Internal electric field steering S-scheme charge transfer in ZnIn2S4/COF boosts H2O2 photosynthesis from water and air for sustainable disinfection

Ensuring the provision of clean and safe water is a fundamental necessity for all human beings^1,2,3. Regrettably, a considerable segment of the world’s population still faces the challenge of not having access to safe drinking water and adequate sanitation facilities. In response to the urgent global water and sanitation crisis, the United Nations (UN) has established Sustainable Development Goal 6 (SDG 6) within the framework of the 2030 Agenda for Sustainable Development, with the objective of ensuring universal access to water and sanitation and promoting their sustainable management^4,5. To achieve this objective, disinfection plays a critical role in eliminating or inactivating pathogens present in water sources. However, conventional water disinfection methods, such as chlorine and ultraviolet disinfection, have been associated with excessive energy consumption, microbial resistance, and the generation of persistent environmental pollution from by-products^6,7.

In recent years, Fenton technology has emerged as a promising alternative for water disinfection, owing to its environmentally friendly attributes^8,9. The effectiveness of Fenton disinfection systems hinges on the involvement of hydrogen peroxide (H₂O₂). Nevertheless, the traditional approach to anthraquinone-based H₂O₂ production is accompanied by notable drawbacks, including substantial energy consumption, utilization of hazardous solvents, and the formation of toxic by-products¹⁰. Additionally, the transportation and storage of H₂O₂ also entail considerable safety concerns that warrant serious attention.

Photocatalytic in situ production of H₂O₂ has garnered significant attention lately due to its inherent selectivity, mild reaction conditions, renewable energy utilization, catalyst recyclability, and scalability^{11,12,13,14,15,16,17,18,19,20,21}. Through the utilization of photosynthesis, this method enables the generation of H₂O₂ exclusively from water and air, without the necessity of sacrificial agents. As a result, it holds great promise as an ideal approach for green and sustainable strategies. Despite significant research efforts dedicated to inorganic photocatalysts like TiO₂, ZnO, and CdS for H₂O₂ photosynthesis, their practical implementation has encountered several obstacles, including limited visible light absorption, rapid charge recombination, inadequate selectivity and exhibit suboptimal efficiency in the production of H₂O₂. To address those obstacles, the scientific community has pivoted its focus toward the exploration of organic photocatalysts. Within the realm of organic photocatalysts, covalent organic frameworks (COFs) have surfaced as one of the most promising options and have demonstrated potential across a spectrum of applications²². COFs represent a novel category of porous organic materials interconnected through robust covalent bonds and have garnered considerable interest due to their unique structural and functional properties, such as exceptional specific surface area, remarkable thermal and chemical stability, as well as superior light absorption capabilities^{23,24,25,26,27,28,29,30}.

While COFs exhibit favorable characteristics that address the limitation of traditional photocatalysts, they still suffer from prompt recombination of photogenerated carriers, resulting in unsatisfactory performance^{31,32,33,34,35,36,37,38}. A viable strategy to enhance the separation of charged carriers in COFs involves the addition of sacrificial agents³⁹. For example, Wang et al. employed ethanol (EtOH) as a sacrificial agent to curb carrier recombination and facilitated the release of protons for photocatalytic H₂O₂ production⁴⁰. Similarly, Kong et al. developed a benzyl alcohol-water system to avoid the decomposition of H₂O₂, with benzyl alcohol serving as a sacrificial agent to improve the utilization rate of electron-holes (e⁻-h⁺)⁴¹. However, it is essential to consider that the use of sacrificial reagents can conflict with the principles of green and sustainable chemistry. Consequently, researchers have been investigating alternative approaches to improve the efficiency of COFs.

