Efficiently Visible-Light Driven Photoelectrocatalytic Oxidation of As(III) at Low Positive Biasing Using Pt/TiO2 Nanotube Electrode

A constant current deposition method was selected to load highly dispersed Pt nanoparticles on TiO2 nanotubes in this paper, to extend the excited spectrum range of TiO2-based photocatalysts to visible light. The morphology, elemental composition, and light absorption capability of as-obtained Pt/TiO2 nanotubes electrodes were characterized by FE-SEM, energy dispersive spectrometer (EDS), X-ray photoelectron spectrometer (XPS), and UV-vis spectrometer. The photocatalytic and photoelectrocatalytic oxidation of As(III) using a Pt/TiO2 nanotube arrays electrode under visible light (λ > 420 nm) irradiation were investigated in a divided anode/cathode electrolytic tank. Compared with pure TiO2 which had no As(III) oxidation capacity under visible light, Pt/TiO2 nanotubes exhibited excellent visible-light photocatalytic performance toward As(III), even at dark condition. In anodic cell, As(III) could be oxidized with high efficiency by photoelectrochemical process with only 1.2 V positive biasing. Experimental results showed that photoelectrocatalytic oxidation process of As(III) could be well described by pseudo-first-order kinetic model. Rate constants depended on initial concentration of As(III), applied bias potential and solution pH. At the same time, it was interesting to find that in cathode cell, As(III) was also continuously oxidized to As(V). Furthermore, high-arsenic groundwater sample (25 m underground) with 0.32 mg/L As(III) and 0.35 mg/L As(V), which was collected from Daying Village, Datong basin, Northern China, could totally transform to As(V) after 200 min under visible light in this system.


Background
Arsenic (As) contamination is widely recognized as a global health problem. The distribution of As(III) and As(V) in natural water depends on the redox potential and pH of water [1]. Compared with As(V), As(III) is generally reported to have a low affinity to the surface of various minerals, because it mainly exists as nonionic H 3 AsO 3 in natural water when pH <9. Nevertheless, As(V) adsorbs easily to solid surfaces, so it is easier to be removed. Since As(III) is more toxic and more difficult to remove than As(V), a pre-oxidation technology by transforming As(III) to As(V) is highly desirable to remove arsenic from water [2].
Kinds of treatment methods have been reported on oxidizing As(III) to As(V), including biological oxidation, chemical oxidation with conventional oxidants, such as chlorine, chlorine dioxide (ClO 2 ), chloroamine (NH 2 Cl), permanganate (MnO 4 − ), manganese oxides, and hydrogen peroxide [3], photo-oxidation using ultraviolet and visible light radiation, and photocatalytic oxidation [4]. Among these techniques, the photocatalytic oxidation of As(III) to As(V) is newly developed and becoming a promising method. Up to now, the photocatalysts used for oxidizing As(III) reported in literatures are TiO 2 [5,6], BiOI [7], and WO 3 [8], and TiO 2 is widely used for As(III) oxidation. Photocatalytic oxidation of As(III) in TiO 2 suspensions has been proved to be an efficient and environmentally acceptable technique [9][10][11][12]. Rapid oxidation from As(III) to As(V) could be realized in TiO 2 suspensions, e.g., a 10 mg/L of As(III) could be totally oxidized to As(V) within minutes under UV irradiation [13]. TiO 2 is limited as an efficient photocatalyst because of its wide band gap (3.2 eV) and high recombination rate of photogenerated electron-hole pairs. So, how to expand the absorption band of TiO 2 -based photocatalysts to visible light range or reduce the recombination of electron-hole pairs are key points in using TiO 2 -based materials as highly efficient photocatalysts.
