Structural reorganization of CuO/ Cu2[Fe(CN)6] nanocomposite: characterization and electrocatalytic effect for the hydrogen peroxide reduction

We report the study on the formation of the Cu2[Fe(CN)6] nanocomposite, which was obtained from copper oxide nanoparticles (CuO NPs) and Prussian Blue precursors. UV-vis analysis indicated that Cu2+ ions are released from CuO NPs, while Fe3+ ions are adsorbed onto the structure of CuO due to a sharp increase in zeta potential (from -30 to 0 mV) after the formation of the Cu2[Fe(CN)6]. Moreover, energy dispersive spectroscopy confi rmed that Fe3+ ions are trapped in the CuO NPs structure. The CuO/ Cu2[Fe(CN)6] nanocomposite exhibited the monoclinic and face-centered cubic phases that correspond to the CuO and Cu2[Fe(CN)6] components. Cyclic voltammetry (CV) for the Nanocomposite modifi ed electrode revealed two well-defi ned redox couples at -0.073 ((E1/2)1) and 0.665 mV ((E1/2)2), attributed to the conversion of Cu2+ to Cu+ and CuFe2+ CuFe3+ pairs, respectively, which is similar to those in the CuO and Cu2[Fe(CN)6] components. Furthermore, the catalytic activity of the nanocomposite towards hydrogen was investigated through CV, where the reduction of H2O2 led to increased currents for the electrochemical process associated with the fi rst redox pair. In contrast, for isolated materials (CuO NPs and Cu2[Fe(CN)6]), there was no signifi cant increase in the current associated with either redox pair.


INTRODUCTION
Nanocomposites can be defi ned as multiphase solid materials in which at least one of the phases is present in the nanometer range (Jelinkova et al. , Rawtani et al. 2019. The use of nanocomposites can impart structural properties and performance capabilities that are signifi cantly different from those of the original components. Several nanomaterials such as magnetite (Santos et al. 2016 Mn, Co, Ni or Fe (Bie et al. 2018). In particular, the most promising PBA compounds include copper hexacyanoferrate nanoparticles, which impart high adsorption capacity (Tao et al. 2019), and undergo electrochemical processes in different media (Baioni et al. 2008). These compounds have the formula M x Cu y [Fe(CN) 6 ].nH 2 O, where copper is usually present in the divalent cationic form, while iron is present in both oxidation states, (II) and (III). In addition to these species, the structure may contain alkali metal cations (M + ) and hydrating water molecules. PB and its derivatives are widely used in electrochemical studies, mainly because of the stability and characteristic redox reactions of the [M 2+ -CN-Fe 3+ ] fragment, which is able to mediate and catalyze electrochemical reactions of peroxide (Chu et al. 2017). However, the use of PB in redox processes is limited mainly to the use of electrolytes with a small hydrodynamic radius (up to 0.1 nm), which are capable of occupying the interstices of its structure, ensuring electroneutrality and enabling electron transfer (Ricci 2005 (Nerush et al. 2019). In this context, devices that measure the concentration of H 2 O 2 in various media are of utmost importance. We report below a novel CuO/Cu 2 [Fe(CN) 6 ] nanocomposite that takes advantage of the assembly process of Cu 2 [Fe(CN) 6 ] and CuO NPs. The formation, structural organization, electronic properties, and electrochemical performance of the nanocomposite are investigated using zetapotential data, ultraviolet−visible (UV−vis) and Fourier-transform infrared (FTIR) spectroscopy, X-ray diffraction (XRD), transmission electron microscopy (TEM), scanning electronic microscopy (SEM), energy dispersive spectroscopy (EDS), and CV measurements.

Chemicals and materials
All reagents were of analytical grade and were used without further purification. In the synthesis of PB, potassium hexacyanoferrate(II) trihydrate PA (K 4 [Fe(CN) 6 ].3H 2 O, Isofarma), iron (III) chloride hexahydrate 99% (FeCl 3 .6H 2 O, Sigma Aldrich), and citric acid PA (C 6 H 8 O 7 , Dinâmica) were used. Copper (II) sulfate pentahydrate PA (CuSO 4 .5H 2 O) and sodium hydroxide 99% (NaOH), purchased from Dinâmica and Sigma-Aldrich, respectively, were used to prepare copper oxide nanoparticles. These reagents were used to synthesize the Cu 2 [Fe(CN) 6 ] complex and the physical mixture containing CuO and PB. The acetate buffer was prepared by titrating acetic acid (CH 3 COOH, Dinâmica) into a potassium hydroxide ³ 85% solution (KOH, Sigma-Aldrich). Hydrogen peroxide (H 2 O 2 , 3%) was purchased from Rioquímica. The water used in the experiments was purified by the Purelab Option-Q (Elga) system with a resistance equal to 18.2 MΩ cm.

