Strongly coupled and highly-compacted zirconium aminobenzenedicarboxylate crystal membranes for accelerating carbon dioxide capture

Abstract

This study presents the fabrication of a UiO-66-NH₂ composite membrane on a porous carbon cloth–polypyrrole substrate (CC–PPy@UiO-66-NH₂) that demonstrates strongly coupled and highly ordered features to enhance CO₂ capture efficiency. This free-standing configuration exhibits a higher adsorption kinetics with a slope of 0.48 mg g⁻¹ min⁻¹ for CO₂ and superior adsorption selectivity, demonstrating that the adsorption kinetics rate of the CC–PPy@UiO-66-NH₂ composite membrane for CO₂ is approximately twice that of bulk UiO-66-NH₂ (0.27 mg g⁻¹ min⁻¹) crystals at a bar.

---

Porous membrane separation technologies offer high throughput, selectivity, and separation efficiency, as well as low energy consumption, positioning them as green and efficient solutions for molecular separation. However, traditional organic/inorganic porous membranes face limitations owing to their fixed pore sizes, complex medium environments, material stability, etc., often failing to achieve effective separation in many applications. Metal–organic frameworks (MOFs) have excellent thermal stability, customizable pore structures, and permanent internal voids, making them uniquely suitable for molecular separation.^1–3 However, MOFs, particularly in nanocrystal form, are typically produced powders with high surface energy, which considerably limits their development and practical application in the field of separation.

The development of MOF hybrid membranes has progressed beyond the design and exploration of powder and single-crystal materials. These membranes integrate the excellent selectivity and permeability of MOF crystals, addressing the challenge of low powder utilization.⁴ The primary challenge now lies in preparing robust highly ordered hybrid membranes that leverage the strong coupling and directional characteristics of individual MOF crystals. Advances in this regard have led to the production of membranes with closely packed structures and stable crystal orientations.^5–7 By controlling these orientations, the MOF channels within the crystals can be aligned in specific directions, endowing MOF membranes with anisotropic properties.^8–10 These properties include enhanced diffusion along preferred directions, preferential orientation of guest molecules, and protection of functional guest molecules. As depicted in Scheme 1, the coupling configuration, interactions, and arrangement density of MOF crystals on the surface of porous substrates considerably influence their separation performance. The configurations of MOF crystals can be categorized into three types: (a) strongly coupled and highly ordered, resulting in high selectivity and permeability; (b) uneven size and random arrangement, leading to low selectivity and permeability; and (c) weakly coupled and loosely arranged, also resulting in low selectivity and permeability.

Scheme 1 Schematic of several porous substrate-supported MOF composite membranes.

UiO-66, a distinguished member of the Zr-MOF family with a three-dimensional topology, has emerged as a promising candidate for membrane-based CO₂ separation and utilization owing to its abundant and interconnected pore channels, high chemical affinity for CO₂, and exceptional thermal stability.^11–15 It has been demonstrated that reducing the thickness of UiO-66 membranes considerably enhances the CO₂/N₂ separation performance, particularly in terms of dynamic kinetics.^16,17 Furthermore, downsizing UiO-66 crystals further improves the CO₂/N₂ separation performances.^18,19 However, the fabrication of high-performance UiO-66 hybrid membranes, which maintain small size and octahedral shape, face many challenges, such as weak interactions between crystals and substrates, irregular arrangement, and a loose state. These disadvantages impede further enhancement of CO₂ selectivity and kinetics. To overcome these challenges, designing a new synthesis route for achieving strongly coupled and highly ordered UiO-66 hybrid membranes has become indispensable. Despite numerous efforts devoted to the configuration and arrangement of UiO-66 crystals supported on porous substrates, advancement with respect to the development of hybrid membranes exhibiting high efficiency remains stagnant.^20–22

