Converging Cooperative Functions into the Nanospace of Covalent Organic Frameworks for Efficient Uranium Extraction from Seawater

Open AccessCCS ChemistryRESEARCH ARTICLE14 Jul 2022Converging Cooperative Functions into the Nanospace of Covalent Organic Frameworks for Efficient Uranium Extraction from Seawater Mengjie Hao, Zhongshan Chen, Xiaolu Liu, Xianhai Liu, Juyao Zhang, Hui Yang, Geoffrey I. N. Waterhouse, Xiangke Wang and Shengqian Ma Mengjie Hao College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Zhongshan Chen College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Xiaolu Liu College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Xianhai Liu College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Juyao Zhang College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Hui Yang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author , Geoffrey I. N. Waterhouse MacDiarmid Institute for Advanced Materials and Nanotechnology, School of Chemical Sciences, The University of Auckland, Auckland 1142 Google Scholar More articles by this author , Xiangke Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206 Google Scholar More articles by this author and Shengqian Ma *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, University of North Texas, Denton, TX 76201 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.022.202201897 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail The extraction of uranium from seawater is challenging though it offers tremendous potential for the sustainable production of nuclear fuel for the energy sector. Herein, we report a new strategy for efficient extraction of uranium from seawater via converging the cooperative functions of adsorption–photocatalysis into the nanospace of covalent organic frameworks (COFs). Functionalization of the organic linkers in the multicomponent COFs allowed exploration of the relationship between material composition and adsorption–photocatalytic activity for uranium extraction. The presence of amidoxime groups in the COFs offered selective binding sites for uranyl, whilst triazine units and bipyridine-Pd groups acted cooperatively to photocatalytically reduce adsorbed U(VI) to a U(IV) solid product (UO2) for facile collection. One of our developed COFs, 4-Pd-AO, displayed exceptional performance in sequestering and reducing uranyl from natural seawater, with a high extraction capacity of 4.62 mg U/g per day (average data) under visible light irradiation. Mechanistic studies revealed that 4-Pd-AO not only reduced adsorbed uranyl(VI) to U(IV)O2, but also generated 1O2 and superoxide radicals under visible light excitation, thus affording excellent antibacterial and antialgal activities (i.e., antibiofouling properties) for sustained efficient uranium extraction performance. This proof-of-concept study establishes multicomponent COFs as promising candidates for efficient uranium extraction from seawater. Download figure Download PowerPoint Introduction Nuclear energy is expected to play an important role in the future decarbonization of the energy sector.1,2 Enriched uranium is the fuel used in most fission-based nuclear reactors. The scarcity of uranium ore reserves on land (∼4.5 million tons) is an obstacle to the long-term future of nuclear energy in beyond-fossil-fuel energy infrastructures.3 However, uranium reserves in seawater are abundant, ∼1000 times those on land.4,5 Accordingly, the discovery of efficient technologies for the extraction of uranium from seawater is seen as a promising pathway for the sustained production of nuclear fuel. To address this need, much effort has been focused on developing sorbents for uranium extraction, with essential requirements of such sorbents being high uranyl adsorption selectivity, rapid adsorption kinetics, large adsorption capacity, and reusability.6 Traditional porous sorbents such