Sustainable chromic acid oxidation: solar-driven recycling of hexavalent chromium ions for quinone production by WO₃ nanosponge photoanodes

Abstract

Photoelectrochemical chromium ion oxidation from the trivalent to the hexavalent state has been demonstrated by using WO₃ nanosponge photoanodes in sulfuric acid electrolytes under simulated sunlight. Faradaic efficiencies of almost 100% were achieved in electrolytes containing high Cr³⁺ (≥10 mM) and low SO₄²⁻ (≤10 mM) concentrations at very low bias potential voltages below 1.1 V_RHE. At the photoanode, the production of Cr⁶⁺ ions competed with the formation of S₂O₈²⁻ ions from HSO₄⁻/SO₄²⁻ ions. At the cathode, hydrogen evolution occurred with 100% faradaic efficiency. Additionally, Cr³⁺ recovery was also observed at the cathode upon switching the electrolytes of the anode and cathode cells after the Cr⁶⁺ production, demonstrating the complete cyclic photoelectrochemical reaction between Cr³⁺ and Cr⁶⁺ ions. Finally, we showed the applicability of chromic acid oxidation to quinone production by using the photoelectrochemically produced Cr⁶⁺ ions in the sulfuric acid solutions. p-Benzenediol was successfully converted into p-benzoquinone with high efficiency. Thus, we have demonstrated the production of high-value-added organic reagents via photoelectrochemically driven chromic acid oxidation with the recycling of Cr⁶⁺ from Cr³⁺ using solar energy.

---

Introduction

Hexavalent chromium ions are an important industrial reagent with a wide range of applications, such as oxidation to produce various organic materials and in Cr metal plating, dyes/pigments, and catalysis.^1,2 However, the high toxicity of Cr⁶⁺ ions due to their strong oxidizing power represents a major limitation.^2–5 In contrast, the Cr³⁺ ion is an essential nutrient for the human body.⁶ Although environmental exposure to Cr⁶⁺ ions through dyes and pigments has significantly diminished over the last few decades as a result of strict environmental laws and regulations,⁷ Cr⁶⁺ ions are still widely used around the world (Fig. S1†). Among the applications of Cr⁶⁺ ions, the production of high-value-added industrial reagents using chromic acid oxidation processes that rely on the strong oxidizing power of hexavalent chromium warrants attention.^8–11

Chromic acid oxidation has been applied in several derivative processes that use not only aqueous sulfuric acid solution but also organic solvents and complexes to form Cr⁶⁺ ions.⁹ These processes are typically used for the conversion of secondary alcohols into ketones and primary alcohols into carboxylic acids. In particular, quinones and their derivatives are valuable products that are essential for the manufacturing of various pharmacologically active reagents.^12,13 Although chromic acid oxidation is an industrial process with unquestionable practical benefits, the fact remains that the safety issues absolutely necessitate the development of alternative ways to use Cr⁶⁺ ions. As Cr⁶⁺ compounds are scarcely found in nature due to their strong oxidizing power, they are generally manufactured industrially via a firing process involving chrome ore and soda ash at production facilities separate from the factories that use them. The main safety issue of Cr⁶⁺ is occupational exposure to the Cr⁶⁺ ions during each of the processes involved in production, transportation, handling, and disposal (ESI†).¹⁴

To allow the safe and efficient use of chromic acid oxidation involving Cr⁶⁺ ions in the production of various high-value-added reagents, we devised a system for the photoelectrochemical (PEC) recycling of Cr⁶⁺ ions using solar energy in a closed reaction vessel. If the trivalent chromium ions produced by the chromic acid oxidation process can be reoxidized to hexavalent chromium ions by sunlight using photoanodes, the toxic Cr⁶⁺ ions could be used as the mediator for subsequent reactions on site and on demand without producing hazardous chromium waste. This design permits the utilization of Cr⁶⁺ ions in production systems while increasing the occupational safety. The use of solar energy means that the final products of the chromic acid oxidation can be obtained with lower energy consumption and additionally allows solar hydrogen generation^15–19 at the cathode when using aqueous electrolyte solutions. Thus, this reaction system has multiple advantages to think about building a sustainable society.

