As a low-energy and green technology, photocatalytic technology has attracted increasing attention in recent decades [1, 2]. This technology can use solar energy to destroy organic and highly toxic macromolecules [3]. Many photocatalysts such as TiO2, ZnO, V2O5, and WO3 [4-7] have the advantages of high photosensitivity, low cost, natural nontoxicity, and environmental friendliness for photocatalytic and photoelectrochemical applications [8]. However, those photocatalysts respond only to ultraviolet light due to their wide band gap. Visible-light catalyst development focused on ion doping, noble metal deposition, and catalyst compounding of traditional catalysts to reduce the width of the band gap [9-11]. That method is highly attractive for exploring new photocatalysts with relatively narrow band gaps and high photocatalytic activity for improving their reaction to visible light [12].
Recently, Bismuth-based semiconductors such as BiOX (X = Cl, Br, and I), Bi2WO6, and Bi2MoO6 have received significant attention due to their unique electronic structures and decent photocatalytic properties [13-15]. Among them, bismuth oxyhalides [BiOX (X = Cl, Br, and I)] have been confirmed as promising photocatalysts for water treatment [16, 17]. Bismuth oxychloride (BiOCl) has shown excellent photocatalytic activity under UV light and has achieved the complete degradation of Rhodamine B (RhB) in 40 min [18]. However, its band gap of 3.2 to 3.5 eV means that it cannot degrade pollutants effectively under visible light [19].
Different atomic ratios of Bi:O:Cl, such as Bi3O4Cl (2.60 eV), Bi12O17Cl2 (2.10 eV), Bi12O15Cl6 (2.86 eV), and Bi24O31Cl10 (2.70 eV), have been synthesized in recent years [20-23]. Those catalysts have narrower band gaps than that of BiOCl and can absorb visible light. Recent research showed that a lower Cl/O ratio might reduce the band gap width of semiconductors and enhance their ability to use visible light [24]. Bi12O17Cl2 has the lowest Cl/O ratio, and its band gap is the narrowest among those catalysts, so it has great potential for improving visible-light utilization efficiency. However, Bi12O17Cl2 has a quick electron-hole recombination [25] that reduces the efficiency of photocatalytic reactions. On the other hand, BiOCl has high stability and a low photoelectron recombination rate. In addition, the structure of BiOCl and Bi12O17Cl2 are similar, and their composite will maintain stability. Thus, combining BiOCl with Bi12O17Cl2 will improve the latter's stability and delay electron-hole recombination, which will improve the photocatalytic ability of the photocatalyst and facilitate the removal of refractory organic pollutants.
However, a BiOCl-Bi12O17Cl2 composite, synthesized by a solvothermal method via adjusting the basicity, showed an irregular morphology and insufficient photocatalytic activity. After 4 h of irradiation, only 21.19% to 73.32% pollutant removal was achieved for various pollutants including dyes and pharmaceuticals [26]. Furthermore, the waste solvent produced during preparation likely caused secondary pollution. To improve the catalytic activity and solve the waste solvent problem, an ultrasonication-assisted hydrothermal method was adopted. Ultrasonic waves can strengthen the mass transfer and thermal effects through cavitation [27]. The ultrasonic hydrothermal method can promote the dispersion of active materials on the surface of the carrier and introduce the active ingredients into the material structure, thereby generating more active parts to improve catalytic activity [28].
In this study, a BiOCl-Bi12O17Cl2 nanocomposite was synthesized using a novel ultrasonic hydrothermal method and was used for a visible-light catalytic reaction. Rh B and ciprofloxacin (CIP) were used as target pollutants. The removal efficiency and catalyst reusability were studied. The mechanism of the catalytic degradation was determined by active species experiments. The aim of this study was to develop a green, stable, and high-efficiency visible-light catalyst for the treatment of refractory organic pollutants.
