Photocatalytic degradation of Reactive Black dye using ZnO–CeO2 nanocomposites

This study presents the photocatalytic efficiency of ZnO–CeO2 nanocomposites for the degradation of a model Reactive Black (RB) dye. Nano-CeO2 was synthesized using cerium nitrate precursor solution via chemical precipitation. Synthesized nano-CeO2 was mixed with ZnO nanoparticles in different mass ratios to obtain ZnO–CeO2 heterojunction photocatalyst. The morphology of the nanocomposites was examined using transmission electron microscope (TEM). X-ray diffraction patterns of the CeO2 corresponded well with (1 1 1) plane of cubic-phase CeO2. The band gap of the ZnO–CeO2 nanocatalyst synthesized was determined to be 3.08 eV, which was lower than that of the pristine CeO2 and ZnO powders, respectively. The results indicate that 1:1 wt. ratio ZnO–CeO2 nanocomposite provides about 85% RB degradation within 90 min under UV light under alkaline pH conditions. Degradation rate of RB dye achieved with ZnO–CeO2 nanocomposite was almost 1.5 times greater than that obtained with pristine ZnO. Increasing CeO2 ratio beyond 1:1 wt. ratio did not significantly increase RB degradation. The results demonstrate that addition of CeO2 to ZnO results in lowering its band gap energy and aids charge carrier separation resulting in enhanced oxidation of RB dye under UV light.


Introduction
Over the last two decades, there has been an increased emphasis on semi-conductor materials as photocatalysts for recalcitrant organic pollutant degradation (Bellardita et al. 2020;Chatterjee and Chakraborty 2021;Kurian 2020;Nagajyothi et al. 2020;Sar et al. 2018). Conventional techniques such as filtration, adsorption, sedimentation, or chemical treatment are either less effective or expensive for the removal of emerging contaminants present in very low concentrations. Also, these techniques do not permanently eliminate the contaminants from the ecosystem, as they most often only transfer the contaminants from one stream to another. On the other hand, biological treatments are very slow and due to recalcitrant nature of the emerging contaminants, it is difficult to degrade them by biological methods (Sharma et al. 2019). Thus, photocatalytic reactions which are rapid, efficient at low pollutant concentrations, and capable of completely mineralizing the contaminants are most preferred.
Photodegradation reaction begins with absorption of light by the photocatalysts resulting in the formation of electron-hole pairs. For decades, photocatalysts such as ZnO and TiO 2 are commonly used in water treatment due to their (1) effectiveness; (2) ease in synthesis; (3) mechanical and thermal stability; (4) high specific surface area; (5) suitable band gap; and (6) persistent generation of photo-generated carriers (Cai et al. 2017;Davis et al. 2019;Guan et al. 2021;Jo and Tayade 2014;Lee et al. 2016;Mahlambi et al. 2015;Ong et al. 2018;Pare et al. 2008). When the semiconductors absorb photons with sufficient energy, their valence band electrons jump to conduction band creating holes in valence band and free electrons in conduction band. The charge carriers thus generated can initiate series of redox reactions on the catalyst surface. Typical mechanism of photocatalytic oxidation is schematically represented in Fig. 1. While the valence band holes generate hydroxyl radical by oxidizing the water molecules adsorbed on the catalyst surface, conduction band electrons generate superoxide radical by reducing the oxygen molecules (Hasija et al. 2021a;Sharma et al. 2019;Sharma et al. , 2022. Alternately, electron-hole pairs can undergo recombination and lose their redox potential. The recombination of electron-hole pair needs to be prevented to maximize the photocatalytic efficiency of the catalyst. A significant body of research has shown that nano-ZnO, typically a hexagonal, wurtzite-type semiconducting oxide, is highly effective in pollutant degradation because of its wide band gap in the near-UV spectral region, and good photocatalytic property (Gu et al. 2016;Lee et al. 2016;Ratshiedana et al. 2021;Sanoop et al. 2016). ZnO can be easily processed at low temperatures to form nano-ZnO, which contains significant amounts of lattice defects and oxygen vacancies in the crystal structures, and provides anisotropic growth which helps in increased photocatalysis (Gu et al. 2016). Even though a large number of photocatalysts had been developed in the recent years, commercial-scale application of photocatalysts had been only limited due to poor efficiencies of these catalysts. Poor light absorption efficiency and recombination tendency of the electron-hole pair generated are the two important reasons for this.