In this work, we used WO₃ photoanodes for effective oxidation reactions under sunlight and firstly revealed that the Cr³⁺ ions were efficiently oxidized by the PEC reaction. We were able to achieve chromium ion oxidation from the trivalent to the hexavalent form on WO₃ nanosponge photoanodes under simulated sunlight with a faradaic efficiency of almost 100% under the optimum conditions. We also demonstrated the applicability of the chromic acid oxidation process to quinone production. By using the photoelectrochemically produced Cr⁶⁺ ions, we confirmed that p-benzenediol was successfully converted into p-benzoquinone, which is used industrially as an oxidizing agent for various organic materials, as a polymerization inhibitor for raw polyester materials, and in dyes and photographic chemicals.^20,21

Experimental procedures

The WO₃ nanosponge photoanodes with high PEC performance were prepared by a nanoparticle/solution hybrid dispersion–deposition process (Fig. S2†).^22,23 A WO₃ nanoparticle/solution hybrid dispersion was prepared by wet ball milling of WO₃ nanocrystal powder (EM Japan) in isopropanol with a tilted rotation planetary ball mill (Planet M, Nagao System) using zirconia balls at 650 rpm with a toluene solution of tungsten phenoxide (Gelest). Polyethylene glycol (PEG; molecular weight, 300; Wako Pure Chemical Industries) was added to the WO₃ dispersion in a 1 : 1 volume ratio. The WO₃ dispersion ink containing PEG was spin coated at 2000 rpm for 10 s onto F-doped SnO₂ (FTO) glass substrates. The films were preheated at 500 °C for 5 min. The coating and preheating processes were repeated several times to increase the film thickness, and the preheated films were then fired at 550 °C for 30 min in air. The phase purity and crystallinity of the WO₃ thin films on the FTO substrates were determined by X-ray diffraction (SmartLab, Rigaku; Fig. S2†). The surface morphology and cross-sectional structure were determined by field emission scanning electron microscopy (FESEM; SU9000, Hitachi).

Linear sweep voltammetry (LSV) for photocurrent measurements and the production tests for Cr⁶⁺ and H₂ were performed under 100 mW cm⁻² AM1.5G simulated sunlight (1 sun) from a 150 W Xe lamp (XES-40S2-CE, San-Ei Electric) with Ar gas bubbling in aqueous Cr₂(SO₄)₃ and H₂SO₄ solutions at 0.4–2.0 V versus the reversible hydrogen electrode (V_RHE) in three-electrode one- or two-compartment cells equipped with a quartz glass window (30 mL of solution was used in each cell compartment). Illumination was carried out from the rear side of the photoanode (Fig. S3†). The 1 sun simulated sunlight was calibrated using a spectroradiometer (Fig. S4;† Soma Optics). A Ag/AgCl reference electrode in saturated KCl solution, a Pt-wire counter electrode, and a Nafion ion-exchange membrane were used, and the scan rate was 5 mV s⁻¹. The measured potential versus Ag/AgCl (V_Ag/AgCl) in the three-electrode system was converted into V_RHE according to the equation V_RHE = V_Ag/AgCl + 0.059 × pH + V⁰_Ag/AgCl, where V⁰_Ag/AgCl = 0.1976 at 25 °C. The applied bias photon-to-current efficiency (ABPE_Cr⁶⁺) values of the WO₃ photoanodes were measured using a two-electrode configuration cell with a Pt-wire counter electrode, according to the equation ABPE_Cr⁶⁺ = J(1.36 − E_CE)η(Cr⁶⁺)/I_AM1.5, where J is the photocurrent density (mA cm⁻²), E_CE is the applied bias voltage versus the counter electrode, η(Cr⁶⁺) is the faradaic efficiency of Cr⁶⁺ production, and I_AM1.5 is the irradiance of AM1.5G simulated sunlight at 100 mW cm⁻². The production of Cr⁶⁺ and S₂O₈²⁻ was evaluated colorimetrically by using UV-vis spectroscopy (UV-3150, Shimadzu; Fig. S5†). The evolved H₂ gas was detected using a gas chromatograph with a thermal conductivity detector (GC-8AIT, Shimadzu).

The reactions using photoelectrochemically evolved Cr⁶⁺ for p-benzoquinone production were carried out by the addition of p-benzenediol to the obtained Cr⁶⁺ test solutions. The production of p-benzoquinone was evaluated both colorimetrically using UV-vis spectroscopy and by gas chromatography-mass spectrometry (GC/MS; 7890B, Agilent and JMS-Q1500GC, JEOL).

Results and discussion

Fig. 1a shows the two-compartment PEC cell setup for the production of Cr⁶⁺ and H₂ using a WO₃ photoanode under simulated sunlight. We used FESEM to confirm that the prepared WO₃ photoanodes possessed a nanosponge morphology²³ with 5–100 nm nanopores and 5 μm thickness (Fig. S6†). To directly use the produced Cr⁶⁺ ions as chromic acid oxidation reagents, we mainly studied the PEC reaction of Cr³⁺ ions in aqueous sulfuric acid solution. Therefore, there are three candidates for oxidation in the anode cell: Cr³⁺ ions, HSO₄⁻ (SO₄²⁻)^19,20 ions, and H₂O. The oxidation potentials for the three reactants are shown in Fig. 1b, and the reactions and electrochemical potentials can be described as follows:

2Cr³⁺ + 7H₂O → Cr₂O²⁻₇ + 14H⁺ + 6e⁻, Cr₂O²⁻₇/Cr³⁺ = +1.36 V vs. RHE (1)

