Self-regulation of Lewis acid sites on FeOCl toward piezo-self-Fenton reaction for continuous hydroxyl radicals generation

Abstract

Clarifying the structure-property relationship is critical to the design of high-performance piezocatalysts. Herein, with iron oxychloride (FeOCl) as the model piezocatalyst, a systematic investigation into the evolution of surface properties by the induced piezopotential is performed. Except for the promotion in carrier dynamics, the induced piezopotential also achieve in-situ regulation of the surface Lewis acid sites. The surface Lewis acidity of FeOCl, as well as other classic piezocatalyst such as BaTiO₃, BiTiO₃, and BiFeO₃, is dynamically enhanced under the application of mechanical stress. Consequently, the activation and consecutive conversion of the O₂ molecule, whose Lewis basicity is weaker than H₂O₂, is realized on the FeOCl piezocatalyst. The heterogeneous Fenton activity of FeOCl catalyst is upgraded into a more advanced piezo-self-Fenton activity, producing hydroxyl radicals efficiently under ultrasonic vibration (556.8 µmol g⁻¹ h⁻¹). The FeOCl piezo-self-Fenton system exhibits exceptional broad-spectrum degradation efficiency, operational stability, and scalability for the treatment of pharmaceutical wastewater. This work highlights the in-situ regulation of surface property by the induced piezopotential as a new feature for the piezocatalysts, which contributes to the enhancement of piezocatalytic activity and even changes the reaction pathway.

Designing efficient piezocatalysts requires understanding how mechanical stress alters surfaces. This study shows stress-induced electric potentials boost surface acidity, activate oxygen, and enable hydrogen-peroxide-free pollutant degradation.

Introduction

Piezocatalysis has emerged as a sustainable and cost-efficient strategy for environmental remediation^1,2. This technique converts the ambient mechanical energy (sound, wave, and vibration) into chemical energy, facilitating the on-demand and on-site energy utilization without the constraint of location and time^3–5. During piezocatalysis, the catalyst with piezoelectricity will generate an induced surface potential in response to the external mechanical stress, which triggers the catalytic reaction^6–8. Improving the piezocatalytic activity is critical to the practical application of this technology. In recent years, researchers have made great efforts in exploring catalysts with strong piezoelectricity and in constructing active sites through material engineering. Nevertheless, an in-depth investigation of the structure-property relationship of the piezocatalyst is still needed, which may guide the rational design of the high-performance piezocatalyst^9,10.

The piezocatalytic activity is the consequence of the induced piezopotential acting on the surface active sites^11–13. Mechanical energy, as well as the surface energy, compensates for the activation barrier of the target reactant and facilitates the catalytic reaction. The positive effect of piezopotential mainly reflects on the promotion of the carrier dynamics¹⁴. For instance, Ding et al. confirmed that the heterogeneous Fenton activity of iron oxychloride (FeOCl) catalyst for bisphenol A degradation was enhanced by 18 times with the induced piezopotential under ultrasonic vibration¹⁵. Our recent study also highlighted the importance of constructing active sites on the surface where the induced piezopotential was located, which may effectively connect the process of piezoresponse and surface catalysis. In addition to facilitating the carrier dynamics, the induced piezopotential can also realize the in-situ regulation of the surface properties, representing another key issue to improving the piezocatalytic activity^16–18. Wang et al. reported the phenomenon of an increase in surface Lewis acid/base sites on the hydroxyapatite piezocatalyst under ultrasonic vibration, providing new insights into the mechanism of piezocatalysis¹⁹. However, the exploration of the in-situ regulation behavior by the piezopotential is still in its infancy. Clarifying the relevant mechanisms will help guide the modification and optimization of piezocatalysts.

Proceeding from the insights above, FeOCl is herein selected as the model catalyst with intrinsic heterogeneous Fenton activity and piezoelectricity. This study focuses on the evolution of surface properties on FeOCl during the application of mechanical stress. Surprisingly, in-situ Lewis acid (LA) site engineering by the induced piezopotential is validated and extended to the common piezocatalysts. More importantly, the in-situ regulation of surface LA sites over FeOCl switches the reaction pathway from heterogeneous Fenton reaction (H₂O₂ → HO^•) to heterogeneous piezo-self-Fenton reaction (O₂ → H₂O₂ → HO^•). The activation and consecutive conversion of the O₂ molecule, whose Lewis basicity is weaker than H₂O₂, is achieved. As a result, efficient HO^• production (556.8 µmol g⁻¹ h⁻¹) over FeOCl from the piezo-self-Fenton process is realized under ultrasonic vibration without H₂O₂ input. This piezo-self-Fenton activity is comparable to the state-of-the-art homogeneous and heterogeneous Fenton activity, showing significant potential in the treatment of antibiotic-laden wastewater. The FeOCl piezo-self-Fenton system demonstrates unparalleled broad-spectrum degradation efficiency against diverse antibiotic pollutants, as validated through both simulated wastewater models and pharmaceutical wastewater. This work emphasizes the in-situ regulation of surface property by the piezoelectric effect of the catalyst as a general phenomenon, offering insightful viewpoints for the optimization of high-performance piezocatalytic systems.

