English

• Ethylene (C₂H₄) is an essential feedstock for the chemical industry, and its downstream chemicals account for the majority of petrochemical products. Presently, three main processes are used to prepare C₂H₄: Naphtha cracking, oxidation and dehydrogenation (ODH) of ethane, and coal-to-ethylene. Nevertheless, in response to increasingly severe environmental issues and cost-cutting, researchers have developed more production processes with lower energy consumption, fewer pollutant emissions, and higher ethylene recovery rates. The oxidative coupling of methane (OCM) for ethylene, which involves a simple and cost-effective process, is a highly promising technique for ethylene production [1]. In the OCM process, the C–H bond is activated by O₂ to generate radicals, which then undergo dehydrogenative coupling to form C₂H₄, while the extensive thermals released in the process may cause over oxidation of C₂H₄ to produce C₂H₂, CO₂, and other byproducts. For the sake of downstream products made from C₂H₄, removing C₂H₄ from other byproducts is critically important. Considering the cost and environmental impact, porous materials show great potential in separating C₂H₄ from light hydrocarbon mixtures. Since their slight difference in quadrupole moments (C₂H₂: 7.2 × 10⁻²⁶ esu cm², CO₂: 4.3 × 10⁻²⁶ esu cm², C₂H₄: 1.5 × 10⁻²⁶ esu cm²) and kinetic sizes (C₂H₂: 3.3 Å, CO₂: 3.3 Å, C₂H₄: 4.2 Å) (Table S1 in Supporting information), one-step purification of C₂H₄ from C₂H₂/CO₂/C₂H₄ ternary mixtures is extremely challenging. Therefore, an effective strategy to address this challenge is to design and synthesize a polar structure, wherein a negative electrostatic potential environment within the pores facilitates the enhancement of adsorption differentiation among gas molecules with distinct polarities. Metal–organic frameworks (MOFs) are a class of porous and inorganic-organic hybrid materials. Currently, the pore size designability and modifiability of MOFs have reached the range of sub-angstrom, thus providing great potential in separating molecules with subtle differences, such as C₂H₂/CO₂ [2–4], C₂H₆/C₂H₄ [5,6], and xylene isomers separation [7–9]. However, most of the reported MOFs have a poor selectivity for both C₂H₂/C₂H₄ and CO₂/C₂H₄ because of their similar physical properties, so it is still difficult to realize one-step purification C₂H₄ from C₂H₂/CO₂/C₂H₄ ternary mixtures [10–13]. In the view of pore structure, reducing the pore size of MOFs is one way to improve selectivity, proper apertures allow the framework to selectively capture small gases while excluding larger gases. Therefore, the design and synthesis of new and polar ultra-microporous MOFs is mandatory considering the kinetic sizes and quadrupole moments of C₂H₂, CO₂, and C₂H₄.

  Zeolite-like metal–organic frameworks (ZMOFs) are a unique branch of MOFs, which exceed zeolites and other MOFs in gas storage and separation application [14–19]. Compared to traditional MOFs, the advantages of ZMOFs primarily lie in their cage-like cavities and ultra-microporous window sizes, which are beneficial for enhancing adsorption capacity and selective adsorption performance. Additionally, the anionic or cationic nature of some ZMOFs can be exploited to tailor the frameworks’ affinity for certain gases, improving separation performance. Compared to other ZMOFs topologies, the sodalite (SOD) net is the most common occurrence, owing to its simple tiling of edge-transitive nets (1121). Specifically, the SOD cage is composed of six four-membered rings (4MR) and eight six-membered rings (6MR), while the sodalite network is a 3D structure formed by the interconnected SOD cages through their coplanar assembly. Through the unremitting efforts of researchers, how to construct the SOD topological ZMOFs has made substantial strides. For example, Eddaoudi's group used the Zr₆ cluster as a tetrahedral building unit (TBU) and judiciously selected the bent ditopic linker with steric hindrance to construct the Zr-sod-ZMOFs [20]. Subsequently, they utilized the centring structure-directing agents (cSDA) with appropriate geometry to prepare 21 types of ZMOFs with SOD topology, during which Fe-sod-ZMOFs-320 possessed the largest pore volume (3.21 cm³/g) of sodalite compounds and exhibited the outstanding gas storage performance [21]. ZIF-8 with SOD topology can be regarded as a potential ideal material for gas separation due to its ultra-microporous pore size (3.4 Å) and high porosity (0.66 cm³/g). However, ZIF-8 shows poor separation ability in slightly different gases (e.g., C₂H₂, CO₂, and C₂H₄) owing to its nonpolar skeleton. To solve this problem, a feasible strategy is to construct an interpenetrated SOD structure that can fine-tune the pore size under sub-angstrom and increase the number of active sites per unit volume. However, rare examples of interpenetrated ZMOFs have been witnessed on account of the trait of zeolite type [22,23], and theoretically, the SOD net is not self-dual and forbidden self-interpenetration. Therefore, it remains a great challenge for the synthesis of interpenetrated SOD-based ZMOFs.

