Efficient Selective Capture of Carbon Dioxide from Nitrogen and Methane Using a Metal-Organic Framework-Based Nanotrap

Abstract

Selective carbon capture from exhaust gas and biogas, which mainly involves the separation of CO₂/N₂ and CO₂/CH₄ mixtures, is of paramount importance for environmental and industrial requirements. Herein, we propose an interesting metal-organic framework-based nanotrap, namely ZnAtzCO₃ (Atz⁻ = 3-amino-1,2,4-triazolate, CO₃²⁻ = carbonate), with a favorable ultramicroporous structure and electrostatic interactions that facilitate efficient capture of CO₂. The structural composition and stability were verified by FTIR, TGA, and PXRD techniques. Particularly, ZnAtzCO₃ demonstrated high CO₂ capacity in a wide range of pressures, with values of 44.8 cm³/g at the typical CO₂ fraction of the flue gas (15 kPa) and 56.0 cm³/g at the CO₂ fraction of the biogas (50 kPa). Moreover, ultrahigh selectivities over CO₂/N₂ (15:85, v:v) and CO₂/CH₄ (50:50, v:v) of 3538 and 151 were achieved, respectively. Molecular simulations suggest that the carbon atom of CO₂ can form strong electrostatic C^δ+···^δ−O-C interactions with four oxygen atoms in the carbonate ligands, while the oxygen atom of CO₂ can interact with the hydrogen atoms in the triazolate ligands through O^δ−···^δ+H-C interactions, which makes ZnAtzCO₃ an optimal nanotrap for CO₂ fixation. Furthermore, breakthrough experiments confirmed excellent real-world separation toward CO₂/N₂ and CO₂/CH₄ mixtures on ZnAtzCO₃, demonstrating its great potential for selective CO₂ capture.

1. Introduction

Global warming, one of the biggest global issues, is causing various long-term disastrous environmental effects, including abnormal climate patterns, rising sea levels, accelerated extinction of species, and shifts in agricultural patterns, which pose severe threats to the survival and development of humanity [1,2,3,4]. As the main source of greenhouse gases, carbon dioxide (CO₂), with a large annual emission (e.g., 36.8 Gt in 2022), contributes to 69.4% of the anthropogenic greenhouse gases [5,6]. Hence, CO₂ capture and utilization is of utmost importance to curb global warming [7]. In particular, electricity generation from the combustion of fossil fuels, resulting in exhaust gas composed mainly of CO₂ and nitrogen (N₂), is the primary source of anthropogenic CO₂ emissions [8,9]. Additionally, biogas, acknowledged as “renewable natural gas”, is a green fuel with the efficient component of methane (CH₄) [10]. Nevertheless, biogas contains a considerable amount of CO₂ that could significantly decrease the calorific value and lead to severe erosion in the equipment. Therefore, efficient selective capture of CO₂ from N₂ and CH₄ is of paramount significance for environmental protection and biogas upgrading.

Benefiting from a plausible gentle operation condition and reduced energy consumption for regeneration, adsorptive separation of CO₂ has been recognized as a promising alternative among the current CO₂ capture techniques [11,12,13,14]. Fundamentally, an ideal adsorbent with high capacity and selectivity for the target molecule is the key to achieving satisfactory separation performance [15,16]. As a class of innovative porous adsorbents, metal-organic frameworks (MOFs) are showing extraordinary versatility in pore tunability and chemical functionalities [17]. Hence, adsorptive separation based on MOFs has been explored in miscellaneous separation circumstances, including carbon capture [8,9,11,18,19,20]. In general, there are several strategies that can enhance the CO₂ capture performance on MOF materials, including (i) incorporation of functionalities (such as open metal sites, Lewis basic sites, and polar functional groups) into the frameworks, (ii) utilization of ultramicropores to maximize the confinement effect, (iii) use of kinetic difference and even the molecular sieving effect, (iv) introduction of structural flexibility, (v) the combination thereof [16,21,22]. Specifically, MOFs constructed by triazolate linkers represent one subclass of MOFs that are comprised of excellent CO₂ capture performance, affordable costs, and good stability under humid conditions [23,24]. For example, by introducing amino groups into the triazolate linkers of the prototype MOF ZnF(TZ), the resultant ZnF(daTZ) demonstrates an appreciable volumetric CO₂ uptake of 75 cm³/cm³ and high CO₂/N₂ equilibrium selectivity of 120 based on the ideal adsorbed solution theory (IAST) as well as excellent CO₂/H₂O kinetic selectivity of 70 [25]. This remarkable separation performance was probably originated from the main adsorptive site at the channel center for CO₂ to afford electrostatic interactions with the amino groups, while H₂O was more likely to locate at the channel corner, according to GCMC simulations. Built from dual ligands of oxalate and 1,2,4-triazolate, zinc-based Calgary Framework 20 (CALF-20) exhibits a high CO₂ capacity of 87.36 cm³/g at atmospheric conditions, excellent selectivities over N₂ and H₂O, and the facility for scalable production, making it the first practical MOF for industrial carbon capture [26,27]. Molecular simulations suggest the CO₂ adsorption location in CALF-20 also lies in the channel center to form interactions with the zinc/nitrogen/carbon atoms in the framework. Likewise, the amine-appended zinc-oxalate-triazolate MOFs demonstrated enhanced CO₂ capacity at low CO₂ concentrations compared to CALF-20, due to higher-density interaction sites and more contracted pore sizes [28,29]. Recently, we prepared a flexible MOF, namely ZnDatzBdc, that showed step-shaped CO₂ isotherm due to breakage/reformation of intra-framework hydrogen bonds and rotation of the phenyl rings, giving rise to an excellent CO₂ theoretical working capacity of 94.9 cm³/cm³ if performed in typical pressure vacuum swing adsorption at 273 K [30].

