Composite material and method of preparing thereof

Abstract

The invention provides a composite material comprising a porous substrate, a polycrystalline material layer disposed on a surface of the porous substrate, wherein the polycrystalline material layer comprises a plurality of intergrown crystals formed from a UiO-66 metal- organic framework (MOF), or derivative thereof and wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks indicative of reo-structured regions within the UiO-66 MOF, or derivative thereof. The invention also provides a method of preparing the composite material comprising the steps of providing the porous substrate and a reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt and immersing the porous substrate in the reaction mixture from step (a) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C for a period of time to provide the composite material.

Description

[0001] COMPOSITE MATERIAL AND METHOD OF PREPARING THEREOF

[0002] FIELD OF INVENTION

[0003] The present invention provides a composite material, more particularly, a composite material comprising a metal-organic framework (MOF), and a method of preparing thereof. The present invention also provides a method of liquid-phase separation using the composite material.

[0004] BACKGROUND

[0005] 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.

[0006] Chemical processes are being improved by new-generation separation technologies featuring low energy expenditures and high processing efficiency. One intriguing candidate is membrane technology, which provides a sustainable, energy-efficient and low-carbon- footprint option for producing high-quality chemical raw materials. Organic solvent nanofiltration (OSN), which capitalizes on solvent-resistant membranes to separate molecules in organic liquids, is of practical and scientific interest. However, although traditional membranes show OSN potential, their performance often suffers from considerable trade-offs between permeance and selectivity. Many of these membranes are also susceptible to performance deterioration over time because of solvent swelling and physical aging of their labile building materials. Challenges in precisely regulating membrane pore size and geometry also limit their separation precision and application diversity. However, the demand for OSN membranes with a sharp and tailorable selectivity capable of discriminating complex non-aqueous mixtures has grown substantially in chemical processes, including catalyst recovery, pharmaceutical upgrading and crude-oil fractionation. State-of-the-art membranes could selectively separate large molecules from complex mixtures, but often fail in the scenario of fine targets due to their large and size-distributed pores. Thus, developing molecular-sieving membranes that can achieve separation of fine and similarly sized molecules with favorable solvent permeance is imperative for separation engineering.

[0007] New materials with regularly arranged and spatially connected micropores provide access to selective molecular-sieving membranes. Recently, high-order structural assembly of building blocks has generated exceptionally selective membranes, but the synthetic protocols can lack universality. Periodic porous crystalline materials, exemplified by metal-organic frameworks (MOFs) and covalent organic frameworks (COFs), offer distinct advantages for advancing separation membranes due to their pore uniformity and robust structures. In the past several years, precisely engineered crystalline porous membranes have played a crucial role in tackling bottleneck issues in many energy and environmental areas. However, the challenges that hamper highly selective molecular sieving persist, including mismatched pore sizes and insufficient crystallinity. By contrast, prominently crystallized MOF membranes with regular subnanometer-sized pores are promising candidates for the effective sieving of fine targets. MOF membranes with designed pore chemistry have demonstrated success in separating industrially valuable gases and ions. Nevertheless, their implementation in liquid-phase molecular separations remains rare, and insights into the fundamentals of mass transport through polycrystalline MOF nanofiltration membranes remain underexplored. In particular, compared with nanometer-thick polymer membranes, existing MOF membranes suffer from large thicknesses (typically micrometers), resulting in inferior solvent permeances far below the benchmark of OSN membranes. Moreover, the design principles for MOF-based OSN membranes to achieve fast and selective molecular sieving have yet to be fully determined.

[0008] Structural engineering of MOFs to finetune the pore structures and chemistry at the atomic level opens up an avenue for membrane design. Zirconium-based UiO-66 and its derivatives have emerged as preferred candidates because of their chemical robustness and high synthetic reproducibility. The archetypical structural defects in UiO-66 include the loss of organic linkers and Zr clusters, which often coexist. Despite numerous studies on creating missing-linker defects in UiO-66 (Tan, K. et al. Defect termination in the UiO-66 family of metalorganic frameworks: the role of water and modulator. J. Am. Chem. Soc. 143, 6328- 6332 (2021 ); Feng, L. et al. Creating hierarchical pores by controlled linker thermolysis in multivariate metal-organic frameworks. J. Am. Chem. Soc. 140, 2363-2372 (2018)), little attention has been paid to missing-cluster defects, which hold potential to artificially manipulate the pore microenvironments. In particular, for UiO-66 with a defective reo topology, its regular, large-sized internal cavities arising from uniformly distributed missingcluster defects could improve solvent permeance at minimal expense of selectivity.

[0009] Thus, there is a need for alternative and / or improved membranes / composite materials and methods of preparing thereof as well as methods of separation, in particular, liquid-phase separation. SUMMARY

[0010] Aspects and embodiments of the current invention will now be described by reference to the following numbered clauses.

[0011] 1 . A composite material comprising: a porous substrate; a polycrystalline material layer disposed on a surface of the porous substrate; wherein the polycrystalline material layer comprises a plurality of intergrown crystals formed from a UiO-66 metal-organic framework (MOF), or derivative thereof; wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks indicative of reo-structured regions within the UiO-66 MOF, or derivative thereof.

[0012] 2. The composite material of clause 1 , wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks at 26 = about 4.3° and about 6.1 ° (e.g. 4.3° ± 0.2° and 6.1 ° ± 0.2°).

[0013] 3. The composite material of clause 2, wherein the powder X-ray diffractogram has further characteristic peaks at 26 = about 7.3° and about 8.6° (e.g. 7.3° ± 0.2° and 8.6° ± 0.2°).

[0014] 4. The composite material of any one of clauses 1 to 3, wherein the polycrystalline material layer is substantially free from intercrystalline gaps.

[0015] 5. The composite material of any one of the preceding clauses, wherein the polycrystalline material layer consists essentially of the plurality of intergrown crystals.

[0016] 6. The composite material of any one of the preceding clauses, wherein the plurality of intergrown crystals have a surface area of from about 500 m2g1to about 2000 m2g1, for example from about 500 m2g1to about 2000 m2g1, from about 500 m2g1to about 1200 m2g1, from about 800 m2g1to about 1100 m2g1, or from about 900 m2g1to about 1000 m2g1. 7. The composite material of any one of the preceding clauses, wherein the plurality of intergrown crystals comprise pores with a pore volume of from about 0.3 cm2g1to about 1 cm2g-1, for example from about 0.4 cm2g1to about 0.6 cm2g1.

[0017] 8. The composite material any one of the preceding clauses, wherein the plurality of intergrown crystals comprise pores with an average pore diameter of from about 1 nm to about 1 .5 nm, for example about 1 .2 nm.

[0018] 9. The composite material of any one of the preceding clauses, wherein the thickness of the polycrystalline material layer is from about 200 nm to about 400 nm, for example from about 200 to about 300 nm (e.g. from about 250 to about 270 nm).

[0019] 10. The composite material of any one of the preceding clauses, wherein the composite material has a surface average roughness of from about 20 nm to about 100 nm, for example from about 20 nm to about 40 nm, about 30 nm to about 40 nm or about 30 nm to about 35 nm.

[0020] 11 . The composite material of any one of the preceding clauses, wherein the composite material displays a permeance of greater than 3.4 L m- 2h-1bar-1for methanol and a dye rejection rate of greater than about 90% for Evans Blue (i.e. tetrasodium (6E,6'E)-6,6- [(3,3'-dimethylbiphenyl-4,4'-diyl)di(1 E)hydrazin-2-yl-1 -ylidene]bis(4-amino-5-oxo-5,6- dihydronaphthalene-1 ,3-disulfonate).

[0021] 12. The composite material of any one of the preceding clauses, wherein the composite material displays a permeance of from about 5 L m- 2h-1bar-1to about 12 L m- 2h1bar-1for methanol, and a dye rejection rate of greater than about 95% for Evans Blue, for example a dye rejection rate of greater than 96%, of greater than 97%, of greater than 98%, or of greater than 99% for Evans Blue.

[0022] 13. The composite material of any one of the preceding clauses, wherein the porous substrate is selected from the group consisting of a ceramic (e.g. alumina), a carbon cloth, a metal, and a metal oxide.

[0023] 14. The composite material of any one of the preceding clauses, wherein the porous substrate is provided in the form of a tube, a mesh, a sheet, or hollow fibers (e.g. porous alumina hollow fibers), or other arrangements that are obtainable by the folding of a tube, a mesh, a sheet, or hollow fibers. 15. The composite material of any one of the preceding clauses, wherein the substrate is an activated alumina substrate.

[0024] 16. The composite material of any one of the preceding clauses, wherein the porous substrate has a first surface and second surface, wherein the polycrystalline material layer is disposed on the first surface, and the second surface is substantially free from the polycrystalline material layer formed from the UiO-66 MOF, or derivative thereof.

[0025] 17. The composite material of any one of the preceding clauses, wherein the composite material is suitable for use as a filter material in organic solvent nanofiltration.

[0026] 18. The composite material of any one of the preceding clauses, wherein the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis using a reaction mixture comprising an organic solvent, 1 ,4-benzene- dicarboxylate (BDC) or derivative thereof, a Zr salt and a Zn salt, and wherein the solvothermal synthesis was performed at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C).

[0027] 19. The composite material of clause 18, wherein after the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis, the composite material was subjected to washing with DMF, methanol, or a mixture thereof.

[0028] 20. The composite material of clause 18 or 19, wherein the Zn salt and the Zr salt are present in the reaction mixture at molar ratio of Zn to Zr of from about 0.1 to about 1 , for example from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5).

[0029] 21. The composite material of any one of clauses 18 to 20, wherein the organic solvent is dimethylformamide (DMF).

[0030] 22. The composite material of any one of clauses 18 to 21 , wherein the derivative of BDC is 2-aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

[0031] 23. The composite material of any one of the preceding clauses, wherein the derivative of UiO-66 MOF is UiO-66-NH2or UiO-66-OH. A method of preparing the composite material of any one of the preceding clauses, the method comprising the steps of:

[0032] (a) providing the porous substrate and a reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;

[0033] (b) immersing the porous substrate in the reaction mixture from step (a) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for a period of time to provide the composite material. The method according to clause 24, wherein step (b) comprises heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for about 12 to about 72 h (e.g. about 24 h) to provide the composite material. The method according to clause 24 or 25, wherein the Zn salt and the Zr salt are present in the reaction mixture in step (a) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5). The method according to any one of clauses 24 to 26, wherein the organic solvent in the reaction mixture of step (a) is DMF. The method according to any one of clauses 24 to 27, wherein the reaction mixture in step (a) further comprises an acid, for example HCI. The method according to any one of clauses 24 to 28, wherein following step (b), the method further comprises the steps of:

[0034] (c) washing the composite material with an organic solvent;

[0035] (d) providing a second reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;

[0036] (e) immersing the composite material in the second reaction mixture from step (d) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. about 200 °C) for a period of time to provide the composite material (e.g. for about 12 to about 72 h, for example, about 24 h); and

[0037] (f) optionally washing the composite material following step (e) with an organic solvent. 30. The method according to clause 29, wherein:

[0038] - the organic solvent in step c) and / or (f) is DMF, methanol, or a mixture thereof; and / or

[0039] - the organic solvent in step d) is DMF; and / or

[0040] - the Zn salt and the Zr salt are present in the second reaction mixture in step (d) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5); and / or

[0041] - the second reaction mixture in step (d) further comprises an acid, for example HOL

[0042] 31 . The method according to any one of clauses 24 to 30, or the composite material of any one of clauses 18 to 23, wherein the Zn salt is Zn(NO3)2.

[0043] 32. The method according to any one of clauses 24 to 31 , or the composite material of any one of clauses 18 to 23 or 31 , wherein the Zr salt is ZrCh.

[0044] 33. The method according to any one of clauses 24 to 32, wherein the derivative of BDC is 2-aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

[0045] 34. A composite material obtained by the method according to any one of clauses 24 to 33.

[0046] 35. A method of removing a solute from a solvent, the method comprising passing the solvent through the composite material of any one of clauses 1 to 23 or through the composite material of clause 34.

[0047] 36. The method according to clause 35, wherein the solute has a molecular weight of greater than about 300 g mol'1.

