English

• Ethylene (C₂H₄), as one of the world's largest chemical products, is the core of the petrochemical industry. Currently, a majority of the world's C₂H₄ is produced by thermal cracking of ethane (C₂H₆) or steam cracking, and C₂H₆ is inevitably existed as a major byproduct [1]. The separation of C₂H₆ from C₂H₄ is thereby a critical process to yield polymer-grade C₂H₄ (> 99.9%) for polymer production [2]. The industrial separation of C₂H₄ from C₂H₆ typically relies on high-pressure cryogenic distillation at temperatures as low as −160 ℃ because of the similar sizes and volatilities between them, representing one of the most energy-intensive processes [3]. Hence, the energy-efficient separation technology for the production of polymer grade C₂H₄ is highly desired and challenging [4].

  Separation by means of adsorption using porous materials is recognized as a promising alternative because of low energy consumption and high efficiency [5-11]. Metal-organic frameworks (MOFs) are porous crystalline materials formed by self-assembly of inorganic metal centers with organic ligands. In comparison to traditional porous materials, MOFs are regarded as excellent C₂H₆/C₂H₄ separation and purification platforms owing to their ability to accurately adjust the framework structures and functionalize the pore surfaces. MOFs for C₂H₆/C₂H₄ separation were divided into C₂H₄-selective MOFs and C₂H₆-selective MOFs [12-17]. In general, the design of C₂H₄-selective MOFs is easier to implement through taking advantage of stronger interactions between unsaturated C₂H₄ molecules and high polar binding centers, such as open metal sites. However, to get C₂H₄ product, the next desorption process is necessary, which is still difficult to produce highly pure C₂H₄ due to co-adsorption of C₂H₆ in materials, so it needs the multiple adsorption-desorption cycles and is also energetically costly [18-23]. On the contrary, C₂H₆-selective MOFs can preferentially capture C₂H₆ impurity to directly produce C₂H₄ at the separation column outlet [24-28], which is more facile and economical process, and can also greatly reduce energy consumption [29]. But C₂H₆-selective MOFs usually exhibit relatively low loading or selectivity due to lacking strong binding sites. So it is imperative to develop new C₂H₆-selective MOFs for efficient C₂H₄ purity [30-35].

  The challenge of separating C₂H₆ and C₂H₄ stems from their very close molecular sizes and boiling points (Table S1 in Supporting information) [36]. Compared to C₂H₄, C₂H₆ has a smaller quadrupole moment (0.65 × 10⁻²⁶ esu cm² vs. 1.50 × 10⁻²⁶ esu cm²) and a larger polarizability (44.7 × 10⁻²⁵ cm³ vs. 42.52 × 10⁻²⁵ cm³), indicating that dispersion and induction interactions would make major contributions in C₂H₆-selective adsorbents [37]. C₂H₆-selective MOFs can be obtained by designing inert pore surfaces or implanting C₂H₆ affinity sites which will preferentially adsorb C₂H₆ over C₂H₄ [17,24,38-42]. For example, Chen et al. reported a microporous MOF, Fe₂(O₂)(dobdc), with iron (Fe)-peroxo sites for preferential binding of C₂H₆ over C₂H₄, showing a separation selectivity of up to 4.4 for C₂H₆/C₂H₄ [24]. By virtue of pore engineering, Li et al. successfully designed an inert ultramicroporous material (Cu(Qc)₂) with optimized pore structure, realizing excellent C₂H₆-selective adsorption behavior [39]. Although some C₂H₆-selective MOFs have been reported, there are still some deficiencies such as typical "trade-off" effect between selectivity and capacity, balancing these issues remains a constant challenge [43].