By now, developing nanotraps with multiple host-guest interactions toward the target molecules offers a feasible strategy to accomplish high adsorption capacity and selectivity, which has been successfully applied in separation circumstances, such as C₂H₂/CO₂ separation [31,32], C₃H₄/C₃H₆ separation [33,34], CH₄/N₂ separation [35], and olefin/paraffin separation [36]. For CO₂ adsorption, if a contracted pore exhibits opposite electrostatics on the adjacent positions and the same electrostatics on its opposite side, it can form strong electrostatic interactions with both the carbon and oxygen atoms in the CO₂ molecule, as shown in Figure 1. Hence, this type of pore can act as a suitable nanotrap for CO₂ fixation, from which a remarkable capacity and selectivity for CO₂ can be achieved.

Herein, we synthesized a novel MOF, namely ZnAtzCO₃, with 3-amino-1,2,4-triazolate (Atz^-) and carbonate (CO₃²⁻) as dual ligands. Inspiringly, ZnAtzCO₃ shows the desired ultramicropores due to the small-sized ligands and suitable crystal structure, which is desirable for the adsorption and separation of small molecules, such as CO₂. Moreover, the favorable electrostatic environment of ZnAtzCO₃ makes it a feasible nanotrap to form multiple host-guest interactions with CO₂, and hence, efficient CO₂ capture from N₂ and CH₄ could be achieved. In particular, at atmospheric temperature, equilibrium isotherms showed high CO₂ capacities of ZnAtzCO₃ in a wide pressure range, with values of 44.8 cm³/g (STP, standard temperature and pressure) at the typical fraction of the flue gas (15 kPa) and 56.0 cm³/g at the fraction of the biogas (50 kPa). Moreover, adsorptive selectivity based on the IAST model indicated that ultra-high CO₂/N₂ and CO₂/CH₄ selectivities of 3538 and 151 were realized at ambient conditions, respectively. The excellence in capacity and selectivity of this MOF-based nanotrap was illustrated by molecular simulations in terms of preferential adsorption sites, binding energy, and adsorption distributions. Furthermore, breakthrough experiments toward the binary mixtures of CO₂/N₂ and CO₂/CH₄ were conducted on ZnAtzCO₃, which verified its efficient dynamic separation performance.

