Constructing donor–acceptor covalent organic frameworks with bex topology for efficient photocatalytic reduction of uranium

Abstract

Effective elimination of uranium from wastewater is crucial for sustainable development in the nuclear industry. Extracting U(VI) by photocatalytic reduction using covalent organic frameworks (COFs) is regarded as an effective and environmentally friendly method. However, COFs with unconventional bex topology for photocatalytic uranium reduction remain largely unexplored. Herein, we synthesized three COFs that have distinct bex topologies, and their performance in the photocatalytic reduction of U(VI) was investigated. Among these COFs, Py-TAPT-COF, in which the triazine unit acts as an electron acceptor and the pyrene unit functions as an electron donor, showed excellent photocatalytic reduction capability, achieving 92.1% reduction of U(VI). Additionally, from a mixture containing 13 competing metal ions, Py-TAPT-COF demonstrated remarkable uranium extraction selectivity and retained 78.8% of its photocatalytic efficiency after four consecutive cycles. Density functional theory (DFT) calculations revealed that the triazine and pyrene units are active photocatalytic reaction sites, which significantly promote the separation of photoelectrons from vacancies.

---

Introduction

Nuclear energy has received growing international attention as a pivotal clean energy source. Uranium is an indispensable fuel for nuclear energy production and is also known as the “food” for the development of the nuclear industry. However, terrestrial uranium resources are not only limited in reserves but also unevenly distributed geographically. By the year 2040, around 87% of the globally identified uranium resources are likely to be consumed.¹ Therefore, exploring new uranium resources is crucial to ensuring a sustainable supply–demand balance in the future. In addition, uranium, which is a radioactive element, exhibits both radiological and chemical toxicity. If its concentration exceeds the permitted limit, it poses a threat to human health and ecological security.^2,3 Therefore, whether motivated by strategic considerations for the sustainable utilization of uranium resources or by the practical need to remediate uranium-contaminated environments, the effective treatment and disposal of uranium-bearing radioactive substances are of paramount importance.

At present, various techniques for treating uranium-containing wastewater have been reported, such as membrane separation, ion exchange, chemical precipitation, adsorption, and photocatalysis.^4–9 Photocatalysis is considered to be an emerging green technology for uranium extraction due to the advantages of safety, low cost, high efficiency, and environmental friendliness. At present, most research on the photocatalytic reduction of uranium focuses on the development of new photocatalysts. Photocatalysts have evolved from early inorganic semiconductor materials represented by TiO₂ to the current organic semiconductor materials such as g-C₃N₄.^10–12 However, most of the reported photocatalytic materials currently exhibit amorphous structures, which not only greatly limit their photocatalytic efficiency, but also pose a huge challenge to accurately revealing the structure–activity relationship of materials. Therefore, the selection or preparation of suitable photocatalytic materials has become a challenging task in the application of efficient photocatalytic reduction of U(VI).

Porous materials such as covalent organic frameworks (COFs), metal–organic frameworks (MOFs), and hydrogen-bonded organic frameworks (HOFs) have been explored and applied for photocatalytic uranium reduction. Lan et al. reported a porous HOF (UPC-H4a) self-assembled from 5,10,15,20-tetra(4-(2,4-diaminotriazine)phenyl) porphyrin via intermolecular hydrogen bonds, and UPC-H4a enables the removal of uranium from radioactive wastewater utilizing visible light.¹³ Additionally, Wang et al. introduced 1,1′-ferrocene dicarboxylic acid (FcDA) to synthesize the photocatalyst Zr-Fc-MOF₃. They pointed out that the incorporation of FcDA reduced the band gap of Zr-Fc-MOF₃, enhanced the separation of electron–hole pairs and improved the carrier transport efficiency.¹⁴ In particular, COFs, synthesized via condensation of various organic monomers, are considered porous crystalline photocatalytic materials due to their superior advantages such as tunable functional groups, well-organized crystal structures, large specific surface areas, and appropriate photoelectric responses.^15,16 Since Wang et al. explicitly demonstrated that a polyoxometalate–organic framework synergistically integrates ligand complexation with photocatalytic reduction for enhanced uranium extraction, extensive studies have subsequently highlighted the potential of COFs as effective photocatalysts for uranium extraction.¹⁷ However, most COFs exhibited a series of shortcomings, including insufficient active centers and poor electron–hole separation, thus leading to low photocatalytic efficiency. Recent research demonstrated that constructing suitable electron donor–acceptor (D–A) structures is a beneficial strategy for establishing electron transport paths, which accelerates the separation of light-generated electrons (e⁻) and holes (h⁺). When a D–A-structured photocatalyst is irradiated, intramolecular electrons are transferred from the donor to the electrophilic acceptor unit, leading to efficient spatial separation of photogenerated charges. This process increases the electron density in the acceptor region, therefore suppressing carrier recombination through the strong binding of excitons. For instance, Zhong et al. synthesized three structurally analogous 2D COFs—Tp-Tapb, Tp-Taz, and TpTt—and proved the occurrence of photoinduced electron transfer in these donor–acceptor (D–A)-configured systems. TpTt COFs demonstrated exceptional uranium adsorption properties (505 mg g⁻¹) and remarkable photocatalytic reduction performance, with a reaction rate constant of 0.22 h⁻¹. The two short-range acceptor units (triazine and ketone) accelerated charge separation and transfer, thus enhancing the performance.¹⁸ In another study, a battery of D–A-structured COFs for photocatalytic uranium reduction was successfully synthesized by Ma et al. Furthermore, they confirmed that local charge distribution modulation within the pore channels promotes both the electron excitation dynamics and the photocatalytic activity.¹⁹

