Double Z-scheme biochar-based g-C3N4/Bi2WO6/Ag3PO4 nanocomposite for efficient removal of antibiotics and synergistic mechanisms
Tongtong Wang, Di Zhang, Hui Shi, Jiyong Zheng, Huixia Wang, Eric Lichtfouse
TL;DR
A new nanocomposite material efficiently removes antibiotics from water using a unique photocatalytic mechanism.
Contribution
The study introduces a biochar-based nanocomposite with a double Z-scheme heterojunction for enhanced photocatalytic antibiotic removal.
Findings
The nanocomposite achieved nearly complete tetracycline removal in 120 minutes with a high removal rate.
It showed over 85% antibiotic removal in real wastewater and 99% bacterial sterilization within 48 hours.
The composite's structure and heteroatoms significantly enhance photocatalytic activity through radical generation.
Abstract
Photocatalysis research has evolved towards increasingly sophisticated structural regulation and material design. The synergistic enhancement of photocatalysis by multi-component semiconductors and biochar warrants detailed investigation. This study introduces an innovative biochar-based g-C3N4/Bi2WO6/Ag3PO4 nanocomposite (CN/Bi/Ag@ACB), which was applied to the efficient removal of antibiotic pollutants represented by tetracycline (TC). Findings reveal that CN/Bi/Ag@ACB forms a double Z-scheme heterojunction, significantly reducing photogenerated carrier recombination. It absorbs light in the 200-800 nm range, with a band gap of 1.91 eV. Under 120 min of illumination, the composite nearly completely removed 50 mg·L-1 of TC, achieving a removal rate of 0.0351 min-1, which is 8.56-13.50 times higher than that of the individual semiconductors. In real wastewater, TC removal exceeded…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 10
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 11- —the Shaanxi Provincial Natural Science Basic Research Program for Youth Project (2025JC-YBQN-457), the Xi'an University of Architecture and Technology National Foundation Incubation Program (X20250047
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Photocatalysis Techniques · TiO2 Photocatalysis and Solar Cells · Advanced oxidation water treatment
Introduction
The expansion of economic activities and the growing global population have exacerbated environmental pollution and energy crises, posing significant challenges across many regions. There is an increasing demand from both the public and governments for effective, sustainable methods to mitigate pollution and promote resource recycling^1^. The application scope of semiconductor photocatalysis has recently broadened, showing immense promise in fields ranging from the oxidative degradation of emerging pollutants in water to photocatalytic hydrogen production^2^. This technology is poised to address the complex issues related to environmental pollution and energy reduction through multidisciplinary approaches. Photocatalysis, in particular, stands out as a green, eco-friendly process that merges photoreaction and catalysis, converting light energy into chemical energy^3^. This process generates photogenerated electron–hole pairs (carriers), which are instrumental in degrading emerging pollutants and addressing energy demands^4^. Developing efficient photocatalysts is essential for the sustained advancement and application of photocatalytic technology. The current research trend focuses on solar energy utilization, driving scientists to concentrate on semiconductor materials responsive to visible light for green development. Among various semiconductors, graphitic carbon nitride (g-C_3_N_4_), bismuth tungstate (Bi_2_WO_6_), and silver phosphate (Ag_3_PO_4_) have garnered attention due to their unique structural, physical, and chemical properties, which facilitate visible light photocatalytic activity in environmental and energy-related applications.
g-C_3_N_4_ is a polymeric, nonmetallic semiconductor with a two-dimensional layered structure akin to graphite, featuring a layer spacing of approximately 0.326 nm, slightly narrower than that of graphite^5^. The conduction band (ECB) of g-C_3_N_4_ consists of C atom p_z_ orbitals positioned at approximately −1.30 eV, while the valence band (EVB) comprises N atom p_z_ orbitals at approximately 1.40 eV. Bi_2_WO_6_ features a layered structure formed by alternating WO_6_ and Bi_2_O_2_ units along the c-axis through intermolecular forces, with O atoms shared between layers. Bi_2_WO_6_ belongs to the orthorhombic crystal system, characteristic of perovskite minerals, with a band gap (Eg) typically ranging from 2.6 to 2.8 eV, and EVB and ECB values of approximately 3.21 and 0.31 eV, respectively^6^. Ag_3_PO_4_, an important N-type semiconductor in the silver family, exhibits a cubic crystal structure capable of absorbing visible light up to <530 nm, with an Eg of around 2.36 to 2.43 eV^7^. Formed PO_4_ tetrahedra decreases Ag–O bonding energy, thus enhancing the surface migration of photoelectrons^8^. Notably, the physico-chemical properties of these semiconductors may exhibit slight variations depending on the preparation process. Photocatalytic performance using pure semiconductors often falls short, with low quantum efficiency, particularly in treating wastewater containing persistent pollutants. The three semiconductors also exhibit notable limitations, including poor visible light utilization^3^, high monomer condensation^9^, susceptibility to agglomeration or photo-corrosion^7^, low specific surface area^10^, and significant exciton effects^11^, all of which contribute to high photogenerated carrier recombination rates^12,13^. These issues hinder the photocatalytic activity from meeting application demands.
To enhance light absorption and photocatalytic efficiency, researchers have proposed various design strategies, such as band structure regulation, morphology control, heterojunction construction, ion doping, and the addition of co-catalysts^12,14^. Notably, constructing heterojunctions by integrating pure semiconductors with carbon materials has emerged as one of the most effective methods for boosting photocatalytic performance^15,16^. Selecting carbon materials with high adsorption capacities can increase reaction sites, significantly enhance conductivity, and optimize band structures^13^. As a prevalent mesoporous carbon material, biochar possesses a substantial specific surface area, with a polar surface that is often negatively charged. Its appeal is further enhanced by cost-effectiveness, oxygen-containing functional groups (OFGs), wide availability from various feedstocks, a polar (typically negatively charged) surface, and chemical stability. Especially, biochar is known to be rich in environmentally persistent free radicals (EPFRs), which can mediate catalytic reactions^5^ or function as co-catalysts due to their defective structure^17^. Consequently, the research on loading semiconductor materials onto biochar or modifying it is rapidly advancing. Biochar-based g-C_3_N_4_, Bi_2_WO_6_, or Ag_3_PO_4_ composites have been developed and applied in pollutant treatment^18^. However, our research indicates that these binary composites still suffer from a high recombination rate of photogenerated carriers, making them suboptimal for antibiotic degradation and prone to inactivation due to photochemical corrosion^5,19^. To further enhance the photocatalytic activity, synthesizing quaternary composite photocatalysts with double-heterojunction structures is gaining importance and is deserving. Kumar et al.^20^ developed a quaternary composite (BiOCl/g-C_3_N_4_/Cu_2_O/Fe_3_O_4_) via the co-precipitation method, which effectively coupled p–n–p junctions, inducing energy band shifts and charge separation within the built-in electric field. Under visible light or natural sunlight, this composite demonstrated effective removal of sulfamethoxazole within a very short time. Similarly, Bagheri et al.^21^ reported carbon-based quaternary composites that exhibited high photocatalytic efficiency in degrading organic pollutants. Ai et al.^22^ reviewed g-C_3_N_4_-related quaternary composites, highlighting their potential to integrate multifunctional synergies across different semiconductors, leading to full sunlight utilization and effective suppression of charge recombination. The photocatalytic advantages of multicomponent systems, particularly quaternary composites, are evident, making their preparation and synergistic photocatalytic mechanisms worthy of in-depth exploration. However, reports on biochar-based quaternary composites remain scarce, with current research still in its preliminary stages.