One such strategy that has shown promise involves the construction of S-scheme heterojunctions by coupling inorganic and organic photocatalysts, which effectively suppress charge recombination and enhance photocatalytic performance^42,43,44. Among various inorganic photocatalysts, ZnIn₂S₄ stands out as a superior choice for constructing S-scheme heterojunctions with COFs in the context of H₂O₂ photosynthesis. This is attributed to its suitable band alignment, excellent light capture ability, robust stability, and non-toxic properties^45,46. Importantly, the conduction band (CB) and Fermi level (E_f) of ZnIn₂S₄ possess more negative values relative to most COFs, thereby enabling the establishment of an efficient S-scheme heterojunction. This facilitates the generation of an internal electric field (IEF), propelling the separation of useful carriers with enhanced redox potential in the ZnIn₂S₄/COF composite while minimizing recombination losses. Although some heterostructures have been exploited for photocatalytic purposes, accurately determining and quantifying the driving force behind photogenerated carrier separation and unraveling the electron migration pathways between the components remain complex challenges in the field.

In this work, TaTp was selected as the COF component of heterojunction owing to its abundant active sites and superior light absorption properties. These characteristics are expected to accelerate the adsorption and activation of reactive substances and produce more electrons, thus facilitating their migration to other components. The heterojunction photocatalyst was constructed by hybridizing TaTp with ZnIn₂S₄ and employed for the in situ inactivation of Escherichia coli (E. coli) through the utilization of generated H₂O₂. Experimental results indicated that without the addition of H₂O₂, the photocatalytic Fenton system completely inactivated 2 × 10⁷ colony forming unit (cfu) mL⁻¹ E. coli by incorporating in situ synthesized H₂O₂ with Fe (II). The S-scheme transport paths of the carriers have been unveiled through in situ X-ray photoelectron spectroscopy (in situ XPS), Kelvin probe force microscopy (KPFM) and electron paramagnetic resonance (EPR). Moreover, the primary mechanism for enhancing the efficiency of carrier separation and prolonging the carrier lifetime was elucidated through the ultraviolet photoelectron spectroscopy (UPS) test, along with the quantified analysis and calculation of the IEF intensity. In conclusion, this research establishes both the theoretical and experimental groundwork for the evolution and development of S-scheme heterojunction. It expands the scope of COF-based photocatalyst studies in the eco-friendly synthesis of H₂O_2, realizing in situ activation and utilization, as well as affording a referable strategy for green, sustainable, and efficient water disinfection technology.

Structure and morphology analyses of photocatalysts

As shown in Fig. 1a, we fabricated the heterojunction photocatalyst via the post-modification method (ZnIn₂S₄ is denoted as ZIS, and the composite is denoted as x% ZIS/TaTp). Subsequently, the crystal structure and internal chemical bonds of the catalysts were identified through X-ray diffraction (XRD) and Fourier-transform infrared spectroscopy (FTIR)⁴⁷. As presented in Fig. 1b, four characteristic diffraction peaks were observed at 5.6°, 9.7°, 14.6° and 26.8° for TaTp⁴⁸. ZIS showed obvious characteristic diffraction peaks at 21.6°, 27.7°, 47.2°, and 52.4°, belonging to the (006), (102), (110), and (116) crystal facets, respectively⁴⁴. Notably, the diffraction peaks of TaTp and ZIS exist simultaneously in the XRD pattern of ZIS/TaTp nanocomposite, without miscellaneous peaks, which indicated that ZIS was resoundingly combined with TaTp. In Fig. 1c, the peak at 1581 cm⁻¹ represented the successful formation of imine (C=N) linkages in TaTp, while the peaks at 1508 cm⁻¹ and 1362 cm⁻¹ confirm the presence of C=N-C bonds in the triazine^48,49. For ZIS, only two peaks located at 1396 cm⁻¹ and 1610 cm⁻¹ were detected, demonstrating the H₂O molecules and hydroxyl groups taken on the surface of the photocatalyst⁵⁰. The characteristic peaks representing TaTp are discernible in the FTIR spectra of ZIS/TaTp, which proves that the structure of TaTp was not damaged during the compounding process.

Illustration of ZIS/TaTp synthesis (a), XRD (b) and FTIR (c) patterns of photocatalysts. SEM images of TaTp (d), ZIS (e), and 20% ZIS/TaTp (f); TEM images of 20% ZIS/TaTp (g–i).