2. Results and Discussion

2.1. Crystal Structure and Pore Properties

Reactions of ZnSO₄ and 3-amino-1H-1,2,4-triazole (HAtz) in a binary solution of DMF/H₂O afforded high-quality crystals of ZnAtzCO₃. The single-crystal X-ray diffraction (SCXRD) measurement indicates that ZnAtzCO₃ belongs to the triclinic crystal system (a = 9.6217 Å, b = 9.6316 Å, c = 16.3408 Å, α = 81.355°, β = 86.938°, γ = 76.093°). Each asymmetric unit contains four zinc atoms, four 3-aminotriazolate ligands, and two triangular carbonate linkers (Figure S1). Because no carbonate was added to the reactants, it is assumed that the carbonate linker originated from the decomposition of the DMF molecule [37]. In addition, the elemental analysis suggests no sulfur element in the framework, which further confirms the existence of the carbonate linker instead of the sulfate or sulfite linker in the framework. As shown in Figure 2a, the zinc atom coordinates with one oxygen atom from the carbonate linker and three nitrogen atoms from three different triazolate rings. Each carbonate linker contains one uncoordinated oxygen atom, which is fixed by the intra-framework hydrogen bonding with the amino group in the triazolate ligand. We tried to obtain isoreticular structures by replacing the HAtz ligand with 1H-1,2,4-triazole (HTz) and 3,5-diamine-1H-1,2,4-triazole (HDatz), but the trial failed. Hence, we assume that these weak intra-framework hydrogen bonds are vital for the structural formation of ZnAtzCO₃. ZnAtzCO₃ can be regarded as a pillared–layered structure by connecting the wavy and continuous zinc-triazolate layers by the carbonate pillars (Figure 2b). Inspiringly, the small size of the carbonate and triazolate ligands gives rise to an ultramicroporous structure desirable for adsorptive separation. Specifically, ZnAtzCO₃ contains two types of zig-zag channels with a cross-section area of 2.9 × 5.1 Ų and 3.5 × 5.1 Ų that interconnect with the adjacent channels through small apertures of 3.0 × 3.9 Ų, 2.2 × 2.8 Ų, and 2.6 × 3.3 Ų. We noticed that a similar form of Zn₂(atz)₂(CO₃) was previously reported by the reaction of Zn(NO₃)₂, NaHCO₃, and HAtz. Zn₂(atz)₂(CO₃) displayed the same connectivity but belonged to another different space group of a Pnma unit cell (a = 9.806 Å, b = 9.3353 Å, c = 16.194 Å, α = β = γ = 90°) [38]. The difference in structure is probably because the degrees of buckling for zinc-triazolate layers can vary significantly under different synthetic conditions. Because the specific crystal structure affects the pore systems and, subsequently, the sorption behavior, the following discussion was carried out with our obtained crystal data.

2.2. Characterizations

The physicochemical behavior of ZnAtzCO₃ was measured to investigate its textural characteristics. Fourier transform infrared reflection (FTIR) patterns were performed to further confirm the existence of the carbonate linker. Figure 3a suggests an intense broad band at 1408 cm⁻¹ and an additional band at 1310 cm⁻¹, corresponding to the asymmetric stretching modes of carbonate [39]. Besides, the medium band at 850 cm⁻¹ was assigned to the bending mode of the carbonate. Figure 3b depicts the powder X-ray diffraction (PXRD) pattern of ZnAtzCO₃, which is identical to that derived from SCXRD measurement, indicative of the high purity of the powder sample. In addition, the activation step did not lead to transformation or decomposition of the structure, as suggested by the well-maintained PXRD patterns. Thermogravimetric (TG) analysis in Figure 3c indicates that ZnAtzCO₃ is stable up to 500 K, and hence, it holds enough thermal stability for adsorptive separation, which usually requires moderate heating for regeneration. The porosity feature was derived from CO₂ sorption isotherms at 195 K, as shown in Figure 3d. The typical type-I CO₂ isotherms indicated the microporous nature of ZnAtzCO₃, giving a Brunauer–Emmett–Teller (BET) surface area of 455.6 m²/g and a micropore volume of 0.196 cm³/g.

2.3. Adsorption Equilibrium Behavior of CO₂, N₂, and CH₄

Figure 4a depicts the pure adsorption isotherms of CO₂, N₂, and CH₄ on ZnAtzCO₃ at 298 K. It is noticed that CO₂ capacity increased sharply at low pressure, while N₂ and CH₄ uptakes showed a slow increment as the pressure rose. Hence, ZnAtzCO₃ showed a significantly higher CO₂ capacity than N₂ and CH₄ between 0–100 kPa. At atmospheric pressure, CO₂ capacity reached as high as 62.8 cm³/g, while this value for N₂ and CH₄ was 4.3 and 14.7 cm³/g, respectively. This distinction in adsorption capacity suggests excellent thermodynamic separation for CO₂/N₂ and CO₂/CH₄. Moreover, ZnAtzCO₃ demonstrated an exceptional CO₂ adsorption capacity of 44.8 cm³/g at 15 kPa, highlighting its promising application prospects in low-concentration CO₂ capture, such as CO₂ elimination from the exhaust gas. Moreover, ZnAtzCO₃ exhibits a high CO₂ capacity of 56.0 cm³/g at 50 kPa, indicative of its potential in CO₂ separation with higher CO₂ concentrations, including CO₂ removal from the biogas.

