Multisite atomic-chlorine-passivation stabilizes perovskite interfaces for efficient H2O2 photosynthesis from seawater

Multisite atomic-chlorine-passivation stabilizes perovskite interfaces for efficient H2O2 photosynthesis from seawater | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Multisite atomic-chlorine-passivation stabilizes perovskite interfaces for efficient H 2 O 2 photosynthesis from seawater Baodui Wang, Genping Meng, Shuai Wei, Ning Li, Yuhui Yin, Bin Dong, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7187821/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Mar, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Lead halide perovskites are promising for artificial photosynthesis but suffer from aqueous instability. Here, we stabilize CsPbI 3 quantum dots (QDs) within a hydrophobic chlorine-functionalized covalent organic framework (COF-Cl) through multisite atomic-chlorine passivation, forming dual Cl–Pb coordination and Cl–I halogen bonding at the interface. This suppresses ionic migration while creating a gas–liquid–solid triphase interface for enhanced O 2 diffusion. The resulting S-scheme heterojunction spatially separates carriers to concurrently drive two-electron oxygen reduction and water oxidation for H 2 O 2 synthesis without sacrificial agents. The system achieves record production rates of 25.29 mmol h -1 g -1 in pure water and 20.37 mmol h -1 g -1 in seawater under visible light, with a solar-to-chemical conversion efficiency of 1.38%. Crucially, it operates stably for 20 h in seawater and produces 11.7 mmol L -1 H 2 O 2 in 10 h under natural sunlight. Mechanistic studies confirm synergistic interfacial charge transfer and dual-reaction pathways via both oxygen reduction and water oxidation. This work establishes a paradigm for robust perovskite-based photocatalysts toward scalable solar-driven chemical synthesis from seawater. Physical sciences/Chemistry/Catalysis/Photocatalysis Physical sciences/Chemistry/Green chemistry/Sustainability Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Hydrogen peroxide (H 2 O 2 ) is an indispensable chemical oxidant with burgeoning applications in water treatment, disinfection, and green synthesis 1 , 2 . However, the incumbent anthraquinone process for H 2 O 2 production is energy-intensive, generates significant waste, and relies on centralized facilities, limiting its accessibility and sustainability. Solar-driven photosynthesis of H 2 O 2 directly from water (H 2 O) and oxygen (O 2 ) offers a compelling decentralized and sustainable alternative 3 . Seawater, constituting over 96% of Earth's hydrosphere, presents an abundant aqueous resource for this process, yet its complex ionic composition and corrosivity pose formidable challenges to photocatalyst stability and performance 4 . The development of robust, seawater-compatible photocatalytic systems is therefore essential to unlock scalable solar H 2 O 2 synthesis. Lead halide perovskites, particularly CsPbI 3 quantum dots (QDs), have emerged as outstanding light harvesters for artificial photosynthesis 5 , 6 . This is attributed to their extensive visible-light absorption, suitable band edges, and high charge-carrier mobility 7 . However, their well-known vulnerability to ionic dissolution and structural degradation in aqueous environments, especially in ion-rich seawater, significantly limits their practical applications 8 . Although conventional encapsulation strategies using polymers or metal oxides offer some degree of barrier protection, they often impede mass transport (such as O 2 diffusion) or fail to completely suppress interfacial ion migration, which is a crucial degradation pathway 9 , 10 . Covalent organic frameworks (COFs) have recently gained attention as protective matrices due to their crystallinity, designable porosity, and functional versatility 11 , 12 . Notably, chlorine-functionalized COFs (COF-Cl) remain underexplored, despite the potential of atomic chlorine sites to form strong, directional bonds with perovskite surfaces through halogen interactions (Cl-I) and coordination (Cl-Pb) 13 , 14 . Such atomic-scale passivation could simultaneously stabilize the perovskite and enhance interfacial charge dynamics. Herein, we introduce a novel interfacial engineering strategy to simultaneously passivate CsPbI 3 QDs against aqueous degradation and enhance its photocatalytic activity for H 2 O 2 production directly from seawater (Scheme 1 ). We rationally design and synthesize COFs featuring precisely positioned, multisite atomic chlorine functionalities. The COF-Cl matrix stabilizes CsPbI 3 QDs through dual halogen bonding (Cl-I) and coordination (Cl-Pb), effectively suppressing ionic migration while creating a hydrophobic triphase interface to enhance O 2 diffusion. Crucially, the S-scheme heterojunction forms between the CsPbI 3 QDs and COF-Cl, enabling efficient spatial separation of charge carriers: electrons localized on the COF-Cl drive the 2e⁻ ORR, while holes retained on the CsPbI 3 QDs drive the WOR. This system achieves record production rates of 25.29 mmol h − 1 g − 1 (pure water) and 20.37 mmol h − 1 g − 1 (natural seawater) with > 90% stability over 20 hours in seawater and operates efficiently under natural sunlight. Through integrated experimental and theoretical analyses, we elucidate the synergistic interfacial mechanisms driving this performance. Our work establishes a paradigm for robust perovskite-based photosynthesis, unlocking scalable solar-driven H 2 O 2 synthesis from abundant seawater. Results Synthesis and characterization of hydrophobic halogenated COFs and CsPbI 3 @COF-X. A series of hydrophobic halogenated covalent organic frameworks (COF-X: Cl, Br, I) were synthesized via Schiff-base condensation between 4,4',4''-(1,3,5-triazine-2,4,6-triyl) trianiline and halogenated 2,5-dihaloterephthalaldehydes (Cl, Br, I) (Figure S2). Highly crystalline materials were obtained in o-dichlorobenzene/n-BuOH/6 M HAc (5:5:1, v/v) at 120°C for 72 h. Structural analysis confirmed successful synthesis and high crystallinity (Figure S3-S5). Powder X-ray diffraction (PXRD) patterns (Figure S6) showed prominent (100) peaks at 2.79°, 2.75°, and 2.70° for COF-Cl, COF-Br, and COF-I, respectively, matching simulated AA stacking models (Figures S3-S5, Table S1 ). Solid-state 13 CNMR revealed the characteristic imine peak (~ 160 ppm, Figure S7). Fourier transform infrared (FTIR) spectra (Figure S8) confirmed Schiff-base formation by the disappearance of precursor -NH 2 (3213, 3327 cm − 1 ) and aldehyde (1688 cm − 1 ) stretches and the appearance of -C = N stretches (1633 cm − 1 ) for all three COF-X 15 . COF-X exhibit mesoporosity with type IV isotherms (Figure S9). Brunauer-Emmett-Teller (BET) surface areas decreased with increasing halogen size: 2289.7 m 2 g - 1 (COF-Cl), 1757.0 m 2 g - 1 (COF-Br), 1150.9 m 2 g - 1 (COF-I), correlating with pore diameters of 3.53 nm, 3.09 nm, and 2.63 nm, respectively (Table S2). Scanning electron microscope (SEM) and transmission electron microscope (TEM) images (Figures S10, S11) revealed uniform fibrous morphologies and layered structures, with visible pores in COF-X HR-TEM (Figure S11b). X-ray photoelectron spectra (XPS) confirmed elemental composition (Figures S12-S14). Water contact angles of 143.0°, 145.5°, and 147.5° (Figure S15) demonstrate increasing hydrophobicity with halogen size. Excellent thermal stability was confirmed by TGA, showing negligible weight loss at 150°C (Figure S16). These results confirm the successful synthesis of crystalline, mesoporous, hydrophobic halogenated COFs. A sequential deposition route was employed to synthesize CsPbI 3 @COF-X composites (Figure S17). First, COF-X matrices were immersed in a PbI 2 precursor solution at 70°C. Residual PbI 2 was removed via centrifugation using DMF/ethanol. The resulting PbI 2 @COF-X intermediates were then dispersed in a CsI-containing DMF solution under ultrasonication. Finally, the reaction mixture was precipitated in toluene at room temperature to obtain a series of CsPbI 3 @COF-X systems, where CsPbI 3 QDs in-situ nucleated and crystallized inside the COF-X matrix (for a full description of the methods, see the Supplementary information). Taking CsPbI 3 @COF-Cl as an example, we demonstrate how the framework and topology of the COF-Cl matrix direct the confined growth of CsPbI 3 QDs. Structural characterization confirmed the successful synthesis of CsPbI 3 @COF-Cl. SEM imaging (Figure S18) revealed clean surfaces, indicating CsPbI 3 QDs are embedded within the COF-Cl matrix. TEM analysis (Fig. 1 a) further shows CsPbI 3 QDs uniformly dispersed in the mesopores of COF-Cl matrix. Isolated CsPbI 3 QDs with an average diameter of 3.2 nm are clearly resolved (Fig. 1 b), consistent with the COF-Cl mesopore size. High-resolution TEM (HRTEM, Fig. 1 c) displays distinct lattice fringes (0.63 nm spacing), matching the (100) planes of α-CsPbI 3 (space group Pm-3m ). The crystalline structure is further confirmed by fast-Fourier transform (FFT) patterns (Fig. 1 d). STEM-HAADF imaging (Fig. 1 e) corroborates the uniform dispersion of 3.2 nm CsPbI 3 QDs within the COF-Cl pores. Elemental mapping (EDX, Fig. 1 f) confirms homogeneous distribution of Cl, Cs, Pb, and I throughout the composite. Hydrophobicity, critical for aqueous applications, was evaluated via contact angle measurements (Fig. 1 g). The CsPbI 3 @COF-Cl system exhibits strong hydrophobicity (139.5°), attributable to the COF-Cl matrix. Consequently, the lightweight composite spreads uniformly on water surfaces (Figures S19a, b), enabling redox reactions at the air-liquid-solid interface of the floating photocatalyst (Figure S20). PXRD analysis (Fig. 2 h) verifies crystallinity, with peaks at 14.1°, 19.9°, 24.5°, and 28.4° corresponding to the (100), (110), (111), and (200) planes of α-CsPbI 3 material (PDF No. 01-080-4039) 16 . N 2 adsorption-desorption isotherms (Fig. 2 i) provide additional evidence for CsPbI 3 QDs encapsulation, showing reduced pore volumes and specific surface area due to in situ CsPbI 3 QDs growth within the mesopores. Collectively, these results confirm the successful construction of the CsPbI 3 @COF-Cl. To evaluate the interfacial interaction between COF-Cl and CsPbI 3 QDs, we employed X-ray photoelectron spectroscopy (XPS), solid-state 13 C nuclear magnetic resonance (NMR) spectroscopy, and time-resolved fluorescence decay spectroscopy (TRFDS) to analyze the electronic structure of constituent elements. XPS analysis confirmed the elemental composition of the CsPbI 3 @COF-Cl system, with distinct peaks for C, N, Cl, Cs, Pb, and I (Figure S21), verifying successful synthesis. Notably, the Pb 4f XPS spectra (Fig. 2 a) revealed a distinct shift toward higher binding energy in CsPbI 3 @COF-Cl compared to pristine CsPbI 3 QDs, indicating interaction between under-coordinated Pb 2+ and Cl atoms 17 , 18 . Similarly, the I 3d spectra (Fig. 2 b) showed a slight shift to higher binding energy, consistent with Cl-I bond formation and interfacial charge transfer between CsPbI 3 QDs and COF-Cl 19 . Conversely, the Cl 2p peaks shifted toward lower binding energy (Fig. 2 c), further evidencing strong chemical interactions. To further investigate the chemical environment changes, solid-state 13 C NMR spectroscopy was performed. The spectrum (Fig. 2 d) revealed a downfield shift of 0.3 ppm for the carbon atoms directly bonded to chlorine in CsPbI 3 @COF-Cl relative to pristine COF-Cl. This shift provides evidence for the formation of halogen-halogen (Cl-I) bonds and coordination interactions (Cl-Pb) at the interface. To directly assess the impact of these interactions on defect passivation, TRFDS measurements were conducted (Fig. 2 e, Table S3). Fluorescence decay curves were fitted to a tri-exponential model, where fast decay correlates with surface/crystal defects and slow decay with carrier transport. Pristine CsPbI 3 QDs exhibited rapid decay (τ 1 = 3.13 ns, τ 2 = 13.68 ns, τ 3 = 62.22 ns; τₐ v = 24.96 ns), attributed to non-radiative recombination from Pb-I antisite defects and vacancies 20 . In contrast, CsPbI 3 @COF-Cl showed significantly slower decay (τ 1 = 9.60 ns, τ 2 = 57.52 ns, τ 3 = 289.41 ns; τₐ v = 174.38 ns), demonstrating that COF-Cl passivates defects via halogen-halogen bonds and coordination interactions, thereby suppressing non-radiative pathways 20 . To elucidate the multi-site bonding mechanism of the halogen-rich porous COF-Cl matrix in detail, we investigated the adsorption modes of the COF-Cl matrix on CsPbI 3 QD surfaces, along with the corresponding binding and formation energies. Computational modeling (Fig. 2 f, S22) reveals that COF-Cl can achieve continuous multi-site bonding. Specifically, there is Cl-Pb coordination on Pb-rich surfaces and Cl-I bonding on I-rich surfaces. The formation energies for continuous multi-site Cl-Pb and Cl-I bonds are highly favorable at -2.76 eV and − 2.56 eV, respectively (Fig. 2 g). Similarly, the binding energies are strong at -1.54 eV and − 1.32 eV, respectively (Fig. 2 h). For comparison, selective removal of Cl functional groups to create discontinuous passivation (Figure S23) resulted in significantly reduced binding and formation energies, confirming that continuous multisite passivation optimally stabilizes the perovskite lattice. Furthermore, molecular dynamics simulations assessed structural stability. The root mean square deviation (RMSD) curves (Fig. 2 i), which measure atomic position fluctuations over time, stabilized after ~ 1 ps, indicating system equilibrium. Throughout the simulation, the RMSD values for systems with continuous multi-site Cl-Pb and Cl-I passivation remained consistently lower than those with discontinuous passivation. This demonstrates minimal atomic displacement and superior structural stability afforded by the continuous multi-site bonding mechanism. Characterizations of the electronic properties and the band edge positions. Based on the ultraviolet-visible diffuse reflection spectra (UV-vis DRS), the COF-X matrix exhibits adsorption peaks (Soret band) around 350–500 nm (Fig. 3 a), attributed to its periodic molecular framework and long-range ordered π-stacking structure. In contrast, CsPbI 3 QDs show a broad absorption band extending to 700 nm, indicative of a narrow bandgap. Notably, upon encapsulation of CsPbI 3 QDs by COF-X to form the CsPbI 3 @COF-X system, the absorption edge of the CsPbI 3 @COF-X system undergoes a significant redshift beyond 700 nm, enabling more efficient solar energy utilization. This redshift clearly demonstrates the synergistic effect between COF-X and CsPbI 3 QDs. Tauc plot analysis (Figure S24) determined the bandgap energies (E g ) of COF-Cl, COF-Br, COF-I, and CsPbI 3 QDs to be 2.50 eV, 2.58 eV, 2.49 eV, and 1.77 eV, respectively, confirming their semiconductor nature. For efficient photocatalytic reactions, the electronic band structure must align with the relevant redox potentials. Valence band XPS spectra (VB-XPS, Figure S25) measured the valence band maximum (E VB ) positions relative to the vacuum level (E VB, XPS ). After conversion to the standard hydrogen electrode (NHE) scale, the E VB, NHE values are: COF-Cl (2.52 eV), COF-Br (2.39 eV), COF-I (2.27 eV), and CsPbI 3 QDs (0.94 eV). The resulting band diagrams are plotted in Fig. 3 b. Crucially, this staggered band alignment satisfies the prerequisite for forming S-scheme heterojunction. The results indicate that CsPbI 3 QDs possess strong reducing capabilities, while the COF-X matrices exhibit high oxidizing capabilities, with COF-Cl