A New Route for the Rapid Synthesis of Metal–Organic Frameworks at Room Temperature

Open AccessCCS ChemistryRESEARCH ARTICLES20 Jul 2022A New Route for the Rapid Synthesis of Metal–Organic Frameworks at Room Temperature Pei Zhang, Xinchen Kang, Liming Tao, Lirong Zheng, Junfeng Xiang, Ran Duan, Jikun Li, Peng Chen, Xueqing Xing, Guang Mo, Zhonghua Wu and Buxing Han Pei Zhang Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Xinchen Kang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 School of Chemistry, University of Chinese Academy of Sciences, Beijing 100049 , Liming Tao Key Laboratory of Science and Technology on Wear and Protection of Materials, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou, Gansu 730000 , Lirong Zheng Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Junfeng Xiang Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Ran Duan Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Jikun Li Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Peng Chen Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 School of Chemistry, University of Chinese Academy of Sciences, Beijing 100049 , Xueqing Xing Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Guang Mo Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Zhonghua Wu Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 and Buxing Han *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Beijing National Laboratory for Molecular Sciences, Key Laboratory of Colloid, Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 School of Chemistry, University of Chinese Academy of Sciences, Beijing 100049 Shanghai Key Laboratory of Green Chemistry and Chemical Processes, School of Chemistry and Molecular Engineering, East China Normal University, Shanghai 200062 https://doi.org/10.31635/ccschem.022.202202155 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Rapid synthesis of metal–organic frameworks (MOFs), especially high-valence MOFs at room temperature without external energy, is a challenging topic. In this work, a stable radical solution has been discovered. Various MOFs with versatile metal nodes and ligands were rapidly synthesized at room temperature in the absence of external energy. Especially, MOFs with conjugated ligands achieved instantaneous architecture (in less than 1 s) and quantitative yield. Radicals in the solution play a crucial role in the accelerated kinetics, and the new radical route paves a cyclic pathway for the MOF synthesis. The mechanism has been thoroughly investigated by electron paramagnetic resonance, in situ proton nuclear magnetic resonance, X-ray absorption spectra, in situ small-angle X-ray scattering-wide-angle X-ray scattering, and density functional theory calculations. Download figure Download PowerPoint Introduction Metal–organic frameworks (MOFs) represent a class of crystalline porous materials with tunable pore size, high surface area, structural diversity, and designability.1–6 To date, a myriad of MOFs prepared by versatile methods have been reported.7–11 Room temperature fabrication of MOFs has profound implications from techno-economic and eco-friendly perspectives,12–16 and quite a few divalent metal-based MOFs such as Cu- and Zn-based MOFs have been synthesized at room temperature.17–21 However, high-valence metal-based MOFs are far more difficult to synthesize at room temperature, let alone with rapid kinetics,22–24 and their synthesis basically requires external energy such as microwaves and electricity.25,26 Aside from external energy, solution is another foremost factor that synthesis of MOFs relies upon, and syntheses are very sensitive to subtle variations in solutions.27 Solvent effect on the kinetics of the formation of MOFs is of great importance.28,29 Exploring powerful solutions to replace external energy to realize both room temperature synthesis and rapid formation of MOFs, especially high-valence metal-based MOFs is very promising. Although creating new solution-based routes for the synthesis of MOFs at room temperature is highly desirable, it is also very challenging. Herein, we propose a new route for the MOF synthesis at room temperature in a stable radical solution. A wealth of MOFs, including NU-1000, UiO-66, UiO-67, PCN-94, Rod-8, MFM-300, and