Impact of Ni(II) coordinatively unsaturated sites and coordinated water molecules on SO₂ adsorption by a MOF with octanuclear metal clusters

Abstract

A Ni-based pyrazolate MOF (NiBDP) is studied for SO₂ adsorption under static conditions, demonstrating a high SO₂ uptake of 8.48 mmol g⁻¹ at 298 K and 1 bar while maintaining a high chemical stability. The influence of Ni(II) coordinatively unsaturated metal sites and coordinated water on the SO₂ adsorption performance of this MOF is investigated by using a combination of experimental techniques, including FTIR and in situ DRIFTS measurements, along with Density Functional Theory calculations. The pore-filling of the SO₂ adsorbates within the material, at the molecular level, is further unravelled through grand Canonical Monte Carlo simulations employing a newly DFT-derived accurate set of force field parameters.

---

Introduction

The standard families of porous materials, e.g., activated carbons, zeolites, and mesoporous silica, have been extensively studied for the adsorption of small molecules within their porous cavities for diverse applications.^1,2 More recently, metal–organic frameworks (MOFs), constructed by metal-containing nodes and organic linkers,³ which can propagate in one to three dimensions, have been shown to represent a milestone in the capture of greenhouse (CH₄ and CO₂)^4–11 and toxic gases (NH₃, NO_x and SO₂).^12,13 The diversity of metal ions and organic ligands that can be used for the construction of MOFs facilitates their architecture design and, hence, the modulation of chemical features in the pore walls (e.g., functionalization) and of pore dimension/shape to optimize the interactions with a given guest molecule.¹⁴ Therefore, a selection of MOFs has emerged as promising candidates in the field of material science for the adsorption of toxic and corrosive gases.¹⁵ Typically, SO₂ has gained a lot of interest since it is the main precursor of acid rain, and the hazardous effects of SO₂ on the environment and on human health are a huge concern. Thus, SO₂ capture by MOFs has been investigated since 2008, when Yaghi et al.,¹⁶ reported the SO₂ adsorption by several MOFs with diverse chemical and structural features. After two decades of worldwide efforts, certain chemical characteristics have been recognized to render MOF materials not only resistant to SO₂ exposure but also with attractive capture performance.^17–19 These included the use of (i) high valent-metal ions, such as Zr⁴⁺, Al³⁺, In³⁺, and Sc³⁺ with carboxylate ligands; (ii) robust metal clusters with strong metal–ligand bonds, for example, the use of pyrazolate or imidazolate ligands and late transition M²⁺ ions; and (iii) the inclusion of coordinately unsaturated metal sites (CUS) or functional groups (X–H), capable of interacting with SO₂ by coordination bonds,²⁰ hydrogen bonds,²¹ and other supramolecular interactions.²²

Nickel(II)-based MOFs are a priori attractive sorbents owing to the good abundance of this metal and their usually high thermal and chemical stabilities. Moreover, from a chemical point of view, Ni-based compounds are known to exhibit interesting properties such as electrocatalytic capabilities,²³ magnetic properties,²⁴ energy storage,²⁵ and detection.²⁶ Although some M²⁺-based MOFs are considered unstable to SO₂ adsorption, the combination of polyazolate ligands and divalent metals increases the MOF structural stability over alkaline or acidic environments.²⁷ MOFs formed by strong Ni²⁺–N(azolate) bonds provide greater chemical and thermal stability compared to their carboxylate analogues. According to Pearson's hard–soft acid–base principle, a highly chemically stable MOF material can be constructed using azolate ligands with low-valent transition metals. The pK_a value for the deprotonation of N–H bonds shows high structure robustness.²⁸ These characteristics make attractive the study of Ni²⁺ pyrazolate-based MOFs for the adsorption of corrosive gases.²⁹

Herein, we investigate a Ni(II)-based MOF, [Ni₈(OH)₄(H₂O)₂(BDP)₆] termed NiBDP (H₂BDP = 1,4-bis(pyrazol-4-yl)benzene), for SO₂ capture. This material is based on octanuclear Ni(II) hydroxo clusters [Ni₈(OH)₄(H₂O)₂], linked by BDP to form [Ni₈(OH)₄(H₂O)₂(BDP)₆] cluster (Fig. 1a). NiBDP material, along with an isoreticular series of MOFs, has been studied earlier for dynamic breakthrough SO₂ adsorption.³⁰ Results showed an SO₂ uptake of 2.02 mmol g⁻¹ at 303 K, and cycling experiments revealed a reversible physisorption process.

Fig. 1 (a) Structure of the octanuclear Ni(II) cluster in [Ni₈(OH)₄(H₂O)₂(BDP)₆]. (b) View of the tetrahedral, and (c) octahedral cages found in the crystal structure of NiBDP MOF. Color code: Ni (green); N (blue); O (red); C (grey) and H (white).

Although this Ni(II)-based MOF was previously investigated under SO₂ (dynamic breakthrough), we decided to explore its SO₂ adsorption capacities under static condition at room temperature and 1 bar, in order to gain a comprehensive understanding on the SO₂ adsorption mechanism. Also, neither experimental exploration in thermodynamic conditions nor a complete understanding of the microscopic mechanism in play have been reported so far in the literature. Thus, we have embarked on the task of coupling advanced experimental and theoretical approaches to revisit the SO₂ adsorption performance of NiBDP in static mode and gain a full picture of the adsorption mechanism at the molecular level. A special attention has been also paid to assess computationally the impact of coordinated water on the SO₂ adsorption isotherms by considering different NiBDP-nH₂O structures (where n = 0, 1, 2, correspond to distinct structure models possessing 100%, 50%, and zero CUS sites per metal node) that mimic scenarios of full, partial or non-evacuation of the CUS sites during the experimental activation of the sample.