Potassium chloride (KCl) was purchased from Sinopharm Chemical Reagent Co., Ltd., China. Sodium hydroxide (NaOH), ethanol, and tert-butyl alcohol (TBA) were purchased from Beijing Chemical Works, China. Bismuth nitrate pentahydrate [(Bi(NO3)3·5H2O)], benzoquinone (BQ), and sodium oxalate (SO) were purchased from Shandong Xi Ya Chemical Industry Co., Ltd. RhB and CIP were purchased from Shanghai Macklin Biochemical Co., Ltd. P25, the typical commercial TiO2, was purchased from Aladdin Industrial Corporation, Shanghai, China. All reagents used were of analytical grade.
The BiOCl-Bi12O17Cl2 catalyst was prepared using an ultrasonication-assisted hydrothermal method. In a typical synthesis process, 1.94 g of Bi(NO3)3·5H2O (4 mmol) was added to 60 ml of aqueous solution with 0.298 g KCl (4 mmol) and stirred for 30 min to get a white suspension. Then, an ultrasonic treatment by an ultrasonic generator (0.4 W/cm3) (FB-1500, Aoran Technology Ltd, China) was conducted for 5 min, which resulted in a uniformly dispersed suspension. After the ultrasonic treatment, a NaOH aqueous solution (1.0 mol/L) was used to adjust the pH of the suspension to 12.6 along with vigorous stirring. The suspension was ultrasonically treated again for 30 min. After the second ultrasonic treatment, the suspension liquid was poured into a Teflon autoclave (100 ml) and kept at 160 ℃ for 24 h. The precipitate was collected and washed several times with deionized water and ethanol. Finally, the product was dried at 80 ℃ for 6 h. The synthesis process is schematically illustrated in Scheme S1. The, BiOCl and Bi12O17Cl2 were prepared in accordance with previous reports [29, 30].
X-ray diffractometer (XRD) measurements of samples were performed on a D/max-rc diffractometer (Rigaku, Japan) using Cu radiation, and the scanned 2θ range was from 10° to 70°. The microstructure of the catalysts was observed using a Hitachi S-4700 scanning electron microscope (SEM). The high-resolution transmission electron microscopy (HRTEM) was investigated by using a Tecnai G2 20 (FEI Co., Holland) microscope operating at an accelerating voltage of 200 kV, and the sample powder was dispersed into ethanol by sonication and a very dilute suspension was dropped onto a nickel grid. The Brunauer-Emmett-Teller (BET) specific surface area measurement of the sample was made by a 3H-2000PS2 specific surface analysis instrument (BeiShiDe Instrument Technology Co. Ltd., China) using N2 adsorption-desorption experiments. The ultraviolet-visible (UV-vis) diffuse reflection spectra (DRS) were measured on a UV-2700 UV-vis spectrophotometer (Shimadzu Corp., Japan) with BaSO4 as a reference. The samples of X-ray photoelectron spectroscopy (XPS) were measured on an EScalab 250Xi spectrometer (Thermo Fisher Scientific, USA). The photoluminescence (PL) spectra of the samples were measured using a Renishaw RM 1000 fluorescence spectrometer (Renishaw, UK) at room temperature.
The photocatalytic activity of the catalysts was evaluated by the degradation of RhB (20 mg/L) and CIP (10mg/L) under visible-light illumination (a 500 W xenon lamp with a 420 nm cut-off filter; the radiation flux was 15.1 mW/cm2). Typically, a given mass of the sample (ranging from 0.1 to 1.2 g/L) was added to a quartz condensation beaker with 200 mL of contaminant solution. Dark stirring was conducted for 1 hour before illumination to ensure the adsorption-desorption equilibrium. Immediately after that, approximately 3 to 4 mL of the reaction solution was taken out at a specified time, and particles were removed using a 0.22-μm nitrocellulose filter. The residual RhB and CIP concentrations in the solution were measured using a TU-1900 UV-vis spectrometer (Beijing Purkinje General Instrument Co. Ltd.) at 554 and 277 nm, respectively.
