Ag3PO4 Deposited on CuBi2O4 to Construct Z-Scheme Photocatalyst with Excellent Visible-Light Catalytic Performance Toward the Degradation of Diclofenac Sodium

CuBi2O4/Ag3PO4 was synthesized through a combination of hydrothermal synthesis and an in situ deposition method with sodium stearate as additives, and their textures were characterized with XRD, XPS, SEM/HRTEM, EDS, UV-Vis, and PL. Then, the photodegradation performance of CuBi2O4/Ag3PO4 toward the degradation of diclofenac sodium (DS) was investigated, and the results indicate that the degradation rate of DS in a CuBi2O4/Ag3PO4 (1:1) system is 0.0143 min−1, which is 3.6 times that in the blank irradiation system. Finally, the photocatalytic mechanism of CuBi2O4/Ag3PO4 was discussed, which follows the Z-Scheme theory, and the performance enhancement of CuBi2O4/Ag3PO4 was attributed to the improved separation efficiency of photogenerated electron–hole pairs.


Introduction
Diclofenac sodium (DS) is one of the most widely used nonsteroidal anti-inflammatory drugs [1,2]. However, most DS will be excreted outside of the body through urine and stool. Since the removal rate of DS after the treatment of conventional wastewater treatment plants (WWTPs) is no more than 20%, most of the remaining DS would be discharged along with effluents into estuaries, rivers, surface water, ground water, and even drinking water [3][4][5]. The DS frequently detected in aquatic environments (estuaries, rivers, surface water, ground water, and even drinking water) has aroused great concern in that it endangers human health and poses great risks to the environment [6,7]. Over the last several years, various technologies encompassing adsorption [7,8], photocatalysis [9], Fenton reagents [10], ozonation [11], etc. have been studied to remove DS. Semiconductor photocatalysis is considered a green environmentally friendly and cost-efficient technology [12,13]. However, many of the known photocatalysts such as TiO 2 [14,15], CeO 2 [16,17], AgCl [18], ZnO [19], etc. have large energy band gaps, making them able to only utilize ultraviolet (UV) light to stimulate their activity. Unfortunately, the region of UV only occupies approximately 4% of the entire solar spectrum, which is just 9.3% of visible light. Considering the above, great efforts have been put in the development of visible-light responsive photocatalysts in recent years [12,13,20].  3 and HNO 3 was prepared to dissolve the Bi(NO 3 ) 3 ·5H 2 O precursor; secondly, the precipitator of NaOH was added into the reaction system. Thirdly, the mixture solution was diluted to a certain volume; fourth, the solution was reacted hydrothermally under 100 • C for 24 h. After the isolation of the produced precipitates by centrifugation, drying of these solids at 60 • C overnight is necessary.

Synthesis of CuBi 2 O 4 /Ag 3 PO 4
The in situ deposition method was used to prepare CuBi 2 O 4 /Ag 3 PO 4 with sodium stearate as additives. The detailed preparation processes of CuBi 2 O 4 /Ag 3 PO 4 for the mass ratio of 1:1 is as follows: 0.1256 g CuBi 2 O 4 was dispersed into 40 mL of ultrapure water, the solution was ultrasound at 100 W for 15 min subsequently, and then a sodium stearate solution (10 mL, the mole ratio of the sodium stearate to Ag + is 2) was added into the reaction system and mechanically agitated for 2 h. After that, an AgNO 3 solution (10 mL, 0.9 mmol) was added to the mixture. After stirring for 30 min, a Na 2 HPO 4 ·12H 2 O solution (20 mL, 0.3 mmol) was added drop by drop. The precipitate was isolated and washed several times with absolute ethanol and distilled water and then dried at 60 • C overnight.

Characterization of Photocatalysts
The morphologies of as-prepared catalysts were characterized with JSM-6360LV (JEOL, Tokyo, Japan) and high-resolution transmission electron microscopy (HRTEM, FEI Tecnai G 2 F20 S-TWIN, Hillsboro, OR, USA). The crystal structures of various catalysts were examined with the X-ray diffraction instrument (XRD, Rigaku D/max 2500 PC, Rigaku, Japan). The surface element composition and chemical state analysis were performed on the X-ray photoelectron spectroscopy (XPS) spectra (ESCALAB 250 spectrometer, Waltham, MA, USA). Energy-dispersive X-ray spectroscopy (EDS) was also used to investigate the types and contents of elements in the materials. The nitrogen adsorption-desorption isotherms were obtained using a NOVA-2200e volumetric analyzer (Quantachrome, Boynton Beach, FL, USA). The surface areas of the samples can also be estimated by the BET model. The optical absorption performance of the photocatalysts was analyzed with UV-vis diffusive reflectance spectra labeled by Shimadzu UV-2550 (Kyoto, Japan). The photoluminescence (PL) spectroscopy was detected with a JASCO, FP-6500 florescence spectrophotometer (Oklahoma City, OK, USA).

