Enhanced Sulfamerazine Removal via Adsorption–Photocatalysis Using Bi2O3–TiO2/PAC Ternary Nanoparticles

The presence of sulfonamides (SAs) in water has received increasing attention due to the risk to ecosystems. The adsorption and photocatalysis performance for sulfamerazine (SMZ) of Bi2O3–TiO2 supported on powdered activated carbon (Bi2O3–TiO2/PAC) nanoparticles was evaluated. The amount of doped Bi2O3 not only influenced the photocatalytic performance but also impacted the adsorption capacity. The adsorption mass transfer mechanism of Bi2O3–TiO2/PAC was elucidated and is further discussed in combination with the photocatalytic mechanism. It was indicated that Bi2O3–TiO2/PAC(10%–700 ◦C) performed best, and the SMZ removal by the adsorption–photocatalysis of Bi2O3–TiO2/PAC(10%–700 ◦C) reached 95.5%. Adsorption onto active sites was a major adsorption step, and external diffusion was assisted. Superoxide radical (•O2−) and hole (h+) were identified as the major reactive oxygen species (ROS) for SMZ removal. Benzene ring fracture, SO2 extrusion and nitrogenated SMZ were proposed as the main pathways for photocatalysis. Meanwhile, alkaline conditions enhanced photocatalytic performance, while contrary effects were observed for adsorption. The adsorption–photocatalysis removal performance for SMZ in lake water was better than that for river water. It can be generalized for the potential application of photocatalysis coupling with adsorption to remove refractory antibiotics in water.


Introduction
Sulfamerazine (SMZ) is one of the most commonly used sulfonamides (SAs) for human and animal infections. The highest levels of SMZ residues have been reported in animal-sourced foods such as eggs, meat and milk [1,2], causing potential adverse effects on food products and ecosystems. SAs and their metabolites cannot degrade completely in traditional wastewater treatment plants, and antibiotic residues also appear in natural waters [3,4]. Various approaches [5] have been developed to remove SAs such as adsorption [6,7], membrane filtration [8,9], electrochemical oxidation [10], biodegradation [11], photocatalysis [12,13], etc. However, adsorption and membrane filtration carry out the transformation of contaminants but cannot achieve the real removal of pollutants. Excessive electrode and energy consumption should be considered for electrochemical oxidation, and the bacteriostatic property of SAs may lead to apparent negative removal during biological wastewater treatment [14]. Among them, solar-driven photocatalysis is widely utilized due to its characteristics of being environmentally friendly, energy-saving and easy to controlling.

Photocatalytic Experiments
Two sets of photocatalytic experiments were conducted. One was performed from the beginning of solar light irradiation, and the other was carried out after adsorption equilibrium in the dark and then irradiation under solar light. A 300 W xenon lamp (CEL-HXF300E7, Ceaulight, Beijing, China) was used to generate simulated solar light irradiation, the light intensity of which was 443.5 mω 2 /cm 2 , measured with an optical power density meter (NP2000, Ceaulight, Beijing, China). A concentration of 20 mg/L of SMZ was continuously stirred in a 0.25 L batch reactor surrounded by a water-cooling device. The distance between the surface of the reaction liquid and light source was approximately 15 cm. Composites at 1.0 g/L were added into the solution for the 120 min photocatalytic experiments. The aliquots were sampled at intervals of 20, 40, 60, 80, 100 and 120 min. The photocatalytic kinetics of the Bi 2 O 3 -TiO 2 /PAC composites followed a pseudo-first-order kinetic model [38], as shown in Equation (1).
where C 0 (mg/L) and C t represent the concentration of SMZ at adsorption equilibrium and at time t, respectively; t (min) means the reaction time, and k (min −1 ) means the kinetic rate constant. The experiments of trapping and studying the influences of water quality parameters on photocatalysis were conducted at adsorption equilibrium. Hydroxyl ( • OH), • O 2 − and h + were captured with IPA, BQ and EDTA-2Na, respectively. Scavenging agents at 2.0 and 5.0 mM were added into the solution for the trapping experiment. Electron paramagnetic resonance (EPR) spectrum measurements for • OH and • O 2 − were investigated using DMPO and methanol, respectively, with sampling through a 100 µL capillary. The initial pH of the solution was adjusted with diluted 0.1 M HCl and NaOH. Three concentrations of anions including HCO 3 − , Cl − and SO 4 2− were set at 0.5, 2.0, and 5.0 mM. The humic acid (HA) concentrations were 0.5, 2.0 and 5.0 mg/L. Surface water was collected from Tonghui River (39 •  The main water quality of the surface water is shown in Table S3. All aliquots were filtered with 0.22 µm syringe filters. All trials were conducted three times, and average values are reported.

