Structural and Dynamic Analysis of Sulphur Dioxide Adsorption in a Series of Zirconium‐Based Metal–Organic Frameworks

Abstract

We report reversible high capacity adsorption of SO₂ in robust Zr‐based metal–organic framework (MOF) materials. Zr‐bptc (H₄bptc=biphenyl‐3,3′,5,5′‐tetracarboxylic acid) shows a high SO₂ uptake of 6.2 mmol g⁻¹ at 0.1 bar and 298 K, reflecting excellent capture capability and removal of SO₂ at low concentration (2500 ppm). Dynamic breakthrough experiments confirm that the introduction of amine, atomically‐dispersed Cu^II or heteroatomic sulphur sites into the pores enhance the capture of SO₂ at low concentrations. The captured SO₂ can be converted quantitatively to a pharmaceutical intermediate, aryl N‐aminosulfonamide, thus converting waste to chemical values. In situ X‐ray diffraction, infrared micro‐spectroscopy and inelastic neutron scattering enable the visualisation of the binding domains of adsorbed SO₂ molecules and host–guest binding dynamics in these materials at the atomic level. Refinement of the pore environment plays a critical role in designing efficient sorbent materials.

A series of Zr‐based metal–organic framework materials have been investigated for reversible SO₂ uptake. The incorporation of atomically‐dispersed Cu^II, amine or heteroatomic sulphur sites enhances the uptake of SO₂ at low concentrations. This work confirms that control of pore environments is an important approach for optimising the adsorption of SO₂ at low concentrations.

Introduction

Sulphur dioxide (SO₂) is an important air pollutant as well as a key chemical feedstock for the synthesis of sulfuric acid and various fine chemicals.[ ¹ , ² , ³ , ⁴ , ⁵ ] State‐of‐the‐art flue‐gas desulphurisation (FGD) technology uses limestone slurry to capture SO₂ effectively, but this is an irreversible process that generates a tremendous amounts of solid waste.[ ⁶ , ⁷ ] Recovery of SO₂ from exhaust gases via reversible adsorptive techniques can promote the development of “waste‐to‐chemical” technologies, but it relies on the development of efficient sorbent materials that not only show high and reversible adsorption of SO₂, but also are highly robust so that regeneration of the sorbent can be achieved for use over many cycles.

Metal–organic framework (MOF) materials have been studied widely for gas adsorption and separation owing to their high surface area and tuneable pore environment.[ ⁸ , ⁹ ] The study of MOF materials as SO₂ reservoirs has seen significant interest recently,[ ¹⁰ , ¹¹ , ¹² ] but only a limited number of MOFs show reversible SO₂ uptake and structural stability upon desorption: for example, Mg‐MOF‐74 (8.6 mmol g⁻¹), ^[13] EDTA‐MOF‐808 (9.8 mmol g⁻¹), ^[14] [Ni(bdc)(ted)_0.5] (10.0 mmol g⁻¹), ^[15] [Zn₂(L)₂(bipy)] (10.9 mmol g⁻¹), ^[16] SIFSIX‐1‐Cu (11.0 mmol g⁻¹), ^[17] ECUT‐111 (11.6 mmol g⁻¹), ^[18] DMOF (13.1 mmol g⁻¹), ^[19] MFM‐300(Sc)@EtOH (13.2 mmol g⁻¹), ^[20] MOF‐808 (15.3 mmol g⁻¹), ^[14] MFM‐170 (17.5 mmol g⁻¹) ^[21] and MIL‐101(Cr)‐4F(1 %) (18.4 mmol g⁻¹) ^[22] all at 298 K and 1 bar of SO₂. MOFs constructed from {Zr₆} clusters are renowned for their high stability.[ ²³ , ²⁴ , ²⁵ ] However, their performance in adsorption of SO₂ has been poorly explored and, to date, only few Zr‐MOFs have shown reversible SO₂ adsorption at 298 K and 1 bar, including MFM‐601 (12.3 mmol g⁻¹) ^[26] and NU‐1000 (10.9 mmol g⁻¹). ^[27]