For the reuse experiments, after one run of catalytic reaction, the catalyst was separated, washed by deionized water and ethanol, and then dried at 80 ℃. Later, the catalyst was dispersed in another fresh 20 mg/L RhB or 10 mg/L CIP aqueous solution for the next run.
For these experiments, the active species involved in the photocatalytic procedure were examined. The superoxide radical (·O2–), holes (h+), and the hydroxyl radical (·OH) were consumed by adding benzoquinone (BQ, 1 mmol/L), sodium oxalate (SO, 1 mmol/L), and tert-butyl alcohol (TBA, 10 mmol/L), respectively [31, 32]. Except for the addition of a scavenger, the active species was detected under the same conditions as the RhB photodegradation reaction. The test of the reactive radicals trapped with electron spin resonance (ESR) was carried out using a Bruker A300 (Bruker, German). Typically, 10 mg of photocatalyst was dissolved in 500 μL of methanol (DMPO-•O2–), after which 45 μL DMPO was added, followed by ultrasonic dispersion for 10 min.
All experiments were repeated three times and the averaged values were reported.
The crystal structures of BiOCl, Bi12O17Cl2, and the BiOCl-Bi12O17Cl2 composite were investigated using XRD analysis. Fig. 1 shows that all peaks of the BiOCl and Bi12O17Cl2 consisted primarily of BiOCl (JCPDS card number 06-0249) and Bi12O17Cl2 (JCPDS card number 37-0702), respectively. The sample prepared using the ultrasonic-assisted hydrothermal method was well indexed to the phases of both BiOCl and Bi12O17Cl2. No impurity peaks were observed over the BiOCl-Bi12O17Cl2 composite, demonstrating that the composite was synthesized successfully.
The chemical state and composition of the BiOCl-Bi12O17Cl2 composite were studied using XPS analysis. Fig. 2 shows the XPS spectra of the composite. It can be seen that the peaks of the C, Bi, O, and Cl elements could be accurately identified. The high-resolution XPS spectra of the sample for Bi 4f (Fig. 2(b)) shows that the banding energies of Bi 4f5/2 and Bi 4f7/2 were 164.9 and 159.6 eV, respectively, corresponding to the characteristics of Bi3+ [33]. Fig. 2(c) shows the Cl 2p spectra. The Cl 2p1/2and Cl 2p3/2 peaks were situated at 200 and 198.3 eV, respectively, which could be associated with Cl at the monovalent oxidation state [26], indicating that the BiOCl-Bi12O17Cl2 composite did not affect the interlayered Cl− [34]. Fig. 2(d) shows that the O 1s peak was deconvoluted into two peaks at 531.5 and 530.3 eV corresponding to the adsorption oxygen (Oads) and lattice oxygen (Olat), respectively [35, 36]. The peak intensity of the lattice oxygen was much higher than that of the adsorption oxygen, which was different from that with BiOCl and Bi12O17Cl2 [37, 38]. Such a difference might have been caused by the interaction between BiOCl and Bi12O17Cl2 in the composite. This would expedite catalytic reactions, as the lattice oxygen plays an important role in catalytic performance and is directly involved in oxygen transport and catalytic reactions [39].
The morphologies of the as-synthesized samples were examined by SEM. Fig. 3 shows SEM images of Bi12O17Cl2 and the BiOCl-Bi12O17Cl2 composite. As observed, the Bi12O17Cl2 exhibited an irregular agglomeration of particles, and the average thickness was 100 to 200 nm. The nanocomposite showed a sheet flower-like structure with a uniformly ordered and dispersed arrangement, and the average thickness was 20 to 50 nm. These characteristics were significantly different from those of Bi12O17Cl2. This could have been due to the cavitation effect of the ultrasonic waves that may have caused a uniform distribution, prevented agglomeration, and reduced the particle size, which could have given the catalyst a larger specific surface area to promote adsorption and degrade pollutants [40, 41]. It has been reported that during the hydrothermal reaction, the thickness of the composite was less than that of the Bi3O4Cl through the crystallization, dissolution and recrystallization procedures [34]. Therefore, we speculate that the difference of morphology between Bi12O17Cl2 and the BiOCl-Bi12O17Cl2 composite was caused by both a hydrothermal reaction and the ultrasound effects. In addition, the lesser thickness of the nanosheets indicated a relatively high catalytic activity, which could be attributed to the high electron mobility and low electron injection barrier on the surface of the sheet [42].