  The adsorption amounts and selectivity for gases in MOFs are crucial for separation, which are closely dependent on the pore environments of MOFs. Therefore, the regulation of pore walls through using functionalized ligands would provide an important approach to enhance the separation performance of MOFs for C₂H₆/C₂H₄ mixtures (Scheme 1). To target MOFs with the preferential binding of C₂H₆ over C₂H₄, we initially designed a C₂H₆-selective MOF [Cu_1.5(BTC)(DPU)_1.5(H₂O)_1.5] (Cu-MOF) from 1, 3, 5-benzenetricarboxylic acid (H₃BTC) and 1, 3-di(pyridin-4-yl)urea (DPU) linkers, which possesses abundant urea groups in pores and will provide stronger multiple interactions with C₂H₆ over C₂H₄. Then according to the isoreticular principle, the -NH₂ groups that could be the potential adsorption sites through hydrogen bonds with gas molecules were embedded into the pores through replacing H₃BTC with NH₂H₃BTC ligands, generating the isomorphic framework Cu-MOF(NH₂), [Cu_1.5(NH₂-BTC)(DPU)_1.5(H₂O)_1.5], with more accessible sites in pores to study the control of pore chemistry for advancing C₂H₆/C₂H₄ separation. It was found that due to the electronegative N/O sites and the coordination between water and metal centers hindering the open metal sites (OMSs), both MOFs exhibited C₂H₆-selective adsorption behaviors. In particular, the introduction of -NH₂ groups in pores provides additional binding sites to more enhance the binding affinity for C₂H₆, leading to increased C₂H₆ and C₂H₄ uptakes and C₂H₆/C₂H₄ selectivity (from 1.4 to 1.8) in Cu-MOF(NH₂) compared to Cu-MOF. The breakthrough experiments also showed that Cu-MOF(NH₂) can obtain highly pure C₂H₄ (≥99.99%) in one step with longer breakthrough interval times and higher C₂H₄ productivity.

  Scheme 1

  

  Scheme 1.  Strategy to boost C₂H₆/C₂H₄ separation through pore functionalization.

  DownLoad: Full-Size Img PowerPoint

  The synthesis details of Cu-MOF and Cu-MOF(NH₂) were given in Supporting information. Single-crystal X-ray diffraction analysis determined that two MOFs exhibited isomorphic network and crystallized in the same trigonal crystal system (Tables S2 and S3 in Supporting information). The crystal structure of NH₂-functionalized Cu-MOF(NH₂) was described representatively. The asymmetric unit contains half a Cu²⁺ ion, half a DPU ligand, two-thirds of deprotonated NH₂-BTC ligand, and half a H₂O coordinated molecule. The Cu²⁺ ion has a distorted square pyramidal geometry constituted by two carboxylate O atoms from two NH₂-BTC and two pyridine N atoms from two DPU on the base plane and one water O atom on the vertex (Fig. 1a). Six Cu²⁺ ions and six NH₂-BTC are alternately connected to form a 48-membered planar ring, which extends outward to form a layer with hexagonal windows (Fig. 1b). Furthermore, the linkages of adjacent layers with DPU as pillars to afford a (3, 4)-connected three-dimensional network, which with large windows is threaded by the other network to afford a two-folded interpenetrated framework. The upper and lower BTC—Cu coordinated motifs in an independent network are supported by three DPU pillars to form a triangular prism-like cage (cage Ⅰ), and which passes through one ring in the layer of the other network to form a cavity with six windows (Fig. 1c). Along the c axis, the other cage (cage Ⅱ) (Fig. 1d) is also formed between two independent networks on both sides of cage Ⅰ. These cavities are connected to form channels with abundant accessible N/O sites (Fig. 1e), in particular the modification of -NH₂ groups in Cu-MOF(NH₂) provides more potential sites in channels compared to Cu-MOF, which may play a positive role in enhancing the separation performance for C₂H₆/C₂H₄ mixtures.

  Figure 1

  

  Figure 1.  (a) Coordination environment of Cu²⁺ ion, (b) layer, (c) cage Ⅰ, (d) cage Ⅱ, and (e) 3D framework in Cu-MOF(NH₂).

  DownLoad: Full-Size Img PowerPoint

  Sample purity was verified by matched powder X-ray diffraction (PXRD) between the measured results of as-synthesized samples and the simulated patterns from single crystal structures (Figs. S1 and S2 in Supporting information). Thermogravimetric analysis (TGA) indicated the initial weight loss of lattice solvent molecules in two MOFs before 110 ℃. After a thermal stable platform, the skeleton began to decompose at 250 ℃ (Figs. S3 and S4 in Supporting information). Fourier transform infrared spectrum (FT-IR) peaks of Cu-MOF(NH₂) at 3306 and 3400 cm⁻¹ are the characteristic peaks of N—H vibration from -NH₂ groups, however, Cu-MOF only possessed one peak at 3307 cm⁻¹ (Figs. S5 and S6 in Supporting information). Meanwhile, Cu-MOF(NH₂) also showed the peak at 1379 cm⁻¹, resulted from the C—N bond between the benzene ring and the amino group (Fig. S6 in Supporting information) [44]. The frameworks of two MOFs were activated by heating at 110 ℃ under vacuum for 3 h for the acetone-exchanged (72 h) samples. It can be found that the coordinated water molecules which correspond to the initial weight loss in activated samples were held (Figs. S3 and S4 in Supporting information). The coordination water molecules exclude the exposed metal ions which are not favorable for C₂H₆ selectivity over C₂H₄.

