Supramolecular framework crystallinity engineering via surface-confined polymerization for enhanced C₃H₆/C₂H₄ separation

Guoliang Liu^a, Fan Li^a, Ze-Jiu Diao^a, Hua-Dong Li^a, Xiaoyu Wu^bc, Hang-Ou Qi^a, Lifeng Ding^b and Lin-Bing Sun*^a
aState Key Laboratory of Materials-Oriented Chemical Engineering, Jiangsu National Synergetic Innovation Center for Advanced Material (SICAM), College of Chemical Engineering, Nanjing Tech University, Nanjing 211816, China. E-mail: lbsun@njtech.edu.cn
^bAdvanced Materials Research Center and Department of Chemistry and Materials Science, School of Science, Xi'an Jiaotong-Liverpool University, Suzhou 215123, China
^cDepartment of Chemical and Biomolecular Engineering, National University of Singapore, 117576, Singapore

First published on 29th July 2025

---

---

1. Introduction

Ethylene (C₂H₄) and propylene (C₃H₆) are essential feedstocks in the petrochemical industry, widely used in the production of plastics, solvents, cosmetics, paints, and pharmaceuticals.¹ Traditionally, these light olefins are produced through the thermal or catalytic cracking of crude oil. However, in recent decades, alternative routes such as the methanol-to-olefin (MTO) process have gained significant attention. The MTO process has emerged as an efficient method for producing light olefins from methanol, which can be derived from coal, natural gas, or biomass.² Despite its advantages, the MTO process generates a mixture of C₂H₄ and C₃H₆, necessitating separation to obtain pure C₂H₄ and C₃H₆ for various industrial applications. Currently, cryogenic distillation is the most commonly used method for separating the C₃H₆/C₂H₄ mixture. However, this process is energy-intensive and costly considering that C₃H₆ and C₂H₄ have similar physicochemical properties.^3–5 Therefore, the development of more efficient and sustainable separation technologies has become a crucial research focus. Adsorption technology leverages the differences in the physical and chemical properties of the gases, such as kinetic diameter, polarity, and polarizability, to achieve separation.^6–15 This method is considered a promising alternative due to its potential for lower energy consumption and higher selectivity.^16–18 The fundamental aspect of adsorptive separation technology lies in the adsorbent, so it is important to develop porous materials with efficient C₃H₆/C₂H₄ separation performance.

Metal–organic cages (MOCs) are coordination compounds constructed from metal clusters and organic ligands, which possess tunable structures and are applicable in different scenarios.^19–27 Recently, the development of supramolecular frameworks (SMFs) formed via hierarchical self-assembly of MOCs on the basis of multiple weak interactions has attracted much attention.^28–31 SMFs feature intrinsic porosity and packed extrinsic porosity depending on the packing style of MOCs, broadening their applications for gas storage and separation.^32,33 It should be mentioned that the packed extrinsic pores are often located around the mesopore scale. This is highly desired for C₃H₆/C₂H₄ separation owing to its significantly facilitated C₃H₆ uptake.³⁴ However, the packed extrinsic porosity tended to be compromised because of amorphous transformation upon removal of the guest solvent molecules due to its hollow core and weak interactions between adjacent MOCs, limiting their applications in C₃H₆/C₂H₄ separation. The isolation of MOCs in mesoporous substrates, such as mesoporous silica, mesoporous carbon, and mesoporous metal–organic frameworks (MOFs), has been exploited to avoid the aggregation of MOCs.³⁵ The pores of MOCs are maintained accessible after isolation imparted by the substrates. However, this strategy can only work with a small amount of MOC and an aggregation will revive the increment of MOC loadings. This will compromise the packed extrinsic porosity of MOC-based SMFs and then gas adsorption performance.³⁶ Given these, developing a strategy to stabilize SMFs with preserved crystallinity, especially for the ones of high porosity, with the target for efficient C₃H₆/C₂H₄ separation is highly desired.

The shells of soft-bodied animals (such as shrimp and crabs) not only protect their internal soft organs but also provide a skeletal-like support function. Inspired by this, we adopted a strategy to preserve framework crystallinity and achieve high porosity upon activation through a surface-confined polymerization of isophorone diisocyanate (IDI) on the crystal surface of NUT-161 to mimic these dual functions (Fig. 1). According to N₂ adsorption and desorption analysis, NUT-161@PolyIDI exhibits a 324% increment in specific surface area compared to the activated NUT-161, indicating a significant improvement in crystallinity as further confirmed via powder X-ray diffraction (PXRD) analysis. The C₃H₆ and C₂H₄ adsorption capacities were systematically evaluated, and NUT-161@PolyIDI displays improved C₃H₆ uptake owing to the preserved packed mesopores, resulting in separation potential (Δq) enhancement compared to NUT-161. Moreover, the dynamic breakthrough studies demonstrate that NUT-161@PolyIDI exhibits a highly efficient C₃H₆/C₂H₄ separation capacity, especially at 273 K, and the separation performance can be well-maintained after three cycles. To understand the separation mechanism, the Grand Canonical Monte Carlo (GCMC) simulation was performed to illustrate the separation performance of NUT-161@PolyIDI imparted by the well-preserved crystallinity and low content of surface-covered polyIDI. The results suggest the framework exhibits multiple C–H⋯π interactions with C₃H₆, preferentially binding to C₃H₆ over C₂H₄. This bio-inspired approach allows for innovative solutions that leverage the efficiency and effectiveness found in nature to address science and engineering challenges.

2. Experimental section

2.1. Preparation of NUT-161

To synthesize NUT-161, 0.06 g of 4,4′-(ethyne-1,2-diyl)dibenzoic acid (H₂-EDBA) and 0.15 g of zirconocene dichloride (Cp₂ZrCl₂) were placed in a 20 mL glass vial. Then, 9 mL of dimethylacetamide (DMA) was added, and the solution was sonicated at 30 °C until it became clear. Subsequently, 450 μL of deionized water and 3 mL of dichloromethane (DCM) were added, and the solution was sonicated again until homogeneous. The sealed vial was then kept at 20 °C for 8 hours. After completion of the reaction, the product was washed sequentially with DMA (15 mL × 3), tetrahydrofuran (THF, 15 mL × 3), DCM (15 mL × 3), and acetone (15 mL × 3), yielding the white crystalline NUT-161 (yield ∼67.6%).