Numerous attempts have been devoted to extend the photo response range of TiO 2 to visible spectral area. For instance, TiO 2 -based photocatalysts were modified by doping with metal cations [14] or nonmetal ions [15], photosensitizing with dyes on the TiO 2 surface, depositing noble metals [16], or coupling with another semiconductor (such as CdS, Fe 2 O 3 , ZnO, and SnO 2 ) [17,18]. Up to now, TiO 2 -based nanoparticles functionalized with Fe [19], γ-Fe 2 O 3 [20], Mn 3 O 4 [21] and MoO x [22] , and sensitized with ruthenium dye [23,24] have been used for arsenite oxidation and all exhibited better photocatalytic oxidation performance for arsenite than pure TiO 2 . Among these studies, the deposition of Pt nanoparticles on TiO 2 was proved to have a high photocatalytic activity [25]. Pt doping of TiO 2 can form the Schottky barrier among the metals and the electronic potential barrier at the metal-semiconductor heterojunction, and the platinized TiO 2 can trap the photogenerated electrons efficiently [26,27]. Furthermore, Pt-doped TiO 2 materials produce significantly higher photocatalytic activity under visible light irradiation, while the photocatalytic activity under UV irradiation is improved slightly.
To reduce the recombination of photogenerated electron-hole pairs, the technique of photoelectrocatalytic oxidation has attracted increasing attention in the field of environmental protection. Photoelectrocatalytic oxidation techniques were first applied in As(III) oxidation under UV irradiation by Fei et al., and the application of an external positive bias voltage on the catalyst could draw the photo-generated electrons away via the external circuit, leaving the holes for oxidation of As(III). Therefore, compared to the photocatalytic process, the probability of the rapid recombination of electron-hole pairs is largely reduced and the photo-oxidation ability for As(III) can be raised in the photoelectrocatalytic process [28]. Later, dye-sensitized photoelectrocatalytic oxidation over nanostructured TiO 2 film electrodes were applied in As(III) transformation under visible light by Li et al. and showed a high photocatalytic activity for As(III) oxidation [23,24,29].
Compared with TiO 2 film as a photocatalytic electrode, TiO 2 nanotube arrays fabricated by electrochemical anodization have been demonstrated to be a promising photoanode because of their good physical and chemical properties, large specific surface area, facile synthesis process, and high stability in acidic and alkaline solutions [30,31]. So, TiO 2 nanotubes are widely used as photoelectric catalytic electrode instead of TiO 2 film, and also usually used as novel and stable support for the noble metal catalysts. It has been demonstrated that Pt dopant can also improve the photoelectrochemical performance of TiO 2 nanotubes under visible light irradiation [32]. Up to now, TiO 2 nanotubes and Pt/ TiO 2 nanotubes have not been used in photocatalytic oxidation for As(III).
In this paper, constant current deposition [32] method was selected to synthesize Pt/TiO 2 nanotubes electrode, which was proved to have smaller band gap and stronger absorption in visible light region. These Pt/TiO 2 nanotubes materials were firstly tried to photoelectrocatalytic oxidation for As(III) in water driven by visible light. The photoelectrochemical oxidation performances of these materials for As(III) separately under visible light and sunlight were tested. The kinetics process of As(III) photoelectrochemical oxidation was analyzed to fit the pseudo-first-order reaction model equation. Real sample from Daying Village, Datong basin, Northern China, with high concentration of As(III) was tried by this system under visible light, and all As(III) was found to be transformed into As(V) in 200 min.

Reagents
All reagents were obtained from Sinopharm Chemical Reagent Co., Ltd. and were the highest grade available. All solutions and subsequent dilutions were prepared using deionized water from a scientific nanopure water purifier (Thermo fisher, America) with a resistivity of less than 0.055 μS/cm. A 1000 mL of As(III) standard solution (1000 mg/L) was prepared by dissolving 1.3203 g of As 2 O 3 in the minimum amount of 4.0 M NaOH and then adjusting pH to 3.0 with 1.0 M H 2 SO 4 .

Instruments
The morphology of the samples was studied with the use of a Hitachi SU8010 field emission scanning electron microscope (FE-SEM).
The analysis of the optical properties was performed on a U-4100 UV-vis spectrophotometer (Hitachi, Japan) in the region of 200-800 nm.