Results

FeOCl catalyst and its intrinsic piezoelectricity

FeOCl nanobelts, as the representative heterogeneous Fenton catalysts, were synthesized via the pyrolysis of FeCl₃·6H₂O^20–22. Supplementary Fig. 1 presents the X-ray diffraction (XRD) pattern of the as-prepared FeOCl. The pattern closely aligns with the standard reference data (JCPDS No. 39-0612)²³, confirming the successful preparation of the FeOCl catalyst. The transmission electron microscopy (TEM) image in Fig. 1a shows that the length and width of the FeOCl nanobelt are approximately 400 and 50 nm, respectively. Energy-dispersive X-ray spectrometry (EDS) mapping images of the sample further confirm that Fe, O, and Cl are uniformly distributed within the nanobelt (Supplementary Fig. 2). Additionally, the atomic force microscopy (AFM) image (Supplementary Fig. 3) indicates a thickness of ~200 nm for the FeOCl nanobelt. The piezoresponse behavior of FeOCl was assessed using piezoresponse force microscopy (PFM). Figure 1b shows a characteristic ~180° phase loop and a butterfly-shaped amplitude-voltage curve collected from the FeOCl nanobelts, demonstrating the intrinsic piezoelectricity. The piezopotential generated by FeOCl was further corroborated by the Kelvin probe force microscopy (KPFM) (Fig. 1c). The surface potential of 18.3 mV provided additional support for the intrinsic piezoelectricity of FeOCl. Electrochemical impedance spectroscopy (EIS) further revealed that the charge‑transfer resistance of FeOCl decreased significantly under ultrasonic vibration (Supplementary Fig. 4), confirming that the induced piezopotential effectively facilitates interfacial electron transfer. The significant transient current response of FeOCl under ultrasonic vibration (Supplementary Fig. 5) also suggests the effective charge migration achieved by the induced piezopotential. These findings indicate that the intrinsic piezoelectricity of FeOCl contributes to the enhancement of carrier dynamics, coinciding with the promotion of the heterogeneous Fenton activity under ultrasonic vibration as previously reported^24–26.

Fig. 1. Evolution of LA sites on the FeOCl piezocatalyst.

a TEM images of FeOCl. b Butterfly-shaped amplitude curve and hysteresis loop of FeOCl. c KPFM image and the surface potential of FeOCl. d Dynamic pyridine adsorption on FeOCl. e Pyridine-FTIR spectra of FeOCl before and after ultrasonic vibration. f Concentration of LA sites on FeOCl derived from the Pyridine-FTIR spectra before and after ultrasonic vibration. US denotes ultrasonic vibration. The error bars represent the standard deviation, calculated from at least two independent experiments.

Validation of in-situ LA site engineering

The influence of induced piezopotential on the surface property of FeOCl catalyst was then studied. Based on the crystal structure, the surface coordinatively unsaturated Fe atoms with Lewis acidity on FeOCl contribute to the intrinsic heterogeneous Fenton activity^27,28.

Focusing on the evolution of LA sites, a series of measurements with pyridine as the probe molecule were performed (Supplementary Fig. 6). For pyridine adsorption over FeOCl (Fig. 1d), a remarkable 8.1-fold increase was attained under ultrasonic vibration compared to the static adsorption. Partial desorption of the adsorbed pyridine was detected when the vibration was stopped, but the adsorption capacity remained higher than that of static adsorption¹⁹. This result indicates a significant increase in the concentration of surface LA sites on FeOCl under the application of mechanical stress. Infrared spectroscopy of pyridine adsorption (Py-IR) was employed to probe the surface acid sites on FeOCl before and after ultrasonic vibration (Fig. 1e). The characteristic absorption bands near at 1450 cm⁻¹ and 1603 cm⁻¹ are attributed to pyridine adsorbed on LA sites²⁹. Compared to the fresh sample, the post-treated FeOCl exhibited stronger peak intensities under identical desorption temperatures, suggesting an overall increase in LA site density. As the temperature was raised, the intensity of these bands gradually decreased. Since the site with stronger Lewis acidity usually requires a higher temperature for pyridine desorption, this trend reveals that there were abundant acid sites with different Lewis acidity on the FeOCl catalyst. Quantitative analysis based on the band near 1450 cm⁻¹ further discloses the effect of induced piezopotential on FeOCl (Fig. 1f). For the fresh FeOCl catalyst, the density of LA sites at 50, 100, and 200 °C was 24.87, 14.01, and 5.95 μmol g⁻¹, respectively. After being subjected to ultrasonic vibration, these values increased to 28.32, 16.80, and 7.34 μmol g⁻¹. Furthermore, the NH₃-TPD profile of FeOCl after ultrasonic vibration reveals an intensified desorption signal compared to the fresh sample, accompanied by a distinct shift of the NH₃ desorption peak from 147 °C to 172 °C (Supplementary Fig. 7). These results collectively affirm that the induced piezopotential not only in-situ increases the concentration of surface LA sites on FeOCl but also strengthens the acidity³⁰. Consistently, O₂-TPD indicates enhanced O₂ adsorption after treatment. Elevated desorption intensity reflects increased adsorption capacity, and the shift of the desorption peak to higher temperatures signifies strengthened O₂-surface interactions (Supplementary Fig. 8). These results demonstrate that the enhanced Lewis acidity directly facilitates O₂ adsorption and activation. Furthermore, the XRD pattern of FeOCl after ultrasonic vibration exhibits no notable changes compared with the fresh sample (Supplementary Fig. 9), indicating that the observed enhancements in surface properties are not due to phase transformation or loss of structural integrity.

