Effect of the Precursor on the Synthesis of ZnO and Its Photocatalytic Activity

Zinc nitrate (ZnON) and zinc acetate (ZnOA) were used as precursors for the synthesis of zinc oxide (ZnO) nanoparticles by the sol–gel method. The ZnO powder was characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), UV–vis diffuse reflectance spectroscopy, X-ray diffraction (UV–Vis DRS), Fourier transform infrared spectroscopy (FTIR), physisorption of nitrogen, and X-ray photoelectron spectroscopy (XPS). On the other hand, the photocatalytic activity of the samples was tested in the degradation of 2,4-Dichlorophenoxyacetic acid (2,4-D) and 2,4-Dichlorophenol (2,4-DCP) under UV-light irradiation. The ZnON and ZnOA showed polycrystalline irregular structures and rod-like morphology with mean sizes of 40 and 99 nm, respectively. The precursor type influenced the bandgap, crystallite size, surface area, total pore volume, and pore diameter. The XPS results showed high contents of C and N in the ZnO samples, and as a consequence, the solids present remarkable differences in the C/N, O/C, and O/Zn atomic ratios, which significantly influenced the physicochemical characteristics. The ZnON and ZnOA exhibit photocatalytic properties against 2,4-D (74.7 and 90.9%, respectively) and 2,4-DCP (78.4 and 86.7%, respectively) and better performance of ZnOA. These results are promising and indicate the potential to use this material as a photocatalyst to degrade organic pesticides.


Introduction
The phenolic compounds are hazardous and highly stable contaminants. Chlorophenols belong to a group of organic compounds containing a benzene ring, -OH group, and chlorine atoms. Within this group, 2,4-DCP and 2,4D stand out as compounds used in different fields [1,2]. The 2,4-DCP and 2,4-D are pesticides widely used in the agro-industry to deal with certain pests which are a threat to agriculture. Although the Federal Drugs Administration (FDA) and World Health Organization (WHO) have banned these products in Mexico, they are still used [3][4][5]. When the concentration of 2,4 D and 2,4-DCP is significant, these compounds can reach the aquifers and lead to intestinal diseases, and their accumulation in humans can cause carcinogenesis after water consumption [6]; the presence of these compounds in water has been reported [7,8]. Furthermore, the Advance

Scanning Electron Microscopy (SEM) Analysis
The micrographs of ZnO nanoparticles are shown in Figure 1. The ZnO N (Figure 1a) showed that nanorods are shaped with agglomerated irregular nanoparticles of diverse size similar to those reported in the literature [33,34]. On the other hand, ZnO A (Figure 1b) exhibited an amorphous rod-like form [35]. The morphology of the nano-ZnO differs significantly depending on the precursor used [36,37]. The addition of NH 4 OH during the synthesis process could affect the nucleation and growth of the ZnO, thereby influencing the crystal formation and morphological structure, which is associated with a supersaturation that affects the nuclei formation [38].

Transmision Electron Microscopy (TEM) Studies
The morphology (TEM), the selected area electron diffraction (SAED) studies, and the mean size particles of ZnO N  (random spherical with an average particle size of 50 nm) [39] and zinc acetate (hexagonal geometric form with a size between 25 and 50 nm) [40] as precursors. Additionally, the SAED imagens obtained for ZnO N and ZnO A (Figure 2b,e) showed ring patterns (intensity and brightness) indicating polycrystalline nanomaterials [25,41]. It has been reported that the crystallinity of ZnO nanoparticles depends on chemical precursors, synthesis method and conditions (pH, solvent, and temperature), and annealing conditions [42].

Transmision Electron Microscopy (TEM) Studies
The morphology (TEM), the selected area electron diffraction (SAED) studies, and the mean size particles of ZnON and ZnOA are shown in Figure 2. The corresponding TEM (Figure 2a,d) image shows that materials exhibited a near-spherical shape for ZnON (Figure 2a), whereas ZnOA showed an elongated rod-like morphology (Figure 2d) with mean particle sizes of 99.28 and 40.94 nm (Figure 2c,f), respectively. Comparable shapes and sizes were previously reported when ZnO nanoparticles were synthesized using zinc nitrate (random spherical with an average particle size of 50 nm) [39] and zinc acetate (hexagonal geometric form with a size between 25 and 50 nm) [40] as precursors. Additionally, the SAED imagens obtained for ZnON and ZnOA (Figure 2b,e) showed ring patterns (intensity and brightness) indicating polycrystalline nanomaterials [25,41]. It has been reported that the crystallinity of ZnO nanoparticles depends on chemical precursors, synthesis method and conditions (pH, solvent, and temperature), and annealing conditions [42].