Experimental

Synthesis of NiBDP

NiBDP was synthesized according to the reported procedure.³¹ In a typical synthesis, 63.1 mg (0.3 mmol) of 4,4′-benzene-1,4-diylbis(1H-pyrazole) (H₂BDP) were dissolved in 16 mL of DMF, and 99.2 mg (0.4 mmol) of Ni(AcO)₂·4H₂O were dissolved in 4 mL of H₂O. The two limpid solutions were mixed and refluxed for 6 h under stirring. The light green solid obtained was filtered and washed with EtOH and Et₂O, yielding 89 mg (66%). IR (KBr) 3433(br), 1655(vs.), 1574(s), 1390(m), 1358(m), 1250(s), 1182(w), 1124(w), 1057(s), 958(s), 845(m), 658(w), 540(w), 505(w) cm⁻¹.

Characterization methods

Detailed information on the instrumental techniques is available in the ESI: Section S1.†

SO₂ sorption measurements

SO₂ adsorption–desorption isotherms were measured in a Dynamic Gravimetric Gas/Vapour Sorption Analyser, DVS vacuum (Surface Measurement Systems Ltd) with a static method. The samples were activated at 423 K under vacuum (1 × 10⁻⁶ bar) for 5 hours. SO₂ adsorption isotherm was carried out at 298 K up to 1 bar.

DFT geometry optimization

Periodic DFT calculations were conducted to optimize the atomic position and cell volume of both the empty NiBDP-nH₂O structures (where n = 0, 1, 2, correspond to different structure models possessing 100%, 50%, and zero CUS sites per metal node) and the associated guest-loaded structures with SO₂ adsorbed onto them using Vienna Ab initio Simulation Package (VASP).³² The core and valence electrons of the interacting elements were treated with projector-augmented-wave (PAW) potentials and plane-wave basis sets.³³ The general gradient approximation (GGA) to the exchange–correlation functional according to Perdew–Burke–Ernzerhof (PBE)³⁴ was used along with a combination of Grimme's empirical dispersion correction with Becke–Johnson damping (DFT-D3/BJ).³⁵ To account for the electronic correlation of d electrons on-site, Hubbard U correction was considered for Ni(II) atoms, with the benchmark U value of 6.4 eV previously validated on different properties of Ni(II) oxide systems as well as Ni(II) MOFs.^36–38Γ point meshes for sampling the Brillouin zone were used along a plane-wave kinetic energy cutoff of 520 eV. The energy and ionic force convergence criteria were set to 1 × 10⁻⁶ eV and 0.02 eV Å⁻¹, respectively. Consequently, the atomic partial charges of the MOF atoms of the DFT-optimized empty MOF structures were derived by employing the density-derived electrostatic and chemical (DDEC) charge partitioning scheme as implemented in the chargemol program.^39,40 These atomic partial charges are provided in Table S1 in the ESI.† The textural properties of NiBDP-nH₂O (n = 0, 1, 2) MOFs, including pore volume (PV), surface area (SA), largest cavity diameter (LCD) and pore limiting diameter (PLD), were calculated by using Zeo++ software (see details for the calculations in the ESI†),⁴¹ and tabulated in Table S2.†

Force field parameterization

The consideration of generic force field for MOF frameworks combined with force field models for the guest molecules is known to dramatically fail to describe specific interactions between coordinative unsaturated MOFs with any guest molecules.^42–46 Therefore, such MOF- CUS sites/guest molecule interactions require a specific FF parameterization based on first-principle quantum calculations. To do so, a 100% CUS containing model of the NiBDP MOF (i.e., NiBDP-0H₂O) was first considered and an approaching path consisting of 10 different configurations associated to different SO₂–Ni CUS (NiBDP-0H₂O) separating distances were constructed while maintaining the same orientation as DFT-optimized minimum energy structure (Fig. S9†). Subsequently, single point energy calculations were performed to deduce the corresponding interaction energies for a SO₂ molecule within the NiBDP-0H₂O pore using the energy expression: E_int = E_{NiBDP-0H2O+SO2} − (E_NiBDP-0H2O + E_SO2), where, E_{NiBDP-0H2O+SO2} corresponds to the single point energies of the SO₂ loaded NiBDP-0H₂O adsorption configurations along the interacting path, while E_NiBDP-0H2O and E_SO2 correspond to the energy associated to an empty NiBDP-0H₂O MOF and an isolated SO₂ molecule, respectively. The resulting interaction energy curve was then fitted to an analytical function combining van der Waals (vdW) and coulombic terms. Here, non-bonded interactions between SO₂ and NiBDP-0H₂O MOF atoms other than Ni(II) sites were modeled with standard Lennard-Jones (LJ) 12-6 potential while a Morse potential was employed to describe the interactions between SO₂ and Ni(II) atoms. The LJ parameters (ε and σ) for the MOF atoms were adopted from the universal force field (UFF),⁴⁷ whereas the SO₂ molecule was represented by the three-site LJ charged model reported by Ketko et al.,⁴⁸ (Tables S3–S5†). LJ cross-terms were then calculated using the Lorentz–Berthelot combination rules. The Ewald summation technique⁴⁹ was employed to compute the long-range coulombic contributions, in which, the atomic partial charges of the MOF are determined using the DDEC charge partitioning scheme as mentioned above, while the charges for SO₂ are obtained from the potential model of Ketko et al.⁴⁸ Finally, the Morse potential parameters, D, α, and r₀ associated with the Ni(II)–O(SO₂)– and Ni(II)–S(SO₂) specific interactions were determined using the General Utility Lattice Program (GULP)⁵⁰ by minimizing the energy differences with respect to the DFT counterpart. The corresponding fitting curve and the parameters are available in the ESI.†

