Field of Invention

The invention relates to a composite material membrane suitable for gas separation processes.

Background

The listing or discussion of a prior-published document in this specification should not necessarily be taken as an acknowledgement that the document is part of the state of the art or is common general knowledge.

Biological membrane systems are known to respond to stimuli. For example, changes in pH, temperature, electric field, light, etc. can induce transitions between ‘open’ and ‘closed conformations to allow or inhibit cellular transport. The extension of these characteristics for the control of permeance and selectivity in other membrane applications, such as gas separation, may redefine industrial processes. There are substantial interests in regulating pore size and/or interlayer distance in membranes as these changes allow separations based on the molecular sieving effect.

Metal-organic frameworks (MOFs) are an important class of crystalline porous materials and they have been applied in gas separation by both adsorption and membrane-based approaches. The molecular building blocks of the MOFs may be adjusted such as to induce a structural change in the MOFs under a certain external stimulus. The response in MOFs to external stimulus can serve as the basis for “smart” separation behaviour. Ligand rotational flexibility has been successfully utilized in some smart MOF membranes such as the defibrillation of soft ZIF-8 membrane under electric fields, the switch between trans and cis configuration of photoresponsive azobenzene containing MOFs by irradiation with ultraviolet or visible light, and the reversed thermo-switchable molecular sieving of MAMS-1 membrane through controlling the interlayer distance under different temperatures. These examples demonstrate the potential of flexibility in membrane separation.

The ‘gate-opening’ phenomenon is an example of a guest-induced responsive behavior, where the adsorbed guest molecules can promote a structural change of the host framework from a ‘closed’ form to an ‘open’ form under a particular pressure exerted by the guest molecule (the gate opening pressure) ( Angew . Chem. Int. Ed. 42, 428-431 (2003)). The gate-opening phenomena of flexible MOFs have been studied in adsorption-based gas separations. Several flexible MOFs such as MIL-53, ZIF-7, and ZIF-8 have been evaluated as adsorbents for various gas separations such as CO2 capture and purification of nature gas (Angew. Chem. Int. Ed. 45, 7751-7754 (2006); J. Am. Chem. Soc. 132, 17704-17706 (2010); and The Journal of Physical Chemistry Letters 4, 3618-3622 (2013)).

Yu et al. reported ultrathin, molecular-sieving graphene oxide (GO) membranes with thickness approaching 1.8 nm, prepared by a facile filtration process. These membranes showed mixture separation selectivities as high as 3400 and 900 for H2/CO2 and H2/N2 mixtures, respectively. However, these membranes showed low permeation flux ( Science , 2013, 342, 95-98).

Yang et al. used porous layered metal-organic framework nanosheets (MONs) with nanometer thickness as building blocks to assemble ultrathin molecular sieve membranes, which achieved H ₂ permeance of up to several thousand gas permeation units (GPUs) with H2/CO2 selectivity greater than 200. As the mechanical properties of the thin separation layer was poor, no transmembrane pressure was applied between the feed and permeate side. ( Science , 2014, 346, 1356-1359).

There is therefore a need for improved materials and membranes that provide effective gas separation.

Summary of Invention

Aspects and embodiments of the invention will now be discussed by reference to the following numbered clauses.

1. A composite material membrane comprising: a substrate having a surface; and a layer attached to the surface of the substrate, the layer comprising: pressure-responsive metal-organic framework nanosheets; and graphene oxide nanosheets, wherein a plurality of the pressure-responsive metal-organic framework nanosheets are layered on top of one another to provide a layered material.

2. The composite material membrane according to Clause 1, wherein the layer attached to the surface of the substrate is from 10 to 500 nm thick on the surface of the substrate. 3. The composite material membrane according to Clause 2, wherein the layer attached to the surface of the substrate is from 50 to 250 nm thick on the surface of the substrate, optionally wherein the layer attached to the surface of the substrate is from 75 to 150 nm thick, such as about 100 nm thick, on the surface of the substrate.

4. The composite material membrane according to any one of the preceding clauses, wherein the pressure-responsive metal-organic framework nanosheets are formed from one or more of the group selected from ZIF-7, MIL-53, Cu(benzene-1 ,4-dicarboxylate)i _/2 (4,4’- bipyridyl), Cu(pyridine-2,3-dicarboxylic acid)( 1,3-bis(4-pyridyl)propane), Zn ₂ (2,4,5-tri(4- pyridyl)-imidazole)2(d-camphorate)2, Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) and solvates thereof (e.g. aqueous solvates, such as Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’- bipyridyl)H ₂ 0).

5. The composite material membrane according to Clause 4, wherein the pressure- responsive metal-organic framework nanosheets are formed from Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) and/or its solvate Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) H ₂ 0).

6. The composite material membrane according to any one of the preceding clauses, wherein each layered material has from 2 to 10 layers, such as from 3 to 7 layers, of pressure- responsive metal-organic framework nanosheets layered on top of one another.

7. The composite material membrane according to Clause 6, wherein each layered material has from 4 to 5 layers of pressure-responsive metal-organic framework nanosheets layered on top of one another.

8. The composite material membrane according to any one of the preceding clauses, wherein each layered material has a thickness of from 2 to 20 nm, such as from 3 to 10 nm, such as about 4.4 nm, optionally wherein the layered material has a uniform thickness.

9. The composite material membrane according to any one of the preceding clauses, wherein each layered material independently has a lateral size of from 1 to 20 pm, such as from 2 to 10 pm, such as about 4 pm.

10. The composite material membrane according to any one of the preceding clauses, wherein the weight-to-weight ratio of pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets is from 10:1 to 100:1. 11. The composite material membrane according to Clause 10, wherein the weight-to- weight ratio of pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets is from 30:1 to 70:1, such as from 40:1 to 60:1.

12. The composite material membrane according to Clause 11, wherein the weight-to- weight ratio of pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets is about 50:1.

13. The composite material membrane according to any one of the preceding clauses, wherein the membrane is defect-free when viewed under scanning electron microscopy following at least one round of an activation process on the composite material membrane.

14. The composite material membrane according to any one of the preceding clauses, wherein the membrane switches from a closed state to an open state when subjected to a pressure of at least about 50 kPa CO2.

15. The composite material membrane according to any one of the preceding clauses, wherein the membrane displays a CO2 permeance of from 800 to 1200, such as from 1050 to 1100 (e.g. about 1076.7) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO2/N2, where CO2 provides 140 kPa of the pressure, optionally wherein the permeation selectivity under said conditions is from 20:1 to 25:1 (e.g. 23.1:1) in favour of CO2 over N2.

16. The composite material membrane according to any one of the preceding clauses, wherein the membrane displays a CO ₂ permeance of from 800 to 1200, such as from 1050 to 1100 (e.g. about 1051) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO ₂ /CH ₄ , where CO ₂ provides 140 kPa of the pressure, optionally wherein the permeation selectivity under said conditions is from 15:1 to 25:1 (e.g. 19.3:1) in favour of CO ₂ over ChU.

