An adsorbate biased dynamic 3D porous framework for inverse CO₂ sieving over C₂H₂

Abstract

Separating carbon dioxide (CO₂) from acetylene (C₂H₂) is one of the most critical and complex industrial separations due to similarities in physicochemical properties and molecular dimensions. Herein, we report a novel Ni-based three-dimensional framework {[Ni₄(μ₃-OH)₂(μ₂-OH₂)₂(1,4-ndc)₃](3H₂O)}_n (1,4-ndc = 1,4-naphthalenedicarboxylate) with a one-dimensional pore channel (3.05 × 3.57 Ų), that perfectly matches with the molecular size of CO₂ and C₂H₂. The dehydrated framework shows structural transformation, decorated with an unsaturated Ni(II) centre and pendant oxygen atoms. The dynamic nature of the framework is evident by displaying a multistep gate opening type CO₂ adsorption at 195, 273, and 298 K, but not for C₂H₂. The real time breakthrough gas separation experiments reveal a rarely attempted inverse CO₂ selectivity over C₂H₂, attributed to open metal sites with a perfect pore aperture. This is supported by crystallographic analysis, in situ spectroscopic inspection, and selectivity approximations. In situ DRIFTS measurements and DFT-based theoretical calculations confirm CO₂ binding sites are coordinatively unsaturated Ni(II) and carboxylate oxygen atoms, and highlight the influence of multiple adsorption sites.

---

Introduction

C₂H₂ holds crucial significance as a monomeric organic feedstock in both chemical and polymer industries, along with extensive use as a fuel in welding apparatus.¹ The production of ultra-pure C₂H₂ is essential for industrial chemical manufacturing, but the presence of impurities such as CO₂ and CH₄ in reactor products from hydrocarbon cracking poses a significant challenge.^2–4 Purifying C₂H₂ and separating it from persistent CO₂ impurities is regarded as one of the most formidable tasks due to the similarities in molecular dimensions (C₂H₂: 3.32 × 3.34 × 5.7 ų and CO₂: 3.18 × 3.33 × 5.36 ų).⁵ The current purification technology necessitates operation under cryogenic temperatures and high pressures, adding complexity to the separation process due to the closely matched boiling points (C₂H₂, 189.3 K; CO₂, 194.7 K).^4,6 Therefore, the aforementioned conventional methods for gas separation, such as solvent extraction or cryogenic distillation, are expensive, inefficient, and highly energy-consuming. In this context, adsorptive separation through porous materials is attracting significant interest as it is more environmentally friendly, cost-effective, and involves lower energy expenses, making it a next-generation separation technology.

Benefiting from precise control over pore accessibility and the pore environment, metal–organic frameworks (MOFs) demonstrate significant potential in the field of gas storage and separation.^7–12 Regrettably, the majority of reported MOFs exhibit a very similar binding affinity for both C₂H₂ and CO₂, with only a few designed specifically for the exclusive separation of C₂H₂ from CO₂.^13,14 In those few successful reports, pores are typically engineered to favor the preferential adsorption of C₂H₂ over CO₂. This is achieved by employing a hydrophobic pore environment, leveraging acid–base interactions with the binding site, and incorporating linkers with highly polarizable π-electrons.^15–19 Nevertheless, in each case, the blowdown step necessitates an extra step involving a deep vacuum to achieve the recovery of ultrapure C₂H₂ due to its strong binding to the pore.²⁰ Hence, achieving inverse selectivity, where CO₂ adsorption is favored over C₂H₂, could eliminate the need for the co-current blowdown step, making the overall separation process more straightforward and energy-efficient.

In this regard, Yang et al. were the first to report the preferential adsorption of CO₂ over C₂H₂ by incorporating both pendant carboxyl and pyridyl groups in the channels.²¹ Recently, Kitagawa and coworkers showcased the selective adsorption of CO₂ based on a flexible PCP (porous coordination polymer) driven by C–H⋯π and π–π interactions. However, it also delimits a significant C₂H₂ uptake of 86.2 mL g⁻¹ STP at 195 K.²² Other approaches involve post-synthetically modifying the pore surface with ions, immobilizing methyl groups to enhance van der Waals interactions, and introducing pendant polar OH groups, among others.^23–25 However, a trend of both CO₂ and C₂H₂ adsorption was observed. The structural flexibility and dynamics of MOFs can potentially enhance shape matching and provide specific binding towards particular gas molecules.^26–34 In this study, our strategy for preferential CO₂ sieving involves introducing an open metal site while maintaining the specific geometry of the pore. In comparison to functional group interactions (from linkers) with CO₂, as discussed earlier, the open metal site interaction with CO₂ is expected to be stronger. This could potentially enable the sieving out of target CO₂ molecules by minimizing the co-adsorption of counter gas (C₂H₂) molecules.