To provide more insights for in-situ LA site engineering, In-situ Raman spectra was employed to investigate the bonding configuration of FeOCl under external mechanical forces. As shown in Fig. 2a, the vibration mode of FeOCl at 224 cm⁻¹ under atmospheric pressure is attributed to the stretching vibration of the Fe-Cl bond, whereas the modes at 291 and 410 cm⁻¹ correspond to the stretching vibration of the Fe-O bond^31,32. As the applied pressure increases, the Raman signals of FeOCl shift toward higher frequencies, although the overall profile of the spectra remains unchanged. This shift indicates a slight shortening of the bonds. When the applied pressure is above 2.2 GPa, a new Raman signal at approximately 300 cm⁻¹ is gradually detected, indicating a reduced symmetry of FeOCl under mechanical forces. The change of lattice symmetry may be the key to the in-situ regulation of surface properties. As shown in Fig. 2b, the magnetic hysteresis loop of FeOCl exhibits negligible changes after ultrasonic vibration. This result implies that the spin state of the Fe center was not affected under the application of mechanical stress, pointing the essence of in-situ LA site engineering to the evolution of electronic structure caused by the induced piezopotential³³.

Fig. 2. Mechanisms of the in-situ LA site engineering by the induced piezopotential.

a In-situ Raman spectra of FeOCl upon different pressures at room temperature. b Magnetization hysteresis (M-H) loops of FeOCl before and after ultrasonic vibration. c Fe 2p XPS spectra of FeOCl before and after ultrasonic vibration. d Normalized Fe K-edge XANES spectra of FeOCl and the reference samples (i.e., Fe foil, FeO, Fe₂O₃, and Fe₃O₄). e The chemical valence of iron for FeOCl and the reference samples. US denotes ultrasonic vibration.

Based on the results above, the valence state of Fe on the FeOCl catalyst is then characterized. The Fe 2p X-ray photoelectron spectroscopy (XPS) spectra of the FeOCl catalysts before and after being subjected to ultrasonic vibration are shown in Fig. 2c. The spectra show three deconvoluted doublet peaks, corresponding to the Fe³⁺ 2p_3/2, Fe³⁺ 2p_1/2, and oscillating satellite signals. Following exposure to ultrasonic vibration, only a slight shift in binding energy is observed, reflecting minor changes in the surface electronic structure. X-ray absorption spectroscopy (XAS) provides additional details of the electronic structure and coordination environment of the FeOCl catalyst at the atomic level. In the Fe K-edge X-ray absorption near-edge structure (XANES) spectra (Fig. 2d), the curve of FeOCl closely aligns with that of Fe₂O₃, indicating that the oxidation state of Fe atoms is predominantly +3^34,35. After ultrasonic vibration, the absorption edge of FeOCl shifts to a lower energy. The result of linear combination fitting (Fig. 2e) confirms that the average oxidation state of Fe on FeOCl varies from +2.89 to +2.88 along with the application of mechanical stress³⁶. The subtle changes observed in the XPS and XANES spectra reflect a reversible modulation of the surface electronic structure induced by the piezopotential, rather than persistent changes in material properties. The results indicate that mechanical stress can dynamically modulate surface properties. Crucially, this minimal change in formal oxidation state indicates that the enhancement in Lewis acidity does not arise from the conventional redox process. Instead, density-of-states (DOS) analysis reveals its fundamental origin in the electronic structure by showing that the d‑band center of the surface Fe atoms shifts upward from −1.72 eV at 0 GPa to −1.47 eV at 2 GPa (Supplementary Fig. 10). This upshift of the d‑band center, a key descriptor in catalysis, shifts the energy of Fe 3 d orbitals closer to the Fermi level. Consequently, the elevated d states strengthen interactions between Fe sites and electron-donating adsorbates, thereby increasing the Lewis acidity of the Fe sites^37,38.

To evaluate the universality of in-situ LA site engineering by the induced piezopotential, we extended our investigation to three classic piezocatalysts BaTiO₃, BiTiO₃, and BiFeO₃. The intrinsic piezoelectricity of these piezocatalysts has been verified by PFM (Supplementary Fig. 11). Similar to FeOCl, all three materials exhibited a significant increase in the capacity of pyridine adsorption (by approximately 5.3, 1.3, and 2.2 times, respectively) under ultrasonic vibration (Supplementary Figs. 12 and 13). Systematic Py-IR analysis (Supplementary Fig. 14) revealed a notable enhancement in the Lewis acidity for all three materials after ultrasonic vibration. Quantitative analysis based on the characteristic peak at 1450 cm⁻¹ (Supplementary Fig. 15) further elucidated the change in LA sites. Following exposure to the mechanical stress, the derived LA site density on BaTiO₃ and BiTiO₃ across the desorption temperatures from 50 °C to 200 °C was all higher than that of the fresh samples. In the case of BiFeO₃, although the change in LA site density was negligible from the Py-IR spectrum at 200 °C, obvious increases were observed from that at 50 °C and 100 °C for the sample after ultrasonic vibration. These results demonstrate the general outcome of in-situ LA site engineering when the piezocatalyst is subjected to mechanical stress.