To evaluate the competitive separation of CO₂/CH₄ and CO₂/N₂ on ZnAtzCO₃, the IAST selectivities were calculated by means of the IAST model, taking into account the composition of CO₂/N₂ (15:85, v:v) and CO₂/CH₄ (50:50,v:v) in the exhaust gas and biogas, respectively [40,41]. By incorporating the dual-site Langmuir–Freundlich (DSLF) parameter (Table S1) into the IAST model [42], the IAST selectivities were obtained and shown in Figure 4b. Significantly, CO₂/CH₄ and CO₂/N₂ selectivities are quite high in the whole pressure range of 0–100 kPa. At ambient pressure, the IAST selectivity for CO₂/CH₄ and CO₂/N₂ reached as high as 3538 and 151, respectively, which are comparable to those benchmark MOFs for selective CO₂ capture through thermodynamic separation [12]. Hence, compromised of excellent CO₂ capacity and selectivity, ZnAtzCO₃ reveals great potential for selective CO₂ capture.

Furthermore, the isosteric heat (Q_st) of CO₂, N₂, and CH₄ on ZnAtzCO₃, which can be derived from their pure adsorption isotherms at various temperatures (273, 288, and 298 K), is a crucial parameter for determining the adsorption interaction strengths. As shown in Figure 5b, the zero-coverage Q_st for the three gases followed an order of CO₂ (32.6 kJ/mol) > CH₄ (22.4 kJ/mol) > N₂ (18.1 kJ/mol), consistent with the dipole moment of the guest molecules (CO₂: 29.1 × 10⁻²⁵ cm³, CH₄: 25.9 × 10⁻²⁵ cm³, N₂: 17.4 × 10⁻²⁵ cm³) [43]. Additionally, it is apparent that the Q_st for all gases remained constant between 0–100 kPa, suggesting homogeneity on the pore surface. From the Q_st result, we speculate that the remarkable CO₂ selectivity is primarily attributed to its highest adsorption enthalpy.

In addition, ZnAtzCO₃ was compared with other MOFs constructed by triazolate linkers on their CO₂ uptake at 15 kPa and 100 kPa, together with the isosteric heat [25,26,29,30,44,45,46,47]. As shown in

Table 1. Comparisons of CO₂ uptakes at 15 kPa and 50 kPa on typical MOFs constructed by triazolate linkers.

MOFs	QCO2 at 15 kPa (cm3/g, STP)	QCO2 at 50 kPa (cm3/g, STP)	Qst (kJ/mol)	T (K)	Ref.
MAF-7	4.5	12.5	25	298	[44]
ZnF(TZ)	6.0	19.1	24	298	[25]
ZnF(daTZ)	21.4	34.1	33	298	[25]
ZnDatzBdc	1.9	5.8	29	298	[30]
CALF-20	53.7	68.3	33.5	303	[26]
ZnAtzOx	60.5	65.0	55	303	[29]
ZU-301	47.7	52.6	39	298	[45]
Zn(FA)(datrz)2	12.3	28.5	24.0	298	[46]
Zn2(TRZ)2(BDC)	23.5	41.4	-	298	[47]
Zn2(TRZ)2(FA)	34.7	73.9	-	298	[47]
ZnAtzCO3	44.8	56.0	32.6	298	This work

, the CO₂ uptakes on ZnAtzCO₃ exceed those of the ZnF(Tz) series, ZnDatzBdc, Zn(FA)(datrz)₂, and Zn₂(TRZ)₂(BDC), and are comparable to that of ZU-301, and slightly lower than those of ZnAtzOx and CALF-20. Hence, ZnAtzCO₃ holds comparatively high CO₂ capacity among these triazolate-based MOFs. Hence, ZnAtzCO₃ can be regarded as a promising adsorbent in combination with good capacity and selectivity toward CO₂.

