Perovskite Zinc Titanate Photocatalysts Synthesized by the Sol–Gel Method and Their Application in the Photocatalytic Degradation of Emerging Contaminants

: In this study, perovskite ZnTiO 3 photocatalysts were fabricated by the sol–gel method. The photocatalytic capability was veriﬁed by the degradation of the emerging contaminant, the antibiotic amoxicillin (AMX). For the preparation, the parameters of the calcination temperature and the additional amount of polyvinylpyrrolidone (PVP) and ammonia are discussed, including the calcining temperature (500, 600, 700, 800 ◦ C), the volume of ammonia (750, 1500, 3000 µ L), and the weight of PVP (3 g and 5 g). The prepared perovskite ZnTiO 3 was characterized by XRD, FESEM, BET, and UV-Vis. It is shown that the perovskite ZnTiO 3 photocatalysts are structurally rod-like and ultraviolet light-responsive. Consequently, the synthesis conditions for fabricating the perovskite ZnTiO 3 photocatalysts with the highest photocatalytic performance were a calcining temperature of 700 ◦ C, an additional ammonia amount of 1500 µ L, and added PVP of 5 g. Moreover, the photocatalytic degradation of perovskite ZnTiO 3 photocatalysts on other pollutants, including the antibiotic tetracycline (TC), methyl orange (MO), and methylene blue (MB) dyes, was also examined. This provides the basis for the application of perovskite ZnTiO 3 as a photocatalyst to decompose emerging contaminants and organic pollutants in wastewater treatment.


Introduction
The use of antibiotics has increased with the development of medicine, fishery, and animal husbandry in recent years. In the aquatic environment, the sequencing proliferation of antibiotic-resistant genes (ARGs) and antibiotic-resistant bacteria (ARBs) has become a global issue. ARBs are transferred to drinking water sources via wastewater emissions in various circumstances, leading to a severe threat to human health, prolonging morbidity and increasing mortality. In fact, ARBs and ARGs have been discovered in various environments, such as aquaculture farms, hospital wastewater, livestock effluent wastewater, surface water, groundwater, wastewater treatment plants, etc. Even worse, ARGs and ARBs have been detected in untreated drinking water sources and even in tap or bottled water [1].
It was found that ARBs are generated due to ADP-ribosyl and glycosyltransferases with the sequential acetylation, phosphorylation, adenylation, nucleotidylation, ribosylation, and glycosylation of the antibiotics [2]. It was also reported that antibiotics cause the evolution of ARBs, even at low concentrations of the antibiotics [3]. Therefore, wastewater treatment to degrade antibiotics becomes more and more urgent.
Pharmaceutical antibiotics are also regarded as being among the emerging contaminants (ECs) [4]. Among the antibiotics, penicillins have played the most critical role in human and veterinary medicine [5]. In most countries, the consumption of penicillins accounts for 50-70% of the total amount of antibiotics [6]. Amoxicillin (AMX) is the most commonly used β-lactam antibiotic due to its broad action. It is often the first candidate for general infections [7]. While AMX is discharged into water resources through improper waste treatment methods, it could become a pollutant that would cause environmental damage. Accordingly, it is critical to develop a green procedure to decompose antibiotics and lower their emissions into natural water sources.
