Highly efficient visible light-induced photocatalytic oxidation of arsenite with nanosized WO3 particles in the presence of Cu2+ and CuO

ABSTRACT Although WO3 appears to be one of the extensively studied photocatalysts, the low response of pure WO3 in aqueous solution under visible light limits its application remarkably. In this work, the enhancement of the efficiency of WO3 for the visible light-driven photocatalytic oxidation of arsenite was explored using Cu2+ ion and CuO as a co-catalyst. While the addition of Cu2+ was found effective for the suppression of dissolution of WO3, the efficiency of CuO appeared to be slightly lower. Significant improvement of the efficiency for the photocatalytic oxidation of As(III) with WO3 was noted when Cu2+ ions and CuO were added. The optimized conditions were WO3 in the presence of 10 mg L−1 Cu2+ ion and 1 wt% CuO coupled with WO3, respectively. The As(III) concentration of 10 mg L−1 could be lowered to less than 0.1 mg L−1 by the photocatalytic treatment. Acidic pH favours the oxidation of arsenite in the presence of Cu2+ whereas basic pH is suitable with CuO. Characterization techniques such as TEM, XPS, XRD and UV–DRS were used to characterize photocatalysts. The reactive species scavenger tests revealed that the photo-induced holes (h+) play a key role in the photocatalytic oxidation process while the effect of •OH is negligible. It was found that As(III) oxidation rate was remarkably suppressed in the nitrogen atmosphere. A mechanism for enhanced photocatalytic oxidation has been proposed based on the results of the reactive species scavenger tests. This research may contribute to the large-scale As(III) oxidation treatment in the groundwater. GRAPHICAL ABSTRACT


