Boosting piezo-photocatalytic performance in antibiotic wastewater treatment through graphitic carbon nitride–zinc oxide based heterojunction engineering

Abstract

Semiconductor photocatalytic technology is confronted with the key challenge of rapid recombination of photogenerated carriers. This work successfully fabricates a flexible g-C₃N₄–ZnO@PVDF composite film via a facile phase inversion method. Its principal innovation lies in the effective coupling of the material's inherent photovoltaic properties with the piezoelectric effect derived from PVDF under mild mechanical agitation (aeration). This synergy establishes a powerful built-in electric field that dramatically enhances the separation and migration of photogenerated charge carriers. The film exhibits superior performance in the degradation of the antibiotic Tetracycline (TC, 20 mg L⁻¹), achieving over 99% removal under optimal conditions. The built-in piezoelectric field not only boosts photocatalytic efficiency but also enables a self-powered purification capability. The system demonstrates excellent stability and reusability, presenting a promising and sustainable strategy for advanced wastewater treatment.

---

1. Introduction

In recent years, antibiotic usage has escalated significantly, leading to an ever-growing fraction of antibiotic-containing wastewater in overall wastewater streams.¹ The major sources of antibiotic contamination in water are wastewater from hospitals,² pharmaceutical manufacturing,³ animal husbandry,⁴ aquaculture, and related sectors.⁵ They are primarily utilized for pest and disease control, preventing infectious diseases, and protecting human health.⁶ Conventional techniques like adsorption, anion exchange, and membrane filtration merely transfer TC without degrading it, inevitably posing risks of secondary pollution.⁷ Moreover, the treatment of antibiotic-laden wastewater faces significant challenges because of the numerous inhibitory substances it contains.^8–12

Photocatalysis belongs to the family of advanced oxidation processes and can harvest solar radiation to generate chemical energy with high efficiency, thereby positioning it as a compelling strategy for environmental restoration.^13,14 Because the reactions proceed under mild conditions, impose minimal ecological burden, and enable complete mineralisation of organic contaminants, this technique has attracted widespread scholarly interest.^15,16

ZnO′ characteristics are low cost, non-toxicity and excellent physicochemical robustness. Yet its photoactivity is chiefly confined to the ultraviolet (UV) region, only 4% of the solar spectrum, severely restricting practical deployment.¹⁷ As a representative study, Helapiyumi Weerathunga et al.¹⁸ prepared ZnO nanoparticles with rod-, plate- and cone-like architectures via a non-hydrolytic route, and every morphology exhibited pronounced photocatalytic performance.

g-C₃N₄ is a low-cost, non-toxic organic semiconductor that can be readily synthesized from abundant precursors; it possesses a suitably sized band gap and remarkable chemical stability.^19–21 However, bulk g-C₃N₄ has a relatively small specific surface area, which limits its photocatalytic performance. This shortcoming can be effectively mitigated by tailoring g-C₃N₄'s morphology and structure to expose more active sites, thereby enhancing its photocatalytic efficiency.²² Zan et al. successfully prepared CNQDs through acid etching ultrasound and formed a heterojunction with BiOBr, achieving efficient degradation of tetracycline and ciprofloxacin as well as H₂O₂ production.²³

Additionally, doping g-C₃N₄ with metallic or non-metallic elements has been explored to adjust its light absorption and utilization, as well as to modify its physicochemical properties.²⁴ Nevertheless, pristine g-C₃N₄ exhibits only a modest quantum efficiency due to the rapid recombination of its photogenerated electron–hole (e⁻/h⁺) pairs.²⁵ To overcome these limitations, g-C₃N₄ is often coupled with a narrow-bandgap semiconductor to extend its light absorption range and facilitate charge separation, thereby enhancing photocatalytic performance.²⁶ Forming a heterojunction between g-C₃N₄ and another semiconductor can further promote interfacial charge transfer and suppress recombination, resulting in higher photocatalytic efficiency.²⁷ For example, Li et al.²⁸ developed a pg-C₃N₄/β-FeOOH S-scheme heterostructure with a special band structure by anchoring porous pg-C₃N₄ on needle-like β-FeOOH, and achieved the simultaneous degradation of organic pollutants through photoelectrocatalytic hydrogen evolution. Polyvinylidene fluoride (PVDF) is a semi-crystalline polymer. PVDF can crystallize into four polymorphic phases (α, β, γ, δ), among which the all-trans β-phase is primarily responsible for PVDF's piezoelectric properties.^29–33 PVDF films in which the β-phase is dominant also exhibit greater fouling resistance than those rich in the α-phase.³⁴

In light of the above, the present study proposes a novel g-C₃N₄–ZnO@PVDF heterojunction film fabricated by phase inversion, which uniquely integrates piezoelectric and photocatalytic functionalities. The system leverages the complementary properties of its components: the piezoresponse of PVDF generates a built-in electric field that directs charge migration and curbs recombination, while the g-C₃N₄/ZnO heterojunction establishes an efficient electron-transfer pathway. This synergy collectively enhances charge-carrier separation and prolongs the lifespan of reactive species, thereby maximizing the overall redox potential.³⁵ Through systematic optimization of parameters such as the composite ratio, catalyst loading, and film thickness, this strategy achieves high-efficiency removal of antibiotic contaminants under combined piezo-photocatalytic conditions. Furthermore, theoretical simulations offer deeper insight into interfacial charge redistribution under mechanical stress, thereby paving the way for scalable development of multifunctional catalytic films for sustainable wastewater treatment. For specific experimental details, please refer to the SI.

2. Results and discussion

2.1 Morphology and composition characterization

Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analyses were performed to observe the sample morphologies. As shown in Fig. 1b, the ZnO sample exhibits a rod-like morphology, whereas Fig. 1c shows g-C₃N₄ appearing as thin flakes. Fig. 1d shows that the pure PVDF film has a porous network structure. Incorporating g-C₃N₄–ZnO particles into PVDF (Fig. 1e) yields a g-C₃N₄–ZnO@PVDF composite film that exhibits enhanced piezoelectric properties and improved pollutant removal efficiency. This composite structure ensures intimate contact between catalyst particles and the surrounding medium, thereby increasing the number of active sites accessible for photocatalytic reactions.

Fig. 1 (a) Schematic illustration of the synthesis process for g-C₃N₄–ZnO@PVDF composite film. SEM images of (b) ZnO nanorods, (c) g-C₃N₄ sheets, (d) porous PVDF film, and (e) g-C₃N₄–ZnO@PVDF composite film. (f) TEM image and (g) HRTEM image of the g-C₃N₄–ZnO@PVDF composite.

SEM elemental mapping (Fig. S1) confirms the presence of C, N, Zn, and O uniformly distributed in the composite film. Fig. 1f presents a TEM image of a g-C₃N₄–ZnO composite, clearly showing ZnO nanorods growing on g-C₃N₄ sheets. High-resolution TEM (Fig. 1g) reveals two distinct lattice fringes: one with d = 0.262 nm corresponding to the (002) plane of ZnO, and another with d = 0.284 nm corresponding to the (100) plane of hexagonal wurtzite ZnO. No distinct lattice fringe from g-C₃N₄ is observed, likely due to the inherently weaker scattering of low-Z elements (C and N).³⁶ The resulting ZnO/g-C₃N₄ heterostructure exhibits enhanced interfacial charge transfer kinetics, as indicated by the well-defined interface, consequently curtailing the pathways through which photogenerated charge carriers can recombine.

