Ultrasound assisted Fenton-like degradation of dyes using copper doped graphitic carbon nitride

With rapid industrial development, large amounts of persistent organic pollutants are discharged into the natural environment and cause serious environmental pollution. Approximately 80% of total global wastewater is released into the environment. The wastewater emission poses serious threats to ecosystem and public health (Mian & Liu 2018). Wastewater pollution has been a global challenge in the field of environmental remediation. Dyes are common organic pollutants in wastewater from textile, printing, leather tanning, and cosmetic industries, and they are toxic to ecosystems and the human body (Chanikya et al. 2021). There is special concern about removal of toxic dyes from wastewater, which is still challenging (Routoula & Patwardhan 2020).

The synthesis of nanoscale catalysts is an attractive strategy for improving catalytic activity. Metals supported on carbon matrix such as activated carbon, biochar, and graphene have been reported (Wu et al. 2020; Wang et al. 2021b). Graphitic carbon nitride (g-C₃N₄), a layered material with a 2D π-conjugated polymeric structure, is a promising matrix for loading active metals. The unique physical structure, excellent chemical stability and tunable electronic properties support the photocatalytic activity in environmental applications (Yang et al. 2020; Hasija et al. 2021). g-C₃N₄ can be easily synthesized through the thermal polymerization of low-cost N-rich precursors such as urea, melamine, dicyandiamide, and thiourea (Jourshabani et al. 2020; Nguyen et al. 2021). Various modification strategies, such as elemental doping, heterojunction, defects engineering and structural regulation, have been reported to improve the properties for potential applications (Chen et al. 2020a). Due to the chemical stability and ligand-field effects, metal doped g-C₃N₄ can be promising candidates for Fenton-like catalysts (Zhu et al. 2019).

To solve the problem of developing effective Fenton-like catalysts, a novel copper doped g-C₃N₄ was designed and facilely synthesized. Methylene blue (MB) was chosen as a target pollutant. Ultrasound-assisted MB degradation was evaluated in the presence of H₂O₂. The objectives of this study are (1) to characterize the catalyst, (2) to determine the catalytic performance, and (3) to discuss the plausible catalytic mechanism.

All used chemicals were analytically pure, including hydrogen peroxide (H₂O₂; Sinopharm Chemical Reagent Co., Ltd), copper chloride (CuCl₂·2H₂O; Sinopharm Chemical Reagent Co., Ltd), melamine (C₃H₆N₆; Macklin Reagent Co., Ltd), nitric acid (HNO₃; Sinopharm Chemical Reagent Co., Ltd), sodium hydroxide (NaOH; Sinopharm Chemical Reagent Co., Ltd), isopropyl alcohol (IPA; C₃H₈O; ALADDIN Reagent (Shanghai) Co., Ltd), methylene blue (C₁₆H₁₈ClN₃S; Tianjin Kemiou Chemicals Co., Ltd), rhodamine B (C₂₈H₃₁ClN₂O₃; Tianjin Guangfu Fine Chemical Research Institute), and acid orange 7 (C₁₆H₁₁N₂NaO₄S; ALADDIN Reagent (Shanghai) Co., Ltd). Ultrapure water from a Millipore Milli-Q water purification system was used in experiments.

The synthesis of catalysts was conducted by metal ion blending and carbonization as shown in Fig. S1 (Supporting Information). First, melamine (2 g) and copper chloride (1.62 g) were dissolved in 10 mL methanol to obtain a uniform solution. Methanol was the chosen solvent because both melamine and copper chloride can be dissolved in it. After removing methanol by stirring at 60 °C, the solid product with uniformly mixed melamine and copper was obtained. Finally, the solid product was calcined in a muffle furnace with limited oxygen. The temperature was elevated to 550 °C with a ramping rate of 5 °C/min and maintained at 550 °C for 120 min. The calcined product was rinsed using ethanol solution (50%) to remove residual substances or by-products of carbonization. After drying at 100 °C for 8 h, the obtained copper doped carbon nitride (Cu-C₃N₄) was used as a catalyst.

