English

• Electronic specialty gases (ESGs), especially fluorinated gases (F-gases), play pivotal roles in the advancement of semiconductor and microelectronics industries [1–7]. Hexafluoroethane (C₂F₆) as an important perfluorinated electronic specialty is highly preferred in the advanced electronics industry for its excellent etching performance, high chemical stability, low global warming potential (GWP), and low production cost [8–10]. In industry, C₂F₆ is predominantly manufactured by fluorinating 1,1,1,2-tetrafluoroethane (CF₃CH₂F) and pentafluoroethane (CF₃CHF₂), so that small amounts of CF₃CH₂F or CF₃CHF₂ would invariably remain in the crude product [11]. Given the extremely similar boiling points (CF₃CH₂F, 246.7 K; CF₃CHF₂, 224.6 K; C₂F₆, 195.1 K) and electronic-grade purity requirement, it is difficult and energy-consuming to obtain high purity C₂F₆ industrially using cryogenic distillation methods [12–15].

  Adsorptive separation process has been increasingly regarded as the most promising gas purification technology due to its high efficiency and low energy consumption [16–18]. With the flourishing of adsorbents in the past decade, numerous tricky separations alkanes/alkenes [19–22], alkynes/alkenes [23–26], aromatics/cyclic aliphatics [27–30] and so on have been achieved. However, in contrast to the above gas mixtures, the carbon atoms in C₂F_6, CF₃CH₂F and CF₃CHF₂ are all sp³ hybridized, with fluorine atoms being the main component of molecules [31–34]. This means that the physico-chemical properties of the three gases are much closer. Presently, no clear separation mechanism has been developed for the separation of C₂F₆ from CF₃CH₂F/CF₃CHF₂/C₂F₆ mixture. Compared to C₂F₆, both CF₃CH₂F and CF₃CHF₂ have hydrogen atoms in their molecules. Owing to the electron-withdrawing effect of the F atom, the C—H bonds in CF₃CH₂F and CF₃CHF₂ are highly polarized and the H atoms are relatively active [35–37]. As a result, the H atoms in CF₃CH₂F and CF₃CHF₂ are good hydrogen-bond donors. As such, constructing adsorbents with high degree of hydrogen-bond acceptor sites should allow for selective adsorption of CF₃CH₂F and CF₃CHF₂ [38–41]. Firmly, in the previous reports, J. Milner, Motkuri and Su et al. have proven that halogen (X), O or N atoms in the frameworks can form X/O/N···H hydrogen bonds with the hydrogen atoms of hydrofluorocarbons to enhance their adsorption [42–44]. Thus, if we functionalize multiple hydrogen bond acceptors in the framework, the selectivities for CF₃CH₂F and CF₃CHF₂ can be similarly improved.

  As we know, an eligible adsorbent should not only have good gas selectivity, but also a high adsorption capacity, so as to achieve efficient separation performance [45–48]. However, adsorbents with narrow channels often possess high gas selectivity but often show low adsorption capacity because of the strict pore size limitations or immobilization of functional sites will drastically reduce their porosity. It is widely known that cage-like materials have high gas uptakes due to their extremely large cavities which can accommodate many gas molecules [49–53]. Consequently, if we can construct a porous material with both cage-like cavities and multiple hydrogen-bond acceptors, high CF₃CH₂F and CF₃CHF₂ adsorption capacity and high CF₃CH₂F/C₂F₆, CF₃CHF₂/C₂F₆ selectivity could be achieved simultaneously (Scheme 1).

  Scheme 1

  

  Scheme 1.  The strategy of immobilizing multiple hydrogen-bond acceptors as the adsorption sites in a cage-like framework to consturct the efficient CF₃CH₂F/CF₃CHF₂-selective adsorbent.

  DownLoad: Full-Size Img PowerPoint

  With the above consideration in mind, we herein report a cage-like metal–organic framework FJI-W20. On its pore surface, multiple hydrogen-bond acceptors such as F, O and N atoms that can provide tight C—H···F, C—H···O and C—H···N interactions with CF₃CH₂F and CF₃CHF₂ molecules are modified. As expected, FJI-W20 exhibits simultaneous preferential and high adsorption for CF₃CH₂F and CF₃CHF₂. Adsorption experiments show that it can absorb 132.9 cm³/g of CF₃CH₂F and 110.1 cm³/g of CF₃CHF₂ but can only absorb 57.9 cm³/g of C₂F₆ at 298 K and 1 bar. Due to the large difference in their uptakes, FJI-W20 delivers selectivity as high as 17.6 and 12.3 for respective CF₃CH₂F/C₂F₆ (5/95, v/v) and CF₃CHF₂/C₂F₆ (5/95, v/v). Breakthrough experiments show that high-purity C₂F₆ can not only be obtained from the binary CF₃CH₂F/C₂F₆ (5/95, v/v) and CF₃CHF₂/C₂F₆ (5/95, v/v) mixtures but also can be directly collected from ternary CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) mixture. After one breakthrough experiment at 298 K and 1 bar, the productivity of C₂F₆ can reach 20.6 mol/kg. Furthermore, FJI-W20 can maintain its good separation performance at different gas flow rates and temperatures. Computational simulation reveals that multiple electronegative atoms on the pore wall of FJI-W20 can generate substantial hydrogen bonding interactions with CF₃CH₂F and CF₃CHF₂, which facilitates the FJI-W20 to achieve an eye-catching CF₃CH₂F/CF₃CHF₂/C₂F₆ separation performance.