X-ray diffraction (XRD) was employed to further determine the crystalline phases present in the g-C₃N₄–ZnO@PVDF sample. As depicted in Fig. 2a, diffraction patterns for pristine ZnO, pristine g-C₃N₄ and their composite were collected in the 2θ range of 5–80° using a 0.5° divergence slit. For ZnO, the diffractogram verifies a hexagonal zincite phase, featuring strong reflections at approximately 31.6°, 34.3°, 36.2°, 47.5°, 56.7°, 62.8°, 68.0°, 69.1°, 72.6° and 77.0° (2θ), which can be indexed to the (100), (002), (101), (102), (110), (103), (112) and (201) planes of hexagonal ZnO.³⁷ All of these reflections coincide with those listed in JCPDS card no. 03-0888 for ZnO, and the absence of extraneous reflections indicates that the as-prepared ZnO is phase-pure. For g-C₃N₄, the signature peaks appear at roughly 12.8° and 27.6° (2θ), corresponding to the (100) and (002) planes,³⁸ respectively, in agreement with JCPDS no. 87-1526. Within the diffraction profile of the g-C₃N₄–ZnO composite, the ZnO reflections dominate and display almost no positional shift, suggesting that embedding g-C₃N₄ leaves the ZnO lattice unaffected and that the composite forms without compromising the crystallographic integrity of either phase.

Fig. 2 (a) XRD patterns of ZnO, g-C₃N₄, and g-C₃N₄–ZnO (3 : 1) composite. High-resolution XPS spectra of (b) Zn 2p, (c) O 1s, and (d) N 1s for the samples.

The FTIR spectra of ZnO, g-C₃N₄ and their composite are presented in Fig. S2a. In the case of pristine g-C₃N₄, intense absorptions between 1200 and 1650 cm⁻¹ arise from C–N stretching vibrations within the heterocyclic network. Specifically, the 1624 cm⁻¹ peak is ascribed to C=N stretching, whereas the 1427 cm⁻¹ band originates from aromatic C–N stretching. Furthermore, the feature at 820 cm⁻¹ stems from out-of-plane bending of tri-s-triazine rings, and the 623 cm⁻¹ band reflects C–N bending vibrations. Both ZnO and the g-C₃N₄–ZnO composite exhibit a 490 cm⁻¹ band corresponding to Zn–O stretching, evidencing ZnO incorporation without appreciable structural disturbance.

X-ray photoelectron spectroscopy (XPS) was subsequently employed to probe the chemical composition and valence states of the photocatalysts. A comparison of survey XPS spectra for ZnO, g-C₃N₄ and the g-C₃N₄–ZnO (3 : 1) composite is provided in Fig. S2b. Pristine ZnO displays only Zn and O signals, whereas g-C₃N₄ exhibits peaks attributable to C and N; all four elements—Zn, O, C and N—are detected in the composite, corroborating that both components are retained. High-resolution C 1s spectra of g-C₃N₄ (Fig. S2c) display three features centred at 284.8, 286.6 and 288.4 eV, which can be assigned to C–C, C–N and sp² N–C=N bonds, respectively.^39,40 For the g-C₃N₄–ZnO (3 : 1) composite, these C 1s peaks move toward lower binding energies relative to those of pure g-C₃N₄, implying electron donation from ZnO to the electron-deficient triazine rings and hence an increased electron density on g-C₃N₄. The Zn 2p spectra are illustrated in Fig. 2b; two peaks located at 1021.9 eV and 1045.0 eV are attributed to Zn 2p_3/2 and Zn 2p_1/2, respectively.⁴¹ Relative to pristine ZnO, the Zn 2p signals of the composite are shifted toward higher binding energies, which points to the generation of Zn–N bonds and attendant electron transfer to nitrogen. The O 1s spectrum (Fig. 2c) can be deconvoluted into two contributions at approximately 530.0 eV and 531.5 eV, assigned to lattice O and surface –OH groups, respectively.^42,43 As illustrated in Fig. 2d, the N 1s spectrum exhibits peaks at 398.6, 400.2, 401.3 and 404.2 eV, corresponding to sp² N in C=N–C, tertiary N in N–(C)₃, amine N in C–N–H and π-excitation, respectively.⁴⁴ Taken together, the binding energies of the composite are displaced with respect to those of the single components, shifts that are attributable to pronounced interfacial interactions and to the heterojunction established between ZnO and g-C₃N₄.

2.2 Piezoelectric characterization

This study employed piezoelectric force microscopy (PFM) to characterize both the piezoelectric properties and local micro-morphology of the g-C₃N₄–ZnO@PVDF film. By applying an external voltage to the sample surface, the resulting vibrational amplitude and phase shifts were measured to characterize the piezoelectric response. The magnitude of the amplitude response directly reflects the strength of the material's piezoelectric performance. In the case of measuring the vertical response only, the amplitude distribution reveals distinct contrast (Fig. 3a). White-highlighted regions correspond to piezoelectric activity, whose ubiquitous presence demonstrates the film's inherent capacity for piezoelectric response across its entirety. Besides, the phase images are described in Fig. 3b, the presence of 180° domains indicates that the sample has excellent piezoelectricity.

Fig. 3 PFM tests of prepared g-C₃N₄–ZnO@PVDF: (a) amplitude image, (b) phase image, (c) butterfly loops and (d) piezoelectric hysteresis.

To gain a more comprehensive understanding of the piezoelectric properties of the prepared films, the piezoelectric hysteresis loops and butterfly curves of g-C₃N₄–ZnO@PVDF film was characterized using PFM. As illustrated in Fig. 3c and d, a representative set of standard amplitude butterfly curves and phase hysteresis loops is observed upon reversing the applied AC voltage from −12 V to 12 V. The g-C₃N₄–ZnO@PVDF exhibits notable piezoelectric properties and demonstrates a distinct 180 degree phase reversal under vertically applied polarization conditions.

2.3 Photochemical and electrochemical characterization

To evaluate the electrochemical performance, cyclic voltammetry was conducted under ambient light for ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF electrodes (in a three-electrode setup with a Hg/HgO reference, Pt counter, and 0.1 M K₃Fe(CN)₆ electrolyte). Photocurrent measurements and EIS were also used to indirectly evaluate charge separation and transfer efficiency during photocatalysis.

Fig. 4a demonstrates that the g-C₃N₄–ZnO@PVDF electrode delivers a markedly larger photocurrent than pristine g-C₃N₄, ZnO, or the binary g-C₃N₄–ZnO composite, reflecting more effective photogenerated carrier separation and accelerated interfacial charge migration within the ternary architecture. These improved charge-transport characteristics indicate that g-C₃N₄–ZnO@PVDF is poised to exhibit enhanced photocatalytic performance.

Fig. 4 (a) Transient photocurrent response curves for ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. (b) Cyclic voltammograms of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. (c) Nyquist plots (EIS) of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. (d) Steady-state PL spectra of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. (e) UV-Vis DRS absorption spectra of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. (f) Tauc plots for band gap determination of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. UPS spectra of (g) ZnO and (h) g-C₃N₄ (He I 21.22 eV).

Fig. 4b shows that all samples (ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF) display both oxidation and reduction peaks in their CV curves, reflecting good reversible redox behavior.⁴⁵ Notably, the g-C₃N₄–ZnO@PVDF electrode delivers the highest current response among them, suggesting the greatest electrochemical activity.

In addition, electrochemical impedance spectroscopy (EIS) was employed to compare the charge transfer resistance of ZnO, g-C₃N₄, g-C₃N₄–ZnO, and g-C₃N₄–ZnO@PVDF. Fig. 4c shows that the diameter of the Nyquist semicircle for the g-C₃N₄–ZnO@PVDF electrode is much smaller than those for ZnO, g-C₃N₄, or g-C₃N₄–ZnO. A smaller semicircle (lower charge-transfer resistance) indicates more efficient charge separation (higher photoelectric conversion efficiency) and a lower electron–hole recombination rate.⁴⁶ These results demonstrate that the g-C₃N₄–ZnO@PVDF heterojunction accelerates charge transfer processes, thereby improving charge separation efficiency.