Herein, we report a systematic structural and dynamic analysis of adsorption of SO₂ in seven robust Zr‐MOFs: UiO‐66, UiO‐66‐NH₂, UiO‐66‐Cu^II, Zr‐DMTDC (H₂DMTDC=3,4‐dimethylthieno[2,3‐b]thiophene‐2,5‐dicarboxylic acid), Zr‐bptc, MFM‐133 and MFM‐422. Compared with UiO‐66, the introduction of amine groups (UiO‐66‐NH₂), thienothiophene groups (Zr‐DMTDC) or atomically‐dispersed Cu^II sites (UiO‐66‐Cu^II) afford 76 %, 47 % and 43 % enhancement of SO₂ uptake at 0.1 bar and 298 K, respectively. Zr‐bptc exhibits an exceptional SO₂ uptake of 6.2 mmol g⁻¹ at 0.1 bar and 298 K and dynamic breakthrough confirms the highly selective capture of SO₂ from a mixture of SO₂/CO₂ (2500 ppm SO₂, 15 % CO₂ diluted in He). In addition, the captured SO₂ in Zr‐bptc can be converted to aryl N‐amino sulphonamide, an important compound in medicinal chemistry, thus fulfilling the “waste‐to‐chemicals” target. MFM‐422 shows a high Brunauer–Emmett–Teller (BET) surface area of 3296 cm² g⁻¹ and an exceptional and reversible uptake of SO₂ of 31.3 mmol g⁻¹ at 1 bar and 273 K. These materials show high stability with full retention of structure and uptake capacities over multiple cycles of adsorption‐desorption of dry SO₂. The adsorption domains and binding dynamics of SO₂ in these MOFs have been studied by in situ synchrotron X‐ray powder diffraction (SXPD), inelastic neutron scattering (INS), and synchrotron infrared micro‐spectroscopy (microFTIR) to provide key insights into the structures and dynamics of high adsorption of SO₂ in these systems.

Results and Discussion

UiO‐66, ^[28] UiO‐66‐NH₂, ^[29] UiO‐66‐Cu^II[30] and Zr‐DMTDC ^[31] are iso‐structural and constructed from 12‐connected {Zr₆(μ₃‐O)₄(μ₃‐OH)₄(OOCR)₁₂} clusters bridged by dicarboxylates to give cubic structures of fcu topology (Figure 1). These structures consist of two types of cages with an octahedral cage (Cage O, diameter of 9–12 Å) connecting to eight tetrahedral cages (Cage T, diameter of 7.3 Å) via triangular faces (Figure 1). The pores of UiO‐66‐NH₂ and Zr‐DMTDC are decorated with free −NH₂ and −S− sites, respectively, affording additional binding sites for guest molecules. In UiO‐66‐Cu^II, defect sites with free −OH/−OH₂ sites in the pore are decorated with open Cu^II sites. Desolvated UiO‐66, UiO‐66‐NH₂, UiO‐66‐Cu^II and Zr‐DMTDC show BET surface areas of 1221, 1037, 1068 and 1345 m² g⁻¹, respectively.

Figure 1.

Views of {Zr₆}‐clusters, linkers and structures of the Zr‐based MOFs used in this study (Zr, aqua; C, grey; O, red; H, white; S, yellow).

Zr‐bptc is built from 12‐connected {Zr₆(μ₃‐O)₄(μ₃‐OH)₄(OOCR)₁₂} clusters and tetracarboxylate ligands in an open framework of ftw topology. ^[32] Desolvated Zr‐bptc consists of cubic cages (cage A) of diameter 12 Å fused to tetrahedral cages (cage B) of 4.5 Å diameter (Figure 1) with a BET surface area of 960 m² g⁻¹. MFM‐133 ^[33] is constructed from 8‐connected {Zr₆(OH)₈(OH)₈(OOCR)₈} clusters and thcb⁴⁻ ligands (H₄thcb=3,3′,5,5′‐tetrakis(4‐carboxyphenyl)‐2,2′,4,4′,6,6′‐hexamethyl‐1,1′‐biphenyl) to form a flu topology. MFM‐133 shows an axially elongated octahedral cage (10.4×10.4×25.9 Å) and a BET surface area of 2156 m² g⁻¹ (Figure 1). A new MOF, MFM‐422, is constructed by linking 8‐connected {Zr₆(OH)₈(OH)₈(OOCR)₈} clusters with the tetratopic ligand 3,3′′,5,5′′‐tetrakis(4‐carboxyphenyl)‐p‐terphenyl (H₄tcpt) to give a neutral framework of sqc topology. MFM‐422 is comprised of a trigonal cage (cage B, diameter of 7.7 Å) and a hexagonal cage (cage A, diameter of 30 Å, Figure 1). Desolvated MFM‐422 shows a BET surface area of 3296 m² g⁻¹ and a high thermal stability up to 500 °C (Figures S43–S44).