as porous carbons,7–12 mesoporous silica,13 and layered inorganic materials14 are limited in their usefulness because of the low density of functional groups that selectively bind U(VI) ions, generally leading to slow adsorption kinetics and poor adsorption capacities. Amorphous porous organic polymers (POPs) are possible candidates, but their performance is hampered by their disordered structures, which block the uranyl chelating sites.15–24 As porous crystalline materials, metal–organic frameworks (MOFs) have been pursued for uranium extraction, stimulated by their high surface areas and well-defined/varied pore environments.25–29 However, the modest stabilities of most MOFs in water limit their practical utility. Compared with traditional porous materials such as POPs, and MOFs, covalent organic frameworks (COFs) show particular promise as uranium adsorbents from seawater due to their uniform porosity, tunable chemical characteristics and functional groups, high surface area, physiochemical stability, and acid–base stability.30–34 A number of works have shown that the introduction of amidoxime chelating groups into COF structures results in enhanced uranium extraction from seawater.19,23,31,32,34–38 However, competitive metal ion adsorption can block some uranyl binding sites, resulting in blockage of the channels of the COFs and a reduction in the uranium adsorption capacity and hindering uranium recovery and adsorbent reuse. Furthermore, biofouling by marine bacteria and algae can also passivate sorbents in seawater, thus lowering the uranyl uptake capacity rapidly and in extreme cases completely suppressing adsorption.39–41 To overcome these issues, the development of materials with fast kinetics, large capacities, high adsorption selectivities, and antibiofouling properties are crucial if COFs are to be of practical use for uranium extraction from seawater. Recently, increasing interest has focused on the synthesis of COFs with photosensitive functional groups that can effectively promote photocatalytic reduction reactions.42–46 This inspired us to develop an adsorption–photocatalytic strategy for uranium extraction from natural seawater. Our targeted strategy involved the synthesis of COFs with a high density of chelating moieties and abundant photoactive sites, thereby achieving effective uranyl binding kinetics, highly selective uranyl adsorption, a high adsorption capacity, uranyl reduction to a solid U(IV) product under visible light, and reactive oxygen species-mediated antibiofouling activity. COFs with such characteristics would solve the critical challenges of extracting uranium from natural seawater. Engineering the chemical microenvironment in multicomponent COF frameworks is critical to realize such broad-spectrum performance. To implement this strategy, we engineered the pore structure of a multicomponent COF to contain amidoxime groups, triazine groups, and bipyridine-metal groups (Figure 1). This functionality was realized by incorporating the relevant functional moieties on the framework linkers. Through detailed experimental studies, the resultant COF (denoted herein as COF 4-Pd-AO) demonstrated excellent uranium extraction performance in spiked-seawater and natural seawater under visible light irradiation. The amidoxime groups imparted hydrophilicity and a high binding affinity for uranyl. The triazine groups and bipyridine-Pd(II) components served as duel photocatalytic active sites, which reduced U(VI) of uranyl to U(IV) in the form of solid UO2 for easy collection. Mechanistic studies revealed that O2•− and 1O2 free radicals generated by the COF under light irradiation could damage the structure of marine bacteria and inhibit colonization by algae, imparting the multicomponent COF with excellent antibiofouling activity. Because of these features, COF 4-Pd-AO delivered a uranium uptake capacity of 