Pore engineering of porous framework materials for efficient SF₆ capture

Abstract

Sulfur hexafluoride (SF₆) is an artificial inert gas widely used in the power and semiconductor industries and is known to be a significant contributor to the greenhouse effect due to its high global warming potential. Porous framework materials (PFMs), including metal–organic frameworks (MOFs), covalent organic frameworks (COFs) and porous aromatic frameworks (PAFs), are well-established SF₆ adsorbents characterized by highly ordered pore structures that can efficiently adsorb SF₆via various electrostatic interactions. Recent advancements have focused on optimizing electrostatic interactions with SF₆ molecules to enhance their SF₆ adsorption performance in industrial settings. However, a comprehensive review on precise control of the structural and chemical properties through pore engineering strategies for PFMs has not been reported yet. This review systematically outlines the structure–activity relationship and the dominant mechanisms of host–guest interaction modes for effective SF₆ capture and discusses pore engineering strategies to achieve efficient SF₆ capture. Specifically, adjusting the pore structure to align with SF₆ molecules, modifying the pore surface to strengthen electrostatic interactions and designing hierarchical pore structures to enhance SF₆ adsorption kinetics are involved. Lastly, the potential of PFMs for SF₆ capture is discussed from perspectives of AI-driven creation of high-performance PFMs, construction of PFMs with enhanced water- and acid-resistance under specialized conditions and shaping methods, aiming to guide the development of advanced SF₆ adsorbents based on PFMs.

---

1. Introduction

With the development of economy and the increasing electricity demand, the electric power industry is progressively moving towards high voltage and ultra-high voltage.^1,2 Insulating materials play a crucial role in ensuring the safe and stable operation of power systems.^3,4 Gas insulation materials are extensively utilized in high-voltage and ultra-high-voltage power equipment due to their advantages of relatively light weight, blast resistance and especially their self-healing ability after insulation damage.^5–8

Sulfur hexafluoride (SF₆) is an artificial insulating gas synthesized by two French chemists, Moissan and Lebeau, in 1900 and has enjoyed a history of over a hundred years.⁹ SF₆ is composed of six fluorine atoms and a central sulfur atom with a kinetic diameter of 5.2 Å. Its octahedral molecular structure exhibits short bond distance and high bond energy, imparting exceptional chemical stability.¹⁰ SF₆ is an electronegative gas that readily attracts free electrons to diminish the ionization process after electric breakdown, making it an excellent insulator and arc extinguisher.¹¹ Consequently, SF₆ has been the predominant choice in power facilities since 1947,¹² including gas insulated switchgear (GIS),¹³ generator circuit breaker (GCB),¹⁴ gas insulated transmission line (GIL)¹⁵ and gas insulated transformer (GIT).¹⁶ Additionally, SF₆ serves as a plasma etchant in the semiconductor industry for silicon-based chip etching.^17,18

SF₆ poses significant environmental concerns despite its numerous advantages, since SF₆ has a global warming potential 23 900 times that of CO₂ and an atmospheric lifetime of up to 3200 years due to its extremely high chemical stability.¹⁹ As an important contributor to the greenhouse effect, the increase of SF₆ concentration would significantly warm up the Earth and cause other environmental problems, which has triggered global efforts to reduce the concentration of SF₆ in the atmosphere.²⁰ The Kyoto Protocol stipulates SF₆ as one of the six major greenhouse gases and advocates for using the SF₆/N₂ mixture to mitigate the environmental impact without compromising its insulating properties.^5,21 However, the emission of SF₆ from aging pipelines and improper storage has still caused it to increase by threefold of the global mean concentration up to 10.5 ppt (t for trillion) in the past 25 years.^10,22 Therefore, an efficient method for selectively capturing SF₆ is needed to avoid its environmental hazards and facilitate its reuse.^9,23,24 The adsorption separation technology with the advantages of easy operation and without SF₆ phase change can capture and recover valuable fluorinated gases from industrial waste gases. Considering the low environmental concentration of SF₆, non-thermal adsorption technology is the most promising and energy-efficient separation technique.^20,25,26

Various porous materials, such as zeolites,^27–29 porous polymer and organic cages,^30–32 activated carbon,^33,34 and porous framework materials (PFMs), have been investigated for the selective capture of SF₆. Among them, PFMs, including metal–organic frameworks (MOFs), covalent organic frameworks (COFs) and porous aromatic frameworks (PAFs), are reported as famous SF₆ adsorbents.^35–40 Their high surface area and tunable pore structure endow high adsorption uptake and selectivity for SF₆via high affinity from electrostatic interaction sites.^41–45 Effective pore engineering promises enhanced SF₆ adsorption performance of PFMs by controlling their characteristics, including pore sizes, surface chemistry and flexibility.^46–49 The burgeoning research demonstrated promising SF₆ separation capabilities of PFMs over the past decade. However, a comprehensive overview focusing on the correlation between the PFM structure and their SF₆ adsorption performance has not been reported till now, which is pivotal for guiding the design and fabrication of PFMs for efficient SF₆ capture.

This review summarized the dominant mechanisms of SF₆ adsorption on PFMs and delved into the latest advancements in PFMs as SF₆ adsorbents, based on pore engineering strategies to enhance their adsorption performance. To begin with, the standards for evaluating the performance of adsorbents and the host–guest interaction modes for effective SF₆ capture were introduced. Subsequently, modulation strategies for pore structure and surface chemistry were discussed for achieving enhanced capture of SF₆. Finally, future challenges and prospects of PFMs as SF₆ adsorbents in wider applications were addressed, aiming to improve their SF₆ adsorption performance in industrial settings.

2. Properties and mechanisms of SF₆ adsorption on PFMs

2.1. Properties

In order to effectively evaluate the performance of PFMs, the following aspects during the adsorption process should be considered.

2.1.1. Adsorption isotherm. An adsorption isotherm refers to the relationship between the gas pressure and the amount of gas molecules at a certain temperature.⁵⁰ It is typically plotted at a constant temperature and demonstrates the adsorption capacity of PFMs for SF₆.⁵¹ The single-component isotherm provides fundamental data on the SF₆ capture capacity of PFMs and determines the maximum adsorption capacity of PFMs. As the basis for studying interactions between adsorbents and gas molecules, adsorption isotherms would be greatly influenced by the structural characteristics of PFMs and therefore cannot be used as the sole evaluation criterion.

2.1.2. Adsorption selectivity. Adsorption selectivity refers to the preferential adsorption capability of PFMs for SF₆ in a multi-component gas mixture, reflecting the ability to selectively separate different gases. There are two methods for calculating selectivity: IAST selectivity and Henry constant selectivity. (i) IAST selectivity can be predicted from the single-component adsorption isotherm using the ideal adsorption solution theory (IAST), which is only applicable to the ideal state of binary mixtures adsorbed uniformly on the surface of the adsorbent.⁵² It also relies on the accurate fitting of the single-component isotherm, which makes it a challenge to fit the adsorption isotherm of a weakly bound component N₂ since PFMs hardly adsorb N₂ at room temperature. Meanwhile, IAST selectivity cannot apply on flexible frameworks due to the door-opening effect caused by gas adsorption.⁵³ (ii) Henry's constant selectivity: Henry's constant is an indicator of the affinity between the host (frameworks) and the guest (gas molecules), describing the amount of gas that is adsorbed on a solid surface at unit pressure. Based on the single-component low-pressure adsorption isotherm, Henry's constant of a gas molecule can be calculated and thereby the upper limit of SF₆/N₂ selectivity.⁵⁴

2.1.3. Isosteric heat of adsorption (Q_st). Isosteric heat of adsorption refers to the heat released during the adsorption process and reflects the strength and stability of the interaction between SF₆ and PFMs.⁵⁵ Measurement of isosteric heat of adsorption is crucial for evaluating the thermodynamic properties of adsorption processes, helping to understand the types of interactions between SF₆ and PFMs. Notably, isosteric heat of adsorption varies with the coverage of SF₆ on the adsorbent surface, generally decreasing with increasing adsorption amount as adsorption becomes more difficult.^56,57

2.1.4. Adsorption kinetics. Gas adsorption kinetics studies the time-dependent behavior of gas molecules during the adsorption process and primarily focuses on the dynamics of diffusion and adsorption rates.⁵⁸ Adsorption kinetics help find out how the adsorption sites are occupied and the diffusion rate of adsorbate molecules on the adsorbent surface, gaining a deeper understanding of the dynamic changes in the interaction between the adsorbent and the adsorbate. The kinetic behavior of adsorbents mainly focuses on the penetration curve and adsorption rate curve, which are crucial for providing fundamental information required for the design and operation of adsorption equipment used in industrial applications.^59,60

2.2. Adsorption mechanisms

2.2.1. Adsorption thermodynamics. Physical adsorption is usually described as weak interactions between the adsorbate and the adsorbent by adsorption thermodynamics, which originates from intermolecular forces. The interactions between PFMs and SF₆ can be roughly divided into Lewis acid–base interaction, F–π interaction, and the induced dipole moment effect.
2.2.1.1 Lewis acid–base interaction. SF₆ with highly electronegative F atoms can act as a Lewis base and interact with Lewis acid sites through coordination. Typical Lewis acid sites include open metal sites (OMS) and hydrogen bonds (H-bonds).^61,62

(i) Open metal sites: according to the Lewis acid–base theory, OMS are a typical kind of Lewis acid sites with unfilled bond orbitals and tend to match paired electrons in SF₆ through the interaction of σ bonds and π bonds.⁶³ For instance, Chowdhury and co-workers found that HKUST-1 (Cu₃(1,3,5-benzenetricarboxylate)₂) contained a certain amount of unsaturated Cu^II sites, which allowed HKUST-1 to adsorb SF₆via strong electrostatic attraction between electropositive Cu^II and electron-rich SF₆.⁶⁴ Experimental results showed that HKUST-1 had a much higher SF₆ adsorption capacity of up to 2.23 mmol g⁻¹ than UiO-66 (0.82 mmol g⁻¹) with similar pore sizes at 0.1 bar.⁶⁵ Subsequently, OMS have been confirmed to enhance the interaction between the adsorbate and adsorbent and may further promote the adsorption through the interaction between the adsorbate and adsorbate.