17. The composite material membrane according to any one of the preceding clauses, wherein the membrane displays a CO2 permeance of from 800 to 1200, such as from 1000 to 1050 (e.g. about 1024) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO2/H2, where CO2 provides 140 kPa of the pressure. 18. The composite material membrane according to any one of the preceding clauses, wherein the pressure-responsive metal-organic framework nanosheets are substantially provided in the form of the layered material.

19. A method of obtaining a gas enriched in CO2 from a mixed gas comprising CO2, comprising the step of supplying a mixed gas under pressure to a composite material described in any one of Clauses 1 to 18 to provide a permeated gas that is enriched in CO2 compared to the mixed gas.

Drawings

Fig. 1. Crystal structure of the flexible MOF (Cu(dhbc) ₂ (bpy) H ₂ 0) and the gate-opening membrane architecture. (A) MOF structure viewed along c axis highlighting the pore window (3.5 x 6.0 A). (B) MOF structure viewed along b axis highlighting the nonporous wall of the 1 D channels along a axis. (C) MOF structure viewed along a axis highlighting the 2D layered structure with pressure-responsive interlayer channels (3.6 ^c 4.2 A). (D) Architecture of the C0 ₂ -gate-opening MOF based membranes by confining MOF nanosheets (MONs) between graphene oxide nanosheets (GONs) so that the large pores along c axis can be mostly covered, and gas molecules are guided to diffuse mainly along a axis, whose pore opening is controlled by C0 ₂ pressure. Note that C0 ₂ molecules serve as keys to open the flexible channels of MONs.

Fig. 2. Preparation and characterizations of few-layer MONs. (A) Scheme of the preparation of few-layer Cu(dhbc)2(bpy) H ₂ 0 MONs at the air/water interface. (B) Optical image of the synthesized MONs. The insert indicates Tyndall effect of the MON suspension in isopropanol. (C) Tapping-mode AFM topographical image of MONs on silicon wafer. The insert is the height profiles of the MONs corresponding to the three marked lines. (D) Low- magnification TEM image of few-layer MONs. (E and F) High-resolution TEM (HRTEM) images of few-layer MONs. The inset is the selected area electron diffraction (SAED) pattern.

(G) The zoom-in HRTEM image overlaid with the crystal structure model along [001] zone axe.

(H) XRD patterns of the MONs in different gas atmospheres including C0 ₂ , N ₂ , and CH ₄ at atmospheric pressure (1 bar). (I) The adsorption-desorption isotherms of C0 ₂ , N ₂ , CH ₄ , and H ₂ on few-layer Cu(dhbc)2(bpy) H ₂ 0 MONs at 298 K (closed, adsorption; open, desorption).

Fig. 3. Membrane preparation and characterizations. (A) Scheme of the preparation of MON@GON membranes by compositing MONs with GONs through pressure-assisted self- assembly (PASA) method. (B) Photo of the bare CX-AI2O3 support (left), MON@GO membrane (middle), and GO membrane (right). (C) Top-view SEM image of the MON@GON-0.1-0.002 membrane. The inset is the lower magnification SEM image of the membrane. (D) Cross- sectional SEM image of the MON@GON-0.1-0.002 membrane. (E) XRD patterns of the MON@GON-0. 1-0.002 membrane at different gas atmospheres. (F) Top-view SEM image of the GOM-0.02 membrane. (G) Cross-sectional SEM image of the GOM-0.02 membrane.

Fig. 4. Pressure-responsive gas separation performance of MON@GON membranes.

(A) Three-cycle CO2 permeance, N2 permeance, and CO2/N2 perm-selectivity of MON@GON- 0.1-0.002 and GOM-0.02 membranes for CO2/N2 mixed gas as a function of CO2 partial pressure. (B) Three-cycle CO2 permeance, ChU permeance, and CO2/CH4 perm-selectivity of MON@GON-0. 1-0.002 and GOM-0.02 membranes for CO2/CH4 mixed gas as a function of CO2 partial pressure. Solid lines represent steps with increasing CO2 partial pressures, and dotted lines represent steps with decreasing CO2 partial pressures.

Fig. 5 depicts a gas separation setup for a flat membrane module.

Fig. 6. XRD patterns of the simulated, as-prepared, and activated MOF. Note that the shifts of XRD peaks to higher values in the activated MOF indicate the removal of solvent molecules and accordingly closure of the 1 D channels along a direction.

Fig. 7. XRD patterns of the activated bulk MOF in different gas atmosphere (CO2, CH ₄ , N ₂ , or H2). Note that the shifts of XRD peaks to lower values in the activated MOF in CO2 indicate the phase transition triggered by CO2 sorption and accordingly opening of the 1D channels along a direction.

Fig. 8. Scheme of the gas permeation direction along a axis in the composite membrane

Fig. 9. 298 K gas sorption isotherms of MONs prepared by freeze-thaw exfoliation (filled, adsorption; open, desorption).

Fig. 10. 298 K gas sorption isotherms of MONs prepared by ball milling and sonication exfoliation (filled, adsorption; open, desorption).

Fig. 11. 77 K N2 sorption isotherm of MONs and the calculated BET surface area (filled, adsorption; open, desorption).

Fig. 12. FTIR spectra of bulk MOF and few-layer MONs prepared at air/water interface. Fig. 13. TGA curves of bulk MOF and few-layer MONs prepared at air/water interface.

Fig. 14. TEM images of GONs.

Fig. 15. AFM images (A1 and B1) and the corresponding height lines (A2 and B2) of GONs. Fig. 16. Scheme of the internal configuration of the PASA setup for the membrane preparation.. Fig. 17. ATR-FTIR spectra of MON membrane, MON@GON membrane and GON membrane. Fig. 18. SEM image of the pure MON membrane.

Fig. 19. CO2/N2 separation performance as a function of membrane thickness.

Fig. 20. CO2/N2 separation performance as a function of MON/GON weight ratio.

Fig. 21. (A) Single-component gas permeation of MON@GON-0.1-0.002 and GOM-0.02 membranes. (B) Ideal selectivity of CO2/N2, CO2/CH4 and H2/CO2 on MON@GON-0.1-0.002 and GOM-0.02 membranes.

Fig. 22. Three-cycle CO2 permeance, H2 permeance and H2/CO2 perm-selectivity of MON@GON-0. 1-0.002 and GOM-0.02 membranes for H2/CO2 mixed gas as a function of CO2 partial pressure. Solid lines represent steps with increasing CO2 partial pressures, and dotted lines represent steps with decreasing CO2 partial pressures.