  In this work, we first proposed a feasible 4 + 4 + 2 strategy to prepare a novel ZMOF (namely JLU-MOF119) with a unique two-fold interpenetrated SOD topology based on the 4 + 2 strategy for ZIF-8 synthesis (Scheme 1). Specifically speaking, based on the 4 + 2 strategy, besides the TBU served by the 4-connected metal node and a 2-connected linker (such as imidazole) to generate the 6MR, a coplanar 4-connected linker is properly chosen to replace some 2-connected linker and serves as the 4MR. In such 4 + 4 + 2 strategy, lengthening of the 4-connected ligand is crucial to construct interpenetrated structures. Given that, following our previous work of JLU-MOF52 (GIS) [24], a 2,5-H₄ABTC ligand with low symmetry was utilized as the 4-c ligand. Such ligand may transform its configuration from tetrahedron (in JLU-MOF52) to parallelogram like 3,5-H₄ABTC in PCN-250 in appropriate conditions [25]. Meanwhile, to increase the number of polar sites, the imidazole in ZIF-8 is replaced by the TAZ ligand. As a result, the prepared JLU-MOF119 possesses the SOD cage with abundant uncoordinated O and N sites, and the ultra-microporous pore size (4.1 Å and 3.3 Å). JLU-MOF119 exhibits one-step purification C₂H₄ from ternary mixtures of C₂H₂/CO₂/C₂H₄, reaching 99.2% purity of C₂H₄, as proved by the breakthrough and recycling experiments.

  Scheme 1

  

  Scheme 1.  Illustration of design strategy of 2-fold interpenetrated SOD topology in JLU-MOF119.

  DownLoad: Full-Size Img PowerPoint

  According to the reported ZMOFs, constructing the SOD ZMOFs involves several strategies: (1) 4 + 2 strategy, i.e., 4-connected single-metal-ion nodes serve as TBU bridge by ditopic linker [26–30], (2) 4 + 4 strategy, i.e., bridge the TBU with a tetrahedral tetratopic linker [31], (3) 4 + 6 strategy, i.e., bridge the TBU with a hexacarboxylic linker [32], (4) face-directed strategy, i.e., the linker serves as the face of the cage [33–36], (5) polyhedral assembly strategy, i.e., the single-metal building blocks {In(COO)₄} as the TBU to construct the SOD-type frameworks and the internal trinuclear clusters {In₃O} form the tetrahedral cages (Fig. S1 in Supporting information) [37,38]. Here, based on the 4 + 2 strategy, a novel 4 + 4 + 2 strategy was proposed and successfully applied to prepare JLU-MOF119 with unique 2-fold interpenetrated SOD net. Specifically, a portion of the imidazole ligands is replaced by 4-connected 2,5-H₄ABTC ligands, while the remaining imidazole ligands are substituted by TAZ ligands, thereby forming the 4 + 4 + 2 strategy. The employment of elongate 4MR results in two types of 6MR and distorted framework, eventually leading to the interpenetrated SOD structure.