Defect-engineered COFs have attracted widespread attention in the field of photocatalysis. In previous work, our group reported a novel strategy which significantly enhanced photocatalytic H₂O₂ production through the rational control of topological defects in COFs.²⁰ Similarly, Li et al. prepared COFs with a controlled proportion of structural defects and significantly enhanced photocatalytic H₂ production.²¹ However, most COFs employed for photocatalytic uranium reduction have common topological structures, including honeycomb (hcb), square (sql), kagome (kgm), and hexagonal (hxl) lattices. In contrast, sub-stoichiometric COFs (also referred to as Type III COFs) with bex topology are still less explored. These COFs contain purposefully synthesized, unreacted functional groups in their structures. Such precise control enables a fundamental investigation of structure–property relationships. Sub-stoichiometric COFs with bex topology could be obtained using a [4 + 3] sub-stoichiometric condensation. For instance, Wang et al. synthesized a pyrene-based 2D-COF (HITMS-COF-11) with bex topology for application in the hydrogen evolution reaction (HER). This COF exhibited an excellent photocatalytic HER rate of up to 25 000 μmol g⁻¹ h⁻¹. The outstanding performance is attributed not only to the broadened photoresponse and optimized generation rate of photo-generated carriers but also to the enhanced hydrophilicity arising from the free amine groups on the pyrene units.²² Zeng et al. reported two novel [4 + 3] imine-linked covalent organic frameworks (COFs), demonstrating that the out-of-plane pendant unreacted aldehyde groups can serve as catalytic centers, which significantly enhanced the transport space and enabled their application in the electrocatalytic production of hydrogen peroxide.²³

Since sub-stoichiometric COFs are rarely employed for photocatalytic uranium reduction, herein we successfully synthesized three D–A COFs with bex topology, namely, Py-TAPA-COF, Py-TAPB-COF, and Py-TAPT-COF. Specifically, the pyrene units (Py) serve as strong electron-donating units, and tris(4-aminophenyl)amine (TAPA), 1,3,5-tris(4-aminophenyl)benzene (TAPB) and 2,4,6-tris(4-aminophenyl)-1,3,5-triazine (TAPT) serve as electron-withdrawing units. Furthermore, by employing a [4 + 3] condensation strategy, we achieved sub-stoichiometric control, thus leading to unreacted aldehyde (–CHO) groups pendant from the framework. These COFs exhibited excellent crystallinity and large surface areas, with quadrilateral and unique pores. Under visible light, the three COFs were evaluated for photocatalytic U(VI) reduction performance. Among these COFs, Py-TAPT-COF exhibited superior photocatalytic reduction performance. It achieved a reduction efficiency of 92.1%, exhibited high uranium selectivity amid 13 competing metal ions, and retained 78.8% of its activity after four cycles. DFT calculations indicate that the HOMO and LUMO distributions are localized on both pyrene and triazine units, promoting intramolecular charge transfer and suggesting their roles as active sites.

Results and discussion

Synthesis and characterization

The synthetic route utilized in this study is depicted in Scheme 1. Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF were synthesized via a Schiff-base reaction catalyzed by imidazole and acetic acid in a mixed solution of n-butanol and o-dichlorobenzene. As shown in Scheme 1, the topological framework of pyrene bearing periodically uncondensed aldehyde groups is fundamentally distinct from that of fully condensed COFs.

Scheme 1 Synthetic routes of Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF.

