Oxygen Vacancies-Rich S-Cheme BiOBr/CdS Heterojunction with Synergetic Effect for Highly Efficient Light Emitting Diode-Driven Pollutants Degradation

Recently, the use of semiconductor-based photocatalytic technology as an effective way to mitigate the environmental crisis attracted considerable interest. Here, the S-scheme BiOBr/CdS heterojunction with abundant oxygen vacancies (Vo-BiOBr/CdS) was prepared by the solvothermal method using ethylene glycol as a solvent. The photocatalytic activity of the heterojunction was investigated by degrading rhodamine B (RhB) and methylene blue (MB) under 5 W light-emitting diode (LED) light. Notably, the degradation rate of RhB and MB reached 97% and 93% in 60 min, respectively, which were better than that of BiOBr, CdS, and BiOBr/CdS. It was due to the construction of the heterojunction and the introduction of Vo, which facilitated the spatial separation of carriers and enhanced the visible-light harvest. The radical trapping experiment suggested that superoxide radicals (·O2−) acted as the main active species. Based on valence balance spectra, Mott-Schottky(M-S) spectra, and DFT theoretical calculations, the photocatalytic mechanism of the S-scheme heterojunction was proposed. This research provides a novel strategy for designing efficient photocatalysts by constructing S-scheme heterojunctions and introducing oxygen vacancies for solving environmental pollution.


Introduction
Nowadays, the environmental deterioration caused by rapid socioeconomic development is becoming increasingly serious [1][2][3]. Photocatalysis as a green technology can effectively solve these problems [4][5][6]. Semiconductor-based photocatalysts can degrade pollutants and transform CO 2 , and N 2 into high-value products through sunlight [7][8][9]. So far, many materials have attracted focus due to their considerable photocatalytic properties, such as TiO 2 , CdS, ZnO, SnO 2 , WO 3 , etc [10][11][12][13][14]. As a promising layered photocatalyst, BiOBr has been extensively investigated in the photocatalytic area because of its good stability. However, the charge carriers in BiOBr are easy to recombine and the light utilization of BiOBr is low. Therefore, constructing a high-efficiency BiOBr-based photocatalyst is still a challenge.

Synthesis
Vo-BiOBr/CdS heterojunction was constructed by a two-step solvothermal method. Firstly, 2.01 g of Cd(NO 3 ) 2 ·4H 2 O and 2.04 g of L-Cysteine were put into a beaker containing 30 mL ethylenediamine and stirred for 1 h. Then the above dispersion was transferred into a 100 mL of Teflon-lined stainless-steel autoclave and heated at 180 • C for 5 h. The resulting yellow product was washed with water and ethanol three times and dried at 60 • C to obtain CdS.

Characterization
XRD pattern was obtained with a Bruker D8 Advance X-ray diffractometer using Cu-Kα radiation (λ = 1.5405 nm). The X-ray photoelectron spectroscopy (XPS, Thermo Scientific ESCALAB 250Xi) was applied to get the chemical status and valence state of samples. The transmission electron microscopy (TEM), selected area electron diffraction (SAED), and energy dispersive X-ray (EDX) mapping and spectrum were measured by an FEI Tecnai G2 F20 S-Twin with a 200 kV accelerating voltage. UV-vis diffuse reflection spectra (UVvis DRS) were performed on a Lambda 650 UV-vis spectrophotometer at wavelengths of 200-800 nm. The photocurrent response, EIS, and M-S curves were measured on the CHI-660E electrochemical workstation.

Photocatalytic Test
The photocatalytic performance of the photocatalysts was assessed by the degradation of RhB (10 mg·L −1 ), and MB (20 mg·L −1 ) under visible light irradiation using a 5 W LED lamp (PerfectLight, λ > 420 nm). Typically, 20 mg of catalysts were put into 40 mL RhB (or MB) solution and stirred for 30 min in the dark. After irradiation, 4 mL of the liquid was sampled and centrifuged. The concentration of the solution was characterized using a UV-vis spectrometer.
The photocatalytic stability of the samples can be tested by cycling experiments of photocatalytic degradation of RhB. After one degradation, the samples were filtered, washed, and dried before repeating the degradation experiment.
To investigate the photocatalytic species, 5 mL 0.1 mmol/L radical scavengers (isopropanol, 1,4-benzoquinone, and KI) were added into the RhB solution in the above photocatalytic process.