  The crystallography data reveals that JLU-MOF119 crystallizes in a trigonal system with R-3 space group (Table S2 in Supporting information). In the skeleton, independent Zn²⁺ and 2,5-H₄ABTC adopt four coordination, and TAZ serves as a ditopic linker. Each Zn²⁺ is coordinated with two oxygen atoms from two individual 2,5-H₄ABTC ligands and two nitrogen atoms from different TAZ, respectively, to form the low-symmetry TBU, ZnN₂O₂ (Fig. 1a and Fig. S2a in Supporting information). Each independent 2,5-H₄ABTC ligand is linked to four Zn²⁺ and two TAZ, generating the 4MR (Fig. 1c and Fig. S2b in Supporting information). Moreover, each carboxyl group of 2,5-H₄ABTC contains an uncoordinated O atom and points into the SOD cage. Each individual TAZ linker is coordinated to two Zn²⁺ cations and then engenders the 6MR-1 (Fig. S2c in Supporting information), the adjacent TAZ linkers are interlaced, and the uncoordinated N atoms of TAZ point toward the center of the window. There are two kinds of 6MR in the framework, one is the regular 6MR-1, and the other is formed by two TAZ, four half of 2,5-H₄ABTC and six Zn²⁺ (Figs. 1c and d). In the 6MR-2, the TAZ serves as the shorter edge, and the half of 2,5-H₄ABTC serves as the longer edge, the length varies between TAZ and 2,5-H₄ABTC resulting in the window of 6MR-2 is bigger than that of 6MR-1. From the topological perspective, the 4MR, 6MR-1, and 6MR-2 further extend to a 3D framework, which has the same underlying SOD topology as ZIF-8 (Fig. 1b), yet distinguishes it from the SOD net. In the JLU-MOF119 framework, the 4MR with low symmetry and represents parallelogram, and other 4MR of ZMOFs with SOD net are mostly square. The twisted 4MR gives rise to the formation of 6MR with different sizes, which causes structural interpenetration and distortion of the framework (Fig. 1b). Thus, there are two types of apertures for JLU-MOF119, one is 6MR-1 forming a hexagonal aperture with size of 4.1 Å × 4.1 Å along the [001] direction and the other has a rectangular channel formed by interpenetrated 6MR-2 with size of 4.8 Å × 3.3 Å (Figs. 1e–g). It is noted that the pore structure of JLU-MOF119 is distinct from ZIF-8 due to the interpenetrated framework, in which ZIF-8 only has one kind of hexagonal aperture and a significant smaller cage of SOD (11.6 Å vs. 18.6 Å), yet the cage of JLU-MOF119 is reduced to 5.9 Å after the interpenetration of the framework (Fig. S3 in Supporting information).

  Figure 1

  

  Figure 1.  Single crystal structure illustration of JLU-MOF119: (a) The Zn²⁺ node, TAZ, and 2,5-H₄ABTC linker. (b) Polyhedral view of one SOD cage and ball-and-stick view of 2-fold interpenetrated framework. (c) Ball-and-stick representations of the 6MR, TBU, and 4MR SBU. (d) Stick representations of the 6MR, TBU, and 4MR SBU. (e) The CPK view of JLU-MOF119 framework. (f) The window size of 6MR. (g) Another type of window size after interpenetration. Color code: Zn, purple; C, gray; O, pink; N, brown.

  DownLoad: Full-Size Img PowerPoint

  Notably, JLU-MOF119 possesses two kinds of ultra-microporous pores as well as better stability in various solvents and thermal stability before 210 ℃ (Figs. S4 and S5 in Supporting information), which endows it with potential adsorption and separation ability in light hydrocarbons mixtures. N₂ could not be adsorbed by JLU-MOF119 for its ultra-microporous window size, thereby, CO₂ adsorption was used to evaluate the porosity of JLU-MOF119 at 195 K. As shown in Fig. 2a, the uptake of JLU-MOF119 is 125 cm³/g at 195 K and 100 kPa. The Brunauer-Emmett-Teller (BET) and Langmuir surface areas of JLU-MOF119 are 387 and 433 m²/g, respectively. Light hydrocarbons are significant raw materials and industrial gases for the chemical industry, and the ultra-microporous window size of JLU-MOF119 prompts us to hypothesize that it has the potential for the separation of light hydrocarbons. Therefore, JLU-MOF119 was performed for single-component adsorption of N₂, CO₂, CH₄, C₂H₆, C₂H₄, C₂H₂, and C₃H₈ (Figs. 2b–d and Fig. S6 in Supporting information) at 273 and 298 K. Apparently, the gases of N₂, CH₄, C₂H₆, and C₃H₈ with bigger kinetic diameters are hardly adsorbed by JLU-MOF119 at 273 and 298 K, the slight uptake is primarily attributed to the adsorption on the surface of JLU-MOF119. Simultaneously, the C₂H₂, CO₂, and C₂H₄ can be adsorbed by JLU-MOF119, these results are consistent with the ultra-microporous pore size, which means that JLU-MOF119 exhibits a respectable size sieving effect. Moreover, the uptake of C₂H₄ (17.37 cm³/g) is lower than that of C₂H₂ (45.61 cm³/g) and CO₂ (31.54 cm³/g) at 1.0 bar and 298 K, which uncovers the possibility of separating C₂H₂/C₂H₄ and CO₂/C₂H₄. Simultaneously, compared to ZIF-8, the advantage of the interpenetrated framework is more obvious in gas separation, which exhibits an increase in the uptake of gases with smaller size (e.g., CO₂ and C₂H₂), while a significant decrease of bigger gas molecules (e.g., C₂H₆, C₃H₈, and C₃H₆) [39–41]. Moreover, the adsorption capacity of JLU-MOF119 for C₂H₂ remains stable after 12 cyclic tests, indicating that JLU-MOF119 has excellent stability and C₂H₂ sorption capacity (Fig. S7 in Supporting information).