Analytic Methods
The concentration of SMZ was detected with a high-performance liquid chromatograph (Agilent 1260 LC) equipped with a C18 column (Waters XBridge, 4.6 mm × 250 mm, 5 µm). The detection wavelength was 270 nm. The mobile phase was methanol and ultrapure water with a 50:50 ratio, and the flow rate was 1 mL/min. The EPR spectra were collected on an electron spin resonance spectrometer (JEOL JES-FA200). The intermediates were identified by UPLC/MS (Agilent 6460 Triple Quad LC/MS) with a C18 column (Agilent Zorbax Eclipse Plus, 2.1 mm × 50 mm, 1.8 µm). The mobile phase of the UPLC/MS was a 70:30 mixture of methanol and ultrapure water (0.1% formic acid) at a flow rate of 0.2 mL/min. Positive electrospray ionization (ESI + ) and negative electrospray ionization (ESI − ) modes were both used for analysis.

Adsorption Kinetics
The adsorption capacity of the Bi 2 O 3 -TiO 2 /PAC composites increased sharply in the first 10 min, and all reached adsorption equilibrium after 30 min ( Figure S1). In the three models, the correlation coefficients (R 2 ) of the pseudo-second-order kinetics were above 0.99 (Table 1), suggesting that the adsorption rate was proportional to the square of the SMZ concentration and the chemical reaction was the vital factor for SMZ adsorption. For the different Bi/Ti molar ratios of the composites, the adsorption rate gradually decreased with an increase in the Bi/Ti molar ratio. Although doping Bi improved the utilization of visible light to enhance the photocatalytic performance, excessive Bi 2 O 3 reduced the adsorption efficiency of the SMZ, which was detrimental to the enrichment of the SMZ at low antibiotic concentrations in the water environment [39]. Comparing with the adsorption kinetics of Bi 2 O 3 -TiO 2 (10%-700 • C) and TiO 2 (700 • C), likewise, the recombination of Bi 2 O 3 could reduce the adsorption capacity of TiO 2 . Therefore, the addition of PAC effectively improved the adsorption performance of Bi 2 O 3 -TiO 2 , which further promoted the photocatalytic efficiency. For Bi 2 O 3 -TiO 2 /PAC composites with different second stage calcination temperatures, as shown in Table S4, the adsorption performance of the second stage calcination at 500 • C was the best. It is worth noting that the difference between the adsorption properties of Bi 2 O 3 -TiO 2 /PAC(10%-600 • C) and Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) was negligible, indicating the effectiveness of N 2 atmosphere for preventing the loss of PAC in the process of high-temperature calcination [40]. The Langmuir and Freundlich isotherm models were used to describe the adsorption process, the R 2 of which, both above 0.95, well described the monolayer adsorption and multilayer adsorption simultaneously existing, as shown in Figure 1. To be specific, the R 2 of the Langmuir model was larger, indicating that monolayer adsorption and surface adsorption predominated. Consistent with the adsorption kinetics, the larger the molar ratio of Bi/Ti, the lower the adsorption capacity of the Bi 2 O 3 -TiO 2 /PAC composites on SMZ. The adsorption performance of Bi 2 O 3 -TiO 2 was less than that of TiO 2 , indicating Bi 2 O 3 occupied part of the lattice gap of TiO 2 . Compared with the adsorption isotherm models of PAC, Bi 2 O 3 -TiO 2 (10%-700 • C) and TiO 2 (700 • C), the adsorption capacity of the Bi 2 O 3 -TiO 2 /PAC composites was lower than that of PAC but higher than that of Bi 2 O 3 -TiO 2 (10%-700 • C) and TiO 2 (700 • C). This revealed that the adsorption capacity of PAC was manifested in the Bi 2 O 3 -TiO 2 /PAC composites. Additionally, the higher the second stage calcination temperature, the lower the adsorption capacity ( Figure S2). A high calcination temperature can improve the adsorption capacity of TiO 2 because of the higher crystallinity [41]; however, the adsorption performance of Bi 2 O 3 -TiO 2 /PAC was contrary. This is another piece of evidence for the adsorption dominance of PAC, since a high temperature would lead to PAC loss.  The Langmuir and Freundlich isotherm models were used to describe the adsorption process, the R 2 of which, both above 0.95, well described the monolayer adsorption and multilayer adsorption simultaneously existing, as shown in Figure 1. To be specific, the R 2 of the Langmuir model was larger, indicating that monolayer adsorption and surface adsorption predominated. Consistent with the adsorption kinetics, the larger the molar ratio of Bi/Ti, the lower the adsorption capacity of the Bi2O3-TiO2/PAC composites on SMZ. The adsorption performance of Bi2O3-TiO2 was less than that of TiO2, indicating Bi2O3 occupied part of the lattice gap of TiO2. Compared with the adsorption isotherm models of PAC, Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C), the adsorption capacity of the Bi2O3-TiO2/PAC composites was lower than that of PAC but higher than that of Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C). This revealed that the adsorption capacity of PAC was manifested in the Bi2O3-TiO2/PAC composites. Additionally, the higher the second stage calcination temperature, the lower the adsorption capacity ( Figure S2). A high calcination temperature can improve the adsorption capacity of TiO2 because of the higher crystallinity [41]; however, the adsorption performance of Bi2O3-TiO2/PAC was contrary. This is another piece of evidence for the adsorption dominance of PAC, since a high temperature would lead to PAC loss.
Water 2020, 12, x FOR PEER REVIEW 6 of 18