showing the strongest oxidizing potential. This favorable energy level alignment within the CsPbI 3 @COF-Cl composite not only maximizes the redox capabilities of the individual components but also facilitates the construction of an efficient S-scheme heterojunction (Fig. 3 c). The CsPbI 3 @COF-Cl system significantly enhances charge transfer and carrier separation. Photoluminescence (PL) spectroscopy and TRFDS show substantially quenched PL intensity and prolonged fluorescence lifetime in CsPbI 3 @COF-Cl compared to pure CsPbI 3 QDs (Fig. 3 d and 2 e), indicating suppressed electron-hole recombination due to spatial charge separation. These properties directly improve photochemical conversion activity. Photocurrent response tests (Fig. 3 e) show faster charge transfer in CsPbI 3 @COF-Cl compared to COF-Cl alone. Pure CsPbI 3 QDs were excluded due to their aqueous instability. Electrochemical impedance spectroscopy (EIS) further confirms reduced charge-transfer resistance in CsPbI 3 @COF-Cl, evidenced by a smaller Nyquist semicircle (Figure S26). The interfacial electron transfer mechanism in the CsPbI 3 @COF-Cl heterojunction was demonstrated by Ultraviolet photoelectron spectroscopy (UPS). The results are displayed in Figure S27 and Figure S28. The work functions of CsPbI 3 QDs and COF-Cl are 5.83 eV and 7.00 eV, respectively. Correspondingly, the Fermi levels of CsPbI 3 QDs and COF-Cl are − 5.83 eV and − 7.00 eV (vs vacuum level), respectively. This work function disparity establishes a higher Fermi level in CsPbI 3 QDs than in COF-Cl, driving electron migration from CsPbI 3 QDs to COF-Cl upon heterojunction formation 21 . Consequently, band bending at the interface induces charge redistribution, establishes a unified Fermi level, and generates an internal electric field (IEF) 22 . Further validation through in-situ XPS under light illumination tracked binding energy shifts in key elements. As shown in Fig. 3 f, light exposure caused the Cl 2p peaks in CsPbI 3 @COF-Cl to shift from 202.2 eV and 200.6 eV to 202.6 eV and 201.0 eV, indicating electron loss from the COF-Cl component. Simultaneously, the Pb 4f peaks shifted from 144.29 eV/139.42 eV and 143.57 eV/138.70 eV to 143.92 eV/139.05 eV and 143.25 eV/138.38 eV (Fig. 3 g), while the I 3d peaks shifted from 632.12 eV/620.62 eV and 630.95 eV/619.45 eV to 632.46 eV/619.96 eV and 630.46 eV/618.96 eV (Fig. 3 h). These shifts collectively confirm electron gain by the CsPbI 3 QDs component. Therefore, photogenerated electrons transfer from COF-Cl to CsPbI 3 QDs under light excitation, directly evidencing the S-scheme charge transfer process. The combined UPS and XPS results support the proposed mechanism (Figure S29). Initially, the Fermi level difference drives the flow of electrons from CsPbI 3 QDs to COF-Cl, thereby forming the IEF. This leads to an upward band bending in CsPbI 3 QDs and a downward band bending in COF-Cl at the interface. Under illumination, the IEF, Coulomb attraction, and band bending promote recombination of COF-Cl conduction band electrons with CsPbI 3 valence band holes. Significantly, this mechanism retains useful electrons in the CB of CsPbI 3 QDs and holes in the VB of COF-Cl, maximizing their redox potential for enhanced photocatalytic oxygen reduction. Full reaction photosynthesis of H 2 O 2 on a three-phase interface. The CsPbI 3 @COF-Cl system demonstrates significant potential for the photocatalytic synthesis of H 2 O 2 , as its energy band structure is sufficiently capable of driving both essential half-reactions: the oxidation of H 2 O (E(H 2 O 2 / H 2 O) = + 1.76 V vs. NHE, pH = 0) and the reduction of O 2 (E( H 2 O 2 /O 2 ) = + 0.68 V vs. NHE, pH = 0) 23 . This indicates its theoretical suitability as an effective photocatalyst for the full photosynthetic reaction producing H 2 O 2 . Cyclic voltammetry (CV) further elucidates these redox capabilities. As shown in Figure S30, the CV curve identifies a reduction potential of -0.94 V (vs. Ag/AgCl, pH = 7), corresponding to -0.33 V vs RHE (pH = 0), assigned to the reduction of O 2 to •O 2 ⁻ . Simultaneously, an oxidation potential of + 1.15 V (vs. Ag/AgCl, pH = 7), equivalent to + 1.76 V vs RHE (pH = 0), is observed for the oxidation of H 2 O to H 2 O 2 . Therefore, the combined theoretical and electrochemical data confirm that the CsPbI 3 @COF-Cl system can effectively drive the photochemical synthesis of H 2 O 2 using only water and atmospheric oxygen. Photocatalytic H 2 O 2 generation was evaluated in both pure water and seawater under visible-light irradiation (λ > 420 nm) and ambient air, without sacrificial agents. Benefiting from its lightweight nature and excellent hydrophobicity, the CsPbI 3 @COF-Cl system readily spreads across the water surface, facilitating redox reactions at this unique air-liquid-solid interface (Figure S20). Under air in pure water, the H 2 O 2 yield showed clear accumulation over time, exhibiting a near-linear relationship between production and irradiation duration for both the COF-Cl matrix alone and the CsPbI 3 @COF-Cl system (Fig. 4 a). This indicates sustained high photosynthetic rates even under prolonged illumination. After 1 h, H 2 O 2 yields reached 15.57 mmol g − 1 h − 1 for COF-Cl and 25.29 mmol g − 1 h − 1 for CsPbI 3 @COF-Cl, respectively (Figure S31). This significant enhancement in the CsPbI 3 @COF-Cl system is attributed to optimized charge separation efficiency achieved through interfacial band engineering. Given this promising performance, we systematically evaluated H 2 O 2 production under seawater and air conditions (Fig. 4 b). After 1 h of photochemical conversion, H 2 O 2 yields of 12.52 mmol g − 1 h − 1 and 20.37 mmol g − 1 h − 1 were achieved for the COF-Cl matrix and the CsPbI 3 @COF-Cl system, respectively (Figure S32). Furthermore, the CsPbI 3 @COF-Cl system demonstrated a notable apparent quantum yield (AQY) of 15.76% at 450 nm (Fig. 4 c) and solar-to-chemical conversion (SCC) efficiency of 1.38%. Cycling stability tests confirmed the system's durability. The CsPbI 3 @COF-Cl system maintained superior H 2 O 2 production performance over five consecutive photocatalytic cycles in seawater under air (Figure S33 and S34). TEM imaging and XRD analysis revealed no significant structural changes after cycling in seawater (Figures S35 and S36), highlighting its robustness. To further verify operational stability in seawater, a continuous 20-h photocatalytic reaction was conducted (Fig. 4 d), demonstrating stable H 2 O 2 production. PXRD measurements during operation further confirmed excellent structural stability under prolonged seawater exposure (Fig. 4 e). To demonstrate practical solar-to-chemical energy conversion, the H 2 O 2 production performance of CsPbI 3 @COF-Cl was evaluated under natural sunlight in Lanzhou City, China (May 1, 2025; Figs. 4 f and S37). Real-time monitoring was carried out to track solar irradiance at 450 nm, system temperature, and H 2 O 2 yield. After 10 hours, an accumulated H 2 O 2 concentration of 11.7 mmol L - 1 was achieved, confirming efficient photocatalytic activity under real-world conditions. Combined with its high efficiency, these findings indicate that the CsPbI 3 @COF-Cl system holds significant promise for large-scale, solar-driven H 2 O 2 production from pure water and natural seawater. This performance is significantly higher than that of most reported photocatalysts (Fig. 4 g and Table S4) 24 – 35 . To elucidate the reaction mechanism of hydrogen peroxide (H 2 O 2 ) generation, electron paramagnetic resonance (EPR) spectroscopy was employed to detect key intermediates during photochemical conversion. Using 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMP) as a spin trap for singlet oxygen ( 1 O 2 ) and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) for superoxide anion radicals (•O 2 ⁻ ), distinct signals for DMPO-•O 2 ⁻ and TEMP- 1 O 2 adducts were observed (Figs. 5 a, 5 b) 36 . The CsPbI 3 @COF-Cl system exhibited significantly stronger intermediate signals than COF-Cl alone. Superoxide (•O 2 ⁻ ), a key intermediate in H 2 O 2 generation via oxygen reduction reactions (ORR), forms through O 2 + e⁻ → •O 2 ⁻ (electrode potential: −0.33 V vs. RHE, pH = 0) and subsequently reacts via •O 2 ⁻ + e⁻ + 2H⁺ → H 2 O 2 . Singlet oxygen ( 1 O 2 ) is generated by •O 2 ⁻ oxidation by photogenerated holes (h⁺): •O 2 ⁻ + h⁺ → 1 O 2 , a pathway consistent with prior reports on photocatalytic H 2 O 2 production 37 . Furthermore, 1 O 2 contributes to H 2 O 2 formation through a two-electron reduction: 1 O 2 + 2e⁻ + 2H⁺ → H 2 O 2 . Alternatively, H 2 O 2 can form directly via a two-electron ORR: O 2 + 2e⁻ + 2H⁺ → H 2 O 2 (0.68 V vs. RHE, pH = 0). The significant EPR signals for both 1 O 2 and • O 2 ⁻ observed in the CsPbI 3 @COF-Cl system demonstrate that H 2 O 2 generation proceeds via both pathways, contributing to its enhanced photosynthetic efficiency. We also explored the contribution of the water oxidation reaction (WOR) to H 2 O 2 production. Previous studies indicate that water can react with holes to form hydroxyl radicals (H 2 O + h⁺ → •OH + H⁺), which dimerize to produce H 2 O 2 (2•OH → H 2 O 2 ) 22 . This process has a standard electrode potential of 2.72 V vs. RHE (pH = 0). However, no •OH signals are detected in the CsPbI 3 @COF-Cl system, as their VB positions are more negative than the standard redox potential of H 2 O/•OH, rendering them thermodynamically incapable of oxidizing H 2 O to generate •OH radicals 21 . Alternatively, H 2 O 2 can be generated directly from water and holes (2H 2 O + 2h⁺ → H 2 O 2 + 2H⁺; E = 1.76 V vs. RHE, pH = 0), suggesting WOR also contributes to H 2 O 2 formation. To validate this hypothesis and explore alternative H 2 O 2 -production pathways, a series of control experiments were performed over CsPbI 3 @COF-Cl system. As shown in Fig. 5 c, p-benzoquinone (p-BQ), tryptophan (Trp), AgNO 3 , and ethanol (EtOH) were employed as scavengers for •O 2 ⁻ , 1 O 2 , e − , and h + , respectively 38 . The addition of p-BQ drastically reduced the H 2 O 2 yield, underscoring the critical role of •O 2 ⁻ in H 2 O 2 generation. Similarly, Trp moderately suppressed H 2 O 2 production, confirming the participation of 1 O 2 in the formation process. Notably, while AgNO 3 and EtOH also diminished H 2 O 2 yields, a significant amount of H 2 O 2 persisted, suggesting that both e − and h + contribute to but are not solely responsible for H 2 O 2 production. To rigorously assess this hypothesis and elucidate potential competing H 2 O 2 formation mechanisms, we conducted systematic control experiments on the CsPbI 3 @COF-Cl photocatalyst. As shown in Fig. 5 c, p-benzoquinone (p-BQ), tryptophan (Trp), AgNO 3 , and ethanol (EtOH) were employed as scavengers for •O 2 ⁻ , 1 O 2 , e − , and h + , respectively. The pronounced suppression of H 2 O 2 generation upon p-BQ introduction establishes •O 2 ⁻ as the dominant reactive oxygen species in the production pathway. Similarly, Trp moderately suppressed H 2 O 2 production, confirming the participation of 1 O 2 in the formation process. Notably, while AgNO 3 and EtOH also diminished H 2 O 2 yields, a significant amount of H 2 O 2 persisted, suggesting that both e − and h + contribute to but are not solely responsible for H 2 O 2 production. Supporting evidence for surface intermediates was obtained through in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) under visible-light irradiation in the presence of H 2 O and O 2 (Fig. 5 d). The appearance of a C-OH vibration peak (1077 cm − 1 ) confirms the presence of adsorbed *OH species, likely formed by H 2 O dissociation 39 . Furthermore, characteristic vibration signals of *OOH (1647 and 1013 cm − 1 ) were detected, demonstrating the activation of adsorbed O 2 by photogenerated electrons (e⁻) to form *OOH intermediates 40 . Moreover, the signals of O-O stretching mode (890–921 cm − 1 ) gradually increased 41 – 43 . Density functional theory (DFT) calculations determined the Gibbs free energy (ΔG) for the adsorption of intermediates (*O 2 , *OOH, *OH) at distinct sites on the CsPbI 3 @COF-Cl system during both the oxygen reduction reaction (ORR) and water oxidation reaction (WOR) pathways for photocatalytic H 2 O 2 synthesis (Fig. 5 e and 5 f). These pathways involve unique intermediates and reaction steps. For the ORR process, O 2 undergoes conversion to *O 2 and *OOH, followed by *OOH protonation to yield H 2 O 2 . As visualized in Fig. 5 e, both *O 2 and *OOH adsorb preferentially on CsPbI 3 QDs. The energy barriers for these steps are − 0.44 eV and − 0.88 eV, respectively, reflecting the synergistic role of the S-scheme heterojunction in reducing the overall reaction energy barrier. Conversely, the WOR pathway involves H 2 O dehydrogenation to *OH (the rate-determining step) followed by *OH dimerization to form H 2 O 2 . Figure 5 f illustrates that *OH adsorbs exclusively on COF-Cl sites, with an energy barrier of 1.72 eV for the initial dehydrogenation step. These results demonstrate that the CsPbI 3 @COF-Cl system favors adsorption of *OOH (ORR) and *OH (WOR) intermediates. Under visible light, photogenerated electrons and holes separate via the S-scheme heterojunction and migrate to CsPbI 3 QDs and COF-Cl, respectively, enabling simultaneous ORR and WOR. On CsPbI 3 QDs, ORR proceeds through *O 2 hydrogenation to *OOH and subsequent protonation to H 2 O 2 . Meanwhile, WOR on COF-Cl involves H 2 O dehydrogenation to *OH (ΔG = 1.72 eV) followed by *OH coupling. Critically, protons released during WOR are consumed in ORR, maintaining charge balance. Consequently, the S-scheme heterojunction in CsPbI 3 @COF-Cl accelerates charge separation, facilitating O 2 reduction by electrons and H 2 O oxidation by holes to generate reactive intermediates, thereby promoting efficient H 2 O 2 production. Conclusion This work establishes a multifunctional interfacial engineering paradigm that stabilizes CsPbI 3 quantum dots within a hydrophobic chlorine-functionalized covalent organic framework (COF-Cl), enabling efficient and durable photosynthesis of H 2 O 2 from seawater. Our strategy leverages multisite atomic-chlorine passivation, simultaneously forming Cl–I halogen bonds and Cl–Pb coordination bonds, to suppress ionic migration and defect-induced recombination. Concurrently, the COF-Cl matrix creates a gas–liquid–solid triphase interface that enhances O 2 diffusion while isolating CsPbI 3 from aqueous degradation. Crucially, the system forms an S-scheme heterojunction that spatially separates charge carriers: electrons accumulate on CsPbI 3 for selective 2e⁻ oxygen reduction (O 2 → H 2 O 2 ), while holes localize on COF-Cl to drive water oxidation (H 2 O → H 2 O 2 ), all without sacrificial agents. The system achieves record H 2 O 2 production rates of 25.29 mmol h − 1 g − 1 in pure water and 20.37 mmol h − 1 g − 1 in natural seawater, surpassing state-of-the-art photocatalysts, with an apparent quantum yield of 15.76% at 450 nm and a solar-to-chemical conversion efficiency of 1.38%. Structural integrity is maintained over multiple cycles, and outdoor testing confirms operational viability under natural sunlight. Mechanistic studies confirm synergistic proton-coupled electron transfer, where holes on COF-Cl oxidize H 2 O to supply protons for O 2 reduction on CsPbI 3 QDs, balancing reaction kinetics. This work pioneers a robust design for perovskite-based artificial photosynthesis in corrosive environments, setting new benchmarks for efficiency and stability in solar-driven chemical production. The multifunctional