Full size image

The microstructure of TaTp, ZIS, and 20% ZIS/TaTp was examined using the scanning electron microscope (SEM) and transmission electron microscope (TEM). As illustrated in Fig. 1d, TaTp displayed a fibrous network structure, while ZIS reflected a micro-meter scale flower-like morphology assembled by irregular nanosheets (Fig. 1e). According to Fig. 1f, ZIS was intimately composited with TaTp in a lamellar structure. The absence of a flower-like configuration in the 20% ZIS/TaTp composite can be attributed to the in situ growth of ZIS onto TaTp through the combined method. The inclusion of TaTp disrupted the orderly assembly of ZIS into the flower-like structure, which precisely indicated the resounding preparation of the hybrid photocatalyst⁵⁰. The TEM observation (Fig. 1g) illustrated the close integration of ZIS and TaTp. Furthermore, the lattice fringe of ZIS, discernible in Fig. 1h and i, had a distinct lattice fringe distance of approximately 0.32 nm, corresponding to the (102) planes of ZIS⁴⁵. The analyses from both SEM and TEM revealed that ZIS and TaTp have been successfully merged into a heterogeneous structure. This tight heterojunction interface is a prerequisite for subsequent electron transport and forms the basis for enhancing photocatalytic performance. The N₂ adsorption and desorption isothermal curves for TaTp, ZIS, and 20% ZIS/TaTp were examined at 77 K to ascertain the specific surface area, with all samples showing type IV adsorption isotherms (Supplementary Fig. 1). Upon binding with ZIS, a minor reduction in the BET surface area and porosity of TaTp was observed (Table S1). This is primarily due to the mesoporous structure of TaTp being able to accommodate ZIS, resulting in the lamellar structure of ZIS partially blocking the pores of TaTp. This occlusion provides additional surface active sites on the catalyst, thereby facilitating the adsorption of reactants and intermediates.

Determining band structures of photocatalysts

UV-Vis diffuse reflectance spectroscopy (DRS) was examined to analyze the light-harvesting capability of the photocatalyst. A noteworthy red-shift in the absorption range was observed for ZIS/TaTp compared to pristine TaTp (Fig. 2a). This indicates that the heterostructures have a superior capacity to broaden the spectrum of light absorption, which is beneficial to improving the photocatalytic performance. According to the Kubelka-Munk equation, the band gaps of TaTp and ZIS were calculated to be 2.28 eV and 2.42 eV, respectively (Supplementary Fig. 2)⁴⁷. As shown in Supplementary Fig. 3, the Mott-Schottky test was conducted to analyze the flat band potential (E_fb) of the prepared photocatalyst. Given that the conduction band (CB) potential is approximately 0.2 V more positive than E_fb, the CB potential of TaTp and ZIS was determined to be −0.58 V and −1.1 V (vs. NHE, pH = 7), respectively. The valence band (VB) potentials of the catalyst were calculated to be 1.7 V and 1.32 V (vs. NHE, pH = 7) based on the formula E_g = E_VB − E_CB³⁵. The band structure alignments of the prepared catalysts are depicted in Fig. 2b. In light of these characterizations and analyses, it can be concluded that the ZIS/TaTp heterostructures significantly extend the photo-absorption region, thus stimulating more photogenerated carriers for the photocatalytic reaction.

UV-Vis diffuse reflection spectra of photocatalysts (a), band alignment of TaTp and ZIS (b), transient absorption spectroscopy of TaTp (c) and 20% ZIS/TaTp (d), decay kinetics of TaTp (e) and 20% ZIS/TaTp (f).

Full size image

Photoelectrochemical properties of photocatalysts

Transient absorption spectroscopy (TAS) was performed to investigate the effective separation and kinetic relaxation of photocatalysts. According to the band structure of the synthesized catalyst, 320 nm was selected as the testing pump pulse light source. Both TaTp and 20% ZIS/TaTp demonstrated positive absorption peaks in the region of 450–750 nm, attributed to the trapped electrons (Fig. 2c, d). Decay kinetic spectra were fitted with a two-exponential function. The average relaxation lifespan (τ_avg) of the carriers in 20% ZIS/TaTp was found to be longer than that of the TaTp carriers (Fig. 2e, f). The lifetime of carriers was distinctly prolonged from 92 ps in TaTp to 243 ps in 20% ZIS/TaTp, chiefly due to the efficient transfer of photogenerated e⁻ within the ZIS to TaTp. These findings suggest that the heterojunction of ZIS and TaTp markedly improved the separation of carriers and effectively extended the relaxation lifetime of photogenerated e⁻, implying a kinetic enhancement of the oxygen reduction reaction (ORR).