Introduction
The presence of high concentration of arsenic in water has become one of the major concerns due to its toxicity and a ubiquitous presence on the earth's crust [1]. Although the major species of arsenic in the environment are As(-III), As(0), As(III) and As(V), the primary forms in water are As(III) (arsenite) and As(V) (arsenate) [2]. Generally, the level of total arsenic in groundwater is higher than that in surface water. Globally, the groundwater of about 107 countries are affected by arsenic contamination [3]. For example, in China [4], the concentration of arsenic in groundwater is reported to be >4000 µg L −1 , and in Bangladesh [5], the level of groundwater was reported up to 14000 µg L −1 . Groundwater is enriched with arsenite and it is reported that As (III) can be reached up to 90% of total arsenic which exists primarily as H 3 AsO 3 at near-neutral pH [6]. Due to the uncharged state, As(III) has a poor affinity towards adsorbents and coagulants, which makes it quite difficult to remove from an aqueous solution [7]. Therefore, traditional adsorption-based arsenic removal methods demonstrate very low efficiency for arsenic removal. In addition, the trivalent arsenic (As(III)) is highly mobile and 25-60 times more toxic compared to As(V) [8]. Considering the above-mentioned factors, the pre-oxidation of arsenite is considered a judicious choice for reducing the toxicity and increasing the removal efficiency of total arsenic in arsenic-contaminated waters. Until now, a number of schemes have been proposed for the oxidation of arsenite such as photocatalysis [9], electrocatalysis [10], chemical process [11] and biological method [12]. However, most of the methods among them are costly, while some methods generate toxic and carcinogenic by-products and secondary pollution [13]. Arsenic oxidation by the photocatalytic method has been recognized as a suitable and environment-friendly green energy technique due to a number of advantages such as high efficiency, the no-or-little requirement for chemicals, less need of electrical energy and applicability of this technique for the treatment of a wide range of arsenic concentration [14][15][16].
Although TiO 2 seems to be an extensively studied photocatalyst among the oxide semiconductors owing to the low cost, stability and reactivity, it is effective only under UV light irradiation which limits its application in large-scale treatment. The visible light-active photocatalysts are highly required for the better exploitation of solar energy as it accounts for the lion share of the solar spectrum (∼42%) [17,18].
Recently, tungsten oxide (WO 3 ) is considered a powerful applicant among the visible light-active photocatalysts because of a number of desired properties such as effective oxidation capacity of valence band (VB) holes, strong absorption of solar radiation, little bandgap (2.4-2.8 eV) and nontoxicity [19][20][21]. Nonetheless, pure WO 3 shows low performance during the photocatalytic treatment due to the quick recombination of photo-generated electron-hole pairs since the excited electron in WO 3 is unable to react with oxygen [22,23]. In order to improve the photocatalytic efficiency of WO 3 , a number of strategies are being practised. Coupling of WO 3 with appropriate co-catalyst is one of the commonly applied techniques which enhance the photocatalytic activity by increasing the efficiency of electron-hole separation. The improvement of photocatalytic performance by using co-catalyst has recently received extensive attraction [24][25][26]. Enhanced photocatalytic activity of WO 3 with metal co-catalysts e.g. Ag, Pt, Au and Pd has been reported by many authors [27][28][29][30]. However, applications of these rare and costly materials as co-catalysts on large-scale are restricted due to their slight availability and high cost.
As a substitute, cost-efficient metal ions and metal oxides have been extensively explored [20,31]. Copper oxide and copper ions may be of significant interest as co-catalysts due to their low cost, easy handling and availability in large quantities. To date, a number of research articles are available on the enhancement of visible-light activity of WO 3 photocatalyst in the addition of Cu 2+ and CuO [32][33][34]. Nevertheless, the primary focus of these research efforts was either the oxidation of organic compounds or the degradation of various dyes. Literature on the efficiency of WO 3 for the photocatalytic oxidation of arsenite is limited. Kim and his co-workers [35] studied the visible-lightinduced photocatalytic oxidation of arsenite with pure WO 3 and present a detailed study on the mechanism of arsenite oxidation. In another study, the same authors [36] examined the degradation/oxidation of six aquatic pollutants in addition to As(III) with platinized WO 3 photocatalyst and illustrated an improved photocatalytic performance for the oxidation of arsenite and the mineralization of organic pollutants. Nonetheless, the oxidation of arsenite with Cu 2+ -WO 3 and CuO/WO 3 system under visible-light irradiation draw little attention and we did not find any such study in literature.
Herein, this work was undertaken to examine the improvement of photocatalytic efficiency of WO 3 for the oxidation of arsenite by coupling with Cu 2+ ion and CuO as co-catalyst under the exposure of visible light. An insight into the mechanism of As(III) oxidation was explored by using various trapping agents.

Preparation of photocatalyst
The CuO/WO 3 nanocomposite was prepared was by using a mechanical mixing method where the homogeneity of components in the mixture was achieved by a mortar. The various compositions of CuO/WO 3 were obtained by homogeneously mixing the calculated amount of CuO with tungsten oxide. Two composites such as 1 wt% CuO/WO 3 and 5 wt% CuO/WO 3 were prepared by adding 5 mg CuO in 495 mg WO 3 and 25 mg CuO in 475 mg WO 3 , respectively. The preparation of Cu 2+ (x mg L −1 ) + WO 3 was carried out by adding an estimated volume Cu 2+ solution (prepared from Cu(NO 3 ) 2 ) into the WO 3 suspension. Four solutions of Cu 2+ + WO 3 were prepared such as Cu 2+ (0.33 mg L −1 ), Cu 2+ (3.3 mg L −1 ), Cu 2+ (10 mg L −1 ) and Cu 2+ (16.5 mg L −1 ).

