Metal–organic frameworks for hydrocarbon separation: design, progress, and challenges

Abstract

High-purity hydrocarbons are of great importance in various fields, including the electronics and chemical industries; however, the separation of hydrocarbons can be difficult because of their structural and chemical similarities. The established industrial practices of separating hydrocarbons usually involve energy-intensive procedures such as low-temperature distillation. Adsorptive separation using porous materials is deemed a promising alternative technology due to its potential to significantly reduce energy consumption. Porous materials, MOFs in particular, offer versatility in terms of their synthesis and structural design that can be facilely customized to meet the requirements of specific applications. By tailoring pore dimensions, surface functionality, and framework flexibility, MOF materials exhibit high selectivity for specific molecules in the separation and purification of hydrocarbons. With the rapid advances in reticular chemistry and crystal engineering, researchers have gained deep insight into the directional stitching of diverse building blocks, paving the way for the design of tailor-made MOF materials. This perspective summarizes three main strategies for developing MOFs for hydrocarbon separation: surface engineering, molecular docking, and size exclusion. In addition, the prospects of MOF materials for hydrocarbon separation are discussed, including future directions of addressing the challenges (barriers) for MOF sorbents to be competitive in practical applications, such as adsorption capacity and selectivity, stability, regenerability, scalability and cost-effectiveness, and industrial implementation.

---

Xiao-Jing Xie

Xiao-Jing Xie joined Prof. Dan Li's group at Jinan University as a PhD student in 2021. She completed her B.S. degree at the College of Chemistry and Materials Science, Jinan University, in 2020. Her current research interests are the design and application of metal–organic frameworks for hydrocarbon separation.

Heng Zeng

Heng Zeng received his PhD degree in chemistry from Jinan University in 2022. His research work is focused on the development of multifunctional metal–organic frameworks and hydrogen-bonding organic frameworks for their applications in gas storage, separation, and sensors.

Weigang Lu

Weigang Lu obtained his PhD in organic chemistry (2002) and stayed as a lecturer at Sun Yat-Sen University (2002–2005). After a visiting scholar stint at the Hong Kong University of Science & Technology (2005–2008), he joined Texas A&M University as an assistant research scientist in Prof. Hong-Cai Zhou's group (2008–2015). He started independent research as a research specialist at Fayetteville State University (2016–2018) and now he is a professor of chemistry at Jinan University (2018–Present). His research interest focuses on the rational design and synthesis of porous materials for energy-related applications.

Dan Li

Dan Li received his BSc from Sun Yat-Sen University in 1984 and then worked at Shantou University. He pursued his PhD at The University of Hong Kong with Professor Chi-Ming Che during 1988–1993. Then he returned to Shantou University and became a Professor in 2001. He moved to Jinan University in Guangzhou in 2016. He is a recipient of the National Science Fund for Distinguished Young Scholars of China in 2008 and a Fellow of The Royal Society of Chemistry (FRSC) in 2014. His research interest is the design and fabrication of supramolecular coordination assemblies and their functions including photoluminescence, porosity and chirality. He has been serving as an Associate Editor of J. Mater. Chem. A and Mater. Adv. since mid-2022.

---

1. Introduction

Hydrocarbons such as ethylene (C₂H₄), propylene (C₃H₆), and xylenes are the most important feedstock materials in the petrochemical industry,¹ with an estimated value exceeding USD 307.58 billion by 2029.^2,3 High-purity hydrocarbons are essential for the downstream production of high-value-added chemical products;⁴ yet current hydrocarbon separation technology is extremely energy-intensive with high energy penalties.⁵ For example, cryogenic distillation coupled with high plate numbers and reflux ratios is often required in industrial separation processes to obtain high-purity hydrocarbons, resulting in associated equipment investment and energy consumption of up to 40–90% of the total process cost.^6,7 Therefore, developing alternative separation technology for hydrocarbon purification with less energy consumption and environmental impact is of practical significance from academic research to industrial implementation.

Adsorption-based separation using porous materials has demonstrated great potential as an eco-friendly and energy-efficient technology, accounting for only 1/3 or less of the energy consumption of cryogenic distillation.^8,9 Meanwhile, adsorption-based separation has also shown its capability of enabling high separation selectivity, allowing for the better implementation of challenging separations. Traditional porous solids (e.g. zeolites¹⁰ and carbon materials¹¹) have been explored as adsorbents in the separation of hydrocarbons, for example, the removal of linear alkanes from their branched isomers with zeolite 5A.¹² However, because of their limited structural diversity and designability, zeolites appear to be inadequate for many other industrially demanding hydrocarbon separations.¹³ To reduce the energy footprint of hydrocarbon separations, scientists and engineers in academia and industry are facing the opportunities and challenges of developing efficient adsorbents to overcome the trade-off between the adsorption capacity and separation coefficient.^14,15

Metal–organic frameworks (MOFs),^16,17 also known as porous coordination polymers (PCPs), are a class of porous crystalline materials assembled from metal nodes/metal clusters (a.k.a. secondary building units, SBUs) and organic linkers. With the rapid advances in reticular chemistry and crystal engineering, researchers have gained deep insight into the directional stitching of diverse building blocks, paving the way for the design and construction of tailor-made MOF materials. Compared to traditional porous solids, MOF materials are known for their exceptional porosity (specific surface area up to 7000 m² g⁻¹) and diverse pore structures, which can be beneficial for sensing,¹⁸ catalysis,¹⁹ biomedicine,²⁰ separation,²¹ and other relevant applications.²² Particularly, MOF materials have shown enormous potential in the separation and purification of hydrocarbons owing to their ultrahigh surface area, tunable pore environment, and structural designability. One of the most notable features of MOFs is post-synthetic modification, which allows for the introduction of additional functionalities or active sites into the MOF structures, further enhancing their performance. Another notable feature of MOFs is their well-defined pore structures, which enables a better understanding in terms of structure–property relationships, and therefore, provides directions to further design MOF materials for higher adsorption selectivity. In addition, framework flexibility was observed in over 200 reported MOF materials, leading to many interesting and unexpected properties contingent upon the applied stimuli (e.g., guest, temperature, pressure, light, electricity, etc.).^23–25

Currently, the design and construction of MOF materials for hydrocarbon separation can be roughly categorized into the following three main strategies: surface engineering, molecular docking, and size exclusion (Fig. 1). Surface engineering is to install certain functional groups on the internal surface of MOF materials to distinguish gas molecules. For example, MOF materials with open metal sites tend to have a stronger binding affinity for olefin than for paraffin. Molecular docking is to build pockets/traps with matched dimensions and electrostatic potentials to preferentially accommodate specific guest molecules, not necessarily the smaller ones. Size exclusion is considered an ideal strategy for hydrocarbon separation, in which smaller molecules are adsorbed and larger ones are completely excluded, offering the highest separation coefficient among the three strategies. This perspective reviews the recent undertakings of the above three strategies in the development of MOF materials for hydrocarbon separation, focusing on their structure–property relationships. In addition, we discuss the prospects of MOF materials for hydrocarbon separation and the existing barriers to moving from academic research to industrial implementation.