The charge transport characteristics of the photocatalysts were evaluated using electrochemical impedance spectroscopy (EIS) and photocurrent measurements³⁵. As illustrated in Fig. 3a, the Nyquist plots of the composite exhibited a smaller radius than that of TaTp. This indicates a significant reduction in carrier migration resistance after the combination with ZIS, leading to the production of more utilizable e⁻ in the photocatalytic reaction. The photocurrent spectra in Fig. 3b revealed that 20% ZIS/TaTp exhibited the highest photocurrent intensity among the photocatalysts, proposing that more photogenerated e⁻-h⁺ pairs were generated in the composites to construct the overall photocatalytic system. The results obtained from both EIS and photocurrent measurements consistently illustrate that 20% ZIS/TaTp possesses exceptional charge transfer capability, thereby enhancing photocatalytic performance. Furthermore, photoluminescence (PL) spectroscopy was performed to measure the interfacial e⁻-h⁺ pairs separation ability^38,47. In Supplementary Fig. 4, the PL intensity of 20% ZIS/TaTp was significantly lower than that of pristine TaTp, revealing that the ZIS and TaTp composites can greatly expedite the separation of photogenerated carriers. Under an argon (Ar) atmosphere, the linear scanning voltammetry (LSV) curves of 20% ZIS/TaTp showed minimal reaction currents, whereas the reduction currents were considerably stronger in an air atmosphere (Supplementary Fig. 5). Notably, in Fig. 3c, 20% ZIS/TaTp demonstrated the highest intensity of reduction current compared to TaTp and ZIS single-component catalysis⁴⁷, manifesting that 20% ZIS/TaTp has the most effective ORR performance.

Electrochemical impedance plots (a), photocurrent curves of as-prepared catalysts (b), linear scanning voltammetry curves of TaTp, ZIS, and 20% ZIS/TaTp under air (c), photosynthesis of H₂O₂ activity by photocatalysts (d), control and trapping experiments of the catalysts on the photosynthesis of H₂O₂ (e), the in situ FTIR of 20% ZIS/TaTp for H₂O₂ synthesis (f).

Full size image

Photocatalytic performance

The photocatalytic performance of the as-prepared catalysts was investigated by the photosynthesis of H₂O₂ under visible light irradiation in H₂O without the need for sacrificial reagents or photosensitizers. This reaction system is economical and environmentally friendly, aligning with the principles of sustainable development and green chemistry. As shown in Fig. 3d, the H₂O₂ production rate of TaTp and ZIS were found to be 425 µmol∙g⁻¹ ∙ h⁻¹ and 82 µmol∙g⁻¹ ∙ h⁻¹, respectively. With the increase of ZIS loading amount, the photosynthetic activity of H₂O₂ presented a tendency of initial increase followed by a decrease^35,37. A suitable combination ratio of ZIS with TaTp can improve the photoelectrochemical properties, thereby enhancing the photocatalytic performance. However, excessive loading of ZIS may lead to the formation of charge recombination centers, which can be detrimental to the photocatalytic reaction. Specifically, the H₂O₂ photosynthesis over 20% ZIS/TaTp reached 1325 µmol∙g⁻¹ ∙ h⁻¹, which is 3.12 times and 16.2 times superior to that of TaTp and ZIS, respectively. Moreover, 20% ZIS/TaTp outperforms many similar photocatalysts reported previously (Table S2). In order to evaluate the photocatalytic stability of 20% ZIS/TaTp, four-cycle experiments were conducted, and the activity of 20% ZIS/TaTp remained above 90% of the initial activity after cycles (Supplementary Fig. 6). Indeed, comparison of the XRD and FTIR spectra before and after the reaction, as shown in Supplementary Fig. 7, revealed no significant changes in the characteristic peaks. Additionally, adding 20% ZIS/COF to various concentrations of commercial H₂O₂ under dark conditions with continuous stirring caused no observable changes in H₂O₂ concentration (Supplementary Fig. 8a). To further study the decomposition behaviors of the samples, the concentration change of 1 mM H₂O₂ was monitored under N₂ bubbling during 1 h of light exposure. As shown in Supplementary Fig. 8b, the decomposition of H₂O₂ by 20% ZIS/TaTp was negligible and significantly slower compared to TaTp and ZIS. These results suggest that the 20% ZIS/TaTp exhibits modified charge transfer pathway, with the photogenerated e⁻ and h⁺ primarily participating in reactions with O₂ and water rather than contributing to the in situ decomposition of H₂O₂ on the material surface. These results indicate that 20% ZIS/TaTp possesses excellent activity and structural stability.