Monte Carlo simulations

The SO₂ adsorption isotherms for all three NiBDP-nH₂O (n = 0, 1, 2) structures were calculated through the GCMC simulations implementing the force field discussed above using the Complex Adsorption and Diffusion Simulation Suite (CADSS) of the program.⁴² Nonetheless, for the NiBDP-2H₂O structure, the SO₂ adsorption isotherm was computed by using generic UFF parameters for all framework atoms, as the system does not contains any CUS sites. We employed a simulation box comprising eight-unit cells (2 × 2 × 2) of the MOFs. The atomic positions of the MOFs' skeletons were held fixed during the adsorption, and Monte Carlo trial moves for the guest molecules. The host–guest nonbonded interactions were computed in real space using a cutoff of 12.0 Å. The long-range electrostatic interactions were calculated using the Ewald summation technique⁴⁹ with an accuracy of 1 × 10⁻⁶. For each pressure point, 1 × 10⁷ and 2 × 10⁷ Monte Carlo steps including insertion/deletion, translation and rotational moves, were considered for equilibration and production, respectively. The Peng–Robinson equation of state was used to determine the gas-phase fugacity for SO₂ molecules.⁵¹ The overall adsorption isotherms were obtained by running GCMC simulations for different fugacities ranging from 10⁻⁴ to 1 bar. Additionally, the enthalpy of SO₂ adsorption at infinite dilution was computed employing Widom's test particle insertion method.⁵²

Results and discussion

Characterization of NiBDP

NiBDP MOF was synthesized according to the reported procedure.³¹ This structure crystallizes in the face cubic centered (fcu) topology with the Fm3̄m space group. Each Ni(II) ion is bonded to three N atoms from different organic linkers, and three of the six μ₄-O observed in the octanuclear cluster (Fig. 1a). It is important to note that two of the six oxygen atoms belong to coordinated water molecules rather than hydroxo ligands.⁵³ The structure contains two types of cages: an octahedral cage (diameter of ∼16 Å) connecting to eight tetrahedral cages (diameter of ∼9 Å) accessible via triangular windows (diameter of ∼6 Å), (Fig. 1b, c and S1†). In the octahedral cage, the OH groups point toward the centre of the pore, which can favor interactions with SO₂ in addition to the Ni(II) sites, as it was demonstrated for other materials, such as the MFM-300 family.⁵⁴ The crystalline structure was confirmed by PXRD (Fig. S2†). Based on N₂ sorption measurements at 77 K (Fig. S3†), the activated NiBDP shows a BET surface area of 1220 m² g⁻¹, which is lower than the previously reported (1730 m² g⁻¹).⁵⁵ The corresponding theoretically calculated surface areas of NiBDP_0H₂O and NiBDP_2H₂O are 2650 m² g⁻¹ and 2480 m² g⁻¹, respectively, suggesting that the CUS sites of the investigated sample are not fully evacuated. In addition, elemental analysis (EA) of NiBDP (Ni₈(OH)₄(H₂O)₂(BDP)₆](H₂O)₂(DMF)₁₀) was conducted. Anal. calc. for NiBDP (2589.90 g mol⁻¹) C, 47.3; H, 5.06; N, 18.4; Ni, 18.1, found C, 46.7; H, 5.10; N, 18.2; Ni, 18.8. Nickel content was determined on the basis of TGA (Fig. S4†). Also, electronic spectra (Fig. S5†) characterization is indicative of the presence of the characteristic d–d transition bands of Ni(II) in a poorly distorted Oh environment (³T_2g(F) ← ³A_2g(F), 10 350 cm⁻¹; ³T_1g(F) ← ³A_2g(F), 15 870 cm⁻¹; ³T_1g(P) ← ³A_2g(F), 25 800 cm⁻¹).

SO₂ sorption in NiBDP

Previously, the SO₂ capture in NiBDP was reported using breakthrough experiments for an N₂/SO₂ mixture. However, under these conditions, a relatively low SO₂ concentration (P_SO2 = 0.025 bar) was considered. NiBDP showed a total uptake of 2.02 mmol g⁻¹ at 303 K. The SO₂ affinity was attributed to the Lewis acid–base interactions with crystal defects.³⁰ However, equilibrium adsorption–desorption isotherms are more appropriate to gain deep insights into the adsorption capacity, adsorption mechanism and to understand the adsorption behavior of certain gases within a porous material.⁵⁶ Thus, the SO₂ adsorption–desorption isotherms were investigated for NiBDP. The corresponding SO₂ sorption isotherm is shown in Fig. 2. A steep profile is observed in the adsorption isotherm from 0 to 0.025 bar. At this stage, SO₂ adsorption reaches a value of 1.85 mmol g⁻¹ and remains almost constant up to 0.45 bar (1.89 mmol g⁻¹). This SO₂ uptake is similar to that reported so far for other MOFs like MFM-133 (1.2 mmol g⁻¹),⁵⁷ MFM-422 (1.8 mmol g⁻¹),⁵⁷ SIFSIX-3-Zn (1.89 mmol g⁻¹),⁵⁸ and ELM-12 (1.95 mmol g⁻¹)⁵⁹ at 0.1 bar implying that NiBDP can be used as an alternative SO₂ adsorbent at low gas pressures. Interestingly, a sharp jump in the SO₂ adsorption capacity from 1.89 mmol g⁻¹ up to 7.22 mmol g⁻¹ is observed at 0.45 bar, which ultimately reaches a total SO₂ uptake of 8.48 mmol g⁻¹ at 1 bar. The latter value is comparable to a variety of MOF materials such as UiO-66-Cu,⁵⁷ MFM-300(In)⁶⁰ and MIL-53(Al)-TDC⁶¹ with a reported SO₂ uptake of 8.2 mmol g⁻¹, 8.28 mmol g⁻¹, and 8.9 mmol g⁻¹, respectively, under similar working conditions (1 bar and 298 K). Considering the experimental BET surface area of NiBDP (1220 m² g⁻¹), the SO₂ uptake of 8.48 mmol g⁻¹ aligns well with other microporous MOFs with similar surface area, for instance, UiO-66, possessing the same fcu topology and a BET surface area of 1221 m² g⁻¹, can adsorb 8.6 mmol g⁻¹ of SO₂ at 298 K and 1 bar.⁵⁷