A scanning electron microscopy (SEM) equipped with energy dispersive X-ray spectroscopy (EDS; FEI Quanta FEG 250) was used to analyze the morphology and surface elements. The N₂ adsorption–desorption isotherm was determined at 77 K to analyze the surface area using a surface area analyzer (Micromeritics, ASAP 2020, USA). Fourier transform infrared spectroscopy (FTIR; Frontier, PerkinElmer, USA) was used to record the FTIR spectrum. X-ray diffractometer (XRD; Ultimate IV, Rigaku Co., Japan) was employed to determine the chemical phases. X-ray photoelectron spectroscopy (XPS; Thermo ESCALAB 250XI, Thermo Scientific, USA) was used to examine the surface elements. An ultraviolet–visible spectrophotometer (UV–vis; UV-2600, Shimadzu, Japan) was employed to measure the absorbance of characteristic peak and UV–vis spectra of dye solution. A JES FA200 spectrometer was used to detect the electron spin resonance (ESR) signals using 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) as the radical-capture reagent.

The research design of this work is schematically shown in Fig. S2. Batch experiments were conducted in 150 mL conical flask at room temperature. The conical flask containing 50 mL MB solution was put in an ultrasonic cleaner (40 kHz, 150 W; XM-3200UVF, Xiaomei Instrument Limited Company, China). Experimental parameters were investigated including catalyst (0–0.6 g/L), H₂O₂ (0–40 mM), pH (3.3–9.9), and MB concentration (10–30 mg/L). Quenching experiments were conducted using IPA as radical scavenger. The reusability of the catalyst was studied for MB degradation through five cycles. Catalytic degradation of different dyes was also evaluated. During the experiments, a solution sample was taken out, centrifugally separated, and immediately determined using an UV–vis spectrophotometer. The calculation of MB removal is described in Supporting Information. The pseudo first-order model has been widely used for kinetic analysis. Reaction rate was examined by the kinetic model (described in Supporting Information).

The morphology and elemental distribution of Cu-C₃N₄ are characterized by SEM-EDS. The g-C₃N₄ is a layered massive particle (Hu et al. 2019a). As shown in Figure 1, Cu-C₃N₄ shows a sheet-like structure similar to that of g-C₃N₄. Aggregated nanosized particles without well-defined morphology were observed, which may be the doped copper oxides. Cu-C₃N₄ also manifests a porous morphology with intersect channels, which resulted from the release of gaseous products in the heating process (Ma et al. 2017; Qin et al. 2019). Fig. S3 manifests the major elements of Cu, C, N, and O. Elemental mapping displays the uniform distribution of C, N, O, and Cu in the material, confirming that Cu is doped on g-C₃N₄ (Hu et al. 2019a).

As shown in Figure 2(a), the N₂ adsorption–desorption isotherm of Cu-C₃N₄ shows a type IV isotherm with an H3 hysteresis loop according to the IUPAC classification, which suggests the slit-shaped pores resulted from aggregated particles (Ma et al. 2017; Jiang et al. 2019). The surface area and pore parameters are listed in Table S1 (Supporting Information). The BET surface area is 7.18 m²/g. The pore size distribution plotted by the BJH method displays a bimodal pore distribution ranging from 2 to 10 nm. The average pore diameter is 12 nm.

Figure 2(b) shows the FTIR spectrum of Cu-C₃N₄. The peaks at 3,508–3,423 cm⁻¹ are related to the stretching vibration of the N-H bond and -OH bond (Wang et al. 2019b). The characteristic peaks in the region ranging from 1,132 to 1,624 cm⁻¹ (1,132, 1,292, 1,403, and 1,624 cm⁻¹) corresponded to the trigonal C-N(-C)-C or the bridging C-NH-C units (Yao et al. 2015). The peak associated with the breathing vibration of the s-triazine units appears at 781 cm⁻¹, lower that of g-C₃N₄ (at around 807 cm⁻¹) (Muniandy et al. 2017). The peak at 549 cm⁻¹ may be ascribed to the stretching vibration of the Cu-N bond, implying that copper species are chemically coordinated to g-C₃N₄ (Dong et al. 2017).