In order to overcome these drawbacks, various approaches including use of scavengers, fine-tuning of band gap, fine-tuning of catalyst morphology, and constructing heterojunctions are examined. Several metal, metal oxides, and non-metal dopants have been developed to tune the band gap of ZnO (Gu et al. 2016;Meshram et al. 2017). Addition of metal and metal oxide impurities on ZnO induces defects in the ZnO nanostructures, reduces the band gap energy levels, and expands its visible light response. Additionally, it can produce traps for photo-generated charge carries, thereby accelerating the charge transfer and inhibiting recombination of electron hole pairs (Bechambi et al. 2016;Caregnato et al. 2020;Jiang et al. 2017;Lang et al. 2016;Meshram et al. 2017;Parangusan et al. 2019).
Recent reports indicate that the heterojunction catalysts are more efficient in achieving charge separation in photocatalysts (Hasija et al. 2021b;Kumar et al. 2021;Patial et al. 2021). In heterojunction catalyst formed by coupling two semiconductors of compatible band gap, charge carriers generated in one semiconductor is transferred to another semiconductor thereby enhancing the lifetime of the electron-hole pair and hence the redox efficiency of the catalyst (Hasija et al. 2021a). It has been demonstrated that the CdS/ZnO heterostructure was more efficient in degradation of Rhodamine B dye compared to pure ZnO nanorod (Adegoke et al. 2019). It has been shown that the control morphology and crystalline structure are easy with bismuth oxyhalide catalyst and that their ability to form Z-scheme type-II heterojunctions has been excellent (Sharma et al. 2019). Complex heterojunctions have also been constructed to form Z-scheme heterojunctions wherein charge transfer for oxidation Fig. 1 Schematic diagram of a typical photocatalytic reaction reduction process takes place in a singular catalyst like manganese indium sulfide/cuprous oxide/silver oxide. Heterojunction sensitized with carbon quantum dots have been used for dye and bacterial removal from potable water. The ability of the catalyst to form O 2 • and h + radicals have been the characteristic indicator for the synthesis of Z-scheme photocatalyst (Sharma et al. 2022). Metal organic framework like GCN with wide range of light absorption has also been explored in the recent years (Hasija et al. 2022) (Patial et al. 2021). Even though such sophisticated systems exist, it is imperative to understand the photocatalytic mechanism and assess the ability of mixed metal oxide heterojunctions for the degradation/synthesis reactions.
In recent years, ceria (CeO 2 ) is the most common rareearth metal used due to its relative abundance, lower band gap (2.90-3.20 eV) compared to ZnO (3.10-3.30 eV), its chemical stability, and tunable band gap (Arul et al. 2012;Bellardita et al. 2020;Davis et al. 2019;Kurian 2020;Meshram et al. 2017;Nagajyothi et al. 2020;Rajendran et al. 2016;Sane et al. 2018;Veedu et al. 2020). The redox pair of Ce 3+ /Ce 4+ acts as an electron scavenger, reducing recombination of electron-hole pair, and reducing the band gap. Upon doping 2 mol% Ce onto ZnO, 100% bisphenol A degradation was reported, which was 5-10 more than that observed for pristine ZnO (Bechambi et al. 2016). Meshram and his co-workers reported 99% crystal violet dye degradation using 4% doped Ce on ZnO within 100 min of sunlight irradiation (Meshram et al. 2017). They also observed a 3-fold increase in photocatalytic activity of the Ce-doped ZnO, when compared to commercial pure ZnO. Maximum photocatalytic activity towards Direct Red 23 dye was observed with 3.28% Ce-doping on ZnO (R. Kumar et al. 2015). Increase in Ce doping beyond 3.28% reduced dye degradation moderately. In other studies, the optimum doping percent of Ce on ZnO nanoparticles was found to be 2-10% (Lang et al. 2016). Rajendran et al. observed a consistent decrease in the first-order degradation rate with increase in CeO 2 beyond 10 % (Rajendran et al. 2016). Upon use of 1:5 CeO 2 :ZnO, 98% Rhodamine B degradation was observed in 180 min, compared to the 85% using the primary oxide, while 10% doping of CeO 2 on ZnO reduced methylene blue concentration by 67% in 150 min, again a 30% increase compared to the primary metal oxide photocatalyst (Lee et al. 2016).
