Construction of 2D/2D Mesoporous WO3/CeO2 Laminated Heterojunctions for Optimized Photocatalytic Performance

Photocatalytic elimination of antibiotics from the environment and drinking water is of great significance for human health. However, the efficiency of photoremoval of antibiotics such as tetracycline is severely limited by the prompt recombination of electron holes and slow charge migration efficacy. Fabrication of low-dimensional heterojunction composites is an efficient method for shortening charge carrier migration distance and enhancing charge transfer efficiency. Herein, 2D/2D mesoporous WO3/CeO2 laminated Z-scheme heterojunctions were successfully prepared using a two-step hydrothermal process. The mesoporous structure of the composites was proved by nitrogen sorption isotherms, in which sorption-desorption hysteresis was observed. The intimate contact and charge transfer mechanism between WO3 nanoplates and CeO2 nanosheets was investigated using high-resolution transmission electron microscopy and X-ray photoelectron spectroscopy measurements, respectively. Photocatalytic tetracycline degradation efficiency was noticeably promoted by the formation of 2D/2D laminated heterojunctions. The improved photocatalytic activity could be attributed to the formation of Z-scheme laminated heterostructure and 2D morphology favoring spatial charge separation, confirmed by various characterizations. The optimized 5WO3/CeO2 (5 wt.% WO3) composites can degrade more than 99% of tetracycline in 80 min, achieving a peak TC photodegradation efficiency of 0.0482 min−1, which is approximately 3.4 times that of pristine CeO2. A Z-scheme mechanism is proposed for photocatalytic tetracycline by from WO3/CeO2 Z-scheme laminated heterojunctions based on the experimental results.


Introduction
Excessive current use of antibiotics has led to the wide distribution of antibiotic residues in the global environment, even in drinking water, which adversely affects human health [1,2]. Thus, the removal of antibiotics is of great importance for life on earth and the sustainable development of human society. Various technologies such as physical purification, chemical degradation, biological degradation, and catalytic elimination have been developed to remove antibiotics from polluted water [3,4]. Among these methods, photocatalysis is one of the most promising routes for residual antibiotic elimination because it is a solar-driven process that does not consume energy [5]. However, the application of photocatalytic antibiotic removal in natural conditions is hindered by the low efficiency of the photocatalytic process. Therefore, research on improving photocatalytic efficiency is significant for antibiotic removal.
Semiconductor-based photocatalytic technology can create electrons and holes under light illumination, which in turn produce superoxide radicals (O 2 − ) and hydroxyl radicals (·OH) in the presence of water and oxygen [6]. These reactive oxygen radicals and photogenerated holes are capable of oxidizing antibiotics into CO 2 and H 2 O [7]. Various types of semiconductors, including TiO 2 -based materials [8], bismuth-based photocatalysts [9], g-C 3 N 4 -based systems [10], heterojunction and surface plasmon resonance enhanced photocatalysts [11], and other materials have been explored for photocatalytic removal of antibiotics [12]. Among these photocatalysts, WO 3 has drawn much attention for visible light pollutant removal because of its narrow band gap [13]. Kim et al. investigated the influence of different inorganic oxidants on the photocatalytic performance of WO 3 for visible light degradation of 4-chlorophenol, demonstrating charge separation enhancement by oxyanions [14]. Nguyen et al. reported that amoxicillin could be degraded by WO 3 under simulated solar irradiation, which followed pseudo-first-order kinetics [15]. In addition, near-infrared light photocatalytic removal of methylene blue was achieved on a longitudinally grown WO 3 nanoparticle surface [16]. However, the photocatalytic activity of WO 3 is severely restricted by the prompt charge recombination and the lack of superoxide radicals because of its low conduction band position [17]. These drawbacks can be overcome by the formation of heterojunctions with metals or other semiconductors with high conduction band levels. The electron-hole pair separation and visible light absorption were much improved by the loading of Ag on WO 3 surfaces, producing excellent sulfanilamide degradation activity [18]. The formation of WO 3 /carbon nanotubes was also reported to effectively suppress the recombination of charge carriers and promote sonophotocatalytic tetracycline removal efficiency [19]. Oxidizing species were proved to be OH and h + without O 2 − in that research [19]. CeO 2 is an attractive candidate for constructing heterojunction photocatalysts with other semiconductors because it is chemically stable, nontoxic, and has proper band positions, where the conduction band is positioned at ca. −0.78 eV (vs. NHE) [19]. Oxygen vacancies were proved to be present on the CeO 2 surface, thereby enhancing charge separation and transfer efficiency [20]. Recently, various heterojunctions such as SnIn 4 S 8 /CeO 2 [21], Ni/CeO 2 [22], CdS/CeO 2 [23], and 3D/0D Cu 3 SnS 4 /CeO 2 heterojunctions [24] have been reported to show significant improvement in photocatalytic performance. Bahadoran et al. reported an S-scheme WO 3 /CeO 2 heterojunction with enhanced visible light photocatalytic sulfamerazine degradation performance [25]. Moreover, the photodegradation activity of indigo carmine dye was much improved by the formation of type II WO 3 /CeO 2 heterojunction nanoparticles [26].