Composite materials are also regarded as excellent photocatalysts to degrade emerging contaminants due to the different synergistic effects of these binary materials. For example, Kanakaraju et al., integrated TiO 2 /zeolite composites to remove AMX with light irradiation at a wavelength of 200~600 nm in 2015. AMX was successfully removed according to the mechanism of adsorption on zeolite and photocatalysis on TiO 2 [39]. On the other hand, in 2016, Belaissa et al., indicated that CuO in the heterosystem of the CuO/TiO 2 photocatalyst acted as a sensitizer to absorb longer wavelength light through the synergistic effect. Accordingly, they effectively degraded AMX by using simulated sunlight illumination [40]. In contrast, Yang et al., fabricated a novel composite photocatalyst that was constructed from a MOF, MIL-68(In)-NH 2 , and graphene oxide (GrO). Based on the semiconductor properties of MOFs and the excellent electron transport performance of GrO, the composite materials exhibited an excellent photocatalytic activity under illumination by a light source with a wavelength of 420 nm. The experimental results showed that MIL-68(In)-NH 2 /GrO had better performance in decomposing AMX than the individual materials alone [41]. Graphite carbon nitride (g-C 3 N 4 ) is also a promising candidate for photocatalysis due to its excellent chemical stability, unique two-dimensional structure, and small energy gap (2.70 eV). Dou et al., utilized g-C 3 N 4 to successfully degrade AMX at different initial concentrations in 2019 [42]. Similarly, in 2019, Mirzaei et al., prepared Fe 3 O 4 /g-C 3 N 4 photocatalysts irradiated under different light sources to decompose AMX in a solution [43].
To the best of our knowledge, there have been few perovskite materials used in the degradation of AMX. In 2018, Haddadou et al., indicated that perovskite Ba(Ti 0.950 Sc 0.025 Nb 0.025 )O 3 photocatalysts showed a large dielectric constant and a wide space charge region, which were helpful in the photocatalytic degradation of AMX [44]. Kong et al., synthesized and characterized perovskite ZnTiO 3 photocatalysts. The photocatalytic performance of the azo dye methyl violet was also verified [45]. Therefore, based on the above discussion, perovskite ZnTiO 3 could serve as a potential photocatalyst to decompose emerging contaminants. In fact, perovskite ZnTiO 3 is also used in a wide array of applications, such as on gas sensors, microwave dielectrics, and sorbents; therefore, it is critical to develop a facile method to fabricate ZnTiO 3 . In this study, AMX is the targeted molecule, and various perovskite ZnTiO 3 photocatalysts were prepared using a simple sol-gel method under various synthesis conditions to verify their photocatalytic performance.

Results
The uncalcinated ZnTiO 3 photocatalysts were submitted to thermogravimetric analysis and differential thermal analysis, as shown in Figure 1. Initially, a weight loss of 19.9% was observed before 200 • C in the first stage, mainly due to the evaporation of the glycol solvent [46]. A weight loss of 40.5% occurred from 200 to 500 • C in the second stage. In the second stage as well, thermal decomposition with a maximum weight loss of 31.4% was observed at 315 • C. This indicated that massive thermal cracking of PVP and dehydroxylation of Ti-OH to TiO 2 occurred in this stage. The sharp peak that appeared at 315.5 • C was attributed to the combustion of organic compounds [47]. The weight loss in the third stage between 500 and 800 • C reached 42.4%, and the weight loss increased by approximately 2%, indicating that the residual organic components had been burned out. Since the TGA curve did not change much after 500 • C, this suggested that the samples had to be calcined above this temperature at least to remove the organic precursors. Accordingly, the thermal post-treatments of the as-prepared ZnTiO 3 samples were set at 500, 600, 700, and 800 • C.  . This also means that the calcination temperature of 500 • C was not enough for the crystallization of The strength of these peaks increased even more rapidly at 700 • C. As the temperature of heat treatment increased further, the cubic phase of ZnTiO 3 was transformed into the hexagonal phase at 800 • C [48]. The characterized hexagonal phase was verified from the 2 θ of 19. On the other hand, the XRD pattern also revealed that the cubic-phase ZnTiO 3 was partially decomposed into cubic Zn 2 TiO 4 at 800 • C. Accordingly, the different calcination temperatures had a significant impact on the crystallinity of ZnTiO 3 . The effect of various additional NH 4 OH amounts during synthesis on the crystallinity is shown in Supplementary Materials Figure S1. The crystallinities of ZnTiO 3 prepared by adding 750 µL and 1500 µL of NH 4 OH during the synthesis seem very similar, indicating mainly cubic ZnTiO 3 (JCPDS 39-0190. When the amount of NH 4 OH was increased to 3000 µL, a small amount of cubic Zn 2 TiO 4 crystal (JCPDS 25-1164) appeared. On the other hand, the effect of adding PVP on the crystallinity was also examined. When the amount of PVP of 3 g was added, primarily cubic ZnTiO 3 appeared along with some cubic Zn 2 TiO 4 . However, increasing the amount of PVP up to 7 g resulted in the viscosity of the solution being too high to obtain a homogeneous solution; therefore, Figure S2 only reveals the XRD patterns of the cases of 3 g and 5 g of PVP.
