Novel Bi₃O₅I₂ Hollow Microsphere and Its Enhanced Photocatalytic Activity

Abstract

A new type of I-deficient bismuth oxyiodide Bi₃O₅I₂ with a hollow morphology was prepared by the solvothermal process. The structure, composition, morphology, optical property and photoelectric property of the as prepared photocatalyst were investigated through some characterization methods. Those characterization results showed that Bi₃O₅I₂ displayed a larger specific surface area, promising band structure and lower recombination of photoinduced carriers than pure BiOI. Bi₃O₅I₂ had a higher photocatalytic activity than BiOI on the decomposition of methyl orange (MO) under simulated solar light irradiation. The superoxide (·O₂⁻) and hole (h⁺) were the dominating active species during the degradation of MO. Its stability and reusability performance showed its great promising application in the degradation of organic pollutant.

1. Introduction

With increasing concerns on environmentally friendly processes, semiconductor photocatalysts have been given widely attention because of their promising applications. Among the multitudinous semiconductor photocatalysts, bismuth oxyiodide (BiOI) with unique layered crystal configuration [1,2,3] and excellent visible light absorption [4,5], is a promising photocatalytic material in the field of energy and environment [6,7,8,9,10]. Nevertheless, the photocatalytic activity of bismuth oxyiodide is far from potential industrial applications because of their relatively high recombination of photogenerated carriers and insufficient reduction and oxidation ability [11,12]. For the sake of increasing the photocatalytic performance of BiOI, measures should be taken to decrease the recombination of photogenerated carriers and improving its reduction and oxidation ability. So far, many measures have been taken to enhance the photocatalytic performance of BiOI, such as regulating morphology [13,14,15,16], exposing special crystal face [17,18,19], doping [20,21,22,23], compositing with other semiconductor [24,25,26,27] and synthesizing other bismuth oxyiodides such as Bi₄O₅I₂ [12,28], Bi₅O₇I [29,30,31] and Bi₇O₉I₃ [32,33].

Aside from BiOI, I-deficient bismuth oxyiodides have the positive oxidation and reduction ability [11,12], leading to its high photocatalytic activity. For example, Zhang et al. discovered that Bi₄O₅I₂ and Bi₅O₇I displayed enhanced photocatalytic activity for degradation of tetracycline hydrochloride, bisphenol A and Rhodamine B [11]. Ye et al. synthesized Bi₄O₅I₂ photocatalyst with excellent activities for photocatalytic H₂ evolution [12]. Generally, I-deficient bismuth oxyiodides were prepared by four methods: (1) Hydrothermal progress using the pH modifier [34], (2) in situ calcination progress of BiOI [11], (3) hydrolytic process of the bismuth molecular precursor [12], (4) chemical etching method for bismuth oxides [35]. However, these methods were time consuming or introduce potential environmental and safety problems. Therefore, a simple method of synthesizing the I-deficient bismuth oxyiodide is needed.

It is generally accepted that the morphology of the catalysts has the close relation with the catalyst performance [36,37]. Photocatalysts with a hollow structure display excellent photocatalytic performance on account of their high specific surface area. In addition, both their high energy conversion efficiencies and their strong light-harvesting capacities contribute to their excellent photocatalytic performance [15,38,39]. For example, Li et al. found that BiOI with hollow structures presented a higher photocatalytic performance than those BiOI with low-dimensional [15]. Until now, the I-deficient bismuth oxyiodide with different morphologies are synthesized by several researchers, including sheet-like morphology [40], rod-like morphology [29] and flower-like morphology [33,41]. However, there are no reports about synthesizing the I-deficient bismuth oxyiodide with hollow structure. Therefore, a hollow structure of the I-deficient bismuth oxyiodide should be synthesized.

In this research, a new type of I-deficient bismuth oxyiodide with a hollow structure was prepared by the hydrothermal progress for the first time. Their structures as well as its photocatalytic ability for the demineralization of methyl orange (MO) were discussed under systematical characterization.