Mechanism of H₂O₂ photosynthesis

To explore the key factors affecting the photosynthesis of H₂O₂, a series of control experiments were conducted with the 20% ZIS/TaTp sample. As a result, there was no detection of H₂O₂ in the system without visible light irradiation or catalyst (Fig. 3e), underscoring the indispensable role of light and catalyst in the photosynthesis of H₂O₂. So as to discern the role of h⁺ in the reaction, photosynthesis experiments were carried out in a sealed reactor with continuous N₂ pumping to negate the effects of dissolved O₂. Despite this, the H₂O₂ yield still reached 142 µmol∙g⁻¹ ∙ h⁻¹, probably due to the oxidation of H₂O by h⁺ to produce a small quantity of O₂, which was then reduced to H₂O₂⁴⁴. This hypothesis was corroborated by measuring the variation of dissolved O₂ in the sealed reaction system, which revealed that 20% ZIS/TaTp can produce O₂ (Supplementary Fig. 9), subsequently contributing to the photosynthesis of H₂O₂. In order to further understand the role of different reactive species, scavengers were introduced into the reaction system. Silver nitrate (AgNO₃), disodium ethylenediamine tetraacetate (EDTA-2Na), p-benzoquinone (BQ), and tert-butanol (TBA) were employed to capture e⁻, h⁺, ·O₂⁻, and ·OH, respectively⁵¹. The significant decrease in reaction activity upon the addition of AgNO₃ and BQ highlights the essential roles of e⁻ and ·O₂⁻ in the process. Conversely, the capture of h⁺ improved the photosynthetic performance of H₂O₂, which could be attributed to the increased utilization of e⁻. To further verify the impact of h⁺ capture, EtOH, and methanol (MeOH) were also introduced into the system, respectively, each resulting in a notable increase in H₂O₂ yield (Supplementary Fig. 10). The addition of TBA resulted in no observable change in the performance, confirming that ·OH was not involved in the process of H₂O₂ production⁵¹.

As shown in Fig. 3f, in situ FTIR spectroscopy was conducted to unveil the principal reaction intermediates involved in the photosynthesis of H₂O₂. One significant observation was the gradual increase in the peak at ~1145 cm⁻¹, which corresponds to the presence of ·O₂⁻ over the irradiation time⁴⁹. Meanwhile, the broad peak at about 910–970 cm⁻¹, associated with the O-O bond also exhibited an upward trend. These observations suggest that the H₂O₂ photosynthesis process was predominantly driven by the visible light-mediated reduction of O₂^52,53. Both in situ FTIR spectroscopy and scavenger experiments collectively verified the primary reaction pathway, indicating that the photosynthesis of H₂O₂ predominantly adhered to the two-step one-electron oxygen reduction reaction (1-e⁻ ORR) pathway. The principal reaction equations (Eqs. 1–5) can be expressed as follows:

The work function (Ф) and Fermi level (E_f) of the catalysts were investigated via UPS⁴⁴. The Ф values for TaTp, ZIS, and 20% ZIS/TaTp were measured to be 4.7 eV, 3.97 eV, and 4.25 eV, respectively, while the corresponding E_f values were determined as 0.2 eV, −0.53 eV, and −0.25 eV (Supplementary Fig. 11). Owing to the differentiation of E_f between TaTp and ZIS, e⁻ in ZIS migrated to TaTp during the contact process in the absence of light irradiation, until the equilibrium of E_f was reached. This e⁻ transfer established an IEF, which played a critical role in the successful formation of the S-scheme heterojunction⁴¹. As previously proposed by Kanata et al., the intensity of the IEF was calculated by Eq. 6^54,55. Where E is the IEF intensity, Vs is the surface photovoltage (SPV), ρ is the surface charge density, Ɛ is the low-frequency dielectric constant, and Ɛ₀ is the permittivity of free space.

Given that Ɛ and Ɛ₀ are two constants, the intensity of the IEF is mainly influenced by the SPV and charge density. After measuring Vs and ρ, we can readily compare the IEF intensity across different photocatalysts. As demonstrated in Fig. 4a and b, all catalysts show a positive SPV signal, thereby confirming their characteristics as n-type semiconductors. Remarkably, 20% ZIS/TaTp manifests a pronounced SPV signal, validating the substantial generation of photogenerated carriers under light irradiation. The transient photocurrent curve is integrated to output a specific charge density value, which is proportional to the number of positive charges deposited on the surface. The charge densities for TaTp, ZIS, and 20% ZIS/TaTp were found to be 731 nC/cm², 586 nC/cm² and 2917 nC/cm², respectively (Fig. 4c). In Fig. 4d, the IEF strength of ZIS is set to “1”, and the calculated IEF strengths are normalized accordingly. Specifically, the IEF strength of TaTp is “2.02”, while that for 20% ZIS/TaTp is “5.54”, which is significantly higher than that of TaTp and ZIS. This suggests a positive correlation between the IEF intensity and the photocatalytic activity.

The surface photovoltage spectra (a), surface photovoltage intensity of prepared samples (b), surface charge density (c), IEF intensity of TaTp, ZIS, and 20% ZIS/TaTp (d), XPS and in situ XPS of TaTp, ZIS, and 20% ZIS/TaTp: N 1s (e), O 1s (f), Zn 2p (g), In 3d (h), and S 2p (i).

Full size image

The charge transfer pathway within the composite was comprehensively investigated by in situ XPS. As shown in Fig. 4e–i and Supplementary Fig. 12, upon complexation of TaTp with ZIS, the N 1s and O 1s peaks shift toward lower binding energy, whereas Zn 2p, In 3d, and S 2p peaks shift toward higher binding energy. This observation indicates the e⁻ flow from ZIS to TaTp, resulting in the formation of a built-in electric field. Furthermore, under visible light irradiation, a reverse shift in binding energy was examined, implying that TaTp can transfer photogenerated e⁻ to ZIS. This unique pathway for the transfer of photogenerated carriers plays a momentous role in boosting photocatalytic performance. The presence of the IEF facilitates the recombination of ineffective e⁻ in the CB of TaTp with ineffective h⁺ in the VB of ZIS. This S-scheme e⁻ transference channel not only enhances the separation efficiency of charge carriers but also preserves the redox capability of the heterojunction. To further demonstrate the heterojunction type, EPR characterization was performed to determine the characteristics of DMPO-·OH, DMPO-·O₂⁻ and DMPO-·OOH. In Supplementary Fig. 13a, a distinct DMPO-·O₂⁻ signal was measured for TaTp, ZIS and 20% ZIS/TaTp. The heterojunction photocatalysis exhibited the strongest DMPO-·O₂⁻ signal, indicating the combination of ZIS and TaTp enhances the O₂-reducing capability. This outstanding reducing ability can be attributed to the efficient transfer of photogenerated e⁻ from TaTp to ZIS. It can be observed that with increasing irradiation time, O₂⁻ evolves into a ·OOH signal, which further supports the pathway of H₂O₂ photosynthesis. (Supplementary Fig. 13b). As shown in Supplementary Fig. 13c, none of the prepared photocatalysts produced a DMPO-·OH signal, as the VB potential was not sufficient to generate ·OH species³⁸. Based on the above analysis, the successful construction of novel S-scheme COF-based heterojunction photocatalysts was demonstrated.