Over the past few decades, in situ growth based on chemical grafting has proven considerably effective for fabricating highly coupled and ordered MOF hybrid membranes through strong covalent bonding and surface confinement effects.^23–25 For example, Tan et al. synthesized carboxyl-functionalized cotton fabric-supported Cu–BTC crystals via a layer-by-layer synthesis route through covalent bonding between Cu–BTC and modified cotton fiber.²⁶ Furthermore, our group has demonstrated that ligands (1,4-benzenedicarboxylic acid, BDC) can be grafted onto the N sites of polypyrrole (PPy) via nucleophilic substitution reactions, resulting in the formation of covalently bonded UiO-66 anchored on PPy.¹⁷ Inspired by these achievements, we expected that strongly coupled and ordered UiO-66 composite membranes may be obtained through a two-step nucleophilic substitution reaction based on bridged molecules. Benefiting from the small-size effect of UiO-66-NH₂ nanocrystals, the as-prepared single-layer UiO-66-NH₂ composite membrane with a superior adsorption kinetics rate was demonstrated in the CO₂ capture test. To the best of our knowledge, taking advantage of the covalent grafting strategy to fabricate strongly coupled and highly ordered UiO-66-NH₂ composite membranes has not yet been reported.

In this study, we fabricated a strongly coupled and highly ordered carbon cloth-supported UiO-66-NH₂ composite membrane with an octahedral shape and small size (Fig. S1–S3, ESI†). The details for the synthesis route of the CC–PPy@UiO-66-NH₂ membrane are provided in the Experimental section. To characterize the microstructure of the in situ grown UiO-66-NH₂ crystals, the pristine CC and bulk UiO-66-NH₂ (denoted as B-UiO-66-NH₂), synthesized using the solvothermal method, were characterized via SEM. As shown in Fig. 1a and b, the CC comprises numerous fibers with an average size of 10 μm. B-UiO-66-NH₂ crystals display a typical octahedral shape with an average size exceeding 300 nm (Fig. 1c). To assess the effect of chemical grafting on UiO-66-NH₂ crystals, we fabricated CC/UiO-66-NH₂ membranes using unmodified CC under the same conditions. These membranes exhibited a black appearance (Fig. 1d). The magnified SEM images presented in Fig. 1e and f reveal an irregular and loose arrangement of unmodified CC-supported UiO-66-NH₂ crystals having an average size of 200 nm. Fig. 1g shows the CC–PPy@UiO-66-NH₂ hybrid membrane, wherein 1,4-dibromobutane ligand molecules were covalently grafted onto the CC–PPy composite substrate. The SEM images presented in Fig. 1h and i display octahedral-shaped UiO-66-NH₂ crystals (100 nm) orderly arranged on the surface of the CC composite substrate, demonstrating that covalent grafting positively effects the growth and configuration of UiO-66-NH₂ crystals.

Fig. 1 Morphological characterizations of CC, unmodified CC/UiO-66-NH₂, covalently grafted CC–PPy@UiO-66-NH₂, and bulk UiO-66-NH₂ (denoted as B-UiO-66-NH₂). (a) Digital photograph and (b) SEM image. (c) SEM image of B-UiO-66-NH₂. (d) Digital photograph and (e) and (f) SEM images. (g) Digital photograph and (h) and (i) SEM images.

The crystal phase and high crystallinity of the as-prepared hybrid membrane were verified via X-ray diffraction (XRD) patterns (Fig. 2a), wherein two characteristic peaks observed at 6° and 8° could be indexed to the (111) and (002) planes, respectively, indicating that chemical grafting and subsequent in situ growth exerted no adverse effects on the UiO-66-NH₂ crystals. Specific surface area is a critical parameter for evaluating the physical properties of hybrid membranes or textiles. Fig. 2b illustrates the N₂ sorption isotherms measured at 77 K. B-UiO-66-NH₂ crystals exhibited a type I isotherm with a hysteresis loop at P/P₀ = 0.85–1.0, indicative of a micro–mesoporous structure with a pore size distribution ranging from 1.5 to 5 nm (Fig. S4, ESI†). The BET surface area of B-UiO-66-NH₂ was determined to be 1258 m² g⁻¹, underscoring its potential for gas separation applications. CC–PPy@UiO-66-NH₂ demonstrated a BET surface area of 48.5 m² g⁻¹, significantly higher than that of pristine CC (2.4 m² g⁻¹), which highlights the substantial contribution of UiO-66-NH₂ nanocrystals to the overall BET surface area of the membrane. With an actual loading of 7 mg of UiO-66-NH₂ in the 215 mg membrane, the adjusted BET surface area of the UiO-66-NH₂ nanocrystals was calculated to be 1280 m² g⁻¹, with an average pore size of 2 nm.