In this work, we report the synthesis and structural characterization of a 3D metal–organic framework {[Ni₄(μ₃-OH)₂(μ₂-OH₂)₂(1,4-ndc)₃](3H₂O)}_n (1), embodied by 1,4-naphthalenedicarboxylate linkers (1,4-ndc). The resulting framework features an accessible 1D pore channel with high polarity and an aperture size of approximately 3.3 Å, facilitating preferential CO₂ binding within the undulating 1-D channels. Consecutive temperature-dependent powder X-ray diffraction (PXRD) studies, conducted immediately after obtaining the vapor adsorption isotherm, collectively offer a more comprehensive understanding of the structural dynamics of the 3D framework (Scheme 1). The desolvated framework 1 (1′) exhibits a gate-opening type multistep adsorption profile for CO₂ at 195, 273, and 298 K but not for C₂H₂, demonstrating selective adsorption of CO₂ over C₂H₂ at ambient temperature (298 K). The real time breakthrough gas separation measurements are performed to elucidate the experimental performance of CO₂vs. C₂H₂ and CO₂vs. CH₄ separation under ambient conditions. Additional investigations were carried out using in situ DRIFTS measurements and DFT calculations to gain a deeper understanding of the CO₂ binding sites on the pore surface.

Scheme 1 Schematic representation of dynamicity in a Ni-MOF and its preferential CO₂ sieving over C₂H₂.

Results and discussion

Crystal structure and framework stability of {[Ni₄(μ₃-OH)₂(μ₂-OH₂)₂(1,4-ndc)₃](3H₂O)}_n (1)

The reaction of the 1,4-H₂ndc ligand with Ni(NO₃)₂ in water under basic pH resulted in the formation of dark green block-shaped crystals of 1. Single-crystal X-ray diffraction analysis indicates that 1 crystallized in the triclinic P1̄ space group. The asymmetric unit consists of four Ni(II) centers, three 1,4-ndc linkers, two μ₃-OH groups, two μ₂-OH₂ molecules, and three guest water molecules (Fig. 1a). The 3D framework adopts a square grid structure, formed through the crosslinking of fused Ni(II) octahedral chains with 1,4-ndc linkers, as depicted in Fig. 1b and c. This structure consists of four independent Ni(II) atoms assembled into two parallel Ni–O–Ni chains. The parallel chains in the structure consist of Ni1 and Ni2 connected with a μ₃-OH group, μ₂-OH₂, and μ-O from carboxylate ligands. On the other hand, Ni3 and Ni4 are connected with carboxylate groups through syn–syn bridging (Fig. 1a–c). Consequently, in the parallel chains, each Ni2 and Ni3 are associated with three different carboxylate oxygen atoms, whereas Ni1 and Ni4 are attached to one μ₃-OH, one μ₂-OH₂, two μ-O, and two carboxylates from four different linkers, forming distorted NiO₆ octahedra to create four alternate triangular arrays. The intrachain Ni1–Ni2 separation distances are 3.306 Å (syn–syn bridging) and 3.774 Å (μ₂-OH₂ bridging), while for Ni3–Ni4, they are 3.274 Å (syn–syn bridging) and 3.805 Å (μ₂-OH₂ bridging). The Ni–O distances with 1,4-ndc oxygen atoms and μ₃-OH are all comparable and fall in the range of 1.974(14)–2.210(16) Å. The O–Ni–O cisoid angles range from 79.50(6)° to 102.92(6)°, and the transoid angles vary from 166.79(7)° to 176.45(6)°, indicating the degree of deviation from the ideal octahedron (Tables S1–S3†). The dicarboxylate group of the 1,4-ndc linkers diverges from the chain and connects to the Ni(II) centers of the adjacent chains, forming a 3D framework with 1-D channels along the a-axis (Fig. 1b) with guest water molecules (Fig. 1d). Excluding the guest molecules, 1 has undulating channels (9.8% void space) along the a-axis with a channel size of 3.05 × 3.57 Ų. The removal of the water molecules would provide the structure with the unsaturated Ni(II) sites, which can capture CO₂ molecules having a quadrupole moment.

Fig. 1 (a) Asymmetric unit of 1 executing different Ni(II) environments. (b) Three-dimensional view along the a direction. (c) View of the bilayer {Ni₂-(μ-OH)(μ-OH₂)} chains supported by 1,4-ndc showing the bridging water and hydroxyl groups. (d) Pore view of 1 along the a direction. (e) Structural transformation of the as-synthesized → desolvated → rehydrated MOF as evident from PXRD analysis. (f) A close-up view of (e) reveals the differences in Bragg's reflection between the different phases.

The PXRD pattern of the as-synthesized 1 showed high crystallinity, and good correspondence with the simulated pattern indicates its purity in the bulk phase (Fig. 1e and f) and this was further supported by thermogravimetric analysis (TGA) (Fig. S1†). TGA of 1 revealed the loss of three water molecules at 127 °C with a weight loss of 5.14%, consistent with the calculated value (5.7%) and this phase is tenable up to 225 °C. At 225–280 °C, further weight loss (expt 3.22%, cald 3.8%) was accomplished due to the release of two bridging water molecules and the dehydrated framework is stable up to 300 °C (Fig. S1†). After complete dehydration, 1 transformed to an orthorhombic 1′ phase [cell parameters of a = 20.449 Å, b = 9.794 Å, c = 8.913 Å, α = β = γ = 90° and V = 1785.23 ų] with approximately 6.7% reduction in cell volume, suggesting overall structural contraction to a denser phase (Fig. 1e, f and Table S4†).³⁰ Further rehydration of 1′ (exposed to water vapor for 48 h) revives the as-synthesized phase as realized from the resemblances of PXRD patterns and indexing of the powder pattern [a = 7.081 Å, b = 13.802 Å, c = 20.864 Å, α = 98.486°, β = 96.655°, γ = 101.691° and V = 1952.52 ų] (Fig. 1e, f and Table S4†). The stability test of framework 1 in various solvents was conducted, including hot water (80 °C), dimethylformamide (DMF), methanol (CH₃OH), and chloroform (CHCl₃). The results, depicted in Fig. S3,† indicate structural stability under these conditions. However, the MOF exhibited structural degradation under acidic (1 M HCl) and basic conditions (1 M KOH).