  Figure 2

  

  Figure 2.  JLU-MOF119 adsorption/desorption isotherms: (a) 195 K CO₂, (b) C₂H₂, (c) CO₂ and (d) C₂H₄ at 273 K (blue) and 298 K (green). (e) Q_st for C₂H₂, CO₂, C₂H₄ of JLU-MOF119. (f) Selectivity for C₂H₂/C₂H₄ and CO₂/C₂H₄ mixtures predicted by IAST at 298 K and 1.0 bar.

  DownLoad: Full-Size Img PowerPoint

  The isosteric heat of adsorption (Q_st) for C₂H₂, CO₂, and C₂H₄ was calculated by the Virial equation to estimate the affinity between gases and the skeleton. As shown in Fig. 2e, the Q_st of C₂H₂, CO₂, and C₂H₄ are 22.6, 21.5, and 19.55 kJ/mol respectively at zero-coverage, indicating a stronger affinity of C₂H₂ and CO₂ than C₂H₄. Additionally, all the values of Q_st are lower than the majority of MOFs, e.g., FJI-H36 (36.1, 29.2, and 29.9 kJ/mol) [42], SNNU-98-Mn (60.2, 29.2, and 30.8 kJ/mol) [43], and ZNU-8 (33.1, 26.6, 25.7 kJ/mol) [4], which implies the low energy consumption for regeneration. The simulation of the adsorption isotherms and Q_st plots for the other gases are shown in Figs. S8–S11, Tables S3 and S4 (Supporting information).

  To assess the separation capacity, the adsorption selectivity was calculated by the ideal adsorption solution theory (IAST) of dual-site Langmuir-Freundlich model from single component adsorption isotherms. The selectivity for C₂H₂/C₂H₄ (50/50, 1/99, v/v) is 3.4 and 3.3 (Fig. 2f), higher than those reported MOFs of Fe-MOF-74 (2.08, 50/50) [44], BSF-1 (2.3, 50/50) [45], TJT-100 (1.8, 50/50) [46], JLU-MOF88 (3, 50/50) (Table S5 in Supporting information) [47]. The selectivity for CO₂/C₂H₄ (50/50, 1/99, v/v) is 2.03 and 2.05, higher than Zn-atz-oba (1.33, 50/50) [13], NKMOF-1-Ni (0.8, 50/50) [48]. Compared to reported MOFs in purification C₂H₄ from ternary mixtures of C₂H₂/CO₂/C₂H₄ (Fig. S12 and Table S6 in Supporting information), such as Zn-atz-oba (C₂H₂/C₂H₄, 1.33, CO₂/C₂H₄, 1.43) [13], Zn-fa-atz (C₂H₂/C₂H₄, 1.5, CO₂/C₂H₄, 1.4) [10], JLU-MOF119 is an effective purification candidate for C₂H₄ from ternary mixtures of C₂H₂/CO₂/C₂H₄ by one step.

  As is well known, compared to C₂H₄, C₂H₂, and CO₂ possess higher electron cloud densities. Therefore, a polar cage surface demonstrates a stronger affinity for C₂H₂ and CO₂. To thoroughly verify the practical application potential of JLU-MOF119 for C₂H₂/CO₂/C₂H₄ separation, JLU-MOF119 was executed the breakthrough experimental tests. The binary mixtures breakthrough experiments of C₂H₂/C₂H₄ and CO₂/C₂H₄ were performed in equimolar at 2 mL/min and 298 K (Figs. 3a and b). In Fig. 3a, the C₂H₄ is first flowed out at 17 min, while the C₂H₂ is eluted at 32 min. The experiment result is consistent with the static adsorption isotherm. Compared to separate C₂H₂/C₂H₄, the mixtures of CO₂ and C₂H₄ have closer quadrupole moments, which means it would be more obstacle to separate CO₂/C₂H₄. As indicated in Fig. 3b, the C₂H₄ is eluted at 17 min, and CO₂ is retained until eluted at 33 min, confirming JLU-MOF119’s exceptional CO₂/C₂H₄ separation performance. In the OCM process, C₂H₂, CO₂, and C₂H₄ are coexistent, thus the breakthrough experiment of C₂H₂/CO₂/C₂H₄ was executed at 3 mL/min and 298 K (Fig. 3c). As shown in Fig. 3c, the C₂H₄ is first flowed out at 14 min, while C₂H₂ and CO₂ are flowed out at 21 min and 22 min. And the purity of C₂H₄ reaches a high value of 99.2% at the outlet before the C₂H₂ and CO₂ flow out further proving JLU-MOF119 achieves one-step purification of C₂H₄ from C₂H₂/CO₂/C₂H₄ ternary mixtures. To better meet the practical application, cyclic tests on JLU-MOF119 were performed to evaluate the regenerative capacity. As shown in Figs. 3d–f, the capacity of JLU-MOF119 to separate C₂H₂/C₂H₄, CO₂/C₂H₄ and C₂H₂/CO₂/C₂H₄ remains virtually unchanged after 4 cycles. Moreover, the consistent C₂H₂ adsorption capacity after 12 cycles indicates that JLU-MOF119 retains its selective C₂H₄ purification performance even after 12 breakthrough experiments (Fig. S7).

