Abstract
We report a heterojunction photocatalyst by coupling CdS with Bi12O17Cl2 via a facile thermal annealing process. Such heterostructure can promote the charge separation between CdS and Bi12O17Cl2, thereby leading to a considerable improvement on the performance for CO2 reduction into valuable CH4 and phenol conversion under visible light exposure. The CdS/Bi12O17Cl2 heterojunction with mass ratio of 1 : 1 (50% CdS/Bi12O17Cl2) shows the highest CH4 production rate (1.28 μmol·h-1·g-1), which is as 9.8 and 5.8 times as the CH4 generation rate of pure Bi12O17Cl2 and CdS (0.13 and 0.22 μmol·h-1·g-1), respectively. Also, the 50% CdS/Bi12O17Cl2 sample shows the highest activity of phenol conversion with a conversion ratio of 92%, which is as 2.2 times and 1.9 times as the conversion ratio of pristine Bi12O17Cl2 (41%) and CdS (48%), respectively. The present study demonstrates the great potential of the CdS/Bi12O17Cl2 heterojunction for efficient charge separation towards increased photocatalytic CH4 production and phenol conversion.
1. Introduction
Nowadays, photocatalytic technology has garnered numerous attentions for solar fuel production and environmental purification [1–5]. After decades of studies, hundreds of photocatalysts have been reported on hydrogen generation [6, 7], CO2 reduction [8–11], NO removal [12], and organic pollutant degradation [13, 14]. Bismuth oxychlorides refer to compounds with different ratios of Bi : O : Cl (e.g., BiOCl, Bi3O4Cl, Bi12O17Cl2, and Bi24O31Cl10) [15–18], which has become one of the research hotspots of photocatalytic applications in recent years. Among these compounds, Bi12O17Cl2 photocatalyst has aroused great interest of scientists due to its excellent visible light absorption capacity, strong oxidation capacity, and small bandgap energy [19–22]. Zhang et al. prepared a two-dimensional BiOCl/Bi12O17Cl2 heterojunction photocatalyst, which improved the photocatalytic removal efficiency of nitric oxide (NO) [23]. Ma et al. prepared a new two-dimensional layered MoS2/Bi12O17Cl2 photocatalyst, which increased the visible light absorption range and accelerated the degradation rate of Rhodamine B (RhB) [24]. He et al. synthesized a novel three-dimensional flower-like Bi12O17Cl2/β-Bi2O3, which enhanced the photocatalytic activity of 4-tert-butyl phenol under visible light irradiation [25]. Thus, the photocatalytic performance of Bi12O17Cl2 can be effectively improved by forming heterojunction structure with other semiconductors.
CdS is narrow bandgap semiconductor with strong visible light capture capability, which can be used in photocatalytic CO2 reduction, hydrogen generation, and pollutant degradation [26–30]. However, the application of CdS is limited by photocorrosion and rapid recombination of photogenerated electron-hole pairs [31, 32]. Studies have shown that the construction of heterojunction photocatalysts can effectively separate photocarriers and significantly improve the photocatalytic performance; the multicomponent photocatalysts made of different semiconductors become a feasible method to improve the photocatalytic efficiency [33–36]. Herein, we report on the fabrication of a visible-light-driven CdS/Bi12O17Cl2 heterostructure for both CO2 reduction and phenol conversion. The results showed that CdS/Bi12O17Cl2 heterojunction could effectively improve the photocatalytic reduction rate of carbon dioxide and the removal ratio of phenol.
2. Experiments
2.1. Synthesis of Catalysts
Bi12O17Cl2 samples were prepared by a hydrothermal method [37]. In this preparation method, Bi(NO3)3·5H2O (9 mmol) and KCl (1.5 mmol) were dissolved in an appropriate amount of ethylene glycol, respectively. Add the KCl solution to the bismuth nitrate solution slowly. The compound was stirred for 0.5 h and then kept 160°C for 15 h in a Teflon-lined stainless-steel autoclave (100 mL). The acquired precipitation was collected, washed, and dried. And the powder was further heated at 550°C for 1 h in a muff furnace. Finally, the product was ground to yield the yellow Bi12O17Cl2.
In a typical synthesis of CdS [38], cadmium acetate dihydrate (6.4 mmol) and thiourea (32 mmol) were first added to deionized water (70 mL) under magnetic stirring for 0.5 h. The compound was then kept 140°C for 24 h in a Teflon-lined stainless-steel autoclave (100 mL). And the product was collected, washed, and dried to obtain the orange powder.
To prepare the CdS/Bi12O17Cl2 heterojunction catalyst, Bi12O17Cl2 (0.3 mmol) and a certain amount of CdS thus obtained were first mixed in 30 mL deionized water under the ultrasonic treatment for 1 h and then dried in a water bath at 80°C. And the mixture was heated at 300°C for 1 h to gain the CdS/Bi12O17Cl2 sample. The complete process of heterojunction preparation is shown in Figure 1. By changing the dosage of CdS, a series of CdS/Bi12O17Cl2 samples were prepared. According to the mass content of cadmium sulfide, the samples were named as 34% CdS/Bi12O17Cl2, 50% CdS/Bi12O17Cl2, and 60% CdS/Bi12O17Cl2, respectively.