Transmision Electron Microscopy (TEM) Studies
The morphology (TEM), the selected area electron diffraction (SAED) studies, and the mean size particles of ZnON and ZnOA are shown in Figure 2. The corresponding TEM (Figure 2a,d) image shows that materials exhibited a near-spherical shape for ZnON (Figure 2a), whereas ZnOA showed an elongated rod-like morphology (Figure 2d) with mean particle sizes of 99.28 and 40.94 nm (Figure 2c,f), respectively. Comparable shapes and sizes were previously reported when ZnO nanoparticles were synthesized using zinc nitrate (random spherical with an average particle size of 50 nm) [39] and zinc acetate (hexagonal geometric form with a size between 25 and 50 nm) [40] as precursors. Additionally, the SAED imagens obtained for ZnON and ZnOA (Figure 2b,e) showed ring patterns (intensity and brightness) indicating polycrystalline nanomaterials [25,41]. It has been reported that the crystallinity of ZnO nanoparticles depends on chemical precursors, synthesis method and conditions (pH, solvent, and temperature), and annealing conditions [42].  Figure 3 shows the UV-Vis spectra obtained for the ZnON and ZnOA nanoparticles. The powders exhibited a strong absorption wavelength of around 327 nm for ZnON and 406 nm for ZnOA, which is typical of ZnO materials [13]. According to Bathia and Verma [43], the UV-absorption peak of nano-ZnO may be shifted due to the difference in crystal size, influenced by the precursor and synthesis method. The band-gap energies of the  Figure 3 shows the UV-Vis spectra obtained for the ZnO N and ZnO A nanoparticles. The powders exhibited a strong absorption wavelength of around 327 nm for ZnO N and 406 nm for ZnO A , which is typical of ZnO materials [13]. According to Bathia and Verma [43], the UV-absorption peak of nano-ZnO may be shifted due to the difference in crystal size, influenced by the precursor and synthesis method. The band-gap energies of the ZnO N and ZnO A nanoparticles were calculated by applying Plank's equation [44], which showed significant differences in the band-gap values, 3.79 eV for ZnO N and 3.05 eV for ZnO A [13,45]. Similar trends have been reported in ZnO nanopowders synthesized by the sol-gel method using zinc nitrate (3.37 eV) [21] and zinc acetate (2.93 eV) as precursors [24]. According to Zeng et al. [46], various defect states are present in ZnO (donor defects such as Zn i

UV-VIS by Diffuse Reflectance Spectroscopy
and V o , and acceptor defects such as Zn" and Zn'), with the Zn interstitials and oxygen vacancies being the predominant ionic defect types influencing the excitation level of ZnO. Additionally, diverse authors argue that factors such as the shape of the nanoparticles, precursors, calcination temperature, and induced defects (degree of open lattice structure) influenced the band-gap values of ZnO nanoparticles [24,40,43].  Figure 3 shows the UV-Vis spectra obtained for the ZnON and ZnOA nanoparticles. The powders exhibited a strong absorption wavelength of around 327 nm for ZnON and 406 nm for ZnOA, which is typical of ZnO materials [13]. According to Bathia and Verma [43], the UV-absorption peak of nano-ZnO may be shifted due to the difference in crystal size, influenced by the precursor and synthesis method. The band-gap energies of the ZnON and ZnOA nanoparticles were calculated by applying Plank's equation [44], which showed significant differences in the band-gap values, 3.79 eV for ZnON and 3.05 eV for ZnOA [13,45]. Similar trends have been reported in ZnO nanopowders synthesized by the sol-gel method using zinc nitrate (3.37 eV) [21] and zinc acetate (2.93 eV) as precursors [24]. According to Zeng et al. [46], various defect states are present in ZnO (donor defects such as Zni •• , Zni • , Zni x , Vo •• , Vo • , and Vo, and acceptor defects such as Zn'' and Zn'), with the Zn interstitials and oxygen vacancies being the predominant ionic defect types influencing the excitation level of ZnO. Additionally, diverse authors argue that factors such as the shape of the nanoparticles, precursors, calcination temperature, and induced defects (degree of open lattice structure) influenced the band-gap values of ZnO nanoparticles [24,40,43].