[0048] 37. The method according to clause 35 or 36, wherein the solvent is water or an organic solvent selected from the group consisting of a monohydric alcohol (e.g. ethanol), a hydrocarbon (e.g. hexane), an ether (e.g. tetrahydrofuran), a ketone (e.g. acetone), and a mixture of two or more thereof.

[0049] BRIEF DESCRIPTION OF DRAWINGS FIG. 1 depicts the structural determination of UiO-66. a, Schematic illustration of Zr cluster, BDC linker, feu - and reo-UIO-66. b, PXRD patterns of the experimental and simulated UiO- 66. c, Relative peak intensity of the (100) plane as a function of Zn / Zr molar ratio for biUiO- 66-100. d, 77 K N2adsorption-desorption isotherms (filled symbols, adsorption; open symbols, desorption), e, Pore size distribution data calculated based on the N2isotherms. Note that the (100) peak intensity of biUiO-66-100 samples was normalized by that of biUiO- 66-100-1.

[0050] FIG. 2 depicts the structural characterization of UiO-66. a, TGA curves of the UiO-66 samples, b, O 1 s XPS spectra of the UiO-66 samples, c, Zn XPS spectra of bi UiO-66- 100-1 . d, EDS mapping of UiO-66-200 and biUiO-66-100-1 without solvent washing.

[0051] FIG. 3 depicts the synthesis and characterization of UiO-66 membranes, a, Schematic illustration of the synthesis of UiO-66 membranes. b,c, Surface SEM images of the biUiO- 66-200-0 (b) and biUiO-66-200-0.4 (c) membranes. d,e, AFM images of the biUiO-66-200-0 (d) and biUiO-66-200-0.4 (e) membranes. f,g, Cross-sectional SEM images of the biUIO-66- 200-0 (f) and biUiO-66-200-0.4 (g) membranes. h,i, Cross-sectional TEM image (h) and EDS mapping (i) of the biUiO-66-200-0.4 membrane, j, PXRD patterns of the biUIO-66-200- 0 and biUiO-66-200-0.4 membranes. Note that the UiO-66 layers in f and g are artificially colored to highlight the contrast.

[0052] FIG. 4 depicts the performance evaluation of UiO-66 membranes, a, Separation performance as a function of Zn / Zr molar ratio (data are presented as mean ± s.d. of n = 3-5 replicates), b, Molecular rejection profiles of biUiO-66 membranes (data are presented as mean ± s.d. of n = 3-5 replicates), c, Performance comparison based on MeOH permeance and dye MWCO. d, Separation performance as a function of operation duration (n = 1 replicate), e, Schematic illustration for selective sieving of fine molecules with similar sizes, f, Structures of the fine probes for molecular sieving. MO1 , mordant orange 1 ; BB, basic blue 17; MO, methyl orange; Pa, 1 ,4-phenylenediamine; NP, p-nitrophenol; PTSA, p- toluenesulfonic acid, g, UV-vis spectra of the selective separation of a mixture of NP and MO. h, Molecular sieving of the NP / MO mixture as a function of operation duration (n = 1 replicate). Details of the performance comparison in c are available in Table 3.

[0053] Fig. 5 depicts the demonstration of the biUiO-66-200-0.2 membrane in practical applications, a, Molecular structures of the studied APIs, b, UV-vis spectra for the separation of rifampicin. Inset: photograph of the feed and filtrate during the separation of rifampicin, c, Rejection rates of various APIs (n = 1 replicate), d, UV-vis spectra for the separation of Ru-BINAP. Inset: molecular structure of Ru-BINAP.

[0054] FIG. 6 depicts the demonstration of a tubular membrane module, a, Schematic of the tubular membrane module, b, Schematic of the continuous crossflow filtration. c,d, Surface (c) and cross-sectional (d) SEM images of the tubular membrane. Note that the UiO-66 layer in d is artificially colored to highlight the contrast, e, Molecular rejection profile of the tubular membrane (data are presented as mean ± s.d. of n = 3 replicates), f, Separation performance as a function of operation duration (n = 1 replicate), g, Radar chart for the qualitative comparison of crystallization, chromatography and membrane separation. Details of the comparison in g are available in Table 7.

[0055] FIG. 7 includes the PXRD patterns of the UiO-66 powders synthesized at various temperatures: (a) 2-theta range of 2.5°-40°: (b) 2-theta range of 2.5°-7°. Note: The synthesized samples were designated as UiO-66-x, in which x stands for the synthetic temperature.

[0056] FIG. 8 includes the PXRD patterns of the biUiO-66 powders synthesized with various Zn / Zr molar ratios: (a) 2-theta range of 2.5°-40°; (b) 2-theta range of 2.5°-7°. Note: The synthesized samples were designated as biUiO-66-x-y, in which x and y stand for the synthetic temperature and the molar ratio of Zn(NO3)2 and ZrCk (Zn / Zr), respectively.

[0057] FIG. 9 includes the PXRD patterns of biUiO-66-100-1 and UiO-66-100 samples synthesized without adding Zn salts.

[0058] FIG. 10 includes the PXRD pattern of biUiO-66-200-1 .

[0059] FIG. 11 depicts (a) Properties of the ligand and metal ions based on Pearson’s hard / soft acid / base theory, (b) PXRD patterns of the UiO-66-100 and biUiO-66-100-1 powders synthesized with Zn(NO3)2 replaced by various metal salts.

[0060] FIG. 12 depicts (a) 77 K N2adsorption-desorption isotherms of UiO-66-85. (b) Pore size distribution of UiO-66-85.

[0061] FIG. 13 depicts the BET surface areas and pore volumes of the UiO-66 samples. FIG. 14 includes the SEM images of (a) UiO-66-200, (b) UiO-66-100, and (c) biUiO-66-100-1 . Note: The UiO-66 samples were grown on alumina support for SEM observations.

[0062] FIG. 15 depicts the mass weight of the raw materials and the synthesized biUiO-66-100 samples as a function of Zn / Zr molar ratio.

[0063] FIG. 16 includes the full-range XPS spectra of UiO-66-200, UiO-66-100, and biUiO-66-100-1 .

[0064] FIG. 17 includes the illustration for the calculation of ideal TGA plateau normalized by the end weight of ZrOs at 800 °C.

[0065] FIG. 18 includes the PXRD patterns of biUiO-66-100-1 subjected to various solvent treatments.

[0066] FIG. 19 includes the full-range XPS spectra of biUiO-66-100-1 with various solvent treatments.

[0067] FIG. 20 includes the PXRD patterns of biUiO-66-100-1 before and after being treated with 0.1 M HCI.

[0068] FIG. 21 includes the PXRD pattern of the biUiO-66-100-1 membrane.

[0069] FIG. 22 includes the surface SEM images of the biUiO-66-100-1 membrane.

[0070] FIG. 23 includes the surface SEM images of the alumina substrate.

[0071] FIG. 24 includes the SEM images of the billiO-66-200-0.2 membrane.

[0072] FIG. 25 includes the cross-sectional SEM images of the (a) billiO-66-200-0, (b) bil)iO-66- 200-0.2, and (c) biUiO-66-200-0.4 membranes.

[0073] FIG. 26 includes the cross-sectional SEM image of the biUiO-66-200-0.2 membrane.

[0074] FIG. 27 includes the cross-sectional TEM image of the biUiO-66-200-0.4 membrane.

[0075] FIG. 28 includes the PXRD patterns of UiO-66-200, biUiO-66-200-0.2, and biUiO-66-200-0.4. FIG. 29 depicts (a) 77 K N2adsorption-desorption isotherms and (b) pore size distribution data of UiO-66-200, biUiO-66-200-0.2, and biUiO-66-200-0.4.

[0076] FIG. 30 includes the TGA curves of UiO-200, biUiO-66-200-0.2, and biUiO-66-200-0.4.

[0077] FIG. 31 includes the PXRD patterns of the biUiO-66-200-0 membrane after solvent treatments.

[0078] FIG. 32 includes the PXRD patterns of the billiO-66-200-0 membrane after being treated with organic acid and alkaline solvents.

[0079] FIG. 33 includes the SEM images of the biUiO-66-200-0 membrane after solvent treatments in (a) water, (b) MeOH, and (c) DMF for 3 months.

[0080] FIG. 34 includes the SEM images of the biUiO-66-200-0 membrane after being treated with (a) MeOH (7 mL) + acetic acid (3 mL) and (b) MeOH (7 mL) + triethylamine (3 mL) for 3 days.

[0081] FIG. 35 includes the SEM images of the biUiO-66-200-0 membrane before and after exposure to flowing solvents, (a) Original, (b) water, (c) MeOH, and (d) DMF.

[0082] FIG. 36 includes the PXRD patterns of the biUiO-66-200-0 membrane before and after exposure to flowing solvents.

[0083] FIG. 37 includes the1H NMR spectra of the UiO-66 and the dried solutions subjected to various flowing solvents.

[0084] FIG. 38 includes the UV-vis spectra for the separation of EB by the alumina substrate.

[0085] FIG. 39 depicts the comparison of the thickness of UiO-66 membranes. Note: Details of the comparison are given in Table 2.

[0086] FIG. 40 depicts the water uptakes of UiO-66-200 and biUiO-66-100-1 .

[0087] FIG. 41 depicts the water contact angles of the biUiO-66-200-0 and biUiO-66-200-0.4 membranes as a function of testing duration FIG. 42 depicts (a, b) Simulation cell of feu- and reo-UiO-66. (c) Mean square displacement of MeOH in feu- and reo-UiO-66.

[0088] FIG. 43 depicts the molecular structures of the dyes used in this work.

[0089] FIG. 44 includes the UV-vis spectra for the separation of (a) MO, (b) CV, (c) AF, (d) OR, and (e) EB by the billiO-66-200-0 membrane.

[0090] FIG. 45 includes the UV-vis spectra for the separation of (a) MO, (b) CV, (c) AF, (d) OR, and (e) EB by the biUiO-66-200-0.2 membrane.

[0091] FIG. 46 includes the UV-vis spectra for the separation of (a) MO, (b) CV, (c) AF, (d) CR, and (e) EB by the biUiO-66-200-0.4 membrane.

[0092] FIG. 47 depicts the prediction of solute rejections by the steric hindrance pore, Verniory, Ferry, and modified Ferry transport models: (a) Predicted and experimental solute rejections as a function of molecular weight; (b) Predicted and experimental solute rejections as a function of molecular weight with the rejection rate in the range of 0-100%.

[0093] FIG. 48 includes the photographs of the biUiO-66-200-0.2 membrane (a) before and (b) after the long-term filtration test. Note: The membrane surface remained clean after long-term filtration, showing insignificant adsorption and fouling.

[0094] FIG. 49 includes the UV-vis spectra of the feed, filtrate, and retentate during the separation of EB. Note: The volumes of the feed, filtrate, and retentate were about 12, 2, and 10 mL, respectively. The adsorption rate was calculated to be -1.3%.

[0095] FIG. 50 includes the UV-vis spectra for the static adsorption of (a) MO, (b) CV, (c) AF, (d) CR, and (e) EB by the biUiO-66-200-0 powders.

[0096] FIG. 51 depicts the zeta potential values of UIO-66 and molecules.

[0097] FIG. 52 depicts the EB rejection of the biUiO-66-200-0.2 membrane under various transmembrane pressures.

[0098] FIG. 53 depicts the comparison of the operation duration of our membrane and others. Note: Details of the comparison are given in Table 4. FIG. 54 includes the molecular structures of the ultrafine molecules used for sieving tests.

[0099] FIG. 55 includes the UV-vis spectra for the molecular sieving of (a) Pa / MO, (b) NP / MO1 , (c) NP / BB, and (d) PTSA / MO mixtures by the biUiO-66-200-0.2 membrane.

[0100] FIG. 56 includes the schematic diagram of the correlation between estimated molecular diameters and window pore size.

[0101] FIG. 57 depicts the comparison of the membrane selectivity toward ultrafine molecules. Note: Details of the comparison are given in Table 5.

[0102] FIG. 58 includes the UV-vis spectra for the separation of (a) curcumin, (b) tetracycline, (c) VB12, and (d) spiramycin.

[0103] FIG. 59 includes (a) Photograph of the tubular membrane module, (b) Photograph of the module details.

[0104] FIG. 60 includes (a-c) Surface and (d) cross-sectional SEM images of the alumina tube.

[0105] FIG. 61 includes the UV-vis spectra of the selective separation of NP / MO mixture through the tubular biUiO-66-200-0.2 membrane.