The structures of Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF were determined by solid-state cross-polarization magic-angle-spinning ¹³C nuclear magnetic resonance (CP/MAS ¹³C NMR) spectroscopy. As shown in Fig. S1–S3, the characteristic chemical shifts at ∼160 ppm corresponded to the –C=N– bonds.^24,25 In addition, the chemical shifts between 120 and 140 ppm were assigned to the existence of aromatic phenyl groups in these three frameworks.²⁶ It is worth noting that a chemical shift at ∼191 ppm was observed in all three COFs. This is attributed to the free uncondensed aldehyde groups from pyrenyl.²⁷ The Fourier transform infrared (FT-IR) spectra of these three COFs showed the successful formation of –C=N– bonds due to the existence of stretches at 1613 cm⁻¹ (Fig. S4). Noticeably, the stretch at 1668 cm⁻¹ was assigned to the free uncondensed aldehyde groups,^28,29 which was in agreement with the ¹³C CP-MAS NMR spectroscopy results. As shown in the thermogravimetric analysis (TGA) curves (Fig. S5), the thermal stability of Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF exceeds 350 °C.

Powder X-ray diffraction (PXRD) analyses were performed to evaluate the crystallinity of Py-TAPA-COF, Py-TAPB-COF, and Py-TAPT-COF. As shown in Fig. 1a–c, all the COFs exhibited good crystallinity. The intense peak around 3.6° (2θ) corresponds to the (100) reflection, indicating the formation of a long-range ordered framework. In addition, the formation of a 2D framework was confirmed by a minor diffraction peak around 25°. Then, the bex crystallographic network topologies of the three COFs were built, and the Pawley refinement was further performed to verify their structures. As shown in Fig. S6–S11, the AA stacking model structure matched well with the refined experimental patterns with very small corresponding residual values (R_wp = 0.88% and R_p = 0.43% for Py-TAPT-COF; R_wp = 1.87% and R_p = 1.11% for Py-TAPB-COF; and R_wp = 0.47% and R_p = 0.25% for Py-TAPA-COF). The obtained cell parameters were a = 53.11 Å, b = 25.44 Å, and c = 3.50 Å for Py-TAPT-COF; a = 54.86 Å, b = 24.27 Å, and c = 3.51 Å for Py-TAPB-COF; and a = 50.50 Å, b = 22.08 Å, and c = 3.50 Å for Py-TAPA-COF, respectively (Tables S1–S3). In this model, the triangular linkers (TAPT, TAPB, and TAPA) connect two Py molecules on both sides. Then, the Py acts as a linker through its diagonal group to connect with the triangular linkers. Consequently, two free –CHO groups from the pyrene units are generated per unit cell. The porosity of the three COFs (Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF) was further investigated through N₂ adsorption–desorption isotherm measurements at 77 K. All three materials exhibited reversible type IV isotherms, confirming their predominantly mesoporous nature (Fig. 1d–f). Brunauer–Emmett–Teller (BET) surface area analysis yielded specific surface areas of 564.24 cm² g⁻¹, 489.60 cm² g⁻¹, and 414.4 cm² g⁻¹ for Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF, respectively. The pore size distributions revealed primary pore diameters centered at 1.92 nm, 1.91 nm, and 1.62 nm for Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF, respectively. These experimental pore sizes are consistent with the theoretical values derived from the corresponding structural models (Fig. S12–S14).

Fig. 1 Structural characterization of the COFs. PXRD patterns of (a) Py-TAPT-COF, (b) Py-TAPB-COF and (c) Py-TAPA-COF. N₂ adsorption–desorption isotherm curves of (d) Py-TAPT-COF, (e) Py-TAPB-COF, and (f) Py-TAPA-COF.

Performance of photocatalytic uranium extraction

The uranium adsorption performance of the three COFs was studied systematically under dark and photocatalytic reduction environments. As shown in Fig. 2a, the three COFs showed almost no adsorption performance for UO₂²⁺ in the dark and reached equilibrium within 1 h. The negligible removal efficiency indicated that they are ineffective as adsorbents.

Fig. 2 The photocatalytic U(VI) reduction performance. (a) Uranium removal performance of Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF under light and dark conditions. (b) Pseudo-second-order kinetic fit for the photocatalytic process of Py-TAPT-COF, Py-TAPB-COF, and Py-TAPA-COF. (c) Extraction efficiency of U in the presence of other competing ions after a 2 h reaction under visible light. (d) Cycling runs for the photochemical reduction of U(VI). (e) U 4f XPS spectra of Py-TAPT-COF loaded with U(VI) under light conditions. (f) SEM and corresponding EDS elemental mapping images of Py-TAPT-COF after photocatalytic reduction of uranium.