The spatial charge transfer in ZIS/TaTp composite was further revealed by KPFM. Supplementary Fig. 14 and Fig. 5a, b shows the representative topographic image of ZIS/TaTp composite and the corresponding surface potential maps in darkness and under illumination. The surface potential difference between TaTp and ZIS before light irradiation demonstrates the formation of a built-in electric field pointing from ZIS to TaTp, which can act as a driver for photogenerated charge migration⁵⁶. After light irradiation, the surface potential of ZIS decreases while that of TaTp increases (Fig. 5c, d), which further suggests that the electrons of TaTp flow to ZIS under visible light irradiation.

KPFM potential images of 20% ZIS/TaTp in dark (a) and under 420 nm light illumination (b), The line-scanning surface potential from TaTp to ZIS (c), The schematic illustration of photoirradiation KPFM (d).

Full size image

In order to provide additional evidence for the S-scheme, Density Functional Theory (DFT) calculations were performed to determine the work function of TaTp and ZIS (Fig. 6a, b). The computed results were generally in alignment with those obtained from UPS. Moreover, the difference in charge density further verified the charge distribution at the interface (Fig. 6c). The yellow region signifies e⁻ aggregation, while the blue region indicates the opposite. At the interface of 20% ZIS/TaTp, the ZIS surface is predominantly characterized by a blue color, revealing the depletion of e⁻. Conversely, the TaTp surface exhibits more yellow regions, indicating the aggregation of e⁻. Examination of the electron localization function (ELF) reveals that e⁻ are distributed throughout all the atoms (Fig. 6d), indicating excellent electrical conductivity of the catalyst for unimpeded e⁻ transfer. As shown in Fig. 6e–g, TaTp and ZIS possess distinct Fermi levels. Upon contact between TaTp and ZIS, the internal e⁻ will migrate to TaTp due to the more negative E_f of ZIS, resulting in a tendency toward equilibrium in the E_f of the two semiconductors. As a result, a built-in electric field, oriented from ZIS to TaTp, is established at the contact interface. Under light irradiation, both TaTp and ZIS undergo excitation, leading to the photogeneration of e⁻-h⁺ pairs within each semiconductor. Driven by the built-in electric field, the e⁻ in the CB of TaTp will recombine with the h⁺ in the VB of ZIS. This process not only significantly improves the efficiency of carrier separation but also enables the composite to maintain carriers in the VB of TaTp and the CB of ZIS with more substantial redox abilities. Based on the aforementioned analysis, the enhancement of H₂O₂ photosynthesis can be ascribed to the conventional S-scheme heterojunction reaction mechanism (Fig. 7).

Work functions of TaTp (a) and ZIS (b) by DFT calculation, charge density difference (c) and electron localization function (d) of 20% ZIS/TaTp, and schematic diagram of photogenerated H₂O₂ over 20% ZIS/TaTp (e–g).

Full size image

Schematic diagram of H₂O₂ photosynthesis by 20% ZIS/TaTp.

Full size image

Photocatalysis-self-Fenton system for water disinfection

The in situ Fenton system was constructed for the disinfection of water bodies, adopting E. coli as the model bacterium. As illustrated in Fig. 8a, a complete eradication of 2 × 10⁷ cfu mL⁻¹ E. coli was achieved within 50 min when Fe(II) was added to activate the in situ generated H₂O₂. The successful establishment of the photocatalysis-self-Fenton system was corroborated through radical trapping experiments. Coumarin was employed as the ·OH trapping agent, resulting in the formation of a highly fluorescent adduct, 7-hydroxycoumarin, with emission properties peaking at ~460 nm. With the introduction of Fe (II) to the reaction system, the PL signal exhibited a progressive increase trend over time (Fig. 8b). Conversely, in the absence of Fe (II) addition, there was no significant enhancement in the fluorescent intensity (Fig. 8c). The fitting analysis of the PL intensity displayed a well-defined linear trend (Supplementary Fig. 15). Furthermore, Fig. 8d illustrates the time-dependent sterilizing effect of 20% ZIS/TaTp + Fe(II), as evidenced by photographs of agar plates inoculated with the reaction solutions over a 50-min period. These results collectively demonstrate the effective integration of H₂O₂ photosynthesis and the Fenton reaction, highlighting the potential for sustainable water disinfection.