Theoretical Calculations
First-principle calculations based on Density functional theory (DFT) were performed using the VASP software Package with plane wave techniques [31]. Perdew-Burke-Ernzerhof (PBE) [32] functional framework was used to describe the exchange-correlation functional interaction. The truncation energy of the Plane Wave Truncation Base Group was 500 eV. The energy band structures and density of states (DOS) of the 2 × 2 × 1 BiOBr and Vo-BiOBr supercell model and the hexagonal CdS cell model were calculated respectively ( Figure S1).

Structure and Morphology
Firstly, the CdS nanorods were synthesized via a solvothermal approach (Figure 1). Secondly, the Vo-BiOBr/CdS was prepared by dispersing the CdS in EG-H 2 O solution containing PVP, KBr, and Bi(NO 3 ) 3 . The oxygen vacancies were introduced by adding ethylene glycol (EG) to the solution [33]. The addition of PVP facilitated the creation of Vo because it tends to bind with the unsaturated Bi atom on the BiOBr surface to decrease the surface energy, which leads to the creation of more Vo [34]. In addition, PVP as a common capping agent can also affect the morphology of the catalyst [35].
Firstly, the CdS nanorods were synthesized via a solvothermal approach (Figure 1). Secondly, the Vo-BiOBr/CdS was prepared by dispersing the CdS in EG-H2O solution containing PVP, KBr, and Bi(NO3)3. The oxygen vacancies were introduced by adding ethylene glycol (EG) to the solution [33]. The addition of PVP facilitated the creation of Vo because it tends to bind with the unsaturated Bi atom on the BiOBr surface to decrease the surface energy, which leads to the creation of more Vo [34]. In addition, PVP as a common capping agent can also affect the morphology of the catalyst [35].  [37]. Notably, diffraction peaks at 21.92° and 44.72° were not observed in BiOBr, which may be due to the introduction of oxygen defects [38]. Furthermore, the intensity of (001) and (102) planes markedly decreased, while that of (110) increased compared with BiOBr. The possible reason was that when the relative intensity of (110)/(001) planes increased, the structure of Vo-BiOBr tended to form a hierarchical structure [39], which can be verified by BiOBr nanoplate with an exposed (001) facet and Vo-BiOBr nanosheets with a (110) facet. The    [37]. Notably, diffraction peaks at 21.92 • and 44.72 • were not observed in BiOBr, which may be due to the introduction of oxygen defects [38]. Furthermore, the intensity of (001) and (102) planes markedly decreased, while that of (110) increased compared with BiOBr. The possible reason was that when the relative intensity of (110)/(001) planes increased, the structure of Vo-BiOBr tended to form a hierarchical structure [39], which can be verified by BiOBr nanoplate with an exposed (001) facet and Vo-BiOBr nanosheets with a (110) facet. The Firstly, the CdS nanorods were synthesized via a solvothermal approach (Figure 1). Secondly, the Vo-BiOBr/CdS was prepared by dispersing the CdS in EG-H2O solution containing PVP, KBr, and Bi(NO3)3. The oxygen vacancies were introduced by adding ethylene glycol (EG) to the solution [33]. The addition of PVP facilitated the creation of Vo because it tends to bind with the unsaturated Bi atom on the BiOBr surface to decrease the surface energy, which leads to the creation of more Vo [34]. In addition, PVP as a common capping agent can also affect the morphology of the catalyst [35].  [37]. Notably, diffraction peaks at 21.92° and 44.72° were not observed in BiOBr, which may be due to the introduction of oxygen defects [38]. Furthermore, the intensity of (001) and (102) planes markedly decreased, while that of (110) increased compared with BiOBr. The possible reason was that when the relative intensity of (110)/(001) planes increased, the structure of Vo-BiOBr tended to form a hierarchical structure [39], which can be verified by BiOBr nanoplate with an exposed (001) facet and Vo-BiOBr nanosheets with a (110) facet. The (002)   The morphology of the photocatalysts was studied by TEM technology. Figure 3a,b showed Vo-BiOBr as nanosheets and CdS as nanorods with a diameter of tens of nanometers. In Figure 3c,d, CdS nanorods were distributed on the surface of Vo-BiOBr nanosheets. According to the HRTEM image of Vo-BiOBr/CdS (Figure 3e), the lattice spacing of 0.281 nm assigned to the (110) plane of Vo-BiOBr and 0.332 nm assigned to the (002) plane of CdS was observed [40,41]. Furthermore, a distinct interface could be found, demonstrating the construction of a heterojunction between Vo-BiOBr and CdS [42]. The obscured, circled lattice fringes were observed (Figure 3e), meaning that defects existed in Vo-BiOBr [43]. In addition, the SAED pattern of Vo-BiOBr/CdS (Figure 3f) showed clear diffraction rings, ing of 0.281 nm assigned to the (110) plane of Vo-BiOBr and 0.332 nm assigned to the (002) plane of CdS was observed [40,41]. Furthermore, a distinct interface could be found, demonstrating the construction of a heterojunction between Vo-BiOBr and CdS [42]. The obscured, circled lattice fringes were observed (Figure 3e), meaning that defects existed in Vo-BiOBr [43]. In addition, the SAED pattern of Vo-BiOBr/CdS (Figure 3f Figure  4b-f. The spectra of Bi 5f exhibited two peaks at 158.78 eV and 164.08 eV (Figure 4b), which can be ascribed to Bi 4f7/2 and Bi 4f5/2, respectively, demonstrating the presence of Bi 3+ . As for Br 3d (Figure 4c), two peaks located at 67.88 eV and 68.93 eV, can be ascribed to Br  (Figure 4b), which can be ascribed to Bi 4f 7/2 and Bi 4f 5/2 , respectively, demonstrating the presence of Bi 3+ . As for Br 3d (Figure 4c), two peaks located at 67.88 eV and 68.93 eV, can be ascribed to Br 3d 5/2 and Br 3d 3/2 , respectively, indicating the chemical states of Br − . In Figure 4d, the spectra of O 1s can be fitted with three peaks at 529.38 eV, 530.83 eV, and 532.33 eV inferring lattice oxygen (Bi-O), oxygen vacancies, and absorbed oxygen, respectively [44]. It is worth noting that Vo-BiOBr/CdS had a larger area for the characteristic peak representing the oxygen vacancy than BiOBr (Table S1), suggesting that the reduction of ethylene glycol is vital to forming Vo. More significantly, the peaks of Bi 4f, Br 3d, and O 1s of Vo-BiOBr moved toward lower binding energy compared with BiOBr ( Figure S2), which proved that the charge density increased because of the local electrons of oxygen vacancies. In Figure 4e, the peaks at 404.68 eV and 411.43 eV matched Cd 3d 5/2 and Cd 3d 3/2 , respectively; the other peaks originated from satellite peaks of Cd 2+ . The peaks at 161.08 eV and 162.48 eV were ascribed to S 2p 3/2 and S 2p 1/2 (Figure 4f), respectively [45]. Notably, the Bi 4f, Br 3d, Cd 3d, S 2p, and O 1s peaks of Vo-BiOBr/CdS shifted to higher binding energy compared with Vo-BiOBr or CdS, which may be due to the changing of the surface-electron density, suggesting the formation of a heterojunction in Vo-BiOBr/CdS. Consequently, when CdS coupled with BiOBr, the transfer channel of carriers would be constructed, causing a charge redistribution between Vo-BiOBr and CdS. This facilitated the charge transfer and enhanced the photocatalytic activity.