Fig. 1 Schematic diagram of the three main strategies for developing MOF materials for hydrocarbon separation. (Top) Reproduced from ref. 34 with permission from copyright 2018, The American Association for the Advancement of Science. (Bottom-left) Reproduced from ref. 46 with permission from copyright 2015, Springer Nature. (Bottom-right) Reproduced from ref. 59 with permission from copyright 2016, The American Association for the Advancement of Science.

2. Design strategies

2.1. Surface engineering

Surface engineering represents an intuitive and effective strategy in the study of MOF materials for hydrocarbon separation. The introduction of functional groups on the internal surface may endow MOF materials with polarized surfaces or electric field gradients to boost adsorbate–adsorbent interaction. Extensive research efforts have been devoted to rational surface engineering by the systematic installation of functional groups via reticular synthesis.

MOF materials with coordinatively unsaturated metal sites (open metal sites) can preferentially adsorb alkenes over alkanes through metal–π interaction and thus have attracted substantial attention for their use in olefin/paraffin separation. As early as 2002, Bülow Martin et al. first reported the separation of C₂H₄/C₂H₆ mixtures with Cu₃(BTC)₂ (HKUST-1), a landmark MOF material with Cu^II open metal sites on the axial positions of copper paddlewheel clusters.²⁶ Detailed theoretical calculations revealed that C₂H₄ is indeed preferentially adsorbed on the Cu^II open metal site via the metal–π interaction.²⁷ In 2012, Jeffrey R. Long et al. examined an iron-based MOF, Fe₂(dobdc) (dobdc: 2,5-dioxido-1,4-benzenedicarboxylate), because of its ability to separate C₂H₄/C₂H₆ and C₃H₆/C₃H₈ at 318 K.²⁸ Fe₂(dobdc) exhibited a strong affinity towards C₂H₂, C₂H₄, and C₃H₆, as evidenced by their steep adsorption curves at low pressures in the adsorption isotherms of these unsaturated hydrocarbons. Neutron powder diffraction (NPD) studies directly revealed that both C₂H₄ and C₃H₆ are bound to the Fe^II open metal sites in side-on mode, with Fe–C distances in the range from 2.42(2) to 2.60(2) Å, while the interactions of C₂H₆ and C₃H₈ with the Fe^II open metal sites are much weaker, as evidenced by the longer Fe–C distances (∼3 Å) (Fig. 2a and b). Equilibrium adsorption experiments demonstrate that Fe₂(dobdc) exhibited large olefin adsorption capacities and high olefin/paraffin selectivities (up to 18 for C₂H₄/C₂H₆ and 15 for C₃H₆/C₃H₈) at 318 K. Breakthrough experiments confirmed that Fe₂(dobdc) is capable of separating an equimolar mixture of C₂H₄/C₂H₆ and C₃H₆/C₃H₈ at 1 bar and 318 K, affording olefins with >99% purity (Fig. 2c and d).

Fig. 2 (a) The structure of Fe₂(dobdc)·2C₂D₄ from neutron powder diffraction data, showing the strong metal–π interactions. (b) Views of the iron centers interacting with C₂D₄, C₂D₆, C₃D₆, and C₃D₈. Orange, red, gray, and blue spheres represent Fe, O, C, and D atoms, respectively. (c) and (d) Experimental breakthrough curves for equimolar C₂H₄/C₂H₆ and C₃H₆/C₃H₈ mixtures on Fe₂(dobdc). Reproduced from ref. 28 with permission from copyright 2012, The American Association for the Advancement of Science.

Since open metal sites can be used to reversibly bind olefin molecules via metal–π interaction, introducing high-density open metal sites is one of the most effective strategies for olefin/paraffin separation.^29–31 Most of the reported MOF materials can offer up to one accessible open metal site per metal center. In 2022, our group reported a microporous MOF (JNU-4) consisting of square-planar mononuclear Cu^II centers and tetrahedral organic linkers, allowing for two accessible binding sites per metal center for C₂H₂ molecules.³² The adsorption isotherm of JNU-4 exhibited a steep adsorption curve at low pressures and an overall adsorption capacity of 222 cm³ g⁻¹ (9.91 mmol g⁻¹) at 298 K and 1 bar. Computational modeling studies revealed that C₂H₂ molecules preferentially occupy both sides of the square-planar metal centers. Breakthrough experiments demonstrated that JNU-4 is capable of achieving by far the largest C₂H₂ capture capacity (7.14 mmol g⁻¹) from an equimolar mixture of C₂H₂/CO₂, and fuel-grade C₂H₂ production (4.69 mmol g⁻¹, ≥98%) upon desorption.

Despite having more accessible binding sites, most of the MOF materials used for hydrocarbon separation have rigid frameworks whose dimensions/pore sizes hardly change upon loading with guest molecules, resulting in a decrease in binding affinity once the strongest adsorption sites are occupied. The flexible MOFs with induced-fit behavior may adaptively match the sizes and shapes of the guest molecules, gradually enhancing the host–guest interaction. Meanwhile MOF materials constructed with infinite 1-periodic metal clusters are known for their exceptional stability. In 2019, our group reported the synthesis of a highly stable MOF (JNU-1) from the reaction of benzotriazole-5-carboxylic acid (H₂btca) and Zn(NO₃)₂·6H₂O.³³ JNU-1 contains 1-periodic Zn clusters with high-density open metal sites oriented toward the one-dimensional (1D) diamond-shaped channels. More importantly, it exhibited an induced-fit adsorption behavior when exposed to C₂H₂ (Fig. 3). C₂H₂ adsorption isotherms displayed step-shaped adsorption curves at low pressures, which gradually disappeared with increasing temperature. Continuous C₂H₂ adsorption/desorption measurements confirmed the robust interaction with C₂H₂, and JNU-1 can only be fully regenerated under a high vacuum and a high temperature (393 K). In situ powder X-ray diffraction (PXRD) measurements implied that the JNU-1 framework can self-adaptively shrink in response to C₂H₂ (Fig. 3d), resulting in a rarely observed continuously increasing C₂H₂ binding affinity with C₂H₂ loading, as evidenced by the calculated C₂H₂ adsorption enthalpy (Q_st) (Fig. 3c). In situ single-crystal X-ray diffraction (SCXRD) and computational simulation revealed that C₂H₂ is initially bound by two unsaturated Zn metal sites via side-on metal–π interactions (Fig. 3a and b) and the channel size of JNU-1 along the b axis is eventually reduced from 12.7 to 10.8 Å, confirming the induced-fit behavior of JNU-1 upon C₂H₂ adsorption.

Fig. 3 (a) Schematic representation of the induced-fit behavior of JNU-1 upon C₂H₂ adsorption. (b) Single-crystal structure of C₂H₂@JNU-1 with C₂H₂ molecules being observed on the primary site (site I, pink) and the secondary site (site II, blue). (c) Comparison of the C₂H₂ adsorption enthalpy (Q_st) of JNU-1 and some selected MOFs. (d) In situ PXRD patterns of JNU-1 at different C₂H₂ concentrations. Reproduced from ref. 33 with permission from copyright 2019, John Wiley and Sons.

