Highly efficient g-C₃N₄ catalysts derived from various precursors for aflatoxin B1 degradation under visible light

Liang Wei^a, Sihong Li^a, Haibo Wang^b, Zimei Mo^b, Junji Chen^a, Meng Li^a, Hongxi Zhang^a, Jing Yang*^a and Xiande Yang*^a
aGuangxi Key Laboratory of Natural Polymer Chemistry and Physics, Nanning Normal University, Nanning 530001, P. R. China
^bGuangxi-ASEAN Food Inspection Center Food Testing Department, No. 42 Foziling Road, Qingxiu District, Nanning City, Guangxi, Nanning 530029, China. E-mail: yangjing1512F@126.com; yangxde@outlook.com

First published on 2nd September 2024

---

---

1 Introduction

Food security is a pivotal issue that must be addressed in the progression of human development. Amid the intricate processes of food production, transportation, and storage, the persistent threat of mycotoxin contamination remains a significant concern. Consuming contaminated food can inflict harmful effects on the human body, and AFB1 stands out as a particularly alarming contaminant among all known mycotoxins.¹ AFB1, a highly toxic carcinogen produced by fungi and parasitic molds, has been categorized as a Group 1 carcinogen by the International Agency for Research on Cancer. It possesses formidable biological toxicity and teratogenicity, posing a grave threat to both humans and animals.² AFB1 can maintain its structural stability under extreme conditions, rendering it difficult for the human body to metabolize or eliminate through the circulatory system.³ Hence, our primary focus lies in discovering a safe and reliable method to degrade AFB1 in food, ensuring food safety and protecting public health.

In recent years, numerous methods have been developed for AFB1 treatment, including UV light irradiation,⁴ gamma ray radiation,⁵ ozone treatment,⁶ and others. Nonetheless, these approaches are often accompanied by substantial drawbacks, including significant energy consumption and suboptimal removal efficiency. In contrast, photocatalytic technology offers a promising solution. This innovative method harnesses the power of sunlight to activate semiconductor materials, generating numerous reactive species that engage in oxidation–reduction reactions with pollutants, effectively breaking down their structure for degradation.⁷ Photocatalytic technology stands out as a low-energy, recyclable, and environmentally friendly approach commonly employed for the elimination of pollutants in various environments.⁸

g-C₃N₄ has garnered considerable attention in photocatalysis for its remarkable attributes, including its simplicity in preparation, cost-effectiveness, and robust thermal stability.^9–11 Notably, the photocatalytic activity of g-C₃N₄ is heavily contingent upon the precursors in its synthesis. The common precursors for the fabrication of g-C₃N₄ encompass thiourea, melamine, urea, and dicyandiamide. These precursors, owing to their variations in chemical element composition, impart significant disparities in the morphology, structural configuration, and ultimately affect the photocatalytic performance of g-C₃N₄.^11,12 For instance, Shirman et al.¹³ explored the photocatalytic hydrogen evolution activity of palladium-loaded g-C₃N₄ derived from distinct precursors, revealing that cyanuric dicyandiamide derived g-C₃N₄ displayed slightly superior catalytic activity than that of urea derived g-C₃N₄. Chand et al.¹² discovered that urea derived g-C₃N₄ possessed a higher catalytic activity for carbon dioxide conversion compared to those derived from melamine, thiourea, and dicyandiamide. Jung et al.¹⁴ found that cyanuric dicyandiamide derived g-C₃N₄/ZnO composites achieved the highest methylene blue photodegradation efficiency. Liang et al.¹⁵ noted significant differences in the photocatalytic reduction of Cr(VI) using g-C₃N₄ synthesized from dicyandiamide, urea, melamine, and thiourea, where urea derived g-C₃N₄ nanosheets emerged as the most efficient catalyst in Cr(VI) reduction. Shao et al.¹⁶ found that the presence of trace sulfur in thiourea derived g-C₃N₄ induced higher asymmetry in its structural units, resulting in the highest activity for piezoelectric reduction of KMnO₄. However, there has been a lack of reports exploring the effect of g-C₃N₄ prepared from varying precursors on the photocatalytic degradation of AFB1.

In this study, urea, melamine, thiourea, and dicyandiamide were used as precursors to synthesize four g-C₃N₄ catalysts via thermal polymerization approach. The comprehensive discussions of the morphologies, structural integrities, electrochemical properties of these g-C₃N₄ catalysts were conducted. Their photocatalytic efficiencies were evaluated through AFB1 degradation under visible light. Subsequently, the products generated from the photocatalytic degradation process were subjected to the UHPLC-Q Exactive MS system, thereby inferring potential AFB1 degradation pathways and proposing relevant mechanisms. This study provides valuable insights into the influence of precursors in g-C₃N₄ synthesis on AFB1 degradation.

2 Experimental sections

2.1 Chemicals

The AFB1 standard was obtained from J&K Scientific. Urea and thiourea were acquired from Guangdong Guanghua Technology Co., Ltd, while melamine and dicyandiamide were purchased from Shanghai Macklin Biochemical Co., Ltd. Analytical-grade isopropyl alcohol (IPA) and 1,4-benzoquinone (BQ, C₆H₄O₂) were provided by Aladdin Biochemical Technology Co. We also procured EDTA-2Na from Tianjin Aopusheng Chemical Co., Ltd. For high-performance liquid chromatography (HPLC) analysis, we utilized HPLC-grade methanol sourced from Thermo Fisher Scientific. Deionized water was used in all experiments.

