Efficient Degradation of Refractory Organics Using Sulfate Radicals Generated Directly from WO 3 Photoelectrode and the Catalytic Reaction of Sulfate

An environment-friendly method of efficiently degrading refractory organics using SO4• generated directly from a WO3 photoelectrode and a catalytic reaction of sulfate was proposed, in which the cycling process of SO4 → SO4• → SO4 was achieved in the treatment of organic pollutants without any other activator and without the continuous addition of sulfate. The results show that the removal efficiency for a typical refractory organics of methyl orange (MO) with 5 mg/L was up to 95% within 80 min, and merely 3% by photolysis and 19% by photocatalysis, respectively, under similar conditions. The rate constant for the disposal of MO at pH 2, in which SO4• instead of HO• is the main oxidizer confirmed by radical scavenger experiment, is up to 5.21 × 10−4 s−1, which was ~6.6 times that (7.89 × 10−5 s−1) under neutral condition, in which HO• is the main oxidizer. The concentration of active persulfate (S2O8, SO5, and SO4•) species at pH 2 was up to 0.38 mM, which was ~16-fold as much as that (0.023 mM) in neutral conditions. The method provides a new approach for the treatment and resource utilization of sulfate wastewater.


Introduction
Peroxydisulfate (S 2 O 8 2− ), as a peroxide that possesses the highest oxidizability, can play a role in various chemical fields [1].Recently, S 2 O 8 2− has received intense scrutiny in water purification and soil remediation since the activation of S 2 O 8 2− is one of the most effective methods of producing sulfate radical (SO 4 − •) for the degradation of organic waste [2].SO 4 − • with one unpaired electron has a high oxidation potential of 2.6~3.2V vs. the normal hydrogen electrode (NHE), and more importantly has a longer lifetime (30-40 µs) than hydroxyl radical (HO•), so it is highly active to react with various contaminant, causing either partial or complete mineralization [3][4][5].Currently, SO 4 − • is commonly produced from the activation of peroxydisulfate via heat [6][7][8], ultraviolet (UV) light [9,10], transition metals [11][12][13], or other means [14].However, there are still some challenges for the application of SO 4 − • with these methods.The first problem is that a large amount of expensive peroxydisulfate must be used to form the highly reactive SO 4 − •, leading to a high water treatment cost.The other is that the efficiency of activated persulfate in practical systems is relatively low because only one SO 4 − • from the activation of the peroxydisulfate can be obtained to oxidize organic compounds, and the other free radical was consumed by the activator, which might cause new secondary environment pollution.Therefore, how to cost-effectively produce and utilize SO 4 − • has become an urgent issue.
Photoelectrocatalytic (PEC) techniques as a promising advanced oxidization technology has attracted much attention in pollutant degradation because the photoelectrodes can generate free radicals under light illumination that can rapidly decompose organic compounds [15].Some photoanode materials, such as TiO 2 [16][17][18][19], WO 3 [20][21][22][23][24][25][26][27][28][29], Fe 2 O 3 [30,31], ZnO [32,33], and BiVO 4 [34][35][36] have been widely investigated because of their considerably high incident light-to-current conversion efficiencies.Among them, WO 3 is a promising material impacting many research fields [37] for its specific properties.For instance, it has a relatively narrow band gap of 2.5-2.7 eV, which can absorb approximately 12% of solar light with a theoretical maximum conversion efficiency of ~6.3%.Additionally, WO 3 possesses a moderate hole diffusion length (~150 nm) that is longer than those of α-Fe 2 O 3 (2-4 nm) and TiO 2 (~100 nm), thus showing inherently good electron transport properties.Especially, the valence band edge of WO 3 is located at approximately 3.0 V vs. NHE [38], which provides a sufficient potential for the oxidation of sulfate to form SO 4 − • under visible light irradiation.This property gives us an inspiration that it is possible to find a new way to produce SO 4 − • using a WO 3 photoanode in a PEC system.
Refractory organics are usually difficult to decompose by organisms in natural conditions.Traditional treatment methods, such as biological methods and conventional physical and chemical methods, do not easily deal with refractory pollutants.This paper proposes a novel environmentally friendly method of efficiently degrading refractory organics using SO 4 − • generated directly from the catalysis of sulfate by a WO 3 photoelectrode, in which SO 4 − • can be favorably generated as a result of photocatalytic oxidation of sulfate, and a cycling process of SO 4 2− → SO 4 − • → SO 4 2− was achieved in the treatment of organic pollutants without the continuous addition of sulfate.SO 4 − • instead of HO• is the main oxidizer in the system, so the degradation of organics is mainly accomplished by SO 4 − •, and no other activator needed to be added.In practice, a large amount of sulfate-rich wastewater is produced from medicine, chemical engineering, cement and dyeing, food, and fertilizer industries [39], which represents a potential threat to water ecological and human health [40].Therefore, the proposed method provides a new way for the disposal and resource utilization of sulfate-rich organic wastewater.