X-ray photoelectron spectroscopy (XPS) analysis was carried out to determine the surface properties of the catalysts using a Physical Electronics PHI model 5700 instrument (a RBD upgraded PHI-5000 C ESCA system, PerkinElmer, America), with Al X-ray source operating at 250 W. The takeoff angle of the sample to analyzer was 45°. Survey spectra were collected at pass energy (PE) of 187.85 eV over a binding energy range from 0 to 1300 eV. High binding energy resolution multiplex data for the individual elements were collected at a PE of 29.55 eV. During all XPS experiments, the pressure inside the vacuum system was maintained at 1 × 10 −9 Pa. Before the above analysis, all samples were dried under vacuum at 80°C overnight. Binding energies were calibrated by using the containment carbon (C1s = 284.6 eV).
To detect concentration of arsenic, an ELAN DRC II ICP-MS (PerkinElmer, America) equipped with an atomizer and a spray chamber was used. The ICP-MS normal operating parameters were as follows: RF power 1100 W, lens voltage 7.25 V, nebulizer gas flow rate 0.98 L/min, auxiliary gas flow rate 1.2 L/min, and plasma gas flow rate 15.00 L/min. Arsenic species were separated by Series 200 HPLC (PerkinElmer, America) with an automatic sample injector and directly introduced into ICP-MS. A C8 chromatographic column (PerkinElmer, America) was used with the mobile phase containing 1 mM tetrabutylammonium hydroxide, 0.05 mM dipotassium EDTA, and 0.05 % methanol (pH 6.8).

Preparation of TiO 2 Nanotubes and Pt/TiO 2 Nanotubes
Titania nanotubular membranes were fabricated from titanium foil of 0.30 mm thickness (99.9 % pure, Erli, China). Prior to membrane fabrication, the titanium foil was polished with abrasive paper for metallograph, and then ultrasonically cleaned with acetone, ethanol, and de-ionized water, separately for 15 min, and then dried. Then the cleaned titanium foil was set into an electrolyte composed with 0.3 wt.% ammonium fluoride and 2 vol.% water in ethylene glycol. Potentiostatic anodization was done at room temperature with titanium foil (2.0 cm × 3.8 cm) as anode and graphite plate (2.5 cm × 4.5 cm) as cathode. A GPC-6030D constant-voltage DC source (GWinstek, China) was used as the voltage source to drive the anodization. After electrochemical anodic oxidation at voltage of 30.0 V for 2 h, 3.7-μm thick layer of aligned amorphous TiO 2 nanotubes with 95 ± 5 nm diameters would be presented on Ti sheet. Prepared TiO 2 nanotubes were ultrasonically cleaned in deionized water for 1-2 min to remove surface debris. Then amorphous TiO 2 nanotubes layers were converted to the anatase phase by annealing at 450°C [33]. SEM images of prepared TiO 2 nanotubes were shown in Fig. 1. Figure 1a shows the morphology of unwashed TiO 2 nanotubes after annealing treatment; b-d give the topand cross-sectional view of washed TiO 2 nanotubes.
Pt electrodeposition was carried out by using a CS300 electrochemical workstation (Koster, China) with a standard three-electrode system. TiO 2 nanotubes served as the working electrode, an Ag/AgCl electrode and a graphite plate electrode served as the reference and counter electrode, respectively. Low negative current density (−0.  Although no obvious Pt particles were observed from the SEM results of Pt/TiO 2 nanotubes (Fig. 2), energy dispersive spectroscopy (EDS) analysis proved that Pt nanoparticles were existed and focused on the tube wall close to the nozzle (Table 1). Furthermore, from EDS results, the deposition amount of Pt nanoparticles were found steadily increased with the applied current density of Pt deposition.
