Composite Magnetic Photocatalyst Bi5O7I/MnxZn1−xFe2O4: Hydrothermal-Roasting Preparation and Excellent Photocatalytic Activity

A new composite magnetic photocatalyst, Bi5O7I/MnxZn1−xFe2O4, prepared by a hydrothermal-roasting method was studied. The photocatalytic properties of Bi5O7I/MnxZn1−xFe2O4 were evaluated by degradation of Rhodamine B (RhB) under simulated sunlight irradiation, and the structures and properties were characterized by X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), ultraviolet-visible light (UV-Vis) diffuse reflectance spectra (DRS), and a vibrating sample magnetometer (VSM). The results indicated that Bi5O7I/MnxZn1−xFe2O4 was an orthorhombic crystal, which was similar to that observed for Bi5O7I. Bi5O7I/MnxZn1−xFe2O4 consisted of irregularly shaped nanosheets that were 40–60 nm thick. The most probable pore size was 24.1 nm and the specific surface area was 7.07 m2/g. Bi5O7I/MnxZn1−xFe2O4 could absorb both ultraviolet and visible light, and the energy gap value was 3.22 eV. The saturation magnetization, coercivity and residual magnetization of Bi5O7I/MnxZn1−xFe2O4 were 3.9 emu/g, 126.6 Oe, and 0.7 emu/g respectively, which could help Bi5O7I/MnxZn1−xFe2O4 be separated and recycled from wastewater under the action of an external magnetic field. The recycling experiments revealed that the average recovery rate of the photocatalyst was 90.1%, and the photocatalytic activity was still more than 81.1% after five cycles.


Introduction
In recent years, with the progress of nanotechnology and the application of photocatalytic technology in environmental pollution treatment [1][2][3], lots of semiconducting metal-oxide nanostructures were widely used for water purification due to their great photocatalytic performance [4,5]. Grottrup [6] applied Bi for doping ZnO, which significantly enhanced its ability in the photocatalytic degradation of methylene blue. Huang [7] indicated that the rate constant of degradation of 17 β-estradiol over N-doped Bi 2 O 3 was 6.3 times that obtained over Bi 2 O 3 . Oppong [8] pointed out that the reason for the better photocatalytic performance of Gd-TiO 2 -graphene oxide (GO) nanocomposites compared to pure TiO 2 was because GO sheets and Gd 3+ ions are excellent co-catalysts and their presence promotes the reaction sites. Currently, bismuth-based nanometer semiconductors are one of the research hotspots in the field of photocatalytic materials due to their unique electronic structure and excellent absorption ability of ultraviolet and visible light [9][10][11][12]. Bi 5 O 7 I, an oxygen-rich bismuth-based nanometer semiconductor [13][14][15], is composed of a Bi 6p orbital at the bottom of the conduction band and Bi 6s, O 2p, and I 5p orbitals at the top of the valence band [16,17]. The Bi 6s and O 2p orbitals can form several dispersed hybrid valence bands, while the I 5p orbital disperses the valence bands further [18,19]. Consequently, the above results accelerate the migration of photo-generated holes and promote the occurrence of oxidation reactions [20,21]. Sun et al. thought that the Bi 5 O 7 and I sections form a unique hierarchical structure successively along the c-axis orderly. As an accelerator for the separation of photo-generated electron-hole pairs, the permanent electrostatic field between the layers can improve the photocatalytic activity of Bi 5 O 7 I [22]. Xia et al. prepared sheets of Bi 5 O 7 I using the calcining method. However, the photocatalytic degradation rate of Bi 5 O 7 I (0.1 g) for Rhodamine B (RhB) solution (100 mL, 10mg/L) was only 52% in 120 min under simulated sunlight irradiation [23]. In order to enhance the photocatalytic activity of Bi 5 O 7 I, some means were mentioned, such as doping and compounding [24][25][26][27]. The photocatalytic degradation rate of Eu (3%)/Bi 5 O 7 I microspheres was 2.8 times that of Bi 5 O 7 I [28], and the photocatalytic degradation efficiency of a graphitic carbon nitride (g-C 3 N 4 ) (10%)-Bi 5 O 7 I nanocomposite was 1.4 times that of Bi 5 O 7 I [29].
Most organic pollutants in wastewater can be degraded using photocatalytic technology, with good degradation effects and no secondary pollution. Nevertheless, the complex process, large energy consumption, and low recovery rate are the main disadvantages of the common recovery method, because photocatalytic materials disperse in wastewater uniformly [30]. Mn x Zn 1−x Fe 2 O 4 is a soft magnetic ferrite material with its own benefits, such as high saturation magnetization, high permeability, low coercive force, low loss, strong stability, and so on [31]. Therefore, composite magnetic photocatalytic materials prepared with Mn x Zn 1−x Fe 2 O 4 as a magnetic matrix can achieve the goal of magnetic recovery, and the heterojunction formed between Mn x Zn 1−x Fe 2 O 4 and the photocatalyst is conducive to enhancing the photocatalytic activity of the composite. For instance, Zhang et al. [32] synthesized a Mn x Zn 1−x Fe 2 O 4 /α-Bi 2 O 3 composite magnetic photocatalyst using the dip-calcination method. However, the recovery rate was only 84.1% under the action of an external magnet, and the degradation time was more than four hours when the degradation rate attained 86.2% after five recoveries. The energy consumption of the preparation process is tremendous, because the precursor of Mn x Zn 1−x Fe 2 O 4 must be calcined at 1200 • C for three hours. In addition, the synthesis of composite magnetic photocatalytic materials using Bi 5 O 7 I as a photocatalyst and Mn x Zn 1−x Fe 2 O 4 as the magnetic matrix is rarely reported.
To overcome these shortcomings, a Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 composite magnetic photocatalyst was prepared using a hydrothermal-roasting method, and the structures and properties were characterized by X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), ultraviolet-visible light (UV-Vis) diffuse reflectance spectra (DRS), and a vibrating sample magnetometer (VSM). In the meantime, the activity and stability of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 were evaluated through the degradation of RhB under simulated sunlight irradiation. Bi5O7I/MnxZn1−xFe2O4 was prepared using a hydrothermal-roasting method. Firstly, 5 mL of ethylene glycol (EG) was dissolved in 35 mL of deionized water with stirring for 10 min to gain solution A. Then, 2 mmol Bi (NO3)3·5H2O and 2 mmol KI were continuously dissolved in solution A while stirring to acquire suspension liquid B. Then, 10 wt.% as-prepared MnxZn1−xFe2O4 was added into suspension liquid B with continuous stirring for 60 min. Afterward, the mixed turbid solution was transferred into a Teflon-lined autoclave for reaction at 160 °C for 12 h, and the resulting precipitates were washed with deionized water several times. Finally, the precipitates were dried at 80 °C for 5 h and roasted at 480 °C for 2 h in a muffle, obtaining Bi5O7I/MnxZn1−xFe2O4. The synthetic process is displayed in Scheme 1.