Although seemingly efficient, the olefin/paraffin separation relying on the open metal sites of MOF materials could be dramatically diminished under humid conditions due to their equally strong (if not more) binding affinity toward the water molecules. On the other hand, introducing polar functional groups with electronegative elements (e.g., N, O, and F) has been widely explored as an alternative approach to facilitate the olefin/paraffin separation. Many MOFs have been explored for the separation of C₂H₆/C₂H₄ based on surface engineering strategies (Table 1). In 2018, inspired by natural metalloenzymes and synthetic compounds for alkane C–H activation, including metal–peroxo, metal–hydroperoxo, and metal–oxo species, Chen et al. reported iron oxidation of the MOF material Fe₂(dobdc),³⁴ and the obtained Fe₂(O₂)(dobdc) exhibited a preferential binding of C₂H₆ over C₂H₄ and the highest C₂H₆/C₂H₄ separation coefficient (4.4) (Fig. 4a). Single-component adsorption isotherm results showed that the C₂H₆ adsorption capacity of Fe₂(O₂)(dobdc) reached 74.3 cm³ g⁻¹ (3.32 mmol g⁻¹) (Fig. 4b), realizing a reversed C₂H₆/C₂H₄ adsorption behaviour as opposed to pristine Fe₂(dobdc). High-resolution neutron powder diffraction (NPD) measurements revealed that C₂D₆ preferentially binds the peroxo site through a C–D⋯O interaction (D⋯O, ∼2.17 to 2.22 Å) (Fig. 4a), with a distance much shorter than the sum of the van der Waals (vdW) radii of the oxygen (1.52 Å) and hydrogen (1.20 Å) atoms. The C₂H₄/C₂H₆ (50/50) breakthrough results confirmed that Fe₂(O₂)(dobdc) is capable of directly producing 0.79 mmol g⁻¹ of C₂H₄ with ≥99.99% purity in a single breakthrough operation (Fig. 4c).

Table 1 Some selected MOFs for C₂H₆/C₂H₄ separation based on the surface engineering strategy

Materials	T (K)	C2H6/C2H4 uptake (mmol g^−1, 1 bar)	IAST selectivity (C2H6/C2H4, 50/50)	Interaction	Ref.
NKMOF-8-Br	298	4.22/3.67	2.65	C–H⋯π	46
NKMOF-8-Me	298	4.82/4.67	1.88	C–H⋯π	46
Cu(Qc)2	298	1.85/0.78	3.4	C–H⋯π	44
IRMOF-8	298	3.6/2.75	1.6	C–H⋯π	39
Fe2(O2)(dobdc)	298	3.4/2.6	4.4	C–H⋯O	34
MUF-15	293	4.69/4.15	1.96	C–H⋯π	38
TJT-100	298	3.66/3.4	1.2	C–H⋯O	47
Ni(IN)2	298	3.05/0.8	2.44	C–H⋯π	37
AzoleTh-1	298	4.47/3.62	1.46	C–H⋯π	43
FJI-H11-Me(des)	298	2.59/2.05	2.09	C–H⋯O/π	45
1a	298	3.63/3.28	2.15	C–H⋯O/π/N	41
Tb-MOF-76(NH2)	298	3.27/2.97	2.05	C–H⋯O/π/N	48
UiO-67-(NH2)2	296	5.32/4.32	1.7	C–H⋯O/π/N	49
Zn-atz-oba	298	2.1/2	1.27	C–H⋯N	50

---

Fig. 4 (a) Structure of Fe₂(O₂)(dobdc) and its interaction with C₂D₆ determined from neutron powder diffraction (NPD) studies. (b) C₂H₆ and C₂H₄ single-component adsorption isotherms of Fe₂(O₂)(dobdc) at 298 K. (c) Experimental column breakthrough curves of Fe₂(O₂)(dobdc) for C₂H₄/C₂H₆ (50/50) mixtures at 298 K. Reproduced from ref. 34 with permission from copyright 2018, The American Association for the Advancement of Science.

The introduced polar functional groups, however, may also interact with water molecules, which could significantly affect their separation potential under humid conditions. For example, the Fe₂(O₂)(dobdc) material must be handled inside a glove box. Recent studies show that nonpolar environments can block moisture from entering inside the frameworks, thus maintaining the C₂H₄/C₂H₆ separation potential even under humid conditions. Considering that the polarizability of alkanes is higher than that of alkenes (C₂H₄: 42.52 × 10⁻²⁵ cm³, C₂H₆: 44.7 × 10⁻²⁵ cm³, C₃H₆: 62.6 × 10⁻²⁵ cm³, and C₃H₈: 63.7 × 10⁻²⁵ cm³),³⁵ introducing nonpolar functional groups (e.g., –methyl, –naphthalene, and –anthracene) on the pore surface could provide a substantial energetic contribution to the preferential adsorption of alkanes over alkenes.^36–43 In 2018, Chen et al. reported two isoreticular ultramicroporous MOF materials, Cu(ina)₂ (Hina = isonicotinic acid) and Cu(Qc)₂ (HQc = quinolone-5-carboxylic acid), for C₂H₆/C₂H₄ separation (Fig. 5a and b).⁴⁴ The pore surfaces of both MOF materials are decorated with aromatic rings, which favour the binding of more polarizable C₂H₆ over C₂H₄. From a structural point of view, Cu(Qc)₂ has a smaller pore aperture and a larger aromatic π system than Cu(ina)₂, which endows Cu(Qc)₂ with further improved C₂H₆/C₂H₄ separation performance as a C₂H₆-selective adsorbent. As expected, C₂H₆ and C₂H₄ adsorption isotherms at 298 K and 1 bar on Cu(Qc)₂ can reach up to 60.0 cm³ cm⁻³ (1.85 mmol g⁻¹) and 25.3 cm³ cm⁻³ (0.78 mmol g⁻¹), respectively (Fig. 5c), resulting in a superior C₂H₆/C₂H₄ uptake ratio of 237%. This value is much higher than that of its isoreticular analog Cu(ina)₂ (105%). High-resolution neutron powder diffraction (NPD) analyses of the C₂D₆-loaded Cu(Qc)₂ revealed that the C₂D₆ molecule is located in a rhombic cavity formed by aromatic rings of ligands from the same layered network via multiple C–D⋯π interactions (Fig. 5d and e). The breakthrough experimental result for an equimolar C₂H₄/C₂H₆ mixture demonstrated a complete separation of C₂H₄ from the C₂H₆/C₂H₄ mixture, affording a C₂H₄ productivity of 587 mmol L⁻¹ (C₂H₄/adsorbent) (Fig. 5f). This strategy was also illustrated in the MOF material FJI-H11-Me(des),⁴⁵ and the accessible surface of FJI-H11-Me(des) is decorated with aromatic rings and alkyl groups of the ligands, resulting in a preferential C₂H₆ adsorption over C₂H₄. The adsorption capacity can reach up to 57.97 cm³ g⁻¹ (2.59 mmol g⁻¹) and 46.65 cm³ g⁻¹ (2.08 mmol g⁻¹) for C₂H₆ and C₂H₄, respectively, at 298 K and 1 bar. In addition, it is worth noting that the separation performance of FJI-H11-Me(des) was not affected even under high relative humidity (RH) due to its hydrophobicity.

Fig. 5 (a) and (b) Crystal structures of Cu(ina)₂ and Cu(Qc)₂. (c) Adsorption isotherms of C₂H₆ and C₂H₄ on Cu(ina)₂ and Cu(Qc)₂ at 298 K. (d) and (e) Neutron diffraction crystal structures of [Cu(Qc)₂]@C₂D₆ and [Cu(Qc)₂]@C₂D₄. (f) Experimental breakthrough curves of Cu(Qc)₂ for an equimolar C₂H₆/C₂H₄ mixture at 298 K and 1.0 bar. Reproduced from ref. 44 with permission from copyright 2018, American Chemical Society.