Steady-state photoluminescence (PL) spectra were acquired for ZnO, g-C₃N₄, g-C₃N₄/ZnO and the g-C₃N₄–ZnO@PVDF hybrid (Fig. 4d) to probe the recombination behaviour of photo-induced carriers. ZnO shows only faint emissions between 400 and 460 nm, attributable to excitonic recombination linked to surface oxygen vacancies and other structural defects within the nanorods. In stark contrast, pristine g-C₃N₄ presents a pronounced PL band centred at 430 nm. When ZnO is integrated with g-C₃N₄, the peak position remains essentially constant, yet the overall PL intensity diminishes sharply. This pronounced quenching reflects a lower rate of radiative e⁻/h⁺ recombination in the g-C₃N₄–ZnO composite, accounting for its superior photocatalytic performance.

The optical absorption properties and band gaps of ZnO, g-C₃N₄, and g-C₃N₄–ZnO were evaluated using UV-Vis diffuse reflectance spectroscopy. According to the DRS spectra (Fig. 4e), ZnO nanorods absorb strongly only in the UV region (up to 360 nm), where as g-C₃N₄ extends its absorption into the visible range (up to 390 nm). For band gap estimation, the Kubelka–Munk function was applied (eqn (1)):⁴⁷

αhν = A(hν − E_g)^n/2 (1)

where α, h, ν, E_g, and A represent the absorption coefficient, Planck's constant, photon frequency, band gap energy, and a proportionality constant, respectively. Here n is determined by the nature of the electronic transition (n = 1 for direct band gap semiconductors and n = 4 for indirect ones). Since both ZnO⁴⁸ and g-C₃N₄ (ref. 49) have direct band gaps, n = 1 was used. Using Tauc plots (Fig. 4f), the band gap energies were determined to be approximately 3.13 eV for ZnO, 2.72 eV for g-C₃N₄, and 2.95 eV for the g-C₃N₄–ZnO composite. The slightly widened band gap of the composite (relative to g-C₃N₄) suggests successful integration of the two semiconductors.

Ultraviolet photoelectron spectroscopy (UPS, He I, 21.22 eV) was employed to pinpoint the valence band (VB) edges of the samples. From the secondary-electron cut-offs in the UPS curves (Fig. 4g–h), E_cutoff values of 20.58 eV for ZnO and 20.01 eV for g-C₃N₄ were obtained. According to Φ = 21.22 eV − E_cutoff,⁵⁰ the corresponding work functions are 3.57 eV (ZnO) and 4.54 eV (g-C₃N₄). These give valence band maximum (VBM, referenced to vacuum) of −7.67 eV for ZnO and −6.22 eV for g-C₃N₄.

Using the relation E_vac = −E_RHE − 4.44 eV⁵¹ and the conversion E_NHE = E_RHE − 0.059 pH,⁵² the VB positions at pH 7 translate to +2.82 V (ZnO) and +1.36 V (g-C₃N₄) versus NHE. Considering optical band gaps of 3.13 eV for ZnO and 2.72 eV for g-C₃N₄, the conduction band minima (CBM) are therefore located at approximately −0.30 V and −1.36 V vs. NHE, respectively.⁵³

We conducted BET (Brunauer–Emmett–Teller) tests on ZnO and g-C₃N₄–ZnO and obtained their N₂ adsorption/desorption isotherms and pore size distributions (Fig. S3). It is worth noting that the curves of two samples displayed type IV isotherms and possessed a classical H3 hysteresis loop which can strongly prove the existence of microporous structure. The BET surface areas of ZnO and g-C₃N₄–ZnO were 10.16 m² g⁻¹ and 70.43 m² g⁻¹, respectively. The g-C₃N₄–ZnO composite material has a larger specific surface area and more pores to absorb pollutants for photocatalytic degradation, and rapid ion diffusion that promotes mass transfer efficiency.

2.4 Evaluation of photocatalytic activity

A series of experiments were conducted to systematically investigate the effects of various factors (including the g-C₃N₄ : ZnO composite ratio, light vs. dark conditions, aeration, catalyst dosage, pollutant concentration, and film thickness) on the performance of g-C₃N₄–ZnO@PVDF. The goal was to determine the most favorable conditions for TC removal. Notably, all the materials are porous and exhibit adsorptive properties. To isolate the effect of adsorption, the experiments included a dark adsorption period: the TC solution was first stirred with the catalyst in the dark until adsorption–desorption equilibrium was reached, and only then was the catalytic reaction initiated under light.

Fig. 5a shows that the piezo-photocatalytic performance of the 3 : 1 g–C₃N₄–ZnO@PVDF film is superior to that of the other materials tested, achieving a TC removal efficiency of 99.15%. This demonstrates that coupling g–C₃N₄–ZnO with a piezoelectric matrix effectively suppresses electron–hole recombination, thereby improving photocatalytic efficiency and boosting the TC degradation rate. Fig. 5h indicates that the g-C₃N₄–ZnO@PVDF film is more hydrophilic than either component alone. Its more hydrophilic surface (lower contact angle) facilitates the diffusion of contaminants to the catalyst interface, thereby further enhancing the TC removal efficiency of g-C₃N₄–ZnO@PVDF.

Fig. 5 Effects of various operational conditions on TC removal efficiency. (a) Comparison of different catalysts under combined light and aeration. (b) Effect of individual vs. combined light and aeration. (c) Performance of g-C₃N₄–ZnO@PVDF with different g-C₃N₄ : ZnO ratios. (d) TC removal efficiency at different initial concentrations. (e) Pseudo-first-order kinetic fit for different TC concentrations. (f) TC removal efficiency for different film thicknesses. (g) Pseudo-first-order kinetic fit for different film thicknesses. (h) Water contact angle (hydrophilicity) of the films.

As shown in Fig. 5b, combining light irradiation with aeration yields a much higher TC degradation rate than using either aeration or illumination alone. Among the single-factor conditions, aeration (piezoelectric catalysis) is more effective than illumination (photocatalysis) alone, while light-only conditions give the lowest removal efficiency (51.68%). These results confirm that the piezoelectric effect induced by aeration (vibrational stress) greatly enhances the decontamination performance of the g-C₃N₄–ZnO@PVDF catalyst.

The influence of the g-C₃N₄ : ZnO ratio on TC degradation was examined (Fig. 5c). Composite films were fabricated with 100 mg of g-C₃N₄–ZnO at different g-C₃N₄ : ZnO mass ratios (1 : 1, 3 : 1, 5 : 1) added to the PVDF casting solution. At 20 mg L⁻¹ TC, the removal efficiencies achieved by the 1 : 1, 3 : 1, and 5 : 1 g-C₃N₄–ZnO@PVDF films were 86.85%, 99.15%, and 91.74%, respectively. These results indicate that although all tested composites show high efficiency, the 3 : 1 g-C₃N₄ : ZnO ratio provides the optimal removal performance.

To evaluate the performance of g-C₃N₄–ZnO@PVDF under different pollutant loads, experiments were conducted at various initial TC concentrations. For these tests, the composite ratio was fixed at 3 : 1 (g-C₃N₄ : ZnO) with 50 mg catalyst in the PVDF matrix, and the resulting films were cast to a thickness of 1.25 mm. According to Fig. 5d, at TC concentrations of 20, 50, and 80 mg L⁻¹, the TC removal efficiencies were 99.15%, 71.92%, and 54.43%, respectively. Fig. 5e presents the first-order kinetic fitting for TC degradation; at 20 mg L⁻¹, the rate constant k is highest, reflecting the best removal performance at this concentration. These results indicate that as the initial pollutant concentration increases, the removal efficiency decreases. Nonetheless, even at higher concentrations, the g-C₃N₄–ZnO@PVDF film still achieves appreciable removal of the contaminant.

