Characterization of the nanoparticles and the nano-composite bimetallic catalyst were done by the analysis of the catalyst yield, and using BET surface area, XRD, FESEM, Raman Spectra, and DLS studies.
Table 1. shows the yield of nanoparticles after 15h of ball milling. The yield of CeO2 was 17.3 g. Yield of ZnO nanoparticle was 17.7g.The size of nanoparticles was found to be 24.89nm for CeO2and 15.9nm for ZnO based on the analysis of DLS and XRD. The efficiencyof the catalytic ozonation process depends to a large degree on the catalyst and its surface areaproperties. Thus the surface area is a crucial factor in catalytic ozonation. The results of surface area analysis areshown in Table 2. It shows that nano-composite bimetallic catalyst was having a surface area of 32.39 m²/g (BET surface area). The pore size was 88°A and Pore volume was greater than 0.071579 cm³/g.
XRD pattern of the CeO2nano-particlessynthesized by ball milling process is shown in Fig. 2. All peaks in the XRD spectra were indexed as (JCPDS- 34–0394) of CeO2. From the analysis of XRD pattern, peak intensity, position and full-width at half-maximum (FWHM) data were determined. The diffraction peaks located at 28.540, 33.080, 47.480, 56.340,59.090and 69.420 belongs to CeO2. This confirmed the synthesized nanoparticle’s purity. There were no other characteristic peaks other than CeO2 peaks. The synthesized CeO2 nanoparticle diameter was calculated using Debye-Scherrer formula (Desai et al., 2020).The average particle size of the sample was found to be 24.89 nm.
Figure 2, shows the XRD patterns of the individual metal oxides, AC and the nanocomposite bimetallic catalyst. The diffraction peaks at 28.540, 47.480, 56.340, 59.090 and 69.420belongs to CeO2and diffraction peaks located at 31.770, 34.430, 47.550and 69.680 were related to ZnO nanoparticles (JCPDS: 65-3411) (Shi et al., 2014). The peaks at 240 and 420 correspond to activated carbon (Song et al., 2017). The composite peaks found in the XRD analysis mostly corresponds to CeO2.The X-ray diffraction patterns of the activated carbon structure showed diffused peaks at 240 and 420. They appeared at narrow angles as fingerprint peaks. The AC structures were highly amorphous in nature and they had heterogeneous surface (Danish et al., 2011).
The structural characterization of pure CeO2 nanoparticle was done using FESEM. Figure 3 shows the FESEM-Image of CeO2, ZnO, AC nanoparticles and AC/CeO2/ZnOnano-composite bimetallic catalyst. The morphological studies showed that CeO2 nanoparticles had uniform agglomerated nanosphere structure.The morphology of the synthesized ZnO nanoparticle was in the form of triangle shaped nano rods like triangle prism and the pores had been created on activated carbon during the activation process of carbon. The pores were partially opened due to an increase in activation temperature from 5000C to 6000C. CeO2 nanoparticles were evenly distributed and ZnO was evenly impregnated on activated carbon.
Figure 4 (a) shows the Raman Spectra of AC/CeO2/ZnO nanocomposites.The Raman spectrum of the nanocomposites exhibited an intense band at 453.79 cm− 1, which is attributed to a symmetrical stretching mode of the CeO2(Maensiri et al., 2014; Reddy et al., 2007).The peaks for ZnO nanoparticles at 95.71 cm− 1 and 585.49 cm− 1 were assigned to the low, high longitudinal optical phonon peak of the ZnO nanoparticles (Du et al., 2005; Song et al., 2019; Damen et al., 1966). The obtained spectra also showed the presence of the band near 1583.56 cm− 1 (G band) typical of more organized graphitic materials and band at 1349.55 cm− 1 (D band) suggested the presence of more defective amorphous carbon structures. The peaks at 1593.55 cm− 1 and 132.71cm− 1 were typical of activated carbon (Nakamizo et al., 1974).
Particle size has a direct influence on material properties such as reactivity and dissolution rate of catalysts. Analyzing the particle size of the catalyst will fetch information on the interaction between catalyst and ozone. Figure 4 (b), represents the graphical representation of Dynamic Light Scattering result. The Particle size of the nano-composite bimetallic catalyst was found to be 453.3 d.nm.