Overall Adsorption-Photocatalysis Performance
SMZ removal was enhanced initially by increasing the Bi/Ti molar ratio and decreased thereafter, while the Bi/Ti proportion reached a high level. Bi2O3-TiO2/PAC(10%-700 °C) exhibited the best removal efficiency from the beginning of solar light irradiation within 120 min, as shown in Figure 4a. When the Bi/Ti molar ratio was lower than the optimal amount, the Bi-O polyhedral was increased with an increase in the Bi/Ti molar ratio. It can prolong the life of carriers and increase the numbers of active sites and substrates adsorbed due to the smaller crystal size. However, when the Bi/Ti molar ratio was higher than its optimum amount, higher concentration of Bi increased the particle agglomeration of Bi2O3, which caused severe nanocrystal heterojunction decline and then affected the adsorption-photocatalysis performance [42].
The coupling removal efficiency for SMZ was affected by co-existing organic matter. The same concentration of SMZ in two bodies of water as in ultrapure water was removed by Bi2O3-TiO2/PAC(10%-700 °C) in 120 min of solar light irradiation without pre-adsorption and solution pH adjustment. It can be observed in Figure 4b that the coupling adsorption-photocatalysis efficiencies for SMZ in the river water and lake water were lower than those of ultrapure water. The removal performance for SMZ in the lake water was better than that in the river water, probably because the organic matter of the river water was more abundant than that of the lake water. The inhibition was mainly caused by HA, fulvic acid (FA), high-molecular-weight polysaccharides and proteins [43]. It has been reported that the inhibition effect was higher for FA than that for HA [44]. They not only competed with SMZ for the reactive oxygen species (ROS) in photocatalysis but also for the adsorption-active sites of Bi2O3-TiO2/PAC [45].

Overall Adsorption-Photocatalysis Performance
SMZ removal was enhanced initially by increasing the Bi/Ti molar ratio and decreased thereafter, while the Bi/Ti proportion reached a high level. Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) exhibited the best removal efficiency from the beginning of solar light irradiation within 120 min, as shown in Figure 4a. When the Bi/Ti molar ratio was lower than the optimal amount, the Bi-O polyhedral was increased with an increase in the Bi/Ti molar ratio. It can prolong the life of carriers and increase the numbers of active sites and substrates adsorbed due to the smaller crystal size. However, when the Bi/Ti molar ratio was higher than its optimum amount, higher concentration of Bi increased the particle agglomeration of Bi 2 O 3 , which caused severe nanocrystal heterojunction decline and then affected the adsorption-photocatalysis performance [42].

Overall Adsorption-Photocatalysis Performance
SMZ removal was enhanced initially by increasing the Bi/Ti molar ratio and decreased thereafter, while the Bi/Ti proportion reached a high level. Bi2O3-TiO2/PAC(10%-700 °C) exhibited the best removal efficiency from the beginning of solar light irradiation within 120 min, as shown in Figure 4a. When the Bi/Ti molar ratio was lower than the optimal amount, the Bi-O polyhedral was increased with an increase in the Bi/Ti molar ratio. It can prolong the life of carriers and increase the numbers of active sites and substrates adsorbed due to the smaller crystal size. However, when the Bi/Ti molar ratio was higher than its optimum amount, higher concentration of Bi increased the particle agglomeration of Bi2O3, which caused severe nanocrystal heterojunction decline and then affected the adsorption-photocatalysis performance [42].
The coupling removal efficiency for SMZ was affected by co-existing organic matter. The same concentration of SMZ in two bodies of water as in ultrapure water was removed by Bi2O3-TiO2/PAC(10%-700 °C) in 120 min of solar light irradiation without pre-adsorption and solution pH adjustment. It can be observed in Figure 4b that the coupling adsorption-photocatalysis efficiencies for SMZ in the river water and lake water were lower than those of ultrapure water. The removal performance for SMZ in the lake water was better than that in the river water, probably because the organic matter of the river water was more abundant than that of the lake water. The inhibition was mainly caused by HA, fulvic acid (FA), high-molecular-weight polysaccharides and proteins [43]. It has been reported that the inhibition effect was higher for FA than that for HA [44]. They not only competed with SMZ for the reactive oxygen species (ROS) in photocatalysis but also for the adsorption-active sites of Bi2O3-TiO2/PAC [45].   The coupling removal efficiency for SMZ was affected by co-existing organic matter. The same concentration of SMZ in two bodies of water as in ultrapure water was removed by Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) in 120 min of solar light irradiation without pre-adsorption and solution pH adjustment. It can be observed in Figure 4b that the coupling adsorption-photocatalysis efficiencies for SMZ in the river water and lake water were lower than those of ultrapure water. The removal performance for SMZ in the lake water was better than that in the river water, probably because the organic matter of the river water was more abundant than that of the lake water. The inhibition was mainly caused by HA, fulvic acid (FA), high-molecular-weight polysaccharides and proteins [43]. It has been reported that the inhibition effect was higher for FA than that for HA [44]. They not only competed with SMZ for the reactive oxygen species (ROS) in photocatalysis but also for the adsorption-active sites of Bi 2 O 3 -TiO 2 /PAC [45].

