Photocatalytic reduction of aqueous chromium(VI) by RuO₂/g-C₃N₄ composite under visible light irradiation

Abstract

Graphitic carbon nitride (g-C₃N₄) has been extensively investigated as a novel nonmetallic visible-light response photocatalyst. However, its uses in photocatalytic reductions were limited because of the sluggish oxygen evolution reaction (OER) and the resulting self-decomposition. In this paper, a ruthenium dioxide loaded g-C₃N₄ composite (RuO₂/g-C₃N₄) was prepared by forced oxidative hydrolysis of ruthenium(III) chloride on the surface of g-C₃N₄ that was obtained by direct condensation polymerization of melamine. Photocatalytic reduction of aqueous Cr(VI) by it under illumination from a 400–410 nm light-emitting diode was examined. It was shown that the Cr(VI) reduction rate was much higher in RuO₂/g-C₃N₄ than in pure g-C₃N₄. Without any sacrificial electron donor and at initial solution pH 2.3, Cr(VI) removal (200 mL and 0.5 mM) was 34% and 76.4% with 0.1 g pure g-C₃N₄ and 0.1 RuO₂ (1.0 wt%)/g-C₃N₄, respectively. The optimum initial solution pH was 2.4. Methanol accelerates while acetone suppresses the Cr(VI) reduction significantly. Ferric ions catalyze the reduction, especially in the later stage. UV-Vis diffusion reflectance spectroscopy and theoretical analysis showed that RuO₂ not only boosts the charge separation but also protects the g-C₃N₄ from decomposition by its extraordinary catalytic action for OER. The used RuO₂/g-C₃N₄ was separated from the solution by microfiltration, with little leaching and residue remaining in the filtrate. The reclaimed RuO₂/g-C₃N₄ was recycled for 5 cycles and no obvious decrease in catalytic activity was observed, indicating its superior potential in industrial applications.

---

1. Introduction

Chromium is a common pollutant in water, where it usually exists in the extremely toxic and carcinogenic hexavalent [Cr(VI)] form.¹ Converting Cr(VI) to much less toxic and precipitable trivalent form [Cr(III)] via chemical reduction is a key step to treat the Cr(VI) contaminated wastewater. However, the chemical reduction needs excess reducing agents and acids to drive the reaction, adding tremendous amounts of neutralizing reagents in the subsequent precipitation.^2,3 Therefore, developing atom-economic processes for Cr(VI) reduction is in great need.

Recently, graphite carbon nitride (g-C₃N₄) has emerged as a novel nonmetallic photocatalyst due to its narrow band gap (2.7 eV), visible light response capability, rich sources of precursors and simple preparation method.⁴ In theory, g-C₃N₄ can be applied for reduction of heavy metal ions because of its highly active conduction band electrons (E_cb = −1.1 eV vs. NHE, slightly dependent on fabrication method and reaction procedure). However, pure g-C₃N₄ exhibits low photocatalytic activity due to its high recombination speed of photo-generated charge carriers and its low conductivity. Constructing g-C₃N₄-based composites is considered as a promising strategy in promoting photocatalytic reductions. Wang et al.⁵ found that the 1D black phosphorus-tubular g-C₃N₄ can remove 94.1% of Cr(VI) with a rate constant of 0.0404 min⁻¹. Ren et al.⁶ constructed a g-C₃N₄/NH₂-UiO-66(Zr) heterojuncter by solvothermal and in situ deposition to effect both Cr(VI) reduction and tetracycline hydrochloride oxidation in aqueous solution, where the photocatalytic removal of Cr(VI) by CU-20 wt% forming heterojunction was 1.86 times that of pure NH₂-UiO-66(Zr) under visible light irradiation. Eslamlu et al.⁷ reported that Sb₂MoO₆ coupled with g-C₃N₄ nano-tubes showed Cr(VI) reduction efficiency of 22 times higher than the bare g-C₃N₄. When MoS₂/g-C₃N₄ was grafted with cyclodextrins, Cr(VI) reduction in the simulated agricultural wastewater was remarkably enhanced.⁸ Mohamed et al.⁹ synthetized mesoporous BiVO₄/2D-g-C₃N₄ heterostructures for superior visible light-driven photocatalytic reduction of Hg(II) in the presence of HCOOH.