Characterization
The structures of the as-prepared products were characterized by X-ray diffraction (XRD; Shimadzu, XRD-6000, Shimadzu, Kyoto, Japan) and Fourier-transform infrared spectroscopy (FTIR; Nicolet iS50, Thermo Fisher Scientific, Waltham, MA, USA). The morphologies and microstructures of the products were observed using scanning electron microscopy (SEM; S4800, Hitachi, Tokyo, Japan) and transmission electron microscopy (TEM; Tecnai G2F20, FEI, Hillsboro, OR, USA). The surfaces and apertures of the products were measured using an automatic multistation surface and aperture analyzer (Quadrasorb 2MP, Quantachrome, Boynton Beach, FL, USA). The element contents in the composite were analyzed by X-ray photoelectron spectroscopy (XPS, ESCALAB250Xi, Thermo Fisher Scientific, Waltham, MA, USA). The optical absorption ability and magnetic performance of the products were identified with UV-Vis diffuse reflectance spectra (UV-Vis DRS; TU1901, Beijing, China) and a vibrating sample magnetometer (VSM; 7410, Lake Shore, Westerville, OH, USA) respectively.

Photocatalytic Evaluation
The photocatalytic properties of the samples were evaluated through the degradation of Rhodamine B (RhB) under simulated sunlight irradiation. Firstly, 0.1 g of photocatalyst and 100 mL of RhB aqueous solution (10 mg·L −1 ) were put into a beaker and stirred for 30 min in the dark to reach the adsorption balance. Then, the mixtures were irradiated with a xenon lamp (CEL-HXF3000, AULTT) of 300 W, and the ultraviolet-visible emission spectrum. Then, 4 mL of the solution was