2.2 Preparation of respective catalysts

The catalysts were synthesized using thermal polymerization method. Specifically, a precise amount of precursor, including urea, melamine, thiourea, or dicyandiamide, was transferred into a ceramic crucible and then positioned in a muffle furnace. The temperature was gradually raised to 550 °C and held constant for 4 h in air. After cooling down to room temperature, the sample was collected and ground thoroughly to obtain g-C₃N₄. The g-C₃N₄ derived from urea, melamine, thiourea, and dicyandiamide were designated as U-CN, M-CN, T-CN, and D-CN, respectively.

2.3 Characterizations

To analyze the crystal phases of the catalysts, X-ray diffraction (XRD, Bruker D8-Advance, Germany) was employed. The morphology and particle size of the catalysts were characterized using field emission scanning electron microscopy (FESEM, SU8010, Hitachi, Japan) and transmission electron microscopy (TEM, JEM 2100F). Additionally, elemental analysis was conducted via X-ray photoelectron spectroscopy (XPS, EscaLab-250Xi, USA). The specific surface area and pore size of the catalysts were determined using the Brunauer–Emmett–Teller (BET) system (BET, Bei-Shi-De 3H-2000PS2, Beijing). To investigate the light absorption property of the catalysts, UV-Vis diffuse reflectance spectroscopy (UV-Vis DRS, UV-3600Plus, Shimadzu, Japan) was utilized. The photoluminescence (PL) spectrum (λ_ex of g-C₃N₄ is 380 nm) of the catalysts was recorded using a fluorescence spectrophotometer (F-4600, Hitachi, Japan). Electron paramagnetic resonance (EPR) tests were conducted on a Bruker EPR EMXplus spectrometer (Bruker, Germany) with DMPO as the spin-trapping agent. Furthermore, the identification and confirmation of degradation products were carried out using a Thermo UltiMate 3000 UHPLC-Q mass spectrometry system (Thermo Fisher Scientific, Waltham, MA, USA).

2.4 Photocatalytic measurement

Using a 300 W Xe lamp coupled with a cut-off filter (λ ≥ 420 nm, PLS-SXE 300, Beijing PerfectLight Co. Ltd., China), the photocatalytic degradation efficacy of an AFB1 aqueous solution (100 mL, 1 μg mL⁻¹) by 2 mg of catalysts was evaluated. Before irradiation, the mixture solution was stirred in darkness for 30 min to attain adsorption–desorption equilibrium. During irradiation, 50 μL of the supernatant were withdrawn every 5 min and diluted to 4 mL using 70% HPLC-grade methanol. This photocatalytic irradiation process spanned 30 min. Analysis was conducted using a high-performance liquid chromatography system (Beijing Mingnike P2000), equipped with a C18 column (254 mm × 4.6 mm × 5 mm, Leena) and a UV-visible detector. The HPLC conditions were set as follows: A mobile phase consisting of 70% methanol in water, a flow rate of 1.0 mL min⁻¹, and a detection wavelength of 365 nm. The AFB1 removal efficiency (Y) was determined utilizing the following equation:

Y = (1 − CtVt/C0V0) × 100%

In this equation, C₀ (μg mL⁻¹) and V₀ (mL) represent the initial AFB1 concentration and volume, respectively. C_t (μg mL⁻¹) and V_t (mL) correspond to the AFB1 concentration and volume after the degradation process.

2.5 Electrochemical measurements

In the CHI660E electrochemical system (CH Instruments Ins.), a three-electrode cell configuration is employed, consisting of an Ag/AgCl reference electrode and a Pt counter electrode. In addition, a 300 W Xe lamp (λ > 400 nm) was used as the light source. The working electrode is prepared as follows: Firstly, 4 mg of the samples is dispersed in 4 mL of ethanol with ultrasonication for 2 h to obtain a homogeneous solution. Subsequently, 50 μL of the above solution is pipetted onto an ITO glass substrate (1.0 cm × 1.5 cm), and this process is repeated three times after each drying cycle. Finally, transient photocurrent response measurements and electrochemical impedance spectroscopy (EIS) are conducted using 0.5 M Na₂SO₄ solution as the electrolyte.

2.6 Photocatalytic stability test

Utilizing U-CN samples which exhibited the optimal photocatalytic performance for stability testing. We conducted four consecutive cycles of photocatalytic degradation experiments on AFB1 by U-CN catalysts. For each cycle, 100 mL of AFB1 solution (1 μg mL⁻¹) was used. After 30 min of photocatalytic reaction, the U-CN catalysts were retrieved and alternately rinsed with deionized water and ethanol. The U-CN catalysts were then dried at 60 °C for subsequent reuse. This process was repeated until four cycles of photocatalytic degradation experiments were completed.

3 Results and discussion

3.1 XRD analysis and FT-IR analysis

The XRD patterns present in Fig. 1(a) carefully illustrate the diffraction peaks of U-CN, M-CN, T-CN, and D-CN, revealing their respective crystallinities and structural characteristics. These XRD patterns reveal that all g-C₃N₄ samples exhibited two remarkably distinct characteristic peaks at 12.8° and 27.4°, precisely aligning with the standard card of g-C₃N₄. Notably, the diffraction peak situated near 12.8° is attributed to the intralayer long-range atomic order, stemming from the hydrogen bonding present in g-C₃N₄.¹⁷ Conversely, the pronounced peak observed around 27.4° corresponds to the interlayer periodic stacking along the c-axis, which is induced by the stacking of conjugated aromatic systems within g-C₃N₄.^18,19 FT-IR analysis was employed to identify the functional groups present in U-CN, M-CN, T-CN, and D-CN. As depicted in Fig. 1(b), the samples derived from four precursors exhibited characteristic FT-IR spectra of g-C₃N₄. Prominent absorption bands were distinctly observed at 811 cm⁻¹, 1200–1700 cm⁻¹, and 3000–3500 cm⁻¹.¹⁶ Specifically, the sharp band near 811 cm⁻¹ corresponds to the stretching vibrations of the distinctive tri-s-triazine units within the g-C₃N₄ structure.^15,20 The absorption peaks spanning around 1200–1700 cm⁻¹ are attributed to the presence of C–N and CN bonds in the aromatic nitrogen–carbon heterocycles.²¹ Furthermore, the distinctive absorption peaks between 3000–3500 cm⁻¹ originate from the stretching vibrations of –OH and –NH groups,²² which can be partially attributed to uncondensed amine groups and water molecules adsorbed on the surface of the catalysts.²³ In summary, Fig. 1 confirmed that g-C₃N₄ samples derived from different precursors were successfully prepared.