  The identical frameworks but different potential adsorption sites between two MOFs provide good platforms for studying the influence of pore environments on gas separation. Two MOFs revealed type-I N₂ adsorption isotherms with the close loadings of about 280 cm³/g at 77 K and 100 kPa as well as the pore sizes of 6.5–8.5 Å (Figs. S7 and S8 in Supporting information). To evaluate the adjustment on C₂H₆/C₂H₄ separation by -NH₂ modification, single-component adsorption isotherms were measured. As shown in Figs. 2a and b, the uptakes of Cu-MOF for C₂H₄ and C₂H₆ at 100 kPa are 110.8 and 109.3 cm³/g at 273 K and 91.6 and 91.5 cm³/g at 298 K, respectively, however, those valves in Cu-MOF(NH₂) are increased to 119.6 and 117.6 cm³/g at 273 K and 98.8 and 101.3 cm³/g at 298 K, respectively. At the same time, the adsorption isotherms of Cu-MOF(NH₂) are moderately steeper than Cu-MOF, indicating that the pore modification by -NH₂ groups makes an increase in adsorption for two gases. Further, the isosteric heat of adsorption (Q_st) was calculated by fitting adsorption isotherms to virial equation, which revealed the affinity of MOFs toward adsorbates (Figs. S9-S12 in Supporting information). As shown in Figs. 2c and d, the initial Q_st values for C₂H₆ of 30.6 kJ/mol in Cu-MOF(NH₂) and 27.6 kJ/mol in Cu-MOF are higher than the corresponding values for C₂H₄ of 26.1 and 23.0 kJ/mol, indicating stronger interactions of two MOFs for C₂H₆. Meanwhile, the slopes of C₂H₆ isotherms are also significantly larger than those of C₂H₄. These results illustrated a stronger affinity of the framework for C₂H₆ than C₂H₄, supporting the C₂H₆-selective behavior in two MOFs. In addition, the adsorption cyclic tests of Cu-MOF(NH₂) showed repeated results with no decrease in adsorption capacity (Fig. S13 in Supporting information).

  Figure 2

  

  Figure 2.  Gas adsorption isotherms of Cu-MOF and Cu-MOF(NH₂) at (a) 273 and (b) 298 K. Q_st plots of (c) Cu-MOF(NH₂) and (d) Cu-MOF.

  DownLoad: Full-Size Img PowerPoint

  After surveying the improvement of C₂H₆-selective adsorption and higher C₂H₆-binding affinity in Cu-MOF(NH₂), ideal adsorbed solution theory (IAST) was utilized to calculate the adsorption selectivity of Cu-MOF(NH₂) for 1/1, 1/9, and 1/15 of C₂H₆/C₂H₄ mixtures and compared to those of Cu-MOF (Figs. S14-S17 in Supporting information). As shown in Fig. 3a, Cu-MOF(NH₂) exhibits the C₂H₆/C₂H₄ selectivity of about 1.8 for equimolar mixtures at 298 K and 100 kPa, which is significantly higher than that of Cu-MOF (1.4), and is equivalent to some top-performing C₂H₆-selective materials, such as CPM-733 (1.75) [45], ZIF-8 (1.7) [46], PCN-250 (1.9) [47], and Zn-atz-ipa (1.7) [40], but greatly surpasses some excellent MOFs, including TJT-100(1.2) [30], Azole-Th-1 (1.46) [25], UPC-612 (1.4) [26], and Ni(bdc)(ted)_0.5 (1.6) [48]. We used the separation potential (∆q) as a further evaluation and screening metric of MOFs [49]. As shown in Fig. 3b, Figs. S18 and S19 (Supporting information), the pure amounts of C₂H₄ recovered from Cu-MOF are 0.66, 1.40, and 1.49 mmol/g for the 1/1, 1/9, and 1/15 mixtures, respectively, however which are greatly increased to 1.24, 2.95, and 3.17 mmol/g for Cu-MOF(NH₂). In addition, taking the 1/15 mixtures as an example, the ∆q of Cu-MOF(NH₂) are higher than many benchmark C₂H₆-selective materials, including MUF-15 (2.97 mmolg) [31], UiO-67-(NH₂)₂ (2.51 mmol/g) [50], and Cu(Qc)₂ (1.73 mmol/g) [39], thus outperforming many reported adsorbents [17,42,47].