Gas adsorption study

The gas adsorption properties were investigated with N₂ at 77 K and CO₂ at 195 K (Fig. 2a). There is no significant N₂ adsorption observed, indicating the non-porous nature of the dehydrated framework. In contrast, 1′ exhibited a gate-opening, multistep CO₂ adsorption profile at 195 K, with a total uptake of 61 cm³ g⁻¹ (equivalent to 11.9 wt% and 2.47 molecules per formula unit). The gated CO₂ uptake can be attributed to the smaller kinetic diameter of CO₂ (3.3 Å) compared to N₂ (3.64 Å) and the quadrupolar nature of CO₂, which facilitates interactions with the pore surface decorated with unsaturated Ni(II) sites and oxygen atoms from the bridging –OH and carboxyl groups.

Fig. 2 (a) N₂ and CO₂ adsorption isotherms of 1′ at 77 and 195 K (inset; log curve for a clear understanding of the close-to-open phase transition). (b) Adsorption isotherms of CO₂ and C₂H₂ at 298 K. (c) Isosteric heat of adsorption of CO₂ for 1′ at different loading concentrations with virial fitting in the inset. (d) Vapor adsorption isotherms of H₂O (298 K) and MeOH (293 K). In the adsorption isotherms, at points P and Q, the adsorbed samples were subjected to PXRD measurement which revealed a similar pattern to the as-synthesized framework after adsorbing water and methanol molecules as shown in Fig. S7.†

The adsorption isotherm showed negligible CO₂ uptake up to P/P₀ ∼ 0.25, followed by an abrupt increase in uptake and then saturating at P/P₀ = 0.99. The non-coincidence of desorption graph with respect to the adsorption isotherm with a broad hysteresis suggested strong confinement of CO₂ molecules inside the pore. CO₂ adsorption isotherms were also measured at 273 and 298 K (Fig. S4a†). The total uptake capacities are 49 cm³ g⁻¹ (9.8 wt% and 2.02 molecules per formula unit, at 273 K) and 37 cm³ g⁻¹ (7.8 wt% and 1.62 molecules per formula unit, at 298 K). The threshold gate opening pressure (P/P₀) shifted to 0.3 to 0.5 at 273 and 298 K, respectively (Fig. S4b†). The gate opening pressure increases with rising measurement temperatures. The isosteric heat of adsorption (Q_st) for CO₂ is calculated using the adsorption isotherms at 273 and 298 K, employing the Clausius–Clapeyron equation.^35–38 This calculation reveals a high value of 42.35 kJ mol⁻¹ at zero loading, indicating a strong adsorbate–adsorbent interaction (Fig. 2c). The value sharply increases at the high loading pressure of CO₂, suggesting structural expansion from a non-porous to an open porous structure. This transition from a non-porous to a porous state is further supported by high-pressure CO₂ adsorption isotherms. At pressures exceeding 1.0 bar, the adsorption profile displays a sharp increase resulting in a total adsorption capacity of 57 cm³ g⁻¹ at 273 K and 56 cm³ g⁻¹ at 298 K, finally reaching the same amount. This observation strongly confirms the CO₂ adsorption capabilities in the open phase. The dynamic pore chemistry and geometry, combined with the appropriate pore aperture of 3.05 × 3.57 Ų, motivated us to assess the adsorption capabilities of 1′ for various small, lightweight hydrocarbons (CH₄, C₂H₂, C₂H₄, C₂H₆) and H₂ under ambient conditions (Fig. 2b, S5a and Table S6†). However, the adsorption of C₂H₄ (4.16 Å) and C₂H₆ (4.44 Å) was excluded due to the inconvenience of size between the pore aperture and their kinetic diameters (Fig. 2b and c). Given that C₂H₂ possesses a rod-shaped anisotropic linear geometry similar to CO₂, with acidic hydrogen at both ends, it was anticipated that 1′ would also serve as a suitable adsorbent for C₂H₂. Surprisingly, 1′ showed minimal adsorption of C₂H₂ (4.1 cm³ g⁻¹), which is notably lower than the CO₂ uptake observed under similar conditions. This suggested that the interactions of C₂H₂ molecules with the framework may not be sufficiently strong to trigger the gate-opening mechanism of 1′. Specific interactions, including Lewis acid–base interactions, electrostatic interactions, and hydrogen bonding, between CO₂ molecules and the pore surface, adorned with unsaturated metal sites and various functional groups from the organic linker, are recognized as pivotal factors in achieving selective molecular sieving.³⁹ The adsorption kinetics of CO₂ was further assessed by fitting them into the Linear Driving Force (LDF) model. The rate constant value at the 2^nd point (P ∼ 4 kPa) is k = 2.21 × 10⁻² s⁻¹ (Fig. S6†). It is worth mentioning that N₂, H₂ and CH₄ molecules have either zero or lower quadrupole moment compared to CO₂ (Table S6†). C₂H₄ and C₂H₆ exhibit larger kinetic diameters compared to CO₂, and they also possess relatively lower quadrupole moments.⁴⁰ C₂H₂ and CO₂ have similar kinetic diameters and quadrupole moments.⁴¹ The large quadrupole moment of CO₂ can interact effectively with the dehydrated framework decorated with the unsaturated Ni(II) sites and oxygen atoms from carboxylate groups which help to diffuse the CO₂ molecule inside the pore.^22,42 This process results in the transformation from a non-porous to a porous state with gated sorption behaviour. In the case of C₂H₂, due to its inverse polarizability, it is expected to bind laterally, which hinders its entry into the pore due to its larger size (5.7 Å) along the z-direction.²² Therefore, for CO₂, it exhibits a discriminatory gate-opening sorption isotherm, which is further stabilized by strong host–guest interactions.