The crystallinity of Cu-C₃N₄ was examined by XRD analysis (Figure 2(c)). The diffraction peak at 2-theta of 27.7° is ascribed to the (002) plane for graphitic materials (PDF#87-1526), suggesting the presence of g-C₃N₄ structure (Wang et al. 2019b). The diffraction peaks at 2-theta of 32.4°, 35.4°, 38.6°, 48.8°, 53.3°, 58.2°, 61.5°, 66.2°, and 67.9°, which are indexed to the (1 1 0), (1 1 −1), (1 1 1), (2 0 −2), (0 2 0), (2 0 2), (1 1 −3), (3 1 −1), and (1 1 3) crystal planes of CuO (PDF#48-1548) (Feng et al. 2019). The peaks at 2-theta of 36.4° and 41.2° are associated with the (1 1 1) and (2 0 0) crystal planes of Cu₂O (PDF#65-3288) (Omrani & Nezamzadeh-Ejhieh 2020). The diffraction peaks at 2-theta of 43.2°, 50.4°, and 74.8° are attributed to the (1 1 1), (2 0 0), and (2 2 0) crystal planes of Cu (PDF#65-9026) (Zhang et al. 2020a).

The signals of C 1s, N 1s, O 1s, and Cu 2p are observed at around 285 eV, 399 eV, 530 eV, and 932 eV in the XPS spectra of Cu-C₃N₄ (Figure 2(d)). As shown in Figure 2(e), the high-resolution XPS spectra of C 1s manifest the peaks at 284.8 eV (C-C), 286.2 eV (C-O), and 288.3 eV (N-C = N) (Zhang et al. 2020b; Song et al. 2021). Figure 2(f) displays the high-resolution N 1s spectra. The peaks at 398.9 eV, 399.6 eV and 400.7 eV are assigned to the sp²-hybridized pyridine nitrogen (C-N = C), tertiary pyrrolic nitrogen (N-C₃), and C-N-H groups (Li et al. 2018a). High-resolution O 1s spectra show peaks at 529.9 eV and 531.5 eV (Figure 2(g)), which are associated to the oxygen coordinated with copper (Cu-O) and the hydroxyl groups (-OH) (Kumar et al. 2014). Figure 2(h) shows the high-resolution XPS spectra of Cu 2p. The binding energy peaks at 932.9 eV and 952.8 eV correspond to Cu⁺ or Cu⁰ (Song et al. 2021). The peaks at 935.1 eV and 954.9 eV are attributed to the presence of Cu²⁺ (Liu & Hensen 2013). The peak at 570.1 eV in the Cu LMM spectrum confirms the presence of Cu⁺. In addition, the shakeup satellite peaks are observed at 942.5 eV and 962.5 eV (Liu & Hensen 2013). The peak fittings of high-resolution Cu 2p spectra reveal the existence of Cu²⁺, Cu⁺, and Cu⁰ in Cu-C₃N₄, which agrees with XRD analysis. Combining the FTIR, XRD, and XPS analysis, it can be concluded that copper is chemically incorporated into the g-C₃N₄ structure, probably embedded in the cavities of g-C₃N₄, or formed copper porphyrins and phthalocyanines (Dong et al. 2017).

It is reported that ultrasound is effective to enhance Fenton-like reactions (Wang et al. 2020a). Fig. S4 shows the comparison of MB degradation under ultrasound and oscillation. MB removal increases progressively with increasing time. The removal is approximately 80% at 90 min under oscillation, while it reaches above 90% at 40 min assisted by ultrasound. The pseudo first-order model has been extensively reported for studying the kinetics of Fenton-like processes (Wang et al. 2021b). Linear fitting of the model offers the apparent rate constant of pseudo first-order kinetics. The rate constant is significantly improved from 0.019 to 0.048 min⁻¹ with the assistance of ultrasound (Fig. S5). The catalytic system of Cu-C₃N₄/H₂O₂/ultrasound was used for subsequent degradation experiments.