Experiments of Photocatalytic Performance
Photocatalytic activity experiments: The photocatalytic activity of as-prepared catalysts was evaluated toward the degradation of diclofenac sodium (DS) as well as dyestuff of Rhodamine B (RhB), Congo red (CR), Methyl Orange (MO), and Methylene Blue (MB). The photoreaction apparatus, equipped with a 300 W Xe lamp and an ultraviolet cutoff filter providing visible light at ≥400 nm, was manufactured by Bilang Biological Science and Technology Co., Ltd., Xi'an, China. The configuration of the apparatus is illustrated in our previous paper [34]. During the experiments, 0.025 g of photocatalyst was added into a 50 mL 10 mg/L DS solution. After the preparation, the adsorption-desorption equilibrium experiment was firstly onset in the dark with magnetically stirred for 30 min. Then, the Xe lamp light was turned on, and the sample was withdrawn at a setting time interval and was filtered with 0.45-µm membrane filters. The pH, concentration of residual DS, and TOC in the filter were detected with the pH analyzer (PHS-3C, Shanghai Jinghong Scientific Instrument Co. Ltd., Shanghai, China), high-performance liquid chromatography (HPLC, Agilent 1260, Santa Clara, CA, USA), and TOC (Shimadzu TOC-V CPH, Kyoto, Japan) analyzer, respectively. As for the photodegradation of dyestuff in the CuBi 2 O 4 /Ag 3 PO 4 (1:1) system, the concentration of dyestuff is 15 mg/L and the catalyst dosage is 0.5 g/L.
Photocatalytic stability experiments: The stability of CuBi 2 O 4 /Ag 3 PO 4 (1:1) was demonstrated with a repeatability test. The specific experimental processes can refer to the photocatalytic activity experiments. After the reaction in each run, the photocatalysts were collected and vacuum-dried overnight at 60 • C for recycling use.

Analysis of Reactive Species
The reactive species formed during the photodegradation process of DS were checked with free radical capture experiments, and tert-butanol (t-BuOH), disodium ethylenediamine tetraacetate (EDTA-Na 2 ), benzoquinone (BZQ) were chose as the scavenger of hydroxyl radical (OH), the hole (h + ), and the superoxide radical (O 2 − ), respectively. The detailed free radical capture experiment processes were similar to the photocatalytic activity experiments.

Morphology and Associated Formation Mechanism of CuBi 2 O 4
Four different experiments under various conditions for the preparation of CuBi 2 O 4 were firstly conducted to investigate the effect of the morphology and structure on the photocatalytic activity. SEM images of the as-prepared CuBi 2 O 4 are shown in Figure 1a-d. As shown in Figure 1a, the CuBi 2 O 4 prepared with 1.2 M NaOH and the step of "dilution" after the addition of NaOH displays well-distributed microsphere structures with diameters of~4 µm, and its surfaces are comprised of nanometer quadrilateral plates with lengths of~250 nm and widths of~200 nm. When the mixture solution of Bi(NO 3 ) 3 ·5H 2 O + Cu(NO 3 ) 2 ·3H 2 O + HNO 3 was diluted before the addition of 1.2 M NaOH, the CuBi 2 O 4 with a relatively compact and smooth surface are obtained, as described in Figure 1b. Figure 1c,d shows the SEM images of CuBi 2 O 4 prepared with 2.4 M NaOH, and the step of "dilution" was after or before the addition of NaOH, respectively. The microsphere structural CuBi 2 O 4 with its surface comprised of tens of nanometers cube was obtained when the 2.4 M NaOH was added before the "dilution", and the CuBi 2 O 4 with nonuniform size and structure were obtained when the 2.4 M NaOH was added after the "dilution". The remarkable influence of NaOH molarity on the morphology and property of CuBi 2 O 4 was also evidenced by others. Wang et al. [35] reported the hedgehog-like CuBi 2 O 4 hierarchical microspheres. Oh et al. [36] reported the three-dimensional spherical CuBi 2 O 4 nan column arrays connected through non-covalent interactions. Patil et al. [37] reported the spherulitic flower morphology of CuBi 2 O 4 . Xie et al. [38] reported the leaf-like structural CuBi 2 O 4 . Moreover, the order of "dilution" mainly influences the mass transfer rate and Gibbs free energy of the reaction system, finally leading to the formation of different structural crystals to achieve the lowest Gibbs free energy [36].