(ii) H-bonds: H atoms in hydrogen-containing functional groups tend to show partial positive charge and become Lewis acid sites to form strong polar H-bonds with F atoms in SF₆.⁶⁶ For example, the electron cloud in the –NH₂ group of UiO-66-NH₂ deflected towards N atoms with higher electronegativity, while the H atoms were almost in a proton state, capable of forming N–H⋯F bonds with SF₆.⁶⁷ As a result, UiO-66-NH₂ showed an increased Henry's constant for SF₆ up to 38.2 mmol g⁻¹ bar⁻¹ compared with that of UiO-66 (22.2), indicating the improvement of the intrinsic interaction between the framework and SF₆.

2.2.1.2. F–π interaction. The F–π interactions are electrostatically driven non-covalent interactions governed by the high electronegativity F atoms.^68,69 The large electron cloud in aromatic molecules of the delocalized π-bonding system can form multiple F–π interactions with F atoms in SF₆, which act as strong π-electron donors.^70,71 The stability of the F–π interaction can be enhanced by increasing the electron deficiency in the aromatic system, thus correlating with the electrostatic surface potential (ESP) of the aromatic system. For example, SBMOF-1 with a pocket-like pore formed by four benzene rings attracted SF₆ through F–π interactions.⁷² This is because the surface potential of the aromatic ring of the SDB ligands (4,4′-sulfonyldibenzoic acid) in SBMOF-1 was positive, while the surface potential of SF₆ was negative.
2.2.1.3. Induced dipole moment effect. The electric field of polar functional groups can cause the deformation of the electron cloud and generate induced dipole moments to polarize SF₆ molecules.⁷³ The S–F bond in SF₆ molecules would be polarized by the induced electric field gradient created by polar functional groups in the framework, generating electron transfer between PFMs and gas molecules. For instance, the entry of strongly electronegative F atoms in fluorinated PAFs (PAF-4F and PAF-8F) caused localized charge separation on the pore surface, making benzene rings become electropositive sites and intensify the adsorption affinity towards electronegative SF₆.⁷⁴

2.2.2. Adsorption kinetics. Kinetic separation is achieved through the difference in diffusion rates of different gas molecules, which is also known as the partial molecular sieving effect. For example, Skarmoutsos et al. reported an interpenetrated MOF SIFSIX-2-Cu-i to separate SF₆ from SF₆/N₂ thoroughly.⁷⁵ The interpenetration of the SIFSIX-2-Cu network ([Cu(4,4′-dipyridylacetylene)₂(SiF₆)]_n) resulted in the formation of the isomer SIFSIX-2-Cu-i, which significantly reduced its channel size from 13.05 Å to 5.15 Å. The steric hindrance resulting from the smaller pore size of the interpenetrating structure compared to the kinetic size of SF₆ completely excluded SF₆ from the MOF, resulting in an “infinite high” selectivity.

On the other hand, PFMs with mesopores can reduce the mass transfer resistance of SF₆ to improve the absorption performance and reduce the time required for SF₆ absorption to reach equilibrium. For example, Chuah et al. synthesized HKUST-1c nanocrystals with a hierarchical pore structure, which exhibited higher SF₆ uptake (4.98 mmol g⁻¹vs. 3.56 mmol g⁻¹ at 25 °C and 1 bar) and better SF₆/N₂ selectivity (70 vs. 50 at 25 °C) compared to bulk HKUST-1 crystals.⁷⁶ This result may indicate that the mesopores in HKUST-1c not only shortened the diffusion length and promoted the transport of SF₆ molecules to active sites but also increased the number of accessible active sites for SF₆ molecules.

3. Pore engineering strategies for enhancing the SF₆ adsorption performance of PFMs

In the process of SF₆ adsorption, SF₆ molecules first diffuse through porous channels to the adsorption sites and then are anchored via various electrostatic interactions. Therefore, the adsorption performance of PFMs would be influenced by pore sizes, pore shape and surface properties.²⁶ Based on the properties and mechanisms of PFMs mentioned before, it can be concluded that PFMs capable of capturing SF₆ efficiently in their pores have a strong tendency to form various electrostatic interactions with SF₆ molecules. The metal node composition and functionality present in ligands are fundamentally important structural parameters that determine the affinity of PFMs to adsorb SF₆. In addition, such PFMs must exhibit appropriate pore apertures and pore shapes matching the kinetic diameters of SF₆. In this section, several useful strategies and approaches from the perspective of pore engineering were given to enhance the interaction between PFMs and SF₆ and achieve effective SF₆ capture.

3.1. Metal–organic frameworks

Metal–organic frameworks (MOFs) are known as a class of relatively new porous materials with a periodic network composed of metal ions/clusters and organic ligands linked via coordination bonds.^77–79 Different from other traditional porous materials such as zeolites and activated carbons, MOFs possess clearly defined and adjustable crystal structures that are more suitable for on-demand targeted design.⁸⁰ Evidently, gaining a better understanding of the structure–property relationship offers great potential for the enhancement of SF₆ adsorption performance on MOFs by structural adjustments via isoreticular chemistry or by the design and construction of new MOF structures via the practice of reticular chemistry.^81,82 This section focused on enhancing pore-matching effects and multiple interactions between the framework and gas molecules to improve the SF₆/N₂ separation of MOFs, and Table 1 summarizes recent advances in MOFs for SF₆ adsorption.

Table 1 The SF₆ adsorption characteristics of metal–organic frameworks

Sample name	BET surface area (m^2 g^−1)	Pore size (Å)	SF6 uptake at 0.1 bar (mmol g^−1)	SF6 uptake at 1.0 bar (mmol g^−1)	SF6/N2 selectivity (1 : 9)	SF6 isosteric heat of adsorption, Qst (kJ mol^−1)	Ref.
UiO-66	1074	9	0.82	1.45	140.0	32.0	64
UiO-66-Br2	616	9	0.78		220	45.0
SBMOF-1	169	5.5		1.14	92.6	34.2	66
V-TBAPy	1370	6.40	1.32	3.28	360.7	30.48	78
Ga-TBAPy	1484	5.90	1.33	3.50	418.5	30.44
V-TCPB	1107	5.38	2.29	3.07	65.6	28.08
Ga-TCPB	1144	7.85	2.25	2.95	55.4	27.18
Cu(peba)2	814	12.0	0.15	2.36	18.2	14.4	79
Ni(pba)2	807	8.2	1.70	3.50	200.6	24.0
Ni(ina)2	470	6.0	2.39	2.84	375.1	33.4
ZIF-8	1540	3.4		0.15	1.7		80
ZIF-70.26-80.74	948	7.2		2.08	40
SNNU-202	1319	9.0		4.94	9	20.0	81
SNNU-203	1522	7.8		5.43	27	20.8
SNNU-204	2170	7.6		6.06	49	21.0
Mg-MOF-74	1631	11.0	2.10	6.42	18.0	32.0	82
Co-MOF-74	1219	11.0	2.00	5.34	34.0	40.0
Zn-MOF-74	992	11.0	1.40	3.90	46.0	26.0
Ni(3-mbpa)2	835	11.9	1.79	2.83	221.0	30.2	83
Co(3-mbpa)2	626	12.5	1.77	3.25	161.0	26.8
Ni(adc)(dabco)0.5	744	5.5	2.23	2.38	919.4	47.6	84
Cu-MOF-NH2	2145	6	3.39	7.88	266	55.2	85

---

3.1.1. Pore size regulation. MOFs offer immense design advantages over other porous materials as their pore size can be readily controlled through exchanging metal ions and organic ligands. The adjustment of pore size can boost both SF₆ selectivity and uptake of MOFs by enhancing the affinity towards SF₆.

The coordination pattern of metal ions profoundly influences the topological structure of MOFs, leading to tunable gas adsorption capability. Åhlén et al. designed a series of MOFs based on TBAPy⁴⁻ (1,3,6,8-tetrakis(4-carboxyphenyl)pyrene) to investigate the impact of metal species on pore size.⁸⁶ Yb- and Tm-TBAPy exhibited similar diamond-shaped channels with a pore size distribution of approximately 6.5 Å, while Hf-TBAPy featured hexagonal channels spanning about 7.2 Å in diameter. Smaller pores allowed for greater potential overlap with SF₆ molecules, thereby enhancing their SF₆ adsorption and resulting in higher adsorption capacities for Yb- and Tm-TBAPy. Experimental results showed that the SF₆ uptake of Yb- and Tm-TBAPy reached 2.33 mmol g⁻¹ and 1.83 mmol g⁻¹ at 1 bar, surpassing that of Hf-TBAPy (1.38 mmol g⁻¹).