Optical Property
The optical performance of as-synthesized samples was examined by UV-vis DRS (Figure 5a). Vo-BiOBr exhibited a stronger absorption of ultraviolet and visible light than that of BiOBr, and the increased absorption intensity in 550-700 nm was evident. This was attributed to the introduction of Vo, which enhanced the absorption intensity of visible

Optical Property
The optical performance of as-synthesized samples was examined by UV-vis DRS (Figure 5a). Vo-BiOBr exhibited a stronger absorption of ultraviolet and visible light than that of BiOBr, and the increased absorption intensity in 550-700 nm was evident. This was attributed to the introduction of Vo, which enhanced the absorption intensity of visible light [46]. When coupling with CdS, Vo-BiOBr/CdS showed an enhanced absorption intensity and an occurrence of a red shift compared with Vo-BiOBr. It indicated that oxygen vacancies and the heterojunction greatly enhanced the visible light absorption ability. The bandgap energies (Eg) of samples could be identified by the following formula (αhυ)n= A(hυ − Eg) (Here, n = 2 for direct semiconductors and n =1/2 for indirect semiconductors) [47]. As for Vo-BiOBr and CdS, the n values were 1/2 and 2, respectively [48]. Hence, the bandgap energies of BiOBr, Vo-BiOBr, and CdS were 2.78, 2.14, and 2.40 eV, respectively ( Figure 5b). Notably, the Eg of Vo-BiOBr was smaller than that of BiOBr, suggesting that oxygen vacancy could narrow the band gap energy [49], which is helpful for the utilization of visible light.