Fig. 23. (A) Comparison of CO2/N2 separation performance with other membranes. (B) Zoom- in area of dotted box in (A). The value in the bracket represents the CO2 partial pressure in unit of bar. (C) Legend of the performance points in (A) and (B) (AIChE Journal 62, 3836-3841 (2016); J. Am.Chem.Soc. 137, 1754-1757 (2015); ACS AppL Mater. Interfaces 9, 5678-5682 (2017); Nat. Mater. 18, 163-168 (2019); Nat. Commun. 8, 2107 (2017); Angew. Chem. Int. Ed. 54, 15483-15487 (2015); Adv. Mater. 29, 1701631 (2017); Nat. Energy 2, 17086 (2017))

Fig. 24. (A) Comparison of CO ₂ /CH ₄ separation performance with other membranes. (B) Zoom-in area of dotted box in (A). The value in the bracket represents the CO ₂ partial pressure in unit of bar. (C) Legend of the performance points in (A) and (B) ( AIChE Journal 62, 3836- 3841 (2016); J. Am.Chem.Soc. 137, 1754-1757 (2015); Angew. Chem. Int. Ed. 54, 15483- 15487 (2015); Adv. Mater. 29, 1701631 (2017); Nat. Energy 2 17086 (2017); J. Am.Chem.Soc. 132, 76-78 (2010); J. Mater Chem. A 2, 1239-1241 (2014); J. Mernbr Sci. 456, 185-191 (2014); J. Mernbr Sci. 533, 1-10 (2017))

Fig. 25. Effect of temperature on the separation performance of MON@GON-0.1-0.002 membrane. The molar ratio of feed gas (CO2/N2) is 1:1. Error bars represent standard deviations of three independent measurements.

Fig. 26. Effect of transmembrane pressure on the separation performance of MON@GON- 0.1-0.002 membrane. The molar ratio of feed gas (CO2/N2) is 1:1. Error bars represent standard deviations of three independent measurements.

Fig. 27. One-cycle CO2 permeance, N ₂ permeance and CO2/N2 perm-selectivity of MON@GON-0. 1-0.002 membrane for humid CO2/N2 mixed gas (relative humidity is 85%) as a function of CO2 partial pressure. Solid line represents step with increasing CO2 partial pressure, and dotted line represents step with decreasing CO2 partial pressure.

Description

It has been surprisingly found that it is possible to provide a composite material that is pressure-responsive with good mechanical properties. These composite materials may display good selectivity for one gas over other gasses, thereby allowing effective separation of a mixture of gases at reduced energy costs.

In a first aspect of the invention, there is provided a composite material membrane comprising: a substrate having a surface; and a layer attached to the surface of the substrate, the layer comprising: pressure-responsive metal-organic framework nanosheets; and graphene oxide nanosheets, wherein a plurality of the pressure-responsive metal-organic framework nanosheets are layered on top of one another to provide a layered material.

The arrangement of the metal-organic framework nanosheets and the graphene oxide nanosheets is not yet fully understood. However, it will be appreciated that the pressure- responsive metal-organic framework nanosheets are generally provided as layered materials (so as to provide the desired pressure-responsive effects). The layers of metal-organic framework nanosheets may be sandwiched between graphene oxide nanolayers, thereby providing units of a nanomaterial that may then be packed together on the surface of the substrate to provide the layer. Alternatively, the packing may be random, such that there is no ordered arrangement of the metal-organic framework nanosheets (i.e. provided as a plurality of layered materials) and the graphene oxide nanosheets. Without wishing to be bound by theory, given the method of formation disclosed herein, it is believed that the packing may be random in nature.

In embodiments herein, the word “comprising” may be interpreted as requiring the features mentioned, but not limiting the presence of other features. Alternatively, the word “comprising” may also relate to the situation where only the components/features listed are intended to be present (e.g. the word “comprising” may be replaced by the phrases “consists of” or “consists essentially of”). It is explicitly contemplated that both the broader and narrower interpretations can be applied to all aspects and embodiments of the present invention. In other words, the word “comprising” and synonyms thereof may be replaced by the phrase “consisting of” or the phrase “consists essentially of’ or synonyms thereof and vice versa.

The phrase, “consists essentially of” and its pseudonyms may be interpreted herein to refer to a material where minor impurities may be present. For example, the material may be greater than or equal to 90% pure, such as greater than 95% pure, such as greater than 97% pure, such as greater than 99% pure, such as greater than 99.9% pure, such as greater than 99.99% pure, such as greater than 99.999% pure, such as 100% pure.

The term “nanosheets” when used herein refers to layered materials with a thickness that is typically below 100 nm, and a lateral size that is greater than 50 nm. Thickness can be measured by atomic force microscopy, and lateral size can be measured by transmission electron microscopy or dynamic light scattering techniques.

When used herein “attached” may refer to direct attachment or to indirect attachment (e.g. through attachment to another material that is directly attached).

When used herein, the term “substrate” refers to any material that may be used to support a layer of the nanomaterial described herein. Examples of suitable substrates that may be mentioned include, but are not limited to inorganic (e.g. alumina oxide) or polymeric materials.

Suitable metal-organic frameworks (MOFs) that may be used to provide nanosheets in the current invention are those that have pressure-responsive features (e.g., with gate-opening pressures). Examples of such metal-organic frameworks include, but are not limited to, ZIF-7 (J. Am. Chem. Soc., 2010, 132, 17704-17706), MIL-53 {Angew. Chem. Int. Ed, 2006, 45, 7751-7754), Cu(benzene-1 ,4-dicarboxylate)i _/2 (4,4’-bpy) (Angew. Chem. Int. Ed., 2003, 42, 428-431), Cu(pydc)(bpp) (F^pydc = pyridine-2, 3-dicarboxylic acid; bpp = 1,3-bis(4- pyridyl)propane) (J. Am. Chem. Soc., 2005, 127, 17152-17153.), Zn2(Htpim)2(d-cam)2 (Htpim = 2,4,5-tri (4-pyridyl)-imidazole, d-cam = d-camphorate) ( Dalton Trans., 2017, 46, 14728- 14732), and Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl). These materials may be provided in pure form or as their solvates (e.g. aqueous solvates). Particular MOFs that may be mentioned herein include Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) and its aqueous solvate.

Each of the MOFs described herein when provided in a plurality of layers may display a unique gating pressure for each gas that the material may come into contact with. The gating pressure of the MOF may be influenced by the ambient temperature of its environment. The gating pressure may also be influenced by the presence of graphene oxide nanosheets. In general, the more strongly a particular gas is adsorbed by a MOF, the lower the gating pressure for that gas. Additionally, the lower the temperature, the lower the gating pressure required. When used herein, the term “gating pressure” is intended to refer to the point where the pressure exerted by a gas causes a shift in the structure of the plurality of layers of the MOF, resulting in a channel through the MOF that the gas can pass through (selectively). Where only a single gas is used, then the gating pressure will be the entire pressure exerted by the gas on the plurality of layers of MOF (or the composite material). When a mixture of gases is used, then the gating pressure for each gas will be determined based on the partial pressure of the gas on the plurality of layers of MOF (or the composite material).

In general, at room temperature (298 K), the gating pressure needed for a gas to pass through the plurality of layers of MOF (or the composite material) will increase in the order from CO2, CH ₄ , to N2. In the example provided herein (where the MOF used is Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) and its aqueous solvate), the gating pressure for CO2 may be about 50 kPa, the gating pressure for CFU may be about 1 ,000 kPa and the gating pressure for N2 may be 5,000 kPa. As will be appreciated, each MOF may have a unique gating pressure for each gas that it is used with, which can be readily determined by the skilled person using the teachings provided herein and their own knowledge of this field.