Characterization
The structures of the as-prepared products were characterized by X-ray diffraction (XRD; Shimadzu, XRD-6000, Shimadzu, Kyoto, Japan) and Fourier-transform infrared spectroscopy (FTIR; Nicolet iS50, Thermo Fisher Scientific, Waltham, MA, USA). The morphologies and microstructures of the products were observed using scanning electron microscopy (SEM; S4800, Hitachi, Tokyo, Japan) and transmission electron microscopy (TEM; Tecnai G2F20, FEI, Hillsboro, OR, USA). The surfaces and apertures of the products were measured using an automatic multistation surface and aperture analyzer (Quadrasorb 2MP, Quantachrome, Boynton Beach, FL, USA). The element contents in the composite were analyzed by X-ray photoelectron spectroscopy (XPS, ESCALAB250Xi, Thermo Fisher Scientific, Waltham, MA, USA). The optical absorption ability and magnetic performance of the products were identified with UV-Vis diffuse reflectance spectra (UV-Vis DRS; TU1901, Beijing, China) and a vibrating sample magnetometer (VSM; 7410, Lake Shore, Westerville, OH, USA) respectively.

Photocatalytic Evaluation
The photocatalytic properties of the samples were evaluated through the degradation of Rhodamine B (RhB) under simulated sunlight irradiation. Firstly, 0.1 g of photocatalyst and 100 mL of RhB aqueous solution (10 mg·L −1 ) were put into a beaker and stirred for 30 min in the dark to reach the adsorption balance. Then, the mixtures were irradiated with a xenon lamp (CEL-HXF3000, AULTT) of 300 W, and the ultraviolet-visible emission spectrum. Then, 4 mL of the solution was withdrawn at set time intervals, before being centrifuged at 4000 rpm for 5 min to get the supernatant. Finally, the characteristic absorbance of RhB was measured using a UV-Vis spectrophotometer. withdrawn at set time intervals, before being centrifuged at 4000 rpm for 5 min to get the supernatant. Finally, the characteristic absorbance of RhB was measured using a UV-Vis spectrophotometer.  The FTIR spectra of MnxZn1−xFe2O4, Bi5O7I, and Bi5O7I/MnxZn1−xFe2O4 are shown in Figure 2. The characteristic peaks at 3434 cm −1 and 2360 cm −1 were attributed to the stretching vibration and bending vibration of the hydroxyl group (-OH) from surface-adsorbed water, respectively [33]. Typical Raman bands of the Fe-O-Fe bond and the stretching vibration of the Zn-O bond were located at 1399 cm −1 and 568 cm −1 , respectively [34]. The intensive signals around 1389 cm −1 , 846 cm −1 , and 610 cm −1 referred to the bending vibration of the Bi-O bond, and 491 cm −1 referred to the stretching vibration of the Bi-O bond [35].     In order to observe the morphology of the materials, the samples were characterized by SEM. As seen from Figure 4, Bi5O7I was composed of irregularly shaped nanosheets, while MnxZn1−xFe2O4 was a spherical particle. The layer thickness of the nanosheets was around 40-60 nm, as shown in Figure 4c. The energy-dispersive X-ray spectroscopy (EDS) spectrum confirmed that the particles in Figure 4c were MnxZn1−xFe2O4., demonstrating the successful creation of the MnxZn1−xFe2O4 and Bi5O7I compound. In order to observe the morphology of the materials, the samples were characterized by SEM. As seen from Figure 4, Bi 5 O 7 I was composed of irregularly shaped nanosheets, while Mn x Zn 1−x Fe 2 O 4 was a spherical particle. The layer thickness of the nanosheets was around 40-60 nm, as shown in Figure 4c. The energy-dispersive X-ray spectroscopy (EDS) spectrum confirmed that the particles in Figure 4c were Mn x Zn 1−x Fe 2 O 4 , demonstrating the successful creation of the Mn x Zn 1−x Fe 2 O 4 and Bi 5 O 7 I compound.    A specific surface analyzer was utilized to research the specific surface area and pore diameter distribution of Bi5O7I/MnxZn1−xFe2O4. From Figure 6, according to the Brunauer isotherm classification method, the adsorption-desorption isotherm belonged to class IV. The pore diameter distribution curve described that the most probable pore size of Bi5O7I/MnxZn1−xFe2O4 was 24.1 nm. Furthermore, the specific surface area of the Bi5O7I/MnxZn1−xFe2O4 sample calculated using the Brunauer-Emmett-Teller (BET) model was 7.07 m 2 /g.   A specific surface analyzer was utilized to research the specific surface area and pore diameter distribution of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 . From Figure 6, according to the Brunauer isotherm classification method, the adsorption-desorption isotherm belonged to class IV. The pore diameter distribution curve described that the most probable pore size of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 was 24.1 nm. Furthermore, the specific surface area of the Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 sample calculated using the Brunauer-Emmett-Teller (BET) model was 7.07 m 2 /g.