  Figure 3

  

  Figure 3.  (a) IAST selectivity of Cu-MOF and Cu-MOF(NH₂) for C₂H₆/C₂H₄ mixtures. (b) Separation potential for C₂H₆/C₂H₄ mixtures (1/15) in different materials. (c) Comparison of C₂H₄ and C₂H₆ uptakes in different materials. (d) Comparison of C₂H₆ uptakes and C₂H₆/C₂H₄ selectivity in different materials.

  DownLoad: Full-Size Img PowerPoint

  Adsorption amount is as equally important as the selectivity, however, due to frequently existed "trade-off" effect between selectivity and adsorption amount for separation in MOFs, it is challenging to simultaneously achieve both high C₂H₆/C₂H₄ selectivity and high C₂H₆ loading in one MOF. Although the C₂H₆ uptake of Cu-MOF(NH₂) (101.3 cm³/g) is not as good as several prominent MOFs, such as CPM-733 (159.6 cm³/g) [45], Ni-MOF 2 (133 cm³/g) [35], PCN-250 (116.7 cm³/g) [47], and ZJU-120a (110 cm³/g) [33], based on different pore environments and framework topologies, but is greatly higher than most reported MOFs, such as MAF-49 (38.8 cm³/g) [38], Cu(Qc)₂ (41.5 cm³/g) [39], ZIF-8 (56 cm³/g) [46], Fe₂(O₂)(dobdc) (74.3 cm³/g) [24], and UPC-612 (80.1 cm³/g) (Fig. 3c) [26]. Taken together, Cu-MOF(NH₂) not only exhibits good C₂H₆/C₂H₄ selectivity but also excellent C₂H₆ uptake, signifying the potential as an efficient material for one-step purity of C₂H₄ from C₂H₆/C₂H₄ mixtures (Fig. 3d).

  To further evaluate the effect of -NH₂ groups on C₂H₆/C₂H₄ separation performance, dynamic breakthrough experiments were conducted in a packed column filled with activated MOFs for C₂H₆/C₂H₄/Ar (5/5/90, 1/9/90, and 1/15/84, v/v/v, flow rate = 7.0 mL/min, Ar as carrier gas) mixtures at 298 K and 100 kPa, respectively. As shown in Fig. 4, two MOFs can separate C₂H₆/C₂H₄ mixtures, whereby Cu-MOF(NH₂) presents a considerably superior performance compared to Cu-MOF, aligning with the predicted results based on gas adsorption amount and IAST selectivity. As shown in Fig. 4a, C₂H₆ can be effectively separated from C₂H₆/C₂H₄ (50/50) mixtures in Cu-MOF, in which a high-purity C₂H₄ (≥99.9%) of 3.77 L/kg productivity is obtained at the outlet. However, Cu-MOF is failed to efficiently eliminate C₂H₆ from 1/9 and 1/15 C₂H₆/C₂H₄ mixtures because of low selectivity for C₂H₆ over C₂H₄, which leads to negligible C₂H₄ productivity and short separation time (Figs. 4b and c). By contrast, for Cu-MOF(NH₂), the ≥99.99% purity of C₂H₄ with 5.81, 8.33, and 30.02 L/kg productivity can be directly recovered from the 1/1, 1/9, and 1/15 C₂H₆/C₂H₄ mixtures in one cycle, respectively. The higher C₂H₄ productivity and longer breakthrough interval times determined by breakthrough experiments indicated that the immobilization of -NH₂ groups in MOF successfully optimized pore confinement, which associated with additional binding sites significantly improved the separation of material for C₂H₆/C₂H₄ mixtures (Fig. 4d). Considering better separation performance of Cu-MOF(NH₂), the breakthrough measurements at different flow rates of C₂H₆/C₂H₄ mixtures (1/15, v/v) were further conducted at 298 K in Cu-MOF(NH₂), it also indicated the complete separation of C₂H₆/C₂H₄ mixtures (Fig. 5a). Notably, as shown in Fig. 5b, Cu-MOF(NH₂) can also be easily regenerated by purging with Ar at 323 K, in which the adsorbed C₂H₄ due to weaker binding affinity is desorbed more quickly than C₂H₆. The breakthrough experiment cycles were performed on equimolar C₂H₆/C₂H₄ mixtures to investigate the reproducibility and recyclability of Cu-MOF(NH₂), which showed no decrease in separation performance (Fig. 5c). In addition, the PXRD measurements revealed that the framework treated with different environments retained structural integrity with no phase change and loss of crystallinity observed (Fig. 5d).