For the avoidance of doubt, unless otherwise stated, any values for CO2 permeance discussed herein are obtained at room temperature (298 K) and reference to a CO2 pressure refers to the partial pressure exerted by CO2 as part of a mixture of gases. The membranes can be installed into a flat membrane module or a hollow fiber membrane module. CC containing mixed gases such as CO2/N2, CO2/CH4 or CO2/H2 with different CO2 partial pressures (0-200 kPa) can be fed to the feed side of the membrane. More particularly, CC>2-containing mixed gases such as CO2/N2 or CO2/CH4 with different CO2 partial pressures (0-200 kPa) can be fed to the feed side of the membrane. Enriched CO2 can be collected from the permeate side. When the CO2 partial pressure is above 50 kPa, there will be a sudden increase in CO2 permeance because of exceeding the CO2 gating-pressure of the MOF layers. As a result, CO2/N2 or CO2/CH4 selectivity can be expected to increase as well because of the pressure-responsive feature of the membrane. A setup scheme using a flat membrane module is shown in Figure 5.

The layer attached to the surface of the substrate may have any suitable thickness. However, without wishing to be bound by theory, the thickness of the layer attached to the surface of the substrate also affect separation performance of the membranes disclosed herein. A higher thickness may improve gas selectivity at the expense of gas permeability (or permeance). In embodiments that may be mentioned herein, the thickness of the layer attached to the surface of the substrate may provide an optimized balance between gas permeance and selectivity. For example, the thickness may be from 10 to 500 nm thick on the surface of the substrate. In further embodiments, the layer attached to the surface of the substrate may be from 50 to 250 nm thick on the surface of the substrate. For example, the layer attached to the surface of the substrate may be from 75 to 150 nm thick, such as about 100 nm thick, on the surface of the substrate.

For the avoidance of doubt, it is explicitly contemplated that where a number of numerical ranges related to the same feature are cited herein, that the end points for each range are intended to be combined in any order to provide further contemplated (and implicitly disclosed) ranges.

Thus, taking the numerical ranges immediately above as an example, there is disclosed a layer having a thickness of: from 10 to 50 nm, from 10 to 75 nm, from 10 to 100 nm, from 10 to 150 nm, from 10 to 250 nm, from 10 to 500 nm; from 50 to 75 nm, from 50 to 100 nm, from 50 to 150 nm, from 50 to 250 nm, from 50 to 500 nm; from 75 to 100 nm, from 75 to 150 nm, from 75 to 250 nm, from 75 to 500 nm; from 100 to 150 nm, from 100 to 250 nm, from 100 to 500 nm; from 150 to 250 nm, from 150 to 500 nm; and from 250 to 500 nm.

As noted above, any suitable material that can be provided as a pressure-sensitive metal- organic framework nanosheet may be used herein. Examples of pressure-responsive metal- organic framework nanosheets that may be mentioned in embodiments of the invention may be selected from one or more of the group consisting of ZIF-7, MIL-53, Cu(benzene-1 ,4- dicarboxylate)i _/2 (4,4’-bipyridyl), Cu(pyridine-2,3-dicarboxylic acid)(1 ,3-bis(4-pyridyl)propane), Zn2(2,4,5-tri(4-pyridyl)-imidazole)2(d-camphorate)2, Cu(2,5-dihydroxybenzoic acid)2(4,4’- bipyridyl) and solvates thereof (e.g. aqueous solvates, such as Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) H ₂ 0). In particular embodiments of the invention that may be mentioned herein, the pressure-responsive metal-organic framework nanosheets may be formed from Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) and/or its solvate Cu(2,5-dihydroxybenzoic acid) ₂ (4,4’-bipyridyl) H ₂ 0).

As will be appreciated, the pressure-responsive metal-organic framework nanosheets form a plurality of layers in order to provide their pressure-gating effects. As such, the pressure- responsive metal-organic framework nanosheets are provided in the composite material in the form of a layered material (i.e. a plurality of layered materials). Each layered material may have any suitable number of pressure-responsive metal-organic framework nanosheets layered on top of one another. For example, each layered material may have from 2 to 10 layers, such as from 3 to 7 layers, of pressure-responsive metal-organic framework nanosheets layered on top of one another. In certain embodiments that may be mentioned herein, each layered material may have from 4 to 5 layers of pressure-responsive metal- organic framework nanosheets layered on top of one another.

As will be appreciated, the thickness of each layered material will be determined in large part by the number of layers of pressure-responsive metal-organic framework nanosheets layered on top of one another. With that in mind, the thickness of each layered material may be from 2 to 20 nm, such as from 3 to 10 nm, such as about 4.4 nm. In certain embodiments that may be mentioned herein, each layered material in the composite material may have a uniform thickness. When used herein, the term “uniform thickness” may refer to layered materials whose thickness is identical or differs fractionally from the other layered materials. For example, the thickness may differ by less than 5%, such as less than 1%, such as less than 0.5%, such as less than 0.1%, such as less than 0.01%. Each layered material (that is, the pressure-responsive metal-organic framework nanosheets layered on top of one another) may independently have a lateral size of from 1 to 20 pm, such as from 2 to 10 pm, such as about 4 pm.

Any suitable weight-to-weight ratio of the pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets may be used in the composite material disclosed herein. While not wishing to be bound by theory, it is believed that a higher weight-to-weight ratio of the pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets may improve the gas permeability at the expense of gas selectivity. In embodiments that may be mentioned herein, the weight-to-weight ratio of pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets may provide an optimized balance between gas permeance and selectivity. For example, the weight-to-weight ratio of pressure-responsive metal-organic framework nanosheets to graphene oxide nanosheets may be from 10:1 to 100:1, more particularly, the weight-to-weight ratio may be from 30:1 to 70:1 , such as from 40:1 to 60:1. Yet more particularly, the weight-to-weight ratio of the pressure-responsive metal-organic framework nanosheets to the graphene oxide nanosheets may be about 50:1.

Without wishing to be bound by theory, it is believed that the graphene oxide nanosheets guide the gas flow direction in the 2D membranes (i.e. the composite materials disclosed herein) to maximize the gas separation performance. In addition, the use of graphene oxide nanosheets may also help provide defect-free 2D membranes. Without wishing to be bound by theory, it is believed that the formation of defect-free 2D membranes disclosed herein may be influenced by the excellent mechanical properties of the graphene oxide nanosheets.

Thus, in certain embodiments of the invention, the composite material membrane disclosed herein may be defect-free when viewed under scanning electron microscopy following at least one round of an activation process on the composite material membrane. Any suitable activation process may be used herein. For example, the activation process may involve subjecting the composite material membrane to a temperature of 393 K under vacuum for a period of time.

As noted above, the composite material membranes disclosed herein may be used for gas separation. As such, the composite material membranes may have particular properties that can be switched on or off depending on the pressures provided by a particular gas. For example, the composite material membrane may be switched from a closed state to an open state when subjected to a pressure of at least about 50 kPa CO2. When the term “at least about” is used herein, the actual value may be somewhat higher or lower than the point value listed. For example, the value may be 10% higher or lower, such as 5 wt% higher or lower or, more particularly, 1% higher or lower.