2.4. Molecular Simulations on the Selective CO₂ Adsorption over N₂ and CH₄

The intrinsic mechanism for the excellent separation performance on ZnAtzCO₃ was illustrated by molecular simulations on the preferential adsorption sites, adsorption density distributions, and interaction energy with the aid of Material Studio 7.0 [48]. To visualize the intrinsic host–guest interactions between ZnAtzCO₃ and the gas molecules, the preferential interaction site was calculated and depicted in Figure 6. Specifically, for the CO₂ molecule, the carbon atom can form four C^δ+···^δ−O-C electrostatic interactions with the oxygen atoms in the carbonate linker, and each oxygen atom can form O^δ−···^δ+H-C electrostatic interactions with the hydrogen atoms in the aminotriazolate rings. The multiple host–guest interactions verify that the favorable electrostatic environment of ZnAtzCO₃ can form an efficient nanotrap for CO₂. It is noticed that the amine group did not form electrostatic interactions with CO₂, which might originate from insufficient contact with CO₂. For CH₄, CH₄ interacts with three oxygen atoms in the carbonate linker, two nitrogen atoms in the amino groups, and one adjacent triazolate ring through dispersion forces. Considering its significantly smaller polarizability and more inert nature, the preferential site displayed weaker affinity for CH₄ than CO₂. Likewise, being the weakest adsorbate, N₂ forms dispersion interactions with three hydrogen atoms in the aminotriazolate linkers and one oxygen atom in the carbonate linker. From the result above, ZnAtzCO₃ shows stronger host–guest interactions with CO₂ compared to N₂ and CH₄.

Figure 7 presents the ambient-temperature adsorption density distributions of the three gases on ZnAtzCO₃ at low (15 kPa) and ambient pressure (100 kPa). At both 15 kPa and 100 kPa, the density distribution for CO₂ was the highest, followed by CH₄ and N₂, which is a valid proof of the significantly higher capacity of CO₂ than N₂ and CH₄. As the pressure rose from 15 kPa to 100 kPa, the adsorption density increased for all gases because, generally, the increment in pressure can provide an increasing driving force for gas adsorption.

In addition, the distinction of the simulated interaction energy was calculated to further confirm the difference in the adsorption enthalpy. As shown in Figure 8, the average energies between ZnAtzCO₃ and the gas molecules were −6.85, −6.06, and −3.50 kcal/mol for CO₂, CH₄, and N₂, respectively, which shows a consistent trend with the Q_st result. Hence, the stronger electrostatic host–guest interactions, apparently higher adsorption density distributions, and larger adsorption energy, comprehensively explain the selective CO₂ adsorption over N₂ and CH₄ on ZnAtzCO₃.

2.5. Dynamic Breakthrough Experiments

For evaluation of the capability for selective CO₂ capture from the exhaust gas and biogas on ZnAtzCO₃, the breakthrough experiments were performed to simulate the dynamic separation performance toward CO₂/N₂ (15:85, v:v) and CO₂/CH₄ (50:50, v:v) at ambient conditions (Figure 9). It is noticed that CH₄ and N₂ were detected shortly after induction of the gas mixtures and reached equilibrium rapidly, confirming their uptake was inappreciable on ZnAtzCO₃. In contrast, CO₂ broke through at 81 min/g and reached equilibrium at 134 min/g for the CO₂/N₂ mixture, and the breakthrough time and equilibrium time for CO₂/CH₄ mixture were 37 and 53 min/g, respectively. In addition, the CO₂ breakthrough time remained constant for five cycles for both mixtures, suggesting their excellent cyclicity in real-world CO₂ capture conditions.

3. Materials and Methods

3.1. Material Sources

Zinc sulfate heptahydrate (ZnSO₄·7H₂O, 99.9%), 3-amino-1H-1,2,4-triazole (HAtz, 99.9%), N,N’-dimethylformamide (DMF, 99.5%), and methanol (MeOH, 99.5%) were supplied by Macklin Biochemical Co., Ltd. (Shanghai, China). The gases employed in this work, including single-component gases (CO₂, CH₄, N₂, and He) of high purity (over 99.9%) and binary gas mixtures of CO₂/N₂ (15:85, v:v) and CO₂/CH₄ (50:50, v:v), were purchased from Kedi Gas Chemical Industry Co., Ltd. (Foshan, China).

