The industrial production processes such as metal smelting, chemical industries, production of pigment, medicine and paper have given rise to heavy metal pollution which are harmful to human beings and other creatures in nature. Among the toxic heavy metal pollutants, the threaten to human being and ecological systems caused by hexavalent chromium (Cr(VI)) has attracted more and more attention since Cr(VI) has high solubility in aqueous solution, substantiated toxicity and potential carcinogenicity. Therefore, it is highly necessary to solve the Cr(VI) presence in wastewater^[1,2]. Since the conventional approaches bring out the problems such as secondary pollution, onerous procedures and high energy consumption, the reduction of Cr(VI) to non-toxic Cr(III) which can be realized by semiconductor has attracted increasing attention with the advantages of low cost, environmental safety and high efficient utilization of solar energy^[3,4]. Semiconductor photocatalytic method for the reduction of Cr(VI) to Cr(III) is ascribed to electron-hole pairs generated by energetic photons which lead to the photo redox process^[5]. TiO₂ photocatalyst or TiO₂ based composites have received great attention because of high photocatalytic activity, favorable chemical stability and low-cost. However, its low photo-responsivity in the visible light hinders its practical utilization efficiency. Therefore, a large number of visible light response photocatalysts is designed and synthesized to improve the photocatalytic efficiency^[6,7].

Among various photocatalysts, Bi-compounds based photocatalysts, such as Bi₂O₃, BiVO₄, Bi₂MoO₆, BiOX (X=Cl, Br, I) and Bi₂O₂CO₃ have been extensively reported with satisfactory photocatalytic performance for the degradation of organic pollutants and removal of Cr(VI) reduction^[8,9,10]. As a member of bismuth family, Bi₅O₇I is currently discovered as visible-light-driven semiconductor with its optical band gap of about 2.8 eV. The characteristic layered crystal structure of Bi₅O₇I from the [Bi₂O₂]²⁺ and double I layers could induce an internal static electric field which is perpendicular to each layer and subsequently beneficial for the separation of photo- generated carriers^[11]. However, Bi₅O₇I suffered from the insufficient light absorption and fast recombination of electron-hole pairs which considerably restricted its extended practical applications in photocatalysis^[12,13]. For practical application, many efforts have been devoted for improving photocatalytic property of Bi₅O₇I, containing designing metal-semiconductor composites, constructing semiconductor composites with proper band gap semiconductors or high specific surface area, such as g-C₃N₄/ Bi₅O₇I, Ag₂O/Bi₅O₇I, AgI/Bi₅O₇I^[14,15,16].

In addition, CuS, as a p-type photocatalyst with narrow band gap exhibits broad absorption spectra in the range from ultraviolet (UV) to near-infrared (NIR) region, has been extensively used as an efficient photocatalyst. For example, Lai, et al. ^[17] reported that CuS decorated BiVO₄ to enhance its photocatalytic property in ciprofloxacin degradation. Naghizadeh, et al. ^[18] reported that the CoFe₂O₄@CuS nanocomposite shows enhanced photocatalytic degradation of penicillin G antibiotic. Ghim, et al. ^[19] prepared ZnMoS₄/CuS p-n heterojunctions which exhibited efficient photocatalytic hydrogen generation. These results indicate that CuS could be a favorable co-catalyst for Bi₅O₇I photocatalyst. Yet there are little reports on the fabrication and photocatalytic reduction activity of Cr(VI) under visible light of CuS/Bi₅O₇I composite.

In this work, novel CuS/Bi₅O₇I composite was fabricated by a facile hydrothermal method. Bi₅O₇I microrods were decorated by CuS nanosheets, which was in favor of convenient and effective regulation of Bi₅O₇I-based composite’s photocatalytic properties. The Cr(VI) aqueous solution is employed as the target degradation contamination to evaluate the photocatalytic performance of CuS/Bi₅O₇I. The photocatalytic experiments results manifested the improved photocatalytic reduction efficiency of CuS/Bi₅O₇I composites for aqueous Cr(VI) under visible light irradiation. Moreover, the possible mechanism of photocatalytic reduction of Cr(VI) by CuS/Bi₅O₇I composites was also explored.

