Crumpled graphene oxide membrane

By forming crumpled GO membranes through controlled deformation on a shrinkable substrate, the permeability-selectivity trade-off is mitigated, achieving high gas transport efficiency and stability for industrial processes.

WO2026235628A1PCT designated stage Publication Date: 2026-11-12NATIONAL UNIVERSITY OF SINGAPORE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/SG2026/050284
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-08
Filing Date
2026-05-05
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Conventional graphene oxide (GO) membranes face a permeability-selectivity trade-off, limiting their scalability and efficiency for industrial gas separation applications due to uncontrolled deformation methods leading to random misalignment and tortuous transport pathways.

Method used

A method involving deposition of GO dispersion on a shrinkable substrate, followed by drying and heating to induce lateral shrinkage, forming a crumpled structure with distinct interlayer spacings and strain-induced out-of-plane deformations, and optionally coating with a protective polymer.

Benefits of technology

The crumpled GO membranes exhibit enhanced gas permeability and selectivity, maintaining structural stability under varying conditions, suitable for industrial applications like hydrogen purification and carbon capture.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure SG2026050284_12112026_PF_FP_ABST
    Figure SG2026050284_12112026_PF_FP_ABST
Patent Text Reader

Abstract

A crumpled graphene oxide (GO) membrane and a method for forming the crumpled GO membrane are provided. In particular, the method enables formation of a crumpled GO membrane having non-uniform interlayer spacings between adjacent GO layers.
Need to check novelty before this filing date? Find Prior Art

Description

CRUMPLED GRAPHENE OXIDE MEMBRANECROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of priority to Singapore patent application no.10202501188X which was filed on 8 May 2025, the contents of which are hereby incorporated by reference in its entirety for all purposes.TECHNICAL FIELD

[0002] This application relates to a crumpled graphene oxide (GO) membrane and a method for forming the crumpled GO membrane. In particular, the method enables formation of a crumpled GO membrane having non-uniform interlayer spacings between adjacent GO layers.BACKGROUND

[0003] Graphene oxide (GO) membranes have attracted considerable attention for gas separation applications due to their ease of fabrication, energy-efficient operation, and ability to provide high selectivity. In particular, lamellar GO membranes comprising stacked GO layers define two-dimensional nanochannels that enable molecular separation based on size and interaction effects. Such membranes have potential for use in a range of applications, including hydrogen purification, carbon capture, ion separation, and nanoreactor systems.

[0004] More recently, those skilled in the art have started employing membrane-based separation technologies in industrial processes, particularly in view of achieving sustainability and decarbonization objectives. Membrane materials proposed by those skilled in the art include graphene, graphene oxide (GO), fullerenes, MXenes, covalent organic frameworks (COFs), and metal-organic frameworks (MOFs), which typically comprise nanochannels having sub-nanometre to nanometre dimensions for selective transport of molecular species.

[0005] Among these materials, GO-based membranes are of particular interest to those skilled in the art due to their tune-able interlayer spacings and ability to form well-defined lamellar structures. In defect-free configurations, nanometre-thick flat GO (fGO) membranes were found to exhibit very high selectivity, for example H2 / CO2selectivity on the order of 1000. However, such lamellar structures inherently impose tortuous transport pathways, as gas molecules must diffuse along and between stacked layers.

[0006] As a result, when the thickness of fGO membranes is increased to levels required for mechanical stability and industrial operation, the permeability of the membranes decreases significantly, often to values below 100 Barrer. This reflects a well-known permeability and selectivity trade-off, in which highly selective membranes exhibit limited gas flux. Consequently, despite their favourable selectivity, conventional lamellar GO membranes are not readily scalable for industrial gas separation processes.

[0007] In view of the restriction mentioned above, various approaches have been proposed by those skilled in the art to modify GO membrane structures in order to improve permeability of such membranes. For example, techniques such as aerosol-assisted capillary compression and spray atomization at elevated temperatures have been employed to form crumpled or disordered GO structures. However, these methods typically rely on uncontrolled capillary and entropic stresses during solvent evaporation. This in turn tends to lead to random misalignment of GO flakes and limited control over the resulting nanochannel structure.