The prepared samples were further characterized by HRTEM. Fig. 4 shows HRTEM images of Bi12O17Cl2 and the BiOCl-Bi12O17Cl2 nanocomposite. As shown in Fig. 4(a), the slim lattice fringes of Bi12O17Cl2, with intervals of approximately 0.3175 nm, corresponded to the (319) plane of Bi12O17Cl2. The slim lattice fringes, with spacing of 0.3177 and 0.2990 nm in Fig. 5(b), were indexed to the (212) plane of BiOCl and the (315) plane of Bi12O17Cl2, respectively. In order to further determine the phase formation, the corresponding Fast Fourier Transform (FFT) image is provided. The insets of Figs. 5(a) and 5(b) exhibited a clear point array and were indexed to the (319) plane of Bi12O17Cl2, the (212) plane of BiOCl, and the (316) plane of Bi12O17Cl2, respectively. These above results demonstrated that the BiOCl-Bi12O17Cl2 heterojunction was successfully synthesized by introducing ultrasound.
Nitrogen adsorption-desorption were used to test the pore structure of the BiOCl-Bi12O17Cl2 composite, and the results are shown in Fig. 5. The adsorption-desorption isotherm can be classified as Type Ⅳ, and the hysteresis loops of Type H3 reveal the formation of slit-like pores due to the aggregations of the flake particles [43]. The hysteresis loop was at a relative pressure (p/p0) between 0.2 and 1.0, indicating that capillary condensation occurred in the mesoporous structure. The average specific surface area of the BiOCl-Bi12O17Cl2 composite was16.491 m2/g, greater than the reported that of the Bi12O17Cl2 [33, 44, 45]. The pore volume and diameter of the composite were 0.031 ml/g and 2.491 nm, respectively. The higher surface area increased the pollutant adsorption capacity of the catalyst, which could have accelerated the reaction rate.
The UV–vis (DRS) of BiOCl, Bi12O17Cl2, and BiOCl-Bi12O17Cl2 are shown in Fig. 6(a). The maximum absorption wavelengths of BiOCl, Bi12O17Cl2, and BiOCl- Bi12O17Cl2 were approximately 370, 510, and 490 nm, respectively, suggesting their ability to absorb UV light and visible light. The light absorption edge of the composite was between that of BiOCl and Bi12O17Cl2.The band gap energies of these catalysts were calculated using the following equation [46]:
where α, A, Eg, and v are the absorption coefficient, a constant, the band gap energy, and the light frequency, respectively. The value of n (n = 1 or 4) is a constant, depending on the characteristic of the semiconductor transition. As shown in Fig. 6(b), the band gap energy was 3.25, 2.20, and 2.39 eV for BiOCl, Bi12O17Cl2, and the composite respectively, which was similar to previous reports [47, 48]. These results indicate that the BiOCl-Bi12O17Cl2 composite had a matching energy band position, which was conducive to photocatalytic degradation of organic pollutants under visible-light illumination.
The migration, transformation, and separation of charge carriers of photoinduced electron holes can be investigated by PL spectra. Generally, a decrease of PL intensity can result in a decrease of recombination rate [49]. As shown in Fig. 7, the peak of the composite was significantly lower than that of Bi12O17Cl2, indicating that the composite had a lower recombination rate of photogenerated carriers. Thus, the PL results manifested that the photocatalytic activity of the composite should be greater than that of Bi12O17Cl2. Therefore, we achieved the combination of the stability of BiOCl and the high visible-light absorption capacity of Bi12O17Cl2.