Synthesis of CuO nanoparticles
The CuO NPs were prepared using the synthetic strategy described by Kamila & Venugopal (2017), with minor changes. In this procedure, 20 mL of 0.1 mol L -1 CuSO 4 .5H 2 O solution (2.0 × 10 -3 mol) was prepared and added slowly to a reaction flask containing 20 mL of 0.2 mol L -1 NaOH solution (4.0 × 10 -3 mol). The product was centrifuged three times for 5 min at 3500 rpm, dried in an oven at 60°C for 2 h, and calcined at 400°C for 4 h.

Characterization
UV−vis spectra of solutions and dispersions of the precursors and reaction products were recorded with an Agilent Technologies Cary 60 UV-Vis spectrophotometer using a quartz cuvette with 1 cm optical path, in the wavelength range 200 to 900 nm. Infrared spectra between 400 and 4000 cm -1 were obtained with a PerkinElmer Spectrum 100 FTIR spectrometer, using pellets composed of 99 mg of KBr and 1 mg of the sample material. The spectra were normalized with maximum and minimum values corresponding to 1.0 and 0.0, respectively. Zeta potentials were used to verify the charges on the nanoparticle surfaces. For these investigations, CuO NPs, Cu 2 [Fe(CN) 6 ], and nanocomposite dispersions of 2.0 ×10 -2 g L -1 of each component were prepared. Separately, a 1 mL aliquot of the heated dispersion containing CuO NPs and Fe 3+ ions, from the synthesis step described in experimental section, was diluted to a concentration of 2.0 ×10 -2 g L -1 CuO. Analyses were performed using Horiba SZ-100 nano-sizer equipment. TEM images were used to investigate the morphology and particle size distribution of the nanoparticles, using a JEOL JEM 2100 microscope (LAB6 filament), with an accelerating voltage of 200 kV. The TEM images were obtained digitally by use of the Image-Pro Plus software package. The crystallinity and polymorphic forms were analyzed by XRD on a Panalytical Empyrean diffractometer with a CoK radiation source, using a 2θ scanning range from 10° to 70°. Finally, the precursor (CuO/Fe 3+ ) was characterized using a Field Emission Scanning Electron Microscope (FESEM) (QUANTA 250 FEI, FEI Company, Eindhoven, The Netherlands), coupled with elemental analysis by EDS (EDAX Apollo X, FEI Company, Eindhoven, The Netherlands). The CV procedures were performed with the potentiostat/galvanostat Autolab PGSTAT128N equipment, coupled to a 3 mL electrolytic cell comprising a 3 mm diameter glassy carbon electrode, a 0.35 cm 2 area platinum counter electrode, and a saturated calomel electrode (SCE) as reference. An acetic acid/acetate buffer with pH 4.1 was used as a support electrolyte, and N 2 gas was bubbled for ten minutes before the analysis. The potential window employed was -0.6 to 1.2 V. To obtain cast films, dispersions (6 mg mL -1 ) of CuO and Cu 2 [Fe(CN) 6 ] were prepared, and 10 µL of each solution was transferred by the drop coating method to the glassy carbon electrode (GCE). The voltammetry experiment on the nanocomposite was performed after collecting 20 µL of the dispersion and drying the cast films at 40°C for approximately 20 min.
