English

• CO₂ emission from the excessive use of fossil fuel has become one of the most serious global environmental problems [1-4]. Efficient capture and utilization of CO₂ is one of the keys to achieve the goal of carbon neutrality [5-13]. The cycloaddition of CO₂ and epoxide, as the key reaction to produce biodegradable polymers and electrolytes for lithium-ion battery, and featuring the atom-utilization efficiency of 100%, is one of the most efficient ways for CO₂ fixation [14-18]. In industry, homogeneous catalysts are prerequisite for this reaction so far, and many efforts have been devoted to developing high-performance heterogeneous catalysts [19-21]. If the catalyst have high activity at room temperature and low CO₂ concentration, the flue gases (ca. 15% CO₂) [22-25] could be directly used as the feedstock, which avoids the energy-consuming processes of CO₂ separation, transportation, pressurization, etc. [26-29].

  Among numerous types of porous materials, metal‒organic frameworks (MOFs) possess many advantages for heterogeneous catalysis including convenient synthesis and modification, well-defined active sites, and uniform pore structures [30,31]. Moreover, MOFs can supply some unique active structures which can be hardly constructed in other types of heterogeneous catalysts [32-34]. Even so, high temperature and/or high pressure are generally needed for CO₂ cycloaddition reactions, especially when low-concentration CO₂ is used as the reactant [35-37], which illustrates the relatively low activities of known MOF catalysts.

  As a kind of famous MOFs with excellent stability and relatively low cost, metal azolate frameworks (MAFs), such as MAF-4/ZIF-8, demonstrate great advantages in adsorption, separation, sensing and more, but exhibit relatively low catalytic performances for CO₂ cycloaddition [17,38], indicating low acidity of the simple tetrahedral metal centres coordinated with nitrogen donors. Recent research revealed that high-valency non-3d metals are promising to modulate the electronic structures of catalytic active centres [39,40]. For example, [Co₄(MoO₄)(eim)₆] (MAF-69-Co-Mo, Heim = 2-ethylimidazole) and [Co₄(WO₄)(eim)₆] (MAF-69-Co-W) showed high electro-catalytic OER performances because the MoO₄^2– and WO₄^2– units can tailor the electronic structures of Co(Ⅱ) centres [41]. In this work, we synthesized their zinc(Ⅱ) analogues for high-performance CO₂ cycloaddition under ambient condition using pure CO₂ and even simulated flue gas as the reaction source.

  Considering that Zn(Ⅱ)-based MOFs usually show higher activity than Co(Ⅱ)-based ones for CO₂ cycloaddition reactions [42-44], two new MOFs, namely [Zn₄(MoO₄)(eim)₆] (denoted as MAF-69-Zn-Mo or 1) and [Zn₄(WO₄)(eim)₆] (denoted as MAF-69-Zn-W or 2), were synthesized according to the synthetic method for MAF-69-Co (Fig. 1a). Both single-crystal X-ray diffraction (SCXRD) and powder X-ray diffraction (PXRD) analyses confirmed that both 1 and 2 are isostructural with MAF-69-Co (Fig. 1, Fig. 2, Figs. S1 and S2, Table S1 in Supporting information). The Zn–N bond lengths in 1 are slightly shorter than those in 2, implying that the electron-withdrawing effect of MoO₄^2– toward Zn(Ⅱ) is stronger than that of WO₄^2–. This result could be further confirmed by Fourier transform infrared (FT-IR) spectra (Fig. S3 in Supporting information) [45]. To study the role of non-3d metal modulation effects, RHO-[Zn(eim)₂] (MAF-6) [46], an analogue of zinc imidazolate, was also synthesized for comparison. PXRD confirmed the phase purity of all microcrystalline samples (Fig. S4 in Supporting information). Scanning electron microscope (SEM) showed particle sizes of ca. 630, 131 and 420 nm for 1, 2 and MAF-6, respectively (Figs. 2b and c, Figs. S5−S8 in Supporting information).