To evaluate the BiOCl-Bi12O17Cl2 composite for its adsorption and visible light photocatalytic activity, several catalysts (BiOCl, Bi12O17Cl2, and P25) were compared. As shown in Fig. 8(a), P25, BiOCl, and Bi12O17Cl2 displayed poor adsorption and photocatalytic performance. In contrast, the BiOCl-Bi12O17Cl2 composite showed excellent adsorption: 67.9% of RhB was adsorbed within 60 min, and the composite adsorption performance was approximately 2.10, 14.50, and 8.95 times that of Bi12O17Cl2, BiOCl, and P25, respectively. During the following catalytic photodegradation, the remaining RhB was completely removed after 30 min. As shown in Fig. 8(b), the composite showed the highest visible light activity, and the reaction rate constant was approximately 8.14, 64.66, and 42.63 times that of Bi12O17Cl2, BiOCl, and P25 respectively.
Similarly, as shown in Fig. 8(b), 34.2% of CIP was adsorbed within 60 min, and the composite adsorption performance was approximately 1.22, 4.44, and 8.34 times that of Bi12O17Cl2, BiOCl, and P25, respectively. From Fig. 8(d), we can see that the composite also showed the highest visible light activity, and the reaction rate constant was approximately 4.94, 11.91, and 36.07 times that of Bi12O17Cl2, BiOCl, and P25 respectively. These results showed that the composite had greater adsorption performance and catalytic properties than those of the other catalysts. These results might have been because the ultrasonic method had the important advantage of distributing the active species, which inhibited the aggregation of particles and made the surface of the particles rougher [50]. This would have helped to increase the specific surface area, adsorption capacity, and catalytic degradation ability of the catalyst. In addition, the removal efficiency of RhB by BiOCl-Bi12O17Cl2 was better than that of CIP, and this phenomenon can be explained by the photosensitization effect of the dyes, which could promote the dye's degradation under visible light.
To investigate the effect of the BiOCl-Bi12O17Cl2 dose, we varied the catalyst dose from 0.1 to 1.2 g/L. As shown in Figs. 9 (a, b, e and f), the adsorption performance, removal efficiency, and reaction rate constant of RhB and CIP gradually improved with an increase of the catalyst dose. This was because by adding more catalyst, more active sites were obtained, which increased the adsorption and catalytic efficiency of RhB and CIP. Considering cost and efficiency, 0.5 and 0.8 g/L of catalyst were used in the following experiments for degrading RhB and CIP, respectively. These doses were lower than those in previous reports [44, 51, 52].
The effects of the initial pH value on the degradation of RhB and CIP were further investigated. As shown in Figs. 9(c) and 9(d), the adsorption and photocatalytic performances of the catalyst were enhanced under acidic conditions, where RhB could have been completely removed in 30 min. When the initial pH was 5, the catalyst showed the highest reaction rate constant. However, the RhB could not be adsorbed by the catalyst under alkaline conditions. As shown in Figs. 9(g) and 9(h), when the CIP pH was 7, the catalyst had the highest adsorption and photocatalytic performances. This confirmed that the degradation processes of RhB and CIP were pH-sensitive, as previously reported [53-55]. Thus, the composite catalyst adsorbed and removed RhB well in a range of neutral and strong acid solutions, while it adsorbed and removed CIP well in a range of strong acid and weak base solutions. This has great significance for practical applications.
For future practical applications, the stability of catalyst should be considered. A recycling photodegradation test was performed to test the durability of BiOCl-Bi12O17Cl2, and the results are shown in Fig. 10. The BiOCl-Bi12O17Cl2 catalyst had good reusability, the RhB contaminant was almost completely removed after only four runs, and the CIP contaminant was almost completely removed after only five runs. Therefore, the composite had excellent reusability.
To determine the main active species in the photodegradation of RhB, active species trapping experiments were carried out. As shown in Fig. 11(a), the reaction was not affected by adding tert-butyl alcohol, a scavenger for ·OH. The degradation of RhB was significantly suppressed with the addition of sodium oxalate (a scavenger for h+) and slightly suppressed with the addition of benzoquinone (a scavenger for ·O2–). These results show that photogenerated holes played a leading role in the photocatalytic decomposition of RhB, while ·O2– promoted the photocatalytic decomposition of RhB.