[0008] In view of the challenges mentioned above, those skilled in the art are constantly striving to design an improved GO membrane structure and a method for forming such a membrane, which enables enhanced gas permeability while maintaining high selectivity, and is suitable for scalable and reproducible fabrication.SUMMARY

[0009] In one aspect, the present disclosure describes a method for forming a crumpled graphene oxide (GO) membrane. The disclosed method comprises the steps of depositing a GO dispersion onto a shrinkable substrate to form a laminated GO film comprising a plurality of stacked GO layers, drying the laminated GO film to form a planar GO membrane, heating the shrinkable substrate to induce lateral shrinkage of the substrate, such that a compressive strain is applied to the GO membrane during shrinkage of the substrate, and deforming the GO membrane, in response to the application of the compressive strain, to form a crumpled structure comprising a plurality of strain-induced out-of-plane deformations, wherein the crumpled structure exhibits at least two distinct interlayer spacings between adjacent GO layers.

[0010] In embodiments of the one aspect, the method further comprises the step of depositing a protective polymer layer onto the crumpled structure of the deformed GO membrane.

[0011] In embodiments of the one aspect, the method further comprises the step of removing the shrinkable substrate to obtain a free-standing crumpled GO membrane. In embodiments, the protective polymer layer may comprise poly(vinyl alcohol) (PVA), and the step of removing the shrinkable substrate may comprise the step of dissolving the substrate using an organic solvent comprising ethyl acetate.

[0012] In embodiments of the one aspect, the laminated GO film has a thickness in a range between 1pm and 3pm. In embodiments, the step of depositing the GO dispersion onto the shrinkable substrate may comprise blade casting the GO dispersion onto the shrinkable substrate. In embodiments, the shrinkable substrate may comprise polystyrene.

[0013] In embodiments of the one aspect, the shrinkable substrate is plasma-treated prior to the step of depositing the GO dispersion onto the shrinkable substrate. In embodiments, the step of heating the shrinkable substrate is performed at a temperature range between 120°C and 160°C.

[0014] In embodiments of the one aspect, the at least two distinct interlayer spacings may comprise a first interlayer spacing having a spacing between 8.0 A and 8.7 A, and a second interlayer spacing having a spacing between 9.0 A and 10.0 A. Further, the crumpled structure may comprise in-plane pores having a size in a range between 2.5 A and 3.5 A. In embodiments, the shrinkable substrate may be shrunk by at least 50%.

[0015] In embodiments of the one aspect, the first interlayer spacing defines a selective region configured for molecular sieving; and the second interlayer spacing defines a transport region configured to facilitate gas diffusion.

[0016] In embodiments of the one aspect, the GO dispersion may comprise GO flakes dispersed in an aqueous solution, the flakes having a lateral size between 0.5 pm and 2.0 pm, and the dispersion having a concentration between 10 mg / ml and 50 mg / ml.