  Figure 1

  

  Figure 1.  (a) Schematic presentation of tailoring d-band electronic structure of Zn by MO₄ modulation. (b) Structure and topology of 1 and 2 (yellow: Mo or W; green: Zn; blue: N; grey: C; red: O).

  DownLoad: Full-Size Img PowerPoint

  Figure 2

  

  Figure 2.  (a) PXRD patterns. (b) SEM image and (c) particle size distribution of 1. (d) High-resolution Zn 2p orbital XPS profiles. (e) NH₃-TPD.

  DownLoad: Full-Size Img PowerPoint

  Both 1 and 2 exhibited excellent thermal and chemical stabilities. Thermogravimetry and PXRD revealed that 1 and 2 can be stable up to at least 400 ℃ (Figs. S9-S11 in Supporting information). PXRD showed that 2 and 1 can be stable in water under pH 1−13 and pH 3−13 for at least one week, respectively (Figs. S12 and S13 in Supporting information). At 298 K and 1 atm (Fig. S14 in Supporting information), the CO₂ uptake of 1 (0.44 mmol/g) is higher than that of 2 (0.37 mmol/g), implying 1 exhibits stronger CO₂ affinity than 2 does, as their void volumes (1: 0.14 cm³/g; 2: 0.13 cm³/g) are highly similar.

  X-ray photoelectron spectroscopy (XPS) was employed to investigate the electronic structures of metal ions, which usually have crucial influence on the catalytic properties (Fig. 2d). In the high-resolution spectra of the 2p orbits of Zn, the orbital binding energies follow 1 (1044.8 and 1021.8 eV) > 2 (1044.7 and 1021.7 eV) > MAF-6 (1044.6 and 1021.6 eV), indicating that a stronger electron-withdrawing property of MO₄^2– and an increased charge density of Zn, consistent with the bond length results (Table S2 in Supporting information). Moreover, in the temperature programmed desorption curves of NH₃, the desorption temperatures and amounts follow 1 > 2 > MAF-6 (Fig. 2e), demonstrating a Lewis acidity sequence of 1 > 2 > MAF-6.

  The CO₂ cycloaddition reaction with epichlorohydrin (ECH) as the reaction substrate and tetrabutylammonium bromide (TBAB) as a co-catalyst at ambient condition (Scheme 1) was employed as a model reaction to evaluate the enhanced catalytic activities of 1. Through Proton nuclear magnetic resonance (¹H-NMR), the reaction process and the corresponding conversion can be determined. In Fig. 3a and Fig. S15 (Supporting information), compound 1 exhibits a 95% and 99% of ECH conversion ratio after 24 and 48 h in pure CO₂, respectively, and thus the reaction in pure CO₂ was fixed as 24 h in subsequent comparison. Notably, compound 1 represents a rare example that can efficiently catalyze CO₂ cycloaddition reactions at room temperature and ambient pressure and such high conversion is among the highest values for MOF catalysts at ambient condition (Table S3 in Supporting information) [17]. Besides, the conversions of ECH are found to be close to 0 when the TBAB was absent. The necessity of TBAB (nucleophilic Br⁻ as Lewis base to attack epoxide and open the ring) illustrates that these MAF catalysts only provide the metal sites as Lewis acid for the reaction (Fig. S16 in Supporting information). Expectedly, compound 1 showed high recyclability for the ECH-CO₂ cycloaddition reaction (Fig. S17 in Supporting information). After five reaction cycles, the conversion can still reach 97% in 24 h, and the recycled frameworks remained intact as verified by PXRD (Fig. S18 in Supporting information). In addition, the reactants were extended to other epoxides to further verify the catalytic activity of 1 (Table S4 in Supporting information). When the reactant was butylene oxide, the conversion ratio also can reach 94% after 24 h. Nevertheless, the conversion ratio gradually decreases for the bulkier epoxides, which may be attributed to the limitation of pore on the diffusion of large reactants.

  Scheme 1

  

  Scheme 1.  Catalytic cycloaddition reaction between ECH and CO₂.