  Figure 4

  

  Figure 4.  Breakthrough curves of MOFs for C₂H₆/C₂H₄ mixtures at 298 K, (a) 1/1, (b) 1/9, and (c) 1/15. (d) Comparison of the comprehensive separation performance for Cu-MOF(NH₂) and Cu-MOF.

  DownLoad: Full-Size Img PowerPoint

  Figure 5

  

  Figure 5.  (a) Breakthrough curves of Cu-MOF(NH₂) for C₂H₆/C₂H₄ (v/v, 1/15) mixtures at different total flow rates at 298 K. (b) Desorption curves of Cu-MOF(NH₂) for C₂H₆/C₂H₄ (v/v, 1/15) mixtures under Ar (7 mL/min) weeping at 323 K. (c) Breakthrough cycles of Cu-MOF(NH₂) for C₂H₆/C₂H₄ mixtures (v/v, 5/5) at 298 K. (d) PXRD patterns of Cu-MOF(NH₂) treated with different environments.

  DownLoad: Full-Size Img PowerPoint

  To in-depth elucidate the origin of C₂H₆-selective behavior enhanced by -NH₂ functionalization, grand canonical Monte Carlo (GCMC) simulations were done to investigate the interactions between the frameworks and gas molecules. As shown in Figs. 6a and b, in Cu-MOF the preferential adsorption sites for C₂H₄ and C₂H₆ are located in cages Ⅰ and Ⅱ, respectively. The C₂H₄ molecule interacts with one pyridine ring and one urea N atom from DPU, and one carboxylate O atom from BTC, through C—H···π and C—H···N/O interactions. By contrast, the C₂H₆ molecule is resided between two BTC linkers and involved in more hydrogen bonds and C—H···π interactions with the phenyl rings. For Cu-MOF(NH₂), the preferential adsorption sites for C₂H₆ and C₂H₄ are both located in cage Ⅱ (Figs. 6c and d). The C₂H₄ molecule forms two C—H···N hydrogen bonds and two C—H···π interactions with the -NH₂ groups and phenyl rings in the upper and lower NH₂-BTC ligands, while the C₂H₆ molecule interacts with two phenyl rings from two NH₂-BTC through forming four C—H···C and one C—H···π interactions with the distances of 2.915–3.041 Å, as well as two stronger C—H···N (H···N = 2.939 and 2.999 Å) interactions compared to C₂H₄ (H···N = 2.996 and 3.187 Å). Therefore, the introduction of -NH₂ groups not only changes the position of preferential adsorption sites, but also makes C₂H₄ and C₂H₆ molecules more closely contact with the pore walls. In particular, there obviously exist more supramolecular contacts between C₂H₆ molecule and pore walls in Cu-MOF(NH₂). Meanwhile, the calculated binding energies for C₂H₆ and C₂H₄ in Cu-MOF(NH₂) (42.5 kJ/mol vs. 32.3 kJ/mol) also showed the higher values compared to Cu-MOF (37.5 kJ/mol vs. 29.4 kJ/mol). These findings are consistent with the better C₂H₆-selectivity for Cu-MOF(NH₂) than Cu-MOF, as found in experimental results.

  Figure 6

  

  Figure 6.  (a, b) C₂H₄ and (c, d) C₂H₆ preferential adsorption sites in Cu-MOF and Cu-MOF(NH₂).

  DownLoad: Full-Size Img PowerPoint

  Furthermore, the more different binding sites were further explored by GCMC simulations at 298 K and 100 kPa to evaluate the binding of C₂H₄ and C₂H₆ molecules in Cu-MOF (Fig. 7) and Cu-MOF(NH₂) (Fig. 8). The detailed data on these interactions was listed in Table S4. The results revealed that the adsorption sites are mainly located in cavities for two MOFs through forming C—H···C/N/O/π interactions. Comparing the interactions between the framework and gas molecules: C₂H₄ in Figs. 7a and b vs. C₂H₆ in Figs. 7c and d for Cu-MOF; C₂H₄ in Figs. 8a and b vs. C₂H₆ in Figs. 8c and d for Cu-MOF(NH₂), it clearly revealed that the interactions between the framework and C₂H₆ are more and stronger than C₂H₄ in two MOFs. In particular, the -NH₂ groups in Cu-MOF(NH₂) make both C₂H₆ and C₂H₄ molecules have closer contacts with the framework, leading to increased C—H···N interactions as a result, and which is more evident for C₂H₆ in pores, agreeing well with the experimental findings of more C₂H₆ capture and higher C₂H₆/C₂H₄ selectivity in Cu-MOF(NH₂) relative to Cu-MOF.