3.2 Morphological structures

As depicted in Fig. 2, FESEM was utilized to analyze the morphological structures of four g-C₃N₄ samples derived from various precursors. In Fig. 2(a), U-CN exhibits a loose mesh-like structure, featuring irregular pore sizes, potentially enhancing its specific surface area. Meanwhile, M-CN presents a large, blocky sheet structure with a relatively smooth surface (Fig. 2(b)). T-CN displays a layered structure adorned with numerous particles of varying sizes scattered across its surface (Fig. 2(c)). D-CN shows a flattened layered structure with a reduced number and size of particles on its surface (Fig. 2(d)). The TEM result in Fig. 2(e) further reveals a loose mesh-like configuration of U-CN, comprising numerous overlapping nanosheets of varying sizes and curvature.

3.3 XPS analysis

The XPS analysis of U-CN, M-CN, T-CN, and D-CN provides their oxidation states and surface chemical composition. As depicted in Fig. 3(a), the XPS survey spectra reveal that the as-synthesized g-C₃N₄ samples primarily consist of C, N, and O elements. Notably, the O elements can be attributed to the adsorption of H₂O on the catalysts’ surface.²⁴ The high-resolution of C 1s XPS spectrum reveals two distinct peaks centered at 288.2 eV and 284.8 eV (Fig. 3(b)). The peak at 288.2 eV corresponds to the sp² N–CN bond, which is indicative of the graphitic carbon nitride structure. In contrast, the peak at 284.8 eV represents the C–C bond, potentially originating from adventitious carbon contamination.²⁵ Moreover, Fig. 3(c) displays the N 1s XPS spectra of the four g-C₃N₄ samples. All four samples exhibit peaks located at 404.2 eV, 401.3 eV, 400.2 eV, and 398.7 eV. These peaks correspond to π excitation,²⁶ –NH₂, tertiary nitrogen N–(C)₃, and C–NC bonds,²⁵ respectively, providing a comprehensive understanding of the nitrogen bonding configurations within the g-C₃N₄ samples. The C 1s and N 1s peaks of U-CN exhibit significant deviations relative to those of M-CN, T-CN, and D-CN, aligning with the findings of XRD analysis. XRD and XPS results confirm that U-CN possesses a lower crystallinity, which in turn affects its electron energy level distribution, ultimately leading to the observed peak shifts.²⁷

3.4 BET analysis

In order to gain a profound comprehension of the specific surface area and pore architecture of four g-C₃N₄ samples, N₂ adsorption–desorption isotherms and pore size distributions were carefully assessed. As illustrated in Fig. 4(a), the N₂ adsorption–desorption isotherms for four g-C₃N₄ samples exhibit characteristic H3 hysteresis loops, signifying a type IV isotherm pattern.²⁸ Fig. 4(b) reveals that all the samples possess mesoporous structures.²⁹ In Table 1, U-CN stands out with a remarkable specific surface area of 100.5 m² g⁻¹ and pore volume of 0.7998 cm³ g⁻¹ compared to the other samples. This enhancement can be attributed to the pyrolysis process of urea, during which a self-sustaining atmosphere is created by the liberation of significant amounts of pyrolysis gases, including NH₃, H₂O, and CO₂. These gases function as soft templates, facilitating the development of a porous structure that enhances the specific surface area and contributes to the formation of a more elaborate porous architecture.³⁰ Fundamentally, a larger specific surface area often correlates with an increased number of reactive sites, which is favourable for photocatalytic degradation.

---

3.5 UV-vis DRS analysis

To investigate the optical properties of U-CN, M-CN, T-CN, and D-CN, their UV-vis DRS absorptions were measured. As depicted in Fig. 5(a), the UV-vis DRS absorption spectra reveal that all g-C₃N₄ samples can exhibit visible light response. Fig. 5(b) further illustrates the corresponding band gaps of respective samples. g-C₃N₄ is direct semiconductor, and its band gap can be calculated via the Kubelka–Munk equation³¹: αhν = A(hν − E_g)^n/2. Where h represents the Planck constant, ν is the photon frequency, A is the proportionality constant, α is the photopic absorption coefficient, and E_g denotes the band gap. Therefore, the band gap for U-CN, M-CN, T-CN, and D-CN was determined to be 2.72 eV, 2.60 eV, 2.54 eV, and 2.47 eV, respectively. Notably, U-CN exhibited a larger band gap compared to the other g-C₃N₄ samples, potentially attributed to quantum size effects.³²

3.6 MS and PL analysis

The photocatalytic activity of semiconductors is intricately linked to their light harvesting capabilities and band gap characteristics. Consequently, we also examined the Mott–Schottky (MS) curves of U-CN, M-CN, T-CN, and D-CN. As evident in Fig. 6(a), the positive slope of the MS curves for all samples suggests that the g-C₃N₄ samples derived from different precursors are n-type semiconductors.^33,34 The flat band potential (E_fb) of U-CN, M-CN, T-CN, and D-CN is −0.65 V, −0.58 V, −0.46 V, and −0.62 V (vs. Ag/AgCl), respectively. Upon reviewing the reported XPS VB spectra of g-C₃N₄, we can deduce that the VB-E_f potentials for U-CN, M-CN, T-CN, and D-CN are 2.28 eV, 2.22 eV, 2.32 eV, and 2.22 eV,¹⁵ respectively. Employing the established methodologies,^33,35 we have calculated the valence band (VB) position for these materials, which is 1.63 V for U-CN, 1.64 V for M-CN, 1.86 V for T-CN, and 1.60 V for D-CN. Furthermore, by combining the calculated band gaps with the empirical formula: E_VB = E_CB + E_g, we determined the conduction band (CB) position of U-CN, M-CN, T-CN, and D-CN to be −1.09 V, −0.96 V, −0.68 V, and −0.87 V, respectively.

