Construction of Water Vapor Stable Ultramicroporous Copper-Based Metal–Organic Framework for Efficient CO₂ Capture

Abstract

It is quite essential to obtain an excellent CO₂ adsorption capacity, CO₂ adsorption selectivity and water vapor stability at the same time for practical CO₂ capture after combustion. Through the combination of ultramicropore and the high density of CO₂-philic sites without OMSs, an ultra-microporous Cu-based metal–organic framework has been designed and synthesized, featuring a high CO₂ capacity (99 cm³ g⁻¹ and 56.6 cm³ g⁻¹ at 273 K and 298 K, respectively), high selectivity over N₂ (118 at a scale of CO₂/N₂ 15/85, 298 K) and excellent water vapor stability, simultaneously. Theoretical calculations indicate that neighboring ketonic O atoms with suitable distance play vital roles in boosting CO₂ selective capture.

1. Introduction

With the development of the industry, human demand for energy is increasing, and fossil fuels are still our main source of energy today. This has led to a dramatic increase in greenhouse gas emissions in the atmosphere. The greenhouse effect caused by excessive emissions of greenhouse gases such as CO₂, CH₄, O₃, NO₂ and Freon is the main cause of global warming. Among them, CO₂, as the greenhouse gas with the highest concentration in the atmosphere, produces about 60% of the total greenhouse effect of all greenhouse gases. Over the past few decades, with the dramatic increase in the global population and rapid economic industrialization, energy consumption has also increased dramatically. Although the rapid development of technology has been accompanied by an increasing variety of energy sources, it is estimated that about 60% of global CO₂ emissions come from fossil fuel power generation [1]. In general, there are strategies to reduce CO₂ emissions in the environment through efficient energy use, the use of hydrogen and renewable energy as alternatives to fossil fuels, and the development of new carbon capture technologies. Thus, the installation of effective CO₂ capture systems in coal-fired or gas-fired power plants, which can selectively remove CO₂ in fluid gas, could greatly reduce global CO₂ emissions. The main CO₂ capture technologies currently used are chemisorption and physical adsorption [2]. Industrially, the current mainstream technology is chemisorption using organic amines. Although this technology is capable of absorbing 98% of CO₂ in the exhaust gas, it has disadvantages such as high energy consumption, large amounts of adsorbent and high corrosiveness. Compared with the traditional chemisorption method using organic amines, physical adsorption technology using porous materials exhibits a great prospect due to its facile regeneration process without corrosion problems [3,4,5,6,7,8,9].

Zeolite molecular sieves are a class of traditional microporous materials, and their microporous environment with a high density of Lewis polar sites gives these materials considerable CO₂ adsorption capacity and selectivity. For example, according to NaY [10], their CO₂ adsorption capacity can reach 5.5 mmol g⁻¹ at 305 K and 1 atm. According to Siriwardane [11], the CO₂ adsorption capacity and selectivity can be maximized by adjusting the specific surface area, metal type and content of the material framework. Unfortunately, zeolite molecular sieves have strong interactions with water molecules and preferentially adsorb water molecules in their presence, which is detrimental to CO₂ capture. Studies have shown that activated carbon has a stronger affinity to CO₂ compared to zeolite while interacting weaker with N₂, thus showing excellent selectivity for CO₂ gas. However, activated carbon is very sensitive to temperature changes, and its CO₂ capture capacity decreases sharply with increasing temperature, while CO₂/N₂ selectivity becomes poor as a result. Du [12] investigated the CO₂ adsorption isotherms of Ajax-activated carbon at three different temperatures at 10 kPa and found that the CO₂ adsorption capacity of this material decreased with increasing temperature. In recent years, a new class of porous material-metal organic frameworks (MOFs), also known as porous coordination polymers (PCPs) and porous coordination networks (PCNs), have been developed into one of the hot spots of research in chemistry and materials. Metal–organic frameworks (MOFs) are considered to be promising materials for CO₂ adsorption compared to inorganic porous materials. The material is an organic–inorganic hybrid material consisting of metal centers (metal ions or metal clusters, SUB) that are self-assembled with organic ligands through coordination bonds to form crystalline porous materials with a periodic network structure. Metal–organic frameworks (MOFs) have attracted a lot of attention as CO₂ capture adsorbents due to their advantages, such as high crystallinity, the precise modularity of their pore size and easy functionality of the pore walls [13,14,15,16].