  Figure 3

  

  Figure 3.  Breakthrough experiments of (a) C₂H₂/C₂H₄ in equimolar at 2 mL/min (v/v = 1/1) and 298 K. (b) CO₂/C₂H₄ in equimolar at 2 mL/min (v/v = 1/1) and 298 K. (c) C₂H₂/CO₂/C₂H₄ in equimolar at 3 mL/min (v/v/v = 1/1/1) and 298 K. Cycles experiments for (d) C₂H₂/C₂H₄, (e) CO₂/C₂H₄ and (f) C₂H₂/CO₂/C₂H₄.

  DownLoad: Full-Size Img PowerPoint

  First-principles based calculations with empirical correction for dispersive interactions were performed to investigate the mechanisms that differentiate C₂H₂, CO₂, and C₂H₄ adsorption in JLU-MOF119 (Figs. 4a–c) [49,50]. In the most plausible adsorption structure (Fig. S13 in Supporting information), C₂H₂ stays within the channel surrounded by Zn(II), TAZ, and dimethyl ammonia cations (DMA) (Fig. 4a), and the corresponding E_ads is 0.75 eV. C₂H₂ stays around the 6MR-2, it interacts mainly with the uncoordinated O of the 2,5-H₄ABTC at H(C₂H₂)−O(2,5-H₄ABTC) distance of 2.31 and 2.61 Å, respectively, and also interacts electrostatically with the TAZ right below it at C(C₂H₂)−N(DMA) and C(C₂H₂)−C(DMA) distances of 3.46 and 3.53 Å, respectively. C₂H₂ is also distorted slightly upon adsorption, the H–C–C and C–C–H angles are distorted from 180° when freestanding to 176.38° and 179.43°, respectively. These, together with the isosurface plot of charge density difference upon C₂H₂ adsorption suggest that dispersive interactions are formed to stabilize C₂H₂ (Fig. 4d).

  Figure 4

  

  Figure 4.  The most plausible adsorption structures (a–c) and corresponding isosurface plots of charge density difference (d–f) of C₂H₂, CO₂ and C₂H₄ in JLU- MOF119 viewed along z axis. In (a–f), the charge accumulation and charge depletion regions are in yellow and blue, respectively, and the isovalues are ±0.0003 a.u.

  DownLoad: Full-Size Img PowerPoint

  Different from the case of C₂H₂, CO₂ would diffuse into the 6MR-2 opening to interact with the H of adjacent and nearly parallelly arranged TAZ ligands to gain additional stability (Fig. 4b), and the calculated E_ads is 0.31 eV. In this configuration, positively charged C(CO₂) interacts electrostatically with an H (2,5-H₄ABTC) and a carboxylic-O of uncoordination from adjacent 2,5-H₄ABTC ligands, and two O(CO₂) also interact with the 2 H of different TAZ ligands. The calculated O(2,5-H₄ABTC)−C(CO₂) and H(2,5-H₄ABTC)−C(CO₂) distances are 2.75 and 3.20 Å, respectively. The calculated H(TAZ)−O(CO₂) distances are 2.89 and 3.29 Å, respectively. The O–C–O angle within CO₂ is also distorted from 180° to 177.89°. The isosurface plot of charge density difference upon CO₂ adsorption also suggests the dispersive nature of CO₂ adsorption with cooperative interactions with both negatively and positively charge centres of adjacent ligands that were previously proposed to account for enhanced CO₂ adsorption in microporous amorphous carbon (Fig. 4e).