The electron–hole separation efficiency in U-CN, M-CN, T-CN, and D-CN samples was thoroughly analyzed via photoluminescence spectroscopy. As depicted in Fig. 6(b), the PL emission peak's intensity at 470 nm follows the sequence of D-CN > M-CN > T-CN > U-CN for the g-C₃N₄ samples derived from respective precursors. Conventionally, a lower PL intensity signifies a higher efficiency in charge carrier separation and a reduced recombination rate, thereby indicating superior photocatalytic activity of the catalyst.^32,36 Notably, U-CN exhibits a lower fluorescence intensity, which can be attributed to its loose mesh-like structure compared to M-CN, T-CN, and D-CN. The loose mesh-like structure facilitates the rapid transfer of electrons to the catalyst surface and accelerates the separation of photogenerated carriers.

3.7 Electrochemical results

As depicted in Fig. 7(a), all g-C₃N₄ catalysts exhibit commendable photocurrent responses, with U-CN showing the highest photocurrent. Notably, U-CN displays the most robust photocurrent response, signifying its exceptional photogenerated carriers’ separation efficiency,³⁷ which aligns with the PL results. In Fig. 7(b), the radius's trend of the EIS Nyquist curves for four g-C₃N₄ samples is as follows: U-CN < M-CN < T-CN < D-CN. Conventionally, a smaller Nyquist curve radius indicates a lower charge transfer resistance in the catalysts, favoring enhanced charge mobility and consequently pointing to a superior separation efficiency of photogenerated carriers.^32,37 These findings further reinforce that U-CN possesses the highest electron–hole pair separation efficiency, which can effectively enhance its photocatalytic degradation performance.

3.8 Photocatalytic performance

The photocatalytic degradation of AFB1 was conducted utilizing U-CN, M-CN, T-CN, and D-CN as catalysts. Given the remarkable stability of AFB1 in aqueous solutions, its concentration remains largely unchanged under visible light irradiation.^38,39 As depicted in Fig. 8(a), the photocatalytic degradation efficiency of AFB1 by various g-C₃N₄ follows the order of U-CN > M-CN > T-CN > D-CN. Notably, U-CN emerges as the most effective catalyst with a remarkable AFB1 removal rate of 93.5% for 100 mL of AFB1 (1 μg mL⁻¹) within 30 min, significantly outperforming M-CN (82.5%), T-CN (78.5%), and D-CN (46.2%). This outstanding photocatalytic performance of U-CN aligns perfectly with the findings from transient photocurrent response, EIS, and PL analysis, confirming its superior efficiency in electron–hole pairs’ separation. Additionally, upon reaching adsorption–desorption equilibrium in darkness, the AFB1 concentration in the U-CN suspension is notably lower than that of the other three samples, due to U-CN's largest specific surface area and higher density of active adsorption sites for the substrate. In general, the AFB1 removal by all g-C₃N₄ samples is mainly due to photocatalytic degradation. To further analyze the degradation kinetics, the kinetic equation ln(C_t/C₀) = kt was applied. Where C_t (mg L⁻¹) represents the AFB1 concentration at time t, C₀ (mg L⁻¹) is the initial AFB1 concentration, and k (min⁻¹) is the first-order rate constant. As presented in Fig. 8(b) and Table 2, the photocatalytic degradation processes of U-CN, M-CN, T-CN, D-CN samples adhere to a pseudo-first-order kinetic model, and sample U-CN has the largest k value of 0.08856 min⁻¹.

---

We delved deeper into the characteristics of U-CN, which exhibits the highest photocatalytic efficiency, examining the mass of U-CN on the photocatalytic degradation of AFB1 (Fig. 9(a)) and the effect of distinct pH values on the degradation rate of AFB1 by U-CN (Fig. 9(c)). Notably, an increase in the mass of U-CN catalyst led to a continuous enhancement in the degradation rate of AFB1. This enhancement can be attributed to the provision of more active reaction sites as the catalyst concentration rises.⁴⁰ However, after surpassing a catalyst mass of 2 mg, further increments in catalyst's concentration yield diminishing returns in terms of enhancing the degradation rate. This is likely due to the fact that an excessive amount of catalyst may hinder the utilization of visible light and mass transfer during the photocatalytic degradation process.⁴¹ Additionally, changing the pH of the solution has a substantial impact on the AFB1 degradation. Since the lactone ring of AFB1 molecules is prone to disruption in alkaline solutions, alkaline conditions were not considered in this study.⁴² Instead, we focused on the impact of pH values ranging from 3 to 7.⁴³ Under these conditions, the degradation rate of AFB1 by U-CN follows this order: pH = 7 > pH = 5 > pH = 3. This trend can be explained by the fact that under acidic conditions, more H⁺ ions tend to eliminate ˙OH radicals.⁴⁴ Their kinetics of photocatalytic reactions both align with a pseudo-first-order kinetic model (Fig. 9(b) and (d)), indicating a consistent relationship between reaction rate and reactant concentration.