Fig. 2 Experimental SO₂ adsorption–desorption isotherm of an activated NiBDP sample (filled green circles = adsorption; open green circles = desorption) at 298 K up to 1 bar.

Interestingly, the adsorption isotherm profile is unusual for a well-defined microporous MOF. In the low-pressure region, a rapid adsorption profile is defined, which can be related to a high interaction between the Ni²⁺ CUS sites and the SO₂ molecule, which can be postulated as a possible main adsorption site for NiBDP and gives rise to adsorbate seeding in the pore structure.⁶¹ Then, a subsequent almost horizontal plateau is observed, which can be associated with the presence of weaker interaction sites such as water molecules and hydroxide groups coordinated to the Ni²⁺ sites. Later, the sudden step at 0.45 bar followed by a linear adsorption (SO₂ saturation up to 1 bar) should be related to an adsorbate clustering, giving rise to pore filling. The observed seeding and clustering adsorption behavior can be related to the water adsorption mechanism unraveled for the MOF-303 system.⁶²

Finally, the retention of NiBDP crystallinity after SO₂ adsorption was corroborated by PXRD analysis (Fig. S2†), demonstrating that the inclusion of polyazolate-containing ligands and borderline soft acid Ni²⁺ in MOFs improves their structural stability against corrosive gases.⁶³

Furthermore, the isosteric enthalpy of adsorption (ΔH) for SO₂ adsorption was calculated at low coverage using the virial method (Fig. S6†).⁶⁴ The calculated value at low SO₂ loading (<0.03 bar) is −41.2 kJ mol⁻¹, which is in good agreement with the value reported in earlier (−40.1 kJ mol⁻¹).³⁰ This relatively high value can be associated with moderately strong SO₂ interaction with NiBDP. Typically, a MOF material with CUS sites usually shows a high ΔH value (−80/−90 kJ mol⁻¹) due to the strong coordination between the metal and the SO₂ molecule.²⁶ The lower value obtained here can be related to a decrease in the availability of Ni²⁺ CUS sites due to the presence of coordinated water.

FTIR and DRIFTS analysis in NiBDP

FTIR experiments were carried out on a NiBDP sample before and after exposure to SO₂ gas and moisture traces to confirm the stability (Fig. S7†). The FTIR spectra of NiBDP show the characteristic bands for a pyrazolate MOF material. The bands related to C=O and C=C are associated with the peaks at 1649 and 1574 cm⁻¹, respectively.⁶⁵ The bands at 1410 and 1385 cm⁻¹ are related to C=N. The bands at 1240 and 510 cm⁻¹ are assigned to N–N bonds in the organic linker and the coordination bonds between Ni and N, respectively.⁶⁶ Also, after SO₂ adsorption show different bands from the region 1800–450 cm⁻¹ (Fig. 3a). Two principal SO₂ bands at 1153 cm⁻¹ and 1303 cm⁻¹ are observed, that are related to asymmetric and symmetric stretching modes of SO₂ molecules within the pores. This type of interaction agrees with SO₂ physisorption on Ni(BDC)(TED)_0.5 with bands at 1144 and 1326 cm⁻¹.⁶⁷ Finally, the formation of hydrolyzed species of SO₂ (i.e., SO₃²⁻ or HSO₃²⁻) due to the presence of water is discarded due to the lack of characteristic IR bands for these ions.

Fig. 3 (a) FT-IR spectra of activated and SO₂-loaded NiBDC material, into 1800–400 cm⁻¹ wavelength region. Gray rectangles show the asymmetric ν_as and symmetric ν_s stretch modes of physically adsorbed SO₂ molecules. (b) IR spectra on activated NiBDP at 70 °C (blue line) and 190 °C (green line) in the region between 2300 and 2000 cm⁻¹ after 5 min of CO flux. (c) Subtracted spectra before and after admittance of CO in the region between 2300 and 2000 cm⁻¹ after 5 min of CO flux.

In addition, in situ DRIFTS measurements were conducted using CO as a probe molecule. This technique is a powerful tool to characterise acid/base sites for different porous materials, including MOFs.⁶⁸ Even though the chemical nature of CO is different to SO₂, this study is useful to gain insight into the possible adsorption sites for the guest molecules, as previously reported⁶⁹ in particular to make the distinction between the interactions with Ni²⁺ CUS sites or the coordinated water molecules. Herein, the CO adsorption was studied by DRIFTS in situ analysis at 298 K after sample activation at two different temperatures, 70 and 190 °C under a He atmosphere.

The IR spectra of the activated sample are shown in Fig. S8.† When the sample was activated at 70 °C, broadband was observed between 3450 and 3200 cm⁻¹. On the other hand, when the sample was activated at 190 °C, this band partially decreased in intensity. This band is associated with water molecules that are not fully desorbed from the Ni²⁺ centers. Fig. 3b shows the spectra after 5 min of a continuous CO flux. The CO gas spectrum is composed of two asymmetric peaks centered at 2173 and 2113 cm⁻¹. When the CO is absorbed by the MOF, these two peaks are also observed, but the shape changes. The principal difference between the samples activated at 70 °C and 190 °C seems to be a peak depicted at 2200 cm⁻¹. However, compared to the blank spectra (Fig. 3b, inset), it is observed that this peak belongs to the sample.