2.2. Characterizations
X-ray diffraction (XRD) patterns were recorded by a Bruker D8 X-ray diffractometer with a Cu Kα radiation source. X-ray photoelectron spectroscopy (XPS) and valence XPS spectra were carried out on a photoelectron spectrometer (VG ESCALAB 210). The microstructure was analyzed by transmission electron microscopy (TEM, JEM-2100F) and high-resolution transmission electron microscopy (HRTEM, JEM-2100F). Ultraviolet-visible (UV-vis) absorption spectra were collected with a UV-vis spectrophotometer (TU-1901). The specific surface areas were measured on a Micromeritics 3Flex apparatus by nitrogen adsorption–desorption isotherms at 77 K.
2.3. Photoelectrochemical Test
An electrochemical workstation (CHI660C) with a three-electrode system was used to examine the photoelectrochemical properties under visible light irradiation. Platinum was chosen as the counter electrode, the saturated calomel electrode was used as the reference electrode, and the working electrode was the thin film formed by coating the sample on ITO glasses. 5 mg sample was ultrasonically dispersed in N, N-dimethylformamide, and the mixture was then uniformly applied to ITO glass with an area of approximately 1 cm2 and dried. The working electrode was ready to be completed. A 300 W Xe lamp was chosen as the light source, and Na2SO4 solution (0.1 mol· L-1) was selected for ion transport.
2.4. Photocatalytic Performance Measurements
The photocatalytic CO2 reduction was carried out in an air-closed reactor (100 mL), which is equipped with a quartz window and two side sampling ports. A 300 W xenon lamp ( nm) was chosen as the light source. High-purity CO2 was pumped into the reactor for 30 minutes to remove O2. The sample (20 mg) with a Pt load of 0.5 wt% was scattered on a 4 cm2 glass substrate. Then, the glass containing samples was placed into the reactor. Ultrapure water (0.2 mL) was injected into the reaction system. Gas product from the reactor was fed into a gas chromatograph (GC-7890A, Agilent) at an interval of 1 h. The gas phase product was firstly separated in a carbon molecular sieve TDX-01 column and then entered the flame ionization detector to detect the content of methane. Qualitative analysis was carried out according to the retention time of the peak on the chromatographic outflow curve, and quantitative analysis was carried out according to the area of the peak.
The photocatalytic conversion of phenol was evaluated under the irradiation of a 300 W Xe lamp ( nm). Firstly, catalysts with mass of 50 mg and 10 mg·L-1 phenol solution were stirred in the dark for 0.5 h. The light source was then turned on, and 5 mL of the reaction solution was taken every 15 minutes for analysis. A UV-vis spectrophotometer (TU-1901) was then used to determine the concentration of phenol solution.
3. Results and Discussion
3.1. Characterization Analysis
The phase composition of Bi12O17Cl2, CdS, and CdS/Bi12O17Cl2 samples was characterized by XRD (Figure 2). All diffraction peaks shown in Figures 2(a) and 2(e) were well matched with those of tetragonal Bi12O17Cl2 (JCPDS No. 37-0702) [39] and CdS (JCPDS No. 77-2306) [38], respectively. Noticeably, no phase changes of CdS and Bi12O17Cl2 and no indication of any new phase generation were observed after a thermal annealing procedure. For CdS/Bi12O17Cl2, peaks corresponding to (100), (002), (101), (110), and (112) facets of CdS are clearly shown in Figure 2, and the peak intensity increased gradually with the increase of CdS content. However, the peak strength of (103) facet was very weak.