2.2. Preparation of NUT-161@PolyIDI

Next, to synthesize NUT-161@PolyIDI, 0.05 g of NUT-161 was placed in a 20 mL glass vial and washed several times with DCM. The excess DCM was removed, leaving a small amount to cover the sample. Then, 1 mL of fresh DCM and 7 μL of IDI were added. The mixture was gently stirred with a pipette to disperse the sample and sealed. The vial was left at room temperature for 24 hours. After sufficient reaction, the product was washed sequentially with DCM (15 mL × 3) and acetone (15 mL × 3) to remove the unreacted IDI, resulting in NUT-161@PolyIDI.

2.3. Gas adsorption measurement

C₃H₆ and C₂H₄ adsorption isotherms were collected on an ASAP 2020 analyzer at 298 K. The adsorption temperature can be well controlled by immersing the sample tube in the water bath. Prior to each measurement, freshly prepared samples were first exchanged with DCM for 2 days. Approximately 50 mg of the sample soaked in DCM was then taken and placed in the quartz tube used for the test, and the solvent was removed using a vacuum pump. Finally, the sample was degassed under a vacuum line of ASAP 2020 at 353 K for 4 h to completely remove the guest molecules in the pores of NUT-161 and NUT-161@PolyIDI to be tested. Ultra-high purity grade C₃H₆ (99.99%) and C₂H₄ (99.99%) were used for gas adsorption measurements.

3. Results and discussion

3.1. Structural and surface properties

Following the successful synthesis of the ligand H₂-EDBA (Fig. S1), a previously unreported metal–organic composite material, NUT-161, was successfully prepared from DMA and DCM solvents containing H₂-EDBA and Cp₂ZrCl₂ in the presence of water. NUT-161 was obtained as cubic crystals, a representative morphology for zirconium-MOC-based SMFs, such as SMF-BDC, SMF-BPDC, PCC-20t, and MSF-1, indicating they may possess an isoreticular structure with the same topology.^37–41 During the process of multiple attempts to grow crystals of appropriate size and high enough diffraction intensity for a single crystal X-ray diffraction study, we can easily obtain the unit cell parameters of NUT-161 (Table S1). We made a comparison between NUT-161 and the reported SMFs. The unit cell length of these SMFs is positively correlated with the ligand length. Specifically, the unit cell length of NUT-161 is slightly larger than the value for PCC-20t, considering the two samples were constructed from ligands with very similar size. This comparison confirms that NUT-161 is isoreticular to each of these SMFs. This indicated that a cationic coordination cage with V₄E₆ (V = vertex, E = edge) topology was formed, consisting of [Cp₃Zr₃O(OH)₃]⁺ clusters at the vertices and deprotonated H₂-EDBA ligands at the edges. Chloride ions (Cl⁻) balance the positive charge, and facilitate hierarchical self-assembly into a high-order 3D SMF (Fig. 2). Within the crystal lattice, the tetrahedral Zr-EDBA cages are interconnected through Cl⁻ mediated charge-assisted O–H⁺⋯Cl⁻⋯⁺H–O hydrogen bonds, thereby forming a (4,8)-connected three-dimensional network that exhibits the flu topology. Notably, the structure of NUT-161 features a quasi-small rhombicuboctahedron, which arises from the suboptimal packing of eight tetrahedral Zr-EDBA cages. This distinctive arrangement gives rise to a highly porous structure, indicative of a sophisticated SMF.

NUT-161 exhibits good solubility in polar solvents, such as methanol, backstopping the tetrahedral cage of NUT-161 bridges with neighboring ones through weak interactions, unlike the MOFs where the metal clusters and organic ligands are linked through coordination bonds. Due to its good solubility, the as-synthesized NUT-161 was first investigated by ¹H NMR (Fig. 3a). The characteristic signals of H₂-EDBA were observed at chemical shifts of 7.57 ppm and 7.98 ppm, consistent with the ¹H NMR spectrum of H₂-EDBA (Fig. S1). Furthermore, the detection of a distinct peak at 6.60 ppm, corresponding to cyclopentadiene (Cp), provides clear evidence for the successful synthesis of NUT-161. To further confirm the exact composition of NUT-161, the sample was analyzed by high-resolution mass spectrometry (HR-MS). It can be observed that three prominent signal peaks are detected at m/z values of 931.9003, 1242.1996, and 1863.2877, corresponding to the +4, +3, and +2 charge states, respectively (Fig. 3b). Analysis of the results shows that the relative molecular mass of Zr-EDBA is approximately 3727 Da, which is in close agreement with the theoretical value. The experimental and simulated isotope patterns show the same normal distribution results and matching peak positions, confirming that the observed signal peaks correspond to the three ion peaks of the target compound of Zr-EDBA, indicating a successful synthesis of the material. The phase purity of NUT-161 was verified through a consistent match between the experimental and simulated PXRD patterns (Fig. 3c). The mismatch between the measured and simulated diffraction peak intensities is chiefly caused by the preferred crystallographic orientation.^42,43 NUT-161 was also examined by Fourier transform infrared (FT-IR) and X-ray photoelectron spectroscopy (XPS) for structural confirmation (Fig. 3d–g, S2, and S3) and the results support the successful synthesis of NUT-161.

The porosity characteristics of the activated NUT-161 were meticulously examined through N₂ adsorption and desorption isotherm analysis conducted at 77 K. Intriguingly, NUT-161 displays a type I isotherm, which is quintessential for microporous materials, and has an N₂ adsorption capacity of around 156 cm³ g⁻¹ at a pressure of 1 bar (Fig. 4a). The Brunauer–Emmett–Teller (BET) specific surface area, ascertained from N₂ adsorption analysis, was found to be 522 m² g⁻¹ (Fig. 4b). Notably, NUT-161 exhibits a lower N₂ uptake capacity and a smaller BET surface area compared to other reported Zr-MOC-based SMFs that typically feature lower porosity. For instance, SMF-BDC, which is constructed using the relatively shorter organic ligand terephthalic acid (H₂-BDC), demonstrates higher values in these parameters. In addition, the pore size distribution (PSD) shows that the pores in NUT-161 are mainly around 2.0 nm (Fig. 4c) and the peak number is not consistent with the intrinsic and extrinsic pores derived from the crystal structure. PXRD analysis of the samples following gas adsorption showed that NUT-161 lost its crystallinity and turned amorphous (Fig. 3c), implying that NUT-161 disintegrated completely upon solvent removal. The observed amorphization of NUT-161 may be ascribed to the synergistic effects of enhanced porosity and reduced density of charge-assisted hydrogen bonding. The N₂ uptake capacity is likely to originate from the intrinsic porosity of the tetrahedral Zr-EDBA cages and the decreased interstitial space due to the inefficient packing of these cages.