Tetracycline (C_22_H_24_N_2_O_8_, TC) is the most widely utilized antibiotic in animal production across several countries, and is also frequently detected at high concentrations in various aquatic environments^23^. Residual TC is resistant to natural degradation and poses significant ecological risks, including the induction of antibiotic-resistant pathogens and antibiotic-resistance genes, which ultimately threaten human health. Given these concerns, TC served as the representative antibiotic pollutant for evaluating the photocatalytic performance. This study aims: (1) to synthesize a biochar-based g-C_3_N_4_/Bi_2_WO_6_/Ag_3_PO_4_ quaternary composite photocatalyst (CN/Bi/Ag@ACB) through a three-step process involving ultrasonic-ball milling, hydrothermal synthesis, and chemical co-precipitation, followed by a comprehensive characterization. We assume that the constructed composites have superior photocatalytic activity. (2) To systematically assess the efficacy of CN/Bi/Ag@ACB, this study investigated the removal rate of TC by quantifying the kinetic reaction, measuring the mineralization efficiency via total organic carbon (TOC) analysis, and determining the role of various environmental variables. The study also assessed the efficacy of CN/Bi/Ag@ACB in removing various antibiotics from real wastewater and its application in the disinfection of S. aureus and E. coli. (3) To explore the photogenerated carrier model of CN/Bi/Ag@ACB, identify the dominant free radical (i.e., reactive oxygen species, ROS), and elucidate the synergistic mechanisms responsible for enhancing photocatalytic activity. (4) To identify the degradation process and intermediate products of TC and further deduce possible degradation pathways. This research aims to provide novel insights into the interfacial composite process and the synergistic photocatalytic mechanisms of multi-component biochar-based nanocomposites, while establishing a theoretical basis for tetracycline antibiotics’ efficient removal and new composite design.
Results and discussions
Basic physical-chemical properties analysis
The preparation process of CN/Bi/Ag@ACB is shown in Fig. 1. The X-ray diffraction (XRD) patterns of CN/Bi/Ag@ACB, shown in Fig. S1a, reveal weak diffraction peaks at 2θ values of 13.1° and 27.2°, corresponding to the (001) and (002) planes of hexagonal g-C_3_N_4_ (JCPDS 87-1526), respectively. The XRD pattern revealed characteristic diffraction peaks corresponding to orthorhombic Bi_2_WO_6_ (JCPDS 39-0256), with the observed 2θ values at 28.3° (131), 32.8° (200), 47.1° (202), and 56.0° (133) confirming the crystalline phase formation. The peaks at 2θ = 20.9°, 29.7°, 33.3°, and 36.3° are matched to the (110), (200), (210), and (211) planes of cubic Ag_3_PO_4_ (JCPDS 06-0505), respectively. The other diffraction peaks are labeled in Fig. S2a. The absence of noticeable diffraction peaks for ACB is due to its amorphous nature or low crystallinity. Therefore, the XRD pattern of CN/Bi/Ag@ACB primarily reflects the presence of g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_, indicating the successful construction of a composite heterojunction. The heterojunction promotes photogenerated charge migration, consequently boosting photocatalytic performance^4^.Fig. 1. Schematic illustration of the synthesis of CN/Bi/Ag@ACB in this work (Some elements in the figure are reproduced with permission from refs. ^17, 20^).
As depicted in Fig. S1b, characteristic peaks from ACB, Ag_3_PO_4_, Bi_2_WO_6_, and g-C_3_N_4_ were identified in CN/Bi/Ag@ACB. Notably, the broad stretching vibration peak at 3420 cm^−1^ corresponds to the −OH groups from the oxygen-containing functional groups in ACB. Fourier-transform infrared (FTIR) spectroscopy identified characteristic absorption bands at approximately 1010 and 560 cm^−1^, corresponding to the antisymmetric and symmetric P–O stretching modes of phosphate groups in the Ag_3_PO_4_. The characteristic peak near 1010 cm^−1^ in CN/Bi/Ag@ACB is broadened due to the overlap with the C–O stretching vibration of ACB at the same position. The presence of Bi_2_WO_6_ is confirmed by absorption bands observed at 439 cm^−1^ (Bi–O stretching) and at 730 cm^−1^ (associated with typical Bi–O–W bridging modes and W–O stretching). Additionally, the characteristic peaks of C–N heterocycles or C–N–C stretching vibrations in the 1200–1700 cm^−1^ range, typically associated with g-C_3_N_4_, were attenuated after the hydrothermal reaction with ACB and Bi_2_WO_6_, and chemical co-precipitation with Ag_3_PO_4_. The C–N heterocycles were primarily evidenced by two distinct vibrational peaks, appearing at 1384 and 1630 cm^−1^.
In Fig. S1c, the typical characteristic peaks of Bi_2_WO_6_ were observed in CN/Bi/Ag@ACB, including the asymmetric and symmetric modes of O–W–O at 798 and 886 cm^−1^, respectively. The asymmetric W–O stretching vibration generates the peak at 710 cm^−1^, while the prominent band at 300 cm^−1^ and the weak band at 150 cm^−1^ originate from vibrational modes involving the Bi–O bonds, BiO_6_ polyhedron, and WO_6_ octahedron^13^. Compared to the characteristic peaks of ACB, the positions of the D band (disordered amorphous carbon), G band (graphitic carbon), and 2D band (related to single or multi-layer graphene) in CN/Bi/Ag@ACB showed no significant shifts. However, the intensities of these peaks were markedly reduced, indicating the integration of g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ into the multilayer graphitized structure of ACB. The ratio of the D-peak intensity (ID) to the G-peak intensity (IG) in the composite is 2.51 (Fig. S2b), which is significantly higher than that of ACB (0.98). This indicates a higher degree of structural defects and atomic disorder in CN/Bi/Ag@ACB. This increase likely contributes to the suppression of photogenerated carrier recombination, enhancing photocatalytic efficiency.
The composite material demonstrates a type IV adsorption-desorption isotherm, as shown in Fig. S1d. This isotherm features a clear hysteresis loop extending across the relative pressure range of 0.1–1.0. This behavior is indicative of a typical mesoporous structure (2–50 nm pore size). More detailed data is provided in Table S1 and S2. The specific surface area (SBET) of CN/Bi/Ag@ACB is measured at 27.94 m^2^ g^−1^. Compared to g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_, CN/Bi/Ag@ACB exhibits a significantly larger SBET and pore size. The pore size distribution, shown in Fig. S3, further reveals the presence of some micropores (<2 nm) within CN/Bi/Ag@ACB, with an average pore size of 1.96 nm. This is primarily attributed to the high specific surface area and porous structure provided by the ACB.
Micromorphology
Combined FESEM (Figs. 2 and S4), HRTEM (Fig. 3), and EDS (Table S1) analyses revealed the materials’ microstructural features. The distinct microstructures of g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ are depicted in Fig. S4a, b, c, respectively, and are integrated within the CN/Bi/Ag@ACB composite (Figs. 2a and S4d). g-C_3_N_4_ appears as a loose, flaky structure stacked in blocks, while Bi_2_WO_6_ forms nanosheets assembled into flower-like sheets or microspheres. Ag_3_PO_4_ is characterized by irregular spherical or cubic particles. ACB features a loose-surfaced, porous honeycomb architecture abundant in pores, where g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ are densely embedded. Elemental mapping of CN/Bi/Ag@ACB (Fig. 2b) and the detailed EDS spectra (Table S1) demonstrate the homogeneous spatial dispersion of all constituent elements (C, O, N, P, Ag, Bi, W, and K), evidencing the effective construction of the g-C_3_N_4_/Bi_2_WO_6_/Ag_3_PO_4_-activated biochar composite.Fig. 2. Field emission scanning electron microscopy (FESEM) and mapping images of CN/Bi/Ag@ACB.a FESEM image of CN/Bi/Ag@ACB, b FESEM-Mapping mixed total image of CN/Bi/Ag@ACB, followed by the distribution maps of each element as C, Ag, Bi, O, K, N, P, and W, respectively.Fig. 3. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images of CN/Bi/Ag@ACB.a TEM image of CN/Bi/Ag@ACB, b, c HRTEM images of CN/Bi/Ag@ACB, and d selected area electron diffraction (SAED) of CN/Bi/Ag@ACB.
In Fig. 3a, ACB within the composite is presented as a thin, wrinkled layer attributed to its amorphous carbon composition, similar to the morphology observed in graphene nanosheets. g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ are anchored onto this thin layer, aggregating into dark lattice spots. Figure 3b reveals lattice spacings of 0.3361 nm (g-C_3_N_4_ (002)), 0.3151 nm (Bi_2_WO_6_ (131)), and 0.2454 nm (Ag_3_PO_4_ (211)). The close contact between g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ nanoparticles with ACB is evident, confirming the formation of a quaternary composite heterojunction. Similar observations in Fig. 3c reveal the presence of the (200) plane of Bi_2_WO_6_ and the (210) plane of Ag_3_PO_4_ in different regions. The selected area electron diffraction pattern of CN/Bi/Ag@ACB (Fig. 3d) demonstrates that the composite photocatalyst is polycrystalline, with identified planes including the (002) of g-C_3_N_4_, the (131) and (200) of Bi_2_WO_6_, and the (210) and (211) of Ag_3_PO_4_.