  Figure 7

  

  Figure 7.  (a, b) C₂H₄ and (c, d) C₂H₆ adsorption sites at 298 K and 100 kPa in Cu-MOF.

  DownLoad: Full-Size Img PowerPoint

  Figure 8

  

  Figure 8.  (a, b) C₂H₄ and (c, d) C₂H₆ adsorption sites at 298 K and 100 kPa in Cu-MOF(NH₂).

  DownLoad: Full-Size Img PowerPoint

  In summary, based on pore decoration strategy, we have effectively tuned the pore environment through installing -NH₂ functionalized groups and obtained an isomorphic Cu-MOF(NH₂) for targeting highly efficient one-step C₂H₄ purification from C₂H₆/C₂H₄ mixtures. The -NH₂ groups enhanced the interactions of the framework with C₂H₆ molecules, which preferentially adsorbed more C₂H₆ from C₂H₆/C₂H₄ mixtures. Cu-MOF(NH₂) with high C₂H₆ uptakes and significant C₂H₆/C₂H₄ selectivity effectively reduced "trade-off" effect in comparison to Cu-MOF. As a result, Cu-MOF(NH₂) produced polymer-grade C₂H₄ (≥99.99%) in one-step separation with a high productivity of 30.02 L/kg, which was 3.3 times higher than that of C₂H₄ produced by Cu-MOF, and a significantly higher purity than Cu-MOF. The utilization of this pore engineering approach will aid in the development and utilization of MOF materials for the challenging separation of critical chemicals.

  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 National Natural Science Foundation of China (Nos. 22371226 and 22371225) and Natural Science Basic Research Program of Shaanxi (No. 2024JC-JCQN-18).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109729.

  
  
• 1. [1]

     S.M. Sadrameli, Fuel 173 (2016) 285–297. doi: 10.1016/j.fuel.2016.01.047

  2. [2]

     R. Sahoo, M.C. Das, Coord. Chem. Rev. 442 (2021) 213998. doi: 10.1016/j.ccr.2021.213998

  3. [3]

     D.S. Sholl, R.P. Lively, Nature 532 (2016) 435–437. doi: 10.1038/532435a

  4. [4]

     S. Chu, Y. Cui, N. Liu, Nat. Mater. 16 (2016) 16–22. doi: 10.1038/nmat4834

  5. [5]

     J.Y. Li, Y. He, Y.C. Zou, et al., Chin. Chem. Lett. 33 (2022) 3017–3020. doi: 10.1016/j.cclet.2021.12.011

  6. [6]

     P.Q. Liao, N.Y. Huang, W.X. Zhang, et al., Science 356 (2017) 1193–1196. doi: 10.1126/science.aam7232

  7. [7]

     H. Zeng, M. Xie, T. Wang, et al., Nature 595 (2021) 542–548. doi: 10.1038/s41586-021-03627-8

  8. [8]

     X.B. Mu, Y.Y. Xue, M.C. Hu, et al., Chin. Chem. Lett. 34 (2023), 107296. doi: 10.1016/j.cclet.2022.03.019

  9. [9]

     S. Zhou, O. Shekhah, A. Ramírez, et al., Nature 606 (2022) 706–712. doi: 10.1038/s41586-022-04763-5

  10. [10]

      F. Xie, H. Wang, J. Li, J. Mater. Chem. A 11 (2023) 12425–12433. doi: 10.1039/d2ta09326j

  11. [11]

      F. Zhang, H. Shang, B. Zhai, et al., Angew. Chem. Int. Ed. 62 (2023) e202316149. doi: 10.1002/anie.202316149

  12. [12]

      M. -H. Yu, H. Fang, H. -L. Huang, et al., Small 19 (2023) 2300821. doi: 10.1002/smll.202300821

  13. [13]

      S. Mukherjee, D. Sensharma, K.J. Chen, et al., Chem. Commun. 56 (2020) 10419–10441. doi: 10.1039/d0cc04645k

  14. [14]

      C.X. Chen, Z.W. Wei, T. Pham, et al., Angew. Chem. Int. Ed. 60 (2021) 9680–9685. doi: 10.1002/anie.202100114