Gravimetric adsorption isotherms of SO₂ have been recorded for these MOFs at 273–298 K and from 0–1 bar (Figures 2a, b, S1–S7 and Table 1). MFM‐422 shows a SO₂ uptake of 31.3 mmol g⁻¹ at 273 K and 1.0 bar, comparable to the record previously achieved by UR3‐MIL‐101(Cr) (36.7 mmol g⁻¹) under the same conditions. ^[34] At 298 K and 1 bar, all 7 MOFs, i.e., UiO‐66, UiO‐66‐NH₂, UiO‐66‐Cu^II, Zr‐DMTDC, Zr‐bptc, MFM‐133 and MFM‐422, show fully reversible uptakes of SO₂ of 8.6, 8.8, 8.2, 9.6, 7.8, 8.9 and 13.6 mmol g⁻¹, respectively (Figures 2b and S1–S7). The multiple cycles of adsorption‐desorption of SO₂ for all samples at 298 K show little change in the capacity, demonstrating excellent stability towards dry SO₂ (Figures S1–S7, S30–S35). The comparable adsorption uptakes of UiO‐66, UiO‐66‐NH₂ and UiO‐66‐Cu^II at 1 bar (8.2–8.8 mmol g⁻¹) suggest that decoration of the pore environment with functional groups or open Cu^II sites has little impact on the total uptake capacity, which is determined primarily by the surface area. The slightly higher uptake of Zr‐DMTDC (9.6 mmol g⁻¹) is consistent with its higher surface area (1345 m² g⁻¹), compared with the other three UiO‐66 materials. In contrast, enhancements in the uptake at 0.1 bar were observed for UiO‐66‐NH₂, UiO‐66‐Cu^II and Zr‐DMTDC, compared with UiO‐66 (uptakes of 3.7, 3.0, 3.1 and 2.1 mmol g⁻¹, respectively, Figure 2c). This demonstrates that the introduction of accessible −NH₂, Cu^II or R‐S‐R sites into the pores can increase the binding strength with SO₂ molecules. Interestingly, Zr‐bptc displays an extremely high uptake of 6.2 mmol g⁻¹ at 0.1 bar and 298 K, suggesting potential for selective adsorption of SO₂ at low concentration. The isosteric heats of adsorption (Q _st) for SO₂ uptake show decreasing values of 45–50, 44–32, 38–34, 32–29, 37–27, 31–27 and 26–19 kJ mol⁻¹ for Zr‐bptc, UiO‐66‐NH₂, UiO‐66‐Cu^II, Zr‐DMTDC, UiO‐66, MFM‐422 and MFM‐133, respectively. Compared with UiO‐66, the materials UiO‐66‐NH₂, UiO‐66‐Cu^II and Zr‐DMTDC show higher values for Q _st, consistent with the enhanced adsorption at low pressure. The relatively low values of Q _st for MFM‐133 and MFM‐422 are consistent with their large pores, reducing the strength of host–guest interactions.

Figure 2.

Gas adsorption, thermodynamic, selectivity and separation data. SO₂ adsorption isotherms at (a) 273 K and 1 bar and (b) 298 K and 1 bar (desorption data are omitted for clarity and shown in Figures S1–S7); (c) SO₂ adsorption isotherms from 0 to 0.1 bar at 298 K; (d) variation of Q _st; (e) CO₂ adsorption isotherms for Zr‐bptc, Zr‐DMTDC and UiO‐66 materials at 298 K and 1 bar; (f) comparison of IAST selectivities of SO₂/CO₂ (1 : 99) for Zr‐bptc, Zr‐DMTDC and UiO‐66 materials at 298 K; (g) breakthrough plots for a SO₂/CO₂ mixture (2500 ppm SO₂, 15 % CO₂ in He, total flow rate: 20 mL min⁻¹) in Zr‐DMTDC and UiO‐66 materials at 298 K (solid line: SO₂; dashed line: CO₂); (h) breakthrough plots for a SO₂/N₂ mixture (2500 ppm SO₂, 75 % N₂ in He, total flow rate: 14 mL min⁻¹) in Zr‐DMTDC and UiO‐66 materials at 298 K (solid line: SO₂; dashed line: N₂); (i) breakthrough plots for a SO₂/N₂ mixture (2500 ppm SO₂, 75 % N₂ in He, total flow rate: 40 mL min⁻¹) and (2500 ppm SO₂, 15 % CO₂ in He, total flow rate: 40 mL min⁻¹) in Zr‐bptc at 298 K (blue: SO₂/N₂ mixture; red: SO₂/CO₂ mixture).