Metal oxides or metal sulfides were usually employed to couple with g-C₃N₄ to improve the photocatalytic efficiency. However, most of the metal oxides or metal sulfides are, due to their chemical nature, not durable enough in corrosive media, such as acidic, alkaline, reducing, oxidizing or chelating conditions which are frequently encountered in real wastewater samples. Stability of the applied catalysts is pivotal in real wastewater treatment, as the decomposition or leaching not only results in cost rising but also polluting the water for treatment. At present, little work concerning the catalyst leaching was reported. When the photogenerated electrons (e⁻) are captured by the heavy metal ions, the holes (h⁺) will inevitably accumulate in the valence band. Although the band potential of g-C₃N₄ (E_VB ∼ 1.5 V) is higher than that of oxygen (O₂), (2H₂O − 4e⁻ → O₂ + 4H⁺, ϕ_O2/H2O = 1.23 V), it is not positive enough to oxidize water to O₂ due to the large kinetic barrier in the four-electron oxygen evolution reaction (OER), which makes OER in conventional g-C₃N₄–metal oxide heterojuncters particularly sluggish. If the accumulated h⁺ were not effectively removed by water molecule, they would oxidize the catalyst itself, leading to the g-C₃N₄ break down.¹⁰ Although g-C₃N₄ can be protected by using sacrificial electron donor (usually organic additives such as alcohols¹⁰), such a strategy works well only when the real wastewater contains both Cr(VI) and organic pollutants. As some Cr(VI) wastewater contains organic pollutants and some not, searching efficient and stable heterostructure-forming units that can not only enhance the separation of photogenerated charge carriers but also promote OER during g-C₃N₄ photocatalytic reduction is of paramount importance. In this regard, ruthenium oxide (RuO₂) is with no doubt an appealing material for forming heterostructures due to its superior chemical stability,^11,12 metallic conductivity [2.0–2.5 × 10⁴ S cm⁻¹], and high catalytic activity for the OER¹³ for its optimal oxygen binding energy. In addition, the precursor of RuO₂ is relatively lower cost as compared to other precious metals. RuO₂ nanoparticles-accommodated g-C₃N₄ for photocatalytic oxidation of trichloroethylene has been reported.¹⁴ However, its action in promoting photocatalytic reductions is rarely reported. In this report, effect and mechanism of RuO₂ on promoting the g-C₃N₄ photocatalytic Cr(VI) reduction was examined and explored. It was found that small amount of RuO₂ deposition on g-C₃N₄ can improve its photocatalytic reduction activity greatly, even without the addition of extra organic additives. The composite is stable in the reaction but also protects the g-C₃N₄ from decomposition, which means that the RuO₂ has multiple effect on the photocatalytic ability and greatly broaden the application of RuO₂/g-C₃N₄ composite in wastewater treatment.

2. Experimental

2.1. Materials

Ruthenium chloride (RuCl₃·3H₂O) and melamine (C₃H₆N₆, 99%) were purchased from Tianjin Meiske Chemical Co., Ltd and Tianjin Zhiyuan Chemical Reagent Co., Ltd respectively. Potassium dichromate (K₂Cr₂O₇), methanol (CH₃OH), 1,5-diphenylcarbazide, acetone (CH₃COCH₃) and tert-butanol ((CH₃)₃COH) were all of analytical reagent grade and used as received. Ultra-pure water for preparation of any solution was obtained from the Milli-Q® system.

2.2. Syntheses of g-C₃N₄ and RuO₂/g-C₃N₄ composite

g-C₃N₄ was prepared by calcining melamine in a muffle furnace.¹⁵ Typically, 10 g of melamine was placed into an alumina crucible with a cover and heated from room temperature with a ramp rate of 10 °C min⁻¹ to 520 °C and then sustained for 4 h in a muffle furnace. After natural cooling, product in the crucible was transferred into a mortar to be grinded into yellow g-C₃N₄ powder (20 min) for later use.

For the synthesis of RuO₂/g-C₃N₄ composite, in situ deposition by forced hydrolysis and oxidation of ruthenium chloride was adopted.¹¹ Briefly, 0.02 g of RuCl₃·3H₂O was dissolved in 30 mL of 60 °C aqueous g-C₃N₄ suspension with stirring for 60 min. The obtained mixture was evaporated to get the gel. The resulting gel was dried in an electrothermal drier at 105 °C for 24 h and then was added into an alumina crucible with a cover and heated from room temperature with a ramp rate of 10 °C min⁻¹ and then sustained at 520 °C for 30 min. Finally, RuO₂/g-C₃N₄ composite was obtained and labeled as RuO₂ (x%)/g-C₃N₄, where “x%” denoted the mass fraction of RuO₂ in the composite.

2.3. Characterization

X-ray diffraction (XRD) patterns of the pure g-C₃N₄ and RuO₂/g-C₃N₄ composite were examined by a Rigaku D/Max-Ultima⁺ diffractometer equipped with Kα radiation of Cu (λ = 0.15418 nm). Transmission electron microscopy (TEM) of g-C₃N₄ and RuO₂/g-C₃N₄ were investigated operating at 200 kV through a JEOL JEM-2100 electron microscope. X-ray photoemission spectroscopy (XPS) analysis was evaluated in a Kratos-AXIS ULTRA DLD equipped with a monochromatic Al Kα X-ray source. UV-Vis diffuse reflectance spectra of g-C₃N₄ and RuO₂/g-C₃N₄ were analyzed on a TU-1901 UV/Vis spectrophotometer with an IS19-1 integrating sphere to collect the diffusing light with BaSO₄ as the reference. FTIR spectra of the RuO₂/g-C₃N₄ KBr mixed disks, were recorded utilizing a Thermo Nicolet is5 (400 to 4000 cm⁻¹). PL spectra were recorded by RF-5301 PC spectrophotometer using a 150 W xenon lamp at λ ∼ 365 nm.

