Abstract
Bifunctionalized NH2-UiO-66-CF3 nanofillers were successfully prepared by grafting - CF3 on NH2-UiO-66 via aldehyde-amine condensation reaction and applied to the preparation of 6FDA-ODA/NH2-UiO-66-CF3 mixed matrix membranes (MMMs). The successful introduction of -NH2 and -CF3 groups was verified by FT-IR, XRD, and XPS characterization, and the integrity of the crystal structure was ensured. Due to the presence of hydrogen bonding, NH2-UiO-66-CF3 showed good interfacial compatibility with the polymer matrix. Due to the introduction of the - CF3 group, the MMMs exhibited excellent thermal stability (T10% > 500°C, Tg = 300°C–305°C), hydrophobicity (contact angle of 92.96–97.71), and mechanical properties (tensile strength of 74–85 MPa). The gas separation performance tests showed that the MMMs exhibited significantly improved gas permeability and selectivity compared to the pure 6FDA-ODA membranes, which was attributed to the fact that the -NH2 and -CF3 groups provided more adsorption sites for gases. In particular, the PI/NUF(10) MMMs showed the best performance, with permeabilities of 42.71, 9.22, and 92.79 Barrer for CO2, O2, and He, respectively, which were 248%, 213%, and 173% higher than those of the pure membranes. Meanwhile, the selectivities of CO2/N2, O2/N2 and He/N2 were improved by 37%, 24%, and 8%, respectively. Close to the Robeson upper limit, demonstrating excellent gas separation potential.
Introduction
The extensive use of non-renewable fossil fuels (e.g., coal, petroleum, and natural gas) has resulted in increasingly severe issues of atmospheric pollution and energy scarcity. As a pivotal technology to address these challenges, membrane-based gas separation has gained widespread application in air separation, natural gas purification, and hydrogen recovery owing to its advantages of high efficiency, low energy consumption, and environmental friendliness.1–5 In membrane separation processes, material selection is critical, with options including inorganic, polymeric, and mixed matrix membranes (MMMs).6–13 In order to break through the bottleneck of organic polymer membranes constrained by the Robeson upper bound,14,15 and overcome the shortcomings of difficult and costly membrane fabrication of inorganic materials, researchers have increasingly focused on MMMs incorporating organic polymers and inorganic porous fillers. The introduction of fillers not only disrupts the close packing of polymer chains and increases the free volume, which in turn enhances membrane permeability,16,17 but also forms interfacial pores and nanocavities in MMMs, providing fast transport channels for gases. MMMs combine the excellent mechanical properties and processability of organic membranes with the high permeability and selectivity of inorganic fillers, demonstrating significant industrial potential. However, poor interfacial compatibility between fillers and polymer matrices remains a major challenge hindering commercial implementation.
Metal-organic frameworks (MOFs) are crystalline porous materials formed by the coordination between metal nodes and organic ligands, exhibiting distinct advantages including high porosity, large specific surface area, and tunable functionality.18–20 Compared with conventional inorganic fillers (e.g., zeolites, graphene, and carbon nanotubes),21–23 MOFs demonstrate superior compatibility with polymer matrices attributed to their organic ligands, rendering them ideal candidates for MMMs fabrication. Particularly, UiO-66 with zirconium metal node demonstrates exceptional chemical and thermal stability while maintaining structural integrity under harsh conditions, positioning it as a preferred material for composite membrane fabrication.24,25 However, the incorporation of crystalline MOFs into polymers often induces interfacial defects due to mismatched phase compatibility. To address this issue, ligand functionalization strategies have been developed.26,27 Amino-functionalized MOFs serve dual roles: (1) providing