Table 1.

Summary of BET surface areas, SO₂ uptakes and Q _st and IAST selectivities in Zr‐MOFs.

MOFs	BET [m2 g−1]	SO2 Uptake [mmol g−1] at 1 bar		SO2 Q st [kJ mol−1]	Selectivity
		298 K	273 K		SO2/CO2 (1/99)	SO2/N2 (1/99)
Zr‐bptc	960	7.8	8.6	45–50	600	>5 000
UiO‐66‐CuII	1068	8.2	9.6	38–34	54	3 100
UiO‐66‐NH2	1037	8.8	10.5	44–32	25	486
Zr‐DMTDC	1345	9.6	11.6	32–29	20	280
UiO‐66	1221	8.6	11.5	37–27	13	208
MFM‐133	2156	8.9	10.7	31–27	–	–
MFM‐422	3296	13.6	31.3	26–19	–	–

Adsorption isotherms of CO₂ and N₂ have also been recorded for Zr‐bptc, UiO‐66‐NH₂, UiO‐66‐Cu^II, Zr‐DMTDC and UiO‐66 to assess the adsorption selectivity (Figures 2e, S8–S12, Table S1). At 298 K, Zr‐bptc displays CO₂ uptakes of 2.5 and 0.82 mmol g⁻¹ at 1.0 and 0.15 bar, respectively. While UiO‐66‐NH₂ and Zr‐ DMTDC display 58 % and 42 % enhancements in the CO₂ uptake at 0.15 bar and 298 K compared with UiO‐66, UiO‐66‐Cu^II shows a reduction of CO₂ uptake of 47 % at 0.15 bar and 298 K (Figures 2e). Thus, the latter has great potential for selective adsorption of SO₂. Analysis of pure‐component isotherms via ideal adsorbed solution theory (IAST) ^[35] affords adsorption selectivities for mixtures of SO₂/CO₂ (1/99) and SO₂/N₂ (1/99) (Figure 2f and S13) for Zr‐bptc, UiO‐66‐NH₂, UiO‐66‐Cu^II, Zr‐DMTDC and UiO‐66. Zr‐bptc displays high selectivities of 600 for SO₂/CO₂ and >5000 for SO₂/N₂; the very high IAST selectivity is subject to uncertainties owing to the extremely low adsorption of N₂. UiO‐66‐Cu^II, UiO‐66‐NH₂, Zr‐DMTDC and UiO‐66 display IAST selectivities for SO₂/CO₂ of 54, 25, 20 and 13, and for SO₂/N₂ of 3100, 486, 280 and 208, respectively. To confirm the selective capture of SO₂ under realistic concentrations, ^[36] fixed‐beds packed with these MOFs were studied by dynamic breakthrough experiments with a mixture of SO₂/CO₂ (2500 ppm SO₂/15 % CO₂ in He) at 298 K and 1.0 bar. UiO‐66, UiO‐66‐NH₂, Zr‐DMTDC and UiO‐66‐Cu^II exhibit retention times for SO₂ in the expected order of 33, 53, 58 and 100 min g⁻¹, respectively (Figure 2g). The same sequence was observed in the separation of the mixture of SO₂/N₂ (2500 ppm/75 %) with retention times of 80, 175, 157 and 175 min g⁻¹ for UiO‐66, UiO‐66‐NH₂, Zr‐DMTDC and UiO‐66‐Cu^II, respectively (Figure 2h). Zr‐bptc shows highly selective retention of SO₂ at 213 and 235 min g⁻¹ for mixtures of SO₂/CO₂ (2500 ppm SO₂/15 % CO₂ in He) and SO₂/N₂ (2500 ppm SO₂/75 % CO₂ in He), respectively (Figure 2i). Thus, the breakthrough results are fully consistent with the isotherm data and confirm the positive role of open Cu^II sites on selective SO₂ adsorption.