In Figure 3, the FTIR spectra of various ZnTiO 3 calcinated at 500, 600, 700, and 800 • C are shown. The absorption peak within the lower wavenumber region (480~750 cm −1 ) was mainly attributed to the Ti-O vibration. Zn-O-Ti groups were also observed at 735 cm −1 [49]. This suggested that crystalline ZnTiO 3 was formed when the calcination temperature was higher than 600 • C, which agreed well with the XRD results. Meanwhile, a broad peak near 3400 cm −1 was attributed to free water or adsorbed water, which was related to the OH stretching. There were several peaks between 845 and 1496 cm −1 , which indicated the existence of PVP residues [50]. Meanwhile, the absorption peak near 1620~1630 cm −1 belonged to the bending vibrations of Ti-OH [51,52], suggesting that ZnTiO 3 is hydrophilic. The peak intensities of the OH stretching and PVP residues gradually decreased with increasing calcination temperature, which indicated that the water content and PVP residues decreased with the increase in the calcination temperature. Subsequently, the effect of NH 4 OH and PVP addition on the FTIR spectra was also examined; however, the difference of the spectra was not significant, and no obvious signals of water or Ti-OH were observed, but Ti-O was, as shown in Figure S3. The SEM images of ZnTiO 3 -500 • C, ZnTiO 3 -600 • C, ZnTiO 3 -700 • C, and ZnTiO 3 -800 • C are shown in Figures 4a-d and 4e-h at magnifications of 10,000× and 30,000×, respectively. In Figure 4a-d, most of the ZnTiO 3 has a sharp rod shape. The one-dimensional morphology could be mainly attributed to the limitation of PVP during synthesis [53]. PVP could assemble concurrently on the surface of ZnTiO 3 , preventing ZnTiO 3 from growing along the radial direction of the rod. After calcination, PVP would be burned out, and mesoporous ZnTiO 3 rods could be formed. The size and length of the ZnTiO 3 rods were approximately 0.5~0.8 µm and 1~4 µm, respectively. As a demonstration, in Figure 4e-h, the cross-sectional shapes of the ZnTiO 3 rods are irregular or hexagonal. At a magnification of 30,000×, it is clearly observed that the surfaces of the ZnTiO 3 -500 • C and ZnTiO 3 -600 • C rods are relatively smooth and distinct. On the contrary, the surfaces of the ZnTiO 3 -700 • C and ZnTiO 3 -800 • C rods are relatively rough and thick, with some connections between the rods. Accordingly, we speculated that the increase in the calcination temperature caused smaller particles to be sintered together into larger particles due to the thermal migration [54]. With the increase of the amount of NH 4 OH, the length of the rod-shaped ZnTiO 3 increased. As shown in Figure S4, the length of the ZnTiO 3 -N750 rods is shorter than those of ZnTiO 3 -N1500 and ZnTiO 3 -N3000. Some of the ZnTiO 3 -N750 had not yet formed a rod shape. When the amount of NH 4 OH reached 1500 µL, most of the obtained ZnTiO 3 -N1500 maintained a rod shape with a length of around 1~4 µm. Compared with ZnTiO 3 -N1500, ZnTiO 3 -N3000 revealed rod shapes with longer lengths and a thicker size. The lengths of some ZnTiO 3 -N3000 could even exceed 9 µm. The possible reason was the increase in the amount of NH 4 OH, which would accelerate ZnTiO 3 formation in a short time during the synthesis. Such a greater amount of ZnTiO 3 caused continual growth simultaneously along the axial and radial directions despite the presence of PVP as the template. Moreover, the influence of the different additional amounts of PVP on the morphology was not apparent. In Figure S5, ZnTiO 3 -P3 only shows a slightly smaller cross-sectional length and width compared to ZnTiO 3 -P5. Generally, the concentration of PVP had a critical impact on the morphology due to the formation of micelles with different morphologies. There was a tendency for the aspect ratio of the crystals to became larger with the increase in the concentration of the PVP in the solution [55]. Besides, a higher PVP concentration might result in a more viscous solution, which could cause the instability of the nanofibers and generate more irregular structures [56]. This also proved that less PVP addition to the ZnTiO 3 -P3 resulted in less agglomeration or bead formation than the case of ZnTiO 3 -P5.