Characterization of WO 3 Photoanodes
The obtained WO 3 films were transparent, uniform, and smooth on an F-doped tin oxide (FTO) substrate.Figure 1 shows a scanning electron microscope (SEM) image and an X-ray diffractometry (XRD) pattern of a typical WO 3 nanoplate array film.As is shown in Figure 1a, the prepared WO 3 films had platelike morphologies, and the average thickness of the plates was ~100 nm.The cross-sectional SEM images in Figure 1a show that the WO 3 films were perpendicular to the substrate and tightly adhered to the FTO substrate and the thickness of the WO 3 films remained almost the same at ∼850 nm.Those results are in agreement with the reported [41].Figure 1b shows the X-ray diffractometry (XRD) patterns of WO 3 nanoplate array films, where (020), (002), and (200) peaks were observed.The (020) and (002) diffraction peaks were relatively low, which indicates that the (020) direction and (002) direction were parallel to the substrate.Furthermore, the (200) direction was vertical to the substrate, as evidenced by the fact that the (200) peak exhibited the highest intensity.

PEC Degradation of Methyl Orange in Sulfate Medium
Figure 2a shows the PEC degradation of 5 mg/L methyl orange (MO) within 80 min by photolysis, electrocatalysis (EC), photocatalysis (PC), and PEC in Na2SO4 solution (pH = 2, regulating the pH value with H2SO4), respectively.As can be seen, the photolytic reaction only resulted in a removal ratio of 3% within 80 min, and 18% of the MO was removed in electrocatalytic (EC) conditions at the same time.The photocatalytic reaction (without bias supply), compared to photolysis, degraded 19% of the MO within 80 min, while the PEC process obtained a removal rate of 95% under the same conditions.The significant advance of PEC over the other methods (photolysis and photocatalysis) is attributed to the electric bias that promoted the separation of photogenerated holes and electrons from recombination, and a more powerful oxidizer was probably produced in the PEC process.
Figure 2b shows the PEC degradation of 5 mg/L MO under different concentrations of Na2SO4 solution (0.1 M, 0.5 M, 0.7 M, 0.8 M, and 1 M) at pH = 2. Apparently, the removal rate of MO increased with the concentration of sulfate.Only 53% of methyl orange was removed in 0.1 M Na2SO4 within 80 min, while the removal rate reached a substantial level, 84%, 90%, and 95%, respectively, when the concentration of sulfate increased from 0.1 M to 0.5 M, 0.7 M, and 0.8 M (1 M).This result suggests that the concentration of sulfate has a direct effect on the degradation of methyl orange, indicating that the degradation of methyl orange is likely not attributed to hydroxyl radicals, but more likely new oxidizing species that are connected with sulfate.As is known, the valence band edge of WO3 is located at approximately 3.0 V vs. NHE [38], theoretically, which could provide a sufficient potential to oxidize sulfate to SO4 − • under visible light irradiation [1,38,42,43].Furthermore, the preferential oxidation of HSO4 − at WO3/electrolyte contacts and the generation of peroxydisulfate (S2O8 2− ) on a WO3 photoanode in sulfuric acid was reported by Lewis et al. [42,43].Therefore, we speculate that the fast degradation of methyl orange in 0.8 M Na2SO4 is possibly due to the active SO4 − • generated on the WO3 surface.With the increase of sulfate concentration, more sulfate oxidation occurred on WO3 photoanondes [38,42,43] and water oxidation is suppressed.Thus, a large amount of SO4 − • was produced such that the removal rate of MO was enhanced.Noting that, in 0.8 M and 1 M sulfate, removal rates were almost identical as shown in Figure 2b, the yield of SO4 − • radicals reached the limit with over 0.8 M sulfate.
Figure 2c presents the degradation results of 5 mg/L MO measured at varying pH levels (pH 0, 1, 2, 3, and 7, the pH value being regulated with H2SO4).Obviously, the removal rate of MO was quicker at lower pH values.As can be seen, 95% of MO was degraded at pH 2.0, while the removal rate of MO was merely 32% at pH 7.0 over 80 min.It was also noted that the removal rate sharply increased as pH value decreased from 3.0 to 2.0.It was found that only 47% of MO was removed at pH 3.0, which is a small increment over the 32% under neutral condition (pH 7.0).However, there was a 100% and 200% increase in removal rate at 2.0 compared to that at 3.0 and 7.0, respectively, indicating that stronger acid conditions are favorable for the formation of SO4 − •.This is because the