  The case of C₂H₄ is similar to that of C₂H₂ (Fig. 4c). Dispersive interactions are formed between C₂H₄ and TAZ ligands. The calculated H(C₂H₄)−N(TAZ) distances are 3.00, 3.05, 3.06, and 3.07 Å, respectively. C₂H₄ is also distorted slightly, the H–C–C angles are distorted from 121.69° when freestanding to 121.87° after adsorption. These together with the charge density difference upon C₂H₄ adsorption suggests that dispersive interactions account for C₂H₄ adsorption (Fig. 4f). The charge transfer accompanying C₂H₄ adsorption is less significant as compared with that of C₂H₂ and CO₂ (Figs. 4d and e), so C₂H₄ adsorption would be weaker. Accordingly, the calculated E_ads of C₂H₄ is 0.26 eV and C₂H₄ adsorption is the weakest among the three gases investigated.

  In summary, a novel 4 + 4 + 2 strategy is designed to construct the first interpenetrated SOD-based ZMOF, JLU-MOF119. The 4 + 4 + 2 strategy is based on the 4 + 2 strategy of constructing ZIF-8, which is formed by using the 4-c ligand to replace a portion of 2-c ligand. Specifically, the former 4 refers to the TBU of ZnN₂O₂, which consists of TAZ and 2,5-H₄ABTC; the latter 4 refers to the 4-c ligand of 2,5-H₄ABTC; and the 2 refers to the 2-c ligand of TAZ. In such framework, the use of elongated and distorted ligand of 2,5-H₄ABTC results in the formation of two types of 6MR, and subsequently assemble into a distorted and interpenetrated SOD framework. JLU-MOF119 possesses abundant uncoordinated N and O sites in the SOD cage and the ultra-microporous pore size, which represents a decent separation capacity of C₂H₂/CO₂/C₂H₄, and the C₂H₄ purity of 99.2% is obtained by dynamic column breakthrough experiments. The simulations and calculations study gives more insight into the unique structure feature and the distinct affinity between gases of C₂H₂, CO₂, and C₂H₄ and the framework, accounting for the excellent selectivity of light hydrocarbons. This work pioneers a new and feasible strategy of SOD topology with interpenetrated structure, and explores new MOF materials applied in the field of multi-component light hydrocarbon separation, which also provides an effective pathway for designing structures with specialized functions.

  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.

  CRediT authorship contribution statement

  Jianyun Li: Writing – original draft, Investigation, Conceptualization. Lifei Zou: Methodology, Investigation, Data curation. Jianwei Cao: Data curation. Zhaohui Shi: Data curation. Zhiwei Liu: Data curation. Lirong Zhang: Writing – review & editing, Supervision. Jiyang Li: Writing – review & editing. Xin Liu: Data curation. Yunling Liu: Writing – review & editing, Funding acquisition, Conceptualization.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 22171100 and U23A20360) and the ‘111 Center’ (No. B17020). X. Liu was supported by National Natural Science Foundation of China (Nos. 21771029, 11811530631, and 21573034). The supercomputer time was provided by High Performance Computing Center at Dalian University of Technology. L. Zou was supported by National Natural Science Foundation of China (No. 22361001).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111189.

  
  
• 1. [1]

     J. Liu, J. Yue, M. Lv, et al., Carbon Res. Conv. 5 (2022) 1–14.

  2. [2]

     J. Wang, Y. Zhang, Y. Su, et al., Nat. Commun. 13 (2022) 200.

  3. [3]

     L. Zhang, T. Xiao, X. Zeng, et al., J. Am. Chem. Soc. 146 (2024) 7341–7351. doi: 10.1021/jacs.3c12032

  4. [4]

     Y. Zhang, W. Sun, B. Luan, et al., Angew. Chem. Int. Ed. 62 (2023) e202309925. doi: 10.1002/anie.202309925

  5. [5]

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

  6. [6]

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

  7. [7]

     J. Zhou, T. Ke, Y. Song, et al., J. Am. Chem. Soc. 144 (2022) 21417–21424. doi: 10.1021/jacs.2c10595

  8. [8]

     X. Cui, Z. Niu, C. Shan, et al., Nat. Commun. 11 (2020) 5456. doi: 10.1038/s41467-020-19209-7

  9. [9]

     X. Li, J. Wang, N. Bai, et al., Nat. Commun. 11 (2020) 4280. doi: 10.1038/s41467-020-17640-4

  10. [10]

      R. Yang, Y. Wang, J. Cao, et al., Nat. Commun. 15 (2024) 804. doi: 10.1038/s41467-024-45081-w

  11. [11]