A high CO₂ adsorption capability, CO₂ adsorption selectivity and water vapor stability are very important prerequisites for ideal adsorbents; however, it is not easy to obtain all of these benefits at the same time. For example, various kinds of open metal sites (OMSs) have been widely introduced into the framework as CO₂-philic sites to boost CO₂ capture capability, and this has afforded an exceptionally high CO₂ adsorption capacity and high selectivity for CO₂ in some famous MOFs such as the MOF-74 series [17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]. However, the activated OMSs could rapidly reabsorb water molecules, thus inevitably resulting in a remarkable decline in CO₂ adsorption capacity in the presence of water vapor, which is commonly seen in fluid gas. Meanwhile, the linkage between the metal junction and the ligand usually becomes fragile with the presence of water vapor, causing concerns about structural stability [32,33,34,35]. For another, ligands with polar groups such as hydroxyl, free carboxyl and uncoordinated N atom, etc., have been adopted to build MOFs with saturated coordinated metal ions (that is, no OMSs) to alleviate the negative effects imposed by water vapor, while the CO₂ selectivity of MOFs bearing such functional sites under ambient conditions is usually low because of their relatively weak CO₂–framework interactions [36,37,38,39,40]. According to recent reports, the selectivity of CO₂ can be greatly improved by adjusting the pore size falling in the ultra-microporous range due to CO₂’s increased affinity in confined space and the molecular sieving effect [41,42,43].

Based on the above considerations, the construction of MOF materials with ultramicropore and the high density of CO₂-philic sites without OMSs has the potential to enhance CO₂ adsorption capacity, improve CO₂ selectivity and strengthen water vapor stability simultaneously. By selecting 5-(1H-1,2,4-triazol-1-yl) isophthalic acid (H₂TIPA) with multiple coordination sites as the organic ligand (Figure 1) and Cu(II) paddle-wheel as the metal node, a new Cu-MOF SXU−5 was synthesized, and the above goals were satisfied all at once. Although the same ligand was used to synthesize a Cu-MOF, our synthesis conditions were different and led to a crystal structure with a different space group [21]. The significance of using the same ligand to explore new Cu-MOF synthesis lies in the potential to systematically investigate the effect of various reaction conditions on the resulting materials while maintaining a constant ligand framework. By keeping the ligand constant, the changes in the properties of the resulting MOFs can be attributed to variations in the synthetic parameters, allowing for the identification of structure–property relationships. Therefore, synthesizing new Cu-MOFs with the same linker can identify new MOF candidates with the desired CO₂ adsorption properties and guide subsequent design and synthesis. Owing to the short distance between two different functional groups from the H₂TIPA ligand, the pore size of SXU−5 falls into the ultramicroporous category (<5–7 Å). As expected, activated SXU−5 reveals a high CO₂ adsorption quality of 99 cm³ g⁻¹ and 56 cm³ g⁻¹ at 273 K and 298 K, respectively. Its IAST selectivity for CO₂ over N₂ is as high as 118 at a CO₂/N₂ proportion of 15/85 at 298 K, which is comparable to some of the best-performing CO₂ capture materials [44,45,46]. Theoretical calculations indicate how neighboring two ketonic O atoms with a suitable distance can play a vital role in boosting CO₂ selectively capture.

2. Materials and Methods

All reagents and solvents were purchased from chemical suppliers without further purification. The ligand H₂TIPA was purchased from Jinan Henghua Co., Ltd. (Shanghai, China) and characterized by ¹H NMR experiments (Figure S1). Thermogravimetric analysis (TGA) was carried out on a NETZSCH STA 449C thermal analyzer at a rate of 10 °C/min under a nitrogen (N₂) atmosphere from 30 to 900 °C. At 40 kV and 40 mA, powder X-ray diffraction (PXRD) experiments were performed on a Rigaku Ultima IV X-ray diffractometer with a scanning rate of 1°/min in the 2θ range of 3–40°, and the diffractometer was equipped with a Cu sealed tube (λ = 0.15406 nm). Based on single crystal diffraction data, simulated PXRD patterns were generated by Mercury 3.9 software.