2.4. Photocatalytic reduction of aqueous Cr(VI)

The photocatalytic reduction activity of g-C₃N₄ and RuO₂/g-C₃N₄ composite was tested for Cr(VI) reduction in an aqueous solution. In each typical run, 0.1 g of the prepared photocatalyst was added to 200 mL of 0.5 mM Cr(VI) solution. The mixture was magnetically stirred without illumination for 30 min to attain the adsorption–desorption equilibrium. The suspension was then illuminated by an LED lamp (40 W) surrounded outside the reactor with wavelength centered at 407.4 nm with full-width at half-maximum (FWHM) of 24.7 nm for 2 h with intensity of 25 mW cm⁻², where the emission spectra of the LED lamp was presented in Fig. 1.

Fig. 1 Emission spectra of the LED lamp used for the photocatalytic reduction of Cr(VI).

During the illumination, 4.0 mL of the suspension was taken out from the reactor every 15 min, followed by filtration with 0.22 μm membrane and then the filtrate was subjected to chemical analyses. Cr(VI) concentration in the filtrate was determined spectrophotometrically at 520 nm where 1,5-diphenylcarbazide was used as the coloration developer. The photocatalytic reduction efficiency (η) of the prepared catalysts was determined as:

where C₀ and C_t is the Cr(VI) concentration at the beginning of illumination and illumination time t (min), respectively. Possible ruthenium leaching from the catalyst was analyzed by an inductively coupled plasma atomic emission spectrometer (ICPAES, Prodigy, Leeman Laboratories). The residue catalysts were evaluated by TOC analysis of the filtrate. Oxygen and nitrogen evolved during the photocatalytic reactions were collected by an air bag and analyzed with a gas analyzer (ST8100A, Smart Sensor Co. Ltd, China).

3. Results and discussions

3.1. Catalyst characterizations

Fig. 2 shows the XRD patterns of the prepared bare g-C₃N₄ and RuO₂/g-C₃N₄ composite, respectively.

Fig. 2 XRD patterns of bare g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ composite.

As shown in Fig. 2, two distinct diffraction peaks were observed for pure g-C₃N₄. The weak low-angle reflection peak at 12.74° (d₁₀₀ = 0.694 nm) was originated from in-planar repeating of tri-s-triazine (melem) unit and the peak centered at 27.42° was attributed to the periodic interlayer-stacking (d₀₀₂ = 0.325 nm) of the polymeric melon, implying successful condensation of melamine and the distinctive graphitic structure of C₃N₄ formed.⁴ XRD pattern of RuO₂ (1.0%)/g-C₃N₄ composite was quite similar to that of bare g-C₃N₄, indicating that the composition and structure of g-C₃N₄ was not altered appreciably with deposition of the RuO₂. In addition, no peaks for RuO₂ were observed, possibly because of its low contents.¹⁴

According to the Scherrer's formula, the thickness (D₀₀₂, nm) of bare g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ composite can be estimated by:

where λ is wavelength of the X-ray used for diffraction (nm), B₀₀₂ is the FWHM of the diffraction peak at 27.42° (rad) and θ is the Bragg angle (°). It can be calculated that the thicknesses of g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ were about 5.9 nm and 5.7 nm, respectively. As the interlayer-stacking of g-C₃N₄ is ca. 0.325 nm, the sheet of the above catalysts consists of 18–19 layers of polymeric melons.¹⁵ Nanosheets usually possess high conductivity, which is beneficial to the charge transfer.¹⁶

FTIR spectra of bare g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ composite were determined as shown in Fig. 3.

Fig. 3 FTIR spectra of bare g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ composite.

The peak at 806 cm⁻¹ is originated to the breathing of the heptazine ring system. The absorption bands between 1200 and 1700 cm⁻¹ indicate the presence of C–N heterocycles. The broad peak at 3164 cm⁻¹ was due to the stretching of the terminal N–H from the uncondensed amine.¹⁴ FTIR spectra of RuO₂/g-C₃N₄ composite are comparable to those of bare g-C₃N₄ in the lower wavenumber range. However, the peak intensities of RuO₂/g-C₃N₄ composite were a little weaker than those of bare g-C₃N₄,¹⁵ indicating loss of N–H bond during the preparation of RuO₂/g-C₃N₄ composite.

Representative TEM images of bare g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ composite were presented in Fig. 4a and b respectively.

Fig. 4 TEM images of pure g-C₃N₄ (left) and RuO₂ (1.0%)/g-C₃N₄ composite (right).

The TEM image of pure g-C₃N₄ indicated a wrinkled-layer structure along with some stacking layers that have a thin sheet and a typical lamellar morphology. The TEM image of RuO₂ (1.0%)/g-C₃N₄ composite revealed that small RuO₂ particles are evenly dispersed on the g-C₃N₄ surface. The size of RuO₂ particles is in the range 10–20 nm. HRTEM images and STEM-EDS mapping of RuO₂ (1.0%)/g-C₃N₄ composite were presented in Fig. S1 and S2,† respectively. It can be clearly observed from Fig. S1† that the lattice fringe of RuO₂ was present in the sample. Fig. S2† indicated the distribution of Ru agrees well with those of both dark filed and bright field mapping but differs slightly from that of oxygen, meaning that the oxygen not only comes from RuO₂ but also from g-C₃N₄ (Fig. 5). These results verified that RuO₂ is tightly contacted with g-C₃N₄, forming the RuO₂/g-C₃N₄ heterojuncters.