PEC Degradation of Methyl Orange in Sulfate Medium
Figure 2a shows the PEC degradation of 5 mg/L methyl orange (MO) within 80 min by photolysis, electrocatalysis (EC), photocatalysis (PC), and PEC in Na 2 SO 4 solution (pH = 2, regulating the pH value with H 2 SO 4 ), respectively.As can be seen, the photolytic reaction only resulted in a removal ratio of 3% within 80 min, and 18% of the MO was removed in electrocatalytic (EC) conditions at the same time.The photocatalytic reaction (without bias supply), compared to photolysis, degraded 19% of the MO within 80 min, while the PEC process obtained a removal rate of 95% under the same conditions.The significant advance of PEC over the other methods (photolysis and photocatalysis) is attributed to the electric bias that promoted the separation of photogenerated holes and electrons from recombination, and a more powerful oxidizer was probably produced in the PEC process.
Figure 2b shows the PEC degradation of 5 mg/L MO under different concentrations of Na 2 SO 4 solution (0.1 M, 0.5 M, 0.7 M, 0.8 M, and 1 M) at pH = 2. Apparently, the removal rate of MO increased with the concentration of sulfate.Only 53% of methyl orange was removed in 0.1 M Na 2 SO 4 within 80 min, while the removal rate reached a substantial level, 84%, 90%, and 95%, respectively, when the concentration of sulfate increased from 0.1 M to 0.5 M, 0.7 M, and 0.8 M (1 M).This result suggests that the concentration of sulfate has a direct effect on the degradation of methyl orange, indicating that the degradation of methyl orange is likely not attributed to hydroxyl radicals, but more likely new oxidizing species that are connected with sulfate.As is known, the valence band edge of WO 3 is located at approximately 3.0 V vs. NHE [38], theoretically, which could provide a sufficient potential to oxidize sulfate to SO 4 − • under visible light irradiation [1,38,42,43].Furthermore, the preferential oxidation of HSO 4 − at WO 3 /electrolyte contacts and the generation of peroxydisulfate (S 2 O 8 2− ) on a WO 3 photoanode in sulfuric acid was reported by Lewis et al. [42,43].Therefore, we speculate that the fast degradation of methyl orange in 0.8 M Na 2 SO 4 is possibly due to the active SO 4 − • generated on the WO 3 surface.With the increase of sulfate concentration, more sulfate oxidation occurred on WO 3 photoanondes [38,42,43] and water oxidation is suppressed.Thus, a large amount of SO 4 − • was produced such that the removal rate of MO was enhanced.Noting that, in 0.8 M and 1 M sulfate, removal rates were almost identical as shown in Figure 2b, the yield of SO 4 − • radicals reached the limit with over 0.8 M sulfate.Figure 2c presents the degradation results of 5 mg/L MO measured at varying pH levels (pH 0, 1, 2, 3, and 7, the pH value being regulated with H 2 SO 4 ).Obviously, the removal rate of MO was quicker at lower pH values.As can be seen, 95% of MO was degraded at pH 2.0, while the removal rate of MO was merely 32% at pH 7.0 over 80 min.It was also noted that the removal rate sharply increased as pH value decreased from 3.0 to 2.0.It was found that only 47% of MO was removed at pH 3.0, which is a small increment over the 32% under neutral condition (pH 7.0).However, there was a 100% and 200% increase in removal rate at 2.0 compared to that at 3.0 and 7.0, respectively, indicating that stronger acid conditions are favorable for the formation of SO 4 − •.This is because the theoretical potential of water oxidation (Equation ( 1)) [43] is higher than that of sulfate oxidation under stronger acid conditions (pH < 2).The sulfate could be oxidized to form SO 4 − • in the PEC system, which subsequently promoted the removal of MO.With the increase in pH value (pH > 2), the potential of water oxidation continuously decreases, while the potential of sulfate oxidation was kept stable (Equation ( 2)) [1,44,45].Therefore, water oxidation effectively competes with oxidation of sulfate at the WO 3 /electrolyte interface and gradually dominates the photoanodic behavior of WO 3 .In this case, the degradation of MO was hindered.Another phenomenon is that a further decreasing pH value (1, 0) did not show significant improvements in the removal rate of methyl orange.It is suggested that the yield of SO 4 − • on WO 3 electrodes reached a maximum level at pH 2.
Catalysts 2017, 7, 346 4 of 12 theoretical potential of water oxidation (Equation ( 1)) [43] is higher than that of sulfate oxidation under stronger acid conditions (pH < 2).The sulfate could be oxidized to form SO4 − • in the PEC system, which subsequently promoted the removal of MO.With the increase in pH value (pH > 2), the potential of water oxidation continuously decreases, while the potential of sulfate oxidation was kept stable (Equation ( 2)) [1,44,45].Therefore, water oxidation effectively competes with oxidation of sulfate at the WO3/electrolyte interface and gradually dominates the photoanodic behavior of WO3.
In this case, the degradation of MO was hindered.Another phenomenon is that a further decreasing pH value (1, 0) did not show significant improvements in the removal rate of methyl orange.It is suggested that the yield of SO4 − • on WO3 electrodes reached a maximum level at pH 2. Additionally, degradation kinetics of MO by photoelectrocatalysis in different sulfate concentrations or in different pH value concentrations were investigated.The results show that the decolorization of MO followed the pseudo-first-order kinetic within 70 min, shown in Figure 3a, which can be expressed by Equation (3): where k is the apparent reaction constant, and C0 and Ct are the initial concentration and the concentration at time t, respectively.Consistent with results in Figure 2b, the apparent rate constant for the photoelectrocatalysis system reached the limit with over 0.8 M sulfate.The apparent rate constant in 1 M Na2SO4 was 5.20 × 10 −4 s −1 , which was ~3.3 times that (1.58 × 10 −4 s −1 ) in 0.1 M Na2SO4.Additionally, degradation kinetics of MO by photoelectrocatalysis in different sulfate concentrations or in different pH value concentrations were investigated.The results show that the decolorization of MO followed the pseudo-first-order kinetic within 70 min, shown in Figure 3a, which can be expressed by Equation (3): where k is the apparent reaction constant, and C 0 and C t are the initial concentration and the concentration at time t, respectively.Consistent with results in Figure 2b, the apparent rate constant for the photoelectrocatalysis system reached the limit with over 0.  Degradation kinetics of MO by photoelectrocatalysis in different pH value concentrations also obeyed pseudo-first-order kinetic within 70 min, shown in Figure 3b.The apparent rate constant at pH 2 is 5.21 × 10 −4 s −1 , which was ~6.6 times that (7.89 × 10 −5 s −1 ) under neutral conditions.

