A facile synthesis of bismuth oxychloride-graphene oxide composite for visible light photocatalysis of aqueous diclofenac sodium

In this study, bismuth oxychloride/graphene oxide (BiOCl-GO) composite was fabricated by facile one pot hydrothermal method. The pure BiOCl and BiOCl-GO composite was characterized by X-ray diffraction, Transmission electron microscopy X-ray photoelectron spectroscopy and UV–Vis diffuse reflectance spectroscopy. The synthesized composite was then assessed for photocatalytic degradation of diclofenac sodium (DCF) in visible as well as direct solar light and UV irradiation. Results indicated that the photocatalytic removal efficiency of DCF was significantly affected by dose of catalysts, pH value and source of light. The results reveled that degradation efficiency of BiOCl-GO for DCF reduced from 100 to 34.4% with the increases in DCF initial concentration from 5 mg L−1 to 25 mg L−1. The solar light degradation of DCF using BiOCl-GO was achieved with apparent rate constant 0.0037 min−1. The effect of scavengers study revealed that superoxide ions and holes were mainly responsible for DCF degradation. The regeneration study indicates that BiOCl-GO composite can be successfully recycled up to the five cycles. The study revealed the effectiveness of one pot hydrothermal method for the fabrication of BiOCl-GO composite.

Scientific RepoRtS | (2020) 10:14191 | https://doi.org/10.1038/s41598-020-71139-y www.nature.com/scientificreports/ are non-biodegradable and toxic in nature, thus, a need for advance treatment process like advance oxidation by photocatalysis to eliminate such pollutants completely from wastewater treatment plants 6 . Bismuth oxychloride is a promising oxyhalide which contains unique electric and optical properties. BiOCl has a band gap of ~ 3.2 eV constructed as a nanostructure 7 and is an economically viable and environmental friendly catalyst. In BiOCl, the electron-hole separation is facilitated by its indirect transition and opened structure. Furthermore, BiOCl does not contain toxic substances, hence, is extensively explored in photocatalytic degradation of pollutants 8 . Recently, efforts have been undertaken to enhance the separation of charges such as band and facet engineering to induce proactivity of visible light [9][10][11] . For water treatment, it is necessary to explore the properties of such materials to increase its photocatalytic activity. Graphene is a carbon-based material having sp 2 -hybridized carbon atoms with a honeycomb structure 12 . Graphene's composites have been used for water splitting through photocatalysis to generate hydrogen as well as for photocatalytic degradation of pollutants due to its high specific surface area and greater mobility of charge particles [13][14][15][16] . Several studies proved that BiOCl is a good photocatalyst but due to fast recombination of photogenerated electron-hole (e − /h + ) pairs, its photocatalytic efficiency is compromised [17][18][19][20][21][22] . Therefore, coupling of BiOCl with GO could be a facile method to separate the photogenerated e − /h + to enhance the visible light photocatalysis properties. Several binary, ternary and quaternary composites based on GO and BiOCl have been reported in the literature which showed the enhanced photocatalytic activity. Lin et al. 12 synthesized the GO/BiOCl thin film for the decomposition of rhodamine B and almost 99% degradation was observed within 90 min. A reduction in the bandgap energy to 2.9 eV was observed due to coupling of GO with BiOCl in comparing to pure BiOCl (3.2 eV). Zhang et al. 19 , synthesized the 3D BiOCl/RGO aerogel photocatalyst for the adsorption and photocatalytic degradation of oxytetracycline. The BiOCl/RGO aerogel showed 1.43 times higher removal of oxytetracycline than pristine BiOCl under the visible light irradiations. A series of quaternary BiO x Cl y /BiO m Br n /BiO p I q /GO composites have been synthesized by hydrothermal method and used as a visible light active catalyst (Eg − 2.18 eV) for the photocatalytic degradation of 2-hydroxybenzoic acid and O 2 ·ˉ, h + , and · OH were responsible of mineralization of the 2-hydroxybenzoic acid 20 . These studies revealed that the coupling to the semiconductor catalyst with GO enhance the interaction of the composite with pollutant and facilitates efficient decomposition of the organic contaminants in wastewaters.
Herein, BiOCl coupled GO composite was synthesized by facile one pot method for the photocatalytic degradation pharmaceutical DCF pollutant in aquatic system under of UV, synthetic visible and solar light irradiation. The role of solution pH, initial DCF concentration and BiOCl-GO mass was evaluated to identify the optimum photocatalysis conditions. The kinetics and photocatalysis mechanism of DCF degradation onto BiOCl-GO were also investigated.

Synthesis of GO, BiOCl and BiOCl-GO.