4.62 mg/g per day (average data) in natural seawater using the adsorption–photocatalytic process. Our COF-based adsorption–photocatalytic strategy thus offers a promising new approach for uranium extraction from seawater. Figure 1 | Adsorption–photocatalytic strategy for uranium extraction. Schematic of a multicomponent COF with chelating amidoxime moieties and photocatalytic active sites and antibiofouling ability as an adsorption–photocatalyst for uranium extraction from seawater. Download figure Download PowerPoint Experimental Methods Synthesis of COF 1 In a 5 mL glass tube, p-phenylenediamine (Pa, 8.6 mg), 2,2′-bipyridine-5,5′-diamine (Bpy, 7.5 mg), and 2,4,6-triformylphloroglucinol (Tp, 16.8 mg) were dissolved in 1.1 mL of a mixed solvent solution containing o-dichlorobenzene (o-DCB)/n-butyl alcohol (n-BuOH)/acetic acid (AcOH, 6 M) in a volume ratio of 5/5/1. The mixture was then frozen in a liquid nitrogen bath and sealed with a gas torch. The tube was then heated at 120 °C for 3 days, after which the product was washed several times with tetrahydrofuran (THF) and acetone, collected by vacuum filtration, and dried under vacuum at 40 °C overnight. Synthesis of COF 2 In a 5 mL glass tube, 2,5-diaminobenzonitrile (Db, 10.6 mg), Bpy (7.5 mg), and Tp (16.8 mg) were dissolved in 1.1 mL of a mixed solvent solution of o-DCB/n-BuOH/AcOH (6 M) in a volume ratio of 5/5/1. The mixture was then frozen in a liquid nitrogen bath and sealed with a gas torch. The tube was then heated at 120 °C for 3 days, after which the product was washed several times with THF and acetone, collected by vacuum filtration, and dried under vacuum at 40 °C overnight. Synthesis of COF 3 In a 5 mL glass tube, [1,1′-biphenyl]-3,3′-dicarbonitrile,4,4′-diamino- (Bpdba, 9.4 mg), Bpy (14.9 mg), and Tp (16.8 mg) were dissolved in 1.1 mL of a mixed solvent solution of o-DCB/n-BuOH/AcOH (6 M) in a volume ratio of 5/5/1. The mixture was then frozen in a liquid nitrogen bath and sealed with a gas torch. The tube was then heated at 120 °C for 3 days, after which the product was washed several times with THF and acetone, collected by vacuum filtration, and dried under vacuum at 40 °C overnight. Synthesis of COF 4 In a 5 mL glass tube, 4,4′,4″-(1,3,5-triazine-2,4,6-triyl)tribenzaldehyde (TATTA, 31.5 mg), Bpy (14.9 mg) and Bpdba (9.4 mg) were dissolved in 1.1 mL of a mixed solvent solution of o-DCB/dimethylacetamide/AcOH (6 M) in a volume ratio of 9/1/1. The mixture was then frozen in a liquid nitrogen bath and sealed with a gas torch. After being heated at 120 °C for 3 days, the product was washed several times with THF and acetone, collected by filtration, and dried under vacuum overnight. Synthesis of COFs 2-AO, 3-AO, and 4-AO For the synthesis of COF 2-AO (AO, amidoxime groups), 0.2 g of COF 2 was dispersed in 30 mL of ethanol, followed by the addition of 0.5 g of NH2OH·HCl and 0.1 mL of trimethylamine. After stirring for 12 h at 75 °C, the product was collected by filtration, washed several times with deionized water, and then finally dried at 40 °C under vacuum. COFs 3-AO and 4-AO were synthesized via a similar synthetic route, using COF 3 and COF 4, respectively. Synthesis of COFs 3-Pd-AO and 4-Pd-AO For the synthesis of COF 3-Pd-AO, 0.1 g of COF 3 was dispersed in 100 mL of acetonitrile, followed by the addition of 63.5 mg of [PdCl2(CH2CN)2]. After heating at 65 °C for 24 h, the product was collected by filtration, washed several times with acetonitrile and methanol, and then finally dried under vacuum to yield COF 3-Pd. Subsequently, COF 3-Pd-AO was obtained by following the above procedure. COF 4-Pd-AO was synthesized via a similar synthetic route, using COF 4 as the starting material. Results and Discussion Synthesis and characterization of the COF adsorption–photocatalysts To render COFs with specific functionality as uranium extraction materials, our initial