Enhanced Photocatalytic Performance
The photocatalytic activity of Vo-BiOBr/CdS was evaluated by the degradation of dyes under LED light irradiation. The photocatalytic process started when the dye molecules were adsorbed onto the photocatalyst surface in the dark to achieve equilibrium. In Figure 6a, the concentration of RhB had no significant change in the absence of photocatalysts, indicating high stability. The CdS and pristine BiOBr without oxygen vacancies exhibited low photodegradation efficiency: 42% and 79% RhB were decomposed after 60 min, respectively. When the BiOBr was coupled to CdS nanorods, the composites (Bi-OBr/CdS) exhibited slightly enhanced photodegradation efficiency, owing to the construction of heterojunction, which promoted the fast separation of photogenerated carriers [50]. Compared with BiOBr, CdS, and BiOBr/CdS, the Vo-BiOBr/CdS heterojunction displayed the highest photocatalytic performance when the mass ratio of CdS to Vo-BiOBr was 0.1 ( Figure S3). Nearly 97% of RhB was degraded when irradiated for 60 min under LED light, suggesting that the synergistic effect of Vo and the heterojunction played a crucial part in promoting photocatalytic activity. Under similar conditions, the Vo-Bi-OBr/CdS displayed a superior photocatalytic performance to that of the reported (Table  S2). The bandgap energies (Eg) of samples could be identified by the following formula (αhυ)n = A(hυ − E g ) (Here, n = 2 for direct semiconductors and n =1/2 for indirect semiconductors) [47]. As for Vo-BiOBr and CdS, the n values were 1/2 and 2, respectively [48]. Hence, the bandgap energies of BiOBr, Vo-BiOBr, and CdS were 2.78, 2.14, and 2.40 eV, respectively ( Figure 5b). Notably, the E g of Vo-BiOBr was smaller than that of BiOBr, suggesting that oxygen vacancy could narrow the band gap energy [49], which is helpful for the utilization of visible light.