2.1. Synthesis of SXU-5

A mixture of Cu(NO₃)₂·3H₂O (0.05 mmol, 12 mg) and H₂TIPA (0.05 mmol, 11.6 mg) was dissolved in H₂O (1 mL), DMF (Dimethyl Formamide, 2 mL) and DMA (Dimethylacetamide, 1 mL) in a 20 mL glass vial. Then, the sealed vial was sonicated for 15 min and heated at 85 °C for 72 h. After cooling naturally to room temperature, blue rod-shaped crystals were obtained after washing with fresh DMA three times (yield, 78%, based on H₂TIPA ligand).

2.2. Gas Adsorption

The Quantachrome instrument and Asiqwin system were used to obtain gas adsorption isotherms. To obtain samples free of guests, approximately 100 mg of fresh SXU−5 was soaked three times in 20 mL of ethanol for 72 h to exchange high-boiling-point guest molecules in its 1D channels, followed by outgassing under a dynamic vacuum at 100 °C for 8 h. The adsorption isotherm of N₂ was observed in a liquid nitrogen bath at 77 K. Pore size distribution (PSD) data were acquired from N₂ isotherms according to the no-local density functional theory (NLDFT) model (Figure S2). The specific surface area of SXU−5 was calculated by the BET and Langmuir model from the N₂ adsorption isotherm at 77 K. Room temperature adsorption isotherms of CO₂, N₂, CH₄, C₂H₂, C₂H₄ and C₂H₆ were measured in an ice-water bath (273 K) and water bath (298 K), respectively.

2.3. Thermal Weight Analysis

Thermogravimetric analysis was performed on a NETZSCH STA 449C thermogravimetric analyzer. An appropriate amount of leach extract with mother liquor was taken and extracted and then ramped up from 35 °C to 900 °C at a ramp rate of 10 °C/min under a N₂ atmosphere. The process of change in the material structure can be characterized by temperature.

2.4. Variable Temperature X-ray Powder Diffraction Testing

The structural changes of samples from 30 °C to 300 °C can be tested in a variable temperature X-ray powder diffractometer. The test conditions were as follows: a scanning range of 5–40 °C and a scanning rate of 1°/min. Information on structural changes at different temperatures was characterized.

2.5. Single-Crystal X-ray Diffraction (SC-XRD) Studies

Single-crystal data of SXU−5 were collected on a Rigaku Oxford single-crystal diffractometer with graphite-monochromated Cu Kα radiation at 77 K. The structure (

Table 1. Crystallographic data and structural refinement summary.

Compound	SXU-5
CCDC	No. 2012534
Empirical	C10H5CuN3O4
Formula weight	294.71
Temperature (K)	100.0(3)
Crystal system	monoclinic
Space group	P21/c
a (Å)	10.8510(3)
b (Å)	11.7772(3)
c (Å)	14.6364(5)
α (°)	90
β (°)	109.695(3)
γ (°)	90
Volume(Å3)	1761.03(9)
Z	4
Dc (g cm−3)	1.112
μ (mm −1)	1.841
F (000)	588
Crystal size (mm3)	0.2 × 0.1 × 0.1
Radiation	CuKα (λ = 1.54184)
Goodness-of-fit on F2	1.098
Final R indexes [I ≥ 2σ (I)] [a]	R1 = 0.0699, wR2 = 0.2140
Final R indexes [all data] [a]	R1 = 0.0771, wR2 = 0.2223
Largest diff. peak/hole/e Å−3	1.74/−1.95

$R_1=\sum \left|\left|F_0\right|-\left|F_c\right|\right|/\sum \left|F_0\right|; w{R}_2={[\sum w{({\left|F_0\right|}^2-|F_c{|}^2)}^2/\sum w{\left(F_0{}^2\right)}^2]}^{1/2}$.

) was solved with SHELXT [47], refined by full-matrix least squares on |F|² by SHELXL,³² and interfaced through the program OLEX2 [48]. All non-H-atoms were refined anisotropically, and H-atoms were fixed in geometrically estimated positions and refined using the riding model unless stated. Highly disordered solvent molecules were treated with the SQUEEZE routine of PLATON [49].