Rietveld refinements of the high‐resolution SXPD data of SO₂‐loaded UiO‐66 [Zr₆O₄(OH)₄(bdc)₆ ⋅ (SO₂)_7.7] reveal two binding sites I and II located in cage T (SO₂/{Zr₆}=5.1) and cage O (SO₂/{Zr₆}=2.6), respectively (Figure 3a). The hydrogen bond [OSO⋅⋅⋅μ₃‐HO=2.32(1) Å] and dipole–dipole interaction [O₂ S⋅⋅⋅phenyl ring=3.69(2) Å] stabilise SO₂ (I) (Figure 3d). SO₂ (II) is stabilised by two hydrogen bonds [OSO⋅⋅⋅H−C=1.58(2), 2.70(6) Å] (Figure 3e). In SO₂‐loaded UiO‐66‐NH₂ [Zr₆O₄(OH)₄(bdc−NH₂)₆ ⋅ (SO₂)_8.1], two binding sites I′ and II′ are observed in cage T (SO₂/{Zr₆}=4.7) and cage O (SO₂/{Zr₆}=3.4), respectively (Figure 3b). Due to the presence of active −NH₂ groups, the adsorbed SO₂ molecules are stabilised strongly by the formation of supramolecular interactions between −NH₂ groups and SO₂ molecules. A dipole–dipole interaction [NH₂⋅⋅⋅SO₂=3.77(9) Å] was identified and works together with an interaction [O₂ S⋅⋅⋅phenyl ring=3.58(1) Å] and hydrogen bonding [OSO⋅⋅⋅μ₃‐HO=2.94(5) Å] that stabilise SO₂ binding at site I′ (Figure 3f). In addition, seven hydrogen bonds were identified [OSO⋅⋅⋅H−C=2.88(1) Å, SO ₂⋅⋅⋅NH₂=1.73(3), 2.43(6), 2.87(7), 3.21(1), 3.30(3) and 3.63(8) Å], which work together with two further dipole–dipole interactions [O₂ S⋅⋅⋅NH₂=2.40(4) and 3.10(7) Å] to stabilise SO₂ at site II′ (Figure 3g). The additional hydrogen bonds and dipole–dipole interactions demonstrate enhanced binding of SO₂ in UiO‐66‐NH₂, consistent with the increased SO₂ adsorption at low pressure.

Figure 3.

Views of binding of SO₂ in (a) UiO‐66 (site I in cage T: lavender; site II in cage O: cyan); (b) UiO‐66‐NH₂ (site I’ in cage T: lavender; site II’ in cage O: orange); (c) Zr‐DMTDC (site I’’, II’’ and III’’ in cage T: teal, lavender and orange, respectively; site IV’ in cage O: blue); (d) UiO‐66 at site I (SO₂: lavender); (e) UiO‐66 at site II (SO₂: cyan); (f) UiO‐66‐NH₂ at site I’ (SO₂: lavender); (g) UiO‐66‐NH₂ at site II’ (SO₂: orange); (h) Zr‐DMTDC at site I’′ (SO₂: teal) and II’’ (SO₂: lavender); (i) Zr‐DMTDC at site III’’ (SO₂: orange); (j) Zr‐DMTDC at site IV’’ (SO₂: blue) (in framework Zr: cyan; O: red; S: yellow; C: dark grey and H: white). All units are quoted in Å.

In SO₂‐loaded Zr‐DMTDC [Zr₆O₄(OH)₄(DMTDC)₂ ⋅ (SO₂)_13.1], four binding sites were revealed (I′′‐IV′′). Sites I′′, II′′ and III′′ are localised in cage T (SO₂/{Zr₆}=4.2, 4.1 and 2.5, respectively) (Figure 3c). The SO₂ molecule at site I′′ is stabilised by the formation of three dipole–dipole interactions [OSO⋅⋅⋅S‐ring=2.40(2) and 3.52(7) Å; O₂ S⋅⋅⋅thiophene ring=3.65(3) Å] (Figure 3h). The SO₂ molecules at site II′′ are stabilised further by four dipole–dipole interactions [OSO⋅⋅⋅S‐ring=2.46(1) and 2.70(4) Å; O₂ S⋅⋅⋅thiophene ring=3.72(3) and 3.72(3) Å] and supramolecular interaction [O₂ S⋅⋅⋅μ₃‐O=3.63(9) Å]. In addition, dipole–dipole interaction between SO₂ at sites I′′ and II′′ [OSO(I′′)⋅⋅⋅SO₂(II′′)=2.81(7) Å] was identified (Figure 3h), which is not observed in either UiO‐66 or UiO‐66‐NH₂ and may result from the slightly enlarged pore size. The formation of dipole–dipole interactions [O₂ S⋅⋅⋅μ₃‐O=3.40(6) Å] and [OSO⋅⋅⋅S‐ring=3.97(6) Å] were identified between SO₂(III′′) and the framework (Figure 3i). Site IV′′ is in the cage O and stabilised by two dipole–dipole interactions [OSO⋅⋅⋅S‐ring=3.43(2) and 3.77(8) Å] and a hydrogen bond [OSO⋅⋅⋅H ₃C=2.42(2) Å] (Figure 3j). This crystallographic study enables direct observation of host–guest interactions, and revealed that the introduction of heteroatom S dominated the supramolecular interactions facilitating the immobilisation of SO₂ at low pressure.