  DownLoad: Full-Size Img PowerPoint

  Figure 3

  

  Figure 3.  (a) Reaction conversions of cycloaddition between CO₂ and ECH under ambient conditions catalyzed by 1 and 2. (b) Comparison of catalytic efficiencies at ambient temperature and pressure of 1, 2 and MAF-6 for CO₂ cycloaddition of ECH in pure CO₂ for 24 h and in simulated flue gas for 48 h, respectively.

  DownLoad: Full-Size Img PowerPoint

  The 24-h conversion of ECH in pure CO₂ catalyzed by 1, 2 and MAF-6 were determined as 95%, 86% and 73%, respectively. And the corresponding turnover frequency (TOF) can be calculated as 19.8, 17.9 and 14.8 h⁻¹, respectively. Furthermore, it is known that the catalytic activities of MAF-4 and [Zn(cbIm)₂] (ZIF-95, HcbIm = 5-chlorobenzimidazole) are based on coordinated defects on the crystal surface, i.e., Zn ion and terminal coordinated hydroxide, whose acidity and basicity are weak so that their performances are low even with TBAB [47,48]. In principle, the MO₄^2– units in 1 and 2 not only increase the acidity of Zn but also decrease the basicity of the terminal hydroxide group. Therefore, they exhibit enhanced catalytic activity toward CO₂ cycloaddition compared with the simple binary Zn-imidazolate framework. Moreover, the high catalytic activity of 1 was highlighted by using simulated flue gas as the CO₂ source. Under such a low CO₂ concentration, the conversion ratio still reached 48% after 24 h and 98% after 48 h (Fig. 3b and Fig. S19 in Supporting information), which are among the highest for MOF catalysts at similar conditions (Table S5 in Supporting information) [49,50]. As a comparison, the 48-h conversion ratios merely reach 84% and 33% for 2 and MAF-6, respectively, also verifying the catalytic activity order of 1 > 2 > MAF-6.

  In order to understand the different catalytic performances of zinc imidazolate centres, we also carried out computational simulations by density functional theory (DFT). We initially calculated the electrostatic potential (ESP) charges of zinc atoms, giving the ESP₁ of 0.498 and ESP_MAF-6 of 0.326, respectively, which agree well with the experimental results of XPS and TPD (1 with more positively charged zinc centre). The partial density of states (PDOS) of d orbital on the active metal sites, was shown in Fig. S20 (Supporting information). The d-orbital centre of 1 was remarkably shifted toward the Fermi level compared with that of MAF-6, demonstrating the electronic modulation effect of MoO₄²⁻ unit toward zinc centre. Furthermore, according to Park et al. [51], the whole reaction process contains ring opening of the three-membered ring and CO₂ insertion to form carbonate ester (Fig. 4 and Fig. S21 in Supporting information), involving two corresponding transient states (TS). According to the calculation results, 1 exhibits a more mitigated energy change thermodynamically during the reaction process, suggesting the optimal electronic structure of Zn centre of 1. Moreover, the rate determined step (RDE) can be confirmed as the ring opening reaction, during which the α-C atom of ECH is attacked by the Br⁻ of TBAB to form the bromo-substituted alkoxide intermediate (TS1), also verifying the necessity of co-catalysts. For MAF-6, the activation energy of TS1 was 116.3 kJ/mol, whereas that of 1 was dramatically decreased to 36.1 kJ/mol, being consistent with the experimental activities. Alternative to MAF-6, the reaction activation energy of TS1 is probably reduced by the unique hydrogen bonding interaction (C-H⋯O = 2.39 Å) between the proton on α-C atom of ECH and the coordinated O atom on Zn centre in 1 (Fig. S22 in Supporting information). These results indicate that the insight of MoO₄²⁻ unit on boosting CO₂ cycloaddition includes optimizing the electronic structure of Zn centre, facilitating the rate-determined ring opening process and minimizing the reaction activation energy.

  Figure 4

  

  Figure 4.  DFT-derived energy profiles of the proposed ECH cycloaddition process catalyzed by 1 and MAF-6 with corresponding structures of the active sites and intermediates of 1 during reaction. The capping ligand, ethyl group and hydrogen atoms are omitted for clarity (yellow: Mo; green: Zn; blue: N; grey: C; red: O; light green: Cl; white: H; pink: Br).