Piezo-self-Fenton activity for HO^• generation

Given above, the induced piezopotential on FeOCl nanobelt not only promotes the carrier dynamics but also enhances the Lewis acidity of surface Fe sites. Beyond the improvement of heterogeneous Fenton activity as previously reported, the in-situ regulation of LA sites on FeOCl also affect the reaction pathway. The activation and conversion of the O₂ molecule with weaker Lewis basicity than H₂O₂ is expected to be achieved on the FeOCl piezocatalyst.

The piezocatalytic activity of FeOCl for H₂O₂ generation was examined in pure water under ultrasonic vibration (100 W, 40 kHz). As shown in Fig. 3a, the H₂O₂ yield reached 1618.8 μmol g^⁻1 h^⁻1, which was comparable to the performance of some classic piezocatalysts^39,40. However, as the reaction progressed, a slowdown in H₂O₂ accumulation was observed. The H₂O₂ concentration reached the equilibrium at 1.3 mmol L⁻¹, with the H₂O₂ generation rate declining from 1618.8 μmol g^⁻1 h^⁻1 in the first hour to 853.6 μmol g^⁻1 h^⁻1 over the next eight hours. This result validates O₂ conversion into H₂O₂ on the FeOCl piezocatalyst, meanwhile suggesting the simultaneous H₂O₂ conversion. During the reaction, the solution pH decreased from ~6.7 to ~4.3 and then slightly increased to ~4.5 under ultrasonic vibration (Supplementary Fig. 16). The intrinsically acidic surface of FeOCl establishes a local proton-rich microenvironment, facilitating the protonation steps during the O₂ to H₂O₂ reduction⁴¹. With terephthalic acid (TA) as a fluorescent probe, the production of HO^• radicals by the FeOCl piezocatalyst was further quantified (Fig. 3b)⁴². Continuous HO^• generation (556.8 µmol g⁻¹ h⁻¹) was confirmed along with the application of mechanical stress. The performance represents the highest level to date (Fig. 3c). The long-term stability of the FeOCl piezo-self-Fenton system was confirmed through cyclic and extended-duration experiments. Under alternating ultrasound (ON for 20 min) and static (OFF for 20 min) conditions, HO^• were generated only during ultrasound exposure and ceased immediately when the ultrasound was turned off, confirming the reversible, piezopotential-driven radical generation. The total HO^• yield accumulated over three cycles (60 min in total) closely matched that obtained under continuous ultrasonic vibration for 1 h, demonstrating the consistent performance under intermittent operation (Supplementary Fig. 17). Furthermore, during the 4 h continuous ultrasonic vibration experiment, the concentration of HO^• exhibited sustained growth, verifying persistent catalytic performance (Supplementary Fig. 18). To be noted, Fe leaching during ultrasonic vibration was also monitored (Supplementary Fig. 19). The concentration was approximately 135 μg L^⁻1, which was insufficient to generate HO^• radicals at such a level through the homogeneous Fenton reaction. To further confirm that the enhanced activity arises from piezopotential-induced surface modulation rather than other ultrasound-related contributions, control experiments were conducted. HO^• generation under ultrasound alone (20.05 μmol L^-1), using leached Fe³⁺ (27.72 μmol L^-1), and from the pyridine-poisoned FeOCl (20.92 μmol L^-1) were all markedly lower than that achieved on FeOCl (111.36 μmol L^-1) (Supplementary Fig. 20). These results confirm that the enhanced activity originates from the in-situ modulation of Lewis acid sites on FeOCl by the piezopotential, rather than from cavitation or Fe leaching. In-situ electron paramagnetic resonance (EPR) analysis was also performed using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as a trapping agent to monitor the evolution of HO^• radicals (Fig. 3d and Supplementary Fig. 21). As shown in Fig. 3e, f, and Supplementary Fig. 22, the signal of DMPO-HO^• adduct emerged at about 36 s and became stable under ultrasonic vibration. Considering the limited lifetime of spin adducts, these results offer additional evidence for the sustained production of HO^• radicals over the FeOCl piezocatalyst with LA sites.

Fig. 3. Piezo-self-Fenton activity of FeOCl.

a Piezocatalytic H₂O₂ generation over FeOCl. b Piezocatalytic HO^• generation over FeOCl. Reaction conditions: [catalyst] = 0.2 g L⁻¹, [TA] = 10 mM, Ultrasonic power = 100 W. c Comparison of the HO^• production rate over the reported state-of-the-art heterogeneous Fenton and self-Fenton systems. The relevant references are listed in Supplementary Table 1 and 2. d Schematic diagram of in-situ EPR system and hyperfine splitting diagrams of the DMPO-HO^• species adduct detected by EPR. e In-situ EPR spectra probing HO^• radicals generated by the FeOCl piezocatalyst. f Quantitative analysis of HO^• radicals derived from the in-situ EPR measurements. Reaction conditions: [Catalyst] = 1 g L⁻¹, V = 5 mL (Ultrapure water), Ultrasonic power = 100 W, V_{capture agent} = 100 μL. g Quenching and atmosphere-controlled experiments of piezocatalytic HO^• generation over FeOCl. Reaction conditions: [catalyst] = 0.2 g L⁻¹, [TA] = [MeOH] = [TBA] = [BQ] = 10 mM, Ultrasonic power = 100 W, [Reaction time] = 1 h. h Schematic illustration for piezocatalytic HO^• generation over FeOCl. The error bars represent the standard deviation, calculated from at least two independent experiments.