Identification of Main Active Radical in the PEC System
As previously discussed, SO4 − • could be the primary active species for the degradation of MO.However, hydroxyl radicals (HO•) originated from the oxidation of water by photogenerated holes is a common oxidant that can react with methyl orange in aqueous solution [43].Therefore, the degradation of methyl orange in this study may have been achieved by SO4 − • and HO•.In order to distinguish the effect of the two free radicals, excess methanol (MA) and tert-butyl alcohol (TBA) were added in the solution, respectively.Methanol is a well-known quenching agent for both sulfate and hydroxyl radicals, and reaction rate constant of methanol with sulfate radical is similar to that with hydroxyl radicals [46].In contrast to MA, tert-butyl alcohol (TBA) with no alpha hydrogen is an effective quenching agent for hydroxyl radicals, which is 1000-fold faster than with sulfate radical [46,47].
Figure 4 presents the removal of MO in 1 M sulfate (pH = 2, 3, and 7) when the MA or TBA was added or not added, respectively.As is shown in Figure 4a, without any radical quenching agent, the removal rate of MO reached 95% but only 30% of MO was removed with the addition of 2 mM MA.This indicates that MA quenched almost all radicals in the solution so that redox was confined to the interface between electrode and electrolyte, thus greatly depressing the degradation of MO.Although the addition of TBA still hindered the degradation of MO, the removal rate increased to 57% compared to 30% with the use of MA.Since TBA is prone to react with hydroxyl radicals as previously mentioned, more SO4 − • could be preserved and reacted with MO.This result confirmed that SO4 − • might be the main oxidant resulting in the degradation of MO.It was noted that the removal rate was not more than 60% in 80 min.We speculated that TBA competed with sulfate for photoholes on WO3 photoanode and that partial TBA could still react with SO4 − •.As is shown in Figure 4b, the degradation of MO (30%) with an addition of 2 mM TBA was better than that (21%) with an addition of 2 mM MA at pH = 3, indicating that SO4 − • still emerged in the WO3 photoanode to react with MO.Under neutral conditions, shown in Figure 4c, the degradation of MO was greatly depressed by MA or TBA, indicating that hydroxyl radicals can be used as the main oxidant in the degradation of MO.Degradation kinetics of MO by photoelectrocatalysis in different pH value concentrations also obeyed pseudo-first-order kinetic within 70 min, shown in Figure 3b.The apparent rate constant at pH 2 is 5.21 × 10 −4 s −1 , which was ~6.6 times that (7.89 × 10 −5 s −1 ) under neutral conditions.