Fig. 5f demonstrates that the film thickness influences the pollutant removal efficiency. Keeping other conditions constant, films of 1.00 mm, 1.25 mm, and 1.50 mm thickness achieved TC removal efficiencies of 70.64%, 97.25%, and 64.88%, respectively. Fig. 5g shows first-order kinetic fits for different thicknesses; notably, the 1.25 mm film exhibits the highest reaction rate constant. These results suggest that the 1.25 mm film provides the optimal catalytic performance. In general, thinner films offer a higher surface-area-to-volume ratio and better contact with the pollutant solution, which enhances the piezo-photocatalytic activity and pollutant removal. Fig. S4 demonstrates the influence of different catalyst loading amounts on the removal efficiency of TC under the condition that other factors remain unchanged, among them, both 50 mg and 75 mg can achieve the best effect, so the 50 mg load is the most cost-effective.

Meanwhile, we evaluated the mineralization degree of the degraded water samples, and the results are shown in the Fig. S5, The 87.70% mineralization rate indicates that g-C₃N₄–ZnO@PVDF the composite film can not only decompose the parent structure of tetracycline, but also deeply oxidize the vast majority of it into harmless inorganic small molecules (such as CO₂ and H₂O), rather than merely converting it into intermediate products.

To verify the cycling performance of g-C₃N₄–ZnO@PVDF, cyclic experiments were carried out using a 3 : 1 g-C₃N₄–ZnO@PVDF film (1.25 mm thickness) treating 20 mg L⁻¹ TC. As shown in Fig. S6, the TC removal efficiency of the g-C₃N₄–ZnO@PVDF film only slightly decreased with increasing number of cycles. This result indicates that the g-C₃N₄–ZnO@PVDF film can withstand repeated use (mechanical and operational stresses) while maintaining a high removal efficiency. Such stability and reusability help overcome the typical challenges associated with separating and recovering traditional powder catalysts.

Under the same conditions as in the original article, we respectively degraded two dyes (rhodamine B and methylene blue) and two antibiotics (ofloxacin and ciprofloxacin), the concentration of these solutions is all 20 mg L⁻¹. We can clearly see from the picture that our material has very good universality (Fig. S7).

Table S4 compares the TC removal efficiencies of g-C₃N₄–ZnO@PVDF with those of similar photocatalysts reported in the literature.^54–62 Fig. S8 further illustrates that the g-C₃N₄–ZnO@PVDF catalyst exhibits competitive performance under visible-light irradiation.

2.5 Theoretical calculations

Fig. S9a illustrates the electrostatic potential (ESP) surface of the TC molecule. Red contours mark electron-rich (negative-potential) zones, blue denotes electron-deficient (positive-potential) areas, and green highlights nearly neutral regions, together portraying the molecule's overall charge landscape. The electron-laden red domains are the most vulnerable to electrophilic radical attack.⁶³

Fig. S9b and c illustrate the frontier-orbital topology of TC, namely its highest occupied (HOMO) and lowest unoccupied (LUMO) molecular orbitals. In the corresponding 3-D isosurface maps, red signifies electron-rich (negatively charged) regions, whereas green denotes electron-poor (positively charged) areas. The HOMO electron density is concentrated chiefly on the aromatic benzene ring, rendering that segment susceptible to electrophilic radical attack. Conversely, the LUMO is situated mainly on the amino, hydroxyl and carbonyl groups, suggesting these functional sites are preferential targets for nucleophilic species such as ˙O₂⁻ and ˙OH.

2.6 Piezoelectric characteristic

The g-C₃N₄–ZnO@PVDF composite is calculated to have a lower work function (Φ) than g-C₃N₄–ZnO (without PVDF). This reduced work function lowers the energy barrier for photogenerated electrons to escape from the catalyst surface to the reaction medium, thereby facilitating more efficient electron transfer to reactants (Fig. 6a and b). According to the density of states (DOS) analysis (Fig. 6c), the valence and conduction bands of g-C₃N₄–ZnO@PVDF are dominated by contributions from Zn and N orbitals. Moreover, the overall electronic state density of the composite increases upon the introduction of PVDF (which brings in H and F atoms).

Fig. 6 Calculated electronic properties of g-C₃N₄–ZnO and g-C₃N₄–ZnO@PVDF. (a and b) Work function plots for g-C₃N₄–ZnO and g-C₃N₄–ZnO@PVDF, respectively. (c) Density of states of the g-C₃N₄–ZnO@PVDF composite. Charge density difference distributions at the g-C₃N₄–ZnO@PVDF interface under (d) compressive strain, (e) no strain, and (f) tensile strain. Optimized structures and 2D charge density difference cross-sections for (g) g-C₃N₄–ZnO and (h) g-C₃N₄–ZnO@PVDF.

Fig. 6d–f show the optimized g-C₃N₄–ZnO@PVDF structures under compressive strain, no strain, and tensile strain, respectively. To investigate charge-transfer behavior under mechanical stress, charge density difference simulations were performed for g-C₃N₄–ZnO@PVDF under different stress conditions (compression, tension, and no stress).⁶⁴ Three-dimensional charge density difference maps were obtained for each case. The simulations indicate that charge accumulation occurs at the interfaces: between PVDF and the g-C₃N₄–ZnO composite, as well as at the g-C₃N₄/ZnO interface. In the charge difference plots, yellow regions denote charge accumulation, whereas cyan regions denote charge depletion.⁶⁵ These simulations indicate a net electron transfer from PVDF into the g-C₃N₄–ZnO phase, and also a charge redistribution between g-C₃N₄ and ZnO (electrons migrating from one to the other), confirming the presence of an internal electric field in the composite.

Bader charge analysis quantifies this effect: at zero stress, PVDF transfers 1.03 electrons to g-C₃N₄–ZnO; under compressive strain, 1.14 e⁻ are transferred; and under tensile strain, 1.26 e⁻ are transferred. This increase in transferred charge under stress confirms that a polarized electric field is generated in g-C₃N₄–ZnO@PVDF, which enhances charge transfer during the piezoelectric process.

The electron localization function (ELF) is used to visualize and quantify the degree of electron localization in molecules and solids. Fig. 6g and h present cross-sectional charge density maps for g-C₃N₄–ZnO and g-C₃N₄–ZnO@PVDF, respectively. In these plots, blue regions indicate charge depletion while red regions indicate charge accumulation. The isolated g-C₃N₄ and ZnO show electron distributions largely localized around atomic sites, indicating that each material alone readily accommodates external electrons. In contrast, the charge density analysis of the g-C₃N₄–ZnO@PVDF composite reveals dual electronic behavior: a pronounced accumulation of charge at the g-C₃N₄/ZnO interface, and a dominant electron migration from the PVDF (particularly its C and F atoms) toward the semiconductor interfaces.⁶⁶

2.7 Derivation of TC purification pathways

To unravel the degradation mechanism of TC on g-C₃N₄–ZnO@PVDF, the treated solutions were examined by high-performance liquid chromatography-mass spectrometry (HPLC-MS), enabling identification of the reaction intermediates. The resulting mass spectra for pristine TC and its degradation products (Fig. S10) underpin the plausible degradation pathways schematically outlined in Fig. 7.

Fig. 7 Proposed degradation pathways for TC removal by the g-C₃N₄–ZnO@PVDF system.

Pathway 1. The TC parent molecule (m/z 455) first undergoes demethylation and dehydroxylation to form intermediate A (m/z 416). Intermediate A then undergoes deamination and further dehydroxylation to yield intermediate B (m/z 381). Next, B undergoes ring cleavage along with additional deamination, producing compound C (m/z 318). Finally, C undergoes deacylation to give product D (m/z 274).