X-ray Diffraction
The diffractograms (  [14,47,48]. On the other hand, the ZnO N showed reflection signals around 22.13 • , 29.04 • , and 44.6 • (2θ), which are representative of Zn(OH) 2 [49][50][51]. Similar peaks were observed in the diffractograms for PEG at 2θ = 26.19 • , 29.16 • , and 44.57 • , which is consistent with previous reports [52,53]. Some authors have suggested than an incomplete transformation of the precursor to ZnO during synthesis can occur [54][55][56]. Table 1 [58] reported a crystallite size and interplanar d-spacing of 0.28 nm and 0.32 nm, respectively, in ZnO synthesized through the wet chemical method using zinc acetate as the precursor. Nonetheless, it has been reported that the crystallite size of ZnO prepared by the sol-gel method using zinc acetate ranged from 11.84 to 24.82 nm in a pH-dependent response [59].

FT-IR Results
The FTIR spectrum of ZnON and ZnOA after calcination were recorded in the range of 4000 to 400 cm −1 ( Figure 5). In general, ZnON and ZnOA exhibited similar absorption regions, in agreement with previous reports [21,60]. Bands at 547 and 558 cm −1 were observed, and the vibrations in these regions are assigned to the Zn-O bond [21]. Signals at 803 and 1055 cm −1 are due to the C=O, C-C, and Zn-OH stretching modes [60][61][62][63]. However, ZnOA showed a noticeable decrease in transmittance at 1055 cm −1 in comparison with ZnON, which could be attributed to the stretching vibration of the residual C-O group present in the salt precursor [64], whereas the peaks around 1023 cm −1 depict the stretching vibration of the C-CH3 bond, associated with the ZnOA precursor [65][66][67].

FT-IR Results
The FTIR spectrum of ZnO N and ZnO A after calcination were recorded in the range of 4000 to 400 cm −1 ( Figure 5). In general, ZnO N and ZnO A exhibited similar absorption regions, in agreement with previous reports [21,60]. Bands at 547 and 558 cm −1 were observed, and the vibrations in these regions are assigned to the Zn-O bond [21]. Signals at 803 and 1055 cm −1 are due to the C=O, C-C, and Zn-OH stretching modes [60][61][62][63]. However, ZnO A showed a noticeable decrease in transmittance at 1055 cm −1 in comparison with ZnO N , which could be attributed to the stretching vibration of the residual C-O group present in the salt precursor [64], whereas the peaks around 1023 cm −1 depict the stretching vibration of the C-CH 3 bond, associated with the ZnO A precursor [65][66][67].

Nitrogen Physisorption Analysis
The N2 physisorption analysis was used to investigate the textural properties (BET surface area (SBET), total pore volume, and average pore size) of the ZnO nanoparticles

Nitrogen Physisorption Analysis
The N 2 physisorption analysis was used to investigate the textural properties (BET surface area (S BET ), total pore volume, and average pore size) of the ZnO nanoparticles (Table 1). Figure 6 shows the isotherms of ZnO N and ZnO A of type IV which are characteristic of microporous materials. The samples present H3-type hysteresis behaviour associated with interconnected micropores with non-uniform sizes or shapes [68].

