Enhanced Visible-Light Photocatalytic Activity of Ag QDs Anchored on CeO2 Nanosheets with a Carbon Coating

Ag quantum dots (QDs) anchored on CeO2 nanosheets with a carbon coating (Ag/CeO2@C) (composites) were prepared via an in situ reduction approach for the photocatalytic degradation of Cr(VI) and tetracycline hydrochloride (TCH) in the visible-light region. The photocatalytic activity of Ag/CeO2@C was greatly affected by carbon content, Ag-doping content, Cr(VI) concentration, pH value, and inorganic ions. Enhanced photocatalytic activity was obtained by Ag/CeO2@C (compared to CeO2 and CeO2@C), of which 3-Ag/CeO2@C-2 with an Ag-doping content of 5.41% presented the best removal efficiency and the most superior stability after five cycles. ·O2− and ·OH radicals were crucial for the photocatalytic capacity of 3-Ag/CeO2@C-2. The combined effect of the surface plasma resonance (SPR) of Ag QDs, an electron trapper of carbon shells, and the redox activity of the Ce(III)/Ce(IV) coupling induced efficient charge transfer and separation, suppressing the recombination of electron–hole pairs.

These peaks were assigned to the (111), (200), (220), (311), (222), (400), (311), and (420) facets of cubic CeO 2 phases (JCPDS No. , respectively [20,23,25]. The peaks of Ag/CeO 2 @C at 38.10 • , 44.32 • , and 64.49 • were ascribed to the (110), (200), and (220) planes of face-centered cubic Ag phases (JCPDS No. 04-0783) [22,23,41]. The diffraction peaks of carbon phases were not detected in the XRD patterns of CeO 2 @C and Ag/CeO 2 @C because they had less carbon content (<5%) and weak amorphous carbon intensity [6,53,54]. With an increase in Ag content, the diffraction peak intensities of Ag (110) and CeO 2 (111) (Figure 1) respectively increased at around 38.10 • and 28.57 • , indicating the high crystallinity of Ag/CeO 2 @C. long-term irradiation, leading to inferior photocatalytic stability [15]. The divided Gaussian peaks at 531.2 eV and 529.6 eV ( Figures 2B, S2B, and S3B) were attributed to the defect oxygen (or oxygen vacancy) and lattice oxygen, respectively [15,20,31]. The varied ratios of defect oxygen (or oxygen vacancy) to lattice oxygen were greatly related to the Ce 3+ content in the as-obtained samples. The concentration of the defect oxygen (or oxygen vacancy) in the O 1s XPS spectrum of the used 3-Ag/CeO2@C-2 ( Figure 2B) was higher than that of a fresh sample. This was due to the released and adsorbed oxygen of CeO2 and the defect structure of Ce 3+ -Ov-Ce 4+ (Ov-oxygen vacancy) under reduction and oxidation conditions [38,42]. The C1 spectra of CeO2@C-2 ( Figure 3C) and fresh and used 3-Ag/CeO2@C-2 ( Figure 2C) were divided into three Gaussian peaks at 288.4 eV, 285.3 eV, and 284.8 eV, which respectively belonged to C=O, C-OH, and C-C/C=C bonds. The splitting peaks at 374.2 eV and 368.19 eV in the Ag 3d XPS spectrum of fresh and used 3-Ag/CeO2@C-2 ( Figure 2D) were assigned to Ag 3d3/2 and Ag 3d5/2, respectively. Although there was no change in the microstructure of used 3-Ag/CeO2@C-2 ( Figure S4), long-term irradiation induced a difference in the C 1s and Ag 3d XPS spectra between the fresh and used samples ( Figure 3C,D), leading to inferior charge transfer and separation [45,47].  XPS was applied to investigate the surface compositions and chemical states of the obtained CeO 2 composites. The Ce 3d spectrum (Figure 2A, Figures S2A and S3A) was split into eight Gaussian peaks. The peaks at 900.7 eV (U) and 882.5 eV (V) were respectively assigned to Ce 3d 3/2 and Ce 3d 5/2 , indicating the formation of Ce 3+ and Ce 4+ [20,21]. The peaks at 902.9 eV (U ) and 884.9 eV (V ) were ascribed to Ce 3+ [23]. The peaks at 916.6 eV (U ), 907.7 eV (U"), 898.2 eV (V ), and 888.6 eV (V") were indexed to Ce 4+ [20]. The relative content of Ce 3+ could be calculated by the following equation (listed in Table S1): The surface atomic compositions of these samples were nonstoichiometric values compared to the theoretical values due to the presence of a Ce 3+ state (Table S1). The Ce 3+ concentrations of CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were, respectively, 