[0106] FIG. 62 includes the photograph of the tubular membrane before and after continuous filtration for 20 days.

[0107] DESCRIPTION

[0108] The present inventors have developed a robust and accurate molecular-sieving membrane created through the topological design of a metal-organic framework (e.g., UiO-66) for use in durable liquid phase separations. The membrane is prepared by crystallizing the metalorganic framework using a bimetallic method, which yields distinctive reo-topology frameworks with periodic missing-cluster defects (e.g., reo-UiO-66). The membrane exhibits improved and robust performance, lasting for over 1 ,500 h, as well as excellent membrane selectivity to accurately discriminate fine complex mixtures with molecular weights below 350 g mol-1. In addition, the membrane demonstrates applications in purifying and recovering high-value pharmaceuticals and catalysts. Thus, in a first aspect of the invention, there is provided a composite material comprising: a porous substrate; a polycrystalline material layer disposed on a surface of the porous substrate; wherein the polycrystalline material layer comprises a plurality of intergrown crystals formed from a UiO-66 metal-organic framework (MOF), or derivative thereof; wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks indicative of reo-structured regions within the UiO-66 MOF, or derivative thereof.

[0109] 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.

[0110] As used herein, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise.

[0111] As used herein, the term “reo-structured regions” refers to regions of the UiO-66 MOF (or derivative thereof) that have ordered missing-cluster defects. The presence of reo-structured regions within UiO-66 MOF (or derivative thereof) may be characterised by a powder X-ray diffractogram with characteristic peaks at 26 between from about 4° to about 6.5°, for example about 4.3° and about 6.1 ° (e.g. 4.3° ± 0.2° and 6.1 ° ± 0.2°). The detection of such characteristic peaks indicate that the reo-structured regions are regularly distributed through the UiO-66 MOF (or derivative thereof).

[0112] Therefore, in certain embodiments, the plurality of intergrown crystals may be characterized by a powder X-ray diffractogram with characteristic peaks at 26 = about 4.3° and about 6.1 ° (e.g. 4.3° ± 0.2° and 6.1 ° ± 0.2°). In more particular embodiments, the powder X-ray diffractogram may have further characteristic peaks at 26 = about 7.3° and about 8.6° (e.g. 7.3° ± 0.2° and 8.6° ± 0.2°). The term “about” as used herein can allow for a degree of variability in a value or range, for example, within 10%, within 5%, within 1%, within 0.5%, within 0.1 %, within 0.05%, within 0.01 %, within 0.005%, or within 0.001% of a stated value or of a stated limit of a range, and includes the exact stated value or range.

[0113] As demonstrated in the Examples section of the present disclosure, the method of preparing the present composite material is capable of producing highly-crystallized and well- intergrown membranes free of obvious grain boundary gaps. As such, in certain embodiments, the polycrystalline material layer may be substantially free from intercrystalline gaps. In addition, in certain embodiments, the polycrystalline material layer may consist essentially of the plurality of intergrown crystals.

[0114] 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.

[0115] In certain embodiments, the plurality of intergrown crystals may have a surface area of from about 500 m2g1to about 2000 m2g1, for example from about 500 m2g1to about 2000 m2g1, from about 500 m2g1to about 1200 m2g1, from about 800 m2g1to about 1100 m2g-1, or from about 900 m2g1to about 1000 m2g1. In certain embodiments, the plurality of intergrown crystals may comprise pores with a pore volume of from about 0.3 cm2g1to about 1 cm2g1, for example from about 0 4 cm2g1to about 0.6 cm2g1. In certain embodiments, the plurality of intergrown crystals may comprise pores with an average pore diameter of from about 1 nm to about 1 .5 nm, for example about 1 .2 nm.

[0116] The surface area of materials disclosed herein is determined from nitrogen adsorption measurements at 77 K using the Brunauer-Emmett-Teller (BET) method. See J. Am. Chem. Soc. 1938, 60, 309-331 , which is incorporated herein by reference. The pore volume and pore size of the material may also be calculated from N2adsorption isotherms measured at 77 K, as determined using BET.

[0117] As used herein, the term “pore size” refers to the cavity size of the reo-structured regions in the UiO-66 MOF, or derivative thereof. 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.

[0118] In certain embodiments, the thickness of the polycrystalline material layer may be from about 200 nm to about 400 nm, for example from about 200 to about 300 nm (e.g. from about 250 to about 270 nm). Methods for measuring the thickness of the polycrystalline material layer are known in the art, and include for example, Scanning Electron Microscopy (SEM).

[0119] In certain embodiments, the composite material may have a surface average roughness of from about 20 nm to about 100 nm, for example from about 20 nm to about 40 nm, about 30 nm to about 40 nm or about 30 nm to about 35 nm. As used herein, the term “surface average roughness” refers to the average of surface heights and depths across a given surface. Methods for measuring surface average roughness of a surface are known in the art, and include for example, Atomic Force Microscopy (AFM).

[0120] As demonstrated in the Examples section of the present disclosure, the present composite material exhibits excellent performance compared to commercial / conventional membranes, including improved permeance and sharp selectivity to fine molecules. As such, in certain embodiments, the composite material may display a permeance of greater than 3.4 L nr2h1bar-1for methanol and a dye rejection rate of greater than about 90% for Evans Blue (i.e. tetrasodium (6E,6'E)-6,6-[(3,3'-dimethylbiphenyl-4,4'-diyl)di(1 E)hydrazin-2-yl-1 -ylidene]bis(4- amino-5-oxo-5,6-dihydronaphthalene-1 ,3-disulfonate). In addition, in certain embodiments, the composite material may display a permeance of from about 5 L m- 2h-1bar-1to about 12 L rrr2h-1bar-1for methanol, and a dye rejection rate of greater than about 95% for Evans Blue, for example a dye rejection rate of greater than 96%, of greater than 97%, of greater than 98%, or of greater than 99% for Evans Blue.

[0121] The term “permeance” refers to the volume of a liquid (e.g. methanol) that can flow through composite material of the present invention per unit area of composite material, per unit time, and per unit of pressure. Typically, the methanol permeance of the composite material of the present invention is expressed in units of L m- 2h-1bar-1, unless otherwise stated.

[0122] The porous substrate can be in any suitable form, which include, but is not limited to, tubes, meshes, sheets and hollow fibers (e.g. porous alumina ceramic hollow fibers), plus other forms that can be obtained by the folding of these primary forms. As such, in certain embodiments, the porous substrate may be provided in the form of a tube, a mesh, a sheet, or hollow fibers (e.g. porous alumina hollow fibers), or other arrangements that are obtainable by the folding of a tube, a mesh, a sheet, or hollow fibers.

[0123] Examples of substrates in the form of sheets include, but are not limited to, carbon film / cloth. Meshes may include, but are not limited to, metal meshes and metal oxide meshes. Hollow fiber structures that may be mentioned herein include, but are not limited to ceramics (e.g. alumina) and zirconia hollow fiber. As such, in certain embodiments, the porous substrate may be selected from the group consisting of a ceramic (e.g. alumina), a carbon cloth, a metal, and a metal oxide.

[0124] It is noted that certain substrates (e.g. carbon films / cloths or stainless steel meshes) can be functionalized by carboxylation or amination, which can facilitate the growth of crystal seeds. In addition, the flexibility of carbon films / cloths as substrates can offer good mechanical properties to the resultant membranes.

[0125] The substrate may be a porous AI2O3 substrate, optionally wherein the substrate is a porous C1-AI2O3 substrate. An example of a suitable porous a-A^Os substrate are porous asymmetric C1-AI2O3 supports with a maximum pore size of 70 nm, a diameter (for the substrate) of 18 mm, and a thickness of 1 mm, which may be purchased from the Fraunhofer Institut fur Keramische Technologien and Systeme (IKTS), Germany. As will be appreciated, materials with different pore sizes may also be used, for example the maximum pore size may be from 30 nm to 200 nm. As described herein, the composite material of the present invention may be formed by solvothermal synthesis of UiO-66 MOF (or derivative thereof) on a surface of a porous substrate using a reaction mixture comprising an organic solvent, 1 ,4-benzene- dicarboxylate (BDC) or derivative thereof, a Zr salt and a Zn salt. Typically, the solvothermal synthesis is performed at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C), and typically under ambient pressure. As such, in certain embodiments, the substrate may be an activated alumina substrate.

[0126] In certain embodiments, the porous substrate may have a first surface and second surface, wherein the polycrystalline material layer is disposed on the first surface, and the second surface is substantially free from the polycrystalline material layer formed from the UiO-66 MOF, or derivative thereof.

[0127] In certain embodiments, the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis using a reaction mixture comprising an organic solvent, 1 ,4-benzene-dicarboxylate (BDC) or derivative thereof, a Zr salt and a Zn salt, and wherein the solvothermal synthesis was performed at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C).

[0128] In certain embodiments, the Zn salt and the Zr salt may be present in the reaction mixture at molar ratio of Zn to Zr of from about 0.1 to about 1 , for example from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5).

[0129] Any suitable organic solvent may be used. In certain embodiments, the organic solvent may be dimethylformamide (DMF).

[0130] As demonstrated in the Examples section of the present disclosure, washing procedures affect the determination of diffraction peaks from reo structures using PXRD testing. As such, in certain embodiments, after the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis, the composite material may be subjected to washing with DMF, methanol, or a mixture thereof.

[0131] Derivatives of 1 ,4-benzene-dicarboxylate (BDC) include compounds comprising a core structure corresponding to 1 ,4-benzene-dicarboxylate wherein the benzene ring includes one or more substituents at the 2, 3, 5 and / or 6 positions. For example, substituents such as halogen (e.g. F or Cl), NH2 and OH. Example of derivatives of BDC include 2- aminobenzene-1 ,4-dicarboxylic acid (ABDC) and 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH). As such, in certain embodiments, the derivative of BDC may be 2- aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

[0132] Derivatives of UiO-66 MOF include MOFs formed from a Zr salt and a derivative of 1 ,4- benzene-dicarboxylate (BDC). Examples of UiO-66 MOF include UIO-66-NH2 or UiO-66-OH. UiO-66-NH2 is formed from a Zr salt and 2-aminobenzene-1 ,4-dicarboxylic acid (ABDC). UiO-66-OH is formed from a Zr salt and 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC- OH). As such, in certain embodiments, the derivative of UiO-66 MOF may be UiO-66-NH2 or UiO-66-OH.

[0133] As demonstrated in the Examples section of the present disclosure, the present composite material is useful for liquid phase separations, including organic solvent nanofiltration. As such, in certain embodiments, the composite material may be used as a filter material in organic solvent nanofiltration. As mentioned above, the present inventors have developed a bimetallic method of topologically designing UiO-66 membranes with rich reo structures by using Zn salts in the crystallization of UiO-66 as a modulator to produce reo-topology frameworks with long-range ordered missing-cluster defects.

[0134] Thus, in a second aspect of the invention, there is provided a method of preparing the composite material of any one of the preceding claims, the method comprising the steps of:

[0135] (a) providing the porous substrate and a reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;

[0136] (b) immersing the porous substrate in the reaction mixture from step (a) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for a period of time to provide the composite material.

[0137] Without wishing to be bound by theory, the addition of soft Lewis-acid metal ions (such as Zn2+, Co2+and Cd2+, see FIG. 11 ) can effectively induce reo structures. In contrast, the use of hard Lewis-acid metal ions (such as Al3+, Fe3+and Zr4+) is ineffective. According to Pearson’s hard / soft acid / base theory, interactions between the hard Lewis-base BDC and hard Lewis-acid metal ions are stronger than those between BDC and soft Lewis-acid metal ions. As such, the presence of weak interactions between BDC and soft Lewis-acid metal ions might appropriately compromise the coordination between BDC and Zr, giving rise to reo structures.

[0138] Furthermore, as demonstrated in the Examples section of the present disclosure, the reo- structured biUiO-66 could be synthesized at a high temperature (e.g. 200 °C, see FIG. 10) with the aid of bimetallic regulation.

[0139] In certain embodiments, step (b) may comprise heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for about 12 to about 72 h (e.g. about 24 h) to provide the composite material.

[0140] In certain embodiments, the Zn salt and the Zr salt may be present in the reaction mixture in step (a) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5).