Subsequently, they designed a series of rod MOFs constructed by high valence metal ions (Ga(III) and V(IV)) with suitable pore size and uniform one-dimensional channels for efficient capture of SF₆ (Fig. 1a and b).⁸³ The rod secondary building units (SBUs) of MOFs coordinated with carboxylic acid groups of the ligands to form one-dimensional rhombohedral channels. The benzene ring of the linkage contracted along the channel direction with pore sizes of about 5.5 Å, which increased the affinity toward SF₆ molecules, resulting in steep adsorption isotherms at low pressure (Fig. 1c). Experimental results showed the SF₆ uptake of V-TCPB and Ga-TCPB (TCPB = 1,2,4,5-tetrakis(4-carboxlatephenyl)benzene) up to 3.0 mmol g⁻¹ with excellent SF₆/N₂ selectivity above 350.

Fig. 1 (a) Synthesis route of Ga- and V-based MOFs with the respective tetratopic H₄TBAPy and H₄TCPB ligands. (b) Structure of Ga-TCPB as viewed along the a-axis showing the presence of two unique pores (colored in green and blue) and the rod SBU running along the a-axis in the structure. (c) SF₆ and N₂ adsorption isotherms recorded at 293 K. Reproduced with permission from ref. 78. Copyright 2023 Royal Society of Chemistry.

The pore size of MOFs can also be easily modified by linearly adjusting the length of the ligand within the existing parent structure, offering viable approaches to strengthen the host–guest interaction and enhance their SF₆ adsorption properties. For example, Wang et al. used a family of bifunctional pyridylcarboxylate ligands with different lengths to create ultra-microporous MOFs for SF₆ selective adsorption at low pressure.⁸⁷ The pore size decreased from 12 to 6 Å with the reduction of the ligand length (Fig. 2a). Notably, the SF₆ adsorption isotherm of Ni(ina)₂ reached a plateau at very low pressure (>0.1 bar), indicating strong interactions between SF₆ molecules and the framework (Fig. 2b). This was attributed to the perfect match between the pore and SF₆ molecules (6 Å vs. 5.2 Å), which allowed the framework to generate multipoint van der Waals interactions with SF₆ through several F–H bonds. As a result, Ni(ina)₂ exhibited an unprecedented SF₆/N₂ selectivity up to 375.1 (Fig. 2c).

Fig. 2 (a) The pore structures of Cu(peba)₂, Ni(pba)₂ and Ni(ina)₂ and the calculated adsorption binding sites of SF₆ in Ni(ina)₂. (b) SF₆ and N₂ adsorption isotherms of Cu(peba)₂, Ni(pba)₂ and Ni(ina)₂ at 298 K. (c) SF₆/N₂ selectivity of the three MOFs calculated at 298 K (SF₆/N₂, 1/9, v/v) using IAST. Reproduced with permission from ref. 79. Copyright 2022 Wiley-VCH.

Introducing multiple ligands into the same framework can adjust the structure of MOFs from the coordination sites and redistribute their pore sizes, thereby enhancing their affinity with SF₆ molecules. For instance, Åhlén and colleagues developed a series of mixed-linker ZIF-7-8s by substituting the 2-methylimidazolate (mIm) ligand in ZIF-8 with a larger benzimidazolate (bIm) ligand to reduce its pore size and enhance the selective adsorption for SF₆.⁸⁸ The inner cage sizes of ZIF-8 were reduced from 12.5 Å to 7.3 Å by adjusting the ratio between bIm and mIm ligands, thereby enhancing the van der Waals interaction between SF₆ molecules and the framework. Among them, ZIF-7_0.2-8_0.8 with a bIm ligand concentration of 20% exhibited notable improvements compared with the original ZIF-8, surging the adsorption capacity from 0.15 mmol g⁻¹ to 2.08 mmol g⁻¹ and SF₆/N₂ selectivity from 1.39 to 40.04.

Subsequently, Li et al. achieved pore-window size control in Cu-MOFs via a mixed-linker strategy.⁸⁴ Pore window partitions were realized by connecting adjacent square metal nodes through ditopic linear linkages, allowing for stepwise modification of pore sizes from 9.0 to 7.6 Å (SNNU-202-204) (SNNU-202 = Cu₆(H₂O)₃(NTB)₄(PYZ)_1.5, SNNU-203 = Cu₁₂(H₂O)₃(NTB)₈(PYZ)₃(BPY)_1.5 and SNNU-204 = Cu₆(NTB)₄(DABCO)_1.5(BPY)_1.5, H₃NTB = nitrilotribenzoic acid, PYZ = pyrazine, BPY = 4,4′-bipyridine and DABCO = 1,4-diazabicyclo[2.2.2]octane) (Fig. 3). The ultra-microporous structure of SNNU-204 formed strong interaction with SF₆ molecules, leading to a significant increase of SF₆/N₂ selectivity from 9 to 49.

Fig. 3 Progressive process of stepwise pore window partitions based on different partitional linkers. Reproduced with permission from ref. 81. Copyright 2024 Elsevier.

3.1.2. Pore environment regulation. The pore microenvironments of MOFs, including open metal sites (OMS), functional groups in ligands and topological structures, are of great importance for the enhancement of SF₆ adsorption and selectivity.
3.1.2.1. Open metal sites. As reported earlier, the species and quantity of OMS in MOFs can greatly affect their adsorption properties towards SF₆via electrostatic attraction. Kim et al. proposed adjusting OMS species to change their SF₆ adsorption properties.⁸⁵ Primarily, in this work, a series of M-MOF-74s (M = Mg, Co, and Zn) with highly dense OMS have been constructed and used to capture SF₆ under low pressure (Fig. 4a). Experimental results showed that the SF₆ adsorption capacity of these three MOFs exhibited the order of Co-MOF-74 > Mg-MOF-74 > Zn-MOF-74. Among them, Co-MOF-74 featuring unsaturated Co^II sites had the highest adsorption uptake and adsorption heat for SF₆ (Fig. 4b and c). This was ascribed to the strongest polarization of Co^II sites, which generated very strong electrostatic interaction with S–F bonds of SF₆. Meanwhile, Co-MOF-74 was proved to possess about 1.5 times higher density of Co^II sites compared to Cu^II sites in HKUST-1,⁸⁹ achieving a SF₆ adsorption capacity of up to 5.34 mmol g⁻¹ at 1 bar.

Fig. 4 (a) The effective removal of guest molecules from Co-MOF-74 during the activation process. Reproduced with permission from Liang et al.⁸⁹ Copyright 2021 Elsevier. (b) Adsorbed SF₆ molecules per metal site at 298 K. (c) Isosteric heat of adsorption (Q_st) for SF₆ adsorption in three MOFs. Reproduced with permission from ref. 82. Copyright 2014 Elsevier.

3.1.2.2. Functional groups of MOF ligands. Functional groups in MOF ligands also play a crucial role in SF₆ adsorption performance through influencing the formation of H-bonds and pore polarizability in MOFs. For instance, introducing functional groups containing H atoms, such as methyl (–CH₃) and amino (–NH₂) groups, into ligands can create more affinity adsorption sites towards SF₆ in MOFs. Zheng et al. implanted a methyl group (–CH₃) into 4-(4-pyridyl)benzoate (pba) ligands to synthesize Ni(3-mpba)₂ (3-mpba = 3-methyl-4-(4-pyridyl)benzoate) (Fig. 5a).⁹⁰ In this system, methyl group in 3-mpba efficiently formed H-bonds with SF₆, which strengthened its SF₆ adsorption via multiple site bindings (Fig. 5b). This increased affinity could also be expressed by the change of their isosteric heat of adsorption (Q_st) and SF₆/N₂ selectivity. Methyl-modified Ni(3-mpba)₂ exhibited 6.2 kJ mol⁻¹ higher enthalpy value of SF₆ compared with its pristine Ni(pba)₂, and its SF₆/N₂ selectivity increased from 160 to 221 (Fig. 5c).

Fig. 5 (a) The coordination geometry and pore structures of Ni(pba)₂ and Ni(3-mpba)₂. (b) Simulation of SF₆ binding sites in Ni(3-mpba)₂. (c) Isosteric heat of adsorption (Q_st) for SF₆ and N₂ adsorption in Ni(pba)₂ and Ni(3-mpba)₂. Reproduced with permission from ref. 83. Copyright 2023 Elsevier.

Additionally, Kim et al. introduced a series of halogen groups (Cl, Br, and I) in UiO-66 to explore the polarizability effect of the adsorbent surface on the adsorption properties of SF₆.⁶⁷ Their adsorption capacity of SF₆ under low pressure is given as follows: UiO-66-Br₂ > UiO-66-I > UiO-66-Br > UiO-66-Cl > UiO-66. It was found that their adsorption affinity increased with the atomic weight of these halogen atoms, which facilitated the van der Waals interaction towards SF₆. Moreover, the introduction of bromine atoms partially blocked its pores and narrowed the pore size from 12 to 9 Å, thereby improving the SF₆/N₂ selectivity from 140 to 220. Chen et al. introduced F atoms into tetrazole ligands to synthesize extensively fluorinated MOFF-5.⁹¹ However, MOFF-5 only exhibited a modest SF₆ adsorption capacity of 0.08 mmol g⁻¹ at 0.1 bar, which was due to the mismatch between the highly polarized environment inside fluorinated pores and SF₆ with low polarity, resulting in the absence of strong binding sites.

3.1.2.3. Topological structure. Topological features of MOFs could affect SF₆ capture through forming multiple sites with the aid of the pore confinement effect.