Enhanced Photocatalytic Performance
The photocatalytic activity of Vo-BiOBr/CdS was evaluated by the degradation of dyes under LED light irradiation. The photocatalytic process started when the dye molecules were adsorbed onto the photocatalyst surface in the dark to achieve equilibrium. In Figure 6a, the concentration of RhB had no significant change in the absence of photocatalysts, indicating high stability. The CdS and pristine BiOBr without oxygen vacancies exhibited low photodegradation efficiency: 42% and 79% RhB were decomposed after 60 min, respectively. When the BiOBr was coupled to CdS nanorods, the composites (BiOBr/CdS) exhibited slightly enhanced photodegradation efficiency, owing to the construction of heterojunction, which promoted the fast separation of photogenerated carriers [50]. Compared with BiOBr, CdS, and BiOBr/CdS, the Vo-BiOBr/CdS heterojunction displayed the highest photocatalytic performance when the mass ratio of CdS to Vo-BiOBr was 0.1 ( Figure S3). Nearly 97% of RhB was degraded when irradiated for 60 min under LED light, suggesting that the synergistic effect of Vo and the heterojunction played a crucial part in promoting photocatalytic activity. Under similar conditions, the Vo-BiOBr/CdS displayed a superior photocatalytic performance to that of the reported (Table S2). To briefly represent the photocatalytic degradation rate toward RhB of samples, the first-order model (ln (Co/C) = kt) was adopted [51], and the fitting results were depicted in Figure 6c. The rate constant (k) of Vo-BiOBr/CdS was calculated to be 0.053 min −1 , about 3.4, 6.1, and 2.2 times as high as those of BiOBr, CdS, and BiOBr/CdS, respectively. Similarly, Vo-BiOBr/CdS showed the highest photocatalytic properties toward the MB ( Figure  6b). The degradation rate by Vo-BiOBr/CdS reached 94% after irradiation for 60 min. In Figure 6d, its corresponding k was 0.041 min −1 , about 9.2, 9.6, and 3.9 times as high as that of BiOBr, CdS, and BiOBr/CdS, respectively. The time-dependent UV-vis spectrum of Vo-BiOBr/CdS for degrading RhB and MB under LED light was illustrated in Figure 6e,f, respectively. A gradual decrease in the characteristic peak intensity of RhB and MB could be observed as time increased, which proved that RhB and MB were degrading [46]. To briefly represent the photocatalytic degradation rate toward RhB of samples, the first-order model (ln (C o /C) = kt) was adopted [51], and the fitting results were depicted in Figure 6c. The rate constant (k) of Vo-BiOBr/CdS was calculated to be 0.053 min −1 , about 3.4, 6.1, and 2.2 times as high as those of BiOBr, CdS, and BiOBr/CdS, respectively. Similarly, Vo-BiOBr/CdS showed the highest photocatalytic properties toward the MB (Figure 6b). The degradation rate by Vo-BiOBr/CdS reached 94% after irradiation for 60 min. In Figure 6d, its corresponding k was 0.041 min −1 , about 9.2, 9.6, and 3.9 times as high as that of BiOBr, CdS, and BiOBr/CdS, respectively. The time-dependent UV-vis spectrum of Vo-BiOBr/CdS for degrading RhB and MB under LED light was illustrated in Figure 6e,f, respectively. A gradual decrease in the characteristic peak intensity of RhB and MB could be observed as time increased, which proved that RhB and MB were degrading [46].
During the wastewater treatment process, the concentration of RhB could affect the degradation rate of photocatalysts. As depicted in Figure 7a, the impact of RhB concentrations on the degradation rate of Vo-BiOBr/CdS was also conducted. With the RhB concentration increasing, the removal rate decreased. After 60 min of irradiation, the degradation rate reached 96%, 90%, and 86% when the initial concentrations were 10 mg/L, 20 mg/L, and 30 mg/L, respectively. The impact of catalyst dosage on removal efficiency was also explored; as illustrated in Figure 7b, the degradation rate was enhanced when the dosage of Vo-BiOBr/CdS was changed from 10 mg to 20 mg. However, when it increased from 20 to 30 and 40 mg, degradation efficiencies just slightly increased. Therefore, 20 mg was the best choice considering the cost and degradation efficiency. During the wastewater treatment process, the concentration of RhB could affect the degradation rate of photocatalysts. As depicted in Figure 7a, the impact of RhB concentrations on the degradation rate of Vo-BiOBr/CdS was also conducted. With the RhB concentration increasing, the removal rate decreased. After 60 min of irradiation, the degradation rate reached 96%, 90%, and 86% when the initial concentrations were 10 mg/L, 20 mg/L, and 30 mg/L, respectively. The impact of catalyst dosage on removal efficiency was also explored; as illustrated in Figure 7b, the degradation rate was enhanced when the dosage of Vo-BiOBr/CdS was changed from 10 mg to 20 mg. However, when it increased from 20 to 30 and 40 mg, degradation efficiencies just slightly increased. Therefore, 20 mg was the best choice considering the cost and degradation efficiency. To investigate the degradation and reduction efficiency for pollutants other than dyes, the photodegradation of TC and photoreduction of Cr (VI) to Cr(III) were performed. The degradation and reduction rates of TC and Cr (VI) by Vo-BiOBr/CdS were 71% and 99%, respectively ( Figure S4). The above results proved that the photocatalysts exhibited efficient performance for different pollutants. In addition, the actual wastewater often contained various pollutants. Hence, mixed solutions containing RhB, Cr 6+ , and MB by volume ratio 1:1:1 were employed to investigate the effect of co-existing contaminants on photodegradation and photoreduction efficiencies. As depicted in Figure 7c,d, Vo-Bi-OBr/CdS exhibited excellent photocatalytic activity toward the mixed solution. It is worth noting that the photodegradation rate toward RhB decreased from 97% to 82%, possibly due to competitive interactions between various pollutants [52].