Photocatalytic oxidation
A batch reactor including a 50-mL cylinder-shaped glass cell was used for the photocatalytic treatment of As(III). In each experiment, 10 mg of catalyst was added into the reactor as well as 30-mL aqueous As(III) solution having 10 mg L −1 . A dilute solution of HCl or NaOH was used to maintain the pH of the suspension to the expected level. Before starting irradiation, the suspension was stirred for 30 min with a magnetic stirrer in the dark to attain adsorption-desorption equilibrium. Between the light source and reactor, a UV cut filter was positioned in order to cut off the transmission of light energy <400 nm. A regular visible LED light that provides irradiance with λ max at around 405 nm was employed to illuminate the suspended solution. In order to carry out treatment under an inert atmosphere, N 2 gas was purged into the solution by bubbling. The role of photo-generated reactive species for the oxidation of As(III) was evaluated by using respective quenching agent of excess concentration (1000 mg L −1 ). After completing photocatalytic treatment, a 20μm membrane filter (Advantec) was employed to separate catalyst and arsenic solution. The individual arsenic species concentration was then analysed from the filtered solution. To ensure reproducibility, every experiment was carried out more than twice.
UV-visible spectrophotometer was used to measure the concentration of As(V) with arsenic-molybdate technique [37]. The concentration of total arsenic ([As] tot = [(As(III)] + [As(V)]) was determined by using a modified arseno-molybdate technique developed by the author [38]. The [As(III)] was estimated by subtracting the concentration of As(V) from the total arsenic concentration.

Characterization of photocatalyst
A transmission electron microscope with accelerating voltage 125 kV (TEM, Hitachi H-7000) was employed for the investigation of the morphology and microstructure of the photocatalysts. The TEM images of pristine WO 3 and CuO are illustrated in Figure 1. The particle sizes of <100 nm for both oxides could be confirmed by the TEM image. A horizontal-type RIGAKU Ultima IV model powdered X-ray diffractometer was used to examine the diffraction patterns of the prepared catalysts where the range of 2θ angle was from 10 o to 80 o and the scan rate was 4 o min −1 with a step size of 0.02°. A CuK α radiation of λ = 0.15406 nm at 40 kV and 50 mA was used as an X-Ray source. In order to carry out XPS (X-Ray photoelectron spectroscopy) measurement, a PHI Quantera SXM photoelectron spectrometer using AlK α radiation source was employed. The surface charge effects were compensated by calibrating the binding energies using the C 1s peak at 284.80 as the reference. The DRS (Diffuse reflectance spectra) of photocatalysts of various compositions were recorded by a Shimadzu UV2450 UV/VIS system coupled with an integrating sphere diffuse reflectance accessory. The spectra were taken using reference material BaSO 4 of over 200-850 nm wavelength range.

Suppression of WO 3 dissolution
The metal oxides may undergo dissolution by various processes for decreasing thickness hydration and the formation of porosity and roughness. The dissolution of WO 3 is related to the point of zero charge (PH pzc ) which is lowest at pH pzc = 2.5 [39]. Below the pH pzc, the dissolution process is aided by the H + ion as presented in Equations 1 and 2 [40].
At around pH 2.5, the dissolution of WO 3 is mostly promoted by the direct attack of water molecules.
Above the pH pzc , the tungsten trioxide dissolution is OH − assisted [41].
The incident of WO 3 dissolution was evaluated by visually assessing the turbidity of the filtrated solution after photocatalytic treatment of As(III) with bare WO 3 , Cu 2+ -WO 3 and CuO/WO 3 at pH around 7. The colour of the filtrated solution was shown in Figure 2. Maximum dissolution of WO 3 was observed when the suspension contains only WO 3 . It was found that with CuO/WO 3 , the addition of CuO was effective for the decrease of WO 3 dissolution in the solution, although slight turbidity could be observed. Also, in the case of Cu 2+ ion, the concentration 10 mg L −1 was effective for the suppression of dissolution and a completely transparent solution could be obtained. The minimization of dissolution by the presence of a co-catalyst has been reported in the literature, for instance, in the case of Dy-doped WO 3 . It was found that Dy-doped WO 3 did not react with water [41]. The reduction of dissolution in the presence of Cu 2+ can be ascribed to the creation of CuWO 4 via the surface  precipitation reaction between Cu 2+ ion and WO 4 2− [42].
The newly formed compound CuWO 4 may be stable in a wide pH range, even under light irradiation and can be present on the surface of WO 3 .