To analyze the difference in CO adsorption between the two activation methods, a subtraction of the spectra before and after the admittance of CO was carried out. Fig. 3c shows the subtracted spectra. It is evident that a peak appears at 2109 cm⁻¹ when the sample is activated at 70 °C. This peak was already assigned to the adsorption of CO over metallic Ni.⁷⁰ Therefore, the DRIFT analysis confirms that both Ni²⁺ CUS sites and coordinated water can interact with the guest molecules in activated NiBDP sample.

SO₂ adsorption simulations

To further gain insights into the effect of Ni²⁺ CUS sites and coordinated water molecules in the SO₂ adsorption, the adsorption of SO₂ in NiBDP with and without evacuated Ni(II) CUS sites were systematically studied through GCMC simulations to fully understand the resulting adsorption behavior. As a starting point, we computed the SO₂ single component adsorption isotherm for NiBDP-2H₂O MOF at 298 K, covering a pressure range from 1 × 10⁻⁴ to 1 bar, and by using the generic UFF potential parameters as the system does not contain CUS sites. As illustrated in Fig. 4, our GCMC simulations result in an SO₂ uptake of 15.2 mmol g⁻¹ for this NiBDP-2H₂O structure at 1 bar, which exceeds the experimentally determined maximum uptake of 8.48 mmol g⁻¹. This deviation is expected due to the difference in the experimental BET surface area (1220 m² g⁻¹) and theoretical N₂ surface area (NiBDP_0H₂O and NiBDP_2H₂O of 2650 m² g⁻¹ and 2484 m² g⁻¹, respectively).

Fig. 4 Comparison of SO₂ adsorption isotherms in NiBDP MOFs, obtained experimentally and calculated by GCMC simulations at 298 K for 3 different structure models containing 0, 1 or 2 water molecules per unit cell, covering a pressure range from 1 × 10⁻⁴ to 1 bar.

This discrepancy further underscores the ambiguity about the proper activation of CUS-containing MOFs, where the number of weakly coordinated water molecules removed from the framework and the creation of CUS sites can vary significantly, influenced by specific experimental conditions and activation procedures. CUS sites are considered the most attractive and predominant adsorption sites for all types of approaching guest molecules, especially under very low-pressure conditions. To further assess the influence of CUS sites present within the NiBDP framework on its SO₂ uptake performance, we further considered two model systems, namely NiBDP-1H₂O (with 50% CUS sites, i.e., one H₂O per metal node) and NiBDP-0H₂O (with 100% CUS sites, no H₂O on the metal nodes). In the literature, it has been widely debated that generic force fields such as UFF⁴⁷ and DREIDING⁷¹ are not capable of accurately predicting the strong host–guest interactions in MOFs containing CUS sites.

Therefore, in the current study, we derived a specific set of Morse type potential parameters through periodic DFT calculations to accurately assess the strength of the pair-wise SO₂–Ni(II) CUS sites interactions using the 100% CUS containing NiBDP-0H₂O model (see parametrization details in the experimental section). Notably, DFT determined minimum energy geometry of SO₂ adsorption configuration in NiBDP-0H₂O framework reveals that an SO₂ molecule interacts with four Ni(II) open metal sites through an oxygen atom, with characteristic separation distances ranging from 2.27 to 2.49 Å (Fig. S10†), accompanied by a DFT-derived binding energy of −118.6 kJ mol⁻¹. As also depicted in Fig. S10,† the second oxygen atom of the SO₂ molecule was found to be oriented toward the nitrogen atom of the MOF framework, maintaining a separation distance of 2.93 Å.

Consequently, GCMC simulations performed utilizing our newly DFT-derived SO₂/Ni(II) force fields effectively reproduce the characteristic SO₂–Ni(II) interacting distances (see Fig. 5a) observed in the DFT equilibrium geometry as demonstrated in the corresponding radial distribution function (RDF) plots calculated under low-pressure conditions (Fig. 6a, S11 and S12†). The steep simulated uptake of SO₂ in NiBDP-0H₂O at the very initial stage of adsorption (Fig. 4) corresponds to the initial binding of each Ni(II) site by 1SO₂ molecule as illustrated in Fig. 5a with a characteristic interacting Ni–O(SO₂) distance of 2.5 Å (as shown by the RDF plotted in Fig. 6a). Fig. 6b evidence an additional interaction between the second oxygen of SO₂ with framework nitrogen atoms (Fig. S10†), the corresponding calculated RDFs between N–O(SO₂) and N–S(SO₂) showing peaks at 2.9 Å and 3.2 Å, respectively. This dual-site adsorption mode is probably the origin of the overestimation of the experimental adsorption uptake in the very low domain of pressure.

Fig. 5 GCMC simulated adsorption sequence of SO₂ in NiBDP-0H₂O at 298 K, and at different pressures. SO₂ atom color codes: red (O), yellow (S). MOF atom color codes: red (O), gray (C), blue (N), white (H), purple (Ni).

Fig. 6 Intermolecular radial pair distribution functions of NiBDP-0H₂O MOF atoms and adsorbed SO₂ molecules were calculated at P = 0.10 bar (a and b), and guest SO₂–SO₂ (c) was calculated at P = 0.50 bar.