To further confirm the composition of prepared Pt/ TiO 2 nanotubes, XPS was introduced to detect the surface composition of samples (as shown in Fig. 3). In Fig. 3a, the two peaks at 458.5 and 464.1 eV were assigned to the Ti (2p 3/2 ) and Ti (2p 1/2 ) states in Pt/ TiO 2 nanotubes, respectively [34]. According to literature, binding energy of Ti 4+ (2p 3/2 ) and Ti 3 + (2p 3/2 ) in titanium dioxide was 459 and 457 eV, respectively. The slight peak shift toward low energy suggested that the existence of small amount of Ti 3+ in the Pt/ TiO 2 nanotubes. The strong peak centered at 529.7 eV corresponded to O(1s) bonded to titanium (Fig. 3b). Compared with the standard O(1s) peak located at 530.0 eV in the XPS spectra of pure TiO 2 samples, the peak exhibited a 0.3 eV shift to lower energy, which were similar with results reported by Xing et al. [35]. Such shift can be attributed to the lack of oxygen in the Pt/TiO 2 nanotubes, and oxygen vacancies will be produced with the generation of Ti 3+ during the preparation process. Ti 3+ and oxygen vacancies could be generated in the anneal process of TiO 2 nanotubes and the Pt deposition process, because in the anneal process of TiO 2 nanotubes oxygen vacancies could not be fully eliminated in current fabrication procedures [36]; at the same time, partial Ti 4+ would be transformed into Ti 3+ during the deposition of Pt and Pb with the interaction between Pt/Pb and TiO 2 [37,38]. The two peaks located at 70.4 and 74.3 eV shown in Fig. 3c could be assigned to Pt (4f 7/2 ) and Pt (4f 5/2 ), respectively [39], which indicated that Pt was deposited on the TiO 2 nanotubes substrate successfully. The binding energy peaks at 70.4 and 74.3 eV are a little higher than that of free Pt   [40]. The two Pt 4f peaks could be divided into four separated peaks attributed to Pt 0 (4f 7/2 ), Pt 2+ (4f 7/2 ), Pt 0 (4f 5/2 ), and Pt 2+ (4f 5/2 ), and it was found that Pt 0 was the dominant species in Pt deposited on TiO 2 nanotubes [41]. In order to determine the photo-absorbance properties, the UV-vis diffuse reflectance spectra (DRS) of pure TiO 2 nanotubes and Pt/TiO 2 nanotubes was analyzed from 200 to 800 nm wavelengths, as shown in Fig. 4. TiO 2 nanotubes exhibited a photo-response in ultraviolet region with wavelengths below 390 nm, which could be attributed to intrinsic band gap of TiO 2 . The weak absorption of TiO 2 nanotubes within the visible light range could be ascribed to the scattering of light caused by pores or cracks in the nanotube arrays or the presence of oxygen vacancies and Ti 3+ species in the synthesized TiO 2 nanotubes. Previous researches [36] indicated that in the anneal process of TiO 2 nanotubes oxygen vacancies and Ti 3+ species could not be fully eliminated with the current fabrication procedures. Ti 3+ species could accelerate the formation of isolated defect energy level below the bottom of the conduction band (CB) of TiO 2 , and also absorbed visible light, which would excite and produce photo-generated electrons transforming from Ti 3+ states to CB of TiO 2 . The weakly visible light absorption of TiO 2 nanotubes further indicated that oxygen vacancies and Ti 3+ species probably occurred during the anneal process of TiO 2 nanotubes. When Pt nanoparticles were loaded on TiO 2 nanotubes, the photoabsorption amount of the catalyst in visible light region increased and the amount of photoabsorption in the ultraviolet light range decreased. This result was similar to the findings of previous investigations [31]. Compared with pure TiO 2 nanotubes, the photosensitivity of Pt/TiO 2 nanotubes in the visible and near visible light wave range increased, because of localized surface plasmon resonance (LSPR) of Pt nanoparticles on the pore-wall of TiO 2 nanotubes. These results proved, when Pt nanoparticles were loaded on TiO 2 nanotubes as inorganic sensitizer, the LSPR of Pt nanoparticles promoted the separate efficiency of  photogenerated charges and extended the range of the excited spectrum. According to XPS spectrum results, it was demonstrated that oxygen vacancies and Ti 3+ species were present in Pt/TiO 2 nanotubes, which induced broad visible light absorption of Pt/ TiO 2 nanotubes. So, Pt/TiO 2 nanotubes can be tested under visible light to oxidate As(III).