Nitrogen Physisorption Analysis
The N2 physisorption analysis was used to investigate the textural properties (BET surface area (SBET), total pore volume, and average pore size) of the ZnO nanoparticles (Table 1). Figure 6 shows the isotherms of ZnON and ZnOA of type IV which are characteristic of microporous materials. The samples present H3-type hysteresis behaviour associated with interconnected micropores with non-uniform sizes or shapes [68]. ZnON and ZnOA showed SBET of 0.85 and 2.24 m 2 g −1 , respectively, which are in the range of those reported by Wang et al. [69]. Hamrouni et al. [70] synthesized ZnO nanoparticles (using C4H6O4Zn*2H2O as a precursor) with a BET surface area of 4.2 m 2 g −1 . It has been reported that the SBET of ZnO could be influenced by the type of precursor [71]. Wang et al. [71] reported a SBET of 91, 81, 79, and 68 m 2 g −1 for ZnO nanoparticles prepared using zinc sulfate, zinc nitrate, zinc acetate, and zinc chloride, respectively. Moreover, both ZnON and ZnOA nanoparticles exhibited differences in the total pore volume (ZnON of 0.01 cm 3 g −1 and ZnOA of 0.02 cm 3 g −1 ) with average pore diameters of 54.82 and 39.10 nm, respectively. Similar trends have been reported in the ZnO synthesized using different precursors (pore volume from 0.38 to 0.57 cm 3 g −1 and average pore size from 19.1 to 33.5 cm 3 g −1 ). According to Veriansyah et al. [42], differences in pore volume and pore diameter between samples could be associated with some superficial agglomerations, blocking the pores on the particle internal surface. ZnO N and ZnO A showed S BET of 0.85 and 2.24 m 2 g −1 , respectively, which are in the range of those reported by Wang et al. [69]. Hamrouni et al. [70] synthesized ZnO nanoparticles (using C 4 H 6 O 4 Zn * 2H 2 O as a precursor) with a BET surface area of 4.2 m 2 g −1 .
It has been reported that the S BET of ZnO could be influenced by the type of precursor [71]. Wang et al. [71] reported a S BET of 91, 81, 79, and 68 m 2 g −1 for ZnO nanoparticles prepared using zinc sulfate, zinc nitrate, zinc acetate, and zinc chloride, respectively. Moreover, both ZnO N and ZnO A nanoparticles exhibited differences in the total pore volume (ZnO N of 0.01 cm 3 g −1 and ZnO A of 0.02 cm 3 g −1 ) with average pore diameters of 54.82 and 39.10 nm, respectively. Similar trends have been reported in the ZnO synthesized using different precursors (pore volume from 0.38 to 0.57 cm 3 g −1 and average pore size from 19.1 to 33.5 cm 3 g −1 ). According to Veriansyah et al. [42], differences in pore volume and pore diameter between samples could be associated with some superficial agglomerations, blocking the pores on the particle internal surface. Figure 7 exhibits the XPS spectra of ZnO N and ZnO A samples and shows the corresponding signals for carbon (C1s), oxygen (O1s), nitrogen (N1s), and zinc (Zn2p). According to the nature of the precursor, it is expected that some N and C atoms would remain solid after calcination. The two peaks characteristic of Zn2p 3/2 and Zn2p 1/2 [72][73][74][75] are shown in Figure 7d, for ZnO N and ZnO A , respectively. These two signals are observed at 1022.04 and 1045.17 eV for the ZnO N sample, whereas those for ZnO A are located at 1021.44 and 1044.52 eV [76]. The Zn2p signals for ZnO N are shifted to higher binding energies than those measured for ZnO A . Moreover, the peaks assigned to C1s are displayed in Figure 7a; that for ZnO N is located at 285.02 eV, and that for ZnO A is placed at 284.9 eV. The difference indicates a shift to higher binding energies when ZnO N is used as a precursor [77]. For the signal corresponding to O1s (Figure 7b), the peaks are located at 533.19 and 530.32 eV for ZnO N and ZnO A, respectively, and belong to O − 2 of the wurtzite crystal structure [78]. The difference demonstrates a move to higher binding energies when using ZnO N [72]. From the XPS results, it is clearly observed that a shift to higher binding energies resulted when using Zn(NO 3 ) 2 * 6H 2 O compared to C 4 H 6 O 4 Zn * 2H 2 O, and this is probably due to the difference in C and N concentrations in both ZnO samples, as discussed below. Moreover, the signal intensity of the C1s is stronger for ZnO A than that observed for ZnO N, indicating the presence of carbon on the surface. The nitrogen signals indicated that nitrogen is present in both samples in different arrangements (Figure 7c). Shimizu et al. [79] noted that the nitrogen signals detected in the XPS study were related to the speciation of ammonium at the hematite/water interface. The authors also assigned the N1s peaks at 401.7 for NH + 4 and at 400.1 eV for NH 3 . In our study, the deconvoluted peaks were located at 393.8, 397.4, 399.2, and 401 eV for both samples and associated with the presence of the NH 3 group from the ammonium hydroxide (NH 4 OH) solution used for the adjustment of pH during the synthesis procedure. The atomic surface composition determined by XPS is shown in Table 2. The C/N atomic ratio was 47.22 and 32.21 for ZnO A and ZnO N , respectively. Furthermore, the O/C atomic ratio was 1.47 and 4.15 for the ZnO A and ZnO N samples, respectively. The small O/C atomic ratio for ZnO A is probably due to the doping of carbon into the ZnO structure which was the result of the zinc acetate precursor, as was suggested by Lavand and Malghe [75]. The presence of high N and C content in both samples after calcination indicates that N and C absorbed species from the reagents used int the synthesis method on the ZnO samples [79,80]. It is also worth noting that the O/Zn ratio varies; it was 1.28 for ZnO A , whereas it was 2.04 for ZnO N . These differences in the C/N, O/C, and O/Zn atomic ratios of the ZnO materials will undoubtedly influence the bandgap energies determined from the UV-Vis spectra.