2HSO⁻₄ → S₂O²⁻₈ + 2H⁺ + 2e⁻, S₂O²⁻₈/HSO⁻₄ = +2.12 V vs. RHE (2)

2H₂O → O₂ + 4H⁺ + 4e⁻, O₂/H₂O = +1.23 V vs. RHE (3)

Fig. 1 (a) Conceptual diagram of the PEC reaction in a 2-compartment cell setup and the cross-sectional FESEM image of WO₃ nanosponge photoanodes. The anolyte (aqueous H₂SO₄ containing Cr³⁺ ions) and catholyte (aqueous H₂SO₄) are separated by a Nafion membrane filter. (b) Energy diagram of the photoelectrode system for the redox reactions of H₂, H₂O, Cr³⁺, and HSO₄⁻.

The PEC reaction depicted in eqn (2) has been previously reported for WO₃ photoanodes in aqueous H₂SO₄ electrolytes under simulated sunlight.^23–26 It is therefore necessary to consider this possible side reaction when studying the intended oxidation reaction shown in eqn (1).

The LSV curve for the electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ (pH = 1.38 and 25 mM SO₄²⁻) is shown in Fig. 2a. The onset potential of the photocurrent was about 0.55 V_RHE, and the J increased up to 2.0 mA cm⁻² at 1.45 V_RHE. Using this solution, the WO₃ photoanode was subjected to continuous illumination with the simulated sunlight at 1.50 V_RHE. Color variation of the solution from bluish-green to greenish-yellow was observed during the illumination. The illumination-time dependence of the absorption spectra is shown in Fig. 2b. With increasing illumination time, the two maxima at 417 nm and 586 nm corresponding to the Cr³⁺ ions decreased and an additional two peaks appeared at 350 nm and 446 nm, which can be assigned to the Cr⁶⁺ ions (Fig. S5†). The applied electric charge at 6 h was 2.79 C, and the η(Cr⁶⁺) was determined to be 83.4%.

Fig. 2 (a) J–V curves under chopped simulated sunlight at 1 sun from the rear side of the WO₃ photoanode in electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄. (b) Illumination-time dependence of the absorption spectra of electrolyte solutions reacted under 1 sun at 1.5 V_RHE.

We have carefully studied the effects of the Cr³⁺ and SO₄²⁻ concentrations in the solutions on the η(Cr⁶⁺). The value of η(Cr⁶⁺) varied almost inversely with the logarithm of the H₂SO₄ concentration and exhibited a sudden increase beyond a Cr³⁺ concentration of 1 mM (Fig. 3a). Fig. 3b shows a map of η(Cr⁶⁺) as functions of the Cr³⁺ and SO₄²⁻ concentrations for the PEC reactions at 1.50 V_RHE. The η(Cr⁶⁺) was found to be strongly enhanced at higher and lower concentrations of Cr³⁺ and SO₄²⁻ ions, respectively. The η(Cr⁶⁺) exceeded 80% at a bias voltage of 1.50 V_RHE above 10 mM Cr³⁺ and below 200 mM SO₄²⁻. Sulfate ions are known to be oxidized to S₂O₈²⁻ during PEC reactions at WO₃ photoanodes;^23–26 therefore, the η(Cr⁶⁺) would be heavily competing with the reaction of SO₄²⁻ ions at the photoanode. We also analyzed the products formed at both electrodes during the PEC reaction in the electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ at 1.50 V_RHE. Hydrogen gas evolution with almost 100% faradaic efficiency was observed at the cathode, while Cr⁶⁺ and S₂O₈²⁻ ions were formed at the photoanode with efficiencies of 81–84% and 12–20%, respectively (Fig. 3c). We confirmed that other potential products such as oxygen were not observed at the anode cell, and the formation of Cr⁶⁺ ions was not due to oxidation by the simultaneously generated S₂O₈²⁻ ions (Fig. S7†). These results strongly indicate that the photocurrent was consumed for the production of Cr⁶⁺ and S₂O₈²⁻ ions at the photoanode in this system. The selectivity of products would be dependent on the adsorptivity of each chemical species on the WO₃ photoanode surface. The Cr³⁺ ([Cr(H₂O)₆]³⁺) and HSO₄⁻ (SO₄²⁻) ions are expected to be preferentially adsorbed at the WO₃ surface around this concentration range, compared to the water molecules.

Fig. 3 (a) η(Cr⁶⁺) as functions of the concentrations of Cr³⁺ (at 10 mM H₂SO₄) and H₂SO₄ (at 10 mM Cr³⁺) at 1.5 V_RHE. (b) η(Cr⁶⁺) contour map as functions of the concentrations of Cr³⁺ and SO₄²⁻ at 1.5 V_RHE. (c) Reaction-time dependence of the faradaic efficiencies for H₂, Cr⁶⁺, and S₂O₈²⁻ generation in electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ at 1.5 V_RHE.