Effect of Water Quality Parameters
3.4.1. Effect of Initial pH SMZ adsorption on Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) was significantly affected by the pH of the solution, which governed the dissociation degree of the functional groups on the composite and changed the surface potential of the material. It was shown that at pH 7.0 of the solution, SMZ was more easily adsorbed on Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) (Figure 5a). Because of two acid-base dissociation points of SMZ, there are three forms of SMZ at different pH values [46]. The neutral molecular form is more readily adsorbed than the other forms [47]. When the pH of the solution was 3.0-7.0, the adsorption removal of SMZ gradually increased, which was due to H + and SMZ competitive adsorption. With the continuous augmentation of pH, the adsorption removal of SMZ declined. The electrostatic repulsion between the anionic forms of SMZ and the Bi 2 O 3 -TiO 2 /PAC(10%-700 • C), as well as the hydrophobic effect on the surface groups of the composite, affected the adsorption performance of the SMZ.
The effect of pH on the photocatalysis of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) was distinguished from that on adsorption. The SMZ photocatalytic efficiency increased from 61.3% to 67.9% as the initial pH increased from 3.0 to 7.0 ( Figure 5b). When the initial pH further increased to 9.0, the removal efficiency continued rising but then dropped at pH 11. H + can consume • O 2 − , which was the main reaction species for the Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) photocatalysis of SMZ; therefore, a high concentration of H + resulted in a lower • O 2 − yield (Equations (2) and (3)). The dissociation constant (pKa) of SMZ is pK a,1 = 1.75 and pK a,2 = 7.35 [48]. When the initial pH of the SMZ solution was higher than 7.35, the ability of sulfonyl-N to donate electrons to SMZ was enhanced due to deprotonation, which was conducive to the degradation of SMZ. However, when the concentration of OH − was higher, the photo-induced carriers h + reacted with OH − and created • OH (Equation (4)). The generated • OH contributed less than h + to the SMZ photocatalytic degradation by Bi 2 O 3 -TiO 2 /PAC(10%-700 • C), reducing the removal efficiency for SMZ.
Water 2020, 12, x FOR PEER REVIEW 8 of 18

Effect of Water Quality Parameters
3.4.1. Effect of Initial pH SMZ adsorption on Bi2O3-TiO2/PAC(10%-700 °C) was significantly affected by the pH of the solution, which governed the dissociation degree of the functional groups on the composite and changed the surface potential of the material. It was shown that at pH 7.0 of the solution, SMZ was more easily adsorbed on Bi2O3-TiO2/PAC(10%-700 °C) (Figure 5a). Because of two acid-base dissociation points of SMZ, there are three forms of SMZ at different pH values [46]. The neutral molecular form is more readily adsorbed than the other forms [47]. When the pH of the solution was 3.0-7.0, the adsorption removal of SMZ gradually increased, which was due to H + and SMZ competitive adsorption. With the continuous augmentation of pH, the adsorption removal of SMZ declined. The electrostatic repulsion between the anionic forms of SMZ and the Bi2O3-TiO2/PAC(10%-700 °C), as well as the hydrophobic effect on the surface groups of the composite, affected the adsorption performance of the SMZ.
The effect of pH on the photocatalysis of Bi2O3-TiO2/PAC(10%-700 °C) was distinguished from that on adsorption. The SMZ photocatalytic efficiency increased from 61.3% to 67.9% as the initial pH increased from 3.0 to 7.0 (Figure 5b). When the initial pH further increased to 9.0, the removal efficiency continued rising but then dropped at pH 11. H + can consume • O2 − , which was the main reaction species for the Bi2O3-TiO2/PAC(10%-700 °C) photocatalysis of SMZ; therefore, a high concentration of H + resulted in a lower • O2 − yield (Equations (2) and (3)). The dissociation constant (pKa) of SMZ is pKa,1 = 1.75 and pKa,2 = 7.35 [48]. When the initial pH of the SMZ solution was higher than 7.35, the ability of sulfonyl-N to donate electrons to SMZ was enhanced due to deprotonation, which was conducive to the degradation of SMZ. However, when the concentration of OH − was higher, the photo-induced carriers h + reacted with OH − and created • OH (Equation (4)). The generated • OH contributed less than h + to the SMZ photocatalytic degradation by Bi2O3-TiO2/PAC(10%-700 °C), reducing the removal efficiency for SMZ.