3.9 Photocatalytic stability test

To evaluate the photostability and reusability of U-CN for AFB1 degradation, this study undertook four consecutive cycles of photocatalytic degradation. Notably, U-CN catalyst sustained a commendable degradation efficiency towards AFB1 after four cycles (Fig. 10(a)), with even the fourth cycle achieving a degradation rate of 85% within 30 min. The slight decline in degradation efficiency throughout the cycling experiments can be attributed to two factors. Firstly, the main reason is that the amount of catalyst (2 mg) is too small, which makes it difficult to recover after the photocatalytic reaction, resulting in a decrease in degradation efficiency. Secondly, the catalyst becomes contaminated by the intermediate products generated during the AFB1 degradation process.⁴⁵ Upon scrutinizing the FT-IR spectra before and after the cycling experiments (Fig. 10(b)), it is evident that the structure of U-CN remains intact after four cycles. This is evident from the absence of notable shifts or variations in the absorption peaks at 811 cm⁻¹ and within the range of 1200–1700 cm⁻¹. Nonetheless, an increase in the intensity of the absorption peak at 3200–3500 cm⁻¹ is discernible, which can be attributed to the adsorption of intermediates generated during the AFB1 degradation process onto the surface of the catalyst. This observation is consistent with the subsequent SEM findings. A comparative analysis of the SEM images before and after four cycles reveals that U-CN's morphology remained largely unchanged (Fig. 10(c) and (d)), indicating that sample U-CN has a good photostability and reusability. However, a notable presence of new granular substances is observed, which can be attributed to the adsorption of AFB1 intermediates onto the surface of U-CN. As can be seen from Table 3, the U-CN catalyst has high efficiency for AFB1 degradation compared to other reported g-C₃N₄-based catalysts. It means that the U-CN catalyst can effectively solve the actual AFB1 contamination problem.

---

3.10 Exploring the photocatalytic degradation mechanism

In this study, scavenging experiments were conducted separately for superoxide anion radicals (˙O₂⁻), photogenerated holes (h⁺), and hydroxyl radicals (˙OH) by introducing 1 mM of benzoquinone (BQ), ethylenediaminetetraacetic acid disodium salt (EDTA-2Na), and isopropanol (IPA). After the addition of BQ, IPA, and EDTA-2Na, the degradation efficiency of AFB1 by U-CN decreased from 93.5% to 56.0%, 78.2%, and 83.2% (Fig. 11(a)), respectively. This indicates that these three types of radicals play a significant role in influencing the photocatalytic performance of U-CN, and ˙O₂⁻ is identified as the primary reactive species. This phenomenon may be caused by the fact that the E_CB of U-CN is more negative than that of E^θ(O₂/˙O₂⁻), which makes it easy for U-CN to generate a large amount of ˙O₂⁻. Under visible light irradiation, U-CN tends to generate electrons (e⁻) in the CB and generate h⁺ in the VB. Subsequently, e⁻ in the CB react with O₂ to form ˙O₂⁻. It is noteworthy that despite the lower E_VB position of U-CN compared to E^θ(˙OH/H₂O), precluding direct ˙OH generation on its VB, the degradation efficiency of U-CN significantly decreases upon IPA addition for ˙OH scavenging. This indicates an indirect reaction between ˙O₂⁻ and H₂O₂ may generate a finite amount of ˙OH during U-CN photocatalysis, as reported previously (O₂ + e⁻ → ˙O₂⁻, ˙O₂⁻ + e⁻ + 2H⁺ → H₂O₂, H₂O₂ + e⁻ → ˙OH + OH⁻).^45,51 These active species then undergo oxidation–reduction reactions with AFB1, ultimately achieving the photocatalytic degradation of AFB1. This process can be succinctly summarized as:

g-C3N4 + hν → g-C3N4 (e− + h+)

O2 + e− → ˙O2−

˙O2− + e− + 2H+ → H2O2

H2O2 + e− → ˙OH + OH−

AFB1 + ˙O2− (or h+ or ˙OH) → small molecules

Utilizing 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as a trapping agent, ESR analysis identified the presence of ˙O₂⁻ and ˙OH during photocatalytic reactions^52,53 (Fig. 11(b) and (c)). Notably, no ESR signal was detected in the absence of light, suggesting that U-CN did not generate ˙O₂⁻ in dark conditions. However, a significant ESR signal emerged upon exposure to visible light irradiation, indicating the generation of ˙O₂⁻ during the photocatalytic process. Similarly, the ESR analysis also revealed the characteristic peak of the ˙OH signal generated by U-CN under visible light irradiation. These results confirmed that the active species of the ˙OH and ˙O₂⁻ can be generated during the photocatalytic process.

To further elucidate the degradation mechanism of AFB1 by U-CN, UHPLC-MS analysis was performed. After visible light irradiation, eight intermediate products are identified (Fig. 12(a)–(d)), including C₁₇H₁₈O₇ (m/z = 334.11), C₁₄H₁₂O₆ (m/z = 276.06), C₁₆H₁₀O₆ (m/z = 298.05), C₁₆H₁₂O₅ (m/z = 284.07), C₁₄H₈O₄ (m/z = 240.04), C₁₁H₁₀O₃ (m/z = 190.06), C₁₁H₁₂O₂ (m/z = 176.08), and C₅H₆O (m/z = 82.04). Previous researches have highlighted that the toxicity and carcinogenicity of AFB1 primarily originate from its double bonds in the furan and lactone rings.^54,55 In general, ˙OH addition reactions tend to occur on unsaturated carbon bonds.⁵⁶ Consequently, the initial degradation product, C₁₇H₁₈O₇, likely forms through addition reactions involving the C8C9 double bond in AFB1's furan ring, as well as reactions with the lactone ring and cyclopentenone group. Upon visible light irradiation, C₁₇H₁₈O₇ undergoes ROS reactions and transform into C₁₄H₁₂O₆. Another possible degradation pathway involves ˙OH addition reactions on various unsaturated carbon bonds in AFB1, particularly those located at C8C9, owing to the strong conjugated π bonds present at C9bC9c, C5aC5, C4C3b, and C3aC11a.³⁷ This pathway leads to the formation of non-toxic or mildly toxic small molecules through a series of reactions, ultimately contributing to the degradation of AFB1 (Fig. 12(e)). Based on the above analysis, the possible photocatalytic mechanism of AFB1 via U-CN under visible light can be observed in Fig. 12(f).

