Enhanced photocatalytic activity of AgBr photocatalyst via constructing heterogeneous junctions with reduced graphene

A series of reduced graphene (rGO) and AgBr composite heterojunction photocatalysts were fabricated through a facile solvothermal method. The rGO/AgBr heterostructures were characterized by XPS, XRD, UV–Vis DRS, SEM, TEM, photoluminescence (PL), and the transient photocurrent responses. The XRD, SEM, XPS, and TEM analyses indicated that the graphene and silver bromide were successfully compounded without other impurities. The UV–Vis DRS exhibited that the rGO/AgBr composites have better optical properties than pure silver bromide. The PL and the transient photocurrent responses demonstrated that the addition of graphene significantly promotes the separation of photogenerated electrons and holes. Subsequently, the photocatalytic activities of rGO/AgBr composites were studied by degrading Rhodamine B (RhB). It turned out that the degradation rate of RhB by the rGO/AgBr heterojunction photocatalysts was significantly higher than that by pure AgBr. Furthermore, to study the photocatalytic degradation mechanism of RhB by rGO/AgBr heterostructures, the trapping experiments were used to identify main active components. This work confirmed that the photocatalytic degradation performance of the catalyst was greatly improved after doping graphene, which provided certain data support for degradation of organic contaminants in water.


Introduction
Rapid industrial developments improved human life quantity, whereas the discharge of pollution has become a compelling threat to the environment [1]. In recent years, enhancing the requirements of environmental protection, especially the treatment of sewage, has a considerable impact on sustaining development of human society [2]. The organic dyes, which are widely used in textile, printing, papermaking, and other industries around the world, are colorants with complex structures, high molecular weight, water solubility, degradation resistance, and potential carcinogenicity [3][4][5]. They will damage the aquatic ecosystem and threaten human health if the organic dyes in the wastewater are directly discharged into the rivers, lakes, and groundwater. Therefore, the organic dyes in the wastewater have become a key issue for the pollution of water resources [1,6]. In recent years, the requirements for environmental protection have been even more stringent, and the treatment of sewage has become a hot spot of social concern. Therefore, removing organic dyes in wastewater has become an urgent and significant research direction. Adsorption and chemical coagulation are techniques often used in wastewater treatment. However, these techniques cause secondary pollution [1].
In order to degrade organic pollutants in water more effectively, photocatalysis, the most promising technology, has been studied by more and more researchers [7][8][9]. Among various semiconductor catalysts, titanium dioxide (TiO 2 ) has attracted the attention from researchers and extensive research has been carried out [10,11]. However, TiO 2 has photocatalytic activity only under ultraviolet light, which accounts for about 2-5% of solar energy. Furthermore, the recombination of massive photogenerated electron holes limits its photocatalytic efficiency [12][13][14]. Therefore, the discovery for new effective visible light responsive photocatalysts is indispensable for future applications [7].
Silver halide (AgX, X = Cl, Br, I), which is a photocatalytic material with a narrow band gap, exhibits great photocatalytic performances for organic pollutant removal in the presence of visible light [15][16][17]. However, pure AgX inevitably forms silver atoms under visible light irradiation because its photogenerated electron tends to bond with an interstitial Ag ? by absorbing incident light, and its photocatalytic efficiency will be inhibited by the photodecomposition [18][19][20][21]. As a consequence, the photoinduced stability of pure AgX could be improved by quickly capturing photogenerated electrons before they combine with silver ions to form silver atoms [20,22,23]. Recently, it has been demonstrated that AgX can enhance stability and catalytic activity due to the formation of heterogeneous junctions by loading different cocatalysts [20,24,25]. For example, Chen et al. [26] synthesized AgBr/Ag 3 PO 4 @natural hematite heterojunction photocatalyst to degrade antibiotics under simulated solar light. In addition, Cao et al. [18] synthesized AgBr/WO 3 complex for the removal of methyl orange in visible light. The results indicated that the heterojunctions display outstanding performance than the single component, which may be attributed to the rapid separation of electron-hole pairs as a result of the existence of heterojunction structure [18,27]. Therefore, finding a suitable cocatalyst to form heterojunction with AgX is a research hotspot to improve photocatalytic performance. Recently, graphene oxide (GO) and reduced graphene oxide (rGO) have drawn more and more attention from researchers because of their high specific surface area and superior electron mobility [28][29][30]. Now, the GO can be mass produced at a lower cost, and there were so many researchers who constructed plenty of different types of semiconductor-graphene (SC-graphene) heterojunction materials to improve their photocatalytic performance [16,31,32], such as TiO 2 /GO [33], Cu 2 O/rGO [34], a-Fe 2 O 3 @GO [35], ZnO/rGO [36], and Bi 2 WO 6 /graphene [37]. As expected, the photocatalytic performances of the composites were improved because the graphene in the composites can promote charge separation and inhibit photogenerated electron-hole pair recombination [38].
Herein, a series of rGO/AgBr heterojunction catalysts were processed via a simple solvothermal treatment. The photocatalytic properties of the rGO/ AgBr hybrid materials were estimated by using RhB as a target organic pollutant, and the kinetics of photocatalytic degradation were studied. Furthermore, the photocatalytic mechanism of rGO/AgBr composite was also investigated.