Effect of Cu 2+ ion and CuO
The influence of the presence of Cu 2+ ion and CuO on the visible-light responsive photocatalytic oxidation of arsenite with tungsten oxide was investigated and is illustrated in Figure 3. Although pure WO 3 could oxidize arsenite, the oxidation efficiency was slightly low (∼55% at 6 h irradiation). On the other hand, the photocatalytic performance of WO 3 was significantly improved in the presence of Cu 2+ ions (Figure 3(a) and Figure S1 in SI). The trend for the oxidation efficiency of As(III) was independent of the concentration of Cu 2+ ion of >0.33 mg L −1 . Since the Cu 2+ concentration of 10 mg L −1 could completely inhibit the dissolution of WO 3 , the optimum Cu 2+ ion concentration became 10 mg L −1 . An analogous incident has been observed for the TiO 2 system with dissolved Cu 2+ ions for the degradation of sucrose [43]. Also, the significant improvement of organic dye degradation performance of WO 3 in the addition of Cu 2+ was accounted for by Arai and his coworkers [44].
The As(III) oxidation efficiency with 1% CuO/WO 3 was found to be better compared with that observed for 5% CuO/WO 3 (Figure 3b and Figure S2 in SI). The composition dependency on the visible-light activity of WO 3 in the presence of other co-catalysts has been extensively cited in earlier works [32,45]. The decrease of photocatalytic oxidation efficiency with the further increase of CuO proportion (5% CuO/WO 3 ) might be due to the recombination of photo-generated electrons and holes [38]. Since, in Figures S3 and S4 in SI, the curve for the efficiency of arsenite oxidation in the ternary system with 1% CuO/WO 3 in the addition of Cu 2+ ion was almost the same as that observed with only 1% CuO/WO 3 revealed that the synergic effect could not be obtained in the ternary system for the addition of both CuO and Cu 2+ ions.
The kinetics for the photocatalytic oxidation of arsenite can be explained by the modified Langmuir −Hinshelwood kinetic model [46]; Here, C 0 and C represent the initial and real-time concentrations of arsenite in mg L −1 , k obs indicates the apparent first-order rate constant in min −1 and t is the irradiation time in min. A plot of Ln(C/C 0 ) versus the treatment time is illustrated in Figure 4. By applying least-square regression analysis, the values of k obs were obtained as shown in Table 1. These values of observed rate constant are significantly higher than similar studies carried out by Moon et al. (2017), where carbon nitride-modified WO 3 was used, and Qin et al. (2016), where Ptmodified TiO 2 was used for the photocatalytic oxidation of As(III) under visible light [47,48]. The visible-lightinduced photocatalytic activity of WO 3 in the addition of Cu 2+ ion was comparatively better relative to those obtained with CuO/WO 3 . The cyclic runs for the oxidation of arsenite with CuO/WO 3 and WO 3 in the addition of Cu 2+ were investigated, and it was found that after five cycles, the efficiency became a little worse (only <2%).