step was to synthesize a three-component COF with one-dimensional (1D) channels. This was achieved via the Schiff-base condensation of Pa, Tp, and Bpy under solvothermal conditions (denoted as COF 1) (Figure 2a). The chemical structure of COF 1 was determined by Fourier transform infrared (FTIR) spectroscopy, solid-state cross-polarization magic angle spinning 13C NMR (13C CP/MAS NMR), and powder X-ray diffraction (PXRD). The disappearance of –NH2 stretching bands at ∼3500–3300 cm−1 and –CHO stretching band at ∼1638 cm−1 in the precursors indicated the formation of imine bonds in the COF 1 structure ( Supporting Information Figures S1 and S2). The solid-state 13C CP/MAS NMR spectrum of the product further showed the characteristic signal for the C–NH group at 159.6 ppm ( Supporting Information Figure S6). A possible theoretical crystalline structural simulation of COF 1 was determined by Materials Studio Software,47 built under a monoclinic P21/m space group with unit cell parameters of a = 27.60 Å, b = 26.78 Å, and c = 3.54 Å, α = β = 90°, and γ = 128.02° ( Supporting Information Tables S1 and S2). Comparing the eclipsed stacking (AA) and staggered stacking (AB) modes showed the experimental data more closely matched the calculated AA stacking mode with an unweighted-profile R factor (Rp) = 1.76% and weighted-profile R factor (Rwp) = 2.34%, suggesting the validity of the computational model (Figure 2b and Supporting Information Table S2). The PXRD pattern of COF 1 showed diffraction peaks at 4.26° and 7.61°, which could readily be assigned to the (010) and (110) Bragg peaks of the P21/m structure, respectively (Figure 2b). Based on these results, we deduced that COF 1 demonstrated a two-dimensional (2D) structure when viewed from the top with hexagonal 1D channels with a theoretical pore size of 2.0 nm (Figures 2a and 2c). The side view showed the layered stacking structure with an interlayer distance of 3.49 Å. Thermogravimetric analysis (TGA) demonstrated that COF 1 was thermally stable up to 450 °C under a nitrogen atmosphere ( Supporting Information Figure S15). Figure 2 | Preparation and characterization of COF 1. (a) Synthetic scheme of COF 1 through the condensation of Bpy, Tp, and Pa. (b) Experimental and simulated PXRD patterns of COF 1. (c) Top and side view of the eclipsed (AA) stacking crystal structure of COF 1. Hydrogen atoms are omitted for clarity. The C, N, and O atoms are represented by gray, blue, and red spheres, respectively. Download figure Download PowerPoint Considering the selective uranyl recognition and coordination abilities of amidoxime groups, we sought to replace the Pa linker in COF 1 with an amidoxime-rich linker to produce the amidoxime functionalized COFs. Based on our previous experience, amidoxime functional groups can be produced by hydrolysis of cyano groups with hydroxylamine. Based on this approach, a cyano-functionalized COF was first prepared by combining Db, Tp, and Bpy, finally producing COF 2 with a monoclinic unit cell with a = 27.25 Å, b = 26.84 Å, and c = 3.53 Å, α = β = 90°, and γ = 129.15° (Figure 3a, route A, Supporting Information Tables S1 and S3). The experimental PXRD pattern closely matched the calculated results based on a model with eclipsed AA stacking (Figure 3b). The sharp peaks observed at 3.98° and 7.76° in the PXRD pattern of COF 2 correspond to the (100) and (110) planes, respectively. The observed peaks at 95.1 ppm and 2219 cm−1 in the solid-state 13C NMR and FTIR spectra, respectively, confirmed the presence of cyano groups (–C≡N) in the structure (Figure 3d and Supporting Information Figure S7). The characterization results and calculated unit cell for COF 2 (Figure 3c) revealed a structure similar to COF 1. However, due to the introduction of cyano groups, COF 2 possessed a smaller pore size of ∼1.9 nm. COF 3, another nitrile-functionalized framework, was