The membranes disclosed herein may display any suitable CO2 permeance. For example, the membrane may display a CO2 permeance of from 800 to 1200, such as from 1050 to 1100 (e.g. about 1076.7) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO2/N2, where CO2 provides 140 kPa of the pressure. Under such conditions, the permeation selectivity may be from 20:1 to 25:1 (e.g. 23.1:1) in favour of CO2 over N2.

In embodiments that may be mentioned herein, the membrane may display a CO2 permeance of from 800 to 1200, such as from 1050 to 1100 (e.g. about 1051) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO2/CH4, where CO2 provides 140 kPa of the pressure. Under such conditions, the permeation selectivity may be from 15:1 to 25:1 (e.g. 19.3:1) in favour of CO2 over CH4.

In embodiments that may be mentioned herein, the membrane may display a CO2 permeance of from 800 to 1200, such as from 1000 to 1050 (e.g. about 1024) gas permeation units when subjected to a pressure of 200 kPa generated by a gas containing CO2/H2, where CO2 provides 140 kPa of the pressure.

As will be appreciated, the pressure-responsive metal-organic framework nanosheets in the composite material membrane are substantially provided in the form of a layered material. Without wishing to be bound by theory, it is believed that the presence of a plurality of layers of pressure-responsive metal-organic framework nanosheets laid on top of one another to create a layered material that provides the desired selectivity and pressure-responsiveness.

There are three major features to the composite materials disclosed herein. These are summarised below.

1. The composite membrane materials require two components (in addition to a substrate): metal-organic framework nanosheets and graphene oxide nanosheets. The graphene oxide nanosheets serve to increase the quality of the membrane by providing mechanical support and may ensure that the composite material membrane is defect-free. 2. The chosen metal-organic framework should be pressure-responsive, e.g., possessing gating-pressures for specific gases.

3. The chosen metal-organic framework should be able to be processed into nanosheets with large lateral size that are suitable to be fabricated into membranes. Examples of a suitable lateral size include, but are not limited to a lateral size of from 1 to 20 pm, such as from 2 to 10 pm, such as about 4 pm.

In a further aspect of the invention, there is also disclosed a method of obtaining a gas enriched in CO2 from a mixed gas comprising CO2, comprising the step of supplying a mixed gas under pressure to a composite material as described hereinbefore to provide a permeated gas that is enriched in CO2 compared to the mixed gas.

In yet a further aspect of the invention, there is provided a method of forming the composite material disclosed hereinbefore. This method may comprise the steps of mixing together graphene oxide nanosheets with pressure-responsive metal-organic framework nanosheets (i.e. substantially in the form of a layered material, where a plurality of the pressure-responsive metal-organic framework nanosheets are layered on top of one another) in a solvent and subjecting a substrate to a suitable filtration process (e.g. a PASA filtration process) to provide the desired composite material. Further details of this process are provided in the examples section hereinbelow.

Further aspects and embodiments of the invention will now be described by the non-limiting examples below.

Examples

Disclosed herein is an ultrathin membrane (thickness: ca. 100 nm) based on a CC>2-gated flexible MOF containing few-layer two-dimensional (2D) MOF nanosheets (MONs) and graphene oxide nanosheets (GONs). This material has been found effective for CO2 separation. In addition, the membrane possesses CC>2-induced gate opening behaviour with a sharp increase of CO2 permeance and CO2/N2 and CO2/CH4 selectivities.

Materials CU(N0 ₃ ) ₂ 3H ₂ 0 was purchased from Merck KGaA, Germany. Both 4,4’-bipyridine (4,4’-bpy, > 98.0%) and 2,5-dihydroxybenzoic acid (Hdhbc, > 98.0%) were purchased from TCI. Graphene oxide (GO) powder with lateral sizes of about 5 pm was purchased from Nanjing XFNANO Materials Tech Co. Ltd. Diethyl ether (AR), isopropanol (AR), and ethanol (AR) were purchased from Fisher Scientific. The deionized water produced by Millipore-Q System (Millipore, Billerica, MA, USA) was used in all experiments. H2 (> 99.999%), CO2 (> 99.999%), N2 (> 99.999%), CH4 (> 99.999%) and Ar (> 99.999%) were purchased from Air Liquide in Singapore. Polytetrafluoroethylene dish (PTFE, 6 cm in diameter) was obtained from Latech Scientific Supply Pte. Ltd, Singapore. Porous CX-AI2O3 disks with diameter of 25 mm, thickness of 1.0 mm and top layer pore size of 70 nm were purchased from Fraunhofer Institute IKTS, Germany. Anodic aluminum oxide (AAO) filters with support ring, pore size of 20 nm, diameter of 25 mm, and thickness of 60 pm were purchased from Whatman.

Characterizations

Crystal structure was characterized by X-ray diffraction (XRD) on an X-ray powder diffractometer (Rigaku MiniFlex 600) equipped with a Cu sealed tube (l = 1.5418 A) at a scan rate of 0.5 min. To confirm the structure phase state of the powder or membrane in different gas atmospheres, the sample was loaded into a gas chamber to conduct the XRD test under different gas atmospheres (CO2, N2, ChU, or H2). Gas sorption isotherms were measured using a surface area and pore size analyser (Micromeritics ASAP 2020). Before the measurements, the samples were activated by degassing under high vacuum at 120 °C for 10 h. High-pressure sorption isotherms were conducted on a Quantachrome iSorb HP1 instrument. Before the measurements, the samples were activated by degassing under high vacuum at 120 °C for 10 h. The morphologies of MOFs and membranes were investigated by field emission scanning electron microscopy (FESEM, JSM-6700, JEOL) under 20 kV. Before observation, all the samples were sputtered with Pt by Sputter Coater (Cressington 208 HR) under a current of 20 mA for 120 s. Transmission electron microscopy (TEM) and selected area electron diffraction (SAED) of samples on carbon-coated TEM grids were conducted on a JEOL JEM- 3010 transmission electron microscope. Atomic force microscopy (AFM) of the nanosheets deposited on silica wafer was conducted using a Bruker Dimension Icon atomic force microscope with tapping mode. Attenuated total reflectance-Flourier transformed infrared (ATR-FTIR) spectra were obtained with a VERTEX 70 FTIR spectrometer (Bruker). Thermogravimetric analyses (TGA) were performed using a thermogravimetric analyser (Shimadzu DTG-60AH) in the temperature range of 20-800 °C with a heating rate of 5 °C/min under a flowing N2 atmosphere (50 mL min ^-1 ). Analysis of MOF

We chose a 2D flexible MOF, Cu(dhbc)2(bpy)· H2O (dhbc: 2,5-dihydroxybenzoic acid; bpy: 4,4'- bipyridyl), developed by Kitagawa et al. in 2003 ( Angew . Chem. Int. Ed. 42, 428-431 (2003)), as the smart membrane material. It has a 2D layered structure, low CC>2-gate-opening pressure (0.55 bar), and facile synthesis. Its structure comprises one-dimensional (1D) Cu- bpy chains further linked by dhbc ligands into a rigid 2D sheet motif (Fig. 1A). When viewed along b axis, the pore size of the channel wall is too small to allow gas molecules to penetrate (Fig. 1 B). The sheet motifs form interlocking ridges and valleys constructed by the dhbc benzene rings, which are mutually interdigitated to create 1D channels along a axis with a cross-sectional channel size of 3.6 ^c 4.2 A (Fig. 1C). The gliding motion of tt-stacked dhbc benzene rings mediates the transition between ‘closed and ‘open’ states, which can be triggered by various adsorbed guests at different pressures. Consequently, the gate-opening behavior of the bulk layered MOF was readily observed in the CO2, CH ₄ , and N ₂ sorption isotherms with gate-opening pressures of about 0.55, 10, and 50 bar for CO2, CH ₄ , and N ₂ , respectively. Besides, the low Brunauer-Emmett-Teller (BET) specific surface area (38 m ² g ^_1 ) of this MOF calculated based on the nitrogen sorption isotherm at 77 K further confirms that it is difficult for nitrogen molecules (molecular size: 3.64 A) to diffuse into the micropores of this MOF.