Influence of pH
The solution pH can significantly influence the nature of the catalyst surface along with the speciation characteristics of arsenic in aqueous media. The relationship between the pH and the photocatalytic activity is supposed to be a complex phenomenon that might be correlated with both the nature of substrate adsorption onto the surface of the photocatalyst and the reaction mechanism [49]. The pH dependency of the visiblelight-driven photocatalytic oxidation onto the surface of bare WO 3 , CuO/WO 3 as well as WO 3 in the addition of Cu 2+ ion (10 mg L −1 ) was examined and the results are presented in Figure 5. The results demonstrate that the efficiency for the photocatalytic oxidation of arsenite was declined with the increase of pH. Since the pH pzc (point of zero charge) for WO 3 is approximately 2.5 [27], the surface of tungsten oxide particles at pH 3 has a slightly negative charge. Because the arsenite species H 3 AsO 3 at pH 3 are neutral, the affinity between arsenic species and the photocatalyst surface at pH 3 is relatively good [23,50]. On the other hand, since at pH 10, the surface of the WO 3 particle is negative and the arsenic species exist as H 2 AsO 3 − , there is an active repulsive force between the photocatalyst surface and As(III). The low photocatalytic efficiency of WO 3 at higher pH (pH 10) in the presence of Cu 2+ ions might be ascribed to the development of Cu(OH) 2 . This phenomenon can be explained by the solubility product constant (K sp ) concept. The K sp value for Cu (OH) 2 is 2.20 × 10 −20 as presented below [51]: K sp = 2.20 × 10 20 (6) As shown in Figure S5 in SI, because the product of [Cu 2 + ] and [OH − ] 2 exceeds solubility product constant (K sp ) at pH 6.1, the newly formed Cu(OH) 2 starts to precipitate at this pH according to the theoretical calculation. With increasing pH at more than 6, the decrease in Cu 2+ concentration in the solution was further confirmed by the colorimetric method (bathocuproine test) [52]. After   the photocatalytic treatment at pH 10, the concentration of a copper ion in the filtrated solution was >1 mg L −1 . The pH trend for As(III) oxidation efficiency with bare WO 3 as well as CuO/WO 3 was different from the case of WO 3 in the presence of Cu 2+ ions ( Figure 6). Although in the photooxidation with bare WO 3 and CuO/WO 3 , the arsenite oxidation efficiencies in an alkaline environment (pH 10) were slightly larger compared with those seen at pH values 3 and 7, the difference was small. A similar pH dependency was mentioned by Qin [48]. Moreover, it was reported by Lee and Choi that the rate of arsenite oxidation at pH 9 was about double compared to the rate observed at pH 3 [53].
Since WO 3 has the valence band (VB) potential +3.1 V NHE [54] and the reduction potential of As(V)/As(III) couple is +0.56 V NHE at pH 0 [55,56], the photoinduced hole has sufficient thermodynamic potential by which it can oxidize As(III) to As(V). As the potential for As(V)/As(III) couple at pH 10 is around −0.3 V, the difference of potential between the valence band E VB and the redox potential E°(As(V)/As(III)) at basic pH (pH 10) will be larger relative to that would be at pH 3 and pH 7 [57]. The slight increase of arsenite oxidation efficiency could be due to the larger driving force at pH 10.

Characteristics of catalyst
The XRD patterns of pristine WO 3 , CuO/WO 3 as well as the catalyst after the photocatalytic experiment with WO 3 in the addition of Cu ion (Cu 2+ -WO 3 ) are shown in Figure S6 in SI. The patterns of X-ray diffraction (XRD) for CuO/WO 3 and the catalyst after the photocatalytic treatment were almost comparable to that of pristine WO 3 . The diffraction peaks for Cu 2+ ion and CuO species could not be spotted perhaps because of low concentration and good dispersion for their species. Furthermore, the peak patterns for the catalyst after the photocatalytic treatment with Cu 2+ -WO 3 were almost the same for various Cu 2+ ion concentrations, indicating that the WO 3 dissolution in the solution was very slight and the addition of Cu 2+ ion was effective for the decreasing of their dissolutions.
The ultraviolet diffuse reflectance spectra (UV-DRS) for bare WO 3 CuO/WO 3 and the catalyst after the photocatalytic oxidation with Cu 2+ -WO 3 are presented as Kubelka-Munk function vs. wavelength in Figure S7 in SI. The peaks for the surface plasmon resonance effect could not be observed in these samples. The absorption of visible light by CuO/WO 3 was larger than the bare tungsten oxide. The Tupac plots of the catalysts are shown in Figure S8 in SI. The similar band gaps (2.7 eV) were obtained for three samples. These phenomena postulate that Cu 2+ was not reduced to Cu(0) and consequently copper metal was not settled down on the WO 3 surface at the time of photocatalytic treatment.
To comprehend the change of catalyst at the time of photocatalytic treatment, the X-ray photoelectron spectroscopic (XPS) spectra were evaluated before and after the treatment. Figure 7 illustrates the XPS survey spectra of bare WO 3 , CuO/WO 3 without arsenic treatment, CuO/ WO 3 after arsenic oxidation and the catalyst after the photocatalytic oxidation with WO 3 + Cu 2+ ion (10 mg L −1 ). The main peaks for W 4f , W 4d , W 4p , O 1s and Cu 2p were observed in the spectrum for as-prepared CuO/ WO 3 . The spectra of CuO/WO 3 and the catalyst (Cu 2+ -WO 3 ) after photocatalytic oxidation were roughly analogous to that obtained from CuO/WO 3 without treatment. The narrow scan XPS spectra of Cu 2p, W 4f and O 1s are illustrated in Figure 8. The peaks of 2p 3/2 Figure 7. XPS survey spectra of pure WO 3 , as prepared 1%CuO/ WO 3 , 1%CuO/WO 3 after treatment and the catalyst after treatment with WO 3 in the presence of Cu 2+ ion (10 mg L −1 ). (933 eV) and 2p 1/2 (953 eV) for as-prepared CuO/WO 3 can be attributed to Cu(II). The similar peaks at 933 and 953 eV could be observed for both CuO/WO 3 after the treatment and the catalyst after the photocatalytic treatment with Cu 2+ -WO 3 . It was reported in previous work [33,45,57]. Therefore, the oxidation of Cu(I) into Cu(II) by dissolved oxygen seems to be very quick and speedy.

