V to 0.74 V indicating that the protons are involved in the RhB oxidation at electrode. The relationship between pH of solution and oxidation peak potential is plotted as in Figure 9(d) illustrating a linear correlation. The value of 0.053 is close to the theoretical Nestian value of 0.0592 (Figure 9(d)) implying that the oxidation reaction of RhB at electrode involves an equal number of electrons and protons.
(3) Effect of Scan Rate. Effect of scan rate on electrochemical signals also provides the information on oxidation mechanism ( Figure 10). Upon increasing scan rates, the cathodic peak current magnitude grows along with the shift of peak potential to more positive value as shown in Figure 10(a) implying a nondiffusion controlled process.
In order to confirm whether the electrooxidation reaction is diffusion controlled or adsorption controlled, the Scheme 2: The proposed mechanism of electrode process for RhB oxidation. 10 Journal of Nanomaterials dependence between Ip and v 1/2 was plotted ( Figure 10(b)).
The linear regression equation of Ip vs v 1/2 is derived as follows: Ip = 0:039 ± 0:008 ð Þ v 1/2 + 0:027 ± 0:0045 ð Þ , r = 0:993 ð3Þ The plot of Ip vs. square root of the scan rate is highly linear (r = 0:993). The value in bracket presents the confidence interval of 95%. With a confidence interval of 95%, the intercept varies from 0.023 to 0.032 which means that the oxidation of RhB at the modified electrode is an adsorption controlled process [14,31].
The dependence of anodic peak potential (Ep) on the scan rate also provides another insight into the mechanism of the process. The number of electrons transferred during the oxidation process can be determined via the relation of peak potential (Ep) and natural logarithm of scan rate expressed by Laviron equation: [32].
where E0 is the formal redox potential, ks is the apparent electron-transfer rate constant, α is the charge transfer coefficient, n is the number of electrons transferred, v is the scan rate (V.s -1 ), and T = 298 K, R = 8:314 J mol K − 1, and F = 96480 C mol − 1.
The value of nð1 − αÞ obtained from the slope of the linear regression line of Eq. (6) is found as 0.99. The value of α is assumed to be 0.5, which is commonly used for a totally irreversible system [33]. Therefore, the number of electrons transferred (n) in the electro-oxidation of RhB is 1.98, which implies that the oxidation mechanisms for RhB can involve two electrons and two protons.
The oxidation mechanism of RhB oxidation on the electrode surface can be illustrated in Scheme 1.
The enhancement of RhB oxidation on ZIF-67/rGO/GCE can be explained by the improvement of RhB absorption on this electrode due to the contribution of various intermolecular interactions between RhB and each individual component of the electrode. The interaction can be electrostatic interaction between ionic RhB with charged ZIF-67/rGO surface, depending on pH, or π-π stacking interactions between 2-methylimidazole ligand of ZIF-67 and the π-electron of benzene ring in RhB moiety, or complexation interaction between Co(II) and carboxyl group (Scheme 2).

Quantitative Determination of RhB Using Differential
Pulse Voltammetry Method (DPV). Differential pulse voltammetry method, which is more sensitive than the CV approach, using a ZIF67/rGO/GCE was investigated to quantify the RhB content in real samples, normally at low concentration. By using a solution containing 21.1 μg/L RhB, the operational parameters were optimized as follows: accumulation potential: -0.2 V; accumulation time: 20s; pulse altitude: 0.06 V; rating potential: 0.008 V ( Figure S1-4).
(1) Interference, Reproducibility, Repeatability, Linear Range and Limit of Detection. Relative standard deviation (RSD) is applied to estimate the effect of some potential interfering species on the detection of RhB in which the RSD = ðIo − IÞ /Io where Io is peak current measured without and with interference. The tolerance limit is defined as the maximum concentration of the interferent that leads to a RSD of over ±5%. This value of some possible inorganic and organic substances in the use of 5.7 μg.L -1 RhB is given in Table 1. For inorganic salts, except for Na 2 SO 4 , tested at 50-fold excess, all other inorganic substances at 100-fold excess show no interference effect on the detection of RhB. Similarly, saccharine (at 100-fold excess), glucose (60-fold excess), sodium benzoate, and saccharine (50-fold excess) do not interfere the signal of RhB.
The reproducibility was investigated with nine individual ZIF-67/rGO-GCEs in the detection of 5.7 μg/L RhB under 11 Journal of Nanomaterials the optimized conditions. It was found that a good reproducibility with the RSD of the current responses of 7.0% can be obtained. The expectable RSD for nine independent electrodes indicates the good reproducibility of the ZIF-67/rGO-GCE in the present work.
The repeatability of ZIF-67/rGO-GCE for DPV was also studied with the solutions of 5.7; 21.1 and 44.1 μg/L RhB. The determination of 5.7, 21.1, and 44.1 μg/L RhB was measured successively for nine times ( Figure S5). The obtained RSD for 5.7, 21.1, and 44.1 μg/L RhB was 6.2, 3.9, and 0.8%, respectively, which are lower than the 1/2 RSD Horwitz predicted [34]. Such reasonable RSD of successive measurements confirm that the ZIF-67/rGO-GCE is able to be repeatedly used for the determination of RhB in either low concentration range or high concentration range.
The LOD, calculated by 3 Sy/b, where Sy is the standard deviation of y-residuals and b is the slope of linearity, is found to be 1.79 μg.L -1 .
The detection performance of the ZIF-67/rGO modified electrode in DPV method was compared with the other methods as shown in Table 2. It could be noticed that the linear range and LOD of RhB at the ZIF-67/rGO-GCE are lower or comparable with those results reported previously. The ZIF-67/rGO modified electrode presents a lower LOD compared to UV-Vis spectrophotoscopy method with solid phase extraction using Sepab eads SP 70 resin, or voltammetric method using silica-pillared zirconium phosphate/nafion composite modified glassy carbon electrode. Overall, the ZIF-67/rGO composite is proved to be an effective electrode modifier for the detection of RhB in aqueous solution.
3.2.4. Real Sample Analysis. In order to assess the application of the proposed method on real samples, two samples of tomato sauce and one of the chili sauce were chosen to use for the spiked experiments in comparison with the results from highperformance liquid chromatography (HPLC). The results are given in Table 3 showing that all samples are free of RhB and the recovery is in the exceptional range of 98-103% with RSD less than 2%. The results obtained by this method are in statistical agreement with those by HPLC method (paired samples t-test, t ð2Þ = 0:99, p = 0:427 > α = 0:05) manifesting that the voltammetric method using ZIF-67/rGO-GCE is satisfactory.

Conclusions
A novel modified electrode for RhB quantifying was developed based on the composite of ZIF-67 and reduced graphene oxide. Owing to synergetic effects from rGO with good electroconductivity properties and ZIF-67 with excellent textural properties, the use of ZIF-67/rGO modified electrode in DPV technique provides the broad linear response range and low LOD of RhB detection. Furthermore, the ZIF-67/rGO-GCE has good stability and reproducibility making the proposed method a prospective application for the determination of RhB in foodstuffs.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that they have no conflicts of interest.