Catalytic activity of CuO, Cu 2 [Fe(CN) 6 ], and CuO/ Cu 2 [Fe(CN) 6 ] modified electrodes The electrocatalytic activity of carbon-modified electrodes of CuO, Cu 2 [Fe(CN) 6 ], and CuO/ Cu 2 [Fe(CN) 6 ] was investigated by CV upon addition of hydrogen peroxide at concentrations ranging from 3.0 × 10 -4 to 3.0 × 10 -3 mol L -1 , with an acetate buffer of pH 4.1. We used a scan rate of 25 mV s -1 at room temperature (25°C).  (2015), the mechanism for PB/ cyclodextrin polymer nanocomposite formation occurs in two steps. First, the Fe 2+ species binds to the OHsites of the polymer chain, and after this, the Fe 2+ species coordinates to the cyanide ligand of the [Fe III (CN) 6 ] 3ion. To explain the formation of the nanocomposite, we conducted a preliminary study. Before adding the [Fe(CN) 6 ] 4species, the zeta potential for the isolated CuO NPs was measured to be -30 mV. However, after the addition of the Fe 3+ species to the mixture, the measured zeta potential increased to 0, suggesting that the Fe 3+ ions were adsorbed on the surface of the CuO NPs. The UV-Vis spectrum of the mixture containing CuO NPs and Fe 3+ ions ( Figure S1 ions from the crystalline structure (into the mixture), which then react with [Fe(CN) 6 ] 4to form Cu 2 [Fe(CN) 6 ]. In another test, potassium thiocyanate (KSCN) was added to a dispersion containing CuO and Fe 3+ species, and heated (at 60°C), in order to investigate the possibility of obtaining the [Fe(SCN)] 2+ complex (K f = 2.0 × 10 6 ). In aqueous solution, this complex is red in color and exhibits a characteristic absorption band at 460 nm in the UV-Vis spectrum ( Figure  S1b). When drops of a 0.1 mol L -1 solution of KSCN were added, a greenish-colored complex was formed, which exhibited an absorption band at 338 nm, characteristic of the Cu(SCN) + ion (K f = 5.6 × 10 3 ). With this experiment, we have clearly demonstrated the inability of Fe 3+ ions to react with [Fe(CN) 6 ] 4-, owing to their attraction toward CuO NPs. Surprisingly, the Cu 2+ ions released by CuO NPs are able to react in solution with ferrocyanide ions, producing the complex Cu 2 [Fe(CN) 6 ].

RESULTS AND DISCUSSION
To reinforce the justification, the elemental composition and morphology resulting from this interaction (CuO + Fe 3+ ) of the solid phase were recorded by SEM ( Figure S2) and EDS ( Figure  S3). The SEM analysis indicates ( Figure S2) the formation of polydisperse particles having cubic shapes. As expected, large amounts of Cu and O from the CuO NPs, and a significant iron fraction as well as some impurities, were observed, as shown in (Table SI-Supplementary Material). The results show that the copper atoms (14.6%) are in the same molar ratio with the iron atoms (16.9%) in the solid phase. The effect of the strong interaction between the CuO NPs and Fe 3+ ions can be rationalized according to Pearson's acid-base theory (Pearson 1963). According to the theory, the O 2anion (which is a hard base) has a greater tendency to bind with hard acids (which are typically highly charged cations with low polarizing power). Analyzing the nature of the Fe 3+ and Cu 2+ ions, these cations are classified as hard and borderline acids, respectively. Thus, it is possible to release copper ions in exchange for iron atoms in the crystalline structure of CuO, with the adsorption of Fe 3+ ions, based on a Pearson acid-base reaction (Pearson 1963). Similarly, the ionic radii of the Cu 2+ and Fe 3+ species are 73 and 64 pm, respectively, with a radius ratio of 1.14 for Cu 2+ /Fe 3+ ; this is very close to unity, a value that would also facilitate confinement of the ferric ions in the CuO unit cell. Finally, the zeta potentials for the Cu 2 [Fe(CN) 6 ] complex and the nanocomposite were -29 mV and +28 mV, respectively, indicating that this structure has stoichiometric defects due to the reaction conditions. It is important to note that the synthesis of the CuO/Cu 2 [Fe(CN) 6 ] nanocomposite was carefully investigated to confirm its reproducibility.

Spectroscopic studies for the formation of CuO/Cu 2 [Fe(CN) 6 ] nanocomposite
In order to investigate the unexpected formation of the CuO/Cu 2 [Fe(CN) 6 ] nanocomposite, we carried out the following spectroscopic studies. The relative absorbance spectra for both the precursor compounds, as well as the nanocomposite, are shown in Figure 1 band gap corresponding to the energy difference between the Cu 3d and O 2p orbitals (Absike et al. 2019). It is interesting to note that the CuO/ Fe 4 [Fe(CN) 6 ] 3 nanocomposite was the expected product, because ferric and ferrocyanide ions were available in the solution to form PB (with a characteristic absorption band at 690 nm in the UV-Vis spectrum). However, the spectrum in curve c clearly indicates that the formation of CuO/[Fe(CN) 6 ] occurs without formation of PB.