Fig. 4a shows the absorption spectra of electrolyte solutions (original concentrations were 1.0 mM Cr³⁺ and 10 mM H₂SO₄) during the PEC reactions under different external biases of 0.9, 1.1, 1.3, and 1.5 V_RHE. Whereas the absorption peaks associated with the Cr⁶⁺ ions were observed with increasing illumination time at all of the bias voltages tested, the resulting peak intensity was strongly dependent on the bias voltage. The peak intensity at 350 nm reached a maximum at 1.1 V_RHE. The J–V curve in this electrolyte was similar to those in 10 mM Cr³⁺ and 10 mM H₂SO₄, although it exhibited a somewhat more gradual increase that could be due to the insufficient supply of Cr³⁺ to the photoanode surface (Fig. S8†). Although the value of J increased from 0.55 V_RHE and saturated at around 1.6 V_RHE, the η(Cr⁶⁺) increased with the decreasing bias voltage. At about 1.0 C, the value of η(Cr⁶⁺) was 10.4% and 12.6% at 1.5 and 1.3 V_RHE, respectively. A significantly higher η(Cr⁶⁺) of 27.5% was observed at 1.1 V_RHE, and the value reached 40.2% at 0.9 V_RHE. This η(Cr⁶⁺) variation was ascribed to the difference in theoretical redox potentials for the evolution of Cr⁶⁺ and S₂O₈²⁻ ions as detailed in Fig. 1b and eqn (1) and (2), indicating that the oxidation of Cr³⁺ is dominant at lower applied bias voltages. The competition between the Cr⁶⁺ and S₂O₈²⁻ ions was weakened by the increase of the initial amount of Cr³⁺ relative to SO₄²⁻. At 10 and 100 mM Cr³⁺, the value of η(Cr⁶⁺) reached almost 100% at 1.1 V_RHE or below (Table 1). The concentrations of Cr⁶⁺ ions evolved in the electrolytes containing 10 and 100 mM Cr³⁺ were 313 μM at 1.1 V_RHE and 432 μM at 1.3 V_RHE, respectively (Fig. 4d and e).

Fig. 4 (a) Illumination-time dependence of the absorption spectra of electrolyte solutions containing 1.0 mM Cr³⁺ and 10 mM H₂SO₄ reacted under 1 sun at 0.9, 1.1, 1.3, and 1.5 V_RHE. (b) η(Cr⁶⁺) as a function of applied electric charge at 0.9, 1.1, 1.3, and 1.5 V_RHE for electrolyte containing 1.0 mM Cr³⁺ and 10 mM H₂SO₄. (c–e) Time dependence of the Cr⁶⁺ concentration produced from (c) 1, (d) 10, and (e) 100 mM Cr³⁺ ions in 10 mM H₂SO₄ at 0.9, 1.1, 1.3, and 1.5 V_RHE.

Table 1 η(Cr⁶⁺) versus the bias potential and Cr³⁺ concentration

Bias/VRHE	η(Cr^6+)/% for 1.0 mM Cr^3+	η(Cr^6+)/% for 10 mM Cr^3+	η(Cr^6+)/% for 100 mM Cr^3+
0.9	40.2	100.2	101.8
1.1	27.5	98.1	99.0
1.3	12.6	91.2	97.0
1.5	10.4	81.3	94.0

---

To evaluate the effect of acidity on the η(Cr⁶⁺), we also studied the PEC reaction in Na₂SO₄ electrolyte. The pH of a solution of 10 mM Cr³⁺ and 10 mM Na₂SO₄ was 2.70, while both the 10 mM Cr³⁺/10 mM Na₂SO₄ and 10 mM Cr³⁺/10 mM H₂SO₄ solutions contained the same concentration of SO₄²⁻ ions (25 mM). The LSV curve using the Na₂SO₄ electrolyte increased gradually compared with that using the sulfuric acid electrolyte, and the photocurrent had not reached saturation even at 2.0 V_RHE (Fig. 5a). A similar onset potential of around 0.55 V_RHE for the two solutions indicates that the pH difference only affected the theoretical flat band potential shift and had almost no effect on the band structure and transport properties of WO₃ itself. The inhibition of Cr³⁺ ([Cr(H₂O)₆]³⁺) supply (adsorption) to the photoanode surface due to the excess Na⁺ ions may be one possible explanation. In contrast, the peak intensities for the Cr⁶⁺ ions in the absorption spectra were clearly stronger than those in the sulfuric acid solution (Fig. 5b) and the η(Cr⁶⁺) at 1.5 V_RHE was significantly higher at almost 100%. In accordance with this η(Cr⁶⁺) of almost 100%, the formation of S₂O₈²⁻ was not detected, and the concentration of Cr⁶⁺ ions generated after 6 h was 94% higher than that generated in the sulfuric acid solution (Fig. 5c).