Wu et al. constructed a benzene encircled SBMOF-1 with four benzene rings that formed a 5.5–8.5 Å pocket-shaped micropore to accommodate SF₆ (5.2 Å) (Fig. 6a).⁷² High amounts of C–H groups in aromatic rings formed multiple van der Waals interactions with the F atoms of SF₆ molecules and efficiently increased synergistic interactions (Fig. 6b). As a result, SBMOF-1 exhibited SF₆ selectivity reaching 727 at 300 ppm.

Fig. 6 (a) Pore structure of SBMOF-1 from the front view. (b) Electrostatic potential of N₂ and SF₆ inside the pore channel of SBMOF-1. (c) SF₆ and N₂ adsorption isotherms of SBMOF-1 at 298 K. Reproduced with permission from ref. 66. Copyright 2022 Elsevier.

Yang and co-workers used naphthalene rings as ligands to construct a splint-like pore in Ni(NDC)(TED)_0.5 (H₂NDC = 1,4-naphthalenedicarboxylic acid, TED = triethylenediamine), which could form a stronger potential overlap with SF₆.⁹² The F–π interactions between naphthalene rings and SF₆ were proved to be stronger than H-bonds. Therefore, Ni(NDC)(TED)_0.5 exhibited a SF₆ selectivity of up to 750 and a SF₆ uptake of up to 2.76 mmol g⁻¹ at 0.1 bar.

In addition, Chang et al. designed a single-molecule SF₆ trap based on a pillar-layered MOF (Ni(adc)(dabco)_0.5) (H₂adc = 9,10-anthracenedicarboxylic acid, dabco = 1,4-diazabicyclo[2.2.2]octane).⁹³ The parallel anthracene rings created 5.5 Å splint-like pores with a giant potential overlap with SF₆, thereby confining gas molecules within the pore (Fig. 7a). Therefore, this single-molecule trap attracted SF₆ through multiple affinity sites and a fluorophilic microenvironment, achieving unprecedented separation selectivity (Fig. 7b). As a result, Ni(adc)(dabco)_0.5 exhibited an adsorption heat of up to 47.6 kJ mol⁻¹ with ultra-high SF₆/N₂ selectivity (919.4) (Fig. 7c).

Fig. 7 (a) Structure description of Ni(adc)(dabco)_0.5. (b) Periodic DFT-optimized adsorption configuration of the SF₆ molecule in the splint-like pore. Atom colors: Ni, cyan; O, red; N, blue; C, gray; H, white. (c) Comparison of the IAST-predicted selectivity of Ni(adc)(dabco)_0.5 with some selected benchmark materials reported in the literature. The molar composition of the bulk mixture: SF₆/N₂ = 10/90; operation conditions: 298 K. Reproduced with permission from ref. 84. Copyright 2022 American Chemical Society.

Ren et al. found that the MOF (Cu-MOF-NH₂) with calix-shaped pores surrounded by benzene rings and multiple annularly pendant amino groups shows excellent SF₆ capture (Fig. 8a).⁹⁴ Such an internal environment endowed the framework with a multi-dimensional confinement towards SF₆via strong H-bonds and van der Waals-type interactions. As a result, Cu-MOF-NH₂ showed a SF₆ adsorption capacity of 3.39 mmol g⁻¹ at 0.1 bar and a SF₆/N₂ selectivity of up to about 266 (Fig. 8b and c).

Fig. 8 (a) Representation of the [calix3]arene analogous microenvironments present in Cu-MOF-NH₂ and IGM-derived isosurfaces of δ^inter_g (isovalue = 0.005 a.u.) for SF₆ in cavity A. (b) Experimental N₂ (blue) and SF₆ (red) single component adsorption isotherms and the IAST-derived SF₆/N₂ selectivity (pink) in Cu-MOF-NH₂ at 298 K. (c) Separation performance of Cu-MOF-NH₂ and various benchmark materials. Reproduced with permission from ref. 85. Copyright 2021 American Chemical Society.

3.1.3. Flexible structure. A portion of MOFs have flexible structures and undergo structural changes when exposed to external stimuli. Variation in temperature and pressure would trigger ligand or metal node rotation and cause thermal expansion of MOFs.^95–97 The flexible nature of molecular cage crystals can break the barrier to SF₆ diffusion into the flexible cage window, changing their original adsorption behavior for SF₆, such as adsorption affinity and adsorption selectivity.

The gate-opening phenomenon of flexible MOFs allows gas molecules having larger sizes to enter their pores, thereby achieving higher adsorption capacity and selectivity of SF₆ under specific pressure and temperature stimuli.⁷⁸ Choi et al. investigated the effect of the gate-opening phenomenon via ligand rotation of ZIF-8 under threshold pressure on selective adsorption properties of SF₆.⁹⁸ The ligands of ZIF-8 rotated at a certain angle induced by high pressure, which increased its pore size from 3.4 to 7.6 Å to enhance its adsorption capacity and form a strong interaction with SF₆. Experimental results showed that the rotated ZIF-8 presented an increase of SF₆ uptake of up to 3.35 mmol g⁻¹ when the pressure was increased to 20 bar. When the pressure was reduced to atmospheric pressure, SF₆ molecules were locked in the shrunk pores of ZIF-8. Finally, SF₆ can be recovered through variable temperature desorption, achieving efficient separation and capture of SF₆ from various gases.

Senkovaska and colleagues developed a type of pillar-layered MOF M₂(bdc)₂(dabco) (M = Co, Ni, bdc = 1,4-benzenedicarboxylate; dabco = diazabicyclo[2.2.2]octane) with flexible M₂ paddle-wheel nodes exhibiting a gate-opening phenomenon (Fig. 9a and b), which showed excellent adsorption performance of SF₆.⁹⁹ As shown in Fig. 9c, MOFs transformed from a narrow pore (np)-phase (5 Å) to a large pore (lp)-phase (7–9 Å) due to the swelling and switching of flexible metal nodes under high pressure stimuli, allowing SF₆ molecules to enter the pores and form strong interactions with MOFs. Meanwhile, Co₂(bdc)₂(dabco) showed a greater phase transition magnitude than Ni₂(bdc)₂(dabco) since the outer electrons of Co were more tolerable to the changes in the coordination environment. Therefore, Co₂(bdc)₂(dabco) reached a high SF₆ uptake of 4.95 mmol g⁻¹ at 15 bar, which was 7.4 times greater than that of Ni₂(bdc)₂(dabco).¹⁰⁰

Fig. 9 (a) Depiction of the building blocks used for the synthesis of M₂(fu-bdc)₂(dabco) materials. (b) Side view of a single cavity of the extended 3D network structure. (c) Schematic illustration of the np → lp phase transition of the M₂(fu-bdc)₂(dabco) materials. Reproduced with permission from ref. 92. Copyright 2018 American Chemical Society.

Other flexible MOFs show thermal expansion features when suffering the temperature variation, and the changing porous structure would affect SF₆ adsorption behaviors. Åhlén et al. reported that CAU-33 (1,2,4,5-tetrakis-(4-carboxyphenyl)benzene) exhibited positive thermal expansion properties, which could increase its pore volume to enhance SF₆ capacity through temperature increase.¹⁰¹ The covalent bonds between metal centers and ligands were weak due to the low coordination number of Bi molecules and therefore could be elongated under heating conditions. Experiment verified that its pore volume could expand by 9% when the temperature was increased to 60 °C, thereby enhancing the adsorption capacity of SF₆ (1.10 mmol g⁻¹) which is twice that of the value observed at 20 °C (0.58 mmol g⁻¹).

Senkovaska et al. found that Zn₄O(dmcpz)₃ (dmcpz = 3,5-dimethyl-4-carboxypyrazolato) exhibited negative thermal expansion properties, which could reduce its pore size to improve the SF₆/N₂ selectivity by increasing the temperature.⁶⁵ The effective length of the ligands was shortened with the lateral vibration of benzene rings during the heating process, thereby reducing the pore size of Zn₄O(dmcpz)₃ and generating stronger van der Waals interactions with SF₆. The pore size of Zn₄O(dmcpz)₃ decreased to 6.5 Å at 60 °C, thereby increasing its SF₆/N₂ selectivity from 44 (at room temperature) to 80.

3.2. Covalent organic frameworks

Another family of porous organic frameworks is covalent organic frameworks (COFs) constructed by covalent combination of multiple organic linkages.¹⁰² They have the similar merits of high porosity and tunable surface properties to MOFs. Differently, their metal-free and fully conjugated structures are beneficial to enhance the chemical stability and surface hydrophobicity and thus expand the SF₆ capture application of humid environments.¹⁰³ Furthermore, the unique covalent structures of COFs are vital to facilitate electron transport efficiency for SF₆ selective capture.¹⁰⁴

Zheng et al. designed a 2-dimensional COF-6 with a graphite-like structure via B–O–C covalent bonds of boronate esters to selectively adsorb SF₆ under low pressure (<0.1 bar).¹⁰⁰ The large aromatic rings endowed the framework with a highly π-conjugated structure, allowing COF-6 to adsorb SF₆ through strong electrostatic interactions. Meanwhile, the graphite-like organic layer partially slipped and generated 6.4 Å grottos along the pore wall of COF-6 and trapped SF₆ molecules with suitable pore confinement. As a result, the two-dimensional COF-6 achieved a saturated adsorption of 3.71 mmol g⁻¹, and its SF₆/N₂ selectivity reached 860, greatly exceeding other previously reported porous materials.