In this work, a CeO 2 -ZnO heterojunction nanocomposite was synthesized by mixing CeO 2 and ZnO nanoparticles at various weight ratios. The morphological characteristics of the ZnO-CeO 2 nanocomposite were determined. Reactive Black (RB) was used as the model dye compound to evaluate photocatalytic activity of the nanocomposite under UV light illumination. With the exception of one study by Rajendran et al. (2016), most studies have used less than 10 wt.% CeO 2 doped ZnO. Therefore, it is crucial to understand the effect of higher CeO 2 content on ZnO and its ability to influence the photocatalytic behavior of the composite. Additionally, the effect of process parameters such as dye concentration, catalyst amount and ZnO:CeO 2 weight ratios, and pH on photocatalytic ability of ZnO-CeO 2 nanocomposite was evaluated to ascertain the efficacy of the nanocomposite under various operating parameters.

Materials
Commercial nano-zinc oxide (ZnO, 99.9%) was purchased from Spectrum India Ltd. Sodium carbonate (Na 2 CO 3 ) was purchased from Paxmy Specialty Chemicals, India. Cerium (IV) ammonium nitrate ((NH 4 ) 2 Ce(NO 3 ) 6 ), Reactive Black dye (RB, C 26 H 21 N 5 Na 4 O 19 S 6 ), NaOH, and H 2 SO 4 were purchased from Avra Synthesis Ltd., India. All chemicals were of reagent grade and utilized as obtained. Double distilled water was used for material synthesis and photocatalytic experiments.

Synthesis of ZnO:ceria composites
Synthesis of cerium oxide (CeO 2 ) nanoparticles was adapted from previous studies (Lang et al. 2016;Ma et al. 2019;Nagajyothi et al. 2020;Sane et al. 2018). In 30 ml distilled water, 10 g cerium ammonium nitrate was added and kept for 30 min of stirring. The as-prepared solution was added dropwise to 20 ml saturated solution of aqueous sodium carbonate with a pH of 9. A white precipitate was formed soon after cerium ammonium nitrate was added. Using stirring, the formed precipitate was re-dissolved. By adding solid sodium carbonate, the ammonium carbonate solution was sustained at a pH of 9. The obtained slurry is centrifuged, and the wet product was calcined to 750°C, and synthesized CeO 2 nanoparticles were stored at room temperature. ZnO:CeO 2 nanocomposites were prepared by adding appropriate quantities of nano-ZnO and nano-CeO 2 particles at varying weight ratios ranging from 1:4 to 2:1, hereby referred to as 1C1Z (1:1 CeO 2 :ZnO), 0.5C1Z (1:2 CeO 2 :ZnO), 0.33C1Z (1:3 CeO 2 :ZnO), 0.25C1Z (1:4 CeO 2 :ZnO), and 2C1Z (2:1 CeO 2 :ZnO).

Material characterization
X-ray diffraction (XRD) analysis of the nanoparticles was documented on a Focus X-ray Diffractometer (Bruker, Germany). Debye-Scherrer formula was used to estimate the crystallite size of nanoparticles. Transmission electron microscopy (TEM, Hitachi) was used to determine the particle size and morphology of ZnO and CeO 2 nanoparticles. UV-visible diffuse reflectance spectra of ZnO and TiO 2 were carried out on a HACH UV-Vis spectrophotometer with a resolution of 5nm in the range of 200-900 nm, and the band gap was determined using the Kubelka-Munk function. Reactive Black dye (RB) was used as the model pollutant to determine the photocatalytic activity of ZnO:CeO 2 mixed oxides. The photodegradation studies were performed in a photochemical reactor setup provided by SAIC India Ltd, Chennai, similar to the unit reported in our earlier study.