alkaline sites for enhanced CO2 adsorption, and (2) forming hydrogen bonds with polymer polar groups to strengthen interfaces. Similarly, fluorinated MOFs concurrently enhance MMMs hydrophobicity, gas selectivity, and thermal stability.28,29 Notably, Pu et al. 30 reported that the NUS-8-NH2/PIM-1 MMM achieved a CO2 permeability of 14,000 Barrer with CO2/N2 selectivity of 30, surpassing the 2008 Robeson upper bound. This enhancement originates from -NH2 groups acting as both CO2 transport promoters and interfacial linkers via hydrogen bonding with PIM’s cyano groups. Zhou et al. 31 further demonstrated that UiO-66-(CF3)2 incorporation improved both separation performance and hydrothermal stability of PIM-1 membranes. These cases validate that MOFs functionalization effectively enhance membrane performance through tailored interfacial engineering. Nevertheless, single-functionalized MOFs exhibit certain limitations in gas separation performance enhancement. Bifunctional MOFs offer multiple advantages including additional gas adsorption sites, improved selectivity and adsorption capacity for various gas species, as well as enhanced membrane stability under diverse operating conditions. Park et al. 32 developed a mixed-ligand strategy to prepare amine- and fluorine-co-functionalized MIL-101(Cr). This bifunctional material exhibited significantly enhanced CO2 separation performance due to its dual adsorption sites. Remarkably, the co-functionalized MIL-101(Cr) retained excellent separation efficiency even in a humid environment, confirming its practical application potential.
In this work, NH2-UiO-66 with optimal pore size is first synthesized and the dual-functionalized NH2-UiO-66-CF3 filler is subsequently prepared by introducing -CF3 groups into the MOF through an aldehyde-amine condensation reaction where -CF3 partially substitutes -NH2 groups. The modified MOF is then incorporated into the 6FDA-ODA matrix, a benchmark material for gas-separation membranes, to prepare 6FDA-ODA/NH2-UiO-66-CF3 MMMs with different MOF loadings. The MMM structures are systematically characterized by FT-IR and XRD, while their thermal stability, hydrophobicity, mechanical properties and gas separation performance are comprehensively evaluated. These analyses reveal the crucial role of NH2-UiO-66-CF3 bifunctionalization in enhancing membrane characteristics, particularly gas separation efficiency, and clarify the underlying mechanisms.
Experimental
Materials
4,4′-(hexafluoroisopropylidene)diphthalic anhydride (6FDA, purity > 98%, Adamas), 4,4′-oxydianiline (ODA, 98%, Adamas), m-Cresol (99%, Shanghai Aladdin Co., Ltd), isoquinoline (97%, Shanghai Aladdin Co., Ltd), Zirconium (IV) chloride (ZrCl4, purity > 98%, Adamas), 2-Aminoterephthalic Acid (BDC-NH2, purity >98%, Adamas), 4-(Trifluoromethyl)Benzaldehyde (4-TFMB, purity > 98%, Adamas), N,N-dimethylformamide (DMF, Sinopharm Chemical Reagent Co., Ltd). Acetic acid (CH3COOH) and ethanol (EtOH), methanol (MeOH) (Xilong Chemical Co., Ltd). All solvents are analytical grade and can be used without further treatment. CO2, O2, N2 and He (Liuzhou Hongrun Gas Co., Ltd.) are experimental gases with purity greater than 99.9%.
Preparation of 6FDA-ODA
6FDA (4.44 g, 0.01 mol) and ODA (2.00 g, 0.01 mol) at a 1:1 molar ratio are placed in a 100.0 mL three-necked flask equipped with mechanical stirring and a condenser. M-Cresol (58.0 mL) is added, and the reaction system is heated to 50°C under nitrogen atmosphere. After complete mixing, isoquinoline (0.3 mL) is added as the catalyst. The thermal imidization proceeds via stepwise heating: first at 80°C for 3 h, then at 120°C for 3 h, followed by 180°C for 3 h, and finally at 200°C for 12 h. Upon reaction completion, the system is cooled to room temperature, the condenser water and nitrogen flow are stopped, and the solution is poured slowly into methanol for precipitation. The precipitated product is collected by filtration and dried under vacuum at 150°C for 12 h to yield fibrous polyimide.