Liao et al. designed RCOF-1 connected by bis-imine (C=N) to further enhance the adsorption of SF₆ through multiple adsorption sites (Fig. 10a).¹⁰⁵ RCOF-1 possessed two adjacent chemical groups of benzene ring and bis-imine and formed a unique intersection angle at the edge of its pore cages. Among them, benzene rings formed F–π interactions at the benzene center and multiple F–H bonds at the edge with the SF₆ molecule. Moreover, its imine group could attract electronegative SF₆ molecules through lone pairs of N atoms at the same time. These sites formed multisite capture towards the SF₆ molecule with a high binding energy of 27.0 kJ mol⁻¹, as illustrated by the green iso-potential surface in Fig. 10b. Its adsorption experiments showed that RCOF-1 had a superior SF₆ adsorption capacity of up to 4.13 mmol g⁻¹ and a relatively high SF₆/N₂ selectivity of up to 125 at 273 K and 100 kPa compared with the state-of-the-art porous materials (Fig. 10c and d).

Fig. 10 (a) Topological structures of RCOF-1. (b) IGMH map of the cluster model consisting of SF₆ and RCOF-1. (c) SF₆ adsorption isotherms of RCOF-1. (d) Calculated IAST SF₆/N₂ (10 : 90) selectivity of RCOF-1. Reproduced with permission from ref. 98. Copyright 2023 Elsevier.

Polarized modification of the COF pore surface can create high SF₆ affinity sites to promote the selectivity for SF₆.¹⁰⁶ For instance, Xi's group proposed the use of fluorinated monomers to construct FCOF-1 to promote SF₆ selective adsorption.¹⁰⁵ Benzene rings and F atoms in FCOF-1 generated 3D multiple adsorption sites and formed strong electrostatic attraction towards SF₆ through π-conjugation and F–F polar interaction. Therefore, FCOF-1 exhibited a SF₆/N₂ selectivity of 22.1 and a SF₆ uptake of 2.21 mmol g⁻¹ at 273 K and 100 kPa.

Similarly, Xi's group introduced a trifluoromethyl group (–CF₃) into BrCOF-2 through post-synthetic modification to improve its SF₆/N₂ selectivity (Fig. 11a).¹⁰⁷ The inclusion of –CF₃ altered the charge distribution in BrCOF-2, making it more effective to polarize SF₆. Furthermore, the introduction of the bulky –CF₃ functional groups reduced the pore size from 22 to 13 Å and further strengthened the binding force towards SF₆. As a result, BrCOF-2-CF₃ displayed an adsorption energy of 24.8 kJ mol⁻¹ for SF₆, increasing by 4.2 kJ mol⁻¹ compared to BrCOF-2. Consequently, BrCOF-2-CF₃ exhibited a SF₆/N₂ selectivity of 50, increasing by 2.8 times that of original BrCOF-2 (Fig. 11c).

Fig. 11 (a) Representation of the synthesis of BrCOFs and the post-synthetic modification via Suzuki–Miyaura cross-coupling. (b) SF₆ and N₂ sorption isotherms at 273 K and 298 K of BrCOF-2 and BrCOF-2-CF₃. (c) Adsorption selectivity of the SF₆/N₂ (10 : 90) mixture in BrCOF-2 and BrCOF-2-CF₃. Reproduced with permission from ref. 100. Copyright 2021 Wiley-VCH.

Later, Xi's group designed COF-300 consisting of rigid C=N double bonds to capture SF₆ in humid air.¹⁰⁵ The covalent bonds composed of light elements (C and N) were less susceptible to be attacked by water molecules because of their short bond length and strong bond energy. Stability experiment results showed that there were no observable changes in the XRD pattern and specific surface area of COF-300 after being placed in humid air for several weeks, indicating the high water stability of COF-300. Simultaneously, the tetrahedral phenyl linkages in COF-300 created a hydrophobic environment, enabling COF-300 to reach a high SF₆ uptake of up to 4.03 mmol g⁻¹.

3.3. Porous aromatic frameworks

Porous aromatic frameworks (PAFs) are a type of porous organic framework composed of exclusively aromatic rigid linkers, which would adsorb SF₆ mainly through enriched F–π interactions.¹⁰⁸ In contrast to COF materials prepared through reversible polymerizations, PAFs are typically synthesized by irreversible cross-coupling of rigid building blocks connected by C–C bonds and are well known for their top-notch stability.¹⁰⁹ These bonds made them structurally stable in the presence of acidic environments and thus showed excellent acid resistance. These properties gained wider applications for SF₆ adsorption compared to MOFs and COFs. Table 2 provides a summary of the performance of COFs and PAFs in SF₆ adsorption.

Table 2 The SF₆ adsorption characteristics of covalent organic frameworks and porous aromatic frameworks

Sample name	BET surface area (m^2 g^−1)	Pore size (Å)	SF6 uptake at 1.0 bar (mmol g^−1)	SF6/N2 selectivity (1 : 9)	SF6 isosteric heat of adsorption, Qst (kJ mol^−1)	Ref.
PAF-4F	1072	9.7	2.46	145.6	30.4	68
PAF-8F	818	10.1	1.90	365.2	30
COF-6	750	9	3.71	850		93
RCOF-1	1139	9	4.13	125	27.0	98
RCOF-2	763	11	2.41	70	25.1
RCOF-3	242	13	0.91	34	21.4
BrCOF-2-CF3	1514	13	2.61	50	24.8	100
New-PAF-1	3957	12.0	2.01	28.5	22.99	103
N–SO3H	1334	9.0	4.66	74.1	29.94
PAF-XJTU-1	2025	10.1	2.32	47.18	24.04	104
PAF-XJTU-2	1924	9.2	2.43	46.64	33.64
PAF-XJTU-3	1499	8.5	2.56	36.76	31.03
PAF-XJTU-4	1419	10.1	1.47	29.39	25.08
PPN0	1480	12.0	0.48	41.6	22.3	105
PPN1	1392	12.0	0.60	51	28.1
PPN2	1081	12.0	0.52	45	24.2

---

Ma's group constructed an HF resistant PAF (New-PAF-1) with a diamond-like structure to capture SF₆ in semiconductor etching exhaust containing the corrosive by-product HF.¹¹⁰ The diamond-like structure of New-PAF-1 allowed sufficient exposure of phenyl rings, enabling the formation of strong electrostatic interaction with SF₆ molecules. Simultaneously, the robust tetrahedral C–C linkage of rigid phenyl rings imparted New-PAF-1 with exceptional acidic stability, rendering it resistant to corrosion from acidic HF. Experimental results indicated that there was no significant change in the structure of New-PAF-1 after soaking in HF solution for 24 h, and its SF₆ uptake reached 2.01 mmol g⁻¹ at 1 bar.

Most PAFs were reported to be typical microporous structures due to their short distance linkage of monophenyl ligands, which was not conducive to SF₆ diffusion in PAFs. Wu et al. proposed the design of a series of PAFs (PAF-XJTU-1 to PAF-XJTU-4) with hierarchical pore structures using longer polyaromatic blocks (Fig. 12a).¹¹¹ As expected, these PAF-XJTUs exhibited micropores concentrated at about 0.9 nm and mesopores at around 2.1 nm, maintaining high affinity and fast diffusion for SF₆ adsorption. Furthermore, the F–π interaction between SF₆ and benzene rings (green area) facilitated the F–F interaction among SF₆ molecules (red area) with the increase of SF₆ loading (Fig. 12b) and enhanced the affinity of PAF-XJTU for SF₆ (Fig. 12c).

Fig. 12 (a) Pore size distribution of PAF-XJTU-1-4. (b) The optimized adsorption structures of different SF₆ loadings in the pore models with pore sizes from 0.82 to 2.15 nm, with the visualization of host–guest and guest–guest interaction energies. (c) Isosteric enthalpy of SF₆ and N₂. Reproduced with permission from ref. 104. Copyright 2023 Elsevier.

Another limitation of using PAFs as SF₆ adsorbents is their inferior SF₆/N₂ selectivity compared to MOFs and COFs. Therefore, researchers adopted a polarized modification strategy to enhance the interaction between SF₆ and PAFs, thereby improving their SF₆ affinity and SF₆/N₂ selectivity. For instance, Chuah et al. designed a series of amine-incorporated PAFs (PPNx) using triphenylamine (TPA) struts and investigated the effect of amine content on SF₆/N₂ selectivity (Fig. 13a).¹¹² The presence of tertiary amines in TPA caused dipole-induced dipole interactions between PPNx and SF₆, leading to increases in heat of adsorption of SF₆ and SF₆/N₂ selectivity that were 1.25 times greater for both PPN1 and PPN2 compared to that of original PPN0 (Fig. 13b and c). However, excessive doping of tertiary amines reduced the specific surface area and micropore volume of PAFs (946–717 m² g⁻¹, 0.442–0.336 cm³ g⁻¹), resulting in a decreased adsorption performance of PPN2 than PPN1.

Fig. 13 (a) Reaction scheme of PPNx. (b) Isosteric heat of adsorption of PPNx copolymers as a function of SF₆ loading. (c) IAST SF₆/N₂ selectivity of PPNx copolymers as a function of pressure at 25 °C. Reproduced with permission from ref. 105. Copyright 2018 Elsevier.