Identification of Main Active Radical in the PEC System
As previously discussed, SO 4 − • could be the primary active species for the degradation of MO.
However, hydroxyl radicals (HO•) originated from the oxidation of water by photogenerated holes is a common oxidant that can react with methyl orange in aqueous solution [43].Therefore, the degradation of methyl orange in this study may have been achieved by SO 4 − • and HO•.In order to distinguish the effect of the two free radicals, excess methanol (MA) and tert-butyl alcohol (TBA) were added in the solution, respectively.Methanol is a well-known quenching agent for both sulfate and hydroxyl radicals, and reaction rate constant of methanol with sulfate radical is similar to that with hydroxyl radicals [46].In contrast to MA, tert-butyl alcohol (TBA) with no alpha hydrogen is an effective quenching agent for hydroxyl radicals, which is 1000-fold faster than with sulfate radical [46,47].Figure 4 presents the removal of MO in 1 M sulfate (pH = 2, 3, and 7) when the MA or TBA was added or not added, respectively.As is shown in Figure 4a, without any radical quenching agent, the removal rate of MO reached 95% but only 30% of MO was removed with the addition of 2 mM MA.This indicates that MA quenched almost all radicals in the solution so that redox was confined to the interface between electrode and electrolyte, thus greatly depressing the degradation of MO.Although the addition of TBA still hindered the degradation of MO, the removal rate increased to 57% compared to 30% with the use of MA.Since TBA is prone to react with hydroxyl radicals as previously mentioned, more SO 4 − • could be preserved and reacted with MO.This result confirmed that SO 4 − • might be the main oxidant resulting in the degradation of MO.It was noted that the removal rate was not more than 60% in 80 min.We speculated that TBA competed with sulfate for photoholes on WO 3 photoanode and that partial TBA could still react with SO 4 − •.As is shown in Figure 4b, the degradation of MO (30%) with an addition of 2 mM TBA was better than that (21%) with an addition of 2 mM MA at pH = 3, indicating that SO 4 − • still emerged in the WO 3 photoanode to react with MO.Under neutral conditions, shown in Figure 4c, the degradation of MO was greatly depressed by MA or TBA, indicating that hydroxyl radicals can be used as the main oxidant in the degradation of MO.SO4 − • plays an important role in the degradation of MO, but it is difficult to direct identify the SO4 − • in the whole process because of a short lifetime of sulfate radical.In fact, SO4 2− can be oxidized to S2O8 2− (Equation ( 2)) [43] and SO4 − • could be an intermediate species that is prior to the formation of S2O8 2− (Equation ( 4)).Therefore, the amount of SO4 − • could be indirectly analyzed by investigating the production of S2O8 2− .In fact, a small part of S2O8 2− in 1.0 M Na2SO4 was hydrolyzed to produce HSO5 − , which has an effect on the determination of S2O8 2− via colorimetry, but the hydrolysis of S2O8 2− is so slow that it can be neglected [43].Thus, we assumed here that the concentration of S2O8 2− was also equal to the total active persulfate (S2O8 2− , SO5 2− , SO4 − •) species, which can be represented by the result determined by colorimetry.
Figure 4d shows the concentration of S2O8 2− with time in various pH values.At different pH values, the amount of S2O8 2− maintains a steady rise, indicating that the yield of S2O8 2− remains relatively constant and decreases gradually over time at the same pH value within 80 min.Simultaneously, pH values have a great effect on the generation of S2O8 2− .The amount of S2O8 2− rapidly decreased as pH increased, which suggests that the photoholes on the surface of the WO3 photoanode tended to oxidize water molecules (Equation ( 1)) instead of the sulfate anion at high pH values [38].
In order to further confirm the preferential formation of S2O8 2− in strong acidic conditions, the photoconversion efficiency for the different species was investigated.According to the I-V curves of the WO3 photoanode in the PEC process, the photoconversion efficiency can be calculated by Equation ( 5) [1,43]:  is so slow that it can be neglected [43].Thus, we assumed here that the concentration of S which suggests that the photoholes on the surface of the WO 3 photoanode tended to oxidize water molecules (Equation ( 1)) instead of the sulfate anion at high pH values [38].
In order to further confirm the preferential formation of S 2 O 8 2− in strong acidic conditions, the photoconversion efficiency for the different species was investigated.According to the I-V curves of the WO 3 photoanode in the PEC process, the photoconversion efficiency can be calculated by Equation ( 5) [1,43]: where J is the photocurrent density, Int is the illumination power density, E is the applied photoanode potential, A − represent a reduced species (H As shown in Figure 5a, the photocurrent of the WO 3 photoanode reached 2.8 mA cm −2 at 2.1 V vs. counter electrode in 1.0 M Na 2 SO 4 solution at pH 2. In this case, the photoconversion efficiency for S 2 O 8 2− reached a maximum of 2.16% at 1.2 V vs. counter electrode (Figure 5b).However, given that O 2 (g) was the product in the WO 3 photoanode, the photoconversion efficiency would merely yield a maximum 0.39% at 0.96 V vs. counter electrode.This suggests that solar light energy, compared to the oxidation of water, is better utilized for the production of S 2 O 8 2− in strong acidic conditions, which is consistent with the trend of MO degradation.
Catalysts 2017, 7, 346 7 of 12 where J is the photocurrent density, Int is the illumination power density, E is the applied photoanode potential, A − represent a reduced species (H2O or HSO4 − ), A represent an oxidized species (O2 or S2O8 2− ), is 1.23 V vs. NHE and E(S2O8 2− /HSO4 − ) is 2.12 V.As shown in Figure 5a, the photocurrent of the WO3 photoanode reached 2.8 mA cm −2 at 2.1 V vs. counter electrode in 1.0 M Na2SO4 solution at pH 2. In this case, the photoconversion efficiency for S2O8 2− reached a maximum of 2.16% at 1.2 V vs. counter electrode (Figure 5b).However, given that O2 (g) was the product in the WO3 photoanode, the photoconversion efficiency would merely yield a maximum 0.39% at 0.96 V vs. counter electrode.This suggests that solar light energy, compared to the oxidation of water, is better utilized for the production of S2O8 2− in strong acidic conditions, which is consistent with the trend of MO degradation.