  FJI-W20 can be afforded as dark orange block-shaped crystals by solvothermal method (Fig. S1 in Supporting information). Single-crystal X-ray experiments reveal that it crystallizes in the I4₁/amd space group (Table S1 in Supporting information). As depicted in Fig. 1a, FJI-W20 is constructed by three components, i.e., BDC—CF₃²⁻ anion, BIPYDZ ligand and [Co₈(μ_3-OH)₄(H₂O)₄(SO₄)₂(COO)₈] cluster (Co₈ cluster). In the Co₈ cluster unit of, Co₁ atom is six-coordinated by two carboxylate oxygen atoms from two BDC—CF₃²⁻ ligands, two oxygen atoms of two μ_3-OH groups, one oxygen atom from SO₄²⁻, one water oxygen atom and one nitrogen atom from BIPYDZ, leading to octahedral geometry [54]. Analogously, the coordination mode of Co₂ atom in this Co₈ cluster is also hexa-coordination, which is completed by two carboxylate oxygen atoms from two BDC—CF₃²⁻ ligands, two oxygen atoms of μ_3-OH groups and two oxygen atoms of SO₄²⁻ anion, respectively. In the structure of FJI-W20, each Co₈ cluster can bind to twelve linker ligands (eight BDC—CF₃²⁻ ligands and four BIPYDZ) (Fig. 1e). This unusual Co₈-based cluster node, in conjunction with two linear ligands BDC—CF₃²⁻ and BIPYDZ, jointly constructs a 3D framework with a unique cage-like structure (Fig. 1b). The solvent accessible volume in fully evacuated FJI-W20 is up to 59.2% [55]. In FJI-W20, there are two kinds of polyhedral cages, that is, small tetrahedral cage (Cage-A) which is completed by three Co₈ clusters, four BDC—CF₃²⁻ ligands and one BIPYDZ ligand and big octahedral cage (Cage-B) which is completed by eight Co₈ clusters, eight BDC—CF₃²⁻ ligands and four BIPYDZ ligands (Figs. 1c and d). Especially, in simultaneous cage-A and cage-B, there are multiple electronegative atoms distributed, such as the F atoms, N atoms and O atoms, which are potential strong hydrogen bond receptors. This indicates that FJI-W20 can provide strong interactions with the gas molecules with the partially positive hydrogen atoms, such as CF₃CH₂F and CF₃CHF₂, thereby increasing the selectivity of CF₃CH₂F/C₂F₆ and CF₃CHF₂/C₂F₆.

  Figure 1

  

  Figure 1.  Single-crystal X-ray structure of FJI-W20. (a) The subunits of FJI-W20. (b) Cage-like structure of FJI-W20. Solvent molecules and hydrogen atoms are omitted for clarity. (c) The structure of tetrahedron cage-A. (d) The structure of octahedron cage-B. (e) The 12-connected Co₈ cluster.

  DownLoad: Full-Size Img PowerPoint

  The PXRD patterns of the as-synthesized FJI-W20 sample are well consistent with the calculated ones from the single crystal, indicating that the prepared FJI-W20 sample has high phase purity (Fig. S2 in Supporting information). Furthermore, FJI-W20 possesses excellent stability in air and common organic solvents, which can be certificated by the well matched PXRD patterns with the fresh sample after imposing FJI-W20 in air for 30 days or various organic solvents for 7 days (Figs. S2 and S4 in Supporting information). What is more, FJI-W20 also demonstrates good thermal stability. The thermogravimetric analysis experiment (TGA) shows that there is no obvious weight loss up to 290 ℃ (Fig. S3 in Supporting information). Variable-temperature PXRD (VT-PXRD) were also carried out and the results show that no significant changes in the PXRD patterns are observed even at 200 ℃ in air atmosphere, implying that the framework can be maintained at such high temperature (Fig. S5 in Supporting information). On the whole, the excellent structural stability endows it with great Given the exceptional stability and large pore of FJI-W20, we next tested its gas adsorption properties. The 77 K N₂ was firstly employed to illustrate its porosity. As revealed in Fig. S6 (Supporting information), it can absorb 419.9 cm³/g of N₂ at 77 K and 1 bar. The calculated BET surface area is up to 1640.4 m²/g and the pore volume is as high as 0.65 cm³/g, which indicates that FJI-W20 can be used as a promising platform for gas adsorption. Furthermore, owing to the cage-like structure and multiple hydrogen bond acceptors on its pore surface, it can be reasonably inferred that FJI-W20 may exhibits great potential for absorbing gas molecules that possess active hydrogen atoms.