Photocatalytic experiments
Experiments were done in a well type reactor of 100 ml capacity, with constant air circulation to ensure uniform mixing. Irradiation was provided using 30 W, 365 nm high pressure xenon long arc lamp. The concentration of RB investigated ranged from 250 mg/L to 1000 mg/L, with varying amounts of the ZnO:CeO 2 mixed oxides. It was ensured that the catalyst was well dispersed in solution, prior to the start of the experiment. Control experiments in dark and non-aerated conditions with and without the catalyst were also performed to determine auto-degradation of RB dye. Parametric evaluation of the photocatalytic process was performed as follows on the best performing ZnO:CeO 2 mixed oxide: (1) the effect of dye concentration ([RB] = 250 − 1000 mg/L) using 1 g/L of nanoparticle; (2) the effect of nanoparticle loading (0.5-1.5 g/L) at RB concentration of 250-1000 mg/L; and (3) the effect of pH. Percent dye degradation was calculated according to Eq. (1). Here C 0 is the initial concentration (mg/L) and C t is the concentration (mg/L) at a given time. The kinetic study data obtained was fitted to pseudo-first-order and pseudo-second-order kinetic equations, commonly used to determine rate constants for photocatalytic processes, as shown in Eqs. (2) and (3), where k is the rate constant. DR-6000 UV-Vis spectrophotometer (HACH India) was used to determine RB dye concentration at a wavelength of 595 nm.
The diffuse reflectance spectra of the ZnO and CeO 2 and 1:1 CeO 2 :ZnO composite (1C1Z) are shown in Fig. 3. The intercept of plot of photon energy (eV) versus (( ) * hν) 1/2 (Kubelka-Munk function) revealed that the band gap energy for ZnO and CeO 2 was 3.12 eV and 3.24 eV, respectively. The band gap energy obtained for CeO 2 in this study was higher than that reported in other studies (Sane et al. 2018;Veedu et al. 2020). Upon addition of the ceria to the ZnO nanoparticles, the band gap energy reduced to 3.08 eV.
It can be noticed that addition of CeO 2 to ZnO helped decrease the band gap energy, which may result in an increased efficiency of the composite photocatalyst. Similar decrease in the band gap energy was observed in other studies (Lang et al. 2016;Rodwihok et al. 2020;Veedu et al. 2020). Luo et al. (2020) noted that with 3% doping of CeO 2 on ZnO, the band gap energy reduced from 3.12 to 3.04 eV, and attributed the decrease to increased oxygen vacancies created upon addition on CeO 2 , a rare-earth metal oxide found to be abundant in oxygen vacancies. The charge carrier separation mechanism observed in ZnO-CeO 2 catalysts reported in previous literatures is schematically represented in Fig. 4. As the valence and conduction bands of CeO 2 are more negative than ZnO band energy levels, in ZnO-CeO 2 heterojunction, the photoinduced electrons (CB) will transfer from CeO 2 to ZnO and holes (VB) will transfer from ZnO to CeO 2 enabling efficient charge carrier separation. Based on the oxidation potentials of ZnO and CeO 2 reported in previous literatures, the generation of the following ROS, namely OH• (~2.8 eV), H 2 O 2 (~1.78 eV), HO 2 • (~1.68 eV), and O 2 (~1.23 eV), is thermodynamically feasible (Yu et al. 2019).