4 Conclusion

In summary, this study investigates the synthesis of four distinct g-C₃N₄ morphologies employing urea, melamine, thiourea, and dicyandiamide as precursors. Furthermore, the photocatalytic properties of these g-C₃N₄ catalysts, derived from varied precursors, were rigorously evaluated during the AFB1 degradation under visible light. Notably, the g-C₃N₄ catalyst (U-CN) derived from urea exhibited a remarkable photocatalytic degradation rate (93.5%) of AFB1 within 30 min, which was attributed to its porous structure, large surface area, and reduced electron–hole pairs’ recombination rate. Furthermore, its consistent performance over four consecutive cycles validates that U-CN catalyst has a good photocatalytic stability. We also provided a potential deactivation pathway for AFB1 degradation by U-CN catalyst. This study underscores the critical role of precursor choice in enhancing the photocatalytic property of g-C₃N₄, thereby laying a theoretical foundation and technological groundwork for designing highly efficient photocatalysts in AFB1 degradation.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22262025, 21862004, U23A2089), Science and Technology Program of Market Supervision Administration of Guangxi Zhuang Autonomous Region (GXSJKJ2024-13), Scientific Research Foundation for Doctor, Nanning Normal University (No. 602021239253), and BAGUI Scholar Program of Guangxi Province of China. The financial support was gratefully appreciated.