To preserve the crystallinity of NUT-161, we treated NUT-161 with IDI for surface-confined polymerization. The resulting sample NUT-161@PolyIDI was checked by various characterizations to confirm the successful introduction of the PolyIDI onto the surface of NUT-161. FT-IR shows that NUT-161@PolyIDI has absorption peaks similar to those of NUT-161 (Fig. 3d). However, as expected, NUT-161@PolyIDI exhibits a new absorption peak around 2910 cm⁻¹, which can be attributed to the stretching vibrations of methylene or methyl groups on the polymer, indicating successful polymer incorporation. XPS analysis provided additional confirmation of the successful coating process (Fig. 3e–g). Upon the introduction of the polymer onto NUT-161, a distinct new peak appeared in the N 1s region at 400.4 eV. At the same time, the Zr 3d peaks of NUT-161 at 184.5 eV and 182.2 eV and the Cl 2p peak at 197.6 eV remained unchanged after the polymer coating, suggesting that polymer incorporation did not compromise the integrity of the NUT-161 structure. The morphology and elemental distribution of NUT-161@PolyIDI were further investigated by scanning electron microscopy (SEM) and energy-dispersive X-ray (EDX) analysis (Fig. 4d, e, S4, and S5). The SEM image shows that the composite material maintains a well-ordered cubic crystal structure and is covered by a thin layer of coarse polymers, whereas no distinct polymer layer was observed on the surface of NUT-161 (Fig. S4). EDX results indicate that the elements C, O, Cl, Zr, and N are evenly distributed throughout the sample (Fig. S5). In particular, the presence of the element N in the material is attributed to the surface-coated polyIDI, providing further evidence of successful polymer coating. Thermogravimetric analysis (TGA) was performed to evaluate the thermal stability of NUT-161@PolyIDI. The result shows that NUT-161@PolyIDI has similar weight loss trends compared to NUT-161 over the test temperature range (Fig. S6 and S7), demonstrating that both materials exhibit good thermal stability and the polymer coating does not significantly impact the thermal stability of the material. The content of PolyIDI was calculated to be about 3.09%. It can be concluded from the analysis as mentioned above that the surface-confined polymerization utilizing IDI to coat the SMFs is feasible.

To verify the polymerization mechanism of IDI on NUT-161, a model reactant p-tolyl isocyanate was selected. Blank control experiments showed that when pure p-tolyl isocyanate was dissolved in DCM and subjected to standard drying treatment, no reaction was observed by ¹H NMR analysis. However, upon the addition of NUT-161 with adsorbed water, characteristic peaks corresponding to urea groups appeared at 8.48 ppm in the ¹H NMR spectrum, confirming the occurrence of the reaction (Fig. S8). Significantly, p-tolyl isocyanate contains merely a single isocyanate group (–NCO), which, after one-step reaction, loses its reactive site and ultimately forms a soluble dimer structure. Conversely, IDI, which features dual-NCO groups, can theoretically undergo continuous polymerization to form extended polymeric networks (Fig. S9). NUT-161 can enrich the IDI monomer and direct the polymerization reaction. As the reaction progresses, the polymer that infiltrates the surface cavities of NUT-161 will gradually obstruct the ingress of monomers. This blockage induces a deceleration or even a complete arrest of the reaction, attributable to a passivation effect. Consequently, this phenomenon promotes the formation of a polyurea coating on the surface of NUT-161.

To verify whether isocyanate undergoes a chemical reaction with the NUT-161, the sample after model reaction investigation was analyzed by ¹H NMR analysis (Fig. S10). Comparative results showed no significant changes in the characteristic peak positions and integral areas of the zirconium cluster and organic ligands, while the two newly emerged peaks could be attributed to the urea product. The sample was also checked via HR-MS and the result is identical to the pristine NUT-161 (Fig. S11). These results further confirm the absence of significant chemical bonding interactions between PolyIDI and NUT-161. The binding of PolyIDI to the NUT-161 surface is primarily achieved through physical interactions.

The PolyIDI provided NUT-161 with improved properties. In particular, the crystallinity for activated NUT-161@PolyIDI was verified by PXRD patterns. Unlike activated NUT-161, the main peak positions for the activated NUT-161@PolyIDI are consistent with the simulated NUT-161 (Fig. 3c), confirming the improved crystallinity after surface-confined polymerization. The PolyIDI chains within the surface cavities of the NUT-161 provide support and thereby enhance the framework stability of NUT-161.⁴⁴ No new peaks appear compared to the simulated ones, suggesting that the PolyIDI is amorphous. The N₂ adsorption–desorption isotherms of NUT-161@PolyIDI showed a type IV isotherm, indicating the presence of mesopores (Fig. 4a). Specifically, the BET surface area and pore volume of NUT-161@PolyIDI were 2213 m² g⁻¹ and 1.43 cm³ g⁻¹, reflecting an increase of 324% and 321%, respectively, compared to NUT-161 (Fig. 4b). The PSD of NUT-161@PolyIDI shows pore sizes around 2.1 nm and 2.7 nm, which are consistent with the intrinsic and extrinsic pores of NUT-161, confirming the preservation of the mesopores (Fig. 4c). The well-preserved high porosity and consistent PSD of NUT-161@PolyIDI after surface-confined polymerization indicate that the surface of the cubic crystals is passivated with PolyIDI and the initially formed PolyIDI inhibits the penetration of monomers into the pore of NUT-161. The SEM images provide further information that the cubic morphology was well-maintained for NUT-161@PolyIDI while NUT-161 was pulverized (Fig. 4d and e), indicating improved mechanical properties. The surface properties of NUT-161 and NUT-161@PolyIDI were then evaluated. The static water contact angle of NUT-161 was observed to be 0°, indicating a pronounced hydrophilicity in accordance with its ionic nature, which facilitates the absorption of moisture from the surrounding air, potentially leading to pulverization and dissolution of frameworks. On the other hand, NUT-161@PolyIDI exhibited a static water contact angle of 125.1°, signifying a shift to a hydrophobic nature upon polymer incorporation (Fig. 4f).