Because of its wide band gap of ca. 3.2 eV, CeO 2 photocatalysts can only be activated by UV irradiation, but possess decent photocatalytic activity for pollutant degradation with the generation of oxidizing radicals including O 2 − and OH [26]. In contrast, WO 3 has a narrow band gap of 2.4-2.8 eV, thus exhibiting potential applications in visible-light-driven photocatalytic reactions [27,28]. However, the photogenerated electrons and holes on the WO 3 surface demonstrate insufficient redox ability due to its band positions [28]. Therefore, the addition of small amounts of WO 3 to CeO 2 photocatalysts could extend the light response range from ultraviolet to the visible light region without significant deactivation. Although type II heterojunctions have been investigated for indigo carmine dye removal [26], to our knowledge, the construction of 2D/2D mesoporous WO 3 /CeO 2 Z-scheme laminated heterojunctions with promoted photocatalytic tetracycline degradation activity has never been reported. We fabricated 2D/2D mesoporous WO 3 /CeO 2 Z-scheme laminated heterojunctions via a facile two-step hydrothermal process. The transfer efficiency of the photoinduced charges was significantly enhanced by the low-dimension morphology and the formation of the Z-scheme laminated heterostructure. The improved charge transfer efficiency and suppressed electron hole recombination were confirmed by various characterizations, including electrochemical impedance spectra, transient photocurrent responses, surface photovoltage spectroscopy, and photoluminescence.

Synthesis of Mesoporous CeO 2 Nanosheets, WO 3 Nanoplates, and WO 3 /CeO 2 Composites
The 2D CeO 2 nanosheets were synthesized according to our previously reported method [29]. Typically, 10 mmol glucose was added to 80 mL deionized water and magnetically stirred for 30 min. Subsequently, 15 mmol C 3 H 5 NO and 5 mmol Ce(NO 3 ) 3 ·6H 2 O were dissolved in the above solution under magnetic stirring for 1 h, and 3 mL NH 3 ·H 2 O (28 wt%) was added drop by drop to the mixture and stirred for 5 h, creating a dark brown gel mixture. The mixture was transferred into a Teflon-lined autoclave and maintained at 180 • C for 72 h. The solid products were collected and washed via centrifugation with water and ethanol after reaction. CeO 2 nanosheets were obtained via thermal treatment at 600 • C for 6 h under a nitrogen atmosphere and subsequent calcination at 400 • C for 4 h under an air atmosphere. The 2D WO 3 nanoplates were fabricated using a slightly modified procedure [30]. In a typical procedure, 0.06 g Na 2 WO 4 ·2H 2 O was first added to 30 mL water. Subsequently, 0.013 g C 6 H 8 O 7 ·H 2 O and 0.04 g C 6 H 12 O 6 were added to the above solution. After 20 min magnetic stirring, 3 mL HCl (6M) was added dropwise to the mixture, and the mixture was transferred to a 50 mL autoclave and heated at 120 • C for 24 h. After the reaction, the product was collected and washed with anhydrous ethanol and deionized water several times by centrifugation. The sample was then heated to 400 • C at a heating rate of 5 • C min −1 and maintained for 2 h to obtain mesoporous WO 3 nanoplates. Mesoporous WO 3 /CeO 2 heterojunctions were prepared by a two-step hydrothermal method. The synthetic procedure of WO 3 /CeO 2 is presented in Scheme 1. First, 0.06 g Na 2 WO 4 ·2H 2 O and 2.3 g prepared CeO 2 were added to 30 mL water to form a transparent solution. Subsequently, 0.013 g C 6 H 8 O 7 ·H 2 O and 0.04 g C 6 H 12 O 6 were added to the above solution. After 20 min magnetic stirring, 3 mL HCl (6M) was added drop by drop to the mixture, and the solution was transferred to a 50 mL autoclave and maintained at 120 • C for 24 h. After the reaction, the product was centrifuged and washed with anhydrous ethanol and deionized water several times to remove surface impurities. The sample was then heated to 400 • C at a heating rate of 5 • C min −1 and kept for 2 h to improve the crystallinity. A yellow powder was obtained, denoted as xWO 3 /CeO 2 (x%: mass ratio for WO 3 ). The 3WO 3 /CeO 2 , 5WO 3 /CeO 2 , and 7WO 3 /CeO 2 heterojunctions were obtained using 1.39 g, 2.3 g, and 3.25 g CeO 2, respectively, during the synthetic procedure.