XPS analysis
The XPS spectra presented in Fig. 4 provide a detailed characterization of the bonded valence structure within CN/Bi/Ag@ACB. The survey spectrum in Fig. 4a reveals that CN/Bi/Ag@ACB predominantly comprises W 4f, P 2p, Bi 4f, C 1s, Ag 3d, N 1s, and O 1s, with atomic percentages of 1.16%, 1.41%, 3.32%, 55.27%, 4.92%, 6.81%, and 27.11%, respectively. Four deconvoluted peaks in the C 1s high-resolution spectra (Fig. 4b) were identified at binding energies of 284.8 eV (C–C), 286.2 eV (C–OH or C–N), 288.4 eV (C=O), and 291.7 eV (π–π interactions)^5^, which are mainly attributed to ACB and g-C_3_N_4_. The N 1s spectra (Fig. 4c) show three deconvoluted peaks at 399.5, 400.5, and 401.6 eV, representing C–N, sp^2^ hybridized nitrogen (C=N–C), and sp^3^ hybridized nitrogen (N–(C)3) bonds derived from g-C_3_N_4_, respectively. These peaks relate to the basic structural units of C_3_N_3_ or C_6_N_7_ rings in g-C_3_N_4_^24^. O 1s spectral features (Fig. 4d) at 529.8 eV (Bi_2_WO_6_/Ag_3_PO_4_ lattice oxygen, Ag–O/Bi–O bonds)^6^ and 532.3 eV (oxygen adsorbates, e.g.,−OH) evidence dual oxygen states. The substantial peak area of lattice oxygen suggests a significant presence of Bi_2_WO_6_ and Ag_3_PO_4_ in the composite. In Fig. 4e, the Ag 3d spectra reveal two prominent peaks at 367.9 and 373.9 eV, corresponding to Ag 3d5/2 and Ag 3d3/2, respectively, which confirm the presence of Ag^+^. Notably, no peaks corresponding to elemental silver were detected, consistent with the XRD findings, indicating that Ag_3_PO_4_ exhibits high crystallinity and has not undergone photo-corrosion. The P 2p spectrum (Fig. 4f) shows a characteristic peak at 133.9 eV, indicative of the P^5+^ state in Ag_3_PO_4_^25^. The high-resolution Bi 4f spectra (Fig. 4g) display peaks at 164.2 and 159.0 eV, corresponding to Bi 4f5/2 and Bi 4f7/2 binding energies, confirming the presence of Bi^3+^. Finally, the W 4f spectra (Fig. 4h) reveal peaks at 37.0 and 34.8 eV, corresponding to W 4f5/2 and W 4f7/2, respectively, confirming the existence of W^6+^ in CN/Bi/Ag@ACB^13^.Fig. 4X-ray photoelectron spectra (XPS) of CN/Bi/Ag@ACB.a Survey, b C 1s, c N 1s, d O 1s, e Ag 3d, f P 2p, g Bi 4f, and h W 4f.
Optical properties
As illustrated in Fig. 5a, the maximum absorption edges for pure g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ are approximately 508, 465, and 556 nm, respectively. ACB integration induces a red shift in CN/Bi/Ag@ACB, enabling strong broadband absorption (200–800 nm) with sustained ~80% absorbance. This value substantially exceeds those of the constituent semiconductors. This enhanced absorption suggests that CN/Bi/Ag@ACB not only broadens the absorption spectrum but also effectively reduces the rate of charge recombination, attributed to the full-band absorption capabilities of ACB and the formation of a quaternary composite heterojunction. The Eg of the prepared materials was calculated using the Kubelka–Munk method. As depicted in Fig. 5b, the direct Eg values for g-C_3_N_4_, Bi_2_WO_6_, Ag_3_PO_4_, and ACB are 2.70, 2.60, 2.30, and 1.91 eV, respectively. The incorporation of ACB significantly reduces the Eg of CN/Bi/Ag@ACB. Figure 5c presents the valence band spectra of CN/Bi/Ag@ACB obtained via XPS, where the valence band potential is referenced to the Fermi level. As shown in Fig. 5d, all samples exhibit distinct absorption characteristics, with CN/Bi/Ag displaying a maximum absorption edge at 484 nm. However, materials composite with biochar demonstrate superior absorption intensity across the entire UV–vis region (200–800 nm) compared to CN/Bi/Ag. Notably, CN/Bi/Ag@ACB maintains the highest absorption intensity and stable absorption characteristics. According to Linic et al.^26^, the valence band potential (EVB, NHE) is converted to the standard hydrogen electrode scale. The conversion formula used is as follows: EVB, NHE = φ + EVB, xps−4.44, where φ represents the work function of the instrument, which was determined to be φ = 4.5 eV in this study. Thus, the EVB for CN/Bi/Ag@ACB is estimated to be 1.38 eV. The ECB is then calculated to be −0.53 eV using Eq. (1). Furthermore, the EVB values for g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ can be estimated using Eq. (2), and their respective ECB values can be derived.
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{g}={E}_{{{\mathrm{VB}}}}-{E}_{{{\mathrm{CB}}}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${E}_{{{\mathrm{VB}}}}=X-{E}^{e}+0.5{E}_{g}$$\end{document}Fig. 5UV–vis/diffuse reflectance spectroscopy (UV–vis/DRS) spectra of prepared photocatalysts.a UV–vis/DRS spectra, b direct band gap based on tauc plot, c XPS valence band of CN/Bi/Ag@ACB, and d comparison of light absorption properties of different composites.
Here, E^e^ (approximately 4.5 eV) and X are the important energy level and electronegativity parameters of this empirical equation, respectively, and detailed physical significance can be obtained by referring to the literature^27^. Based on the reported X value of 4.64 for g-C_3_N_4_^20^, its valence and conduction band positions are estimated at EVB = 1.49 eV and ECB = −1.21 eV. Similarly, Bi_2_WO_6_ exhibits EVB = 3.01 eV and ECB = 0.41 eV^28^; while Ag_3_PO_4_ shows EVB = 2.61 eV and ECB = 0.31 eV^27^.
Photocatalytic performance
As systematically evaluated in Fig. 6, the TC removal efficiency showed that ACB adsorption reached equilibrium within 60 min and remained stable thereafter (Fig. 6a). The dark reaction process primarily involves physical adsorption, where macromolecular pollutants are not tightly bound by chemical bonds but are instead adsorbed onto the surface or diffuse into the porous structure of ACB through physical forces, mainly Van der Waals interactions. The adsorption-desorption process reaches dynamic equilibrium within 1 h, stabilizing the concentration of macromolecular organic pollutants in the reaction solution. Additionally, due to the limited SBET of g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_, their adsorption capacity for TC is minimal, leading to rapid attainment of adsorption-desorption equilibrium during the dark reaction. Upon 120 min of photoreaction, the TC removal rate by CN/Bi/Ag@ACB reached 99.04%, indicating almost complete removal. In contrast, the TC removal rates by ACB, g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ were only 26.58%, 30.12%, 35.87%, and 41.02%, respectively. Notably, CN/Bi/Ag@ACB significantly enhanced TC removal during the photoreaction phase, primarily due to the improved visible light degradation capability of the composite. Figure 6b presents a pseudo-first-order kinetic constant of CN/Bi/Ag@ACB (0.0351 min^−1^, 90 min), surpassing Ag_3_PO_4_, Bi_2_WO_6_, and g-C_3_N_4_ by factors of 8.56 times (0.0040 min^−1^), 11.70 times (0.0030 min^−1^), and 13.50 times (0.0026 min^−1^), respectively. This confirms its superior visible-light photocatalytic activity. To further assess the mineralization capacity of these photocatalysts on TC, the TOC removal rate was measured. Figure 6c shows that CN/Bi/Ag@ACB achieved a TOC removal rate of 67.74% for TC within 90 min, whereas Ag_3_PO_4_, Bi_2_WO_6_, and g-C_3_N_4_ were 29.46%, 23.63%, and 20.66%, respectively. Thus, CN/Bi/Ag@ACB exhibits 2.3–3.3-fold higher TOC removal rates than pure semiconductors, suggesting that TC undergoes gradual degradation and mineralization.Fig. 6. Photocatalytic performance of prepared materials.a Adsorption-photocatalytic degradation of TC using the prepared photocatalysts, b pseudo-first-order kinetic curves of TC by the prepared photocatalysts, and c total organic carbon (TOC) removal efficiency. (Experimental conditions: initial concentrations of TC at 50 mg L^−1^, photocatalysts dosage of 1 g L^−1^, initial pH of ~7.0, 500 W xenon lamp).