The introduction of polar functional groups can alter the charge distribution of PAFs to a certain extent to enhance their electrostatic interaction towards SF₆. For instance, Zhang et al. proposed fluorinating PAFs (PAF-4F and PAF-8F) by using fluorinated building units and successfully constructed the induced electric field gradients.⁷⁴ The entry of strongly electronegative F atoms caused localized charge separation on the pore surface of PAFs. This made benzene rings become electropositive sites (red) (Fig. 14a) and formed an interaction angle with adjacent benzene rings, which efficiently intensified the adsorption affinity towards electronegative SF₆ (blue) (Fig. 14b). PAF-8F displayed a SF₆/N₂ selectivity of up to 365.22 and established a new benchmark for PAFs, which was attributed to its strong induced electric field gradient caused by high fluorine content (Fig. 14c).

Fig. 14 (a) Simulated models of PAF-4F@gas and PAF-8F@gas. (b) The electrostatic potential maps of PAF-4F@gas and PAF-8F@gas. (c) The IAST selectivity of PAF-4F and PAF-8F to the SF₆/N₂ gas mixture at 1 atm. Reproduced with permission from ref. 68. Copyright 2022 American Chemical Society.

Moreover, Wang et al. also developed a sulfonate-grafted PAF (N–SO₃H) to achieve specific capturing of SF₆.¹¹⁰ Sulfonate-grafted N–SO₃H exhibited significantly increased charge density difference from SF₆ compared with the original New-PAF-1, as illustrated in the blue and green areas in Fig. 15a. DFT calculations revealed that the total atomic charge of SF₆ changed from neutral to negative after binding in N–SO₃H, suggesting that part of the electrons of the framework were transferred to SF₆via the –SO₃H group and formed a bonding state. By comparison, the original New-PAF-1 adsorbed SF₆ only through weak interactions dominated by van der Waals forces, while the modified N–SO₃H adsorbed SF₆ through charge transfer effects in a synergistic manner, as evidenced by the increase in the heat of adsorption of SF₆ from 22.99 kJ mol⁻¹ to 29.94 kJ mol⁻¹. As a result, N–SO₃H displayed a remarkable SF₆/N₂(O₂) selectivity of up to 250 in a ternary gas mixture (N₂ : O₂ = 79% : 21%), which was 3.4 times greater than that of the original New-PAF-1 (Fig. 15c).

Fig. 15 (a) Schematic representation of New-PAF-1 and N–SO₃H. (b) Charge density difference graph of SF₆ absorbed on New-PAF-1 and on N–SO₃H. (c) IAST selectivity of SF₆ in a ternary mixture for New-PAF-1 and N–SO₃H at 298 K and 100 kPa. Reproduced with permission from ref. 103. Copyright 2022 American Chemical Society.

4. Conclusion and outlook

Although PFMs have been proven to be promising SF₆ adsorbents, they are still facing significant limitations under practical conditions, hindering their widespread application. Addressing these challenges, we propose the following possible research directions to enhance their SF₆ adsorption performance in industrial settings:

(1) AI-driven creation of high-performance PFMs: currently, the design of PFMs heavily relies on experimental data and computational work, which is not feasible to evaluate the adsorption performance of PFMs using purely experimental methods. Leveraging fast deep learning of artificial intelligence (AI), it is easy to systematically explore the impact of metal species, molecular characteristics of ligand groups and coordination modes on the surface morphology, crystal and pore structures of PFMs, helping to develop materials with a high specific surface area. Furthermore, machine learning algorithms and AI reasoning can be employed to study the design and construction methods of various adsorption sites on PFMs to predict the structure of PFMs with high selectivity and adsorption capacity.

(2) Construction of PFMs with high water- and acid-resistance: SF₆ often coexists with acidic by-products like HF and water vapor in industrial waste gases, which would lead to the collapse and destruction of the framework structure during the adsorption process, resulting in a decline in adsorption performance. Strategies including surface modification of PFMs can be used to enhance the stability of the material in high-humidity environments, such as introducing hydrophobic groups or encapsulating the material in a hydrophobic coating. Utilizing high-valence metal nodes or constructing diamond topological structures can bolster the structural stability of PFMs against an acidic atmosphere.¹¹³ Furthermore, researchers should investigate the influence of humidity and acidic gases on SF₆ adsorption, optimizing adsorption sites on PFMs through adjusting the hydrophobicity or introducing specific functional groups to enhance the specificity and avoid competitive adsorption.

(3) Shaping of high SF₆ adsorption capacity PFMs: traditional shaping methods often lead to pore blockage or structural collapse of PFMs, diminishing their SF₆ adsorption performance.¹¹⁴ Shaping by in situ growth such as using a gelation freeze-drying method to obtain a PFM gel or in situ crystallization on a substrate to obtain a PFM film can preserve the specific surface area and avoid the loss of SF₆ adsorption capacity of PFMs.¹¹⁵ In addition, monomicelle self-assembly or growth of PFMs on custom-shaped porous scaffolds can create orderly mesoporous and macroporous channels to inhibit pore blockage and further enhance the SF₆ mass transfer speed, improving the adsorption/desorption efficiency of PFMs.

(4) Construction of PFMs for SF₆ capture via kinetic adsorption: kinetic adsorption driven by the inherent differences between SF₆ and N₂ molecules in diffusion dynamics is believed to improve the energy efficiency of the separation process, which does not involve strong host–guest interactions and is conducive to the regeneration of adsorbents.¹¹⁶ The diffusion rate of gas molecules mainly depends on the structural characteristics of PFMs and thus could be regulated by adjusting the crystal size and growth orientation. For example, surface-coordinated MOF thin films with preferred growth orientation can be grown by the liquid-phase epitaxial method, enhancing the intragranular diffusion of SF₆ to promote kinetic adsorption.¹¹⁷

(5) Green and scale-up synthesis of economically feasible PFMs: in addition to SF₆ adsorption performance, the industrial application of PFMs also needs to consider factors such as economic feasibility, scale-up and green synthesis.^118,119 However, most of the reported synthesis methods of PFMs involve expensive organic linkers and toxic organic solvents (such as N,N-dimethylformamide) and cannot be synthesized on a large scale, making PFMs unsuitable for large-scale applications.^120,121 Choosing mechano-synthesis or using supercritical fluids and ionic liquids as solvents can reduce or eliminate the use of organic solvents, achieving cost reduction and a green synthesis process.¹²² Besides, the use of continuous flow chemistry can improve the scalability of PFMs, which is attractive in industrial processes considering the efficiency and consistency.¹²³

Porous framework materials (PFMs) with adjustable pores and abundant adsorption sites have been used as ideal materials for capturing SF₆, which can efficiently adsorb SF₆ through Lewis acid–base interactions, F–π interactions and induced dipole moment effects. This review summarized the latest progress of PFMs for SF₆ adsorption based on the adsorption mechanism and discussed various modification strategies to improve their adsorption performance through pore engineering. Finally, possible research directions for the future development of PFMs were proposed, hoping to provide assistance for the development of SF₆ adsorbents based on PFMs.

Data availability

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Author contributions

Xiaoxuan Sun: writing – original draft, investigation. Liqin Zhou: writing – review & editing, investigation, conceptualization. Jianmin Chen: writing – review & editing, supervision, conceptualization. Zhaowei Jia: writing – review & editing, methodology. Zhongxing Zhao: writing – review & editing, supervision, funding acquisition. Zhenxia Zhao: writing – review & editing, conceptualization, funding acquisition.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. U23A20120), Natural Science Foundation of Guangxi Province (No. AD24010026), and National Natural Science of China (No. 22178073, 22268004 and 31401629). We thank Jiabao Wang from Shiyanjia Lab (https://www.shiyanjia.com/) and Jiaqi Li from SCI-GO (https://www.sci-go.com/) for providing the instrument for experimental testing.