Fig. 5 (a) J–V curves under chopped simulated sunlight at 1 sun in electrolyte containing 10 mM Cr³⁺ and 10 mM Na₂SO₄. (b) Illumination-time dependence of the absorption spectra of electrolyte solutions reacted under 1 sun at 1.5 V_RHE. (c) Time dependence of Cr⁶⁺ and S₂O₈²⁻ concentrations produced in the 10 mM H₂SO₄ and Na₂SO₄ electrolytes. The initial Cr³⁺ concentration and potential bias were 10 mM and 1.5 V_RHE, respectively. (d) Map of the η(Cr⁶⁺) and SO₄²⁻ concentration as functions of the Cr³⁺ concentration and pH at 1.5 V_RHE.

Fig. 5d shows the η(Cr⁶⁺) and SO₄²⁻ concentration as functions of the pH and Cr³⁺ concentration. The η(Cr⁶⁺) was actually enhanced by the low acidity of the Cr³⁺/Na₂SO₄ electrolyte, but it was not simply correlated with the pH value. The very high η(Cr⁶⁺) of 100% in 10 mM Cr³⁺ and 10 mM Na₂SO₄ (pH = 2.70 and 25 mM SO₄²⁻) was clearly lowered to 76.5% in 10 mM Cr³⁺ and 1.0 M Na₂SO₄ (pH = 3.15 and 1.015 M SO₄²⁻). As such, the SO₄²⁻ concentration in this PEC reaction system cannot be ignored. The results of the PEC reaction in the Cr³⁺/Na₂SO₄ electrolyte also support the aforementioned key factor for the formation of Cr⁶⁺ ions: the η(Cr⁶⁺) faced strong competition with the production of S₂O₈²⁻ from the sulfate ions. The faradaic efficiency is strongly correlated with both the ratio of concentrations and the differences from each theoretical oxidation potential. The electrolyte effect of H₂SO₄/Na₂SO₄ on the shape of the J–V curves was not large.²⁷ Therefore, it is necessary to discuss the correlation between H⁺/Na⁺ and Cr³⁺ and the surface properties of the WO₃ photoanodes in this PEC reaction, because it could be affected by the cationic (Cr³⁺ ([Cr(H₂O)₆]³⁺)) and anionic (SO₄²⁻/HSO₄⁻) adsorption dependences.

Fig. 6a shows the J–V curve for the electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ using a one-compartment cell. The J–V curve was very similar to that obtained using the two-compartment cell; however, the Cr⁶⁺ generation at 1.5 V_RHE was quite low compared with the latter case (Fig. 2). The absorption peak increased slightly even at 6 h (Fig. 6c), and the η(Cr⁶⁺) was determined to be 9.2%. This low η(Cr⁶⁺) would be better treated as an apparent value of the efficiency. Cr⁶⁺ ions are known to be reduced at the cathode;^28–32 therefore, both oxidation and reduction reactions of chromium ions can occur at the photoanode and cathode, resulting in the apparent low η(Cr⁶⁺). Combinations of TiO₂-based photoanodes and noble metal cathodes have generally shown that the Cr⁶⁺ reduction is dominant in PEC reaction systems using a one-compartment cell.²⁸ Therefore, the use of WO₃ photoanodes with strong oxidizability (valence band potential deeper than that of the TiO₂)^33–35 would provide an important advantage for obtaining Cr⁶⁺ ions by the PEC reaction. Nevertheless, it is necessary to employ the two-compartment cell setup for extracting Cr⁶⁺ ions efficiently.

Fig. 6 (a) J–V curves under chopped simulated sunlight at 1 sun in electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ from the rear side of the WO₃ photoanode using a one-compartment cell. (b) J–V curves without photo-illumination in electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ using the FTO anode and a two-compartment cell. (c) Time dependences of the absorption spectra of electrolytes containing 10 mM Cr³⁺ and 10 mM H₂SO₄ at 1.5 V_RHE for the experimental setups shown in (a) and (b).

We also carried out an electrochemical reaction for Cr⁶⁺ generation. Fig. 6b shows the J–V curve in the electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄ using the two-compartment cell and FTO anode without photo-illumination during the measurement. The onset potential for the current was about 2.05 V_RHE and the current grew linearly above 2.5 V_RHE. The amount of Cr⁶⁺ ions generated in the electrochemical reaction depended on the applied bias. At 2.6 V_RHE, which is a much higher bias potential than that applied in the PEC reactions, the amount of Cr⁶⁺ ions generated was only 50–60% of that in the PEC reaction, and the η(Cr⁶⁺) remained at 40%. It has also been reported that the η(Cr⁶⁺) remained at 80% even at the much higher bias potential of 3.5 V.² These results indicate that the PEC reaction using the WO₃ photoanode can produce Cr⁶⁺ ions very efficiently compared with the electrochemical process.