[0017] In embodiments of the another aspect, a crumpled graphene oxide (GO) membrane is disclosed. The disclosed crumpled GO membrane comprises a plurality of stacked GO layers arranged in a laminated structure that is deformed into a crumpled structure having a plurality of strain-induced out-of-plane deformations, wherein the crumpled structure exhibits at least two distinct interlayer spacings between adjacent GO layers.BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Various embodiments of the present disclosure are described below with reference to the following drawings:Figure 1 illustrates a block diagram showing the process for forming a crumpled GO membrane in accordance with embodiments of the present disclosure;Figure 2 illustrates a block diagram showing the process for preparing the crumpled GO membrane illustrated in Figure 1 for measurement;Figures 3a and 3b illustrate schematics comparing the gas transport pathways of a crumpled and flat surface GO membrane;Figure 4 illustrates an exemplary casting blade used to deposit the GO dispersion onto the shrinkable substrate in accordance with embodiments of the present disclosure;Figure 5 illustrates the water contact angle of polystyrene substrate before and after the substrate has been plasma treated;Figure 6 illustrates SEM image of the surfaces and cross sections of the crumpled GO membrane;Figure 7 illustrates SEM cross section images of crumpled GO membranes labelled as cGO1, cGO2, cGO3, cGO4, cGO5 and flat GO membrane fGO1;Figure 8 illustrates attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectra that display the characteristic peaks in all membranes, including O-H (3225 cm-1), C=O (1786 cm-1), C=C (1620 cm'1), and C-0 (1068 cm'1) bonds which belonged to the GO membrane; Figure 9 illustrates digital images of free-standing robust crumpled GO membranes that were bent using tweezers;Figure 10a illustrates a digital image of selected regions on top and bottom surface of the cGO4 membrane for XRD mapping measurements and the XRD spectra of the selected regions on top and bottom surfaces of cGO4 membrane.Figure 10b illustrates the tensile strength and elongation of the cGO membranes;Figure 11a illustrates the pure He, N2 and CH4permeability through the cGO membranes having different roughness at a temperature of 21 °C and a pressure of 1 bar;Figure 11b illustrates pure H2 and CO2 permeability and H2 / CO2 selectivity of the cGO membranes at 21 °C and 1 bar;Figure 12a illustrates a comparison of H2 and CO2 permeability between flat - fGO4 and crumpled membrane - cGO4;Figure 12b illustrates a comparison of H2 and CO2 permeability between flat - fGO1 and crumpled membrane - cGO4;Figure 13 illustrates H2 / CO2 separation performance of flat and crumpled membranes compared with the benchmark of gas separation membranes;Figure 14 illustrates a plot showing the pure-gas permeability through cGO4 membranes as a function of the gas kinetic diameter at 21 °C and 1 bar;Figure 15a illustrates a plot showing the pure and equimolar mixed gas permeation tests of cG04 under different temperatures and 1 bar;Figure 15b illustrates a plot showing the H2 / CO2 separation performance of cGO4 membranesin the temperature range between -20 °C and 80 °C;Figure 15c illustrates a plot showing the pure gas permeation tests of cGO4 under humidified environments and 1 bar pressure;Figure 16 illustrates plots showing the pure and equimolar mixed gas permeation tests of cGO4 under increased and decreased temperatures and under increased and decreased humidity environments;Figure 17a illustrates plots of the XRD spectra of cGO4 membranes in different humidity conditions;Figure 17b illustrates plots of the interlayer spacing of cGO4 membranes in different humidity conditions;Figure 18a illustrates plots showing the long-term stability test of separating equimolar dry H2 / CO2 mixture of cGO4 at 21 °C and 1 bar;Figure 18b illustrates plots showing the long-term stability test of separating equimolar humidified H2 / CO2 mixture (RH 88%) of cGO4 at 21 °C and 1 bar;Figure 19 illustrates plots showing the effect of H2 concentration in the feed gas on the separation performance of cGO4 membrane with mixed H2 / CO2 gas feeding;Figure 20 illustrates plots showing the permeability and selectivity of H2 / CO2 mixture with 0.9 H2 molar ratio at elevated pressures, and the total cost with the CO2 removal rate of 1 kg / h from H2 for one-hour operation under different feed pressure;Figure 21a illustrates bar graphs showing the required membrane areas of fGO and cGO to reach H2 permeate purity >99.5 vol.% with the CO2 removal rate of 1 kg / h under varying feed pressure;Figure 21b illustrates bar graphs showing the H2 purity in the permeate separated according to GO and cGO under varying feed pressure;Figure 22 illustrates a cross-sectional schematic of an exemplary membrane holder in accordance with embodiments of the disclosure;Figure 23a illustrates SEM images (left) and confocal microscopy images (right) of the cGO and fGO membranes - cGO1, cGO2, cGO3, cGO4, cGO5, and fGO1;Figure 23b illustrates a plot of tensile strength and elongation of the fGO membranes plotted against the thickness of fGO membranes;Figure 23c illustrates a plot showing the correlation between the arithmetical mean height of the surface (Sa) and curvature x surface area index (CAI) of cGO membranes plotted against the thickness of fGO membranes;Figure 24 illustrates an SEM image showing macroscopic defects on the cGO5 membranes due to the overstretch of the GO flakes;Figure 25a illustrates plots of the XRD spectra for cGO4 and fGO1 membranes;Figure 25b illustrates plots of the interlayer spacing of the cGO and fGO1 membranes calculated from the XRD fitting data;Figure 25c illustrates plots of the XRD spectra for various cGO membranes;Figure 26a illustrates the random configuration of the molecules in a domain with decreased interlayer distance where the oxidized and graphene-like regions intersect each other;Figure 26b illustrates the ordered configuration of the molecules in a domain with increased interlayer spacing where there is a fully oxidized domain surrounded by a graphene-like region; Figure 27 illustrates the chemical composition of cGO as analysed using XPS data from O1s and C1s spectra;Figure 28 illustrates Raman spectroscopy-derived IDI / IG ratio as plotted for cGO membranes; Figure 29 illustrates plots showing the pore size distribution of cGO1, cGO2, cGO3, cGO4, and cGO5 membranes;Figure 30 illustrates TEM and STEM images of cGO membranes on silicon wafer prepared by focused ion beam (FIB);Figure 31 illustrates plots of the diffusivity of H2 and CO2 in cGO membranes across different thickness at 21 °C and 1 bar;Figure 32 illustrates plots of the solubility of H2 and CO2 in cGO membranes across different thickness at 21 °C and 1 bar; andFigure 33 illustrates a flowchart of a process for forming a crumpled GO membrane in accordance with embodiments of the disclosure.DETAILED DESCRIPTION