synthesized with the same general synthetic strategy, using combinations of Bpdba, Tp, and Bpy (Figure 3a, route B). FTIR, solid-state 13C CP/MAS NMR, and PXRD analysis were used to establish the probable structure of COF 3 (2D eclipsed AA stacking, with hexagonal 1D channels approximately 2.6 nm in diameter) (Figures 3f and 3g and Supporting Information Figure S8 and Tables S1 and S4). TGA analysis showed COF 2 and COF 3 were stable up to 450 °C and 400 °C, respectively, by heating under N2 ( Supporting Information Figures S16 and S17). Chemical transformation of the cyano groups in these COFs into amidoxime groups with hydroxylamine in ethanol produced COF 2-AO and COF 3-AO, respectively (Figures 3a, 3e, and 3i). The absence of a nitrile stretch (∼2219 cm−1) in the FTIR spectra of COF 2-AO and COF 3-AO accords well with the successful modification of cyano groups to amidoxime groups (Figures 3d and 3h). Solid-state 13C NMR analyses further confirmed this transformation, evidenced by the disappearance of the cyano groups peak at 95.1 (COF 2) and 97.5 ppm (COF 3) and the emergence of C=N signals (from amidoxime groups) at 166.7 and 160.5 ppm from the generated amidoxime groups in the spectra of COF 2-AO and COF 3-AO, respectively ( Supporting Information Figures S7 and S8). PXRD patterns showed the parent structures were retained after the amidoxime functionalization (Figures 3b and 3f). Figure 3 | Preparation and characterization of COFs 2, 3, 2-AO, 3-AO. (a) Synthetic schemes of COF 2 and COF 3, and their corresponding post-synthetic modification to chemically transformation cyano groups to amidoxime groups, yielding COF 2-AO, and COF 3-AO, respectively. (b) Experimental and simulated PXRD profiles. (c) Graphic view of the eclipsed AA stacking structure of COF 2. (d) FTIR spectra of COF 2 and COF 2-AO. (e) Graphic view of COF 2-AO. (f) Experimental and simulated PXRD profiles. (g) Graphic view of COF 3. (h) FTIR spectra of COF 3 and COF 3-AO. (i) Graphic view of COF 3-AO. Hydrogen atoms are omitted for clarity. The C, N, and O atoms are represented by gray, blue, and red spheres, respectively. Download figure Download PowerPoint Aside from acting simply as uranyl adsorbents, COF materials are able to act as photocatalysts for the reductive deposition of uranyl as solid products, such as UO2, thereby offering a practical technology for uranium extraction and collection.31 This inspired us to construct an adsorption–photocatalysis COF system to achieve efficient uranium extraction from seawater. Triazine48–51 and [Pd(ligand)xCl2]46,52,53 components are commonly used as photosensitizing agents in the literature. Accordingly, we aimed to install both of these photosensitizing groups into a COF framework. To achieve this, as schematically illustrated in Figure 4a, a triazine containing molecule TATTA was used instead of Tp in the COF 3 synthesis. This produced COF 4. The structure of COF 4 was determined from PXRD measurements in conjunction with structural calculations and Pawley refinements using Materials Studio Software.47 The experimental data were a close match for the calculated data (negligible difference, Rp, 1.21%; Rwp, 1.84%), indicating phase purity of COF 4 and an AA stacking mode (Figures 4b and 4c and Supporting Information Tables S1 and S5). The compound possesses a 2D layer structure with 3.6 nm honeycomb-like open channels (Figures 4a and 4c). FTIR spectroscopy showed characteristic peaks at 2219 and 1619 cm−1, confirming the presence of cyano groups and imine C=N bonds, respectively, in the COF 4 structure ( Supporting Information Figure S4). The 13C CP/MAS NMR showed a peak at 94.6 ppm attributed to C≡N (Figure 4d). TGA showed that COF 4 possesses very high thermal stability up to 500 °C in N2 ( Supporting Information Figure S18). The obtained COF was then stirred with [PdCl2(CH2CN)2] in acetonitrile to produce