Absorption Light Ability and Magnetic Properties
The UV-Vis DRS and the (ahv) 1/2 −hv curve are shown in Figure 7. The largest absorption wavelengths of Bi 5 O 7 I and Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 were 470 nm and 600 nm, respectively, illustrating that both could absorb ultraviolet and visible light, and that adding Mn x Zn 1−x Fe 2 O 4 extended the range of absorption light. Moreover, the band-gap energy of the samples could be obtained using Equation (1) [22].
where α, h, v, and E g are the absorption coefficient, Planck constant, light frequency, and bandgap width, respectively. A is a constant, and n depends on the transition type of the semiconductor optical carriers (direct transition, n = 1; indirect transition, n = 4). According to the plots of (ahv) 1/2 −hv, the band-gap energies of Bi 5 O 7 I and Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 were determined to be 3.27 eV and 3.22 eV, respectively. In general, the magnetic performance of catalysts determines the recovery efficiency. Therefore, the magnetic hysteresis loops of the samples were measured. Figure 8b depicts that the saturation magnetization (Ms), coercivity (Hc), and residual magnetization (Mr) of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 were 3.9 emu/g, 126.6 Oe, and 0.7 emu/g, respectively. Compared with Mn x Zn 1−x Fe 2 O 4 , Ms declined because the mass ratio of magnetic materials in the photocatalyst was only 10%. It is worth noting that Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 is easy to magnetize or demagnetize, and its hysteresis loss was small in the alternating magnetic field. Figure 8d shows that the particles of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 in the suspension (the right bottle) moved to the magnet rapidly when a magnet was placed close to the bottle, and the suspension became clear after 4 min. However, the suspension containing Bi 5 O 7 I (the left bottle) did not show any obvious change under the same conditions. From the above analysis, it was determined that the composite photocatalyst has great magnetic separation capabilities.
where α, h, v, and Eg are the absorption coefficient, Planck constant, light frequency, and bandgap width, respectively. A is a constant, and n depends on the transition type of the semiconductor optical carriers (direct transition, n = 1; indirect transition, n = 4). According to the plots of (ahv) 1/2 −hv, the band-gap energies of Bi5O7I and Bi5O7I/MnxZn1−xFe2O4 were determined to be 3.27 eV and 3.22 eV, respectively. In general, the magnetic performance of catalysts determines the recovery efficiency. Therefore, the magnetic hysteresis loops of the samples were measured. Figure 8b depicts that the saturation magnetization (Ms), coercivity (Hc), and residual magnetization (Mr) of Bi5O7I/MnxZn1−xFe2O4 were 3.9 emu/g, 126.6 Oe, and 0.7 emu/g, respectively. Compared with MnxZn1−xFe2O4, Ms declined because the mass ratio of magnetic materials in the photocatalyst was only 10%. It is worth noting that Bi5O7I/MnxZn1−xFe2O4 is easy to magnetize or demagnetize, and its hysteresis loss was small in the alternating magnetic field. Figure 8d shows that the particles of Bi5O7I/MnxZn1−xFe2O4 in the suspension (the right bottle) moved to the magnet rapidly when a magnet was placed close to the bottle, and the suspension became clear after 4 min. However, the suspension containing Bi5O7I (the left bottle) did not show any obvious change under the same conditions. From the above analysis, it was determined that the composite photocatalyst has great magnetic separation capabilities.