The PEC Degradation Mechanism of MO
Theoretically, SO4 2− can be oxidized to S2O8 2− (Equation ( 2)) in a WO3 photoanode at any pH value.However, the theoretical voltage of water oxidation to form O2 (Equation ( 1)) is affected by the concentration of hydrogen ions.The high pH value (pH > 2) is conducive to the evolution of oxygen from water oxidation.Equation (1) dominates the oxidative processes at WO3 photoanodes in 1.0 M Na2SO4 at high pH values (pH > 2).This is consistent with the low production of S2O8 2− when pH is 3 or 7, shown in Figure 6.Therefore, at a high pH value (pH > 2), the decolorization of methyl orange can be attributed to hydroxyl radical (HO•) from water oxidation by photogenerated holes (h + ) via the following equations: MO + HO• → products + H2O.
Sulfate, instead of water, was oxidized by photogenerated holes (h + ) at low pH levels.A reasonable schematic mechanism for efficiently removing MO using a WO3 electrode in a PEC system is shown in Figure 6.The production of S2O8 2− at pH 2 is approximately six times that at pH 3. Prior to the peroxydisulfate (S2O8 2− ), the photogenerated SO4 − • can initially be formed as intermediates in the WO3 photoanode [42,48].Subsequently, the generated SO4 − • can directly react with MO in Na2SO4 solution and revert to SO4 2− (Equations ( 8) and ( 9)).Additionally, from the dynamic viewpoint, the existence of hydrogen ion is beneficial to the oxidation of SO4 − • to form  .This is consistent with the low production of S 2 O 8 2− when pH is 3 or 7, shown in Figure 6.Therefore, at a high pH value (pH > 2), the decolorization of methyl orange can be attributed to hydroxyl radical (HO•) from water oxidation by photogenerated holes (h + ) via the following equations: Sulfate, instead of water, was oxidized by photogenerated holes (h + ) at low pH levels.A reasonable schematic mechanism for efficiently removing MO using a WO 3 electrode in a PEC system is shown in Figure 6 8) and ( 9)).Additionally, from the dynamic viewpoint, the existence of hydrogen ion is beneficial to the oxidation of SO HSO4 − in the case of low pH, which accelerates the reaction between sulfate radical and organic compounds.

Recycling Performance
For environmental applications, the long-term operation of the PEC system is also a key property.According to the above assumption, firstly, SO4 2− transforms into SO4 − •, and SO4 − • can react with MO and be reduced to SO4 2− .Therefore, the SO4 2− undergoes a cycle in the oxidation process.With these considerations in mind, we expected that MO can be continuously degraded in the proposed PEC method.Figure 7 shows three successive MO degradation experiments using a WO3 photoanode in 1 M Na2SO4 (pH = 2).At 80 min intervals, 5 mg L −1 MO was added to the system again without additional SO4 2− .It can be seen from Figure 7 that MO degradation efficiency in the three tests were 95%, 94%, and 95%, respectively.In other words, MO can be continuously removed in this PEC system without adding chemical additives.

Recycling Performance
For environmental applications, the long-term operation of the PEC system is also a key property.According to the above assumption, firstly, SO With these considerations in mind, we expected that MO can be continuously degraded in the proposed PEC method.Figure 7 shows three successive MO degradation experiments using a WO 3 photoanode in 1 M Na 2 SO 4 (pH = 2).At 80 min intervals, 5 mg L −1 MO was added to the system again without additional SO 4 2− .It can be seen from Figure 7 that MO degradation efficiency in the three tests were 95%, 94%, and 95%, respectively.In other words, MO can be continuously removed in this PEC system without adding chemical additives.HSO4 − in the case of low pH, which accelerates the reaction between sulfate radical and organic compounds.