XPS Spectroscopy
sor [77]. For the signal corresponding to O1s (Figure 7b), the peaks are located at 533. 19 and 530.32 eV for ZnON and ZnOA, respectively, and belong to O 2 of the wurtzite crystal structure [78]. The difference demonstrates a move to higher binding energies when using ZnON [72]. From the XPS results, it is clearly observed that a shift to higher binding energies resulted when using Zn(NO3)2*6H2O compared to C4H6O4Zn*2H2O, and this is probably due to the difference in C and N concentrations in both ZnO samples, as discussed below. Moreover, the signal intensity of the C1s is stronger for ZnOA than that observed for ZnON, indicating the presence of carbon on the surface. The nitrogen signals indicated that nitrogen is present in both samples in different arrangements (Figure 7c). Shimizu et al. [79] noted that the nitrogen signals detected in the XPS study were related to the speciation of ammonium at the hematite/water interface. The authors also assigned the N1s peaks at 401.7 for NH 4 + and at 400.1 eV for NH3. In our study, the deconvoluted peaks were located at 393.8, 397.4, 399.2, and 401 eV for both samples and associated with the presence of the NH3 group from the ammonium hydroxide (NH4OH) solution used for the adjustment of pH during the synthesis procedure. The atomic surface composition determined by XPS is shown in