Pathway 2. Alternatively, TC (m/z 455) can first undergo deamination and dehydroxylation to form intermediate E (m/z 416). E then l000oses an acyl group, yielding intermediate F (m/z 388). Compound F subsequently undergoes demethylation, another deamination, and further dehydroxylation to form intermediate G (m/z 326). Finally, removal of additional methyl and hydroxyl groups from G produces compound H (m/z 280).

Upon further attack by reactive species (e.g.,˙OH), the ring structures in intermediates D and H undergo successive opening reactions. This breaks them down into smaller organic fragments I (m/z 114), J (m/z 106), and K (m/z 100) as various functional groups are cleaved off. With continued photocatalysis, these small molecules are eventually mineralized into innocuous end-products (CO₂ and H₂O).

2.8 Determination of intermediates and piezoelectric photocatalytic mechanism

To clarify the piezo-photocatalytic mechanism and identify the key reactive species, radical scavenging experiments were performed. Specifically, ammonium oxalate (AO, 1 mM) was added as an ˙OH scavenger, p-benzoquinone (BQ, 1 mM) as an ˙O₂⁻ scavenger, and isopropanol (IPA, 1 mL) as a hole (h⁺) scavenger. Fig. 8a shows that adding BQ or AO markedly inhibits the TC degradation efficiency, whereas adding IPA has a comparatively smaller effect. Thus, ˙OH and ˙O₂⁻ are identified as the primary reactive species responsible for TC degradation in this system, with the latter playing a more significant role.

Fig. 8 (a) Effect of different scavengers on TC removal rate: none, AO (˙OH scavenger), BQ (˙O₂⁻ scavenger), and IPA (h⁺ scavenger). (b) EPR spectrum for ˙OH (DMPO-˙OH) and (c) for ˙O₂⁻ (DMPO-˙O₂⁻) after 5 min reaction (light + aeration). (d) Schematic mechanism of piezoelectric-photocatalytic TC degradation by g-C₃N₄–ZnO@PVDF.

Electron paramagnetic resonance (EPR) spin-trap experiments using DMPO were carried out to detect ROS in the g-C₃N₄–ZnO@PVDF system. As shown in Fig. 8b and c, at the start of the reaction (0 min) no EPR signals were observed. After 5 minutes of reaction, characteristic DMPO-˙OH and DMPO-˙O₂⁻ adduct signals appear under both light and aeration conditions. This confirms that ˙OH and ˙O₂⁻ radicals are generated and play a dominant role in TC degradation, by accelerating charge carrier separation and enhancing the overall catalytic efficiency.

Fig. 8d illustrates the proposed piezoelectric-photocatalytic mechanism for the g-C₃N₄–ZnO@PVDF system. Mechanical agitation from aeration applies stress on PVDF, inducing a piezoelectric charge separation in PVDF that injects electrons into the conduction band (CB) of ZnO. Meanwhile, under light irradiation, ZnO generates electron–hole pairs; the photogenerated electrons travel upward to the ZnO CB and the holes remain in the ZnO VB (the VB edge of ZnO is at +2.82 V vs. NHE and the CB edge at −0.30 V, whereas g-C₃N₄ has a VB at +1.36 V and a CB at −1.36 V vs. NHE). EPR results confirm the presence of both ˙OH and ˙O₂⁻ in the system. Because the CB of ZnO (−0.30 V vs. NHE) is slightly less negative than the O₂/˙O₂⁻ reduction potential (−0.33 V), electrons in ZnO's CB cannot readily reduce O₂ to ˙O₂⁻. In contrast, the CB of g-C₃N₄ (−1.36 V vs. NHE) is much more negative, so electrons in g-C₃N₄'s CB can efficiently reduce O₂ to ˙O₂⁻, which contributes to TC degradation. Moreover, the H₂O/˙OH oxidation potential (+2.72 V vs. NHE) lies between the VB potentials of g-C₃N₄ (+1.36 V) and ZnO (+2.82 V). Therefore, holes in the ZnO VB are sufficiently oxidizing to convert H₂O or OH⁻ into ˙OH radicals (capable of oxidizing TC), whereas holes in g-C₃N₄'s VB are not. In the Z-scheme heterojunction between g-C₃N₄ and ZnO, electrons from ZnO's CB migrate to and recombine with holes in g-C₃N₄'s VB. This charge relocation effectively separates the remaining electrons and holes (preventing their recombination within each semiconductor), thereby improving overall charge separation efficiency.

3. Conclusions

In summary, we have developed a novel g-C₃N₄–ZnO@PVDF film that leverages a piezo-photocatalytic synergy to overcome the high recombination rate of conventional semiconductor catalysts. By ingeniously coupling the piezoelectrical property of PVDF with a direct Z-scheme heterojunction, the composite achieves an unprecedented synergistic effect. It demonstrates a remarkable 99.15% tetracycline degradation efficiency, far surpassing individual catalytic processes. The Z-scheme heterojunction between g-C₃N₄ and ZnO facilitated efficient spatial charge separation, with ˙O₂⁻ and ˙OH identified as the dominant reactive species in the degradation process. The composite film's robust cycling stability (>90% efficiency retention over five cycles) effectively resolves the persistent issue of catalyst loss and reuse. Finally, a comprehensive mechanism for the Z-scheme heterojunction piezo-photocatalytic degradation of TC was proposed. Overall, this study provides pivotal insights that advance the paradigm of piezo-photocatalysis, laying the groundwork for a readily scalable technology to address the critical challenge of antibiotic-laden wastewater.

Author contributions

Jiajun Li: investigation, methodology, formal analysis, data curation, writing – original draft. Bing Yang: investigation, methodology, software, formal analysis, data curation, writing – original draft. Ran Deng: formal analysis, data curation, writing – original draft. Tingting Yu: conceptualization, supervision, conceptualization, writing – review & editing, funding acquisition. Tao Yang: formal analysis, data curation, writing – original draft. Wenbin Chen: investigation, methodology, formal analysis. Jizhou Jiang: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, writing – review & editing.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

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

Supplementary Information (SI): experimental details, physical characterizations, FTIR spectra, XPS results, N₂ sorption isotherms and pore size distributions, TOC removal rate, cyclic experiment, universal experiment, performance comparison, details of theoretical. calculations, M–S plots https://doi.or./10.1039/d5ta07439h.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (No. 22506066, 62004143), the Key Project of Scientific Research Plan of Hubei Provincial Department of Education (No. D20241501), the Natural Science Foundation of Jiangsu Province (No. BK20181074), the China Postdoctoral Science Foundation (No. 2021M691327), the Jiangsu Postdoctoral Science Foundation (No. 2021K313C), the Lianyungang Postdoctoral Research Foundation and Postgraduate Research & Practice Innovation Program of Jiangsu Province (No. KYCX23_3460), the Postgraduate Research & Practice Innovation Program of Jiangsu Ocean University (No. KYCX2024-37), Jiangsu Ocean University talent introduction start-up fund (No. KQ18005) along with the Innovation Project of Engineering Research Center of Phosphorus Resources Development and Utilization of Ministry of Education (No. LCX202404), the Key Research and Development Project of Lianyungang City (Social Development) (No. SF2516), the Lianyungang City's Challenge and Response (Technology Transfer) project (No. CA202201).