The surface chemical composition of Bi12O17Cl2, CdS, and 50% CdS/Bi12O17Cl2 samples was measured by XPS. The binding energies of Bi 4f (Figure 3(a)) of the 50% CdS/Bi12O17Cl2 (158.9 and 164.2 eV) were shifted slightly to high value than those of pure Bi12O17Cl2 (158.6 and 163.9 eV) [39]. For pure Bi12O17Cl2, the XPS spectrum of O 1s (Figure 3(b)) could be divided into two peaks with binding energies of 529.2 eV and 530.9 eV, corresponding to the Bi-O bond and the O-H bond of H2O that was adsorbed on the surface, respectively [37]. And the binding energy of O 1s XPS spectrum in the 50% CdS/Bi12O17Cl2 sample (529.6 and 531.2 eV) was higher than that in pure Bi12O17Cl2. Furthermore, the binding energies of Cl 2p (197.9 and 199.5 eV) for the 50% CdS/Bi12O17Cl2 sample (Figure 3(c)) were shifted slightly to high value than those of pure Bi12O17Cl2 (197.6 and 199.1 eV) [40]. For pure CdS, the binding energies of S 2p (Figure 3(d)) were 161.1 and 162.3 eV, and the binding energies of Cd 3d (Figure 3(e)) were 404.4 and 411.2 eV [41]. In the 50% CdS/Bi12O17Cl2 sample, the binding energies of S 2p (160.1 and 161.3 eV) and Cd 3d (403.5 and 410.2 eV) were shifted to a lower level. The shift of binding energy is resultant from the strong interaction between Bi12O17Cl2 and CdS [42, 43], which is in accordance with the HRTEM results as discussed below.

(a)

(b)

(c)

(d)

(e)
The intimate contact between CdS and Bi12O17Cl2 component is highly essential for charge transfer across the interfaces of CdS/Bi12O17Cl2. TEM and HRTEM were carried out to visualize such contact in the samples. Figures 4(a)–4(c) showed the TEM image of Bi12O17Cl2, CdS, and 50% CdS/Bi12O17Cl2, respectively. It can be observed that Bi12O17Cl2 was an irregular block-shaped structure accumulated by nanometer slices (Figure 4(a)). In contrast, the CdS particles thus obtained were of typical microspheres with a diameter of ~250 nm (Figure 4(b)). HRTEM image in Figure 4(f) illustrated the clear lattice streaks, the spacing of 0.585 nm corresponded to (006) facet of Bi12O17Cl2, and the spacing of 0.336 nm corresponded to (002) facet of CdS [23, 38]. This result was well consistent with the HRTEM results of pristine Bi12O17Cl2 (Figure 4(d)) and CdS (Figure 4(e)). Thus, a heterogeneous interface was formed between CdS and Bi12O17Cl2 which was conducive to charge transfer.

(a)

(b)

(c)

(d)

(e)

(f)
The Raman spectra were measured to research the chemical bonding information of Bi12O17Cl2, CdS, and 50% CdS/Bi12O17Cl2 (Figure 5). For the Bi12O17Cl2 phase, the bands situated at about 96 and 165 cm−1 could be attributed to the and modes of external Bi-Cl stretching patterns [43]. For the CdS phase, two peaks at 298 and 595 cm−1 were attributed to the first-order and second-order longitudinal optical modes, respectively [44]. In the pattern of 50% CdS/Bi12O17Cl2, both the characteristic peaks of Bi12O17Cl2 and CdS could be observed. In addition, the peak at 165 cm−1 was obviously shifted to the right, indicating that CdS affected the Bi-Cl stretching patterns of Bi12O17Cl2, which could prove the existence of bonds between CdS and Bi12O17Cl2 [37].

3.2. Light Absorption and Band Structure Analysis
Figure 6 shows the light absorption patterns and bandgap energies of the samples. The UV-vis absorption spectra (Figure 6(a)) of CdS/Bi12O17Cl2 heterostructures mainly exhibited the absorption feature of CdS owing to the relatively high content (34%-60%) of CdS in the heterostructures. Interestingly, CdS/Bi12O17Cl2 heterostructures showed an absorption edge at ~550 nm, although the CdS content in CdS/Bi12O17Cl2 heterostructures increased. This result indicated that the light would be mainly absorbed by CdS during the photocatalytic reactions. The bandgap energies of these samples were calculated following the formula where , , , , and represent the light absorption coefficient, Planck constant, frequency, bandgap energy, and constant, respectively [45]. Bi12O17Cl2 is an indirect bandgap semiconductor, while CdS belongs to direct bandgap semiconductors, so their values are 4 and 1, respectively [46–48]. As shown in Figures 6(b) and 6(c), the bandgaps of CdS, 60% CdS/Bi12O17Cl2, 50% CdS/Bi12O17Cl2, 34% CdS/Bi12O17Cl2, and Bi12O17Cl2 were 2.26, 2.29, 2.32, 2.35, and 2.33 eV, respectively.