The information of the particle sizes observed from the SEM and element analysis results obtained by EDS is shown in Table 1. ZnTiO 3 photocatalysts calcinated at different temperatures resulted in the detection of O, Zn, and Ti, which proved the formation of ZnTiO 3 . Moreover, the Zn/Ti atomic ratios of ZnTiO 3 -600 • C and ZnTiO 3 -700 • C were close to 1:1, which was consistent with the main composition of ZnTiO 3 detected by XRD [53]. In comparison, the Zn/Ti atomic ratios of ZnTiO 3 -800 • C were close to 4:3, which might be derived from the multiple crystalline phases proven by XRD. On the other hand, the effects of NH 4 OH and PVP on the composition ratio were not noticeable, while the calcination temperature was fixed at 700 • C. Accordingly, the Zn/Ti atomic ratios of ZnTiO 3 samples were still close to 1:1. To determine the specific surface area and pore properties of the prepared ZnTiO 3 photocatalysts, N 2 adsorption-desorption analysis was carried out. In Figure 5a, the N 2 adsorption-desorption isotherms of all ZnTiO 3 photocatalysts are confirmed as type IV isotherms. As the calcinating temperature increased, the hysteresis loop gradually right-shifted to a higher relative pressure. Besides, the closed area of the hysteresis loop became narrower with the increase in the calcination temperature [57]. Figure 5b reveals the pore size distribution of the ZnTiO 3 photocatalysts as mainly from 2 to 50 nm, owing to their mesoporous properties. It is interesting to note that the addition of NH 4 OH and PVP did not affect the pore structure and specific surface area. The variance of the N 2 adsorption-desorption isotherms among the different ZnTiO 3 photocatalysts prepared with different amounts of NH 4 OH and PVP was not significant, as shown in Figure S6.
This suggested that the additional amounts of NH 4 OH and PVP might be saturated, so that the pore structures of the samples would not be affected. Table 2 brings together all the ZnTiO 3 samples. It can be observed that as the calcining temperature increased, the specific surface area value dropped sharply. The main reason was the shrinkage of the air gaps between the particles and the agglomeration of ZnTiO 3 during sintering [58].   The UV-Vis analysis of the ZnTiO 3 photocatalysts is illustrated in Figure 6. ZnTiO 3 -500 • C, ZnTiO 3 -600 • C, ZnTiO 3 -700 • C, and ZnTiO 3 -800 • C had their maximum absorption at the wavelengths of 300 nm, 286 nm, 286 nm, and 290 nm, respectively. This indicated that as the calcination temperatures increased, the UV-Vis absorption exposed a blue shift toward shorter wavelengths. By plotting the Tauc curve of (Ahυ) 2 vs. hυ [59], the bandgaps of the ZnTiO 3 photocatalysts were obtained as shown in the embedded graph of Figure 6, where A is the light absorbance and hυ is the energy of a photon. Hence, the energy gaps of ZnTiO 3 -500 • C, ZnTiO 3 -600 • C, ZnTiO 3 -700 • C, and ZnTiO 3 -800 • C were 3.54, 3.63, 3.72, and 3.75 eV, respectively. This indicated that the increasing calcination temperature led to increasing grain size; therefore, the band-gaps of ZnTiO 3 became large due to the quantum effect [60]. Meanwhile, the effects of NH 4 OH or PVP addition were not significant, as shown in Figure S7. The band-gaps of ZnTiO 3 with different NH 4 OH and PVP modifications were around 3.72 eV.