To better understand the nature of the pore surface, solvent vapour (water; H₂O (298 K) and methanol; CH₃OH (293 K)) adsorption has been carried out with activated 1 (1′). Notably, the H₂O vapor (with a kinetic diameter of 2.64 Å) adsorption isotherm for 1′ displayed a rapid and steep uptake in the low-humidity range, reaching a saturation uptake of 128 cm³ g⁻¹, which is equivalent to 5.1 molecules per formula unit and similar to the H₂O molecules (bridging and guest) present in the as-synthesized framework (Fig. 2d). This steep uptake in a relatively low-pressure region suggests a strong interaction between H₂O molecules and the pore surface, which is decorated with unsaturated Ni(II) centers, as well as oxygen atoms from the bridging –OH and –OCO groups from the 1,4-ndc linker. As anticipated, the pore exhibited less affinity for CH₃OH molecules, exhibiting a distinct gate-opening-type profile due to their larger size (3.62 Å) and lower polarity compared to H₂O (2.64 Å).

This observation is further supported by the βE₀ values obtained from the Dubinin–Radushkevich (D–R) equation^43,44 which reveal the adsorbate–adsorbent interactions. The values are found to be 7.8 and 2.5 kJ mol⁻¹ for H₂O and CH₃OH, respectively. To gain deeper insights into the transition in porosity upon guest loading, we conducted PXRD measurements on samples collected at two points, P and Q, during H₂O and CH₃OH vapor adsorption, respectively (Fig. S7†). The PXRD patterns at these points closely matched the diffraction patterns of the as-synthesized structure (1), indicating the expansion of the framework from the activated shrunk phase (1′) back to its original state (1, open-pore phase) to accommodate guest molecules. The adsorption isotherm and the observed structural transformation undeniably support the notion that the hydrophilic pore surface is biased for polar molecules, such as H₂O and CO₂.

In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) with CO₂ on 1′

To understand the mode of interactions, we have performed in situ DRIFTS with CO₂ over 1′. This experiment has been performed with an evenly coated activated sample on the glass disc placed in a removable vacuum tight sample compartment (evaluable/purgeable) against an MCT (mercury–cadmium–telluride) detector in an FT-IR (Vertex 70B) spectrometer. The sample compartment, made of zinc selenide (ZnSe), is resistant to high temperature and pressure. In this case, the CO₂ concentration in the chamber has been regulated through an MFC (mass flow controller). At first, the IR peaks attributed exclusively to molecular CO₂ were identified with increased dosing of only CO₂ (99.995% dry) in the chamber, in the absence of sample (Fig. S8–S10†). Further, the activated MOF was recorded in the chamber without CO₂ flow (Fig. S9†). To isolate the spectrum attributed exclusively to the CO₂⋯MOF interaction, the spectra from previous experiments were utilized as background references. For signals related to CO₂⋯MOF interaction, the CO₂ flow in the chamber was gradually increased under similar conditions. The emerging peaks became prominent with higher CO₂ dosing which is indicative of CO₂ interactions with the MOF. As depicted in Fig. 3, with a gradual increase in CO₂ flow, prominent IR features emerged when comparing the blue background (BG) spectrum to the other spectra in the dataset. As illustrated in Fig. 3 and S9,† the ν₃ mode of CO₂ antisymmetric stretching was observed in a narrow spectral interval of 2383–2300 cm⁻¹ (Table S7†).^45–49 Therefore, the appearance of strong bands when the sample is subjected to CO₂ pressure can be directly correlated with the interaction of CO₂ molecules with the available sites in the framework. Two strong asymmetric vibrations were observed at 2352 and 2331 cm⁻¹ (with a difference of ∼20 cm⁻¹), when only CO₂ was passed without the MOF. When CO₂ is flowed over activated 1′, four peaks are prominently observed with different intensities suggesting different modes of CO₂ interaction with the functional sites. The vibration frequencies observed are 2383 and 2362 cm⁻¹ (with a difference of 20 cm⁻¹) and 2340 and 2312 cm⁻¹ (with a difference of 28 cm⁻¹). According to the literature, all these vibrational frequencies can be correlated with the ν₃(CO₂) asymmetric modes of O–C–O interacting with open metal site Ni^II (Fig. 3).^45,50 The relative shifting of these peaks may originate from varying strengths of interactions between CO₂ and open metal sites (Ni^δ+⋯O^δ−–C–O type interaction). Likewise, the two weak bands at 1645 and 1455 cm⁻¹ may be attributed to carboxylate stretching frequencies.^51,52 Notably, activated 1′, in the absence of CO₂, showed carboxylate stretching frequencies of 1576 and 1369 cm⁻¹, indicating changes in the chemical environment of the 1,4-ndc-carboxylate group when dozed with CO₂ pressure (Fig. S9†). Several other bands were also observed in the range between 3734 and 3600 cm⁻¹, and their interpretation is complex. We believe that the presence of multiple peaks, such as those at 3734, 3723, and 3695 cm⁻¹, can be attributed to both OH vibrational frequencies and CO₂ overtones (ν₁ + ν₂).^53–55 The graph clearly illustrates that the peak at 3723 cm⁻¹ increases with the rising CO₂ flow in comparison to other peaks. Hence, in accordance with the available literature, this phenomenon can be attributed to either (1) CO₂ overtones or (2) the presence of bicarbonate (OOC⋯OH^δ−–M^δ+) in that specific region.^45,51,52,56 The in situ IR study confirmed that CO₂ adsorption occurs within the pores of 1′, facilitated primarily by Ni^II⋯O^δ−CO interaction with possible additional interaction sites, which is evident in the gated multistep CO₂ adsorption profile at 298 K.