3.2. Synthesis of ZnAtzCO₃

HAtz (84.0 mg, 1 mmol) was dispersed in DMF (2 mL) in an autoclave, followed by addition of an aqueous solution (8 mL) of dissolved ZnSO₄·7H₂O (287.6 mg, 1 mmol) under stirring. After the suspension was mixed by stirring for 15 min, the autoclave was sealed, kept at 423 K for 60 h, and subsequently slowly cooled to ambient temperature. The resultant colorless crystals were recovered by filtration, washed with deionized H₂O (2 × 20 mL) and MeOH (3 × 20 mL) to remove the unreacted reactants, and then dried in air. The activated ZnAtzCO₃ was prepared by heating to remove the contaminants such as DMF, H₂O, and the gases adsorbed from air. CCDC number: 2297804. Crystal Data for Zn₄C₁₀H₁₂N₁₆O₆ (M = 713.92 g/mol): triclincic, space group P-1 (no. 2), a = 9.6217(2) Å, b = 9.6316(2) Å, c = 16.3408(4) Å, α = 81.355(2) °, β = 86.938(2)°, γ = 76.093(2)°, V = 1453.01(6) ų, Z = 2, T = 150.00(10) K, μ(CuKα) = 0.790 mm⁻¹, D_calc = 1.632 g/cm³, 5103 reflections measured (4.735° ≤ 2Θ ≤ 66.988°), 4762 unique (Rint = 0.0381, Rsigma = 0.0300) which were used in all calculations. The final R₁ was 0.0623 (I > 2σ(I)) and wR₂ was 0.1667 (all data). Elemental analysis (wt%) calculated for Zn₄(Atz)₄(CO₃)₂ (Zn₄C₁₀H₁₂N₁₆O₆): C, 16.83; H, 1.69; N, 31.40; S: 0.00. Found: C, 17.25; H, 2.05; N, 30.5; S, 0.00.

3.3. Characterizations

SCXRD analysis was carried out on a Rigaku Oxford Diffraction (Rigaku, Tokyo, Japan) with a hybrid pixel array detector. Reflections combined with SHELXL corresponding to the crystal class were employed for the calculation of statistics and refinement to solve the non-hydrogen atoms, while the locations and numbers of all hydrogen atoms were calculated theoretically. Elemental analysis was performed on a Vario EL elemental analyzer (Elementar, Langenselbold, Germany) in the CHNS mode. FTIR spectroscopy in the range of 1800–400 cm⁻¹ was recorded on a Thermo Scientific iN10 (Thermo Fisher Scientific, Waltham, MA, USA) microscope with potassium bromide as the matrix. PXRD patterns were collected on a Bruker D8 Advance diffractometer (Bruker, Mannheim, Germany). TG measurements of the as-synthesized ZnAtzCO₃ were performed on a TGA 550 thermal gravimetric analyzer (Thermo Fisher Scientific, Waltham, MA, USA), and the sample was heated from 303 K to 973 K at a ramping rate of 10 K/min under flowing nitrogen.

3.4. Single-Component Gas Sorption Isotherm Measurements

Single-component sorption isotherm measurements between 0–100 kPa were carried out on 3Flex (Micromeritics, Norcross, GA, USA) at various temperatures. In the preparation process, approximately 100 mg of ZnAtzCO₃ was activated under dynamic vacuum for 6 h at 393 K to afford a guest-free sample. During the test, the sample tube was placed in a thermostatic environment by using the ice-acetone bath (195 K) or circulating water bath (288 K, 298 K, and 313 K) to maintain a constant operational temperature.

3.5. Adsorption Selectivity Based on IAST Model

Before the IAST selectivity calculation, the experimental isotherms of CO₂, N₂, and CH₄ require accurate fitting to a mathematical model. In this work, the DSLF equation, based on the assumption that two types of adsorption sites are present in the structure, was selected and described to describe the adsorption equilibrium of the single-component gases Equation (1) [30,42].

$$q=q_{e1}\frac{k_1{p}^{t_1}}{1+k_1{p}^{t_1}}+q_{e2}\frac{k_2{p}^{t_2}}{1+k_2{p}^{t_2}}$$

(1)

where p is the specific pressure when the gas phase and adsorbed phase reach a steady state; q_ei is the saturated uptake of site i; k_i is the affinity coefficients of site i; t_i represents the divergence from an absolute homogeneous surface on site i.

The IAST selectivity for CO₂/N₂ and CO₂/CH₄ on ZnAtzCO₃ can be derived according to Equation (2) [40].

$$S_{AB}=\frac{x_A/x_B}{y_A/y_B}$$

(2)

where x_i and y_i refer to the volume fractions of component i in the adsorbed phase and the gas phase, separately.