3.2. Adsorption performance

The above characterizations have shown that the extrinsic mesopores generated by the ordered tetrahedral cage packing are preserved upon the introduction of PolyIDI. To further investigate the effect of these mesopores on the adsorption performance of C₃H₆ and C₂H₄, we performed adsorption tests on both NUT-161 and NUT-161@PolyIDI at 298 and 273 K (Fig. 5a and b). The adsorption amounts of C₃H₆ and C₂H₄ on NUT-161 are 51.9 and 27.8 cm³ g⁻¹ at 298 K, respectively, while for NUT-161@PolyIDI, they are 131.9 and 32.4 cm³ g⁻¹, respectively. Interestingly, the C₂H₄ adsorption on NUT-161@PolyIDI remains almost unchanged, compared to NUT-161, while the C₃H₆ adsorption increases 2.5 times (Fig. 5c). Notably, NUT-161@PolyIDI achieves a high C₃H₆ adsorption capacity at 298 K, reaching values exceeding those of SMFs including HOFs, such as HOF-NBDA,⁴⁵ and also comparable to some of the most effective MOFs, including MAC-4 (127.0 cm³ g⁻¹),⁴⁶ Zn-BPZ-TATB (114.0 cm³ g⁻¹),⁴⁷ Cd-dtzip-H₂O (112.0 cm³ g⁻¹),⁸ UPC-33 (94.3 cm³ g⁻¹),⁴⁸ [Zn₂(oba)₂(dmimpym)](76.0 cm³ g⁻¹),⁴⁹ and NEM-7-Cu (75.5 cm³ g⁻¹).⁵⁰ The adsorption amounts of C₃H₆ and C₂H₄ on NUT-161 are 64.8 and 38.9 cm³ g⁻¹, respectively, while for NUT-161@PolyIDI, the adsorption amounts are 226.3 and 51.5 cm³ g⁻¹ at 273 K, respectively. The adsorption amounts of C₃H₆ and C₂H₄ on NUT-161@PolyIDI are 249% and 32% higher than those on NUT-161, respectively. This further demonstrates that the polymer coating improves the crystallinity of NUT-161 with well-maintained extrinsic mesopores of NUT-161, which significantly enhances its C₃H₆ adsorption capacity.^34,51 Additionally, the polyIDI component of the material, which consists of methyl, methylene, and urea groups, exhibits preferential interaction with C₃H₆ over C₂H₄. This preference is driven by the formation of C–H⋯O and C–H⋯N hydrogen bonds, as well as van der Waals interactions.⁴⁷ The presence of a higher number of hydrogen atoms in C₃H₆ facilitates these interactions, thereby enhancing its adsorption capacity. After multiple cycles, the adsorption capacity of C₃H₆ remained stable (Fig. S12). Meanwhile, the PXRD analysis of NUT-161@PolyIDI after gas adsorption exhibited distinct diffraction peaks similar to the simulated ones (Fig. S13) and the N₂ adsorption–desorption isotherms remained well-preserved (Fig. S14), suggesting efficient C₃H₆/C₂H₄ separation performance can be achieved. The coverage-dependent adsorption enthalpy (Q_st), derived from single–component isotherms at different temperatures, reveals that NUT-161 exhibits a Q_st value of 33.1 kJ mol⁻¹ for C₃H₆ at near-zero surface coverage, which is significantly higher than the corresponding value of 24.8 kJ mol⁻¹ for C₂H₄ (Fig. S15). This result indicates a stronger interaction between the framework and C₃H₆ molecules. Importantly, NUT-161@PolyIDI not only demonstrates a higher adsorption capacity than NUT-161 but also exhibits lower Q_st values, specifically 22.1 kJ mol⁻¹ for C₃H₆ and 17.9 kJ mol⁻¹ for C₂H₄. These lower enthalpy values suggest that C₃H₆ desorption from NUT-161@PolyIDI can be achieved with reduced energy input, thereby improving the ease of adsorbent regeneration and maintaining its adsorption performance over repeated cycles.

We calculated the Δq of C₃H₆/C₂H₄ for both materials at 298 and 273 K. The Δq of C₃H₆/C₂H₄ for NUT-161 at 298 K is only 1.6 mmol g⁻¹, while that of NUT-161@PolyIDI reaches 2.7 mmol g⁻¹, which is a 69% increase compared to that of NUT-161 (Table S2 and Fig. 5d). This indicates that the presence of packed extrinsic mesopores is generally beneficial for the C₃H₆/C₂H₄ separation, confirming that the strategy of enhancing the C₃H₆/C₂H₄ separation performance by coating SMFs with polymers is successful. Notably, when the temperature is decreased to 273 K, the Δq of NUT-161 increases to 2.1 mmol g⁻¹, an improvement of 31%. In contrast, the Δq of NUT-161@PolyIDI increases by 85% and reaches 5.0 mmol g⁻¹. This indicates that NUT-161@PolyIDI has a more pronounced advantage in separating the C₃H₆/C₂H₄ mixture under low temperature conditions. The substantial increase in Δq for NUT-161@PolyIDI is consistent with the significant increase in C₃H₆ adsorption, which is considerably higher than that of other adsorbents (Table S3) such as Zn-BPZ-TATB (3.74 mmol g⁻¹),⁴⁷ NEM-7-Cu (2.8 mmol g⁻¹),⁵⁰ Zn-BPZ-SA (1.92 mmol g⁻¹),⁵² and srl-MOF (0.92 mmol g⁻¹).⁵³

To assess the feasibility of using the material in a fixed bed for the separation of C₃H₆ and C₂H₄, dynamic breakthrough experiments were conducted under controlled conditions (298 and 273 K, 1 bar). The experiments utilized a 50/50 (v/v) C₃H₆/C₂H₄ gas mixture, representative of typical MTO product compositions (Fig. 5e and f). For NUT-161, C₂H₄ breakthrough was observed at 60 min g⁻¹, followed by C₃H₆ at 100 min g⁻¹. Notably, NUT-161@PolyIDI demonstrated superior breakthrough performance. Specifically, C₂H₄ reached saturation and eluted at 85 min g⁻¹, while the breakthrough time for C₃H₆ was significantly extended to 155 min g⁻¹. This prolonged retention time for C₃H₆ relative to C₂H₄ in NUT-161@PolyIDI enabled an enhanced C₃H₆/C₂H₄ separation performance, which is consistent with theoretical predictions of Δq. It is worth mentioning that the separation performance of NUT-161@PolyIDI at 273 K showed a significant leap owing to the well-maintained mesopores facilitating C₃H₆ uptake while the performance of NUT-161 was almost unchanged. Furthermore, the breakthrough results remained almost unchanged after three consecutive cycles, indicating excellent stability and recyclability of NUT-161@PolyIDI. These findings underscore the ability of the material not only to maintain its performance over repeated use but also achieve robust separation of C₃H₆ and C₂H₄.