Fig. 2 (a) X-ray diffraction patterns of pristine CC, B-UiO-66-NH₂, and CC–PPy@UiO-66-NH₂. (b) Specific surface area of pristine CC, B-UiO-66-NH₂, and CC–PPy@UiO-66-NH₂ measured at 77 K. (c) The adsorption performances and (d) the adsorption kinetic rate of pristine CC, B-UiO-66-NH₂, and CC–PPy@UiO-66-NH₂ for 0.5 mM methyl orange (MO) solution, respectively.

The thermal stability of the CC–PPy@UiO-66-NH₂ was assessed using thermogravimetric analysis (TGA) under N₂ flow. Both the CC–PPy@UiO-66-NH₂ hybrid membrane and B-UiO-66-NH₂ exhibited minimal weight loss from room temperature (25 °C) up to 300 °C (Fig. S5, ESI†), confirming their exceptional thermal stability. The adsorption performance of the CC–PPy@UiO-66-NH₂ membrane was evaluated using a 0.5 mM methyl orange (MO) solution, showing that both CC–PPy@UiO-66-NH₂ and B-UiO-66-NH₂ have excellent adsorption capabilities for MO (Fig. 2c). Notably, the adsorption kinetics of the CC–PPy@UiO-66-NH₂ membrane reached 5 μM s⁻¹, approximately four times faster than that of B-UiO-66-NH₂, which was 1.3 μM s⁻¹ (Fig. 2d). Due to the robust structure of the CC, the CC–PPy@UiO-66-NH₂ hybrid membrane demonstrated remarkable flexibility, capable of enduring 360° bending and multiple twisting actions without damage (Fig. S6, ESI†).

The CO₂ capture and separation performance of the CC–PPy@UiO-66-NH₂ hybrid membrane was assessed using adsorption kinetics curves and a vacuum vapor/gas sorption method at various temperatures. To elucidate the contributions of UiO-66-NH₂ nanocrystals grown in situ on carbon fibers, this hybrid membrane is referred to as M-UiO-66-NH₂, and the corresponding data were normalized based on the weight percentage of UiO-66-NH₂. Fig. 3a presents the CO₂ adsorption isotherms for three samples measured at 273 K. It was found that the pristine CC exhibited the lowest CO₂ adsorption capacity (0.2 cm³ g⁻¹), which can be attributed to its low porosity and insufficient adsorption sites. In contrast, M-UiO-66-NH₂ demonstrated a high CO₂ uptake capacity of 98 cm³ g⁻¹, comparable to that of B-UiO-66-NH₂ (105 cm³ g⁻¹). This indicates that the covalently grafted UiO-66-NH₂ nanoparticles retain the complete topological structure and abundant –NH₂ sites, contributing almost entirely to the total CO₂ adsorption capacity of the hybrid membrane. As shown in Fig. 3b, both the M-UiO-66-NH₂ and B-UiO-66-NH₂ displayed relatively low N₂ adsorption capacities at 273 K, which is capable of improving the CO₂/N₂ separation potential. Notably, M-UiO-66-NH₂ showed higher ideal adsorbed solution theory (IAST) selectivity compared to B-UiO-66-NH₂, particularly at higher pressures with an IAST of 35.2, which is advantageous for accelerating CO₂/N₂ separation (Fig. 3c). The superior IAST selectivity of M-UiO-66-NH₂ at high pressures can be attributed to the ordered arrangement of UiO-66-NH₂ nanocrystals, which not only restricts N₂ diffusion into the pores, but also enhances stronger quadrupolar interactions with CO₂.¹⁸ This structural arrangement has been shown to confer a preferential binding affinity toward CO₂, resulting in higher IAST selectivity under elevated pressures.

Fig. 3 CO₂ separation performance of the B-UiO-66-NH₂ and CC–PPy@UiO-66-NH₂ hybrid membrane (denoted as M-UiO-66-NH₂). (a) CO₂ adsorption isotherms of the three samples. (b) Details of CO₂ and N₂ adsorption isotherms measured at 273 K, and (c) the corresponding CO₂/N₂ IAST selectivity for B-UiO-66-NH₂ and M-UiO-66-NH₂. (d) CO₂ adsorption isotherms of M-UiO-66-NH₂ at 273, 298, and 323 K. (e) Adsorption capacity–time (adsorption kinetics) curves, and (f) statistical adsorption kinetics curves (note: averages of adsorption capacity for each pressure value) for B-UiO-66-NH₂ and M-UiO-66-NH₂ using a vacuum vapor/gas sorption method at 273 K.