Photocatalytic degradation of Reactive Black dye
Control experiments were performed to evaluate the auto-degradation of RB dye. As seen in Fig. 5A, degradation was negligible under dark conditions. In the presence of UV light and aeration, but absence of the catalyst, less than 5% degradation was observed. This is most likely due to photolytic degradation of the dye (Kuo and Ho 2001). Addition of CeO 2 NPs to either an aerated or non-aerated dye solution showed insignificant dye degradation, suggesting no adsorption of RB on CeO 2 (Fig. 5A). However, when ZnO was added to both an aerated and non-aerated dye solution, in dark conditions, up to 7% dye removal was observed, indicating sorption. Figure 5B and C and Fig. 6 show the photocatalytic ability of various ZnO:CeO 2 wt. ratios on RB degradation at [RB] = 1000 mg/L and 1 g/L catalyst, and Table 1 shows the percent degradation. Within 90 min, > 60% degradation was observed with the use of different weight ratios of CeO 2 and ZnO nanoparticles. The absorbance spectra of RB decreased significantly within the first 15 min upon use of CeO 2 -ZnO nanocomposites, Fig. 2 Characterization of as-prepared CeO 2 and ZnO: a, b TEM images of as-prepared CeO 2 ; c, d TEM images of as-prepared ZnO; e, f XRD patterns of CeO 2 and ZnO with near complete degradation within 1 h (Fig. 6). The degradation reached near equilibrium within 70 min and significantly reduced degradation rate was observed beyond reaction time of 70 min. The percent degradation was highest for 1C1Z (87.7%), while it was 80.4, 66.1, 68.2, and 85.4% for 1C2Z, 0.33C1Z, 0.25C1Z, and 2C1Z respectively. Decreasing the amount of CeO 2 in the ZnO-CeO 2 heterojunction nanocomposite appeared to reduce RB degradation rates significantly (Table 1). Among the various wt. ratios of the composites investigated, IC1Z provided the maximum dye degradation. It is hypothesized that as the CeO 2 content is increased, efficient separation of electron-hole pair occurs favoring redox reactions on the surface, and hence increased percent degradation was observed.
Typically, photocatalytic degradation kinetics can best be described using first-order rate kinetics. Figure 5C shows the plot of ln (C/C 0 ) vs. time for the various ZnO:CeO 2 nanocomposites. The first-order plot clearly indicated that ZnO:CeO 2 at 1:1 wt. ratio (1C1Z) performed the best among the various photocatalysts (Fig. 5C). The first-order rate constant determined for 1C1Z was almost 3-fold higher than that for ZnO, and almost similar to pure CeO 2 at 1000 mg/L RB concentration and 1 g/L catalyst loading (Table 1). Increasing ZnO content in the composite decreased the first-order rate constant, while increasing CeO 2 content increased it, largely due to the decrease in the band gap energy upon addition of CeO 2 . The data was also fitted to a secondorder rate expression (Fig. 7). The data fitted well to the pseudo-second-order rate expression as well (Eq. 3, Table 2) at 1000 mg/L RB concentration and 1 g/L catalyst loading (R 2 : 0.93-1.00). Sane et al. (2018) reported better fit of a second-order model to the kinetic data on the degradation of reactive dyes using CeO 2 . The rate constants reported in their study ranged from 0.0085 to 0.016 L/mg min for the CeO 2 photocatalyst investigated for reactive dyes in the concentration range of 10-100 mg/L, slightly higher values of rate constant to the values obtained in our study.

Effect of initial dye concentration
The effect of initial dye concentration on degradation rates was examined. Figures 8 and 9 show the kinetic data and the respective first-order kinetic plots for RB  concentration of 500 mg/L ( Fig. 8A and B) and 250 mg/L ( Fig. 9A and B), at 1 g/L catalyst loading, respectively. In about 75 min, greater than 90% degradation was observed for various weight ratios of CeO 2 and ZnO nanoparticles. Similar to that observed for 1000 mg/L RB concentration, decreasing CeO 2 content in the nanocomposite decreased percent degradation, albeit marginally for an initial concentration of 500 mg/L (Fig. 8). The degradation kinetics clearly fitted to first-order rate expression with high R 2 values, but appeared to be poorer fits for second-order rate expressions (Table 2). As the initial RB concentration is lowered to 250 mg//L, percent removal increased to near 100% (Fig. 9A). Similar to the previous observation, the data fitted better to first-order rate expressions, rather than second-order rate expressions (Fig. 9B). Similarly, increasing CeO 2 fraction on ZnO did not significantly improve the percent degradation since near complete degradation was observed for all photocatalysts used within 45 min (Fig. 9A). Equilibrium was also obtained within 45 min with negligible change in rate of degradation after 30 min. Although degradation rate was slower for CeO 2 and 1C1Z initially, the degradation rate of 1C1Z after 45 min was comparable with that of ZnO photocatalysts. Decrease in the initial RB concentration resulted in faster degradation, as observed by a 2-5-fold increase in the first-order rate constant (Table 2). For 1C1Z, the rate constant amplified almost 6 times from 0.021 to 0.136 min −1 as RB concentration decreased from 1000 to 250 mg/L. It was observed generally that as initial RB concentration decrease, the rate constant and therefore the initial rate (kC o ) increased. For the same number of reactive sites on the surface, decrease in [RB] moieties due to decreasing initial concentration results in increased adsorption on the surface and increased degradation. At higher concentration, there is competition for the reactive sites; hence, the initial rate is lowered. In all the experiments with varying RB concentrations and ZnO:CeO 2 weight ratios, 1C1Z exhibited the highest degradation rate, with the first-order rate constant almost twice as high as that of pure ZnO.