To further confirm the existence of ·O2– radical species, the ESR technique was performed. As shown in Fig. 11b, no ESR signals could be detected in the dark, but the DMPO-·O2– signal could be observed under visible-light irradiation. This result further suggesting that ·O2– are generated in the photocatalytic reaction system.
Based on the above analysis, the possible processes for RhB degradation can be described in Eqs. (2) – (6):
The valence band (VB) and conduction band (CB) energies of BiOCl and Bi12O17Cl2 were evaluated by using Eqs. (7) and (8) [15].
where X, Eg, and Ee are the electronegativity of the semiconductor, the band gap energy, and the energy of free electrons on the hydrogen scale, respectively. Here, X is calculated from the electronegativity of the constituent atoms, and Ee is the energy of free electrons on the hydrogen scale (approximately 4.5 eV). The X values of BiOCl and Bi12O17Cl2 are approximately 6.64 and 6.31 eV. The band gaps of BiOCl and Bi12O17Cl2 were 3.25 and 2.2 eV (Fig. 6(b)). For BiOCl, EVB and ECB were estimated to be 3.77 and 0.52 eV, and those of Bi12O17Cl2 were estimated to be 2.91 and 0.71 eV in this study. The Fermi level of BiOCl was located close to the VB, while the Fermi level of Bi12O17Cl2 was close to the CB [26]. The coupling of BiOCl and Bi12O17Cl2 enabled the Fermi level to shift to the same level [15]; thus, the VB of BiOCl shifted to an energy level lower than that of Bi12O17Cl2 (Fig. 12(a)).
Based on the preceding discussion, a potential photocatalytic mechanism of the BiOCl-Bi12O17Cl2 composite is proposed. As shown in Fig. 12(b), when exposed to visible light, charge carriers in the BiOCl-Bi12O17Cl2 composite were excited. The band gap width of the catalyst and the reactive species can be obtained from Fig. 6(b) and Fig. 11 respectively. Because the CB position of BiOCl was more negative, the photoexcited electron of BiOCl in the CB would have been transferred to the CB of Bi12O17Cl2 [33, 57]. In contrast, the valence band hole (h+) of Bi12O17Cl2 was transferred to the valence band of BiOCl, which had lower energy [26]. The electrons in the CB of Bi12O17Cl2 could further react with oxygen absorbed on the semiconductor surface and produced superoxide radicals (·O2–) [32]. The produced h+ and ·O2– have strong oxidation capacity, so organic pollutants can be oxidized into small molecules: CO2 and H2O.
In summary, a BiOCl-Bi12O17Cl2 nanocomposite was synthesized by a novel one-step ultrasound-assisted hydrothermal method. SEM and BTE results showed that the nanocomposite had good dispersion and a flower-like structure with a high specific surface area and an average thickness of 20 to 50 nm. UV-vis diffuse reflection spectra and PL spectra analyses showed that the nanocomposite had good visible-light absorption, and a much lower photogenerated electron-hole recombination rate than that of Bi12O17Cl2. The catalyst's kinetic constants of RhB (/CIP) were approximately 8.14(/4.94), 64.66(/11.91), and 42.63(/36.07) times those of Bi12O17Cl2, BiOCl, and P25, respectively. Suitable morphology, structure, and photoelectric properties were the reasons for the excellent photocatalytic performance of the nanocomposite photocatalyst. Further, it showed a wide pH range for application and an excellent reusability. Mechanism studies showed that the main reactive species were photogenerated holes and ·O2–. This research developed a green, stable, and high-efficiency visible-light photocatalyst, which is a valuable contribution to the area of BiOCl-based photocatalysts.
2020,
Vol. 41 Issue (3): 464-473 DOI: S1872-2067(19)63474-1 
PDF
2020,
Vol. 41 Issue (3): 464-473 DOI: S1872-2067(19)63474-1 PDF