The photocurrent response and EIS measurements were performed to study a better photocatalytic performance of Vo-BiOBr/CdS compared with BiOBr, CdS, and BiOBr/CdS. To investigate the degradation and reduction efficiency for pollutants other than dyes, the photodegradation of TC and photoreduction of Cr (VI) to Cr(III) were performed. The degradation and reduction rates of TC and Cr (VI) by Vo-BiOBr/CdS were 71% and 99%, respectively ( Figure S4). The above results proved that the photocatalysts exhibited efficient performance for different pollutants. In addition, the actual wastewater often contained various pollutants. Hence, mixed solutions containing RhB, Cr 6+ , and MB by volume ratio 1:1:1 were employed to investigate the effect of co-existing contaminants on photodegradation and photoreduction efficiencies. As depicted in Figure 7c,d, Vo-BiOBr/CdS exhibited excellent photocatalytic activity toward the mixed solution. It is worth noting that the photodegradation rate toward RhB decreased from 97% to 82%, possibly due to competitive interactions between various pollutants [52].
The photocurrent response and EIS measurements were performed to study a better photocatalytic performance of Vo-BiOBr/CdS compared with BiOBr, CdS, and BiOBr/CdS. As illustrated in Figure 8a, the Vo-BiOBr/CdS showed higher photocurrent density compared with BiOBr/CdS, indicating that introducing Vo could facilitate the separation of charge carriers. Additionally, the photocurrent density of BiOBr/CdS was higher than that of CdS and BiOBr, demonstrating that heterostructures could also facilitate the splitting of charge carriers. Subsequently, the quantum efficiencies of photocatalysts were studied by EIS. In Figure 8b, Vo-BiOBr/CdS showed the smallest radius among the as-prepared photocatalysts in the Nyquist plots, representing smoother interface charge transfer paths and enhanced mobility [53]. Hence, the increased photocatalytic performance could attribute to the synergistic effect between the oxygen vacancies and the heterojunction, which accelerated the fast separation of charge carriers. The cycle experiments of RhB on Vo-BiOBr/CdS were performed to examine the stability of the photocatalyst (Figure 8c) and the degradation efficiency slightly decreased, indicating that Vo-BiOBr/CdS had good stability. aterials 2023, 13, x FOR PEER REVIEW 11 of 19 As illustrated in Figure 8a, the Vo-BiOBr/CdS showed higher photocurrent density compared with BiOBr/CdS, indicating that introducing Vo could facilitate the separation of charge carriers. Additionally, the photocurrent density of BiOBr/CdS was higher than that of CdS and BiOBr, demonstrating that heterostructures could also facilitate the splitting of charge carriers. Subsequently, the quantum efficiencies of photocatalysts were studied by EIS. In Figure 8b, Vo-BiOBr/CdS showed the smallest radius among the as-prepared photocatalysts in the Nyquist plots, representing smoother interface charge transfer paths and enhanced mobility [53]. Hence, the increased photocatalytic performance could attribute to the synergistic effect between the oxygen vacancies and the heterojunction, which accelerated the fast separation of charge carriers. The cycle experiments of RhB on Vo-BiOBr/CdS were performed to examine the stability of the photocatalyst (Figure 8c) and the degradation efficiency slightly decreased, indicating that Vo-BiOBr/CdS had good stability. The primary active species in the degradation of RhB on Vo-BiOBr/CdS was discussed by radical trapping experiments. 1,4-benzoquinone (BQ), isopropanol (IPA), and potassium Iodide (KI) were used as the quenchers of superoxide radicals (·O2 − ), hydroxyl radicals (·OH), and holes (h + ), respectively [54]. The dosage of each quencher was maintained equally. As depicted in Figure 8d, the removal rate for RhB significantly decreased in the presence of BQ, which indicated that ·O2 − radicals were the main active species [53].
In conclusion, the ·O2 − was the decisive active species. The primary active species in the degradation of RhB on Vo-BiOBr/CdS was discussed by radical trapping experiments. 1,4-benzoquinone (BQ), isopropanol (IPA), and potassium Iodide (KI) were used as the quenchers of superoxide radicals (·O 2 − ), hydroxyl radicals (·OH), and holes (h + ), respectively [54]. The dosage of each quencher was maintained equally. As depicted in Figure 8d, the removal rate for RhB significantly decreased in the presence of BQ, which indicated that ·O 2 − radicals were the main active species [53]. In conclusion, the ·O 2 − was the decisive active species.