Materials
All reagents were purchased from Kermel, and all of them were of analytical pure (C 99.8%) without further purification.

Synthesis of GO
GO was fabricated by the modified Hummers method. In a 1000 mL flask, 70 mL H 2 SO 4 (98%), 3.0 g graphite powder, and 1.5 g NaNO 3 were added and stirred vigorously in an ice bath. With vigorous stirring and temperature at 0°C-4°C, 9 g KMnO 4 was added uniformly and slowly until the solution turned green. Then magnetically stirred for 40 min under the condition of oil bath (40°C). 150 mL pure water was added to the above solution by inches and heated to 95°C for 15 min before removing the flask from the oil bath. The precipitation obtained by centrifugation was washed with hydrochloric acid (1:10) and pure water, then centrifuged for 30 min after 30 min of ultrasound, and the supernatant was retained and the precipitation (agglomerated GO) was removed, then ultrasonicated and centrifuged (1000 r min -1 ) again, and the solution was divided into three layers. Finally, the precipitation and supernatant were removed, and the intermediate layer was retained. The concentration of GO was 0.007 g mL -1 .

Synthesis of rGO/AgBr
The rGO/AgBr composite photocatalysts were synthesized by a facile solvothermal method and the experimental procedure is illustrated in Scheme 1. First, 0.3397 g AgNO 3 was added to 20 mL ethylene glycol under magnetic stirring conditions. Then GO solution of 0.265, 0.53, 0.802, 1.075 mL was added, respectively, and the design was 0.5, 1, 1.5, and 2% rGO/AgBr, respectively. In addition, the ethylene glycol solution of potassium bromide was added into the above dispersion, in which potassium bromide was excessive to ensure that there were enough halide ions to precipitate silver ions. After 30 min magnetic stirring, the precursor solution was transferred into the Teflon-lined stainless steel autoclave and heated at 160°C for 4 h. The obtained precipitation was washed with absolute ethanol and deionized water to remove ionic residue. Finally, the samples were dried at 60°C and were denoted as rGO-A (A = 0.5, 1, 1.5, and 2.0). The same process was used to prepare pure AgBr without GO.

Characterization
The X-ray diffraction (XRD) patterns were identified by an X-ray diffractometer (D/MAX-r A, Rigaku, Japan). The morphologies of the samples were observed by scanning electron microscopy (SEM, Sigma 500, zeiss, Germany). The chemical compositions and valence state of the composites were investigated by an X-ray photoemission spectroscopy (XPS, Thermo escalab 250Xi). UV-Vis diffuse reflection spectra (UV-Vis DRS) of the samples were performed by using a UV-Vis spectrophotometer (Caly 5000, Agilent, USA). Fourier transform infrared spectra (FT-IR) were detected by FT-IR spectrophotometer (Nicolet iS50, Thermo, USA). The PL spectroscopy was measured at the excitation wavelength of 420 nm on F-4600 Fluorescence Spectrophotometer. Photoelectrochemical measurement was made on an electrochemical analyzer (CHI660E).