passivation strategy, integrating atomic-scale defect control, tailored interfaces, and directional charge separation, paves the way for scalable solar fuel synthesis from abundant seawater resources. Methods Synthesis of 2,4,6-tris(4-aminophenyl)-1,3,5-triazine (TAPT). A solution was prepared by adding 2.0 g (17.0 mmol) of 4-aminobenzonitrile to a 50 mL round-bottom flask and cooling it in an ice bath at 0°C for 30 minutes. Under a nitrogen atmosphere, 5.0 mL (55.5 mmol) of trifluoromethanesulfonic acid was slowly added dropwise, maintaining the temperature at 0°C throughout the addition. Upon completion, the reaction mixture was poured into 50 mL of deionized water. The mixture was then neutralized to pH 7 using a 2 mol·L − 1 NaOH solution, resulting in the formation of a pale yellow precipitate. This precipitate was collected by filtration and purified through multiple washes with distilled water to afford the TAPT product. Synthesis of COF-X (X = Cl, Br, I) Frameworks. The COF-X frameworks were synthesized by dissolving TAPT (0.16 mmol, 56.7 mg) and the appropriate dialdehyde monomer (0.24 mmol) in a pressure-resistant glass tube containing a 1:1 (v/v) mixture of o-dichlorobenzene and n-butanol (1.0 mL each). The dialdehyde monomers used were: 2,5-dichloroterephthalaldehyde (PDA-Cl, 48.7 mg) for COF-Cl, 2,5-dibromoterephthalaldehyde (PDA-Br, 70.1 mg) for COF-Br, and 2,5-diiodoterephthalaldehyde (PDA-I, 92.6 mg) for COF-I. The mixture was sonicated for 5 minutes to enhance dissolution. Subsequently, 0.2 mL of 6 mol·L − 1 acetic acid solution was added quickly, followed by brief sonication (10 seconds). The glass tube was then subjected to three freeze-pump-thaw cycles under liquid nitrogen and vacuum, sealed, and heated in an oil bath at 120°C for 72 hours. After cooling to room temperature, the resulting yellow product was collected by centrifugation, thoroughly washed with tetrahydrofuran and methanol, and purified by Soxhlet extraction to yield the COF-X framework. Preparation of CsPbI 3 @COF-X Artificial Photosynthetic Systems. For each CsPbI 3 @COF-X system, 50 mg of COF-X (X = Cl, Br, I) powder was immersed in 2.5 mL of a 0.04 mol·L − 1 PbI 2 solution in N,N-dimethylformamide (DMF). The suspension was sonicated for uniform dispersion and reacted at 70°C for 2 hours. The resulting solid, designated as PbI 2 @COF-X, was collected and washed repeatedly with DMF and ethanol to remove surface-adsorbed PbI 2 . This intermediate was then immersed in 2.5 mL of a 0.02 mol·L − 1 CsI solution in DMF for 1 hour. Following this, 10 µL of oleic acid (OA) and 5 µL of oleylamine (OM) were added to stabilize the precursor mixture. Toluene was then added to induce the formation of CsPbI 3 quantum dots (QDs) within the pores of the COF-X framework. The final CsPbI 3 @COF-X material was collected by centrifugation, washed thoroughly with toluene three times, and dried. Photochemical H 2 O 2 Production Test under Simulated Sunlight. The photocatalytic performance for H 2 O 2 production was evaluated using a 5 mg sample of the CsPbI 3 @COF-X artificial photosynthetic system dispersed in a double-layered reaction beaker containing 20 mL of either deionized water or seawater (0.5 g·L − 1 ). The reaction mixture was gently shaken to ensure uniform dispersion of the photocatalyst on the water surface. A thermostatic circulating water system maintained the reaction temperature at 25°C. Photoreactions were conducted in air using a quartz and Pyrex glass hybrid reaction cell. Irradiation was provided by a 300 W xenon lamp equipped with a 420 nm cutoff filter (λ ≥ 420 nm). Throughout the reaction, the temperature was maintained at 25°C using a stirrer and the circulating water system. At 15-minute intervals, a 1.5 mL aliquot of the reaction solution was extracted, centrifuged, and filtered through a 0.22 µm membrane to remove the photocatalyst. The hydrogen peroxide concentration in the filtrate was quantified using the iodometric method. Photoelectrochemical and Electrochemical Characterization. Photoelectrochemical and electrochemical properties were characterized using a standard three-electrode system with an Ag/AgCl reference electrode, a platinum sheet counter electrode, and an FTO conductive glass working electrode coated with the catalyst. The working electrode was prepared by dispersing 5 mg of the photocatalytic material in 0.5 mL of an ethanol/isopropanol mixture (3:1 v/v), sonicating for 30 minutes, adding 20 µL of Nafion solution, and sonicating for an additional 30 minutes. A 50 µL aliquot of the resulting catalyst slurry was uniformly coated onto the FTO glass and dried at 60°C for 2 hours. A 300 W xenon lamp with a 400 nm cutoff filter served as the light source. Transient Photocurrent Response: Measured in 0.1 mol·L − 1 Na 2 SO 4 electrolyte under light illumination. Mott-Schottky (MS) Analysis: Conducted in 0.1 mol·L − 1 Na 2 SO 4 electrolyte at frequencies of 500 Hz, 1000 Hz, and 1500 Hz, scanning the potential from − 1.6 V to + 1.6 V vs. Ag/AgCl. Electrochemical Impedance Spectroscopy (EIS): Performed in 0.1 mol·L − 1 Na 2 SO 4 electrolyte over a frequency range of 10 6 to 10 3 Hz. Declarations A cknowledgments The work was financially supported by the National Natural Science Foundation of China (22377047, 22221001), Science and Technology Leading Talent Foundation of Gansu Province (24RCKB003), and the Special Funds of the National Natural Science Foundation of China (Grant No. 223B2204). Author contributions B.D.W. designed and conceived the whole project. G.P.M., S.W. and H. W. carried out the materials synthesis, experimental test, and data analysis. G.P.M. and B.D.W. wrote the paper together. B.D.W. and H. W. revised the manuscript. N. L. and G.P.M. conducted theoretical calculations. G.P.M, S.W., Y.H.Y., B. D. and G.W.H. performed the electron microscopy tests. All the authors contributed to the discussion during the whole project. Additional information Supplementary information is available in the online version of the paper. Reprints and permissions information is available online at www.nature.com/reprints. Correspondence and requests for materials should be addressed to Baodui Wang. Competing financial interests The authors declare no competing financial interests. References Zhang C , et al. Stable and high-yield hydrogen peroxide electrosynthesis from seawater. Nat. Sustain. 8 , 542-552 (2025). Xia C, Xia Y, Zhu P, Fan L, Wang H. Direct electrosynthesis of pure aqueous H 2 O 2 solutions up to 20% by weight using a solid electrolyte. Science 366 , 226-231 (2019). Campos-Martín JM, Blanco-Brieva G, Fierro JLG. Hydrogen peroxide synthesis: an outlook beyond the anthraquinone process. Angew. Chem. Int. Ed. 45 , 6962-6984 (2006). Wang W , et al. Photothermal-enabled single-atom catalysts for high-efficiency hydrogen peroxide photosynthesis from natural seawater. Nat. Commun. 14 , 2493 (2023). Huang W , et al. Intermarrying MOF Glass and Lead Halide Perovskites for Artificial Photosynthesis. J. Am. Chem. Soc. 174 , 3195-3205 (2024). Zhao W , et al. Perovskite-Doped Modulated Color-Selective Photosynaptic Transistors for Target Object Recognition. Nano lett. 24 , 9937-9945 (2024). Sun K , et al. Three-dimensional direct lithography of stable perovskite nanocrystals in glass. Science 375 , 307-310 (2022). Hou J , et al. Liquid-phase sintering of lead halide perovskites and metal-organic framework glasses. Science 374 , 621-625 (2021). Tian T , et al. Large-area waterproofand durable perovskite luminescent textiles. Nat. Commun. 14 , 234 (2023). Fan X , et al. Brightened Bicomponent Perovskite Nanocomposite Based on Forster Resonance Energy Transfer for Micro-LED Displays. Adv. Mater. 35 , e2300834 (2023). Jiang W , et al. Axial alignment of covalent organic framework membranes for giant osmotic energy harvesting. Nat. Sustain. 8 , 446-455 (2025). Suleman S , et al. Turning on Singlet Oxygen Generation by Outer-Sphere Microenvironment Modulation in Porphyrinic Covalent Organic Frameworks for Photocatalytic Oxidation. Angew. Chem. Int. Ed. 63 , e202314988 (2024). Li P , et al. Development of an energy-dense and high-power Li-Cl 2 battery using reversible interhalogen bonds. Chem 10 , 352-364 (2023). Wu J , et al. Regioselective Multisite Atomic-Chlorine Passivation Enables Efficient and Stable Perovskite Solar Cells. J. Am. Chem. Soc. 145 , 5872-5879 (2023). Liu Y , et al. Halogen atom-induced local asymmetric electron in covalent organic frameworks boosts photosynthesis of hydrogen peroxide from water and air. Matter 8 , 102076 (2025). Zhang L , et al. Deep-red emissive colloidal lead-based triiodide perovskite/telluride nanoscale heterostructures with reduced surface defects and enhanced stability for indoor lighting applications. Nano Energy 90 , 106506 (2021). Wang A , et al. Halogenated Hole Selective Contact Enhances Interfacial Weak Bonding of Perovskite Solar Cells. Advanced Energy Materials 14 , 2400640 (2024). He J , et al. Competitive Formation Mechanism for Bidentate Passivation of Halogen Vacancies in Perovskite Based on 6‐Chloropurine. Small 20 , 2305127 (2023). Fu X , et al. Halogen-halogen bonds enable improved long-term operational stability of mixed-halide perovskite photovoltaics. Chem 7 , 3131-3143 (2021). Wang P , et al. Ultrastable Perovskite-Zeolite Composite Enabled by Encapsulation and In Situ Passivation. Angew. Chem. Int. Ed. 59 , 23100-23106 (2020). Qiu J , et al. COF/In 2 S 3 S-Scheme Photocatalyst with Enhanced Light Absorption and H H 2 O 2 -Production Activity and fs-TA Investigation. Adv. Mater. 36 , e2400288 (2024). Li F, Yue X, Liao Y, Qiao L, Lv K, Xiang Q. Understanding the unique S-scheme charge migration in triazine/heptazine crystalline carbon nitride homojunction. Nat. Commun. 14 , 3901 (2023). Zhang Y , et al. H 2 O 2 generation from O 2 and H 2 O on a near-infrared absorbing porphyrin supramolecular photocatalyst. Nat. Energy 8 , 361-371 (2023). Cao D, Du J, Li J, Sun Q, Guan J, Liu J. Hydroxyl-Functionalized Donor-Acceptor Covalent Organic Frameworks for Efficient Photocatalytic Hydrogen Peroxide Production under Visible Light. ACS Catal. 15 , 3584-3594 (2025). Yu Z, Yu F, Xu M, Feng S, Qiu J, Hua J. Amidoxime-Functionalized sp 2 -Carbon-Conjugated Covalent Organic Frameworks for Overall Photocatalytic Hydrogen Peroxide Production. Adv. Sci. 12 , e2415194 (2025). Shi X , et al. Interfacial Engineering of β‐Ketoenamine‐Based COFs/Urea‐Linked Perylene Diimide for Overall Photosynthesis of H 2 O 2 in Seawater. Adv. Funct. Mater. 35 , 2414755 (2024). Yue JY, Pan ZX, Zhang RZ, Xu Q, Yang P, Tang B. One‐Pot Synthesis of Fully Conjugated Covalent Organic Frameworks via the Pictet–Spengler Reaction for Boosting H 2 O 2 Photogeneration in Real Seawater. Adv. Funct. Mater. 35 , 2421514 (2025). Hou Y, Liu F, Liang J, Li Z, Zhou P, Tong M. Building a Confluence Charge Transfer Pathway in COFs for Highly Efficient Photosynthesis of Hydrogen Peroxide from Water and Air. Angew. Chem. Int. Ed. 64 , e202505621 (2025). Yue JY , et al. Thiophene-Containing Covalent Organic Frameworks for Overall Photocatalytic H 2 O 2 Synthesis in Water and Seawater. Angew. Chem. Int. Ed. 62 , e202309624 (2023). Huang Y , et al. Nb 2 C MXene quantum dots modulate built-in electric field within heterostructures for efficient solar-to-H2O2 conversion from seawater. Appl. Catal. B Environ. 371 , 125262 (2025). Yue J-Y , et al. Overall solar H 2 O 2 generation in water and real seawater by covalent organic frameworks with kgd topology. Chem. Eng. J 504 , 158983 (2025). Tang B , et al. Photocatalytic Synthesis of Hydrogen Peroxide by Rhenium Modified Metal-Organic Frameworks Incorporating Bianthracene Ligands. J. Am. Chem. Soc. , 10.1021/jacs.1025c00603 (2025). Hu ZY, Liu T, Yang YR, An AK, Liew KM, Li WW. Binder-Free Immobilization of Photocatalyst on Membrane Surface for Efficient Photocatalytic H 2 O 2 Production and Water Decontamination. Nanomicro Lett. 17 , 301 (2025). Ju Y , et al. Regulating the electronic structure of covalent organic frameworks via heterocyclic isomers for highly efficient photocatalytic H 2 O 2 generation. Nat. Commun. 16 , 5658 (2025). Zhou J , et al. Covalent Organic Frameworks with Constrained Palladium Nanoclusters Lock the Enol-Keto Tautomerism Enhancing Hydrogen Peroxide Photosynthesis. Small 21 , e2409006 (2025). Chen J , et al. Redox-Mediated TEMPO-Based Donor-Acceptor Covalent Organic Framework for Efficient Photo-Induced Hydrogen Peroxide Generation. Angew. Chem. Int. Ed. 64 , e202500924 (2025). Gu M, Yang Y, Zhang L, Zhu B, Liang G, Yu J. Efficient sacrificial-agent-free solar H 2 O 2 production over all-inorganic S-scheme composites. Appl. Catal. B Environ. 324 , 122227 (2023). Luo Y , et al. Sulfone-Modified Covalent Organic Frameworks Enabling Efficient Photocatalytic Hydrogen Peroxide Generation via One-Step Two-Electron O 2 Reduction. Angew. Chem. Int. Ed. 62 , e202305355 (2023). Zhou S , et al. Chloromethylation Modified Pyranonitrile-Based Conjugated Microporous Polymers for Selective One-Step Two-Electron O 2 Reduction to H 2 O 2 . Angew. Chem. Int. Ed. , 10.1002/anie.202508436 (2025). Wang W , et al. Graphene Quantum Dot‐Modified Mn 0.2 Cd 0.8 S for Efficient Overall Photosynthesis of H 2 O 2 . Adv. Funct. Mater. , 10.1002/adfm.202422307 (2025). Xu T , et al. Constructing Photocatalytic Covalent Organic Frameworks with Aliphatic Linkers. J. Am. Chem. Soc. 146 , 20107-20115 (2024). Li Z , et al. Covalent Organic Frameworks for Boosting H 2 O 2 Photosynthesis via the Synergy of Multiple Charge Transfer Channels and Polarized Field. Angew. Chem. Int. Ed. 64 , e202420218 (2025). Liu R , et al. Linkage-engineered donor–acceptor covalent organic frameworks for optimal photosynthesis of hydrogen peroxide from water and air. Nat. Catal. 