(a)

(b)

(c)
3.3. BET-BJH
The N2 adsorption-desorption isotherms and pore size distribution of all samples are shown in Figure 7. The specific surface areas were measured to be 4, 20, 13, 22, and 16 m2·g-1, and the pore sizes were 11.7, 3.8, 14.6, 16.1, and 16.0 nm for Bi12O17Cl2, CdS, 34% CdS/Bi12O17Cl2, 50% CdS/Bi12O17Cl2, and 60% CdS/Bi12O17Cl2, respectively. It can be observed that the specific surface area and pore size of CdS/Bi12O17Cl2 samples increased at first and then decreased with the increase of CdS content.

(a)

(b)
3.4. Photocatalytic Activity and Stability
We compared the photocatalytic CO2 reduction activity of samples under visible light exposure, revealing the contribution of the CdS/Bi12O17Cl2 heterojunction (Figure 8). Bi12O17Cl2 and CdS offered a very low activity for CH4 generation, with a rate of 0.13 and 0.22 μmol·h-1·g-1, respectively. Very encouraging, CdS/Bi12O17Cl2 heterojunctions resulted in a highly increased performance for CH4 generation. The CH4 generation rate for 34% CdS/Bi12O17Cl2 (0.65 μmol·h-1·g-1) was as 5.0 and 3.0 times as the CH4 generation rate of pristine Bi12O17Cl2 and CdS, respectively. The highest CH4 generation rate was 1.28 μmol·h-1·g-1 on the 50% CdS/Bi12O17Cl2 sample. In comparison, a physical mixture of CdS and Bi12O17Cl2 with a mass ratio of 1 : 1 only had a CH4 generation rate of 0.29 μmol·h-1·g-1. The above results validated that 50% CdS/Bi12O17Cl2 heterostructure could efficiently separate the photon-generated carriers in CdS and Bi12O17Cl2 and thus lead to enhanced CH4 generation. Further increasing of CdS content led to a decreased CH4 generation rate of 1.02 μmol·h-1·g-1 (60% CdS/Bi12O17Cl2). The photocatalytic performance of CdS was better than that of Bi12O17Cl2, which may be due to the larger specific surface area, better visible light absorption performance, and smaller bandgaps of CdS than Bi12O17Cl2.

Furthermore, we tested the photocatalytic conversion of phenol on CdS/Bi12O17Cl2 samples since the band energy positions of CdS and Bi12O17Cl2 were suitable for oxidation reactions as discussed below. As shown in Figure 9, phenol was hard to be converted under visible light irradiation as exemplified by the blank experiment. After 2.5 h photocatalytic reaction, the conversion ratios of phenol for Bi12O17Cl2, CdS, 34% CdS/Bi12O17Cl2, 50% CdS/Bi12O17Cl2, 60% CdS/Bi12O17Cl2, and mechanical mixing 50% CdS/Bi12O17Cl2 were about 41%, 48%, 69%, 92%, 81%, and 59%, respectively. Also, Bi12O17Cl2 and CdS exhibited weak activity for phenol conversion, as occurred for CH4 generation. However, upon coupling CdS with Bi12O17Cl2, an evidently increased rate for phenol conversion was realized. Phenol could be almost totally converted on 50% CdS/Bi12O17Cl2 heterostructure in less than 3 hours. By contrast, the simple mixture of CdS and Bi12O17Cl2 at this weight ratio only converted about 59% phenol under the same condition. As shown in Figure 10, the photodegradation process was consistent with the pseudo-first-order kinetic model, as expressed by the following equation: where represents the apparent rate constant calculated from the pseudo-first-order reaction kinetics model and and are the concentration at time of and zero, respectively [33]. As shown in Figure 10, the apparent rate constants for Bi12O17Cl2, CdS, 34% CdS/Bi12O17Cl2, 50% CdS/Bi12O17Cl2, 60% CdS/Bi12O17Cl2, and mechanical mixing 50% CdS/Bi12O17Cl2 were 0.0036, 0.0043, 0.0078, 0.0151, 0.0116, and 0.0059 min-1, respectively. Obviously, 50% CdS/Bi12O17Cl2 revealed the highest photocatalytic activity for phenol conversion. In addition, the removal ratio of chemical oxygen demand (COD) of phenol solution (10 mg·L-1, 50 mL) with 50 mg 50% CdS/Bi12O17Cl2 reached 30% after 4 h illumination (a 300 W xenon lamp was used as the light source).