Photocatalytic experiments
The photocatalytic activity of the photocatalysts was determined by degradation of RhB solution (0.02 g L -1 ) under 300 W xenon lamp (k C 420 nm) irradiation. In the specific experiment, 0.05 g photocatalyst was added into 50 mL RhB solution and stirred magnetically in the darkness for 20 min to ensure the equilibrium of adsorption and desorption. During the irradiation process, 4 mL suspension was taken out at regular intervals (10 min), and the catalyst was filtered out with 0.22 lm membrane filter, and the filtrate was determined by UV-Vis spectrophotometer (UV-2450 Shimadzu, Japan). Photocatalytic performance can be described by the following equation:

Scheme 1 Preparation of rGO/AgBr composite material
where D is the removal rate, C 0 and C t are the initial concentration of RhB solution and the concentration after t min illumination, respectively.
3 Results and discussions 3.1 Phase structure and morphology of photocatalysts The XRD patterns of GO, AgBr, and rGO/AgBr composites with different rGO contents are shown in Fig. 1a, which showed the crystal phase, purity, and crystallinity of the prepared samples.  [39]. As shown in Fig. 1a, there were no characteristic peaks of GO and rGO in the composites, which may be due to the low content of GO. At the same time, no other impurity peaks were detected in all composite materials, which confirmed the high purity of AgBr and composites. In addition, all photocatalysts had sharp characteristic diffraction peaks, indicating that all catalysts had high crystallinity. To further prove the composition of the materials, FT-IR was used to verify the existence of characteristic chemical bonds, and the FT-IR results proved the existence of rGO in the composite material. As shown in Fig. 1b, the typical peaks at 1400 cm -1 and 1231 cm -1 were corresponding to the C-O and C-O-C bonds. Compared with GO in the literature [40], the oxygen functional groups were fewer and weaker, which means that GO was reduced to rGO during the preparation process. X-ray photoelectron spectroscopy was investigated to further study the element composition and chemical state of composite materials. From the complete spectrum shown in Fig. 2a, it could be observed that the elements on the surface of the material were Ag, Br, O, and C, corresponding to AgBr and rGO of the composite material, which meant that the rGO/AgBr successfully composited. In addition, the high-resolution spectra in Fig. 2b-d provided a clearer understanding of the chemical state of these elements. As shown in Fig. 2b, the three peaks of C 1s were 285.05 eV, 286.42 eV, and 288.90 eV, corresponding to C-C/C=C, C-O, and O-C=O. The peak intensity of oxygen-containing functional groups was lower, that is, the content of oxygen-containing functional groups was less, which also proved that the graphene was reduced. This was the same as the FT-IR result shown in Fig. 1b. The peaks at 68.09 eV and 69.14 eV in Fig. 2c corresponded to Br 3d 5/2 and Br 3d 3/2 of Br 3d. In Fig. 2d, Ag 3d had characteristic peaks at 367.62 eV and 373.62 eV, corresponding to Ag 3d 5/2 and Ag 3d 3/2 . It indicated that the catalyst of rGO/ AgBr was successfully synthesized during the preparation process according to above analysis.

The morphology analysis
The surface morphology of pure AgBr and rGA-1 samples was studied by SEM and TEM. Figure 3a-b shows the SEM images of pure AgBr, indicating that the pure AgBr photocatalyst particles were irregular block morphology with uneven particle size and a diameter of about 1-4 lm. In addition, Fig. 3c-d shows the prepared rGA-1 samples. As shown in Fig. 3c, pure AgBr particles were uniformly covered by graphene, and the film-like substance was graphene. These images showed that graphene had been successfully composited with pure AgBr photocatalyst. Compared with pure AgBr, the diameter of AgBr composited with graphene was about 0.5-1.5 lm, which were more uniform and smaller than pure AgBr particles. In contrast, the specific surface area increased, and contact area with pollutants also increased.

Optical absorption property
The optical properties of the different photocatalysts could be characterized by UV-Vis diffuse reflectance spectra, and the spectra of the as-prepared samples are demonstrated in Fig. 4a, which showed that the significant light absorption in visible range less than 480 nm was revealed by the pure AgBr. The spectra of the rGA-A (A = 0.5, 1.0, 1.5, and 2.0) samples which contained graphene revealed the larger absorption edge visible light range and the wavelength had an obvious redshift. The results showed that the rGO/AgBr photocatalyst exhibited higher utilization efficiency of visible light than pure AgBr photocatalyst. It was consistent with the results of photocatalytic degradation experiments, indicating that the enhanced photocatalytic activity of AgBr composites is attributed to the addition of GO in the photocatalysts. The band gap energy of AgBr could be evaluated by the equation below: In this equation, a, hv, A, and E g are the absorption coefficient, light frequency, a constant, and band gap energy, respectively. Furthermore, n depended on the type of semiconductor, the value of the direct band gap semiconductor was , and the indirect band gap semiconductor was 2. According to previous review, AgBr was an indirect semiconductor, so the value of n is 2 [22]. Thus, as shown in Fig. 4b, it could be inferred that the E g of AgBr was 2.40 eV.