[0141] Any suitable organic solvent may be used. In certain embodiments, the organic solvent in the reaction mixture of step (a) may be DMF. In certain embodiments, the reaction mixture in step (a) may further comprise an acid, for example HCI.

[0142] The obtained membrane from steps (a) and (b) may be used washed for secondary growth. As such, in certain embodiments, following step (b), the method may further comprise the steps of:

[0143] (c) washing the composite material with an organic solvent;

[0144] (d) providing a second reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;

[0145] (e) immersing the composite material in the second reaction mixture from step (d) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. about 200 °C) for a period of time to provide the composite material (e.g. for about 12 to about 72 h, for example, about 24 h); and

[0146] (f) optionally washing the composite material following step (e) with an organic solvent.

[0147] In more particular embodiments, one or more of the following may apply:

[0148] - the organic solvent in step c) and / or (f) is DMF, methanol, or a mixture thereof; and / or

[0149] - the organic solvent in step d) is DMF; and / or

[0150] - the Zn salt and the Zr salt are present in the second reaction mixture in step (d) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5); and / or

[0151] - the second reaction mixture in step (d) further comprises an acid, for example HCI.

[0152] Any suitable Zn salt may be used. In certain embodiments, the Zn salt may be Zn(NO3)2. Any suitable Zr salt may be used. In certain embodiments, the Zr salt may be ZrCI4.

[0153] As mentioned above, the derivative of BDC is 2-aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

[0154] In a third aspect of the invention, there is provided a composite material obtained by the method disclosed hereinbefore. As details of the composite material have already been described above, they are omitted here for brevity. As demonstrated in the Examples section of the present disclosure, the present composite material is useful for liquid phase separations, including organic solvent nanofiltration. Thus, in a fourth aspect of the invention, there is provided a removing a solute from a solvent, the method comprising passing the solvent through the composite material as disclosed hereinbefore.

[0155] Advantageously, the present composite material is able to achieve precise sieving of small molecules with molecular weights below 350 g mol-1. As such, in certain embodiments, the solute may have a molecular weight of greater than about 300 g mol-1.

[0156] In certain embodiments, the solvent may be water or an organic solvent selected from the group consisting of a monohydric alcohol (e.g. ethanol), a hydrocarbon (e.g. hexane), an ether (e.g. tetrahydrofuran), a ketone (e.g. acetone), and a mixture of two or more thereof.

[0157] The invention has been described broadly and generically herein. Those of ordinary skill in the art will readily appreciate that all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and / or configurations will depend upon the specific application or applications for which the teachings of the present invention is / are used. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. It is, therefore, to be understood that the foregoing embodiments are presented by way of example only and that, within the scope of the statements of invention and equivalents thereto, the invention may be practiced otherwise than as specifically described. The present invention is directed to each individual feature, system, article, material, kit, and / or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and / or methods, if such features, systems, articles, materials, kits, and / or methods are not mutually inconsistent, is included within the scope of the present invention. Further, each of the narrower species and subgeneric groupings falling within the generic disclosure also form part of the invention. This includes the generic description of the invention with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

[0158] Further aspects and embodiments of the invention will now be discussed by reference to the following non-limiting examples.

[0159] EXAMPLES Materials and Methods

[0160] Materials: All the chemicals and reagents were obtained from commercial suppliers and used as received. 1 ,4-Dicarboxybenzene (BDC, 99%), p-toluenesulfonic acid (PTSA, 98%), 1 ,4-phenylenediamine (Pa, 98%), p-nitrophenol (NP, 99%), mordant orange 1 (MO1 , 98%), methyl orange (98%), crystal violet (CV, 98%), Congo red (OR, 98%), Evans blue (EB, 80%), basic blue 17 (BB, 98%), triethylamine (99%), acetic acid (99.5%), and hydrochloric acid (HCI, 37%) were provided by Tokyo Chemical Industry. Zirconium(IV) chloride (ZrCk, 98%) was supplied by Aladdin. Zinc nitrate hexahydrate (Zn(NO3)2-6H2O, 98%), zinc chloride (ZnCh, 98%), zinc acetate (Zn(CH3COO)2, 98%), zinc carbonate basic (ZnCO3, >58%), cobalt nitrate hexahydrate (Co(N03)2-6H20, 98%), cadmium nitrate tetrahydrate (Cd(NO3)2-4H2O, 98%), aluminum nitrate nonahydrate (AI(NO3)3-9H2O, 98%), and iron nitrate nonahydrate (Fe(NO3)3-9H2O, 98%) were obtained from Sigma-Aldrich. [(R)-2,2'- bis(diphenylphosphino)-1 ,1 '-binaphthyl]ruthenium(ll) dichloride (Ru-BINAP, 90%) and active pharmaceutical ingredients (APIs), including spiramycin (90%), tetracycline (95%), rifampicin (98%), curcumin (97%), and Vitamin B12 (VB12, 95%), were supplied by Tokyo Chemical Industry. N,N-dimethylformamide (DMF, 99.5%) and methanol (MeOH, 99.9%) were purchased from Fisher Chemical. Porous asymmetric alumina supports with a top y-alumina layer (pore size: ~50-200 nm) were obtained from FoShan Yirun Ceramic New Material Co. Ltd. Porous asymmetric alumina tubes were purchased from Nanjing Membrane Materials Industrial Technology Research Institute Co, Ltd. Deionized (DI) water produced by Elga Micromeg Wall Mounted Deionizer System (Cartridge type MC:DS) was used in all experiments.

[0161] Characterizations: Scanning electron microscopy (SEM) measurements were performed on a JEOL JSM-7610FPIus scanning electron microscope at an accelerating voltage of 5 kV. The samples were sputter-coated with a thin layer of Pt by a sputter coater (Cressington 208 HR) at a current of 20 mA for 120 s to prevent charging. Atomic force microscopy (AFM) measurements were carried out on a Broker Dimension Icon AFM microscope. Powder X- ray diffraction (PXRD) patterns were recorded on a Rigaku MiniFlex 600 X-ray powder diffractometer with Cu Ka radiation (A = 1 .5406 A) at a scan rate of 2° min1. N2adsorptiondesorption isotherms were obtained using a Micromeritics ASAP 2020 surface area and porosity analyzer at 77 K. The samples were degassed at 120 °C for 24 h before the measurements. X-ray photoelectron spectroscopy (XPS) measurements were performed on a Kratos Axis Ultra DLD X-ray photoelectron spectrometer with an excitation source of Al Ka = 1486.71 eV. The binding energies were calibrated by referencing C 1 s to 284.5 eV. Thermogravimetric analysis (TGA) was performed on a Shimadzu DTG-60AH analyzer from room temperature to 800 °C at a heating rate of 10 °C min1under a constant and simultaneous flow of air (50 mL min1). Zeta potential values of the molecules and UiO-66 in MeOH were tested using a Malvern Zetasizer Nano ZSP analyzer. The cross-sectional membrane was prepared by FEI DA300 focused ion beam (FIB) using Ga ion. The sample was coated with a protective layer of Pt before FIB cutting. Transmission electron microscopy (TEM) analyses were performed using a 200 kV microscope (Tecnai X-TWIN, FEI). Dynamic water contact angles were measured by a contact angle goniometer (OCA 15EC, Dataphysics). Water sorption tests of powders were conducted by a dynamic vapor sorption analyzer (IGAsorp, Hiden Isochema) at room temperature.

[0162] Synthesis of UiO-66 powders: BDC (2 mmol, 332 mg) and ZrCk (1 mmol, 233 mg) were charged into a 20-mL glass vial, followed by the addition of DMF (12 mL) and HCI (0.2 mL). The prepared mixture was ultrasonically treated to form a clear and transparent solution. Subsequently, the solution was transferred into a Teflon lining with a stainless-steel cover. The autoclave was placed in a preheated oven with designated temperatures for 24 h. Finally, the generated white powders were collected through centrifugation, thoroughly washed with DMF and MeOH for several days, and then dried at 85 °C under vacuum for 24 h. The neat UiO-66 solids free of solvent washing were also prepared to study washing treatments. It is noted that fresh HCI should be used for the synthesis.

[0163] Activation of alumina supports: The alumina supports were immersed in 2 M HCI aqueous solution for 24 h. The treated alumina supports were thoroughly washed with DI water and vacuum dried at 85 °C for at least 24 h.

[0164] Synthesis of tubular biUiO-66 membranes: The original tubes with a length of 10 cm were washed with water and ethanol, followed by vacuum drying at 80 °C overnight. The surfaces at both ends of the tubes were sealed with glaze. BDC (15.6 mmol, 2.5896 g) and metal salts (7.8 mmol) with a Zn / Zr molar ratio of 0.2 were charged into a glass bottle, followed by the addition of DMF (143 mL) and HCI (1.3 mL). The prepared mixture was ultrasonically treated to form a clear and transparent solution. Subsequently, the solution was transferred into a Teflon lining with a stainless-steel cover. A homemade Teflon shelf preloaded with 5 pieces of tubes was introduced into the autoclave lining. The autoclave was placed in a 200 °C oven for 24 h. The obtained membranes were fully washed with DMF for secondary growth, which was performed the same as the first growth. The synthesized tubular membranes were completely washed with DMF and MeOH for several days before tests. It is noted that fresh HCI should be used for the synthesis. Stability tests of biUiO-66 membranes: The bil)iO-66-200-0 membrane was soaked in DI water, MeOH, and DMF for 3 months to probe the solvent stability. The membrane was also subjected to 10-mL MeOH solutions containing 3 mL acetic acid or triethylamine. After these treatments, the membrane was thoroughly washed with MeOH and vacuum dried at 85 °C for 24 h before further characterization. The biUiO-66-200-0 membrane was also assembled into a crossflow filtration cell with circulating solvents at a flow rate of -30 mL min'1. The solutions produced from each solvent were completely dried and analyzed. Quantitative elemental analysis was performed with inductively coupled plasma-optical emission spectrometry (ICP-OES) on a Perkin Elmer Avio 500 ICP spectrometer. 1 H nuclear magnetic resonance (NMR) spectroscopy was performed on a Bruker Avance 400 MHz NMR spectrometer (DRX400).

[0165] Performance evaluation of biUIO-66 membranes: Pure solvent permeance and organic molecule rejection were tested to evaluate membrane performance. The synthesized membranes were sealed into a stainless-steel dead-end filtration cell with an effective area of 0.78 cm2. The filtration tests were conducted under stirring at 300-400 rpm to weaken the concentration polarization. The volumes of the used feed and the collected filtrate were at least 50 mL and 10 mL, respectively. The tubular membrane with an effective area of -18.8 cm2was assembled into a crossflow filtration module, where the feed solution circulated continuously. Before filtration tests, the membranes were pre-pressed at 2 bar for at least 2 h to achieve a steady condition. The solvent permeation flux (J, L rm2h-1) and permeance (P, L nr2h-1bar-1) of membranes were calculated by the following equations:

[0166] J = V / (A At) (1 )

[0167] P = J / AP (2) where V (L) is the volume of the collected permeate, A (m2) is the effective membrane area, At (h) is the permeation duration, and AP (bar) is the transmembrane pressure.

[0168] Dye molecules and APIs dissolved in MeOH with a concentration of 10 ppm and 25 ppm, respectively, were used as the feed to evaluate organic solvent nanofiltration performance. The filtration tests were conducted under stirring at 300-400 rpm to weaken the concentration polarization. The dye and API concentrations were analyzed by a UV-vis absorption spectrometer (Cary 60, Agilent). The rejection rate (R, %) and adsorption rate (Ads, %) were calculated by the following equations:

[0169] R = (3)

[0170] Ads (4) where Cf, Cr, and Cprepresent the solute concentrations of the feed, retentate, and permeate, respectively; Vf, Vr, and Vprepresent the volumes of the feed, retentate, and permeate, respectively.

[0171] The long-term stability of separation performance was evaluated by filtrating the EB and MO solutions using the biUiO-66-200-0.2 membrane at 2 bar. The feed solution was refreshed termly to minimize concentration polarization at the membrane surface. The mechanical stability of the biUiO-66-200-0.2 membrane was assessed by filtrating the EB solution under transmembrane pressures from 2 to 5 bar. To evaluate the static sorption of UiO-66, the biUiO-66-200-0 powders (10 mg) were immersed in the dye solutions (10 ppm, 10 mL) for 24 h. The concentrations of the solutions before and after sorption were determined by UV-vis spectroscopy.