3. Results and Discussion

3.1. Description of Structure

Single-crystal X-ray structure analysis indicates that SXU−5 crystallizes in the monoclinal space group P2₁/c based on binuclear paddle-wheel Cu secondary building units (SBUs) and TIPA²⁻ organic ligands. As depicted in Figure 1a, each binuclear Cu₂ SBU comprises eight O atoms from four distinct carboxylate groups and two axially located imine N atoms from two different 1,2,4-triazole groups, representing a distorted octahedral configuration. Each TIPA²⁻ adopts three-coordinated distorted triangle geometry and is coordinated to three different Cu₂ SBUs. The Cu-O and Cu-N bond lengths fall within the normal scope of 2.11(8)–2.398(12) Å and 2.71(2)–2.83(2) Å, respectively [50,51,52]. There are one-dimensional (1D) rhombic zigzag passages along the crystallographic c-axis (Figure 2b,c) with a pore limiting diameter (PLD) of approx 3.6 Å based on Zeo++_V0.3 software analysis [53], which falls between the dynamic diameters of carbon dioxide (3.4 Å) and nitrogen molecules (3.64 Å), suggesting a strong molecular sieving effect. The total solvent accessible volume for SXU−5 was approximately 45% based on the PLATON calculation using a probe radius of 1.2 Å, indicating its high porosity [49]. Notably, the 1D channels of SXU−5 were decorated with a high density of uncoordinated N atoms (3.77 mol L⁻¹) [40], which are known for their CO₂-philic characteristics. Topologically, SXU−5 can be simplified with a Schlafli symbol of (4.6²)₂(4².6¹⁰.8³) (apo-topology), as the metal node and organic linker are seen as six- and three-connected nodes, respectively (Figure 2d).

3.2. Stability

The phase purity of as-synthesized SXU−5 was demonstrated by a powder X-ray diffraction (PXRD) experiment with experimental and simulated patterns in good agreement (Figure 3a). Thermogravimetric analysis (TGA) was first used to evaluate its thermal stability displaying two-step weightlessness (Figure S3). The first step of weightlessness from 25 °C to 270 °C was mainly attributed to the removal of guest molecules in the lattice (calcd 35%). After 300 °C, continual weight loss appeared due to the collapse of the framework and the decomposition of organic ligands. Varied temperature PXRD (VT-PXRD) patterns revealed a good agreement with a simulated pattern, while structural changes appear with the peak around 2θ = 15°, which shifted to higher angles after 150 °C, presumably due to its structural flexibility (Figure 3c). High-temperature SC-XRD or powder pattern Rietveld refinements are good methods to enhance our understanding of structural changes in the sample. Unfortunately, our laboratory does not have the capability to perform high-temperature SC-XRD measurements at these temperatures and also lacks the expertise to perform Rietveld refinements. The main frame structure maintains stability in the temperature range of 200 °C to 250 °C, implying its good thermal stability. Moreover, the main diffraction peaks of as-synthesized SXU−5 samples after immersion in different solvents, including diethyl ether, N-propanol, N-hexane and isopropanol for 24 h, matched well with the simulated result except for the lowered diffraction intensities of low-angle peaks around 12/13 degrees, indicating that the X-ray was scattered by the adsorbed molecules (Figure 3a).

Considering the fact that exhaust gas emissions from coal-fired power plants usually include water vapor (5–7%), adsorbents for the practical post-combustion of CO₂ capture should be steady enough toward water vapor [54,55,56,57], and it is necessary to explore the stability of SXU−5 when subjected to water molecule attacks before the CO₂ adsorption test. After exposure to a laboratory atmosphere at an ambient humidity (about 70%) for 7, 14 and even 30 days, the PXRD peaks exhibited no obvious change, revealing its excellent water vapor stability (Figure 3b) and indicating the potential for capturing CO₂ in the flue gas. Generally, many reported Cu carboxylate MOFs are prone to collapse in the case of water vapor, and just a few well-known MOFs exhibit practical water vapor stability [32,58,59,60]. The improved water vapor stability of SXU−5 may be explained from a structural viewpoint in that the coordination sites of binuclear copper are fully occupied by O atoms and N atoms from a 1, 2, 4-triazole moiety with large steric hindrance, which can protect the metal nod from water molecule attacks. We have conducted PXRD tests on the SXU-5 samples before and after immersion in water for 3 h. The PXRD patterns of the water-soaked samples before and after soaking are shown in Figure S4. It can be observed that after immersion, the diffraction peaks of the samples became broader, and their intensity decreased. Furthermore, most of the diffraction peaks did not match those of the as-synthesized sample, indicating that the structure collapsed after direct immersion in water.