The proper dosage of catalyst needs to be determined since it is closely related to operation cost for application. Figure 3(a) shows MB removal in the presence of different Cu-C₃N₄ dosage. MB removal is notably improved when increasing Cu-C₃N₄ from 0 to 0.1 g/L. When Cu-C₃N₄ is increased from 0.1 to 0.4 g/L, the removal is slightly enhanced. MB removal is nearly unchanged with further increasing Cu-C₃N₄ dosage. The apparent rate constant is 0.004 min⁻¹, 0.029 min⁻¹, 0.042 min⁻¹, 0.081 min⁻¹, and 0.087 min⁻¹, respectively (Figure 3(b)). The improved MB degradation associated with catalyst dosage was reported by Qin et al. (2018). The improvement of MB removal with increasing Cu-C₃N₄ is due to introducing more active sites, benefiting the generation of radical species that promote the degradation reactions (Wang et al. 2021b). MB removal is not further enhanced with excessive catalyst, probably because of particle agglomeration of the catalyst and undesirable scavenging radicals (Qin et al. 2018). 0.4 g/L of Cu-C₃N₄ was selected as optimal dosage for MB removal.

Solution pH is a key parameter for the Fenton-like process, and it influences the oxidation potential of reactive radicals, the stability of H₂O₂, and the reactivity of catalysts (Sun et al. 2021). The effect of solution pH on MB removal is shown in Figure 5(a). MB degradation remains almost the same over broad pH regions of 3.3–9.9. Although there exists a small difference in the apparent rate constant (Figure 5(b)), effective MB removal is achieved within 40 min under different pH value. This can be ascribed to the wide operation pH of copper ion. Generally, iron-based catalysts suffer from the narrow operation pH (Yin et al. 2020; de Melo Costa-Serge et al. 2021). Dye degradation at neutral pH has been highlighted by many researchers (Šuligoj et al. 2020; Wang et al. 2020b). Effective MB degradation obtained in a broad pH range is superior to the conventional Fenton process and iron-based catalysts.

The dye concentration in real wastewater varies with time and sources, and thus it is required to evaluate the effect of initial dye concentration. Figure 6(a) displays MB degradation with different initial MB concentration. MB removal increases with increasing reaction time. Effective degradation of MB solution with higher concentration can be obtained by extending reaction time, suggesting the feasibility of MB removal with different concentration. As observed in Figure 6(b), the apparent rate constant is reduced from 0.0810 to 0.0225 min⁻¹ when the initial MB concentration increases from 10 to 30 mg/L. The removal is decreased with an increase in MB concentration. This can be ascribed to the lack of enough reactive radicals (Wang et al. 2020b). Additionally, high MB concentration generates more degradation intermediates, which potentially poison catalyst surface (Wang et al. 2019c).

To evaluate the practicability of real application, the reusability of Cu-C₃N₄ was determined in the catalytic system. MB removal is slightly decreased (approximate 7%) for four runs (Figure 7). Obvious decline in MB removal is seen for fifth run, but effective MB removal can be obtained by prolonging reaction time (60 min). The results suggest the good reusability for application. Cu-C₃N₄ as Fenton-like catalyst has outstanding advantages such as high activity, broad operation pH, and good reusability. The process is promising for potential application to wastewater treatment.