Fig. 3 In situ DRIFTS of CO₂ adsorption in 1′ studied under ambient conditions. The X is the flow rate of CO₂ dosing inside the chamber to develop increased concentration inside the sample chamber. X = 10 L per hour. The schematic of CO₂–MOF interaction as anticipated from the adsorption, PXRD and DRIFTS study.

Theoretical calculations

We conducted density functional theory (DFT) calculations to investigate the interactions between the dehydrated framework and gas (CO₂ and C₂H₂) molecules. A fragmented model consisting of four Ni(II) centers (Ni1–Ni4) without a bridging water molecule was considered for DFT calculations (Table S10†). In the fragmented model Ni1 and Ni2 are the unsaturated metal sites, which would serve as the potential adsorption sites for CO₂ and C₂H₂.⁵⁰ Accordingly, CO₂ and C₂H₂ adsorbed models were designed for DFT calculations (Fig. 4 and Tables S8 and S9†). The results revealed that the CO₂ molecule could be bound to the Ni(II) centers (Ni1 and Ni2) through its O atom in an end-on fashion. The strong binding of the CO₂ molecule with Ni1 and Ni2 could be characterized by effective bond distances of 2.73 Å (O1–Ni1) and 2.84 Å (O1–Ni2), as depicted in Fig. 4a.⁵⁷ Additionally, weak interactions between the electron deficient carbon atom of CO₂ and two adjacent carboxylate oxygen atoms of the framework (2.96 and 2.87 Å) were also observed resulting in a slight deviation of CO₂ bond angle to 177.8°, which showed overall strong binding and stabilization of CO₂ molecules within the framework (Fig. 4a).^50,58 In contrast to this, C₂H₂ barely has any interaction with any Ni(II) or oxo-sites in the framework.

Fig. 4 Density functional theory (DFT) results showing (a) interactions between CO₂ and unsaturated Ni(II) sites along with carboxylate oxygen atoms and (b) no such interactions between C₂H₂ and the framework.

C₂H₂ could be located in a spatially isolated floating position, with a distance of ∼3.4 Å from Ni(II) sites (Fig. 4b and Table S9†). Moreover, the binding energy of the CO₂ and C₂H₂ molecules with the MOF was calculated to quantify the respective binding affinities. As anticipated, CO₂ exhibited a higher binding energy of −42.5 kJ mol⁻¹, significantly higher than that of C₂H₂ (ΔE = −5.8 kJ mol⁻¹), consistent with the earlier observations.³⁸ Therefore, the theoretical findings strongly support the experimentally observed selective uptake of CO₂ compared to C₂H₂. The Ni^II⋯OCO interaction observed in DRIFTS experiments aligns well with the predictions of DFT calculations, confirming a strong Ni^II⋯OCO interaction.

Adsorptive separation based on dynamic breakthrough measurement

In the following step, to project the selectivity of 1′ in the real time separation of industrially crucial CO₂/H₂, CO₂/N₂, CO₂/CH₄, and CO₂/C₂H₂ gas mixtures, a calculation was conducted using the ideal adsorbed solution theory (IAST) after fitting isotherms to the dual-site Langmuir–Freundlich equation at ambient temperature (298 K) (Fig. S11†). As depicted in Fig. 5, the adsorption selectivity for gas mixtures of CO₂/CH₄ and CO₂/C₂H₂ at 110 kPa and 298 K was calculated to be 1.33 and 1.5, respectively. This value is comparable to other reported values, and the uptake ratio is reasonably higher (∼9), comparable to a few other literature reports at saturated vapor pressure.^{15,21,22,59,60} The real-time breakthrough separation was performed by passing similar binary compositions through a packed column bed of 1′ with a 2.8–2.2 mL min⁻¹ flow rate (Fig. S12†). The breakthrough separation reveals that 1′ can achieve satisfactory separation of a CO₂/C₂H₂ (0.5 : 0.5; v/v) mixture, where C₂H₂ was first eluted and quickly approached purity without any detectable CO₂ (Fig. 6, S12†).