Photocatalytic bacterial inactivation experiments under different conditions (a), fluorescence spectra of 7-hydroxycoumarin in 20% ZIS/TaTp system with and without Fe(II) (b, c), photos of the agar plates inoculated with reaction solutions from 20% ZIS/TaTp + Fe(II) at different sampling times (d).

Full size image

Materials

All chemicals utilized in this work were bought from Sinopharm Chemical Reagent Co. Ltd. and were directly employed for experiments after purchase without any treatment. 1,3,5-triformyl phloroglucinol, 1,3,5-tris(4-aminophenyl) triazine, mesitylene, 1,4-dioxane, acetic acid, tetrahydrofuran, acetone, Zn(CH₃COO)₂ ∙ 2H₂O, InCl₃ ∙ 4H₂O, thioacetamide.

Preparation of ZIS/TaTp

The ZIS/TaTp heterojunction was prepared as follows:

Synthesis of TaTp. TaTp was fabricated via a solvothermal method. First, 31.5 mg (0.15 mmol) of 1,3,5-triformyl phloroglucinol, 53.1 mg (0.15 mmol) of 1,3,5-tris(4-aminophenyl) triazine, 0.515 mL of mesitylene and 3.085 mL of 1,4-dioxane were added into a 10 mL autoclave and stirred for 30 min. Afterward, 0.6 mL of 3 M acetic acid was injected into the autoclave. After a few minutes of sonication, the autoclave was heated at 120 °C in an oven for 72 h. Upon completion of the reaction and cooling to room temperature, the obtained mixture was washed with methanol, tetrahydrofuran, and acetone. Lastly, the catalyst was dried in a vacuum desiccator at 100 °C for 12 h to obtain TaTp.

Synthesis of ZIS/TaTp heterojunction. ZIS/TaTp catalysts were fabricated through a hydrothermal approach. Typically, to obtain 20% ZIS/TaTp, 27.5 mg (0.125 mmol) of Zn(CH₃COO)₂ ∙ 2H₂O, 73.3 mg (0.25 mmol) of InCl₃ ∙ 4H₂O, 46.9 mg (0.625 mmol) thioacetamide, and 300 mg of TaTp were dissolved in 50 mL deionized (DI) water. The mixture was then sonicated for 60 min. The autoclave was subsequently sealed and heated at 180 °C in an oven for 24 h. The obtained powder samples were washed thoroughly with DI water and ethanol. Finally, the samples were dried at 60 °C for 12 h. By adjusting the mass of the precursor, other products with different load ratios (10% ZIS/TaTp, 30% ZIS/TaTp, and 40% ZIS/TaTp) were synthesized. ZnIn₂S₄ was synthesized accordingly without the addition of TaTp.

Details regarding the material characterizations, photoelectrochemical measurements and theoretical calculations can be referred to our previous publications^21,31,32.

Photosynthesis and activation of H₂O₂

The photosynthesis of H₂O₂ was carried out under the irradiation of a 300 W Xenon lamp with a 420 nm cut-off filter. For the experiment, 10 mg of photocatalyst was added to a beaker containing 50 mL of ultrapure water without other sacrificial reagents or photosensitizers. Prior to irradiation, the mixture was stirred for 30 min under dark conditions to disperse the catalyst in the solution. Every 10 min, 3 mL of the reaction liquid was collected, and the catalyst in the solution was removed by filtration. The H₂O₂ activation experiment was conducted in the same condition with the addition of Fe (II).

Bacteria inactivation experiments

E. coli was utilized as the model bacteria to evaluate the photocatalytic performance of the catalysts. Generally, 500 μL of 2 × 10⁹ cfu mL⁻¹ bacterial solution was added to a beaker containing 49.5 mL of ultrapure water. At regular intervals, samples were taken and spread on plates, and the number of viable bacteria was determined using the plate count method.