The above results clearly demonstrated that the thermal annealing could ensure activity for phenol conversion. Moreover, 50% CdS/Bi12O17Cl2 heterostructure exhibited the highly stable activity for both CH4 generation and phenol conversion. Only little change was observed on the reaction rates for CH4 generation and phenol conversion in three consecutive experiments, as shown in Figure 11, illustrating the excellent durability of 50% CdS/Bi12O17Cl2 for long-term reaction.

3.5. Photocatalytic Mechanism Analysis
The active substances for the conversion of phenol on CdS/Bi12O17Cl2 were studied by free radical capture experiments, in which ammonium oxalate monohydrate (AO) was selected as the trapping agent for holes, isopropanol (IPA) for ·OH, and benzoquinone (BQ) for ·O2−, respectively. As illustrated in Figure 12, the addition of BQ and AO could greatly hinder the phenol conversion under visible light exposure, indicating that ·O2- and h+ are two prime active substances for phenol conversion. In comparison, only a slight decrease of phenol conversion occurred when IPA was added to the reaction suspension, thereby excluding ·OH as the active substance in present case.

The increased photocatalytic activity for CH4 generation and phenol conversion could be attributed to the enhanced interfacial charge transfer between CdS and Bi12O17Cl2. This efficient charge separation was verified by the photoelectrochemical tests. As illustrated in Figure 13(a), the 50% CdS/Bi12O17Cl2 sample provided the highest transient photocurrent compared to pristine CdS and Bi12O17Cl2, in spite of relatively low value. In addition, the low interface impedance in the 50% CdS/Bi12O17Cl2 sample was also favorable for charge transfer, as exemplified by the small arc radius shown in Figure 13(b).

(a)

(b)
The band energy positions for CdS and Bi12O17Cl2 were determined by the XPS valence spectra. As shown in Figure 14, the maximum valence band (VB) value of Bi12O17Cl2 was 1.42 eV, while that of CdS was 1.19 eV. The bandgaps of pristine Bi12O17Cl2 and CdS were 2.33 and 2.26 eV, respectively. Therefore, the conduction band (CB) minimum was -1.07 eV for CdS and it was -0.91 eV for Bi12O17Cl2. The electrons in the VB are excited by light to transition to the CB, as shown in Figure 15. Since the CB of Bi12O17Cl2 is lower than that of CdS, the photoexcited electrons of CdS would transfer to Bi12O17Cl2, and the hole would transfer reversely, namely, from Bi12O17Cl2 to CdS. Such charge transfer can efficiently separate the electron-hole pair, thus enhancing the photocatalytic property for CH4 generation and phenol conversion. The electrons accumulated in Bi12O17Cl2 would reduce CO2 and H2O to CH4. Meanwhile, photoexcited electrons can also react with oxygen to produce superoxide radical ·O2-. Superoxide radicals and holes have strong oxidation ability, which can convert phenol into small molecules such as H2O and CO2.


4. Conclusions
In summary, the CdS/Bi12O17Cl2 heterojunction has been successfully synthesized by means of thermal annealing. The results of photocatalytic activity experiments showed that CdS/Bi12O17Cl2 heterojunctions could considerably improve the performance for CO2 reduction to CH4 and phenol conversion under visible light exposure. In addition, 50% CdS/Bi12O17Cl2 showed the highest photocatalytic activity with a CH4 generation of 1.28 μmol·h-1·g-1 and phenol removal ratio of 92% after 2.5 h of light irradiation. The activity enhancement is resultant from the effective division of photogenerated electron-hole pairs across the CdS/Bi12O17Cl2 interfaces. The study offers a new insight on exploring effective heterostructure photocatalysts for fuel manufacture and organic pollutant conversion.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Authors’ Contributions
Dandan Hu and Yumin Cui contributed equally to this work.
Acknowledgments
Thanks are due to the support of National Natural Science Foundation in China (21807012), Natural Science Foundation of Anhui Province in China (gxgwfx2018059, KJ2020ZD47), Key Research and Development Programs of Anhui Province in 2021 (202104i07020011), Natural Science Research Projects of Fuyang Normal University of China (2017FSKJ09), Horizontal Cooperation Project of Fuyang municipal government and Fuyang Normal University (XDHX2016002, XDHX201711, XDHXPT201702, and XDHX201716), Scientific Research Innovation Team of Fuyang Normal University (kytd201707), and Innovative training program for College Students (201910371021, S201910371022, and S201910371022) in China. We would also like to thank the university engineering technology research center of biomass conversion and antipollution control of Anhui province platform for supporting this work.