[0172] Simulation: Triclinic simulation boxes (a=b=c=20.9 A and a=p=y=90 for reo-UiO-66, a’=b’=c’=20.77 A and a’=P’=y’=90° for fcu-UiO-66) were used for simulation. Periodic boundary conditions (PBC) were applied in the xyz direction to simulate the experimentally dynamic behavior of MeOH in UiO-66. In all the dynamic simulations, the atom interactions and charges were calculated by UFF force field and QEq method, respectively. The long- range tail of the Lennard-Jones (LJ) potential was truncated at 8.5 A, in which the discontinuity of the interacting potential at the truncation point was smoothed by the cubic spline method with a spline width of 1 A and a buffer width of 0.5 A. To prevent the mass influence on diffusivity calculation, we inserted one MeOH molecule into UiO-66 structures through Sorption models, in which 1 x107microstates were generated in total by the Metropolis sampling. The first 1 x106microstates were used for equilibrium, while the other 9x106microstates were used for statistics. The dynamic simulation was performed at 10 ns scale and 298 K through NVT ensemble.

[0173] Determination of missing-linker defects: TGA analysis of the UIO-66 samples was performed with the assumption that the residue in TGA tests is ZrO2. To ensure this, the TGA tests were run up to 800 °C under a constant flow of synthetic air. Such conditions should promise the complete combustion of organics and the conversion of Zr clusters to ZrO2. Thus, the reaction for the complete combustion of defect-free dehydroxylated UiO-66, Zr6O6(BDC)6, can be described as follows:

[0174] Zr6O6(BDC)5(s) + 45O2(g) 6ZrO2+ 48CO2(g) + 12H2O (g) Here, the molar mass of Zr6O6(BDC)6is 1628.03 g mol-1, which is 2.202 times higher than that of the residual ZrO2(739.34 g mol-1). Therefore, when the end weight (WEnd) in TGA tests is normalized to 100%, the ideal TGA plateau (WIdea.Plat.) of dehydroxylated UIO-66 should reach 220.2%.

[0175] To determine the relative linker deficiency, we further assumed that the average composition of UiO-66 can be expressed as follows: Zr6O6+x(BDC)6-xwhere x stands for the relative linker deficiency per Zr6formula unit.

[0176] The number of BDC linkers in the ideal Zr6formula unit is 6. Therefore, the weight contribution per linker (WLinker) can be estimated as follows: WLinker= (220.2-100) / 6 = 20.03%

[0177] Thus, the actual number (N) of BDC linkers per Zr6formula unit can be calculated as below:

[0178] N = 6-X = ( WExp.Plat.-WEnd) / WLinker(5) where WExp.Plat.is the experimental TGA plateau.

[0179] Therefore, the value of x, the defect concentration (Cdef), and the practical connectivity (n) can be calculated as follows:

[0180] X = 6-N = 6-(WExp Plat -WEnd) / WLinker(6)Cdef. = x / 6 (7) n = 12x(1 -Cdef ) (8)

[0181] Prediction of molecular rejections: The steric hindrance pore, Verniory, Ferry and modified Ferry transport models were applied to predict the rejection of solutes. Given the highly uniform window pore of UiO-66, the effective sieving pore size (dp) was assumed to be 0.6 nm. The estimation of the solute diameter (ds) can be found in the Supplementary Discussion.

[0182] In the steric hindrance pore model, the reflection coefficient owas calculated as o = 1 - HFSF(9) (10) (1 1 ) where HFrepresents the effect of pore walls and SFrepresents the effect of steric hindrance during transport through nanopores.

[0183] In the Verniory model, the frictional drag force g(η ) is incorporated into the calculation of the reflection coefficient .

[0184] (13)

[0185] (14)

[0186] In the Ferry model, with the assumption of uniform pore size and a parabolic velocity dependence in the transport channels, the reflection coefficient σ is calculated as o = 1 - (η (η - 2))2(15)

[0187] Nevertheless, in a confined environment, the terminal velocity of a solute relative to solvent molecules may diverge from that in free space, giving rise to hydrodynamic Iag69. Here, the viscous drag force, G, acting on a solute because of its proximity to the pore wall can be described as

[0188] G = -6πμ ds(K1 U - K2V) (16) where / J is the solvent viscosity, K) and K2are drag coefficients, U is the solute velocity with respect to a reference, and V is the solvent velocity with respect to the same reference.

[0189] At steady state, and further examination of the ratio reveals a close fit to the exponential function exp(-αη2), which represents the influence of steric hindrance during convective transport. Thus, the modified Ferry model describes the calculation of the reflection coefficient a as σ = 1 — [(η (η — 2))2] exp(-αη2) (17) where a is a dimensionless constant, and we optimized the value of or to 1 .3 here.

[0190] Supplementary Discussion: We sought to establish a correlation between molecular rejections and membrane pore size. To this end, the effective solute size in organic solvents should be estimated first. Based on the Stokes-Einstein and Wilke-Chang equations, we can define a constant ratio of the solute diameters (ds) in two solvents as follows: where k is the Boltzmann’s constant, 0 is the solvent association parameter, Msis the solvent molecular weight (g mol-1), and 7Sis the solute molar volume at the boiling point (m3mol '). 0 values of water and MeOH are 2.6 and 1.9, respectively. The ratio of the effective solute diameter in MeOH relative to that in water was thus calculated to be 0.877. Together with the empirical relationship presented by Bruggen,s1the effective solute diameter in MeOH can be estimated as follows: ds= 0.057(Mw)0.438(19) where Mwis the solute molecular weight (g mol-1).

[0191] The Spiegler-Kedem equation, which combines convection and diffusion transport theory, was adopted to describe the solute rejection (P) as follows: where σ is the reflection coefficient (%), Pcis the solute permeability by diffusion h-1, Jvis the volume flux (L m-2h-1), ΔP is the transmembrane pressure (bar), and Δπ is the osmotic pressure difference (bar). The solute transport in non-aqueous porous membranes through nanofiltration is dominated by convectionS2Thus, the magnitude of Jvgreatly surpasses that of Pc, resulting in a value of F that approaches 0. Therefore, the rejection rate R closely approximates the reflection coefficient σ.S1

[0192] Next, four transport models, including steric hindrance pore, Verniory, Ferry, and modified Ferry models, were adopted to predict the reflection coefficient σ of the small molecules used in this work. In stark contrast to the other models with poor fits, the modified Ferry model yielded an excellent fit to the experimental data (FIG. 47). Notably, the original Ferry model with an assumption of uniform pores provided a rough fit showing an analogous tendency to the experimental rejection curve. This assumption aligns well with the highly ordered sieving channels of polycrystalline UiO-66 membranes. The fitting degree is significantly improved using the modified Ferry model with the inclusion of hydrodynamic lag. This indicates slower transport of solutes relative to solvents in UiO-66 membranes mainly due to the steric hindrance. Therefore, by taking the solvent-solute-membrane interactions into consideration, the results discussed above suggest the feasibility of predicting solute rejections in polycrystalline MOF nanofiltration membranes with uniform pores. 51. Geens, J., Boussu, K., Vandecasteele, C. & Van der Bruggen, B. Modelling of solute transport in non-aqueous nanofiltration. J. Membr. Sci. 281 , 139-148 (2006).

[0193] 52. Geens, J., Hillen, A., Bettens, B., Van der Bruggen, B. & Vandecasteele, C. Solute transport in non-aqueous nanofiltration: effect of membrane material. J. Chem. Technol. Biotechnol. 80, 1371 -1377 (2005).

[0194] Example 1 : Structural characteristics of reo-UiO-66

[0195] BDC (2 mmol, 332 mg) and metal salts (1 mmol) with a molar ratio of Zn(NO3)2.6H2O:ZrCl4ranging from 0 to 1 were charged into a 20-ml glass vial followed by the addition of DMF (12 ml) and hydrochloric acid (HCI, 0.2 ml). The prepared mixture was ultrasonically treated to form a clear and transparent solution. Subsequently, the solution was transferred into a Teflon lining with a stainless-steel cover. The autoclave was placed in a preheated oven with designated temperatures for 24 h. Finally, the generated white powders were collected through centrifugation and thoroughly washed with DMF and MeOH for several days, followed by drying at 85 °C under vacuum for 24 h. UiO-66-100 powders were synthesized following the above procedures but without adding Zn(NO3)2.6H2O as a control. The neat biUiO-66 solids free of any solvent washing were also prepared to study washing treatments. It is noted that fresh HCI should be used for the synthesis.

[0196] Results and Discussion:

[0197] Perfect UiO-66 presents a face-centered cubic (feu) network comprising six-centered zirconium oxyhydroxide clusters linked by 1 ,4-dicarboxybenzene (BDC) bridges. Meanwhile, orderly distributed missing-cluster defects can afford a unique reo topological structure (FIG. 1a). Previous attempts to generate missing-cluster defects have resulted in irregular arrangements with insufficient defects (Lee, T. H. et al. Defect engineering in metal-organic frameworks towards advanced mixed matrix membranes for efficient propylene / propane separation. Angew. Chem. Int. Ed. 60, 13081 -13088 (2021 )). Here, we have integrated a bimetallic strategy with temperature regulation to synthesize UiO-66 with prominently reo- structured defects. The proposed method is characterized by its ability to directly generate reo-UiO-66 without post-synthetic treatments such as acid washing. Specifically, zinc nitrate (Zn(NO3)2) and zirconium chloride (ZrCl4) were used as the bimetallic salts for the synthesis of biUiO-66-x-y, in which x and y stand for the synthetic temperature and the molar ratio of Zn(NO3)2 and ZrCl4(Zn / Zr). We also synthesized UiO-66-x solids without bimetallic regulation as control samples. Powder X-ray diffraction (PXRD) patterns reveal high crystallinity of UiO-66 solids with an intense diffraction peak at 7.4° (FIG. 1 b). In sharp contrast to UiO-66-200, we observed distinct diffraction peaks at 4.4° and 6.1° in the PXRD pattern of biUiO-66-100-1 . The newly generated peaks stem from diffraction of the (100) and (110) crystal facets, suggesting correlated reo structures in billiO-66 (Liu, L. M. et al. Imaging defects and their evolution in a metal-organic framework at sub-unit-cell resolution. Nat. Chem. 11 , 622-628 (2019)). The synthetic temperature was found to influence the generation of reo structures (FIG. 7). Synthesis at appropriately low temperatures, such as 85 and 100 °C, can yield reo-structured UiO-66 without damaging the crystallinity. The use of Zn(NO3)2during synthesis largely promoted the growth of reo structures and allowed accurate control over the detectivity that arises from reo structures (FIG. 1c and FIG. 8). The UiO-66-100 samples synthesized with reduced Zr / BDC molar ratios exhibited inferior diffraction intensity from the reo structure, supporting the contribution of Zn in promoting reo structures FIG. 9). Moreover, reo-structured biUiO-66 could be synthesized at a high temperature with the aid of bimetallic regulation (FIG. 10). We examined the structure of the biUiO-66-100-1 samples synthesized with diverse metal sources (FIG. 11 ). The results show that the addition of soft Lewis-acid metal ions can effectively induce reo structures. By contrast, the use of hard Lewis-acid metal ions is ineffective. According to Pearson’s hard / soft acid / base theory, interactions between the hard Lewis-base BDC and hard Lewisacid metal ions are stronger than those between BDC and soft Lewis-acid metal ions. We thus speculate that the presence of weak interactions between BDC and soft Lewis-acid metal ions might appropriately compromise the coordination between BDC and Zr, giving rise to reo structures.

[0198] The 77 K N2sorption tests were conducted to study the aperture features of UiO-66 powders. Low-temperature synthesis resulted in UiO-66 with increased Brunauer-Emmett-Teller (BET) surface areas and pore volumes (FIG. 1d and FIG. 12 and FIG. 13). biUiO-66-100-1 synergistically regulated by temperature and bimetallic synthesis afforded a notable BET surface area and pore volume of 977.9 m2g-1and 0.55 cm3g-1, respectively. Analysis of the isotherms by nonlocal density functional theory (NLDFT) recognized two typical pore types in UiO-66-200 (FIG. 1 e), referring to the window pore and internal feu cavity of perfectly crystallized UiO-66 frameworks. By contrast, narrowly distributed large pores with a width centered at 1.2 nm emerged in UiO-66-100. The presence of these large pores became notable in biUiO-66-100-1. Together with the attenuation of the feu cavity, we attribute these well-defined large pores to a regular deficiency of Zr clusters with the formation of a reo cavity, which rationalizes the increase in BET surface areas and pore volumes. Moreover, as visualized by scanning electron microscopy (SEM), the evolution of UiO-66 from the feu to reo topologies resulted in substantial morphological changes in the crystals (FIG. 14).