2. Results

2.1. Molecular Formula Characterization and Phase Transformation

The energy dispersive spectrometry (EDS) apparatus was employed to investigate the chemical composition on the surface of the as prepared I-deficient bismuth oxyiodide. As shown in Figure 1, the I-deficient bismuth oxyiodide contained Bi, O and I elements. Meanwhile, the associated EDS elemental maps confirmed that these elements were distributed homogeneously. As shown in Table S1, the average Bi:O:I ratio in the surface of the I-deficient bismuth oxyiodide was 3:4.7:2, in conformity with the molecular formula Bi₃O₅I₂ (3:5:2). This result indicated that its chemical composition possibly was Bi₃O₅I₂. Meanwhile, the average Bi:O:I ratio of the pure BiOI are 1:1.1:1 (Table S2), which is close to the ideal ratio for BiOI (1:1:1). This indicated that the result of the EDS was convincing.

The inductively coupled plasma (ICP) analysis showed that the percentage of the Bi element in the bulk bismuth oxyiodide was 64.95%, in conformity with the molecular formula Bi₃O₅I₂ (65%). Consequently, based on the analysis of ICP and EDS, the molecular formula of the as prepared I-deficient BiOI can possibly be determined as Bi₃O₅I₂. Therefore, the I-deficient bismuth oxyiodide was labelled as Bi₃O₅I₂ in the following.

A thermogravimetric (TG) measurement was carried out to evaluate the phase transformation of Bi₃O₅I₂ and BiOI and the result was shown in Figure 2. In Figure 2, the TG curve of the Bi₃O₅I₂ sample was composed by three mass loss steps. The first mass loss step (250–360 °C) was attributed to the transformation from Bi₃O₅I₂ to Bi₄O₅I₂, whereas the second mass loss step (400–520 °C) was ascribed to the transformation from Bi₄O₅I₂ to Bi₅O₇I, the third step (580–760 °C) was ascribed to the transformation from Bi₅O₇I to Bi₂O₃, which can be illustrated as Equation (1). The weight remaining of each phase transformation step in Equation (1) was calculated to be 91.3%, 80.2% and 73.8%, respectively, which is very similar to the experimental value (91.7%, 80.1% and 73.2%). This result shows that the molecular formula Bi₃O₅I₂ is correct. According to the results of EDX, ICP and TG, the as prepared I-deficient bismuth oxyiodide can be determined as Bi₃O₅I₂. The phase transformation of BiOI

$${\mathrm{Bi}}_3{\mathrm{O}}_5{\mathrm{I}}_2\stackrel{91.3\%}{\to }\frac{3}{4}B{i}_4{O}_5{I}_2\stackrel{80.2\%}{\to }\frac{3}{5}B{i}_5{O}_{7}I\stackrel{73.8\%}{\to }\frac{3}{2}B{i}_2{O}_3$$

(1)

2.2. Structure Investigation and Morphology Observation

The crystal information of Bi₃O₅I₂ and BiOI were characterized by the X-ray diffraction (XRD), which was shown in Figure 3. The main diffraction peaks of the as prepared Bi₃O₅I₂ (Figure 3a) could be clearly seen at 2θ of 10.1, 20.5°, 29.2°, 31.7°, 45.7°,52.6°, 55.1° and 64.3°, respectively. According to the JCPDS file 73–2062, the diffraction peaks at 9.7°, 29.3°, 31.9°, 45.4°, 51.5° and 55.4°are corresponding to the (001), (012), (110), (020), (114) and (122) of BiOI, respectively, which is different with the main diffraction peaks of the as prepared Bi₃O₅I₂.

The typical scanning electron microscope (SEM) and transmission electron microscopy (TEM) were shown in Figure 4. It was clearly seen that Bi₃O₅I₂ had a hollow microsphere morphology with an average diameter of 1–3 μm and the entire sphere like structures were composed of numerous nanoplates (Figure 4a,b). In Figure 4c, there were clearly a contrast between the dark boundary and the relatively bright center, which further confirmed their hollow nature [42]. In Figure 4d, clear lattice fringes showed that the lattice spacing were 0.312 nm and 0.339 nm, which might be the (102) and (101) plane of Bi₃O₅I₂.