Second, Fig. 5 illustrates that at a moderate pressure range, SO₂ molecules predominantly populate the region in the vicinity of the organic ligands, giving rise to an adsorbate clustering effect. Finally, at higher pressures, SO₂ molecules further cluster giving rise to the filling of the entire pore volume of the NiBDP structure. This behavior justifies the unusual step shape of the SO₂ isotherm. The overall adsorption behavior of NiBDP-1H₂O and NiBDP-2H₂O is illustrated in Fig. S13 and S14,† respectively. Fig. 6c displays the RDF for SO₂/SO₂ interaction in NiBDP-0H₂O at 298 K. The first characteristic peak was observed at 3.3 Å between S(SO₂)–O(SO₂), and the second characteristic peak was observed between O(SO₂)–O(SO₂) at 3.6 Å. These contacts are followed by S(SO₂)–S(SO₂) at 4.4 Å. RDF analysis of SO₂–SO₂ interactions pointed out that SO₂ molecules are distributed well around the NiBDP-0H₂O pore volume.

Furthermore, the enthalpy of adsorption calculated at very low coverage SO₂ adsorption is found to be −113.0 kJ mol⁻¹ in excellent agreement with the DFT calculated binding energy. These attributes provide a clear indication of achieving an accurate adsorption isotherm at low pressure using this newly derived force field. Indeed, as shown in Fig. 4, the GCMC calculations effectively reproduce the shape of the experimental SO₂ adsorption isotherm at low pressure (1 × 10⁻⁴ bar). NiBDP-2H₂O shows a distinct SO₂ adsorption isotherm profile at low pressure associated with a simulated adsorption enthalpy of −45 kJ. mol⁻¹, more in line with the experimental data that suggests that the activated sample still contains a large concentration of coordinated water. As expected, NiBDP-1H₂O is in-between.

A deeper analysis of these simulated isotherms at very low-pressure regions (1 × 10⁻⁴ to 0.1 bar) unravels the SO₂ adsorption mechanism in all NiBDP-nH₂O (n = 0, 1, 2) systems. However, depending on the availability of CUS sites in NiBDP frameworks, the amount of SO₂ uptake varies, with NiBDP-2H₂O, NiBDP-1H₂O and NiBDP-0H₂O showing 0.00 mmol g⁻¹, 0.55 mmol g⁻¹ and 1.12 mmol g⁻¹ at 1 × 10⁻⁴ bar, respectively. This highlights that the presence of CUS sites in the NiBDP framework enhances SO₂ uptake at a very low-pressure domain. Nevertheless, for all three NiBDP MOFs, our GCMC-predicted SO₂ isotherms exceed the experimental results above 0.2 bar, which may be attributed to differences in the experimentally and theoretically calculated BET surface area (see above).

Conclusions

In summary, a Ni(II)-based pyrazolate MOF, NiBDP, has been investigated for SO₂ adsorption under static conditions. This material exhibits exceptional chemical stability and a maximum SO₂ adsorption capacity of 8.48 mmol g⁻¹ at 298 K and 1 bar and an unusual step isotherm. Molecular simulations reveal that the SO₂ uptake of this material strongly depends on the concentration of Ni CUS sites and the presence of coordinated water molecules with stepwise seeding and clustering of adsorbate molecules. FTIR and in situ DRIFTS experiments corroborated the presence of coordinated water molecules and the availability of Ni(II) CUS sites in the structure simultaneously. A microscopic picture of the SO₂ adsorption mechanism has thus been gained in line with the unusual adsorption isotherm profile exhibited by this material. Thereby, the study highlights the relevance of robust Ni-pyrazolate MOF materials for high SO₂ stability.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

I. A. I. thanks PAPIIT UNAM (IN201123), México, for financial support. J. L. O. thanks CONAHCYT for the PhD fellowship (1003953). We thank U. Winnberg (Euro Health) for scientific discussions and G. Ibarra-Winnberg for scientific encouragement. The authors are grateful to Spanish MCIN/AEI/10.13039/501100011033 (Projects PID2020-113608RB-I00; TED2021-129886B-C41).