Photocatalytic Activity Tests
Photocatalytic activities for As(III) oxidation were conducted in a 50-ml quartz beaker. The initial As(III) concentration was fixed at 3.4 mg/L, and the pH was adjusted with H 2 SO 4 or NaOH solution to the desired value. Prior to As(III) oxidation, TiO 2 nanotubes was added in the solution and kept for 30 min to allow equilibrium adsorption of arsenite on TiO 2 nanotubes. UV light irradiation was applied by a 175-W high-pressure mercury lamp, and visible light source was a 300-W halogen lamp (Philips, Holland) equipped with a wavelength cutoff filter for λ ≤ 420 nm. Water samples were withdrawn by a 1.0 mL pipette intermittently during photoreaction and filtered through 0.22-μm PTFE filters (Millipore). Duplicate or triplicate experiments were performed for each set.

Photoelectrocatalytic Activity Tests
Photoelectrocatalytic oxidation of As(III) was performed in a self-made divided electrolytic tank (Fig. 5a). The anode tank and the cathode tank were isolated, and formed a circuit by a salt bridge. The CS300 electrochemical workstation (Koster, China) was employed to provide constant positive bias voltages, meanwhile, recorded the corresponding current. The Pt/TiO 2 nanotubes served as working electrodes, with 2.0 × 3.8 cm 2 area. A saturated calomel electrode and a graphite rod served as reference electrode and auxiliary electrode, respectively. A 50.0 mL electrolyte was comprised of 0.1 M Na 2 SO 4 (as supporting electrolyte) and As(III) with 2.0, 2.8, 3.4, 4.0, 5.0, and 6.0 mg/L initial concentration. Prior to As(III) oxidation, Pt/TiO 2 nanotubes working electrode was kept in the electrolyte under darkness for 30 min to ensure adsorption equilibrium. The light source was provided by the 300-W halogen lamp (Philips, Holland) in full wavelength range with illumination intensity around 453 mW cm −2 (Fig. 5b). The photocatalytic activity under visible light irradiation (the 300-W halogen lamp) was tested with a cutoff filter to get rid of UV irradiation below 420 nm. To avoid the heating effect caused by the infrared irradiation, the quartz cell was cooled down by circulating water.

Photocatalytic Oxidation for As(III) by Pt/TiO 2 Nanotubes Prepared with Different Current Density and Pt Deposition Time
The effect of Pt loading time on photocatalytic oxidation As(III) was tested at 25°C constant temperature in 3.4 mg/L As(III) solution under visible light for 360 min as shown in Fig. 6a. When the Pt loading time were 2.5, 5, 10, and 20 min, the percentages of final dissolved As(V) in system were 77.3, 83.9, 88.1, and 85.9, respectively. Results showed that with the increase of Pt loading time, the oxidation rate of As(III) first increased, then decreased with extensive loading. Considering the economic reason and As(III) oxidation efficiency, the optimal Pt loading time in the following experiments was focused at 5 min. Figure 6b shows the effect of Pt deposition current density in As(III) photocatalytic oxidation process. When current density of the Pt deposition was increased from 0.2 to 0.8 mA cm −2 , percentages of final dissolved As(V) were varied from 74.1 to 83.8 %. Concentration of generated As(V) first increased with the increase of applied current density of Pt deposition when it was below 0.4 mA cm −2 . This is because both the valence state of Pt loaded on Pt/TiO 2 nanotubes, and the deposition quantity will increase with the applied current density. Fig. 5 a Setup of the photoelectrocatalytic system: 1 CS300 electrochemical workstation, 2 halogen lamp, 3 optical filter, 4 salt bridge, 5 graphite rod, 6 TiO 2 nanotubes electrode, 7 reference electrode; b spectral distribution of the 300-W halogen lamp with and without optical filter Photocatalytic activity of platinized TiO 2 was arranged in the order of Pt (0)/TiO 2 > PtOx (II, IV)/TiO 2 > bare TiO 2 [42]. When applied current density increased to 0.5 mA cm −2 , the photocatalytic ability of Pt/TiO 2 nanotubes for As(III) oxidation was kept stable. To the following experiments, the applied current density was kept at 0.5 mA cm −2 .