References

1. W. Zhang, Y. Wu, Y. Li, S. Chen, Y. Fu, Z. Zhang, T. Yan, S. Wang and H. Ma, Macromolecules, 2022, 55, 1435–1444.
2. L. Ye, Y. Jin, K. Wang, W. Chen, F. Wang and B. Dai, Appl. Energy, 2023, 351, 121866.
3. S. D'Auria, A. M. Pourrahimi, A. Favero, P. Neuteboom, X. Xu, S. Haraguchi, M. Bek, R. Kádár, E. Dalcanale, R. Pinalli, C. Müller and J. Vachon, Adv. Funct. Mater., 2023, 33, 2301878.
4. Y. Yang, Z. M. Dang, Q. Li and J. He, Adv. Sci., 2020, 7, 202002131.
5. M. Rabie and C. M. Franck, Environ. Sci. Technol., 2018, 52, 369–380.
6. E. Guney and O. Ozgonenel, Energies, 2021, 14, 3698.
7. T. Magier, M. Tenzer and H. Koch, IEEE Trans. Power Delivery, 2018, 33, 440–446.
8. Q. Han, L. Chen, I. Cotton and J. Khan, IEEE Trans. Transp. Electrif., 2023, 9, 3590–3600.
9. Z. Cui, Y. Li, S. Xiao, S. Tian, J. Tang, Y. Hao and X. Zhang, Sci. Total Environ., 2024, 906, 107347.
10. X. Fang, X. Hu, G. Janssens-Maenhout, J. Wu, J. Han, S. Su, J. Zhang and J. Hu, Environ. Sci. Technol., 2013, 47, 3848–3855.
11. D. Arbutina, K. Stanković, L. Perazić and M. Pejović, Int. J. Electr. Power Energy Syst., 2019, 104, 436–442.
12. X. Yao, J. Wang, S. Ai, Z. Liu, Y. Geng and Z. Hao, Engineering, 2022, 13, 164–177.
13. H. Zhang, Y. Jung Oh, E. Tsegay Tikue, H. Il Lee, S. Kyung Kang and P. Soo Lee, Sep. Purif. Technol., 2022, 303, 122223.
14. J. Liu, G. M. Huang, Z. Ma and Y. Geng, IEEE Trans. Smart Grid, 2011, 2, 254–264.
15. A. Tena-Colunga, L. Eduardo Pérez-Rocha, J. Avilés and C. Cordero-Macías, J. Build. Eng., 2015, 4, 21–40.
16. P. Wang, W. Chen, J. Wang, J. Tang, Y. Shi and F. Wan, Anal. Chem., 2020, 92, 5969–5977.
17. H. M. Lee, M. B. Chang and K. Y. Wu, J. Air Waste Manage. Assoc., 2012, 54, 960–970.
18. R. S. Pessoa, H. S. Maciel, G. Petraconi, M. Massi and A. S. da Silva Sobrinho, Appl. Surf. Sci., 2008, 255, 749–751.
19. M. P. Sulbaek Andersen, M. Kyte, S. T. Andersen, C. J. Nielsen and O. J. Nielsen, Environ. Sci. Technol., 2017, 51, 1321–1329.
20. R. Bai, X. Song, W. Yan and J. Yu, Natl. Sci. Rev., 2022, 9, nwac064.
21. S. A. Montzka, E. J. Dlugokencky and J. H. Butler, Nature, 2011, 476, 43–50.
22. T. C. Rolfe and A. L. Rice, Nature, 2020, 583, 171.
23. S. Xiao, X. Zhang, J. Tang and S. Liu, Energy Rep., 2018, 4, 486–496.
24. Y. Dai, Q. Li, X. Ruan, Y. Hou, X. Jiang, X. Yan, G. He, F. Meng and Z. Wang, J. Membr. Sci., 2019, 577, 258–265.
25. D.-D. Zhou, X.-W. Zhang, Z.-W. Mo, Y.-Z. Xu, X.-Y. Tian, Y. Li, X.-M. Chen and J.-P. Zhang, EnergyChem, 2019, 1, 100016.
26. C. Y. Chuah, Y. Lee and T.-H. Bae, Chem. Eng. J., 2021, 404, 126577.
27. C. Y. Chuah, S. Yu, K. Na and T.-H. Bae, J. Ind. Eng. Chem., 2018, 62, 64–71.
28. W. Bo, S. Ren, L. You, X. Meng, Y. Ji, X. Peng, M. Zhu, C. Zhang, X. Wang and X. Gu, J. Membr. Sci., 2024, 690, 122200.
29. Y. J. Oh, S. K. Kang, A. H. Lee, S. Park, S. Kim, J. Choi and P. S. Lee, Microporous Mesoporous Mater., 2023, 356, 112590.
30. S. Kitagawa, R. Kitaura and S. i. Noro, Angew. Chem., Int. Ed., 2004, 43, 2334–2375.
31. I. A. Riddell, M. M. J. Smulders, J. K. Clegg and J. R. Nitschke, Chem. Commun., 2011, 47, 457–459.
32. K. Tian, S. M. Elbert, X. Y. Hu, T. Kirschbaum, W. S. Zhang, F. Rominger, R. R. Schröder and M. Mastalerz, Adv. Mater., 2022, 34, 2202290.
33. X. Peng, J. M. Vicent-Luna and Q. Jin, ACS Appl. Mater. Interfaces, 2020, 12, 20044–20055.
34. X. Huang, F. Chen, H. Sun, W. Xia, Z. Zhang, Q. Yang, Y. Yang, Q. Ren and Z. Bao, Sep. Purif. Technol., 2022, 292, 121059.
35. Y. Yin, Y. Zhang, X. Zhou, B. Gui, G. Cai, J. Sun and C. Wang, J. Am. Chem. Soc., 2023, 145, 22329–22334.
36. Y. He, X. Cao, Z. Zhang, Z. Jiang, H. Huang, S. Zhang, Q. Han and C. Zhong, Ind. Eng. Chem. Res., 2023, 62, 7642–7649.
37. L. Huang and D. Cao, J. Mater. Chem. A, 2013, 1, 9433.
38. P. Bhanja, K. Bhunia, S. K. Das, D. Pradhan, R. Kimura, Y. Hijikata, S. Irle and A. Bhaumik, ChemSusChem, 2017, 10, 921–929.
39. S. Chatterjee, P. Bhanja, D. Ghosh, P. Kumar, S. Kanti Das, S. Dalapati and A. Bhaumik, ChemSusChem, 2020, 14, 408–416.
40. S. Ruidas, B. Mohanty, P. Bhanja, E. S. Erakulan, R. Thapa, P. Das, A. Chowdhury, S. K. Mandal, B. K. Jena and A. Bhaumik, ChemSusChem, 2021, 14, 5057–5064.
41. T. Wang, M. Chang, T. Yan, Y. Ying, Q. Yang and D. Liu, Ind. Eng. Chem. Res., 2021, 60, 5976–5983.
42. M.-B. Kim, T.-U. Yoon, D.-Y. Hong, S.-Y. Kim, S.-J. Lee, S.-I. Kim, S.-K. Lee, J.-S. Chang and Y.-S. Bae, Chem. Eng. J., 2015, 276, 315–321.
43. M. I. Nandasiri, S. R. Jambovane, B. P. McGrail, H. T. Schaef and S. K. Nune, Coord. Chem. Rev., 2016, 311, 38–52.
44. J. Wang and S. Zhuang, Coord. Chem. Rev., 2019, 400, 213046.
45. Y. Yuan, Y. Yang and G. Zhu, EnergyChem, 2020, 2, 100037.
46. Y. Hu, L. Wang, R. Nan, N. Xu, Y. Jiang, D. Wang, T. Yan, D. Liu, Y. Zhang and B. Chen, Chem. Eng. J., 2023, 471, 145656.
47. J.-W. Yan, S.-Q. Gang, Z.-Y. Liu, H.-Y. Xu, R. Wang and J.-L. Du, Sep. Purif. Technol., 2023, 327, 24929.
48. S. Jeoung, S. Kim, M. Kim and H. R. Moon, Coord. Chem. Rev., 2020, 420, 213377.
49. Y. Yang, M. Faheem, L. Wang, Q. Meng, H. Sha, N. Yang, Y. Yuan and G. Zhu, ACS Cent. Sci., 2018, 4, 748–754.
50. J. Xie, Y. Liang, Q. Zou, Z. Wang and X. Li, Energy Fuels, 2019, 33, 1910–1921.
51. L.-Q. Yang, J. Yu, P. Zhang, Y. Wang, W.-Y. Yuan and Q.-G. Zhai, Sep. Purif. Technol., 2023, 327, 124930.
52. Y. Wu, D. Yuan, D. He, J. Xing, S. Zeng, S. Xu, Y. Xu and Z. Liu, Angew. Chem., Int. Ed., 2019, 58, 10241–10244.
53. G. Fraux, A. Boutin, A. H. Fuchs and F.-X. Coudert, Adsorption, 2018, 24, 233–241.
54. H.-Y. Ren and X.-M. Zhang, Energy Fuels, 2015, 30, 526–530.
55. R. Krishna and J. M. van Baten, J. Phys. Chem. C, 2012, 116, 23556–23568.
56. J. A. Mason, M. Veenstra and J. R. Long, Chem. Sci., 2014, 5, 32–51.
57. Y. Tian and J. Wu, Langmuir, 2017, 33, 996–1003.
58. A. I. Skoulidas and D. S. Sholl, J. Phys. Chem. B, 2002, 106, 5058–5067.
59. J.-R. Li, R. J. Kuppler and H.-C. Zhou, Chem. Soc. Rev., 2009, 38, 1477–1504.
60. J. Wu, X. Zhu, F. Yang, T. Ge and R. Wang, Chem. Eng. J., 2021, 425, 131409.
61. Ü. Kökçam-Demir, A. Goldman, L. Esrafili, M. Gharib, A. Morsali, O. Weingart and C. Janiak, Chem. Soc. Rev., 2020, 49, 2751–2798.
62. F. Yi, Y. Chen, Z. Tao, C. Hu, X. Yi, A. Zheng, X. Wen, Y. Yun, Y. Yang and Y. Li, J. Catal., 2019, 380, 204–214.
63. F. Zhang, Y. Zhang, L. Ma, J. Yang and J. Li, ACS Sustain. Chem. Eng., 2022, 10, 9359–9368.