Band Structure and Density of States by DFT Calculation
The band structure of BiOBr, Vo-BiOBr, and CdS were simulated through DFT calculation ( Figure 9). The CBM and the VBM were not at the same point, suggesting that BiOBr had an indirect bandgap (Figure 9a). The CdS was a direct band gap semiconductor (Figure 9c). The band gap values of BiOBr and CdS were identified to be 2.24 eV and 1.13 eV, respectively, which was narrower than the experimental values of 2.78 eV and 2.4 eV (may be due to the shortcoming of the DFT calculation). To study the impact of oxygen vacancies, we conducted DFT theoretical calculations to obtain the calculated band structure of Vo-BiOBr (Figure 9e). It is worth noting that a new defect level was created by oxygen vacancies compared with BiOBr. The defect levels facilitate the excitation of electrons and transfer to the CB, thus reducing the band gap of BiOBr.

Band Structure and Density of States by DFT Calculation
The band structure of BiOBr, Vo-BiOBr, and CdS were simulated through DFT calculation ( Figure 9). The CBM and the VBM were not at the same point, suggesting that BiOBr had an indirect bandgap (Figure 9a). The CdS was a direct band gap semiconductor (Figure 9c). The band gap values of BiOBr and CdS were identified to be 2.24 eV and 1.13 eV, respectively, which was narrower than the experimental values of 2.78 eV and 2.4 eV (may be due to the shortcoming of the DFT calculation). To study the impact of oxygen vacancies, we conducted DFT theoretical calculations to obtain the calculated band structure of Vo-BiOBr (Figure 9e). It is worth noting that a new defect level was created by oxygen vacancies compared with BiOBr. The defect levels facilitate the excitation of electrons and transfer to the CB, thus reducing the band gap of BiOBr.  Moreover, we calculated the DOS of BiOBr, Vo-BiOBr, and CdS. Figure 9b shows that the CBM was mainly dominated by Bi 6p orbitals, while the VBM was occupied by O 2p and Br 4p orbitals of BiOBr. The CBM and VBM were both dominated by S 3p orbitals of CdS (Figure 9d). The DOS for Vo-BiOBr showed the presence of defect levels (Figure 9f), in agreement with the results of the band structure.