Morphology and Associated Formation Mechanism of CuBi2O4
Four different experiments under various conditions for the preparation of CuBi2O4 were firstly conducted to investigate the effect of the morphology and structure on the photocatalytic activity. SEM images of the as-prepared CuBi2O4 are shown in Figure 1a-d. As shown in Figure 1a, the CuBi2O4 prepared with 1.2 M NaOH and the step of "dilution" after the addition of NaOH displays welldistributed microsphere structures with diameters of ~4 µm, and its surfaces are comprised of nanometer quadrilateral plates with lengths of ~250 nm and widths of ~200 nm. When the mixture solution of Bi(NO3)3·5H2O + Cu(NO3)2·3H2O + HNO3 was diluted before the addition of 1.2 M NaOH, the CuBi2O4 with a relatively compact and smooth surface are obtained, as described in Figure 1b. Figure 1c,d shows the SEM images of CuBi2O4 prepared with 2.4 M NaOH, and the step of "dilution" was after or before the addition of NaOH, respectively. The microsphere structural CuBi2O4 with its surface comprised of tens of nanometers cube was obtained when the 2.4 M NaOH was added before the "dilution", and the CuBi2O4 with nonuniform size and structure were obtained when the 2.4 M NaOH was added after the "dilution". The remarkable influence of NaOH molarity on the morphology and property of CuBi2O4 was also evidenced by others. Wang et al. [35] reported the hedgehog-like CuBi2O4 hierarchical microspheres. Oh et al. [36] reported the three-dimensional spherical CuBi2O4 nan column arrays connected through non-covalent interactions. Patil et al. [37] reported the spherulitic flower morphology of CuBi2O4. Xie et al. [38] reported the leaf-like structural CuBi2O4. Moreover, the order of "dilution" mainly influences the mass transfer rate and Gibbs free energy of the reaction system, finally leading to the formation of different structural crystals to achieve the lowest Gibbs free energy [36]. On the whole, the mechanism of CuBi2O4 formation involves three consecutive stages: First, the dissolution of Bi(NO3)3·5H2O in HNO3, as described by the Equation (1); second, the precipitation of Bi(OH)3 and Cu(OH)2 because the solubility product constants of Bi(OH)3 and Cu(OH)2 are 2.9 × 10 −7 and 4.8 × 10 −20 (25 °C), respectively, so when the precipitating agent of NaOH was added into the reaction system, the blue flocculent precipitates of Cu(OH)2 formed first, followed by the formation of white precipitates Bi(OH)3, which can be described by Equations (2) and (3); and third, the On the whole, the mechanism of CuBi 2 O 4 formation involves three consecutive stages: First, the dissolution of Bi(NO 3 ) 3 ·5H 2 O in HNO 3 , as described by the Equation (1); second, the precipitation of Bi(OH) 3 and Cu(OH) 2 because the solubility product constants of Bi(OH) 3 and Cu(OH) 2 are 2.9 × 10 −7 Nanomaterials 2019, 9, 959 5 of 15 and 4.8 × 10 −20 (25 • C), respectively, so when the precipitating agent of NaOH was added into the reaction system, the blue flocculent precipitates of Cu(OH) 2 formed first, followed by the formation of white precipitates Bi(OH) 3 , which can be described by Equations (2) and (3); and third, the formation and Ostwald-ripening process of CuBi 2 O 4 , in which the Bi(OH) 3 would lose one water molecule first to become partial yellow bismuth hydroxide BiO(OH) under 100 • C (see Equation (4)) and reacted with Cu(OH) 2 to form CuBi 2 O 4 (see Equation (5)) and then the as-formed CuBi 2 O 4 crystals gave birth to different morphologies via Ostwald ripening.
The products of CuBi 2 O 4 with different morphologies shown in Figure 1a-d were recorded as "Product a", "Product b", "Product c", and "Product d", respectively. Their photocatalytic activity toward DS degradation are described in Figure 1e. As can be seen, all the CuBi 2 O 4 exhibited more enhanced photocatalytic activity than the self-photodegradation of DS and "Product a" showed the highest catalytic activity, which is consistent with the variation of specific surface area (see Figure 1f). Therefore, "Product a" was used to further synthesize the composite CuBi 2 O 4 /Ag 3 PO 4 .

Morphology
The morphology of the as-prepared catalysts are characterized by SEM, and the results are shown in Figure 2. Figure 2a shows the SEM image of Ag 3 PO 4 , which exhibits an irregular polyhedron structure with an average diameter of approximately 400 nm. Compared with the morphology of CuBi 2 O 4 shown in Figure 1a, the surface of composite CuBi 2 O 4 /Ag 3 PO 4 (1:1) becomes rough (see Figure 2b), which is because the nano-particulate Ag 3 PO 4 are attached onto the surface and occupies the nano-sheet gap of the CuBi 2 O 4 s surface. However, the size of Ag 3 PO 4 in the composite is tailored to be approximately~100 nm, mainly because of the inhibition of the nano-sheet gap of the CuBi 2 O 4 s surface on the nucleation, growth, and crystallization processes of Ag 3 PO 4 . Figure  enhanced photocatalytic activity than the self-photodegradation of DS and "Product a" showed the highest catalytic activity, which is consistent with the variation of specific surface area (see Figure 1f). Therefore, "Product a" was used to further synthesize the composite CuBi2O4/Ag3PO4. The morphology of the as-prepared catalysts are characterized by SEM, and the results are shown in Figure 2. Figure 2a shows the SEM image of Ag3PO4, which exhibits an irregular polyhedron structure with an average diameter of approximately 400 nm. Compared with the morphology of CuBi2O4 shown in Figure 1a, the surface of composite CuBi2O4/Ag3PO4 (1:1) becomes rough (see Figure 2b), which is because the nano-particulate Ag3PO4 are attached onto the surface and occupies the nano-sheet gap of the CuBi2O4′s surface. However, the size of Ag3PO4 in the composite is tailored to be approximately ~100 nm, mainly because of the inhibition of the nano-sheet gap of the CuBi2O4′s surface on the nucleation, growth, and crystallization processes of Ag3PO4. Figure 2c shows the lattice fringe image of CuBi2O4/Ag3PO4 (1:1). The observed lattice fringes of 0.267 nm and 0.240 nm