Fig. 5 XPS survey spectra of g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄ (upper), high resolution XPS of RuO₂ (1.0%)/g-C₃N₄ (278–294 eV, middle) and Ru 3d scan of RuO₂ (1.0%)/g-C₃N₄ (lower).

To further identify the chemical structure of g-C₃N₄ and RuO₂/g-C₃N₄, XPS analyses were performed and given in Fig. 5. The XPS survey spectra of g-C₃N₄ confirm that g-C₃N₄ is mainly composed of carbon and nitrogen, with a small amount of oxygen. The existence of oxygen in g-C₃N₄ is likely due to oxidation during the condensation polymerization. It is noteworthy that the relative intensity of nitrogen peak in g-C₃N₄ is higher than that in RuO₂/g-C₃N₄. The higher nitrogen content in g-C₃N₄ can be attributed to the preservation of –NH₂ and NH groups. The peaks located at 280.62 eV and 282.53 eV belong to 3d_5/2 of RuO₂.¹⁴ The theoretical peaks of Ru 3d_3/2 should be observed at 284.88 eV and 286.78 eV, however as they are partially overlapped with those from C 1s, making it difficult to differentiate them, which is consistent with the results of Hwang et al.¹⁷ RuO₂ exists in Ru⁴⁺ state in RuO₂ (1.0%)/g-C₃N₄, indicating the successful deposition of RuO₂ on the g-C₃N₄ surface.

UV-Vis absorption spectra of the pure g-C₃N₄ and the RuO₂/g-C₃N₄ composite at various concentrations were illustrated in Fig. 6. It can be observed from Fig. 6 that all of them were capable of visible light absorption. In general, absorption of the RuO₂/g-C₃N₄ composite was stronger than the bare g-C₃N₄, especially in the visible range. However, the absorption of RuO₂/g-C₃N₄ composite was weaker than bare g-C₃N₄ in the range 397–434 nm, which is different from those reported earlier.¹⁴ Outside the wavelength range, the absorption decreases with increasing RuO₂ loading.^15,16

Fig. 6 UV-Vis absorption spectra of pure g-C₃N₄ and RuO₂/g-C₃N₄ composite.

3.2. Photocatalytic reduction of Cr(VI) by g-C₃N₄ and RuO₂/g-C₃N₄

Variations in Cr(VI) concentration with illumination time in the presence of 0.1 g of pure g-C₃N₄, pure RuO₂ and RuO₂ (1.0)/g-C₃N₄ composite were shown in Fig. 7, respectively.

Fig. 7 Photocatalytic reductions of Cr(VI) by pure g-C₃N₄, pure RuO₂ and RuO₂ (1.0%)/g-C₃N₄ composite (solution volume, 200 mL; catalyst, 0.1 g; initial Cr(VI) concentration, 0.5 mM; initial pH, 2.3; light source, LED (410 nm, 40 W)).

As seen in Fig. 7, Cr(VI) concentration decreased smoothly with the illumination time. However, the Cr(VI) concentration decreased much more rapidly in the case of RuO₂/g-C₃N₄ composite than pure g-C₃N₄ or RuO₂. After 90 min of illumination, the Cr(VI) removal is ca. 76.4% for RuO₂ (1.0%)/g-C₃N₄ and 33.8% for pure g-C₃N₄, while it is negligible for pure RuO₂ (<5%). The Cr(VI) reduction rate (calculated at the initial stage) with RuO₂ (1.0%)/g-C₃N₄ composite is 4.6 times that of the g-C₃N₄, which obviously confirms the promoting effect of RuO₂ in photocatalytic reduction activity of g-C₃N₄. It is noted that the Cr(VI) removal is much less when the combination of RuO₂ and g-C₃N₄ was used to reduce Cr(VI), illustrating the successful preparation of heterojunction. In addition, Fig. 7 also showed that the Cr(VI) removal due to adsorption and the direct photolysis were both negligible.^18,19 Therefore, the present experiments do not consider the direct photolysis of Cr(VI) in the subsequent experiments. When the illumination time is increased to 150 min, the Cr(VI) is below the detection limit (not shown in the figure) in the case of RuO₂ (1.0%)/g-C₃N₄, which means that Cr(VI) can be totally reduced without any sacrificial electron donor.

3.3. Effect of initial pH on Cr(VI) photocatalytic reduction

The natural pH of the Cr(VI) solution prepared by dissolving K₂Cr₂O₇ in ultrapure water is about 5.5, at which little Cr(VI) reduction was observed under photocatalytic conditions. Therefore, we studied Cr(VI) reduction in acidic medium. Fig. 8 shows Cr(VI) photocatalytic reductions in the presence of 0.1 g RuO₂ (1.0%)/g-C₃N₄ composite under different initial pH (pH₀) values.

Fig. 8 Photocatalytic reduction of Cr(VI) with RuO₂ (1.0%)/g-C₃N₄ under different pH₀ values (solution volume, 200 mL; catalyst, 0.1 g; initial Cr(VI) concentration, 0.5 mM; light source, LED (410 nm, 40 W)).