The heterogeneous Fenton activity of FeOCl was also evaluated. For H₂O₂ decomposition over FeOCl (Supplementary Fig. 23), the concentration of H₂O₂ decreased from 5 mM to 1.1 mM over ten hours. This process was accelerated under ultrasonic vibration. The Fenton activity was assessed using 3,3,5,5-tetramethylbenzidine dihydrochloride (TMB) as the substrate. As shown in Supplementary Fig. 24, under ultrasonic vibration, the increasing absorbance of oxidized TMB (ox-TMB) at 652 nm implied the continuous Fenton reaction in the absence of H₂O₂^43,44. Supplementary Fig. 25 shows that the reaction in the presence of H₂O₂ was significantly enhanced under ultrasonic vibration, with the kinetics of the reaction accelerating, further supporting the promotion of the heterogeneous Fenton activity by mechanical stress⁴⁵.

Collectively, efficient O₂ conversion into H₂O₂ and the subsequent heterogeneous Fenton reaction were realized over FeOCl under ultrasonic vibration. With the in-situ regulation of surface LA sites by the induced piezopotential, the heterogeneous Fenton activity of FeOCl catalyst was upgraded into a more advanced piezo-self-Fenton activity⁴⁶. Bypassing the H₂O₂ input endows the FeOCl piezo-self-Fenton system with more possibilities in practical application.

Mechanism of the piezo-self-Fenton reaction

Tracing the reaction pathway elucidated the catalytic mechanism driven by the piezopotential induced regulation of LA sites. Various scavengers were employed to probe the generation of H₂O₂ and HO^• by the FeOCl piezo-self-Fenton system. Specifically, p-benzoquinone (BQ) was used to scavenge the superoxide radicals (O₂^•⁻), tert-butanol (TBA) was used to scavenge the HO^• radicals, and ethylenediaminetetraacetic acid disodium salt (EDTA-2Na) was used to scavenge holes (h⁺). As shown in Supplementary Fig. 26, H₂O₂ generation in the FeOCl piezo-self-Fenton system was inhibited by 87% in the presence of BQ but slightly affected by the introduction of TBA and EDTA-2Na. Furthermore, the experiment under N₂ atmosphere revealed significant inhibition of H₂O₂ production. These results indicated that the two-electron reduction of O₂ was responsible for H₂O₂ generation. In-situ EPR analysis monitoring the O₂^•⁻ radicals was also performed. Under ultrasonic vibration, the signal of DMPO- O₂^•⁻ adduct was readily detected, increasing rapidly and reaching the equilibrium (Supplementary Figs. 27 and 28). Thus, the O₂^•⁻ radicals are critical intermediates in O₂ conversion and the subsequent Fenton reaction. Driven by the enhanced LA sites, the activation of the Lewis basic O₂ molecule represents a key initial step in this pathway, promoting its electron transfer and reduction. The result of quenching experiments focusing on the production of HO^• radicals is shown in Fig. 3g. Similar to the case of H₂O₂ generation, the production of HO^• radicals was significantly inhibited in the presence of BQ or under the N₂ atmosphere. When MeOH was introduced into the FeOCl piezo-self-Fenton system, the efficiency was barely affected. However, HO^• production was suppressed by 83.88% along with the addition of TBA. This result may be ascribed to the different affinities of MeOH and TBA toward the surface of FeOCl. TBA with a strong affinity to the surface may quench the newly-produced HO^• radicals at the solid-liquid interface, while MeOH with an affinity poorer than TBA may eliminate the HO^• radicals diffused into the bulk solution. The result suggests that the HO^• radicals generated by the FeOCl piezo-self-Fenton system were centered at the solid-liquid interface. Since most of the HO^• radicals in heterogeneous Fenton systems undergo spontaneous deactivation during diffusion, the near-surface HO^• radicals produced by the FeOCl piezo-self-Fenton system may effectively participate in the oxidation of pollutants (Fig. 3h)^47,48.