so on, have been rapidly prepared at room temperature in the radical solution in the absence external conditions. In particular, MOFs with conjugated ligands exhibit instantaneous formation (<1 s) with quantitative yield. Insight into the mechanism for MOF synthesis in the radical solution has been proposed based on the in-depth interrogation of electron paramagnetic resonance (EPR), in situ proton nuclear magnetic resonance (1H NMR), X-ray absorption spectra (XAS), in situ small-angle X-ray scattering (SAXS)-wide-angle X-ray scattering (WAXS), and density functional theory (DFT) calculations. Results and Discussion Rapid synthesis of NU-1000 in the radical solution at room temperature Room temperature synthesis of Zr-MOFs is relatively difficult due to its slow nucleation and crystallization.30–32 Thus, synthesis of NU-1000 was studied as an example. The structure of the ligand, 1,3,6,8-tetrakis(p-benzoicacid)pyrene (H4TBAPy), is shown in Supporting Information Figure S1. Solution 1 containing 25 mL dimethylformamide (DMF), 10 mL acetic acid (HAc) and 2 g ionic liquid (IL) 1-octyl-3-methylimidazolium bromide (OmimBr, Supporting Information Figure S2) was prepared. OmimBr was the supporting electrolyte and radical precursor. Solution 2 was prepared by the electrolysis of Solution 1 for 30 min (Figure 1a). MOFs were synthesized by the simple adding of metal salt followed by ligand to Solution 2 at room temperature (Figure 1b). The step for the MOF synthesis was conducted in the absence of external energy. Solution 2 became cloudy very quickly after adding precursors, and 30 min later the yield of MOF achieved 92% ( Supporting Information Table S1). Figure 1 | Schematic diagram for the MOF synthesis in the radical solution. (a) Preparation of Solution 2; (b) MOF synthesis in Solution 2. Download figure Download PowerPoint The as-prepared material at room temperature is denoted as NU-1000-a, the structure and phase purity of which were confirmed by powder X-ray diffraction (PXRD, Figure 2a). For comparison, NU-1000 was synthesized in a traditional solvothermal reaction (denoted as NU-1000-t) for 48 h at 120 °C.33 The structure, morphology, and Brunauer–Emmett–Teller (BET) surface area of NU-1000-a is similar to that of NU-1000-t (Figures 2a–2f), illuminating that NU-1000-a is similar to that synthesized by the solvothermal method. NU-1000-a obtains smaller particle size than NU-1000-t due to the shorter reaction time for crystallization. Compared with the solvothermal method, this strategy not only shortens reaction time from days to minutes but also lowers the reaction temperature from high temperature (∼120 °C) to room temperature. Figure 2 | Characterizations of NU-1000. (a) PXRD patterns; (b) N2 adsorption–desorption isotherms; (c) SEM image of NU-1000-a; (d) SEM image of NU-1000-t; (e) TEM image of NU-1000-a; (f) TEM image of NU-1000-t. The scale bars of (c) and (d) are 200 nm. Download figure Download PowerPoint Radicals in the solution Due to the absence of external energy during MOF synthesis, these fascinating results are attributed to the synthetic solution. OmimBr has a relatively low electrochemical window, and the Omim+ in OmimBr is possibly reduced to Omim• under negative voltages.34 EPR is the most efficient and direct technique to characterize radicals.35,36 Here, N-tert-Butyl-alpha-phenylnitrone (PBN), an efficient spin trap for trapping carbon-centered radicals,37 was used to identify the radicals produced in the solution. EPR signals appear after electrolysis of Solution 1 for 1 min and become more intense as electrolysis proceeds (Figure 3a). The triplet-of-doublets pattern arises from hyperfine coupling to the nitroxide 14N and the β-hydrogen in the nitroxyl adduct radical ( Supporting Information Figure S3). The EPR signals are assigned to carbon-centered radicals trapped by PBN, according to the EPR simulation parameters (A14N and AβH) of Solution 2 ( Supporting Information Figure S4 and Table S2).37,38 C2 of the imidazole ring obtains one electron to generate an Omim• radical at the cathode.39,40 Meanwhile, Br− at the anode is oxidized to Br2 to achieve charge balance at the anode ( Supporting Information Figure S5). Figure 