Synthesis of Mesoporous CeO2 Nanosheets, WO3 Nanoplates, and WO3/CeO2 Composites
The 2D CeO2 nanosheets were synthesized according to our previously reported method [29]. Typically, 10 mmol glucose was added to 80 mL deionized water and magnetically stirred for 30 min. Subsequently, 15 mmol C3H5NO and 5 mmol Ce(NO3)3·6H2O were dissolved in the above solution under magnetic stirring for 1 h, and 3 mL NH3·H2O (28 wt%) was added drop by drop to the mixture and stirred for 5 h, creating a dark brown gel mixture. The mixture was transferred into a Teflon-lined autoclave and maintained at 180 °C for 72 h. The solid products were collected and washed via centrifugation with water and ethanol after reaction. CeO2 nanosheets were obtained via thermal treatment at 600 °C for 6 h under a nitrogen atmosphere and subsequent calcination at 400 °C for 4 h under an air atmosphere. The 2D WO3 nanoplates were fabricated using a slightly modified procedure [30]. In a typical procedure, 0.06 g Na2WO4·2H2O was first added to 30 mL water. Subsequently, 0.013 g C6H8O7·H2O and 0.04 g C6H12O6 were added to the above solution. After 20 min magnetic stirring, 3 mL HCl (6M) was added dropwise to the mixture, and the mixture was transferred to a 50 mL autoclave and heated at 120 °C for 24 h. After the reaction, the product was collected and washed with anhydrous ethanol and deionized water several times by centrifugation. The sample was then heated to 400 °C at a heating rate of 5 °C min −1 and maintained for 2 h to obtain mesoporous WO3 nanoplates. Mesoporous WO3/CeO2 heterojunctions were prepared by a two-step hydrothermal method. The synthetic procedure of WO3/CeO2 is presented in Scheme 1. First, 0.06 g Na2WO4·2H2O and 2.3 g prepared CeO2 were added to 30 mL water to form a transparent solution. Subsequently, 0.013 g C6H8O7·H2O and 0.04 g C6H12O6 were added to the above solution. After 20 min magnetic stirring, 3 mL HCl (6M) was added drop by drop to the mixture, and the solution was transferred to a 50 mL autoclave and maintained at 120 ℃ for 24 h. After the reaction, the product was centrifuged and washed with anhydrous ethanol and deionized water several times to remove surface impurities. The sample was then heated to 400 ℃ at a heating rate of 5 ℃ min −1 and kept for 2 h to improve the crystallinity. A yellow powder was obtained, denoted as xWO3/CeO2 (x%: mass ratio for WO3). The 3WO3/CeO2, 5WO3/CeO2, and 7WO3/CeO2 heterojunctions were obtained using 1.39 g, 2.3 g, and 3.25 g CeO2, respectively, during the synthetic procedure.

Characterizations
X-ray diffraction (XRD) measurements were performed on a Bruker D8 Advance diffractometer (Bruker, Karlsruhe, Germany) using Cu Kα radiation (λ = 1.5406 Å, 40 kV, 40 mA). The step size and scan range were 0.02 • , and 5-80 • , respectively. The UV-vis absorption spectra were obtained using a UV-visible spectrophotometer (Shimadzu UV-2550, Shimadzu, Tokyo, Japan). A PerkinElmer Lambda 950 UV-vis spectrophotometer (PerkinElmer, Waltham, MA, USA) was used to measure the diffuse reflectance spectrum (DRS) in the wavelength range of 200-1800 nm. Materials morphology was analyzed by scanning electron microscopy (SEM) (Hitachi S-4800, Hitachi, Tokyo, Japan) and transmission electron microscopy (TEM) (JEOLJEM-2100, Jeol, Tokyo, Japan). X-ray photoelectron spectroscopy (XPS) measurements were conducted on a Kratos-Axis ULTRA DLD electron spectrometer with an AlKα (1253.6 eV) excitation source. The results were calibrated using the standard C 1s peak (284.6 eV). Nitrogen sorption measurements were performed with a TriStar 3020 ii automatic analyzer. The Brunaeur-Emmet-Teller (BET) equation and the adsorption branch of BET isotherms were used to obtain the specific surface area and pore size distribution, respectively. The surface photovoltage spectroscopy (SPS) was tested on a homemade instrument in the wavelength range of 200-600 nm. The photoluminescence spectra (PL) were measured using a PELS 55 instrument with a selected excitation wavelength of 275 nm. The photoelectric characteristics of the samples were analyzed on a Princeton VersaSTAT workstation (Princeton Applied Research, Oak Ridge, TN, USA) with a three-electrode system, using a catalyst as the photoanode, platinum as the cathode, Ag/AgCl as the reference electrode, and 0.5 M Na 2 SO 4 solution as the electrolyte, respectively.

Photocatalytic Activity Measurements
The photocatalytic activity of CeO 2 and WO 3 /CeO 2 were measured in the tetracycline (TC) degradation process. First, 100 mg of CeO 2 or WO 3 /CeO 2 was added to 100 mL TC solution (10 mg L −1 ). The suspension was kept in the dark for 20 min and analyzed to evaluate the adsorption properties of the sample. Subsequently, the mixture was illuminated under a 300 W Xe lamp (PerfectLight, Beijing, China) with an AM 1.5G filter. The suspension was sampled every 20 min, 5 times in total. The TC concentration was determined by UV-vis spectroscopy at the wavelength of 357 nm.