Overall, surface-confined polymerization endows NUT-161@PolyIDI with improved C₃H₆ adsorption uptake, Δq, and breakthrough performance. The improved separation performance suggests that surface-confined polymerization is a feasible strategy for constructing adsorbents with improved separation performance.

3.3. Proposed adsorption mechanism

The crystal structure of porous materials not only determines their experimental adsorption and separation properties but also enables the elucidation of adsorption mechanisms through computational simulations like GCMC, which is essential for guiding the synthesis of high-performance adsorptive separation materials. Due to the improved crystallinity conferred by polymer coating and the low polymer content, GCMC simulations were conducted on NUT-161 to investigate the adsorption sites and interaction discrepancies between these gas molecules and the framework at the molecular level (Fig. 6a–f). Both intrinsic and extrinsic pores accommodate C₂H₄ and C₃H₆ molecules and the molecular population of both gases in the extrinsic pores increases with pressure increment across the studied pressure range (0.1, 0.4, and 1.0 bar), indicating the well-preserved extrinsic pores are beneficial for gas uptake. A denser molecular population is observed for C₃H₆ compared to C₂H₄ and we attribute this difference to relatively stronger host–guest interactions experienced by C₃H₆ as corroborated by the Q_st of adsorption (Fig. S15). The preferential adsorption sites of both gases are predominantly situated near the edge-vertex basket within the tetrahedral Zr-EDBA cages and within the interstitial regions formed by linkers in the packing extrinsic cage. The sites within the intrinsic tetrahedral cage exhibit notably higher adsorption enthalpies (−24.20 kJ mol⁻¹ for C₂H₄ and −30.64 kJ mol⁻¹ for C₃H₆) compared to those in the larger cage (−11.78 kJ mol⁻¹ for C₂H₄ and −18.03 kJ mol⁻¹ for C₃H₆). We envisage that the higher adsorption affinity of C₃H₆ in both adsorption sites results from enhanced C–H⋯π interactions. Especially, the C₃H₆ molecule is bound to three aromatic rings originating from three organic ligands within the tetrahedral cages, through five C–H⋯π interactions with the distances of 2.596–4.241 Å and bound to two aromatic rings and two –CC– bonds originating from two organic ligands within the large cages, through four C–H⋯π interactions with the distances of 3.213–3.602 Å (Fig. S16). In contrast, the C₂H₄ molecule only shows four C–H⋯π interactions with the distances of 2.853–3.373 Å with three aromatic rings within the tetrahedral cages and four C–H⋯π interactions with the distances of 3.023–3.888 Å with two aromatic rings and two –CC– bonds within the large cages (Fig. S17). As a result, the non-polar pore surfaces of NUT-161 lead to the formation of multiple supramolecular interactions between C₃H₆ molecules and the framework, resulting in a strong affinity for C₃H₆.

4. Conclusions

In summary, we for the first time report a new SMF, NUT-161 of high porosity, and reinforce its stability through a surface-confined polymerization of isophorone diisocyanate (IDI) on the crystal surface of NUT-161. The resulting composite NUT-161@PolyIDI exhibits improved framework crystallinity with a 324% increment in specific surface area. NUT-161@PolyIDI shows improved C₃H₆ uptake with an increment of 249% at 273 K, compared to NUT-161 due to the preserved extrinsic mesopores, resulting in a Δq enhancement compared to NUT-161. Besides, the dynamic breakthrough studies demonstrate that NUT-161@PolyIDI exhibits a highly efficient separation capacity for C₃H₆/C₂H₄ separation and the separation performance can be well-maintained after three cycles. Furthermore, GCMC simulation was performed to illustrate the separation performance of NUT-161@PolyIDI owing to its well-preserved crystallinity and low content of surface-covered PolyIDI, and the results suggest that the framework exhibits multiple C–H⋯π interactions with C₃H₆ and preferentially binds to C₃H₆ over C₂H₄. The strategy demonstrated in this work may open an avenue to screen SMF-based adsorbents with good separation performance for applications in various scenarios.

Author contributions

Guoliang Liu: conceptualization, investigation, data curation, visualization, validation, writing – original draft, and writing – review & editing. Fan Li: data curation and visualization. Ze-Jiu Diao: validation. Hua-Dong Li: visualization. Xiaoyu Wu: Simulation. Hang-Ou Qi: validation. Lifeng Ding: Sources. Lin-Bing Sun: conceptualization, methodology, writing, writing – review & editing, funding acquisition, and supervision.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI.

Supplementary information available: Additional materials and methods with instrument details, separation potential calculation, GCMC simulation, ¹H NMR, XPS, SEM images, TGA, polymerization mechanism, adsorption cycles, PXRD, N₂ adsorption–desorption isotherms, Q_st, and comparison of performance with literature reports. See DOI: https://doi.org/10.1039/d5ta03617h.

Acknowledgements

We acknowledge the financial support of this work by the Natural Science Foundation of Jiangsu Province (BK20231270), the National Science Fund for Distinguished Young Scholars (22125804), and the National Natural Science Foundation of China (U24A20534).