The surface chemical compositions of the Bi₃O₅I₂ and the pure BiOI, as well as the chemical states of them, were investigated by the X-ray photoelectron spectroscopy (XPS), which was shown in Figure 5. In Figure 5a, it could be seen that the surfaces of Bi₃O₅I₂ and BiOI consisted of Bi, O, I and C elements (C may come from the reference sample), indicating a high purity of Bi₃O₅I₂ and BiOI. In Figure 5b, the binding energy values of the Bi 4f spectrum of the BiOI was about 158.6 eV and 163.9 eV, which correspond to Bi 4f_7/2 and Bi 4f_5/2. It indicated the presence of Bi³⁺ in the sample [43]. As for the Bi₃O₅I₂, the peak of Bi 4f_7/2 (or 4f_5/2) peak could be decomposed into two bimodal peaks of 158.5 eV and 159.2 eV (or 163.8 eV and 164.5 eV), which could be ascribed to Bi³⁺ and Bi⁵⁺, respectively [44]. The O 1s spectrum of BiOI (presented in Figure 5c) had two peaks. One was at 530.8 eV and the other was at 532.9 eV. The peak at 530.8 eV should be attributed to the Bi-O bonds the existed in BiOI. The peak at 532.9 eV should be attributed to the O-H bonds on the surface of BiOI [43]. The O 1s spectrum of Bi₃O₅I₂ had three peaks at 530.7 eV, 531.4 eV and 532.9 eV. The peak at 530.7 eV belongs to the Bi³⁺-O bonds in the Bi₃O₅I₂ lattice [45]. The peak at 531.4 eV belongs to the Bi⁵⁺-O bonds in the Bi₃O₅I₂ lattice [43]. In addition, the peak at 532.9 eV belongs to the O-H bonds that existed on the Bi₃O₅I₂ surface. As shown in Figure 5d, the two peaks can be observed at 630.1 eV and 618.6 eV. They were ascribed to I 3d_5/2 and I 3d_3/2, which could be assigned to the monovalent oxidation state of the I elements in BiOI [43]. It should be noted that the peaks for I 3d of the Bi₃O₅I₂ samples (630.4 eV and 618.9 eV) presented a 0.3 eV shift to higher binding energy. This implied that the I in Bi₃O₅I₂ possibly existed at a different ambient chemical environment [43]. The XPS character showed that the Bi in the Bi₃O₅I₂ includes Bi³⁺ and Bi⁵⁺.

As shown in Figure 6, the specific surface area and the pore size distribution of Bi₃O₅I₂ and BiOI were analyzed by the N₂ absorption and desorption analysis. It was clear to see that all the isotherms were type IV, which exhibited the presence of mesoporous [43]. The hysteresis loop located at the range of 0.5–1.0 P/P₀ were ascribed to type H₃ in the IUPAC classification, indicating that the mesoporous could be considered as accumulated pores of the sample nanosheets [43]. Moreover, the hysteresis loop shifts approach P/P₀ = 1, implying the existence of macroporous (>50 nm). The BET specific surface area of the Bi₃O₅I₂ was 26.81 m²/g, while the surface area of BiOI was 13 m²/g, possibly due to its hollow microsphere structure. The PSD curves showed that the pore size range of Bi₃O₅I₂ and BiOI was 2–140 nm. The main pore size of Bi₃O₅I₂ was 21 nm, while the main pore size of BIOI was 39 nm. To our best knowledge, the larger specific surface area can lead to the increase of photocatalytic reaction sites [37]. Therefore, it could be concluded that the Bi₃O₅I₂ might have a higher photocatalytic activity than BiOI.