Morphology and Microstructure
The crystal structures of CeO 2 and WO 3 /CeO 2 were analyzed by X-ray diffraction (XRD). Figure 1a and Figure S1 show the XRD profiles of the CeO 2 and WO 3 /CeO 2 heterojunction photocatalysts. The characteristic peaks of CeO 2 (JCPDS No. 34-0394) and WO 3 (JCPDS No. 20-1323) can be identified in the samples of CeO 2 and WO 3 [31,32]. The characteristic peaks of WO 3 are not observed in the XRD patterns of 3WO 3 /CeO 2 and 5WO 3 /CeO 2 because the loading amounts of WO 3 are too low to be detected. The characteristic peaks of both CeO 2 and WO 3 can be identified in the XRD pattern of 7WO 3 /CeO 2 , indicating the successful formation of WO 3 /CeO 2 composites. The subsequent evaluation of photocatalytic TC degradation performance revealed that 5WO 3 /CeO 2 (5 wt%) demonstrated the best photocatalytic activity. Thus, 5WO 3 /CeO 2 was used for all the characterizations. The light absorption properties of the as-made samples were analyzed by UV-vis spectroscopy, as shown in Figure 1b. CeO 2 exhibited strong absorption in the ultraviolet range (200-400 nm) but limited visible light absorption ability. WO 3 showed strong absorption in the visible light range, thereby enhancing the light absorption ability of the WO 3 /CeO 2 heterojunctions. Figure S2 presents the UV-vis diffuse reflection spectra of the samples with different WO 3 contents. The visible light absorption ability of WO 3 /CeO 2 increases with the WO 3 content. Figure 1c,d describes the N 2 sorption isotherms and the corresponding BJH pore size distribution plots of CeO 2 , WO 3 , and WO 3 /CeO 2 . The BET surface area, average pore size, and total pore volume of the samples are presented in Table S1. Typical type IV hysteresis loops can be observed in all three samples, indicating that mesopores are formed in the synthesized samples, which is consistent with the average pore size data. Materials with mesoporous pores offer enhanced accessibility and superior mass transfer [33,34], thus having the possibility to enhance the photocatalytic redox reac-Nanomaterials 2023, 13, 1798 5 of 15 tion. The average crystallite diameters (D) of the as-prepared samples were obtained by the Scherrer equation [35]: λ, β, and θ are the X-ray wavelength, the full width at half maximum of the peak, and the Bragg's diffraction angle of the peak, respectively. The crystallite size, surface area, pore volume, and main pore size are presented in Table 1.  The morphology and microscopic structure of WO3, CeO2, and WO3/CeO2 were studied by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), as demonstrated in Figure 2. The CeO2 microspheres (diameter: ca. 4.5 µm) consist of uniform CeO2 nanosheets, providing a 2D platform for the formation of laminated heterojunctions. WO3 nanoplates are uniformly dispersed on the surface of CeO2 nanosheets after combination with CeO2 under hydrothermal conditions. Obvious lattice fringes of λ, β, and θ are the X-ray wavelength, the full width at half maximum of the peak, and the Bragg's diffraction angle of the peak, respectively. The crystallite size, surface area, pore volume, and main pore size are presented in Table 1. The morphology and microscopic structure of WO 3 , CeO 2 , and WO 3 /CeO 2 were studied by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), as demonstrated in Figure 2. The CeO 2 microspheres (diameter: ca. 4.5 µm) consist of uniform CeO 2 nanosheets, providing a 2D platform for the formation of laminated heterojunctions. WO 3 nanoplates are uniformly dispersed on the surface of CeO 2 nanosheets after combination with CeO 2 under hydrothermal conditions. Obvious lattice fringes of CeO 2 and WO 3 can be identified by a high-resolution TEM image (Figure 2f), indicating the high crystallinity of the formed WO 3 /CeO 2 composites. The spacing of 0.38 nm is ascribed to the (002) plane of WO 3 , while the lattice fringes of 0.317 nm and 0.286 nm are attributed to the (111) plane and (200) plane of CeO 2 , respectively. The microstructure of the WO 3 /CeO 2 composites revealed by TEM is in good agreement with the XRD data. In addition, the high-resolution image of WO 3 /CeO 2 shows that the lattice fringes of WO 3 and CeO 2 are in close contact with each other, suggesting the successful construction of heterojunction photocatalysts.