It is demonstrated from Fig. 8 that the Cr(VI) reduction proceeds more rapidly in lower pH₀. However, the trend is reversed at pH₀ 2.4. After 90 min of illumination, the Cr(VI) removal is 40% at pH₀ 3.0 and increases to 95% at pH₀ 2.4 and drops to 88% at pH₀ 2.2. The above phenomena can be explained as follows.

In the present experimental conditions, Cr(VI) exists mainly in the form of dichromate (Cr₂O₇²⁻) and hydrogen chromate (HCrO₄⁻). Their mutual relations can be described by reactions (3) and (4).^20,21

Cr₂O₇²⁻ + H₂O ⇄ 2HCrO₄⁻ K₃ = 0.22 M (3)

HCrO₄⁻ ⇄ CrO₄²⁻ + H⁺ K₄ = 3.2 × 10⁻⁷ M (4)

Over 96% of Cr(VI) is present in the form HCrO₄⁻ and ca. 4% is in the form Cr₂O₇²⁻ within pH range 2.0–5.0 at 0.5 mM Cr(VI). As the reaction between HCrO₄⁻ and e⁻ is promoted by H⁺ (reaction (5)), the Cr(VI) reduction rate increases as the solution pH decreases.²¹

HCrO₄⁻ + e⁻ + H⁺ → H₂CrO₄⁻ (Cr(V)) (5)

At high concentrations of H⁺, completing reaction between H⁺ and e⁻ (reaction (6)) prevails, which leads to less e⁻ available for Cr(VI) reduction. As a result, the Cr(VI) reduction rate decreased with further decreasing pH₀:⁸

2H⁺ + 2e⁻ → H₂ (6)

On the other hand, when using sulfuric acid to lower the solution pH, hydrogen sulfate ion will be inevitably formed, and it will react with HCrO₄⁻ to form CrSO₇²⁻, which decreased the effective HCrO₄⁻ concentration and the Cr(VI) reduction would slow down in highly acidic condition (eqn (7)):²¹

HCrO₄⁻ + HSO₄⁻ ⇄ CrSO₇²⁻ + H₂O K₅ = 0.42 M⁻¹ (7)

To further elucidate the pH role in Cr(VI) reduction, variations of solution pH during illumination in the presence and absence of Cr(VI) are presented in Fig. 9.

Fig. 9 Variations of pH during illumination with and without Cr(VI) (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

It can be seen from Fig. 9 that the pH of solution containing Cr(VI) increases apparently with illumination time. However, without Cr(VI), pH of the solution changes little. This can be explained by the fact that the Cr(VI) reduction consumes H⁺, as the overall Cr(VI) reduction stoichiometry can be represented by reactions (8) and (9):

HCrO₄⁻ + 3e⁻ + 7H⁺ → Cr³⁺ + 4H₂O (8)

Reactions (8) and (9) need the involvement of H⁺ from the dynamic aspect. As shown in Fig. 8, Cr(VI) reduction proceeded fastest at pH₀ 2.4, and pH₀ 2.4 is chosen as the optimum pH in the following investigations.

3.4. Effect of electron and hole scavengers on Cr(VI) reduction

In order to further improve the reduction efficiency of Cr(VI), small amount of organic additives was added to the solution to explore whether the organic solvent will inhibit or promote the reduction of Cr(VI). Fig. 10 shows the Cr(VI) concentration variations in the presence of methanol and acetone, respectively.

Fig. 10 Photocatalytic reductions of Cr(VI) with RuO₂ (1.0%)/g-C₃N₄ in the presence of methanol and acetone (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

It is clearly shown from Fig. 10 that acetone suppresses while methanol promotes the Cr(VI) reduction dramatically. Cr(VI) removal rate was more than doubled and three quarters decreased in the presences of 0.05 M methanol and acetone, respectively. The above phenomena can be explained as follows.

Upon illumination, e⁻ and h⁺ were simultaneously produced. The h⁺ can re-oxidize the Cr(III) back to Cr(VI) as the oxidation potential of h⁺ (ca. 1.5 V) is 0.2 V higher than that of Cr(VI) (ca. 1.3 V, reaction (9)). In addition, the h⁺ can recombine with e⁻ to decrease the number of e⁻ available for Cr(VI) reduction. The most desirable way is to convert the h⁺ into organic radicals possessing reduction potentials in the range of −1 to −2 V and are thus capable of reducing Cr(VI) to lower oxidation states, which not only inhibits the re-oxidation of Cr(III) and increases the concentration of e⁻, but also converts the oxidizing h⁺ to the reducing ones:²²

As reaction (10) goes, more e⁻ are available for Cr(VI) reduction. Consequently, the Cr(VI) reduction rate increased in the presence of methanol. On the other hand, as acetone is very stable and cannot be oxidized by the h⁺ but can react with e⁻ as the following:^23–25

As reaction (11) involved in the process, less e⁻ is available for Cr(VI) reduction and the Cr(VI) reduction rate decreased in the presence of acetone.

3.5. Effect of Fe(III) on Cr(VI) photocatalytic reduction

Fe(III) is an ubiquitous and nontoxic elements in earth, and investigate its action on photocatalytic reduction of Cr(VI) is of great significance. Fig. 11 compares the Cr(VI) concentration variations in the presence and absence of 0.05 mM Fe(III), respectively.