Synthesis of MOFs
The synthesis of NH2-UiO-66 powder followed a literature method with modifications. 33 First, ZrCl4 (1 mmol) and BDC-NH2 (1 mmol) were dispersed in a mixed solvent (20.0 mL of DMF and 5.0 mL of acetic acid, total volume 25.0 mL) via ultrasonication, followed by stirring until a homogeneous mixture was obtained. The solution was then transferred to a Teflon-lined autoclave and heated at 120°C for 24 h. After cooling to room temperature, the solid product was collected by centrifugation and washed repeatedly with DMF and methanol. Finally, the powder was dried at 120°C for 24 h to yield NH2-UiO-66.
NH2-UiO-66-CF3 was synthesized via post-synthesis modification (PSM) 34 as follows: first, NH2-UiO-66 was activated under vacuum at 150°C for 24 h to remove residual solvent molecules from the pores. Subsequently, 0.40 g of the activated NH2-UiO-66 was dispersed in 60.0 mL of ethanol via ultrasonication and stirred until homogeneous. Then, 3.0 mL of 4-trifluoromethylbenzaldehyde (4-TFMB) was added to the mixture, and the reaction was refluxed at 80°C for 12 h under nitrogen protection. After cooling to room temperature, the solid product was collected by centrifugation and washed thoroughly with methanol several times to remove unreacted reagents. Finally, the product was dried under vacuum at 80°C for 24 h to yield NH2-UiO-66-CF3 powder.
Fabrication of membranes
The gas separation membranes were fabricated by a solution-casting method (Scheme 1). Five membrane compositions with filler loadings of 0, 5, 8, 10 and 15 wt% were prepared. The detailed procedure was as follows: first, the NH2-UiO-66-CF3 filler was ultrasonically dispersed in DMF, and then the filler dispersion was added into the 6FDA-ODA solution dissolved in DMF, followed by prolonged ultrasonic stirring to ensure homogeneous dispersion. The resulting mixture was doctor-bladed onto a leveled glass substrate, with solvent evaporation at 80°C. After cooling to room temperature, the membrane was peeled off using lukewarm water. Finally, the membrane was vacuum-dried at 150°C for 12 h. The mixed matrix membranes are hereafter denoted as PI/NUF(x), where x represents the filler loading (calculated using equation (1)). Synthesis of 6FDA-ODA and NH2-UiO-66-CF3; Preparation of PI/NUF(x) mixed matrix membranes.
Characterizations
The crystalline properties of MOF fillers and membranes were characterized by X-ray diffraction (XRD, X’Pert PRO X, PANalytical B.V.). Fourier transform infrared spectrometry (FT-IR, Nicolette 6700-NXR, Thermo Nicolet Corporation) and X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, USA Thermo Electron Corporation) were used to characterize the MOF fillers and membrane structural properties. Thermal properties were evaluated through thermogravimetric analysis (TGA, STA 449 C, NETZSCH) and differential scanning calorimetry (DSC, DSC 214, NETZSCH) under N2 atmosphere (10°C/min heating rate, TGA: RT-800°C, DSC: 30°C–400°C). The MOF filler was vacuum degassed at 150°C for 12 h, and then the N2 adsorption-desorption isotherms of the filler were measured by a specific surface area and pore size analyzer (TriStar II 3020, Micromeritics). The specific surface area of the sample was calculated using the Brunauer-Emmett-Teller (BET) method. The micromorphology of MOF powder as well as membrane was observed by scanning electron microscopy (SEM, Hitachi S-4800, Hitachi). Surface hydrophobicity was assessed through water contact angle measurements (JY PHb, Chengde Youte Instrument Co.). An electronic universal testing machine (CMT2103, MTS Corporation) with a maximum tensile load of 5 kN was used to test the mechanical properties of the films.