AMX degradation on ZnTiO 3 was implemented in two steps, dark adsorption and photocatalysis. Thus, Figure 7a,b expresses the concentration ratios of C/C 0 and C/C L0 , which indicate the performance of dark adsorption and photocatalysis, respectively. Before the LEDs were turned on, the AMX concentration was decreased due to the dark adsorption of AMX by the various ZnTiO 3 . In Figure 7a, ZnTiO 3 -500 • C could diminish the AMX by 20%. This might be attributed to the excellent specific surface area of ZnTiO 3 -500 • C compared to that of the other photocatalysts. After 50 min of dark adsorption, the adsorption reached an equilibrium, and the LEDs were turned on to conduct the photocatalytic reaction for 180 min. After light irradiation, the AMX concentration decreased with the irradiation time. Figure 7b reveals the photocatalytic degradation ratios to be about 47.9, 53.3, 63.8, and 45.4% for ZnTiO 3 -500 • C, ZnTiO 3 -600 • C, ZnTiO 3 -700 • C, and ZnTiO 3 -800 • C, respectively.  Although ZnTiO 3 -500 • C had the smallest band-gap and the highest specific surface area, its photocatalytic performance was not the best, which might be related to the poor crystalline form revealed by XRD. While the calcination temperature reached 600~700 • C, the cubic ZnTiO 3 became the dominant crystalline phase and showed a good photocatalytic activity. Moreover, ZnTiO 3 -700 • C showed better crystallinity in the cubic phase compared to ZnTiO 3 -600 • C, resulting in better photocatalytic performance. ZnTiO 3 -800 • C appeared as hexagonal ZnTiO 3 , leading to a decrease in photocatalytic performance, despite another study advocating that hexagonal ZnTiO 3 exhibited the strongest photocatalytic performance [45]. The poor photocatalytic performance of ZnTiO 3 -800 • C might be attributed to its extremely low specific surface area, which would lower the possibility of surface adsorption and reaction. Unfortunately, the tuning of various additional amounts of NH 4 OH and PVP would not affect the photocatalytic performance of AMX degradation, as shown in Figure S8a,b, respectively. Their similar performance might be due to their similar specific surface area and degree of crystallinity. Therefore, we could conclude that the crystallinity and specific surface area of ZnTiO 3 are the primary factors affecting the photocatalytic performance in the degradation of AMX. Supplementary Materials Table S1 provides a comparison of the activities of different photocatalysts in the degradation of AMX with the supporting data. This suggested that ZnTiO 3 -700 • C is a simple photocatalysts with the potential to degrade AMX under low-intensity UV light irradiation.
The possible photocatalysis mechanism is shown in Figure 8. After light irradiation, light-excited electrons and holes would react with O 2 and OH − to form·O 2 − and OH radicals to attack the β-lactam ring of AMX, followed by the concentration degradation, which could be observed by the absorbance of the AMX solution. Moreover, the photocatalytic degradation of AMX using P25 was conducted to serve as a standard case. Following a similar photocatalytic process, the AMX concentration slightly increased for the case of P25 during dark adsorption, as shown in Figure S9. This might result from the incomplete separation of AMX and P25, in which the absorption peaks, AMX: 227.5 nm and P25: <380 nm, overlapped and interfered mutually. Therefore, the separation process should be improved. Despite this issue, obviously, P25 reached the saturation of photocatalytic performance at around 90 min of light irradiation. Meanwhile, ZnTiO 3 -700 • C still maintained its effective photocatalytic activity. Accordingly, ZnTiO 3 -700 • C might be a better potential photocatalyst than P25. Moreover, the photocatalysts after the reaction were examined by XRD in order to check the stability of ZnTiO 3 -700 • C. We found that the main peaks of the XRD signals were the same, suggesting that the ZnTiO 3 -700 • C photocatalyst