As shown in Fig. S5, the removal performance of 12 prepared materials for high-concentration TC. During the dark reaction phase, all materials exhibited limited physical adsorption. Upon illumination, the CN/Bi/Ag@ACB composite exhibited the steepest removal curve slope, achieving near-complete TC removal within 120 min. Its performance significantly surpassed all pure semiconductor materials, CN/Bi/Ag, biochar-based ternary composite, and biochar-based binary composite. This result visually confirms that the quaternary system designed in this study exhibits optimal photocatalytic activity through multicomponent synergy coupled with biochar. Additionally, the biochar-based ternary composites (CN/Bi@ACB, Bi/Ag@ACB, CN/Ag@ACB) demonstrated generally better TC removal performance than the biochar-based binary composites (Ag@ACB, Bi@ACB, CN@ACB). All biochar-based composites showed higher removal efficiency than the CN/Bi/Ag without biochar addition and the pure semiconductor materials. This indicates that loading biochar significantly enhances the photocatalytic performance, suggesting that the removal mechanism of biochar-based composites involves not only adsorption by the carrier and photocatalytic degradation by the semiconductor, but also a synergistic adsorption-photocatalytic degradation mechanism.
To further investigate the effects of different mass ratios of biochar, g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_ on adsorption-photocatalytic activity, CN/Bi/Ag@ACB composites with varying loading ratios were prepared. The results are presented in Fig. S6. It is noteworthy that this study set a mass ratio of g-C_3_N_4_, Bi_2_WO_6_, Ag_3_PO_4_, and ACB in the CN/Bi/Ag@ACB composite to 1:1:1:1 as the baseline. The corresponding CN/Bi/Ag@ACB (2:1:1:1, i.e., doubled g-C_3_N_4_ component), CN/Bi/Ag@ACB (1:2:1:1, i.e., doubled Bi_2_WO_6_ component), CN/Bi/Ag@ACB (1:1:2:1, i.e., doubled Ag_3_PO_4_ component), and CN/Bi/Ag@ACB (1:1:1:0.5, i.e., halved ACB component) were prepared and tested. The results still showed that CN/Bi/Ag@ACB (the default quality ratio is 1:1:1:1) had the best photocatalytic activity. The reason may be explained by the synergistic improvement of the adsorption and photocatalytic activity of biochar as a carrier. The risk of photocorrosion and Ag^+^ leaching that may be associated with an excess of Ag_3_PO_4_, the underutilization of visible light when there is an excess of Bi_2_WO_6_, and the poor adsorption and agglomeration that may occur when there is an excess of g-C_3_N_4_. In our view, an appropriate quantity of biochar not only enhances adsorption and electron transfer but also avoids reducing activity through light shading or excessive dilution of the semiconductor. A reasonable ratio (1:1:1:1) is essential for constructing an effective double Z-scheme potential gradient and inhibiting photocorrosion.
Additionally, the removal performance for norfloxacin (NOR) and chloramphenicol (CAP) is illustrated in Fig. S7. After 120 min of photoreaction, CN/Bi/Ag@ACB achieved a NOR removal rate of 98.01% (Fig. S7a), while the removal rates for Ag_3_PO_4_, Bi_2_WO_6_, and g-C_3_N_4_ were 35.45%, 30.35%, and 28.48%, respectively. Similarly, as shown in Fig. S7b, CN/Bi/Ag@ACB reached a CAP removal rate of 98.84% after 120 min of photoreaction, significantly outperforming Ag_3_PO_4_ (32.16%), Bi_2_WO_6_ (23.63%), and g-C_3_N_4_ (25.78%) for CAP removal. These results clearly demonstrate that CN/Bi/Ag@ACB has a pronounced removal effect on the three antibiotics at high concentrations, with photocatalytic degradation playing a much more significant role than adsorption. The adsorption-desorption dynamic equilibrium for NOR and CAP by ACB is similar to that observed for TC, indicating that ACB’s removal effect on these antibiotics is limited.
Various environmental impacts
Environmental factors can significantly influence the photocatalytic activity of CN/Bi/Ag@ACB. Figure S8a demonstrates a declining TC removal efficiency for CN/Bi/Ag@ACB with increasing initial concentrations (20–60 mg L^−1^), where the reduction magnitude attenuates at higher concentrations. Notably, when the initial TC concentration is below 20 mg L^−1^, adsorption predominates. However, when the concentration exceeds 50 mg L^−1^, the photocatalytic degradation process slows down, likely due to the active sites being heavily occupied by TC molecules, leading to reduced interfacial contact. The TC removal rate by CN/Bi/Ag@ACB increases with the solution’s pH, though the rate of increase diminishes at higher pH levels (Fig. S8b). The removal rate of TC at pH 9.01 is slightly higher than at pH 10.94 and 6.86, possibly because −OH participates in photochemical reactions in alkaline conditions, generating more ROS. Conversely, at pH 2.94, the removal rate is 15.54% lower than at pH 6.86, suggesting that CN/Bi/Ag@ACB possesses some acid-base buffering capacity and maintains high degradation activity even in strongly acidic environments. Figure S8c shows that the TC removal rate by CN/Bi/Ag@ACB increases with higher dosages of the photocatalyst. The impact of coexisting ions on TC removal is presented in Fig. S8d. Compared to the control treatment, NaCl, NaNO_3_, and Na_2_SO_4_ treatments inhibited the removal efficiency of TC, with the most significant reduction being 10.76%. This effect may be related to the strong anti-interference ability of CN/Bi/Ag@ACB. Interestingly, the presence of Na_2_CO_3_ enhances the removal efficiency by approximately 3.57% compared to the control. This improvement is likely due to Na_2_CO_3_’s alkaline nature, which creates a slightly alkaline environment that can boost TC removal. Hailili et al.^29^ reported enhanced visible-light absorption of TC under alkaline versus acidic conditions, correlating with accelerated photodegradation efficacy in alkaline environments.
Photostability and reusability
Building on the typical adsorption-photocatalytic degradation experiments, five consecutive recycling tests were conducted to evaluate the reusability and photostability of CN/Bi/Ag@ACB. As indicated in Fig. S9a, after 5 cycles, the removal rate of CN/Bi/Ag@ACB for TC was only 3.88% lower than in the initial cycle. This indicates that CN/Bi/Ag@ACB is capable of nearly complete removal of TC at the reaction interface, subsequently releasing active sites, which allows the composite material to “self-purify” and be reused.
Further investigation into the photostability of CN/Bi/Ag@ACB involved recovering the samples for XRD (Fig. S9b) and FTIR (Fig. S9c) analysis. XRD and FTIR analyses confirm the structural integrity of CN/Bi/Ag@ACB, with negligible variations observed in both crystalline phases and functional groups after cyclic testing, indicating that CN/Bi/Ag@ACB maintains good stability and reusability throughout the photocatalytic process. However, after the recycling tests, FTIR analysis revealed the attenuated intensity of key vibrational modes: the broad −OH stretch at 3420 cm^−1^, along with characteristic absorptions at 730 cm^−1^ (Bi–O bonds, W–O stretching, and W–O–W bridging) and 560 cm^−1^ (P–O stretching). This reduction is related to the re-adsorption of degradation intermediates of TC by CN/Bi/Ag@ACB, leading to the shielding of −OH, W–O, Bi–O, or P–O bonds.