7 , 195-206 (2024). Scheme Scheme 1 is available in the Supplementary Files section. Additional Declarations There is NO Competing Interest. Supplementary Files SupportingInformation.docx Supporting Information GraphicalAbstract.docx Scheme1.docx Cite Share Download PDF Status: Published Journal Publication published 15 Mar, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7187821","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":509912202,"identity":"2872835b-a90d-458d-bb5e-ec3687199151","order_by":0,"name":"Baodui Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYJACZgY2Gxkwi4cELWk8JGs5TIIWg+NnD78uKDvPIz8jgfHB2zYGeXOCWs7kpVnPOHebh3FGArPh3DYGw50NBLSYHcgxM+Ztu83DLJHAJs3bxpBgcICQlvNvQFrO8bBJJLD/Jk7LjRzjx7xtB3h4gLYwE6XF/sYbM+YZ55J5JHgeNkvOOSdhuIGQFsn+HOPPBWV2cvLtyQc/vCmzkSdoCxCwSUBoxgYgIUFYPRAwfyBK2SgYBaNgFIxcAADxPjmOpxwTPwAAAABJRU5ErkJggg==","orcid":"","institution":"Lanzhou University","correspondingAuthor":true,"prefix":"","firstName":"Baodui","middleName":"","lastName":"Wang","suffix":""},{"id":509912203,"identity":"2f5dfbd7-8b62-478f-a6e5-9cad4d80a7bf","order_by":1,"name":"Genping Meng","email":"","orcid":"https://orcid.org/0000-0002-3531-5235","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Genping","middleName":"","lastName":"Meng","suffix":""},{"id":509912204,"identity":"1f2afe01-bce1-4e73-85f2-4981e100c508","order_by":2,"name":"Shuai Wei","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Shuai","middleName":"","lastName":"Wei","suffix":""},{"id":509912205,"identity":"966c644d-e879-40b1-a348-1632520e6257","order_by":3,"name":"Ning Li","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Ning","middleName":"","lastName":"Li","suffix":""},{"id":509912206,"identity":"d01a58af-883d-45a0-86bb-6ba311b6a5c2","order_by":4,"name":"Yuhui Yin","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yuhui","middleName":"","lastName":"Yin","suffix":""},{"id":509912207,"identity":"237a3c90-78e1-4145-b746-0d6d9d9a41b3","order_by":5,"name":"Bin Dong","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Bin","middleName":"","lastName":"Dong","suffix":""},{"id":509912208,"identity":"92fce992-9545-460d-b7eb-bd9d94b0130d","order_by":6,"name":"Shihao Sun","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Shihao","middleName":"","lastName":"Sun","suffix":""},{"id":509912209,"identity":"8b3ecc7a-bb36-44aa-b7d2-33c65bc05995","order_by":7,"name":"Guowen Hu","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Guowen","middleName":"","lastName":"Hu","suffix":""},{"id":509912210,"identity":"e3945ca4-30c5-4bba-93c8-946b49734f1a","order_by":8,"name":"Hao Wang","email":"","orcid":"","institution":"Lanzhou University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-07-22 13:57:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7187821/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7187821/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-70503-2","type":"published","date":"2026-03-15T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":91509949,"identity":"32aa642d-90f1-4de6-80e0-00946f859c87","added_by":"auto","created_at":"2025-09-17 08:42:18","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":456541,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStructural and Physicochemical Characterizations of the CsPbI3@COF-Cl System. \u003c/strong\u003e(a, b) TEM images showing CsPbI\u003csub\u003e3\u003c/sub\u003e QDs (black dots) confined within the COF-Cl matrix. (c) HRTEM image revealing the lattice spacing of an individual confined CsPbI\u003csub\u003e3\u003c/sub\u003e QD. (d) Fast-Fourier transform (FFT) pattern corresponding to the CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, confirming their high crystallinity. (e) HAADF-STEM image of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. Inset: Size distribution histogram of the confined CsPbI\u003csub\u003e3\u003c/sub\u003e QDs. (f) EDX elemental mapping showing the spatial distribution of Cl, Cs, Pb, and I within the same region. (g) Contact angle measurements on a pressed tablet of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl composite. (h) PXRD pattern of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (i) N\u003csub\u003e2\u003c/sub\u003e sorption isotherms at 77 K and corresponding Brunauer-Emmett-Teller (BET) surface area plots for pristine COF-Cl and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl composite.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/046c8309a5d50d3e2bb33506.png"},{"id":91507478,"identity":"901709e0-6029-4b6d-9b1c-25cc73de259a","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":276065,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvaluation of interaction mechanisms between the COF-Cl matrix and CsPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e QDs. \u003c/strong\u003e(a) High-resolution Pb 4f XPS spectra of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and the CsPbI3@COF-Cl system. (b) High-resolution I 3d XPS spectra of CsPbI3 QDs and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (c) High-resolution Cl 2p XPS spectra of the COF-Cl matrix and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (d) Solid-state \u003csup\u003e13\u003c/sup\u003eC NMR spectra of the COF-Cl matrix and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (e) Time-resolved photoluminescence decay (TRPL) spectra of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. White curves represent biexponential fitting curves. (f) Adsorption models of the COF-Cl matrix showing continuous multisite bonding on Pb-rich and I-rich surfaces of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs. (g) Simulated formation energies for continuous and discontinuous COF-Cl binding on the PbI-terminated CsPbI\u003csub\u003e3\u003c/sub\u003e QD surface. (h) Simulated binding energies for continuous and discontinuous COF-Cl binding on the PbI-terminated CsPbI\u003csub\u003e3\u003c/sub\u003e QD surface. (i) Root mean square deviation (RMSD) over time from molecular dynamics simulations.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/83c3aac45a47fc2f10ac78a0.png"},{"id":91507474,"identity":"e521257f-fde0-4849-a5a2-e5b1383f4538","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":269113,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharacterizations of the optical properties for CsPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e QDs, COF-X matrix and CsPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e@COF-X system.\u003c/strong\u003e (a) UV-vis DRS spectra of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, COF-X matrix and CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X system. (b) Electronic band structures for CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-X matrix (versus NHE). (c) Schematics of photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production over CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (d) PL spectra of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (e) Photocurrent response curves of COF-Cl matrix and CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X system. (f) XPS Cl 2p spectra of CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (g) XPS Pb 4f spectra CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (h) XPS I 3d spectra CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/f77e327c30b0369d61e8439a.png"},{"id":91507473,"identity":"b070e780-0600-4741-8274-29c8dd711226","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":153710,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhotocatalytic H\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e production over COF-Cl and CsPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e@COF-Cl\u003c/strong\u003e. (a) Kinetic curves of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation in pure water under air. (b) Kinetic curves of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation in seawater under air. (c) Apparent quantum yield (AQY) of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. (d) Long-term stability test: Continuous H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production over CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl under visible-light irradiation for 20 h. (e) XRD patterns of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system before and after the 20 h reaction in (d), confirming structural stability. (f) Outdoor experiment: H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yield, system temperature, and light intensity recorded over time for CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl. (g) Performance comparison of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system with recently reported photocatalysts for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production in pure water and seawater.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/5a676a2e8770f871bdedb325.png"},{"id":91507470,"identity":"30808803-d6f6-47c3-b447-89321185838f","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":143980,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMechanistic investigations of the photocatalytic H\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e production over the CsPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e@COF-Cl system.\u003c/strong\u003e (a) EPR spectra using DMPO as a spin-trapping agent for •O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e under dark conditions and visible light illumination. (b) EPR spectra using TEMP as a spin-trapping agent for \u003csup\u003e1\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e under dark conditions and visible light illumination. (c) H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e\u003cstrong\u003e \u003c/strong\u003eyields in the presence of specific scavengers: p-benzoquinone (p-BQ, •O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e scavenger), L-tryptophan (Trp, \u003csup\u003e1\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e scavenger), silver nitrate (AgNO\u003csub\u003e3\u003c/sub\u003e, electron (e⁻) scavenger), and ethanol (EtOH, hole (h⁺) scavenger). (d) Time-course in situ diffuse reflectance Fourier transform infrared spectroscopy (DRIFTS) of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system under visible-light irradiation in an O\u003csub\u003e2\u003c/sub\u003e atmosphere. Free energy diagram for photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e synthesis via the (e) direct 2e⁻ oxygen reduction reaction (ORR) pathway and (f) direct 2e⁻ water oxidation reaction (WOR) pathway on the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/8e2c039934c6242fca1d6aab.png"},{"id":108391173,"identity":"c0c91cd3-95ff-4599-acc9-1e03478d5577","added_by":"auto","created_at":"2026-05-04 07:06:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1729911,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/d0780667-531e-4696-9703-c7fa9c12615b.pdf"},{"id":91507477,"identity":"4e9ae816-1f68-4812-ba2a-68171fac5107","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5315293,"visible":true,"origin":"","legend":"Supporting Information","description":"","filename":"SupportingInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/5b84902986e5cf49511f0128.docx"},{"id":91507472,"identity":"6600e18a-09f5-4d01-be8a-37659174df32","added_by":"auto","created_at":"2025-09-17 08:34:18","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":134407,"visible":true,"origin":"","legend":"","description":"","filename":"GraphicalAbstract.docx","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/a5f5271015fd317b2bfbd57a.docx"},{"id":91509950,"identity":"a20534d3-0a17-4e95-8328-ff8eb042645e","added_by":"auto","created_at":"2025-09-17 08:42:18","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":468285,"visible":true,"origin":"","legend":"","description":"","filename":"Scheme1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7187821/v1/45ca77a6b28db43ba015a045.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"\u003cp\u003eMultisite atomic-chlorine-passivation stabilizes perovskite interfaces for efficient H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e photosynthesis from seawater\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHydrogen peroxide (H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e) is an indispensable chemical oxidant with burgeoning applications in water treatment, disinfection, and green synthesis\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. However, the incumbent anthraquinone process for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production is energy-intensive, generates significant waste, and relies on centralized facilities, limiting its accessibility and sustainability. Solar-driven photosynthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e directly from water (H\u003csub\u003e2\u003c/sub\u003eO) and oxygen (O\u003csub\u003e2\u003c/sub\u003e) offers a compelling decentralized and sustainable alternative\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Seawater, constituting over 96% of Earth's hydrosphere, presents an abundant aqueous resource for this process, yet its complex ionic composition and corrosivity pose formidable challenges to photocatalyst stability and performance\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. The development of robust, seawater-compatible photocatalytic systems is therefore essential to unlock scalable solar H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e synthesis.\u003c/p\u003e\u003cp\u003eLead halide perovskites, particularly CsPbI\u003csub\u003e3\u003c/sub\u003e quantum dots (QDs), have emerged as outstanding light harvesters for artificial photosynthesis\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. This is attributed to their extensive visible-light absorption, suitable band edges, and high charge-carrier mobility\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. However, their well-known vulnerability to ionic dissolution and structural degradation in aqueous environments, especially in ion-rich seawater, significantly limits their practical applications\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Although conventional encapsulation strategies using polymers or metal oxides offer some degree of barrier protection, they often impede mass transport (such as O\u003csub\u003e2\u003c/sub\u003e diffusion) or fail to completely suppress interfacial ion migration, which is a crucial degradation pathway\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Covalent organic frameworks (COFs) have recently gained attention as protective matrices due to their crystallinity, designable porosity, and functional versatility\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Notably, chlorine-functionalized COFs (COF-Cl) remain underexplored, despite the potential of atomic chlorine sites to form strong, directional bonds with perovskite surfaces through halogen interactions (Cl-I) and coordination (Cl-Pb)\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Such atomic-scale passivation could simultaneously stabilize the perovskite and enhance interfacial charge dynamics.\u003c/p\u003e\u003cp\u003eHerein, we introduce a novel interfacial engineering strategy to simultaneously passivate CsPbI\u003csub\u003e3\u003c/sub\u003e QDs against aqueous degradation and enhance its photocatalytic activity for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production directly from seawater (Scheme \u003cspan refid=\"Sch1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We rationally design and synthesize COFs featuring precisely positioned, multisite atomic chlorine functionalities. The COF-Cl matrix stabilizes CsPbI\u003csub\u003e3\u003c/sub\u003e QDs through dual halogen bonding (Cl-I) and coordination (Cl-Pb), effectively suppressing ionic migration while creating a hydrophobic triphase interface to enhance O\u003csub\u003e2\u003c/sub\u003e diffusion. Crucially, the S-scheme heterojunction forms between the CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-Cl, enabling efficient spatial separation of charge carriers: electrons localized on the COF-Cl drive the 2e⁻ ORR, while holes retained on the CsPbI\u003csub\u003e3\u003c/sub\u003e QDs drive the WOR. This system achieves record production rates of 25.29 mmol h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (pure water) and 20.37 mmol h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (natural seawater) with \u0026gt;\u0026thinsp;90% stability over 20 hours in seawater and operates efficiently under natural sunlight. Through integrated experimental and theoretical analyses, we elucidate the synergistic interfacial mechanisms driving this performance. Our work establishes a paradigm for robust perovskite-based photosynthesis, unlocking scalable solar-driven H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e synthesis from abundant seawater.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003eSynthesis and characterization of hydrophobic halogenated COFs and CsPbI\u003c/b\u003e\u003csub\u003e\u003cb\u003e3\u003c/b\u003e\u003c/sub\u003e\u003cb\u003e@COF-X.