The dependences of η(Cr⁶⁺) on the bias potential and electrolyte composition in the PEC reaction are summarized in Fig. 7a. The η(Cr⁶⁺) showed very high values at high Cr³⁺ concentrations above 10 mM and reached almost 100% below 1.1 V_RHE. The amounts of Cr⁶⁺ produced at 1.5 V_RHE were 3.9 and 6.5 μmol cm⁻² h⁻¹ at 10 and 100 mM Cr³⁺, respectively (Fig. 7b). For the electrolyte containing 10 mM Cr³⁺ and 10 mM H₂SO₄, the ABPE_Cr⁶⁺ exhibited a maximum of 0.62% at 0.7 V versus the counter electrode in the two-electrode cell (Fig. S9†). The suppression of S₂O₈²⁻ production by the use of Na₂SO₄ electrolyte instead of sulfuric acid was very effective in enhancing the η(Cr⁶⁺). The Cr⁶⁺ production was 8.6 μmol cm⁻² h⁻¹ at 1.5 V_RHE and 10 mM Cr³⁺. This value was 3.6 times higher than the electrochemical production at the same Cr³⁺ concentration and a much higher bias potential (2.6 V_RHE).

Fig. 7 Bias potential dependences of (a) η(Cr⁶⁺) and (b) Cr⁶⁺ production activity for each electrolyte and several experimental setups. 1CC and EC represent the one-compartment cell setup at 1 sun and the electrochemical reaction without photoillumination, respectively. The other experiments were carried out at 1 sun using the two-compartment cell.

We also examined the backward reaction from the produced Cr⁶⁺ to Cr³⁺ (Fig. 8). The Cr⁶⁺-containing solution from the anode cell of the PEC reaction after 6 h (the original solution was 10 mM Cr³⁺ and 10 mM H₂SO₄) was switched with the sulfuric acid from the cathode cell (10 mM H₂SO₄). The absorption peak corresponding to the Cr⁶⁺ ions decreased with increasing time, and the calculated faradaic efficiency for conversion from the hexavalent to the trivalent form was very high (85%) and comparable with the forward PEC reaction to obtain the Cr⁶⁺. This result demonstrates that on-demand Cr⁶⁺ ion production is feasible and that Cr ions can be safely stored as the Cr³⁺ form in the reaction vessel (Fig. S10†).

Fig. 8 Illumination-time dependence of the absorption spectra of electrolyte solutions (10 mM Cr³⁺ and 10 mM H₂SO₄) reacted under 1 sun at 1.5 V_RHE in the anode cell for 6 h (upper panel), followed by the cathode reaction for 6 h after switching electrolytes between the anode and cathode (lower panel).

Finally, we carried out synthetic experiments for p-benzoquinone production using the Cr⁶⁺ ions obtained from the PEC reaction. As such, 10–2000 μM of p-benzenediol was added to the photoelectrochemically produced 298.5 μM Cr⁶⁺ in 10 mM H₂SO₄ solution. Fig. 9a shows the p-benzenediol input dependence of the absorption spectra. With increasing p-benzenediol addition, the absorption peak corresponding to the Cr⁶⁺ ions at 350 nm decreased and the absorption peak corresponding to the p-benzoquinone at 245 nm increased. The production of p-benzoquinone was also qualitatively confirmed by GC/MS analysis. In the reaction solution, we did not detect by-products. The absorption peak of Cr⁶⁺ at 350 nm disappeared when p-benzenediol concentrations of 500 μM or above were added, and the absorption peak corresponding to excess p-benzenediol started to appear at 290 nm. From the absorption spectra, the concentrations of Cr⁶⁺, p-benzenediol, and p-benzoquinone were determined, as shown in Fig. 9b and S11.† The Cr⁶⁺ ions were almost stoichiometrically consumed by the p-benzenediol addition. This indicates that the two hydroxyl groups of the p-benzenediol were successively oxidized by the Cr⁶⁺ ions in 10 mM H₂SO₄ and the produced intermediate Cr⁴⁺ ions (ESI†). From the initial concentration of Cr⁶⁺ used (298.5 μM), 85.7% and 98.9% of the Cr⁶⁺ ions were converted into Cr³⁺ with the addition of 300 or 500 μM of p-benzenediol, respectively. The amount of p-benzoquinone produced corresponded to the amount of Cr⁶⁺ consumed, indicating that the chromic acid oxidation process proceeded very effectively in this reaction system. The obtained p-benzoquinone was easily extracted by a conventional two-phase sampling method using appropriate solvents such as diethyl ether.

Fig. 9 (a) Absorption spectra of Cr⁶⁺/H₂SO₄ solution with the addition of p-benzenediol. (b) The consumption rate of Cr⁶⁺ and the concentrations of each species with the addition of p-benzenediol. (c) The absorption spectra of Cr⁶⁺/Na₂SO₄ solution with the addition of p-benzenediol. (d) The consumption rate of Cr⁶⁺ and the concentrations of each species with the addition of p-benzenediol. c_D, c_Q, and c_D-add represent the calculated concentrations of p-benzenediol and p-benzoquinone and the concentration of added p-benzenediol, respectively.