Systematic DFT calculations provided additional insights into the role of LA sites in piezocatalytic mechanism. Starting from the interaction between the FeOCl piezocatalyst and the O₂ molecule, the evolution of charge density difference and adsorption energy is exhibited in Fig. 4a, b. At ambient pressure, the interaction between FeOCl piezocatalyst and the O₂ molecule resulted in a 0.17 e of electron transfer from FeOCl to O₂ as well as an O₂ adsorption energy of −0.69 eV. Along with the application of stress (2 GPa), a more significant electron transfer (0.25 e) was observed, accompanied by a more favorable adsorption energy (−2.55 eV), demonstrating the enhanced capability of LA sites under piezopotential to activate O₂. These results imply that the electron donation from FeOCl to O₂ could be enhanced by the application of mechanical stress, facilitating O₂ adsorption and activation. Crystal Orbital Hamiltonian Population (COHP) analysis provides additional support (Fig. 4c). The interaction between FeOCl and the O₂ molecule yielded a more negative ICOHP under 2 GPa than at ambient pressure, which suggests a stronger Fe-O₂ interaction, further corroborating the role of LA sites in strengthening the adsorbate-catalyst interaction. Similar promotion in the adsorption and activation of H₂O₂ was also verified. For the interaction between FeOCl and the H₂O₂ molecule (Supplementary Figs. 29 and 30), an electron transfer of 0.06 e from FeOCl to H₂O₂, as well as a H₂O₂ adsorption energy of −0.12 eV, was obtained at ambient pressure. These values changed into 0.1 e and −0.98 eV, respectively, under 2 GPa. Meanwhile, corresponding COHP calculations (Supplementary Fig. 31) also yielded a more negative ICOHP under 2 GPa compared to the ambient pressure⁴⁹. Based on these findings, the Gibbs free energy diagram for the consecutive O₂ conversion into H₂O₂ and HO^• radical over the FeOCl piezocatalyst was simulated (Fig. 4d). Key elementary steps include: (i) O₂ adsorption; (ii) *O₂ protonation and reduction into *OOH; (iii) *OOH protonation and reduction into *H₂O₂; and (iv) *H₂O₂ reduction into HO^• radical⁵⁰. Under the application of mechanical stress, consecutive O₂ conversion on the FeOCl piezocatalyst was thermodynamically favorable, exhibiting reaction barriers lower than those under ambient pressure in each elementary step. Overall, steps (i)-(iv) define a consecutive three-electron reduction pathway from O₂ to HO^•. Further electrochemical validation using rotating ring‑disk electrode (RRDE) measurements (Supplementary Fig. 32) shows that the treatment of ultrasonic vibration substantially enhances the O₂ reduction current on FeOCl and shifts the electron transfer number toward 3, confirming the feasibility of efficient O₂ to HO^• conversion via this pathway. Based on the analyses above, the LA sites on FeOCl, modulated in-situ by the induced piezopotential, facilitated efficient O₂ activation and consecutive conversion (Fig. 4e).

Fig. 4. Piezocatalytic mechanisms.

a Charge density difference between FeOCl and the adsorbed O₂ molecule at 0 GPa and 2 GPa. b O₂ adsorption energy and the charge transfer between FeOCl and O₂ at 0 GPa and 2 GPa. c COHP diagrams between the Fe atom and the O atom for FeOCl and O₂ at 0 GPa and 2 GPa. d Gibbs free energy profiles for O₂ activation and HO^• formation on FeOCl at 0 GPa and 2 GPa. e Schematic illustration of consecutive O₂ conversion over FeOCl under ultrasonic vibration.

Piezo-self-Fenton system for antibiotic-laden wastewater treatment

The potential of the FeOCl piezo-self-Fenton system in wastewater treatment was fully examined, using sulfamethazine (SMZ) as a representative model pollutant. As shown in Fig. 5a, the FeOCl piezo-self-Fenton system demonstrated 99% SMZ degradation within 60 min under ultrasonic vibration. This performance was even better than that of the FeOCl/H₂O₂ heterogeneous Fenton system, in which the H₂O₂ concentration of 15 and 20 mM resulted in SMZ degradation of 60% and 93%, respectively. This superior activity underscores the efficiency of the LA sites in driving the reaction pathway. For SMZ degradation, the FeOCl piezo-self-Fenton system outperformed the heterogeneous Fenton systems employing classic Fe-based catalysts (Fe₂O₃, Fe₃O₄, CoFe₂O₄, NiFe₂O₄, ZnFe₂O₄, CuFe₂O₄, Fe-BC, Fe-C₃N₄) and was comparable to the Fe²⁺/H₂O₂ homogeneous Fenton system (Fig. 5b and Supplementary Fig. 33). The level of total organic carbon (TOC) removal by the FeOCl piezo-self-Fenton system was also considerable in comparison with that of the FeOCl/H₂O₂ heterogeneous Fenton system and the Fe²⁺/H₂O₂ homogeneous Fenton system (Fig. 5c). This superior performance originates from the piezopotential-driven self-regulation of surface Lewis acid sites, which facilitates the efficient consecutive conversion of O₂ to H₂O₂ and then to hydroxyl radicals (HO^•). The generated HO^• radicals are primarily responsible for the rapid and complete degradation of SMZ (Supplementary Fig. 34).

Fig. 5. FeOCl piezo-self-Fenton system for water purification.

a SMZ degradation by the FeOCl piezo-self-Fenton and FeOCl/H₂O₂ systems. b Comparison of the degradation efficiency of SMZ by the FeOCl piezo-self-Fenton, heterogeneous Fenton systems with different Fe-based catalysts, and Fe²⁺/H₂O₂ systems. c Comparison of the TOC removal of SMZ by the FeOCl piezo-self-Fenton, FeOCl/H₂O₂, and Fe²⁺/H₂O₂ systems. Reaction conditions: [catalyst] = 0.5 g L⁻¹, [Pollutant] = 10 mg L⁻¹, [Reaction time] = 1 h. d Life cycle assessment evaluating the relative environmental impact for treating 1 kg SMZ in the FeOCl piezo-self-Fenton, nZVI heterogeneous Fenton, and homogeneous Fenton systems. e Degradation of various organic pollutants (tetracycline (TC), diclofenac (DCF), phenol (PhOH), sulfamethoxazole (SMX), carbamazepine (CBZ), ibuprofen (IBU), norfloxacin (NOR), and ciprofloxacin (CIP) by the FeOCl piezo-self-Fenton system. Reaction conditions: [catalyst] = 0.5 g L⁻¹, [Pollutant] = 10 mg L⁻¹, Ultrasonic power = 100 W, [Reaction time] = 1 h. f Fixed-bed continuous-flow operation for pollutant removal by the FeOCl@GAC piezo-self-Fenton system. The FeOCl catalyst loaded onto granular activated carbon (GAC) by impregnation and calcination, then loaded into a fixed-bed column. g Continuous-flow operation for pollutant removal over 13 h by the FeOCl piezo-self-Fenton system. h EEM spectra of collected pharmaceutical wastewater before and after treatment. US denotes ultrasonic vibration. The error bars represent the standard deviation, calculated from at least two independent experiments.