3.3. Permanent Porosity

To assess the permanent porosity of SXU−5, an activation process was carried out as follows: firstly, about 100 mg of the synthesized SXU−5 was washed several times with DMF (N,N-dimethylformamide) by ultrasonication and soaked in 20 mL of ethanol for 3 days, exchanging high boiling point guest molecules in its 1D channels. In order to exchange the high boiling point solvent in the frame more thoroughly, we changed the fresh ethanol solvent every 24 h. Then, the solvent exchanged SXU−5 was outgassed at 100 °C for 8 h to produce desolvated SXU−5. This process, called activation, is designed to displace the ligand through a high boiling point solvent in the framework with a low boiling point solvent and then remove the low boiling point solvent from the pore by vacuum drying to achieve a pore that is free of guest molecules. Compared with the as-synthesized sample, the PXRD pattern of desolvated SXU−5 showed no obvious change, revealing that the structure was preserved after the activation process (Figure 3b). The N₂ adsorption isotherm of desolvated SXU−5 at 77 K was obtained with typical type-I isotherm characteristics, indicating its microporous nature (Figure 4a). Based on the adsorption data in the low-pressure range, the surface areas of BET and Langmuir were calculated to be 635 m² g⁻¹ and 883 m² g⁻¹, respectively. The pore volume was estimated to be 0.34 cm³ g⁻¹ according to the saturation uptake of N₂ (222 cm³ g⁻¹), which is consistent with the results of single crystal data (0.4 cm³ g⁻¹), further confirming the sample’s good crystallinity and purity.

3.4. Gas Adsorption and Separation

With the establishment of permanent porosity, the water vapor stability and high density of non-coordinated N atoms were distributed on the pore surface of SXU−5, inspiring us to further discover its application in selective gas adsorption and separation, especially for CO₂. Low-pressure single-component adsorption isotherms of CO₂, N₂, CH₄, C₂H₂, C₂H₄ and C₂H₆ were measured at 273 K and 298 K separately. As shown in Figure 4b,c, SXU−5 had a maximum CO₂ uptake of 99 (273 K) and 56 cm³ g⁻¹ (298 K) at 1 bar. Although the maximum adsorption of CO₂ at an ambient condition was lower than those of MOFs with various OMSs, such as the MOF-74 series, it can still compare favorably with many famous MOFs without OMSs, such as MAF-23 (56.1 cm³ g⁻¹) [45], Zn(ain)₂·(DMF) (51.3 cm³ g⁻¹) [61], IFMC-1 (60.3 cm³g⁻¹) [62] and IRMOF-74-III-(CH₂NH₂)₂ (67.2 cm³g⁻¹) [36] (for more comparisons, please see Table S1). This saturation uptake is an important indicator of adsorbents, and further experiments were performed to investigate their selective gas adsorption properties.

The affinity toward guest gas molecules can be indicated by the heat of adsorption (Qst), which was computed from experimental isotherms (calculation details please refer to supporting information) [63]. On the basis of the adsorption isotherms at 273 K and 298 K (Figure S5), the calculated Qst of SXU−5 toward CO₂ was significantly increased from 15 kJ mol⁻¹ to 43.6 kJ mol⁻¹ with CO₂ uptake increasing from 0 to 111 mg g⁻¹ (Figure 4d). This growing trend indicates that gradually increasing strong CO₂–CO₂ interactions occur inside the narrow pores, which has been observed in previously reported ultra-microporous MOF materials [64,65,66,67]. By contrast, the maximum uptakes of N₂ are 6.88 cm³ g⁻¹ and 3.5 cm³ g⁻¹ at 273 K and 298 K, separately, which are much less than those of CO₂, revealing their distinct selectivity above N₂. The ideal absorbed solution theory (IAST) was then adopted to compute the absorption selectivity of CO₂/N₂ under different temperatures [68]. Surprisingly, the selectivity for CO₂/N₂ (0.15/0.85) at 1 atm and 298 K was as high as 118 (Figure 4e), which is higher than many famous MOFs with excellent CO₂/N₂ selectivity under the same conditions, such as Co-MOF-74 (25) [69], BUT-11 (31.5) [70], and ZIF-69 (20) [71], indicating the bright prospect of SXU−5 for practical CO₂ capture and separation. Meanwhile, the IAST selectivity of SXU-5 (CO₂/N₂, 0.15/0.85) at 273 K was as high as 67 (Figure S6). Simulated results from different proportions of CO₂ and N₂ exhibited similar trends at 298 K and 273 K (Figure 4e, Figures S7 and S8), showing a high selectivity toward CO₂ in a wide range. It is worth mentioning that the selectivity was further increased to a higher value of 129 at 298 K when the CO₂/N₂ proportion increased to 0.2/0.8. The above results confirm the considerable application potential of SXU−5 in practical CO₂ capture, for which the concentration of CO₂ in the wasting gas tends to fluctuate around a certain concentration proportion of 15%.