with high-temperature calcination is stable in the photocatalytic reaction. The changes of (a) C/C0 and (b) C/CL0 for various ZnTiO3 calcinated t 500~800 °C with th reaction time to degrade AMX. The possible photocatalysis mechanism is shown in Figure 8. After light irradiation light-excited electrons and holes would react with O2 and OHto form·O2and OH radicals to attack the β-lactam ring of AMX, followed by the concentration degradation which could be observed by the absorbance of the AMX solution. Moreover, th photocatalytic degradation of AMX using P25 was conducted to serve as a standard case Following a similar photocatalytic process, the AMX concentration slightly increased fo the case of P25 during dark adsorption, as shown in Figure S9. This might result from th incomplete separation of AMX and P25, in which the absorption peaks, AMX: 227.5 nm and P25: <380 nm, overlapped and interfered mutually. Therefore, the separation proces should be improved. Despite this issue, obviously, P25 reached the saturation o photocatalytic performance at around 90 min of light irradiation. Meanwhile, ZnTiO3-70 °C still maintained its effective photocatalytic activity. Accordingly, ZnTiO3-700 °C migh be a better potential photocatalyst than P25. Moreover, the photocatalysts after th reaction were examined by XRD in order to check the stability of ZnTiO3-700 °C. We foun that the main peaks of the XRD signals were the same, suggesting that the ZnTiO3-700 °C photocatalyst with high-temperature calcination is stable in the photocatalytic reaction. Furthermore, the photocatalytic degradation of various pollutants, including TC MB, and MO solutions and AMX in the presence of the ZnTiO3-700 °C photocatalyst, wa also performed. The concentrations of these pollutants were identified by UV-Vis at th Furthermore, the photocatalytic degradation of various pollutants, including TC, MB, and MO solutions and AMX in the presence of the ZnTiO 3 -700 • C photocatalyst, was also performed. The concentrations of these pollutants were identified by UV-Vis at the wavelengths of 357 nm for TC, 664 nm for MB, and 464 nm for MO. Their initial concentrations were individually set at 10 ppm. Afterward, dark adsorption was also examined for 50 min, followed by a 180 min photocatalytic reaction. Under 3 h of UV light irradiation, the concentration ratio decreased with time. As shown in Figure 9, the ZnTiO 3 -700 • C photocatalyst could remove TC, MB, MO, and AMX at overall ratios of 94.8, 57.4, 42.6, and 63.8%, respectively. Interestingly, ZnTiO 3 -700 • C demonstrated a good dark adsorption capability, which might be attributed to the attractive interaction among the five polar hydroxyl groups of TC and the surficial OH groups of ZnTiO 3 -700 • C. In Figure 9a, the TC concentration decreased up to 17% during the dark absorption in the presence of ZnTiO 3 -700 • C. After the appropriate dark adsorption, the ZnTiO 3 -700 • C photocatalyst subsequently exhibited outstanding photocatalytic performance on TC degradation, as shown in Figure 9b. On the contrary, AMX, MB, and MO have more nonpolar branches with carbon-containing chemical structures. These are not conducive to entering the active sites on the mesoporous catalysts, which reduces the photodegradation effect. In order to compare the photocatalytic performance of various ZnTiO 3 calcinated at 500-800 • C, the first-order kinetics model [61] was adopted to analyze the effect of ZnTiO 3 prepared at various calcination temperatures on the photocatalytic degradation of AMX, as shown in Figure 10a. As expected, ZnTiO 3 -700 • C exhibited the highest first-order rate constant of 0.0049 min −1 compared to the other ZnTiO 3 prepared at 500, 600, and 800 • C. Furthermore, the photocatalytic degradation of TC, MB, and MO using ZnTiO 3 -700 • C met the first-order kinetics model with rate constants of 0.0181, 0.0048, and 0.0031, respectively, the detailed information of which is exposure in Figure 10b and summarized in Table 3.