FTIR spectra (Figure 2) & Erasmus 2018). Medium intensity bands at 1605 and 1612 cm -1 were assigned to the O-H deformation modes. Additional bands observed at 595 cm -1 and 500 cm -1 were assigned to Fe-C stretching and Fe-C deformation modes, respectively. The FTIR spectrum for the CuO species (Figure 2a) was found to exhibit three main bands at 1114, 600 and 503 cm -1 , which were assigned to δ Cu-O, ν sym Cu-O, and ν asym Cu-O, respectively, where ν and δ denote stretching and angular deformation modes, respectively (Arun et al. 2015, Zhang et al. 2008). The main stretching and deformation bands seen in the FTIR spectra of the PBA and CuO isolated materials were also observed in the FTIR spectrum of the nanocomposite, indicating that the title nanocomposite was formed by addition of Cu 2 [Fe(CN) 6 ] and CuO components. Table I represents our attempt to assign the various A reduction in crystallinity in the composite was observed from the widening of the highest intensity peak, as estimated by calculating the ratio of width to half peak height of the peak at 39.1° for CuO NPs and for the nanocomposite, indicating a reduction of 55% in the signal corresponding to the crystallographic plane (111) for the CuO component in the nanocomposite. In addition, a decrease in the composite crystallinity was observed using the peak intensity of the crystallographic plane (220) at 17.9 ° for Cu 2 [Fe(CN) 6 ], with a relative reduction of 90% for this signal. Additionally, the broadening Table I of the signals in the composite is attributed to the adsorption and subsequent entrapment of Fe 3 + ions in the CuO NPs unit cell, as verified by Li et al. (2010). Finally, a high intensity peak was observed at 28.6° in the nanocomposite diffractogram, which can probably be associated with KCl formed during the synthesis. There were no signals corresponding to the iron oxide or hydroxide species. The size, shape, and organized entities at the molecular level can be analyzed by TEM. The TEM image for CuO NPs (Figure 3a) exhibited rodlike particles of length 20-30 nm, as reported by Nakhaeepour et al. (2019). These small particles form a supramolecular aggregate with a size range of 100 nm to 1 µm, making it difficult to construct a histogram in order to estimate the particle size distribution. The aggregates formed can be attributed to the absence of surfactants that could stabilize the system (Silva et al. 2013) and to nanoparticles associated with a high surface charge (zeta potential ~ 30 mV), as well as to solvent evaporation in the preparation of the sample for TEM analysis (Lopes 2018). Figure 3b shows that the Cu 2 [Fe(CN) 6 ] complex exhibited well-defined spherical shapes with a polydisperse distribution, as observed by Kim et al. (2017). A histogram of the distribution of these particle sizes (shown in Figure S5a) indicated the presence of particles with a mean diameter of ca. 3.8 nm (using data from 100 nanoparticles). These nanoparticles exhibit small sizes because of their high surface charges (zeta potential -29 mV). In the TEM images for the nanocomposite (Figure 3c), both rod-like structures and spherical shapes, corresponding respectively to the CuO and Cu 2 [Fe(CN) 6 ] species, were observed. A histogram ( Figure S5b) of the particle size distribution (using data from 100 nanoparticles) indicated a mean diameter of 5.81 nm. The decrease in size observed for the CuO NPs could be due to interaction with Fe 3+ species during the synthesis of the Cu 2 [Fe(CN) 6 ] complex in the nanocomposite. The reduced size observed for the CuO NPs is in good agreement with the results reported by Bhuvaneshwari et al. (2016) and Nithya et al. (2019), for CuO NPs after interacting with Cr 3+ and Ni 2 + ions, respectively. On the other hand, the Cu 2 [Fe(CN) 6 ] nanoparticles obtained in the nanocomposite showed a significant increase in size (from 3.3 nm in the starting material to 5.81 nm in the composite) but maintained the same spherical shape. It has been proposed that the increase in size of these structures could be due to the higher temperature used in nanocomposite synthesis, which favors the fusion of nanoparticles (Qu et al. 2006, Azam et al. 2012, Xu et al. 2012. Electrochemical studies of CuO, Cu2[Fe(CN)6], AND CuO/Cu2[Fe(CN)6] modified electrodes CV was employed to study the electrochemical properties of CuO, Cu 2 [Fe(CN) 6 ], and CuO/ Cu 2 [Fe(CN) 6 ] modified electrodes between -0.6 and 1.2 V (vs SCE) in a 0.1 mol L −1 acetate buffer with a pH of 4.1, using a glassy carbon working electrode. The voltammogram for CuO NPs (Figure 