X-ray diffraction (XRD) results indicate that there are two pronounced peak shifts of the (1 0 - 2) and (1 0 2) lattice planes to a higher angle region between 15° and 20° as a result of dehydration (Fig. 6), suggesting a transition from ‘open’ to ‘closed’ state. Consequently, XRD characterization can serve as a clear approach to provide direct insights regarding the state of this flexible MOF, i.e., in ‘open’ or ‘closed’ form. When exposing the bulk MOF powders in different gas atmospheres (including CO2, N2, CH ₄ , and H2) in an enclosed sample chamber (1 bar of absolute pressure), the XRD results show that only CO2 atmosphere can trigger the structural change of this MOF into ‘open’ state (Fig. 7).

From the structural analysis, two issues are inherent in the fabrication of pure 2D Cu(dhbc)2(bpy) MOF membranes: (1) The windows size of 3.5 ^c 6.0 A within ab planes is too large to sieve small gases, let alone to realize the pressure-controlled gate-opening separation; (2) The volume change caused by structural transformation of gate-opening MOFs will lead to the generation of non-selective membrane defects, which is detrimental to the separation performance. Herein we propose a solution of compositing few-layer MONs with GONs. Without wishing to be bound by theory, it is believed that the large windows of MONs along c axis can be mostly covered by GONs. It is believed that in such arrangement, gas molecules will be guided to diffuse mainly along a axis, whose pore opening can be controlled by C0 ₂ pressure so that the gate-opening effect would occur (Fig. 1 D, Fig. 7).

General Procedure 1 : Bottom-up method for preparation of few-layer MONs (Cu(dhbc) ₂ (bpy)

H ₂ 0 nanosheets) at air/water interface

Few-layer metal-organic framework nanosheets (MONs) were prepared by a bottom-up method adapted from Nat. Mater. 9, 565-571 (2010), where the reaction occurred at an air- water interface at room temperature.

In a typical procedure, Cu(N0 ₃ )2-3H ₂ 0 aqueous solution (20 mmol-L ^-1 , 8 ml_) was poured into a polytetrafluoroethylene dish (6 cm in diameter) as the bottom phase. Diethyl ether solutions containing 4,4’-bpy (2.5 mmol-L· ¹ , 25 pl_) and Hdhbc (10 mmol-L· ¹ , 25 mI_) were layered consecutively on the aqueous bottom phase. The polytetrafluoroethylene dish was covered by a glass cover, and the reaction was left for 3 days. The few-layer MONs formed at the air/water interface were collected by draining the underlying aqueous solution with a syringe. The MONs were re-dispersed in isopropanol at a concentration of 0.02 mg mL· ¹ for membrane preparation.

Example 1 : Characterization and sorption properties of MONs

Few-layer Cu(dhbc) ₂ (bpy)-H ₂ 0 MONS were prepared in accordance with General Procedure 1 (Fig. 2A). The obtained nanosheets were dispersed in isopropanol (0.02 mg mL· ¹ ), forming a colloidal suspension with a clear Tyndall effect (insert in Fig. 2B).

Characterization

A layered morphology could be clearly seen from the optical image (Fig. 2B) and atomic force microscopy (AFM) image (Fig. 2C) with a lateral size of about 4 pm and uniform thickness of approximately 4.4 nm (corresponding to 4-5 layers). Low-magnification transmission electron microscopy (TEM) result also confirmed the layered morphology (Fig. 2D). The MOF crystallinity was observed in the few-layer MONs, as demonstrated by high-resolution TEM (HRTEM) and selected area electron diffraction (SAED) as shown in Fig. 2, E and F. In addition, the zoom-in atomic arrangement in the HRTEM image is consistent with the crystal structure model along [001] zone axe (Fig. 2G), revealing the successful preparation of Cu(dhbc) ₂ (bpy)-H ₂ 0 MONs at the air/water interface.

Sorption properties

The structural transition and sorption properties caused by gate-opening of the obtained MONs were characterized by XRD and gas sorption tests. The XRD pattern of MONs is consistent with the simulated and bulk Cu(dhbc)2(bpy) H ₂ 0 XRD patterns, confirming consistent structures between MONs and the bulk MOF (Fig. 2H). Identical to the bulk MOF, there are also peak shifts of crystal planes (1 0 -2) and (1 0 2) in MONs to a higher angle region between 15° and 20° as a result of the CO2 exposure, indicating a structural transformation from ‘closed to ‘open’ form. In comparison, N2 and CFU cannot trigger this phase transition under similar conditions. CO2 sorption isotherms confirm that the gate opening pressure of CO2 is about 0.55 bar in the MONs, and there is no gate-opening phenomenon of N ₂ and CH ₄ with negligible gas uptakes within the tested pressure range (below 1 bar, Fig. 2I). Other tests including 77 K N2 sorption (Fig. 11), Fourier transform infrared (FTIR) spectroscopy (Fig. 12), and thermogravimetric analysis (TGA, Fig. 13) all confirm the structure and behaviour of the MONs are in line with the bulk MOF.

Comparative Example 1 : Preparation of few-layer Cu(dhbc)2(bpy) H ₂ 0 MONs by freeze- thaw or ball milling and sonication exfoliation, and effect on sorption properties

Bulk Cu(dhbc) ₂ (bpy)-H ₂ 0 MOF crystals were prepared according to a reported method (Angew. Chem. Int. Ed. 2003, 42, 428-431). The MOF crystals were dispersed in isopropanol with a concentration of 1 mg-mL· ¹ and heated in a hot water bath (353 K) for 5 min, and then immediately frozen in liquid nitrogen bath (77 K) until it was in a completely frozen state. The frozen solid was subsequently thawed in hot water bath, and the freeze-thaw cycle was repeated 5 times. The exfoliated MONs were collected from the middle one-third MOF suspension after being placed undisturbed for two weeks. The exfoliation yield was about 2% with a concentration of 0.02 mg-mL· ¹ through weighing a certain volume of MOF suspension after being deposited on AAO support. The gas sorption isotherms of the resulting MONS prepared by freeze-thaw exfoliation method is depicted in Fig. 9.