The CO₂ adsorption capacities of M-UiO-66-NH₂ at various temperatures were measured to evaluate its practicality. As shown in Fig. 3d, M-UiO-66-NH₂ exhibited a high CO₂ uptake capacity of 52 cm³ g⁻¹ at 298 K (room temperature). Even at 323 K, M-UiO-66-NH₂ also maintained a relatively high CO₂ adsorption capacity of 41 cm³ g⁻¹, surpassing that of B-UiO-66-NH₂, which was 40 cm³ g⁻¹, as shown in Fig. S7 (ESI†). It is recognized that the size of MOF nanoparticles significantly influences their gas separation performance, particularly the gas adsorption kinetics.¹⁶ In this study, the vacuum vapor/gas sorption method was employed to analyze the adsorption kinetics of the hybrid membrane. It was observed that both M-UiO-66-NH₂ and B-UiO-66-NH₂ demonstrated increased CO₂ adsorption capacities with the increasing of relative pressure (Fig. S8, ESI†), confirming their excellent CO₂ capture performance. Fig. 3e displays the CO₂ adsorption capacity–time curves across eight pressure values. Compared to B-UiO-66-NH₂, M-UiO-66-NH₂ reached system weight equilibrium more rapidly at each relative pressure level, attributed to the faster diffusion of CO₂ into the pores of the smaller UiO-66-NH₂ framework. This rapid diffusion is likely due to reduced physical or chemical resistance encountered by CO₂ molecules entering the pores, facilitating quicker adsorption equilibrium in the smaller M-UiO-66-NH₂ crystals. Further analysis is shown in Fig. 3f, where the average adsorption capacity–time curves are plotted. The slope of these curves represents the adsorption kinetics, with M-UiO-66-NH₂ exhibiting a slope of 0.48 mg g⁻¹min⁻¹, higher than the 0.27 mg g⁻¹min⁻¹ observed for B-UiO-66-NH₂. These results indicate that the CC–PPy@UiO-66-NH₂ hybrid membrane with small-sized nanocrystals not only retains the intrinsic adsorption capacity but also achieves faster adsorption equilibrium compared to bulk UiO-66-NH₂ crystals.

To evaluate the stability of this hybrid membrane under open flame conditions, infrared thermal imaging was used to monitor temperature changes on the side far from the fire source (denoted as the nightside), using a lighted candle as the fire source (see details in Fig. S9, S10 and Videos S1–S4, ESI†). These results indicate that densely packed UiO-66-NH₂ nanocrystals effectively reduced the heat transfer rate through the carbon fibers in the CC. Furthermore, we anticipate that the CC–PPy@UiO-66-NH₂ hybrid membrane will be effective for CO₂ separation in flue gases and for preventing the ingestion of thick smoke at fire scenes.

It has been well accepted that the spatial orientation of MOF crystals plays a substantial role in regulating the separation efficiency of MOF hybrid membranes. To investigate the effect of bridged molecules on the spatial orientation of UIO-66-NH₂ crystals, molecular dynamics simulation, based on the molecular plane model of PPy, was adopted to elucidate the relationships between molecular structure and spatial orientation (Fig. S11, ESI†). These results indicate that brominated alkanes with short chains and asymmetry can regulate the spatial orientation of the BDC-NH₂ molecule. The above theoretical calculations can be used to guide the synthesis of MOF crystal membranes in future work.

In summary, we have demonstrated that the strongly coupled and highly ordered CC–PPy@UiO-66-NH₂ hybrid membrane was fabricated using a universal covalent grafting method. The as-prepared hybrid membrane exhibited excellent CO₂ capture performances (high IAST selectivity and fast adsorption kinetics) while maintaining intrinsic adsorption capacity. Benefiting from its structural superiority, this hybrid membrane exhibited good separation performance for thick smoke, highlighting its considerable potential application in fire accidents.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and its ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank the Jiangsu Specially-Appointed Professor Program and the National Natural Science Foundation of China (22402090) and Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJCX24_1991) and Large Instruments Open Foundation of Nantong University (KFJN2458) for financial support.