Actual wastewater treatment
As depicted in Fig. S9d, CN/Bi/Ag@ACB provided rapid purification (>99%) of TC within 10 min in surface waters from the Weihe and Yellow Rivers, Shaanxi Province, China. Moreover, after 180 min, TC removal efficiencies in Wastewater A and B were 86.0% and 90.9%, respectively. As outlined in Table S3, these industrial wastewaters also contained other antibiotics, such as NOR and CAP. In wastewater A, the CN/Bi/Ag@ACB material demonstrated significant removal efficiencies, achieving 55.7% for NOR and 50.9% for CAP (Fig. S7c), respectively. Similarly, in Wastewater B, NOR and CAP removal rates were 38.7% and 71.5% (Fig. S7c), respectively. These results demonstrate that CN/Bi/Ag@ACB can simultaneously photodegrade TC, NOR, and CAP within the same photocatalytic reaction system. Notably, the composite material exhibited a more pronounced removal effect on TC, which is likely attributed to the specific structure and photosensitivity of TC.
Disinfection effect
To ensure the quality of treated water, some organizations have established mandatory disinfection standards for wastewater treatment. The disinfection efficacy of CN/Bi/Ag@ACB is illustrated in Fig. S10. The disinfection effects of various photocatalysts on E. coli (Fig. S10a) after 12 h of culture showed the following order: CN/Bi/Ag@ACB > Ag_3_PO_4_ (10 min of light) > Bi_2_WO_6_ > Ag_3_PO_4_ (10 min of dark) > g-C_3_N_4_. This trend was consistent in plates cultured for 24 and 48 h, indicating that these materials significantly inhibit E. coli growth compared to the blank control, exhibiting strong disinfecting capabilities. Notably, the disinfection efficacy of CN/Bi/Ag@ACB surpasses that of Ag_3_PO_4_, Bi_2_WO_6_, and g-C_3_N_4_. Colony counts, estimated through microscopic observation and UV–vis spectrophotometry (OD = 600), revealed that the sterilization rates of CN/Bi/Ag@ACB exceeded 99% at both 12 and 24 h, with almost no visible E. coli colonies on the plates. Moreover, similar disinfection results were observed against S. aureus (Fig. S10b). CN/Bi/Ag@ACB treatments showed virtually no bacterial presence at 12, 24, or 48 h of incubation. The estimated sterilization rate of CN/Bi/Ag@ACB on S. aureus at 48 h was approximately 99% compared to the blank control. The mechanism behind this sterilization effect likely involves the generation of ROS via photocatalytic reactions, the nano-interface effect of the composite, and the antibacterial action of dissolved Ag^+^ ions^19^.
The average leaching concentrations of Ag^+^, Bi^3+^, and W^6+^ of CN/Bi/Ag@ACB were detected by ICP-MS as 0.514 mg L^−1^, 0.096 mg L^−1^, and 21.708 μg L^−1^, respectively. Combined with the above recyclability and photostability analyses, such a low concentration of heavy metal leaching basically does not affect the crystal structure of CN/Bi/Ag@ACB. This also implies that CN/Bi/Ag@ACB has good photostability with low metal leaching risk. Based on the analysis of the leaching test data, the average leaching concentration of Ag^+^ in Ag@ACB was shown as 4.595 mg L^−1^. The average leaching concentration of Ag^+^ in the Ag_3_PO_4_ reaction solution was shown as 8.122 mg L^−1^. This strongly suggested that the biochar-supported heterojunction structure helped to inhibit severe photocorrosion and excessive Ag^+^ leaching of Ag_3_PO_4_. CN/Bi/Ag@ACB achieved lower levels of Ag^+^ release while maintaining high sterilization activity. As shown in Fig. S10, comparison reveals that Ag_3_PO_4_ also exhibits bactericidal activity under dark conditions. This implies that the disinfection effect of Ag^+^ was dominant under dark reaction conditions. However, the disinfection effect of CN/Bi/Ag@ACB was slightly higher than that of Ag_3_PO_4_ in the light condition, which suggests that a faster or longer-lasting sterilization effect was achieved at a lower Ag content of CN/Bi/Ag@ACB. This indicates that the synergistic effect between ROS generation during photocatalytic reactions and Ag^+^ play a more critical role, indirectly confirming the significant contribution of the engineered heterojunction to enhancing disinfection efficacy. Admittedly, the dissolution of Ag^+^ made a significant contribution to the sterilization performance. However, CN/Bi/Ag@ACB enhanced the utilization efficiency of photogenerated carriers and ROS generation efficiency through the heterojunction, thereby achieving superior or more stable sterilization effects at the same Ag content.
Photogenerated carrier behavior
As depicted in Fig. 7a, the steady-state photoluminescence (PL) spectroscopy intensity follows the order: CN/Bi/Ag@ACB < Bi_2_WO_6_ < Ag_3_PO_4_ < g-C_3_N_4_ < ACB, indicating the composite’s unique structure significantly improves charge carrier separation^5^. ACB exhibits the highest PL intensity, likely due to surface defects in the biochar that impede electron mobility, increasing the likelihood of giant exciton formation through electron–hole recombination^17^. In Fig. 7b, CN/Bi/Ag@ACB demonstrates the highest photocurrent response among all samples. As the irradiation time extends, the photocurrent density of CN/Bi/Ag@ACB increases. Whereas the photocurrent densities of pure Ag_3_PO_4_ and Bi_2_WO_6_ slightly decrease, and g-C_3_N_4_ remains nearly constant. The superior charge carrier separation in CN/Bi/Ag@ACB, surpassing pure semiconductors, arises from synergistic heterojunction effects and enhanced charge transport properties. As shown in Fig. 7c, the Nyquist plot reveals that CN/Bi/Ag@ACB has a smaller radius compared to pure semiconductors, indicating lower resistance for photogenerated charge carriers. This reduction in resistance facilitates enhanced charge separation and rapid migration at the interface. Collectively, these findings demonstrate that CN/Bi/Ag@ACB achieves the highest carrier transfer and separation efficiency, with Ag_3_PO_4_, Bi_2_WO_6_, g-C_3_N_4_, and ACB working synergistically to boost photocatalytic activity.Fig. 7. Photoluminescence spectroscopy and electrochemical analysis.a Steady-state photoluminescence (PL) spectra, b transient photocurrent responses, c electrochemical impedance spectroscopy (EIS) Nyquist profiles, and d time-resolved photoluminescence (TRPL) spectra.
The photocarrier separation characteristics of the corresponding biochar-based binary composite materials compared with pure semiconductors can be referenced in our previously published papers^5,13,19^. To further investigate the key role of biochar and the heterojunction of CN/Bi/Ag@ACB, time-resolved photoluminescence (TRPL) spectroscopy was employed. The TRPL curves (Fig. 7d) and corresponding fitted data (Table S4 and Fig. S11) reveal that the fluorescence lifetimes of both CN/Bi/Ag@ACB and the ternary control materials conform to a double-exponential model. Comparative analysis revealed that CN/Bi/Ag@ACB exhibited the longest average fluorescence lifetime (τA) of 9.07 ns, significantly exceeding that of the CN/Bi/Ag composite without biochar (7.17 ns) and the biochar-based ternary control samples (7.48–7.90 ns). Short radiative lifetime (τ1) and long radiative lifetime (τ2) also follow similar trends. This directly evidences that the synergistic interaction between the multicomponent heterojunction constructed within CN/Bi/Ag@ACB and the biochar carrier most effectively suppresses the nonradiative recombination of photogenerated electron–hole pairs^30^. Consequently, it significantly enhances the separation and utilization efficiency of photoexcited carriers, providing crucial kinetic evidence for its superior photocatalytic performance.