CeO2 and WO3 can be identified by a high-resolution TEM image (Figure 2f), indicating the high crystallinity of the formed WO3/CeO2 composites. The spacing of 0.38 nm is ascribed to the (002) plane of WO3, while the lattice fringes of 0.317 nm and 0.286 nm are attributed to the (111) plane and (200) plane of CeO2, respectively. The microstructure of the WO3/CeO2 composites revealed by TEM is in good agreement with the XRD data. In addition, the high-resolution image of WO3/CeO2 shows that the lattice fringes of WO3 and CeO2 are in close contact with each other, suggesting the successful construction of heterojunction photocatalysts.  Figure 3 presents the XPS spectra of CeO2 and WO3/CeO2 composites. All the binding energy positions in the XPS measurements were calibrated using C 1s as the reference at 284.8 eV. The full XPS survey spectra of CeO2 and WO3/CeO2 composites are illustrated in Figure 3a, while the analysis of the high-resolution XPS spectra for W, Ce, and O are presented in Figure 3b-d. The presence of W, O, and Ce elements in the XPS survey spectrum of WO3/CeO2 provides evidence that the WO3/CeO2 composites were successfully constructed without impurities. The binding energy at 35.7 and 37.8 eV can be ascribed to W 4f7/2 and W 4f5/2, confirming the presence of W 6+ in the WO3/CeO2 composites [36]. Upon the formation of the WO3/CeO2 composites, the above two peaks of W 4f shifted to lower energy levels, indicating electron transfer occurs because of the energy level difference. Thus, the electron density on the surface of WO3 in the composites increased. As shown in Figure 3c, two kinds of spin-orbit interaction, including Ce 3d3/2 (XPS peaks are denoted as u or u-derivatives) and Ce 3d5/2 (XPS peaks are denoted as v or v-derivatives), can be clearly observed in the high-resolution Ce 3d XPS spectra. The peaks located at 901.  [25,29]. Moreover, the Ce 3d, and O 1s main peaks of the WO3/CeO2 composites demonstrate migration after the  Figure 3 presents the XPS spectra of CeO 2 and WO 3 /CeO 2 composites. All the binding energy positions in the XPS measurements were calibrated using C 1s as the reference at 284.8 eV. The full XPS survey spectra of CeO 2 and WO 3 /CeO 2 composites are illustrated in Figure 3a, while the analysis of the high-resolution XPS spectra for W, Ce, and O are presented in Figure 3b-d. The presence of W, O, and Ce elements in the XPS survey spectrum of WO 3 /CeO 2 provides evidence that the WO 3 /CeO 2 composites were successfully constructed without impurities. The binding energy at 35.7 and 37.8 eV can be ascribed to W 4f 7/2 and W 4f 5/2 , confirming the presence of W 6+ in the WO 3 /CeO 2 composites [36]. Upon the formation of the WO 3 /CeO 2 composites, the above two peaks of W 4f shifted to lower energy levels, indicating electron transfer occurs because of the energy level difference. Thus, the electron density on the surface of WO 3 in the composites increased. As shown in Figure 3c, two kinds of spin-orbit interaction, including Ce 3d 3/2 (XPS peaks are denoted as u or u-derivatives) and Ce 3d 5/2 (XPS peaks are denoted as v or v-derivatives), can be clearly observed in the high-resolution Ce 3d XPS spectra. The peaks located at 901.  [25,29]. Moreover, the Ce 3d, and O 1s main peaks of the WO 3 /CeO 2 composites demonstrate migration after the construction of composites, suggesting the change of chemical environment and the existence of strong interaction between WO 3 and CeO 2 .  The energy band gaps (Eg) for WO3 and CeO2 were obtained by using the Kubella-Munk formula (αhv = A(hv-Eg) η/2 ), where A, α, h, and η, are the constant, the absorption coefficient close to the absorption edge, the Planck constant, and the coefficient of the semiconductor, respectively. The value of coefficient η of CeO2 and WO3 is 4, since both semiconductors have indirect band gaps [37,38]. The Eg values of the as-synthesized CeO2 and WO3 samples were determined to be 2.62 eV and 2.58 eV, respectively, by the Kubella-Munk formula plots derived from the UV-Vis spectra, as presented in Figure 4a,c. The valence band (VB) maximum positions of the materials were normally measured by the valence band XPS spectra [39]. As displayed in Figure 4b,d, the surface valence band edges of CeO2 and WO3 were estimated to be 2.1 eV and 2.67 eV, respectively, by drawing the tangent line [40]. Based on the empirical equation ECB = EVB − Eg (ECB, EVB, and Eg are the conduction band, valence band positions, and the band gap of semiconductors, respectively), and the results of the obtained band gaps and valence band maximum positions, the conduction band (CB) minimum positions of CeO2 and WO3 are calculated to be −0.52 eV and 0.09 eV, respectively. The energy scale is normally demonstrated in volts using the normal hydrogen electrode (NHE) as a reference [41]. The Fermi levels of WO3 and CeO2 were reported to be 0.57 and −0.353 eV, respectively [42,43]. It should be noted that the band bending direction in the interface of composites could be significantly influenced by the Fermi level of the two semiconductors [43]. The charge transfer process of WO3/CeO2 heterojunction is assumed to be a Z-scheme mechanism, which is proved and analyzed below. Therefore, the band edge of CeO2 with a high Fermi level is bent upward while the The energy band gaps (E g ) for WO 3 and CeO 2 were obtained by using