Component and Surface Property
The crystallinity and component of the as-prepared materials were characterized by XRD, and the results are shown in Figure 3. It can be clearly seen that the diffraction peaks in the XRD pattern of Ag 3 PO 4 can be indexed to the phase of Ag 3 PO 4 (JCPDS No. 06-0505) [39]. Moreover, the reflections at 2θ = 20.76, 29

Component and Surface Property
The crystallinity and component of the as-prepared materials were characterized by XRD, and the results are shown in Figure 3. It can be clearly seen that the diffraction peaks in the XRD pattern of Ag3PO4 can be indexed to the phase of Ag3PO4 (JCPDS No. 06-0505) [39]. Moreover, the reflections at 2θ = 20.76, 29 [40]. Furthermore, the XRD pattern of CuBi2O4/Ag3PO4 (1:1) suggests that the material obtained just contains CuBi2O4 and Ag3PO4, with the absence of any other substances, indicating that Ag3PO4 couples with CuBi2O4 mainly through the physical effects but not the chemical reaction. The chemical compositions and surface chemical states of the CuBi2O4/Ag3PO4 (1:1) composite photocatalyst were further confirmed by the X-ray photoelectron spectroscopy (XPS), as shown in Figure 4. The survey spectrum of CuBi2O4/Ag3PO4 (1:1), shown in Figure 4a, indicates the presence of Cu, Bi, Ag, P, and O. Figure 4b presents the Cu 2p region. It can be clearly observed that the two main peaks at 934.25 eV and 954.18 eV with a spin-orbit splitting of 20.13 eV are assigned to the binding energies of Cu 2p3/2 and Cu 2p1/2, and the two satellite peaks observed at 942.07 eV and 963.25 eV further confirm the Cu 2+ valence state [41]. Figure 4c shows the core level of the Bi 4f spectra; two bands at 159.13 eV and 164.44 eV corresponding to the binding energies of Bi 4f7/2 and Bi 4f5/2, respectively, are observed. The results are consistent with the studies reported by other researchers   Figure 4b presents the Cu 2p region. It can be clearly observed that the two main peaks at 934.25 eV and 954.18 eV with a spin-orbit splitting of 20.13 eV are assigned to the binding energies of Cu 2p 3/2 and Cu 2p 1/2 , and the two satellite peaks observed at 942.07 eV and 963.25 eV further confirm the Cu 2+ valence state [41]. Figure 4c shows the core level of the Bi 4f spectra; two bands at 159.13 eV and 164.44 eV corresponding to the binding energies of Bi 4f 7/2 and Bi 4f 5/2 , respectively, are observed. The results are consistent with the studies reported by other researchers [36]. Figure 4d shows the Ag 3d region, including two highly intense peaks at 368.16 eV and 374.17 eV, which can be ascribed to the Ag 3d 5/2 and Ag 3d 3/2 binding energies [42,43], respectively. For the high-resolution XPS spectrum of P 2p, shown in Figure 4e, only one peak with the binding energy of 132.90 eV is observed, and it is extremely similar to the 132.40 eV evolved in the PO 4 3− [44]. In Figure 4f,

Optical Absorption Property
The optical absorption properties of the as-prepared catalysts were determined by the UV-Vis diffusive reflectance analyzer, and the spectra are shown in Figure 5a. Both of the pure CuBi2O4 and Ag3PO4 exhibit strong absorbances in the UV and visible light regions, and their absorbance boundaries are 530 nm and >800 nm, respectively. The results are consistent with the former discovery by other teams [21,45]. In the case of the CuBi2O4/Ag3PO4 (1:1) composite, the absorbance in the visible light regions is much higher than that of the pure Ag3PO4. This property can make a positive contribution to the photocatalytic activity of the composite because a more efficient utilization of solar energy could be achieved [46]. In addition, according to the Kubelka-Munk function and the plot of (αhv) 2 vs hv (shown in Figure 5b) [46,47], the band gaps (Eg) of Ag3PO4, CuBi2O4, and CuBi2O4/Ag3PO4 (1:1) were estimated to be 2.42 eV, 1.72 eV, and 2.01 eV, respectively. The band gap energy of CuBi2O4/Ag3PO4 (1:1) is obviously narrower than that of bare Ag3PO4, which can conclude that the formed composite is more easily excited by visible light and that the utilization ratio of visible light is enhanced. Furthermore, the band-edge potentials of the conduction band (ECB) and valence band (EVB) could be calculated from the equations [47]: EVB = X − E C + 0.5Eg and ECB = X − E C − 0.5Eg, where X is the geometric mean of the electronegativity of the constituent atoms (5.96 eV for Ag3PO4 and 4.59 eV for CuBi2O4) [26,48] and E C is the energy of the free electrons on the hydrogen scale (about 4.5 eV). Thus, the EVB and ECB of Ag3PO4 can be estimated to be 2.67 eV/NHE and 0.25 eV/NHE, while the EVB and ECB of CuBi2O4 can be estimated to be 0.95 and −0.77 eV/NHE, respectively.