Photoelectrochemical properties
Photoluminescence spectra analysis can be used to analyze the separation efficiency of photoelectron and holes in photocatalyst. The combination of electron-hole pair will produce certain fluorescence intensity. The higher the combination rate of electron-hole pair is, the higher the fluorescence intensity can be detected, and the worse the photocatalytic performance is. On the contrary, low fluorescence intensity indicates better photocatalytic performance. To further explore the influence of graphene on the catalytic performance of silver bromide, the obtained catalysts were studied by photoluminescence spectra analysis. As shown in Fig. 4c, the photoluminescence intensity of rGA-1 was significantly weaker than that of pure silver bromide when the excitation wavelength was 325 nm, indicating that the addition of graphene significantly inhibited the recombination rate of photocarriers and improved the separation efficiency of photogenerated electrons and holes, which was conducive to the improvement of photocatalytic activity.
Transient photocurrent response spectroscopy was used to evaluate the efficiency of photoelectron separation and migration. For the transient photocurrent response results, high photoelectron separation efficiency was represented by high photocurrent intensity, which indicated good photocatalytic activity. The transient photocurrent response diagram of AgBr and rGA-1 is displayed in Fig. 4d. As shown in the figure, compared with pure AgBr, the photocurrent intensity of the composite was stronger, indicating that the photoelectron separation efficiency of the composite was better, which further proved that the addition of graphene was conducive to photoelectron separation and transfer in catalyst, thus contributing to the improvement of photocatalytic efficiency. It was also in keeping with the previous PL results.

Photocatalytic activity
To determine the photocatalytic activity of the AgBr and a series of rGO/AgBr composites, RhB was used to investigate the photodegradation rate of these photocatalysts under the irradiation of a 300 W xenon lamp (k C 420 nm). As shown in Fig. 5a, the degradation rate of rGA-0.5, rGA-1.0, rGA-1.5, and rGA-2.0 was almost 100.0% after 40 min irradiation from the xenon lamp, while the degradation rate of AgBr was only 54.8%, which meant that compositing with rGO on the surface of AgBr improved the performance of photocatalyst. The sample with the highest degradation rate was rGA-1, which indicated that too much or too little rGO would affect the degradation efficiency. It may be that insufficient rGO leads to low electron transfer efficiency, while too much rGO may affect the visible light absorption and utilization of AgBr. Besides, too much rGO might form recombination centers, causing some electrons to recombine with holes on the surface of rGO.
In order to further explore the photocatalytic activity and stability of prepared materials, the degradation of RhB under different light conditions and cycling experiments were carried out, as shown in Fig. 5e, f. The results showed that the degradation effect under UV lamp was better than that of xenon lamp and solar radiation, and the effect of solar radiation was slightly worse which was completely degraded in 40 min. It may be due to weak intensity of solar energy. According to cycle experiment shown in Fig. 5f, the degradation rate can still maintain 79% after three cycles, which suggested that the prepared photocatalyst was stable for practical application. The reason for decrease in degradation may be due to that the Ag was not stable enough after the long irradiation and some photocatalysts were lost in the process of cycling experiments.
To study the degradation rate of RhB by prepared catalysts, the kinetics of RhB degradation was studied by using a quasi-first-order kinetic model. The pseudo-first-order kinetic model was as follows: where C t and C 0 represent the concentration of RhB at reaction time at t and 0; t and k ap are the reaction time and the reaction rate constant, respectively. It can be observed from Fig. 5b that the reaction rate constant of composite photocatalyst increased obviously compared with pure silver bromide. The rate constants k of rGA-0.5, rGA-1.0, rGA-1.5 and rGA-2.0 were 0.1377 min -1 , 0.2413 min -1 , 0.1287 min -1 , and 0.1712 min -1 , respectively, which were several times of the pure silver bromide (0.0148 min -1 ). It indicated that the photocatalytic degradation rate of AgBr composited with graphene greatly increased. Among them, rGA-1.0 had the largest reaction rate.