Catalytic ozonation shows great advantages in removing the refractory organics present in water, and is expected to become a powerful and valuable technology in water treatment.Themechanism of catalytic ozonation is based on ozone decomposition reactions followed by the generation of hydroxyl radicals. The metal ions accelerate the decomposition of ozone to produce the •O2, and then electron of •O2 transfers to O3. This is followed by the formation of •O3, and •OH. Figure 5 represents the impact of pH, catalyst dosage and time on catalytic ozonation of BPA and TOC removal. It was found that when pH increases from 6 to 8 at catalyst dosage of 500µg/L and ozone rate 4g/h, the trend of TOC removal in oxidation process increased. When pH was 8, maximum TOC removal was observed within 35 minutes. Increasing the pH from 8 to 10 showed a decreasing pattern in TOC removal. The possible reason for showing maximum removal at pH 8 was because of the generation of more hydroxyl radicals that randomly reacted with BPA, and a decrease in TOC removal was due to clogging of hydroxyl radicals at higher pH (Wang et al., 2019).
Catalyst dosage is a significant aspect in catalytic ozonation. The catalyst surface and type of catalyst also plays a key role in heterogeneous catalytic ozonation. The catalyst dosage selected for the study was in the range of 250 µg/L to 750 µg/L. When the catalyst dosage increased from 250µg/L to 500µg/L at pH 8, 60 minutes and ozone rate 4 g/h, the TOC removal increased. Further increase in catalyst dosage did not show competent increase in TOC removal. At 500µg/L catalyst dosage, 61% TOC removal was achieved within a time of 60 minutes.
At pH 8, Ozone rate 4 g/h and 500 µg/L of catalyst dosage, maximum TOC removal was achieved within 60 minutes. This was because maximum ozone molecules reacted with the catalyst surface within this time. The Ozone molecules decomposed to hydroxyl radicals at the catalyst surface and reacted with the BPA.
Comparison of catalytic ozonation and non-catalytic ozonation
From Fig. 6, it is evident that the TOC removal efficiency of catalytic ozonation is high compared to non-catalytic ozonation. Non-catalytic ozonation achieved only 36% of TOC removal, while catalytic ozonation achieved 61% TOC removal. The increased efficiency was due to the formation of hydroxyl radicals by ozone decomposition on the surface of the nano-composite bimetallic catalyst. The available surface area of AC/CeO2/ZnO nano-composites prompted minimization of the diffusion limitations allowing the rapid adsorption and desorption of ozone molecules dissolved in water.
Figure 7 shows the FTIR spectra of AC/CeO2/ZnO nano-composite bimetallic catalyst before and after catalytic ozonation. The band due to the stretching frequency of Ce-O is below 785 cm− 1 which means that the stretching band at 551.93cm− 1 and 774.23 cm− 1 belongs to Ce-O stretch.The “scissor” bending of H-O-H broad absorption band located at 1596.26 cm− 1is associated with water (Jiang et al., 2016).The absorption band located around 3777.69 cm− 1 corresponds to the O-H stretching vibration of residual water and hydroxyl groups. The stretching at1225.23 cm− 1 can be attributed to the O-H vibration in absorbed water on the sample surface. The stretching frequency of Ce-O can be seen at 767.83 cm− 1also. The FT-IR peaks at 1589.98cm− 1, 1231.63cm− 1, 1039.71cm− 1, 1064cm− 1, 952cm− 1 and 767.83cm− 1 were similar to those of commercial CeO2 powders (Shen et al., 2013) and CeO2 nanoparticles (Phoka et al., 2009).The band at 767.83 cm− 1 corresponds to (Ce-O) metal-oxygen bond (Kumar et al., 2013). The small and weak stretching at 1210.83 is ascribed to C-O in carboxylic acid. The weak stretching at 1596.26 is assigned to carbonyl C = O present in esters, aldehydes, ketonic groups and acetyl derivatives. The small stretching at 2362.32 belongs to weak C ≡ C band of alkynes (Rother et al, 2016).