Figure 8(a) displays the UV–vis spectra of MB solution. The peak at 664 nm is assigned to the chromophore groups (C = S and C = N bonds) in MB molecules (Fig. S7) (Wang et al. 2020b). The rapid decrease of the peak at 664 nm is observed with slight shift. This indicates that MB degradation is predominated by chromophore cleavage in the early stage under experimental conditions (He et al. 2009). The hypsochromic shift of the characteristic peak at later stage suggests the possible pathway of N-de-ethylation due to the auxochromic property of the N-ethyl group (Mitoraj et al. 2018). The degradation intermediates may undergo other reactions including hydroxylation, aromatic ring opening and mineralization (Zhou et al. 2015). MB degradation can be verified by the color change of MB solution (Fig. S8). The dark blue color becomes gradually colorless, and complete discoloration is achieved within 40 min. The color change of MB solution is consistent to the time-dependent UV–vis spectra.

The radicals are responsible for MB degradation. As reported by previous studies, •OH is the dominant oxidant in Fenton-like processes (Wang et al. 2019a; Xin et al. 2021). To verify the role of •OH in the Cu-C₃N₄/H₂O₂/ultrasound system, IPA is used as radical scavenger. Fig. S9 displays MB removal, which is lowered in the presence of IPA, and which is attributed to the scavenging effect. The existence of •OH is further confirmed by ESR spectrum using DMPO as a probe. A strong 4-fold characteristic peak of typical DMPO-trapped •OH adduct with an intensity ratio of 1:2:2:1 is observed in Figure 8(b) (Xin et al. 2021). It should be noted that besides •OH, other radicals cannot be ruled out for MB degradation. Based on above analysis, the plausible catalytic mechanism in Cu-C₃N₄/H₂O₂/ultrasound system is schematically shown in Figure 9.

Degradation of different dyes including azo dye (AO7) and triphenylmethane dye (RhB) were further conducted. Figure 10(a) manifests the UV–vis spectra of RhB. The maximum peak at 554 nm is ascribed to the N-ethyl group of RhB (Wang et al. 2021b). Effective degradation of RhB is obtained within 30 min as verified by the sharp decline of peak at 554 nm. The pathway of RhB degradation involves chromophore groups and N-de-ethylation (Wang et al. 2020a). Similar to MB degradation, the degradation of RhB is dominated by chromophore cleavage, while N-de-ethylation succeeded. Figure 10(b) shows that the typical peak at 484 nm of AO7 significantly declined within 40 min, validating the effective degradation of AO7 in the catalytic system (Wang et al. 2019c). Unlike MB degradation, obvious shift of the characteristic peak occurs in the initial stage, suggesting that degradation is not controlled by chromophore cleavage. The results verify that different dyes can be effectively degraded in the Cu-C₃N₄/H₂O₂/ultrasound system.

As verified by SEM, XRD, FTIR, and XPS, Cu-C₃N₄ contains copper species of Cu, Cu₂O and CuO, which are uniformly doped on the surface of porous g-C₃N₄. Enhanced MB degradation is observed by ultrasound. MB degradation in the Cu-C₃N₄/H₂O₂/ultrasound system is described by pseudo first-order kinetics. MB degradation is influenced by catalyst dosage, H₂O₂ concentration, solution pH, and initial MB concentration. Under conditions of Cu-C₃N₄ 0.4 g/L, H₂O₂ 20 mM, solution pH 6.7, MB 10 mg/L, and ultrasound, MB degradation of 96% is obtained within 30 min, and the corresponding rate constant is 0.099 min⁻¹. Effective MB degradation over a broad pH range (3.3–9.9) is superior to the conventional Fenton process or iron-based catalysts. Cu-C₃N₄ has good reusability as Fenton-like catalysts. Quenching test and ESR spectrum confirmed the existence of •OH in the catalytic system. Ultrasonically improved H₂O₂ activation by Cu-C₃N₄ for the generation of •OH radical is the dominant mechanism for MB degradation. The degradation of RhB and AO7 is also effectively achieved. The commercial application of the catalytic system for wastewater treatment requires more efforts. Nevertheless, this work provides insights into designing Fenton-like catalysts and wastewater treatment.

This work was supported by the National Natural Science Foundation of China (51804276); and the Excellent Youth Scientist of China Association for Science and Technology.

The authors have no competing interests to declare.

Data cannot be made publicly available; readers should contact the corresponding author for details.