Effect of Inorganic Anions
The addition of 0.5 mM HCO3 − improved the photocatalytic efficiency for SMZ. When the concentration of HCO3 − increased above 0.5 mM, the photocatalytic efficiency decreased but remained higher than that without HCO3 − (Figure 6a). HCO3 − is an important scavenger for • OH

Effect of Inorganic Anions
The addition of 0.5 mM HCO 3 − improved the photocatalytic efficiency for SMZ. When the concentration of HCO 3 − increased above 0.5 mM, the photocatalytic efficiency decreased but remained higher than that without HCO 3 − (Figure 6a). HCO 3 − is an important scavenger for • OH (Equations (5) and (6)) [49]. Since • OH was not the dominant factor for photocatalysis and the resulting • CO 3 − may be involved in SMZ degradation, the photocatalytic performance for SMZ was enhanced by adding a small amount of HCO 3 − . Meanwhile, the addition of HCO 3 − increased the pH of the solution from 7 to 8.5, which also promoted the photocatalytic degradation of SMZ by Bi 2 O 3 -TiO 2 /PAC(10%-700 • C).
In the presence of Cl − and SO 4 2− , as shown in Figure 6b,c, the removal efficiency for SMZ of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) slightly decreased with the different concentrations of Cl − and SO 4 2− . Both these anions can react with h + and • OH (Equations (7) and (8) could be attached to the surface of TiO 2 via van der Waals forces and hydrogen bonds, and replace some of the • OH of TiO 2 by a ligand exchange mechanism [50].
Water 2020, 12, x FOR PEER REVIEW 9 of 18 (Equations (5) and (6)) [49]. Since • OH was not the dominant factor for photocatalysis and the resulting • CO3 − may be involved in SMZ degradation, the photocatalytic performance for SMZ was enhanced by adding a small amount of HCO3 − . Meanwhile, the addition of HCO3 − increased the pH of the solution from 7 to 8.5, which also promoted the photocatalytic degradation of SMZ by Bi2O3-TiO2/PAC(10%-700 °C).
In the presence of Cl − and SO4 2− , as shown in Figure 6b,c, the removal efficiency for SMZ of Bi2O3-TiO2/PAC(10%-700 °C) slightly decreased with the different concentrations of Cl − and SO4 2− . Both these anions can react with h + and • OH (Equations (7) and (8)), resulting in a decrease in photocatalytic performance. It should be stated that Cl − was less influential than SO4 2− for photocatalysis. The 5.0 mM addition of SO4 2− led to a 17.8% decrease in photocatalytic efficiency compared to 0 mM SO4 2− in the solution at the initial 20 min. A previous study has reported that SO4 2− could be attached to the surface of TiO2 via van der Waals forces and hydrogen bonds, and replace some of the • OH of TiO2 by a ligand exchange mechanism [50].

Effect of HA
The presence of HA inhibited the photocatalytic degradation of SMZ; with an increase in the HA concentration from 0 to 5.0 mg/L, the photocatalytic efficiency for SMZ decreased by 9.1% (Figure 6d). Firstly, HA competed with SMZ for the ROS in the photocatalysis. Furthermore, the photons were contended by HA and Bi 2 O 3 -TiO 2 /PAC(10%-700 • C), resulting in the production of fewer active oxygen species. However, the SMZ photocatalytic efficiency was up to 58.8%, even at a 5.0 mg/L HA concentration, indicating the photocatalytic stability and availability of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C).
Water 2020, 12, x FOR PEER REVIEW 10 of 18 6d). Firstly, HA competed with SMZ for the ROS in the photocatalysis. Furthermore, the photons were contended by HA and Bi2O3-TiO2/PAC(10%-700 °C), resulting in the production of fewer active oxygen species. However, the SMZ photocatalytic efficiency was up to 58.8%, even at a 5.0 mg/L HA concentration, indicating the photocatalytic stability and availability of Bi2O3-TiO2/PAC(10%-700 °C).

BET Surface Area and Pore Size Distribution Analysis
According to the IUPAC classification, the BET isotherm of the Bi 2 O 3 -TiO 2 /PAC matches the type IV isotherm (Figure 8). The majority pore size of the composite was 2-50 nm, which indicated that substantial mesoporous structures exist. The pore size distribution range of Bi 2 O 3 -TiO 2 /PAC(12%-700 • C) was significantly larger than that of Bi 2 O 3 -TiO 2 /PAC(8%-700 • C) and Bi 2 O 3 -TiO 2 /PAC(10%-700 • C). As the Bi/Ti molar ratio increased, the pore volume decreased (Table 2), which further proved that Bi 2 O 3 influenced the formation of pores. Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) exhibited a larger BET surface area than Bi 2 O 3 -TiO 2 (10%-700 • C) and TiO 2 (700 • C), which was due to the addition of PAC. The pore volumes of PAC and Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) were 0.654 and 0.277 cm 3 /g, respectively, demonstrating that some of the Bi 2 O 3 -TiO 2 nanoparticles were deposited into the PAC pores. From the N 2 absorption-desorption isotherm analysis and pore size distribution in Figure 9, it can be inferred that parts of the pores in PAC were occupied by Bi 2 O 3 -TiO 2 , causing the aperture range to move to the right.
Water 2020, 12, x FOR PEER REVIEW 11 of 18