Briefly, Bi(NO₃)·5H₂O (4.85 g) was dissolved in deionized water (200 mL), ultrasonically dispersed for 30 min and kept stirring for 60 min and then 5 mL of acetic acid was added to obtain suspension A.

KI (1.66 g) was dissolved in deionized water (10 mL) to obtain homogeneous solution which was added to the suspension A dropwise and kept magnetic stirring for 30 min to obtain a black suspension B. After that, 4 mL ammonia water was added dropwise to suspension C until the brick red precipitation appeared. After magnetic stirring at 85 ℃ for 60 min, the BiOI red precipitate was obtained by centrifugation, and fully dried at 70 ℃. The BiOI red powder was then ultrasonically dispersed in a mixture solvent of ethanol and deionized water (25 mL/25 mL) for 30 min to obtain a suspension. Next, NaOH solution (25 mL 0.2 mmol/L) was added to the above suspension with vigorous stirring at 70 ℃ for 60 min. Finally, the white Bi₅O₇I sample was washed by deionized water and ethanol for several times and dried at 70 ℃ overnight.

CuS decorated Bi₅O₇I with different amounts of CuS was prepared as following: CuSO₄·5H₂O, thiourea and trisodium citrate dihydrate was dissolved in deionized water (30 mL). After stirring for 20 min, Bi₅O₇I powder were ultrasonically dispersed in the above solution for 30 min. The suspension was then transferred into a 50 mL Teflon-lined stainless autoclave and reacted at 160 ℃ for 10 h. Finally, the CuS/Bi₅O₇I composite was collected and washed by deionized water, and then dried at 60 ℃ for 10 h. CuS/Bi₅O₇I composites were denoted as CB-x, where x is the mass percentage (wt%) of CuS, and named as CB-10, CB-15 and CB-20, respectively. The fabrication process of CuS/Bi₅O₇I composites was presented in Scheme 1. By contrast, pure CuS were also prepared by the procedure described above without the addition of Bi₅O₇I.

The phase structure of the samples was determined by the Bruker D8 Advance powder X-ray diffractometer with Cu-Kα irradiation (λ=0.154056 nm). The morphologies and microstructural properties were obtained by field emission scanning electron microscope (FESEM, JSM-7100F, JEOL) and high-resolution transmission electron microscope (HRTEM, JEM 2010F). UV-Vis diffuse reflectance spectra were carried out on a UV-Vis spectrophotometer (UV-3600, Shimadzu) with an integrating sphere attachment. QDS- MP-30 volumetric gas adsorption instrument (Quantachrome, USA) was employed to measure the Brunauer-Emmett- Teller (BET) surface area. X-ray photoelectron spectroscopy (XPS) on an ESCALAB 250 xi apparatus was conducted to reveal the elemental composition and surface states of CuS/Bi₅O₇I composite. The photoluminescence (PL) spectra were determined by a fluorescence spectrophotometer (PerkinElmer, LS-55) at room temperature. A Zahner CIMPS electrochemical workstation (Germany) was employed to perform the electrochemical impedance spectra (EIS) measurement in a three-electrode system with Pt wire and Ag/AgCl electrode as the counter and reference electrodes, respectively. Pure Bi₅O₇I, CuS or CB-15 composites spin-coated ITO with the size of 1 cm×1 cm was used as working electrode. The measurement was carried out in 0.2 mol/L Na₂SO₄ aqueous solution.