Effect of catalyst loading
To further evaluate the photocatalytic process, kinetic experiments were performed at varying catalytic loadings using the best performing photocatalyst, 1C1Z, and its efficacy compared to that of pristine nano-ZnO and nano-CeO 2 . Figure 10a-c show the kinetic data for catalyst loading of 0.5 g/L (RB concentration = 250-1000 mg/L) and Fig. 10d-f show the kinetic data for catalyst loading of 1.5 g/L (RB concentration = 250-1000 mg/L). Table 3 presents the percent degradation and constants to the first-order data fit. For 250 and 500 gm/L RB concentration (Fig. 10), there was no appreciable difference in degradation rate among the three photocatalysts investigated, ZnO and CeO 2 and 1C1Z. Near complete degradation was observed within 60 min, particularly at low concentrations. At low catalyst loading (0.5 g/L) and high RB concentration, the degradation rates were on RB degradation at RB=1000 mg/L and catalyst = 1 g/L; and C data fit to first-order rate expression lower. Equilibrium was not attained within 50 min for 1000 mg/L and 500 mg/L RB concentration, while it was observed with 250 mg/L RB concentration. The percent degradation was 35-40% for 1000 mg/L, 60-65% for 500 mg/L, and almost 100 % for 250 mg/L, at 0.5 g/L. The percent degradation was 55-60% for 1000 mg/L, 85-95% for 500 mg/L, and almost 100 % for 250 mg/L, at 1.5 g/L. As the catalyst loading was increased, degradation rates were faster as evidenced by the near complete completion in lesser reaction times. Increasing catalyst loading  Table 1 First-order rate constant and percent degradation for various ZnO:CeO 2 composites Catalyst loading = 1 g/L Photocatalyst 1000 mg/L 500 mg/L 250 mg/L increased percent removals and degradation rates as there is an expected rise in the number of reactive sites. The data was fitted to first-order rate expression, and the rate constant deduced. The first-order rate constant for degradation of dye by ZnO was observed to be 0.008 min −1 (R 2 =0.953), while it was 0.01 min −1 (R 2 = 0.983) and 0.007 min −1 (R 2 =0.988) using CeO 2 and 1C1Z respectively for RB concentration of 1000 mg/L. Similarly, for 500 mg/L RB dye concertation, the first-order rate constants were found to be 0.027 min −1 (R 2 = 0.993), 0.027 min −1 (R 2 = 0.992), and 0.019 (R 2 = 0.984) for ZnO, CeO 2 , and 1C1Z respectively. Correspondingly, the first-order rate constants for 250 mg/L RB dye concentration provided the values of 0.106 min −1 (R 2 = 0.933), 0.07 min −1 (R 2 = 0.978), and 0.04 (R 2 = 0.99) for ZnO, CeO 2 , and 1C1Z respectively. Table 3 presents the rate constant for a catalyst loading of 1.5 g/L. It can be seen that increasing loading three times increased the rate constant also 3-fold, for 1000 mg/L. However, when the initial RB concentration was 250 mg/L, the rate constant remained constant or marginally increased. This suggested that RB concentration was the rate-controlling factor when the number of reactive sites was in abundance. The results from this study were compared to results from studies wherein binary metal oxide photocatalysts were used for dye degradation. As seen in Table 4, when ZnO:CeO 2 was used in a ratio of 1:5 for Rhodamine degradation, the efficiency was 98%, albeit at low dye concentration (24 mg/L). Similar efficiencies were reported when the base material ZnO was doped with other metal oxides such as Bi 2 O 3 and CdO. It has also been noted that doping a second metal oxide on ZnO significantly improved degradation efficiencies, as is the case here.