Fig. 11 Photocatalytic reductions of Cr(VI) with RuO₂ (1.0%)/g-C₃N₄ in the presence and absence of 0.05 mM Fe(III) (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

It can be clearly observed from Fig. 11 that at the initial stage (<20 min), Fe(III) displayed little effect on the reduction of Cr(VI). However, the effect becomes apparent at the later stage. The Cr(VI) removal of Cr(VI) at 75 min can reach 100% with 0.05 mM Fe(III), where it is only 80% without it. When Fe(III) was added, it was reduced to Fe²⁺ by the e⁻, and then the resulting Fe²⁺ reduces Cr(VI) to Cr(III) and next circle begins, which means that Fe(III) can be used as a co-catalyzer.³

Fe(III) + e⁻ → Fe²⁺ (12)

3Fe²⁺ + Cr(VI) → 3Fe(III) + Cr(III) (13)

It is noted that as isopropanol ultimately generated by the reaction (11) cannot reduce Cr(VI), the reduction of Cr(VI) is suppressed. Therefore, although both Fe(III) and acetone belong to electron scavengers, they showed opposing effects on Cr(VI) reductions.

3.6. Mechanism of photocatalytic reduction of Cr(VI) by RuO₂/g-C₃N₄

To elucidate the mechanism in Cr(VI) reduction, evolution of UV-Vis absorption spectra of the solution during photocatalytic reduction are presented in Fig. 12.

Fig. 12 Evolution of UV-Vis absorption spectra of the solution during photocatalytic reduction of Cr(VI) (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

As shown in Fig. 12, the peak height at about 350 nm, which is characteristic of the absorption band of Cr(VI) species,²¹ together with that at 285 nm, gradually decreases with the illumination time. In general, Cr(VI) would undergo a series of intermediate processes before being finally converted to Cr(III). As shown in Fig. 12, no new bands appeared during the reduction, indicating that the possible Cr(IV) and Cr(IV) were too short-lived to be detected by the present technique.^26,27 In the experiment, color of the solution changed from orange to light yellow and finally to colorless, which was verified by the decrease in absorption band from 400 to 450 nm.

According to ref. 14, g-C₃N₄ and RuO₂ are both n-type semiconductors. The values of E_CB and E_VB for pure g-C₃N₄ are −1.125 eV and +1.585 eV,²⁸ respectively. E_CB and E_VB of RuO₂ can be calculated according to the formula (14) and (15):²⁹

E_CB = E_VB − E_g (14)

E_VB = X − E_e + 0.5E_g (15)

where X is the absolute electronegativity of RuO₂ (6.35 eV),²⁷ E_e is the vacuum electron level corresponding to the standard hydrogen (4.5 eV). E_g of RuO₂ was determined to be 2.32 eV.¹⁴ It can be calculated that E_CB and E_CB of RuO₂ are 3.01 eV and 0.69 eV respectively. Because the Fermi level (E_F) of n-type semiconductor is very close to its E_CB, the Fermi levels of g-C₃N₄ and RuO₂ are about −1.125 eV and −0.69 eV. As the E_F of g-C₃N₄ is lower than that of RuO₂, when g-C₃N₄ and RuO₂ contact each other, the electrons on g-C₃N₄ automatically flow to the conduction band of RuO₂, leaving positive charges, thus forming a built-in electric field between RuO₂ and g-C₃N₄ (g-C₃N₄ pointing to RuO₂). When g-C₃N₄ absorbs visible light, electron hole pairs (e⁻ + h⁺) were generated. Under the action of built-in electric field, the e⁻ move towards the solution and are captured by Cr(VI) in the solution, resulting in the reduction of Cr(VI) and suppress the electron hole recombination. Such explanation was documented by Fig. 13.

Fig. 13 Fluorescence emission spectra of g-C₃N₄ and RuO₂ (1.0%)/g-C₃N₄.

It can be shown in Fig. 13 that the emission intensity of RuO₂/g-C₃N₄ was much lower than g-C₃N₄, showing the suppressed recombination of photo-carriers.

In addition, the absorption of RuO₂/g-C₃N₄ is slightly weaker than pure g-C₃N₄ near 410 nm as shown in Fig. 6, indicating that the role of RuO₂ belongs to the enhancement of charge separation rather than the enhancement of light absorption, which also indicates that the mechanism speculation is reasonable. The process can be demonstrated in Fig. 14:

Fig. 14 Mechanism of photocatalytic reduction of Cr(VI) by RuO₂/g-C₃N₄.

On the other hand, the loaded RuO₂ is able to take up the h⁺ from g-C₃N₄, exhibiting functionality as efficient O₂ evolution sites. The reaction sequence for O₂ evolution can be expressed as follows:³⁰

RuO₂ + 2h⁺ → RuO₂²⁺ (16)

RuO₂²⁺ + H₂O → RuO₃ + 2H⁺ (17)

2RuO₃ → 2RuO₂ + O₂(g) (18)

In order to confirm the above assumptions, time courses of O₂ and N₂ evolution during the illumination were examined and the results are presented in Fig. 15.