2.2. Molecular docking

Molecular docking is known in biological processes for recognizing and placing specific molecules at the receptor sites of proteins. For some MOF materials, their structures can be characterized by pocket-like cavities, which may preferentially accommodate guest molecules of matching sizes and shapes, not necessarily the smaller ones. The rational design of molecular pockets or traps can endow MOFs with great potential for hydrocarbon separation.

In 2015, Chen and Zhang et al. reported a microporous MOF, MAF-49, featuring 1D zigzag channels decorated with multiple nitrogen-containing groups (Fig. 6a).⁵¹ The single-component gas adsorption isotherms of MAF-49 showed a higher uptake of C₂H₆ than C₂H₄ at 316 K (Fig. 6b) and an exceptionally high C₂H₆/C₂H₄ selectivity of 9. In situ SCXRD and computational simulation revealed that C₂H₆ forms six C–H⋯N interactions within the pockets on the channels, while C₂H₄ forms four C–H⋯N interactions (Fig. 6c and d). The excellent C₂H₆/C₂H₄ separation potential of MAF-49 was verified by breakthrough experiments on a C₂H₆/C₂H₄ mixed gas. This work has successfully demonstrated that the introduction of multiple hydrogen-bonding acceptors on the pore surface of molecular pockets is an effective strategy to discriminate paraffin from olefin. It should be pointed out that, despite having high C₂H₆/C₂H₄ selectivity, the molecule-pocket-based MAF-49 exhibited a limited adsorption capacity of C₂H₆ (1.7 mmol g⁻¹), probably because every molecular pocket can only accommodate one molecule. Furthermore, considering its highly electronegative environment of the molecular pocket, MAF-49 may not be able to maintain its C₂H₆/C₂H₄ separation capacity under humid conditions due to the competition of water molecules for the adsorption sites.

Fig. 6 (a) Crystal structure of MAF-49, showing a water molecule in the molecular pocket. (b) Single-component adsorption isotherms of C₂H₆, C₂H₄, C₂H₂, and CO₂ on MAF-49 at 316 K. Crystal structure of (c) C₂H₆@MAF-49 and (d) C₂H₄@MAF-49, highlighting the multiple binding interactions. Reproduced from ref. 51 with permission from copyright 2015, Springer Nature.

In 2019, our group reported a microporous MOF (JNU-2)⁵² with cage-like cavities interconnected through tailored apertures (Fig. 7a). The aperture size of 3.7 Å is in the domain of the kinetic diameters of C₂H₄ and C₂H₆ molecules. Indeed, C₂H₄ and C₂H₆ adsorption isotherms showed that JNU-2 exhibited a preferential binding of C₂H₆ over C₂H₄. The adsorption capacity of C₂H₆ can reach up to 4.19 mmol g⁻¹ at 298 K and 1 bar (Fig. 7b), indicating that the large cavities of JNU-2 can be translated into large adsorption capacity, and this value is higher than that of most of the benchmark MOF materials. Molecular modeling studies revealed that C₂H₆ forms four weak hydrogen bonds with oxygen atoms at the aperture, while C₂H₄ forms only two weak hydrogen bonds (Fig. 7c and d). Breakthrough experiments under dry conditions showed that about 21.2 L of C₂H₄ with over 99.99% purity could be retrieved from a C₂H₆/C₂H₄ (50/50) mixture for 1 kg of JNU-2, which is higher than most of the benchmark C₂H₆-selective adsorbents, including MUF-15 (6.6 L kg⁻¹) and ZIF-8 (1.2 L kg⁻¹) (Fig. 7e and f). Furthermore, breakthrough experiments under humid conditions demonstrated that JNU-2 maintained its exceptional separation potential for C₂H₆/C₂H₄ mixtures. Overall, JNU-2 can achieve both large C₂H₆ adsorption capacity and a high C₂H₆/C₂H₄ separation coefficient even under humid conditions due to its cage-interconnected structure and balanced hydrophilicity/hydrophobicity.⁵³ The apertures connecting the cages function as screening sites for C₂H₆ molecules to preferentially go through, while the internal cavities provide a large surface for adsorption capacity. By constructing MOF materials with large cavities interconnected by tailor-made apertures, this work has successfully exemplified a promising strategy for hydrocarbon separation with balanced adsorption capacity and separation coefficient.

Fig. 7 (a) Connolly surface representation of the cage-interconnected structure of JNU-2. (b) Single-component adsorption isotherms of C₂H₆ and C₂H₄ on JNU-2 at 298 K. Host–guest interactions of (c) C₂H₆ and (d) C₂H₄ with JNU-2 at the aperture. (e) Comparison of the breakthrough curves of JNU-2 and other MOF materials for an equimolar C₂H₆/C₂H₄ mixture at a flow rate of 1.0 mL min⁻¹ and 298 K. (f) Comparison of C₂H₄ productivity estimated from the breakthrough curves of ZIF-8, MUF-15, and JNU-2. Reproduced from ref. 52 with permission from copyright 2019, American Chemical Society.

Considering that the molecule pockets/traps are located right on the diffusion pathways for most MOF materials, the adsorption equilibrium could be difficult to reach as the matched molecules have to pass through many apertures, leading to slow adsorption/desorption kinetics and thus less energy efficiency. In 2021, our group reported an orthogonal-array dynamic molecular sieving material, JNU-3a, enabling both large separation capacity and fast adsorption–desorption kinetics for C₃H₆/C₃H₈ separation.⁵⁴ SCXRD analysis revealed that JNU-3a featured a 1D channel with orthogonal-array molecular pockets on both sides opening to the channel through ‘gourd-shaped’ apertures. The size of the aperture is approximately 3.7 Å, which seemed impossible for C₃H₆ (4.0 Å) and C₃H₈ (4.3 Å) to go through (Fig. 8a–c). Yet, in situ SCXRD studies on gas@JNU-3a observed tilting and rotation of the aromatic ring on the organic linker, facilitating the guest molecules such as C₃H₆ and C₃H₈ to open the ‘gourd-shaped’ apertures and enter into the pockets (Fig. 8d). As such, both C₃H₆ and C₃H₈ displayed stepwise adsorption, and the gate-opening pressure of C₃H₈ is about 0.52 bar at 303 K, approximately five times higher than that of C₃H₆, offering a high selectivity of C₃H₆ over C₃H₈ (Fig. 8e). Breakthrough experiments under dry conditions demonstrated that 34.2 L of C₃H₆ with over 99.5% purity could be obtained from an equimolar C₃H₆/C₃H₈ mixture at a flow rate of 1.0 mL min⁻¹ for 1 kg of JNU-3a in a single adsorption/desorption cycle. Moreover, 53.5 L and 44.9 L of C₃H₆ with over 99.5% purity could be obtained from an equimolar C₃H₆/C₃H₈ mixture for 1 kg of JNU-3a under 0% RH and 50% RH conditions, respectively, at a flow rate of 6.0 mL min⁻¹. Owing to the orthogonal-array dynamic molecular sieving mechanism, the diffusion kinetics of C₃H₆ and C₃H₈ on JNU-3a are at least one order of magnitude faster than those of the two reported best-performing sieving materials (KAUST-7 and Y-abtc) for C₃H₆/C₃H₈ mixtures. JNU-3a represented next-generation molecular sieving materials with intrinsically fast adsorption–desorption kinetics, applicable not only to hydrocarbon separation but also to other adsorptive separations.