Fig. 5 IAST selectivity approximation studied from a 298 K adsorption isotherm by fitting it in a dual-site Langmuir–Freundlich (DSLF) equation.

Fig. 6 (a) The bi-cyclic dynamic breakthrough separation of equimolar C₂H₂/CO₂ for 1′ and (b) respective concentration of C₂H₂ in the outlet (plotting cycle 2 with 0 s start time) considering the flow rate through the bed of 2.8–2.2 mL min⁻¹.

For CO₂/C₂H₂ separation, CO₂ was retained in the column bed sufficiently longer compared to C₂H₂ (300 s after the initiation of feed gas dosing) until the material reached its saturation uptake (Fig. 6 and Table S13†). Upon saturation, CO₂ broke through, and corresponding outlet gas mixtures quickly reached equimolar proportions. Thereafter, by stopping feed mixtures and flowing a purge of carrier gas (He here) through the packed column the retained amount of CO₂ could be removed with high purity by making the column regenerated for subsequent separations. Most of the reported frameworks are selective towards C₂H₂ and as a result polymer-grade C₂H₂ can be only eluted in the downstream desorption phase of fixed bed adsorptive operation. Unlike the usual technique, fine matching of the dynamic pore of 1′ allows CO₂ sieving exclusively as pure C₂H₂ can be produced during the adsorption phase, leading to the separation becoming easier and more cost-effective. The breakthrough separation experiments of other gas mixtures such as CO₂/H₂ (0.5 : 0.5, v/v), CO₂/N₂ (0.85 : 0.15, v/v) and CO₂/CH₄ (0.5 : 0.5, v/v) were also performed where H₂, N₂ and CH₄ were first eluted and quickly approached high purity without any detectable CO₂ (Fig. S12†).

Conclusions

In conclusion, a contemporary three-dimensional Ni-based framework has been synthesized, characterized and studied for separating industrially important gas mixtures with similar size and physical properties. The structural flexibility and pore environment have been thoroughly investigated by several diffraction and adsorption tools (H₂O and MeOH). The framework exhibits multistep gated adsorption exclusively towards CO₂ due to likeliness in size, polarizability and alignment in the pore governed by unsaturated metal and carboxylate(O)⋯CO₂ interactions, further confirmed by in situ DRIFTS analysis and theoretical investigation. Such discriminatory sorption of CO₂ makes this framework highly selective from its industrial spin-off components in particular C₂H₂ and CH₄. The practical separation feasibility of this framework is further tested under continuous flowing conditions by using CO₂/C₂H₂, CO₂/CH₄, CO₂/N₂ and CO₂/H₂ gas mixtures at ambient temperature. Therefore, these results will provide a new strategy for designing and utilizing a dynamic porous framework for biased CO₂ capture and an inverse separation from the most important industrial analogue C₂H₂.

Data availability

All associated data are in the ESI† or deposited with CCDC (2310563).

Author contributions

N. S., S. L. and R. J. have performed synthesis and basic characterizations. N. S. and R. J. have analysed the single crystal structure, while N. S. has executed advanced characterizations such as sorption dependent PXRD and adsorption isotherms both in vapor and gaseous phases. S. L. has carried out hydrocarbon adsorption isotherms, selectivity study, breakthrough analysis and co-written the manuscript with N. S. S. L. and A. D. have organized the in situ DRIFTS instrumental assembly and performed spectroscopic measurements under varied conditions. F. A. R. has performed theoretical calculation. T. K. M. has supervised the overall work.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

All the authors thankfully acknowledge SERB, Dept. of Science and Technology (DST), Council of Scientific and Industrial Research (CSIR), Govt. of India for financial support (project no. SPR/2021/000592, CRG/2019/005951). The SAMat, ICMS, SSL research facility and Sheikh Saqr senior fellowship (T. K. M.) are also gratefully acknowledged.