2.3. Optical Property and Photoelectric Property of the Photocatalysts

The UV-vis diffuse reflectance spectroscopy of Bi₃O₅I₂ and BiOI was shown in Figure 7a. In comparison with BiOI, Bi₃O₅I₂ had a clearly blue shift in the absorption edge, indicating Bi₃O₅I₂ had a higher energy gap (Eg). This agrees with the results of the I-deficient bismuth oxyiodide reported in the previous literature [46,47]. Meanwhile, the DRS displayed that the absorption band edge of Bi₃O₅I₂ was 586 nm, which indicated that Bi₃O₅I₂ had a good visible light absorption [29].

To estimate the Eg value of Bi₃O₅I₂, a classical Tauc approach was introduced. According to the equation αhν = K(hν − Eg)^n/2 (α, hν, K and Eg are the absorption coefficient, the energy of irradiation light, a constant and band gap energy, respectively), the Eg value can be calculated [43]. Considering BiOI and Bi₃O₅I₂, were indirect transition semiconductor, the value of n is 4 [31,32,34]. The band gaps of Bi₃O₅I₂ was calculated to be 2.02 eV and higher than that of BiOI (1.89 eV). In order to determine the relative positions of the valence band edge (VB), the VB-XPS spectra of Bi₃O₅I₂ and BiOI were measured and shown in Figure 7b. The VB edge of Bi₃O₅I₂ and BiOI was 1.50 eV and 1.42 eV. The conduction band (CB) edge of Bi₃O₅I₂ and BiOI could be obtained using the equation of E_CB = E_VB − Eg. Therefore, the CB positions of Bi₃O₅I₂ and BiOI were −0.72 eV and −0.47 eV. The suitable bang energy of Bi₃O₅I₂ compared to BiOI implied that Bi₃O₅I₂ has a stronger oxidation and reduction ability than BiOI [43]. As we all know, the electron in the less positive conduction band edge (CB) has higher reduction ability and the hole in the more positive valence band edge (VB) has higher oxidization ability. As we all know, the electron in the less positive conduction band edge (CB) has higher reduction ability and the hole in the more positive valence band edge (VB) has higher oxidization ability. Therefore, Bi₃O₅I₂ with the less positive CB position and more positive VB position had the promising photacatalytic activity than BiOI. The energy band structure of Bi₃O₅I₂ and BiOI were shown in Figure S2. The conduction band edge (CB) and the valence band edge (VB) of Bi₃O₅I₂ were −0.72 eV and 1.50 eV (versus normal hydrogen electrode (NHE)). Since the CB value of Bi₃O₅I₂ was less positive than E⁰(O₂/•O₂⁻) (−0.046 eV vs. NHE) and the VB value of Bi₃O₅I₂ was more positive than E⁰(•OH/OH⁻) (2.38 eV vs. NHE) [48], the photogenerated electrons can reduce O₂ to •O₂⁻ and the photogenerated holes cannot oxidize OH⁻ to •OH. Meanwhile, the VB value of Bi₃O₅I₂ was more positive than the redox potential of MO (1.48 eV versus NHE) [49], the photogenerated holes can oxidize MO directly. The result of active species experiments shown in Figure 11 further confirm the above results. Based on the band structure analysis, Bi₃O₅I₂ would exhibit effective photocatalytic performance, owing to its strong oxidation and reduction ability.

The transient photocurrent responses and electrochemical impedance spectroscopy (EIS) were further explored to estimate the generation and recombination rate of photogenerated electron-hole pairs, which was displayed in Figure 8. In Figure 8a, it was worth noting that the photocurrent density of the Bi₃O₅I₂ electrode is much higher than BiOI for each switch-on and -off event. The enhanced photocurrent response of the Bi₃O₅I₂ indicated a lower recombination of photogenerated electron-hole pairs than pure BiOI [50]. The photocurrent response mainly depends on the generation rate of photoinduced electron-hole pairs and the recombination rate of photogenerated electron-hole pairs. Meanwhile, the recombination rate of photogenerated electron-hole pairs were the rate controlling step of the photocurrent response. Just shown in Figure 7a, BiOI had a superior light harvesting than the I-deficient bismuth oxyiodides. It indicates that BiOI might have a higher generation rate of photoinduced electron-hole pairs than Bi₃O₅I₂. However, it should be noted that the recombination rate of photogenerated electron-hole pairs of BiOI was much higher than Bi₃O₅I₂. Therefore, the photocurrent response of Bi₃O₅I₂ is stronger than that of BiOI. The result of the electrochemical impedance spectroscopy (EIS) of BiOI and Bi₃O₅I₂ shown in Figure 8b can verify this result. Bi₃O₅I₂ had a smaller arc radius of the EIS Nyquist plot than BiOI, suggesting an effective separation of photoinduced electron−hole pairs and fast interfacial charge transfer than BiOI [11]. In conclusion, Bi₃O₅I₂ has a more efficient photocatalytic performance than BiOI.

