All the chemicals used were of analytical grade. Zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O, 98%) was purchased from Dynamic and copper nitrate trihydrate (Cu(NO₃)₂·3H₂O, 99%) was purchased from Vetec (Sigma-Aldrich). All solutions were prepared with deionized water. Natural Cassava starch in the form of colloidal suspension was used as fuel. Aloe vera leaves were harvested in the São José region of Parana-Brazil. To obtain the extract of Aloe gel, about 200 g of Aloe vera leaves were washed with deionized water and the internal mucilaginous gel was extracted. Afterward, the mucilaginous gel was crushed using a pistil and a ceramic mortar to obtain the complete extract. Finally, the solution was washed, filtered and the resulting Aloe vera gel broth extract was stored under refrigeration (2°C).

Two different routes, both easy to reproduce, were used for synthesizing Zinc oxide nanoparticles. In the route I (green synthesis), adapted from Sangeetha et al. (2011), Aloe vera (AL) gel broth extracts at the concentration (90%) were prepared with distilled water, the volume was made up to 100 ml. Subsequently, zinc nitrate (9.40 g) was dissolved in the aloe extract solution under constant magnetic stirring (120 min.) and left at rest for 12 h. The suspension was calcined in a muffle furnace at temperatures (750 °C) for 1 h. In the route II (gelatinization method): first starch (ST) was extracted from 100 g of natural Cassava starch in 300 ml of distilled water under mechanical stirring for 2 h. It was then sieved, and in the colloidal starch suspension was added 9.40 g of zinc nitrate. After 60 min of mechanical stirring (600 rpm), the suspension was calcined in a muffle furnace at a temperature of 750 °C for 1 h (Primo et al., 2019). The ZnO nanoparticles obtained were named Zn-AL (route I) and Zn-ST (route II).

X-ray powder diffraction profile was performed at the Brazilian Synchrotron Light Laboratory (LNLS, using XRD1 beamline, 12 keV energy, λ = 1.033 Å, 2θ of 0°-80°) (Carvalho et al., 2016). Scanning electron microscopy images were recorded using a JEOL-JSM-7500F Field Emission Scanning Electron Microscope operated 15 kV, the spatial resolution was 2.5 nm. The Raman spectra were recorded using a Micro-Raman system, Senterra Bruker Optik GmbH), λ = 532 nm, laser power 5 mW, time 10 s, resolution 4 cm⁻¹. The optical diffuse reflectance was measured (UV-VIS-NIR Spectrophotometer CARYb5G, Varian) in the range of 300–800 nm. Zeta potential was recorded using ZETASIZER NANO ZS90 (MALVERN), model ZEN 1,010 at 25°C. The zeta potentials of the nanoparticles were determined from their electrophoretic mobilities according to Smoluchowski's equation (O'Brien and Hunter, 1981); the pH of these nanoparticles was adjusted between 3 and 11 using HCl or NaOH solutions. The chemical composition was evaluated by X-ray photoelectron spectroscopy (XPS) (Versaprobe PHI 5,000 from Physical Electronics, equipped with a monochromatic Al Kα X-ray source). The XPS spectra were collected at a take-off angle of 45° with respect to the electron energy analyzer and the spot size was 200 μm. Pass energy (PE) of 20 eV was used for the high-energy resolution spectra (Zn 2p, O 1s, and C 1s). The spectra were analyzed using the CASA-XPS software.

The metal ion solutions were analyzed using a Varian^TM SpectrAA® 220 Flame Atomic Absorption Spectrometer (FAAS). The FAAS was equipped with an air-acetylene burner. The hollow cathode lamp was set at 4 mA and the analytical wavelength was adjusted at 324.8 nm. The slit size was adjusted to 0.2 nm. The standard curve was drawn by using copper standard solutions. After the adsorption, the ZnO nanoparticles were characterized concerning their composition by energy dispersive X-ray spectrometer (EDX) from Shimadzu, model EDX-7000, containing a Rh tube, operating at 50 and 15 W. The crystalline phases were identified by powder X-ray diffraction (XRD) performed on a Bruker model D2 Phaser with Cu Kα radiation (λ = 1.5418 Å), with scan in 2θ from 10° to 90° and step rate of 0.2°/s. The zeta potential was recorded using ZETASIZER NANO ZS90 from MALVERN, model ZEN 1010. The electronic spectra of the powdered pigments samples were measured on the range of 400–900 nm with a UV-Vis Ocean Optics spectrophotometer model USB-2000.

To investigate the efficiency of the ZnO nanoparticles as adsorbents for the removal of copper metal ions from water, an adsorption test was performed. The parameters: contact time; initial pH and initial metal ion concentration were investigated. The adsorption experiments were carried out in conical flasks containing 25 mL of copper solution with an initial concentration ranging from 40 to 120 mg L⁻¹. To this end, 250 mg of the ZnO particles were added, and the solutions were kept under continuous shaking for 240 min in a heating bath at 25°C. To study the adsorption kinetics and the pH parameters, 50 mg L⁻¹ of a solution containing Cu (II) and the same amount of ZnO particles was prepared; its pH was adjusted using 0.1 HCl and 0.1 NaOH solutions. The resulting solutions were centrifuged at 1,200 rpm for 15 min. The ion concentration measurements were performed before the adsorption test without the presence of the adsorbent and after 4 h of adsorption in a flame atomic absorption spectrometer (FAAS).