Subsequently, the photocatalytic uranium separation performance of Py-TAPA-COF, Py-TAPB-COF, and Py-TAPT-COF was systematically evaluated under ambient air and visible light conditions. Before the light was switched on, the adsorption equilibrium was tested in the dark with uranium solution. Uranium extraction kinetics were investigated under a 300 W xenon lamp after 5 mL of methanol was added as a hole scavenger. Under visible light irradiation, the removal efficiency was significantly enhanced. Following irradiation for 2 hours, Py-TAPT-COF was found to exhibit superior photocatalytic performance for uranium reduction, with a U(IV) separation efficiency of up to 92.1% (924.69 mg g⁻¹). In contrast, Py-TAPB-COF and Py-TAPA-COF yielded efficiencies of 74.8% (754.50 mg g⁻¹) and 72.6% (728.91 mg g⁻¹), respectively (Fig. 2a and Fig. S15). The results of kinetic analysis showed that the photocatalytic reduction follows pseudo-second-order kinetics (Fig. 2b, Fig. S16 and Table S4), indicating a chemical removal process. In other words, the photocatalytic process is the primary mechanism for uranium extraction by the three COFs. Crucially, the donor–acceptor (D–A) structure of Py-TAPT-COF, with pyrene being a strong electron donor and triazine being a strong electron acceptor, promotes charge separation and subsequent photocatalytic activity.

Based on its efficient photocatalytic reduction performance, we further studied the selectivity and cycling stability of the Py-TAPT-COF. Authentic wastewater was used which contained uranium and several competing metal ions, including monovalent ions (Na⁺, K⁺, and Cs⁺), divalent ions (Ca²⁺, Al³⁺, Co²⁺, Zn²⁺, Sr²⁺, Mn²⁺, and Cu²⁺), and trivalent ions (Cr³⁺ and Cd³⁺). These components may potentially hinder U(VI) photocatalytic reduction. Fig. 2c shows that Py-TAPT-COF can remove nearly 65% of U(VI), and the removal efficiency is much higher than that of the other metal ions (all below 15%). The findings indicate that Py-TAPT-COF was highly selective for U(VI) reduction in the presence of various competing ions.

The ability to regenerate and recycle photocatalysts is an essential concern for practical applications. Therefore, the reusability of Py-TAPT-COF was also investigated in this work. As illustrated in Fig. 2d, no significant loss of performance was observed over four consecutive cycles, with the material retaining 78.8% of its initial removal capacity. This confirms the outstanding recyclability of Py-TAPT-COF. The XPS spectra of Py-TAPT-COF before and after uranium sorption are shown in Fig. S17. The spectrum of uranium-laden Py-TAPT-COF-U exhibits additional peaks corresponding to uranium (U 4f), revealing successful uranium capture. Characteristic U 4f signals were detected at binding energies of 391.8 eV and 380.2 eV in the high-resolution spectrum of the post-photocatalytic sample, confirming successful uranium enrichment (Fig. 2e).^30,31

The morphology of Py-TAPT-COF was studied via scanning electron microscopy (SEM). Prior to photocatalytic reduction, Py-TAPT-COF was shown to consist of aggregated rod-like particles (Fig. S18). After the photocatalytic reduction of U(VI), the sample surface was covered with numerous inhomogeneous nano-flocculent and flake-like deposits. Meanwhile, the corresponding energy-dispersive spectroscopy (EDS) mapping clearly indicated that uranium was uniformly distributed across the material surface (Fig. 2f), further confirming the successful enrichment of uranium on Py-TAPT-COF.

Optical and electrochemical properties of COFs

The charge generation and transfer properties of the three COFs were characterized using a combination of optical and electrochemical techniques. First, the absorption ability for photons was evaluated, as broad optical absorption can enhance solar energy conversion efficiency. Fig. 3a shows the UV-vis diffuse reflectance spectra (DRS) of the three COFs. It is evident that all the COFs exhibit a visible-light response. In particular, the Py-TAPT-COF demonstrates a broad absorption range, which can be attributed to its strong electron-accepting units. This suggests that Py-TAPT-COF possesses superior light-harvesting efficiency compared to the other two COFs. The band gaps (E_g) of the three COFs were calculated from the Kubelka–Munk-transformed function to be 2.18 eV for Py-TAPT-COF, 2.33 eV for Py-TAPB-COF, and 2.38 eV for Py-TAPA-COF, respectively. The narrower E_g of Py-TAPT-COF enables the generation of more photo-generated carriers, which is helpful for enhancing photocatalytic activity.³²

Fig. 3 The optical and electrochemical characterization data of the COFs. (a) UV-visible diffuse reflectance spectra (DRS) of the COFs. (b) The band gap energies of the COFs. (c) Schematic energy band diagrams of the COFs. (d) Transient photocurrent densities of the COFs. (e) EIS curves of the COFs. (f) Steady-state photoluminescence spectra of the COFs.