Fig. 15 Time courses of O₂ and N₂ evolution during the illumination (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

It is demonstrated from Fig. 15 that RuO₂ (1.0%)/g-C₃N₄ composite exhibited high activity for O₂ evolution with no N₂ evolution. This further indicated that RuO₂ loading not only boost the separation of the charge carriers but also protects the decomposition of g-C₃N₄.

3.7. Recyclability and stability test

To evaluate the reusability and stability of the samples, five successive cycles of experiments were performed for Cr(VI) reduction. After completion of each reduction reaction (90 min), the catalyst in the solution was filtered with 0.22 μm filter membrane, washed with ethanol and water for 3 times respectively, put it into the oven, and bake at 60 °C for 24 hours for the next experiment. Namely, the used sample after each cycle was collected through centrifugation and rinsed three times with ethanol and ultrapure water. Subsequently, the washed powder was dried for 8 h (60 °C) to conduct next photocatalytic cycle. Moreover, the XRD of the samples after six cycles was characterized and compared to that of the fresh samples. A total of 5 cycles were carried out. The reduction ratio of Cr(VI) at each final cycle is shown in Fig. 16.

Fig. 16 Cycle stability test of RuO₂ (1.0%)/g-C₃N₄ composite (solution volume, 200 mL; RuO₂ (1.0%)/g-C₃N₄ composite, 0.1 g; initial Cr(VI) concentration, 0.5 mM; pH₀, 2.4; light source, LED (410 nm, 40 W)).

It can be seen from Fig. 16 that the Cr(VI) reduction ratio with recycled RuO₂ (1.0%)/g-C₃N₄ still maintains more than 80% after five repeated experiments, indicating that the photocatalytic stability of RuO₂ (1.0%)/g-C₃N₄ is still very good and can be reused. ICP-AES test showed no leaching of Ru form RuO₂ (1.0%)/g-C₃N₄. In addition, TOC analysis also showed no apparent carbon and nitrogen increase in the solution, indicating that RuO₂ (1.0%)/g-C₃N₄ is very stable during the photocatalysis.

4. Conclusions

RuO₂/g-C₃N₄ composite can be easily prepared by in situ forced hydrolysis and oxidation and is able to catalytically photo-reduce Cr(VI) effectively even in the absence of sacrificial electron donor. The optimum initial pH for the reduction is 2.4. Acetone inhibits the reduction and methanol promotes the reduction. In addition, a small amount of Fe(III) catalyzes the reduction of Cr(VI), especially in the later stage.

The deposited RuO₂ particles not only functionalized as an effective charge separator for Cr(VI) reduction, but also protects the g-C₃N₄ from self-decomposition through catalyzing O₂ evolution, which is indispensable for the reduction without presence of electron donors. Little leaching and residue remained in the solution further proves its potential application in real wastewater treatment process.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article.

Author contributions

Conceptualization, Y. Liu; methodology, Y. Liu; software, Y. Liu; validation, Y. Liu; formal analysis, Y. Liu; investigation, X. Du and Z. Huang; resources, Y. Liu; data curation, Y. Liu; writing—original draft preparation, Y. Liu; writing—review and editing, Y. Liu; visualization, Y. Liu; supervision, Y. Liu; project administration, Y. Liu; funding acquisition, Y. Liu. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This research was funded by the National Natural Science Foundation of China grant number [11005014, 11675031] and the APC was funded by [11005014].