12.15%, 14.45%, and 16.54%. The charge compensation may have been responsible for the increase in Ce 3+ content of CeO 2 @C-2 and 3-Ag/CeO 2 @C-2. It was noticed that the Ce 3+ content (5.81%) of used 3-Ag/CeO 2 @C-2 after five cycles was lower than that of a fresh sample. This was attributed to the reduced charge compensation under long-term irradiation, leading to inferior photocatalytic stability [15]. The divided Gaussian peaks at 531.2 eV and 529.6 eV ( Figure 2B, Figures S2B and S3B ) were attributed to the defect oxygen (or oxygen vacancy) and lattice oxygen, respectively [15,20,31]. The varied ratios of defect oxygen (or oxygen vacancy) to lattice oxygen were greatly related to the Ce 3+ content in the as-obtained samples. The concentration of the defect oxygen (or oxygen vacancy) in the O 1s XPS spectrum of the used 3-Ag/CeO 2 @C-2 ( Figure 2B) was higher than that of a fresh sample. This was due to the released and adsorbed oxygen of CeO 2 and the defect structure of Ce 3+ -O v -Ce 4+ (O v -oxygen vacancy) under reduction and oxidation conditions [38,42]. The C1 spectra of CeO 2 @C-2 ( Figure 3C) and fresh and used 3-Ag/CeO 2 @C-2 ( Figure 2C) were divided into three Gaussian peaks at 288.4 eV, 285.3 eV, and 284.8 eV, which respectively belonged to C=O, C-OH, and C-C/C=C bonds. The splitting peaks at 374.2 eV and 368.19 eV in the Ag 3d XPS spectrum of fresh and used 3-Ag/CeO 2 @C-2 ( Figure 2D) were assigned to Ag 3d 3/2 and Ag 3d 5/2 , respectively. Although there was no change in the microstructure of used 3-Ag/CeO 2 @C-2 ( Figure S4), long-term irradiation induced a difference in the C 1s and Ag 3d XPS spectra between the fresh and used samples ( Figure 3C,D), leading to inferior charge transfer and separation [45,47]. The detailed microstructure and surface morphology of CeO2, CeO2@C, and Ag/CeO2@C were obtained by SEM and TEM. All of these samples were irregular nanosheets with thicknesses ranging from 10 nm to 20 nm ( Figures S4, S5, and 3A,B). Compared to the CeO2 precursor ( Figure S5A,B), there were many more fragments detected in the CeO2 after it was treated at high temperature ( Figure  S5C,D). With the assistance of carbon coating, fewer fragments and agglomerated nanosheets were detected in CeO2@C ( Figure S5E-G) and Ag/CeO2@C ( Figure S6), which was further confirmed by the TEM images ( Figure S7). With an increase in carbon content, the agglomeration of CeO2@C gradually disappeared and even formed single nanosheets, while the specific surface area of CeO2@C decreased (Table S2). Although the nanosheet structure of Ag/CeO2@C was not affected by the Ag doping, the fragment content increased with increasing Ag-doped content due to the damage effect of NaBH4 during the in situ reduction process ( Figure S6). In addition, the specific surface area of Ag-doped CeO2@C-2 deceased with increasing Ag-doped content (Table S2). As is shown in Figure 3, 3-Ag/CeO2@C-2 had irregular and fragmented nanosheets ( Figure 3A,B), and ultrafine Ag nanoparticles with a diameter of around 3 nm (dark section in the red circle) were anchored on the CeO2@C-2 ( Figure 3C,D and Figure S7E,F), which was evidenced by the HRTEM images ( Figure 3E,F). The spacing distances between neighboring lattice fringes of the (111) plane of cubic CeO2 and the (111) facet of cubic Ag ( Figure 3E,F) were, respectively, 0.312 nm and 0.24 nm, which agreed with the XRD pattern of 3-Ag/CeO2@C-2. The elemental distribution of 3-Ag/CeO2@C-2 was further obtained by the elemental mapping images. As is shown in Figure 3G The detailed microstructure and surface morphology of CeO 2 , CeO 2 @C, and Ag/CeO 2 @C were obtained by SEM and TEM. All of these samples were irregular nanosheets with thicknesses ranging from 10 nm to 20 nm ( Figures S4 and S5, and Figure 3A,B). Compared to the CeO 2 precursor ( Figure S5A,B), there were many more fragments detected in the CeO 2 after it was treated at high temperature ( Figure S5C,D). With the assistance of carbon coating, fewer fragments and agglomerated nanosheets were detected in CeO 2 @C ( Figure S5E-G) and Ag/CeO 2 @C ( Figure S6), which was