  DownLoad: Full-Size Img PowerPoint

  In summary, two non-3d metal integrated zinc imidazolate frameworks with high thermal and chemical stabilities were successfully synthesized. Spectroscopic characterizations and DFT calculations revealed that the non-3d metal oxides MO₄^2– units can effectively tailor the electronic structure of Zn(Ⅱ) ion, resulting that 1 exhibited the highest catalytic activity for cycloaddition between CO₂ and ECH under room temperature with low concentration CO₂ or simulated flue gas, which is very important for practical application in reducing the energy consumption. Our findings provide a new thought to the design of coordination catalysts for efficient CO₂ conversion.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 22090061, 21731007, 21890380 and 22161021) and the Guangdong Pearl River Talents Program (No. 2017BT01C161). C.-T. He acknowledges the support of Jiangxi Province (No. jxsq2018106041). Prof. Jie-Peng Zhang is appreciated for his helpful suggestions and discussion.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.cclet.2022.107814.

  
  
• 1. [1]

     L. Yang, C. Zhang, X. Yu, et al., Natl. Sci. Rev. 8 (2021) nwab104. doi: 10.1093/nsr/nwab104

  2. [2]

     X. Ren, S. Liu, H. Li, et al., Sci. China Chem. 63 (2020) 1727–1733. doi: 10.1007/s11426-020-9847-9

  3. [3]

     S.C. Peter, ACS Energy Lett. 3 (2018) 1557–1561. doi: 10.1021/acsenergylett.8b00878

  4. [4]

     J. Liu, C. Chen, K. Zhang, L. Zhang, Chin. Chem. Lett. 32 (2021) 649–659. doi: 10.1016/j.cclet.2020.07.040

  5. [5]

     M. Ding, R.W. Flaig, H.L. Jiang, O.M. Yaghi, Chem. Soc. Rev. 48 (2019) 2783–2828. doi: 10.1039/C8CS00829A

  6. [6]

     W. Gao, S. Liang, R. Wang, et al., Chem. Soc. Rev. 49 (2020) 8584–8686. doi: 10.1039/D0CS00025F

  7. [7]

     D.D. Zhou, X.W. Zhang, Z.W. Mo, et al., Energy Chem. 1 (2019) 100016. doi: 10.1016/j.enchem.2019.100016

  8. [8]

     Q. Wang, H. Pfeiffer, R. Amal, D. O’Hare, React. Chem. Eng. 7 (2022) 487–489. doi: 10.1039/D2RE90007F

  9. [9]

     K. Jiang, P. Ashworth, S. Zhang, et al., Renew. Sust. Energy Rev. 119 (2020) 109601. doi: 10.1016/j.rser.2019.109601

  10. [10]

      Y. Shi, J. Zhao, H. Xu, S.L. Hou, B. Zhao, Sci. China Chem. 64 (2021) 1316–1322. doi: 10.1007/s11426-021-1006-9

  11. [11]

      N. Wang, R.K. Miao, G. Lee, et al., SmartMat 2 (2021) 12–16. doi: 10.1002/smm2.1018

  12. [12]

      C. Yang, Z. Gao, D. Wang, et al., Sci. China Mater. 65 (2022) 155–162. doi: 10.1007/s40843-021-1749-5

  13. [13]

      X. Dong, Z. Xin, D. He, et al., Chin. Chem. Lett. 34 (2023) 107459. doi: 10.1016/j.cclet.2022.04.057

  14. [14]

      M. Liu, X. Wang, Y. Jiang, J. Sun, M. Arai, Catal. Rev. 61 (2019) 214–269. doi: 10.1080/01614940.2018.1550243

  15. [15]

      G. Yuan, Y. Lei, X. Meng, et al., Inorg. Chem. Front. 9 (2022) 1208–1216. doi: 10.1039/D1QI01643A

  16. [16]