[0019] The following detailed description is made with reference to the accompanying drawings, showing details and embodiments of the present disclosure for the purposes of illustration. Features that are described in the context of an embodiment may correspondingly be applicable to the same or similar features in the other embodiments, even if not explicitly described in these other embodiments. Additions and / or combinations and / or alternatives as described for a feature in the context of an embodiment may correspondingly be applicable to the same or similar feature in the other embodiments.

[0020] In the context of various embodiments, the articles “a,” “an” and “the” as used with regard to a feature or element include a reference to one or more of the features or elements.

[0021] In the context of various embodiments, the term “about” or “approximately” as applied to a numeric value encompasses the exact value and a reasonable variance as generally understood in the relevant technical field, e.g., within 10% of the specified value.

[0022] As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items.

[0023] As used herein, “comprising” means including, but not limited to, whatever follows the word “comprising.” Thus, use of the term “comprising” indicates that the listed elements are required or mandatory, but that other elements are optional and may or may not be present.

[0024] As used herein, “consisting of’ means including, and limited to, whatever follows the phrase “consisting of.” Thus, use of the phrase “consisting of’ indicates that the listed elements are required or mandatory, and that no other elements may be present.

[0025] In the context of various embodiments, the term “disposed on" relates to the placement or deposition of one material or layer onto the surface of another and may involve one or more types of deposition techniques.

[0026] In the context of various embodiments, the directional terms mentioned herein, such as “above” and “below” or “upper” and “lower” refer to directions as described with reference to the drawings. Therefore, the directional terms are only used for illustration and are not meant to limit the present disclosure.

[0027] Additionally, for the sake of brevity, extensive explanations of conventional techniques of fabricating semiconductor devices and integrated circuits are not described in detail herein. The tasks and processes described herein may also be integrated into a more comprehensive procedure with extra steps of features that are not elaborated upon in this document. Specifically, certain processes of fabricating semiconductor devices are well known to one skilled in the art hence, such processes will be omitted entirely.

[0028] The present disclosure describes a method for enhancing both permeability and selectivity of graphene oxide (GO) membranes while maintaining structural stability. In particular, the method enables formation of crumpled GO membranes having controlled gas diffusion pathways defined by a multidomain anisotropic structure. The resulting crumpled GO membranes exhibit significantly improved gas transport properties compared to interlayer spacings d of conventional planar or flat lamellar GO membranes. As a non-limiting example, the crumpled GO membranes demonstrate a hydrogen (H2) permeability of approximately 2.1 x 104Barrer, which is substantially higher than that of flat lamellar GO membranes, typically having permeability values below 100 Barrer. At the same time, the membranes maintain high selectivity, for example an H2 / CO2selectivity of approximately 91, thereby exceeding the performance of existing membrane technologies. The crumpled GO membranes disclosed herein were also found to exhibit robust performance under harsh environmental conditions, including in temperatures as low as -20°C and relative humidity levels of up to 96%. Such stability is particularly advantageous for industrial applications, where membranes are required to operate under variable and demanding conditions.