Possible photocatalytic mechanism
In order to study the reaction mechanism underlying the photocatalytic degradation of RhB by rGA-1, the main active components generated in the degradation process were determined by trapping experiments. Here, p-benzoquinone (BQ), CTAB, and isopropanol (IPA) were used as the scavengers for O 2 -, h ? , and ÁOH, respectively. As shown in Fig. 5c, the photocatalytic degradation efficiencies of rGA-1.0 to RhB were significantly inhibited when BQ and CTAB were added, indicating that O 2 and h ? played major roles. In addition, when IPA was added, the photocatalytic degradation rate of RhB by rGA-1 was partially reduced, indicating that ÁOH also played a role in the degradation process, but it was not the main active substance in the degradation process. This result of trapping experiments showed that the O 2 and h ? were the major species, while ÁOH played a supporting role.
Electron spin resonance (ESR) spectroscopy with DMPO as radical scavenger was adopted to further confirm the role of certain active specie in the degradation process. As shown in Fig. 6a a typical six-peak signal was detected under visible light which could be assigned to the characteristic signals of DMPO-ÁO 2 adducts, whereas no signal was found in dark condition, which confirmed that ÁO 2 could be generated in the process of photocatalytic degradation. Meanwhile, the four relatively weak characteristic peaks (1:2:2:1 quartet pattern) of DMPO-ÁOH were detected under visible light irradiation as displayed in Fig. 6b, which meant ÁOH also played a role in degradation process. The results were consistent with the trapping experiments. In order to further study the possible photocatalytic degradation of organic pollutants by rGA-1 and the stability of the catalyst during the degradation process, XRD analysis was performed on the rGA-1 catalyst after the photocatalytic experiment (Fig. 5d). The characteristic peak at 2h = 38.116°of the (111) crystal plane of metallic silver appeared after photocatalytic degradation, indicating that during the catalytic degradation process, rGA-1 produced a small amount of Ag under visible light. The surface plasmon resonance effect of these silver nanoparticles may contribute to the photocatalytic degradation of organic pollutants.
The CB energy level and the VB energy level of AgBr could be calculated according to the following equation: where E VB , E CB , X, E c , and E g are the value band edge of semiconductor, the conduction band edge of semiconductor, the absolute electronegativity of the semiconductor. Therefore, the E CB and the E VB value of AgBr are 0.11 eV and 2.51 eV. Through the previous analysis and calculation, the possible photocatalytic mechanism for the photocatalytic degradation RhB by rGO/AgBr heterojunction photocatalyst is indicated in Scheme 2.
First, under the irradiation of visible light, the electrons in the valence band of AgBr are excited and transferred to the conduction band, leaving holes in the valence band. At the same time, the metallic silver is excited by visible light, triggering the surface plasmon resonance effect. Usually, due to the narrow band gap, electron-hole pairs are rapidly recombined, and only a small part of them participates in the photocatalytic process. However, due to the excellent performance of graphene's rapid photoelectron transfer, a large number of photogenerated electrons are transferred to rGO, which is combined with O 2 on the surface of rGO and converted into superoxide anion radicals. It has strong oxidation ability and can effectively degrade RhB adsorbed on rGO. In addition, due to electron transfer, the electrons and holes are separated and the holes remain in the valence band. The holes in the valence band that also have strong oxidizing ability can directly oxidize and decompose the Rhodamine B adsorbed on the surface of the catalyst.

Conclusion
The rGO/AgBr heterojunction photocatalysts, as efficient visible-light-driven photocatalysts, were prepared by a facile solvothermal method. The reduced graphene and AgBr were successfully composited without other impurities, and the composites Scheme 2 the possible photocatalytic mechanism for the photocatalytic degradation RhB by rGO/ AgBr photocatalyst had better optical properties than pure silver bromide. The photocatalytic degradation experiment showed that the degradation rate of Rhodamine B by the rGO/AgBr photocatalysts (almost 100%) was significantly higher than that by pure AgBr (54.8%) after 40 min irradiation, identifying that compositing rGO on the surface of AgBr particles contributes to enhancing the photocatalytic activity. The kinetics experiments showed that the degradation rate of 1% rGO/AgBr heterojunction catalysts was the highest, indicating that it is the most suitable for degrading Rhodamine B. Furthermore, the trapping experiments verified that the main active species in the photocatalytic degradation of organic pollutants by rGO/AgBr heterostructures were ÁO 2and h ? . In brief, this experiment proved that the photocatalytic degradation rate of Rhodamine B in water was greatly improved after compositing with GO, which was beneficial to the practical application of removing organic pollutants in water.