BET Surface Area and Pore Size Distribution Analysis
According to the IUPAC classification, the BET isotherm of the Bi2O3-TiO2/PAC matches the type IV isotherm (Figure 8). The majority pore size of the composite was 2-50 nm, which indicated that substantial mesoporous structures exist. The pore size distribution range of Bi2O3-TiO2/PAC(12%-700 °C) was significantly larger than that of Bi2O3-TiO2/PAC(8%-700 °C) and Bi2O3-TiO2/PAC(10%-700 °C). As the Bi/Ti molar ratio increased, the pore volume decreased (Table 2), which further proved that Bi2O3 influenced the formation of pores. Bi2O3-TiO2/PAC(10%-700 °C) exhibited a larger BET surface area than Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C), which was due to the addition of PAC. The pore volumes of PAC and Bi2O3-TiO2/PAC(10%-700 °C) were 0.654 and 0.277 cm 3 /g, respectively, demonstrating that some of the Bi2O3-TiO2 nanoparticles were deposited into the PAC pores. From the N2 absorption-desorption isotherm analysis and pore size distribution in Figure 9, it can be inferred that parts of the pores in PAC were occupied by Bi2O3-TiO2, causing the aperture range to move to the right.

UV-Vis Diffuse Reflectance Spectrum Analysis
The absorption spectrum of Bi2O3-TiO2/PAC exhibited significant red-shifts to the visible − light region compared with Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C), indicating the combination of Bi2O3-TiO2/PAC extended light absorption to visible light region and improved the light harvesting. The photophysical properties of Bi2O3-TiO2/PAC(10%-700 °C) revealed a stronger photo-absorption band in both ultraviolet and visible light than the others (Figure 10). The band gap of Bi2O3-TiO2/PAC(10%-700 °C) was 2.58 eV as calculated by the Tauc-plot method [57], which was the minimum among all.

BET Surface Area and Pore Size Distribution Analysis
According to the IUPAC classification, the BET isotherm of the Bi2O3-TiO2/PAC matches the type IV isotherm (Figure 8). The majority pore size of the composite was 2-50 nm, which indicated that substantial mesoporous structures exist. The pore size distribution range of Bi2O3-TiO2/PAC(12%-700 °C) was significantly larger than that of Bi2O3-TiO2/PAC(8%-700 °C) and Bi2O3-TiO2/PAC(10%-700 °C). As the Bi/Ti molar ratio increased, the pore volume decreased (Table 2), which further proved that Bi2O3 influenced the formation of pores. Bi2O3-TiO2/PAC(10%-700 °C) exhibited a larger BET surface area than Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C), which was due to the addition of PAC. The pore volumes of PAC and Bi2O3-TiO2/PAC(10%-700 °C) were 0.654 and 0.277 cm 3 /g, respectively, demonstrating that some of the Bi2O3-TiO2 nanoparticles were deposited into the PAC pores. From the N2 absorption-desorption isotherm analysis and pore size distribution in Figure 9, it can be inferred that parts of the pores in PAC were occupied by Bi2O3-TiO2, causing the aperture range to move to the right.

UV-Vis Diffuse Reflectance Spectrum Analysis
The absorption spectrum of Bi2O3-TiO2/PAC exhibited significant red-shifts to the visible − light region compared with Bi2O3-TiO2(10%-700 °C) and TiO2(700 °C), indicating the combination of Bi2O3-TiO2/PAC extended light absorption to visible light region and improved the light harvesting. The photophysical properties of Bi2O3-TiO2/PAC(10%-700 °C) revealed a stronger photo-absorption band in both ultraviolet and visible light than the others (Figure 10). The band gap of Bi2O3-TiO2/PAC(10%-700 °C) was 2.58 eV as calculated by the Tauc-plot method [57], which was the minimum among all.  (Figure 10). The band gap of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) was 2.58 eV as calculated by the Tauc-plot method [57], which was the minimum among all. The formation of the heterostructure caused by the Bi doping and PAC loading reduced the band gap energy of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) [58]. It has been proposed that the energy band lies between the Ti 4+ 3d band and Bi 3+ 6s band [59].
Water 2020, 12, x FOR PEER REVIEW 12 of 18 The formation of the heterostructure caused by the Bi doping and PAC loading reduced the band gap energy of Bi2O3-TiO2/PAC(10%-700 °C) [58]. It has been proposed that the energy band lies between the Ti 4+ 3d band and Bi 3+ 6s band [59]. The surface morphology and microstructure of Bi2O3-TiO2/PAC(10%-700 °C) was revealed by SEM-EDS and HRTEM images, shown in Figure S5. Bi2O3 and TiO2 particles were attached to the surface and pores of the PAC. PAC accounted for a larger proportion in the Bi2O3-TiO2/PAC(10%-700 °C). It was obvious that Bi2O3 and TiO2 were reasonably well supported on PAC, and parts of distinct pores manifested the retained adsorption properties of the PAC. The coexistence of Bi2O3, anatase TiO2 and rutile TiO2, by matching to the lattice fringes in the as-synthesized composite, was certified.