Photo-oxidation mechanism
In order to evaluate the reaction mechanism, it is crucial to identify the reactive species (ions or radicals) accountable for As(III) oxidation during the photocatalytic experiment. Several reactive species can be produced when photocatalyst is exposed to visible light including  photo-induced holes and the intermediates like •OH, O 2 · and H 2 O 2 . The H 2 O 2 species cannot play a noteworthy role in the oxidation of arsenite [35]. Hydrogen peroxide species could be confirmed during the treatment by the 4--amino antipyrine colorimetric method with the enzyme. The quenching experiments using different quenching agents are reported to be useful for the identification of major photo-generated oxidative species during the reaction. For example, the role of the photo-generated hole can be confirmed by using hole scavenger such as ammonium oxalate and ethylenediaminetetraacetic acid disodium (EDTA) [58,59] while the effect of hydroxyl radical could be understood by using •OH scavenger such as isopropyl alcohol (IPA) and tert-butyl alcohol (TBA) [60]. Tracing agents were added in remarkably excess concentration (1000 mg L −1 ).
How •OH radical plays its role during photocatalytic oxidation of arsenite is not clear yet. Debatable reports have been accounted by researchers on the role of hydroxyl radical in the case of As(III) oxidation [60,61]. A comparative investigation of the photocatalytic oxidation of arsenite with and without hydroxyl radical scavenger (IPA and TBA) was carried out to verify the role of •OH. As presented in Figure 9a, the curves for oxidation efficiency with Cu 2+ -WO 3 in the presence of TBA were almost the same as those observed without radical scavengers. Although in the case of IPA, the oxidation efficiency increased very slightly, the increasing efficiency was neglectable. The similar phenomena were also noted with CuO/WO 3 (Figure 9b). The little change in the oxidation rate in the presence of hydroxyl radical quenchers ultimately proves that the role of •OH radical for the photocatalytic oxidation of arsenite in the current condition is insignificant. These observations are in close accordance with that mentioned by Kim and his coworkers [49].
The photocatalytic activity with WO 3 in the addition of Cu 2+ ion was noticeably suppressed with EDTA and ammonium oxalate (Figure 9c). A similar result for both scavengers was also recorded with CuO/WO 3 (Figure 9d). The oxidation efficiencies in the presence of EDTA for Cu 2+ -WO 3 and CuO/WO 3 were worse relative to those obtained with TBA. Some authors in their earlier studies also reported similar results [51]. These results may lead to the conclusion that photo-induced holes have a key role in the photocatalytic oxidation of As(III).
The efficiency of photocatalytic oxidation of As(III) may be influenced by the presence of dissolved oxygen in aqueous media. The role of dissolved oxygen was assessed by conducting an arsenic oxidation experiment in air and inert (Nitrogen) environment and the results are shown in Figure 10. It is evidenced from the figure that for both the cases of WO 3 in the addition of Cu 2+ and CuO/WO 3 , the oxidation efficiency under nitrogen atmosphere became worse in contrast with those achieved under air environment. This fact demonstrates the involvement of dissolved oxygen in the photocatalytic reaction. When the photocatalytic reaction is carried out with WO 3 in the presence of Cu 2+ , the dissolved oxygen reacts with reduced copper(I) to regenerate Cu(II) ion which helps to enhance oxidation reaction [56]. The comparable incident was also reported in the previous study with CuO/WO 3 [45].