References

1. A. Jallow, H. Xie and X. Tang, et al., Worldwide aflatoxin contamination of agricultural products and foods: From occurrence to control, Compr. Rev. Food Sci. Food Saf., 2021, 20(3), 2332–2381 CrossRef CAS PubMed .
2. D. Conradt, M. A. Schätzle and J. Haas, et al., New insights into the conversion of versicolorin A in the biosynthesis of aflatoxin B1, J. Am. Chem. Soc., 2015, 137(34), 10867–10869 CrossRef CAS PubMed .
3. Z. Peng, L. Chen and Y. Zhu, et al., Current major degradation methods for aflatoxins: A review, Trends Food Sci. Technol., 2018, 80, 155–166 CrossRef CAS .
4. M. Farzaneh, Z. Q. Shi and A. Ghassempour, et al., Aflatoxin B1 degradation by Bacillus subtilis UTBSP1 isolated from pistachio nuts of Iran, Food Control, 2012, 23(1), 100–106 CrossRef CAS .
5. Z. Zhang, Q. Xie and L. Che, Effects of gamma irradiation on aflatoxin B1 levels in soybean and on the properties of soybean and soybean oil, Appl. Radiat. Isot., 2018, 139, 224–230 CrossRef CAS PubMed .
6. P. Murugesan, D. K. Brunda and J. A. Moses, et al., Photolytic and photocatalytic detoxification of mycotoxins in foods, Food Control, 2021, 123, 107748 CrossRef CAS .
7. L. Wei, S. Chen and X. Yang, et al., Synthesis of visible light driven spherical Bi₂MoO₆ structure for aflatoxin B1 photodegradation, Inorg. Chem. Commun., 2024, 160, 111886 CrossRef CAS .
8. G. Li, B. Jiang and X. Li, et al., C₆₀/Bi₂TiO₄F₂ heterojunction photocatalysts with enhanced visible-light activity for environmental remediation, ACS Appl. Mater. Interfaces, 2013, 5(15), 7190–7197 CrossRef CAS PubMed .
9. M. Chen, M. Sun and X. Cao, et al., Progress in preparation, identification and photocatalytic application of defective g-C₃N₄, Coord. Chem. Rev., 2024, 510, 215849 CrossRef CAS .
10. Y. Qin, Y. Li and J. Yang, et al., Multiobjective optimization of g-C₃N₄/Cd-Zn₃In₂S₆ heterojunction for high-efficiency photocatalytic hydrogen evolution, New J. Chem., 2024, 48, 10863–10872 RSC .
11. Y. Chen, Y. Qin and J. Yang, et al., Functionalized graphite carbon nitride nanofluid membranes for enhanced osmotic energy harvesting, J. Electroanal. Chem., 2024, 952, 117967 CrossRef CAS .
12. H. Chand, A. Kumar and S. Goswami, et al., Comparison of catalytic activity of graphitic carbon nitrides derived from different precursors for carbon dioxide conversion, Fuel, 2024, 357, 129757 CrossRef CAS .
13. R. Shirman, A. Bahuguna and Y. Sasson, Effect of precursor on the hydrogen evolution activity and recyclability of Pd-Supported graphitic carbon nitride, Int. J. Hydrogen Energy, 2021, 46(73), 36210–36220 CrossRef CAS .
14. H. Jung, T. T. Pham and E. W. Shin, Effect of g-C₃N₄ precursors on the morphological structures of g-C₃N₄/ZnO composite photocatalysts, J. Alloys Compd., 2019, 788, 1084–1092 CrossRef CAS .
15. J. Liang, C. Jing and J. Wang, et al., Photocatalytic reduction of Cr(VI) over g-C₃N₄ photocatalysts synthesized by different precursors, Molecules, 2021, 26(22), 7054 CrossRef CAS PubMed .
16. Y. Shao, C. Liu and H. Ma, et al., Piezocatalytic performance difference of graphitic carbon nitride (g-C₃N₄) derived from different precursors, Chem. Phys. Lett., 2022, 801, 139748 CrossRef CAS .
17. F. Fina, S. K. Callear and G. M. Carins, et al., Structural investigation of graphitic carbon nitride via XRD and neutron diffraction, Chem. Mater., 2015, 27(7), 2612–2618 CrossRef CAS .
18. H. Wei, Q. Zhang and Y. Zhang, et al., Enhancement of the Cr(VI) adsorption and photocatalytic reduction activity of g-C₃N₄ by hydrothermal treatment in HNO3 aqueous solution, Appl. Catal., A, 2016, 521, 9–18 CrossRef CAS .
19. D. J. Martin, K. Qiu and S. A. Shevlin, et al., Highly efficient photocatalytic H₂ evolution from water using visible light and structure-controlled graphitic carbon nitride, Angew. Chem., Int. Ed., 2014, 53(35), 9240–9245 CrossRef CAS PubMed .
20. P. Niu, M. Qiao and Y. Li, et al., Distinctive defects engineering in graphitic carbon nitride for greatly extended visible light photocatalytic hydrogen evolution, Nano Energy, 2018, 44, 73–81 CrossRef CAS .
21. T. Jiang, Z. Wang and G. Wei, et al., Defective high-crystallinity g-C₃N₄ heterostructures by double-end modulation for photocatalysis, ACS Energy Lett., 2024, 9(4), 1915–1922 CrossRef CAS .
22. Q. Qiao, W. Q. Huang and Y. Y. Li, et al., In situ construction of 2D direct Z-scheme g-C₃N₄/g-C₃N₄ homojunction with high photocatalytic activity, J. Mater. Sci., 2018, 53(23), 15882–15894 CrossRef CAS .
23. H. L. Lee, Z. Sofer and V. Mazánek, et al., Graphitic carbon nitride: Effects of various precursors on the structural, morphological and electrochemical sensing properties, Appl. Mater. Today, 2017, 8, 150–162 CrossRef .
24. X. Zhang, X. Xie and H. Wang, et al., Enhanced photoresponsive ultrathin graphitic-phase C₃N₄ nanosheets for bioimaging, J. Am. Chem. Soc., 2013, 135(1), 18–21 CrossRef CAS PubMed .
25. L. Jiang, X. Yuan and G. Zeng, et al., Phosphorus- and sulfur-codoped g-C₃N₄: facile preparation, mechanism insight, and application as efficient photocatalyst for tetracycline and methyl orange degradation under visible light irradiation, ACS Sustainable Chem. Eng., 2017, 5(7), 5831–5841 CrossRef CAS .
26. J. Liu, T. Zhang and Z. Wang, et al., Simple pyrolysis of urea into graphitic carbon nitride with recyclable adsorption and photocatalytic activity, J. Mater. Chem., 2011, 21, 14398–14401 RSC .
27. H. Estrade-szwarckopf, XPS photoemission in carbonaceous materials: A “defect” peak beside the graphitic asymmetric peak, Carbon, 2004, 42(8–9), 1713–1721 CrossRef CAS .
28. J. Zhu, Y. Xiong and X. Mu, et al., Cobalt-doped g-C₃N₄ nanosheets for one-pot synthesis of imines under mild conditions, ACS Appl. Nano Mater., 2023, 6(7), 5127–5135 CrossRef CAS .
29. J. Jiang, Z. Xiong and H. Wang, et al., Sulfur-doped g-C₃N₄/g-C₃N₄ isotype step-scheme heterojunction for photocatalytic H₂ evolution, J. Mater. Sci. Technol., 2022, 118, 15–24 CrossRef CAS .
30. Y. Zheng, Z. Zhang and C. Li, A comparison of graphitic carbon nitrides synthesized from different precursors through pyrolysis, J. Photochem. Photobiol., A, 2017, 332, 32–44 CrossRef .