  Based on this inference, we immediately afterwards carried out the adsorption experiments for CF₃CH₂F, CF₃CHF₂ and C₂F₆. As anticipated, FJI-W20 showcases simultaneous high uptakes for CF₃CH₂F and CF₃CHF₂ in the wide temperature range. As shown in Figs. 2a–c, at 273 K and 100 kPa, the adsorption amounts for CF₃CH₂F and CF₃CHF₂ are up to 150.0 and 128.0 cm³/g, respectively. When elevating the temperature to 298 K, the saturated uptakes of CF₃CH₂F and CF₃CHF₂ are 132.9 and 110.1 cm³/g. And the adsorption capacity remained almost perfectly consistent in the five consecutive adsorption cycles (Figs. S7–S9 in Supporting information). Notably, even the temperature is increased to 313 K, FJI-W20 can still absorb 124.9 cm³/g of CF₃CH₂F and 101.5 cm³/g of CF₃CHF₂. These results demonstrate that the cage based structure of FJI-W20 promotes its gas absorption capacity, which is unmatched by narrow pore materials such as Ni(pba)₂ that we previously reported [11]. Furthermore, comparing with the high storage capacity of CF₃CH₂F and CF₃CHF₂, FJI-W20 shows relatively low uptake for C₂F₆. The saturated uptakes of C₂F₆ by FJI-W20 at 273, 298 and 313 K are only 87.1, 57.9 and 41.6 cm³/g, which is much lower than that of CF₃CHF₂ and CF₃CH₂F. More importantly, at low pressure region, the adsorption curves of C₂F₆ increase particularly gently, reflecting that the framework can hardly generate strong interaction with C₂F₆ [56–58]. Evidently, the adsorption results reveal that FJI-W20 shows great difference between fluoroethane such as CF₃CH₂F, CF₃CHF₂ and C₂F₆. And the apparently strong affinity for CF₃CH₂F and CF₃CHF₂ relative to C₂F₆ suggests that multiple hydrogen bond receptors in FJI-W20 may favour binding of hydrofluorocarbons.

  Figure 2

  

  Figure 2.  Single-component adsorption isotherms of CF₃CH₂F, CF₃CHF₂ and C₂F₆ at (a) 273 K, (b) 298 K and (c) 313 K. (d) The Q_st for CF₃CH₂F, CF₃CHF₂ and C₂F₆ respectively. (e) IAST selectivity of FJI-W20 for CF₃CH₂F/C₂F₆ (5/95) and CF₃CHF₂/C₂F₆ (5/95). (f) Comparison of CF₃CH₂F/C₂F₆ (5/95) and CF₃CHF₂/C₂F₆ (5/95) selectivity of FJI-W20 and other materials.

  DownLoad: Full-Size Img PowerPoint

  In order to visually illustrate that FJI-W20 exhibits stronger interactions on CF₃CH₂F, CF₃CHF₂ than C₂F₆, we calculated the adsorption heats (Q_st) for CF₃CH₂F, CF₃CHF₂ and C₂F₆ with the adsorption isotherms at 273 and 298 K. As demonstrated in Fig. 2d and Figs. S13-S15 (Supporting information), the Q_st values at zero loading for CF₃CH₂F, CF₃CHF₂ are 43.6 and 42.2 kJ/mol, respectively. In contrast, the corresponding Q_st for C₂F₆ is 23.5 kJ/mol, indicating its lower affinity relative to CF₃CH₂F, CF₃CHF₂. This difference may be attributed to the presence of electronegative atoms in the structure of FJI-W20, which enhances the framework-hydrofluorocarbon interactions. Given the remarkable affinity for CF₃CH₂F and CF₃CHF₂, we judge that it has the ability to purify C₂F₆ from CF₃CH₂F/CF₃CHF₂/C₂F₆ mixture. To validate this inference, ideal adsorbed solution theory (IAST) has been employed to determine the IAST selectivity (Figs. S10–S12 in Supporting information). As shown in Figs. 2e and f, the IAST selectivities for CF₃CH₂F/C₂F₆ (5/95) and CF₃CHF₂/C₂F₆ (5/95) can reach 17.6 and 12.3 at 298 K and 1 bar, which are much higher than other porous materials such as some commercial molecular sieves (12.7/6.3 for β molecular sieve and 13.0/7.1 for TS-1) and the cage-like porous materials without recognized sites (8.6/3.9 for HKUST-1, 8.4/4.8 for UiO-66 and 10.4/4.8 for FJI-H8) (Figs. S16–S37 and Table S2 in Supporting information). And this result sufficiently proves that FJI-W20 can be used as an efficient adsorbent to separate CF₃CH₂F/C₂F₆ and CF₃CHF₂/C₂F₆ and completely indicates that the immobilization of multiple hydrogen-bond acceptor sites in a cage-like metal-organic framework is a verifiable strategy for efficient C₂F₆ purification.