Adsorption Mechanism
The adsorption mass transfer mechanism includes external diffusion, internal diffusion and adsorption on active sites. The mass transfer mechanism can be summarized by an adsorption kinetic study [60]. The pseudo-second-order kinetic model was related to the availability of adsorption sites on the surface of the adsorbent [31]. It was found that the mass transfer adsorption of Bi2O3-TiO2/PAC(10%-700 °C) met the Boyd's external diffusion equation and Langmuir kinetic model, as shown in Table 3, while the k value of the Langmuir kinetics model was lower. In the assumptions of these two models, the slowest adsorption step for each is required. Therefore, it is comprehensively judged that the adsorption mechanism of Bi2O3-TiO2/PAC(10%-700 °C) is mainly adsorption onto active sites as well as assistance by external diffusion, and internal diffusion rarely exists. The adsorption process was influenced with the interaction controlled by the exchange and share of e − between Bi2O3-TiO2/PAC composites and SMZ [61,62].  It was obvious that Bi 2 O 3 and TiO 2 were reasonably well supported on PAC, and parts of distinct pores manifested the retained adsorption properties of the PAC. The coexistence of Bi 2 O 3 , anatase TiO 2 and rutile TiO 2 , by matching to the lattice fringes in the as-synthesized composite, was certified.

Adsorption Mechanism
The adsorption mass transfer mechanism includes external diffusion, internal diffusion and adsorption on active sites. The mass transfer mechanism can be summarized by an adsorption kinetic study [60]. The pseudo-second-order kinetic model was related to the availability of adsorption sites on the surface of the adsorbent [31]. It was found that the mass transfer adsorption of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) met the Boyd's external diffusion equation and Langmuir kinetic model, as shown in Table 3, while the k value of the Langmuir kinetics model was lower. In the assumptions of these two models, the slowest adsorption step for each is required. Therefore, it is comprehensively judged that the adsorption mechanism of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) is mainly adsorption onto active sites as well as assistance by external diffusion, and internal diffusion rarely exists. The adsorption process was influenced with the interaction controlled by the exchange and share of e − between Bi 2 O 3 -TiO 2 /PAC composites and SMZ [61,62].

Photocatalytic Mechanism
Trapping experiments show that both • O 2 − and h + played a greater role, while • OH was weaker than these two radicals ( Figure 11). As for SMZ removal, the experiments of trapping similarly identified • O 2 − and h + as the main radicals. Moreover, the SMZ removal under solar light irradiation without composite was only 3.6%; thus, the photolysis by solar light irradiation can be neglected. The mineralization degree indicated by TOC/TOC 0 reached 0.351 in 120 min of solar irradiation at adsorption equilibrium ( Figure S6), confirming that SMZ decomposed into small fragments.
Water 2020, 12, x FOR PEER REVIEW 13 of 18

Photocatalytic Mechanism
Trapping experiments show that both • O2 − and h + played a greater role, while • OH was weaker than these two radicals ( Figure 11). As for SMZ removal, the experiments of trapping similarly identified • O2 − and h + as the main radicals. Moreover, the SMZ removal under solar light irradiation without composite was only 3.6%; thus, the photolysis by solar light irradiation can be neglected. The mineralization degree indicated by TOC/TOC0 reached 0.351 in 120 min of solar irradiation at adsorption equilibrium ( Figure S6), confirming that SMZ decomposed into small fragments.  Possible SMZ photocatalytic pathways for Bi2O3-TiO2/PAC(10%-700 °C) were proposed. The specific chemical formulas and SMZ chromatographic data are shown in Table S5 and Figure S7. A speculative reaction mechanism is shown in Figure 12-that the first path was the structural damage caused in the benzene ring fracture, which agrees well with the findings of others [63,64]. The second path was SO2 extrusion, leading to the appearance of intermediate product B, which is a common phenomenon in the degradation process for SAs [65]. Further attack of the sulfonamide bond of B produced E. The third path was the nitrogenated SMZ being formed by the oxidation of the amino group of the benzene ring. After that, the intermediates were mineralized into fracture rings and simple compounds, such as CO2, H2O and NH4 + [66]. Possible SMZ photocatalytic pathways for Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) were proposed. The specific chemical formulas and SMZ chromatographic data are shown in Table S5 and Figure S7. A speculative reaction mechanism is shown in Figure 12-that the first path was the structural damage caused in the benzene ring fracture, which agrees well with the findings of others [63,64]. The second path was SO 2 extrusion, leading to the appearance of intermediate product B, which is a common phenomenon in the degradation process for SAs [65]. Further attack of the sulfonamide bond of B produced E. The third path was the nitrogenated SMZ being formed by the oxidation of the amino group of the benzene ring. After that, the intermediates were mineralized into fracture rings and simple compounds, such as CO 2  caused in the benzene ring fracture, which agrees well with the findings of others [63,64]. The second path was SO2 extrusion, leading to the appearance of intermediate product B, which is a common phenomenon in the degradation process for SAs [65]. Further attack of the sulfonamide bond of B produced E. The third path was the nitrogenated SMZ being formed by the oxidation of the amino group of the benzene ring. After that, the intermediates were mineralized into fracture rings and simple compounds, such as CO2, H2O and NH4 + [66].