a Pd(II)-functionalized material. A final amidoximation process with hydroxylamine yielded COF 4-Pd-AO, with the amidoxime groups located in large pores (Figures 4a, 4b, 4d, and 4e). PXRD confirmed a high crystallinity was retained after the functionalization steps (Figure 4b). High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and corresponding elemental mapping images together with X-ray photoelectron spectroscopy (XPS) indicated that Pd and Cl were successfully incorporated into the COF framework (Figure 4f and Supporting Information Figures S20, S28–S30). Pd 3d5/2 and Pd 3d3/2 XPS peaks were observed at 338.4 and 343.8 eV, respectively, consistent with literature values reported for palladium in bipyridine-Pd(II) structures.53 Figure 4 | Preparation and characterizations of COF 4 and COF 4-Pd-AO. (a) Synthetic scheme of COF 4 and corresponding post-synthetically modified COF 4-Pd-AO. (b) Experimental and simulated PXRD profiles. (c) Graphic view of the eclipsed AA stacking structure of COF 4. (d) Solid-state 13C CP/MAS NMR spectra of COF 4 and COF 4-Pd-AO. (e) Top and side view of COF 4-Pd-AO. (f) HAADF-STEM and corresponding elemental mapping images of COF 4-Pd-AO. Hydrogen atoms are omitted for clarity. The C, N, O, Pd, and Cl atoms are represented by gray, blue, red, magenta, and olive spheres, respectively. Download figure Download PowerPoint To validate the utility of COF 4-Pd-AO as a high-performance adsorption−photocatalyst for uranyl adsorption and photocatalytic U(VI) reduction-deposition, we also synthesized several reference materials (COF 3-Pd-AO and COF 4-AO). The detailed synthetic procedures and characterizations of these additional COFs are provided in Supporting Information Figures S3, S5, S9, S10, and S19. Porosity, chemical stability, and photoelectric properties First, the stabilities of the various synthesized multicomponent COFs were studied under different harsh conditions. After 24 h of immersion in HCl (pH 3), NaOH (pH 12), and natural seawater, the COFs remained intact with no framework collapse or undesirable phase transitions occurring, as evidenced by the PXRD patterns of the treated COFs being the same with the as-prepared COFs (Figure 5a and Supporting Information Figures S11–S14). Nitrogen adsorption–desorption isotherms were measured at 77 K to evaluate the surface areas and porosities of the multicomponent COFs. The isotherms showed large N2 uptakes at P/P0 values below 0.1, and a more gradual uptake at a pressure between 0.1 to 0.95 (P/P0), indicating that both micropores and small mesopores were present in the COFs, consistent with their crystal structures (Figure 5b and Supporting Information Figures S21, S23, S25). The Brunauer–Emmett–Teller (BET) surface areas and pore size distributions of the COFs are summarized in Supporting Information Table S6 and Figures S21–S27. The BET surface area calculated for COF 4-Pd-AO was 989 m2/g, lower than that of COF 4-AO (1094 m2/g) and COF 4 (1932 m2/g), demonstrating the retention of significant porosity after PdCl2 functionalization. The as-synthesized COFs were next characterized by UV–vis spectroscopy in diffuse reflectance mode to determine their optical properties. The spectra for COFs 4, 4-AO, 4-Pd-AO, and 3-Pd-AO showed strong absorption bands in the visible region with absorption onsets ranging from 520 nm to 600 nm ( Supporting Information Figure S31). The optical band gaps were estimated to be 2.10, 2.19, 1.89, and 2.00 eV for COFs 4, 4-AO 4-Pd-AO, and 3-Pd-AO, respectively (Figure 5c). COF 4-Pd-AO showed the lowest band gap, indicating it needed the least energy to achieve an effective separation of charge carriers. The internal resistances of the COFs were evaluated by electrochemical impedance spectroscopy (EIS) analysis. COF 4-Pd-AO displayed a smaller semicircular diameter in EIS Nyquist curves compared with COFs 4, 