The comparable SMZ degradation performance of the FeOCl piezo-self-Fenton system to that of the traditional homogenous and heterogeneous Fenton system highlights its remarkable potential for practical application. To further assess its sustainability and environmental impacts, a life cycle assessment (LCA) was conducted, and the FeOCl piezo-self-Fenton process was quantitatively compared with the Fe²⁺/H₂O₂ homogeneous Fenton system and the classic nanoscale zero-valent iron (nZVI) heterogeneous Fenton system (nZVI/H₂O₂) (Supplementary Fig. 35 and Supplementary Table S3). The entire wastewater treatment process at the laboratory scale and the key impacts relevant to current wastewater treatment processes, such as water consumption, resource scarcity, ecotoxicity, and eutrophication, were considered^51,52. As depicted in Fig. 5d, the FeOCl piezo-self-Fenton system demonstrated the lowest environmental impact scores across all impact categories among the three systems. In addition, transportation can also be reduced since the process does not require H₂O₂ input. Therefore, the FeOCl piezo-self-Fenton system provides distinct environmental advantages over the FeOCl/H₂O₂ heterogeneous Fenton system and the Fe²⁺/H₂O₂ homogeneous Fenton system.

The practicability of the FeOCl piezo-self-Fenton system was further evaluated. After five cycles, the system maintained an SMZ degradation efficiency of 90% (Supplementary Fig. 36). The performance of the FeOCl piezo-self-Fenton system was barely affected by the anions commonly found in realistic scenarios (Supplementary Fig. 37), showing robust tolerance to the interfering anions even at high concentrations. The pH sensitivity of the system was also validated across a wide pH range of 3–9 (Supplementary Fig. 38). Particularly in near-neutral aqueous solutions, the system demonstrated superior degradation performance. Additionally, the versatility of the system was evident in the degradation of various organic micropollutants, including tetracycline (TC), diclofenac (DCF), phenol (PhOH), sulfamethoxazole (SMX), carbamazepine (CBZ), ibuprofen (IBU), norfloxacin (NOR), and ciprofloxacin (CIP) (Fig. 5e and Supplementary Table S4). In the control experiments, the degradation achieved under ultrasonic vibration alone was minimal, in contrast to the high efficiencies of the FeOCl piezo-self-Fenton system (Supplementary Figs. 39 and 40). Scalable preparation of FeOCl-loaded granular activated carbon (FeOCl@GAC) catalysts was then performed (Supplementary Figs. 41 and 42), and a continuous-flow reaction device was constructed to evaluate the potential of this system for practical applications. Figure 5f shows that the FeOCl@GAC piezo-self-Fenton system achieved a gradual reduction in the chromaticity of pharmaceutical wastewater after treatment, with a high TOC removal percentage of 30.17% (Supplementary Fig. 43). As shown in Fig. 5g, the degradation efficiency of SMZ remained above 99% for 13 hours with a 5 L input of SMZ solution. In the treatment of pharmaceutical wastewater, the excitation-emission matrix (EEM) spectra (Fig. 5h) revealed that the soluble microbial byproducts and humic acid-like organic matter were effectively removed, with a significant reduction in the fluorescence intensity of the detected peaks. These results demonstrated the great potential of the piezo-self-Fenton system with in-situ LA site engineering in practical wastewater purification⁵³.

The self‑regulating phenomenon reported here is fundamentally driven by mechanical stress. This mechanism offers opportunities for implementation using widely available, low-frequency mechanical energy, such as water flow or ambient vibration. Future efforts will therefore focus on designing flow-or pressure-driven reactor systems to achieve this piezo-self-Fenton process for practical, energy-efficient water treatment applications.

Discussion

In summary, systematic experimental investigations coupled with theoretical analyses have fully validated the in-situ LA site engineering as a new feature achieved by the induced piezopotential, which is suitable for the common piezocatalysts. Specifically for the FeOCl piezocatalyst, the induced piezopotential realized the enhancement of surface Lewis acidity except for the promotion of carrier dynamics. As a result, the activation and conversion of the O₂ molecule, whose Lewis basicity is weaker than H₂O₂, is realized on the LA sites of FeOCl piezocatalyst. The heterogeneous Fenton activity of FeOCl catalyst is upgraded into a more advanced piezo-self-Fenton activity, producing HO^• efficiently under ultrasonic vibration (556.8 µmol g⁻¹ h⁻¹). For SMZ degradation, the performance of the FeOCl piezo-self-Fenton was comparable to the FeOCl/H₂O₂ heterogeneous Fenton system and the Fe²⁺/H₂O₂ homogeneous Fenton system. Moreover, the FeOCl piezo-self-Fenton system demonstrated unparalleled broad-spectrum degradation efficiency against diverse organic pollutants, as validated through both simulated wastewater models and real-world antibiotic-laden effluents. This work offered insightful viewpoints for the structure-property relationship of piezocatalysts as well as the optimization of high-performance piezocatalytic systems.