Optical Absorption Property
The optical absorption properties of the as-prepared catalysts were determined by the UV-Vis diffusive reflectance analyzer, and the spectra are shown in Figure 5a. Both of the pure CuBi 2 O 4 and Ag 3 PO 4 exhibit strong absorbances in the UV and visible light regions, and their absorbance boundaries are 530 nm and >800 nm, respectively. The results are consistent with the former discovery by other teams [21,45]. In the case of the CuBi 2 O 4 /Ag 3 PO 4 (1:1) composite, the absorbance in the visible light regions is much higher than that of the pure Ag 3 PO 4 . This property can make a positive contribution to the photocatalytic activity of the composite because a more efficient utilization of solar energy could be achieved [46]. In addition, according to the Kubelka-Munk function and the plot of (αhv) 2 vs hv (shown in Figure 5b) [46,47], the band gaps (Eg) of Ag 3 PO 4 , CuBi 2 O 4 , and CuBi 2 O 4 /Ag 3 PO 4 (1:1) were estimated to be 2.42 eV, 1.72 eV, and 2.01 eV, respectively. The band gap energy of CuBi 2 O 4 /Ag 3 PO 4 (1:1) is obviously narrower than that of bare Ag 3 PO 4 , which can conclude that the formed composite is more easily excited by visible light and that the utilization ratio of visible light is enhanced. Furthermore, the band-edge potentials of the conduction band (E CB ) and valence band (E VB ) could be calculated from the equations [47]: E VB = X − E C + 0.5E g and E CB = X − E C − 0.5E g , where X is the geometric mean of the electronegativity of the constituent atoms (5.96 eV for Ag 3 PO 4 and 4.59 eV for CuBi 2 O 4 ) [26,48] and E C is the energy of the free electrons on the hydrogen scale (about 4.5 eV). Thus, the E VB and E CB of Ag 3 PO 4 can be estimated to be 2.67 eV/NHE and 0.25 eV/NHE, while the E VB and E CB of CuBi 2 O 4 can be estimated to be 0.95 and −0.77 eV/NHE, respectively. ratio of visible light is enhanced. Furthermore, the band-edge potentials of the conduction band (ECB) and valence band (EVB) could be calculated from the equations [47]: EVB = X − E C + 0.5Eg and ECB = X − E C − 0.5Eg, where X is the geometric mean of the electronegativity of the constituent atoms (5.96 eV for Ag3PO4 and 4.59 eV for CuBi2O4) [26,48] and E C is the energy of the free electrons on the hydrogen scale (about 4.5 eV). Thus, the EVB and ECB of Ag3PO4 can be estimated to be 2.67 eV/NHE and 0.25 eV/NHE, while the EVB and ECB of CuBi2O4 can be estimated to be 0.95 and −0.77 eV/NHE, respectively.