References

1. L. Feyzi, N. Rahemi and S. Allahyari, Efficient degradation of tetracycline in aqueous solution using a coupled S-scheme ZnO/g-C₃N₄/zeolite P supported catalyst with water falling film plasma reactor, Process Saf. Environ. Prot., 2022, 161, 827–847,  DOI:10.1016/j.psep.2022.03.076.
2. A. J. Browne, M. G. Chipeta, G. Haines-Woodhouse, E. P. A. Kumaran, B. H. K. Hamadani, S. Zaraa, N. J. Henry, A. Deshpande, R. C. Reiner, N. P. J. Day, A. D. Lopez, S. Dunachie, C. E. Moore, A. Stergachis, S. I. Hay and C. Dolecek, Global antibiotic consumption and usage in humans, 2000–18: a spatial modelling study, Lancet Planet. Health, 2021, 5, e893–e904,  DOI:10.1016/S2542-5196(21)00280-1.
3. K. Bharathi, G. Bakiyaraj, J. Archana and M. Navaneethan, p-n ZnCr₂O₄/Bi₂O₃ heterojunction as an effective photocatalytic degradation of MB and CIP, Opt. Mater., 2025, 162, 116862,  DOI:10.1016/j.optmat.2025.116862.
4. C. Nannou, K. N. Maroulas, C. Tsamtzidou, K. Ladomenou and G. Z. Kyzas, Photocatalytic degradation of veterinary antibiotics in wastewaters: A review, Sci. Total Environ., 2025, 966, 178765,  DOI:10.1016/j.scitotenv.2025.178765.
5. L. Meng, Y. Zhao, X. Li, Y. Kong, J. Guo and M. Liu, The effect of bacterial functional characteristics on the spread of antibiotic resistance genes in expanded granular sludge bed reactor treating the antibiotic wastewater, Ecotoxicol. Environ. Saf., 2021, 225, 112714,  DOI:10.1016/j.ecoenv.2021.112714.
6. P. Kovalakova, L. Cizmas, M. Feng, T. J. McDonald, B. Marsalek and V. K. Sharma, Oxidation of antibiotics by ferrate(VI) in water: Evaluation of their removal efficiency and toxicity changes, Chemosphere, 2021, 277, 130365,  DOI:10.1016/j.chemosphere.2021.130365.
7. M. Liu, Y. Wan, Y. Wang, J. Xu and X. Li, Robust photoelectrocatalytic degradation of antibiotics by organic-inorganic PDISA/Bi₂WO₆ S-scheme heterojunction membrane, J. Environ. Chem. Eng., 2024, 12, 112328,  DOI:10.1016/j.jece.2024.112328.
8. B. O. Orimolade, A. O. Oladipo, A. O. Idris, F. Usisipho, S. Azizi, M. Maaza, S. L. Lebelo and B. B. Mamba, Advancements in electrochemical technologies for the removal of fluoroquinolone antibiotics in wastewater: A review, Sci. Total Environ., 2023, 881, 163522,  DOI:10.1016/j.scitotenv.2023.163522.
9. S. Li, P. L. Show, H. H. Ngo and S.-H. Ho, Algae-mediated antibiotic wastewater treatment: A critical review, Environ. Sci. Ecotechnology, 2022, 9, 100145,  DOI:10.1016/j.ese.2022.100145.
10. J. Ma, Y. Cui, A. Li, X. Zou, C. Ma and Z. Chen, Antibiotics and antibiotic resistance genes from wastewater treated in constructed wetlands, Ecol. Eng., 2022, 177, 106548,  DOI:10.1016/j.ecoleng.2022.106548.
11. S. Li, Y. Wu, H. Zheng, H. Li, Y. Zheng, J. Nan, J. Ma, D. Nagarajan and J.-S. Chang, Antibiotics degradation by advanced oxidation process (AOPs): Recent advances in ecotoxicity and antibiotic-resistance genes induction of degradation products, Chemosphere, 2023, 311, 136977,  DOI:10.1016/j.chemosphere.2022.136977.
12. X. Bian, F. Li, J. Zhang, M. Zhong, Y. Yang and S. Khan, Photocatalytic degradation of tetracycline antibiotics in swine wastewater using Fe³⁺-loaded NaBiO₃ coupled with sodium persulfate, Catal. Commun., 2023, 174, 106579,  DOI:10.1016/j.catcom.2022.106579.
13. F. Li, Y. Anjarsari, J. Wang, R. Azzahiidah, J. Jiang, J. Zou, K. Xiang and H. Ma, Arramel. Modulation of the lattice structure of 2D carbon-based materials for improving photo/electric properties, Carbon Lett., 2023, 33, 1321–1331,  DOI:10.1007/s42823-022-00380-4.
14. L. Xiao, W. Ren, S. Shen, M. Chen, R. Liao, Y. Zhou and X. Li, Enhancing photocatalytic hydrogen evolution through electronic structure and wettability adjustment of ZnIn₂S₄/Bi₂O₃ S-scheme heterojunction, Acta Phys.-Chim. Sin., 2024, 40(8), 2308036,  DOI:10.3866/PKU.WHXB202308036.
15. Y. Jin, J. Long, X. Ma, T. Zhou, Z. Zhang, H. Lin, J. Long and X. Wang, Synthesis of caged iodine-modified ZnO nanomaterials and study on their visible light photocatalytic antibacterial properties, Appl. Catal., B, 2019, 256, 117873,  DOI:10.1016/j.apcatb.2019.117873.
16. Z. Zhang, X. Chang and A. Chen, Determination of chemical oxygen demand based on photoelectrocatalysis of nanoporous TiO₂ electrodes, Sens. Actuators, B, 2016, 223, 664–670,  DOI:10.1016/j.snb.2015.10.001.
17. N. A. Chopan and H.-T.-N. Chishti, Polypyrrole-decorated ZnO/g-C₃N₄ S-scheme photocatalyst for rhodamine B dye degradation: Mechanism and antibacterial activity, Mater. Today Chem., 2023, 32, 101643,  DOI:10.1016/j.mtchem.2023.101643.
18. H. Weerathunga, C. Tang, A. J. Brock, S. Sarina, T. Wang, Q. Liu, H.-Y. Zhu, A. Du and E. R. Waclawik, Nanostructure shape-effects in ZnO heterogeneous photocatalysis, J. Colloid Interface Sci., 2022, 606, 588–599,  DOI:10.1016/j.jcis.2021.08.052.
19. L. Wang, T. Yang, B. Feng, X. Xu, Y. Shen, Z. Li, Arramel and J. Jiang, Constructing dual electron transfer channels to accelerate CO₂ photoreduction guided by machine learning and first-principles calculation, Chin. J. Catal., 2023, 54, 265–277,  DOI:10.1016/S1872-2067(23)64546-2.
20. E. Cui, Y. Lu, Z. Li, Z. Chen, C. Ge and J. Jiang, Interfacial B-O bonding modulated S-scheme B-doped N-deficient C₃N₄/O-doped-C₃N₅ for efficient photocatalytic overall water splitting, Chin. Chem. Lett., 2025, 36, 110288,  DOI:10.1016/j.cclet.2024.110288.
21. D. Wen, N. Wang, J. Peng, T. Majima and J. Jiang, Unique Cu^x+/Cu⁰ active-site switches in Cu-loaded g-C₃N₄ nanosheets for efficient photocatalytic CO₂ reduction, J. Mater. Sci. Technol., 2025, 226, 93–108,  DOI:10.1016/j.jmst.2024.12.010.
22. Q. Wang, L. Zhang, Y. Guo, M. Shen, M. Wang, B. Li and J. Shi, Multifunctional 2D porous g-C₃N₄ nanosheets hybridized with 3D hierarchical TiO₂ microflowers for selective dye adsorption, antibiotic degradation and CO₂ reduction, Chem. Eng. J., 2020, 396, 125347,  DOI:10.1016/j.cej.2020.125347.