3 | Related characterisations of radicals in the solution. (a) The dependence of EPR spectra of the solution on electrolysis time. (b) Variations of EPR spectra of Solution 2 with time. (c) Plots of Δδ of investigated hydrogens on the Omim• radical from 1H NMR spectra versus ZrCl4 contents in Solution 2; concentrations of ZrCl4 is 3.4 mM. (d) Independent gradient model of the interaction between Zr4+/H4TBAPy and the Omim• radical. Download figure Download PowerPoint The in situ 1H NMR spectrum was employed to study the changes in chemical shifts of hydrogens (δH) on investigated carbons of Omim+ ( Supporting Information Figure S6). δH on C2 shifts the most obviously in the shielded direction, and hydrogens at more distant carbons from C2 show more imperceptible upfield shifts, validating the production of radicals at C2 ( Supporting Information Figure S7). The shift of δH on C2 becomes slower after electrolysis for 30 min ( Supporting Information Figure S7), which signifies that the production of radicals tends to reach equilibrium due to the reversible redox.40 The concentration of radicals increases rapidly with the increase of reduction time from 0 to 30 min and then more slowly from 30 to 90 min, consistent with the EPR results in Figure 3a. After Solution 2 was placed in air for the desired time, PBN was added to investigate the stability of the Omim• radical in the solution. The EPR spectra of Solution 2 remain the same over time, evincing the high stability of the Omim• radical (Figure 3b), which agrees very well with the invariable δH on C2 in the 1H NMR spectra ( Supporting Information Figure S8). This indicates that the Omim• radical can be stabilized for at least 24 h. Rapid synthesis of NU-1000-a was also achieved in this solution at room temperature ( Supporting Information Figure S9), further suggestive of the stability and efficiency of the Omim• radical. Because it is feasible to aggregate the imidazolium cation, the as-produced Omim• radicals are protected in the aggregates to obtain a high stability. In Solution 2, the Omim• radical and Br2 are the only species that differ from Solution 1. NU-1000 cannot be generated in Solution 1 as well as Solution 1 with the desired amount of Br2 at room temperature, corroborating that the Omim• radical plays an important role in MOF formation. Several other ILs, such as 1-octyl-3-methylimidazolium chloride (OmimCl), 1-octyl-3-methylimidazolium tetrafluoroborate (OmimBF4), and 1-octyl-3-methylimidazolium hexafluor-ophosphate (OmimPF6), were used to replace OmimBr to prepare radical solutions. Only OmimCl-containing solution could generate radicals after electrolysis as characterized by EPR ( Supporting Information Figure S10), and NU-1000 could only be synthesized in this solution ( Supporting Information Figure S11). These control experiments confirm the crucial role of the Omim• radical for the rapid synthesis of MOFs at room temperature. Interactions between MOF precursors and radicals in solution The EPR spectra of Solution 2 is maintained after the addition of ZrCl4 and H4TBAPy, respectively, which indicates that these precursors are unable to quench the Omim• radical and predicates the suitability of Solution 2 as the medium for MOF synthesis ( Supporting Information Figure S12). The influence of precursors was further investigated by 1H NMR. ZrCl4 exerts considerable influence on δH on C2 (Figure 3c and Supporting Information Figure S13), and the increase of ZrCl4 concentration renders a linearly escalating downfield shift. A similar tendency is found from the shift of δH on C4∼C7. DFT calculations were further conducted to provide the computational interactions between MOF precursors and the Omim• radical. Because the charge distribution of the Omim• radical is uneven, Zr4+ exhibits very strong interaction (34.12 eV) with the imidazole ring of the Omim• radical due to the Coulomb interaction (Figure 3d), consistent with the 1H NMR results. The binding energy between TBAPy4− and the Omim• radical is 2.12 eV, owing to the strong π–π interaction between the conjugated pyrene ring of TBAPy4− and the Omim• radical (Figure 3d). Mechanism for the rapid synthesis of MOFs at room temperature The coordination behavior