Fig. 8 (a) Pore structure of JNU-3 viewed along the b axis showing the molecular pockets (turquoise) and the 1D channel (yellow). (b) Cross-sectional view of the Connolly surface of the void in (a), where a probe radius of 0.8 Å was used to visualize the molecular pockets opening to the 1D channel. (c) Close-up view of the “gourd-shaped” aperture connecting the pocket to the channel. (d) Connolly surface of the “gourd-shaped” apertures in C₃H₆@JNU-3a and C₃H₈@JNU-3a. (e) Single-component adsorption/desorption isotherms of C₃H₆, C₃H₈, and C₃H₆/C₃H₈ (50 : 50) on JNU-3a at 303 K. (f) Breakthrough curves of an equimolar mixture of C₃H₆/C₃H₈ (1.0 mL min⁻¹) on JNU-3a, followed by desorption curves under helium gas (10.0 mL min⁻¹) sweeping at 303 K. (g) Breakthrough curves of an equimolar mixture of C₃H₆/C₃H₈ (6.0 mL min⁻¹) on JNU-3a under 0% RH and 50% RH conditions, followed by desorption curves under helium gas (10.0 mL min⁻¹) sweeping at 303 K. Reproduced from ref. 54 with permission from copyright 2021, Springer Nature.

2.3. Size exclusion

Size exclusion is considered to be one of the most promising separation strategies for hydrocarbon separation due to its potential to completely block larger molecules and therefore provide an infinite selectivity.^55–58 In addition, the rigid molecular sieving MOF materials can effectively avoid the co-adsorption of analogous alkanes that are commonly observed in MOF materials with open metal sites. However, the molecular-size difference is 0.42 Å for C₃H₈ and C₃H₆ and only 0.28 Å for C₂H₆ and C₂H₄. It is challenging to design and construct MOF materials with the right aperture sizes as molecular sieves for targeted hydrocarbon separation.

In 2018, Chen et al. reported the synthesis of a microporous MOF material, Ca(C₄O₄)(H₂O) (UTSA-280), from calcium nitrate and squaric acid under mild aqueous conditions. UTSA-280 possesses a rigid 1D channel with an aperture large enough for C₂H₄ inclusion but not for C₂H₆ (Fig. 9a and b).⁵⁹ The cross-sectional area of the rigid 1D channel is about 14.4 Ų, which falls between the minimum cross-sectional areas of C₂H₄ (13.7 Ų) and C₂H₆ (15.5 Ų). C₂H₆ adsorption isotherms confirmed that the adsorption of C₂H₆ at 298 K and 1 bar is a mere 0.098 mmol g⁻¹, indicating the complete blocking of C₂H₆. On the other hand, the adsorption of C₂H₄ could reach up to 2.5 mmol g⁻¹ under similar conditions (Fig. 9c). UTSA-280 set the record for the C₂H₄/C₂H₆ separation coefficient (>10 000), which is orders of magnitude higher than those of previously reported MOF materials, such as NOTT-300 (48.7) and Mg-gallate (52). SCXRD studies of the C₂H₄-loaded UTSA-280 coupled with density functional theory (DFT) calculations revealed that the C=C axis of the C₂H₄ molecule is aligned along the channel, interacting with the surroundings through C–H⋯O hydrogen bonding, π⋯π stacking, and vdW interactions. In contrast, C₂H₆ cannot enter into the channels because of its slightly oversized molecular dimension. The capture capacity of C₂H₄ can reach up to 1.86 mmol g⁻¹ from an equimolar C₂H₄/C₂H₆ mixture, which is slightly lower than the equilibrium adsorption capacity of UTSA-280 at 0.5 bar (2.05 mol kg⁻¹). Moreover, UTSA-280 can be synthesized on a kilogram scale under environmentally friendly conditions, and the obtained material maintains excellent separation performance (Fig. 9d).

Fig. 9 (a) View of the crystal structure and packing diagram of C₂H₄@UTSA-280. (b) Schematic illustration of size/shape sieving based on the minimum cross-sectional area of C₂H₄ and C₂H₆ molecules. (c) Single-component adsorption/desorption isotherms of C₂H₄ and C₂H₆ at 298 K. (d) Comparison of experimental breakthrough curves for a 50/50 C₂H₄/C₂H₆ mixture on UTSA-280 synthesized on different scales. Reproduced from ref. 59 with permission from copyright 2018, Springer Nature.

In the same year, Ren et al. reported a family of gallate-based MOF materials, M-gallate (M = Ni, Mg, Co), with interconnected zigzag channels (3.47 × 4.85, 3.56 × 4.84, and 3.69 × 4.95 Ų for Ni, Mg, Co-gallate, respectively),⁶⁰ which are between the molecular sizes of C₂H₆ (3.81 × 4.08 × 4.82 ų) and C₂H₄ (3.28 × 4.18 × 4.84 ų). Adsorption isotherms revealed that all M-gallate adsorbed only negligible amounts of C₂H₆ (<0.31 mmol g⁻¹) at 298 K up to 1 bar, indicating that C₂H₆ could be excluded from entering the pore channels. Experimental breakthrough tests for an equimolar C₂H₄/C₂H₆ mixture indicated that C₂H₆ was eluted through the packed column quickly and high-purity C₂H₄ could be obtained upon desorption. Neutron powder diffraction (NPD) experiments confirmed the adsorption site of C₂D₄ in Mg-gallate (Fig. 10). However, it should be pointed out that, in both cases, desorption is required in order to obtain C₂H₄, and its purity could be lower than expected due to the dead space in the packed column.

Fig. 10 (a) Perspective view of M-gallate along the c-axis showing a triangular main channel and regular branching channels abutting the main channel. (b) The zigzag channels of M-gallate. (c) Schematic diagram of M-gallate for C₂H₄/C₂H₆ separation. (d) Gas sorption isotherms of C₂H₄ (red) and C₂H₆ (blue) in Co-gallate at 298 K. (e) Experimental breakthrough curves at 273 K for M-gallate C₂H₄/C₂H₆ (50/50 mixture) separation with a flow rate of 0.5 mL min⁻¹. (f) Neutron diffraction crystal structure of Mg-gallate 0.485C₂D₄ at 200 K. Reproduced from ref. 60 with permission from copyright 2018, John Wiley and Sons.