      J. Hu, H. Liu, B. Han, Sci. China Chem. 61 (2018) 1486–1493. doi: 10.1007/s11426-018-9396-3

  17. [17]

      J. Liang, Y.B. Huang, R. Cao, Coord. Chem. Rev. 378 (2019) 32–65. doi: 10.1016/j.ccr.2017.11.013

  18. [18]

      H. Hu, D. Zhang, H. Liu, et al., Chin. Chem. Lett. 32 (2021) 557–560. doi: 10.1016/j.cclet.2020.02.028

  19. [19]

      G.N. Bondarenko, E.G. Dvurechenskaya, O.G. Ganina, F. Alonso, I.P. Beletskaya, Appl. Catal. B: Environ. 254 (2019) 380–390. doi: 10.1016/j.apcatb.2019.04.024

  20. [20]

      A. Rehman, V.C. Eze, M.F.M.G. Resul, A. Harvey, J. Energy Chem. 37 (2019) 35–42. doi: 10.1016/j.jechem.2018.11.017

  21. [21]

      L.G. Ding, B.J. Yao, F. Li, et al., J. Mater. Chem. A 7 (2019) 4689–4698. doi: 10.1039/C8TA12046C

  22. [22]

      T.M. McDonald, W.R. Lee, J.A. Mason, et al., J. Am. Chem. Soc. 134 (2012) 7056–7065. doi: 10.1021/ja300034j

  23. [23]

      O.H. Animasahun, M.N. Khan, C.J. Peters, Mol. Phys. 116 (2018) 3434–3445. doi: 10.1080/00268976.2018.1509145

  24. [24]

      D. Lv, J. Chen, K. Yang, et al., Chem. Eng. J. 375 (2019) 122074. doi: 10.1016/j.cej.2019.122074

  25. [25]

      H. Lv, R. Sa, P. Li, et al., Sci. China Chem. 63 (2020) 1289–1294. doi: 10.1007/s11426-020-9801-3

  26. [26]

      Y. Cheng, J. Hou, P. Kang, ACS Energy Lett. 6 (2021) 3352–3358. doi: 10.1021/acsenergylett.1c01553

  27. [27]

      J. Chen, H. Li, M. Zhong, Q. Yang, Green Chem. 18 (2016) 6493–6500. doi: 10.1039/C6GC02367C

  28. [28]

      H. Zhong, J. Gao, R. Sa, et al., ChemSusChem 13 (2020) 6323–6329.

  29. [29]

      F. Duan, X. Liu, D. Qu, B. Li, L. Wu, CCS Chem. 3 (2021) 2676–2687. doi: 10.31635/ccschem.020.202000498

  30. [30]

      C. Duan, K. Liang, J. Lin, et al., Sci. China Mater. 65 (2022) 298–320. doi: 10.1007/s40843-021-1910-2

  31. [31]

      G. Si, X. Kong, T. He, et al., Chin. Chem. Lett. 32 (2021) 918–922. doi: 10.1016/j.cclet.2020.07.023

  32. [32]

      J. Liang, Y.Q. Xie, X.S. Wang, et al., Chem. Commun. 54 (2018) 342–345. doi: 10.1039/C7CC08630J

  33. [33]

      Y. Peng, W. Yang, Sci. China Chem. 62 (2019) 1561–1575. doi: 10.1007/s11426-019-9575-8

  34. [34]

      X.J. Hu, Z.X. Li, H. Xue, et al., CCS Chem. 2 (2020) 616–622. doi: 10.31635/ccschem.019.201900040

  35. [35]

      B. Ugale, S. Kumar, T.J. Dhilip Kumar, C.M. Nagaraja, Inorg. Chem. 58 (2019) 3925–3936. doi: 10.1021/acs.inorgchem.8b03612

  36. [36]

      W. Zhang, F. Ma, L. Ma, Y. Zhou, J. Wang, ChemSusChem 13 (2020) 341–350. doi: 10.1002/cssc.201902952

  37. [37]

      S.H. Kim, R. Babu, D.W. Kim, W. Lee, D.W. Park, Chin. J. Catal. 39 (2018) 1311–1319. doi: 10.1016/S1872-2067(17)63005-5