In SO₂‐loaded Zr‐bptc, [Zr₆O₄(OH)₄(bptc)₃ ⋅ (SO₂)_5.8], six binding sites were revealed (I–VI) (Figure 4a). Sites I, II, III, IV and V are localised in cage A with SO₂/{Zr₆} ratios of 1.7, 1.1, 1.0, 0.61 and 0.96, respectively, and Site VI located at cage B with a SO₂/{Zr₆} ratio of 0.43. SO₂ molecules at site I were stabilised by Zr^IV sites [OSO⋅⋅⋅Zr=3.16(2) Å] and by two hydrogen bonds [OSO⋅⋅⋅H−C=2.79(8) Å and OSO⋅⋅⋅μ₃‐OH=2.36(5) Å] (Figure 4b). SO₂ molecules at site II are immobilised by three dipole–dipole interactions [OSO(II)⋅⋅⋅SO₂(I)=2.06(8) and 3.07(2) Å, OSO(I)⋅⋅⋅SO₂(II)=3.28(1) Å] with SO₂ at site I (Figure 4b). SO₂ molecules at site III are stabilised by dipole–dipole interactions [O₂ S(III)⋅⋅⋅OSO(V)=2.99(7) Å, OSO(III)⋅⋅⋅SO₂(V)=3.13(5) Å] with SO₂ at site V immobilised by dipole–dipole interactions [O₂ S⋅⋅⋅phenyl ring=3.72(5) Å] and two‐fold electrostatic interactions [OSO⋅⋅⋅H−C=2.78(1) and 2.93(9) Å] (Figure 4c). SO₂ molecules at site IV are stabilised by a weak hydrogen bond [OSO⋅⋅⋅μ₃‐HO=3.96(5) Å] and dipole–dipole interaction [OSO(IV)⋅⋅⋅SO₂(VI)=3.83(5) Å] with SO₂ at site VI (Figure 4d). SO₂ molecules at site VI sit at the centre of cage B and are immobilised by five hydrogen bonds [OSO⋅⋅⋅H−C=1.88(6), 1.88(6), 2.23(5), 2.23(5) and 2.69(7) Å] and two dipole–dipole interactions [O₂ S⋅⋅⋅OOC=3.07(2) and 3.07(2) Å] (Figure 4d). In contrast to the host–guest binding observed in SO₂‐loaded UiO‐66 type systems, Zr^IV sites, the strong hydrogen bonding at site I, unique dipole–dipole interactions between SO₂(VI) and carboxylic groups and multiple strong hydrogen bonding at site VI jointly facilitate the exceptional SO₂ uptake at low pressure.

Figure 4.

Views of the binding sites of SO₂ in (a) Zr‐bptc (site I, II, III, IV and V in cage A: rose, yellow, dark green, green and orange, respectively; site VI in cage B: teal); (b) Zr‐bptc at sites I (SO₂: rose) and II (SO₂: yellow); (c) Zr‐bptc at site III (SO₂: dark green) and V (SO₂: orange); (d) Zr‐bptc at site IV (SO₂: green) and VI (SO₂: teal) (in framework Zr: cyan; O: red; S: yellow; C: dark grey and H: white). All units are quoted in Å.