Effective separation of photo-generated electron −hole pair is a key requirement for higher activity under photocatalytic treatment. The electrons positioned in the valence band of WO 3 that are energized during the photocatalytic experiment and jumped to the conduction band enable to create a condition where holes are populated in low-energy valence band and electrons in the high-energy conduction band. The reduction of dissolved O 2 by the conduction band electron of WO 3 surface is thermodynamically unfavourable based on electrochemical consideration since the conduction band potential of WO 3  . This unfavourable situation is responsible for the large recombination of visible-light-induced electron-hole pair which may ultimately result poor in photocatalytic activity. Nevertheless, the effective separation of electron and hole in the presence of CuO can be achieved in two pathways (Scheme 1(a)). In the first pathway, the separation can be accomplished via the Z-scheme mechanism as Figure 10. Effect of dissolved oxygen on the photocatalytic oxidation of As(III) with WO 3 in the presence of Cu 2+ ion and CuO/ WO 3 Initial As(III) concentration: 10 mg L −1 , pH: 7.
presented in pattern 1 of Scheme 1(a) and the second pathway by shifting the conduction band electron of WO 3 to CuO as illustrated in pattern 2 (Scheme 1(a)). Initially, the Cu(II) ion of CuO could be reduced to Cu (I) by accepting an electron from the conduction band of WO 3 [33]. Once reduced to Cu(I), the reduction is not further proceeded to Cu(0) rather can reg to Cu(II) by dissolved oxygen in the aqueous solution [36]. Since the single-electron reduction of O 2 on CuO is not favoured on thermodynamical consideration, the CuO co-catalyst on the WO 3 generate surface may assist the reduction of oxygen by means of a multiple−electron reaction pathway [52]. This synergic effect between WO 3 and CuO can be effective for the suppression of electron-hole recombination which consequently enhances the efficiency of charge separation. The charge separation may also be achieved in the presence of Cu 2+ ions in the reaction media as shown in Scheme 1(b). From the thermodynamic consideration, because of low reduction potential (E°(Cu 2+ /Cu + ) = +0.16 V NHE ), the reduction of Cu 2+ ion by the conduction band electron of WO 3 is not allowed. However, it was reported in previous studies that the reaction might take place by the high-energy electrons that existed in the upper conduction band of WO 3 [61]. The regeneration of Cu(I) to Cu(II) may be occurred by the reaction with dissolved oxygen. Accordingly, the addition of Cu 2+ ion was useful for the suppression of the recombination of electron-hole pairs and the improvement of oxidation efficiency.

Conclusions
The presence of Cu 2+ ion and CuO were able to increase the arsenite oxidation rate constant to 3.6 and 2.3 times better compared to pure WO 3 , respectively. The dissolution of WO 3 was significantly decreased due to the addition of Cu 2+ . Although, the rate of As(III) oxidation was faster in acidic pH conditions in the case of the Cu 2+ -WO 3 system, basic pH was favourable in CuO/ WO 3 system. The study of the reaction mechanism demonstrates that photogenerated holes and dissolved oxygen are the primary species responsible for the photocatalytic oxidation of As(III). The developed technique might be a promising method for the visiblelight-driven photocatalytic oxidation of As(III).

Disclosure statement
No potential conflict of interest was reported by the author(s).

Data availability statement
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.