Notes and references

1. P. J. Stang and F. Diederich, Modern acetylene chemistry, John Wiley & Sons, 2008.
2. J.-W. Zhang, M.-C. Hu, S.-N. Li, Y.-C. Jiang, P. Qu and Q.-G. Zhai, Chem. Commun., 2018, 54, 2012–2015.
3. Y. Ye, Z. Ma, R.-B. Lin, R. Krishna, W. Zhou, Q. Lin, Z. Zhang, S. Xiang and B. Chen, J. Am. Chem. Soc., 2019, 141, 4130–4136.
4. R. Matsuda, R. Kitaura, S. Kitagawa, Y. Kubota, R. V. Belosludov, T. C. Kobayashi, H. Sakamoto, T. Chiba, M. Takata, Y. Kawazoe and Y. Mita, Nature, 2005, 436, 238–241.
5. C. R. Reid and K. M. Thomas, Langmuir, 1999, 15, 3206–3218.
6. M. Bühl and G. Wipff, ChemPhysChem, 2011, 12, 3095–3105.
7. S. Roy, A. Chakraborty and T. K. Maji, Coord. Chem. Rev., 2014, 273–274, 139–164.
8. H. Li, K. Wang, Y. Sun, C. T. Lollar, J. Li and H.-C. Zhou, Mater. Today, 2018, 21, 108–121.
9. A. Hazra, A. Jain, M. S. Deenadayalan, S. A. Adalikwu and T. K. Maji, Inorg. Chem., 2020, 59, 9055–9064.
10. A. Chakraborty, S. Roy, M. Eswaramoorthy and T. K. Maji, J. Mater. Chem. A, 2017, 5, 8423–8430.
11. A. Hazra, S. Jana, S. Bonakala, S. Balasubramanian and T. K. Maji, Chem. Commun., 2017, 53, 4907–4910.
12. R. Haldar, N. Sikdar and T. K. Maji, Mater. Today, 2015, 18, 97–116.
13. J.-P. Zhang and X.-M. Chen, J. Am. Chem. Soc., 2009, 131, 5516–5521.
14. K.-J. Chen, H. S. Scott, D. G. Madden, T. Pham, A. Kumar, A. Bajpai, M. Lusi, K. A. Forrest, B. Space, J. J. Perry and M. J. Zaworotko, Chem, 2016, 1, 753–765.
15. R.-B. Lin, L. Li, H. Wu, H. Arman, B. Li, R.-G. Lin, W. Zhou and B. Chen, J. Am. Chem. Soc., 2017, 139, 8022–8028.
16. A. Luna-Triguero, J. M. Vicent-Luna, R. M. Madero-Castro, P. Gómez-Álvarez and S. Calero, ACS Appl. Mater. Interfaces, 2019, 11, 31499–31507.
17. W. Fan, S. Yuan, W. Wang, L. Feng, X. Liu, X. Zhang, X. Wang, Z. Kang, F. Dai, D. Yuan, D. Sun and H.-C. Zhou, J. Am. Chem. Soc., 2020, 142, 8728–8737.
18. H.-G. Hao, Y.-F. Zhao, D.-M. Chen, J.-M. Yu, K. Tan, S. Ma, Y. Chabal, Z.-M. Zhang, J.-M. Dou, Z.-H. Xiao, G. Day, H.-C. Zhou and T.-B. Lu, Angew. Chem., Int. Ed., 2018, 57, 16067–16071.
19. J. Wang, Y. Zhang, Y. Su, X. Liu, P. Zhang, R.-B. Lin, S. Chen, Q. Deng, Z. Zeng, S. Deng and B. Chen, Nat. Commun., 2022, 13, 200.
20. R. Krishna, ACS Omega, 2020, 5, 16987–17004.
21. W. Yang, A. J. Davies, X. Lin, M. Suyetin, R. Matsuda, A. J. Blake, C. Wilson, W. Lewis, J. E. Parker, C. C. Tang, M. W. George, P. Hubberstey, S. Kitagawa, H. Sakamoto, E. Bichoutskaia, N. R. Champness, S. Yang and M. Schröder, Chem. Sci., 2012, 3, 2993–2999.
22. M. L. Foo, R. Matsuda, Y. Hijikata, R. Krishna, H. Sato, S. Horike, A. Hori, J. Duan, Y. Sato, Y. Kubota, M. Takata and S. Kitagawa, J. Am. Chem. Soc., 2016, 138, 3022–3030.
23. S.-Q. Yang, R. Krishna, H. Chen, L. Li, L. Zhou, Y.-F. An, F.-Y. Zhang, Q. Zhang, Y.-H. Zhang, W. Li, T.-L. Hu and X.-H. Bu, J. Am. Chem. Soc., 2023, 145, 13901–13911.
24. Y. Shi, Y. Xie, H. Cui, Y. Ye, H. Wu, W. Zhou, H. Arman, R.-B. Lin and B. Chen, Adv. Mater., 2021, 33, 2105880.
25. Y. Xie, H. Cui, H. Wu, R.-B. Lin, W. Zhou and B. Chen, Angew. Chem., Int. Ed., 2021, 60, 9604–9609.
26. S. Krause, N. Hosono and S. Kitagawa, Angew. Chem., Int. Ed., 2020, 59, 15325–15341.
27. A. Schneemann, V. Bon, I. Schwedler, I. Senkovska, S. Kaskel and R. A. Fischer, Chem. Soc. Rev., 2014, 43, 6062–6096.
28. T. K. Maji, R. Matsuda and S. Kitagawa, Nat. Mater., 2007, 6, 142–148.
29. J. D. Evans, V. Bon, I. Senkovska, H.-C. Lee and S. Kaskel, Nat. Commun., 2020, 11, 2690.
30. Y. Takashima, V. M. Martínez, S. Furukawa, M. Kondo, S. Shimomura, H. Uehara, M. Nakahama, K. Sugimoto and S. Kitagawa, Nat. Commun., 2011, 2, 168.