The photocatalytic properties of the as-synthesized photocatalysts were accessed for the reduction of Cr(VI) to Cr(III) and performed in aqueous solution with concentration of 10 mg/L. The pH of the mixture solution should be controlled at about 3 with H₂SO₄ solutions. In the photocatalytic activity evaluation experiment, 100 mg sample was added to 150 mL K₂Cr₂O₇ aqueous solutions and kept magnetic stirring for 30 min to reach adsorption/desorption equilibrium. Then, the illumination from a 300 W Xe lamp equipped with 420 nm cutoff filter to make sure the photocatalytic reduction reaction take place under the visible-light irradiation. During the irradiation, 3 mL suspensions were sucked out at given time intervals. After removing the solid photocatalyst by centrifuging and filtration, the concentration of Cr(VI) in the filtrate was measured according to the diphenylcarbazide colorimetric method determined by Shimadzu UV-3600 spectrophotometer. The photocatalytic reduction activity was evaluated by the time-depending C/C₀, where C is the remained Cr(VI) concentration and C₀ is the initial Cr(VI) concentration.

X-ray diffraction patterns of as-prepared Bi₅O₇I, CuS and CuS/Bi₅O₇I composites are presented in Fig. 1. It can be seen that XRD peaks at 2θ=27.9°, 28.9°, 30.9°, 33°, 45.9°, 53.6° and 56.1° correspond to (312), (113), (004), (204), (604), (224), (316) and (912) crystal planes of orthorhombic Bi₅O₇I, respectively. The synthesized Bi₅O₇I is well-crystallized, and all of the peaks are indexed to the standard card (JCPDS 40-0548) without other impurity phase^[15]. The diffractions for pure CuS almost exactly match the standard cards JCPDS 06-0464^[17]. For CuS/Bi₅O₇I composites, no characteristic diffraction peaks of CuS were detected when the content of CuS is lower than 15wt%. However, the character peak of CuS at 2θ=31.8° could be observed in the CB-20 composite.

SEM images of the pure Bi₅O₇I and CuS/Bi₅O₇I composites are presented in Fig. 2. It shows that the synthesized Bi₅O₇I exhibits a microrod-like morphology with length of about 3-4 μm and diameters of 200-300 nm. The surface of Bi₅O₇I microrods are smooth which can provide a favorable environment for assembling the co-photocatalyst. After introducing CuS, Bi₅O₇I microrods are covered by thin CuS nanosheets. The CuS nanosheets grow discretely along the axis of Bi₅O₇I microrods. In addition, the nanosheets decorated on Bi₅O₇I microrods become denser with the increasing of CuS mass percentage. Furthermore, the TEM image of CB-15 composite in Fig. 3(a) confirms the SEM results, which illustrates that the surface of Bi₅O₇I microrods is fully and densely covered by irregular shaped CuS nanosheets with strong linkages in the composite. By analyzing the HRTEM images in Fig. 3(b), the lattice spacing of 0.319 nm corresponds to the (312) plane of orthorhombic Bi₅O₇I, and the well crystalline lattice fringe with interplanar spacing of 0.304 nm can be ascribed to (102) plane of CuS, respectively. The interface between the Bi₅O₇I and CuS lattices clearly indicates the formation of CuS/Bi₅O₇I heterojunctions, which is beneficial for the effective separation of photogenerated electron-hole pairs, and thus enhances the photocatalytic activity^[20,21].

Chemical states and elemental compositions of CB-15 composite were estimated by XPS spectra. It can be observed in Fig. 4(a), the general survey spectrum reveals that the as-prepared CB-15 composite consists of Bi, I, O, Cu and S elements. The signals of C1s located at about 285.3 eV was used for calibration. Fig. 4(b) shows the Bi4f spectrum of CB-15 composite with two peaks at 164.3 and 159.06 eV, which correspond to Bi4f_5/2 and Bi4f_7/2, respectively, demonstrating the characteristics of Bi³⁺ ions in the Bi₅O₇I^[22]. The high-resolution spectra of I3d show two evident peaks at 618.8 and 630.4 eV (Fig. 4(c)), which can be attributed to I3d_5/2 and I3d_3/2, ensuring the monovalent oxidation state of iodine. The O1s spectrum (Fig. 4(d)) was fitted into three peaks centered at 529.6, 531.0 and 532.4 eV, which may be assigned to Bi-O, O-H, and adsorbed H₂O, respectively^[23]. In addition, the Cu2p_3/2 and Cu2p_1/2 peaks are located at 932.3 and 952.4 eV (Fig. 4(e)), which are characteristic of the Cu²⁺ in the CuS material^[24]. As for S2s (Fig. 4(f)), a broad peak at 225.0 eV can be attributed to the metal sulfide^[25].