Fig. 9c shows the p-benzenediol input dependence of the absorption spectra for p-benzoquinone production using the Cr⁶⁺ ions obtained from the PEC reaction in the 10 mM Na₂SO₄ electrolyte. The initial concentration of Cr⁶⁺ was 409.7 μM. Although we confirmed that the p-benzenediol was converted into p-benzoquinone by the Cr⁶⁺ ions as well as in the case of the H₂SO₄ solution discussed above, Cr⁶⁺ ions remained even with the addition of 2000 μM of p-benzenediol. The conversion rate of Cr⁶⁺ with the addition of 1000 μM of p-benzenediol (2.4 times higher concentration than the initial concentration of Cr⁶⁺ ions) was still only 73.7%. These results suggest that the chromic acid oxidation in the Na₂SO₄ solution was less efficient than that in the H₂SO₄ solution. With the same input of p-benzenediol (300 μM), 34% more p-benzoquinone was obtained in the H₂SO₄ solution than in the Na₂SO₄ solution, despite the lower initial concentration of Cr⁶⁺ ions. This means that the H₂SO₄ electrolyte is also effective for producing quinones in the chromic acid oxidation process using the photoelectrochemically produced Cr⁶⁺ ions, although the η(Cr⁶⁺) is slightly lower compared with that observed using the Na₂SO₄ electrolyte.

Thus, we have demonstrated the efficient PEC oxidation of Cr³⁺ to Cr⁶⁺, the chromic acid oxidation for obtaining quinones, and the PEC reduction of Cr⁶⁺ to Cr³⁺ ions. The Cr ion redox mediator can be used for oxidation processes of a wide variety of organics by the PEC reaction due to its high oxidizability, compared to the other PEC redox systems, e.g., Ni²⁺/Ni³⁺ mediator.³⁶ When the external bias voltage is supplied by a solar cell, high-value-added organic reagents can be produced on site and on demand via the chromic acid oxidation by the solar-energy-driven continuous PEC recycling of Cr³⁺ to Cr⁶⁺. This PEC reaction system with Cr ions would also be effective for green energy production, in terms of not only the hydrogen gas evolved from the counter electrode but also the possibility of a new redox flow cell with high energy density using the multi-electron redox reaction of Cr³⁺/Cr⁶⁺ ions. In addition, maintaining a safe stock of Cr ions will be possible by the PEC reduction of Cr⁶⁺. All of these reactions can feasibly be performed in a single closed reaction vessel, permitting the safe use of Cr⁶⁺ ions. Therefore, this system enables the chromic acid oxidation process to be employed for the production of high-value-added reagents without the generation of hazardous chromium waste (Fig. 10), allowing the great industrial value of chromium ions to be exploited once again.

Fig. 10 Schematic conceptual diagram for the cyclic solar-energy-driven use of Cr ions between Cr³⁺/Cr⁶⁺ and the production of high-value-added organic reagents using the chromic acid oxidation.

Conclusions

The PEC oxidation of chromium ions from the trivalent to the hexavalent form has been demonstrated for the first time by using WO₃ nanosponge photoanodes in sulfuric acid electrolyte under simulated sunlight. Faradaic efficiencies of almost 100% were achieved using electrolytes containing high Cr³⁺ (≥10 mM) and low SO₄²⁻ (≤10 mM) concentrations at very low bias potential voltages below 1.1 V_RHE, as the production of Cr⁶⁺ ions was able to outcompete the production of S₂O₈²⁻ ions from HSO₄⁻/SO₄²⁻ ions at the photoanode. In the Na₂SO₄ electrolyte, the faradaic efficiency was 100% even at 1.5 V_RHE. At the cathode, hydrogen evolution occurred with 100% faradaic efficiency. Moreover, the recovery of Cr³⁺ was also confirmed at the cathode by switching the electrolytes of the anode and cathode cells after the Cr⁶⁺ production, indicating that the complete cyclic PEC reaction between the Cr³⁺ and Cr⁶⁺ ions is possible. Finally, we have demonstrated the applicability of chromic acid oxidation to quinone production by using the photoelectrochemically produced Cr⁶⁺ ions in sulfuric acid solution. As such, p-benzenediol was successfully converted into p-benzoquinone with high efficiency. We have therefore shown that high-value-added reagents can be produced via photoelectrochemically driven chromic acid oxidation with the recycling of Cr⁶⁺ from Cr³⁺ using solar energy. All of these reactions can be performed in a single closed reaction vessel, permitting the very safe use of Cr⁶⁺ ions and suggesting a renewed role for chromium ions in future industrial applications.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank Dr S. Iguchi and Dr Y. Miseki for valuable discussions. This research was partially supported by the International Joint Research Program for Innovative Energy Technology of the Ministry of Economy, Trade and Industry (METI).