Pure gas separation evaluation
The gas permeability coefficients (P) of pure CO2, O2, He, and N2 were measured by the constant volume/variable pressure method (Labthink VAC-V1, Jinan, China) at 35°C with 0.4 MPa feed pressure. Prior to testing, the dried membrane was mounted between the test chambers. After tightening the upper chamber, the system underwent sequential evacuation: (1) lower chamber evacuation for 60 s, followed by (2) full-system evacuation for ≥3 h to achieve required vacuum levels. The upper chamber was then pressurized with test gas to establish a constant transmembrane pressure difference, driving gas permeation from the high- to low-pressure side. Triplicate measurements were conducted for each membrane under identical conditions to ensure data reproducibility, with results averaged to minimize experimental error.
Permeability (P) is calculated as follows:
The ideal selectivity (αA/B) of a pair of gases is the ratio of their permeability coefficients, which can be expressed as:
Results and discussion
Synthesis and structural characterization of NH2-UiO-66-CF3
Figure 1 illustrated the XRD patterns of NH2-UiO-66 and NH2-UiO-66-CF3. Both samples exhibited diffraction profiles matching the simulated patterns,
35
and prominent characteristic peaks appeared at the (111) and (002) crystal planes (2θ = 7.4°, 8.5°), indicating that the crystal structure of NH2-UiO-66 was successfully synthesized. After trifluoromethylation, NH2-UiO-66-CF3 maintained the parent topology while preserving high crystallinity. SEM characterization (Figure S1) revealed well-defined octahedral nanoparticles (average size: 200 nm) with sharp edges and intact surfaces, demonstrating structural preservation during trifluoromethylation.
36
The high consistency of this topological morphology combined with XRD data conclusively demonstrated that the trifluoromethyl modification process did not lead to the reconstruction of the MOF crystal structure, providing a material basis for the subsequent interface compatibility study. XRD spectra of NH2-UiO-66, NH2-UiO-66-CF3.
In the FTIR spectrum (Figure S2), NH2-UiO-66 and NH2-UiO-66-CF3 exhibited asymmetric and symmetrical stretching vibration peaks of - C = O at 1690 cm−1 and 1380 cm−1, as well as longitudinal and transverse vibration peaks of Zr-O bonds at 768 cm−1 and 669 cm−1. Notably, in the NH2-UiO-66-CF3 curve, the -NH2 characteristic absorption peak near 3500 cm-1 was significantly weakened, and a new -CF3 characteristic absorption peak appeared at 1060 cm−1, which indicated that some -NH2 groups were replaced by -CF3. In addition, the full spectrum of NH2-UiO-66-CF3 and the fine spectrum of elements were measured by XPS to further determine its chemical structure composition. It can be seen from the wide-scan XPS spectrum of Figure 2(a) that NH2-UiO-66-CF3 is composed of five elements: C, N, O, F and Zr, which also indicated that -CF3 was successfully grafted onto NH2-UiO-66. Figure 2(b) showed that the high-resolution XPS spectra of C 1s were convoluted into C-C (284.40 eV), C=N (285.76 eV), O-C=O (288.83 eV) and C-F (292.96 eV) bonds. The N 1s spectrum (Figure 2(c)) showed characteristic peaks for -NH2 (399.43 eV) and C=N (400.43 eV). The presence of the C-F bond was further confirmed by the F 1s peak at 688.07 eV (Figure 2(d)). All experimental results showed that 4-TFMB was successfully grafted onto NH2-UiO-66 through aldehyde amine condensation reaction, while the -NH2 group did not completely disappear, and finally the bifunctional NH2-UiO-66-CF3 was obtained. XPS of NH2-UiO-66-CF3 (a) full spectrum; (b) C 1s; (c) N 1s; (d) F 1s.
The N2 adsorption and desorption curves of both NH2-UiO-66 and NH2-UiO-66-CF3 (Figure 3) exhibited type I and type IV isotherms, indicating that hierarchical porous MOF materials with co-existing microporous and mesoporous sites were successfully prepared by the acetic acid modifier. The presence of defects led to the exposure of more unsaturated sites, which enhanced its potential application in gas separation. The BET specific surface area of NH2-UiO-66 was 648 m2 g−1 (Table S1), representing a significant increase compared to the literature sample without acetic acid (575 m2 g−1).