References

1. R. J. Conk, S. Hanna, J. X. Shi, J. Yang, N. R. Ciccia, L. Qi, B. J. Bloomer, S. Heuvel, T. Wills, J. Su, A. T. Bell and J. F. Hartwig, Catalytic Deconstruction of Waste Polyethylene with Ethylene to Form Propylene, Science, 2022, 377, 1561–1566 CrossRef CAS PubMed .
2. Y. Wu, J. Han, W. Zhang, Z. Yu, K. Wang, X. Fang, Y. Wei and Z. Liu, Combined Strategies Enable Highly Selective Light Olefins and para-Xylene Production on Single Catalyst Bed, J. Am. Chem. Soc., 2024, 146, 8086–8097 CrossRef CAS PubMed .
3. J. Cui, Z. Zhang, L. Yang, J. Hu, A. Jin, Z. Yang, Y. Zhao, B. Meng, Y. Zhou, J. Wang, Y. Su, J. Wang, X. Cui and H. Xing, A Molecular Sieve with Ultrafast Adsorption Kinetics for Propylene Separation, Science, 2024, 383, 179–183 CrossRef CAS PubMed .
4. H. Zeng, M. Xie, T. Wang, R.-J. Wei, X.-J. Xie, Y. Zhao, W. Lu and D. Li, Orthogonal-Array Dynamic Molecular Sieving of Propylene/Propane Mixtures, Nature, 2021, 595, 542–548 CrossRef CAS PubMed .
5. L.-P. Guo, R.-S. Liu, J. Qian, G.-P. Hao, J. Guo, H. Wu, F. Wang and A.-H. Lu, Surface Sieving Carbon Skins for Propylene and Propane Separation, Nat. Chem. Eng., 2024, 1, 411–420 CrossRef .
6. R.-B. Lin, Z. Zhang and B. Chen, Achieving High Performance Metal–Organic Framework Materials through Pore Engineering, Acc. Chem. Res., 2021, 54, 3362–3376 CrossRef CAS PubMed .
7. J. Li, X. Han, X. Kang, Y. Chen, S. Xu, G. L. Smith, E. Tillotson, Y. Cheng, L. J. McCormick McPherson, S. J. Teat, S. Rudić, A. J. Ramirez-Cuesta, S. J. Haigh, M. Schröder and S. Yang, Purification of Propylene and Ethylene by a Robust Metal–Organic Framework Mediated by Host–Guest Interactions, Angew. Chem., Int. Ed., 2021, 60, 15541–15547 CrossRef CAS PubMed .
8. L. Zhang, R.-C. Gao, X. Liu, B. Zhang, L. Hou and Y.-Y. Wang, Unique H₂O Vortex in One New MOF Material Strengthening C₃H₆ Adsorption and Unprecedented Purification from C₂H₄/C₃H₈/C₃H₆ Mixtures, Adv. Funct. Mater., 2025, 35, 2420927 CrossRef CAS .
9. W. Fan, X. Wang, X. Zhang, X. Liu, Y. Wang, Z. Kang, F. Dai, B. Xu, R. Wang and D. Sun, Fine-Tuning the Pore Environment of the Microporous Cu-MOF for High Propylene Storage and Efficient Separation of Light Hydrocarbons, ACS Cent. Sci., 2019, 5, 1261–1268 CrossRef CAS PubMed .
10. Z. Di, Z. Ji, C. Chen, R. Krishna, D. Yuan, M. Hong and M. Wu, Efficient Separation of Methanol-to-Olefins Products Using a Metal-Organic Framework with Supramolecular Binding Sites, Chem. Eng. J., 2024, 493, 152442 CrossRef CAS .
11. R. Das, H. Li, H. A. Evans, Z. Deng, D. Zhao and A. K. Cheetham, Hydrophobic Metal–Formate Composites for Efficient CO₂ Capture, J. Am. Chem. Soc., 2025, 147, 8377–8385 CrossRef CAS PubMed .
12. Y. Han, L. Wang, Y. Zhang and B. Chen, Metal–Organic Frameworks for One-Step Ethylene Purification from Multi-Component Hydrocarbon Mixtures, Coord. Chem. Rev., 2025, 523, 216291 CrossRef CAS .
13. Y. Peng, H. Xiong, P. Zhang, Z. Zhao, X. Liu, S. Tang, Y. Liu, Z. Zhu, W. Zhou, Z. Deng, J. Liu, Y. Zhong, Z. Wu, J. Chen, Z. Zhou, S. Chen, S. Deng and J. Wang, Interaction-Selective Molecular Sieving Adsorbent for Direct Separation of Ethylene from Senary C₂-C₄ Olefin/Paraffin Mixture, Nat. Commun., 2024, 15, 625 CrossRef CAS PubMed .
14. H. Li, C. Chen, Q. Li, X. J. Kong, Y. Liu, Z. Ji, S. Zou, M. Hong and M. Wu, An Ultra-stable Supramolecular Framework Based on Consecutive Side-by-side Hydrogen Bonds for One-step C₂H₄/C₂H₆ Separation, Angew. Chem., Int. Ed., 2024, 63, e202401754 CrossRef CAS PubMed .
15. L. Zhang, B. Yu, M. Wang, Y. Chen, Y. Wang, L.-B. Sun, Y.-B. Zhang, Z. Zhang, J. Li and L. Li, Ethane Triggered Gate-Opening in a Flexible-Robust Metal-Organic Framework for Ultra-High Purity Ethylene Purification, Angew. Chem., Int. Ed., 2025, 64, e202418853 CrossRef CAS PubMed .
16. L. Wang, W. Sun, Y. Zhang, N. Xu, R. Krishna, J. Hu, Y. Jiang, Y. He and H. Xing, Interpenetration Symmetry Control Within Ultramicroporous Robust Boron Cluster Hybrid MOFs for Benchmark Purification of Acetylene from Carbon Dioxide, Angew. Chem., Int. Ed., 2021, 60, 22865–22870 CrossRef CAS PubMed .
17. Y. Jiang, Y. Hu, B. Luan, L. Wang, R. Krishna, H. Ni, X. Hu and Y. Zhang, Benchmark Single-Step Ethylene Purification from Ternary Mixtures by a Customized Fluorinated Anion-Embedded MOF, Nat. Commun., 2023, 14, 401 CrossRef PubMed .