further confirmed by the TEM images ( Figure S7). With an increase in carbon content, the agglomeration of CeO 2 @C gradually disappeared and even formed single nanosheets, while the specific surface area of CeO 2 @C decreased (Table S2). Although the nanosheet structure of Ag/CeO 2 @C was not affected by the Ag doping, the fragment content increased with increasing Ag-doped content due to the damage effect of NaBH 4 during the in situ reduction process ( Figure S6). In addition, the specific surface area of Ag-doped CeO2@C-2 deceased with increasing Ag-doped content (Table S2). As is shown in Figure 3, 3-Ag/CeO 2 @C-2 had irregular and fragmented nanosheets ( Figure 3A,B), and ultrafine Ag nanoparticles with a diameter of around 3 nm (dark section in the red circle) were anchored on the CeO 2 @C-2 ( Figure 3C,D and Figure S7E,F), which was evidenced by the HRTEM images ( Figure 3E,F). The spacing distances between neighboring lattice fringes of the (111) plane of cubic CeO 2 and the (111) facet of cubic Ag ( Figure 3E,F) were, respectively, 0.312 nm and 0.24 nm, which agreed with the XRD pattern of 3-Ag/CeO 2 @C-2. The elemental distribution of 3-Ag/CeO 2 @C-2 was further obtained by the elemental mapping images. As is shown in Figure 3G-J, there were Ce, O, C, and Ag elements existing in the obtained sample and discontinuous and monodisperse distributions of Ag elements on the bulk surface, indicating the formation of Ag QDs anchored in CeO 2 @C nanosheets. The molecular structures of the obtained CeO2, CeO2@C, and Ag/CeO2@C were obtained by FT-IR, as shown in Figures S8 and S9. The peaks at 3440 and 1640 cm −1 were ascribed to the stretching vibration and bending vibration of the O-H group of absorbed water and surface hydroxyl [9,11]. The peak at 1539 cm −1 was attributed to the H-O-H bending vibration of water molecules. The band peaks around 2921 cm −1 , 2847 cm −1 , and 1377 cm −1 were due to the bending vibration of the C-H group. The peaks at 2362 cm −1 and 2340 cm −1 could be assigned to the stretching vibrations of C=O groups of adsorbed CO2 in the air. The peaks at 676 cm −1 , 567 cm −1 , and 475 cm −1 were attributed to the vibration of metal oxygen bonds [41]. The optical properties of CeO2, CeO2@C-2, and 3-Ag/CeO2@C-2 were obtained by UV-Vis DRS, as shown in Figure 4. Compared to CeO2 and CeO2@C-2, 3-Ag/CeO2@C-2 exhibited strong visible-light-harvesting capacity due to the SPR effect of Ag QDs [18,22,23]. According to the plot of (αhv) 1/2 versus (hv), the calculated band gap energy of 3-Ag/CeO2@C-2 (2.47 eV) was lower than those of CeO2 (2.61 eV) and CeO2@C-2 (2.86 eV). The photoluminescence (PL) spectra ( Figure 5) indicated that the PL peak intensity of 3-Ag/CeO2@C-2 was also weaker than those of CeO2 and CeO2@C-2 after an excitation at a 300-nm wavelength [25]. On the basis of the standard quantum efficiency of 100% formed from the absorbance at the excitation wavelength and the photoluminescence intensity, the estimated fluorescence efficiency of 3-Ag/CeO2@C-2 (13.21%) was higher than those of CeO2 (5.72%) and CeO2@C-2 (8.36%), meaning a lower recombination of charge carriers over 3-Ag/CeO2@C-2. The broad emission band around 350-550 nm was responsible for Ce 3+ ions and oxide defects in CeO2. In addition, the combined effect of carbon coating and Ag QD-doping The molecular structures of the obtained CeO 2 , CeO 2 @C, and Ag/CeO 2 @C were obtained by FT-IR, as shown in Figures S8 and S9. The peaks at 3440 and 1640 cm −1 were ascribed to the stretching vibration and bending vibration of the O-H group of absorbed water and surface hydroxyl [9,11]. The peak at 1539 cm −1 was attributed to the H-O-H bending vibration of water molecules. The band peaks around 2921 cm −1 , 2847 cm −1 , and 1377 cm −1 were due to the bending vibration of the C-H group. The peaks at 2362 cm −1 and 2340 cm −1 could be assigned to the stretching vibrations of C=O groups of adsorbed CO 2 in the air. The peaks at 676 cm −1 , 567 cm −1 , and 475 cm −1 were attributed to the vibration of metal oxygen bonds [41]. The optical properties of CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were