Adsorption-Photocatalysis Mechanism
The coupling adsorption-photocatalysis mechanism of Bi 2 O 3 -TiO 2 /PAC in removing SMZ under solar light irradiation is illustrated in Figure 13. For the Bi 2 O 3 -TiO 2 /PAC system, SMZ was mainly adsorbed on the active sites of the Bi 2 O 3 -TiO 2 /PAC. Meanwhile, the ROS produced by the composite were non-selective and attacked functional groups of SMZ. A heterogeneous junction was formed by doping Bi 2 O 3 with TiO 2 , which greatly reduced the band gap of the photocatalysts and provided good conditions for the separation of photo-generated carriers, greatly improving the photocatalytic performance under visible light. The adsorption capacity of PAC ensured persistent high levels of pollutants for the photocatalytic system [5]. Additionally, the combination of PAC could improve the adsorption performance and recovery rate of photocatalysts, which was of great theoretical and engineering significance for an appropriate adsorption-photocatalysis composite.
Water 2020, 12, x FOR PEER REVIEW 14 of 18

Adsorption-Photocatalysis Mechanism
The coupling adsorption-photocatalysis mechanism of Bi2O3-TiO2/PAC in removing SMZ under solar light irradiation is illustrated in Figure 13. For the Bi2O3-TiO2/PAC system, SMZ was mainly adsorbed on the active sites of the Bi2O3-TiO2/PAC. Meanwhile, the ROS produced by the composite were non-selective and attacked functional groups of SMZ. A heterogeneous junction was formed by doping Bi2O3 with TiO2, which greatly reduced the band gap of the photocatalysts and provided good conditions for the separation of photo-generated carriers, greatly improving the photocatalytic performance under visible light. The adsorption capacity of PAC ensured persistent high levels of pollutants for the photocatalytic system [5]. Additionally, the combination of PAC could improve the adsorption performance and recovery rate of photocatalysts, which was of great theoretical and engineering significance for an appropriate adsorption-photocatalysis composite.

Conclusions
Under a controlled Bi/Ti molar ratio, Bi2O3-TiO2/PAC ternary nanoparticles exhibiting adsorption-photocatalysis are capable of removing SMZ in water. The adsorption capacity of PAC

Conclusions
Under a controlled Bi/Ti molar ratio, Bi 2 O 3 -TiO 2 /PAC ternary nanoparticles exhibiting adsorption-photocatalysis are capable of removing SMZ in water. The adsorption capacity of PAC ensured persistent high levels of pollutants for the photocatalysis system. Adsorption onto active sites assisted by external diffusion is the main mechanism of adsorption. • O 2 − and h + were identified as the major ROS in photocatalysis. Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) exhibited the best SMZ removal performance, at 95.5% under directly solar light irradiation. The crystal size of Bi 2 O 3 -TiO 2 /PAC(10%-700 • C) was larger than that of Bi 2 O 3 -TiO 2 /PAC(8%-700 • C) and Bi 2 O 3 -TiO 2 /PAC(12%-700 • C). The visible light response was enhanced because of heterostructures, a narrowed band gap and an excellent mixture ratio of anatase to rutile in the phase of TiO 2 . As the Bi/Ti molar ratio increased, the pore volume decreased. The removal performance for SMZ in lake water was better than that for river water. Under acidic conditions, photocatalysis performance was greatly reduced. When the pH was above 7, the photocatalytic effect was enhanced while the adsorption performance declined. Besides, the increase in HCO 3 − improved the photocatalytic efficiency for SMZ, while the addition of Cl − , SO 4 2− or HA had negative effects on the photocatalytic efficiency for SMZ. Benzene ring fracture, SO 2 extrusion and nitrogenated SMZ were proposed as the main pathways of SMZ degradation. Antibiotics have been identified as a particular category of trace chemical contaminants. The Bi 2 O 3 -TiO 2 /PAC adsorption-photocatalysis system has exhibited great efficiency for the removal of one of the typical refractory antibiotics. The outcomes of this study provide significant insights into the adsorption-photocatalysis system and potentially benefits the control of antibiotic pollutants in water.
Author Contributions: X.Z., J.L. and H.W. were responsible for the experimental work. X.L. and Y.Y. supervised the laboratory work. Z.Z. led the research. N.W. and Y.S. provided technical support for this work. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China (51978006).