the Kubella-Munk formula (αhv = A(hv − E g ) η/2 ), where A, α, h, and η, are the constant, the absorption coefficient close to the absorption edge, the Planck constant, and the coefficient of the semiconductor, respectively. The value of coefficient η of CeO 2 and WO 3 is 4, since both semiconductors have indirect band gaps [37,38]. The E g values of the as-synthesized CeO 2 and WO 3 samples were determined to be 2.62 eV and 2.58 eV, respectively, by the Kubella-Munk formula plots derived from the UV-Vis spectra, as presented in Figure 4a,c. The valence band (VB) maximum positions of the materials were normally measured by the valence band XPS spectra [39]. As displayed in Figure 4b,d, the surface valence band edges of CeO 2 and WO 3 were estimated to be 2.1 eV and 2.67 eV, respectively, by drawing the tangent line [40]. Based on the empirical equation E CB = E VB − E g (E CB , E VB , and E g are the conduction band, valence band positions, and the band gap of semiconductors, respectively), and the results of the obtained band gaps and valence band maximum positions, the conduction band (CB) minimum positions of CeO 2 and WO 3 are calculated to be −0.52 eV and 0.09 eV, respectively. The energy scale is normally demonstrated in volts using the normal hydrogen electrode (NHE) as a reference [41]. The Fermi levels of WO 3 and CeO 2 were reported to be 0.57 and −0.353 eV, respectively [42,43]. It should be noted that the band bending direction in the interface of composites could be significantly influenced by the Fermi level of the two semiconductors [43]. The charge transfer process of WO 3 /CeO 2 heterojunction is assumed to be a Z-scheme mechanism, which is proved and analyzed below. Therefore, the band edge of CeO 2 with a high Fermi level is bent upward while the band edge of WO 3 with a low Fermi level is bent downward, because the electrons migrate from the high Fermi level to the low Fermi level [43]. A built-in electric field in the heterojunction interface could be formed because of the curved band edges and charge Nanomaterials 2023, 13, 1798 8 of 15 transfer equilibrium between CeO 2 and WO 3 [44]. Thus, the WO 3 /CeO 2 heterojunction is probably fabricated with a staggered band alignment, as shown in Figure 4e. 1 Figure 4. Band gap energies (a,c) and valance band XPS spectra (b,d) and band alignments (e) of WO 3 and CeO 2 , respectively.

Photocatalytic Activity
The photocatalytic activity of different samples in the tetracycline degradation process was evaluated under AM 1.5G illumination. Photocatalytic TC removal efficiency in the presence of the heterojunction photocatalysts with different amounts of WO 3 was measured. As shown in Figure 5a, the photocatalytic TC degradation efficiency of CeO 2 was much improved after the construction of the heterostructure with WO 3 . The photocatalytic efficiency for TC degradation increases with the amount of WO 3 . Under light illumination, 5WO 3 /CeO 2 (5 wt.% WO 3 ) heterojunctions exhibit the highest photocatalytic efficiency, at 99.1% within 80 min for TC removal. The enhanced photocatalytic efficiency of the heterojunctions compared to pristine WO 3 or CeO 2 is mainly attributed to the formation of the laminated heterostructure and the mesoporous structure. The heterojunction interface could inhibit the photogenerated charge recombination rate, thereby facilitating the charge transfer and the generation of oxidizing radicals. The mesoporous structure presented a relatively high absorption capacity for tetracycline. Moreover, a 2D structure could shorten the distance of charge transfer in the heterojunction interface. However, 7WO 3 /CeO 2 (7 wt.% WO 3 ) composites showed a decreased TC removal efficiency (97.9% within 80 min) under identical reaction conditions. Despite the enhanced charge separation and transfer efficiency at the interface of the heterojunctions, excess amounts of WO 3 would result in the coverage of active sites of CeO 2 , thus decreasing the photocatalytic activity. Therefore, 5WO 3 /CeO 2 composites present the proper tradeoff between enhanced charge separation efficiency and active site coverage. The photocatalytic TC removal efficiency of individual WO 3 has been measured to be 64.9% within 80 min, which is much lower than that of CeO 2 . As shown in Figure 5b, the photocatalytic TC degradation rate of CeO 2 decreased quickly with reaction time because of the prompt photoinduced electron-hole combination, reaching a final removal efficiency of 83.9%. Owing to the reduced electron-hole recombination rate, 5WO 3 /CeO 2 composites demonstrated accelerated TC removal efficiency compared to pristine CeO 2 . As illustrated in Figure 5c, the photocatalytic TC degradation rates were well-matched with a pseudo-first-order kinetic model. The kinetic constant (K) of 5WO 3 /CeO 2 heterojunctions for TC photodegradation achieved 0.0482 min −1 , which was approximately 3.4 times that of pristine CeO 2 . To verify the outcome of the TC degradation performance, the total organic carbon (TOC) concentration after the reaction in the presence of WO 3 /CeO 2 was measured using a carbon analyzer (TOC 5000A Shimadzu), as illustrated in Table S2. The TOC concentration after the reaction photocatalyzed by 5WO 3 /CeO 2 heterojunctions was determined to be 1.6 mg L −1 , which is in agreement with the UV-Vis spectroscopy measurements.