\u003c/b\u003e\u003c/p\u003e\u003cp\u003eA series of hydrophobic halogenated covalent organic frameworks (COF-X: Cl, Br, I) were synthesized via Schiff-base condensation between 4,4',4''-(1,3,5-triazine-2,4,6-triyl) trianiline and halogenated 2,5-dihaloterephthalaldehydes (Cl, Br, I) (Figure S2). Highly crystalline materials were obtained in o-dichlorobenzene/n-BuOH/6 M HAc (5:5:1, v/v) at 120\u0026deg;C for 72 h. Structural analysis confirmed successful synthesis and high crystallinity (Figure S3-S5). Powder X-ray diffraction (PXRD) patterns (Figure S6) showed prominent (100) peaks at 2.79\u0026deg;, 2.75\u0026deg;, and 2.70\u0026deg; for COF-Cl, COF-Br, and COF-I, respectively, matching simulated AA stacking models (Figures S3-S5, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Solid-state \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003eCNMR revealed the characteristic imine peak (~\u0026thinsp;160 ppm, Figure S7). Fourier transform infrared (FTIR) spectra (Figure S8) confirmed Schiff-base formation by the disappearance of precursor -NH\u003csub\u003e2\u003c/sub\u003e (3213, 3327 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and aldehyde (1688 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) stretches and the appearance of -C\u0026thinsp;=\u0026thinsp;N stretches (1633 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) for all three COF-X\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. COF-X exhibit mesoporosity with type IV isotherms (Figure S9). Brunauer-Emmett-Teller (BET) surface areas decreased with increasing halogen size: 2289.7 m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e g\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e (COF-Cl), 1757.0 m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e g\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e (COF-Br), 1150.9 m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e g\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e (COF-I), correlating with pore diameters of 3.53 nm, 3.09 nm, and 2.63 nm, respectively (Table S2). Scanning electron microscope (SEM) and transmission electron microscope (TEM) images (Figures S10, S11) revealed uniform fibrous morphologies and layered structures, with visible pores in COF-X HR-TEM (Figure S11b). X-ray photoelectron spectra (XPS) confirmed elemental composition (Figures S12-S14). Water contact angles of 143.0\u0026deg;, 145.5\u0026deg;, and 147.5\u0026deg; (Figure S15) demonstrate increasing hydrophobicity with halogen size. Excellent thermal stability was confirmed by TGA, showing negligible weight loss at 150\u0026deg;C (Figure S16). These results confirm the successful synthesis of crystalline, mesoporous, hydrophobic halogenated COFs.\u003c/p\u003e\u003cp\u003eA sequential deposition route was employed to synthesize CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X composites (Figure S17). First, COF-X matrices were immersed in a PbI\u003csub\u003e2\u003c/sub\u003e precursor solution at 70\u0026deg;C. Residual PbI\u003csub\u003e2\u003c/sub\u003e was removed via centrifugation using DMF/ethanol. The resulting PbI\u003csub\u003e2\u003c/sub\u003e@COF-X intermediates were then dispersed in a CsI-containing DMF solution under ultrasonication. Finally, the reaction mixture was precipitated in toluene at room temperature to obtain a series of CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X systems, where CsPbI\u003csub\u003e3\u003c/sub\u003e QDs in-situ nucleated and crystallized inside the COF-X matrix (for a full description of the methods, see the Supplementary information). Taking CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl as an example, we demonstrate how the framework and topology of the COF-Cl matrix direct the confined growth of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs. Structural characterization confirmed the successful synthesis of CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl. SEM imaging (Figure S18) revealed clean surfaces, indicating CsPbI\u003csub\u003e3\u003c/sub\u003e QDs are embedded within the COF-Cl matrix. TEM analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) further shows CsPbI\u003csub\u003e3\u003c/sub\u003e QDs uniformly dispersed in the mesopores of COF-Cl matrix. Isolated CsPbI\u003csub\u003e3\u003c/sub\u003e QDs with an average diameter of 3.2 nm are clearly resolved (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), consistent with the COF-Cl mesopore size. High-resolution TEM (HRTEM, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec) displays distinct lattice fringes (0.63 nm spacing), matching the (100) planes of α-CsPbI\u003csub\u003e3\u003c/sub\u003e (space group \u003cem\u003ePm-3m\u003c/em\u003e). The crystalline structure is further confirmed by fast-Fourier transform (FFT) patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). STEM-HAADF imaging (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee) corroborates the uniform dispersion of 3.2 nm CsPbI\u003csub\u003e3\u003c/sub\u003e QDs within the COF-Cl pores. Elemental mapping (EDX, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef) confirms homogeneous distribution of Cl, Cs, Pb, and I throughout the composite. Hydrophobicity, critical for aqueous applications, was evaluated via contact angle measurements (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg). The CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system exhibits strong hydrophobicity (139.5\u0026deg;), attributable to the COF-Cl matrix. Consequently, the lightweight composite spreads uniformly on water surfaces (Figures S19a, b), enabling redox reactions at the air-liquid-solid interface of the floating photocatalyst (Figure S20). PXRD analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh) verifies crystallinity, with peaks at 14.1\u0026deg;, 19.9\u0026deg;, 24.5\u0026deg;, and 28.4\u0026deg; corresponding to the (100), (110), (111), and (200) planes of α-CsPbI\u003csub\u003e3\u003c/sub\u003e material (PDF No. 01-080-4039)\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. N\u003csub\u003e2\u003c/sub\u003e adsorption-desorption isotherms (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ei) provide additional evidence for CsPbI\u003csub\u003e3\u003c/sub\u003e QDs encapsulation, showing reduced pore volumes and specific surface area due to in situ CsPbI\u003csub\u003e3\u003c/sub\u003e QDs growth within the mesopores. Collectively, these results confirm the successful construction of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo evaluate the interfacial interaction between COF-Cl and CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, we employed X-ray photoelectron spectroscopy (XPS), solid-state \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003eC nuclear magnetic resonance (NMR) spectroscopy, and time-resolved fluorescence decay spectroscopy (TRFDS) to analyze the electronic structure of constituent elements. XPS analysis confirmed the elemental composition of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system, with distinct peaks for C, N, Cl, Cs, Pb, and I (Figure S21), verifying successful synthesis. Notably, the Pb 4f XPS spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea) revealed a distinct shift toward higher binding energy in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl compared to pristine CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, indicating interaction between under-coordinated Pb\u003csup\u003e2+\u003c/sup\u003e and Cl atoms\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Similarly, the I 3d spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb) showed a slight shift to higher binding energy, consistent with Cl-I bond formation and interfacial charge transfer between CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-Cl\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Conversely, the Cl 2p peaks shifted toward lower binding energy (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), further evidencing strong chemical interactions. To further investigate the chemical environment changes, solid-state \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003eC NMR spectroscopy was performed. The spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed) revealed a downfield shift of 0.3 ppm for the carbon atoms directly bonded to chlorine in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl relative to pristine COF-Cl. This shift provides evidence for the formation of halogen-halogen (Cl-I) bonds and coordination interactions (Cl-Pb) at the interface. To directly assess the impact of these interactions on defect passivation, TRFDS measurements were conducted (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee, Table S3). Fluorescence decay curves were fitted to a tri-exponential model, where fast decay correlates with surface/crystal defects and slow decay with carrier transport. Pristine CsPbI\u003csub\u003e3\u003c/sub\u003e QDs exhibited rapid decay (τ\u003csub\u003e1\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;3.13 ns, τ\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;13.68 ns, τ\u003csub\u003e3\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;62.22 ns; τₐ\u003csub\u003ev\u003c/sub\u003e = 24.96 ns), attributed to non-radiative recombination from Pb-I antisite defects and vacancies\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. In contrast, CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl showed significantly slower decay (τ\u003csub\u003e1\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;9.60 ns, τ\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;57.52 ns, τ\u003csub\u003e3\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;289.41 ns; τₐ\u003csub\u003ev\u003c/sub\u003e = 174.38 ns), demonstrating that COF-Cl passivates defects via halogen-halogen bonds and coordination interactions, thereby suppressing non-radiative pathways\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eTo elucidate the multi-site bonding mechanism of the halogen-rich porous COF-Cl matrix in detail, we investigated the adsorption modes of the COF-Cl matrix on CsPbI\u003csub\u003e3\u003c/sub\u003e QD surfaces, along with the corresponding binding and formation energies. Computational modeling (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, S22) reveals that COF-Cl can achieve continuous multi-site bonding. Specifically, there is Cl-Pb coordination on Pb-rich surfaces and Cl-I bonding on I-rich surfaces. The formation energies for continuous multi-site Cl-Pb and Cl-I bonds are highly favorable at -2.76 eV and \u0026minus;\u0026thinsp;2.56 eV, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg). Similarly, the binding energies are strong at -1.54 eV and \u0026minus;\u0026thinsp;1.32 eV, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh). For comparison, selective removal of Cl functional groups to create discontinuous passivation (Figure S23) resulted in significantly reduced binding and formation energies, confirming that continuous multisite passivation optimally stabilizes the perovskite lattice. Furthermore, molecular dynamics simulations assessed structural stability. The root mean square deviation (RMSD) curves (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ei), which measure atomic position fluctuations over time, stabilized after ~\u0026thinsp;1 ps, indicating system equilibrium. Throughout the simulation, the RMSD values for systems with continuous multi-site Cl-Pb and Cl-I passivation remained consistently lower than those with discontinuous passivation. This demonstrates minimal atomic displacement and superior structural stability afforded by the continuous multi-site bonding mechanism.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eCharacterizations of the electronic properties and the band edge positions.\u003c/b\u003e\u003c/p\u003e\u003cp\u003eBased on the ultraviolet-visible diffuse reflection spectra (UV-vis DRS), the COF-X matrix exhibits adsorption peaks (Soret band) around 350\u0026ndash;500 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea), attributed to its periodic molecular framework and long-range ordered π-stacking structure. In contrast, CsPbI\u003csub\u003e3\u003c/sub\u003e QDs show a broad absorption band extending to 700 nm, indicative of a narrow bandgap. Notably, upon encapsulation of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs by COF-X to form the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X system, the absorption edge of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X system undergoes a significant redshift beyond 700 nm, enabling more efficient solar energy utilization. This redshift clearly demonstrates the synergistic effect between COF-X and CsPbI\u003csub\u003e3\u003c/sub\u003e QDs. Tauc plot analysis (Figure S24) determined the bandgap energies (E\u003csub\u003eg\u003c/sub\u003e) of COF-Cl, COF-Br, COF-I, and CsPbI\u003csub\u003e3\u003c/sub\u003e QDs to be 2.50 eV, 2.58 eV, 2.49 eV, and 1.77 eV, respectively, confirming their semiconductor nature. For efficient photocatalytic reactions, the electronic band structure must align with the relevant redox potentials. Valence band XPS spectra (VB-XPS, Figure S25) measured the valence band maximum (E\u003csub\u003eVB\u003c/sub\u003e) positions relative to the vacuum level (E\u003csub\u003eVB, XPS\u003c/sub\u003e). After conversion to the standard hydrogen electrode (NHE) scale, the E\u003csub\u003eVB, NHE\u003c/sub\u003e values are: COF-Cl (2.52 eV), COF-Br (2.39 eV), COF-I (2.27 eV), and CsPbI\u003csub\u003e3\u003c/sub\u003e QDs (0.94 eV). The resulting band diagrams are plotted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb. Crucially, this staggered band alignment satisfies the prerequisite for forming S-scheme heterojunction. The results indicate that CsPbI\u003csub\u003e3\u003c/sub\u003e QDs possess strong reducing capabilities, while the COF-X matrices exhibit high oxidizing capabilities, with COF-Cl showing the strongest oxidizing potential. This favorable energy level alignment within the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl composite not only maximizes the redox capabilities of the individual components but also facilitates the construction of an efficient S-scheme heterojunction (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). The CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system significantly enhances charge transfer and carrier separation. Photoluminescence (PL) spectroscopy and TRFDS show substantially quenched PL intensity and prolonged fluorescence lifetime in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl compared to pure CsPbI\u003csub\u003e3\u003c/sub\u003e QDs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee), indicating suppressed electron-hole recombination due to spatial charge separation. These properties directly improve photochemical conversion activity. Photocurrent response tests (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee) show faster charge transfer in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl compared to COF-Cl alone. Pure CsPbI\u003csub\u003e3\u003c/sub\u003e QDs were excluded due to their aqueous instability. Electrochemical impedance spectroscopy (EIS) further confirms reduced charge-transfer resistance in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl, evidenced by a smaller Nyquist semicircle (Figure S26).