18. Y. Jiang, J. Hu, L. Wang, W. Sun, N. Xu, R. Krishna, S. Duttwyler, X. Cui, H. Xing and Y. Zhang, Comprehensive Pore Tuning in an Ultrastable Fluorinated Anion Cross-Linked Cage-Like MOF for Simultaneous Benchmark Propyne Recovery and Propylene Purification, Angew. Chem., Int. Ed., 2022, 61, e202200947 CrossRef CAS PubMed .
19. S. Du, J. Lyu, Z. Yu, K. Su, W. Wang, X. Zhang and D. Yuan, Improving the Cellular Internalization of Zr(IV) Nanocages by Tuning Hydrophilicity and Lipophilicity, Angew. Chem., Int. Ed., 2025, 64, e202419602 CrossRef CAS PubMed .
20. X.-Q. Guo, P. Yu, L.-P. Zhou, S.-J. Hu, X.-F. Duan, L.-X. Cai, L. Bao, X. Lu and Q.-F. Sun, Low-Symmetry Coordination Cages Enable Recognition Specificity and Selective Enrichment of Higher Fullerene Isomers, Nat. Synth., 2025, 4, 359–369 CrossRef CAS .
21. S. Bai and Y.-F. Han, Metal-N-Heterocyclic Carbene Chemistry Directed toward Metallosupramolecular Synthesis and Beyond, Acc. Chem. Res., 2023, 56, 1213–1227 CrossRef CAS PubMed .
22. M. K. Dinker, K. Zhao, Z. Dai, L. Ding, X.-Q. Liu and L.-B. Sun, Porous Liquids Responsive to Light, Angew. Chem., Int. Ed., 2022, 61, e202212326 CrossRef CAS PubMed .
23. Y. Liu, H.-Y. Jin, M.-M. Li, M. Zuo, M. Kumar Dinker, J. Kou, J. Yan, L. Ding and L.-B. Sun, Improving Light-Responsive Efficiency of Type II Porous Liquid by Tailoring the Functionality of Host, Angew. Chem., Int. Ed., 2025, 64, e202501191 CrossRef CAS PubMed .
24. Y.-L. Lu, Y.-P. Wang, K. Wu, M. Pan and C.-Y. Su, Activating Metal–Organic Cages by Incorporating Functional M(ImPhen)₃ Metalloligands: From Structural Design to Applications, Acc. Chem. Res., 2024, 57, 3277–3291 CrossRef CAS PubMed .
25. X. Tang, C. Meng, N. Rampal, A. Li, X. Chen, W. Gong, H. Jiang, D. Fairen-Jimenez, Y. Cui and Y. Liu, Homochiral Porous Metal–Organic Polyhedra with Multiple Kinds of Vertices, J. Am. Chem. Soc., 2023, 145, 2561–2571 CrossRef CAS PubMed .
26. D. Luo, Z.-J. Yuan, L.-J. Ping, X.-W. Zhu, J. Zheng, C.-W. Zhou, X.-C. Zhou, X.-P. Zhou and D. Li, Tailor-Made Pd_nL_2n Metal-Organic Cages through Covalent Post-Synthetic Modification, Angew. Chem., Int. Ed., 2023, 62, e202216977 CrossRef CAS PubMed .
27. C.-W. Zhou, X.-Z. Wang, M. Xie, R.-Q. Xia, D. Luo, Z.-X. Lian, G.-H. Ning, W. Lu, X.-P. Zhou and D. Li, A Self-Assembled Capsule for Propylene/Propane Separation, Angew. Chem., Int. Ed., 2023, 62, e202315020 CrossRef CAS PubMed .
28. E. Han, Y.-Q. Li, T. Wu, Q. Bai, Z. Zhang, J. Yuan, W. Liu, D. Liu, Y. Li and P. Wang, Chiral Metal Self-Assemblies of Zirconium-Tetrahedra and Their Second Harmonic Generation Activity, Angew. Chem., Int. Ed., 2025, 64, e202420223 CrossRef CAS PubMed .
29. Z.-M. Cao, G.-L. Li, Z.-Y. Di, C. Chen, L.-Y. Meng, M. Wu, W. Wang, Z. Zhuo, X.-J. Kong, M. Hong and Y.-G. Huang, From a Metal–Organic Square to a Robust and Regenerable Supramolecular Self-assembly for Methane Purification, Angew. Chem., Int. Ed., 2022, 61, e202210012 CrossRef CAS PubMed .
30. W. He, Y. Yu, K. Iizuka, H. Takezawa and M. Fujita, Supramolecular Coordination Cages as Crystalline Sponges through a Symmetry Mismatch Strategy, Nat. Chem., 2025, 17, 653–662 CrossRef CAS PubMed .
31. S. Tokuda and S. Furukawa, Three-Dimensional van der Waals Open Frameworks, Nat. Chem., 2025, 17, 672–678 CrossRef CAS PubMed .
32. Z. Yang, S. B. Peh, S. Xi, Y. Lu, Q. Liu and D. Zhao, Packing Engineering of Zirconium Metal-Organic Cages in Mixed Matrix Membranes for CO₂/CH₄ Separation, Angew. Chem., Int. Ed., 2025, 64, e202418098 CrossRef CAS PubMed .
33. T. Yang, X.-M. Li, Y. Liu, C. Wei, C. Gu, M. Zuo, T. Guo, G. Liu, L. Ding, X.-Q. Liu and L.-B. Sun, Improving Stability, Crystallinity, and Photo-Responsiveness of Supramolecular Frameworks by Surface Polymerization, Adv. Funct. Mater., 2024, 34, 2404869 CrossRef CAS .
34. L.-Z. Qin, X.-H. Xiong, S.-H. Wang, L. Zhang, L.-L. Meng, L. Yan, Y.-N. Fan, T.-A. Yan, D.-H. Liu, Z.-W. Wei and C.-Y. Su, MIL-101-Cr/Fe/Fe-NH₂ for Efficient Separation of CH4 and C3H8 from Simulated Natural Gas, ACS Appl. Mater. Interfaces, 2022, 14, 45444–45450 CrossRef CAS PubMed .
35. Y. Jiang, J. Park, P. Tan, L. Feng, X.-Q. Liu, L.-B. Sun and H.-C. Zhou, Maximizing Photoresponsive Efficiency by Isolating Metal–Organic Polyhedra into Confined Nanoscaled Spaces, J. Am. Chem. Soc., 2019, 141, 8221–8227 CrossRef CAS PubMed .