is explored by XAS.41,42 ZrCl4-containing Solution 2 exhibits a stronger intensity of the white line from Zr K-edge X-ray absorption near-edge spectrum (XANES) than ZrCl4-containing DMF and ZrCl4-containing Solution 1 (Figure 4a). Zr–O and Zr···Zr (Zr–O–Zr) bonds in the regions of 1–2 Å and 2.9–3.6 Å (Figure 4b), respectively, were observed in R-space from the k3-weighted Fourier transform of the extended X-ray absorption fine structure (EXAFS) spectra, indicative of instantaneous formation of Zr4+-based coordination species in Solution 2. The coordination structure observed in Solution 2 is similar to that of NU-1000 by fitting Zr K-edge EXAFS spectra ( Supporting Information Figure S14 and Table S3). As the modulator HAc is the sole species to coordinate with ZrCl4 in the absence of ligand, Zr···Zr and Zr–O bonds derive from the coordination between ZrCl4 and HAc in Solution 2. Figure 4 | Mechanism study for the formation of NU-1000-a. (a) Zr K-edge XANES spectra. (b) EXAFS spectra in R-space. (c) In situ SAXS of the solution system. (d) In situ WAXS of the solution system. (e) Variation of PXRD patterns on time after the precipitation of NU-1000-a. (f) Variation of BET surface area and morphology (insets are SEM images) on time after the precipitation of NU-1000-a. The concentration of ZrCl4 in (a) and (b) is 10 mM. Download figure Download PowerPoint 1H NMR results verify the coordination behavior as well. After the addition of ZrCl4 to Solution 2, the peak assigned to the methyl of HAc from 1H NMR spectrum shifts toward the deshielded direction, which shifts more obviously with the increase of ZrCl4 concentration ( Supporting Information Figure S15). In contrast, no variation is observed after the addition of ZrCl4 to Solution 1 ( Supporting Information Figure S15), suggesting that the Omim• radical promotes the coordination between Zr4+ and Ac−. Thus, the Zr-Ac cluster, produced by the coordination of Zr4+ and Ac− is instantaneously formed as soon as the ZrCl4 is added to Solution 2. The unobservable peak of H4TBAPy in the 1H NMR spectrum after its introduction to Solution 2 ( Supporting Information Figure S16) confirms an instantaneous replacement of Ac− in the Zr-Ac cluster by H4TBAPy to form NU-1000 unit. The EPR signal of the Omim• radical in Solution 2 is maintained during the formation of MOFs ( Supporting Information Figure S17), elucidating that the the Omim• radical is stable. The kinetics of the MOF formation was studied by in situ SAXS–WAXS techniques. As soon as precursors of NU-1000-a were added into Solution 2, much stronger scattering signals were generated from in situ SAXS patterns (Figure 4c), illuminating that the nuclei formation of NU-1000 is instantaneous. After 30 min, NU-1000-a precipitated from Solution 2, resulting in the steeper SAXS scattering curve. Meanwhile, WAXS peaks of NU-1000-a appear (Figure 4d), indicating the formation of a well-ordered structure. The SAXS and WAXS curves remain the same with the increase of reaction time after the precipitation of NU-1000-a (Figures 4c and 4d), suggesting that the solid does not undergo further growth and crystallization. The variation of PXRD patterns, scanning electron microscopy (SEM) images, and the BET surface area of NU-1000-a over time is negligible, as shown in Figures 4e and 4f, indicating that the structure, morphology, and porosity of MOFs are maintained after their formation. We can conclude that the radical-containing Solution 2 favors the coordination between Zr4+ and modulator Ac− as well as the replacement of Ac− in the Zr-Ac cluster in terms of the aforementioned experiments. The mechanism for the rapid synthesis of Zr-MOFs at room temperature is proposed based upon a radical route (Figure 5). Initially, Omim+ cation in Solution 1 obtains one electron to be changed into the Omim• radical in Solution 2. After the introduction of ZrCl4 to Solution 2, radicals are concentrated by Zr4+ cations by virtue of the Coulomb force. As radicals are electroneutral, no bond is formed between Zr4+ cations and radicals. However, the local radical microenvironment can increase the magnetic shielding of the oxygen nucleus, which makes it easier for the oxygen to donate the lone-pair electrons.43,44 