  In order to further study the mechanism of FJI-W20 preferentially adsorbing CF₃CH₂F and CF₃CHF₂, theoretical calculations were carried out. As we know, compared with C₂F₆, CF₃CH₂F and CF₃CHF₂ contain not only F atoms, but also partially positive H atoms. The presence of H atoms enables CF₃CH₂F and CF₃CHF₂ to have more interactions with the skeleton. Therefore, the Q_st values of CF₃CH₂F and CF₃CHF₂ in FJI-W20 are higher than C₂F₆. According to Figs. 3a and b, the calculated results clearly reveal that CF₃CH₂F and CF₃CHF₂ molecules are tightly bound in the corners of the polyhedral hole that possessing abundant F/O/N environment. Specifically shown in Fig. 3a, CF₃CH₂F possesses eight interactions with FJI-W20, including four C—H···F hydrogen bonds (D_H···F = 2.69, 2.83, 3.27 and 3.30 Å), three C—H···π interactions (D_C···C = 3.58, 3.89 and 3.98 Å) and one C—H···O interactions (D_H···O = 3.60 Å) [59–63]. As depicted in Fig. 3b, each CF₃CHF₂ molecule interacts with the framework through two C—H···F (D_H···F = 2.96 and 3.95 Å), one C—H···N (D_H···N = 3.08 Å), one C—H···O (D_H···O = 3.78 Å) and three C—H···π (D_C···C = 3.66, 3.76 and 4.07 Å) supramolecular interactions. These evidences validate that the electronegative sites on the pores are critical for capturing CF₃CH₂F and CF₃CHF₂ molecules [64–67]. As a sharp contrast, there are complete F atoms in C₂F₆ molecule and it can only interact with the framework by C—H···F hydrogen bonds (D_H···F = 2.67, 2.79, 2.82, 3.42 and 3.48 Å) (Fig. 3c). Correspondingly, the calculated static binding energies of CF₃CH₂F, CF₃CH₂F and C₂F₆ are 52.12, 47.71 and 32.09 kJ/mol, respectively, which is in well agreement with the experimental trend of Q_st under zero coverage. These simulation results demonstrate the rationality of FJI-W20 as a purified C₂F₆ adsorbent.

  Figure 3

  

  Figure 3.  The calculated preferential adsorption sites for (a) CF₃CH₂F, (b) CF₃CHF₂ and (c) C₂F₆ in FJI-W20. The C—H···F, C—H···O and C—H···N hydrogen bonds are respectively shown in dashed blue, dashed green and dashed orange lines. The C—H···π interactions between gas molecule and the pore surface are shown in dashed red lines.

  DownLoad: Full-Size Img PowerPoint

  Based on the above compelling rationales, we next conducted dynamic breakthrough experiments to investigate the practical performance of FJI-W20 for the purification of C₂F₆. Firstly, the binary gas mixture of CF₃CH₂F/C₂F₆ and CF₃CHF₂/C₂F₆ were employed to evaluate the C₂F₆ purification potentiality. As shown in Figs. 4a and b, FJI-W20 can effectively purify C₂F₆ from CF₃CH₂F/C₂F₆ and CF₃CHF₂/C₂F₆ respectively. In Fig. 4a, it can be seen that the C₂F₆ effluxes rapidly at 30.9 min whereas the signal of CF₃CH₂F emerges after 391.2 min. High-purity C₂F₆ (≥99.99%) can be collected continuously during 360.3 min with an ultrahigh yield of 35.7 mol/kg. Similarly, FJI-W20 can also achieve efficient CF₃CHF₂/C₂F₆ separation (Fig. 4b). At 298 K and with the flow rate of 2.2 mL/min, high purity of C₂F₆ can be ceaselessly obtained from 30.9 min to 297.8 min and the corresponding C₂F₆ yield is as high as 25.9 mol/kg. The purification yield of C₂F₆ from CF₃CHF₂/C₂F₆ is slightly lower than from the CF₃CH₂F/C₂F₆ and this is owing to the lower adsorption capacity of FJI-W20 for CF₃CHF₂.

  Figure 4

  

  Figure 4.  The breakthrough experiments of (a) CF₃CH₂F/C₂F₆ (5/95, v/v), (b) CF₃CHF₂/C₂F₆ (5/95, v/v) and (c) CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at 298 K. (d) Six-cycle C₂F₆ yields for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at 298 K with gas flow rate of 1.7 mL/min. (e) The breakthrough experiments of CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) with the different flow rates. (f) C₂F₆ productivity for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) under different flow rates. (g) The breakthrough experiments of CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at differrent temperatures. (h) C₂F₆ productivity for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at different temperatures. (i) Breakthrough experiment for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) after exposure to air for 7 days.