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe interfacial electron transfer mechanism in the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl heterojunction was demonstrated by Ultraviolet photoelectron spectroscopy (UPS). The results are displayed in Figure S27 and Figure S28. The work functions of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-Cl are 5.83 eV and 7.00 eV, respectively. Correspondingly, the Fermi levels of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-Cl are \u0026minus;\u0026thinsp;5.83 eV and \u0026minus;\u0026thinsp;7.00 eV (vs vacuum level), respectively. This work function disparity establishes a higher Fermi level in CsPbI\u003csub\u003e3\u003c/sub\u003e QDs than in COF-Cl, driving electron migration from CsPbI\u003csub\u003e3\u003c/sub\u003e QDs to COF-Cl upon heterojunction formation\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Consequently, band bending at the interface induces charge redistribution, establishes a unified Fermi level, and generates an internal electric field (IEF)\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Further validation through \u003cem\u003ein-situ\u003c/em\u003e XPS under light illumination tracked binding energy shifts in key elements. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, light exposure caused the Cl 2p peaks in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl to shift from 202.2 eV and 200.6 eV to 202.6 eV and 201.0 eV, indicating electron loss from the COF-Cl component. Simultaneously, the Pb 4f peaks shifted from 144.29 eV/139.42 eV and 143.57 eV/138.70 eV to 143.92 eV/139.05 eV and 143.25 eV/138.38 eV (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg), while the I 3d peaks shifted from 632.12 eV/620.62 eV and 630.95 eV/619.45 eV to 632.46 eV/619.96 eV and 630.46 eV/618.96 eV (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh). These shifts collectively confirm electron gain by the CsPbI\u003csub\u003e3\u003c/sub\u003e QDs component. Therefore, photogenerated electrons transfer from COF-Cl to CsPbI\u003csub\u003e3\u003c/sub\u003e QDs under light excitation, directly evidencing the S-scheme charge transfer process. The combined UPS and XPS results support the proposed mechanism (Figure S29). Initially, the Fermi level difference drives the flow of electrons from CsPbI\u003csub\u003e3\u003c/sub\u003e QDs to COF-Cl, thereby forming the IEF. This leads to an upward band bending in CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and a downward band bending in COF-Cl at the interface. Under illumination, the IEF, Coulomb attraction, and band bending promote recombination of COF-Cl conduction band electrons with CsPbI\u003csub\u003e3\u003c/sub\u003e valence band holes. Significantly, this mechanism retains useful electrons in the CB of CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and holes in the VB of COF-Cl, maximizing their redox potential for enhanced photocatalytic oxygen reduction.\u003c/p\u003e\u003cp\u003e\u003cb\u003eFull reaction photosynthesis of H\u003c/b\u003e\u003csub\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sub\u003e\u003cb\u003eO\u003c/b\u003e\u003csub\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sub\u003e \u003cb\u003eon a three-phase interface.\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system demonstrates significant potential for the photocatalytic synthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e, as its energy band structure is sufficiently capable of driving both essential half-reactions: the oxidation of H\u003csub\u003e2\u003c/sub\u003eO (E(H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e/ H\u003csub\u003e2\u003c/sub\u003eO)\u0026thinsp;=\u0026thinsp;+\u0026thinsp;1.76 V vs. NHE, pH\u0026thinsp;=\u0026thinsp;0) and the reduction of O\u003csub\u003e2\u003c/sub\u003e (E( H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e/O\u003csub\u003e2\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;+\u0026thinsp;0.68 V vs. NHE, pH\u0026thinsp;=\u0026thinsp;0)\u003csup\u003e23\u003c/sup\u003e. This indicates its theoretical suitability as an effective photocatalyst for the full photosynthetic reaction producing H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Cyclic voltammetry (CV) further elucidates these redox capabilities. As shown in Figure S30, the CV curve identifies a reduction potential of -0.94 V (vs. Ag/AgCl, pH\u0026thinsp;=\u0026thinsp;7), corresponding to -0.33 V vs RHE (pH\u0026thinsp;=\u0026thinsp;0), assigned to the reduction of O\u003csub\u003e2\u003c/sub\u003e to \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e. Simultaneously, an oxidation potential of +\u0026thinsp;1.15 V (vs. Ag/AgCl, pH\u0026thinsp;=\u0026thinsp;7), equivalent to +\u0026thinsp;1.76 V vs RHE (pH\u0026thinsp;=\u0026thinsp;0), is observed for the oxidation of H\u003csub\u003e2\u003c/sub\u003eO to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Therefore, the combined theoretical and electrochemical data confirm that the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system can effectively drive the photochemical synthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e using only water and atmospheric oxygen.\u003c/p\u003e\u003cp\u003ePhotocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation was evaluated in both pure water and seawater under visible-light irradiation (λ\u0026thinsp;\u0026gt;\u0026thinsp;420 nm) and ambient air, without sacrificial agents. Benefiting from its lightweight nature and excellent hydrophobicity, the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system readily spreads across the water surface, facilitating redox reactions at this unique air-liquid-solid interface (Figure S20). Under air in pure water, the H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yield showed clear accumulation over time, exhibiting a near-linear relationship between production and irradiation duration for both the COF-Cl matrix alone and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). This indicates sustained high photosynthetic rates even under prolonged illumination. After 1 h, H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yields reached 15.57 mmol g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for COF-Cl and 25.29 mmol g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl, respectively (Figure S31). This significant enhancement in the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system is attributed to optimized charge separation efficiency achieved through interfacial band engineering. Given this promising performance, we systematically evaluated H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production under seawater and air conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). After 1 h of photochemical conversion, H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yields of 12.52 mmol g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 20.37 mmol g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e were achieved for the COF-Cl matrix and the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system, respectively (Figure S32). Furthermore, the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system demonstrated a notable apparent quantum yield (AQY) of 15.76% at 450 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec) and solar-to-chemical conversion (SCC) efficiency of 1.38%.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eCycling stability tests confirmed the system's durability. The CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system maintained superior H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production performance over five consecutive photocatalytic cycles in seawater under air (Figure S33 and S34). TEM imaging and XRD analysis revealed no significant structural changes after cycling in seawater (Figures S35 and S36), highlighting its robustness. To further verify operational stability in seawater, a continuous 20-h photocatalytic reaction was conducted (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), demonstrating stable H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production. PXRD measurements during operation further confirmed excellent structural stability under prolonged seawater exposure (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo demonstrate practical solar-to-chemical energy conversion, the H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production performance of CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl was evaluated under natural sunlight in Lanzhou City, China (May 1, 2025; Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef and S37). Real-time monitoring was carried out to track solar irradiance at 450 nm, system temperature, and H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yield. After 10 hours, an accumulated H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e concentration of 11.7 mmol L\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e was achieved, confirming efficient photocatalytic activity under real-world conditions. Combined with its high efficiency, these findings indicate that the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system holds significant promise for large-scale, solar-driven H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production from pure water and natural seawater. This performance is significantly higher than that of most reported photocatalysts (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg and Table S4)\u003csup\u003e\u003cspan additionalcitationids=\"CR25 CR26 CR27 CR28 CR29 CR30 CR31 CR32 CR33 CR34\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eTo elucidate the reaction mechanism of hydrogen peroxide (H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e) generation, electron paramagnetic resonance (EPR) spectroscopy was employed to detect key intermediates during photochemical conversion. Using 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMP) as a spin trap for singlet oxygen (\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e) and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) for superoxide anion radicals (\u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e), distinct signals for DMPO-\u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e and TEMP-\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e adducts were observed (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb)\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. The CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system exhibited significantly stronger intermediate signals than COF-Cl alone. Superoxide (\u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e), a key intermediate in H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation via oxygen reduction reactions (ORR), forms through O\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;e⁻ \u0026rarr; \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e (electrode potential: \u0026minus;0.33 V vs. RHE, pH\u0026thinsp;=\u0026thinsp;0) and subsequently reacts via \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e + e⁻ + 2H⁺ \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Singlet oxygen (\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e) is generated by \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e oxidation by photogenerated holes (h⁺): \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e + h⁺ \u0026rarr; \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e, a pathway consistent with prior reports on photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Furthermore, \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e contributes to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e formation through a two-electron reduction: \u003csup\u003e1\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e + 2e⁻ + 2H⁺ \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Alternatively, H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e can form directly via a two-electron ORR: O\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;2e⁻ + 2H⁺ \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e (0.68 V vs. RHE, pH\u0026thinsp;=\u0026thinsp;0). The significant EPR signals for both \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e and \u0026bull; O\u003csub\u003e2\u003c/sub\u003e⁻ observed in the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system demonstrate that H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation proceeds via both pathways, contributing to its enhanced photosynthetic efficiency. We also explored the contribution of the water oxidation reaction (WOR) to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production. Previous studies indicate that water can react with holes to form hydroxyl radicals (H\u003csub\u003e2\u003c/sub\u003eO\u0026thinsp;+\u0026thinsp;h⁺ \u0026rarr; \u0026bull;OH\u0026thinsp;+\u0026thinsp;H⁺), which dimerize to produce H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e (2\u0026bull;OH \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e)\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. This process has a standard electrode potential of 2.72 V vs. RHE (pH\u0026thinsp;=\u0026thinsp;0). However, no \u0026bull;OH signals are detected in the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system, as their VB positions are more negative than the standard redox potential of H\u003csub\u003e2\u003c/sub\u003eO/\u0026bull;OH, rendering them thermodynamically incapable of oxidizing H\u003csub\u003e2\u003c/sub\u003eO to generate \u0026bull;OH radicals\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Alternatively, H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e can be generated directly from water and holes (2H\u003csub\u003e2\u003c/sub\u003eO\u0026thinsp;+\u0026thinsp;2h⁺ \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;2H⁺; E\u0026thinsp;=\u0026thinsp;1.76 V vs. RHE, pH\u0026thinsp;=\u0026thinsp;0), suggesting WOR also contributes to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e formation.\u003c/p\u003e\u003cp\u003eTo validate this hypothesis and explore alternative H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e-production pathways, a series of control experiments were performed over CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, p-benzoquinone (p-BQ), tryptophan (Trp), AgNO\u003csub\u003e3\u003c/sub\u003e, and ethanol (EtOH) were employed as scavengers for \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e, \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e, e\u003csup\u003e\u0026minus;\u003c/sup\u003e, and h\u003csup\u003e+\u003c/sup\u003e, respectively\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. The addition of p-BQ drastically reduced the H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yield, underscoring the critical role of \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e in H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation. Similarly, Trp moderately suppressed H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production, confirming the participation of \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e in the formation process. Notably, while AgNO\u003csub\u003e3\u003c/sub\u003e and EtOH also diminished H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yields, a significant amount of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e persisted, suggesting that both e\u003csup\u003e\u0026minus;\u003c/sup\u003e and h\u003csup\u003e+\u003c/sup\u003e contribute to but are not solely responsible for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production. To rigorously assess this hypothesis and elucidate potential competing H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e formation mechanisms, we conducted systematic control experiments on the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl photocatalyst. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, p-benzoquinone (p-BQ), tryptophan (Trp), AgNO\u003csub\u003e3\u003c/sub\u003e, and ethanol (EtOH) were employed as scavengers for \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e, \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e, e\u003csup\u003e\u0026minus;\u003c/sup\u003e, and h\u003csup\u003e+\u003c/sup\u003e, respectively. The pronounced suppression of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation upon p-BQ introduction establishes \u0026bull;O\u003csub\u003e2\u003c/sub\u003e\u003csup\u003e⁻\u003c/sup\u003e as the dominant reactive oxygen species in the production pathway. Similarly, Trp moderately suppressed H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production, confirming the participation of \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eO\u003csub\u003e2\u003c/sub\u003e in the formation process. Notably, while AgNO\u003csub\u003e3\u003c/sub\u003e and EtOH also diminished H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e yields, a significant amount of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e persisted, suggesting that both e\u003csup\u003e\u0026minus;\u003c/sup\u003e and h\u003csup\u003e+\u003c/sup\u003e contribute to but are not solely responsible for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production.\u003c/p\u003e\u003cp\u003eSupporting evidence for surface intermediates was obtained through \u003cem\u003ein-situ\u003c/em\u003e diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) under visible-light irradiation in the presence of H\u003csub\u003e2\u003c/sub\u003eO and O\u003csub\u003e2\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). The appearance of a C-OH vibration peak (1077 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) confirms the presence of adsorbed *OH species, likely formed by H\u003csub\u003e2\u003c/sub\u003eO dissociation\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Furthermore, characteristic vibration signals of *OOH (1647 and 1013 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) were detected, demonstrating the activation of adsorbed O\u003csub\u003e2\u003c/sub\u003e by photogenerated electrons (e⁻) to form *OOH intermediates\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Moreover, the signals of O-O stretching mode (890\u0026ndash;921 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) gradually increased\u003csup\u003e\u003cspan additionalcitationids=\"CR42\" citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eDensity functional theory (DFT) calculations determined the Gibbs free energy (ΔG) for the adsorption of intermediates (*O\u003csub\u003e2\u003c/sub\u003e, *OOH, *OH) at distinct sites on the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system during both the oxygen reduction reaction (ORR) and water oxidation reaction (WOR) pathways for photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e synthesis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef). These pathways involve unique intermediates and reaction steps. For the ORR process, O\u003csub\u003e2\u003c/sub\u003e undergoes conversion to *O\u003csub\u003e2\u003c/sub\u003e and *OOH, followed by *OOH protonation to yield H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. As visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee, both *O\u003csub\u003e2\u003c/sub\u003e and *OOH adsorb preferentially on CsPbI\u003csub\u003e3\u003c/sub\u003e QDs. The energy barriers for these steps are \u0026minus;\u0026thinsp;0.44 eV and \u0026minus;\u0026thinsp;0.88 eV, respectively, reflecting the synergistic role of the S-scheme heterojunction in reducing the overall reaction energy barrier. Conversely, the WOR pathway involves H\u003csub\u003e2\u003c/sub\u003eO dehydrogenation to *OH (the rate-determining step) followed by *OH dimerization to form H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef illustrates that *OH adsorbs exclusively on COF-Cl sites, with an energy barrier of 1.72 eV for the initial dehydrogenation step. These results demonstrate that the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl system favors adsorption of *OOH (ORR) and *OH (WOR) intermediates. Under visible light, photogenerated electrons and holes separate via the S-scheme heterojunction and migrate to CsPbI\u003csub\u003e3\u003c/sub\u003e QDs and COF-Cl, respectively, enabling simultaneous ORR and WOR. On CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, ORR proceeds through *O\u003csub\u003e2\u003c/sub\u003e hydrogenation to *OOH and subsequent protonation to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. Meanwhile, WOR on COF-Cl involves H\u003csub\u003e2\u003c/sub\u003eO dehydrogenation to *OH (ΔG\u0026thinsp;=\u0026thinsp;1.72 eV) followed by *OH coupling. Critically, protons released during WOR are consumed in ORR, maintaining charge balance. Consequently, the S-scheme heterojunction in CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-Cl accelerates charge separation, facilitating O\u003csub\u003e2\u003c/sub\u003e reduction by electrons and H\u003csub\u003e2\u003c/sub\u003eO oxidation by holes to generate reactive intermediates, thereby promoting efficient H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis work establishes a multifunctional interfacial engineering paradigm that stabilizes CsPbI\u003csub\u003e3\u003c/sub\u003e quantum dots within a hydrophobic chlorine-functionalized covalent organic framework (COF-Cl), enabling efficient and durable photosynthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e from seawater. Our strategy leverages multisite atomic-chlorine passivation, simultaneously forming Cl\u0026ndash;I halogen bonds and Cl\u0026ndash;Pb coordination bonds, to suppress ionic migration and defect-induced recombination. Concurrently, the COF-Cl matrix creates a gas\u0026ndash;liquid\u0026ndash;solid triphase interface that enhances O\u003csub\u003e2\u003c/sub\u003e diffusion while isolating CsPbI\u003csub\u003e3\u003c/sub\u003e from aqueous degradation. Crucially, the system forms an S-scheme heterojunction that spatially separates charge carriers: electrons accumulate on CsPbI\u003csub\u003e3\u003c/sub\u003e for selective 2e⁻ oxygen reduction (O\u003csub\u003e2\u003c/sub\u003e \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e), while holes localize on COF-Cl to drive water oxidation (H\u003csub\u003e2\u003c/sub\u003eO \u0026rarr; H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e), all without sacrificial agents. The system achieves record H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production rates of 25.29 mmol h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in pure water and 20.37 mmol h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in natural seawater, surpassing state-of-the-art photocatalysts, with an apparent quantum yield of 15.76% at 450 nm and a solar-to-chemical conversion efficiency of 1.38%. Structural integrity is maintained over multiple cycles, and outdoor testing confirms operational viability under natural sunlight. Mechanistic studies confirm synergistic proton-coupled electron transfer, where holes on COF-Cl oxidize H\u003csub\u003e2\u003c/sub\u003eO to supply protons for O\u003csub\u003e2\u003c/sub\u003e reduction on CsPbI\u003csub\u003e3\u003c/sub\u003e QDs, balancing reaction kinetics. This work pioneers a robust design for perovskite-based artificial photosynthesis in corrosive environments, setting new benchmarks for efficiency and stability in solar-driven chemical production. The multifunctional passivation strategy, integrating atomic-scale defect control, tailored interfaces, and directional charge separation, paves the way for scalable solar fuel synthesis from abundant seawater resources.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cb\u003eSynthesis of 2,4,6-tris(4-aminophenyl)-1,3,5-triazine (TAPT).\u003c/b\u003e A solution was prepared by adding 2.0 g (17.0 mmol) of 4-aminobenzonitrile to a 50 mL round-bottom flask and cooling it in an ice bath at 0\u0026deg;C for 30 minutes. Under a nitrogen atmosphere, 5.0 mL (55.5 mmol) of trifluoromethanesulfonic acid was slowly added dropwise, maintaining the temperature at 0\u0026deg;C throughout the addition. Upon completion, the reaction mixture was poured into 50 mL of deionized water. The mixture was then neutralized to pH 7 using a 2 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e NaOH solution, resulting in the formation of a pale yellow precipitate. This precipitate was collected by filtration and purified through multiple washes with distilled water to afford the TAPT product.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSynthesis of COF-X (X\u0026thinsp;=\u0026thinsp;Cl, Br, I) Frameworks.\u003c/b\u003e The COF-X frameworks were synthesized by dissolving TAPT (0.16 mmol, 56.7 mg) and the appropriate dialdehyde monomer (0.24 mmol) in a pressure-resistant glass tube containing a 1:1 (v/v) mixture of o-dichlorobenzene and n-butanol (1.0 mL each). The dialdehyde monomers used were: 2,5-dichloroterephthalaldehyde (PDA-Cl, 48.7 mg) for COF-Cl, 2,5-dibromoterephthalaldehyde (PDA-Br, 70.1 mg) for COF-Br, and 2,5-diiodoterephthalaldehyde (PDA-I, 92.6 mg) for COF-I. The mixture was sonicated for 5 minutes to enhance dissolution. Subsequently, 0.2 mL of 6 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e acetic acid solution was added quickly, followed by brief sonication (10 seconds). The glass tube was then subjected to three freeze-pump-thaw cycles under liquid nitrogen and vacuum, sealed, and heated in an oil bath at 120\u0026deg;C for 72 hours. After cooling to room temperature, the resulting yellow product was collected by centrifugation, thoroughly washed with tetrahydrofuran and methanol, and purified by Soxhlet extraction to yield the COF-X framework.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePreparation of CsPbI\u003c/b\u003e\u003csub\u003e\u003cb\u003e3\u003c/b\u003e\u003c/sub\u003e\u003cb\u003e@COF-X Artificial Photosynthetic Systems.\u003c/b\u003e For each CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X system, 50 mg of COF-X (X\u0026thinsp;=\u0026thinsp;Cl, Br, I) powder was immersed in 2.5 mL of a 0.04 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e PbI\u003csub\u003e2\u003c/sub\u003e solution in N,N-dimethylformamide (DMF). The suspension was sonicated for uniform dispersion and reacted at 70\u0026deg;C for 2 hours. The resulting solid, designated as PbI\u003csub\u003e2\u003c/sub\u003e@COF-X, was collected and washed repeatedly with DMF and ethanol to remove surface-adsorbed PbI\u003csub\u003e2\u003c/sub\u003e. This intermediate was then immersed in 2.5 mL of a 0.02 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e CsI solution in DMF for 1 hour. Following this, 10 \u0026micro;L of oleic acid (OA) and 5 \u0026micro;L of oleylamine (OM) were added to stabilize the precursor mixture. Toluene was then added to induce the formation of CsPbI\u003csub\u003e3\u003c/sub\u003e quantum dots (QDs) within the pores of the COF-X framework. The final CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X material was collected by centrifugation, washed thoroughly with toluene three times, and dried.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePhotochemical H\u003c/b\u003e\u003csub\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sub\u003e\u003cb\u003eO\u003c/b\u003e\u003csub\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sub\u003e \u003cb\u003eProduction Test under Simulated Sunlight.\u003c/b\u003e The photocatalytic performance for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production was evaluated using a 5 mg sample of the CsPbI\u003csub\u003e3\u003c/sub\u003e@COF-X artificial photosynthetic system dispersed in a double-layered reaction beaker containing 20 mL of either deionized water or seawater (0.5 g\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). The reaction mixture was gently shaken to ensure uniform dispersion of the photocatalyst on the water surface. A thermostatic circulating water system maintained the reaction temperature at 25\u0026deg;C. Photoreactions were conducted in air using a quartz and Pyrex glass hybrid reaction cell. Irradiation was provided by a 300 W xenon lamp equipped with a 420 nm cutoff filter (λ\u0026thinsp;\u0026ge;\u0026thinsp;420 nm). Throughout the reaction, the temperature was maintained at 25\u0026deg;C using a stirrer and the circulating water system. At 15-minute intervals, a 1.5 mL aliquot of the reaction solution was extracted, centrifuged, and filtered through a 0.22 \u0026micro;m membrane to remove the photocatalyst. The hydrogen peroxide concentration in the filtrate was quantified using the iodometric method.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePhotoelectrochemical and Electrochemical Characterization.\u003c/b\u003e Photoelectrochemical and electrochemical properties were characterized using a standard three-electrode system with an Ag/AgCl reference electrode, a platinum sheet counter electrode, and an FTO conductive glass working electrode coated with the catalyst. The working electrode was prepared by dispersing 5 mg of the photocatalytic material in 0.5 mL of an ethanol/isopropanol mixture (3:1 v/v), sonicating for 30 minutes, adding 20 \u0026micro;L of Nafion solution, and sonicating for an additional 30 minutes. A 50 \u0026micro;L aliquot of the resulting catalyst slurry was uniformly coated onto the FTO glass and dried at 60\u0026deg;C for 2 hours. A 300 W xenon lamp with a 400 nm cutoff filter served as the light source. Transient Photocurrent Response: Measured in 0.1 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte under light illumination. Mott-Schottky (MS) Analysis: Conducted in 0.1 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte at frequencies of 500 Hz, 1000 Hz, and 1500 Hz, scanning the potential from \u0026minus;\u0026thinsp;1.6 V to +\u0026thinsp;1.6 V vs. Ag/AgCl. Electrochemical Impedance Spectroscopy (EIS): Performed in 0.1 mol\u0026middot;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e electrolyte over a frequency range of 10\u003csup\u003e6\u003c/sup\u003e to 10\u003csup\u003e3\u003c/sup\u003e Hz.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003cstrong\u003ecknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe work was financially supported by the National Natural Science Foundation of China (22377047, 22221001), Science and Technology Leading Talent Foundation of Gansu Province (24RCKB003), and the Special Funds of the National Natural Science Foundation of China (Grant No. 223B2204).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eB.D.W. designed and conceived the whole project. G.P.M., S.W. and H. W. carried out the materials synthesis, experimental test, and data analysis. G.P.M. and B.D.W. wrote the paper together. B.D.W. and H. W. revised the manuscript. N. L. and G.P.M. conducted theoretical calculations. G.P.M, S.W., Y.H.Y., B. D. and G.W.H. performed the electron microscopy tests. All the authors contributed to the discussion during the whole project.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary information\u003c/strong\u003e is available in the online version of the paper. Reprints and permissions information is available online at www.nature.com/reprints.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrespondence and requests for materials\u003c/strong\u003e should be addressed to Baodui Wang.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting financial interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing financial interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eZhang C\u003cem\u003e, et al.