2.4. Photocatalytic Performance of the Photocatalysts

The photocatalytic ability of Bi₃O₅I₂ were measured on the decomposing of MO (20 mg/L) in water under the simulated visible light, which was shown in Figure 9. In Figure 9a, the maximum absorption of MO shifted from 464 nm to 420 nm in the presence of Bi₃O₅I₂ after irradiation for 180 min [37]. Meanwhile, the color of MO changed from yellow to colorless during the photocatalytic reaction, which indicated that MO was degraded by Bi₃O₅I₂ under the simulated solar light irradiation.

The variation of the MO content with the irradiation time was shown in Figure 9b. It can be seen that 81% of MO was photodegraded by Bi₃O₅I₂ while only 11% of MO was degraded by BiOI. This confirmed that Bi₃O₅I₂ present higher photoactivities than BiOI. Moreover, it was much higher than most of the activity data over other similar catalysts reported in previous works, as compared in Table S3.

The quantum efficiency of the photocatalytic activity is important for the photocatalysts. However, there is little information about the quantity of electrons needed to degrade a methyl orange molecule in literatures. Additionally it is also difficult to measure it in our Lab. In order to calculate the quantum efficiency (QE) of Bi₃O₅I₂ and BiOI, the following equation about the apparent QE is put forward according to the literature and the calculated process is listed below in detail [12].

The photocatalytic reaction conditions: 100 mL, 20 mg/L methyl orange (MO 327.334 g/mol) solvent, the moles of MO is 6.11 × 10⁻⁶ mol. A light (λ ~ 420 nm) was used and the light intensity was 300 mW·cm⁻². The energy can be obtained on the surface (1.3 cm⁻²) of the reaction mixture solution. We assumed that it takes m electrons to degrade a methyl orange molecule. n % of MO was degraded after the t hour. The apparent QE detected under the same photocatalytic reaction condition and calculated to the equation as below:

$$\begin{array}{l}\mathrm{QE}=\frac{\mathrm{number} \mathrm{of} \mathrm{reacted} \mathrm{eletrons}}{\mathrm{number} \mathrm{of} \mathrm{incident} \mathrm{photons}}\times 100\%=\frac{6.11\times {10}^{-6}mol\times m\times n\%}{\raisebox{1ex}{$E$}\!\left/ \!\raisebox{-1ex}{$h\nu $}\right.} =\frac{6.11\times {10}^{-6}mol\times m\times n\%\times hc}{E\lambda }\\ =\frac{6.11\times {10}^{-6}\times m\times n\%\times 6.02\times {10}^{23}\times 6.63\times {10}^{-34}\times 3\times {10}^8}{300\times 1.3\times {10}^{-3}\times t\times 3600\times 420\times {10}^{-9}}\\ =0.00124\frac{m\times n\%}{t}\end{array}$$

When m = 1, it can be found that the QE of Bi₃O₅I₂ and BiOI were 0.033% and 0.005% at 420 nm.