Fig. 10 depicts the gas sorption isotherms of MONs prepared by a reported ball milling and sonication exfoliation method (Science 2014, 346, 1356-1359). The above results indicate that traditional top-down exfoliation methods such as mechanical milling, ultrasound, and freeze-thaw are time and labour consuming with low yield, often leading to poor control on nanosheet size, and sometimes even structural and functional damage on nanosheets.

General Procedure 2: Preparation of MON(3)GON membranes

Firstly, a suspension of graphene oxide nanosheets (GONs, 0.01 mg mL· ¹ ) was prepared by sonicating graphene oxide (GO) powder dispersed in water for 30 min. The obtained GONs have thicknesses of about 0.8-1 nm and lateral dimensions of about 5 pm (Figs. 14 and 15), confirming their single-layer feature.

A pressure-assisted self-assembly (PASA) filtration technique was used to construct MON@GON membranes (Fig. 16). Certain volumes of MON suspension (prepared according to General Procedure 1) and GON suspension (0.05-2.0 ml_) were mixed in isopropanol to form a total suspension with a volume of 100 ml_, which was shaken for 5 min to reach a homogeneous mixing. A support of anodic aluminium oxide (AAO, pore size: 20 nm, diameter: 25 mm, Whatman) or CX-AI2O3 (top layer pore size: 70 nm, diameter: 25 mm, Fraunhofer Institute IKTS, German) was installed in the membrane-preparation setup. The AAO substrate was used for membrane characterization due to its small thickness, while the CX-AI2O3 substrate was used for membrane performance evaluation as it is more mechanically stable. Both substrates have pore sizes that are much larger than that of the composite layer. PASA filtration processes were operated at a constant pressure difference of 1 bar until all the suspension was filtrated out. The obtained membrane was dried successively at room temperature for 12 h and 393 K for 12 h in a vacuum oven. The resulting composite membranes were denoted as MON@GON-x-y, where x and y represent the mass of MONs and GONs (in mg) used in the membrane fabrication, respectively.

Pure MON membrane and GON membrane were fabricated using the same procedure except for the absence of GONs and MONs, respectively.

Example 2: Preparation and Characterization of MON@GON membranes

A pressure-assisted self-assembly (PASA, Fig. 16) method as described in General Procedure 2 was used to fabricate the composite membrane containing MONs and GONs (Fig. 3A). The fabricated membranes were denoted as MON@GON-x-y, where x and y represent the mass of MONs and GONs (in mg) used in the membrane fabrication, respectively.

The PASA approach achieved uniform surface coverage of MONs and GONs as evidenced by the even brown colouration on the originally white AI2O3 substrate. Furthermore, the MON@GON was visually darker than the pristine GO membrane (Fig. 3B). Other characterizations such as attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) also demonstrate the successful preparation of MON@GON membranes (Fig. 17). The top-view SEM images of MON@GON-0.1-0.002 membrane at both high and low magnifications show an absence of obvious defects (Fig. 3C). Based on the cross-sectional FESEM images, the thickness of the MON@GON-0.1-0.002 membrane is around 100 nm, indicating successful preparation of ultrathin membranes (Fig. 3D).

We estimated the relative composition of the MON@GON-0.1-0.002 membrane by preparing a pure GO membrane. Using ten times the mass of GO used for the MON@GON-0.1-0.002 membrane, a defect-free membrane was obtained with a cross-sectional thickness of 80 nm (Fig. 3, F and G). This suggests that 8 nm of the measured thickness for MON@GON-O.1- 0.002 derives from GONs, with the remainder (ca. 92 nm) ascribed to the stacking of MONs. Considering the individual thicknesses of MON and GON (4.4 and 0.8 nm, respectively), the layer numbers of MON and GON in MON@GON-0.1-0.002 membrane were estimated to be about 20 and 10, respectively. Thus, we readily conclude that the composition of the MON@GON-0. 1-0.002 membrane is dominated by MONs. The XRD results indicate that the MON@GON membrane exhibits MON-derived responsive behavior toward CO2 (Fig. 3E).

Comparative Example 2: Preparation and Characterization of pure MON membrane

A pure MON membrane was also fabricated in accordance with General Procedure 2 except that GONs was not added. Obvious cracks can be easily identified in the pure MON membrane upon activation (Fig. 18). Activation involves subjecting the pure MON membrane under high vacuum at 120 °C for 10 h. We postulate that the volume contraction of the flexible MONs during activation leads to these cracks. Without wishing to be bound by theory, it is believed that the GON component suppress the formation of defects during structural transformation in the MON@GON membranes and increase the chance of fabricating defect-free membranes.

Example 3: Pressure-response gas separation performance of MON@GON membranes A Wicke-Kallenbach setup was used to evaluate the single gas permeation and mixed gas separation performance of the membranes fabricated in Example 2 and Comparative Example 2. Notably, an absolute pressure difference of 1 bar across the membrane was ensured during the tests to avoid any possible back flow of the sweep Ar gas. High gas permeance (CO ₂ permeance > 40,000 GPU (Gas Permeation Unit; 1 GPU = 3.348x10 ¹⁰ mol nr ² s ^_1 Pa ^-1 )) with no selectivity for any of the studied gas pairs was confirmed using bare porous CX-AI ₂ O ₃ support, which is because of its relatively large pore size (70 nm).

Gas separation performance test

Both single gas permeation and mixed gas separation performance were tested using a homemade Wicke-Kallenbach gas permeation apparatus described previously (Nat. Commun. 8, 14460 (2017); J. Am.Chem.Soc. 142, 4472-4480 (2020)). A silicon rubber gasket was used to seal the membranes and the permeation cell. In order to avoid the potential damage of membrane sample when directly contacting the silicon rubber gasket, the membrane was masked with an aluminium gasket with exposing a 10-mm-diameter hole in the center. The volumetric flow rates of feed gas were controlled by mass flow controllers (MFC, D07-26C, SevenStar, China). Argon with a volumetric flow rate of 50 ml_ min ^-1 was controlled by MFC (D07-19C) and used as the sweep gas. In both single-component gas permeation tests and mixed-gas separation tests, the trans-membrane pressure difference was kept at 1 bar by using a back-pressure regulator at the retentate side, i.e. , the feed pressure was fixed at 2 bar of absolute pressure, and the sweep pressure was 1 bar. The molar concentrations of permeate side gas were analysed by gas chromatograph (GC-2014, Shimadzu) with two TCD detectors. The permeation data were recorded at steady state when the composition concentrations of permeate side gas analysed by GC were constant. Every data point was tested at least 3 times to ensure reproducibility, and the relative error was below 15%.

For single gas permeation test, the gas test sequence was H2, N2, CFU, and CO2. After one gas permeation test, an activation process (393 K under vacuum) was implemented. For mixed gas separation test, a CO2 partial pressure ranging from 0-2.0 bar was set by tuning the volumetric flow rate of each single component in the gas pairs. Meanwhile, high gas flow rates (total 60 ml_ min ^·1 ) were used to ensure the membrane stage cut (denoted as the flow- rate ratio of permeate gas to feed gas) was always below 1%.