Mott–Schottky plots of these three COFs indicate that they belong to an N-type semiconductor as a result of positive slopes, where electrons were the primary carriers (Fig. S19–S21). The flat-band potentials of Py-TAPT-COF, Py-TAPB-COF and Py-TAPA-COF were fitted to be −1.56 V, −1.45 V, and −1.39 V (vs. Ag/AgCl), respectively. Typically, the conduction band (CB) and its flat-band potential were similar. Therefore, the CB of Py-TAPT-COF, Py-TAPB-COF and Py-TAPA-COF was estimated to be −1.36 V, −1.25 V and −1.19 V (vs. NHE), respectively. According to the bandgap energy equation (E_g = E_VB − E_CB), the valence band (VB) potentials of Py-TAPT-COF, Py-TAPB-COF and Py-TAPA-COF were calculated to be 0.82 V, 1.08 V, and 1.19 V (vs. NHE), respectively. On account of these results, the energy band positions were plotted in Fig. 3c. As shown in Fig. 3c, the CB positions of all these COFs are more negative than the U(VI)/U(IV) redox couple (0.41 V versus NHE), indicating the ability of these COFs to reduce hexavalent uranium through photocatalysis.^33,34

Subsequently, the ability of electron transfer was further confirmed by photocurrent experiments. As shown in Fig. 3d, Py-TAPT-COF exhibits the most remarkable photocurrent response, indicating superior charge separation capability and stability under illumination. Furthermore, electrochemical impedance spectroscopy (EIS) measurements of the three COFs revealed that Py-TAPT-COF possesses the smallest electron transfer resistance (Fig. 3e).³⁵ Besides, fluorescence intensity is regarded as an essential measurement metric, which reflects the composite efficiency of photoelectron–hole pairs. All three COFs exhibit an emission peak at ∼615 nm in their photoluminescence (PL) spectra. However, Py-TAPT-COF shows the weakest PL intensity, suggesting superior charge carrier separation efficiency, prolonged carrier lifetime, and enhanced photocatalytic performance (Fig. 3f).³⁶

Possible photocatalytic U(VI) extraction mechanism

To elucidate the mechanism of photocatalytic uranium extraction, radical scavenger experiments were conducted (Fig. 4a). As is known, silver nitrate (AgNO₃), p-benzoquinone (BQ) and isopropanol (IPA) were used as scavengers for e⁻, ˙O₂⁻ and ˙OH, respectively.³⁷ Experimental results demonstrated that the addition of AgNO₃ significantly diminished the photocatalytic activity of Py-TAPT-COF for U(VI) reduction to 38.4%, further confirming the critical role of e⁻. Similarly, the addition of p-BQ significantly inhibited the photocatalytic U(VI) removal efficiency, which was reduced to 33%. The marked suppression efficiency in the presence of AgNO₃ and p-BQ suggests that e⁻ and ˙O₂⁻ are pivotal active species responsible for photocatalysis of uranium. In contrast, with the addition of isopropanol (IPA) as a scavenger of hydroxyl radicals (˙OH), the U(VI) photocatalytic reduction performance did not change obviously. This indicates that ˙OH has little influence on the photocatalytic performance of hexavalent uranium. In addition, the U(VI) removal efficiency could be improved substantially by using methanol as a selective scavenger. This is because methanol effectively scavenges photogenerated holes (h⁺), suppressing electron–hole recombination and consequently promoting the availability of photogenerated electrons for the reduction. Consequently, the photocatalytic uranium extraction efficiency was substantially improved.

Fig. 4 Experimental and DFT calculation of the photocatalytic reaction mechanism. (a) The photoreduction experiments of Py-TAPT-COF with different scavenging agents added. (b) HOMO and (c) LUMO orbital distribution diagrams of the unit cell structure of Py-TAPT-COF. (d) Schematic illustration of the photoreduction of U(VI) on Py-TAPT-COF.

In addition, the reaction mechanism was further investigated using DFT calculations. Analysis of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) distributions in the COF units showed that the pyrene and triazine units can serve as potential active sites for redox reactions. The pyrene units contributed to the HOMO of Py-TAPT-COF, while the triazine units contributed to the LUMO (Fig. 4b and c). The electrostatic potential (ESP) of Py-TAPT-COF indicates that the positive potential regions are concentrated on the pyrene units, while the negative potential regions are centered on the triazine units (Fig. S22). Therefore, Py-TAPT-COF exhibits distinct orbital separation and an electron push–pull effect, demonstrating typical donor–acceptor (D–A) structural characteristics. The D–A structure enhances intramolecular charge transfer efficiency within the framework and creates precisely positioned catalytic centers. In addition, the polarity of the dangling aldehyde groups promoted the effective separation of photogenerated electron–hole pairs, leading to the high photocatalytic performance.³⁸