The binding dynamics of adsorption of SO₂ (0–1 bar) in the UiO‐66 type systems have been analysed by in situ synchrotron infrared micro‐spectroscopy. For all the MOFs, clear binding of SO₂ to the hydroxyl group is observed with a red shift of the −OH stretching vibration at ≈3671 cm⁻¹ by 86, 95, 83 and 82 cm⁻¹ in UiO‐66, UiO‐66‐NH₂, UiO‐66‐Cu^II and Zr‐DMTDC, respectively (Figures 5ai–di). There is clear evidence for enhanced interaction of SO₂ via hydrogen bonding to −NH₂ groups ([NH ₂⋅⋅⋅OSO], the characteristic NH₂ band shifting from 3490 to 3502 cm⁻¹ and 3396 to 3386 cm⁻¹ (Figure 5b). Dipole interactions are observed to the Cu^II sites ([OSO⋅⋅⋅⋅Cu−OH], with the characteristic Cu‐OH band shifting from 719 to 732 cm⁻¹. Formation of [OSO⋅⋅⋅Cu−OH₂] interactions leads to a new band at 657 cm⁻¹ assigned to the CuO stretching vibration ^[37] (Figure 5cii), while the thiophene system leads to [OSO⋅⋅⋅S] interactions with a characteristic shift in the S−C stretch from 1664 to 1658 cm^−1[38] (Figure 5dii). The displacement and cooperative binding of SO₂ and CO₂ was investigated in UiO‐66‐Cu^II. The ν(μ ₃‐OH) mode was monitored to examine the displacement of bound CO₂ by SO₂ (Figures 5ciii, S26). Upon loading of CO₂ to 1 bar, the peak areas for the ν(μ ₃‐OH) stretch corresponding to the bare and CO₂‐loaded materials are approximately equal. Due to weak interaction between CO₂ and the μ ₃‐OH group, the bare μ ₃‐OH band is not fully depleted but a new peak at 3643 cm⁻¹ appears and is assigned to the [OH⋅⋅⋅OCO] band (Figure 5ciii).

Figure 5.

(a) (i). IR spectra showing the ν(μ ₃‐OH) and ν(‐CH) stretching region for UiO‐66 at various loadings of SO₂; (ii–iii) views of corresponding structures. (b) (i) IR spectra showing the ν(μ ₃‐OH), ν(‐NH₂) and ν(‐CH) stretching region for UiO‐66‐NH₂ at various loadings of SO₂; (ii‐iii) views of corresponding structures; (c) IR spectra of (i) ν(μ ₃‐OH), (ii) ν(Cu−OH) and ν(Cu−OH₂) stretching region for UiO‐66‐Cu^II at 2 % (blue) and 100 % (orange) loading of SO₂ (other loadings are omitted for clarity and shown in the Supplementary Information Figures S21–S22); iii) IR spectra of the ν(μ ₃‐OH) stretch region for bare (black), 100% CO₂‐loading (purple) and 30% SO₂‐loading for CO₂ displacement (red) in UiO‐66‐Cu^II; (d) IR spectra of (i) ν(μ ₃‐OH) and (ii) ν(S−C) stretching region for Zr‐DMTDC at various loadings of SO₂; (iii). Views of corresponding structures (in various SO₂‐loading experiments: black: bare MOF, red: 1% SO₂‐loading, blue: 2% SO₂‐loading, green: 5% SO₂‐loading, violet: 10% SO₂‐loading, dark yellow: 20% SO₂‐loading, cyan: 40% SO₂‐loading, light wine: 60% SO₂‐loading, wine: 80% SO₂‐loading, orange: 100% SO₂‐loading).

Upon stepwise dosing of the CO₂‐saturated material with SO₂ (i.e., tuning the SO₂/CO₂ mixture from 0/100 to 100/0 while maintaining a total pressure of 1.0 bar), there is a steady change in the ν(μ ₃‐OH) region that includes new bands appearing in a similar manner to the pure SO₂ experiment, indicating that bound CO₂ does not impede SO₂ adsorption (Figure S26). Upon 30 % SO₂‐loading, the characteristic [OH⋅⋅⋅OCO] band has fully disappeared showing that SO₂ readily displaces bound CO₂ in the pore as a result of stronger binding. Hence, selective capture of SO₂ from a mixture of SO₂/CO₂ can be achieved as demonstrated in separation experiments. Furthermore, 40 %, 45 % and 50 % SO₂‐loadings fully displace CO₂ in UiO‐66‐NH₂, Zr‐DMTDC and UiO‐66, respectively (Figures S24–27). The competitive binding studies of SO₂/CO₂ further confirm enhanced SO₂ binding in the decorated MOFs. The decreasing partial pressure of SO₂ on full displacement of CO₂ in UiO‐66‐Cu^II, UiO‐66‐NH₂ and Zr‐DMTDC is consistent with that observed in static and dynamic adsorption studies.