References

1. M. Sai Bhargava Reddy, D. Ponnamma, K. K. Sadasivuni, B. Kumar and A. M. Abdullah, RSC Adv., 2021, 11, 12658–12681.
2. S. He, J. Yang, S. Liu, X. Wang, X. Che, M. Wang and J. Qiu, Chem. Eng. J., 2023, 454, 140202.
3. H. Li, M. Eddaoudi, M. O'Keeffe and O. M. Yaghi, Nature, 1999, 402, 276–279.
4. X. Zheng, H. Zhang, L.-M. Yang and E. Ganz, Cryst. Growth Des., 2021, 21, 2474–2480.
5. H. Zhang, X. Zheng, L.-M. Yang and E. Ganz, Inorg. Chem., 2021, 60, 2656–2662.
6. H. Zhang, L.-M. Yang, H. Pan and E. Ganz, Phys. Chem. Chem. Phys., 2020, 22, 24614–24623.
7. H. Zhang, L.-M. Yang and E. Ganz, Langmuir, 2020, 36, 14104–14112.
8. H. Zhang, C. Shang, L.-M. Yang and E. Ganz, Inorg. Chem., 2020, 59, 16665–16671.
9. H. Zhang, L.-M. Yang, H. Pan and E. Ganz, Cryst. Growth Des., 2020, 20, 6337–6345.
10. H. Zhang, L.-M. Yang and E. Ganz, ACS Sustain. Chem. Eng., 2020, 8, 14616–14626.
11. H. Zhang, L.-M. Yang and E. Ganz, ACS Appl. Mater. Interfaces, 2020, 12, 18533–18540.
12. T. Islamoglu, Z. Chen, M. C. Wasson, C. T. Buru, K. O. Kirlikovali, U. Afrin, M. R. Mian and O. K. Farha, Chem. Rev., 2020, 120, 8130–8160.
13. V. B. López-Cervantes, J. L. Obeso, A. Yañez-Aulestia, A. Islas-Jácome, C. Leyva, E. González-Zamora, E. Sánchez-González and I. A. Ibarra, Chem. Commun., 2023, 59, 10343–10359.
14. S. He, J. Yang, S. Liu, X. Wang and J. Qiu, Adv. Funct. Mater., 2023, 2314133.
15. M. Woellner, S. Hausdorf, N. Klein, P. Mueller, M. W. Smith and S. Kaskel, Adv. Mater., 2018, 30, 1704679.
16. D. Britt, D. Tranchemontagne and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 11623–11627.
17. N. S. Bobbitt, M. L. Mendonca, A. J. Howarth, T. Islamoglu, J. T. Hupp, O. K. Farha and R. Q. Snurr, Chem. Soc. Rev., 2017, 46, 3357–3385.
18. E. Martínez-Ahumada, M. L. Díaz-Ramírez, M. de J. Velásquez-Hernández, V. Jancik and I. A. Ibarra, Chem. Sci., 2021, 12, 6772–6799.
19. A. J. Rieth, A. M. Wright and M. Dincă, Nat. Rev. Mater., 2019, 4, 708–725.
20. K. Tan, S. Zuluaga, H. Wang, P. Canepa, K. Soliman, J. Cure, J. Li, T. Thonhauser and Y. J. Chabal, Chem. Mater., 2017, 29, 4227–4235.
21. K. Tan, S. Zuluaga, Q. Gong, Y. Gao, N. Nijem, J. Li, T. Thonhauser and Y. J. Chabal, Chem. Mater., 2015, 27, 2203–2217.
22. R. Colorado-Peralta, J. María Rivera-Villanueva, J. Manuel Mora-Hernández, D. Morales-Morales and L. Ángel Alfonso-Herrera, Polyhedron, 2022, 224, 115995.
23. C. Bao, Q. Niu, Z.-A. Chen, X. Cao, H. Wang and W. Lu, RSC Adv., 2019, 9, 29474–29481.
24. M. Kurmoo, H. Kumagai, M. Akita-Tanaka, K. Inoue and S. Takagi, Inorg. Chem., 2006, 45, 1627–1637.
25. S. Zheng, X. Li, B. Yan, Q. Hu, Y. Xu, X. Xiao, H. Xue and H. Pang, Adv. Energy Mater., 2017, 7, 1602733.
26. V. B. López-Cervantes, D. W. Kim, J. L. Obeso, E. Martínez-Ahumada, Y. A. Amador-Sánchez, E. Sánchez-González, C. Leyva, C. S. Hong, I. A. Ibarra and D. Solis-Ibarra, Nanoscale, 2023, 15, 12471–12475.
27. C. Pettinari, A. Tăbăcaru and S. Galli, Coord. Chem. Rev., 2016, 307, 1–31.
28. T. He, Z. Huang, S. Yuan, X.-L. Lv, X.-J. Kong, X. Zou, H.-C. Zhou and J.-R. Li, J. Am. Chem. Soc., 2020, 142, 13491–13499.
29. A. J. Rieth and M. Dincă, J. Am. Chem. Soc., 2018, 140, 3461–3466.
30. L. M. Rodríguez-Albelo, E. López-Maya, S. Hamad, A. R. Ruiz-Salvador, S. Calero and J. A. R. Navarro, Nat. Commun., 2017, 8, 14457.
31. N. M. Padial, E. Quartapelle Procopio, C. Montoro, E. López, J. E. Oltra, V. Colombo, A. Maspero, N. Masciocchi, S. Galli, I. Senkovska, S. Kaskel, E. Barea and J. A. R. Navarro, Angew. Chem., Int. Ed., 2013, 52, 8290–8294.
32. J. Hafner, J. Comput. Chem., 2008, 29, 2044–2078.
33. P. E. Blöchl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
34. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
35. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
36. A. S. Palakkal and R. S. Pillai, J. Phys. Chem. C, 2020, 124, 16975–16989.
37. M. R. Tchalala, P. M. Bhatt, K. N. Chappanda, S. R. Tavares, K. Adil, Y. Belmabkhout, A. Shkurenko, A. Cadiau, N. Heymans, G. De Weireld, G. Maurin, K. N. Salama and M. Eddaoudi, Nat. Commun., 2019, 10, 1328.
38. L. Wang, T. Maxisch and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 73, 195107.
39. T. A. Manz and N. G. Limas, RSC Adv., 2016, 6, 47771–47801.
40. T. A. Manz and D. S. Sholl, J. Chem. Theory Comput., 2010, 6, 2455–2468.