31. T. K. Maji, K. Uemura, H.-C. Chang, R. Matsuda and S. Kitagawa, Angew. Chem., Int. Ed., 2004, 43, 3269–3272.
32. S. Bhattacharyya and T. K. Maji, Coord. Chem. Rev., 2022, 469, 214645.
33. N. Sikdar, S. Bonakala, R. Haldar, S. Balasubramanian and T. K. Maji, Chem.–Eur. J., 2016, 22, 6059–6070.
34. N. Sikdar, A. Hazra and T. K. Maji, Inorg. Chem., 2014, 53, 5993–6002.
35. A. Nuhnen and C. Janiak, Dalton Trans., 2020, 49, 10295–10307.
36. L. W. Bingel and K. S. Walton, Langmuir, 2023, 39, 4475–4482.
37. M.-H. Yu, B. Space, D. Franz, W. Zhou, C. He, L. Li, R. Krishna, Z. Chang, W. Li, T.-L. Hu and X.-H. Bu, J. Am. Chem. Soc., 2019, 141, 17703–17712.
38. S. Laha, N. Dwarkanath, A. Sharma, D. Rambabu, S. Balasubramanian and T. K. Maji, Chem. Sci., 2022, 13, 7172–7180.
39. M. Chen, S. Chen, W. Chen, B. E. G. Lucier, Y. Zhang, A. Zheng and Y. Huang, Chem. Mater., 2018, 30, 3613–3617.
40. J.-R. Li, R. J. Kuppler and H.-C. Zhou, Chem. Soc. Rev., 2009, 38, 1477–1504.
41. J. M. Junquera-Hernández, J. Sánchez-Marín and D. Maynau, Chem. Phys. Lett., 2002, 359, 343–348.
42. K. A. Moltved and K. P. Kepp, J. Phys. Chem. C, 2019, 123, 18432–18444.
43. N. D. Hutson and R. T. Yang, Adsorption, 1997, 3, 189–195.
44. T. K. Maji, G. Mostafa, R. Matsuda and S. Kitagawa, J. Am. Chem. Soc., 2005, 127, 17152–17153.
45. K. I. Hadjiivanov, D. A. Panayotov, M. Y. Mihaylov, E. Z. Ivanova, K. K. Chakarova, S. M. Andonova and N. L. Drenchev, Chem. Rev., 2021, 121, 1286–1424.
46. M. Mihaylov, K. Chakarova, S. Andonova, N. Drenchev, E. Ivanova, A. Sabetghadam, B. Seoane, J. Gascon, F. Kapteijn and K. Hadjiivanov, J. Phys. Chem. C, 2016, 120, 23584–23595.
47. S. Andonova, S. S. Akbari, F. Karadaş, I. Spassova, D. Paneva and K. Hadjiivanov, J. CO2 Util., 2021, 50, 101593.
48. N. R. Walker, R. S. Walters, G. A. Grieves and M. A. Duncan, J. Chem. Phys., 2004, 121, 10498–10507.
49. W. Chen, Z.-R. Li, D. Wu, F.-L. Gu, X.-Y. Hao, B.-Q. Wang, R.-J. Li and C.-C. Sun, J. Chem. Phys., 2004, 121, 10489–10494.
50. P. D. C. Dietzel, R. E. Johnsen, H. Fjellvåg, S. Bordiga, E. Groppo, S. Chavan and R. Blom, Chem. Commun., 2008, 41, 5125–5127.
51. W. Zhou, Z. Ma, S. Guo, M. Wang, J. Wang, M. Xia, L. Jia, B. Hou, D. Li and Y. Zhao, Appl. Surf. Sci., 2018, 427, 867–873.
52. L. Proaño, E. Tello, M. A. Arellano-Trevino, S. Wang, R. J. Farrauto and M. Cobo, Appl. Surf. Sci., 2019, 479, 25–30.
53. S. Eckle, H.-G. Anfang and R. J. Behm, J. Phys. Chem. C, 2011, 115, 1361–1367.
54. G. Garbarino, D. Bellotti, E. Finocchio, L. Magistri and G. Busca, Catal. Today, 2016, 277, 21–28.
55. Q. Guan, X. Cheng, R. Li and W. Li, J. Catal., 2013, 299, 1–9.
56. J. Baltrusaitis, J. H. Jensen and V. H. Grassian, J. Phys. Chem. B, 2006, 110, 12005–12016.
57. S. A. FitzGerald, J. M. Schloss, C. J. Pierce, B. Thompson, J. L. C. Rowsell, K. Yu and J. R. Schmidt, J. Phys. Chem. C, 2015, 119, 5293–5300.
58. K. Jayaramulu, S. K. Reddy, A. Hazra, S. Balasubramanian and T. K. Maji, Inorg. Chem., 2012, 51, 7103–7111.
59. J. Lee, C. Y. Chuah, J. Kim, Y. Kim, N. Ko, Y. Seo, K. Kim, T. H. Bae and E. Lee, Angew. Chem., Int. Ed., 2018, 57, 7869–7873.
60. Y. Du, Y. Chen, Y. Wang, C. He, J. Yang, L. Li and J. Li, Sep. Purif. Technol., 2021, 256, 117749.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental details, crystallographic details, TGA, adsorption isotherms, and DRIFTS data. CCDC 2310563. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3sc06611h

‡ These authors contributed equally.

§ Present address: Department of Chemistry, School of Science, Gandhi Institute of Technology and Management (GITAM), Hyderabad 502329, India.

---