In situ INS, coupled with DFT calculations, enables the visualization of binding dynamics for SO₂‐loaded Zr‐bptc. Seven major changes in the INS spectra were observed on the adsorption of SO₂ in Zr‐bptc (Figure 6a). Peaks I‐III occur at low energy transfer (<60 meV) and Peaks IV‐VII at the high energy region (80–120 meV). Peak I (8.3 meV) is assigned to the flapping mode of the aromatic ring and peaks II (19.3 meV) and III (29.6 meV) are due to aromatic deformation. Peaks IV (83.7 meV) and VII (118 meV) are assigned to C−H out‐of‐plane bending modes with H moving in the same direction and opposite directions, respectively. The changes in peaks I, II, III, IV and VII suggest interactions between adsorbed SO₂ molecules and the aromatic moieties, consistent with the crystallographic analyses (Figures 6b–d). Peaks V (92.9 meV) and VI (106.5 meV) are assigned to μ₃‐OH wagging and twisting, respectively, and their changes support the formation of hydrogen bonds [OSO⋅⋅⋅μ₃‐HO] that were observed in the crystallographic analysis (Figure 6b and 6d).

Figure 6.

(a) Comparison of the difference plots for experimental and DFT‐calculated INS spectra of bare and SO₂‐loaded Zr‐bptc. No scale factor was used for the DFT calculations. S, dynamic structure factor; Q, difference between incoming and outgoing wave vector; ω, the energy change experienced by the sample; (b–d) Views of corresponding structures.

Unlike FGD technology, where SO₂ is bound permanently to sorbent materials to form solid inorganic wastes, the SO₂ captured by these Zr‐MOFs remains available to undergo chemical transformation to valuable products. Here, a proof‐of‐concept experiment on aminosulfonylation[ ³⁹ , ⁴⁰ ] using the SO₂‐loaded Zr‐bptc was conducted, and quantitative conversion of the captured SO₂ was achieved to give 4‐methoxyl‐aryldiazonium tetrafluoroborate in 85 % yield (Scheme 1). Upon regeneration, Zr‐bptc can be used for at least 3 cycles without any change in the crystal structure or porosity of the material (Figure S46 and Table S5), thus demonstrating its great potential of the capture and conversion of waste SO₂ to fine chemicals.

Scheme 1.

Conversion of captured SO₂ in Zr‐bptc for the synthesis of fine chemicals; 1 equivalent of 4‐methoxy‐aryldiazonium tetrafluoroborate and 5 equivalents of amine and SO₂‐loaded Zr‐bptc were reacted at room temperature for 1 h.

Conclusion

Powerful drivers exist for the development of new regenerable sorbents for SO₂ to enable its recovery from exhaust gases and conversion into chemical feedstocks. The highly corrosive and reactive nature of SO₂ leads generally to severe structural degradation of sorbent materials. We report the positive impacts on low pressure SO₂ uptake by introducing functional groups and atomically‐dispersed Cu^II sites into a family of Zr‐MOFs. Owing to the confined metal–ligand cages in Zr‐bptc, an exceptional uptake of SO₂ (6.2 mmol g⁻¹) was observed at 0.1 bar and 298 K. Furthermore, the captured SO₂ in Zr‐bptc can be converted readily into fine chemicals, paving new pathways to “waste‐to‐chemicals” technologies. In situ SXPD, microFTIR and INS studies, coupled with DFT calculations, unravel the molecular details of host–guest binding that result in the enhancement of SO₂ adsorption at low pressure in these materials. These studies confirm that control of pore environments is an important approach for improving the adsorption of SO₂.

Associated Content

Additional crystallographic information, gas adsorption data, thermogravimetric analysis, density function theory (DFT) calculations and breakthrough data are available in the Supporting Information. The crystal structures of [Zr₆(OH)₈(OH)₈(tcpt)₂], [Zr₆O₄(OH)₄(bptc)₃ ⋅ (SO₂)_5.8], [Zr₆O₄(OH)₄(DMTDC)₆ ⋅ (SO₂)_13.1], [Zr₆O₄(OH)₄(bdc)₆ ⋅ (SO₂)_7.7] and [Zr₆O₄(OH)₄(bdc−NH₂)₆ ⋅ (SO₂)_8.1] are available free of charge from the Cambridge Crystallographic Data Centre (Deposition Numbers 2132832, 2151090, 2151089, 2151088 and 2151087).

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

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, S. Yang, Angew. Chem. Int. Ed. 2022, 61, e202207259; Angew. Chem. 2022, 134, e202207259.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.