[0199] We conducted a gravimetric analysis of the resulting UiO-66 materials and compared it to the synthetic recipe. The bimetallic modulation caused an evident mass loss for the synthesized biUiO-66-100 solids (FIG. 15). We speculated that Zn was minimally crystallized in the framework of UiO-66 by considering the conspicuous mass reduction with elevation of the Zn / Zr molar ratio. X-ray photoelectron spectroscopy (XPS) confirmed negligible amounts of Zn in bil)iO-66-100-1 (FIG. 16). The creation of missing-cluster defects typically produces loss of organic linkers. Thermogravimetric analysis (TGA) was performed to evaluate the missing-linker defects. FIG. 2a presents typical TGA curves for UiO-66, which mainly include three stages: (1 ) solvent evaporation, (2) removal of structural water molecules and compensating ligands and (3) decomposition of the organic parts. The decomposition temperature was chosen to be -385 °C for the missing-linker defect calculations because of the relative TGA plateau at this temperature. For an ideal structure of UiO-66 without losing linkers, Zr nodes are completely coordinated with BDC linkers to give an ideal plateau weight, Widea.piat., of 220.2% when normalized by the final weight of zirconium oxide (ZrO2; FIG. 17). As expected, UiO-66-200 displayed an experimental plateau at -219.5%, corresponding to its perfect structure. However, the TGA curves of the temperature-regulated UiO-66 (UiO-66- 100 and UiO-66-85) and biUiO-66-100-1 deviated from the ideal plateau, indicating the presence of missing-linker defects. The linker connectivity was further estimated according to the TGA results. biUiO-66-100-1 was calculated to have a connectivity of 10.78, which falls within the connectivity range of perfect fcu-structured UiO-66 (12) and reo-structured UiO-66 (8), suggesting the coexistence of feu and reo topologies. XPS measurements were adopted to gain in-depth insights into the structure of these UiO-66 samples. In the O 1 s signal of UiO-66, three characteristic peaks assigned to the specific oxygen-based species at the Zr6-oxo node could be observed, pointing to Zr-O-Zr (530.5 eV), Zr-O-C (532 eV) and -COOH (533.4 eV) (FIG. 2b). The Zr-O-Zr and -COOH signals indicate the existence of Zr clusters and BDC linkers, respectively, which is promising for the detection of their percentages. The low relative density of Zr-O-Zr signal in biUiO-66-100-1 is consistent with the missing-cluster defects discussed before.

[0200] Washing procedures affect the determination of diffraction peaks from reo structures using PXRD testing. The removal of / V, / V-dimethylformamide (DMF) by thorough solvent exchange with methanol (MeOH) contributed to exposing the characteristic diffraction peaks in the PXRD pattern of bi UiO-66- 100-1 (FIG. 18). To study the effect of Zn clusters during washing, we performed XPS analysis on biUiO-66-100-1 subjected to various washing procedures (FIG. 19). The XPS spectra of Zn signals in FIG. 2c support that negligible Zn is involved in biUiO-66-100-1 , irrespective of the washing procedures. Energy-dispersive X-ray spectroscopy (EDS) mapping of the powders free of solvent washing visually excluded the presence of Zn (FIG. 2d). The lack of an increase in the detectivity of missing clusters after acid treatments also confirms the absence of Zn in biUiO-66-100-1 (FIG. 20). These results complementarily verify that there is no involvement of Zn in UiO-66 frameworks after crystallization.

[0201] Example 2: reo-UiO-66 membranes and performance

[0202] BDC (1.2 mmol, 199.2 mg) and metal salts (0.6 mmol) with a molar ratio of Zn(NO3)2.6H2O:ZrCl4ranging from 0 to 1 were charged into a 20-ml glass vial followed by the addition of DMF (1 1 ml) and HCI (0.1 ml). The prepared mixture was ultrasonically treated to form a clear and transparent solution. Subsequently, the solution was transferred into a Teflon lining with a stainless-steel cover. A homemade Teflon shelf preloaded with two pieces of activated alumina supports was introduced into the autoclave lining. The autoclave was placed in a preheated oven at designated temperatures for 24 h. The obtained membranes were fully washed with DMF for secondary growth. For the secondary growth, BDC (0.6 mmol, 99.6 mg) and metal salts (0.3 mmol) with a corresponding molar ratio of Zn(NO3)2.6H2O:ZrCl4ranging from 0 to 1 were charged into a 20-ml glass vial followed by the addition of DMF (11 ml) and HCI (0.1 ml). The residual procedures were similar to that of the first growth. The synthesized membranes were completely washed with DMF and MeOH for several days, followed by drying at 85 °C under vacuum for 24 h before further use. It is noted that fresh HCI should be used for the synthesis.

[0203] Results and Discussion:

[0204] Thin UiO-66 membranes were crystallized with missing-cluster defects on an activated alumina support using solvothermal growth (FIG. 3a). Although membranes crystallized at 100 °C showed notable diffraction peaks from the correlated reo structure, the nonselective defects prevent their use in separation applications (FIG. 21 and FIG. 22). Elevating the synthesis temperature to 200 °C produced highly crystallized and well-intergrown membranes free of obvious grain boundary gaps (FIG. 3b and FIG. 3c and FIG. 23 and FIG. 24). The distinct contrast of the crystallographic texture imaged by atomic force microscopy (AFM) arises from the presence of missing-cluster defects in the biUiO-66-200-0.4 membrane (FIG. 3d and FIG. 3e). The structural alteration exhibits a negligible effect on the membrane root-mean-squared roughness (Rq), which was determined to be -33.1 and -35.9 nm for the billiO-66-200-0 and biUiG-66-200-0.4 membranes, respectively. Imaging the cross-section of these membranes by SEM shows a composite structure with continuously grown UiO-66 polycrystals on top of alumina supports (FIG. 25). The magnified cross-sectional SEM images reveal the intergrowth of UiO-66 crystals, which possess similar thicknesses of -250-270 nm (FIG. 3f and FIG. 3g and FIG. 26). Cross-sectional transmission electron microscopy (TEM) observation of the biUiO-66-200-0.4 membrane confirmed a thin and continuous UiO-66 layer without intercrystalline gaps (FIG. 3h and FIG. 27). EDS tests of the cross-section detected typical C and Zr signals from UiO-66 frameworks without the presence of Zn (FIG. 3i). In addition, the results present a clear line of demarcation between UiO-66 and alumina supports. Despite their thinness, the UiO-66 membranes still offer excellent crystallinity, as confirmed by PXRD analysis (FIG. 3j). Consistent with the PXRD patterns of the powders (FIG. 28), the biUiO-66-200-0.4 membrane presents characteristic diffraction peaks in the 20 range of 4-6° (FIG. 3j, inset), indicative of reo-structured regions. The 77 K N2sorption tests disclose narrow size distributions of the reo cavity in the biUiO- 66-200 samples (FIG. 29). According to the TGA results (FIG. 30), the connectivities of biUiG-66-200-0.2 and bil)iO-66-200-0.4 were estimated to be 10.87 and 9.96, respectively, agreeing with the co-occurrence of feu and reo topologies.

[0205] To evaluate the liquid-phase separation capability of the membranes, we immersed them in common solvents for three months. PXRD patterns recorded on the treated membranes show intense diffraction peaks arising from the grown UiO-66 (FIG. 31 ), indicating the preserved crystalline structures. The membrane retained its crystalline structure even after being soaked in organic acid and base solutions for three days (FIG. 32). Microscopic observations supported the structural integrity of these membranes (FIG. 33 and FIG. 34). We further evaluated the membrane stability under flowing liquids. SEM images and PXRD patterns revealed an intact polycrystalline structure of the membrane after exposure to the flowing solvents (FIG. 35 and FIG. 36). Importantly, the solutions collected after solvent circulation showed no trace of either Zr or BDC (FIG. 37 and Table 1), confirming the high robustness of the grown UiO-66 membrane. To quantify the separation performance of the UiO-66 membranes, we measured MeOH permeance and Evans blue (EB, Mw = 960.8 g mol-1) rejection using a dead-end filtration cell. It is noted that the alumina substrate exhibited a MeOH permeance of 5,634.9 L m-2h-1bar1and an EB rejection of -2.1 %, giving minimal mass transport resistance and negligible contribution to separation (FIG. 38). The biUiO-66-200-0 membrane showed a MeOH permeance of 3.4 L m-2h-1bar1and impressive EB rejection as high as 99.0% (FIG. 4a). While having comparable rejection performance, the permeance of the neat UiO-66 membrane exceeds that of previous micrometer-thick UiO-66 membranes (Cai, Y. H. et al. Polycrystalline zirconium metalorganic framework membranes supported on flexible carbon cloth for organic solvent nanofiltration. J. Membr. Sci. 615, 118551 (2020)). This is related to the small thickness of the UiO-66 layers (FIG. 39 and Table 2), which reduces the mass transfer distance, facilitating solvent permeation. The biUiO-66-200-0.2 membrane yielded an approximately doubled MeOH permeance compared to that of the neat counterpart, while affording an EB rejection of up to 99.2%. When the Zn / Zr molar ratio increased to 0.4, the MeOH permeance and EB rejection reached 1 1 .2 L m-2h-1bar1and 97.8%, respectively. This prominent improvement in MeOH permeance at a minor cost to EB rejection benefits from the created missing-cluster defects in UiO-66 (Teesdale, J. J., Lee, M. J., Lu, R. X. & Smith, Z. P. Uncertainty in composite membranes: from defect engineering to film processing. J. Am. Chem. Soc. 145, 830-840 (2023)). Specifically, the reo-structured internal cavities with an enlarged pore size of ~1.2 nm contributed to solvent permeation (FIG. 40 and FIG. 41 ). However, they could still be effective in rejecting EB molecules (2.71 nm x 0.76 nm). Moreover, modeling studies quantified that the transport rate of MeOH in reo-UiO-66 was 4.7 times higher than that of fcu-UiO-66 (FIG. 42), comparable to the experimental value of -3.3.

[0206] Table 1. Zr content of the solutions produced from flowing solvents.

[0207] Table 2. Details for the comparison of UiO-66 membrane thicknesses.

[0208] S3. Wu, F. C. et al. Synthesis of stable UiO-66 membranes for pervaporation separation of methanol / methyl tert-butyl ether mixtures by secondary growth. J. Membr. Sci. 544, 342-350 (2017).

[0209] 54. Friebe, S., Geppert, B., Steinbach, F. & Caro, J. Metal-organic framework UiO-66 layer: a highly oriented membrane with good selectivity and hydrogen permeance. ACS Appl. Mater. Interfaces 9, 12878-12885 (2017).

[0210] 55. Wang, X. R. et al. Improving water-treatment performance of zirconium metal organic framework membranes by postsynthetic defect healing. ACS Appl. Mater. Interfaces 9, 37848-37855 (2017). 56. Liu, X. L., Demir, N. K., Wu, Z. T. & Li, K. Highly water-stable zirconium metal organic framework UiO-66 membranes supported on alumina hollow fibers for desalination. J. Am. Chem. Soc. 137, 6999-7002 (2015).

[0211] 57. Yan, J. H. et al. Room-temperature synthesis of defect-engineered zirconium-MOF membrane enabling superior CO2 / N2 selectivity with zirconium-oxo cluster source. J. Membr. Sci. 653, 120496 (2022).

[0212] 58. Liu, X. L., Wang, C. H., Wang, B. & Li, K. Novel organic-dehydration membranes prepared from zirconium metal-organic frameworks. Adv. Fund. Mater. 27, 160431 1 (2017).

[0213] 59. Cai, Y. H. et al. Polycrystalline zirconium metal-organic framework membranes supported on flexible carbon cloth for organic solvent nanofiltration. J. Membr. Sci. 615, 118551 (2020).