By contrast, the design and construction of MOF materials as molecular sieves for C₃H₆/C₃H₈ separation are more challenging due to their increased molecular flexibility.^61–63 Eddaoudi et al. reported the assembly of a fluorinated MOF, NbOFFIVE-1-Ni (KAUST-7),⁶⁴ with (NbOF₅)²⁻ as pillars interconnecting the Ni(II)–pyrazine square-grid layers. The resulting 1D channel enables the complete molecular exclusion of C₃H₈ from C₃H₆. Compared to (SiF₆)²⁻-pillared isomer SIFSIX-3-Ni, (NbOF₅)²⁻ is a relatively larger spacer, leading to shorter C–H⋯F interaction distances between the anionic pillar and pyrazine, which hinders the rotation of the pyrazine ligand and reduces the pore aperture from 4.75 to 3.04 Å (Fig. 10a and b). Gas adsorption data indicated that the reduced pore size of KAUST-7 can still allow C₃H₆ to diffuse into the channels, but not C₃H₈. The adsorption profile of an equimolar C₃H₆/C₃H₈ mixture is in good agreement with that of pure C₃H₆ (Fig. 11c). Furthermore, both calorimetric and gravimetric measurements were carried out, and an almost negligible heat of adsorption of C₃H₈ was observed, confirming the molecular exclusion of C₃H₈. Breakthrough experiments were performed for an equimolar C₃H₆/C₃H₈ mixture to further confirm the separation potential of KAUST-7 (Fig. 11d). In addition, a number of custom-made frameworks have been designed for C₃H₆/C₃H₈ separation by precisely adjusting the pore size around 4.3–4.7 Å, from which the large C₃H₈ molecule can be excluded (Table 2). For example, Li and co-workers reported a microporous metal–organic framework Y₆(OH)₈(abtc)₃(H₂O)₆(DMA)₂ (Y-abtc). Benefiting from the optimal pore window size, Y-abtc can exclusively adsorb C₃H₆ with fast kinetics at ambient temperature.^65,66 Breakthrough measurements confirmed that polymer-grade propylene can be produced with a purity of 99.5% from a mixture of C₃H₆/C₃H₈ (5/95, v/v).

Fig. 11 Comparison of the building block arrangement of (a) KAUST-7 and (b) SIFSIX-3-Ni. (c) Adsorption isotherms of C₃H₆, C₃H₈, and an equimolar mixture of C₃H₆/C₃H₈ on KAUST-7 at 298 K. (d) Breakthrough curves for an equimolar C₃H₆/C₃H₈ mixture on KAUST-7 at 298 K. Reproduced from ref. 64 with permission from copyright 2016, The American Association for the Advancement of Science.

Table 2 A list of selected MOFs for separation of C₃H₆/C₃H₈ based on size exclusion

Materials	T (K)	C3H6 uptake (mmol g^−1, 1 bar)	C3H8 uptake (mmol g^−1, 1 bar)	Ref.
Co-gallate	298	1.79	0.14	67
HIAM-301	298	3.16	<0.3	62
KAUST-7	298	1.41	0.04	64
Y-bptc	298	1.95	0	65
UTSA-400	298	1.84	0.05	68
Y-dbai	298	2.57	0.1	69

---

3. Conclusion and outlook

MOF materials are some of the most promising porous solids for adsorption applications due to their exceptional structural diversity, tunable pore dimension, and easy surface functionalization. This perspective summarizes the latest progress in the development of MOF materials for the selective adsorption of hydrocarbons, and three main strategies to design MOF materials for hydrocarbon separation: (1) surface engineering, installing functional groups on the pore surface to discriminate one hydrocarbon molecule from the others; (2) molecular docking, building molecular pockets to selectively trap guest molecules of particular sizes and shapes; (3) size exclusion, tuning the pore apertures precisely for the inclusion of only small-sized molecules.

The detailed MOF structures and their dynamic interactions with hydrocarbon molecules are critical to their performance in hydrocarbon separation. An in-depth understanding of the structure–property relationship can guide future MOF research in constructing specific structures and functions for the targeted separation. With the rapid development of in situ characterization techniques, it is now possible to study the dynamic interactions between the hosts (MOFs) and guests (gas molecules) at the molecular level. For example, in situ X-ray diffraction (XRD) or neutron diffraction allows researchers to see how the MOF structures are affected by the changes in temperatures, pressures, or gas compositions, providing insight into their stability, phase transitions, and host–guest interactions. In situ spectroscopy techniques such as FT-IR, Raman, and UV-vis can supply information about the electronic and vibrational properties of the MOF materials upon the inclusion of guest molecules, revealing the dynamic mechanisms of the adsorption/desorption processes. In situ microscopy and imaging techniques, such as transmission electron microscopy (TEM) coupled with gas adsorption, can produce high-resolution images with information on the distribution and location of guest molecules within the MOF materials. Computational methods, such as density functional theory (DFT), molecular dynamics (MD), and Monte Carlo simulations, can be used to calculate the binding energies, host–guest interactions, and dynamics of guest molecules within the MOF frameworks.

Currently, the study of MOF materials for separation applications is still largely in the lab stage. Achieving large adsorption capacity and high separation selectivity simultaneously seems to be a common goal for scientists and engineers in the development of MOF materials. For MOF sorbents to be competitive with conventional cryogenic processes for separations, there are also other challenges (barriers) that need to be addressed, including stability, regenerability, scalability and cost-effectiveness, and industrial implementation. An overall performance in these aspects would be ideal for the MOF sorbents. As for the future directions, we believe that these barriers these should be addressed collectively instead of individually.

(1) Adsorption capacity and selectivity: one of the promising strategies to balance selectivity and adsorption capacity for hydrocarbon separation is designing MOFs with large cavities interconnected by tailor-made apertures.^70–73

(2) Stability: future research should focus on the development of robust and stable MOFs,⁷⁴ such as the design of high-valency metal–carboxylate frameworks and low-valency metal–azolate frameworks.⁷⁵

(3) Regenerability: understanding adsorption/desorption mechanisms in MOFs is critical to developing effective regeneration techniques. By studying the interactions between the hydrocarbon molecules and the MOF structure at the molecular level, researchers can identify optimal conditions and methods for efficient desorption.

(4) Scalability and cost-effectiveness: the current practice of synthesizing MOF materials typically involves high-boiling-point solvents, which is not only costly but also environmentally unfriendly. More effort should be put into selecting inexpensive ligands and metal salts for water-mediated or non-mediated synthesis. Meanwhile, researchers should also explore scalable synthesis techniques, including continuous flow processes and microwave-assisted methods.

(5) Industrial implementation: bridging the gap between laboratory-scale research and industrial-scale applications is crucial. Collaboration between researchers, industry partners, and policy-makers can facilitate the deployment of MOF-based separation technologies in real-world settings.

Industrial crude hydrocarbon products often contain a variety of impurities, which may complicate the purification process. For example, if we intend to remove two or more components, they would inevitably compete for the adsorption sites on the diffusion channels of the MOF materials, and one of them could be quickly pushed out, resulting in less efficient separation. Meanwhile, the desired component may not necessarily be the largest or smallest in terms of size; the one-step separation of medium-sized molecules is a significant challenge with tremendous implications for hydrocarbon separation, which may be achievable through the cooperative action of thermodynamics and kinetics. Different separation scenarios require MOF materials with different pores/channels and functions. Given the tunability of MOF structures in their pore dimensions and surface chemistry, adsorptive separation with MOF materials may indeed hold promise as an innovative technology for hydrocarbon separation, with the potential to replace the current energy-intensive practices.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 21731002, 21975104, 22271120, 22301102, and 22150004), Guangdong Basic and Applied Basic Research Foundation (2023A1515010952), Innovation Team Project in Guangdong Colleges and Universities (2021KCXTD009), and National Postdoctoral Program for Innovative Talent (BX20220132).