Based on these analyses, we propose the following mechanism for the photoreduction of U(VI) by Py-TAPT-COF (Fig. 4c). Firstly, U(VI) was captured by the skeleton of Py-TAPT-COF due to the existence of triazine groups. Secondly, the e⁻ and h⁺ (e⁻/h⁺) were generated in Py-TAPT-COF with the D–A structure under illumination. Subsequently, the effective separation of e⁻/h⁺ pairs enabled the rapid transfer of electrons to U(VI), facilitating its further reduction to tetravalent uranium (U(IV)) and the formation of uranium dioxide (UO₂).³⁹ The LUMO energy level of Py-TAPT-COF (−1.36 V vs. SHE) is more negative than the standard redox potential E(O₂/˙O₂⁻) = −0.33 V vs. SHE.⁴⁰ This provides the intrinsic driving force for Py-TAPT-COF to reduce dissolved oxygen (O₂) to the ˙O₂⁻ reactive intermediate, facilitating the photoreduction of U(VI) to U(IV). Photogenerated electrons react with dissolved oxygen to generate ˙O₂⁻, while photogenerated holes were quenched by methanol.

Conclusions

In summary, three COFs with unconventional bex topology were designed for photocatalytic uranium enrichment. Among them, Py-TAPT-COF achieved 92.1% U(VI) photoreduction under visible light, exhibiting outstanding performance in the photocatalytic reduction of uranium. DFT simulations revealed a dual-site redox mechanism, where the HOMO (localized on pyrene units) and LUMO (centered on triazine rings) act cooperatively to enhance charge transfer, while polar aldehyde groups facilitate electron–hole separation. Our approach demonstrates the considerable potential of bex topology COFs as a versatile platform for photocatalytic applications.

Author contributions

Zhang L. L.: writing, carried out experiment, data curation, conceptualization, and DFT calculations. Wang L.: writing – review & editing, funding acquisition, formal analysis, supervision, and conceptualization. Luo F.: writing – review & editing, funding acquisition, validation, and supervision.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

All the data supporting this article have been included in the main text and the supplementary information (SI). Supplementary information: materials, measurements, experimental details, supplementary figures and crystallographic information. See DOI: https://doi.org/10.1039/d6qi00101g.

Acknowledgements

This work was supported financially by the National Natural Science Foundation of China (22506017) and the Open Project Program of the National Key Laboratory of Uranium Resources Exploration-Mining and Nuclear Remote Sensing (2025QZ-YYZ-05). The authors extend their gratitude to Ms Yan Yuwei (from Scientific Compass https://www.shiyanjia.com) for providing invaluable assistance with the XPS analysis.