  DownLoad: Full-Size Img PowerPoint

  Motivated by the confirmed separation performance of binary mixtures for FJI-W20, we further assessed the C₂F₆ purification capacity from ternary C₂F₆-containing gas mixture of CF₃CH₂F/CF₃CHF₂/C₂F₆. As demonstrated in Fig. 4c, FJI-W20 can likewise realize high-efficiency separation of C₂F₆ from CF₃CH₂F/CF₃CHF₂/C₂F₆ in the breakthrough experiments, in which C₂F₆ first breaks the packed column, while CF₃CHF₂ and CF₃CH₂F are retained in adsorbent for a period of time. From the breakthrough curve, it can be seen that the sequence of gas outflow at the outlet is C₂F₆, CF₃CHF₂ and CF₃CH₂F, the result is coincident with the gas adsorption amounts. After one three-component separation experiment, about 20.6 mol/kg of high-purity C₂F₆ can be collected. Furthermore, six consecutive breakthrough experiments on FJI-W20 show that the breakthrough times and productivities of C₂F₆ express no obvious loss (Fig. 4d), indicating that FJI-W20 possesses excellent reusability in C₂F₆ purification process. Moreover, in consideration of the practical industrial situations, the adsorbent should keep good separation performance at different gas flow rates and high temperatures. Thus, in order to examine the separation capacity of FJI-W20 in near-actual conditions, we conducted the breakthrough experiments under different flow rates and high temperatures. As shown in Figs. 4e and f, FJI-W20 could maintain good C₂F₆ purification ability within the flow rates of 1.7–3.8 mL/min and the C₂F₆ productivities can even remain roughly unchanged, which implies that the separation performance of FJI-W20 could not be influenced by the flow rates. More than that, FJI-W20 as well works well under high temperatures. As depicted in Fig. 4g, when we gradually elevated the testing temperature, the separation time of C₂F₆ only shows slightly decrease. For instance, the temperature is increased to 318 K, the separation time of C₂F₆ is ca. 182.1 min and the corresponding C₂F₆ productivity is ca. 13.8 mol/kg (Fig. 4h). Remarkably, after exposure to air for 7 days, the separation performance is almost the same as that of fresh samples, which is rarely seen in previous reports (Fig. 4i). On the whole, above results suggest that FJI-W20 can realize good separation performance under the actual situations.

  In summary, herein we report a cage-based framework FJI-W20 whose pore surface is functionalized with multiple electronegative F, O and N atoms as hydrogen-bond acceptors. Cage-like pores and multiple hydrogen-bond sites between the framework and hydrofluoroethanes ensure selectively capture of CF₃CH₂F and CF₃CHF₂ with high uptakes. For example, at 298 K and 1 bar, FJI-W20 can absorb 132.9 cm³/g of CF₃CH₂F and 110.1 cm³/g of CF₃CHF₂ but only 57.9 cm³/g of C₂F₆. The selectivity of CF₃CH₂F/C₂F₆ (5/95, v/v) and CF₃CH₂F/C₂F₆ (5/95, v/v) can reach 17.6 and 12.3. Separation experiments indicate that FJI-W20 can purify C₂F₆ with high purity from ternary CF₃CH₂F/CF₃CHF₂/C₂F₆ mixtures. After one separation cycle, 20.6 mol/kg of high-purity C₂F₆ can be directly obtained from CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) mixture. Additionally, at different gas flow rates and temperatures high-purity C₂F₆ can also be obtained in one step. This work illustrates that our strategy of immobilizing multiple hydrogen-bond acceptor sites in a cage-like framework is highly successful and effective for designing novel porous materials of purifying of C₂F₆ from CF₃CH₂F and CF₃CHF₂.

  CRediT authorship contribution statement

  Yongqin Zhu: Writing – original draft, Software, Investigation, Formal analysis, Data curation. Zhenyu Ji: Writing – original draft, Conceptualization. Yunzhe Zhou: Project administration, Formal analysis. Mingyan Wu: Writing – review & editing, Funding acquisition.

  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 is supported by the National Natural Science Foundation of China (NSFC, No. 22271282) and the Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDB1170000). This work is also supported by the Self-deployment Project Research Program of Haixi Institutes, Chinese Academy of Sciences (No. CXZX-2022-JQ04) as well as Fujian Science & Technology Innovation Laboratory for Optoelectronic Information of China (No. 2021ZR120).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111478.