Based on the above analysis, the high photocatalytic activity of Bi₃O₅I₂ should be ascribed to its hollow structure, suitable band structure and efficient separation of photogenerated electron–hole pairs. Firstly, with a hole in microspheres, Bi₃O₅I₂ has a larger surface area (26.81 m²/g) than BiOI (13.00 m²/g). With a higher surface area, Bi₃O₅I₂ had more active sites and absorb more MO on its surface. Moreover, the high surface area of hollow structures could promote the transfer of photoinduced carriers to the surface and enhance the decomposition of MO [37]. Secondly, Bi₃O₅I₂ had a higher CB position and lower VB location than BiOI, which enable the photogenerated electron and hole of the Bi₃O₅I₂ strong oxidation capability [9]. Thirdly, the photocurrent spectra showed that Bi₃O₅I₂ had a lower recombination of photoinduced carriers under the simulated light irradiation. Therefore, Bi₃O₅I₂ should have an enhanced photocatalytic performance since the photocatalytic activity of the photocatalyst mainly depend on the charge carrier generation and separation [9].

2.5. The Lifetime eValuation of the Photocatalysts

The reusability of the Bi₃O₅I₂ material was investigated by recycling and reusing the catalyst (in Figure 10). Figure 10a shows no distinct loss in the photocatalytic activity when MO was degraded for the fourth time. Therefore, the Bi₃O₅I₂ exhibited an excellent stability during the degradation of MO.

In order to evaluate the structural stability of Bi₃O₅I₂, the fresh and used Bi₃O₅I₂ were characterized by the XRD. The result was shown in Figure 10b. Obviously, there was no apparent change for the recycled Bi₃O₅I₂ compared with the fresh Bi₃O₅I₂, indicating the Bi₃O₅I₂ are stable and reusable. Owing to its good photocatalytic activity as well as its stable and reusable character, Bi₃O₅I₂ shows great potential for the photocatalytic degradation of organic pollutants in waste water.

2.6. The Possible Photacatalytic Mechanism

The probe mechanism Bi₃O₅I₂ was further evaluated by the radicals and holes trapping experiments and the result was shown in Figure 11. Benzoquinone (BQ) was used as the quenching agent of superoxide radical (•O₂⁻). Potassium iodide (KI) was used as the scavenger of hole (h⁺) and hydroxyl radical (•OH). Tert-butyl alcohol (TBA) was used as the quenching agent of •OH. Their concentrations in the MO solution were 0.1 mM/L. Figure 11 showed that the degradation of MO was obviously decreased from 81% to 55% in the presence of KI, indicating that h⁺ or •OH played an important role for degrading MO [51]. While, the addition of TBA did not influence the MO degradation efficiency, indicating that •OH has no effect on the degrading MO. Therefore, it could be inferred that h⁺ was the major active species for the degradation of MO. When BQ was added, the MO degradation efficiency decreased significantly from 81% to 29%, suggesting that •O₂⁻ played the key role for degrading MO. In conclusion, •O₂⁻ was the primary active species and h⁺ was the main active species in the degradation of MO.

In conclusion of the analysis above, a possible photocatalytic program was proposed, which was displayed in Scheme 1. Firstly, the MO molecules were absorbed on the surface of the Bi₃O₅I₂. Secondly, when the Bi₃O₅I₂ was irradiated by simulated visible light, the photogengenerated electron would transfer from the valence band to the conduction band, leaving h⁺ in the valence band and electrons in the conduction band. Thirdly, h⁺ are capable of oxidizing MO directly and the electrons were also believed to be capable of benefiting the oxidation process through reduction of the absorbed O₂ into •O₂⁻. As a result, the MO molecular was demineralized to CO₂, H₂O and other inorganic substances by the •O₂⁻ and h⁺.

3. Materials and Method

3.1. Materials

The bismuth nitrate pentahydrate (99.5 %, Sinopharm Chemical Reagent Co., Ltd. Shanghai, China), potassium iodide(99.5 %, Sinopharm Chemical Reagent Co., Ltd. Shanghai, China), ethyl alcohol and ethylene glycol (99.0 %, Sinopharm Chemical Reagent Co., Ltd. Shanghai, China) were used as received. The distilled water was synthesized by the water purification system.