Methods

Piezocatalytic H₂O₂ production

All the procedures were consistent if not specified. Generally, an open 300 mL beaker was used as the reactor, positioned above the ultrasonic transducer in a KS-2200DB ultrasonic cleaner (Kunshan Jielimei, China). The power and frequency of the equipment were 100 W and 40 kHz, respectively. The temperature of the water bath was controlled at room temperature by running tap water, and the water level was equal to the liquid level in the beaker. The system was stirred at 400 rpm by a JJ-1B-200W agitating machine (Xinrui, China) during the whole test. 20 mg of FeOCl catalyst was dispersed into 100 mL of deionized water under stirring. The initial pH of the solution was ~6.7 (unadjusted). Then the ultrasonic vibration was exerted on the suspension. The concentration of H₂O₂ was determined by the iodide spectrophotometric method. For details, 1 mL of freshly prepared KI reagent A (0.4 M KI, 0.05 M NaOH, 1.6 × 10⁻⁴ M (NH₄)₆Mo₇O₂₄·4H₂O) and 1 mL of reagent B (0.1 M KHC₈H₄O₄) were mixed with 1 mL of the above samples.

HO^• measurement and quantification

The HO^• radicals were quantified using a fluorescence method with terephthalic acid (TA) as the probe, which could react with HO^• to generate 2-hydroxyterephthalic acid, displaying strong fluorescence at 430 nm. The 2-hydroxyterephthalic acid concentration was measured by a fluorescence spectrophotometer (Hitachi F-4600) with an excitation wavelength of 315 nm.

All the procedures were consistent if not specified. Generally, an open 300 mL beaker was used as the reactor, positioned above the ultrasonic transducer in a KS-2200DB ultrasonic cleaner (Kunshan Jielimei, China). The power and frequency of the equipment were 100 W and 40 kHz, respectively. The temperature of the water bath was controlled at room temperature by running tap water, and the water level was equal to the liquid level in the beaker. The system was stirred at 400 rpm by a JJ-1B-200W agitating machine (Xinrui, China) during the whole test. 20 mg of FeOCl catalyst was dispersed into 100 mL of 10 mM terephthalic acid solution under stirring. The reaction was conducted at the initial solution pH of ~5.6 (unadjusted). Then the ultrasonic vibration was exerted on the suspension. The concentration of HO^• was determined by a fluorescence spectrophotometer (Hitachi F-4600) with an excitation wavelength of 315 nm.

Piezo-self-Fenton degradation of SMZ

All the procedures were consistent if not specified. Generally, an open 300 mL beaker was used as the reactor, positioned above the ultrasonic transducer in a KS-2200DB ultrasonic cleaner (Kunshan Jielimei, China). The power and frequency of the equipment were 100 W and 40 kHz, respectively. The temperature of the water bath was controlled at room temperature by running tap water, and the water level was equal to the liquid level in the beaker. The system was stirred at 400 rpm by a JJ-1B-200W agitating machine (Xinrui, China) during the whole test. 50 mg of FeOCl catalyst was dispersed into 100 mL 10 mgL^-1 SMZ solution under stirring. The initial pH of the SMZ solution was ~6.3 (unadjusted). Samples were periodically withdrawn, quenched with excess Na₂S₂O₃ to terminate the catalytic reaction, and filtered using a 0.22 µm fiber filter. The concentration of organic pollutants was determined by a liquid chromatograph (HPLC, Shimadzu LC-16 with a C18 column, Poroshell 120 SB-C18, 100 mm × 2.1 mm, 2.7 µm) equipped with a UV detector. The removal efficiency was represented by the relative concentration (C/C₀). C₀ and C are the initial and reaction concentrations.

Trials for wastewater treatment

20 g of FeOCl@GAC catalyst was pre-loaded into the U-tube of the continuous-flow reaction device, and the collected pharmaceutical wastewater was pumped through a peristaltic pump. After that, ultrasonic vibration was turned on, and 5 mL of treated wastewater was withdrawn at 1-h intervals, and filtered through a 0.22 μm membrane needle filter for further analysis. Finally, the TOC and EEM spectra of wastewater were used to evaluate the practicality of the FeOCl piezo-self-Fenton system.

Life cycle assessment

Life cycle assessment (LCA) is a standardized approach for evaluating and comparing environmental performance among various products and systems^54,55. Here, we applied LCA using the openLCA (version 2.2.0) to compare the environmental impacts of three Fenton systems (i.e., FeOCl piezo-self-Fenton system, nZVI heterogeneous Fenton system, homogeneous Fenton system) for treating 1 kg SMZ. The foreground inventory of materials (e.g., FeSO₄, H₂O₂, HCl, FeCl₃·6H₂O), electricity, transport, and waste emissions in the three systems were estimated based on our experimental data and literature⁵⁶. The background inventory was selected from the ecoinvent database (version 3.8).

Author contributions

M.Z. designed the research. H.D. completed most of the experiments and theoretical calculations. M.Z. and Y.Z. carried out analysis and wrote the paper. W.L., Y.Z. and Z.L. helped with analysis and characterizations. S.L. performed life cycle assessment. M.Z. supervised the project. All authors discussed the results and commented on the paper.

Peer review

Peer review information

Nature Communications thanks Indrajit Sinha, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The data supporting the findings of this study are included within the main text, the Supplementary Information and the Source data files. All data are available from the corresponding author upon request. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70327-0.