  15. [15]

      R.B. Lin, Z. Zhang, B. Chen, Acc. Chem. Res. 54 (2021) 3362–3376. doi: 10.1021/acs.accounts.1c00328

  16. [16]

      H. Wang, D. Luo, E. Velasco, et al., J. Mater. Chem. A 9 (2021) 20874–20896. doi: 10.1039/d1ta04096k

  17. [17]

      G.D. Wang, R. Krishna, Y.Z. Li, et al., Angew. Chem. Int. Ed. 61 (2022) e202213015. doi: 10.1002/anie.202213015

  18. [18]

      S. Aguado, G. Bergeret, C. Daniel, et al., J. Am. Chem. Soc. 134 (2012) 14635–14637. doi: 10.1021/ja305663k

  19. [19]

      B. Li, Y. Zhang, R. Krishna, et al., J. Am. Chem. Soc. 136 (2014) 8654–8660. doi: 10.1021/ja502119z

  20. [20]

      J.E. Bachman, M.T. Kapelewski, D.A. Reed, et al., J. Am. Chem. Soc. 139 (2017) 15363–15370. doi: 10.1021/jacs.7b06397

  21. [21]

      Z. Bao, J. Wang, Z. Zhang, et al., Angew. Chem. Int. Ed. 57 (2018) 16020–16025. doi: 10.1002/anie.201808716

  22. [22]

      R.B. Lin, L. Li, H.L. Zhou, et al., Nat. Mater. 17 (2018) 1128–1133. doi: 10.1038/s41563-018-0206-2

  23. [23]

      L. Zhang, L. Li, E. Hu, et al., Adv. Sci. 7 (2019) 1901918. doi: 10.1002/advs.201901918

  24. [24]

      L. Li, R.B. Lin, R. Krishna, et al., Science 362 (2018) 443–446. doi: 10.1126/science.aat0586

  25. [25]

      Z. Xu, X. Xiong, J. Xiong, et al., Nat. Commun. 11 (2020) 3163. doi: 10.1038/s41467-020-16960-9

  26. [26]

      Y. Wang, C. Hao, W. Fan, et al., Angew. Chem. Int. Ed. 60 (2021) 11350–11358. doi: 10.1002/anie.202100782

  27. [27]

      G.D. Wang, Y.Z. Li, W.J. Shi, et al., Angew. Chem. Int. Ed. 61 (2022) e202205427. doi: 10.1002/anie.202205427

  28. [28]

      H. Sun, F. Chen, R. Chen, et al., Small 19 (2023) 2208182. doi: 10.1002/smll.202208182

  29. [29]

      W. Liang, Y. Wu, H. Xiao, et al., AIChE J. 64 (2018) 3390–3399. doi: 10.1002/aic.16182

  30. [30]

      H.G. Hao, Y.F. Zhao, D.M. Chen, et al., Angew. Chem. Int. Ed. 57 (2018) 16067–16071. doi: 10.1002/anie.201809884

  31. [31]

      O.T. Qazvini, R. Babarao, Z.L. Shi, et al., J. Am. Chem. Soc. 141 (2019) 5014–5020. doi: 10.1021/jacs.9b00913

  32. [32]

      H. Zeng, X.J. Xie, M. Xie, et al., J. Am. Chem. Soc. 141 (2019) 20390–20396. doi: 10.1021/jacs.9b10923

  33. [33]

      J. Pei, J.X. Wang, K. Shao, et al., J. Mater. Chem. A 8 (2020) 3613–3620. doi: 10.1039/c9ta12671f

  34. [34]

      X.J. Xie, H. Zeng, W. Lu, et al., J. Mater. Chem. A 11 (2023) 20459–20469. doi: 10.1039/d3ta03852a

  35. [35]

      Y. Ye, Y. Xie, Y. Shi, et al., Angew. Chem. Int. Ed. 62 (2023) e202302564. doi: 10.1002/anie.202302564

  36. [36]

      S.Q. Yang, T.L. Hu, Coord. Chem. Rev. 468 (2022) 214628. doi: 10.1016/j.ccr.2022.214628

  37. [37]

      J.R. Li, R.J. Kuppler, H.C. Zhou, Chem. Soc. Rev. 38 (2009) 1477–1504. doi: 10.1039/b802426j

  38. [38]

      P.Q. Liao, W.X. Zhang, J.P. Zhang, et al., Nat. Commun. 6 (2015) 8697. doi: 10.1038/ncomms9697