In the FTIR spectrum of nanocomposites, the absorption at 1601.28 cm− 1was assigned to the C = C stretching of activated carbon (Allwaret al., 2012; Rother et al., 2016). The absorption curve at 1001.97 cm− 1 belonged to the asymmetry vibration of Zn–O. The absorption curve at 812.72 cm− 1 was ascribed to the Zn-O stretching of ZnO (Xiong et al., 2006). The FTIR spectra confirmed the presence of nanocomposites and the absence of impurities in both the precursors and the prepared composite materials.
The FTIR spectra of AC/CeO2/ZnO nanocomposite obtained after the catalytic ozonation process confirmed the degradation of BPA and the formation of intermediates. The O-H stretching vibration at 3443.35cm− 1was attributed to the phenolic group. The stretching between the ranges of wave numbers2800-3200 cm− 1 were attributed to C-H stretching. Thepeaks at 1476.45cm− 1to 1670.81cm− 1wave numbers represented C-O andC-OH bonds of carboxylic groups (Ren et al., 2012). The peak at 1013.10 cm− 1is ascribed to the shift of skeletal vibration ofC(CH3)2 group of BPA (Sahre et al., 2006). The peaks with wave numbers less than 1000 cm− 1represents the para-di-substituted and mono-substituted and/or ortho-di-substituted compounds (Jang andWilkie, 2004). There was also an indication of the formation of polyphenols such as resorcinol (Jyoti et al., 2016). From the above observations, it can be inferred that AC/CeO2/ZnO nanocomposites facilitated the production of •OH and degradation of BPA.
The removal of BPA was analysed using HPLC and it is presented in the Fig. 8. At 60 minutes of catalytic ozonation, BPA concentration decreased about 97%. During catalytic ozonation, the degradation of BPA produced several low molecular weight organic acids which lead to the decrease in the initial solution pH. In order to better understand the BPA degradation during catalytic ozonation, pH of the solution after treatment was estimated with respect to different initial pH. The initial pH of the solutions were 5, 6, 8, 9, and 11 and after treatment the pH of the solutions were 4.81, 4.12, 5.26, 7.57 and 10.17 respectively. The incomplete removal of TOC indicated the possibility of theformation of intermediates.
LC-MS/LC-Q-TOF analysis was performed by comparing the chromatogram of BPA with those of the aliquots taken at different ozonation times. All samples were subjected to similar derivatisation procedure as mentioned inThalamadaiKaruppiah and Bhaskar Raju (2009). Figure 9(C) shows the LC-MS/LC-Q-TOF chromatogram of degradation byproducts. Identification of degradation byproducts was carried out based on fragmentation patterns in the mass spectrum and/or by comparing the mass spectrum with the library available in the instrument database. The proposed five aromatic degradation byproducts are given in Table 3. Researchers had reported the formation of hydroxylated BPA byproducts such as monohydroxylated BPA, dihydroxylated BPA and their quinones. The phenyl moiety based compounds such as p-isopropenyl and p-isopropyl phenol, p-hydroxyacetophenone, etc would have formed(Katsumata et al., 2004; Poerschmann et al., 2010; Olmez-hanci et al., 2013). Acidic compounds (responsible for pH decrease) such as formic, acetic, oxalic, succinic and fumaric acids were also reported (Katsumata et al., 2004; Olmez et al., 2015). Other studies (Poerschmann et al.,2010; Olmez et al.,2015) pointed to the formation of coupling byproducts with higher molecular weight than BPA.
The proposed fragmentation pathway of BPA by catalytic ozonation is displayed in Fig. 9. The •OH radicals ruptured the BPA mainly through two attack sites which were the bond that held the two aromatic rings together and aromatic ring itself. The •OH radicals attacked the methyl bond between the two aromatic rings of BPA and demethylation occurred by hydrogenation and dehydrogenation. The •OH radicals also attacked aromatic ring structure breaking it through hydroxylation and dehydroxylation. Once the ring structures were broken further rupture of the ring structure and hydrocarbon bonds occurred through the same oxidation and reduction reactions.