A facile one pot hydrothermal method was used for the synthesis of BiOCl-GO composite. Initially, 60 mg GO was dissolved in 200 mL of 0.6 M HCl solution and stirred for 2 h. Thereafter, 4.0 g NaBiO 3 was added to the GO solution and left for 16 h for continuous stirring at 25 °C. The obtained materials were transferred into hydrothermal reactor and heated at 140 °C for 16 h. The obtained material was washed with deionized water, acetone and ethanol to remove the excess amount of HCl and dried at 105 °C for 8.0 h. A similar method was used for the synthesis of BiOCl in the absence of GO. Hummer's method was used for the preparation of GO 23 . Characterization. XRD analysis was performed to find out the crystalline nature of catalysts BiOCl and BiOCl-GO on Bruker axis D8 model. Transmission electron microscopic (TEM) analysis was conducted on HITACHI-H-770 with acceleration voltages of 100 kV. X-ray photoelectron spectroscopy (XPS) analysis was accomplished on standard Omicron system equipped with monochromatic Al Kα 1,486.7 eV X-ray source operated at 15 keV. Absorption spectrometry was performed using Shimadzu UV-2550 UV-Vis spectrophotometer.

Photocatalytic experiments. The photocatalytic experiments were accomplished using BiOCl and
BiOCl-GO composite in visible light. For that purpose, 100 mL of different concentrations of DCF (5, 15 and 25 mg L −1 ) were taken in 250 mL Pyrex breaker and added 1.0 g L −1 of the catalyst. Before exposure to light, adsorption-desorption for 30 min achieved by stirring DCF and catalyst solutions. The DCF molecules were photocatalytically degraded using different light sources including UV mercury lamps (12 lamps × 8 W each), white cool visible lamps (spectral irradiance 17.45 mW cm −2 ; λ = 390-700 nm) and visible region of solar light (intensity 22,000 lx ≈ 17.38 mW cm −2 ). The suspension was then exposed to respective light sources for 180 min under continuous stirring throughout the experiment, 5 mL aliquots of DCF solution were taken out after the intervals of 30 min and analysed using UV-Vis spectrophotometer at λ max 276 nm. Percentage degradation of DCF was calculated by applying following equation: where C o refers to initial concentration of DCF while C t refers to concentration of DCF at time t.

Results and discussion
Characterization. The XRD patterns of BiOCl and BiOCl-GO are shown in Fig. 1. Results suggested that diffraction peaks of BiOCl were coinciding with standard (JCPDS # 00-006-0249) reference material 21 . The result revealed a highly pure tetragonal structure for BiOCl. Strong characteristic peak appeared at 11.75° which was indexed as (001), shown in layers of [Cl-Bi-O-Bi-O-Cl] stacking structure present along c-axis 20 . The XRD pattern of BiOCl-GO composite is almost same as BiOCl with slight change in the peak position and intensity. Generally, GO characteristic peak for (002) appeared at 11.7° which is overlapping with the BiOCl peak (001) at 11.75 in BiOCl-GO composite 24 . The morphology of BiOCl and BiOCl-GO were identified with the help of Transmission electron microscopy. It was clearly indicated in Fig. 2a, that BiOCl particles are sheet like structures. Figure 2b illustrates the well distibuted GO sheets over BiOCl. Sun 21 also synthesized distinct plate like BiOCl from solvothermal process having   20 . The XPS analysis results revealed that atomic carbon content in BiOCl and BiOCl-GO composite was 28.68 and 38.3%, respectively. The associated resolved peaks for Bi 4f, Cl 2p, C 1s and O 1s are provided in Fig. 3b-e respectively. The binding energy values, and atomic percentages of the identified elements are also listed down in Table 1. The peak located at 159.71 eV corresponds to Bi 4f in the BiOCl-GO composite 25 . In comparison with the pure BiOCl, the Bi 4f peak of BiOCl-GO is shifted towards a slightly higher binding energy due to the interaction among BiOCl and GO 25,26 . In BiOCl-GO composite, the peak for Cl 2p was identified at 198.5 eV, belongs to the characteristic peak of Cl − in BiOCl. Similarly, O1s peak of BiOCl-GO composite was identified at 530.47 eV, corresponding to O 2− of Bi-O bond in BiOCl. Furthermore, the peak recognized at 284.8 eV was ascribed for C 1s in BiOCl-GO which indicated C-C bond with sp 2 orbitals 26 . A difference in the binding energy about 5.3 eV and 5.31 eV was observed between Bi 4f 5/2 and Bi 4f 7/2 in both BiOCl and BiOCl-GO composite, respectively, which is revealing the presence of Bi in the chemical state of Bi 3+ in both materials 21 .