3.6. Isosteric Heat (Qst) Calculation

The experimental adsorption isotherms of CO₂, N₂, and CH₄ on ZnAtzCO₃ at various temperatures were fitted to the Virial equation, Equation (3) [49], and the parameters are shown in Table S2.

$$\ln \left(P\right)=\ln \left(\mathrm{N}\right)+\left(\frac{1}{T}\right){\sum }_{i=0}^m{a}_i{N}^i+{\sum }_{j=0}^n{b}_j{N}^j$$

(3)

where p is the pressure, N is the gas capacity, T is the absolute temperature, a_i and b_j refer to the corresponding parameter in the Virial equation, while m and n refer to the number required for the accurate fitting of the Virial equation.

Subsequently, the corresponding Q_st was figured out by substituting the parameter in the Virial equation into the Clausius–Clapeyron equation, Equation (4) [50].

$$Q_{st}=-R{\sum }_{i=0}^m{a}_i{N}^i$$

(4)

where R is short for the ideal gas constant, 8.314 J/mol/K.

3.7. Breakthrough Experiments

Dynamic separation experiments of CO₂/N₂ (15:85, v:v) and CO₂/CH₄ (50:50, v:v) mixtures were performed on self-assembly breakthrough equipment (Figure S3). A small-scale adsorption column was prepared by loading approximately 600 mg of the activated ZnAtzCO₃ in a stainless-steel column (Φ 50 × 150 mm). For activation, the column packed with the sample was heated at 393 K for two hours to eliminate the adsorbed contaminants. Subsequently, the column was inserted into the breakthrough equipment and purged by He flow (10 mL/min) at ambient conditions until the baseline was flattened. Finally, the gas was shifted to CO₂/N₂ or CO₂/CH₄ at a flow rate of 3 mL/min. The outlet component was monitored on a thermal conductivity detector (TCD) until the outlet composition reached that of the feed gas, which suggested the breakthrough column reached equilibrium. The adsorption column was recovered by purging He flow at 373 K to liberate the adsorbed gas molecules in the cyclability test.

3.8. Simulation Details

The molecular simulations on the adsorption mechanism were calculated by utilizing the Materials Studio 7.0 software [50]. First, the structures of ZnAtzCO₃ and the adsorbates were optimized with the aid of the Forcite and Dmol3 modules. The adsorption characteristics, including the optimal adsorption sites, adsorption density distribution, and stabilized adsorption energy, were simulated in the Sorption module with the Metropolis Monte Carlo method. The adsorption behavior of the guest molecules on ZnAtzCO₃ was described by several motion types, including exchange, conformation, rotation, translation, and regeneration. The Ewald and atom-based methods were employed to depict the electrostatic interaction and Van der Waals interactions between the structure and the guest molecules, respectively. The cutoff for the Metropolis Monte Carlo simulation was set as 12.5 Å. One gas molecule was randomly inserted into the framework in the Location task in the Sorption module, with 1 × 10⁵ steps for equilibrium and production, separately.

4. Conclusions

In conclusion, we propose an interesting type of MOF-based nanotrap, namely ZnAtzCO₃, for efficient selective capture of CO₂ from N₂ and CH₄. The favorable electrostatic environment and narrow pore geometry of ZnAtzCO₃ show stronger interaction with CO₂ than N₂ and CH₄. Specifically, ZnAtzCO₃ accomplished high CO₂ capacities with values of 74.0 cm³/cm³ at the fraction of the flue gas (15 kPa) and 91.4 cm³/cm³ at the fraction of the biogas (50 kPa), together with ultra-high CO₂/N₂ and CO₂/CH₄ selectivities of 3538 and 151 at ambient conditions, respectively. This excellent separation performance was comprehensively explained by molecular simulations, which suggests that the carbon atom of CO₂ can form strong electrostatic C^δ+···^δ−O-C interactions with the oxygen atoms in the carbonate ligand and the oxygen atom of CO₂ can interact with the hydrogen atoms in the triazolate ligand through O^δ−···^δ+H-C interactions, enabling ZnAtzCO₃ as an optimal nanotrap for CO₂ fixation. Moreover, breakthrough experiments confirm excellent dynamic separation toward CO₂/N₂ and CO₂/CH₄ on ZnAtzCO₃, highlighting its potential for selective CO₂ capture. Furthermore, constructing suitable nanotraps with optimal electrostatic environment and pore geometry is worthy of further exploration in other separation circumstances.