Notes and references

1. A. Lennartson, Nat. Chem., 2014, 6, 942.
2. G. Anger, J. Halstenberg, K. Hochgeschwender, C. Scherhag, U. Korallus, H. Knopf, P. Schmidt and M. Ohlinger, Chromium Compounds, in Ullmann's Encyclopedia of Industrial Chemistry, J. Wiley & Sons, New York, 2000.
3. M. Owlad, M. K. Aroua, W. A. W. Daud and S. Baroutian, Water, Air, Soil Pollut., 2009, 200, 59.
4. S. Langard, Am. J. Ind. Med., 1990, 17, 189.
5. M. Cieślak-Golonka, Polyhedron, 1996, 15, 3667.
6. W. Mertz, J. Nutr., 1993, 123, 1993.
7. S. Langard and T. Vigander, Br. J. Ind. Med., 1983, 40, 71.
8. H. C. Brown and C. P. Garg, J. Am. Chem. Soc., 1961, 83, 2952.
9. F. H. Westheimer, Chem. Rev., 1949, 45, 419.
10. G. Cainelli, G. Cardillo, M. Orena and S. Sandri, J. Am. Chem. Soc., 1976, 98, 6737.
11. F. Holloway, M. Cohen and F. H. Westheimer, J. Am. Chem. Soc., 1951, 73, 65.
12. L. F. Fieser, J. Am. Chem. Soc., 1948, 70, 3165.
13. K. W. Wellington, RSC Adv., 2015, 5, 20309.
14. D. G. Barceloux, J. Toxicol., Clin. Toxicol., 1999, 37, 173.
15. T. Takata, A. Tanaka, M. Hara, J. N. Kondo and K. Domen, Catal. Today, 1998, 44, 17.
16. A. Steinfeld, Sol. Energy, 2005, 78, 603.
17. D. Gust, T. A. Moore and A. L. Moore, Acc. Chem. Res., 2009, 42, 1890.
18. T. Nakajima, T. Nakamura, K. Shinoda and T. Tsuchiya, J. Mater. Chem. A, 2014, 2, 6762.
19. R. Saito, Y. Miseki and K. Sayama, Chem. Commun., 2012, 48, 3833.
20. U. Kammel and T. Schulzke, Chem. Ing. Tech., 1995, 67, 495.
21. T.-K. Yang and C.-Y. Shen, 1,4-Benzoquinone, in Encyclopedia of Reagents for Organic Synthesis, ed. L. Paquette, J. Wiley & Sons, New York, 2004.
22. T. Nakajima, T. Kitamura and T. Tsuchiya, Appl. Catal., B, 2011, 108–109, 47.
23. T. Nakajima, A. Hagino, T. Nakamura, T. Tsuchiya and K. Sayama, J. Mater. Chem. A, 2016, 4, 17809.
24. Q. Mi, A. Zhanaidarova, B. S. Brunschwig, H. B. Gray and N. S. Lewis, Energy Environ. Sci., 2012, 5, 5694.
25. K. Fuku, N. Wang, Y. Miseki, T. Funaki and K. Sayama, ChemSusChem, 2015, 8, 1593.
26. S. H. Ahn, J. Zhao, J. H. Kim and X. Zheng, Electrochim. Acta, 2017, 244, 184.
27. J. C. Hill and K.-S. Choi, J. Phys. Chem. C, 2012, 116, 7612.
28. C. Liu, Y. Ding, W. Wu and Y. Teng, Chem. Eng. J., 2016, 306, 22.
29. Y. Zhang, Q. Wang, J. Lu, Q. Wang and Y. Cong, Chemosphere, 2016, 162, 55.
30. X. Feng, J. Shang and T. Zhu, Electrochim. Acta, 2016, 188, 752.
31. H.-T. Hsu, S.-S. Chen, Y.-F. Tang and H.-C. Hsi, J. Hazard. Mater., 2013, 248–249, 97.
32. H. Song, J. Shang, J. Ye and Q. Li, Thin Solid Films, 2014, 551, 158.
33. M. Miyauchi, A. Nakajima, T. Watanabe and K. Hashimoto, Chem. Mater., 2002, 14, 4714.
34. P. S. Kumar, J. Sundaramurthy, S. Sundarrajan, V. J. Babu, G. Singh, S. I. Allakhverdiev and S. Ramakrishna, Energy Environ. Sci., 2014, 7, 3192.
35. B. D. Alexander, P. J. Kulesza, I. Rutkowska, R. Solarska and J. Augustynski, J. Mater. Chem., 2008, 18, 2298.
36. G. Wang, Y. Ling, X. Lu, T. Zhai, F. Qian, Y. Tong and Y. Li, Nanoscale, 2013, 5, 4129.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ta08001h

---