Degradation of Organic Compounds
The absorbance spectra in the photodegradation of 2,4-D and 2,4-DCP exhibited three characteristic absorption bands centered at 210.5, 245, and 283 nm for 2,4-D, and 210.5, 245, and 285 nm for 2,4-DCP. The primary and secondary transitions were due to the aromatic group, while the third band belongs to the n → π* transition, which is attributed to the C-Cl bond [6]. The shoulder absorption corresponding to 2,4-D (283 nm) and 2,4-DCP (285 nm) showed a significant decrease as the irradiation time increased (360 min), indicating the degradation of 2,4-D and 2,4-DCP, respectively, which is associated with the decomposition of the organic compounds in both catalysts studied [81]. Figure 8 shows the degradation activities of ZnO N and ZnO A . In general, ZnO A exhibited better photocatalytic performance (p < 0.05) in the degradation of 2,4-D (90.9%) and 2,4-DCP (86.7%) than ZnO N (74.7 and 78.4%, respectively), and the photolysis presented a photocatalytic activity of 40.1 and 73.6%, respectively (Table 3). Rodriguez-Mata et al. [29] found that the ZnO (zinc acetate as precursor) was synthesized by the urea precipitation method and showed a degradation of 38.86% in 2,4-D. Similar trends were described by Meenakshi and Sivasamy [45], who reported that ZnO (zinc acetate as precursor) prepared by the sol-gel method exhibited a photodegradation of 70% under UV and visible light irradiations. Similarly, Ba-Abbad et al. [81] reported a degradation efficiency of 2,4-DCP > 96% using a commercial ZnO catalyst under UV-light irradiation after 120 min of exposure in a pesticide-dependent response. Moreover, the performance of ZnO as a catalyst for the photodegradation of a commercial herbicide (dimethylamine salt of 2,4dichlorophenoxyacetic acid) has been investigated by Burbano et al. [82], who reported a photocatalytic degradation of 38% after 8 h of treatment, whereas in the absence of the catalyst (photolysis), a pesticide degradation of 6% was observed. These results indicate that in the presence of ZnO nanoparticles, photolysis and photocatalysis, which further decompose the reaction intermediates, occurred simultaneously [83].    Kinetic study of the photocatalytic degradation of 2,4-D and 2,4-DCP was performed using the Langmuir Hinshelwood (LH) kinetics model. Additionally, Figure 9 shows the kinetic profile of the pseudo-first order reaction for both pollutants. Moreover, the time required for half of the organic compound present in the solution irradiated with UV-light to degrade was calculated [84]. The reaction half-life (t 1 /2) can be calculated with Equation (13) by replacing C by C 0 /2 Equation (1):  All values are mean ± standard deviation of three determinations. Different letters in each column indicate significant statistical differences between treatments (p < 0.05).
Kinetic study of the photocatalytic degradation of 2,4-D and 2,4-DCP was performed using the Langmuir Hinshelwood (LH) kinetics model. Additionally, Figure 9 shows the kinetic profile of the pseudo-first order reaction for both pollutants. Moreover, the time required for half of the organic compound present in the solution irradiated with UV-light to degrade was calculated [84]. The reaction half-life (t½) can be calculated with Equation (13) by replacing C by C0/2 Equation (1):  Table 4 show the apparent first-order kinetic constant according to the LH kinetics model (k), and the reaction half-life (t 1 /2) of 2,4-D and 2,4-DCP in the photocatalytic degradation showed a much longer tailing. These were obtained from linearization of Ln(C/C 0 ) vs. t [85] with high coefficients of correlation. It can be seen how k changes due to the type of precursor used for the ZnO synthesis. This is a typical phenomenon that occurs in nanomaterials due to the saturation of photocatalytic active centers.  Table 5 shows the different correlation coefficients under different kinetic models, where the highest photocatalyst concentration exerted a negative effect because the ZnO N nanoparticles had high concentrations of PEG and inhibited the UV irradiation pass, limiting the photocatalytic process.

Mineralization of Organic Compounds
The mineralization of the organic pollutants 2,4-D and 2,4-DCP was carried out through total organic carbon (TOC) analysis (Figure 10a,b). In general, the mineralization of the pollutant is dependent on the irradiation time and the photocatalyst precursor.  Table 5 shows the different correlation coefficients under different kinetic models, where the highest photocatalyst concentration exerted a negative effect because the ZnON nanoparticles had high concentrations of PEG and inhibited the UV irradiation pass, limiting the photocatalytic process.