37
This improvement was mainly attributed to the strategy of adjusting the concentration of acetic acid to control the particle size, which effectively increased the specific surface area of the MOF. The slight decrease in the specific surface area and microporous volume of NH2-UiO-66-CF3 compared to NH2-UiO-66 was attributed to the modification of 4-TFMB that resulted in the clogging of some of the pores. However, F atoms were shown to enhance host-guest interactions, leading to significantly higher adsorption capacities for CO2 and CH4 than NH2-UiO-66.
33
Additionally, the reduced pore volume (VMicro = 0.23 cm3 g−1) of NH2-UiO-66-CF3 created pore channels more favorable for He (2.60 Å) diffusion, making it a highly promising porous material for He separation. N2 adsorption-desorption isotherms of NH2-UiO-66 and NH2-UiO-66-CF3.
The effect of functionalization modifications on the thermal stability of the materials was systematically evaluated by thermogravimetric analysis (TGA) (Figure 4). Both MOF fillers, NH2-UiO-66 and NH2-UiO-66-CF3, exhibited a three-stage weight loss: the first stage (30°C–150°C) corresponded to the loss of free water molecules, the second stage was attributed to the removal of solvent guest molecules (e.g., DMF) and bond energy changes induced by - NH2/-CF3 group, and the final stage at about 350°C for the weight loss due to the combustion of 2-aminoterephthalic acid and decomposition of the MOF structure.
38
NH2-UiO-66-CF3 exhibited a significantly enhanced thermal stability, with a higher decomposition onset temperature and a slower weight loss rate during framework decomposition compared to the parent material. This enhanced effectivity of thermal stability was attributed to the strong electronegativity of trifluoromethyl groups in synergy with Zr-O bonds. The difference in the second weight loss region further confirmed the successful grafting of -CF3 groups, providing thermodynamic stability for composite membranes under high-temperature conditions. TGA curves of NH2-UiO-66 and NH2-UiO-66-CF3.
Synthesis and characterization of PI/NH2-UiO-66-CF3 MMMs
The membrane structure was analyzed by XRD pattern (Figure 5). The pure 6FDA-ODA matrix membrane showed a typical amorphous broadening diffraction peak at 2θ = 15°, while all PI/NUF(x) MMMs clearly displayed sharp characteristic peaks of the MOF fillers at 7.4°(111) and 8.5°(002) while retaining this feature.
39
As the NH2-UiO-66-CF3 loading increased from 5 wt% to 15 wt%, the characteristic peak intensity of MOF increased, confirming the successful incorporation of functionalized UiO-66 fillers without crystal structure degradation during MMMs preparation. The FT-IR spectra of the membranes were shown in Figure 6, with asymmetric and symmetric stretching vibration peaks of C=O at 1784 cm−1 and 1727 cm−1, stretching vibration peaks of C-N, C-O, and C-F at 1380 cm−1, 1154 cm−1 and 1113 cm−1, respectively, and a bending vibration peak of the imide ring at 723 cm−1, which confirmed the successful synthesis of high-temperature polyimide after thermoimination.40,41 Notably, the C=O vibrational peak of 1784 cm−1 gradually weakened with higher filler content, likely due to hydrogen bonding between -NH2 of MOF nanoparticles and C=O of PI.
42
Additionally, the appearance of a Zr-O stretching vibration at 660 cm−1, with intensity proportional to filler content, further evidenced NH2-UiO-66-CF3 incorporation into the polymer. XRD curves of 6FDA-ODA and PI/NUF/(x) MMMs. FT-IR spectra of 6FDA-ODA and PI/NUF/(x) MMMs.