4a) exhibited a quasi-reversible redox pair, with E 1/2 centered at -0.078 V and a peakto-peak separation (∆Ep) of 0.123 V at a scan rate of 25 mV s −1 , which was assigned to Cu 2+ /Cu + conversion (Ba et al. 2016). The electron transfer at the CuO-modified electrode was evaluated after obtaining voltammograms at different scan rates (10 to 500 mV s -1 ). It was observed that an increase in the scan rate caused a linear increase in the faradaic currents of the anodic and cathodic peaks, suggesting that electron transfer governs the electrochemical process. The linear correlation between faradaic current and v 1/2 with correlation coefficient values of 0.9864 and 0.9851 (Figure 4b) demonstrates that the CuO electron transfer is limited by a diffusion process. As illustrated in Figure 4c, Cu 2 [Fe(CN) 6 ] showed a redox-reversible pair with E 1/2 centered at 0.646 V and a peak-to-peak separation (∆Ep) of 0.143 V at 25 mV s -1 , which is characteristic of the CuFe 2+ /CuFe 3+ pair (Wang et al. 2011, Zheng  (Figure 4d) was observed to be linear at scan rates up to 500 mV s -1 , indicating a diffusion-controlled process. The voltammogram of the CuO/Cu 2 [Fe(CN) 6 ] nanocomposite ( Figure 4e) exhibited two welldefined redox couples at -0.073 ((E 1/2 ) 1 ) and 0.665 mV ((E 1/2 ) 2 ), which are attributed to the conversion of Cu 2+ to Cu + and CuFe 2+ to CuFe 3+ pairs, respectively, which are similar values to those for the CuO and Cu 2 [Fe(CN) 6 ] precursors. For these electrochemical processes, the peakto-peak separations were 0.088 and 0.150 V, respectively, at 25 mV s -1 . By comparing CuO and Cu 2 [Fe(CN) 6 ], we observed an increase in the reversibility for Cu 2+ /Cu + and CuFe 2+ /CuFe 3+ pairs. As observed for the CuO NPs and Cu 2 [Fe (CN) 6 ] isolated structures, the nanocomposite exhibited a linear profile in the current versus v 1/2 plot (Figure 4f), indicating a predominantly diffusion-controlled mechanism.
The cyclic voltammograms shown in Figure  5a indicate the well-known electrochemical processes at -0.092 ((E 1/2 ) 1 ) and 0.644 mV ((E 1/2 ) 2 ) (ascribed to the Cu 2+ /Cu + and CuFe 2+ / CuFe 3+ couples, respectively) for the CuO/ Cu 2 [Fe(CN) 6 ]-modified electrode, in the presence of H 2 O 2 at concentrations between 3.0 × 10 -4 and 3.0 × 10 -3 mol L -1 . The electrode exhibited an electrocatalytic effect toward H 2 O 2 , with a linear increase in the faradaic current of reduction after H 2 O 2 addition. In contrast, for the CuO NPs ( Figure 5b) and Cu 2 [Fe(CN) 6 ] (Figure 5c) electrodes, which were used as experimental controls, there was a decrease in the faradaic currents with the addition of peroxide, possibly because there are no interaction sites with H 2 O 2 on the active surface of these electrodes. This is attributed to reactions of the CuO NPs and Cu 2 [Fe(CN) 6 ] complexes with the peroxide, which would lead to a decrease in the number of electroactive species. Adverse effects are observed in the redox (Cu 2+ /Cu + ) processes of CuO NP structures. Studies by Kamyabi et al. (2017) indicated that the oxidation process is favored with increasing H 2 O 2 concentration, while Gao & Liu (2015b) observed that increased faradaic currents for oxidation and reduction are dependent on peroxide concentration. We observed a substantial increase in reduction processes (Figure 5c), as described by Song et al. (2010). The catalytic effect of peroxide reduction promoted a higher current intensity of the Cu 2+ / Cu + pair from the nanocomposite. The significant catalytic effect exhibited by the nanocomposite can be attributed to the decreased crystallinity of this structure, as confirmed by XRD studies

CONCLUSIONS
In the present study, we propose a mechanism for the formation of the CuO/Cu 2 [Fe(CN) 6 ] nanocomposite based on the adsorption of Fe 3 + ions in the CuO structure, with subsequent release of Cu 2+ ions, as observed by EDS, UV-Vis spectra, and zeta potential analysis. Different synthetic routes using the Fe 3+ and [Fe(CN) 6 ] 2precursors led to the formation of PB, but under the reported experimental conditions, the formation of the CuO/Cu 2 [Fe(CN) 6 ] nanocomposite was favored, probably due to the high affinity of the Cu 2+ species for the cyanide ligand. The nanocomposite showed two well-defined Cu + /Cu 2+ and CuFe 2+ to CuFe 3+ redox couples in the CV. Additionally, a catalytic effect toward H 2 O 2 reduction was observed with increased peak currents for the Cu + /Cu 2+ redox process, in contrast to the CuO and Cu 2 [Fe(CN) 6 ] precursors, in which only a slightly enhanced effect for oxidation was observed.