      Y. Jiang, Y. Hu, B. Luan, et al., Nat. Commun. 14 (2023) 401. doi: 10.1177/02676591211043705

  12. [12]

      Q. Dong, Y. Huang, K. Hyeon-Deuk, et al., Adv. Fun. Mater. 32 (2022) 2203745. doi: 10.1002/adfm.202203745

  13. [13]

      J. Cao, S. Mukherjee, T. Pham, et al., Nat. Commun. 12 (2021) 6507. doi: 10.1038/s41467-021-26473-8

  14. [14]

      M. Eddaoudi, D. Sava, J. Eubank, et al., Chem. Soc. Rev. 44 (2015) 228–249. doi: 10.1039/C4CS00230J

  15. [15]

      Y. Tan, F. Wang, J. Zhang, Chem. Soc. Rev. 47 (2018) 2130–2144. doi: 10.1039/c7cs00782e

  16. [16]

      A. Phan, C. Doonan, F. Uribe-Romo, et al., Acc. Chem. Res. 43 (2010), 58–67. doi: 10.1021/ar900116g

  17. [17]

      F. Yi, H. Yang, X. Zhao, et al., Angew. Chem. Int. Ed. 58 (2019) 2889–2892. doi: 10.1002/anie.201900106

  18. [18]

      J. Li, D. Sava, V. Guillerm, et al., Chem 10 (2024) 567–577. doi: 10.1016/j.chempr.2023.09.026

  19. [19]

      K. Park, Z. Ni, A. Cote, et al., Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 10186–10191. doi: 10.1073/pnas.0602439103

  20. [20]

      N. Alsadun, G. Mouchaham, V. Guillerm, et al., J. Am. Chem. Soc. 142 (2020) 20547–20553. doi: 10.1021/jacs.0c10007

  21. [21]

      M. Barsukova, A. Sapianik, V. Guillerm, et al., Nat. Synth. 3 (2024) 33–46.

  22. [22]

      W. Xie, W. He, S. Li, et al., Chem. Eur. J. 22 (2016) 17298–17304. doi: 10.1002/chem.201603487

  23. [23]

      Y. Liu, M. Grzywa, M. Weil, et al., J. Solid State Chem. 183 (2010) 208–217. doi: 10.1166/sl.2010.1228

  24. [24]

      L. Zou, X. Sun, J. Yuan, et al., Inorg. Chem. 57 (2018) 10679–10684. doi: 10.1021/acs.inorgchem.8b01330

  25. [25]

      D. Feng, K. Wang, Z. Wei, et al., Nat. Commun. 5 (2014) 5723. doi: 10.1038/ncomms6723

  26. [26]

      R. Banerjee, A. Phan, B. Wang, et al., Science 319 (2008) 939–943. doi: 10.1126/science.1152516

  27. [27]

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

  28. [28]

      Y. Liu, V. Kravtsov, R. Larsen, et al., Chem. Commun. 14 (2006) 1488–1490. doi: 10.1039/b600188m

  29. [29]

      B. Abrahams, M. Haywood, R. Robson, et al., Angew. Chem. Int. Ed. 42 (2003) 1112–1115. doi: 10.1002/anie.200390292

  30. [30]

      L. Tabares, J. Navarro, J. Salas, J. Am. Chem. Soc. 123 (2001) 383–387. doi: 10.1021/ja002624a

  31. [31]

      L. Sun, H. Xing, Z. Liang, et al., Chem. Commun. 49 (2013) 11155–11157. doi: 10.1039/c3cc43383h

  32. [32]

      W. Du, Z. Zhu, Y. Bai, et al., Chem. Commun. 54 (2018) 5972–5975. doi: 10.1039/c8cc02193g

  33. [33]

      M. Dinca, W. Han, Y. Liu, et al., Angew. Chem. Int. Ed. 46 (2007) 1419–1422. doi: 10.1002/anie.200604362

  34. [34]

      Y. Tan, Y. He, J. Zhang, Chem. Commun. 47 (2011) 10647–10649. doi: 10.1039/c1cc14118j

  35. [35]

      T. Mcdonald, D. D'Alessandro, R. Krishna, et al., Chem. Sci. 2 (2011) 2022–2028. doi: 10.1039/c1sc00354b

  36. [36]

      M. Feng, X. Zhou, X. Wang, et al., ACS Appl. Mater. Interfaces 15 (2023) 11837–11844. doi: 10.1021/acsami.2c22725

  37. [37]