3.2. Synthesis of the Bi₃O₅I₂ and BiOI

In a typical synthesis route, 0.004 mol of bismuth nitrate pentahydrate and 0.004 mol of potassium iodide (KI) was dissolved in 40 mL of ethylene glycol under vigorously stirring for 0.5 h, respectively. After that, the KI solution was added dropwise into the solution which contained bismuth nitrate. Then, the mixture solution was stirred vigorously for 0.5 h at 25 °C and transferred into a 100 mL Teflon-lined stainless autoclave. The Teflon-lined stainless autoclave was heated to 160 °C and maintained for 12 h in a homogeneous reactor and air-cooled to room temperature. The synthesized precipitates were centrifuged and washed three times by ethanol and deionized water, and dried at 80 °C for 12 h in air. The sample was labelled as the I-deficient bismuth oxyiodide. Meanwhile, BiOI was prepared with the method above except using distilled water as the solvent of KI.

3.3. Photocatalyst Characterization

The samples are tested by the X-ray diffraction (XRD) with a D8 ADVANCE X-ray diffractometer (Cu Kα radiation, λ = 1.54178 Å Karlsruhe, German). A field emission scanning electron microscope (FE-SEM, JSM-7001F, Kyoto, Japan) was used to characterize the morphology of the samples. The FE-SEM was equipped with an energy dispersive analysis system of the X-ray spectroscope (EDS) as well as an elemental mapping system. The images of the transmission electron microscopy (TEM) for the samples were from a JEOL (JEM-2100F, Kyoto, Japan) operating at 200 kV. The N₂ adsorption and desorption isotherms were investigated by a Tristar II 3020 sorptometer (Atlanta, GA, USA). The operating condition is at 77 K with the samples degassed at 100 °C in vacuo. The XPS spectrum was obtained with a KRATOS X-ray photoelectron spectroscopy(AXIS ULTRA DLD, Manchester, UK). The diffuse reflectance UV–vis absorbance spectroscopy were recorded by using a spectrophotometer (Shimadzu UV-2550, Kyoto, Japan) BaSO₄ as the reference. The thermogravimetric analysis (TG) was carried out with a Netzsch STA 449C (Free State of Bavaria, German). In addition, the samples were heated with 10 °C/min from 25 °C to 900 °C in air atmosphere. The content of Bi was estimated by an Inductive Coupled Plasma emission spectrometer (Thermo iCAP6300, Waltham, MA, USA). The CHI 660E electrochemical system (Shanghai, China) was used for photoelectrochemical and electrochemical measurements and the sample preparation process was the same with the previous report [52].

3.4. Test of Photocatalytic Activity

To evaluate the photocatalytic activities of the as prepared samples, the photocatalytic degradation of MO was used under simulated visible light with a radiation source of the 300 W Xe lamp (PLS-SXE 300, Beijing, China). A 250 mL home-made reactor combined with a cooling water system was used. In each experiment, 100 mL of the MO solution (20 mg/L) was put in the reactor. A 25 mg catalyst was added into the MO solution. After that, the mixture of the MO solution and 25 mg catalyst placed in the dark was stirred for 60 min. Then, the mixture was exposed to simulated solar light with stirring. During the process, a 5 mL mixture was sampled at 30 min intervals. In addition, the concentration of MO in the mixture was measured by a UV–visible spectrophotometer. The spectrophotometer is the UV-2550 of Shimadzu, Japan, operating at 464 nm with water as a reference.

4. Conclusions

In summary, a new I-deficient bismuth oxyiodide Bi₃O₅I₂ with a hollow structure was successfully prepared by an ethylene glycol-assisted solvothermal process. The average diameter of the Bi₃O₅I₂ hollow microspheres was 1–3 μm, and the hollow nature contribute to the larger BET specific surface area than BiOI. The Bi₃O₅I₂ presented a higher photocatalytic performance on the degradation of MO under simulated solar light irradiation than BiOI did. The enhanced photocatalytic activities of Bi₃O₅I₂ could be contributed to its hollow structure, its appropriate band-gap and its low recombination rate of photogenerated carriers. The •O₂⁻ and h⁺ was the main active specie in the degradation of MO. The result of the stable and reusable performance indicated that Bi₃O₅I₂ showed great potential for the photocatalytic degradation of organic pollutants in waste water.