\u003c/em\u003e Stable and high-yield hydrogen peroxide electrosynthesis from seawater. \u003cem\u003eNat. Sustain.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 542-552 (2025).\u003c/li\u003e\n\u003cli\u003eXia C, Xia Y, Zhu P, Fan L, Wang H. Direct electrosynthesis of pure aqueous H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e solutions up to 20% by weight using a solid electrolyte. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e366\u003c/strong\u003e, 226-231 (2019).\u003c/li\u003e\n\u003cli\u003eCampos-Mart\u0026iacute;n JM, Blanco-Brieva G, Fierro JLG. Hydrogen peroxide synthesis: an outlook beyond the anthraquinone process. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 6962-6984 (2006).\u003c/li\u003e\n\u003cli\u003eWang W\u003cem\u003e, et al.\u003c/em\u003e Photothermal-enabled single-atom catalysts for high-efficiency hydrogen peroxide photosynthesis from natural seawater. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 2493 (2023).\u003c/li\u003e\n\u003cli\u003eHuang W\u003cem\u003e, et al.\u003c/em\u003e Intermarrying MOF Glass and Lead Halide Perovskites for Artificial Photosynthesis. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e174\u003c/strong\u003e, 3195-3205 (2024).\u003c/li\u003e\n\u003cli\u003eZhao W\u003cem\u003e, et al.\u003c/em\u003e Perovskite-Doped Modulated Color-Selective Photosynaptic Transistors for Target Object Recognition. \u003cem\u003eNano lett.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 9937-9945 (2024).\u003c/li\u003e\n\u003cli\u003eSun K\u003cem\u003e, et al.\u003c/em\u003e Three-dimensional direct lithography of stable perovskite nanocrystals in glass. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e375\u003c/strong\u003e, 307-310 (2022).\u003c/li\u003e\n\u003cli\u003eHou J\u003cem\u003e, et al.\u003c/em\u003e Liquid-phase sintering of lead halide perovskites and metal-organic framework glasses. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e374\u003c/strong\u003e, 621-625 (2021).\u003c/li\u003e\n\u003cli\u003eTian T\u003cem\u003e, et al.\u003c/em\u003e Large-area waterproofand durable perovskite luminescent textiles. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 234 (2023).\u003c/li\u003e\n\u003cli\u003eFan X\u003cem\u003e, et al.\u003c/em\u003e Brightened Bicomponent Perovskite Nanocomposite Based on Forster Resonance Energy Transfer for Micro-LED Displays. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, e2300834 (2023).\u003c/li\u003e\n\u003cli\u003eJiang W\u003cem\u003e, et al.\u003c/em\u003e Axial alignment of covalent organic framework membranes for giant osmotic energy harvesting. \u003cem\u003eNat. Sustain.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 446-455 (2025).\u003c/li\u003e\n\u003cli\u003eSuleman S\u003cem\u003e, et al.\u003c/em\u003e Turning on Singlet Oxygen Generation by Outer-Sphere Microenvironment Modulation in Porphyrinic Covalent Organic Frameworks for Photocatalytic Oxidation. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e63\u003c/strong\u003e, e202314988 (2024).\u003c/li\u003e\n\u003cli\u003eLi P\u003cem\u003e, et al.\u003c/em\u003e Development of an energy-dense and high-power Li-Cl\u003csub\u003e2\u003c/sub\u003e battery using reversible interhalogen bonds. \u003cem\u003eChem\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 352-364 (2023).\u003c/li\u003e\n\u003cli\u003eWu J\u003cem\u003e, et al.\u003c/em\u003e Regioselective Multisite Atomic-Chlorine Passivation Enables Efficient and Stable Perovskite Solar Cells. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e145\u003c/strong\u003e, 5872-5879 (2023).\u003c/li\u003e\n\u003cli\u003eLiu Y\u003cem\u003e, et al.\u003c/em\u003e Halogen atom-induced local asymmetric electron in covalent organic frameworks boosts photosynthesis of hydrogen peroxide from water and air. \u003cem\u003eMatter\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 102076 (2025).\u003c/li\u003e\n\u003cli\u003eZhang L\u003cem\u003e, et al.\u003c/em\u003e Deep-red emissive colloidal lead-based triiodide perovskite/telluride nanoscale heterostructures with reduced surface defects and enhanced stability for indoor lighting applications. \u003cem\u003eNano Energy\u003c/em\u003e \u003cstrong\u003e90\u003c/strong\u003e, 106506 (2021).\u003c/li\u003e\n\u003cli\u003eWang A\u003cem\u003e, et al.\u003c/em\u003e Halogenated Hole Selective Contact Enhances Interfacial Weak Bonding of Perovskite Solar Cells. \u003cem\u003eAdvanced Energy Materials\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 2400640 (2024).\u003c/li\u003e\n\u003cli\u003eHe J\u003cem\u003e, et al.\u003c/em\u003e Competitive Formation Mechanism for Bidentate Passivation of Halogen Vacancies in Perovskite Based on 6‐Chloropurine. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 2305127 (2023).\u003c/li\u003e\n\u003cli\u003eFu X\u003cem\u003e, et al.\u003c/em\u003e Halogen-halogen bonds enable improved long-term operational stability of mixed-halide perovskite photovoltaics. \u003cem\u003eChem\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 3131-3143 (2021).\u003c/li\u003e\n\u003cli\u003eWang P\u003cem\u003e, et al.\u003c/em\u003e Ultrastable Perovskite-Zeolite Composite Enabled by Encapsulation and In Situ Passivation. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e, 23100-23106 (2020).\u003c/li\u003e\n\u003cli\u003eQiu J\u003cem\u003e, et al.\u003c/em\u003e COF/In\u003csub\u003e2\u003c/sub\u003eS\u003csub\u003e3\u003c/sub\u003e S-Scheme Photocatalyst with Enhanced Light Absorption and H H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e-Production Activity and fs-TA Investigation. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, e2400288 (2024).\u003c/li\u003e\n\u003cli\u003eLi F, Yue X, Liao Y, Qiao L, Lv K, Xiang Q. Understanding the unique S-scheme charge migration in triazine/heptazine crystalline carbon nitride homojunction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 3901 (2023).\u003c/li\u003e\n\u003cli\u003eZhang Y\u003cem\u003e, et al.\u003c/em\u003e H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation from O\u003csub\u003e2\u003c/sub\u003e and H\u003csub\u003e2\u003c/sub\u003eO on a near-infrared absorbing porphyrin supramolecular photocatalyst. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 361-371 (2023).\u003c/li\u003e\n\u003cli\u003eCao D, Du J, Li J, Sun Q, Guan J, Liu J. Hydroxyl-Functionalized Donor-Acceptor Covalent Organic Frameworks for Efficient Photocatalytic Hydrogen Peroxide Production under Visible Light. \u003cem\u003eACS Catal.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 3584-3594 (2025).\u003c/li\u003e\n\u003cli\u003eYu Z, Yu F, Xu M, Feng S, Qiu J, Hua J. Amidoxime-Functionalized sp\u003csup\u003e2\u003c/sup\u003e-Carbon-Conjugated Covalent Organic Frameworks for Overall Photocatalytic Hydrogen Peroxide Production. \u003cem\u003eAdv. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, e2415194 (2025).\u003c/li\u003e\n\u003cli\u003eShi X\u003cem\u003e, et al.\u003c/em\u003e Interfacial Engineering of \u0026beta;‐Ketoenamine‐Based COFs/Urea‐Linked Perylene Diimide for Overall Photosynthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in Seawater. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 2414755 (2024).\u003c/li\u003e\n\u003cli\u003eYue JY, Pan ZX, Zhang RZ, Xu Q, Yang P, Tang B. One‐Pot Synthesis of Fully Conjugated Covalent Organic Frameworks via the Pictet\u0026ndash;Spengler Reaction for Boosting H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e Photogeneration in Real Seawater. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 2421514 (2025).\u003c/li\u003e\n\u003cli\u003eHou Y, Liu F, Liang J, Li Z, Zhou P, Tong M. Building a Confluence Charge Transfer Pathway in COFs for Highly Efficient Photosynthesis of Hydrogen Peroxide from Water and Air. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, e202505621 (2025).\u003c/li\u003e\n\u003cli\u003eYue JY\u003cem\u003e, et al.\u003c/em\u003e Thiophene-Containing Covalent Organic Frameworks for Overall Photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e Synthesis in Water and Seawater. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, e202309624 (2023).\u003c/li\u003e\n\u003cli\u003eHuang Y\u003cem\u003e, et al.\u003c/em\u003e Nb\u003csub\u003e2\u003c/sub\u003eC MXene quantum dots modulate built-in electric field within heterostructures for efficient solar-to-H2O2 conversion from seawater. \u003cem\u003eAppl. Catal. B Environ.\u003c/em\u003e \u003cstrong\u003e371\u003c/strong\u003e, 125262 (2025).\u003c/li\u003e\n\u003cli\u003eYue J-Y\u003cem\u003e, et al.\u003c/em\u003e Overall solar H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation in water and real seawater by covalent organic frameworks with kgd topology. \u003cem\u003eChem. Eng. J\u003c/em\u003e \u003cstrong\u003e504\u003c/strong\u003e, 158983 (2025).\u003c/li\u003e\n\u003cli\u003eTang B\u003cem\u003e, et al.\u003c/em\u003e Photocatalytic Synthesis of Hydrogen Peroxide by Rhenium Modified Metal-Organic Frameworks Incorporating Bianthracene Ligands. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e, 10.1021/jacs.1025c00603 (2025).\u003c/li\u003e\n\u003cli\u003eHu ZY, Liu T, Yang YR, An AK, Liew KM, Li WW. Binder-Free Immobilization of Photocatalyst on Membrane Surface for Efficient Photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e Production and Water Decontamination. \u003cem\u003eNanomicro Lett.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 301 (2025).\u003c/li\u003e\n\u003cli\u003eJu Y\u003cem\u003e, et al.\u003c/em\u003e Regulating the electronic structure of covalent organic frameworks via heterocyclic isomers for highly efficient photocatalytic H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e generation. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 5658 (2025).\u003c/li\u003e\n\u003cli\u003eZhou J\u003cem\u003e, et al.\u003c/em\u003e Covalent Organic Frameworks with Constrained Palladium Nanoclusters Lock the Enol-Keto Tautomerism Enhancing Hydrogen Peroxide Photosynthesis. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, e2409006 (2025).\u003c/li\u003e\n\u003cli\u003eChen J\u003cem\u003e, et al.\u003c/em\u003e Redox-Mediated TEMPO-Based Donor-Acceptor Covalent Organic Framework for Efficient Photo-Induced Hydrogen Peroxide Generation. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, e202500924 (2025).\u003c/li\u003e\n\u003cli\u003eGu M, Yang Y, Zhang L, Zhu B, Liang G, Yu J. Efficient sacrificial-agent-free solar H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e production over all-inorganic S-scheme composites. \u003cem\u003eAppl. Catal. B Environ.\u003c/em\u003e \u003cstrong\u003e324\u003c/strong\u003e, 122227 (2023).\u003c/li\u003e\n\u003cli\u003eLuo Y\u003cem\u003e, et al.\u003c/em\u003e Sulfone-Modified Covalent Organic Frameworks Enabling Efficient Photocatalytic Hydrogen Peroxide Generation via One-Step Two-Electron O\u003csub\u003e2\u003c/sub\u003e Reduction. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, e202305355 (2023).\u003c/li\u003e\n\u003cli\u003eZhou S\u003cem\u003e, et al.\u003c/em\u003e Chloromethylation Modified Pyranonitrile-Based Conjugated Microporous Polymers for Selective One-Step Two-Electron O\u003csub\u003e2\u003c/sub\u003e Reduction to H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e, 10.1002/anie.202508436 (2025).\u003c/li\u003e\n\u003cli\u003eWang W\u003cem\u003e, et al.\u003c/em\u003e Graphene Quantum Dot‐Modified Mn\u003csub\u003e0.2\u003c/sub\u003eCd\u003csub\u003e0.8\u003c/sub\u003eS for Efficient Overall Photosynthesis of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e, 10.1002/adfm.202422307 (2025).\u003c/li\u003e\n\u003cli\u003eXu T\u003cem\u003e, et al.\u003c/em\u003e Constructing Photocatalytic Covalent Organic Frameworks with Aliphatic Linkers. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e146\u003c/strong\u003e, 20107-20115 (2024).\u003c/li\u003e\n\u003cli\u003eLi Z\u003cem\u003e, et al.\u003c/em\u003e Covalent Organic Frameworks for Boosting H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e Photosynthesis via the Synergy of Multiple Charge Transfer Channels and Polarized Field. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, e202420218 (2025).\u003c/li\u003e\n\u003cli\u003eLiu R\u003cem\u003e, et al.\u003c/em\u003e Linkage-engineered donor\u0026ndash;acceptor covalent organic frameworks for optimal photosynthesis of hydrogen peroxide from water and air. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 195-206 (2024).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Scheme","content":"\u003cp\u003eScheme 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7187821/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7187821/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLead halide perovskites are promising for artificial photosynthesis but suffer from aqueous instability. Here, we stabilize CsPbI\u003csub\u003e3\u003c/sub\u003e quantum dots (QDs) within a hydrophobic chlorine-functionalized covalent organic framework (COF-Cl) through multisite atomic-chlorine passivation, forming dual Cl–Pb coordination and Cl–I halogen bonding at the interface. This suppresses ionic migration while creating a gas–liquid–solid triphase interface for enhanced O\u003csub\u003e2\u003c/sub\u003e diffusion. The resulting S-scheme heterojunction spatially separates carriers to concurrently drive two-electron oxygen reduction and water oxidation for H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e synthesis without sacrificial agents. The system achieves record production rates of 25.29 mmol h\u003csup\u003e-1\u003c/sup\u003e g\u003csup\u003e-1\u003c/sup\u003e in pure water and 20.37 mmol h\u003csup\u003e-1\u003c/sup\u003e g\u003csup\u003e-1\u003c/sup\u003e in seawater under visible light, with a solar-to-chemical conversion efficiency of 1.38%. Crucially, it operates stably for 20 h in seawater and produces 11.7 mmol L\u003csup\u003e-1\u003c/sup\u003e H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in 10 h under natural sunlight. Mechanistic studies confirm synergistic interfacial charge transfer and dual-reaction pathways via both oxygen reduction and water oxidation. This work establishes a paradigm for robust perovskite-based photocatalysts toward scalable solar-driven chemical synthesis from seawater.\u003c/p\u003e","manuscriptTitle":"Multisite atomic-chlorine-passivation stabilizes perovskite interfaces for efficient H2O2 photosynthesis from seawater","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-17 08:34:13","doi":"10.21203/rs.3.rs-7187821/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5c2feb6a-9494-45b0-9e41-5b820dab5358","owner":[],"postedDate":"September 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":54861349,"name":"Physical sciences/Chemistry/Catalysis/Photocatalysis"},{"id":54861350,"name":"Physical sciences/Chemistry/Green chemistry/Sustainability"}],"tags":[],"updatedAt":"2026-05-04T07:06:51+00:00","versionOfRecord":{"articleIdentity":"rs-7187821","link":"https://doi.org/10.1038/s41467-026-70503-2","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-03-15 04:00:00","publishedOnDateReadable":"March 15th, 2026"},"versionCreatedAt":"2025-09-17 08:34:13","video":"","vorDoi":"10.1038/s41467-026-70503-2","vorDoiUrl":"https://doi.org/10.1038/s41467-026-70503-2","workflowStages":[]},"version":"v1","identity":"rs-7187821","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7187821","identity":"rs-7187821","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}