The optical absorption properties of the prepared samples were investigated by UV-Vis diffuse reflectance spectrometer and presented in Fig. 5. Fig. 5(a) indicated that pure Bi₅O₇I microrods has the absorption edge of about 400 nm with a band gap of 2.92 eV calculated based on Kubelka-Munk theory (Fig. 5(b))^[17]. The absorption edge of pure CuS is at approximately 740 nm, and the corresponding band gap energy is 1.68 eV. Compared with pure Bi₅O₇I microrods, light absorption of CuS/Bi₅O₇I composites expands to the range of visible light. Moreover, the absorption intensity increased after CuS decorated on Bi₅O₇I microrods. The results of diffuse reflectance spectrometer denote that the decoration of CuS nanosheet on Bi₅O₇I microrods is helpful for the improved visible-light harvesting and photocatalytic activity of CuS/Bi₅O₇I composites^[26].

Nitrogen adsorption-desorption isotherms analysis was conducted to investigate the specific surface areas of Bi₅O₇I and CB-15 sample. As shown in Fig. 5(c), the curves of the samples displayed a typical IV feature with a hysteresis loop extended from p/p₀ = 0.5 to p/p₀ = 1 which indicated mesoporous structures of the samples. The BET surface area of CB-15 is 8.77 m²/g, which is 2.46 times of that of pure Bi₅O₇I (3.56 m²/g) because the CuS nanosheet decorated Bi₅O₇I microrods have more stretched surfaces compared with Bi₅O₇I microrods. The higher surface area of CB-15 composite is usually closely related to more adsorption sites exposed under irradiation and increased adsorption capability which is, therefore, in favor of the enhanced photocatalytic activity.

Photoluminescence (PL) spectra and electrochemical impedance spectroscopy were further employed to analyze the separation efficiency and recombination rate of photoexcited electrons and holes of Bi₅O₇I and CuS/Bi₅O₇I composites. It can be observed from Fig. 6(a) that the PL spectra of the as-prepared photocatalysts exhibit an obvious emission peak at about 420 nm. The CuS/Bi₅O₇I composites exhibit obviously lower intensity than pure Bi₅O₇I, suggesting that the decoration of CuS nanosheets on Bi₅O₇I microrods can depress the recombination of photogenerated electrons and holes effectively ^[27]. The faster charge separation would result in increased lifetime of charge carriers and enhanced transfer efficiency of the interfacial charge. Moreover, CB-15 composite shows the lowest intensity, which is consistent with the result of photocatalytic experiment.

Electrochemical impedance spectroscopy (EIS) was conducted to testify the charge migration and recommendation process of the photo-induced carriers in CuS/Bi₅O₇I and at the interface of the composite. The EIS Nyquist plots of Bi₅O₇I and CB-15 composite as illustrated in Fig. 6(b) show that CB-15 composite possesses smaller arc radius. Generally, the smaller radius implies a lower charge transfer resistance^[21], which suggests that Bi₅O₇I coupled with a certain amount of CuS is in favor of the separation of photo-generated charge carriers and the interfacial charge transportation^[11,28]. Therefore, the EIS results demonstrate that Bi₅O₇I coupled with a certain amount of CuS is in favor of the separation of photo-generated charge carriers and the interfacial charge transportation.