Notes and references

1. D. S. Sholl and R. P. Lively, Nature, 2016, 532, 435–437.
2. https://www.databridgemarketresearch.com/reports/global-propylene-market .
3. D. J. Safarik and R. B. Eldridge, Ind. Eng. Chem. Res., 1998, 37, 2571–2581.
4. S. M. Sadrameli, Fuel, 2015, 140, 102–115.
5. S. Kitagawa, Angew. Chem., Int. Ed., 2015, 54, 10686–10687.
6. E. Worrell, D. Phylipsen, D. Einstein and N. Martin, Energy Use and Energy Intensity of the U.S. Chemical Industry, Lawrence Berkeley National Laboratory, Berkeley, CA, 2000.
7. S. Chu, Y. Cui and N. Liu, Nat. Mater., 2017, 16, 16–22.
8. H. Li, K. Wang, Y. Sun, C. T. Lollar, J. Li and H.-C. Zhou, Mater. Today, 2018, 21, 108–121.
9. J. Kim, L. C. Lin, R. L. Martin, J. A. Swisher, M. Haranczyk and B. Smit, Langmuir, 2012, 28, 11914–11919.
10. P. J. Bereciartua, Á. Cantín, A. Corma, J. L. Jordá, M. Palomino, F. Rey, S. Valencia, E. W. Corcoran Jr, P. Kortunov, P. I. Ravikovitch, A. Burton, C. Yoon, Y. Wang, C. Paur, J. Guzman, A. R. Bishop and G. L. Casty, Science, 2017, 358, 1068–1071.
11. B.-U. Choi, D.-K. Choi, Y.-W. Lee, B.-K. Lee and S.-H. Kim, J. Chem. Eng. Data, 2003, 48, 603–607.
12. Q. Gong, L. Yu, J. Ding, S. Zhang, Y. Bo, K. Chi, H. Wang, Q. Xia, S. He and J. Li, Sep. Purif. Technol., 2022, 294, 121219.
13. M. Hartmann, A. G. Machoke and W. Schwieger, Chem. Soc. Rev., 2016, 45, 3313–3330.
14. W. G. Cui, T. L. Hu and X. H. Bu, Adv. Mater., 2020, 32, 1806445.
15. Z. Bao, G. Chang, H. Xing, R. Krishna, Q. Ren and B. Chen, Energy Environ. Sci., 2016, 9, 3612–3641.
16. H. Furukawa, K. E Cordova, M. O'Keeffe and O. M. Yaghi, Science, 2013, 341, 1230444.
17. W.-X. Zhang, P.-Q. Liao, R.-B. Lin, Y.-S. Wei, M.-H. Zeng and X.-M. Chen, Coord. Chem. Rev., 2015, 293, 263–278.
18. E. A. Dolgopolova, A. M. Rice, C. R. Martin and N. B. Shustova, Chem. Soc. Rev., 2018, 47, 4710–4728.
19. W. G. Cui, G. Y. Zhang, T. L. Hu and X. H. Bu, Coord. Chem. Rev., 2019, 387, 79–120.
20. M. Giménez-Marqués, T. Hidalgo, C. Serre and P. Horcajada, Coord. Chem. Rev., 2016, 307, 342–360.
21. X. Zhao, Y. Wang, D. S. Li, X. Bu and P. Feng, Adv. Mater., 2018, 30, 1705189.
22. K. Adil, Y. Belmabkhout, R. S. Pillai, A. Cadiau, P. M. Bhatt, A. H. Assen, G. Maurin and M. Eddaoudi, Chem. Soc. Rev., 2017, 46, 3402–3430.
23. J. P. Zhang, H. L. Zhou, D. D. Zhou, P. Q. Liao and X.-M. Chen, Natl. Sci. Rev., 2018, 5, 907–919.
24. G. Fereyand and C. Serre, Chem. Soc. Rev., 2009, 38, 1380–1399.
25. Q. Chen, S. Xian, X. Dong, Y. Liu, H. Wang, D. H. Olson, L. J. Williams, Y. Han, X.-H. Bu and J. Li, Angew. Chem., Int. Ed., 2021, 133, 10687–10691.
26. Q. M. Wang, D. M. Shen, M. Bülow, M. L. Lau, S. G. Deng, F. R. Fitch, N. O. Lemcoff and J. Semanscin, Microporous Mesoporous Mater., 2002, 55, 217–230.
27. W. You, Y. Liu, J. D. Howe, D. Tang and D. S. Sholl, J. Phys. Chem. C, 2018, 122, 27486–27494.
28. E. D. Bloch, W. L. Queen, R. Krishna, J. M. Zadrozny, C. M. Brown and J. R. Long, Science, 2012, 335, 1606–1610.
29. Z. Zhang, S. B. Peh, Y. Wang, C. Kang, W. Fan and D. Zhao, Angew. Chem., Int. Ed., 2020, 132, 19089–19094.
30. J. Pei, K. Shao, J. X. Wang, Y. Yang, Y. Cui, R. Krishna, B. Li and G. Qian, Adv. Mater., 2020, 32, 1908275.
31. L. Zhang, K. Jiang, L. Yang, L. Li, E. Hu, L. Yang, K. Shao, H. Xing, Y. Cui, Y. Yang, B. Li, B. Chen and G. Qian, Angew. Chem., Int. Ed., 2021, 133, 16131–16138.
32. H. Zeng, X. J. Xie, Y. Wang, D. Luo, R.-J. Wei, W. Lu and D. Li, Chem. Sci., 2022, 13, 12876–12882.
33. H. Zeng, M. Xie, Y. L. Huang, Y. Zhao, X. J. Xie, J.-P. Bai, M.-Y. Wang, R. Krishna, W. Lu and D. Li, Angew. Chem., Int. Ed., 2019, 58, 8515–8519.
34. L. Li, R. B. Lin, R. Krishna, H. Li, S. Xiang, H. Wu, J. Li, W. Zhou and B. Chen, Science, 2018, 362, 443–446.
35. J. R. Li, R. J. Kuppler and H. C. Zhou, Chem. Soc. Rev., 2009, 38, 1477–1504.
36. R. T. Yang, Adsorbents: Fundamentals and Applications, Wiley, Hoboken, 2003.
37. M. Kang, S. Yoon, S. Ga, D. W. Kang, S. Han, J. H. Choe, H. Kim, D. W. Kim, Y. G. Chung and C. S. Hong, Adv. Sci., 2021, 8, 2004940.
38. O. T. Qazvini, R. Babarao, Z.-L. Shi, Y.-B. Zhang and S. G. Telfer, J. Am. Chem. Soc., 2019, 141, 5014–5020.
39. J. Pires, M. L. Pinto and V. K. Saini, ACS Appl. Mater. Interfaces, 2014, 6, 12093–12099.
40. S. Jiang, L. Li, L. Guo, C. Song, Q. Yang, Z. Zhang, Y. Yang, Q. Ren and Z. Bao, Sci. China: Chem., 2021, 64, 666–672.
41. G. D. Wang, Y. Z. Li, W. J. Shi, H. Lei, Y.-Y. Wang and Z. Zhu, Angew. Chem., Int. Ed., 2022, 134, e202205427.