Photocatalytic Ability Comparison Between Naked TiO 2 Nanotubes and Pt/TiO 2 Nanotubes
To prove the function of Pt for photocatalysis, oxidation abilities of As(III) were compared between TiO 2 and Pt/ TiO 2 nanotubes under visible light irradiation or visible light irradiation with 1.2 V positive biasing. From Fig. 7a, we could find that under visible light, no As(III) was oxidized by TiO 2 nanotubes, no matter if 1.2 V positive biasing was applied. While, under ultraviolet light, 82.0 % of As(III) could be oxidized to As(V) after 30 min. This means, only under ultraviolet light condition, TiO 2 nanotubes have photocatalytic oxidation ability for As(III).
To Pt/TiO 2 nanotubes electrodes prepared at 0.5 mA cm −2 with 5 min, they displayed high photocatalytic and photoelectrocatalytic oxidation activity for As(III) (Fig. 7 b). To avoid the influence of adsorption  effect of Pt/TiO 2 nanotubes, 30 min equilibrium adsorption was first operated before catalytic experiments, which made As(III) concentration decrease from 3.41 to 3.20 mg/L. When both electrochemical and photocatalytic processes were simultaneously applied, 94.2 % of As(III) could be oxidized in 280 min. This value was 13.5 % higher than the only photocatalytic oxidation process. Fabricated Pt nanoparticles were acted as electron traps, which could enhance the separation of electron-hole pairs, and the external positive biasing drove electron (e − ) to cathode, then the recombination of electron-hole pairs could be further reduced. So, more holes could cause stronger direct (h + ) oxidation or indirect (HO·) oxidation for As(III) on anode Pt/TiO 2 nanotubes electrode, which was the reason why the As(III) oxidation rate on anode in the photoelectrocatalytic process was higher than that in the photocatalytic process. In addition, an interesting phenomenon was found on Pt/TiO 2 nanotubes, As(III) even could be oxidized in dark condition. And 17.4 % of As(III) could be converted into As(V) in 280 min, this could be induced by catalytic effect of platinum itself, and O 2 activation on Pt nanoparticles might be responsible for this dark activity. When 1.2 V of positive biasing was applied under dark condition, the oxidation efficiency of As(III) to As(V) was hardly improved.
With detection solution in cathode cell, it was interesting to find that As(III) was also continuously transformed into As(V) during the photoelectrocatalytic process at reduction potential. When 1.2 V of positive bias potential was applied to this system, the conversion rate in cathode cell was 66.4 % after 280 min (Fig. 7d). Figure 8a showed the oxidation rate of As(III) increased obviously with the rise of initial As(III) concentration.

Effect of Initial As(III) Concentration on Photoelectrocatalytic Result of Pt/TiO 2 Nanotubes Electrode
Furthermore, the kinetics simulation curves of As(III) photoelectrocatalytic oxidation were summarized and presented in Fig. 8b. All oxidation reactions were well fitted in the pseudo-first-order kinetics model.
where C is the concentration of As(III) (mg/L), t is the reaction time (min), and k app is the apparent first order reaction constant (min −1 ). The values of k app and the regression correlation coefficient R 2 of Pt/TiO 2 nanotubes for As(III) photoelectrocatalytic oxidation were listed in Table 2. Correlation coefficient (R 2 ) values of pseudofirst-order kinetic model with different initial As(III) concentration were all more than 0.980. This meant that photoelectrocatalytic oxidation process of As(III) on Pt/TiO 2 nanotubes obeyed pseudo-first-order kinetics equation.