36. T. Yang, C. Yu, C. Gu, Y. Liu, H. Wen, X.-Q. Liu, G. Liu, L. Ding and L.-B. Sun, Visible Light-Mediated Supramolecular Framework for Tunable CO₂ Adsorption, Chem. Eng. J., 2024, 486, 150254 CrossRef CAS .
37. G. Liu, Z. Ju, D. Yuan and M. Hong, In Situ Construction of a Coordination Zirconocene Tetrahedron, Inorg. Chem., 2013, 52, 13815–13817 CrossRef CAS PubMed .
38. G. Liu, Z. Yang, M. Zhou, Y. Wang, D. Yuan and D. Zhao, Heterogeneous Postassembly Modification of Zirconium Metal–Organic Cages in Supramolecular Frameworks, Chem. Commun., 2021, 57, 6276–6279 RSC .
39. M. Zhou, G. Liu, Z. Ju, K. Su, S. Du, Y. Tan and D. Yuan, Hydrogen-Bonded Framework Isomers Based on Zr-Metal Organic Cage: Connectivity, Stability, and Porosity, Cryst. Growth Des., 2020, 20, 4127–4134 CrossRef CAS .
40. G. Liu, M. Zhou, K. Su, R. Babarao, D. Yuan and M. Hong, Stabilizing the Extrinsic Porosity in Metal–Organic Cages-Based Supramolecular Framework by In Situ Catalytic Polymerization, CCS Chem., 2020, 3, 1382–1390 CrossRef .
41. Z. Xiao, H. F. Drake, Y. H. Rezenom, P. Cai and H.-C. Zhou, Structural Manipulation of a Zirconocene-Based Porous Coordination Cage Using External and Host–Guest Stimuli, Small Struct., 2022, 3, 2100133 CrossRef CAS .
42. G. Liu, X. Zhang, Y. Di Yuan, H. Yuan, N. Li, Y. Ying, S. B. Peh, Y. Wang, Y. Cheng, Y. Cai, Z. Gu, H. Cai and D. Zhao, Thin-Film Nanocomposite Membranes Containing Water-Stable Zirconium Metal–Organic Cages for Desalination, ACS Mater. Lett., 2021, 3, 268–274 CrossRef CAS .
43. D. Nam, J. Kim, E. Hwang, J. Nam, H. Jeong, T.-H. Kwon and W. Choe, Multivariate Porous Platform Based on Metal-Organic Polyhedra with Controllable Functionality Assembly, Matter, 2021, 4, 2460–2473 CrossRef CAS .
44. L. Peng, S. Yang, S. Jawahery, S. M. Moosavi, A. J. Huckaba, M. Asgari, E. Oveisi, M. K. Nazeeruddin, B. Smit and W. L. Queen, Preserving Porosity of Mesoporous Metal–Organic Frameworks through the Introduction of Polymer Guests, J. Am. Chem. Soc., 2019, 141, 12397–12405 CrossRef CAS PubMed .
45. Y. Zhou, C. Chen, Z. Ji, R. Krishna, Z. Di, D. Yuan and M. Wu, A Layered Hydrogen-Bonded Organic Framework with C₃H₆-Preferred Pores for Efficient One-Step Purification of Methanol-to-Olefins (MTO) Products, ACS Mater. Lett., 2024, 6, 1388–1395 CrossRef CAS .
46. G.-D. Wang, Y.-Z. Li, R. Krishna, W.-Y. Zhang, L. Hou, Y.-Y. Wang and Z. Zhu, Scalable Synthesis of Robust MOF for Challenging Ethylene Purification and Propylene Recovery with Record Productivity, Angew. Chem., Int. Ed., 2024, 63, e202319978 CrossRef CAS PubMed .
47. G.-D. Wang, Y.-Z. Li, W.-J. Shi, L. Hou, Y.-Y. Wang and Z. Zhu, Active Sites Decorated Nonpolar Pore-Based MOF for One-step Acquisition of C₂H₄ and Recovery of C₃H₆, Angew. Chem., Int. Ed., 2023, 62, e202311654 CrossRef CAS PubMed .
48. W. Fan, Y. Wang, Q. Zhang, A. Kirchon, Z. Xiao, L. Zhang, F. Dai, R. Wang and D. Sun, An Amino-Functionalized Metal-Organic Framework, Based on a Rare Ba₁₂(COO)₁₈(NO₃)₂ Cluster, for Efficient C₃/C₂/C₁ Separation and Preferential Catalytic Performance, Chem.–Eur. J., 2018, 24, 2137–2143 CrossRef CAS PubMed .
49. Y.-Z. Li, G.-D. Wang, R. Krishna, Q. Yin, D. Zhao, J. Qi, Y. Sui and L. Hou, A Separation MOF with O/N Active Sites in Nonpolar Pore for One-Step C₂H₄ Purification from C₂H₆ or C₃H₆ Mixtures, Chem. Eng. J., 2023, 466, 143056 CrossRef CAS .
50. X. Liu, C. Hao, J. Li, Y. Wang, Y. Hou, X. Li, L. Zhao, H. Zhu and W. Guo, An Anionic Metal–Organic Framework: Metathesis of Zinc(II) with Copper(II) for Efficient C₃/C₂ Hydrocarbon and Organic Dye Separation, Inorg. Chem. Front., 2018, 5, 2898–2905 RSC .
51. D. Lv, Z. Liu, F. Xu, H. Wu, W. Yuan, J. Yan, H. Xi, X. Chen and Q. Xia, A Ni-Based Metal-Organic Framework with Super-High C₃H₈ Uptake for Adsorptive Separation of Light Alkanes, Sep. Purif. Technol., 2021, 266, 118198 CrossRef CAS .
52. G.-D. Wang, R. Krishna, Y.-Z. Li, Y.-Y. Ma, L. Hou, Y.-Y. Wang and Z. Zhu, Rational Construction of Ultrahigh Thermal Stable MOF for Efficient Separation of MTO Products and Natural Gas, ACS Mater. Lett., 2023, 5, 1091–1099 CrossRef CAS .
53. H. Fang, B. Zheng, Z.-H. Zhang, H.-X. Li, D.-X. Xue and J. Bai, Ligand-Conformer-Induced Formation of Zirconium–Organic Framework for Methane Storage and MTO Product Separation, Angew. Chem., Int. Ed., 2021, 60, 16521–16528 CrossRef CAS PubMed .

This journal is © The Royal Society of Chemistry 2025