Notes and references

1. A. Betard and R. A. Fischer, Chem. Rev., 2012, 112, 1055–1183.
2. T. Wen, G. Quan, B. Niu, Y. Zhou, Y. Zhao, C. Lu, X. Pan and C. Wu, Small, 2021, 17, 2005064.
3. Q. Li, F. Yan and J. Texter, Chem. Rev., 2024, 124, 3813–3931.
4. J. Yan, Y. Sun, T. Ji, C. Zhang, L. Liu and Y. Liu, J. Membrane Sci., 2022, 653, 120496.
5. D. L. Gin and R. D. Noble, Science, 2011, 332, 674–676.
6. S. Hermes, F. Schröder, R. Chelmowski, C. Wöll and R. A. Fischer, J. Am. Chem. Soc., 2005, 127, 13744.
7. H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke and J. Caro, J. Am. Chem. Soc., 2009, 131, 16000–16001.
8. S. Panda, S. Kundu, P. Malika and R. Haldar, Chem. Sci., 2024, 15, 2586–2592.
9. H. Sun, F. Wang, X. Li, J. Caro, H. Meng, N. Wang and Q.-F. An, Angew. Chem., Int. Ed., 2023, 62, e202300262.
10. W.-J. Li, J. Liu, Z.-H. Sun, T.-F. Liu, J. Lü, S.-Y. Gao, C. He, R. Cao and J.-H. Luo, Nat. Commun., 2016, 7, 11830.
11. S. Friebe, B. Geppert, F. Steinbach and J. Caro, ACS Appl. Mater. Interface, 2017, 9, 12878–12885.
12. F. Liang, S. Wei, L. Xue and S. Yan, Green Carbon, 2024, 2, 252–261.
13. S. Fakhraie, H. Rajabi and A. Rashidi, Chem. Eng. J., 2023, 452, 139001.
14. J. Zhu, Y. Zhang, Z. Chen, Z. Zhang, X. Tian, M. Huang, X. Bai, X. Wang, Y. Zhu and H. Jiang, Nat. Commun., 2024, 15, 1565.
15. M. Chen, K. Chang, Y. Zhang, Z. Zhang, Y. Dong, X. Qiu, H. Jiang, Y. Zhu and J. Zhu, Angew. Chem., Int. Ed., 2023, 62, e202305530.
16. L. Luo, C. Xu, W. Shi, Q. Liu, Y. Ou-Yang, J. Qian, Y. Wang and Q. Li, J. Phys. Chem. Lett., 2023, 14, 8437–8443.
17. Q. Li, H. Zhu, Y. Tang, P. Zhu, H. Ma, C. Ge and F. Yan, Chem. Commun., 2019, 55, 12108–12111.
18. Y. Sun, J. Yan, Y. Gao, T. Ji, S. Chen, C. Wang, P. Lu, Y. Li and Y. Liu, Angew. Chem., Int. Ed., 2024, 62, e202216697.
19. G. Liu, Y. Guo, C. Chen, Y. Lu, G. Chen, G. Liu, Y. Han, W. Jin and N. Xu, Nat. Mater., 2023, 22, 769–776.
20. S. Rasoul, S. Mahmoud and P. Vahid, Polym. Sci., Ser. A, 2022, 64, 781–793.
21. X. Li, G. Jiang, M. Jian, C. Zhao, J. Hou, A. W. Thornton, X. Zhang, J. Z. Liu, B. D. Freeman, H. Wang, L. Jiang and H. Zhang, Nat. Commun., 2023, 14, 286.
22. L.-C. Jheng, J. Park, H. W. Yoon and F.-C. Chang, Sep. Purif. Technol., 2023, 310, 123–126.
23. K. Barcus, P.-A. Lin, Y. Zhou, G. Arya and S. M. Cohen, ACS Nano, 2022, 16, 18168–18177.
24. M. Taddei, G. M. Schukraft, M. E. A. Warwick, D. Tiana, M. J. McPherson, D. R. Jones and C. Petit, J. Mater. Chem. A, 2019, 7, 23781–23786.
25. X. Pei and W. Song, Langmuir, 2023, 39, 15046–15054.
26. W. Li, Y. Zhang, Z. Yu, T. Zhu, J. Kang, K. Liu, Z. Li and S. C. Tan, ACS Nano, 2022, 16, 14779–14791.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cp03259d

---