The Influence of Applied Bias Potentials on the Photoelectrocatalytic Oxidation of As(III)
The bias potential is an important parameter in the process of photoelectrocatalytic activity. Photo-generated electrons on Pt/TiO 2 nanotubes electrode could be driven to the counter electrode with positive potential. So, bias potentials ranged from 0.0 to 2.0 V were monitored over 280 min in photoelectrocatalytic oxidation treatment (Fig. 9). Figure 9a illustrated that As(III) oxidation rate increased with the increase of applied positive bias. With applied 0.0 to 2.0 V positive bias, the plot of C t versus time was fitted with exponential decay equation with all R 2 exceeding 0.980. The apparent As(III) oxidation rate constants were varied from 7.2 × 10 −3 to 12.4 × 10 −3 min −1 . Applying a positive biasing to the Pt/TiO 2 nanotubes electrode can transfer the photo-generated electrons away from photo-generated Fig. 8 a As(V) concentration in anodic cell and b kinetics simulation of As(III) oxidation in the photoelectrocatalytic oxidation process with varies initial As(III) concentration holes on the Pt/TiO 2 nanotubes electrode via the external circuit; thus, the recombination of photogenerated electron-hole pairs is minimized [43,44].
In Fig. 9a, it was found that in the range of studied positive biasing, rate constant linearly went up with the increase of applied bias voltage. With the positive bias voltage increasing, more and more photogenerated electrons moved to counter electrode. As a result, the photo-generated electrons and holes were well separated, thus more hydroxyl radicals (HO·) could be produced by H 2 O oxidized in the holes [45]. Which species (h + , HO· and ·O 2 − ) was mainly responsible for the oxidation from As(III) to As(V) in the UV/TiO 2 system, different opinions had been proposed on this issue. But so far, it still remained as a controversial issue [46]. So, on the anode Pt/TiO 2 nanotubes electrode, it was in dispute whether h + or HO· was responsible for As(III) oxidation.
At the same time, the oxidation rate of As(III) to As(V) on graphite rod in cathodic cell was also verified to increase with applied potential (Fig. 9b). After 280 min, As(V) concentration were 0, 0.84, 0.96, 1.69, 2.15, 3.02, and 3.39 mg/L when system was applied with 0.0, 0.3, 0.6, 0.9, 1.2, 1.5, and 2.0 V positive bias potentials. As(III) in cathodic cell could be completely transformed into As(V) after 280 min when system was applied with 2.0 V positive bias potential. Leng et al. found that the main product of oxygen reduction reaction was hydrogen peroxide (H 2 O 2 ) on the cathode graphite in aqueous solution with pH from 2 to 12 [47]. Nevertheless, ·O 2 − was only stable in concentrated alkaline solutions or aprotic media. So, it was inferred that eon cathode electrode could be trapped by the surface absorded O 2 to generate H 2 O 2 as Eq.
The Effect of Solution pH Value on As(III) Oxidation The influence of initial pH on photocatalytic oxidation of As(III) in anodic cell was shown in Fig. 10a. The oxidation rate in alkaline environment was mostly a little faster than the rate in acid environment. The result differed from that of Lee and Choi [48] who found that the initial oxidation rate at pH 9 was about twice as fast as the rate at pH 3, and also differed with the result of Bissen et al. [13] and Sharma et al. [49] who both found that the small increase in the oxidation of As(III) with increasing pH was within experimental error. The potential of As(V)/As(III) couple is much less positive than the valence band potential of TiO 2 , so, the photo-generated holes have enough thermodynamic potential to oxidize As(III) to As(V) [50]. The potential of As(V)/As(III) couple in alkaline environment is lower than that in acid, which may be contributed to the increase of the oxidation rate with the increase of pH.  The pH change during the photoelectrocatalytic oxidation of As(III) in aqueous solution was displayed in Fig. 10b. During the electrochemical experiment, pH decreased with irradiation time during both photocatalysis and photoelectrocatalysis experiments. The decrease of pH in the photoelectrocatalysis experiment was faster than that in the photocatalysis.