The amount of Cu²⁺ ion adsorbed at the end of the adsorption experiment and the ion percentage removal (%) by the ZnO nanoparticles were calculated applying Equations (1, 2), respectively:

where q is the amount of ion adsorbed by the adsorbent in mg g⁻¹, C_o is the initial ion concentration in contact with the adsorbent (mg.L⁻¹), C_f is the ion concentration (mg.L⁻¹) after the batch adsorption process, m (g) is the mass of adsorbent and V (L) is the volume of ion solution.

To check the stability of the nanoparticles a method adapted from (Rafiq et al., 2014) was used. Thus, 50 mL of simulated sample was treated separately with 250 mg of ZnO synthesized. The initial pH of the experiment was 4 or 6 and the contents were allowed to remain in contact for 240 min while maintaining the temperature at 25°C. After centrifugation and filtration, the residue was washed with deionized water followed by oven drying at 60°C.

Figure 1A presents the X-ray diffractograms of the zinc oxides nanoparticles obtained after heat treatment at 750°C for 60 min. The crystalline phase is identified by the presence of the characteristic peaks of the Wurtzite ZnO phase (Kisi and Elcombe, 1989), belonging to the compact hexagonal system with a space group P63mc to the crystallographic chart [JCPDS, #PDF01-070-8070]. Additional peaks were not detected, evidencing that the single-phase ZnO was successfully obtained regardless of the synthesis route used and the precursors were completely decomposed. The XRD patterns allowed to determine the average crystallite size of the ZnO nanoparticles, estimated by Scherrer's equation {D = 0.9λ/(B cosθ)} (Hedayati et al., 2015), with the average size of 43.3 nm for Zn-AL and 44.9 nm for Zn-ST. According to these results, the crystalline size is affected by the polysaccharide used in the synthesis, at the same calcination temperature.

Figure 1B shows the Raman spectra of the samples indicating the characteristic wurtzite phase peaks, corroborating with the XRD patterns. The predominant bands are at 99 cm⁻¹ (mode E₂low) and 437 cm⁻¹ (mode E₂high). The E₂low mode is attributed to the vibrations of zinc sublattice in ZnO and E₂high mode is assigned to the oxygen vibration (Cuscó et al., 2007; Stanković et al., 2012), the strong E₂high mode indicates the high crystallinity of the oxide (Jothilakshmi et al., 2009), the same vibrational mode has been identified for the zinc oxide nanoparticles obtained via the Starch-assisted synthetic route, reported by Carp et al. (2015). The bands at 380 and 408 cm⁻¹ correspond to the first-order optical modes A₁(TO) and E₁(TO), bands at 202 and 330 cm⁻¹ are characteristic of second-order modes 2E₂low and E₂high—E₂low, caused by multiphonon processes. The bands located at 573 and 584 cm-1 are assigned to A1(LO) and E1(LO) modes, these bands are associated to the presence of structural defects in the ZnO structure, being the E₁(LO) mode strongly affected (Cuscó et al., 2007).

Figures 2A,B shows the SEM images for Zn-AL, which consists of pseudo-spheres, and non-uniform hexagonal particles. For Zn-ST, uniform spherical particles are formed (Figures 2C,D). The two samples show particle aggregation, related to the self-assembly effect (Khorsand Zak et al., 2013). The Zn-AL particles tend to agglomerate in plaques (Figure 2A), this was attributed to the Aloe vera gel acting as a sacrifice complexant in the formation of the ZnO nanoparticles during the combustion (Kavyashree et al., 2015). The two synthesis routes (Aloe vera and Cassava starch) have polysaccharides as fuel for the formation of ZnO nanoparticles; their formation mechanism can be described by the “egg-box” model (Kavyashree et al., 2015). Therefore, their difference in morphology can be associated with the complex polymeric network of each polysaccharide. Aloe vera gel consists of a combination of organic chains, such as soluble polysaccharides, monosaccharides, proteins, amino acids, among others (Choi and Chung, 2003). The colloidal suspension of Cassava starch is more homogeneous and less complex because it consists basically of amylopectin and amylose leading to the formation of uniform particles, since the Zn (II) ions occupy the “egg-box” more efficiently, with more regular distance. The Aloe vera gel presents a greater variation in its natural components than Cassava starch, affecting directly the shape and reproducibility of ZnO nanoparticles.