obtained by UV-Vis DRS, as shown in Figure 4. Compared to CeO 2 and CeO 2 @C-2, 3-Ag/CeO 2 @C-2 exhibited strong visible-light-harvesting capacity due to the SPR effect of Ag QDs [18,22,23]. According to the plot of (αhv) 1/2 versus (hv), the calculated band gap energy of 3-Ag/CeO 2 @C-2 (2.47 eV) was lower than those of CeO 2 (2.61 eV) and CeO 2 @C-2 (2.86 eV). The photoluminescence (PL) spectra ( Figure 5) indicated that the PL peak intensity of 3-Ag/CeO 2 @C-2 was also weaker than those of CeO 2 and CeO 2 @C-2 after an excitation at a 300-nm wavelength [25]. On the basis of the standard quantum efficiency of 100% formed from the absorbance at the excitation wavelength and the photoluminescence intensity, the estimated fluorescence efficiency of 3-Ag/CeO 2 @C-2 (13.21%) was higher than those of CeO 2 (5.72%) and CeO 2 @C-2 (8.36%), meaning a lower recombination of charge carriers over 3-Ag/CeO 2 @C-2. The broad emission band around 350-550 nm was responsible for Ce 3+ ions and oxide defects in CeO 2 . In addition, the combined effect of carbon coating and Ag QD-doping was helpful for efficient charge transfer and high resistance to the recombination of electron-hole pairs [41,45]. was helpful for efficient charge transfer and high resistance to the recombination of electron-hole pairs [41,45].  The separation efficiency of photocatalytic electron-hole pairs was evaluated by electrochemical impedance spectroscopy (EIS), in which the arc radius represented the transfer rate of the photocatalyst charge. In contrast to CeO2 and CeO2@C-2, 3-Ag/CeO2@C-2 exhibited a higher transient photocurrent under visible-light irradiation ( Figure 6A) and a smaller arc radius of electrochemical impedance ( Figure 6B). The small radius of 3-Ag/CeO2@C-2 suggested that the low resistance was suitable for the efficient separation of charge carriers in the obtained 3-Ag/CeO2@C-2 photocatalyst. Although a large specific surface area is favorable in reducing the diffusion length of charge carriers and further prompting charge transfer, the SPR effect of Ag QDs and a tight interface between carbon and CeO2 nanosheets are crucial for the enhanced separation of charge carriers and the restrained recombination of photoexcited electron-hole pairs [47,48]. Mott-Schottky curves of the obtained CeO2, CeO2@C-2, and 3-Ag/CeO2@C-2 were performed for an evaluation of the semiconductor type and flat band potentials (Vfb), where the Vfb values of these samples could be obtained from the x intercept by prolonging the linear part of the Mott-Schottky curves on the potential axis ( Figure  S10A  was helpful for efficient charge transfer and high resistance to the recombination of electron-hole pairs [41,45].  The separation efficiency of photocatalytic electron-hole pairs was evaluated by electrochemical impedance spectroscopy (EIS), in which the arc radius represented the transfer rate of the photocatalyst charge. In contrast to CeO2 and CeO2@C-2, 3-Ag/CeO2@C-2 exhibited a higher transient photocurrent under visible-light irradiation ( Figure 6A) and a smaller arc radius of electrochemical impedance ( Figure 6B). The small radius of 3-Ag/CeO2@C-2 suggested that the low resistance was suitable for the efficient separation of charge carriers in the obtained 3-Ag/CeO2@C-2 photocatalyst. Although a large specific surface area is favorable in reducing the diffusion length of charge carriers and further prompting charge transfer, the SPR effect of Ag QDs and a tight interface between carbon and CeO2 nanosheets are crucial for the enhanced separation of charge carriers and the restrained recombination of photoexcited electron-hole pairs [47,48] The separation efficiency of photocatalytic electron-hole pairs was evaluated by electrochemical impedance spectroscopy (EIS), in which the arc radius represented the transfer rate of the photocatalyst charge. In contrast to CeO 2 and CeO 2 @C-2, 3-Ag/CeO 2 @C-2 exhibited a higher transient photocurrent under visible-light irradiation ( Figure 6A) and a smaller arc radius of electrochemical impedance ( Figure 6B). The small radius of 3-Ag/CeO 2 @C-2 suggested that the low resistance was suitable for the efficient separation of charge carriers in the obtained 3-Ag/CeO 2 @C-2 photocatalyst. Although a large specific surface area is favorable in reducing the diffusion length of charge carriers and further prompting charge transfer, the SPR effect of Ag QDs and a tight interface between carbon and CeO 2 nanosheets are crucial for the enhanced separation of charge carriers and the restrained recombination of photoexcited electron-hole pairs [47,48]. Mott-Schottky curves of the obtained CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were performed for an evaluation of the semiconductor type and flat band potentials (V fb ), where the V fb values of these samples could be obtained from the x intercept by prolonging the linear part of the Mott-Schottky curves on the potential axis ( Figure S10A). The V fb values of CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were −0.83 V, −0.65 V, and −0.58 V versus a KCl-saturated calomel electrode, respectively. Hence, the valence band values of CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were −0.59 V, −0.41 V, and −0.34 V, respectively. According to the calculated band gap energy, the conduction band values of CeO 2 , CeO 2 @C-2, and 3-Ag/CeO 2 @C-2 were 2.27 V, 2.20 V, and 2.13 V, respectively. Compared to 3-Ag/CeO 2 @C-2 alone with a bias voltage of 1.0 V, the addition of visible-light irradiation with a light-power intensity of 240 mW cm −2 (λ > 420 nm) could enhance the current response ( Figure S10B), indicating efficient photoinduced charge in the visible light region.  CeO2-based composites were performed for the visible-light-driven photocatalytic reduction of Cr(VI) ions and the photodegradation of tetracycline hydrochloride (TCH). Compared to CeO2 nanosheets, carbon-coated composites exhibited better adsorption-photocatalytic activity under the same conditions, especially CeO2@C-2 ( Figure S11). The enhanced adsorption-photocatalysis behaviors of CeO2@C were due to sufficient active sites of carbon shells and the oxygen vacancy of CeO2 cores. In addition, this was ascribed to the trap effect of carbon shells for enhanced charge transfer and efficient charge carrier separation [9,49,50]. With the assistance of carbon shells, the excited electrons could efficiently escape from the conduction band (CB) of CeO2 to amorphous carbon, achieving the separation of electron-hole pairs and restraining their recombination. The SPR effect of Ag QDs could further strengthen the photocatalytic activity of CeO2@C-2 ( Figure 7A). The photocatalytic activity of Ag/CeO2@C-2 for Cr(VI) removal increased and then decreased with an increase in Ag-doping content. In addition, Ag/CeO2@C-2 composites also exhibited excellent photocatalytic activity for the visible-light-driven photodegradation of TCH compared to CeO2@C-2, as shown in Figure S12. A similar tendency in the effects of Ag-doping content on photocatalytic activity was obtained for the removal of TCH in the visible-light region. Among these Ag QDs doped composites, CeO2@C-2 and 3-Ag/CeO2@C-2 with an Ag-doping content of 5.41% presented the best photocatalytic activity. Due to the limited active sites, an excess of Cr(VI) ions could not efficiently access the active sites, leading to inferior photocatalytic efficiency. Hence, the removal efficiency of 3-Ag/CeO2@C-2 decreased with increasing concentrations of Cr(VI) ions (ranging from 10 mg L −1 to 40 mg·L −1 (Figure 7B)). The optimum pH value facilitated the reaction between hydroxyl (OH − ) (or H + ) ions and radical species (such as eand h + ) to generate ·O2 − and ·OH radicals [20,47]. In a photocatalytic reaction system, H + ions can react with ·O2 − to form ·OOH radicals, and ·OOH can react with H + ions to generate H2O2 [3]. Subsequently, H2O2 is likely to react with eto form ·OH and OH − , of which OH − ions are scavenged by h + to produce ·OH [9]. In addition, ·OH is also generated from the reaction between H2O and h + [15]. An excess of OH − ions can quench the above chain reactions, leading to inferior photocatalytic activity ( Figure 7C).