64. P. Chowdhury, C. Bikkina, D. Meister, F. Dreisbach and S. Gumma, Microporous Mesoporous Mater., 2009, 117, 406–413.
65. I. Senkovska, E. Barea, J. A. R. Navarro and S. Kaskel, Microporous Mesoporous Mater., 2012, 156, 115–120.
66. S. J. Grabowski, Molecules, 2020, 25(20), 4668.
67. M.-B. Kim, K.-M. Kim, T.-H. Kim, T.-U. Yoon, E.-J. Kim, J.-H. Kim and Y.-S. Bae, Chem. Eng. J., 2018, 339, 223–229.
68. N. Liu, X. Lv, B. Xiao, D. Kuzuhara, P. Mei, N. Aratani, H. Yamada, F. Qiu, J. Pan and S. Xue, Dalton Trans., 2022, 51, 9606–9610.
69. E. Blum, J. Zhang, J. Zaluski, D. E. Einstein, E. E. Korshin, A. Kubas, A. Gruzman, G. P. Tochtrop, P. D. Kiser and K. Palczewski, J. Med. Chem., 2021, 64, 8287–8302.
70. P. Li, J. M. Maier, E. C. Vik, C. J. Yehl, B. E. Dial, A. E. Rickher, M. D. Smith, P. J. Pellechia and K. D. Shimizu, Angew. Chem., Int. Ed., 2017, 56, 7209–7212.
71. L. Guo, M. Tian, Z. Li, Q. Wang, Q. Wu, L. Hao and C. Wang, Food Chem., 2023, 427, 136674.
72. Y. Wu, S. Wang, W. Zhang, S. Chen, Z. Zhang, B. Yang, S. Li and H. Ma, Sep. Purif. Technol., 2022, 289, 120739.
73. M. Wan, H. Yue, J. Notarangelo, H. Liu and F. Che, JACS Au, 2022, 2, 1338–1349.
74. W. Zhang, Y. Li, S. Wang, Y. Wu, S. Chen, Y. Fu, W. Ma, Z. Zhang and H. Ma, ACS Appl. Mater. Interfaces, 2022, 14, 35126–35137.
75. I. Skarmoutsos, M. Eddaoudi and G. Maurin, Microporous Mesoporous Mater., 2019, 281, 44–49.
76. C. Y. Chuah, K. Goh and T.-H. Bae, J. Phys. Chem. C, 2017, 121, 6748–6755.
77. A. Kirchon, L. Feng, H. F. Drake, E. A. Joseph and H.-C. Zhou, Chem. Soc. Rev., 2018, 47, 8611–8638.
78. Z.-J. Lin, J. Lü, M. Hong and R. Cao, Chem. Soc. Rev., 2014, 43, 5867–5895.
79. F. P. Kinik, A. Ortega-Guerrero, D. Ongari, C. P. Ireland and B. Smit, Chem. Soc. Rev., 2021, 50, 3143–3177.
80. G. Wu, J. Ma, S. Li, J. Li, X. Wang, Z. Zhang and L. Chen, J. Mater. Chem. A, 2023, 11, 6747–6771.
81. J. Li, P. M. Bhatt, J. Li, M. Eddaoudi and Y. Liu, Adv. Mater., 2020, 32, 2002563.
82. W. Fan, X. Zhang, Z. Kang, X. Liu and D. Sun, Coord. Chem. Rev., 2021, 443, 213968.
83. M. Åhlén, Y. Zhou, D. Hedbom, H. S. Cho, M. Strømme, O. Terasaki and O. Cheung, J. Mater. Chem. A, 2023, 11, 26435–26441.
84. Y.-P. Li, J.-J. Ni, S. Li, J. Wang, Z.-Y. Sui and X.-F. Xu, J. Solid State Chem., 2024, 329, 124554.
85. M.-B. Kim, S.-J. Lee, C. Y. Lee and Y.-S. Bae, Microporous Mesoporous Mater., 2014, 190, 356–361.
86. M. Åhlén, F. M. Amombo Noa, L. Öhrström, D. Hedbom, M. Strømme and O. Cheung, Microporous Mesoporous Mater., 2022, 343, 112161.
87. S. M. Wang, X. T. Mu, H. R. Liu, S. T. Zheng and Q. Y. Yang, Angew. Chem., Int. Ed., 2022, 61, e202207066.
88. M. Åhlén, A. Jaworski, M. Strømme and O. Cheung, Chem. Eng. J., 2021, 422, 130117.
89. X. Liang, P. Wang, C. Li, M. Yuan, Q. Shi and J. Dong, Microporous Mesoporous Mater., 2021, 320, 111109.
90. S.-T. Zheng, R.-Y. Jiang, Y. Jiang, S. Ni, G.-W. Guan, S.-Q. Shao, Y.-C. Wang, S.-M. Wang and Q.-Y. Yang, Sep. Purif. Technol., 2023, 318, 123957.
91. T. H. Chen, I. Popov, W. Kaveevivitchai, Y. C. Chuang, Y. S. Chen, A. J. Jacobson and O. Š. Miljanić, Angew. Chem., Int. Ed., 2015, 54, 13902–13906.
92. M. Yang, M. Chang, T. Yan and D. Liu, Sep. Purif. Technol., 2022, 295, 121340.
93. M. Chang, T. Yan, Y. Wei, J.-X. Wang, D. Liu and J.-F. Chen, Chem. Mater., 2022, 34, 9134–9143.
94. J. Ren, M. Chang, W. Zeng, Y. Xia, D. Liu, G. Maurin and Q. Yang, Chem. Mater., 2021, 33, 5108–5114.
95. J. H. Lee, S. Jeoung, Y. G. Chung and H. R. Moon, Coord. Chem. Rev., 2019, 389, 161–188.
96. P. Zhao, S. C. E. Tsang and D. Fairen-Jimenez, Cell Rep. Phys. Sci., 2021, 2, 100544.
97. X. Zhang, Y.-L. Zhao, X.-Y. Li, X. Bai, Q. Chen and J.-R. Li, J. Am. Chem. Soc., 2024, 146, 19303–19309.
98. J. Choi, J. Im, K. Noh, J. Kim, T. Vogt and Y. Lee, J. Phys. Chem. C, 2019, 123, 25769–25774.
99. A. Schneemann, P. Vervoorts, I. Hante, M. Tu, S. Wannapaiboon, C. Sternemann, M. Paulus, D. C. F. Wieland, S. Henke and R. A. Fischer, Chem. Mater., 2018, 30, 1667–1676.
100. X. Zheng, Y. Shen, S. Wang, K. Huang and D. Cao, Chin. J. Chem. Eng., 2021, 39, 88–95.
101. M. Åhlén, E. Kapaca, D. Hedbom, T. Willhammar, M. Strømme and O. Cheung, Microporous Mesoporous Mater., 2022, 329, 111548.
102. U. Díaz and A. Corma, Coord. Chem. Rev., 2016, 311, 85–124.
103. T. Li, Y. Pan, B. Shao, X. Zhang, T. Wu, Q. He, M. He, L. Ge, L. Zhou, S. Liu, X. Zheng, J. Ye and Z. Liu, Adv. Funct. Mater., 2023, 33, 2304990.
104. J. Yang, F. Kang, X. Wang and Q. Zhang, Mater. Horizons, 2022, 9, 121–146.
105. Q. Liao, H. Xu, C. Ke, Y. Zhang, Q. Han, Y. Zhang, Y. Xu, D. Wang and K. Xi, Sep. Purif. Technol., 2023, 306, 122595.
106. I. Ahmed and S. H. Jhung, Coord. Chem. Rev., 2021, 441, 213989.
107. Q. Liao, C. Ke, X. Huang, D. Wang, Q. Han, Y. Zhang, Y. Zhang and K. Xi, Angew. Chem., Int. Ed., 2020, 60, 1411–1416.
108. Y. Yuan and G. Zhu, ACS Cent. Sci., 2019, 5, 409–418.
109. Y. Yuan, Y. Yang and G. Zhu, ACS Cent. Sci., 2020, 6, 1082–1094.
110. S. Wang, Y. Wu, Y. Zhang, Z. Zhang, W. Zhang, X. Li, W. Ma and H. Ma, Langmuir, 2022, 38, 8667–8676.
111. Y. Wu, X. Li, Y. Li, C. Tao, Y. Tang, S. Wang, Y. Fu, W. Ma, W. Zhang and H. Ma, Sep. Purif. Technol., 2023, 315, 123657.
112. C. Y. Chuah, Y. Yang and T.-H. Bae, Microporous Mesoporous Mater., 2018, 272, 232–240.
113. R. Liu, Q.-Y. Liu, R. Krishna, W. Wang, C.-T. He and Y.-L. Wang, Inorg. Chem., 2019, 58, 5089–5095.
114. J. Zheng, X. Cui, Q. Yang, Q. Ren, Y. Yang and H. Xing, Chem. Eng. J., 2018, 354, 1075–1082.
115. Q. Ma, T. Zhang and B. Wang, Matter, 2022, 5, 1070–1091.
116. Q. Ding, Z. Zhang, P. Zhang, J. Wang, X. Cui, C.-H. He, S. Deng and H. Xing, Chem. Eng. J., 2022, 434, 134784.
117. Y.-H. Xiao, Y.-B. Tian, Z.-G. Gu and J. Zhang, EnergyChem, 2021, 3, 100065.
118. A. M. Wright, M. T. Kapelewski, S. Marx, O. K. Farha and W. Morris, Nat. Mater., 2024 DOI:10.1038/s41563-024-01947-4.
119. E. Wu, X.-W. Gu, D. Liu, X. Zhang, H. Wu, W. Zhou, G. Qian and B. Li, Nat. Commun., 2023, 14, 6146.
120. H. M. Wen, C. Yu, M. Liu, C. Lin, B. Zhao, H. Wu, W. Zhou, B. Chen and J. Hu, Angew. Chem., Int. Ed., 2023, 62, e202309108.
121. D. Liu, J. Pei, X. Zhang, X. W. Gu, H. M. Wen, B. Chen, G. Qian and B. Li, Angew. Chem., Int. Ed., 2023, 62, e202218590.
122. S. Głowniak, B. Szczęśniak, J. Choma and M. Jaroniec, Mater. Today, 2021, 46, 109–124.
123. S. Bose, D. Sengupta, T. M. Rayder, X. Wang, K. O. Kirlikovali, A. K. Sekizkardes, T. Islamoglu and O. K. Farha, Adv. Funct. Mater., 2023, 2307478.

---