  
  
• 1. [1]

     A.J. Sicard, R.T. Baker, Chem. Rev. 120 (2020) 9164–9303. doi: 10.1021/acs.chemrev.9b00719

  2. [2]

     Z. Zhang, B. Tan, P. Wang, X. Cui, H. Xing, AlChE J. 66 (2020) e16236. doi: 10.1002/aic.16236

  3. [3]

     W.T. Tsai, H.P. Chen, W.Y. Hsien, J. Loss Prev. Process Ind. 15 (2002) 65–75. doi: 10.1016/S0950-4230(01)00067-5

  4. [4]

     J.R. Clark, J. Chem. Educ. 83 (2006) 857. doi: 10.1021/ed083p857

  5. [5]

     M.B. Chang, J.S. Chang, Ind. Eng. Chem. Res. 45 (2006) 4101–4109. doi: 10.1021/ie051227b

  6. [6]

     X. Zhang, Y.L. Zhao, X.Y. Li, et al., J. Am. Chem. Soc. 146 (2024) 19303–19309. doi: 10.1021/jacs.4c05075

  7. [7]

     H.Y. Ruan, X.Q. Wu, C.L. Liao, et al., Angew. Chem. Int. Ed. 64 (2025) e202419302. doi: 10.1002/anie.202419302

  8. [8]

     J. He, H. Meng, X. Wang, Y. Xu, L. Feng, Sens. Actuator. B: Chem. 418 (2024) 136335. doi: 10.1016/j.snb.2024.136335

  9. [9]

     J. Zhu, J. Hu, H. Xiao, et al., Sep. Purif. Technol. 331 (2024) 125614. doi: 10.1016/j.seppur.2023.125614

  10. [10]

      G. Xu, T. Ke, R. Fan, et al., Adv. Sci. 12 (2025) 2411083. doi: 10.1002/advs.202411083

  11. [11]

      Y. Zhu, Y. Zhou, Z. Ji, W. Zhang, M. Wu, Sep. Purif. Technol. 353 (2025) 128373. doi: 10.1016/j.seppur.2024.128373

  12. [12]

      Y.L. Zhao, X. Zhang, M.Z. Li, J.R. Li, Chem. Soc. Rev. 53 (2024) 2056–2098. doi: 10.1039/d3cs00285c

  13. [13]

      H. Wang, W.P. Lustig, J. Li, Chem. Soc. Rev. 47 (2018) 4729–4756. doi: 10.1039/c7cs00885f

  14. [14]

      D.K.J.A. Wanigarathna, J. Gao, B. Liu, Mater. Adv. 1 (2020) 310–320. doi: 10.1039/d0ma00083c

  15. [15]

      W. Xia, Z. Zhou, L. Sheng, et al., Sci. Bull. 70 (2024) 232–240.

  16. [16]

      M. Wu, Z. Ji, Y. Zhou, et al., Angew. Chem. Int. Ed. 63 (2024) e202319674. doi: 10.1002/anie.202319674

  17. [17]

      Z. Ji, Q. Li, Y. Zhou, et al., Angew. Chem. Int. Ed. 63 (2024) e202411175. doi: 10.1002/anie.202411175

  18. [18]

      Z. Ji, Y. Zhou, Y. Zhu, et al., ACS Mater. Lett. 7 (2025) 837–844. doi: 10.1021/acsmaterialslett.4c02383

  19. [19]

      Y. Zhou, C. Chen, R. Krishna, et al., Angew. Chem. Int. Ed. 62 (2023) e202305041. doi: 10.1002/anie.202305041

  20. [20]

      Z. Di, C. Liu, J. Pang, et al., Angew. Chem. Int. Ed. 61 (2022) e202210343. doi: 10.1002/anie.202210343

  21. [21]

      X. Huang, F. Chen, H. Sun, et al., J. Am. Chem. Soc. 146 (2024) 617–626. doi: 10.1021/jacs.3c10495

  22. [22]

      Y. Wang, Z.J. Jiang, D.R. Wang, W. Lu, D. Li, J. Am. Chem. Soc. 146 (2024) 6955–6961. doi: 10.1021/jacs.3c14610

  23. [23]

      M. Liu, X.W. Gu, C. Lin, et al., J. Mater. Chem. A 13 (2024) 1034–1041.

  24. [24]

      Y. He, R. Krishna, B. Chen, Energy Environ. Sci. 5 (2012) 9107–9120. doi: 10.1039/c2ee22858k

  25. [25]

      M.Y. Gao, A.A. Bezrukov, B.Q. Song, et al., J. Am. Chem. Soc. 145 (2023) 11837–11845. doi: 10.1021/jacs.3c03505

  26. [26]

      T. Ke, Q. Wang, J. Shen, et al., Angew. Chem. Int. Ed. 59 (2020) 12725–12730. doi: 10.1002/anie.202003421

  27. [27]

      C. Chen, H. Guan, H. Li, et al., Angew. Chem. Int. Ed. 61 (2022) e202201646. doi: 10.1002/anie.202201646

  28. [28]

      F. Xie, L. Chen, E.M.C. Morales, et al., Nat. Commun. 15 (2024) 2240. doi: 10.1038/s41467-024-46556-6

  29. [29]

      R.C. Zhao, L.H. Xie, X.M. Liu, et al., J. Am. Chem. Soc. 147 (2024) 2467–2475.