  39. [39]

      R.B. Lin, H. Wu, L. Li, et al., J. Am. Chem. Soc. 140 (2018) 12940–12946. doi: 10.1021/jacs.8b07563

  40. [40]

      K.J. Chen, D.G. Madden, S. Mukherjee, et al., Science 366 (2019) 241–246. doi: 10.1126/science.aax8666

  41. [41]

      J. Liu, J. Miao, H. Wang, et al., AIChE J. 69 (2023) e18021. doi: 10.1002/aic.18021

  42. [42]

      S.M. Wang, H.R. Liu, S.T. Zheng, et al., Sep. Purif. Technol. 304 (2023) 122378. doi: 10.1016/j.seppur.2022.122378

  43. [43]

      X. Lin, Y. Yang, X. Wang, et al., Sep. Purif. Technol. 330 (2024) 125252. doi: 10.1016/j.seppur.2023.125252

  44. [44]

      K. Zhang, F. Chu, Y. Hu, et al., Chin. Chem. Lett. 34 (2023) 107766. doi: 10.1016/j.cclet.2022.107766

  45. [45]

      H. Yang, Y. Wang, R. Krishna, et al., J. Am. Chem. Soc. 142 (2020) 2222–2227. doi: 10.1021/jacs.9b12924

  46. [46]

      U. Böhme, B. Barth, C. Paula, et al., Langmuir 29 (2013) 8592–8600. doi: 10.1021/la401471g

  47. [47]

      Y. Chen, Z. Qiao, H. Wu, et al., Chem. Eng. Sci. 175 (2018) 110–117. doi: 10.1016/j.ces.2017.09.032

  48. [48]

      W. Liang, F. Xu, X. Zhou, et al., Chem. Eng. Sci. 148 (2016) 275–281. doi: 10.1016/j.ces.2016.04.016

  49. [49]

      R. Krishna, ACS Omega 5 (2020) 16987–17004. doi: 10.1021/acsomega.0c02218

  50. [50]

      X.W. Gu, J.X. Wang, E. Wu, et al., J. Am. Chem. Soc. 144 (2022) 2614–2623. doi: 10.1021/jacs.1c10973

• 

  Scheme 1  Strategy to boost C₂H₆/C₂H₄ separation through pore functionalization.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (a) Coordination environment of Cu²⁺ ion, (b) layer, (c) cage Ⅰ, (d) cage Ⅱ, and (e) 3D framework in Cu-MOF(NH₂).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Gas adsorption isotherms of Cu-MOF and Cu-MOF(NH₂) at (a) 273 and (b) 298 K. Q_st plots of (c) Cu-MOF(NH₂) and (d) Cu-MOF.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) IAST selectivity of Cu-MOF and Cu-MOF(NH₂) for C₂H₆/C₂H₄ mixtures. (b) Separation potential for C₂H₆/C₂H₄ mixtures (1/15) in different materials. (c) Comparison of C₂H₄ and C₂H₆ uptakes in different materials. (d) Comparison of C₂H₆ uptakes and C₂H₆/C₂H₄ selectivity in different materials.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Breakthrough curves of MOFs for C₂H₆/C₂H₄ mixtures at 298 K, (a) 1/1, (b) 1/9, and (c) 1/15. (d) Comparison of the comprehensive separation performance for Cu-MOF(NH₂) and Cu-MOF.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Breakthrough curves of Cu-MOF(NH₂) for C₂H₆/C₂H₄ (v/v, 1/15) mixtures at different total flow rates at 298 K. (b) Desorption curves of Cu-MOF(NH₂) for C₂H₆/C₂H₄ (v/v, 1/15) mixtures under Ar (7 mL/min) weeping at 323 K. (c) Breakthrough cycles of Cu-MOF(NH₂) for C₂H₆/C₂H₄ mixtures (v/v, 5/5) at 298 K. (d) PXRD patterns of Cu-MOF(NH₂) treated with different environments.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a, b) C₂H₄ and (c, d) C₂H₆ preferential adsorption sites in Cu-MOF and Cu-MOF(NH₂).

  下载: 全尺寸图片 幻灯片
  

  Figure 7  (a, b) C₂H₄ and (c, d) C₂H₆ adsorption sites at 298 K and 100 kPa in Cu-MOF.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  (a, b) C₂H₄ and (c, d) C₂H₆ adsorption sites at 298 K and 100 kPa in Cu-MOF(NH₂).

  下载: 全尺寸图片 幻灯片
•