23. Z. Zan, X. Li, X. Gao, J. Huang, Y. Luo and L. Han, 0D/2D carbon nitride quantum dots (CNQDs)/BiOBr S-scheme heterojunction for robust photocatalytic degradation and H₂O₂ production, Acta Phys.-Chim. Sin., 2023, 39(6), 2209016,  DOI:10.3866/PKU.WHXB202209016.
24. X. Zhou, C. Zhao, J. Chen and L. Chen, Influence of B, Zn, and B-Zn doping on electronic structure and optical properties of g-C₃N₄ photocatalyst: A first-principles study, Results Phys., 2021, 26, 104338,  DOI:10.1016/j.rinp.2021.104338.
25. Q. Chen, W. Yang, J. Zhu, L. Fu, D. Li and L. Zhou, Enhanced visible light photocatalytic activity of g-C₃N₄ decorated ZrO₂-x nanotubes heterostructure for degradation of tetracycline hydrochloride, J. Hazard. Mater., 2020, 384, 121275,  DOI:10.1016/j.jhazmat.2019.121275.
26. M. MuthuKathija, R. M. Muthukrishnan, D. Renuka Devee, S. M. A. Kader, V. Rama and M. S. M. Badhusha, A novel biogenic method to synthesis a ternary (ZnO-Ag)/g-C₃N₄ nanocomposite with an enhanced photocatalytic and antibacterial activities, Inorg. Chem. Commun., 2023, 154, 110877,  DOI:10.1016/j.inoche.2023.110877.
27. X. Lu, Y. Jin, X. Zhang, G. Xu, D. Wang, J. Lv, Z. Zheng and Y. Wu, Controllable synthesis of graphitic C₃N₄/ultrathin MoS₂ nanosheet hybrid nanostructures with enhanced photocatalytic performance, Dalton Trans., 2016, 45, 15406–15414,  10.1039/c6dt02247b.
28. X. Li, T. Han, Y. Zhou, Y. Xie, Y. Luo, J. Huang, Z. Chen and F. Deng, Photoelectrocatalytic hydrogen evolution and synchronous degradation of organic pollutants by pg-C₃N₄/β-FeOOH S-scheme heterojunction, Sci. China Technol. Sci., 2024, 67, 1238–1252,  DOI:10.1007/s11431-023-2604-x.
29. J. Cai, N. Hu, L. Wu, Y. Liu, Y. Li, H. Ning, X. Liu and L. Lin, Preparing carbon black/graphene/PVDF-HFP hybrid composite films of high piezoelectricity for energy harvesting technology, Composites, Part A, 2019, 121, 223–231,  DOI:10.1016/j.compositesa.2019.03.031.
30. R. Kotuniak and W. Bal, Kinetics of Cu(II) complexation by ATCUN/NTS and related peptides: a gold mine of novel ideas for copper biology, Dalton Trans., 2022, 51, 14–26,  10.1039/d1dt02878b.
31. W. Ma, B. Yao, W. Zhang, Y. He, Y. Yu and J. Niu, Fabrication of PVDF-based piezocatalytic active membrane with enhanced oxytetracycline degradation efficiency through embedding few-layer E-MoS₂ nanosheets, Chem. Eng. J., 2021, 415, 129000,  DOI:10.1016/j.cej.2021.129000.
32. L. Paredes, S. Murgolo, H. Dzinun, M. H. Dzarfan Othman, A. F. Ismail, M. Carballa and G. Mascolo, Application of immobilized TiO₂ on PVDF dual layer hollow fibre membrane to improve the photocatalytic removal of pharmaceuticals in different water matrices, Appl. Catal., B, 2019, 240, 9–18,  DOI:10.1016/j.apcatb.2018.08.067.
33. L. Lin, S. Tang, X. S. Wang, X. Sun, Z. Han and Y. Chen, Accumulation mechanism of tetracycline hydrochloride from aqueous solutions by nylon microplastics, Environ. Technol. Innovat., 2020, 18, 100750,  DOI:10.1016/j.eti.2020.100750.
34. Y. Zhang, L. Ye, W. Zhao, L. Chen, M. Zhang, G. Yang and H. Zhang, Antifouling mechanism of the additive-free β-PVDF membrane in water purification process: Relating the surface electron donor monopolarity to membrane-foulant interactions, J. Membr. Sci., 2020, 601, 117873,  DOI:10.1016/j.memsci.2020.117873.
35. X. Li, Y. Wan, F. Deng, Y. Zhou, P. Chen, F. Dong and J. Jiang, Advances in Z-scheme and S-scheme heterojunctions for photocatalytic and photoelectrocatalytic, Chin. Chem. Lett., 2025, 36, 111418,  DOI:10.1016/j.cclet.2025.111418.
36. H. Osman, Z. Su and X. Ma, Efficient photocatalytic degradation of Rhodamine B dye using ZnO/graphitic C₃N₄ nanocomposites synthesized by microwave, Environ. Chem. Lett., 2017, 15, 435–441,  DOI:10.1007/s10311-017-0604-8.
37. H. Zhao, W. Ge, X. Li, P. Wang, J. Bai, B. Kang, Z. Zhang, P. Wang and X. Jing, Mesoporous ZnO: In nanosheets grown on g-C₃N₄ for improved detection of NO₂, J. Alloys Compd., 2023, 959, 170596,  DOI:10.1016/j.jallcom.2023.170596.
38. X. Li, H. Zhang, J. Luo, Z. Feng and J. Huang, Hydrothermal synthesized novel nanoporous g-C₃N₄/MnTiO₃ heterojunction with direct Z-scheme mechanism, Electrochim. Acta, 2017, 258, 998–1007,  DOI:10.1016/j.electacta.2017.11.151.
39. D. Zhu and Q. Zhou, Nitrogen doped g-C₃N₄ with the extremely narrow band gap for excellent photocatalytic activities under visible light, Appl. Catal., B, 2021, 281, 119474,  DOI:10.1016/j.apcatb.2020.119474.
40. L. Huang, D. Bao, J. Li, X. Jiang and X. Sun, Construction of Au modified direct Z-scheme g-C₃N₄/defective ZnO heterostructure with stable high-performance for tetracycline degradation, Appl. Surf. Sci., 2021, 555, 149696,  DOI:10.1016/j.apsusc.2021.149696.
41. D. Shen, X. Li, C. Ma, Y. Zhou, L. Sun, S. Yin, P. Huo and H. Wang, Synthesized Z-scheme photocatalyst ZnO/g-C₃N₄ for enhanced photocatalytic reduction of CO₂, New J. Chem., 2020, 44, 16390–16399,  10.1039/D0NJ02270E.
42. F. Liu, X. Chen, X. Wang, Y. Han, X. Song, J. Tian, X. He and H. Cui, Fabrication of 1D Zn₂SnO₄ nanowire and 2D ZnO nanosheet hybrid hierarchical structures for use in triethylamine gas sensors, Sens. Actuators, B, 2019, 291, 155–163,  DOI:10.1016/j.snb.2019.04.009.
43. J. Zhang, H. Lu, L. Zhang, D. Leng, Y. Zhang, W. Wang, Y. Gao, H. Lu, J. Gao, G. Zhu, Z. Yang and C. Wang, Metal–organic framework-derived ZnO hollow nanocages functionalized with nanoscale Ag catalysts for enhanced ethanol sensing properties, Sens. Actuators, B, 2019, 291, 458–469,  DOI:10.1016/j.snb.2019.04.058.
44. H. Wang, J. Bai, M. Dai, K. Liu, Y. Liu, L. Zhou, F. Liu, F. Liu, Y. Gao, X. Yan and L. Geyu, Visible light activated excellent NO₂ sensing based on 2D/2D ZnO/g-C₃N₄ heterojunction composites, Sens. Actuators, B, 2020, 304, 127287,  DOI:10.1016/j.snb.2019.127287.
45. J. Zhao, X. Guo, W. Sun, J. Wang, J. Jiang and H. Pang, Hierarchical bimetallic nanoparticles embedded nitrogen-doped carbon derived from ZIF-67 composites for efficient electrochemical nitrite sensing, Carbon Lett., 2025, 35, 635–643,  DOI:10.1007/s42823-024-00807-0.