Inorganic ions such as chlorine (Cl − ), sulfite (SO3 2− ), sulfate (SO4 2− ), and phosphate (H2PO4 − ) could affect the photocatalytic activity of 3-Ag/CeO2@C-2 in Cr(VI) removal. As is shown in Figure 7D, inorganic ions could restrain photocatalytic activity compared to 3-Ag/CeO2@C-2 alone in Cr(VI) removal under the same conditions. The removal efficiency of Cr(VI) was remarkably inhibited by H2PO4 − in comparison to Cl − , SO4 2− , and SO3 2− . The impeding effect of inorganic ions was due to ·OH scavengers and reduced active sites [9,10]. On the one hand, inorganic ions served as ·OH scavengers and competed with Cr(VI) ions for ·OH radicals. Although the ion radicals were suitable for the CeO 2 -based composites were performed for the visible-light-driven photocatalytic reduction of Cr(VI) ions and the photodegradation of tetracycline hydrochloride (TCH). Compared to CeO 2 nanosheets, carbon-coated composites exhibited better adsorption-photocatalytic activity under the same conditions, especially CeO 2 @C-2 ( Figure S11). The enhanced adsorption-photocatalysis behaviors of CeO 2 @C were due to sufficient active sites of carbon shells and the oxygen vacancy of CeO 2 cores. In addition, this was ascribed to the trap effect of carbon shells for enhanced charge transfer and efficient charge carrier separation [9,49,50]. With the assistance of carbon shells, the excited electrons could efficiently escape from the conduction band (CB) of CeO 2 to amorphous carbon, achieving the separation of electron-hole pairs and restraining their recombination. The SPR effect of Ag QDs could further strengthen the photocatalytic activity of CeO 2 @C-2 ( Figure 7A). The photocatalytic activity of Ag/CeO 2 @C-2 for Cr(VI) removal increased and then decreased with an increase in Ag-doping content. In addition, Ag/CeO 2 @C-2 composites also exhibited excellent photocatalytic activity for the visible-light-driven photodegradation of TCH compared to CeO 2 @C-2, as shown in Figure S12. A similar tendency in the effects of Ag-doping content on photocatalytic activity was obtained for the removal of TCH in the visible-light region. Among these Ag QDs doped composites, CeO 2 @C-2 and 3-Ag/CeO 2 @C-2 with an Ag-doping content of 5.41% presented the best photocatalytic activity. Due to the limited active sites, an excess of Cr(VI) ions could not efficiently access the active sites, leading to inferior photocatalytic efficiency. Hence, the removal efficiency of 3-Ag/CeO 2 @C-2 decreased with increasing concentrations of Cr(VI) ions (ranging from 10 mg L −1 to 40 mg·L −1 (Figure 7B)). The optimum pH value facilitated the reaction between hydroxyl (OH − ) (or H + ) ions and radical species (such as e − and h + ) to generate ·O 2 − and ·OH radicals [20,47]. In a photocatalytic reaction system, H + ions can react with ·O 2 − to form ·OOH radicals, and ·OOH can react with H + ions to generate H 2 O 2 [3]. Subsequently, H 2 O 2 is likely to react with e − to form ·OH and OH − , of which OH − ions are scavenged by h + to produce ·OH [9]. In addition, ·OH is also generated from the reaction between H 2 O and h + [15]. An excess of OH − ions can quench the above chain reactions, leading to inferior photocatalytic activity ( Figure 7C).