References

1. Organisation for Economic Co-operation and Development and Nuclear Energy Agency, Uranium 2020 : Resources, Production and Demand. OECD Publishing, 2021.
2. A. S. Katz, The Chemistry and toxicology of depleted uranium, Toxics, 2014, 2, 50–78.
3. Y. Liu, Y. Yuan, S. Ni, J. Liu, S. Xie and Y. Liu, Construction of g-C₃N₄/Ag/TiO₂ Z-scheme photocatalyst and its improved photocatalytic U(VI) reduction application in water, Water Sci. Technol., 2022, 85, 2639–2651.
4. F. Chen, M. Lv, Y. Ye, S. Miao, X. Tang, Y. Liu, B. Liang, Z. Qin, Y. Chen, Z. He and Y. Wang, Insights on uranium removal by ion exchange columns: The deactivation mechanisms, and an overlooked biological pathway, Chem. Eng. J., 2022, 434, 134708.
5. S. Zhou, F. Dong and Y. Qin, High efficiency uranium(VI) removal from wastewater by strong alkaline ion exchange fiber: effect and characteristic, Polymers, 2023, 15, 279.
6. M.-L. Chu, C.-P. Tung and M.-C. Shieh, A Study on triple-membrane-separator (TMS) process to treat aqueous effluents containing uranium, Sep. Sci. Technol., 1990, 25, 1339–1348.
7. Y. Zhang, X. Liu, Y. Liu, J. Feng and K. Jiang, Efficient uranium(VI) separation based on a layered 2D Ti₃C₂Tx/hydroxyapatite hybrid membrane, Chem. Eng. J., 2025, 503, 158707.
8. Y. Huang, S. Yue, A. Li, H. Sun, Y. Wang, Q. Bu, B. Liu and G. Han, Precipitation-flotation process for molybdenum and uranium separation from wastewater, Metals, 2024, 14, 1231.
9. Y.-G. Zhao, L.-H. Chen, M.-L. Ye, W.-S. Su, C. Lei, X. J. Jin and Y. Lu, U (VI) removal on polymer adsorbents: Recent development and future challenges, Crit. Rev. Environ. Sci. Technol., 2025, 55, 264–286.
10. T. Chen, K. Yu, C. Dong, X. Yuan, X. Gong, J. Lian, X. Cao, M. Li, L. Zhou, B. Hu, R. He, W. Zhu and X. Wang, Advanced photocatalysts for uranium extraction: elaborate design and future perspectives, Coord. Chem. Rev., 2022, 467, 214615.
11. S. G. K. Patro, A. Hota and A. O. Salau, Photo-catalytic reduction of destructive U(VI) from uranium-defiled wastewater: An overview, Water, Air, Soil Pollut., 2024, 235, 279.
12. S. Zhang, L. Chen, Z. Qu, F. Zhai, X. Yin, D. Zhang, Y. Shen, H. Li, W. Liu, S. Mei, G. Ji, C. Zhang, X. Dai, Z. Chai and S. Wang, Confining Ti-oxo clusters in covalent organic framework micropores for photocatalytic reduction of the dominant uranium species in seawater, Chem, 2023, 9, 3172–3184.
13. P. Wu, X. Yin, Y. Zhao, F. Li, Y. Yang, N. Liu, J. Liao and T. Lan, Porphyrin-based hydrogen-bonded organic framework for visible light driven photocatalytic removal of U(VI) from real low-level radioactive wastewater, J. Hazard. Mater., 2023, 459, 132179.
14. S. Zhao, T. Feng, M. Cao, T. Liu, Y. Yuan and N. Wang, Ferrocene-based 2D metal-organic framework nanosheet for highly efficient photocatalytic seawater uranium recovery, Chem. Eng. J., 2024, 498, 155228.
15. C. Wen, Y. Yao, L. Meng, E. Duan, M. Wang, Z. Chen and X. Wang, Photocatalytic and electrocatalytic extraction of uranium by COFs: A Review, Ind. Eng. Chem. Res., 2023, 62, 18230–18250.
16. S. Manzoor, M. A. Younis, Y. Yao, Q.-u.-n. Tariq, B. Zhang, B. Tian, L. Yan and C. Qiu, Schiff base COFs for photocatalysis: from fundamentals to applications, Coord. Chem. Rev., 2025, 541, 216840.
17. H. Zhang, W. Liu, A. Li, D. Zhang, X. Li, F. Zhai, L. Chen, L. Chen, Y. Wang and S. Wang, Three mechanisms in one material: uranium capture by a polyoxometalate-organic framework through combined complexation, chemical reduction, and photocatalytic reduction, Angew. Chem., Int. Ed., 2019, 58, 16110–16114.
18. X. Zhong, Q. Ling, Z. Ren and B. Hu, Immobilization of U(VI) onto covalent organic frameworks with the different periodic structure by photocatalytic reduction, Appl. Catal., B, 2023, 326, 122398.
19. H. Yang, M. Hao, Y. Xie, X. Liu, Y. Liu, Z. Chen, X. Wang, G. I. N. Waterhouse and S. Ma, Tuning local charge distribution in multicomponent covalent organic frameworks for dramatically enhanced photocatalytic uranium extraction, Angew. Chem., Int. Ed., 2023, 62, e202303129.
20. L. Guo, Z. Huang, L. Gong and F. Luo, Boosting photosynthesis of H₂O₂ using topological defects in covalent organic framework for selective extraction of uranium through an innovative mineralization strategy, CCS Chem., 2025, 7, 3146–3159.
21. J. Wang, X.-X. Tian, L. Yu, D. J. Young, W.-B. Wang, H.-Y. Li and H.-X. Li, Engineering structural defects into a covalent organic framework for enhanced photocatalytic activity, J. Mater. Chem. A, 2021, 9, 25474–25479.