[0029] Accordingly, the disclosed cGO membranes provide improved durability, scalability, and process compatibility, while effectively mitigating the conventional permeability / selectivity trade-off associated with lamellar GO membranes. As a result, the disclosed membranes offer a viable and efficient solution for gas separation applications, including hydrogen purification and carbon capture.

[0030] It is useful to note that molecular transport phenomena in nanomechanically modified GO structures have not been extensively characterized by those skilled in the art. In the present disclosure, it is demonstrated that in crumpled GO (cGO) membranes, the corrugated structure appears to alter the configuration and size of transport channels, thereby enabling simultaneous enhancement of gas permeability and selectivity. In contrast to conventional flat lamellar GO membranes, in which interlayer channels extend substantially parallel to the membrane plane, the introduction of wrinkling and folding results in a structural rearrangement that produces nanochannels of differing sizes and characteristics. In particular, larger nanochannels facilitate rapid gas uptake and release, while narrower channels define selective regions that enable molecular sieving.

[0031] Additionally, the presence of at least two distinct interlayer spacings within the cGO membrane were found to be consistent with predictions from classical elasticity theory and is supported by experimental observations. In particular, the at least two distinct interlayerspacings comprise a first interlayer spacing (di) corresponding to a relatively smaller interlayer distance, and a second interlayer spacing (d2) corresponding to a relatively larger interlayer distance. When bending energy is taken into account, inhomogeneous in-plane deformation gives rise to spatial variations in interlayer spacing. The formation of discrete interlayer spacings is attributed to realignment of functional groups within the GO layers and associated confinement of water molecules, which stabilizes specific structural configurations. Experimental data (which will be described in greater detail in the later sections) is consistent with predicted behaviour of confined water in ordered and disordered oxidized domains, thereby providing an integrated understanding of the relationship between mechanical deformation, interlayer structure, and gas transport properties in GO membranes.

[0032] Figure 1 illustrates a process for preparing a cGO membrane in accordance with embodiments of the present disclosure. Process 100 for fabricating the cGO membrane begins with the deposition of a GO dispersion comprising GO flakes onto a shrinkable substrate to form a laminated GO film comprising a plurality of stacked GO layers. This takes place at step 102. As a non-limiting example, a 30 mg / ml aqueous GO dispersion comprising GO flakes having a lateral size of approximately 1 pm is deposited onto a plasma-treated polystyrene (PS) substrate by blade casting. The deposited GO dispersion is then dried to form a flat GO membrane (fGO membrane) or a planar GO membrane having a thickness in a range between 1 pm and 3pm, or preferably between 1 pm and 2.6pm.

[0033] At step 104, the fGO membrane is then subjected to heating to induce lateral shrinkage of the shrinkable substrate, thereby applying a compressive strain to the GO membrane. In one implementation, the GO membrane disposed on the PS substrate is heated to approximately 140°C for approximately 10 minutes in a convection oven, causing thermal shrinkage of the PS substrate. During shrinkage, the substrate contracts laterally and compresses the fGO membrane, resulting in deformation of the fGO membrane to form a crumpled structure comprising a plurality of strain-induced out-of-plane deformations. The formation of the crumpled GO membrane is then confirmed by scanning electron microscope (SEM) measurements and attenuated total reflection-Fourier transform infrared (ATR-FTIR) analysis.

[0034] Following thermal treatment, the fGO membrane retracts to approximately one-third of its original lateral dimensions or half its original lateral dimensions. In embodiments of the disclosure, the shrinkable substrate may then be subsequently removed, for example by dissolving the PS substrate, to obtain a free-standing flexible crumpled GO membrane. Thistakes place at step 106. The described process is compatible with scalable manufacturing techniques and may be integrated into automated production systems for membrane-based applications, including gas separation and decarbonization.

[0035] For preparation of gas permeation samples, the free-standing crumpled GO membrane may then be laminated to form a sealed structure suitable for testing. Figure 2 illustrates process 200 for preparing cGO membranes for gas permeation analysis. In particular, at step 202, the crumpled GO membrane is laminated with a flexible polyethersulfone (PES) support and an impermeable perforated laminating pouch to provide an airtight assembly. The crumpled GO membrane is then positioned below the plane of the laminating pouch to preserve the crumpled structure during lamination. At step 204, a UV-sensitive resin is applied to seal any gaps between the crumpled GO membrane and the laminating pouch, and the resin is subsequently cured at step 206 under ultraviolet (UV) irradiation to complete the sample preparation for gas permeation measurements.