Thus,the electron-donating ability of oxygen on Ac− is improved, resulting in the acceleration of the coordination between Zr4+ and Ac− to form Zr–Ac. Ligands interact with radicals via the π–π interaction,45 which results in rich electron density around the oxygen in the ligands. Thus, oxygen on the carboxyl in ligands exhibits stronger electron-donating ability than that in Ac−, leading to the rapid replacement of Ac− in Zr–Ac by ligand anions. Ligand has stronger π–π interactions with the Omim• radical, and thus the replacement of Ac− by the ligand is faster. After a long time, the Omim• radical loses the electron and changes into Omim+. The whole process is a cyclic pathway. Figure 5 | Route for the synthesis of Zr-MOFs assisted by the Omim• radical. Download figure Download PowerPoint In brief, coordination, the first step and a rate-determining step for the formation of MOFs, greatly influences the kinetics for the MOF synthesis. The Omim• radical is concentrated by metal ions to form a radical-rich microenvironment, and it enables an instantaneous coordination between metal ions and modulators in solution. The replacement of modulators by ligands is accelerated by radicals as well. In this process, the Omim• radical can be considered as a catalyst that drastically reduces the activation energy for the coordination, resulting in the rapid kinetics and room temperature synthesis. Universality and tunability of this strategy Upon completion of the synthesis of NU-1000-a, the Omim• radicals are preserved, and MOFs can also be synthesized in the recycled solution ( Supporting Information Figures S18 and S19), illustrating that the Omim• radical contributes to a cyclic pathway. The recyclability of the radical solution indicates that the new route for MOF synthesis is sustainable. Various MOFs, including UiO-66, UiO-67, PCN-94, Rod-8, MFM-300 (Ga), and so on, with versatile metal nodes and ligands (structures are shown in Supporting Information Figure S2), have been rapidly synthesized at room temperature by using this method (Figure 1). Metal salts and ligands used as well as the starting time for the generation of different MOFs in Solution 2 are shown in Supporting Information Table S1. PXRD patterns of these MOFs confirm that the new route is efficient to realize rapid MOF synthesis at room temperature ( Supporting Information Figures S20–S26). MOFs with more conjugated ligands ( Supporting Information Figure S27) achieve more rapid kinetics, owing to the stronger π–π interaction between conjugated ligands and radicals. The reaction time can also be shortened by reducing the content of HAc. For example, NU-1000-b could be instantaneously generated (<1 s) after the introduction of its precursor and achieve a quantitative yield ( Supporting Information Table S1). Due to the extremely rapid kinetics, the instantaneously synthesized MOFs exhibit low crystallinity and small particle size of less than 20 nm ( Supporting Information Figure S28), which is beneficial for heterogeneous catalysis ( Supporting Information Table S4 and Figures S29 and S30). Additionally, since NU-1000-b can be obtained with a yield of 100%, Solution 2 is recovered after the removal of product and used to synthesize NU-1000-b. The yield was not changed in five cycles, further proving the cyclic stability. Conclusion Multitudinous MOFs can be rapidly synthesized at room temperature in a novel radical solution without external energy. The stable Omim• radical in the solution greatly accelerates the coordination between metal centers and ligands to achieve rapid synthesis at room temperature. The radical solution simplifies the strategy for MOF synthesis tremendously and easily realizes the tunable structure of MOFs. The universal, tunable, and sustainable radical-based route has great potential for applications, and we believe that this creative radical-assisted strategy can also be used in some other reactions and material fabrications. Supporting Information Supporting Information is available and includes material characterizations, DFT calculations, Meerwein–Ponndorf–Verley reactions, and supplementary figures and tables. Conflict of Interest There is no conflict of interest to