41. T. F. Willems, C. H. Rycroft, M. Kazi, J. C. Meza and M. Haranczyk, Microporous Mesoporous Mater., 2012, 149, 134–141.
42. K. Gopalsamy, M. Wahiduzzaman, A. Daouli, M. Badawi and G. Maurin, ACS Appl. Nano Mater., 2022, 5, 15123–15132.
43. E. Haldoupis, J. Borycz, H. Shi, K. D. Vogiatzis, P. Bai, W. L. Queen, L. Gagliardi and J. I. Siepmann, J. Phys. Chem. C, 2015, 119, 16058–16071.
44. J. A. Gee and D. S. Sholl, J. Phys. Chem. C, 2015, 119, 16920–16926.
45. H. Fang, H. Demir, P. Kamakoti and D. S. Sholl, J. Mater. Chem. A, 2014, 2, 274–291.
46. E. Martínez-Ahumada, D. won Kim, M. Wahiduzzaman, P. Carmona-Monroy, A. López-Olvera, D. R. Williams, V. Martis, H. A. Lara-García, S. López-Morales, D. Solis-Ibarra, G. Maurin, I. A. Ibarra and C. S. Hong, J. Mater. Chem. A, 2022, 10, 18636–18643.
47. A. K. Rappe, C. J. Casewit, K. S. Colwell, W. A. Goddard and W. M. Skiff, J. Am. Chem. Soc., 1992, 114, 10024–10035.
48. M. H. Ketko, G. Kamath and J. J. Potoff, J. Phys. Chem. B, 2011, 115, 4949–4954.
49. P. P. Ewald, Ann. Phys., 1921, 369, 253–287.
50. J. D. Gale and A. L. Rohl, Mol. Simul., 2003, 29, 291–341.
51. D.-Y. Peng and D. B. Robinson, Ind. Eng. Chem. Fundam., 1976, 15, 59–64.
52. B. Widom, J. Chem. Phys., 1963, 39, 2808–2812.
53. N. Masciocchi, S. Galli, V. Colombo, A. Maspero, G. Palmisano, B. Seyyedi, C. Lamberti and S. Bordiga, J. Am. Chem. Soc., 2010, 132, 7902–7904.
54. S. Yang, J. Sun, A. J. Ramirez-Cuesta, S. K. Callear, W. I. F. David, D. P. Anderson, R. Newby, A. J. Blake, J. E. Parker, C. C. Tang and M. Schröder, Nat. Chem., 2012, 4, 887–894.
55. E. López-Maya, C. Montoro, V. Colombo, E. Barea and J. A. R. Navarro, Adv. Funct. Mater., 2014, 24, 6130–6135.
56. K. V. Kumar, S. Gadipelli, B. Wood, K. A. Ramisetty, A. A. Stewart, C. A. Howard, D. J. L. Brett and F. Rodriguez-Reinoso, J. Mater. Chem. A, 2019, 7, 10104–10137.
57. J. Li, G. L. Smith, Y. Chen, Y. Ma, M. Kippax-Jones, M. Fan, W. Lu, M. D. Frogley, G. Cinque, S. J. Day, S. P. Thompson, Y. Cheng, L. L. Daemen, A. J. Ramirez-Cuesta, M. Schröder and S. Yang, Angew. Chem., Int. Ed., 2022, 61, e202207259.
58. X. Cui, Q. Yang, L. Yang, R. Krishna, Z. Zhang, Z. Bao, H. Wu, Q. Ren, W. Zhou, B. Chen and H. Xing, Adv. Mater., 2017, 29, 1606929.
59. Y. Zhang, P. Zhang, W. Yu, J. Zhang, J. Huang, J. Wang, M. Xu, Q. Deng, Z. Zeng and S. Deng, ACS Appl. Mater. Interfaces, 2019, 11, 10680–10688.
60. M. Savage, Y. Cheng, T. L. Easun, J. E. Eyley, S. P. Argent, M. R. Warren, W. Lewis, C. Murray, C. C. Tang, M. D. Frogley, G. Cinque, J. Sun, S. Rudić, R. T. Murden, M. J. Benham, A. N. Fitch, A. J. Blake, A. J. Ramirez-Cuesta, S. Yang and M. Schröder, Adv. Mater., 2016, 28, 8705–8711.
61. A. López-Olvera, J. A. Zárate, E. Martínez-Ahumada, D. Fan, M. L. Díaz-Ramírez, P. A. Sáenz-Cavazos, V. Martis, D. R. Williams, E. Sánchez-González, G. Maurin and I. A. Ibarra, ACS Appl. Mater. Interfaces, 2021, 13, 39363–39370.
62. N. Hanikel, X. Pei, S. Chheda, H. Lyu, W. Jeong, J. Sauer, L. Gagliardi and O. M. Yaghi, Science, 2021, 374, 454–459.
63. V. Colombo, S. Galli, H. J. Choi, G. D. Han, A. Maspero, G. Palmisano, N. Masciocchi and J. R. Long, Chem. Sci., 2011, 2, 1311.
64. A. Nuhnen and C. Janiak, Dalton Trans., 2020, 49, 10295–10307.
65. F. Salles, G. Maurin, C. Serre, P. L. Llewellyn, C. Knöfel, H. J. Choi, Y. Filinchuk, L. Oliviero, A. Vimont, J. R. Long and G. Férey, J. Am. Chem. Soc., 2010, 132, 13782–13788.
66. 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.
67. K. Tan, P. Canepa, Q. Gong, J. Liu, D. H. Johnson, A. Dyevoich, P. K. Thallapally, T. Thonhauser, J. Li and Y. J. Chabal, Chem. Mater., 2013, 25, 4653–4662.
68. V. Bon, E. Brunner, A. Pöppl and S. Kaskel, Adv. Funct. Mater., 2020, 30, 1907847.
69. E. Martínez-Ahumada, M. L. Díaz-Ramírez, H. A. Lara-García, D. R. Williams, V. Martis, V. Jancik, E. Lima and I. A. Ibarra, J. Mater. Chem. A, 2020, 8, 11515–11520.
70. C. C. Zhang, J. Shi, S. Hartlaub, J. P. Palamara, I. Petrovic and B. Yilmaz, Catal. Commun., 2021, 150, 106273.
71. S. L. Mayo, B. D. Olafson and W. A. Goddard, J. Phys. Chem., 1990, 94, 8897–8909.

---

Footnotes

† Electronic supplementary information (ESI) available: Instrumental techniques and characterization. See DOI: https://doi.org/10.1039/d3ta07582f

‡ These authors contributed equally to this manuscript.

---