The gas permeance (R, GPU, 1 GPU = 3.348 ^c 10 ¹⁰ mol/m ² sPa) of the component gas / was calculated by equation (1), where L/, is the molar flow rate of the gas / (mol/s), A is the membrane effective area (m ² ), and DR, is the partial pressure of the component / (Pa).

The ideal selectivity or mixed gas perm-selectivity (a) of gas pair ///was calculated by equation

(2) where P, and P ₇ are the permeance of gas / and gas j, respectively.

Effect of membrane thickness and MON/GON weight ratio on separation performance

MON@GON-0. 1:0.0033 (MON/GON mass ratio of 30) and MON@GON-0.1:0.0014 (MON/GON mass ratio of 70) membranes were prepared according to General Procedure 2 by varying the mass of MONs and GONs added accordingly.

Membranes of different thickness were prepared according to General Procedure 2 with a MON/GON mass ratio of 50 but with different total mass of MON and GON. When the membrane thickness is 75 nm, the MON and GON mass are 0.075 mg and 0.0015 mg, respectively. When the membrane thickness is 130 nm, the MON and GON mass are 0.13 mg and 0.0026 mg, respectively.

As shown by the gas separation performance results (Fig. 19 and 20), the MON@GON-O.1- 0.002 membrane exhibited an optimized balance between gas permeance and selectivity. In separation tests where the membrane thickness was varied, the MON/GON weight ratio was fixed at 50, and the volume ratio of feed CO2/N2 mixed gas is 1:1. In separation tests where the MON/GON weight ratio was varied, the membrane thickness was kept at about 100 nm, and the volume ratio of feed CO2/N2 mixed gas is 1:1.

Single-component gas permeation separation performance

A sharp cut-off between H2 and CO2 can be observed in both MON@GON-0.1-0.002 and GOM-0.02 (Fig. 21A). This cut-off can be attributed to the difference in kinetic diameter of the two gases. A second cut-off between CO2 versus N2 and CFU was observed in MON@GON- 0.1-0.002. High CO2 permeance (1144 GPU) and high ideal selectivities for CO2/N2 (28.6) and CO2/CH4 (23.8) were obtained in MON@GON-0.1-0.002 (Fig. 21 B), making this membrane a promising candidate for C0 ₂ -related gas separations such as post-combustion carbon capture (CO2/N2 separation) and natural gas upgrading (CO2/CH4 separation). Without wishing to be bound by theory, it is believed that the high CO2 permeance of this membrane come from the CO2- induced phase transition such that the gas diffusion channels along a direction of the MON component is fully opened under 2.0 bar of pure CO2. The permeance of N2 and CH4 are very low in both membranes (MON@GON-0.1-0.002 and GOM-0.02) because of their large molecular size and inability to trigger phase transition under 2.0 bar of feed pressure.

Effect of CO ₂ partial pressure on mixed-qas separation performance

The pressure responsive gas separation performance of MON@GON-0.1-0.002 was further evaluated using mixed gas feed with various CO2 partial pressures. Keeping the total pressure constant at 2.0 bar, the CO2 partial pressure was increased from 0.1 to 2.0 bar, and then decreased back to 0.1 bar to complete one test cycle. In total, three consecutive cycles were performed each for CO2/N2 and CO2/CH4 separation (Fig. 4). The control GOM-0.02 membrane barely showed any dependence of permeance and perm-selectivity on CO2 partial pressure. In contrast, MON@GON-0.1-0.002 membrane underwent a sudden jump of CO2 permeance when the CO2 partial pressure was increased above 0.5 bar, which is also the gate-opening pressure of CO2 for the few-layer MONs. Thus, the opening of dynamic 1 D MOF channels along a direction under interaction with CO2 led to enhanced gas permeance. Consequently, when CO2 partial pressure was increased from 0.1 to 1.4 bar, the CO2 permeance of MON@GON-0.1-0.002 membrane jumped from 173.8 to 1076.7 GPU for CO2/N2 pair, and from 188 to 1051 GPU for CO2/CH4 pair. Meanwhile, although permeance of CH4 or N2 also increased due to phase transition, the magnitude of increase was substantially lower (Fig. 4). Overall, the perm-selectivities increased from 4.4 to 23.1 for CO2/N2 pair, and from 4 to 19.3 for CO2/CH4 pair.

Interestingly, at the decreasing CO2 partial pressure step, there is an unusual hysteresis phenomenon for CO2 permeance as well as CO2/CH4 and CO2/N2 perm-selectivities for the MON@GON-0. 1-0.002 membrane. Such hysteresis loops are similar to CO2 sorption isotherms of the few-layer MONs, and can be ascribed to the stabilizing effect of the adsorbed CO2 molecules to the ‘open’ form of the MON so that it can survive even below the gating pressure (ca. 0.5 bar of CO2 partial pressure). Similar C0 ₂ -induced gating effect was observed in the H2/CO2 pair, leading to increased CO2 permeance (from 136.1 to 1024 GPU) but reduced H2/CO2 perm-selectivity (from 11.7 to 2.4, Fig. 22). The gas separation performance and hysteresis loops could be repeated in the continuous three-cycle runs, confirming the reversible flexibility and stability of MON@GON membranes. Because of the C0 ₂ -inducted gating effect, the CO2/N2 and CO2/CH4 separation performance of the MON@GON-0.1-0.002 membrane can be substantially improved to approach the Robeson upper bounds (2008 revision, Figs. 23 and 24).

Effects of temperature, pressure, and humidity on separation performance

As shown in Fig. 25-27, MON@GON membranes display stable performance under various conditions and thus have great potential in practical applications.

• As the temperature increases, both CO2 and N2 permeance increases due to the enhanced diffusion activated at higher temperatures. However, CO2 permeance increases slower than that of N ₂ , causing a slight decrease of CO2/N2 perm-selectivity (Fig. 25).

• The membrane is stable when transmembrane pressure is below 1.5 bar, but started to fail at 2.0 bar and above (Fig. 26).

• When humid mixed gas (relative humidity: 85%) was used as the feed, water molecules blocked the channel pores, resulting in reduced gas permeance (Fig. 27).

Conclusion

In summary, we herein demonstrate the first C0 ₂ -gated ultrathin 2D membrane by compositing few-layer 2D MONs and GONs for C0 ₂ -related gas separations. Large and well- structured MONs were obtained by an interfacial reaction, which can be easily scaled-up. Without wishing to be bound by theory, it is believed that the use of GONs guide the gas flow directions in the 2D membranes for maximized gate-opening performance. In addition, it is believed that the chance of fabricating defect-free 2D membranes can be greatly increased in the presence of GONs. The optimized membrane assembled with well-aligned MONs and GONs by PASA method exhibited an unprecedented C0 ₂ -responsive gas separation performance with reversible hysteretic cycles, which can be attributed to the C02-responsive ‘gate-opening/closing’ property of the few-layer MONs. This work extends the application of flexible MOFs in membrane separation, and suggests the possibility of fabricating ‘smart membranes’ as molecular switches and gates for various applications.