The chemical environment of the zinc and oxygen atoms were analyzed using X-ray photoelectron spectroscopy (XPS). The O 1s and Zn 2p XPS core-level spectra are shown in Figure 3. The binding energy of the XPS data was calibrated using the C 1s peak at 284.6 eV (Das et al., 2010). The O 1s spectrum was fitted with three components centered on 530.2 ± 0.1, 531.4 ± 0.6, and 532.3 ± 0.7 eV, for both samples (Figure 3A). The low binding energy component located at 530.2 ± 0.1 eV is attributed to O²⁻ ions participating in the Zn-O bond in the wurtzite structure of the hexagonal Zn²⁺ ions of ZnO (Chen et al., 2000; Al-Gaashani et al., 2013). The component centered at 531.4 ± 0.6 is associated with photoelectrons emitted from O²⁻ ions in oxygen-deficient regions in the matrix of ZnO (Chen et al., 2000). The high binding energy component located at 532.7 ± 0.7 is reported to be associated with oxygen species adsorbed on the surface of the ZnO, such a -CO₃, H₂O, or O₂ (Sangeetha et al., 2011; Visinescu et al., 2011). The Zn 2p high-resolution XPS spectra show the 2p doublet (Figure 3B) with components centered at 1020.6 eV (Zn 2p_3/2) and 1043.5 eV (Zn 2p_1/2). For both samples, the binding energy difference between these core levels is 23.0 eV, reference value denoting the presence of zinc in Zn²⁺ oxidation state (Chen et al., 2000; Das et al., 2010), the chemical state is confirmed by the Zn LMM Auger data (Figure 3C).

Figure 4 shows the optical characterization of the ZnO nanoparticles synthesized using Aloe vera (Route I) and Cassava starch (Route II). It can be observed in Figure 4A an increase in the reflectance at wavelengths larger than 380 nm, this can be attributed to the direct band-gap of ZnO due to the electron transitions from the valence band to the conduction band (O_2p Zn_3d) (Kavyashree et al., 2015), with a lower percentage of reflectance for Zn-AL (~65%). The band energy gaps (E_GAP) of the samples were calculated using the Kubelka-Munk method (Cuscó et al., 2007), the E_GAP were determined by linear extrapolation of the curve [F(R) x E]2 vs. energy (E) in (Figure 4B), with values: 3.24 eV (Zn-ST) and 3.18 eV (Zn-AL), similar values have been reported in (Khorsand Zak et al., 2013; Carp et al., 2015) for zinc oxides obtained with Starch. The variation in the optical gap of the ZnO nanoparticles can be associated with a variation in the average particle size and morphology. The synthesized ZnO nanoparticles exhibit a slight red shift in the absorption edge (Figure 4A), this increase in the response range toward the visible radiation region can be explored in the future as photocatalysts with visible light activity (Stanković et al., 2012).

After the adsorption assay, the samples at pH 4 and 6 were dried at 60 °C in an oven-dry for 12 h and characterized. The chemical composition data (EDXRF) evince the incorporation of the Cu (II) ions in both Zn-AL and Zn-ST nanoparticles following their surface adsorption (Table 5).

Figure 12 shows the XRD patterns recorded on the ZnO nanoparticles before and after the Cu (II) metal ions removal. It can be seen that the diffraction peaks recorded on the samples after the adsorption correspond to the majority of peaks of the ZnO hexagonal Wurtzite crystal phase (XRD recorded on the Zn-AL and Zn-ST samples before adsorption). However, in the XRD patterns recorded after adsorption, the CuO phase can be observed in both routes (I and II) samples and for different pHs. The diffraction peaks of CuO were indexed to the monoclinic Tenorite crystal phase of CuO (JCPDS, #PDF 96-901-6327). Moreover, similar to the CuO Bragg peaks the ZnO peaks slightly shift to lower diffraction angles compared to that of the ZnO recorded before adsorption, indicating the substitution of Zn²⁺ by Cu²⁺ ions in the crystal lattice (Mukhtar et al., 2012). According to (Wang et al., 2010), the hydrated Cu (II) or Cu(H2O)62+ can react with the OH⁻ groups and form Cu-O weak bounds through a Lewis interaction, in addition, the adsorbed copper ions can partially hydrolyze leading to the formation of Cu-OH and, consequently, the formation of Cu-O-Cu on the surface of ZnO, thus denoting the Tenorite phase of CuO formed on the surface of ZnO. The crystallite sizes after the adsorption were calculated using Scherrer's equation (Hedayati et al., 2015), the obtained value were: 16.31 nm (Zn-AL, pH 4), and 36.22 nm (Zn-AL, pH 6); and 20.18 nm (Zn-ST, pH 4), and 31.89 nm (Zn-ST, pH 6).

Figure 13 shows the UV-Vis absorbance spectra of ZnO samples before and after Cu (II) adsorption. A broad peak in the visible region centered at 720 nm can be observed after the adsorption of copper. This peak can be assigned to the Cu (II) d-d transition (Li et al., 2013), indicating, the adsorption of Cu (II) ions by the ZnO nanoparticles (Zn-ST and Zn-AL), verifying the XRD results.