Recycling Performance
For environmental applications, the long-term operation of the PEC system is also a key property.According to the above assumption, firstly, SO4 2− transforms into SO4 − •, and SO4 − • can react with MO and be reduced to SO4 2− .Therefore, the SO4 2− undergoes a cycle in the oxidation process.With these considerations in mind, we expected that MO can be continuously degraded in the proposed PEC method.Figure 7 shows three successive MO degradation experiments using a WO3 photoanode in 1 M Na2SO4 (pH = 2).At 80 min intervals, 5 mg L −1 MO was added to the system again without additional SO4 2− .It can be seen from Figure 7 that MO degradation efficiency in the three tests were 95%, 94%, and 95%, respectively.In other words, MO can be continuously removed in this PEC system without adding chemical additives.

Preparation of WO 3 Photoanodes
The WO 3 photoanodes used in this work were prepared using a method reported elsewhere [41].In a typical process, the SnO 2 :F-coated glass slides (FTO) of 3 × 3 cm was pretreated by washing it under ultrasonication with acetone and deionized water for 30 min.A standard precursor solution was prepared as follows.Firstly, 0.4 g of Na 2 WO 4 •2H 2 O, 0.08 g of H 2 C 2 O 4 , and 0.16 g CO(NH 2 ) 2 were dissolved in 33 mL of deionized water (pH 8.6), and 9 mL of HCl (37%) added to this solution with strong agitation for about 1 min to form a suspension.Then, 8 mL of H 2 O 2 (37%) was added into this suspension with strong agitation for 10 min to form a clear solution.Last, 30 mL of ethanol was added with more strong agitation for 10 min to obtain a clear and stable solution.The FTO glass after pretreatment was dipped into precursor solution with the FTO side facing down.The synthesis was carried out at 94 • C in a constant temperature heating device for 240 min and then allowed to cool naturally.Then, the as-prepared film was rinsed with deionized water for 1 min and dried at 50 • C for 10 h.Next, the as-prepared film was annealed in a muffle at 500 • C for 2 h and a cooling rate of 1 • C/min to obtain WO 3 photoanodes.

Apparatus and Methods
The morphologies and microstructures of WO 3 photoanodes were performed by FE-SEM (Sirion200, Philips, Amsterdam, The Netherlands).The crystal phase of the samples was characterized with X-ray diffractometry (XRD) (AXS-8 Advance, Bruker, Karlsruhe, Germany).In a two electrode system, linear sweep voltammetry for WO 3 films photocurrent measurements was investigated under 100 mW cm −2 simulated sunlight from an Xe lamp in 1.0 M Na 2 SO 4 (pH = 2) at 0.4 -2.12 V vs. counter electrode.
A photoelectrocatalysis (PEC) system was constructed using an as-prepared nano-WO 3 photoanode and a Pt-black cathode in Na 2 SO 4 solution under light illumination for the degradation of 5 mg/L MO.All of the photoelectrochemical experiments of MO degradation were performed at 1.5 V of bias voltage, in 50 mL of a quartz glass reactor containing 40 mL of sodium sulfate electrolyte, with only approximately 1 cm 2 WO 3 photoanodes exposed to under 100 mW cm −2 AM1.5G simulated sunlight from an Xe lamp.The degradation of 5 mg/L MO within 80 min by photolysis, electrocatalysis (EC), photocatalysis, and PEC in Na 2 SO 4 medium were investigated.The photolysis experiment was carried out to determine the degradation rate of MO only under visible light irradiation.The electrocatalysis (EC) experiment was performed in a photoelectrocatalysis system, where 1.5 V vs. counter electrode of bias voltage was applied to the WO 3 /FTO electrode in darkness.The photocatalysis (PC) experiment was performed in photoelectrocatalysis system but with no bias voltage compared with photoelectrocatalysis.The effects of sulfate concentrations and pH value on the degradation of methyl orange were also investigated in this photoelectrocatalytic system.Active species resulting in the catalytic oxidation of MO was identified.In addition, the practical ability of the photoelectrochemical system to yield active persulfate (S 2 O 8 2− , SO 5 2− , and SO 4 − •) species was evaluated according to the production of peroxydisulfate (S 2 O 8 2− ).The peroxydisulfate (S 2 O 8 2− ) quantity was measured by colorimetry using UV-VIS spectroscopy at 310 nm in 1 M sodium sulfate solution with different pH values.The sample solution was stirred with an equivalent amount of 0.01 M FeSO 4 and 1 M H 2 SO 4 solution to measure the optical absorbance spectra at 310 nm, and concentration was evaluated according standard curve [49].