  30. [30]

      Y. Han, Y. Chen, Y. Ma, et al., Chem. 9 (2023) 3532–3543. doi: 10.1016/j.chempr.2023.07.016

  31. [31]

      W. Xia, Y. Yang, L. Sheng, et al., Sci. Adv. 10 (2024) eadj6473. doi: 10.1126/sciadv.adj6473

  32. [32]

      A.R. Ravishankara, S. Solomon, A.A. Turnipseed, R.F. Warren, Science 259 (1993) 194–199. doi: 10.1126/science.259.5092.194

  33. [33]

      S. Wang, L. Zhou, H. Qin, et al., Molecules 29 (2024) 1721. doi: 10.3390/molecules29081721

  34. [34]

      R.K. Motkuri, H.V.R. Annapureddy, M. Vijaykumar, et al., Nat. Commun. 5 (2014) 4368. doi: 10.1038/ncomms5368

  35. [35]

      T.-H. Chen, I. Popov, W. Kaveevivitchai, et al., Angew. Chem. Int. Ed. 54 (2015) 13902–13906. doi: 10.1002/anie.201505149

  36. [36]

      J.W. Huang, J.J. Peng, X. Wei, et al., AICHE J. 69 (2023) e18027. doi: 10.1002/aic.18027

  37. [37]

      J. Zheng, D. Barpaga, O.Y. Gutierrez, et al., ACS Appl. Energ. Mater. 1 (2018) 5853–5858. doi: 10.1021/acsaem.8b01282

  38. [38]

      Y. Wang, F. Zhang, Y. Yang, et al., J. Colloid Interface Sci. 687 (2025) 439–448. doi: 10.1016/j.jcis.2025.02.088

  39. [39]

      W. Gong, Y. Geng, P. Gao, et al., J. Am. Chem. Soc. 146 (2024) 21806–21814. doi: 10.1021/jacs.4c06046

  40. [40]

      M. Sadakiyo, T. Yamada, H. Kitagawa, J. Am. Chem. Soc. 133 (2011) 11050–11053. doi: 10.1021/ja203291n

  41. [41]

      H. Wang, Y. Zhao, F. Zhang, et al., Sci. Adv. 7, (2021) eabg0396. doi: 10.1126/sciadv.abg0396

  42. [42]

      X.H. Xiong, L. Song, W. Wang, et al., Adv. Sci. 11 (2024) 2308123. doi: 10.1002/advs.202308123

  43. [43]

      R.T. Jerozal, J. Kim, C. Ma, et al., J. Am. Chem. Soc. 147 (2025) 7127–7136. doi: 10.1021/jacs.5c00393

  44. [44]

      J. Zheng, D. Barpaga, B.A. Trump, et al., J. Am. Chem. Soc. 142 (2020) 3002–3012. doi: 10.1021/jacs.9b11963

  45. [45]

      M. Zheng, W. Xue, T. Yan, et al., Angew. Chem. Int. Ed. 63 (2024) e202401770. doi: 10.1002/anie.202401770

  46. [46]

      H. Li, C. Liu, C. Chen, et al., Angew. Chem. Int. Ed. 60 (2021) 7547–7552. doi: 10.1002/anie.202013988

  47. [47]

      Z. Di, C. Liu, J. Pang, C. Chen, et al., Angew. Chem. Int. Ed. 60 (2021) 10828–10832. doi: 10.1002/anie.202101907

  48. [48]

      Y. Zhang, Y. Han, B. Luan, et al., J. Am. Chem. Soc. 146 (2024) 17220–17229. doi: 10.1021/jacs.4c03442

  49. [49]

      Y. Jiang, L. Wang, J. Hu, et al., Adv. Mater. 36 (2024) 2311140. doi: 10.1002/adma.202311140

  50. [50]

      X.J. Xie, H. Zeng, Y.L. Huang, et al., Chem 11 (2025) 102339. doi: 10.1016/j.chempr.2024.10.006

  51. [51]

      X.H. Xiong, Z.W. Wei, W. Wang, L.L. Meng, C.Y. Su, J. Am. Chem. Soc. 145 (2023) 14354–14364. doi: 10.1021/jacs.3c03309

  52. [52]

      J. Cui, Z. Zhang, L. Yang, et al., Science 383 (2024) 179–183. doi: 10.1126/science.abn8418

  53. [53]

      Y.J. Tian, C. Deng, L. Zhao, et al., Nat. Chem. 17 (2025) 141–147. doi: 10.1038/s41557-024-01672-0