4-AO, and 3-Pd-AO, indicating a smaller interfacial charge-transfer resistance (Figure 5d). Moreover, the conduction-band position of each COF was estimated by measuring the flat band potential (Efb) via Mott–Schottky plots. COF 4-Pd-AO exhibited a negative Efb of −1.04 V versus a Ag/AgCl electrode, much lower than that of COFs 4 (−0.57 V), 4-AO (−0.6 V), and 3-Pd-AO (−0.52 V) (Figure 5e). The positive slopes in Figure 5e suggest these COFs are n-type semiconductors. Generally, the conduction band (CB) of n-type semiconductors is equal to the flat band potential.54 Using the results of the flat band potential measurements and the optical band gaps, the positions of the CBs and valence bands (VBs) in the different COFs were estimated: COFs 4 (CB = −0.37 eV, VB = 1.73 eV), COF 4-AO (CB = −0.40 eV, VB = 1.79 eV), COF 4-Pd-AO (CB = −0.84 eV, VB = 1.05 eV), and COF 3-Pd-AO (CB = −0.32 eV, VB = 1.68 eV) (Figure 5f). The photocurrent density curves for COF 3-Pd-AO and COF 4-Pd-AO showed that photogenerated charges were created under visible light irradiation, thus validating our strategy to introduce photoactive components in the COFs ( Supporting Information Figure S32). The photoresponse for COF 4-Pd-AO was larger than that for 3-Pd-AO, which is explained by the former containing triazine and bipyridyl-Pd(II) sensitizers, whereas the latter only contained bipyridyl-Pd(II) moieties. Obviously, the CB positions of the COFs satisfy the requirement for U(VI) reduction to U(IV) [0.411 V vs normal hydrogen electrode (NHE)], suggesting these COFs could be utilized as an effective adsorption–photocatalyst platform for uranium extraction from seawater.55 Figure 5 | Characterization of synthesized COFs. (a) PXRD patterns of COF 4-Pd-AO after treatment under different conditions. (b) N2 sorption isotherms measured at 77 K for COFs 4, 4-AO, and 4-Pd-AO. (c) Estimated band gap of COFs 4, 4-AO, 4-Pd−AO, and 3−Pd−AO. (d) Electrochemical impedance spectra (EIS) of COFs 4, 4−AO, 4−Pd-AO, and 3-Pd-AO. (e) Mott−Schottky plots for COFs 4, 4-AO, 4-Pd-AO, and 3-Pd-AO. (f) Band alignment of COFs 4, 4-AO, 4-Pd-AO, and 3-Pd-AO. Download figure Download PowerPoint The characterization studies above revealed the successful synthesis of a series of multicomponent COFs, with COF 4-Pd-AO possessing all of the following advantages: (1) excellent stability in acid, basic, and seawater media; (2) imine bonds, amidoxime groups, and [Pd(bpy)2Cl2] groups that imparted hydrophilicity to the COF; (3) abundant amidoxime groups as uranyl-binding sites; (4) triazine functional groups providing photocatalytic activity; (5) bipyridine-Pd(II) sites as secondary photosensitizing sites to boost photocatalytic performance. As a proof-of-concept, we next conducted a series of experiments to assess the uranium extraction performance of the various multicomponent COFs we had synthesized. Physicochemical adsorption studies The uranyl adsorption properties of the different COF materials were initially studied in spiked-seawater solutions at pH 8.1. Equilibrium adsorption capacities were determined by varying the concentrations of uranium from 0 to ∼20 ppm at a fixed sorbent concentration of 0.05 mg/mL. After adsorption equilibrium was attained, the tested capacity per gram of sorbent for COFs 4-AO, 4-Pd-AO, 3-AO, 3-Pd-AO, and 2-AO, were 155.7, 150.8, 146.7, 134.4, and 63.2 mg/g, respectively (Figure 6a and Supporting Information Figures S33 and S34). COFs 4-AO and 4-Pd-AO (as well as COFs 3-AO and 3-Pd-AO) exhibited similar capacities, suggesting the coordinated Pd(II) did not block any uranyl adsorption sites. The equilibrium adsorption data were fitted well by a Langmuir model with correlation coefficients ( Supporting Information Figures S33 and S34 and Table S7). The amidoxime functionalized COFs 4-AO, 4-Pd-AO, 3-AO, 3-Pd-AO, and 2-AO showed rapid adsorption rates for uranyl (Figure