The dispersion of MOF in polymer solution greatly influenced the gas separation performance of MMMs. SEM analysis of 6FDA-ODA and composite membrane cross-sections at different loadings revealed that all MMMs exhibited a concave morphology (Figure 7), characteristic of PI polymer chain accumulation. Octahedral MOF particles were dispersed in the MMMs, and the number of dispersed particles increased with higher loadings of NH2-UiO-66-CF3. However, agglomeration became noticeable at elevated loadings, slightly reducing filler-polymer affinity. Despite the progressively rougher cross-sectional morphology, no interfacial defects were observed, which was attributed to the fact that the presence of bifunctional groups in NH2-UiO-66-CF3 enhanced the interaction with PI, thereby improving the interfacial compatibility. SEM cross-section images of (a) pristine 6FDA-ODA membrane and the PI/NUF(x) MMMs with the loading of (b) 5 wt%, (c) 8 wt%, and (d) 10 wt%, (e) 15 wt%, (f) partial enlargement of c.
The thermal stability of 6FDA-ODA and PI/NUF(x) membranes was characterized by TGA, as illustrated in Figure 8. The TGA curves of the composite and matrix membranes displayed similar two-stage degradation: the loss of solvent and decomposition of NH2-UiO-66-CF3 up to 500°C, and the decomposition and the carbonization of the PI in the second phase. The maximum weight loss rates of all MMMs occurred above 500°C and the residual carbon rates were higher than 50% (Table S2), indicating that the MMMs have good thermal stability. However, PI/NUF(15) showed a slightly lower char yield (52%) than the matrix membrane (55%), attributed to the aggregation of NH2-UiO-66-CF3 in the polymer membrane at higher filler loadings, potentially weakening the filler-PI interactions.
43
This phenomenon was further verified by DSC test. As shown in Figure 9, the glass transition temperature (Tg) varied between 299°C and 305°C, first increasing and then decreasing. All MMMs had only one Tg, while the addition of fillers limited the movement of polymer chains, resulting in higher Tg for all MMMs than pure membranes.
44
Although PI/NUF (15) showed a decrease in Tg due to the weakening of polymer-filler interactions caused by filler agglomeration,
45
there was no phase separation, indicating good interfacial compatibility between NH2-UiO-66-CF3 and PI. TGA curves of 6FDA-ODA and PI/NUF(x) MMMs. DSC curves of 6FDA-ODA and PI/NUF(x) MMMs.

The water contact angle test (θw) was performed on the pure membrane 6FDA-ODA and its MMMs to evaluate hydrophobicity, where larger θw values indicated greater membrane hydrophobicity. As shown in Figure 10, 6FDA-ODA membrane (θw = 98.85°) exhibited inherent hydrophobicity due to its -(CF3)2 groups.
46
With the increase of the bifunctionalized NH2-UiO-66-CF3 loading, θw showed a gradual decrease, but the decrease slowed down significantly, only slightly from 97.71 to 92.96. This was attributed to the strong hydrophobic effect of -CF3 in the MOF filler partially offsetting the hydrophilicity of -NH2. However, the hydrophobic enhancement effect was somewhat limited by the insufficient substitution of the -CF3 group. Notably, all PI/NUF(x) membranes maintained θw > 90°, demonstrating superior water-resistant properties for gas separation applications. This characteristic effectively minimized water molecule interference, thereby enhancing gas separation efficiency and long-term stability. Contact angles of 6FDA-ODA and PI/NUF(x) MMMs.
The mechanical properties of the 6FDA-ODA matrix membrane and its mixed matrix membranes (MMMs) were evaluated through tensile testing (Figure 11). The tensile strengths of all membranes were in the range of 74–94 MPa, the elongation at break was in the range of 6.07%–7.24%, and the Young’s modulus was in the range of 1.54–1.91 GPa (Table S3). The MMMs showed slightly reduced tensile strength and elongation at break compared to the pure 6FDA-ODA membrane, primarily due to the incorporation of rigid NH2-UiO-66-CF3 filler that both restricted polymer chain mobility and disrupted chain regularity. However, the Young’s modulus of the MMMs increased compared to the 6FDA-ODA membrane, because of the formation of strong hydrogen bonds between - NH2 and the polymer, as well as hydrophobic interactions, dipole-dipole interactions and van der Waals forces between -CF3 and 6FDA-ODA. These intermolecular interactions restricted polymer chain mobility, resulting in enhanced film rigidity.47,48 Among them, PI/NUF(10) membrane exhibited the largest Young’s modulus, indicating that the MMM containing 10 wt% NH2-UiO-66-CF3 had the most excellent interfacial interaction force. Stress-strain curves of 6FDA-ODA and PI/NUF(x) MMMs.