Photocatalytic Mechanism
As illustrated in Figure 10a,b, the Mott-Schottky (M-S) spectra of both CdS and Vo-BiOBr were positive, meaning that they are both n-type semiconductors [55]. The flat band (E fb ) potential was measured to be −0.18 and −1.11 eV (vs NHE) respectively. The conduction band (CB) potential (E CB ) of Vo-BiOBr and CdS were identified to be −0.18 and −1.11 eV, respectively, because the E CB of the n-type semiconductor was nearly equal to E fb [56]. Furthermore, the corresponding valance band potential (E VB ) can be calculated by the formula E VB = E CB + E g [57]. Here, E g is band gap energy, identified as 2.14 and 2.4 eV for Vo-BiOBr and CdS by UV-vis DRS. Therefore, the calculated E VB for Vo-BiOBr and CdS were 1.96 and 1.29 eV, respectively. The energy gap between the Fermi level (E f ) and the VB could be identified by the valance band-XPS spectra of Vo-BiOBr and CdS (Figure 10c,d) [58], suggesting that the E f of Vo-BiOBr and CdS were 0.32 and −0.16 eV, respectively.
Based on the previous analyses, the photocatalytic mechanism of Vo-BiOBr/CdS was studied. According to the results of M-S spectra, UV-vis DRS, and XPS valance band, the band structure, as illustrated in Figure 10e, shows that if the composite was a type-II heterojunction, the photoexcited electrons migrated from the CB of CdS to that of Vo-BiOBr and accumulated in the CB of Vo-BiOBr, and the photoexcited holes could migrate from the VB of Vo-BiOBr to CdS and accumulate in the VB of CdS. However, the E CB of Vo-BiOBr was lower than O 2 /O 2 − (−0.33 eV). As a result, the electrons in the CB of Vo-BiOBr could not reduce dissolved oxygen to ·O 2 − , which contradicted the hypothesis that ·O 2 − was the main active species. According to the above analysis, the photocatalysts could not obey the type-II heterojunction, and the more suitable photocatalytic mechanism was the S-scheme heterojunction. Superoxide radicals had strong oxidative activity, and they could oxidize organic matter to carbon dioxide and water. The main reactions of photocatalytic degradation were as follows [59]: Vo-BiOBr/CdS + hv → Vo-BiOBr (e − + h + )/CdS (e − + h + ) Vo-BiOBr (e − ) + CdS (h + ) →Vo-BiOBr/CdS (recombination) To further identify the charge migration pathway of Vo-BiOBr/CdS heterojunction, the work functions of Vo-BiOBr and CdS were calculated. In Figure 11a,b, the electrostatic potentials of Vo-BiOBr (001) and CdS (001) planes were identified to be 3.38 eV and 3.01 eV respectively. The work function of Vo-BiOBr was larger than that of CdS, indicating that electrons could transfer from CdS to Vo-BiOBr spontaneously.
In the S-scheme heterojunction, the Vo-BiOBr presented a bigger Φ and lower E f than CdS. Therefore, the S-scheme heterojunction mechanism was proposed ( Figure 11c). The E f of Vo-BiOBr was lower than that of CdS, so when the two semiconductors contacted each other, electrons would migrate from CdS to Vo-BiOBr until their E f reached equilibrium [60]. This process resulted in the construction of an internal electric field (IEF) directing from CdS to Vo-BiOBr. Under the LED irradiation, electrons were exited from VB to CB in Vo-BiOBr and CdS, respectively. IEF promoted the recombination of photoexcited electrons in the CB of Vo-BiOBr and holes in the VB of CdS [61]. Moreover, the rest of the electrons and holes possessing strong redox potential would be retained. Therefore, the ·O 2 − formation process mainly depended on electrons in the CdS on account of the S-scheme mechanism, but the ·OH could not be produced since the E VB of Vo-BiOBr was higher than H 2 O/·OH (2.39 eV). In conclusion, this S-scheme heterojunction promoted charge separation and retained a strong redox capability. Based on the previous analyses, the photocatalytic mechanism of Vo-BiOBr/CdS was studied. According to the results of M-S spectra, UV-vis DRS, and XPS valance band, the band structure, as illustrated in Figure 10e, shows that if the composite was a type-II heterojunction, the photoexcited electrons migrated from the CB of CdS to that of Vo-BiOBr and accumulated in the CB of Vo-BiOBr, and the photoexcited holes could migrate from the VB of Vo-BiOBr to CdS and accumulate in the VB of CdS. However, the ECB of Vo- To further identify the charge migration pathway of Vo-BiOBr/CdS heterojunction, the work functions of Vo-BiOBr and CdS were calculated. In Figure 11a,b, the electrostatic potentials of Vo-BiOBr (001) and CdS (001) planes were identified to be 3.38 eV and 3.01 eV respectively. The work function of Vo-BiOBr was larger than that of CdS, indicating that electrons could transfer from CdS to Vo-BiOBr spontaneously. In the S-scheme heterojunction, the Vo-BiOBr presented a bigger Φ and lower Ef than CdS. Therefore, the S-scheme heterojunction mechanism was proposed ( Figure 11c). The Ef of Vo-BiOBr was lower than that of CdS, so when the two semiconductors contacted each other, electrons would migrate from CdS to Vo-BiOBr until their Ef reached equilibrium [60]. This process resulted in the construction of an internal electric field (IEF) directing from CdS to Vo-BiOBr. Under the LED irradiation, electrons were exited from VB to

Conclusions
In conclusion, the S-scheme Vo-BiOBr/CdS heterojunction was synthesized by the solvothermal approach. The obtained S-scheme Vo-BiOBr/CdS heterojunction displayed RhB and MB removal efficiencies of 97 % and 94%, respectively. Moreover, the composite showed a good degradation and reduction efficiency for RhB-Cr(VI)-MB mixed solutions. A significant enhancement of the photocatalytic activity can be ascribed to the synergistic effect of the heterojunction and Vo. The results of the radical trapping experiment, valence band-XPS, and work function proved the formation of the S-scheme heterojunction. Finally, the four cycles of RhB degradation on Vo-BiOBr/CdS proved its good stability. This study provides a novel insight into designing highly efficient photocatalysts for environmental remediation.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nano13050830/s1, Figure S1 presents the simulated structures of BiOBr, CdS, and Vo-BiOBr. Table S1 exhibits the surface fraction of O element by XPS spectra. Figure S2 shows high-resolution XPS spectra of Bi 4f (a), Br 3d (b), O 1s (c) of BiOBr, and Vo-BiOBr. Figure S3 displays the removal efficiency of RhB over Vo-BiOBr/CdS with a different mass ratio. Table S2 compares the photocatalytic activity of Vo-BiOBr/CdS with other related catalysts reported in the literature. Figure S4 presents the removal efficiency of TC and reduction efficiencies of Cr 6+