The influence of pH on the speciation of arsenic oxoanions can easily be obtained from Eq. (3) to (8), acid dissociation constants are also given.
In the photocatalytic process, As(III) was oxidized to As(V), meanwhile, protons were produced, and the reaction could be described as In photoelectrocatalytic process, the anode oxidation caused O 2 evolution in anode cell, which produced protons and contributed to the further reduce of the pH during photoelectrocatalytic activity.

Photoelectrocatalytic Oxidation of As(III) by Natural Sunlight
To further extend the photoelectrocatalytic applicability of Pt/TiO 2 nanotube arrays electrode in a more practical situation, the photoelectrocatalytic oxidation ability was also evaluated under natural sunlight. Figure 11a showed the evolution of As(III)/As(V) concentration  with irradiation time with initial 3.4 mg/L As(III) and 1.2 V positive bias potential under natural sunlight irradiation. The photoelectrocatalytic oxidation process under natural sunlight accorded with zero-order kinetics law (C t = − 0.0114t + 2.9971(R 2 = 0.973)), different with process under visible light which followed firstorder kinetics law. As(III) could be completely oxidized after 280 min under natural sunlight irradiation. While, to visible light, it needed 360 min to totally oxidize As(III). This might be because of the effect of ultraviolet light in sunlight.

Photoelectrocatalytic Oxidation of As(III) in High Arsenic Groundwater Sample Under Visible Light
The photoelectrocatalytic activity of real groundwater sample with high arsenic was also investigated by Pt/ TiO 2 nanotubes arrays electrode under visible light. The high arsenic underground water sample (25 m underground) was collected from Daying Village, Datong basin, Northern China, concentrations of main ions in this water sample was determined as follows: Na + 324.2 mg/L, Ca 2+  conductivity. After 30 min for equilibrium adsorption, 9.5 % of As(III) and 14.2 % of As(V) were adsorbed on Pt/TiO 2 nanotube arrays. The conversion of As(III) during the photoelectrocatalytic process with 1.2 V of positive bias potential was found to follow a zero-order kinetics law as C t = − 0.0015t + 0.3013 (R 2 = 0.981), and As(III) in groundwater sample was totally transformed into As(V) in 200 min under visible light. This proved Pt/TiO 2 nanotubes electrode was an efficient photoelectrocatalytic material for As(III) oxidation and could be used in high arsenic underground water pretreatment (Fig. 12).

Conclusions
To promote photocatalytic oxidation of As(III) by TiO 2 materials under visible light, Pt/TiO 2 nanotubes were introduced via a constant current deposition method. First, the prepared nanotubes were detected by SEM, XPS to prove the existing of Pt, and UV-vis diffuse reflectance spectra results proved that Pt could promote separate efficiency and extend the excited spectrum range. Pt/TiO 2 nanotubes presented more efficient photoelectrocatalytic oxidation performance for As(III) than TiO 2 nanotubes. Even under dark condition, it was also useful in As(III) photoelectrocatalytic oxidation. Furthermore, the photoelectrocatalytic oxidation process under visible light was found to obey pseudo-first-order kinetics. The prominent conversion from As(III) to As(V) in cathodic cell also occurred because of the production of H 2 O 2 from electrons trapping by O 2 on cathode. While to natural sunlight, the oxidation of 3.4 mg/L As(III) on Pt/ TiO 2 nanotubes electrode with 1.2 V applied voltage followed zero-order kinetics law, and its oxidation rate was slightly higher than that of As(III) under visible light. As(III) in real groundwater sample could be totally transformed into As(V) in 200 min by Pt/TiO 2 nanotubes electrode with 1.2 V under visible light and also accorded with zero-order kinetics.