31. Y. Xu and S. P. Gao, Band gap of C₃N₄ in the GW approximation, Int. J. Hydrogen Energy, 2012, 37(15), 11072–11080 CrossRef CAS .
32. P. Wu, J. Wang and J. Zhao, et al., Structure defects in g-C₃N₄ limit visible light driven hydrogen evolution and photovoltage, J. Mater. Chem. A, 2014, 2(47), 20338–20344 RSC .
33. Y. Liu, C. Bian and Y. Li, et al., Aminobenzaldehyde convelently modified graphitic carbon nitride photocatalyst through Schiff base reaction: Regulating electronic structure and improving visible-light-driven photocatalytic activity for moxifloxacin degradation, J. Colloid Interface Sci., 2023, 630, 867–878 CrossRef CAS PubMed .
34. C. Bian, Y. Wang and Y. Yi, et al., Enhanced photocatalytic activity of S-doped graphitic carbon nitride hollow microspheres: Synergistic effect, high-concentration antibiotic elimination and antibacterial behavior, J. Colloid Interface Sci., 2023, 643, 256–266 CrossRef CAS PubMed .
35. S. Shao, X. Liu and R. Wang, et al., Morphologic and microstructural modulation of graphitic carbon nitride through EDTA-2Na mediated supramolecular self-assembly route: Enhanced visible-light-driven photocatalytic activity for antibiotic degradation, Appl. Surf. Sci., 2024, 669, 160501 CrossRef CAS .
36. Z. Wu, Y. Xue and X. He, et al., Surfactants-assisted preparation of BiVO₄ with novel morphologies via microwave method and CdS decoration for enhanced photocatalytic properties, J. Hazard. Mater., 2020, 387, 122019 CrossRef CAS PubMed .
37. L. Wei, S. Chen and X. Yang, et al., Synthesis of visible light driven spherical Bi₂MoO₆ structure for aflatoxin B1 photodegradation, Inorg. Chem. Commun., 2024, 160, 111886 CrossRef CAS .
38. M. Zalfani, Z. Y. Hu and W. B. Yu, et al., BiVO₄/3DOM TiO₂ nanocomposites: Effect of BiVO4 as highly efficient visible light sensitizer for highly improved visible light photocatalytic activity in the degradation of dye pollutants, Appl. Catal., B, 2017, 205, 121–132 CrossRef CAS .
39. K. Ren, W. Zhao and Z. Zhai, et al., 2D/0D Bi₂MoO₆ nanosheets/BN quantum dots photocatalysts with enhanced charge separation for efficient elimination of antibiotic, Appl. Surf. Sci., 2021, 562, 150144 CrossRef CAS .
40. D. Li, Y. Xiao and M. Pu, et al., A metal-free protonated g-C₃N₄ as an effective sodium percarbonate activator at ambient pH conditions: Efficiency, stability and mechanism, Mater. Chem. Phys., 2019, 231, 225–232 CrossRef CAS .
41. X. Li and X. Gan, Photo-Fenton degradation of multiple pharmaceuticals at low concentrations via Cu-doped-graphitic carbon nitride (g-C₃N₄) under simulated solar irradiation at a wide pH range, J. Environ. Chem. Eng., 2022, 10(5), 108290 CrossRef CAS .
42. P. Torres, M. Guzmán-Ortiz and B. Ramírez-Wong, Revising the role of pH and thermal treatments in aflatoxin content reduction during the tortilla and deep frying processes, J. Agric. Food Chem., 2001, 49(6), 2825–2829 CrossRef CAS PubMed .
43. Y. Zhang, Y. Sun and Y. Man, et al., Highly efficient adsorption and catalytic degradation of aflatoxin B1 by a novel porous carbon material derived from Fe-doped ZIF-8, Chem. Eng. J., 2022, 440, 135723 CrossRef CAS .
44. G. Pan and Z. Sun, Cu-doped g-C₃N₄ catalyst with stable Cu⁰ and Cu⁺ for enhanced amoxicillin degradation by heterogeneous electro-Fenton process at neutral pH, Chemosphere, 2021, 283, 131257 CrossRef CAS PubMed .
45. J. Zhang, Y. Hu and X. Jiang, et al., Design of a direct Z-scheme photocatalyst: Preparation and characterization of Bi₂O₃/g-C₃N₄ with high visible light activity, J. Hazard. Mater., 2014, 280, 713–722 CrossRef CAS PubMed .
46. J. Mao, L. Zhang and H. Wang, et al., Facile fabrication of nanosized graphitic carbon nitride sheets with efficient charge separation for mitigation of toxic pollutant, Chem. Eng. J., 2018, 342, 30–40 CrossRef CAS .
47. J. Mao, Q. Zhang and P. Li, et al., Geometric architecture design of ternary composites based on dispersive WO₃ nanowires for enhanced visible-light-driven activity of refractory pollutant degradation, Chem. Eng. J., 2018, 334, 2568–2578 CrossRef CAS .
48. D. Sun, J. Mao and L. Cheng, et al., Magnetic g-C₃N₄/NiFe₂O₄ composite with enhanced activity on photocatalytic disinfection of Aspergillus flavus, Chem. Eng. J., 2021, 418, 129417 CrossRef CAS .
49. L. Yao, C. Sun and H. Lin, et al., Enhancement of AFB1 Removal Efficiency via Adsorption/Photocatalysis Synergy Using Surface-Modified Electrospun PCL-g-C₃N₄/CQDs Membranes, Biomolecules, 2023, 13, 550 CrossRef CAS PubMed .
50. R. Song, L. Yao and C. Sun, et al., Electrospun Membranes Anchored with g-C₃N₄/MoS₂ for Highly Efficient Photocatalytic Degradation of Aflatoxin B1 under Visible Light, Toxins, 2023, 15(2), 133 CrossRef CAS PubMed .
51. F. Dong, Z. Wang and Y. Li, et al., Immobilization of Polymeric g-C₃N₄ on Structured Ceramic Foam for Efficient Visible Light Photocatalytic Air Purification with Real Indoor Illumination, Environ. Sci. Technol., 2014, 48(17), 10345–10353 CrossRef CAS PubMed .
52. D. Wang, H. Shen and L. Guo, et al., Porous BiOBr/Bi₂MoO₆ heterostructures for highly selective adsorption of methylene blue, ACS Omega, 2016, 1(4), 566–577 CrossRef CAS PubMed .
53. F. Yue, C. Wang and W. Duan, et al., Selenium vacancies regulate d-band centers in Ni₃Se₄ toward paired electrolysis in anion-exchange membrane electrolyzers for upgrading N-containing compounds, Sci. China: Chem., 2023, 66(7), 2109–2120 CrossRef CAS .
54. C. Yang, Y. Zhang and F. Yue, et al., Co doping regulating electronic structure of Bi₂MoO₆ to construct dual active sites for photocatalytic nitrogen fixation, Appl. Catal., B, 2023, 338, 123057 CrossRef CAS .
55. N. Benkerroum, Chronic and acute toxicities of aflatoxins: Mechanisms of action, Int. J. Environ. Res. Public Health, 2020, 17(2), 423 CrossRef CAS PubMed .
56. X. Geng, N. Wang and Y. Gao, et al., A novel HPLC flow cell integrated UV light emitting diode induced fluorescence detector as alternative for sensitive determination of aflatoxins, Anal. Chim. Acta, 2018, 1033, 81–86 CrossRef CAS PubMed .

This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2024