[0214] To gain a deeper understanding of the correlation between structure and performance, we used various small molecules as probes to assess the selectivity of these membranes (FIG. 43-46). The rejection profiles depicted in FIG. 4b show an identical molecular selectivity for the biUiG-66-200-0 and biUiG-66-200-0.2 membranes, which both produced rejection rates of above 98% for the probes with molecular weights above 300 g mol-1. This molecular selectivity exceeds that of the previous UiO-66 nanofiltration membranes (Shangkum, G. Y., Chammingkwan, P., Trinh, D. X. & Taniike, T. Design of a semi-continuous selective layer based on deposition of UiO-66 nanoparticles for nanofiltration. Membranes 8, 129 (2018); Ma, D., Han, G., Gao, Z. F. & Chen, S. B. Continuous UiO-66-type metal-organic framework thin film on polymeric support for organic solvent nanofiltration. ACS Appl. Mater. Interfaces 11 , 45290-45300 (2019)). Although the selectivity towards low-molecular-weight probes decreased — possibly due to the formation of continuous reo-structured channels with ample missing-cluster defects — the biUiO-66-200-0.4 membrane still achieved a dye molecular weight cutoff (MWCO) of -490 g mol-1. Notably, the biUiO-66-200-0.4 membrane meets the benchmark of the state-of-the-art OSN membranes (FIG. 4c and Table 3), revealing its potential for effective separations in organic solvents. Various mass-transport models were utilized to correlate molecular rejections with membrane pore size (Supplementary Discussion and FIG. 47). Among these, the modified Ferry model, which incorporates hydrodynamic lag due to steric hindrance, provided the best fit. This finding could offer valuable guidance for designing polycrystalline membranes and predicting their separation performance. The molecular-separation mechanism was then investigated. The uncontaminated membrane surface after filtration, combined with the negligible adsorption rate (~1 .3%) and static adsorption tests of the UiO-66 powders, eliminates the possibility of a contribution from adsorption (FIG. 48-50). In addition, both UiO-66 and most of the tested molecules had weak electronegativity (FIG. 51 ). More importantly, the aforementioned study on the molecular-transport models suggests strong size-dependent separation through our membranes. These findings indicate that the separation primarily relies on size-dependent exclusion, with electrostatic repulsion playing a relatively minor role. Table 3. Details of the membrane performance comparison.

[0215] S10. Shi, G. M., Farahani, M., Liu, J. Y. & Chung, T. S. Separation of Vegetable Oil Compounds and Solvent Recovery Using Commercial Organic Solvent Nanofiltration Membranes. J. Membr. Sci. 588, 117202 (2019).

[0216] 51 1. Fu, W. M. et al. A high-flux organic solvent nanofiltration membrane with binaphthol- based rigid-flexible microporous structures. J. Mater. Chem. A 9, 7180-7189 (2021 ).

[0217] 512. Li, Y., Zhu, J. Y., Li, S., Guo, Z. & Van der Bruggen, B. Flexible aliphatic-aromatic polyamide thin film composite membrane for highly efficient organic solvent nanofiltration. ACS Appt. Mater. Interfaces 12, 31962-31974 (2020).

[0218] 513. Jimenez-Solomon, M. F., Song, Q. L., Jelfs, K. E., Munoz-Ibanez, M. & Livingston, A. G. Polymer nanofilms with enhanced microporosity by interfacial polymerization. Nat. Mater. 15, 760-767 (2016).

[0219] 514. Ali, Z. et al. Finely tuned submicroporous thin-film molecular sieve membranes for highly efficient fluid separations. Adv. Mater. 32, 2001 132 (2020).

[0220] 515. Li, Y., Li, S., Zhu, J. Y., Volodine, A. & Van der Bruggen, B. Controllable synthesis of a chemically stable molecular sieving nanofilm for highly efficient organic solvent nanofiltration. Chemical science 11 , 4263-4271 (2020).

[0221] 516. Huang, Y. Z., Li, S. L., Fu, Z. X., Gong, G. H. & Hu, Y. X. Preparation of microporous organic solvent nanofiltration (OSN) composite membrane from a novel tris-phenol monomer. Sep. Purif. Technol. 301 , 121985 (2022).

[0222] 517. Fu, W. M. et al. Ultra-thin microporous membranes based on macrocyclic pillar[n]arene for efficient organic solvent nanofiltration. J. Membr. Sci. 655, 120583 (2022).

[0223] 518. Li, X., Li, C., Goh, K., Chong, T. H. & Wang, R. Layer-by-layer aided p-cyclodextrin nanofilm for precise organic solvent nanofiltration. J. Membr. Sci. 652, 120466 (2022).

[0224] 519. Liu, J. T., Hua, D., Zhang, Y., Japip, S. & Chung, T. S. Precise molecular sieving architectures with Janus pathways for both polar and nonpolar molecules. Adv. Mater. 30, 1705933 (2018).

[0225] 520. Li, S. L. et al. 2,2 '-Biphenol-based ultrathin microporous nanofilms for highly efficient molecular sieving separation. Angew. Chem., Int. Ed. 61 , e202212816 (2022).

[0226] 521. Huang, T. F. et al. Molecularly-porous ultrathin membranes for highly selective organic solvent nanofiltration. Nat. Commun. 11 , 5882 (2020).

[0227] 522. He, X. et al. Controlling the selectivity of conjugated microporous polymer membrane for efficient organic solvent nanofiltration. Adv. Fund. Mater. 29, 1900134 (2019). S23. Huang, T. F., Puspasari, T., Nunes, S. P. & Peinemann, K. V. Ultrathin 2D-layered cyclodextrin membranes for high- performance organic solvent nanofiltration. Adv. Fund. Mater. 30, 1906797 (2020).

[0228] Performance stability and durability are crucial factors in determining the practicality of separation membranes. Intriguingly, the EB rejection rate of the biUiO-66-200-0.2 membrane remained almost unchanged on elevating the transmembrane pressure from 2 to 5 bar (FIG. 52). This pressure-resistant feature originates from the synergy of the rigid UiO-66 frameworks and robust ceramic supports. There is typically a deterioration of performance as a result of solvent swelling and physical ageing of building materials in conventional polymer membranes. Membranes composed of UiO-66 could overcome these limitations with their underlying well-crystallized frameworks. As expected, the biUiO-66-200-0.2 membrane, with its crystalline framework and rigid sieving channels, was found to exhibit exceptional long-term stability in the filtration of small molecules, for over 1 ,500 h (FIG. 4d). This operational durability greatly exceeds those of other reported membranes (FIG. 53 and Table 4), highlighting the prominent stability of our polycrystalline membranes. Altogether, the robust performance with respect to mechanical and working stability endows our membrane with the potential for long-term use.

[0229] Table 4. Details for the comparison of operation durations.

[0230] 524. He, P. P. et al. In-situ growth of double-layered polyaniline composite membrane for organic solvent nanofiltration. Chem. Eng. J. 420, 129338 (2021 ).

[0231] 525. Wu, M. B. et al. Lysozyme membranes promoted by hydrophobic substrates for ultrafast and precise organic solvent nanofiltration. Nano Lett. 20, 8760-8767 (2020). 526. Nie, L. et al. Realizing small-flake graphene oxide membranes for ultrafast sizedependent organic solvent nanofiltration. Sci. Adv. 6, eaaz9184 (2020).

[0232] 527. Liang, B. et al. Microporous membranes comprising conjugated polymers with rigid backbones enable ultrafast organic-solvent nanofiltration. Nat. Chem. 10, 961 -967 (2018).

[0233] 528. Zhang, L. Y., Zhang, M. C., Liu, G. P., Jin, W. Q. & Li, X. Y. Fungal cell wall-graphene oxide microcomposite membrane for organic solvent nanofiltration. Adv. Funct. Mater. 31 , 21001 10 (2021).

[0234] 529. Shen, L. et al. Highly porous nanofiber-supported monolayer graphene membranes for ultrafast organic solvent nanofiltration. Sci. Adv. 7, eabg6263 (2021 ).

[0235] 530. Zhang, Y. Q. et al. Robust natural nanocomposites realizing unprecedented ultrafast precise molecular separations. Mater. Today 36, 40-47 (2020).

[0236] 531. Shi, X. et al. Design of three-dimensional covalent organic framework membranes for fast and robust organic solvent nanofiltration. Angew. Chem., Int. Ed. 61 , 6202207559 (2022).

[0237] Example 3: Molecular-sieving tests

[0238] The biUiO-66-200-0.2 membrane was selected for molecular-sieving tests. Methanol-based solutions containing diverse molecular pairs were separately prepared as feed for sieving at 2 bar. The feed and filtrate were analyzed by a UV-vis absorption spectrometer (Cary 60, Agilent). The long-term molecular-sieving test was conducted on the separation of an NP / MO mixture. The separation factor (a) in molecular-sieving tests was calculated according to a = (CNP-Fi / CMO-Fi) / (CNP-Fe / CMO-Fe) (23) where CNP-Fiand CMO-Fiare the concentrations of NP and MO in the filtrate, respectively, andCNP-Feand CMO-Feare the concentrations of NP and MO in the feed, respectively.

[0239] Results and Discussion:

[0240] The selective separation of complex solutes with molecular weights below 350 g mol-1is recognized as a long-standing challenge. Previous breakthroughs in separation membranes were limited to the discrimination of large molecules, mainly because the pore sizes were large and the pore uniformity was suboptimal (Wang, H. J. et al. Organic molecular sieve membranes for chemical separations. Chem. Soc. Rev. 50, 5468-5516 (2021)). UiO-66, however, has periodically arranged channels consisting of angstrom-sized windows and nanometer-sized cavities, and the size-exclusion gate provided by the ~6-A pore window may offer high selectivity. Meanwhile, the internal reo-structured cavity promises the desired permeability, as shown in FIG. 4e. A series of fine molecules were used to estimate the sieving capability of the biUiO-66-200-0.2 membrane (FIG. 4f and FIG. 54). FIG. 4g presents the ultraviolet-visible (UV-vis) spectral results for separating methyl orange (MO, 14.4 A x 4.3 A) and p-nitrophenol (NP, 6.7 A x 4.2 A). The peaks at 421 and 31 1 nm refer to the characteristic signals of MO and NP, respectively. The filtrate produced by the membrane showed no MO signal, but the intense characteristic peak of NP, indicating excellent molecular-sieving ability. This sieving ability towards fine molecules was then more widely demonstrated by the accurate separation of diverse binary complex mixtures (FIG. 55). This universality makes our membrane attractive for precisely distinguishing molecules with fine and similar dimensions. Importantly, the sieving performance continued for over 500 h without an appreciable loss of selectivity (FIG. 4h), which is promising for achieving continuous sieving. When correlating the estimated diameters of the solvated molecules with the membrane pore size, it is evident that the window pore acts as the sieving gate, achieving exceptional separation precision to within 3 A (FIG. 56). This membrane selectivity towards low-molecular-weight targets outperforms those of other membranes reported in the literature (FIG. 57 and Table 5). These findings highlight the potency of MOFs in the design of polycrystalline membranes to sieve fine molecules with small size differences in organic liquids. We anticipate that the structural design of MOF membranes could unlock their potential for practical separation demands associated with complicated mixtures and isomers.

[0241] Table 5. Details for the comparison of membrane selectivity.

[0242] S32. Yang, Q. et al. Ultrathin graphene-based membrane with precise molecular sieving and ultrafast solvent permeation. Nat. Mater. 16, 1 198-1202 (2017).

[0243] 533. Jiang, Z. et al. Aligned macrocycle pores in ultrathin films for accurate molecular sieving. Nature 609, 58-64 (2022).

[0244] 534. Shi, X. S. et al. Flexible and robust three-dimensional covalent organic framework membranes for precise separations under extreme conditions. Nano Lett. 21 , 8355-8362 (2021 ).

[0245] 535. Zhang, S. X. et al. Ultrathin microporous metal-organic network membranes for molecular separation. J. Mater. Chem. A 9, 25531 -25538 (2021).