Apart from the CO₂ capture of the post-combustion fluid, the purification of natural gas, which is mainly composed of methane for high-efficiency utilization, is essential as impurities such as CO₂, C₂H₆, C₂H₄, C₂H₂, etc., greatly reduce its quality [72,73,74]. SXU−5 exhibits a very limited uptake towards CH₄, which is considerably lower than the corresponding uptakes of commonly seen natural gas impurities at 273 K and 298 K (Figure 4b,c). As shown in Figure 4c, SXU−5 displays gradient uptakes for CO₂ (56 cm³ g⁻¹), C₂H₂ (49.5 cm³ g⁻¹), C₂H₄ (45.5 cm³ g⁻¹), C₂H₆ (27.44 cm³ g⁻¹) and CH₄ (11.6 cm³ g⁻¹), implying that SXU−5 can be a promising adsorbent candidate for natural gas decontamination. Taking CO₂ and C₂H₂ as impurities, for example, the predicted IAST selectivity for equimolar CO₂/CH₄ and C₂H₂/CH₄ as a feature of pressure is presented in Figure 4f. With increased pressure, both the CO₂/CH₄ and C₂H₂/CH₄ selectivities decrease in the low-pressure region and then gradually reach stabilization in the range of 36.1–11.5 and 28.5–9.7 at 298 K, respectively. IAST C₂H₂/CH₄ selectivity lies in the upper class compared with those of reported MOFs [75,76,77,78]. More importantly, SXU−5 can selectively adsorb a variety of commonly seen natural gas impurities at the same time, which is very important when simplifying the purification facility, indicating its huge potential in natural gas purification.

To understand the interaction between CO₂ frameworks and selectivity for carbon adsorption for SXU−5, we performed grand canonical Monte Carlo (GCMC) in Materials Studio 8.0 using their crystallographic data as initial models. Compared with the Cu-MOF reported earlier, the research on SXU−5 focused more on the CO₂ capture mechanism on a molecular level. As shown in Figure 5a, one CO₂ molecule was preferentially adsorbed between two neighboring ketonic atoms in SXU−5 with O–O distances of 2.85 Å and 2.89 Å at low pressure. Interestingly, different from the extensively accepted viewpoint that uncoordinated N atoms could bind the CO₂ preferentially, our calculation results indicate that two adjacent ketonic O atoms exhibit a stronger affinity with CO₂ molecules. This phenomenon might be due to the weak basic nature of the ketonic O atoms [79,80], exhibiting an affinity for acidic oxides such as CO₂ and their suitable distance, which could corporately contribute to enhancing MOF–CO₂ interactions. In addition, the snapshots for CO₂ adsorption in SXU−5 at different pressures (Figure 5b–d) clearly indicate that CO₂ molecules were first adsorbed around ketonic atoms in the pores. Once the ketonic O atoms were fully occupied, the CO₂ molecules could fill up the 1D channel with C-O distance from different CO₂ molecules that were smaller than the sum of their Van der Waals radius (Figure S9), contributing to increased Qst values for the enhanced CO₂–CO₂ interactions. Therefore, the excellent selectivity of CO₂ adsorption for SXU−5 could be due to the fact that the ketonic O atoms have the suitable distance to provide enhanced CO₂ binding interactions.

4. Conclusions

In conclusion, through the combination of ultramicropore and the high density of CO₂-philic sites without OMSs, we successfully constructed a new ultra-microporous Cu-MOF exhibiting good water vapor stability, high CO₂ uptake capacity (99 cm³ g⁻¹ and 56.6 cm³ g⁻¹ at 273 K and 298 K, respectively) and high selectivity over N₂ (118 at a scale of CO₂/N₂ 15/85, 298 K) simultaneously. More importantly, this work demonstrates that two neighboring ketonic O atoms with a suitable distance play vital roles in boosting CO₂ selectively capture and can be referenced for the future design of CO₂ capture materials.