References

1. M. Ding and X. Shi, Molecular mechanisms of Cr(VI)-induced carcinogenesis, Mol. Cell. Biochem., 2002, 234–235, 293–300.
2. L. Wang and X. Jiang, Plasma-induced reduction of chromium(VI) in an aqueous solution, Environ. Sci. Technol., 2008, 42(22), 8492–8497.
3. L. E. Eary and R. Dhanpat, Chromate removal from aqueous wastes by reduction with ferrous ion, Environ. Sci. Technol., 1988, 22, 972–977.
4. X. Wang, K. Maeda and A. Thomas, et al., A metal-free polymeric photocatalyst for hydrogen production from water under visible light, Nat. Mater., 2009, 8(1), 76–80.
5. W. Wang, Q. Niu and G. Zeng, et al., 1D porous tubular g-C₃N₄ capture black phosphorus QDs as 1D/0D metal-free photocatalysts for oxytetracycline hydrochloride degradation and hexavalent chromium reduction, Appl. Catal., B, 2020, 273, 119051.
6. J. Ren, S. Lv and S. Wang, et al., Construction of efficient g-C₃N₄/NH₂-UiO-66(Zr) heterojunction photocatalysts for wastewater purification, Sep. Purif. Technol., 2021, 274, 118973.
7. P. H. Eslamlu, A. H. Yangjeh and S. A. Khaneghah, et al., Integration g-C₃N₄ nanotubes and Sb₂MoO₆ nanoparticles: impressive photoactivity for tetracycline degradation, Cr(VI) reduction, and organic dyes removals under visible light, Adv. Powder Technol., 2021, 32, 2322–2335.
8. Q. Zhong, H. Lan and M. Zhang, et al., Preparation of heterostructure g-C₃N₄/ZnO nanorods for high photocatalytic activity on different pollutants (MB, RhB, Cr(VI) and eosin), Ceram. Int., 2020, 46(8), 12192–12199.
9. R. M. Mohamed and A. A. Ismail, Mesoporous BiVO₄/2D-g-C₃N₄ heterostructures for superior visible light-driven photocatalytic reduction of Hg(II) ions, Ceram. Int., 2021, 472, 26063–26073.
10. K. Maeda, X. Wang and Y. Nishihara, et al., Photocatalytic activities of graphitic carbon nitride powder for water reduction and oxidation under visible light, J. Phys. Chem. C, 2009, 113(12), 4940–4947.
11. T. P. Luxton, M. J. Eick and K. G. Scheckel, Characterization and dissolution properties of ruthenium oxides, J. Colloid Interface Sci., 2011, 359(1), 30–39.
12. Y. Lee, J. Suntivich and K. J. May, et al., Synthesis and activities of rutile IrO₂ and RuO₂ nanoparticles for oxygen evolution in acid and alkaline solutions, J. Phys. Chem. Lett., 2012, 3(3), 399–404.
13. K. Kalyanasundaram and M. Grtzel, Cyclic cleavage of water into H₂ and O₂ by visible light with coupled redox catalysts, Angew. Chem., Int. Ed., 1979, 18(9), 701–702.
14. L. A. Al-Hajji, F. M. Alshareef and A. A. Ismail, et al., RuO₂ nanoparticles-accommodated graphitic carbon nitride for significant enhancement in photocatalytic oxidation of trichloroethylene, Opt. Mater., 2022, 125, 112086.
15. B. Li, K. Nie, Y. Zhang, L. Yi, Y. Yuan, S. Chong, Z. Liu and W. Huang, Engineering single-layer hollow structure of transition metal dichalcogenides with high 1T-phase purity for hydrogen evolution reaction, Adv. Mater., 2023, 35(46), 2303285.
16. Z. Liu, K. Nie, X. Qu, X. Li, B. Li, Y. Yuan, S. Chong, P. Liu, Y. Li, Z. Yin and W. Huang, General bottom-up colloidal synthesis of nano-monolayer transition-metal dichalcogenides with high 1T′-phase purity, J. Am. Chem. Soc., 2022, 144, 4863–4873.
17. J. Y. Hwang, M. F. El-Kady and Y. Wang, Direct preparation and processing of graphene/RuO₂ nanocomposite electrodes for high-performance capacitive energy storage, Nano Energy, 2015, 18, 57–70.
18. P. Niu, L. Zhang and G. Liu, et al., Graphene-like carbon nitride nanosheets for improved photocatalytic activities, Adv. Funct. Mater., 2012, 22(22), 4763–4770.
19. H. Sun and Y. Liu, Efficient Adsorption of Azo Dye Acid Brilliant Red on Graphite Carbon Nitride in Aqueous Solution, ACS Omega, 2024, 9, 28626–28636.
20. J. Y. Tong and E. L. King, A spectrophotometric investigation of the equilibria existing in acidic solutions of chromium(VI), J. Am. Chem. Soc., 1953, 75(24), 6180–6186.
21. G. P. Haight Jr, D. C. Richardson and N. H. Coburn, A spectrophotometric study of equilibria involving mononuclear chromium(VI) species in solutions of various acids, Inorg. Chem., 1964, 3(12), 1777–1780.
22. Y. Shiraishi, S. Kanazawa and Y. Sugano, et al., Highly selective production of hydrogen peroxide on graphitic carbon nitride (g-C₃N₄) photocatalyst activated by visible light, ACS Catal., 2014, 4(3), 774–780.
23. Y. Maham and G. R. Freeman, Effect of solvent structure on electron reactivity: 2-propanol/water mixtures, Can. J. Chem., 1988, 66, 1706–1711.
24. C. D. Jonah, J. R. Miller and M. S. Matheson, The reaction of the precursor of the hydrated electron with electron scavengers, J. Phys. Chem., 1977, 81, 1618–1622.
25. A. M. El-Nahas, J. W. Bozzelli, J. M. Simmie, M. V. Navarro, G. Black and H. J. Curran, Thermochemistry of acetonyl and related radicals, J. Phys. Chem. A, 2006, 110(50), 13618–13623.
26. G. V. Buxton and F. Djouider, Disproportionation of Cr^V generated by the radiation-induced reduction of Cr^VI in aqueous solution containing formate: a pulse radiolysis study, J. Chem. Soc., Faraday Trans., 1996, 92(21), 4173–4176.
27. G. V. Buxton, F. Djouider and D. A. Lynch, et al., Oxidation of Cr^III to Cr^VI initiated by OH and SO₄⁻ in acidic aqueous solution A pulse radiolysis study, J. Chem. Soc., Faraday Trans., 1997, 93(24), 4265–4268.
28. P. Praus, On electronegativity of graphitic carbon nitride, Carbon, 2021, 172, 729–732.
29. R. G. Pearson, Absolute electronegativity and hardness: application to inorganic chemistry, Inorg. Chem., 1988, 27(4), 734–740.
30. S. Barison, D. Barreca and S. Daolio, et al., Influence of electrochemical processing on the composition and microstructure of chemical-vapor deposited Ru and RuO₂ nanocrystalline films, J. Mater. Chem., 2002, 12(5), 1511–1518.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ra00883b

---