shown in Figure 8, the photocatalytic activity of 3-Ag/CeO2@C-2 remained slightly changed after five cycles. However, the mass loss of 3-Ag/CeO2@C-2 in the sedimentation and transferring processes induced a decrease in photocatalytic activity in each cycle test. SEM images ( Figure S4) confirmed that the structure of the used 3-Ag/CeO2@C-2 remained unchanged after five cycles of photocatalytic reactions. However, the surface compositions and chemical states of the used 3-Ag/CeO2@C-2 ( Figure  2) were different from the fresh composites due to long-term photocorrosion.  As is shown in Figure 9, ESR signals of photoinduced radicals such as ·OH and ·O2 − were clearly obtained, in which the intensities increased with an increase in the irradiation time of visible light. Quenching testing ( Figure S13) indicated that the photocatalytic reaction was suppressed by tert- could affect the photocatalytic activity of 3-Ag/CeO 2 @C-2 in Cr(VI) removal. As is shown in Figure 7D, inorganic ions could restrain photocatalytic activity compared to 3-Ag/CeO 2 @C-2 alone in Cr(VI) removal under the same conditions. The removal efficiency of Cr(VI) was remarkably inhibited by H 2 PO 4 − in comparison to Cl − , SO 4 2− , and SO 3 2− . The impeding effect of inorganic ions was due to ·OH scavengers and reduced active sites [9,10]. On the one hand, inorganic ions served as ·OH scavengers and competed with Cr(VI) ions for ·OH radicals. Although the ion radicals were suitable for the oxidizing pollutants, their low oxidoreductive potential induced inferior photocatalytic rates compared to ·OH, especially in acidic conditions [13]. On the other hand, inorganic ions could adsorb on the surface of 3-Ag/CeO 2 @C-2, leading to decreased surface-active sites for Cr(VI) ions. As is shown in Figure 8, the photocatalytic activity of 3-Ag/CeO 2 @C-2 remained slightly changed after five cycles. However, the mass loss of 3-Ag/CeO 2 @C-2 in the sedimentation and transferring processes induced a decrease in photocatalytic activity in each cycle test. SEM images ( Figure S4) confirmed that the structure of the used 3-Ag/CeO 2 @C-2 remained unchanged after five cycles of photocatalytic reactions. However, the surface compositions and chemical states of the used 3-Ag/CeO 2 @C-2 ( Figure 2) were different from the fresh composites due to long-term photocorrosion.
As is shown in Figure 9, ESR signals of photoinduced radicals such as ·OH and ·O 2 − were clearly obtained, in which the intensities increased with an increase in the irradiation time of visible light. Quenching testing ( Figure S13) indicated that the photocatalytic reaction was suppressed by tert-butyl alcohol (t-BuOH), benzoquinone (BQ), and ethylenediaminetetraacetic acid disodium salt (EDTA-2Na), especially p-BQ and t-BuOH. It was concluded that the ·OH and ·O 2 − radicals were vital for the photocatalytic activity of 3-Ag/CeO 2 @C-2. The enhanced photocatalytic capacity was due to the SPR effect of Ag QDs, an electron trapper of carbon shells, and the redox activity of the Ce(III)/Ce(IV) coupling. In other words, the photocatalytic activity was related to the amount of Ag + serving as an electron acceptor (Ag 2+ ↔ Ag 0 ) and/or a hole donor (Ag 2+ ↔ Ag + ) and the concentration of Ce 3+ and oxygen vacancies, which could promote the localization of charge carriers and prolong the separation of electron-hole pairs via trapping at energy levels close to the valance band or conduction band [13]. Previous work has suggested that the temperature-programmed reduction (TPR) of peaks belonging to the conversion of Ag 2+ and Ag + into Ag 0 is obtained at 405 K and 421 K, respectively [55]. Hence, a reaction between e − /h + pairs and Ag 0 /Ag + /Ag 2+ ions is likely to promote the photoreduction of Cr(VI) ions. The possible photocatalytic mechanism of the 3-Ag/CeO 2 @C-2 heterojunction is proposed in Figure 10.  (6) and (7)) [56,57]. In addition, OH− ions were adsorbed on the h + of the valance band to generate an ·OH radical. These radicals were responsible for the efficient photocatalytic reduction of Cr(VI) into Cr(III) ions (Equations (8)- (13)), which was confirmed by the Cr 2p XPS spectrum of the used 3-Ag/CeO 2 @C-2 after five cycles of photocatalytic reaction ( Figure S14). The splitting peaks at 588.2 eV and 579.2 eV were indexed to Cr(VI), and the peaks at 586.2 eV and 576.4 eV were indexed to Cr(III), indicating a conversion of Cr(VI) ions into Cr(III) ions according to the electron transfer process [56]. Under acidic conditions, Cr(VI) ions could react with electrons and ·OH to generate Cr(III) ions (Equations (8)- (11)). Under alkaline conditions, low H + ions could react with Cr(VI) ions to form Cr 3+ ions (Equation (12)) and further generate Cr(OH) 3 loaded on the surface of 3-Ag/CeO 2 @C-2 bulks (Equation (13)), leading to inferior light adsorption capacity [58][59][60]. In addition, TCH could react with ·OH and h + to form small molecules, which was evaluated by UPLC-MS ( Figure S15). Under visible-light irradiation, TCH molecules were degraded via N-C bond cleavage and hydroxylation, and then these intermediates could react with the ·OH radical to destroy C2-C3 double bonds and eliminate NH 3 : they subsequently disintegrated into small molecules and even H 2 O and CO 2 through demethylation, deamination, and ·OH attack [55,56]. Ag QDs doping and carbon shells could serve as an interfacial charge transfer medium and a recombination center, accelerating the photocatalytic reaction: 14H  As is shown in Figure 9, ESR signals of photoinduced radicals such as ·OH and ·O2 − were clearly obtained, in which the intensities increased with an increase in the irradiation time of visible light. Quenching testing ( Figure S13) indicated that the photocatalytic reaction was suppressed by tert-

Conclusions
Ag/CeO2@C nanosheets presented better photocatalytic activity than did CeO2 and CeO2@C for Cr(VI) removal in the visible-light region. The photocatalytic activity of Ag/CeO2@C increased and then decreased with an increase in carbon content, Ag doping content, and pH value. Inorganic ions and Cr(VI) content had a negative effect on the removal efficiency of Cr(VI) under the same

Conclusions
Ag/CeO2@C nanosheets presented better photocatalytic activity than did CeO2 and CeO2@C for Cr(VI) removal in the visible-light region. The photocatalytic activity of Ag/CeO2@C increased and then decreased with an increase in carbon content, Ag doping content, and pH value. Inorganic ions and Cr(VI) content had a negative effect on the removal efficiency of Cr(VI) under the same conditions. The best removal efficiency and the most superior photocatalytic stability after five cycles were achieved by 3-Ag/CeO2@C-2 in the visible-light-driven removal of Cr(VI) ions as well as TCH. The SPR effect of Ag QDs, an electron trapper of carbon shells, and the redox activity of the

Conclusions
Ag/CeO 2 @C nanosheets presented better photocatalytic activity than did CeO 2 and CeO 2 @C for Cr(VI) removal in the visible-light region. The photocatalytic activity of Ag/CeO 2 @C increased and then decreased with an increase in carbon content, Ag doping content, and pH value. Inorganic ions and Cr(VI) content had a negative effect on the removal efficiency of Cr(VI) under the same conditions. The best removal efficiency and the most superior photocatalytic stability after five cycles were achieved by 3-Ag/CeO 2 @C-2 in the visible-light-driven removal of Cr(VI) ions as well as TCH. The SPR effect of Ag QDs, an electron trapper of carbon shells, and the redox activity of the Ce(III)/Ce(IV) coupling played a vital role in the transfer and separation of charge carriers. •O 2 − and •OH radicals were the primary active species of 3-Ag/CeO 2 @C-2 in the photocatalytic system.