Photodegradation Performance of DS in Different Catalysts Systems
In order to estimate the photocatalytic activity of the as-prepared materials, DS was selected as a target pollutant. Figure 6a indicates the photodegradation efficiency of DS in different photocatalytic systems. In the blank irradiation system, the self-photodegradation efficiency of DS during the 120 min of reaction is 37.81%. Under the same conditions, the degradation efficiencies of DS in the pure CuBi 2 O 4 and Ag 3 PO 4 photocatalytic system are 67.12% and 77.84%, respectively. While in the photocatalytic system of CuBi 2 O 4 /Ag 3 PO 4 (1:1), the degradation efficiency of DS reaches 85.45%. This suggests that the addition of catalysts promotes the degradation of DS and that there is a synergistic effect between CuBi 2 O 4 and Ag 3 PO 4 . A pseudo-first-order kinetic model was employed to fit the experimental data of DS degradation in different photocatalytic systems, and the results are shown in Table 1. It can be seen that the degradation rate constants of DS are 0.0041 min −1 , 0.0084 min −1 , 0.0112 min −1 , and 0.0069 min −1 in the photocatalytic system of blank irradiation, CuBi 2 O 4 , Ag 3 PO 4 , and commercial TiO 2 , respectively. The low photocatalytic activity of commercial TiO 2 is mainly due to its weak absorption of visible light, resulting in a small number of active carriers. As for the photocatalytic systems of composite CuBi 2 O 4 /Ag 3 PO 4 , the rate constant of DS increases firstly and then decreases with the further increase of Ag 3 PO 4 content in the composites, which is consistent with the variation of specific surface area (see Table 1). The main reason is that the deposition of excessive Ag 3 PO 4 on the surface of CuBi 2 O 4 will block the gap between the nanosheets on the surface of CuBi 2 O 4 , which is not conducive to the diffusion of reactants and products. Also, when the mass ratio of CuBi 2 O 4 and Ag 3 PO 4 in the composite is 1:1, the rate constant of 0.0143 min −1 can be obtained, which is approximately 3.6 times that of the values in the blank irradiation system. To understand the mineralization rate of DS in different photocatalytic systems, the TOC concentrations of the reacted DS solutions were detected and the removal efficiency are described in Figure 6b. In the blank irradiation system, the mineralization rate of DS is 14.36%. When the catalysts of CuBi 2 O 4 , Ag 3 PO 4 , and CuBi 2 O 4 /Ag 3 PO 4 (1:1) were introduced into the reaction system, the mineralization rates of DS increased to be 35.43%, 49.62%, and 57.33%, respectively. Figure 6c shows the change of the solution pH as reaction time goes on at different photocatalytic systems.
In general, the pH of the solution decreases, but there is a fluctuation at a certain time interval in each photocatalytic system, which can be explained by forming some acidic intermediates, and they are decomposed as the reaction time is prolonged. Figure 6d shows the removal efficiency of various dyes in the catalytic system of CuBi 2 O 4 /Ag 3 PO 4 (1:1). It can be seen that, within 20 min, 15 mg/L of RhB and MB can be completely decomposed by 0.5 g/L of the composite photocatalyst, while the removal efficiency of CR and MO are 85.38% and 94.24%, respectively. The results show that the as-prepared CuBi 2 O 4 /Ag 3 PO 4 exhibits excellent photocatalytic activity for the degradation of dyes as well. in the photocatalytic system of CuBi2O4/Ag3PO4 (1:1), the degradation efficiency of DS reaches 85.45%. This suggests that the addition of catalysts promotes the degradation of DS and that there is a synergistic effect between CuBi2O4 and Ag3PO4. A pseudo-first-order kinetic model was employed to fit the experimental data of DS degradation in different photocatalytic systems, and the results are shown in Table 1. It can be seen that the degradation rate constants of DS are 0.0041 min −1 , 0.0084 min −1 , 0.0112 min −1 , and 0.0069 min −1 in the photocatalytic system of blank irradiation, CuBi2O4, Ag3PO4, and commercial TiO2, respectively. The low photocatalytic activity of commercial TiO2 is mainly due to its weak absorption of visible light, resulting in a small number of active carriers. As for the photocatalytic systems of composite CuBi2O4/Ag3PO4, the rate constant of DS increases firstly and then decreases with the further increase of Ag3PO4 content in the composites, which is consistent with the variation of specific surface area (see Table 1). The main reason is that the deposition of excessive Ag3PO4 on the surface of CuBi2O4 will block the gap between the nanosheets on the surface of CuBi2O4, which is not conducive to the diffusion of reactants and products. Also, when the mass ratio of CuBi2O4 and Ag3PO4 in the composite is 1:1, the rate constant of 0.0143 min −1 can be obtained, which is approximately 3.6 times that of the values in the blank irradiation system. To understand the mineralization rate of DS in different photocatalytic systems, the TOC concentrations of the reacted DS solutions were detected and the removal efficiency are described in Figure 6b. In the blank irradiation system, the mineralization rate of DS is 14.36%. When the catalysts of CuBi2O4, Ag3PO4, and CuBi2O4/Ag3PO4 (1:1) were introduced into the reaction system, the mineralization rates of DS increased to be 35.43%, 49.62%, and 57.33%, respectively. Figure 6c shows the change of the solution pH as reaction time goes on at different photocatalytic systems. In general, The stability of the photocatalyst is a very important parameter with regard to practical applications. To further investigate the performance stability of the prepared CuBi 2 O 4 , Ag 3 PO 4 , and CuBi 2 O 4 /Ag 3 PO 4 (1:1), the cycling degradation experiments of DS were carried out and the results are shown in Figure 7a. It is found that the degradation efficiency of DS in the pure CuBi 2 O 4 photocatalytic system reduced from 67.12% to 63.05% after a 5-time repeated reaction, while the degradation efficiency of DS in the pure Ag 3 PO 4 and CuBi 2 O 4 /Ag 3 PO 4 (1:1) photocatalytic system reduced by a large extent, i.e., 18.51% and 11.99%. Furthermore, the corresponding XRD results shown in Figure 7b suggest that there is a negligible change about the phase structure of CuBi 2 O 4 sample after the repeated photocatalytic reactions, further indicating the photocatalytic stability of the CuBi 2 O 4 . However, compared with the XRD spectra of the fresh Ag 3 PO 4 and CuBi 2 O 4 /Ag 3 PO 4 (1:1), the diffraction peaks readily indexed as the metallic silver (Ag 0 ) emerge in the XRD spectra of the reacted Ag 3 PO 4 and CuBi 2 O 4 /Ag 3 PO 4 (1:1). In addition, the appearance of Ag 0 in the reacted Ag 3 PO 4 and CuBi 2 O 4 /Ag 3 PO 4 (1:1) can be further confirmed by the XPS. Figure 7c shows the high-resolution Ag 3d XPS spectrum of CuBi 2 O 4 /Ag 3 PO 4 (1:1) used five times. Based on the Ag 3d XPS spectrum of the fresh CuBi 2 O 4 /Ag 3 PO 4 (1:1) (see Figure 4d), the two individual peaks are further divided into four different peaks and the two main peaks at binding energies of 368.16 eV and 374.17 eV can be attributed to the Ag + of Ag 3 PO 4 , whereas the peaks at 368.32 eV and 374.40 eV can be attributed to the metallic silver (Ag 0 ). From the result, it is clear that the Ag 0 should have been formed on the surface of the reacted CuBi 2 O 4 /Ag 3 PO 4 (1:1) and that the proportion of the peaks relative to Ag 0 occupy 5.52% of the total fitting peak area. However, the Ag 3d XPS spectrum of the reacted Ag 3 PO 4 shown in Figure 7d indicates that the formed Ag 0 in the surface of Ag 3 PO 4 is 11.78%, which is much higher than that in the reacted CuBi 2 O 4 /Ag 3 PO 4 (1:1), further implying the enhanced stability of the CuBi 2 O 4 /Ag 3 PO 4 (1:1) in comparison to pure Ag 3 PO 4 .
CuBi2O4/Ag3PO4 (1:1). It can be seen that, within 20 min, 15 mg/L of RhB and MB can be completely decomposed by 0.5 g/L of the composite photocatalyst, while the removal efficiency of CR and MO are 85.38% and 94.24%, respectively. The results show that the as-prepared CuBi2O4/Ag3PO4 exhibits excellent photocatalytic activity for the degradation of dyes as well.  The stability of the photocatalyst is a very important parameter with regard to practical applications. To further investigate the performance stability of the prepared CuBi2O4, Ag3PO4, and CuBi2O4/Ag3PO4 (1:1), the cycling degradation experiments of DS were carried out and the results are shown in Figure 7a. It is found that the degradation efficiency of DS in the pure CuBi2O4 photocatalytic system reduced from 67.12% to 63.05% after a 5-time repeated reaction, while the degradation efficiency of DS in the pure Ag3PO4 and CuBi2O4/Ag3PO4 (1:1) photocatalytic system reduced by a large extent, i.e., 18.51% and 11.99%. Furthermore, the corresponding XRD results shown in Figure 7b suggest that there is a negligible change about the phase structure of CuBi2O4 sample after the repeated photocatalytic reactions, further indicating the photocatalytic stability of the CuBi2O4. However, compared with the XRD spectra of the fresh Ag3PO4 and CuBi2O4/Ag3PO4