[0246] Example 4: Pharmaceutical and catalyst recovery

[0247] To reveal the feasibility of our membrane in practical applications, we explored the separation of active pharmaceutical ingredients (APIs) that are prevalent in the supply chains of the pharmaceutical industry (FIG. 5a). Among these APIs, curcumin is increasingly in demand, because it has a vast scope of pharmacological actions with negligible biotoxicity. Spectral analysis and digital images unambiguously showed efficient separations of the APIs (FIG. 5b and FIG. 58). The solute rejection (%) for curcumin, tetracycline, rifampicin, vitamin B12 and spiramycin was measured to be 99.3%, 100%, 99.9%, 99.6%, and 94.2%, respectively (FIG. 5c). Thus, the designed membrane shows potential for energy-saving and high-quality production of active ingredients in the pharmaceutical industry. Molecular- separation membranes have also shown promise for recycling precious catalysts. As a proof of concept, we performed the recovery of [(R)-2,2'-bis(diphenylphosphino)-1, 1'- binaphthyl]ruthenium(l I) dichloride (Ru-BINAP), which has been widely used for catalytic hydrogenation reactions. The membrane exhibited Ru-BINAP rejection of up to 96% (FIG. 5d), suggesting its potential in this emerging application field. Overall, these results disclose the practicality and versatility of our membrane in different separation applications.

[0248] Example 5: Tubular membrane module and crossflow filtration

[0249] Finally, we turned to the design and fabrication of large-size membrane modules for practical applications. We synthesized biUiO-66-200-0.2 membranes on a commercially available, cheap alumina tube and designed a crossflow filtration module with an effective membrane area of 18.8 cm2(FIG. 6a and FIG. 59). The laboratory fabrication cost of our tubular membrane was estimated to be ~S$10,851 m- 2(S$, Singapore dollar; Table 6). This cost can be largely reduced to about '387 S$ rm2when using industrial materials. Moreover, the price of alumina substrates dominates this fabrication cost, which can be further reduced by bulk ordering. As illustrated in FIG. 6b, the feed solution was circulated by a piston pump, and a counterbalance valve was used to regulate the transmembrane pressure under crossflow filtration. The tubular biUiC-66-200-0.2 membrane exhibited surface and cross- sectional morphologies similar to those of the small membrane disk (FIG. 6c and FIG. 6d and FIG. 60). Its MeOH permeance was measured to be 4.6 L m-2h-1bar1. Analyses of the separation performance for various fine solutes revealed a sharp molecular rejection curve comparable to the result discussed earlier (FIG. 6e). The tubular membrane further demonstrated excellent molecular-sieving ability to separate MO and NF in a binary system (FIG. 61 ). FIG. 6f shows membrane performance over the period of operation. The rejections of MO and EB were uncompromised after continuous operation for 20 days, indicating the stability of this tubular membrane module. Moreover, the tubular membrane remained uncontaminated after long-term filtration due to the excellent repulsion of the fine solutes (FIG. 62). Overall, we have demonstrated a prototypical tubular membrane module for continuous crossflow filtration, showcasing the practical utility of our membrane. We are also developing multichannel modules capable of accommodating multiple membrane tubes. This design will further improve separation efficiency. When comparing our membrane technology with existing methods, it is clear that each technology offers unique advantages and inherent limitations (FIG. 6g and Table 7). The choice of method depends on factors such as the compound properties, targeted purity levels, scalability requirements and process compatibility with production. A rational integration of these methods can be explored to maximize energy and capital efficiency with improved sustainability. Table 6. Cost analysis for the fabrication of our tubular membrane. Note: S$ stands for Singapore dollar.

[0250] Table 7. Details for the comparison of crystallization, chromatography, and membrane for the pharmaceutical industry.

[0251] Conclusion

[0252] In summary, we have demonstrated the synthesis of reo-structured UiO-66 featuring regularly distributed missing-cluster defects by means of a facile bimetallic method. Such a specific reo-UiO-66 with angstrom-sized windows and enlarged internal cavities was crystallized into robust and thin membranes capable of liquid-phase molecular separations. Benefiting from the created reo structures, the synthetic membranes exhibited improved solvent permeance at little expense of molecular selectivity. The membrane pore size and solute rejections were quantitatively correlated by the modified Ferry transport model, providing a paradigm to predict molecular rejections through non-aqueous nanofiltration. Highly crystalline frameworks allow efficient separation of fine molecules in MeOH for over 1 ,500 h, which is promising for practical applications. The membrane further performed accurate sieving of molecule pairs with fine and similar sizes. More importantly, pharmaceuticals and catalysts of practical value could be effectually purified and recovered by our membrane. We finally demonstrated the efficacy of our tubular membrane module for continuous crossflow operation. The present work provides a viable strategy to develop competitive membrane technology based on intrinsically crystalline porous materials for sustainable chemical separation in liquids.

Claims

CLAIMS1 . A composite material comprising: a porous substrate; a polycrystalline material layer disposed on a surface of the porous substrate; wherein the polycrystalline material layer comprises a plurality of intergrown crystals formed from a UiO-66 metal-organic framework (MOF), or derivative thereof; wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks indicative of reo-structured regions within the UiO-66 MOF, or derivative thereof.

2. The composite material of claim 1 , wherein the plurality of intergrown crystals are characterized by a powder X-ray diffractogram with characteristic peaks at 26 = about 4.3° and about 6.1 ° (e.g. 4.3° ± 0.2° and 6.1 ° ± 0.2°).

3. The composite material of claim 2, wherein the powder X-ray diffractogram has further characteristic peaks at 26 = about 7.3° and about 8.6° (e.g. 7.3° ± 0.2° and 8.6° ± 0.2°).

4. The composite material of any one of claims 1 to 3, wherein the polycrystalline material layer is substantially free from intercrystalline gaps.

5. The composite material of any one of the preceding claims, wherein the polycrystalline material layer consists essentially of the plurality of intergrown crystals.

6. The composite material of any one of the preceding claims, wherein the plurality of intergrown crystals have a surface area of from about 500 m2g1to about 2000 m2g1, for example from about 500 m2g1to about 2000 m2g1, from about 500 m2g1to about 1200 m2g1, from about 800 m2g1to about 1100 m2g1, or from about 900 m2g1to about 1000 m2g1.

7. The composite material of any one of the preceding claims, wherein the plurality of intergrown crystals comprise pores with a pore volume of from about 0.3 cm2g1to about 1 cm2g-1, for example from about 0.4 cm2g1to about 0.6 cm2g1.

8. The composite material any one of the preceding claims, wherein the plurality of intergrown crystals comprise pores with an average pore diameter of from about 1 nm to about 1 .5 nm, for example about 1 .2 nm.

9. The composite material of any one of the preceding claims, wherein the thickness of the polycrystalline material layer is from about 200 nm to about 400 nm, for example from about 200 to about 300 nm (e.g. from about 250 to about 270 nm).

10. The composite material of any one of the preceding claims, wherein the composite material has a surface average roughness of from about 20 nm to about 100 nm, for example from about 20 nm to about 40 nm, about 30 nm to about 40 nm or about 30 nm to about 35 nm.11 . The composite material of any one of the preceding claims, wherein the composite material displays a permeance of greater than 3.4 L m- 2h-1bar-1for methanol and a dye rejection rate of greater than about 90% for Evans Blue (i.e. tetrasodium (6E,6'E)-6,6- [(3,3'-dimethylbiphenyl-4,4'-diyl)di(1 E)hydrazin-2-yl-1 -ylidene]bis(4-amino-5-oxo-5,6- dihydronaphthalene-1 ,3-disulfonate).

12. The composite material of any one of the preceding claims, wherein the composite material displays a permeance of from about 5 L m- 2h-1bar-1to about 12 L m- 2h-1bar-1for methanol, and a dye rejection rate of greater than about 95% for Evans Blue, for example a dye rejection rate of greater than 96%, of greater than 97%, of greater than 98%, or of greater than 99% for Evans Blue.

13. The composite material of any one of the preceding claims, wherein the porous substrate is selected from the group consisting of a ceramic (e.g. alumina), a carbon cloth, a metal, and a metal oxide.

14. The composite material of any one of the preceding claims, wherein the porous substrate is provided in the form of a tube, a mesh, a sheet, or hollow fibers (e.g. porous alumina hollow fibers), or other arrangements that are obtainable by the folding of a tube, a mesh, a sheet, or hollow fibers.

15. The composite material of any one of the preceding claims, wherein the substrate is an activated alumina substrate.

16. The composite material of any one of the preceding claims, wherein the porous substrate has a first surface and second surface, wherein the polycrystalline material layer is disposed on the first surface, and the second surface is substantially free from the polycrystalline material layer formed from the UiO-66 MOF, or derivative thereof.

17. The composite material of any one of the preceding claims, wherein the composite material is suitable for use as a filter material in organic solvent nanofiltration.

18. The composite material of any one of the preceding claims, wherein the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis using a reaction mixture comprising an organic solvent, 1 ,4-benzene- dicarboxylate (BDC) or derivative thereof, a Zr salt and a Zn salt, and wherein the solvothermal synthesis was performed at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C).

19. The composite material of claim 18, wherein after the polycrystalline material layer was formed on the surface of the porous substrate by solvothermal synthesis, the composite material was subjected to washing with DMF, methanol, or a mixture thereof.

20. The composite material of claim 18 or 19, wherein the Zn salt and the Zr salt are present in the reaction mixture at molar ratio of Zn to Zr of from about 0.1 to about 1 , for example from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5).

21. The composite material of any one of claims 18 to 20, wherein the organic solvent is dimethylformamide (DMF).

22. The composite material of any one of claims 18 to 21 , wherein the derivative of BDC is 2-aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

23. The composite material of any one of the preceding claims, wherein the derivative of UiO-66 MOF is UiO-66-NH2or UiO-66-OH.

24. A method of preparing the composite material of any one of the preceding claims, the method comprising the steps of:(a) providing the porous substrate and a reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;(b) immersing the porous substrate in the reaction mixture from step (a) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for a period of time to provide the composite material.

25. The method according to claim 24, wherein step (b) comprises heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. 200 °C) for about 12 to about 72 h (e.g. about 24 h) to provide the composite material.

26. The method according to claim 24 or 25, wherein the Zn salt and the Zr salt are present in the reaction mixture in step (a) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5).

27. The method according to any one of claims 24 to 26, wherein the organic solvent in the reaction mixture of step (a) is DMF.

28. The method according to any one of claims 24 to 27, wherein the reaction mixture in step (a) further comprises an acid, for example HCI.

29. The method according to any one of claims 24 to 28, wherein following step (b), the method further comprises the steps of:(c) washing the composite material with an organic solvent;(d) providing a second reaction mixture comprising an organic solvent, BDC or derivative thereof, a Zr salt, and a Zn salt;(e) immersing the composite material in the second reaction mixture from step (d) and heating the resulting mixture at a temperature of from about 150 °C to about 250 °C (e.g. about 200 °C) for a period of time to provide the composite material (e.g. for about 12 to about 72 h, for example, about 24 h); and(f) optionally washing the composite material following step (e) with an organic solvent.

30. The method according to claim 29, wherein:- the organic solvent in step c) and / or (f) is DMF, methanol, or a mixture thereof; and / or- the organic solvent in step d) is DMF; and / or- the Zn salt and the Zr salt are present in the second reaction mixture in step (d) at a molar ratio of Zn to Zr from about 0.1 to about 1 , for example, from about 0.1 to about 0.6 (e.g. about 0.2, about 0.3, about 0.4 or about 0.5); and / or- the second reaction mixture in step (d) further comprises an acid, for example HCI.31 . The method according to any one of claims 24 to 30, or the composite material of any one of claims 18 to 23, wherein the Zn salt is Zn(NO3)2.

32. The method according to any one of claims 24 to 31 , or the composite material of any one of claims 18 to 23 or 31 , wherein the Zr salt is ZrCl433. The method according to any one of claims 24 to 32, wherein the derivative of BDC is 2- aminobenzene-1 ,4-dicarboxylic acid (ABDC) or 2-hydroxy-1 ,4-benzenedicarboxylic acid (H2BDC-OH).

34. A composite material obtained by the method according to any one of claims 24 to 33.

35. A method of removing a solute from a solvent, the method comprising passing the solvent through the composite material of any one of claims 1 to 23 or through the composite material of claim 34.

36. The method according to claim 35, wherein the solute has a molecular weight of greater than about 300 g mol-1.

37. The method according to claim 35 or 36, wherein the solvent is water or an organic solvent selected from the group consisting of a monohydric alcohol (e.g. ethanol), a hydrocarbon (e.g. hexane), an ether (e.g. tetrahydrofuran), a ketone (e.g. acetone), and a mixture of two or more thereof.

Images