On the basis of the analysis from PL spectra and EIS, the improved photocatalytic reduction of Cr(VI) by the CuS/Bi₅O₇I composite can be attributed to the synergism of the greater surface areas, broader light adsorption and higher separation and transfer efficiency of charge carrier with the CuS decoration on Bi₅O₇I microrods.

Cr(VI) aqueous solutions were used to assess the photocatalytic activity of the CuS/Bi₅O₇I composites, as well as pure Bi₅O₇I and pure CuS. The experiments were carried out under visible light illumination and the results are shown in Fig. 7(a). As can be observed that pure Bi₅O₇I shows low photocatalytic activity with only 12% efficiency after 4 h of visible-light irradiation which can be attributed to the weak absorption in the visible light range. The photoreduction efficiency of Cr(VI) is about 45% in the presence of pure CuS. Under the identical condition, the removal efficiency of Cr(VI) by CuS/Bi₅O₇I was noticeably enhanced, which could be attributed to the higher separation efficiency of photo-excited electrons and holes for the CuS/Bi₅O₇I. On the other hand, it can be obviously found that the photocatalytic activities of CuS/Bi₅O₇I are dependent on CuS content. With the CuS content increasing from 10wt% to 15wt%, the photocatalytic activity reached the maximum with efficiency of about 92%. However, photocatalytic activity decreased with the CuS content further increasing to 20wt%, which indicated that the proper CuS proportion was conducive to the transfer of electrons and holes at interfaces^[29]. The photocatalytic reduction kinetic is generally applied to depict the process, and the results are plotted in Fig. 7(b). The photocatalytic reduction process accords to pseudo-first-order kinetics, ln(C₀/C_t)=kt, where C₀ is the initial concentration of the Cr(VI) solution, C_t is the concentration of Cr(VI) at time t, and the slope k is the apparent reaction rate constant. It is apparent than that the apparent reaction rate constants of CuS/Bi₅O₇I composites are greater that of as-prepared Bi₅O₇I. k value of CB-15 is 0.0054 min^-1, which confirmed its best photocatalytic performance.

Based on the experimental results and discussion above, the photo-induced electron-hole pairs separation process during the reduction of Cr(VI) by CuS/Bi₅O₇I composite under visible light irradiation was shown in Fig. 8. In general, upon visible illumination, the electrons of Bi₅O₇I and CuS on the valence band (VB) are excited to conduction band (CB), generating holes on the VB which form large number of valence band.

The CB edge potential of Bi₅O₇I is more negative than that of CuS, and the VB position of CuS is positive than that of Bi₅O₇I. As a result, the photoinduced electrons on the Bi₅O₇I nanosheet surface can jump easily to the conduction band of CuS resulting in the reduction of Cr₂O₇^2- to Cr³⁺. By this means, the photo-excited electrons and holes are separated effectively in the CuS/Bi₅O₇I composites and electron-hole recombination behavior is depressed, leading to higher photocatalytic reduction efficiency of Cr(VI) to Cr(III). Consequently, the as-fabricated CuS/Bi₅O₇I composite photocatalyst shows enhanced photocatalytic performance than Bi₅O₇I and CuS.

In conclusion, CuS nanosheets decorated Bi₅O₇I composites have been successfully synthesized via a facile solution method. CuS/Bi₅O₇I composites exhibit a satisfactory strengthened photocatalytic efficiency for the removal of aqueous Cr(VI) comparing with pure Bi₅O₇I and CuS under visible-light illumination. The photocatalytic reduction rate of Cr(VI) in the presence of CuS/Bi₅O₇I composite with 15wt% CuS is about 20 times of pristine Bi₅O₇I and 3 times of pure CuS. The enhanced photocatalytic activity of CuS/Bi₅O₇I composites is owing to the synergistic effect of the greater surface areas, broader light absorption region, improved charge separation and transfer efficiency of photo-excited electron-hole pairs. This work is helpful for the construction of visible light responsive photocatalytic systems for the practical application involving the treatment of environmental remediation and water purification.