Identification and analysis of active species
The sequence of TC removal efficiency by CN/Bi/Ag@ACB at 120 min, as presented in Fig. 8a, ranks the scavengers as follows: ammonium oxalate (C_2_H_8_N_2_O_4_, abbreviated as AO) yielded the lowest efficiency, followed by p-benzoquinone (C_6_H_4_O_2_, BQ), then tert-butyl alcohol (C_4_H_10_O, TBA), then manganese (II) acetate (C_4_H_6_MnO_4_, MA), with the highest efficiency observed in the absence of any scavenger. The significant reduction in removal rates with AO and BQ treatments suggests that the photocatalytic performance is primarily influenced by the following active species: h^+^ > **·**O_2_^−^ > **·**OH > e^−^ > no scavenger. This indicates that h^+^ and **·**O_2_^−^ are the dominant active species, while **·OH also plays a notable role, and e^−^ has minimal impact on the degradation process. Further verification was conducted using electron spin resonance (ESR) measurements. In the dark (Fig. 8b, c), no significant signals for DMPO-·OH (four-fold peaks) or DMPO-·O_2_^−^ (six-fold peaks) were observed. Following 10 min of exposure, distinct signals for DMPO-·OH and DMPO-·**O_2_^−^ emerged, demonstrating that CN/Bi/Ag@ACB generates **·**OH and **·O_2_^–^ radicals under visible-light irradiation. Notably, the signal intensities of DMPO-·OH and DMPO-·**O_2_^−^ were significantly enhanced in CN/Bi/Ag@ACB compared to the pure semiconductor, which can be attributed to the defective structure of biochar and the activation of EPFRs^5^. The high intensity of **·**O_2_^−^, produced by the reduction of O_2_ via a single electron, suggests that CN/Bi/Ag@ACB has an enhanced capacity for molecular oxygen activation^31^. Moreover, the non-radical generation of CN/Bi/Ag@ACB was also characterized, involving photogenerated holes and singlet oxygen (^1^O_2_). Figure 8d shows that photogenerated holes were captured by 1-hydroxy-3-carboxy-2,2,5,5-tetramethylpyrrolidine (CPH), resulting in a distinct CPH-h^+^ signal with a characteristic triple peak (1:1:1 ratio) across all photocatalysts^29^. This observation confirms the successful excitation of electrons from EVB to ECB of the photocatalysts, leading to the effective formation of photoinduced h^+^ and electron–hole pairs. These findings highlight the critical role of reactive species in the degradation of TC. However, the CPH-h^+^ signals were relatively weak in Bi_2_WO_6_ and g-C_3_N_4_, and the introduction of ACB did not significantly enhance h^+^ generation. Therefore, it can be inferred that h^+^ and **·**O_2_^−^ are primarily generated by the semiconductor components, while **·**OH is mainly contributed by the ACB composite. As shown in Fig. 8e, no ^1^O_2_ signal was detected during the dark reaction. After 10 min of exposure, both CN/Bi/Ag@ACB and CN/Bi/Ag produced ^1^O_2_ signals. Moreover, the singlet oxygen signal intensity generated by CN/Bi/Ag@ACB is significantly higher than that of CN/Bi/Ag. Furthermore, the previous research conducted an in-depth analysis of the ESR signals on the biochar surface^5^, confirming the presence of a lot of EPFRs on the ACB surface. These EPFRs gradually convert into **·**OH and **·**O_2_^−^ radicals, promoting the substantial generation of ROS. Furthermore, the **·**OH and **·**O_2_^−^ signals (Fig. S12) in composites with and without biochar addition were further investigated. It was found that the **·**OH and **·**O_2_^−^ signal intensities generated by CN/Bi/Ag@ACB were significantly higher than those of CN/Bi/Ag. This comparison demonstrates that the introduction of biochar not only promotes non-radical pathways (^1^O_2_) through its surface defects or EPFRs, but also enhances the separation of photogenerated carriers, thereby increasing radical yield. This elucidates its crucial synergistic role in both the radical and non-radical dual pathway.Fig. 8. Active species analysis and Mott–Schottky plot.a Active radical species trapping experiments for the photocatalytic degradation of TC under visible light irradiation, b–e electron spin resonance (ESR) spectra of radical adducts trapped by DMPO/CPH (•OH, •O_2_^−^, photogenerated hole, and singlet oxygen, respectively) in the dark and with the visible light irradiation of 10 min (b, c, d, e, respectively), and f Mott–Schottky plot of CN/Bi/Ag@ACB.
Synergistic mechanisms
Building on the previous analysis, the potential charge transfer pathways and photocatalytic degradation mechanisms of CN/Bi/Ag@ACB are illustrated in Fig. 9. Driven by solute-solvent interactions, TC diffuses to the CN/Bi/Ag@ACB surface and is subsequently adsorbed onto active sites via combined physical and chemical mechanisms, leveraging the ACB carrier’s large S_BET_, OFGs, surface charge and porosity, and π–π conjugation^16,32^. Heterogeneous photocatalysis then commences at the TC-CN/Bi/Ag@ACB interface, generating ROS responsible for the progressive degradation of TC^10^. With the ECB position of −0.53 eV, CN/Bi/Ag@ACB exhibits stronger reducing capability than the O_2_/·O_2_^−^ redox couple (−0.33 eV), thereby promoting ·O_2_^−^ formation in the photocatalytic system^33^. Conversely, the EVB position is +1.38 eV, which is lower than the OH^−^/·OH redox potential (+2.4 eV), hindering the formation of ·OH. This explains why ·O_2_^−^ emerges as the dominant active species. In the CN/Bi/Ag@ACB system, ·OH primarily originates from the EPFRs in ACB and may also derive from the EVB of Bi_2_WO_6_ and Ag_3_PO_4_^34^. As degradation progresses, the intermediates detach from the CN/Bi/Ag@ACB surface, freeing active sites to reabsorb TC and its degraded intermediates, thus continuing the photocatalytic process. The ROS generated during this process is highly oxidative, eventually mineralizing TC into H_2_O and CO_2_.Fig. 9. Schematic illustration of the possible synergistic mechanisms and the separation mode of the photogenerated carriers.
According to solid energy band theory and the synthesis sequence, the possible modes of photogenerated carrier separation in CN/Bi/Ag@ACB are depicted in Fig. 9. Specifically, holes (h^+^) in the EVB of Ag_3_PO_4_ (+2.61 eV) and Bi_2_WO_6_ (+3.01 eV) migrate to the EVB of g-C_3_N_4_ (+1.49 eV), while electrons (e^−^) in the ECB of g-C_3_N_4_ (−1.21 eV) transfer to the ECB of Ag_3_PO_4_ (+0.31 eV) and Bi_2_WO_6_ (+0.41 eV). However, in a typical type-II heterojunction, e^−^ would migrate to the ECB of Ag_3_PO_4_ and Bi_2_WO_6_, leading to reduction, while h^+^ would migrate to the EVB of g-C_3_N_4_ (+1.49 eV), where oxidation occurs. The calculated band positions suggest an inability of Ag_3_PO_4_/Bi_2_WO_6_ to generate ·O_2_^−^ and of g-C_3_N_4_ to produce ·OH. This is inconsistent with the ·O_2_^−^/·OH activity evidenced by radical trapping experiments. Such inconsistency was also observed by Ma et al.^10^, who suggest that a double Z-scheme eterojunction pathway can effectively resolve this issue. This mechanism not only spatially separates charge carriers but also retains a higher redox capacity^35^. Consequently, a double Z-scheme eterojunction is proposed for the CN/Bi/Ag@ACB system, allowing for the simultaneous generation of ·O_2_^−^ and ·OH. This configuration enhances the rapid separation of photogenerated electron–hole pairs, as supported by the electrochemical, PL, and TRPL spectra data shown in Fig. 7.