[0036] Figure 3a illustrates a crumpled GO membrane 302 comprising a crumpled structure formed by strain-induced out-of-plane deformations, resulting in a structural rearrangement of the stacked GO layers. Unlike flat graphene oxide (fGO) membranes having a lamellar structure in which interlayer channels extend substantially parallel to the membrane plane, this rearrangement 304 produces a plurality of nanochannels having at least two distinct interlayer spacings and differing functional characteristics. As illustrated in schematic 304 of Figure 3a, it can be seen that the crumpled GO membrane comprises regions defining relatively larger interlayer spacings that facilitate rapid gas uptake and release, and regions defining relatively smaller interlayer spacings that provide selective transport pathways for molecular sieving.

[0037] In contrast, as illustrated in Figure 3b, conventional flat GO membranes 306 comprise a planar lamellar structure 308 in which transport occurs through substantially parallel interlayer channels, resulting in limited permeability due to tortuous diffusion paths. In the crumpled GO membrane, the presence of the at least two distinct interlayer spacings enable functional separation of transport and selectivity, wherein regions having larger interlayer spacing define transport pathways, and regions having smaller interlayer spacing define selective filtering pathways. Accordingly, the crumpled GO membrane provides enhanced gas permeability while maintaining high selectivity relative to flat lamellar GO membranes.

[0038] Figure 4 illustrates an example of a blade casting apparatus used for depositing a GO dispersion onto a substrate in accordance with embodiments of the present disclosure. The apparatus comprises casting blade 402 positioned above a substrate, the blade being configured to define a controlled gap between the blade and the substrate surface. As shown in magnified view 404, the gap determines the thickness of the deposited GO layer by regulating the volume of GO dispersion passing beneath the casting blade during the casting process. The apparatus further includes adjustable elements configured to control the position of the casting blade relative to the substrate, thereby enabling precise control of the gap and ensuring uniform formation of a laminated GO film.

[0039] Figure 5 illustrates the effect of plasma treatment on the surface wettability of a polystyrene (PS) substrate, as characterised by water contact angle (WCA) measurements. As shown in image 502 of Figure 5, the untreated PS substrate exhibits a relatively high contact angle of approximately 86.9°, indicating a hydrophobic surface. In contrast, as shown in image 504 of Figure 5, the plasma-treated PS substrate was found to exhibit a significantly reduced contact angle of approximately 22.6°, indicating increased surface wettability. The reduction in contact angle demonstrates that plasma treatment modifies the surface properties of the PS substrate, thereby improving its suitability for deposition of a GO dispersion. In particular, the increased wettability facilitates uniform spreading of the GO dispersion during blade casting, as described with reference to Figure 4, thereby promoting formation of a uniform laminated GO film. Furthermore, the improved spreading of the GO dispersion enables more consistent control of the film thickness defined by the gap between the casting blade and the substrate resulting in a more uniform laminated GO film.

[0040] The morphological differences between the flat GO membrane and the crumpled GO membrane are illustrated by scanning electron microscopy (SEM) images shown in Figure 6. In particular, SEM images 602 and 606 illustrate cross-sectional views of the crumpled GO membrane while SEM image 604 illustrates a surface view of the crumpled GO membrane. It can be seen from these SEM images that the stacked GO layers are deformed into a crumpled structure comprising a plurality of strain-induced out-of-plane deformations, resulting in a highly corrugated and folded morphology. These deformations give rise to a non-uniform surface topology and variation in interlayer spacing across the membrane. In contrast, SEM image 608 shows the top view and SEM image 610 shows the cross-section view of the flat GO membrane, which exhibits a substantially smooth surface and a lamellar structure with stacked GO layers arranged in a substantially parallel configuration. These SEM images demonstrate that the crumpled GO membrane exhibits a significantly altered morphologyrelative to the flat or planar GO membrane, consistent with the formation of the crumpled structure defined by strain-induced deformation.