22. T. Banerjee, F. Haase, S. Trenker, B. P. Biswal, G. Savasci, V. Duppel, I. Moudrakovski, C. Ochsenfeld and B. V. Lotsch, Sub-stoichiometric 2D covalent organic frameworks from tri- and tetratopic linkers, Nat. Commun., 2019, 10, 2689.
23. S. Zheng, Z. Ouyang, M. Liu, S. Bi, G. Liu, X. Li, Q. Xu and G. Zeng, Construction of dangling and staggered stacking aldehyde in covalent organic frameworks for 2e-oxygen reduction reaction, Carbon Neutralization, 2024, 3, 415–422.
24. S. T. Emmerling, R. Schuldt, S. Bette, L. Yao, R. E. Dinnebier, J. Kästner and B. V. Lotsch, Interlayer interactions as design tool for large-pore COFs, J. Am. Chem. Soc., 2021, 143, 15711–15722.
25. H. Wang, W.-N. Jiao, W.-D. Zhu, S. Huang, X.-C. Lin, T. Chen, Y. Fan, F. Chen, H.-S. Xu, M. Pan and C.-Y. Su, Stepwise post-modification of pyridine-imine COFs for enhanced hydrolytic stability and proton conductivity, Angew. Chem., Int. Ed., 2025, e202511559.
26. M. Lu, M. Zhang, J. Liu, T.-Y. Yu, J. N. Chang, L.-J. Shang, S.-L. Li and Y.-Q. Lan, Confining and highly dispersing single polyoxometalate clusters in covalent organic frameworks by covalent linkages for CO₂ photoreduction, J. Am. Chem. Soc., 2022, 144, 1861–1871.
27. S. Maiti, J. K. Sharma, J. Ling, D. Tietje-Mckinney, M. P. Heaney, T. Runčevski, M. A. Addicoat, F. D'Souza, P. J. Milner and A. Das, Emissive substoichiometric covalent organic frameworks for water sensing and harvesting, Macromol. Rapid Commun., 2023, 44, 2200751.
28. H. Li, Z. Zhou, T. Ma, K. Wang, H. Zhang, A. H. Alawadhi and O. M. Yaghi, Bonding of polyethylenimine in covalent organic frameworks for CO₂ capture from air, J. Am. Chem. Soc., 2024, 146, 35486–35492.
29. A. Larrieu, S. K. Das, A. Viterisi and L. Billon, Self-assembly-induced colloidal covalent organic frameworks toward photocatalytic H₂ evolution: A surfactant-free concept, Small, 2025, 2505636.
30. X. Zhong, Z. Ren, Q. Ling and B. Hu, Adsorption-photocatalysis processes: The performance and mechanism of a bifunctional covalent organic framework for removing uranium ions from water, Appl. Surf. Sci., 2022, 597, 153621.
31. X. Wu, K. Yu, Y. He, X. Cao, T. Chen and W. Zhu, Artificial photosynthetic assemblies constructed by NH₂-UiO-66 decorated with spatially separated dual cocatalysts for Enhanced photocatalytic uranium reduction, Inorg. Chem., 2024, 63, 24141–24149.
32. C. Zhu, C. Gong, D. Cao, L.-L. Ma, D. Liu, L. Zhang, Y. Li, Y. Peng and G. Yuan, Cobalt-metalated 1D perylene diimide carbon-organic framework for enhanced photocatalytic α-C(sp³)-H activation and CO₂ reduction, Angew. Chem., Int. Ed., 2025, 64, e202504348.
33. Y.-K. Zhang, L. Zhao, A. O. Terent'ev and L.-N. He, Dynamic pyridinethiol ligand shuttling within iron-anchored covalent organic frameworks boosts CO₂ photoreduction, J. Mater. Chem. A, 2025, 13, 1407–1419.
34. Z. Chen, J. Wang, M. Hao, Y. Xie, X. Liu, H. Yang, G. I. N. Waterhouse, X. Wang and S. Ma, Tuning excited state electronic structure and charge transport in covalent organic frameworks for enhanced photocatalytic performance, Nat. Commun., 2023, 14, 1106.
35. Z. Li, J. Jiao, W. Fu, K. Gao, X. Peng, Z. Wang, H. Zhuo, C. Yang, M. Yang, G. Chang, L. Yang, X. Zheng, Y. Yan, F. Chen, M. Zhang, Z. Meng and X. Shang, Integration of perylene diimide into a covalent organic framework for photocatalytic oxidation, Angew. Chem., Int. Ed., 2024, 63, e202412977.
36. Z. Chen, J. Guo, F.-H. Song, S.-D. Wang, S. A. C. Carabineiro, S.-X. Ouyang and L.-L. Wen, Rational regulation of dipole polarization in donor-acceptor covalent organic frameworks for enhanced photocatalytic efficiency, ACS Catal., 2025, 15, 8284–8296.
37. W. Zhang, B. Wang, H. Cui, Q. Wan, B. Yi and H. Yang, Unveiling the exciton dissociation dynamics steered by built-in electric fields in conjugated microporous polymers for photoreduction of uranium(VI) from seawater, J. Colloid Interface Sci., 2024, 662, 377–390.
38. Y. Liu, W.-K. Han, W. Chi, Y. Mao, Y. Jiang, X. Yan and Z.-G. Gu, Substoichiometric covalent organic frameworks with uncondensed aldehyde for highly efficient hydrogen peroxide photosynthesis in pure water, Appl. Catal., B, 2023, 331, 122691.
39. P. He, L. Zhang, L. Wu, S. Xiao, X. Ren, R. He, X. Yang, R. Liu and T. Duan, Synergy of oxygen vacancies and thermoelectric effect enhances uranium(VI) photoreduction, Appl. Catal., B, 2023, 322, 122087.
40. L. Yang, W. Lv, Y. Wang and Y. Wang, Carbon nanotube incorporation and framework protonation-regulated energy band for enhanced photocatalytic hydrogen peroxide production of COF, Nano Res., 2025, 18, 94907024.

---