To further validate the energy band positions of these materials, Mott–Schottky (M–S) tests were conducted on CN/Bi/Ag@ACB and the control material to provide an in-depth analysis of the flat-band potential (Efb) and charge carrier type in the semiconductor material. As shown in Fig. S13, the curves for all pure semiconductor materials (g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_) exhibit a positive slope, confirming that these three semiconductors are all n-type semiconductors. Fitting yields the Efb of −1.00, 0.52, and 0.50 eV for g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_, respectively. Typically, for n-type semiconductors, the Efb is approximately 0.1–0.3 eV below the ECB (with a smaller absolute value)^36^. This aligns with the results calculated in the Optical properties section: −1.21, 0.41, and 0.31 eV for g-C_3_N_4_, Bi_2_WO_6_, and Ag_3_PO_4_, respectively. Based on the band calculation Eq. (1), this also indirectly confirms that the valence band positions of these three semiconductors are reasonable. However, in this study, the M–S curves of many ternary composites exhibited a significant positive shift, showing both positive and negative slopes, indicating the presence of n–p heterojunctions (Figs. 8f and S13). This may be related to the introduction of ACB and the synthesis process, suggesting the formation of effective interfacial charge redistribution and band bending within the composites^36^. As shown in Fig. 8f, CN/Bi/Ag@ACB exhibited an n–p heterojunction with n-type and p-type Efb values of −0.29 and 1.15 eV, respectively. Typically, for p-type semiconductors, Efb is located approximately 0.1–0.3 eV below the valence band (with a smaller absolute value). This aligns with the positions of the ECB (−0.53 eV) and EVB (1.38 eV) for CN/Bi/Ag@ACB analyzed in the optical properties section. Furthermore, the slope of the M–S curve is proportional to the photogenerated carrier concentration of the material. Comparing Figs. 8f and S13g reveals that CN/Bi/Ag@ACB exhibited a steeper slope relative to CN/Bi/Ag, suggesting that the introduction of biochar enhances the concentration of photo-generated carriers. This finding is consistent with the analysis of TRPL results (Fig. 7d and Table S4). Therefore, the CN/Bi/Ag@ACB photogenerated carrier pathway is a typical double Z-scheme heterojunction. To further confirm this at the microscopic level, photodeposition experiments were conducted using Ag ions (representing electrons) and Pb ions (representing holes). As shown in Fig. S14a, the Ag, C, and N elemental mapping images overlap, indicating that Ag ions primarily deposited on the g-C_3_N_4_ surface. This suggests that photogenerated electrons are predominantly enriched on g-C_3_N_4_. Similarly, the mapping images of Pb, Bi, W, O, Ag, P, and O elements show high overlap (Fig. S14b), indicating that Pb ions primarily deposited on the Bi_2_WO_6_ or Ag_3_PO_4_ surfaces. This suggests that photogenerated holes are predominantly enriched on Bi_2_WO_6_ or Ag_3_PO_4_. Additionally, Ag and Pb are also distributed on the ACB, visually demonstrating charge transfer between multiple components via the ACB, which strongly supports the double Z-scheme heterojunction. The above discussion further validates this double Z-scheme heterojunction. This pathway spatially separates the reduction and oxidation centers, not only suppressing electron–hole recombination but also preserving a strong redox potential. This provides a basis for the synergistic generation of ROS, thereby significantly enhancing photocatalytic performance.
The synergistic mechanisms behind the enhanced photocatalysis of CN/Bi/Ag@ACB likely involve several key contributions. (1) Special structure and surface properties of ACB: within the double Z-scheme model, ACB’s excellent electrical conductivity reduces the overall resistance of CN/Bi/Ag@ACB and facilitates the transfer of photogenerated electrons^35^. ACB can function as both an electron acceptor and donor, enabling efficient electron transfer to the surfaces of Ag_3_PO_4_, g-C_3_N_4_, and Bi_2_WO_6_. Additionally, excess electrons from the pure semiconductors can be transferred to ACB, further enhancing the photogenerated carriers’ separation^31^. The graphitized, defective structure of ACB aids in capturing photogenerated electrons, preventing their recombination with holes. The π–π interaction between biochar and g-C_3_N_4_ mediates electron delocalization, thereby facilitating electron transport along this pathway^5^. This improves the separation of electron–hole pairs and redistributes charge density at the interface, leading to the formation of ·O_2_^−^ through reactions with O_2_, thereby enhancing photocatalytic activity^19^. Additionally, oxygen vacancies and EPFRs in biochar may further contribute to the generation of ROS under irradiation^16^. The introduction of ACB also optimizes the band structure and increases visible light absorption (Fig. 5a, d), resulting in more charge carriers being generated upon excitation, which boosts photocatalytic activity. Furthermore, the porous structure and surface rich in oxygen-containing functional groups of biochar give it high specific surface area adsorption sites^13^, which facilitates the effective mass transfer of TC and solvent oxygen in the adsorption stage and photocatalytic process. Additionally, the composite of biochar into the ternary semiconductor system can optimize the energy band structure and form a double Z-scheme heterojunction, which not only enhances the visible light absorption but also effectively suppresses the photogenerated carrier separation efficiency (Fig. 7)^19^. Moreover, biochar may act as a co-catalyst to enhance the reaction process of photocatalysis^5,16^. Whether biochar is a semiconductor material is not conclusive, which is closely related to its preparation process and raw materials. In our opinion, biochar is a complex mixture of carbon systems, in which the percentages of aromatic hybridized carbon, graphitized carbon, amorphous carbon, functional groups, defective structures, etc., may affect its properties to become a co-catalyst or a semiconductor^34^. These remain to be further studied in depth. It is predictable that an increase in the percentage of graphitized carbon in the biochar system may significantly enhance the photogenerated electron transfer properties as the pyrolysis temperature of the biomass increases. Finally, biochar has a photothermal effect^37^, with strong light absorption (infrared light) causing a localized temperature rise and an increase in the rate of the promoted photocatalytic reaction. This is attributed to the black appearance of the biochar and the biomass-derived micro-configuration that contributes to the absorption and scattering of light, thereby enhancing light capture efficiency.
(2) Heteroatoms in the composite: the presence of heteroatoms enhances photocatalytic activity through both radical and nonradical mechanisms. Beyond the previously mentioned radical-enhancing effects^38^, nonradical pathways also play a significant role, as supported by recent literature^39,40^. Huang et al.^41^ suggested that the tightly coupled three-dimensional structure connected by Ag facilitates charge transfer between g-C_3_N_4_ and Bi_2_WO_6_, promoting photogenerated carrier separation and enhancing photocatalytic efficiency. Miao et al.^31^ further proposed that certain heteroatoms, such as N and P, might act as charge transport bridges between the semiconductors, accelerating the recombination of electrons in the ECB of Ag_3_PO_4_ with holes in the EVB of g-C_3_N_4_. Studies by Chong et al.^42^ and Dou et al.^9^ have reported that Bi_2_WO_6_ and g-C_3_N_4_ composites can form Z-scheme and S-scheme charge separation modes. While the presence of Ag in this study likely accelerated the spatial separation of photogenerated electrons, no pure Ag was detected in CN/Bi/Ag@ACB, suggesting that this phenomenon warrants further investigation. Moreover, the photocatalytic activity was significantly enhanced by both non-radical processes (generation of singlet oxygen and photogenerated holes, Fig. 8)^38,39^ and heteroatom doping (N and P elements, surface plasmon resonance effect of Ag)^41^. Therefore, the unique properties of biochar synergize the double Z-scheme heterojunction to achieve more efficient translocation and inhibit recombination of photogenerated carriers. The above mechanism analysis is further supported by the characterization data (Figs. 5, 7 and 8), experimental data (Figs. 6 and S5–S7), and actual wastewater treatment data (Figs S7c and S9d), which demonstrates that composite biochar can enhance photovoltage density, reduce charge transfer impedance, prolong photogenerated carrier lifetime, and strengthen key radical signals. In summary, the role of biochar has evolved from a “simple carrier or adsorbent” to a “key functional component that participates in and regulates charge transport within the double Z scheme model.”
Degradation process of tetracycline
The 3D excitation-emission matrix (EEM) spectra for residual TC solutions sampled at different time points are displayed in Fig. S15. Initially, the 3D EEM signal of the TC solution (Fig. S15a) is notably weak, consistent with previous studies^43^, which have demonstrated that TC exhibits fluorescence quenching, resulting in low fluorescence efficiency. This quenching effect is primarily attributed to the presence of electron-withdrawing groups (e.g., carbonyl groups) within the TC molecule, which significantly diminishes its fluorescence efficiency. As the photocatalytic reaction progresses (Fig. S15a–e), the fluorescence center gradually shifts towards the blue along both the emission and excitation axes. This blue shift likely reflects the disruption of condensed aromatic systems and concomitant scission of macromolecules into lower-molecular-weight fragments^43^. After 10 min of visible light exposure, fluorescence spectra reveal a dominant peak at Ex/Em = 200–250/34–420 nm and a minor peak in the region of Ex/Em = 270–300/350–375 nm (Fig. S15b), corresponding to the fluorescence regions of fulvic acid and humic acid, respectively. As the irradiation time increases to 30 and 60 min (Fig. S15c, d), the shape and intensity of these characteristic fluorescence peaks become more pronounced, with a noticeable blue shift. This shift is likely due to the partial decomposition of macromolecules into smaller fragments and the breakdown of condensed aromatic compounds, leading to a reduction in the number of aromatic rings and a decrease in the extent of the π-electron system^44^. After 90 min of photocatalysis, the intensities of these fluorescence peaks decrease significantly (Fig. S15e). The strong fluorescence peak intensity drops from above 10,000 to below 5000, and the weak fluorescence signal nearly vanishes, indicating that the humic acid and fulvic acid have been degraded. These intermediates may be converted into each other and could also be fully mineralized, releasing CO_2_ and H_2_O. This observation aligns with the TOC results, confirming that CN/Bi/Ag@ACB exhibits excellent mineralization capability in the degradation of TC.