Synthesis of ZnTiO 3
The ZnTiO 3 photocatalysts were fabricated by the sol-gel process by adding a PVP template. First, 1.2 g of zinc acetate dihydrate, as the precursor of ZnTiO 3 , providing the Zn source, was dissolved in 100 mL of MEG to obtain Solution A. Similarly, 1.65 mL of TTIP served as the Ti source for ZnTiO 3 and was mixed into 100 mL of MEG to obtain Solution B. Solution C comprised 5 or 7 g PVP as the template, 750, 1500, or 3000 µL NH 4 OH as the alkaline source, and 120 mL MEG as the solvent. Next, Solution B was added gently into Solution A under continuous stirring, to obtain Solution D. Then, Solution D was mixed by stirring with Solution C in the constant-temperature reactor, which controlled the temperature at 50 • C using a water bath. The mixtures were continuously agitated at a fixed temperature of 50 • C for 20 h. Afterward, a milky white solution was formed according to the chemical reaction as follows: After this, the milky white solution was centrifuged at 3200× g for 30 min, followed by removing the supernatant and obtaining the sediment. The sediment was washed out from the bottom of the centrifuge tube with ethanol. The sediment and ethanol solution was poured into a flat-bottom flask and then concentrated at 60 • C in a vertical rotary vacuum concentrator. When there was no more solvent gathered in the collection bottle of the concentrator, 160 mL of ethanol was further added into the flask and well mixed with the precipitation to gain a thick white liquid. The thick liquid was subsequently placed in a crucible for vacuum drying at 60 • C for 10 h. The dried samples were then calcinated at the designed temperature of 500, 600, 700, and 800 • C for 3 h. Ultimately, the ZnTiO 3 photocatalysts were obtained by grinding these flaky and calcinated powders.

Characterizations
The crystal phase of the ZnTiO 3 photocatalysts was elucidated by the X-ray diffraction patterns (XRD) using D8 ADVANCE (Bruker, Germany). The wavelength of the CuKα target was 0.15406 nm, excited by 40 kV of working voltage. The scan rate was 4 • /min, and the range of the scanning angle was from 20 • to 80 • . The infrared absorption pattern was examined using Spectrum One (Perkin Elmer, USA, Waltham, MA, USA) as the Fourier transform infrared spectroscope (FTIR). The morphology and composition of the photocatalysts were observed by employing JSM-6701F (JEOL, Tokyo, Japan) as the scanning electron microscope (SEM), equipped with INCA X-act (Oxford Instruments, Oxford, U.K.) as the energy dispersion spectroscope (EDS). The N 2 adsorption-desorption isotherm curves of the photocatalysts at 77 K were inspected to reveal the property of the specific surface area using a physisorption analyzer (ASAP 2020 PLUS, USA) and Brunauer-Emmett-Teller (BET) method. The diffuse reflectance UV-Vis spectra at 200~800 nm were measured by applying a UV-Vis spectrophotometer (V-670, JASCO, Pfungstadt, Germany). The thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were also performed using the simultaneous thermogravimetric analyzer (SDT 2960, TA Instruments, USA).

Photocatalytic AMX Degradation
The removal of AMX was achieved with two processes, dark adsorption and photocatalytic degradation. Firstly, 50 mg of the ZnTiO 3 photocatalysts was suspended with continuous stirring in a 100 mL AMX solution, with an initial concentration of 10 ppm (C 0 ). Then, the dark adsorption process of AMX was carried out on the ZnTiO 3 photocatalysts covered by a black box. During the dark adsorption, 3 mL AMX solution was taken out at the 30th, 40th, and 50th minute for sampling. After that, the adsorption and desorption were regarded as reaching an equilibrium, and the concentration (C L0 ) was noted. Subsequently, UV LEDs (8 W × 4) within a wavelength range of 280~320 nm (G8T5E, Sankyo-Denki, Tokyo, Japan) were turned on to irradiate the AMX solution to facilitate the photocatalytic degradation. After 30 min of irradiation, the 3 mL AMX solution was sampled to inspect the concentration (C) at the 30th, 60th, 90th, 120th, 150th, and 180th minute. Then, each 3 mL sample of the AMX solution was centrifuged at 3200× g for 15 min to remove the ZnTiO 3 sediment. The supernatant solution was detected by employing a UV-Vis spectroscope to identify the concentration of AMX, which showed a characterized absorption peak at~228 nm in the spectrum. This wavelength is the absorption of the βlactam ring of the pharmacophore of AMX [62], and the β-lactam ring would be destroyed by the photocatalytic reaction.