  54. [54]

      L. Hou, W.X. Zhang, J.P. Zhang, et al., Chem. Commun. 46 (2010) 6311–6313. doi: 10.1039/c0cc01106a

  55. [55]

      A.L. Spek, J. Appl. Crystallogr. 36 (2003) 7–13. doi: 10.1107/S0021889802022112

  56. [56]

      W. Fan, X. Wang, X. Zhang, et al., ACS Central Sci. 5 (2019) 1261–1268. doi: 10.1021/acscentsci.9b00423

  57. [57]

      Y. Zhou, C. Chen, Z. Ji, et al., ACS Mater. Lett. 6 (2024) 1388–1395. doi: 10.1021/acsmaterialslett.4c00111

  58. [58]

      Z.Y. Ji, Y. Fan, M. Wu, M. Hong, Chem. Commun. 57 (2021) 3785–3788. doi: 10.1039/d1cc00014d

  59. [59]

      T. Steiner, W. Saenger, J. Am. Chem. Soc. 114 (1992) 10146–10154. doi: 10.1021/ja00052a009

  60. [60]

      M.S. Thompson, T.J. Meyer, J. Am. Chem. Soc. 104 (1982) 5070–5076. doi: 10.1021/ja00383a013

  61. [61]

      D.Z. Veljkovic, G.V. Janjic and S.D. Zaric, CrystEngComm 13 (2011) 5005–5010. doi: 10.1039/c1ce05065f

  62. [62]

      G.R. Desiraju, Chem. Commun. (2005) 2995–3001. doi: 10.1039/b504372g

  63. [63]

      N.A. Porter, B. Giese, D.P. Curran, Acc. Chem. Res. 24 (1991) 296–304. doi: 10.1021/ar00010a003

  64. [64]

      W.T. Tsai, H.P. Chen, W.Y. Hsien, J. Loss Prev. Process Ind. 15 (2002) 65–75. doi: 10.1016/S0950-4230(01)00067-5

  65. [65]

      Z. Zhang, S. Zhang, X. Liu, et al., J. Am. Chem. Soc. 146 (2024) 31213–31220. doi: 10.1021/jacs.4c11470

  66. [66]

      T.H. Chen, I. Popov, W. Kaveevivitchai, et al., Nat. Commun. 5 (2014) 5131. doi: 10.1038/ncomms6131

  67. [67]

      W. Zhang, Y. Li, Y. Wu, et al., Sep. Purif. Technol. 287 (2022) 120561. doi: 10.1016/j.seppur.2022.120561

• 

  Scheme 1  The strategy of immobilizing multiple hydrogen-bond acceptors as the adsorption sites in a cage-like framework to consturct the efficient CF₃CH₂F/CF₃CHF₂-selective adsorbent.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Single-crystal X-ray structure of FJI-W20. (a) The subunits of FJI-W20. (b) Cage-like structure of FJI-W20. Solvent molecules and hydrogen atoms are omitted for clarity. (c) The structure of tetrahedron cage-A. (d) The structure of octahedron cage-B. (e) The 12-connected Co₈ cluster.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Single-component adsorption isotherms of CF₃CH₂F, CF₃CHF₂ and C₂F₆ at (a) 273 K, (b) 298 K and (c) 313 K. (d) The Q_st for CF₃CH₂F, CF₃CHF₂ and C₂F₆ respectively. (e) IAST selectivity of FJI-W20 for CF₃CH₂F/C₂F₆ (5/95) and CF₃CHF₂/C₂F₆ (5/95). (f) Comparison of CF₃CH₂F/C₂F₆ (5/95) and CF₃CHF₂/C₂F₆ (5/95) selectivity of FJI-W20 and other materials.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The calculated preferential adsorption sites for (a) CF₃CH₂F, (b) CF₃CHF₂ and (c) C₂F₆ in FJI-W20. The C—H···F, C—H···O and C—H···N hydrogen bonds are respectively shown in dashed blue, dashed green and dashed orange lines. The C—H···π interactions between gas molecule and the pore surface are shown in dashed red lines.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The breakthrough experiments of (a) CF₃CH₂F/C₂F₆ (5/95, v/v), (b) CF₃CHF₂/C₂F₆ (5/95, v/v) and (c) CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at 298 K. (d) Six-cycle C₂F₆ yields for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at 298 K with gas flow rate of 1.7 mL/min. (e) The breakthrough experiments of CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) with the different flow rates. (f) C₂F₆ productivity for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) under different flow rates. (g) The breakthrough experiments of CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at differrent temperatures. (h) C₂F₆ productivity for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) at different temperatures. (i) Breakthrough experiment for CF₃CH₂F/CF₃CHF₂/C₂F₆ (5/5/90, v/v/v) after exposure to air for 7 days.

  下载: 全尺寸图片 幻灯片
•