[0041] It is useful to note that among the cGO membranes prepared using laminated GO films having different thicknesses, the membranes denoted as cGO1, cGO2, cGO3, cGO4, and cGO5 correspond to laminated GO films having thicknesses of approximately 1 ± 0.1μm, 1.3 ± 0.1μm, 1.6 ± 0.1μm, 2.4 ± 0.1μm , and 2.6 ± 0.1μm, respectively. Figure 7 illustrates SEM images showing cross-sectional morphologies of the crumpled GO membranes formed from these laminated GO films, together with a comparative flat GO membrane. In particular, SEM images 702, 704, 706, 708, and 710 correspond to the crumpled GO membranes cGO1, cGO2, cGO3, cGO4, and cGO5, respectively, each exhibiting a crumpled structure comprising a plurality of strain-induced out-of-plane deformations of stacked GO layers. The images demonstrate that, as the thickness of the laminated GO film increases, the resulting crumpled structure exhibits progressively larger and more pronounced folds and corrugations. In contrast, SEM image 712 illustrates a flat GO membrane (fGO1), which exhibits a substantially smooth lamellar structure with stacked GO layers arranged in a substantially parallel configuration and without the presence of strain-induced out-of-plane deformations. The comparison highlights the structural differences between flat and crumpled GO membranes and illustrates the effect of thickness on the formation of the crumpled structure.

[0042] As mentioned in the earlier sections, the deformation of the fGO membrane under compressive strain results in the formation of the crumpled structure, which is further characterised using attenuated total reflection-Fourier transform infrared (ATR-FTIR) analysis, as illustrated in Figure 8. The ATR-FTIR spectra for the crumpled GO membranes cGO1 to cGO5 exhibit characteristic absorption bands corresponding to functional groups present in the GO layers, including O-H stretching vibrations, C=O functional groups, and C-0 bonding. Variations in the intensity and distribution of these absorption bands across the cGO samples indicate changes in the chemical structure associated with the deformation process. In particular, the compressive strain and resulting crumpling are associated with modification and partial redistribution of oxygen-containing functional groups within the GO layers, thereby reflecting structural and chemical changes accompanying formation of the crumpled GO membrane.

[0043] After thermal treatment, the membrane retracts to one-third of its original size. Once the PS template is dissolved, a free-standing flexible crumpled membrane is then obtained asillustrated in images 902, 903, 904, and 905 shown in Figure 9. The streamlined nature of the process allows for easy integration into existing manufacturing lines, enabling robotization and automation for industries focused on gas separation, decarbonization, and other membranebased technologies.

[0044] Figure 10a illustrates characterization of the cGO membrane cGO4 using X-ray diffraction (XRD) mapping. In particular, the left-hand image 1002 shows a digital image identifying selected regions on the top and bottom surfaces of the cGO4 membrane for XRD measurements, while the right-hand image 1004 shows corresponding XRD spectra obtained from these selected regions. The XRD spectra exhibit characteristic peaks corresponding to a first interlayer spacing (d1) and a second interlayer spacing (d2), as defined in previous sections. To recap, the first interlayer spacing (d1), corresponds to a relatively smaller interlayer distance, and the second interlayer spacing (d2), corresponds to a relatively larger interlayer distance. These interlayer spacings are consistently observed across the selected regions, indicating that the crumpled structure comprises a plurality of regions exhibiting at least two distinct interlayer spacings.

[0045] As can be seen in Figure 10a, plot 1004 illustrates X-ray diffraction (XRD) spectra obtained from selected regions on the top surface and the bottom surface of the cGO membrane. The spectra correspond to multiple measurement locations, labelled as Top1-Top3 and Bottom 1-Bottom3, and are plotted as intensity (a.u.) as a function of diffraction angle (29). Each spectrum exhibits two distinct peaks corresponding to a first interlayer spacing (d1) and a second interlayer spacing (d2). The peak associated with the first interlayer spacing (d1) corresponds to a relatively smaller interlayer distance, while the peak associated with the second interlayer spacing (d2) corresponds to a relatively larger interlayer distance. The consistent presence and relative positions of the d1and d2peaks across both the top and bottom measurement regions indicate that the crumpled GO membrane exhibits a substantially uniform distribution of the at least two distinct interlayer spacings throughout the membrane thickness.