Conclusions
In this study, ZnTiO 3 photocatalysts were successfully fabricated using the sol-gel process. The influence of the calcination temperature, NH 4 OH addition, and PVP addition during the synthesis on their characteristics was discussed. Furthermore, the photocatalytic performances of the ZnTiO 3 photocatalysts in decomposing emerging pollutants, such as AMX, TC, MB, and MO, were also examined. The results showed that the different calcination temperatures significantly affected the photocatalytic performance in the degradation of the AMX. Under a low calcination temperature of 500 • C, the resulting photocatalyst showed a poor crystalline phase. As the calcination temperature increased to 600 • C, an obvious cubic ZnTiO 3 phase appeared. As the calcination temperature reached 700 • C, the photocatalysts had a sharp cubic ZnTiO 3 phase and showed an excellent photocatalytic degradation of the AMX. However, some hexagonal ZnTiO 3 appeared while calcinating the samples at 800 • C, leading to a smaller specific surface area and less satisfactory photocatalytic performance. Based on setting the calcination temperature at 700 • C, no matter the amount of NH 4 OH or PVP added during the synthesis, the effect on the pore properties and crystal phase was not significant. This indicated that the photocatalytic performance was related to the pore properties and crystal phase of the cubic ZnTiO 3 .
Moreover, the synthesized ZnTiO 3 -700 • C met the type IV isotherm and revealed the mesoporous property with the pore size distribution from 2~50 nm. Accordingly, the ZnTiO 3 -700 • C photocatalyst was employed to remove different organic pollutants, including TC, MB, and MO, under UVB light irradiation. Among these pollutants, the degradation effect of TC in the presence of ZnTiO 3 -700 • C could reach 94.8%. It was speculated that there are multipolar OH groups on TC, which would attract the OH groups on the surface of the ZnTiO 3 -700 • C photocatalyst. Subsequently, it is easier for TC to be adsorbed into the pores of the ZnTiO 3 -700 • C photocatalyst, followed by the contact of the active sites with the photocatalysts. To sum up, the calcination temperature was the primary factor influencing the pore structure and crystal phase of the ZnTiO 3 -700 • C photocatalyst, which showed great potential for the photocatalytic degradation of AMX, TC, MO, and MB under the irradiation of low-intensity UV light.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/catal11070854/s1, Figure S1: XRD patterns of ZnTiO 3 photocatalysts (synthesized by adding 5 g PVP) under the calcination temperature of 700 • C with various amounts of NH 4 OH addition (750, 1500, and 3000 µL); Figure S2. The XRD patterns of ZnTiO 3 photocatalysts prepared by adding 3 g and 5 g PVP under the calcination temperature of 700 • C with NH 4 OH addition of 1500 µL; Figure S3. FTIR spectra of various ZnTiO 3 prepared by various conditions; Figure S4. (a)-(d) and (d)-(f) the SEM images at the magnification of 10,000× and 30,000× of ZnTiO 3 -N750, ZnTiO 3 -N1500, and ZnTiO 3 -3000; Figure S5. the SEM images at the magnification of 30,000× of (a) ZnTiO 3 -P3 and (b) ZnTiO 3 -P5; Figure S6. N2 adsorption-desorption isotherms of ZnTiO 3 prepared various NH 4 OH and PVP modification; Figure S7. UV-Vis result of ZnTiO 3 prepared by various amounts of NH 4 OH and PVP modification; Figure S8. The changes of C/C L0 with reaction time (a) for various amounts of NH 4 OH addition and (b) for various amounts of PVP addition; Figure S9. The changes of C/C 0 with reaction time by using P25 and ZnTiO 3 -700 • C; Table S1. the comparison of activities over different photocatalysts to degrade AMX.