The reusability of the optimized photocatalysts (5WO 3 /CeO 2 ) was evaluated in 10 cycling tests, and the TC photoremoval efficiency is presented in Figure 5d. The heterojunction composites exhibited stable TC elimination efficiency under light illumination, which was almost 100% maintained after ten consecutive cycling measurements, suggesting that the as-synthesized composites possess relatively high photocatalytic stability in the TC elimination process. Table 2 presents the photocatalytic tetracycline degradation performance of 2D/2D mesoporous WO 3 /CeO 2 compared with reported CeO 2 -based materials. It can be clearly seen that the as-synthesized WO 3 /CeO 2 laminated heterojunctions exhibit superior photocatalytic TC removal activity compared to other photocatalysts. Moreover, the cycling tests of photocatalytic TC degradation in this study were performed 10 times without apparent photocatalyst deactivation, while only 4 or 5 cycling tests were conducted in the previous studies.  To get insights into the photocatalytic process of TC removal, photoelectrochemical measurements were conducted and presented in Figure 6 to study the separation and transfer efficiency of photoinduced charge carriers on the photocatalyst surface. Figure 6a shows the electrochemical impedance spectra (EIS) measurements of the as-prepared samples, which were conducted at the open-circuit potential of the as-made photocatalysts. The diameter of the obtained semicircle of WO3/CeO2 composites is clearly smaller than that of the pristine CeO2, indicating that composites with heterostructures have lower charge transfer resistance. The transient photocurrent responses of CeO2 and WO3/CeO2  To get insights into the photocatalytic process of TC removal, photoelectrochemical measurements were conducted and presented in Figure 6 to study the separation and transfer efficiency of photoinduced charge carriers on the photocatalyst surface. Figure 6a shows the electrochemical impedance spectra (EIS) measurements of the as-prepared samples, which were conducted at the open-circuit potential of the as-made photocatalysts. The diameter of the obtained semicircle of WO 3 /CeO 2 composites is clearly smaller than that of the pristine CeO 2 , indicating that composites with heterostructures have lower charge transfer resistance. The transient photocurrent responses of CeO 2 and WO 3 /CeO 2 were measured under intermittent light illumination, as illustrated in Figure 6b. The photocurrent density of WO 3 /CeO 2 was much higher than that of CeO 2 , indicating that the WO 3 /CeO 2 laminated heterojunction composites exhibited slower charge recombination rate and stronger electron-hole separation ability. In addition, repeatable photocurrent responses in every illumination cycle can be observed, showing the excellent stability of the constructed heterojunctions. Figure 6c presents the surface photovoltage spectroscopy (SPV) data of the samples. The SPV measurements of WO 3 /CeO 2 demonstrate much-increased surface voltage under light irradiation compared to CeO 2 , showing the accelerated charge transfer efficiency on the heterojunction surface [53]. The suppressed-charge recombination rate was also examined using PL emission measurements, as shown in Figure 6d. The PL spectra of WO 3 /CeO 2 show clear photoluminescence quenching compared to pure CeO 2 at an excitation wavelength of 365 nm. This result demonstrates that the efficient suppression of photogenerated electron-hole recombination was achieved via the formation of heterojunction interfaces in WO 3 /CeO 2 composites. All the above measurements provided solid evidence for the enhanced charge separation and transfer efficiency and suppressed electron-hole recombination, which improves photocatalytic activity for TC degradation. were measured under intermittent light illumination, as illustrated in Figure 6b. The photocurrent density of WO3/CeO2 was much higher than that of CeO2, indicating that the WO3/CeO2 laminated heterojunction composites exhibited slower charge recombination rate and stronger electron-hole separation ability. In addition, repeatable photocurrent responses in every illumination cycle can be observed, showing the excellent stability of the constructed heterojunctions. Figure 6c presents the surface photovoltage spectroscopy (SPV) data of the samples. The SPV measurements of WO3/CeO2 demonstrate much-increased surface voltage under light irradiation compared to CeO2, showing the accelerated charge transfer efficiency on the heterojunction surface [53]. The suppressed-charge recombination rate was also examined using PL emission measurements, as shown in Figure  6d. The PL spectra of WO3/CeO2 show clear photoluminescence quenching compared to pure CeO2 at an excitation wavelength of 365 nm. This result demonstrates that the efficient suppression of photogenerated electron-hole recombination was achieved via the formation of heterojunction interfaces in WO3/CeO2 composites. All the above measurements provided solid evidence for the enhanced charge separation and transfer efficiency and suppressed electron-hole recombination, which improves photocatalytic activity for TC degradation. All obtained results for TC photodegradation and heterojunction characterizations led to the proposed mechanism for photocatalytic TC elimination on the WO3/CeO2 surface, as illustrated in Figure 7. As previously mentioned, the band alignment of WO3/CeO2 is assumed to be staggered according to the measured and calculated band positions. The electrons are excited under light irradiation and migrated from the valence band to the conduction band, thereby leaving holes with positive charges in the valence band. If type II heterojunctions were formed in