  38. [38]

      C.M. Miralda, E.E. Macias, M. Zhu, P. Ratnasamy, M.A. Carreon, ACS Catal. 2 (2012) 180–183. doi: 10.1021/cs200638h

  39. [39]

      B. Zhang, L. Wang, Z. Cao, et al., Nat. Catal. 3 (2020) 985–992. doi: 10.1038/s41929-020-00525-6

  40. [40]

      F. Wang, Z.S. Liu, H. Yang, Y.X. Tan, J. Zhang, Angew. Chem. Int. Ed. 50 (2011) 450–453. doi: 10.1002/anie.201005917

  41. [41]

      Y.T. Xu, Z.M. Ye, J.W. Ye, et al., Angew. Chem. Int. Ed. 58 (2019) 139–143. doi: 10.1002/anie.201809144

  42. [42]

      B. Mousavi, S. Chaemchuen, B. Moosavi, et al., Chem. Open. 6 (2017) 674–680.

  43. [43]

      J. Sirijaraensre, New J. Chem. 43 (2019) 11692–11700. doi: 10.1039/C9NJ02566A

  44. [44]

      R. Zou, P.Z. Li, Y.F. Zeng, et al., Small 12 (2016) 2334–2343. doi: 10.1002/smll.201503741

  45. [45]

      E. Shamsaei, Z.X. Low, X. Lin, et al., Chem. Commun. 51 (2015) 11474–11477. doi: 10.1039/C5CC03537F

  46. [46]

      X.C. Huang, Y.Y. Lin, J.P. Zhang, X.M. Chen, Angew. Chem. Int. Ed. 45 (2006) 1557–1559. doi: 10.1002/anie.200503778

  47. [47]

      K.M. Bhin, J. Tharun, K.R. Roshan, et al., J. CO2 Util. 17 (2017) 112–118. doi: 10.1016/j.jcou.2016.12.001

  48. [48]

      R.R. Kuruppathparambil, R. Babu, H.M. Jeong, et al., Green Chem. 18 (2016) 6349–6356. doi: 10.1039/C6GC01614F

  49. [49]

      R. Das, D. Muthukumar, R.S. Pillai, C.M. Nagaraja, Chem. Eur. J. 26 (2020) 17445–17454. doi: 10.1002/chem.202002688

  50. [50]

      J. Gu, X. Sun, X. Liu, et al., Inorg. Chem. Front. 7 (2020) 4517–4526. doi: 10.1039/D0QI00938E

  51. [51]

      A.C. Kathalikkattil, R. Babu, R.K. Roshan, et al., J. Mater. Chem. A 3 (2015) 22636–22647. doi: 10.1039/C5TA05688H

• 

  Figure 1  (a) Schematic presentation of tailoring d-band electronic structure of Zn by MO₄ modulation. (b) Structure and topology of 1 and 2 (yellow: Mo or W; green: Zn; blue: N; grey: C; red: O).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) PXRD patterns. (b) SEM image and (c) particle size distribution of 1. (d) High-resolution Zn 2p orbital XPS profiles. (e) NH₃-TPD.

  下载: 全尺寸图片 幻灯片
  

  Scheme 1  Catalytic cycloaddition reaction between ECH and CO₂.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Reaction conversions of cycloaddition between CO₂ and ECH under ambient conditions catalyzed by 1 and 2. (b) Comparison of catalytic efficiencies at ambient temperature and pressure of 1, 2 and MAF-6 for CO₂ cycloaddition of ECH in pure CO₂ for 24 h and in simulated flue gas for 48 h, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  DFT-derived energy profiles of the proposed ECH cycloaddition process catalyzed by 1 and MAF-6 with corresponding structures of the active sites and intermediates of 1 during reaction. The capping ligand, ethyl group and hydrogen atoms are omitted for clarity (yellow: Mo; green: Zn; blue: N; grey: C; red: O; light green: Cl; white: H; pink: Br).

  下载: 全尺寸图片 幻灯片
•