Degradation pathways of tetracycline
High-performance liquid chromatography-mass spectrometry (HPLC-MS) was employed to identify possible intermediates in the photocatalytic degradation of TC. Figure S16 provides the detailed mass spectra of these intermediates, and Table S5 lists the resulting degradation products. Figure 10 presents postulated TC degradation pathways derived from the experimental data and 3D EEM analysis. Following dehydration in pathway I, TC converts to P1, identified at m/z = 427. P1 is then photochemically oxidized, resulting in the removal of a methyl group to form P2 (m/z = 413). Following this, the N-methyl group is removed, leading to deamination and cleavage of the carbon ring to yield P3 (m/z = 346). Finally, P3 loses a carbonyl group to produce P4 (m/z = 317). In pathway II, TC is decomposed into P5 (m/z = 477) through multiple hydroxylation steps facilitated by ·OH oxidation. P5 subsequently undergoes photocatalytic oxidation, leading to the loss of H_2_O and the formation of P6 (m/z = 459). P6 then undergoes deamidation, producing P7 (m/z = 400), which is further deaminated and methylated to yield P8 (m/z = 343). Pathway III begins with the destruction of multiple methyl groups by ROS oxidation, producing P9 (m/z = 403). P9 is further degraded into P10 (m/z = 337) through a series of reactions, including deamidation, deamination, hydroxylation, carbonylation, addition reactions, and ring-opening. Due to the instability of the enol structure in P10, it converts into the more stable ketone form, producing P11 (m/z = 337), an isomer of P10. P12 (m/z = 297) is then formed when P11 loses CH_2_O and C_2_H_4_O groups. As the TC structure continues to degrade, the reaction mixture becomes increasingly complex due to the formation of small molecular intermediates, primarily resulting from ring-opening and cleavage of the central carbon structure. These intermediates include compounds with m/z values of 249, 247, 228, 201, 218, 195, and 106. Under the continuous action of ROS generated in the CN/Bi/Ag@ACB system, these small organic molecules are gradually mineralized into CO_2_ and H_2_O, completing the degradation process of TC. The TOC analysis further supports that a significant portion of organic carbon is converted into inorganic carbon through photocatalytic oxidation, highlighting the efficacy of CN/Bi/Ag@ACB in fully degrading TC.Fig. 10. The possible photocatalytic degradation pathways of TC by CN/Bi/Ag@ACB.
Conclusions
In summary, this study successfully developed a novel quaternary composite, CN/Bi/Ag@ACB, featuring a double Z-scheme heterojunction. The degradation pathway and efficacy of CN/Bi/Ag@ACB on TC were investigated, along with its potential environmental applications, to elucidate the synergistic enhancements in its photocatalytic mechanism. Characterization results revealed that CN/Bi/Ag@ACB possesses microporous and mesoporous structures, a rich elemental composition, OFGs, and defect structures. The composite exhibits strong absorbance across the 200–800 nm range, with an average absorbance of 60%. With the conduction band at −0.53 eV and valence band at 1.38 eV, CN/Bi/Ag@ACB exhibits suppressed photogenerated carrier recombination. Near-complete elimination of 50 mg L^−1^ TC was attained by CN/Bi/Ag@ACB after 120 min irradiation, displaying a 0.0351 min^−1^ rate constant that exceeds those of Ag_3_PO_4_ (8.56 times), Bi_2_WO_6_ (11.70 times), and g-C_3_N_4_ (13.50 times). The TOC removal rate reached 67.74%, which is 2.3–3.4 times that of the pure semiconductors. Additionally, the composite demonstrated a sterilization rate of approximately 99% for E. coli and S. aureus within 48 h. CN/Bi/Ag@ACB exhibited strong anti-interference capabilities and excellent reusability, with only a 3.88% decrease in TC removal efficiency after five cycles, while maintaining structural integrity. The removal rate of TC in wastewater exceeded 85.95%, and the composite was also effective in degrading other antibiotics. Radical species, including ·O_2_^−^, h^+^, and ·OH radicals, constitute the primary photocatalytic mechanism. Furthermore, the photocatalytic enhancement is attributed to ACB’s distinctive structural and surface characteristics coupled with heteroatom incorporation in the composite. The unique properties of biochar synergize with the double Z-scheme heterojunction to enable more efficient migration of photo-generated carriers in CN/Bi/Ag@ACB, thereby enhancing radical yield. This study elucidates the key enhancing role of biochar in both radical and non-radical pathways. The degradation process of TC is linked to the production, increase, and subsequent reduction of the fluorescence regions associated with fulvic acid and humic acid. Three possible degradation pathways were identified, involving dehydration, demethylation, deamination, hydroxylation, deamidation, carbonylation, addition reactions, ring-opening, and central carbon cleavage of the TC molecule. These findings deliver dual contributions: guiding innovative double Z-scheme photocatalyst design, and advancing theoretical frameworks for synergy-driven mechanisms in multi-component semiconductors and biochar systems.
Methods
Chemical reagents
Tetracycline, Bi(NO_3_)3·5H_2_O, Na_2_WO_4_·2H_2_O, AgNO_3_, Na_2_HPO_4_, KOH, HNO_3_, melamine (C_3_H_6_N_6_), norfloxacin (C_16_H_18_FN_3_O_3_, NOR), chloramphenicol (C_11_H_12_Cl_2_N_2_O_5_, CAP), and other common reagents were procured from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). These reagents require no further purification, and the water used is all double-distilled water.
Preparation of quaternary composite photocatalyst
The activated biochar (ACB) was prepared through pyrolysis (650 °C for 3 h, N_2_ atmosphere) of Caragana korshinskii biomass sourced from Guyuan City, PR China, followed by ultrasonic and KOH activation treatment. The g-C_3_N_4_ was prepared through thermal polycondensation of melamine, with these specific parameters detailed in the Supplementary Methods.
This is for the preparation of CN/Bi/Ag@ACB. ACB (1.0 g) and g-C_3_N_4_ (0.55 g) underwent ultrasonic treatment for 24 h, followed by 1 h of ball milling, and then the resultant composite was placed in beaker A. In beaker B, 1.50 g of Bi(NO_3_)3·5H_2_O was dissolved in 50 mL of distilled water containing 5 mL of HNO_3_, and the solution was ultrasonicated for 20 min. Following this, 0.5 g of Na_2_WO_4_·2H_2_O was added to beaker B. The contents of beakers A and B were then combined, ultrasonicated for 1 h, adjusted to a neutral pH, and stirred for 10 min. The resulting mixture was placed in a Teflon-lined autoclave and subjected to hydrothermal reaction at 200 °C for 24 h. After the reaction, the mixture was filtered using a 0.45 μm membrane to obtain a moist solid powder. Next, AgNO_3_ solution (0.2 M, 18 mL) was added to the solid powder, followed by 1 h of ultrasonication and 1 h of stirring. Subsequently, Na_2_HPO_4_ solution (0.2 M, 7 mL) was added dropwise in a dark environment, and the system was ultrasonicated for another hour and stirred for 1 h. Finally, the was collected by filtration, dried at 80 °C for 12 h, and sieved through a 0.15 mm mesh to obtain CN/Bi/Ag@ACB. The corresponding control materials were also prepared.
Batch tests and material properties analysis
The performance of CN/Bi/Ag@ACB in removing TC was systematically assessed via standard adsorption–photocatalytic degradation experiments, including both model aqueous systems and real wastewater samples. The disinfection efficacy of the composites was evaluated using the spread plate method. All experiments were conducted with three parallel tests and corresponding control setups. Three-dimensional EEM spectroscopy and HPLC-MS were used to explore TC degradation processes and pathways. These classical experimental methods or characterization techniques have been widely used and reported, and are therefore simplified here. Comprehensive best testing steps, all material characterizations, electrochemical tests, and analytical methods are summarized in the Supplementary Methods.
Supplementary information
Supplementary Information
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Djurišić, A. B., He, Y. & Ng, A. Visible-light photocatalysts: prospects and challenges. APL Mater.8, 030903 (2020).