the as-synthesized WO3/CeO2 composites, the excited All obtained results for TC photodegradation and heterojunction characterizations led to the proposed mechanism for photocatalytic TC elimination on the WO 3 /CeO 2 surface, as illustrated in Figure 7. As previously mentioned, the band alignment of WO 3 /CeO 2 is assumed to be staggered according to the measured and calculated band positions. The electrons are excited under light irradiation and migrated from the valence band to the conduction band, thereby leaving holes with positive charges in the valence band. If type II heterojunctions were formed in the as-synthesized WO 3 /CeO 2 composites, the excited electrons would transfer from the conduction band of CeO 2 to that of WO 3 . The photogenerated holes migrate from the valence band of WO 3 to that of CeO 2 . However, the photoinduced electrons in the conduction band of WO 3 could not reduce oxygen into superoxide radicals because of the more positive CB position of WO 3 (0.50 eV vs. NHE) compared to the O 2 /·O 2 − potential (−0.33 V vs. NHE) [54]. Likewise, the photogenerated holes in the VB of both WO 3 (2.62 eV vs. NHE) and CeO 2 (2.1 eV vs. NHE) could not oxidize H 2 O to form ·OH, due to the more negative VB potential of the two materials compared to the H 2 O/·OH potential (2.72 V vs. NHE) [54]. In that scenario, O 2 − and ·OH cannot be generated in the TC degradation process. Thus, the typical type II heterojunction mechanism was impractical in the WO 3 /CeO 2 composites. Accordingly, the Z-scheme charge transfer mechanism was applied to the WO 3 /CeO 2 heterojunction, as shown in  Figure 2, as 2D photocatalysts have significant advantages in charge-transfer kinetics [55]. Thanks to the 2D morphology and the formation of the Z-scheme laminated heterojunction, the spatial charge separation and transfer efficiency are much enhanced under light illumination. Surface-adsorbed oxygen can capture photogenerated electrons to form superoxide radicals (O 2 − ). Tetracycline is eventually oxidized into CO 2 and water by the photoinduced radicals and holes on WO 3 /CeO 2 surface. However, this mechanism is merely an assumption based on the ·OH and·O 2 − formation mechanism. The detection of reactive species such as ·O 2 − and the TC oxidation products will be performed in further research.  [54]. In that scenario, O2 -and ·OH cannot be generated in the TC degradation process. Thus, the typical type II heterojunction mechanism was impractical in the WO3/CeO2 composites. Accordingly, the Z-scheme charge transfer mechanism was applied to the WO3/CeO2 heterojunction, as shown in Figure 7.
Electrons moved from WO3 (CB) to CeO2 (VB). The electrons accumulated in the CB of CeO2 easily oxidize O2 into O2 − . Photogenerated holes in WO3 (VB) and·O2 -would degrade TC simultaneously under light irradiation. Both CeO2 and WO3 are two-dimensional (2D) materials, as illustrated in Figure 2, as 2D photocatalysts have significant advantages in charge-transfer kinetics [55]. Thanks to the 2D morphology and the formation of the Zscheme laminated heterojunction, the spatial charge separation and transfer efficiency are much enhanced under light illumination. Surface-adsorbed oxygen can capture photogenerated electrons to form superoxide radicals (O2 − ). Tetracycline is eventually oxidized into CO2 and water by the photoinduced radicals and holes on WO3/CeO2 surface. However, this mechanism is merely an assumption based on the ·OH and·O2 -formation mechanism. The detection of reactive species such as ·O2 -and the TC oxidation products will be performed in further research.

Conclusions
In summary, 2D/2D mesoporous WO3/CeO2 photocatalysts were constructed via a facile two-step hydrothermal process. The heterojunction structure was confirmed by high-resolution transmission electron microscopy (HRTEM). The light absorption ability of the composites was improved because of the presence of WO3. Based on the measured and calculated band positions and the Fermi level positions, the band alignment of WO3/CeO2 was assumed to be a staggered Z-scheme heterojunction. The charge separation and migration efficiency were much enhanced in the mesoporous WO3/CeO2 heterojunction interface, as confirmed by electrochemical impedance spectra, transient photo-

Conclusions
In summary, 2D/2D mesoporous WO 3 /CeO 2 photocatalysts were constructed via a facile two-step hydrothermal process. The heterojunction structure was confirmed by high-resolution transmission electron microscopy (HRTEM). The light absorption ability of the composites was improved because of the presence of WO 3 . Based on the measured and calculated band positions and the Fermi level positions, the band alignment of WO 3 /CeO 2 was assumed to be a staggered Z-scheme heterojunction. The charge separation and migration efficiency were much enhanced in the mesoporous WO 3 /CeO 2 heterojunction interface, as confirmed by electrochemical impedance spectra, transient photocurrent responses, and surface photovoltage spectroscopy. The suppression of photogenerated charge recombination was proved by photoluminescence. Thanks to the much enhanced charge migration and separation efficiency, the optimized 5WO 3 /CeO 2 (5 wt.% WO 3 ) Z-scheme laminated heterojunction photocatalysts demonstrated tetracycline photodegradation efficiency of 99.1% within 80 min, achieving a peak TC photodegradation efficiency of 0.0482 min −1 , which is approximately 3.4 times that of pristine CeO 2 . Under light irradiation, the fabricated photocatalysts maintained high TC removal efficiency for 10 cycles. The 2D/2D heterostructure proposed here provides the possibility of rare-earth-metal-based composites for efficient, stable, and reusable water purification.