42. B. Zhu, J. Cao, S. Mukherjee, T. Pham, T. Zhang, T. Wang, X. Jiang, K. A. Forrest and M. J. Zaworotko, J. Am. Chem. Soc., 2021, 143, 1485–1492.
43. Z. Xu, X. Xiong, J. Xiong, R. Krishna, L. Li, Y. Fan, F. Luo and B. Chen, Nat. Commun., 2020, 11, 3163.
44. R. B. Lin, H. Wu, L. Li, X.-L. Tang, Z. Li, J. Gao, H. Cui, W. Zhou and B. Chen, J. Am. Chem. Soc., 2018, 140, 12940–12946.
45. Z. Di, C. Liu, J. Pang, S. Zou, Z. Ji, F. Hu, C. Chen, D. Yuan, M. Hong and M. Wu, Angew. Chem., Int. Ed., 2022, 134, e202210343.
46. S. B. Geng, E. Lin and X. Li, J. Am. Chem. Soc., 2021, 143, 8654–8660.
47. H.-G. Hao, Y.-F. Zhao, D.-M. Chen, J.-M. Yu, K. Tan, S. Ma, Y. Chabal, Z.-M. Zhang, J.-M. Dou, Z.-H. Xiao, G. Day, H.-C. Zhou and T.-B. Lu, Angew. Chem., Int. Ed., 2018, 130, 16299–16303.
48. G. D. Wang, R. Krishna, Y. Z. Li, W. J. Shi, L. Hou, Y. Y. Wang and Z. H. Zhu, Angew. Chem., Int. Ed., 2022, 61, e202213015.
49. X. W. Gu, J. X. Wang, E. Wu, H. Wu, W. Zhou, G. Qian, B. Chen and B. Li, J. Am. Chem. Soc., 2022, 144, 2614–2623.
50. J. W. Cao, S. Mukherjee, T. Pham, Y. Wang, T. Wang, T. Zhang, X. Jiang, H. J. Tang, K. A. Forrest, B. Space, M. J. Zaworotko and K.-J. Chen, Nat. Commun., 2021, 12, 6507.
51. P. Q. Liao, W. X. Zhang, J. P. Zhang and X.-M. Chen, Nat. Commun., 2015, 6, 8697.
52. H. Zeng, X.-J. Xie, M. Xie, Y. Huang, D. Luo, T. Wang, Y. Zhao, W. Lu and D. Li, J. Am. Chem. Soc., 2019, 141, 20390–20396.
53. X.-J. Xie, H. Zeng, M. Xie, W. Chen, G.-F. Hua, W. Lu and D. Li, Chem. Eng. J., 2022, 427, 132033.
54. H. Zeng, M. Xie, T. Wang, R.-J. Wei, X.-J. Xie, Y. Zhao, W. Lu and D. Li, Nature, 2021, 595, 542–548.
55. T.-L. Hu, H. Wang, B. Li, R. Krishna, H. Wu, W. Zhou, Y. Zhao, Y. Han, X. Wang, W. Zhu, Z. Yao, S. Xiang and B. Chen, Nat. Commun., 2015, 6, 7328.
56. B. Li, X. Cui, D. O'Nolan, H.-M. Wen, M. Jiang, R. Krishna, H. Wu, R.-B. Lin, Y.-S. Chen, D. Yuan, H. Xing, W. Zhou, Q. Ren, G. Qian, M. J. Zaworotko and B. Chen, Adv. Mater., 2017, 29, 1704210.
57. R. B. Lin, L. Li, H. Wu, H. Arman, B. Li, R.-G. Lin and B. Chen, J. Am. Chem. Soc., 2017, 139, 8022–8028.
58. Y. Xie, Y. Shi, E. M. Cedeño Morales, A. E. Karch, B. Wang, H. Arman, K. Tan and B. Chen, J. Am. Chem. Soc., 2023, 145, 2386–2394.
59. R. B. Lin, L. Li, H. L. Zhou, H. Wu, C. He, S. Li, R. Krishna, J. Li, W. Zhou and B. Chen, Nat. Mater., 2018, 17, 1128–1133.
60. Z. Bao, J. Wang, Z. Zhang, H. Xing, Q. Yang, Y. Yang, H. Wu, R. Krishna, W. Zhou, B. Chen and Q. Ren, Angew. Chem., Int. Ed., 2018, 130, 16252–16257.
61. B. Liang, X. Zhang, Y. Xie, R.-B. Lin, R. Krishna, H. Cui, Z. Li, Y. Shi, H. Wu, W. Zhou and B. Chen, J. Am. Chem. Soc., 2020, 142, 17795–17801.
62. L. Yu, X. Han, H. Wang, S. Ullah, Q. Xia, W. Li, J. Li, I. da Silva, P. Manuel, S. Rudić, Y. Cheng, S. Yang, T. Thonhauser and J. Li, J. Am. Chem. Soc., 2021, 143, 19300–19305.
63. H. Wang, X. Dong, V. Colombo, Q. Wang, Y. Liu, W. Liu, X.-L. Wang, X.-Y. Huang, D. M. Proserpio, A. Sironi, Y. Han and J. Li, Adv. Mater., 2018, 30, 1805088.
64. A. Cadiau, K. Adil, P. M. Bhatt, Y. Belmabkhout and M. Eddaoudi, Science, 2016, 353, 137–140.
65. H. Wang, X. Dong, V. Colombo, Q. Wang, Y. Liu, W. Liu, X.-L. Wang, X.-Y. Huang, D. M. Proserpio, A. Sironi, Y. Han and J. Li, Adv. Mater., 2018, 30, 1805088.
66. E. Velasco, S. Xian, L. Yu, H. Wang and J. Li, Sep. Purif. Technol., 2022, 282, 120010.
67. B. Liang, X. Zhang, Y. Xie, R.-B. Lin, R. Krishna, H. Cui, Z. Li, Y. Shi, H. Wu, W. Zhou and B. Chen, J. Am. Chem. Soc., 2020, 142, 17795–17801.
68. Y. Xie, Y. Shi, E. M. C. Morales, A. E. Karch, B. Wang, H. Arman, K. Tan and B. Chen, J. Am. Chem. Soc., 2023, 145, 2386–2394.
69. S. Tu, L. Yu, Y. Wu, Y. Chen, H. Wu, L. Wang, B. Liu, X. Zhou, J. Xiao and Q. Xia, AIChE J., 2022, 68, e17551.
70. L. Wang, W. Xue, H. Zhu, X. Guo, H. Huang and C. Zhong, Angew. Chem., Int. Ed., 2023, 135, e202218596.
71. L. Yu, S. Ullah, H. Wang, Q. Xia, T. Thonhauser and J. Li, Angew. Chem., Int. Ed., 2022, 134, e202211359.
72. X. Huang, S. Jiang, D. Ma, J. Xie, X. Feng and B. Wang, Angew. Chem., Int. Ed., 2023, 62, e202303671.
73. X.-P. Fu, Y.-L. Wang, X.-F. Zhang, Z. Zhang, C.-T. He and Q.-Y. Liu, CCS Chem., 2022, 4, 3416–3425.
74. M. Ding and H. L. Jiang, CCS Chem., 2021, 3, 2740–2748.
75. S. Yuan, L. Feng, K. Wang, J. Pang, M. Bosch, C. Lollar, Y. Sun, J. Qin, X. Yang, P. Zhang, Q. Wang, L. Zou, Y. Zhang, L. Zhang, Y. Fang, J. Li and H.-C. Zhou, Adv. Mater., 2018, 30, 1704303.

---