Mineralization of Organic Compounds
The mineralization of the organic pollutants 2,4-D and 2,4-DCP was carried out through total organic carbon (TOC) analysis (Figure 10a,b). In general, the mineralization of the pollutant is dependent on the irradiation time and the photocatalyst precursor. According to these data, the degradation rate of both pollutants is higher than the mineralization process. Ramos-Ramírez et al. [6] mentioned that this phenomenon is attributed to the different absorption bands detected in the UV-Vis study, particularly for the aromatic group located at 210.5 nm. They suggested that the absence of oxygen limits According to these data, the degradation rate of both pollutants is higher than the mineralization process. Ramos-Ramírez et al. [6] mentioned that this phenomenon is attributed to the different absorption bands detected in the UV-Vis study, particularly for the aromatic group located at 210.5 nm. They suggested that the absence of oxygen limits the formation of superoxide radicals, which promotes partial dechlorination of organic pollutants; therefore, the aromatic structure of phenol present in both studied pollutants is maintained, avoiding the total degradation and mineralization. Furthermore, Wang et al. [69] reported that a small size in nanoparticles equates to a larger specific surface area, giving rise to a more significant number of active sites. Table 6 shows the results of data analysis for 2,4-D and 2,4-DCP. In the first case, we can see that the use of different precursors to obtain ZnO nanoparticles does not represent differences between treatments, with both showing very similar results. On the other hand, the treatments using catalysts compared to the photolysis process represent statistical differences (p < 0.05). However, the results of the analysis of mineralization of the 2,4-DCP present statistical differences (p < 0.05) between treatments. Macías-Sánchez et al. [86] used an aqueous solution of a mixture of 20 mg L −1 2,4-D and 5 mg L −1 picloram adjusted to pH 7 under visible light radiation ≥ 400 nm and found a mineralization efficiency of 0.2%. Barik and Gogate [87] investigated sonocatalytic degradation of 2,4-DCP using the process of ultrasonic irradiation in combination with ZnO nanoparticles and reported achieving, under conditions of initial concentration of 20 mg L −1 of pollutant and with a dose of 0.1 g L −1 of ZnO, a TOC removal of 24.6%.
The possible photocatalytic degradation mechanism of 2,4-D and 2,4-DCP under UV light irradiation is described below: Pesticides + e − cb → Reduction products (3) Equations (4) • are considered oxidizing agents that promote the mineralization of pollutants and generate CO 2 and produce H 2 O, which are considered final degradation products [29,88]. According to Djebbar and Sehili [89], about 70% of the conversion of 2,4-D involves hydroxyl radicals and about 30% can be associated with the capture of positive holes by adsorbed 2,4-D on semiconductors.

ZnO Nanoparticles Synthesis
ZnO nanoparticles were synthesized by the sol-gel method using Zn(NO 3 ) 2 * 6H 2 O and C 4 H 6 O 4 Zn * 2H 2 O as precursors [90] with some modifications. For this purpose, 14 g of Zn(NO 3 ) 2 * 6H 2 O was dissolved in 140 mL of ethanol in a three-mouth flask. In addition, 2 g of PEG was added to the solution (as a structure directing agent) under continuous magnetic stirring until the PEG dissolved completely [91]. Then, a few drops of NH 4 OH (1 M) were added to adjust the pH to 7 in the solutions. The solution was heated under reflux at 80 • C for two h under magnetic stirring. After this period, the solution was cooled down to −4 • C and aged for six h under cold storage. The gel was dried at 100 • C and annealed at 500 • C for 240 min in a static air atmosphere (heating rate of 2 • C min −1 ). A similar procedure was followed for the synthesis of ZnO nanoparticles using C 4

Scanning Electron Microscopy
The morphology of the materials was observed by scanning electron microscopy (Tescan, MIRA3 LMU, London, UK) operated at 20 kV.

Transmission Electron Microscopy
High-resolution images were obtained using a transmission electron microscopy (Jeol, JEM ARM200F, Boston, MA, USA) operated at 200 kV, and the images were analyzed using specialized software (Gatan Micrograph v. 3.7.0, Gatan Inc., Pleasanton, CA, USA).

UV-Vis Diffuse Reflactance Spectroscopy (UV-Vis DRS)
The absorption spectra of the materials were acquired by a UV-Vis DRS (Shimadzu UV-2600, Tokyo, Japan) provided with an integration sphere suitable for diffuse reflectance studies. The UV-Vis DRS spectra were obtained from a wavelength of 200 to 900 nm. From the plot, the band gap energy was calculated using Plank's Equation (8) [44].
where energy (E g ) = band gap energy (eV) and wavelength (ň) absorption peak value.

X-ray Diffraction
The X-ray powder diffraction patterns were acquired using an XRD Panalytical diffractometer (Empyrean, Almelo, Netherland) equipped with Cu Kα radiation (ň = 0.154 nm). Data were collected from 10 • to 90 • (2θ) with a scan rate of 0.02 • /0.2 s. The average crystal size was determined using the Scherrer Equation (9).
where D is the crystal size, k is the form factor (0.89), ň is the wavelength of Cu Kα radiation (0.154 nm), β is the width evaluated at mid-high of the most intense diffraction peak, and θ is the Bragg angle. The inter-planar distance (d) can also be evaluated from Bragg's law (10).