Gas separation performance
The permeability of the prepared membranes to four pure gases, CO2, O2, He, and N2, was measured using the constant volume-variable pressure method at 35°C and 0.4 MPa feed pressure. The trend of gas permeability versus selectivity was observed in Figure 12. He had the largest permeability among the four gases, because it had the smallest molecular size (2.60 Å) and the highest diffusion coefficient, which facilitated transport through the membrane’s microporous structure. With the increase of NH2-UiO-66-CF3 content, the gas permeability gradually improved, mainly attributed to the MOF filler introducing additional pores that enhanced gas diffusion while simultaneously increasing the membrane’s free volume through disruption of molecular chain stacking, thereby resulting in smoother gas transport.
49
The permeation performance of the composite membranes reached a maximum when the loading of NH2-UiO-66-CF3 reached 10 wt%, with permeabilities of 42.71 Barrer for CO2, 9.22 Barrer for O2, 92.79 Barrer for He, and 1.77 Barrer for N2 (Table 1). Compared to the pure matrix membrane 6FDA-ODA, the permeability coefficients of CO2 were enhanced by 248%, O2 by 213%, and He by 173%. Meanwhile the selectivities of CO2/N2, O2/N2 and He/N2 improved by 37%, 24% and 8%, respectively. This indicated that the appropriate amount of MOF packing could effectively enhance the permeability and selectivity of the membranes. However, when the filler loading was further increased to 15 wt%, the gas permeability showed a decreasing trend, which was attributed to the possible aggregation of excess NH2-UiO-66-CF3 in the membranes, leading to the blockage of some of the pores or the formation of disordered filler stacking, which in turn reduced the gas diffusion rate.
50
Nevertheless, the permeation rates of the PI/NUF (15) MMM still maintained significantly higher than the 6FDA-ODA matrix membrane, demonstrating that the composite membranes could preserve better permeability performance even at higher loadings. It was noteworthy that the gas selectivity of MMMs began to show a declining trend before the filler content reached 15 wt%. This phenomenon may be attributed to nanoparticle agglomeration during the filler loading process, resulting in uneven dispersion within the membrane. Such agglomeration could potentially create non-selective voids that preferentially increase N2 permeation, thereby reducing overall selectivity. (a) Gas permeability (b) gas selectivity of pure 6FDA-ODA membranes and PI/NUF(x) MMMs at different loadings. Comparison of the gas separation properties of MMMs reported in this work with those reported in the literature. Note. 1 Barrer = 10−10 [cm3(STP)·cm]/(cm2s·cmHg), Ideal Selectivity α = P(A)/P(B), A and B are two different pure gases.
In addition, the introduction of NH2-UiO-66-CF3 significantly enhanced CO2 permeability and CO2/N2 selectivity. This enhancement primarily resulted from strong interactions between CO2 molecules and polar groups (-NH2 and -CF3) in the MOF fillers. The N, O, and F elements provided alkaline active sites for CO2 adsorption, while C-F and N-H bonds generated dipole-quadrupole interactions with CO2 molecules, substantially increasing CO2 solubility in the membrane.38,51 Thus, despite CO2 having a larger kinetic diameter than He, the MMMs with 10 wt% MOF loading demonstrated optimal enhancement in both CO2 permeability and CO2/N2 selectivity compared to the pristine 6FDA-ODA matrix membrane, showing remarkable improvement of 248% and 37%, respectively. By optimizing the NH2-UiO-66-CF3 loading, the composite membranes achieved not only enhanced overall gas separation performance but also superior CO2 separation efficiency while maintaining satisfactory mechanical properties, demonstrating great potential for gas separation applications.