The light absorption propreties of the BiOCl and BiOCl-GO composite were detrmined using UV-Vis diffuse reflectance spectroscopy. The UV-Vis DRS (Fig. 4a) shows that pure BiOCl has very strong absorption in the UV region with a sharp edge around 372 nm. While BiOCl-GO composite showed the greater improvement in the absorption of the light in the visible region. As shown in Fig. 4a, a red shift in the absorption edge of www.nature.com/scientificreports/ BiOCl-GO composite is detected in comparison to pure BiOCl 11 . These results reveleing that couplong the GO with BiOCl enhaced the optical absorption properties of pure BiOCl and a bnd gap shift from 3.33 to 3.08 eV was observed (Fig. 4b).
Photocatalysis. The adsorption experiment was conducted in dark for 3 h using BiOCl and BiOCl-GO catalysts. Figure 5 shows that about 17.2% and 11.9% of the pollutant were removed within 60 min by BiOCl and BiOCl-GO, respectively. Thereafter, equilibrium was established, and no further adsorption was observed.
The results were similar to the study conducted by Mendez et al. 27 , where only 14% of DCF was removed by TiO 2 . Another study was in a close agreement with the finding where 5% of DCF was adsorbed onto the catalyst in first 30 min and the adsorption remained constant for the next 240 min 28 Figure 5 shows the BiOCl/DCF photocatalysis in UV light with complete degradation within 90 min while BiOCl-GO/DCF photocatalytic degradation was completed within 60 min. These results revealed that incorporation of GO with BiOCl played a positive role in the photocatalysis. However, the DCF degradation using BiOCl-GO under solar light and visible light irradiation was 34.5% and 22.9%, respectively. The large difference in the photocatalytic efficiency of the BiOCl-GO composite under UV, solar and synthetic visible light is mainly due strong photolytic interaction of DCF with UV light. The UV light (λ max -254 nm) energy is much higher than visible light (λ max > 400 nm). At greater light intensity, the production of electron hole pair generation is higher, and a large amount of reactive species are generated which are involved in degradation [33][34][35] . Moreover, UV light photolysis (without catalyst) degrades almost 34% DCF which is also responsible for higher DCF degradation under UV photocatalysis process. Therefore, BiOCl and BiOCl-GO showed higher DCF degradation in UV irradiation in compassion to solar and synthetic visible light. Although, cost and safety hazards of UV light in the catalytic process mandate that natural solar light photocatalysis is a viable option. Therefore, experimental conditions must be optimized to find the optimum photocatalytic degradation of DCF using BiOCl-GO composite in the solar light.
The effect of DCF concentration onto the photocatalysis using BiOCl-GO composite was studied at 5, 15 and 25 mg L −1 and the results are depicted in the Fig. 7. The results showed that the degradation rate is inversely proportional to initial concentrations. About 100%, 38.1% and 34.3% degradation were observed at 5, 15 and 25 mg L −1 concentrations, respectively. With the increase in DCF initial concentrations, the removal rate decreased due to lower photon absorption by the photocatalyst because of the magnified number of DCF molecules that hampers the active radical production, hence decreasing the degradation efficiency 1 . Additionally, when the substrate concentration is higher, the surface of the BiOCl-GO composite may be accumulated by the DCF molecules that could result in the reduced surface area for degradation of the pollutant 36 .  The mass of catalyst is one vital parameter which regulates the efficiency of the catalysis process. The degradation of DCF was increased with the increase in mass of BiOCl-GO photocatalyst as shown in the Fig. 8 and  11.1%, 34.3%, 36.7% and 50.8% of DCF photodegradation was found at 0.5, 1, 1.5 and 2 g L −1 BiOCl-GO mass, respectively. The degradation behavior due to increased catalyst load may be ascribed to the presence of greater number of active sites for pollutant along with the higher absorbance of light by BiOCl-GO catalyst resulted in the generation of higher number of reactive species (super oxide and hydroxyl species) that are responsible for photocatalytic degradation 37 .
In the photocatalytic processes, solution pH plays a pivotal role due to the change in the surface charge and functionality of the catalyst and pollutants. Herein, the degradation of the DCF on to BiOCl-GO composite was investigated at the pH 5, 7 and 9 pH because DCF is poorly soluble in water at the lower pH (below pH 3) 38,39 . The results depicted in Fig. 9, revealing that acidic medium is the most suitable for the photocatalytic decomposition of the DCF and maximum degradation was observed at pH 5. The higher degradation of DCF at slightly acidic pH is due to the good interaction between the protonated BiOCl-GO composite and the anionic DCF (pKa-4.2) 39 . Conversely, the presence of anionic DCF in alkaline medium and the negatively charged catalyst resists the interaction of DCF, results in low photocatalytic degradation as shown in the Fig. 9.