FTIR Spectroscopy
The FTIR spectra for the material were recorded with an FTIR (Thermo Fisher Scientific, Nicolet iS5, Tokyo, Japan) spectrophotometer using attenuated total reflectance (ATR) with a diamond waveguide (XR model). A detector of fast recovery deuterated triglycine sulfate (DTGS) (standard) was used for the analysis. The spectra were recorded at room temperature, with 24 scans and 4 cm −1 of resolution. Samples were recorded from 4000 cm −1 to 500 cm −1 .

Nitrogen Physisorption Analysis
The ZnO textural properties (surface area, pore volume, and pore size) were determined by nitrogen adsorption-desorption with a Micromeritics (TriStar II Plus, Norcross, GA, USA). The samples were degassed at 200 • C for 2 h under a vacuum. Nitrogen adsorption isotherms were measured at liquid nitrogen temperature (77 K) with nitrogen pressures ranging from 10 -6 to 1.0 P/P 0 . The specific surface area was obtained by the Brunauer-Emett-Teller method (BET) and the pore size distribution following the Barret-Joyner-Halenda (BJH) method.

XPS Spectroscopy
The surface composition of the powder was evaluated by X-ray photoelectron spectroscopy (XPS) using a spectrometer (Thermo Scientific K Alpha, Tokyo, Japan) operated at 1487 eV. Narrow scans with 60 eV Epass with a point size of 400 mm were acquired. To minimize the effect of charge, the samples were supported on an indium foil.

Photocatalytic Activity
The photocatalytic properties of the ZnO nanoparticles were studied through the degradation of 2,4-D and 2,4-DCP in a typical kinetic reaction experiment. A glass batch reactor provided with a cooling medium, and a UV-lamp of 1 mW cm −2 placed inside a quartz tube was used. The reactor was loaded with 200 mL of deionized water polluted with 80 mg L −1 of 2,4-D or 80 mg L −1 2,4-DCP and 100 mg L −1 of ZnO as a catalyst. The degradation curves were obtained by measuring the evolution of the absorbance maximum (at 283 and 285 nm for 2,4-D and 2,4-DCP, respectively) as a function of time (every 30 min up to 360 min) using a UV-Vis (Shimadzu UV-2600, Tokyo, Japan) spectrophotometer with scans from 200 to 900 nm. Furthermore, the total organic carbon in the samples was measured using TOC-LCSN equipment (Shimadzu, model) and calculated by applying the Equation (11): where TOC is the amount of total organic carbon (mg L −1 ), TC is the amount of total carbon (mg L −1 ), and IC is the inorganic carbon (mg L −1 ) in an aqueous solution.
The quantitative analysis of the reaction kinetics of the degradation of organic compounds can be obtained by fitting the experimental data using the Langmuir-Hinshelwood model and expressed by the following Equation (12): where k is the rate constant, K is the equilibrium constant, and c is the concentration of the organic pollutant. Nonetheless, Equation (1) can be simplified due to the low concentration of pollutant Kc < 1 to adapt to the form of a first-order apparent rate Equation (13): where k App is the apparent pseudo-first-order rate constant (min −1 ), C and C 0 are the final and initial organic compound concentrations (mg L −1 ), and t is the reaction time (min).

Data Analysis
The results of the degradation of organic pollutants were subjected to a statistical analysis applying a One-Way ANOVA and the Tukey test to analyze the differences between samples (p < 0.05). Statistical analyses were obtained with the help of Statistic software (V. 10.0 Statsoft, Tulsa, OK, USA).

Conclusions
Zinc oxide nanoparticles using zinc nitrate and zinc acetate as precursors were successfully synthesized following the sol-gel method. The morphology, crystalline structure, and physicochemical properties depend on the zinc oxide precursor. The final C/N, O/C, and O/Zn atomic ratios in the zinc oxide samples varied according to the precursor and as a consequence, a difference in the band gap values was determined. Our experiments showed that the ZnO A catalyst was better than the ZnO N sample in the photocatalytic degradation of both 2,4-DCP and 2,4-D under UV-light irradiation, which could be associated with the crystallinity, crystallite size, specific surface area, and type of ZnO precursor. The experimental results also demonstrate that both photolysis and photocatalysis contributed to the degradation of both organic pesticides.