Photocatalytic Activity
The photocatalytic activity of the as-prepared samples was investigated through photocatalytic degradation experiments of RhB. From Figure 9a, pure Bi5O7I presented the highest photocatalytic activity, whereby 97.6% of RhB was degraded within 120 min. The degradation properties of 5%, 10%, 15%, 20%, and 25% MnxZn1−xFe2O4/Bi5O7I were 96.3%, 96.7%, 93.8%, 82.3%, and 72.7%, respectively. In addition, the first-order kinetics model given by equation ln(C0/C) = kt was applied to quantitatively understand the reaction kinetics, where C0 (mg·L −1 ) is the initial concentration of RhB solution, C (mg·L −1 ) is the concentration in aqueous solution at time t, and k (min −1 ) is the apparent first-order kinetic constant [36]. The degradation constants (in Figure 9b) were calculated to be 0.0284, 0.0252, 0.0274, 0.0217, 0.0138, and 0.0108 min −1 for pure Bi5O7I, and 5%, 10%, 15%, 20%, and 25% MnxZn1−xFe2O4/Bi5O7I samples, respectively. Compared with Bi5O7I, the photocatalytic activity of the compounds declined for two reasons. On one hand, MnxZn1−xFe2O4 became the recombination center of the photogenerated electron (e − ) and hole (h + ), which reduced the lifetime of the photogenerated carriers. On the other hand, adding MnxZn1−xFe2O4 decreased the amount of catalyst in the compound.

Stability and Recycling Ability
The stability and the recycling ability of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 were studied through recycling experiments. After each reaction, the photocatalyst was separated by an external magnet and then washed with deionized water before being dried at 80 • C for 3 h. The recycling experiments show that the average recovery rate was 90.1%. Figure 10 shows that the degradation rate was still more than 81.1% after five reuses. The experimental results indicate that the composite magnetic photocatalyst can be reused several times with excellent stability.

Stability and Recycling Ability
The stability and the recycling ability of Bi5O7I/MnxZn1−xFe2O4 were studied through recycling experiments. After each reaction, the photocatalyst was separated by an external magnet and then washed with deionized water before being dried at 80 °C for 3 h. The recycling experiments show that the average recovery rate was 90.1%. Figure 10 shows that the degradation rate was still more than 81.1% after five reuses. The experimental results indicate that the composite magnetic photocatalyst can be reused several times with excellent stability.

Photocatalytic Mechanism
There are a few reasons why the composite magnetic photocatalyst has good photocatalytic activity. Firstly, the existence of a porous structure is helpful for improving the transfer efficiency of the photogenerated electron and hole. In this structure, the distance from where photogenerated charge is generated to the semiconductor surface is shortened, which can effectively reduce the

Photocatalytic Mechanism
There are a few reasons why the composite magnetic photocatalyst has good photocatalytic activity. Firstly, the existence of a porous structure is helpful for improving the transfer efficiency of the photogenerated electron and hole. In this structure, the distance from where photogenerated charge is generated to the semiconductor surface is shortened, which can effectively reduce the recombination of the photogenerated electron and hole [37].
Secondly, the band structure of the photocatalyst is beneficial for producing a hydroxyl radical (•OH). Figure 11 illustrates the proposed mechanism for the photocatalytic activity of Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 with RhB. As is known, •OH is the major active substance in the photocatalytic degradation of organic pollutants [38]. The adsorbed water is oxidized to •OH by a hole when the valence band top has more positive redox potential than that of •OH/H 2 O (+2.27 eV). The position of the conduction band bottom (E CB ) can be obtained using Equation (2) [39].
where X is the absolute electronegativity of the semiconductor oxide, E C is the potential energy of the free electron in a standard hydrogen electrode (~4.5 eV), and E g is the band gap of the semiconductor oxide. The position of the conduction band bottom for the photocatalyst was determined to be 0.04 eV. Therefore, the position of the valence band top was 3.31 eV, which is sufficient to turn OH − into •OH through oxidation.
Nanomaterials 2019, 9, x FOR PEER REVIEW 12 of 14 Figure 11. Schematic of the possible reaction mechanism of the photocatalytic procedure.

Conclusions
The composite magnetic photocatalyst Bi5O7I/MnxZn1−xFe2O4 was prepared using a hydrothermal-roasting method. This is convenient for mass production in the future because of its simple process and low cost. According to the analysis results of XRD, FTIR, XPS, SEM, and TEM, Bi5O7I and MnxZn1−xFe2O4 were successfully combined. Bi5O7I/MnxZn1−xFe2O4 is a mesoporous material, able to absorb ultraviolet and visible light. Meanwhile, Bi5O7I/MnxZn1−xFe2O4 is a soft magnetic material with great magnetic induction intensity. The photocatalytic degradation and recycling experiments revealed that Bi5O7I/MnxZn1−xFe2O4 has good photocatalytic activity and stability.

Conclusions
The composite magnetic photocatalyst Bi 5 O 7 I/Mn x Zn 1−x Fe 2 O 4 was prepared using a hydrothermalroasting method. This is convenient for mass production in the future because of its simple process and low cost.