The heterogeneous photocatalysis degradation of DCF was found to be more relevant to kinetic model pseudo-first order as shown in Eq. 2 40 .
where, C represented as concentration of DCF at time t, C o denotes as initial concentration of DCF and k (min -1 ) is apparent rate constant. Kinetic plots for each parameter (initial concentration, dose, and pH) were obtained by ln(C 0 /C t ) vs time (Fig. S1). Apparent rate constant values obtained from plots and R 2 values are represented in   The results revealed that with the increase in catalyst dose of BiOCl-GO from 0.5 to 2 g L −1 , the apparent rate constant increases from 0.0003 to 0.0042 min -1 , respectively. For varied pH conditions the highest rate constant was observed at pH 5 with the value of 0.0037 min −1 . These results suggested that the rate of DCF degradation was higher in acidic condition.
Photocatalysis mechanism. The photocatalytic degradation mechanism of DCF on to BiOCl-GO catalyst is shown in Fig. 10. The direct band gap energy of the BiOCl was found to be around 3.33 eV as reported in other studies 17 . Due to large band gap energy, BiOCl is not visible light active catalyst. After incorporation of BiOCl with GO, the composite showed band gap shifts to 3.08 eV exhibiting good catalytic properties. Under the solar light illumination, the photo-induced electrons from the conduction band of the BiOCl moved to its valance band 41 . Thereafter, these electrons are trapped by the graphene oxide which restricts the electron-hole recombination in the catalyst. The photogenerated electrons and holes are responsible for the production of active radical spaces as shown in Fig. 10a, which are in turn responsible for the decomposition and mineralization of the DCF in the solution. Moreover, to identify the role of the reactive species i.e., electrons, holes, superoxide radicals and hydroxyl ions on the DCF degradation, the photocatalysis experiments were conducted in the presence of scavengers. The p-benzoquinone (PBQ), isopropanol (IP) and triethanolamine (TEOA) were employed for superoxide, hydroxyl radicals and holes scavengers, respectively. The photocatalyst without any scavenger showed around 49% degradation of DCF in sunlight. Adequate amount of TEOA was added to different replicates of DCF solution containing the photocatalyst which exhibited decline in photocatalytic activity up to 18.5%. Similarly, the addition of IP and PBQ also diminished the photocatalytic activity of BiOCl-GO in the order no-scavenger > PBQ > TEOA > IP as shown in Fig. 10b. The significant decline in the photocatalytic activity with addition of BQ, TEOA and IP suggest that the photocatalysis of DCF over BiOCl-GO is affected by holes, super oxides and hydroxyl radicals with major contribution from superoxide ions and holes 42,43 . Reusability of BiOCl-GO composite. The stability and reusability of the spent BiOCl-GO composite was studied up to the five cycles. Prior to the regeneration study, the spent BiOCl-GO composite was washed with de-ionized water and dried at 105 °C. The results for the regeneration studies are depicted in Fig. 10c. The obtained results revealed that a slight reduction in the photocatalytic efficiency of the BiOCl-GO composite for DCF after the fifth cycle. These results clearly demonstrated that BiOCl-GO composite is stable and can be reused for the multiple times for DCF degradation.
Comparative photocatalysis. To estimate the efficiency of the BiOCl-GO composite, its photocatalytic activity was compared with literature studies. The highest solar light photocatalytic activity for 25 mg L −1 Diclofenac was achieved with BiOCl-GO composite compared with previously studies, as shown in Table 3. These results reveal that BiOCl can effectively remove Diclofenac from water. The higher percentage activities of other catalysts may be attributed to either exposure to UV light where photolysis plays a stronger role than photocatalysis 27 or variation in operational conditions such as lower initial pollutant concentration or use of noble metal dopants 44,45 . conclusion It is concluded from this study that BiOCl and BiOCl-GO were successfully synthesized by a facile one pot hydrothermal method. The photocatalytic activity of BiOCl-GO was systematically examined for the degradation of micro pollutant diclofenac sodium. In comparison to BiOCl, the BiOCl-GO composite showed complete degradation of DCF within 60 min in UV light for 25 mg L −1 DCF. Solar and synthetic visible light mediated www.nature.com/scientificreports/ experiments also favored superior activity of BiOCl-GO towards DCF with higher pseudo first order kinetic constant value 0.0386 min −1 . The cost and safety hazards of UV light in the catalytic process mandate that natural solar light photocatalysis a viable option. This study showed a great potential to degrade pharmaceutical pollutants under solar irradiation for enhanced removal of micropollutants from the aqueous environments.