      S. Zheng, J. Bu, Y. Li, et al., J. Am. Chem. Soc. 132 (2010) 17062–17064. doi: 10.1021/ja106903p

  38. [38]

      S. Zheng, T. Wu, F. Zuo, et al., J. Am. Chem. Soc. 134 (2012) 1934–1937. doi: 10.1021/ja209800x

  39. [39]

      S. Xiang, W. Zhou, J. Gallegos, et al., J. Am. Chem. Soc. 131 (2009) 12415–12419. doi: 10.1021/ja904782h

  40. [40]

      J. McEwen, J. Hayman, A. Yazaydin, Chem. Phys. 412 (2013) 72–76. doi: 10.1016/j.chemphys.2012.12.012

  41. [41]

      L. Lyu, H. Wu, L. Li, et al., Chem. Ing. Tech. 94 (2022) 119–127. doi: 10.1002/cite.202000259

  42. [42]

      J. Tian, Q. Chen, F. Jiang, et al., Angew. Chem. Int. Ed. 62 (2023) e202215253. doi: 10.1002/anie.202215253

  43. [43]

      J. Wang, S. Fan, H. Li, et al., Angew. Chem. Int. Ed. 62 (2023) e202217839. doi: 10.1002/anie.202217839

  44. [44]

      E. Bloch, W. Queen, R. Krishna, et al., Science 335 (2012) 1606–1610. doi: 10.1126/science.1217544

  45. [45]

      Y. Zhang, L. Yang, L. Wang, et al., Angew. Chem. Int. Ed. 131 (2019) 8229–8234. doi: 10.1002/ange.201903600

  46. [46]

      H. Hao, Y. Zhao, D. Chen, et al., Angew. Chem. Int. Ed. 130 (2018) 16299–16303. doi: 10.1002/ange.201809884

  47. [47]

      T. Xu, P. Zhang, F. Cui, et al., Adv. Mater. 35 (2023) 2204553. doi: 10.1002/adma.202204553

  48. [48]

      Y. Peng, T. Pham, P. Li, et al., Angew. Chem. Int. Ed. 57 (2018) 10971–10975. doi: 10.1002/anie.201806732

  49. [49]

      G. Kresse, J. Furthmuller, Phys. Rev. B 54 (1996) 11169–11186. doi: 10.1103/PhysRevB.54.11169

  50. [50]

      E. Johnson, A. Becke, J. Chem. Phys. 124 (2006) 174104. doi: 10.1063/1.2190220

• 

  Scheme 1  Illustration of design strategy of 2-fold interpenetrated SOD topology in JLU-MOF119.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Single crystal structure illustration of JLU-MOF119: (a) The Zn²⁺ node, TAZ, and 2,5-H₄ABTC linker. (b) Polyhedral view of one SOD cage and ball-and-stick view of 2-fold interpenetrated framework. (c) Ball-and-stick representations of the 6MR, TBU, and 4MR SBU. (d) Stick representations of the 6MR, TBU, and 4MR SBU. (e) The CPK view of JLU-MOF119 framework. (f) The window size of 6MR. (g) Another type of window size after interpenetration. Color code: Zn, purple; C, gray; O, pink; N, brown.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  JLU-MOF119 adsorption/desorption isotherms: (a) 195 K CO₂, (b) C₂H₂, (c) CO₂ and (d) C₂H₄ at 273 K (blue) and 298 K (green). (e) Q_st for C₂H₂, CO₂, C₂H₄ of JLU-MOF119. (f) Selectivity for C₂H₂/C₂H₄ and CO₂/C₂H₄ mixtures predicted by IAST at 298 K and 1.0 bar.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Breakthrough experiments of (a) C₂H₂/C₂H₄ in equimolar at 2 mL/min (v/v = 1/1) and 298 K. (b) CO₂/C₂H₄ in equimolar at 2 mL/min (v/v = 1/1) and 298 K. (c) C₂H₂/CO₂/C₂H₄ in equimolar at 3 mL/min (v/v/v = 1/1/1) and 298 K. Cycles experiments for (d) C₂H₂/C₂H₄, (e) CO₂/C₂H₄ and (f) C₂H₂/CO₂/C₂H₄.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The most plausible adsorption structures (a–c) and corresponding isosurface plots of charge density difference (d–f) of C₂H₂, CO₂ and C₂H₄ in JLU- MOF119 viewed along z axis. In (a–f), the charge accumulation and charge depletion regions are in yellow and blue, respectively, and the isovalues are ±0.0003 a.u.

  下载: 全尺寸图片 幻灯片
•