report. Acknowledgments The authors thank the National Natural Science Foundation of China (grant no. 22073104), the Beijing Natural Science Foundation (grant no. 2222043), the National Key Research and Development Program of China (grant nos. 2017YFA0403101, 2017YFA0403003, and 2017YFA0403102), the National Natural Science Foundation of China (grant nos. 21890761, 21733011, and 21533011), the Beijing Municipal Science & Technology Commission (grant no. Z191100007219009), and the Chinese Academy of Sciences (grant no. QYZDY-SSW-SLH013). The XAS (1W1B) and SAXS (1W2A) measurements were performed at the Beijing Synchrotron Radiation Facility, China. References 1. Hou B. S.; Qin C.; Sun C. Y.; Wang X. L.; Su Z. M.Stepwise Construction of Multivariate Metal-Organic Frameworks from a Predesigned Zr16 Cluster.CCS Chem.2021, 3, 287–293. Abstract, Google Scholar 2. Feng L.; Wang K. Y.; Lv X. L.; Metal-Organic and Google Scholar Chen Z. Wang Zhang L.; L.; K. Y.; Zhang Han and in a Metal-Organic at Google Scholar Wang K. C.; L.; Feng L.; Synthesis of Metal–Organic Google Scholar Wang X. S.; Metal-Organic and Google Scholar S.; C.; C.; K. S.; K. in and of Metal-Organic Google Scholar Z. Y.; Li B. Zhang Synthesis of Metal-Organic a Google Scholar Hou B. S.; Qin C.; Sun C. Y.; Wang X. L.; Su Z. M.Stepwise Construction of Multivariate Metal–Organic Frameworks from a Predesigned Zr16 Cluster.CCS Chem.2021, 3, 287–293. Abstract, Google Scholar Chen Y.; Wang in the Synthesis and of Metal-Organic Google Scholar Li Y.; Li C.; Wang Y.; Sun Li Y.; Su B. in [email protected] for High 2, Google Scholar of Metal-Organic Frameworks to Various MOF and Google Scholar C.; of Metal-Organic Frameworks by Google Scholar L.; L.; C. C. as an for Room Temperature Synthesis and of Metal-Organic Google Scholar Zhang X. Synthesis of Metal-Organic Frameworks with and for Google Scholar L.; S.; Synthesis of Metal-Organic in and as and Google Scholar L.; C.; Room Synthesis of Metal-Organic Frameworks with High 3, Google Scholar and of Metal-Organic via a at Room Google Scholar Duan C. Y.; Li Wu Y.; Synthesis of Metal-Organic Frameworks with High Google Scholar Zhang L.; X. Y.; X. Zhang B. Han B. Zhang Y.; Xiang and Synthesis of Metal-Organic in Google Scholar Temperature Synthesis of Metal-Organic and Google Scholar Li S.; Wang Wang Z. S.; Chen Y.; Zhang Z. Synthesis of Metal-Organic Frameworks for Google Scholar S.; Zhang Synthesis of Metal Metal-Organic Google Scholar S.; Synthesis of Metal-Organic Google Scholar C. Li Zhang Temperature Synthesis of an Metal-Organic for Google Scholar Sun X. Li Y.; Zhang X. Zhang and Synthesis of with Google Scholar S.; Y.; Feng Zhang Synthesis of with China Google Scholar Wang C.; for the and of Metal-Organic Google Scholar A New of Google Scholar Li X. Y.; and of Metal-Organic and Google Scholar Chen Z. Wang X. C. Room and Synthesis of Metal-Organic Frameworks for Chemical Google Scholar Synthesis of and of of Google Scholar X. Zhang L.; Xiang Zheng Zhang Wu Z. Li Z. Mo Y.; L.; C. C.; X. Zhang B. Han B. the of Metal-Organic Google Scholar S.; C.; by in a Metal-Organic Google Scholar and of in 1. Google Scholar K. C.; S.; and Google Scholar L.; L.; and of in 2. Google Scholar of 3, Google Scholar of EPR Google Scholar C. for the of in Google Scholar Zhang X. Y.; L.; of to in Google Scholar Y.; Y.; of and to over on Google Scholar of with EXAFS and Google Scholar C.; C. of and in and on the NMR Chemical in Google Scholar B. Chen Li Y.; Wang from a Metal-Organic for Google Scholar C.; C.; K. C.; Molecular 2, Google Scholar Information & Chinese Chemical temperature authors thank the National Natural Science Foundation of China (grant no. 22073104), the Beijing Natural Science Foundation (grant no. 2222043), the National Key Research and Development Program of China (grant nos. 2017YFA0403101, 2017YFA0403003, and 2017YFA0403102), the National Natural Science Foundation of China (grant nos. 21890761, 21733011, and 21533011), the Beijing Municipal Science & Technology Commission (grant no. Z191100007219009), and the Chinese Academy of Sciences (grant no. QYZDY-SSW-SLH013). The XAS (1W1B) and SAXS (1W2A) measurements were performed at the Beijing Synchrotron Radiation Facility, China.