Photocatalytic Mechanism of Different Catalysts
Photoluminescence (PL) is a useful tool for obtaining information about the photogenerated electron-hole recombination property of materials, which helps us to understand the photocatalytic mechanism of the catalysts. Therefore, the PL spectra of the as-prepared CuBi 2 O 4 , Ag 3 PO 4 , and CuBi 2 O 4 /Ag 3 PO 4 (1:1) were first recorded at room temperature with a excitation wavelength of 500 nm, and the spectra are shown in Figure 8a. Obviously, the PL intensity of the CuBi 2 O 4 /Ag 3 PO 4 (1:1) composite is much lower than that of the pure CuBi 2 O 4 and Ag 3 PO 4 , indicating that the combination of CuBi 2 O 4 and Ag 3 PO 4 has effectively improved the photogenerated carriers' separation. Therefore, though a small amount of Ag 0 has formed on the surface of CuBi 2 O 4 /Ag 3 PO 4 (1:1), its photocatalytic performance is still enhanced compared with the pure CuBi 2 O 4 and Ag 3 PO 4 .
Moreover, the free radical capture experiments were designed to investigate the difference of main reactive species in the photocatalysis systems of blank irradiation, CuBi 2 O 4 , Ag 3 PO 4 , and CuBi 2 O 4 /Ag 3 PO 4 (1:1). Generally speaking, when the scavenger is added into the reaction system, if the degradation efficiency of the target contaminant do not change significantly, it indicates the weak effect of the captured reactive species on the photodegradation of the target pollutant, while if the degradation efficiency of the target contaminant is inhibited in a great degree, it indicates the captured reactive specie is involved in the photodegradation of the target pollutant and that the contribution of the reactive species is greater with the increased inhibition efficiency. (1:1), the diffraction peaks readily indexed as the metallic silver (Ag 0 ) emerge in the XRD spectra of the reacted Ag3PO4 and CuBi2O4/Ag3PO4 (1:1). In addition, the appearance of Ag 0 in the reacted Ag3PO4 and CuBi2O4/Ag3PO4 (1:1) can be further confirmed by the XPS. Figure 7c shows the highresolution Ag 3d XPS spectrum of CuBi2O4/Ag3PO4 (1:1) used five times. Based on the Ag 3d XPS spectrum of the fresh CuBi2O4/Ag3PO4 (1:1) (see Figure 4d), the two individual peaks are further divided into four different peaks and the two main peaks at binding energies of 368.16 eV and 374.17 eV can be attributed to the Ag + of Ag3PO4, whereas the peaks at 368.32 eV and 374.40 eV can be attributed to the metallic silver (Ag 0 ). From the result, it is clear that the Ag 0 should have been formed on the surface of the reacted CuBi2O4/Ag3PO4 (1:1) and that the proportion of the peaks relative to Ag 0 occupy 5.52% of the total fitting peak area. However, the Ag 3d XPS spectrum of the reacted Ag3PO4 shown in Figure 7d indicates that the formed Ag 0 in the surface of Ag3PO4 is 11.78%, which is much higher than that in the reacted CuBi2O4/Ag3PO4 (1:1), further implying the enhanced stability of the CuBi2O4/Ag3PO4 (1:1) in comparison to pure Ag3PO4. Photoluminescence (PL) is a useful tool for obtaining information about the photogenerated electron-hole recombination property of materials, which helps us to understand the photocatalytic mechanism of the catalysts. Therefore, the PL spectra of the as-prepared CuBi2O4, Ag3PO4, and CuBi2O4/Ag3PO4 (1:1) were first recorded at room temperature with a excitation wavelength of 500 nm, and the spectra are shown in Figure 8a. Obviously, the PL intensity of the CuBi2O4/Ag3PO4 (1:1) composite is much lower than that of the pure CuBi2O4 and Ag3PO4, indicating that the combination of CuBi2O4 and Ag3PO4 has effectively improved the photogenerated carriers' separation. Therefore, though a small amount of Ag 0 has formed on the surface of CuBi2O4/Ag3PO4 (1:1), its photocatalytic performance is still enhanced compared with the pure CuBi2O4 and Ag3PO4. On the basis of the results described above, the photolysis mechanisms of the blank irradiation, CuBi 2 O 4 , Ag 3 PO 4 , and CuBi 2 O 4 /Ag 3 PO 4 reaction systems were proposed. In the blank irradiation system, DS should first absorb actinic photons and change to be DS *, then two processes may occur: (1) the direct photodegradation process of DS * and (2) the photooxidation process of DS; that is, the DS * reacts with the dissolved O 2 to form O 2 − , which is subsequently involved in the degradation of DS. Because of the free radical capture experiment suggesting the O 2 − has a significant effect on the DS degradation in the blank irradiation system, the photooxidation process of DS is the dominated process. This speculation has also been put forward in a previous paper [49], and the transition of photogenerated electrons and holes is described in Figure 9a. In the CuBi 2 O 4 photocatalytic system, the electron-hole (e − -h + ) pairs will generate under the visible light irradiation and, then, the e − and h + will induce to form other active radicals (such as O 2 − and OH) during the separation/migration processes.