46. J. Zou, S. Wu, Y. Liu, Y. Sun, Y. Cao, J.-P. Hsu, A. T. Shen Wee and J. Jiang, An ultra-sensitive electrochemical sensor based on 2D g-C₃N₄/CuO nanocomposites for dopamine detection, Carbon, 2018, 130, 652–663,  DOI:10.1016/j.carbon.2018.01.008.
47. I. Ahmad, S. Shukrullah, M. Y. Naz, F. K. Alsaif, S. Alsulamy, Y. Khan, N. R. Khalid and W. Q. Khan, Construction of Ag-modified ZnO–CeO₂ 2D-1D S-scheme heterojunction photocatalyst with phenomenal photocatalytic performance for H₂ evolution, Mater. Sci. Semicond. Process., 2023, 159, 107392,  DOI:10.1016/j.mssp.2023.107392.
48. J. Li, Y. Wang, Y. Wang, Y. Guo, S. Zhang, H. Song, X. Li, Q. Gao, W. Shang, S. Hu, H. Zheng and X. Li, MXene Ti₃C₂ decorated g-C₃N₄/ZnO photocatalysts with improved photocatalytic performance for CO₂ reduction, Nano Mater. Sci., 2023, 5, 237–245,  DOI:10.1016/j.nanoms.2023.02.003.
49. S. Li, J. Chen, W. Jiang, Y. Liu, Y. Ge and J. Liu, Facile construction of flower-like bismuth oxybromide/bismuth oxide formate p–n heterojunctions with significantly enhanced photocatalytic performance under visible light, J. Colloid Interface Sci., 2019, 548, 12–19,  DOI:10.1016/j.jcis.2019.04.024.
50. Y. Shi, G. Zhan, H. Li, X. Wang, X. Liu, L. Shi, K. Wei, C. Ling, Z. Li, H. Wang, C. Mao, X. Liu and L. Zhang, Simultaneous manipulation of bulk excitons and surface defects for ultrastable and highly selective CO₂ photoreduction, Adv. Mater., 2021, 33, e2100143,  DOI:10.1002/adma.202100143.
51. D. Zhao, Y. Wang, C.-L. Dong, Y.-C. Huang, J. Chen, F. Xue, S. Shen and L. Guo, Boron-doped nitrogen-deficient carbon nitride-based Z-scheme heterostructures for photocatalytic overall water splitting, Nat. Energy, 2021, 6, 388–397,  DOI:10.1038/s41560-021-00795-9.
52. A. Goyal and M. T. M. Koper, The Interrelated Effect of Cations and Electrolyte pH on the Hydrogen Evolution Reaction on Gold Electrodes in Alkaline Media, Angew. Chem., Int. Ed., 2021, 60, 13452–13462,  DOI:10.1002/anie.202102803.
53. T. Yu, C. Yang, R. Deng, Y. Zhang, J. Li, T. Yang and J. Jiang, Piezoelectric-enhanced BiOBr/UiO-66-NH₂(Hf) photocatalysts for purification of mixed pollutants, Sci. China: Technol. Sci., 2025, 68, 1520206,  DOI:10.1007/s11431-024-2884-5.
54. X. Wang, X. Zhang, Y. Xie, Y. shi, C. Zhao, G. Li, Y. Zhang and H. Shi, Construction of novel biochar/g-C₃N₄/BiOBr heterojunction with S-scheme mechanism for efficient photocatalytic degradation of tetracycline, J. Environ. Chem. Eng., 2024, 12, 113469,  DOI:10.1016/j.jece.2024.113469.
55. W. Zhang, P. Xing, C. Zhang, J. Zhang, X. Hu, L. Zhao and Y. He, Facile synthesis of strontium molybdate coupled g-C₃N₄ composite for effective tetracycline and dyes degradation under visible light, Adv. Powder Technol., 2022, 33, 103573,  DOI:10.1016/j.apt.2022.103573.
56. L. Huang, D. Bao, J. Li, X. Jiang and X. Sun, Construction of Au modified direct Z-scheme g-C₃N₄/defective ZnO heterostructure with stable high-performance for tetracycline degradation, Appl. Surf. Sci., 2021, 555, 149696,  DOI:10.1016/j.apsusc.2021.149696.
57. L. Wang, X. Wang, X. Xu, F. Wu and Y. Ge, Construction of ZnO quantum dots anchored in oxygen-doped g-C₃N₄: A novel Z-type heterojunction with boosting photocatalytic performance for tetracycline degradation, J. Mol. Struct., 2025, 1328, 141382,  DOI:10.1016/j.molstruc.2025.141382.
58. X. Yuan, S. Qu, X. Huang, X. Xue, C. Yuan, S. Wang, L. Wei and P. Cai, Design of core-shelled g-C₃N₄@ZIF-8 photocatalyst with enhanced tetracycline adsorption for boosting photocatalytic degradation, Chem. Eng. J., 2021, 416, 129148,  DOI:10.1016/j.cej.2021.129148.
59. K. Jia, G. Liu, D.-N. Lang, S.-F. Chen, C. Yang, R.-L. Wu, W. Wang and J.-D. Wang, Degradation of tetracycline by visible light over ZnO nanophotocatalyst, J. Taiwan Inst. Chem. Eng., 2022, 136, 104422,  DOI:10.1016/j.jtice.2022.104422.
60. Q. Yu, Z. Li, Z. Wei, M. Ding and H. Zhang, A novel ZnO/FeOCl composite as a photo-Fenton catalyst for degradation tetracycline under visible light, J. Mater. Sci.: Mater. Electron., 2024, 35, 1752,  DOI:10.1007/s10854-024-13497-1.
61. X. Zheng, Y. Liu, X. Liu, Q. Li and Y. Zheng, A novel PVDF-TiO₂@g-C₃N₄ composite electrospun fiber for efficient photocatalytic degradation of tetracycline under visible light irradiation, Ecotoxicol. Environ. Saf., 2021, 210, 111866,  DOI:10.1016/j.ecoenv.2020.111866.
62. Y. Cui, Z. Wang, J. Zheng, B. Li, Y. Yan and M. Meng, Fabrication of silver vanadate quantum dots/reduced graphene oxide/graphitic carbon nitride Z-scheme heterostructure modified polyvinylidene fluoride self-cleaning membrane for enhancing photocatalysis and mechanism insight, J. Colloid Interface Sci., 2022, 614, 677–689,  DOI:10.1016/j.jcis.2022.01.008.
63. Z. Teng, N. Yang, H. Lv, S. Wang, M. Hu, C. Wang, D. Wang and G. Wang, Edge-functionalized g-C₃N₄ nanosheets as a highly efficient metal-free photocatalyst for safe drinking water, Chem, 2019, 5, 664–680,  DOI:10.1016/j.chempr.2018.12.009.
64. W. Zhao, J. Li, B. Dai, Z. Cheng, J. Xu, K. Ma, L. Zhang, N. Sheng, G. Mao, H. Wu, K. Wei and D. Y. C. Leung, Simultaneous removal of tetracycline and Cr(VI) by a novel three-dimensional AgI/BiVO₄ p-n junction photocatalyst and insight into the photocatalytic mechanism, Chem. Eng. J., 2019, 369, 716–725,  DOI:10.1016/j.cej.2019.03.115.
65. J. Zhang, Z. Li, B. Liu, M. Chen, Y. Zhou and M. Zhu, Insights into the role of C-S-C bond in C₃N₅ for photocatalytic NO deep oxidation: Experimental and DFT exploration, Appl. Catal., B, 2023, 328, 122522,  DOI:10.1016/j.apcatb.2023.122522.
66. B. Yang, T. Yu, C. Yang, W. Wu and J. Jiang, Piezoelectric-enhanced MoS₂@PVDF-HFP composite for photocatalytic degradation of tetracycline, J. Alloys Compd., 2025, 1010, 177413,  DOI:10.1016/j.jallcom.2024.177413.

---

Footnote

† These authors contributed equally to the article.

---