Conclusions
In this paper, a novel green method was successfully established to highly efficiently remove refractory organics using SO 4 − • generated directly from a WO 3 photoelectrode and the catalytic reaction of sulfate without the addition of any other activator.The results show that the removal efficiency for MO with 5 mg/L was up to 95% within 80 min in the PEC system, and merely 3% by photolysis and 19% by photocatalysis, respectively, under similar conditions.The rate constant for the disposal of MO at pH 2, in which SO 4 − • instead of HO• was the main oxidizer, confirmed by a radical scavenger test, was up to 5.21 × 10 −4 s −1 , which was ~6.6 times that (7.89 × 10 −5 s −1 ) under neutral conditions, in which HO• was the main oxidizer.Additional, the concentration of active persulfate (S 2 O 8 2− , SO 5 2− , SO 4 − •) species at pH 2 was up to 0.38 mM, which was ~16 times that in neutral conditions (0.023 mM).The proposed method could be a new, effective, environmentally friendly approach for the treatment and resource utilization of sulfate wastewater.

Figure 4 .
Figure 4. Removal of methyl orange with time and different additives to a 1.0 M Na2SO4 aqueous solution at (a) pH = 2. (b) pH = 3 and (c) pH = 7.The initial conditions: MO 5 mg L −1 , Na2SO4 1 M, applied 1.5 V vs. CE.(d) The concentration of S2O8 2− with the function of time at different pH values (involving 40 mL of 1 M Na2SO4 aqueous solution).

Figure 4 .
Figure 4. Removal of methyl orange with time and different additives to a 1.0 M Na 2 SO 4 aqueous solution at (a) pH = 2. (b) pH = 3 and (c) pH = 7.The initial conditions: MO 5 mg L −1 , Na 2 SO 4 1 M, applied 1.5 V vs. CE.(d) The concentration of S 2 O 8 2− with the function of time at different pH values 40 mL of 1 M Na 2 SO 4 aqueous solution).

SO 4 −
• plays an important role in the degradation of MO, but it is difficult to direct identify the SO 4 − • in the whole process because of a short lifetime of sulfate radical.In fact, SO 4 2− can be oxidized to S 2 O 8 2− (Equation (2)) [43] and SO 4 − • could be an intermediate species that is prior to the formation of S 2 O 8 2− (Equation (4)).Therefore, the amount of SO 4 − • could be indirectly analyzed by investigating the production of S 2 O 8 2− .In fact, a small part of S 2 O 8 2− in 1.0 M Na 2 SO 4 was hydrolyzed to produce HSO 5 − , which has an effect on the determination of S 2 O 8 2− via colorimetry, but the hydrolysis of S 2 O 8 2−

Figure 5 .
Figure 5. (a) The photocurrent and (b) the applied bias photo-current efficiency in a 1.0 M Na2SO4 aqueous solution (pH = 2) by using a WO3 photoanode under illumination of simulated solar light.

Figure 5 .
Figure 5. (a) The photocurrent and (b) the applied bias photo-current efficiency in a 1.0 M Na 2 SO 4 aqueous solution (pH = 2) by using a WO 3 photoanode under illumination of simulated solar light.

2. 4 .
The PEC Degradation Mechanism of MO Theoretically, SO 4 2− can be oxidized to S 2 O 8 2− (Equation (2)) in a WO 3 photoanode at any pH value.However, the theoretical voltage of water oxidation to form O 2 (Equation (1)) is affected by the concentration of hydrogen ions.The high pH value (pH > 2) is conducive to the evolution of oxygen from water oxidation.Equation (1) dominates the oxidative processes at WO 3 photoanodes in 1.0 M Na 2 SO 4 at high pH values (pH > 2)

Figure 6 .
Figure 6.Schematic illustration of efficiently degrading MO with WO3 electrode in Na2SO4.

Figure 7 .
Figure 7. MO degradation during a series of three tests with additional 5 mg L −1 MO at 80 min intervals.The initial conditions: MO 5 mg L −1 , Na2SO4 1 M L −1 , applied 1.5 V vs. CE and pH = 2.

Figure 6 .
Figure 6.Schematic illustration of efficiently degrading MO with WO 3 electrode in Na 2 SO 4 .

4 2 −
transforms into SO 4 − •, and SO 4 − • can react with MO and be reduced to SO 4 2− .Therefore, the SO 4 2− undergoes a cycle in the oxidation process.

Figure 6 .
Figure 6.Schematic illustration of efficiently degrading MO with WO3 electrode in Na2SO4.

Figure 7 .
Figure 7. MO degradation during a series of three tests with additional 5 mg L −1 MO at 80 min intervals.The initial conditions: MO 5 mg L −1 , Na2SO4 1 M L −1 , applied 1.5 V vs. CE and pH = 2.

Figure 7 .
Figure 7. MO degradation during a series of three tests with additional 5 mg L −1 MO at 80 min intervals.The initial conditions: MO 5 mg L −1 , Na 2 SO 4 1 M L −1 , applied 1.5 V vs. CE and pH = 2.
4 − • to form HSO 4− in the case of low pH, which accelerates the reaction between sulfate radical and organic compounds.