Photocatalytic Mechanism of Different Catalysts
Due to the oxidation potential of OH − /OH (1.99 eV) [27] being much higher than the potential of the valence band of CuBi 2 O 4 (E VB = 0.95 eV), the OH cannot be induced by the h + . However, the O 2 absorbed on the surface of the catalyst can be induced to be O 2 − , owing to the lower position of the standard redox potential of O 2 /O 2 − (0.13 eV) [50] than the potential of the conduction band (E CB = −0.77 eV). Then, the partial O 2 •− can be further induced to be OH (see Figure 9b). Therefore, the separated h + and the induced O 2 •− /OH • are involved in the degradation of DS. As for the photocatalytic system of Ag 3 PO 4 , the transition of photogenerated eand h + are described in Figure 9c  h + from the VB of Ag 3 PO 4 will be strong. From the free radical capture experiments, the effect of h + on the photodegradation of DS is extremely weak, suggesting that h + is mainly involved in the formation of OH. The detailed transition of photogenerated eand h + in a photocatalytic system of CuBi 2 O 4 /Ag 3 PO 4 is represented in Figure 9d, and its photocatalytic mechanism is consistent with the Z-Scheme theory [28]. of CuBi2O4/Ag3PO4 is represented in Figure 9d, and its photocatalytic mechanism is consistent with the Z-Scheme theory [28].

Conclusions
Ag3PO4 was deposited on CuBi2O4 to construct Z-scheme photocatalyst, and it was used for DS degradation. CuBi2O4/Ag3PO4 exhibited an excellent photocatalytic performance under the system of visible-light irradiation. Free radical capture experiments suggest the active species of O2 − and OH contribute to the degradation of DS. Moreover, Z-Scheme theory can be used to explain the catalytic mechanism of CuBi2O4/Ag3PO4, and the formed Ag 0 during the reaction process serves as the recombination center for the photogenerated efrom CB of Ag3PO4 and the h + from VB of CuBi2O4, which improved the separation of photogenerated carriers, finally leading to the enhancement of the catalytic performance.
Author Contributions: X.C. designed and performed the experiments and drafted the manuscript. R.Z. and N.L.

Conclusions
Ag 3 PO 4 was deposited on CuBi 2 O 4 to construct Z-scheme photocatalyst, and it was used for DS degradation. CuBi 2 O 4 /Ag 3 PO 4 exhibited an excellent photocatalytic performance under the system of visible-light irradiation. Free radical capture experiments suggest the active species of O 2 − and OH contribute to the degradation of DS. Moreover, Z-Scheme theory can be used to explain the catalytic mechanism of CuBi 2 O 4 /Ag 3 PO 4 , and the formed Ag 0 during the reaction process serves as the recombination center for the photogenerated efrom CB of Ag 3 PO 4 and the h + from VB of CuBi 2 O 4 , which improved the separation of photogenerated carriers, finally leading to the enhancement of the catalytic performance.