Effect of pH on RB degradation
The effect of pH on RB degradation was evaluated on the best performing photocatalyst, 1C1Z, at various solution pH and its efficacy compared to that of pristine nano-ZnO and nano-CeO 2 . Here, the effect of pH was performed at 1000 mg/L dye concentration and 1 g/L catalyst loading (Fig. 11). In the acidic pH region, degradation efficiencies were very low (< 30%) for all the photocatalysts investigated. However, as pH was increased to 11, higher degradation of 99%, 98%, and 47% was observed for 1CIZ, ZnO, and CeO 2 respectively. Previous research has shown that photocatalytic reactions are pH-dependent, primarily caused by the surface charge of the catalyst and the molecular structure of the organic contaminant. The increase in percent degradation at higher pH can be attributed to the increase in the number of hydroxyl ions. The pKa value for Reactive Black dye is 3.8 and 6.9. The pH at point of zero charge (pH PZC ) for CeO 2 has been reported to be 6.9 while that for ZnO has been 8.7-9.7 (Meshram et al. 2017). Since  RB is an anionic dye, electrostatic attraction between the dye molecule and catalyst surface at pH < pH PZC results in the dye removal at acidic pH. Since the pH PZC shifts towards more alkaline pH in a ZnO:CeO 2 mixture, higher percent removals are expected. Sane et al. (2018) reported that activity of CeO 2 for dye degradation was higher at neutral pH and followed the order: 7 > 9.2 > 4. However, when ZnO-based catalysts were used for degradation of Rhodamine B dye, higher pH provided the best degradation (Saffari et al. 2020). Highest percentage removal of dye was observed at a pH of 11. Among all the photocatalyst studied, 1C1Z performed the best, clearly indicating that the addition of CeO 2 to ZnO significantly improved the photocatalytic activity of the composite.
As demonstrated above, ZnO-CeO 2 used in this study can serve as a potential heterojunction photocatalyst for degradation of dyes and hence can be tried for similar environmental contaminants. It is important to note that the overall economy of a photocatalytic process depends on feasibility of catalyst recovery and reuse. Also, several regulating agencies do not permit the presence of photocatalysts in treated water. Thus, recovery of the catalysts is very much essential. The catalysts used in this study could be easily separated from the solution due to the difference in specific gravity. Nevertheless, incorporation magnetic property is an important and most prominent approach to enable recovery and reuse of the catalysts (Shekofteh-Gohari et al. 2018). It can be considered for future research.

Conclusion
In the present research, ZnO-CeO 2 heterojunction nanocomposites were synthesized and evaluated for their photocatalytic activity towards the degradation of Reactive Black dye. Addition of CeO 2 to ZnO decreased the band gap of native ZnO up to 10% CeO 2 loading. The band gap energy was lowest for the composite with 1:1 weight ratio Fig. 8 A Effect of CeO 2 content in ZnO:CeO 2 composite on RB degradation for RB=500 mg/L and catalyst = 1 g/L; and B data fit to first-order rate expression Fig. 9 A Effect of CeO 2 content in ZnO:CeO 2 composite on RB degradation for RB=250 mg/L and catalyst = 1 g/L; and B data fit to first-order rate expression of ZnO to CeO 2 and maximum dye degradation (85%) was attained with this composite with an initial dye concentration of 1000 mg/L. Increasing CeO 2 content beyond 1:1 wt. ratio did not have significant effect on the catalyst both in terms of change in band gap energy and dye degradation efficiency. Fitting of experimental data to the kinetic models revealed that the kinetic data fitted well with the first-order model. The effect of catalyst loading showed that the increase in catalyst loading showed enhanced effect in the degradation performance, nevertheless at the higher concentration of RB exhibited decreased removal performance. The results demonstrate that the ZnO-CeO 2 heterojunction catalyst exhibits a better degradation potential for the reactive dye suggesting that it could be potential catalyst for other contaminants as well. Moreover, to improve the economics of the overall process, catalyst recovery strategies such as formulation of magnetic particles need to be explored.