Figure 13 demonstrated the gas separation performance of PI/NUF(x) MMMs in comparison with other reported mixed matrix membranes relative to the Robeson upper bound. The results showed that incorporating NH2-UiO-66-CF3 into 6FDA-ODA significantly enhanced both gas permeability and selectivity, with performance gradually approaching the Robeson upper bound. Among the tested membranes, PI/NUF(10) membrane exhibited the most prominent performance for CO2/N2, He/N2, and He/CO2 gas pairs, closely approaching the upper bound, while PI/NUF(8) membrane is infinitely close to the upper bound for O2/N2 separation. These performance variations resulted from differences in gas physicochemical properties, including molecular size, solubility parameters, and diffusion coefficients. Compared with literature reports (Table 1), this study successfully overcame the traditional trade-off effect through bifunctionalized NH2-UiO-66-CF3 incorporation, achieving simultaneous enhancement in both permeability and selectivity. These findings demonstrated that bifunctionalized MOF fillers represented an effective strategy for optimizing MMMs performance and showed significant potential for industrial gas separation applications. 6FDA-ODA, PI/NUF(x) MMMs of (a) CO2/N2, (b) O2/N2, (c) He/N2, (d) He/CO2 separation performance compared to the Robeson upper bound.
Conclusion
In this study, hierarchical porous NH2-UiO-66 with defects was first synthesized using an acetic acid modifier, followed by successful preparation of bifunctionalized NH2-UiO-66-CF3 nanoparticles through -CF3 grafting via an aldehyde-amine condensation reaction, achieving simultaneous coexistence of -NH2 and -CF3 groups. This bifunctional design conferred multiple advantages to the materials, as demonstrated through thermogravimetric analysis, contact angle measurements, and tensile tests, which confirmed that the PI/NUF(x) MMMs exhibited good thermal stability, hydrophobicity and mechanical properties. The strong hydrogen bonding between -NH2 groups and 6FDA-ODA significantly enhanced the filler-matrix interfacial affinity, while the hierarchical pore structure effectively strengthened the molecular sieving effect. Furthermore, the combined effects of (1) dipole-quadrupole interactions between N-H/C-F bonds and CO2 molecules, and (2) Lewis basic active sites provided by N and F atoms, synergistically enhanced both permeability and selectivity for CO2, O2, and He in PI/NUF(x) MMMs, overcoming the traditional permeability-selectivity trade-off limitation. Through this “dual adsorption sites with hierarchical pores” synergistic mechanism, the bifunctionalized NH2-UiO-66-CF3 demonstrated significant potential for separation applications, offering an efficient new approach for carbon capture from flue gas and air separation processes.
Supplemental Material
Supplemental Material - Bifunctionalized NH2-UiO-66-CF3 mixed matrix membranes for gas separation
Supplemental Material for Bifunctionalized NH2-UiO-66-CF3 mixed matrix membranes for gas separation by Zhiqing Liang, Zhongying Liu, Xiuyuan Wei, Hui Wu, Chanjuan Liu and Xiaohua Huang in High Performance Polymers.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research: This work was supported by the National Natural Science Foundation of China (51963007), Guangxi Science and Technology Plan Project (GuikeAD25069100), Guangxi Natural Science Foundation (2021GXNSFAA075016, 2020GXNSFAA159009, 2020GXNSFAA159006), Independent Research Project of Guangxi Colleges and Universities Key Laboratory of Natural and Biomedical Polymer Materials (24POLY-AA-02), Opening Fund of Key Laboratory of New Processing Technology for Nonferrous Metal & Materials, Ministry of Education/Guangxi Key Laboratory of Optical and Electronic Materials and Devices (No. 22KF8).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