[0046] Figure 10b illustrates the mechanical properties of the cGO membranes, showing tensile strength and elongation at break for membranes cGO1 to cGO5. As shown, the crumpled GO membranes exhibit tensile strengths in a range of approximately 0.3 MPa to 0.5 MPa and elongation at break in a range of approximately 20% to 30%. The data further indicates that, with increasing thickness of the laminated GO film, the tensile strength generally decreases while the elongation increases, reflecting the influence of the crumpled structureon mechanical behaviour. Despite deformation during the crumpling process, the membranes maintain robust mechanical properties.

[0047] The retention of tensile strength and elongation following formation of the crumpled structure demonstrates that the crumpled GO membranes possess sufficient mechanical integrity for handling and integration into membrane-based systems. These mechanical characteristics indicate that the fabrication process yields membranes that are scalable, reproducible, and durable, thereby supporting their suitability for practical applications, including gas separation and related industrial processes.

[0048] Figure 11a illustrates gas permeability of the cGO membranes cGO1 to cGO5 for different gases, including helium (He), methane (CH4), and nitrogen (N2), as a function of membrane thickness at 21 °C and 1 bar pressure, and Figure 11b illustrates pure H2and CO2permeability and H2 / CO2selectivity of the cGO membranes at 21 °C and 1 bar pressure as a function of membrane thickness. The data shown in these two figures demonstrate that gas permeability increases with increasing thickness from cGO1 to cGO4, reaching a maximum at cGO4, and subsequently decreases for cGO5.

[0049] In particular, the cGO4 membrane exhibits the highest permeability across the tested gases, indicating that the crumpled structure formed from a laminated GO film of approximately 2.4μm thickness provides an optimal balance between structural integrity and transport pathways. The observed increase in permeability is attributed to the formation of a crumpled structure comprising a plurality of strain-induced out-of-plane deformations and at least two distinct interlayer spacings, which together facilitate enhanced gas transport. The subsequent decrease in permeability observed for cGO5 suggests that excessive thickness may lead to structural irregularities that reduce effective transport pathways. Accordingly, the results demonstrate that membrane thickness is a key parameter in controlling permeability and selectivity of the crumpled GO membranes. The gas separation performance of cGO membranes in pure gas permeation tests is set out in Table 1 below.Table 1

[0050] Figure 12a illustrates a comparison of H2and CO2permeability between flat graphene oxide (GO) membranes and crumpled GO membranes. In particular, flat GO membrane fGO4 exhibits permeability values for both H2and CO2that are close to the detection limit, indicating significantly restricted gas transport through the lamellar structure. In contrast, crumpled GO membrane cGO4 exhibits substantially increased permeability for both H2and CO2, with H2permeability reaching values on the order of 104Barrer and CO2permeability also showing a corresponding increase. The comparison demonstrates that deformation of the flat GO membrane into a crumpled structure comprising a plurality of strain-induced out-of-plane deformations and at least two distinct interlayer spacings results in a significant enhancement in gas transport. Accordingly, the crumpled GO membrane provides markedly improved permeability relative to the flat GO membrane, while maintaining separation capability between H2and CO2.

[0051] Figure 12b illustrates a comparison of H2 and CO2 permeability between flat GO membrane fGO1 and crumpled GO membrane cGO4. The data demonstrates that the crumpled GO membrane exhibits substantially higher permeability for both H2 and CO2 relative to the flat GO membrane. In particular, the fGO1 membrane exhibits relatively low permeability values for both gases, whereas the cGO4 membrane exhibits permeability values increased by several orders of magnitude, consistent with the formation of a crumpled structure comprising a plurality of strain-induced out-of-plane deformations and at least two distinct interlayer spacings. The permeability data is presented as mean values with associated standard deviation (SD), wherein each bar represents measurements obtained from n = 6 independent replicates corresponding to independently fabricated membranes from separate synthesis batches, without pooling of technical replicates. The consistency of the measurements across independent replicates further confirms the reproducibility of the crumpled GO membrane fabrication process and the reliability of the observed enhancement in gas transport properties. The results show that the permeability of the cGO4 membrane is almost four orders of magnitude higher (H2 permeability - 2.1*104Barrer) - to that of the 1-pm-thick fGOI (H2 permeability < 10 Barrer) membrane.

[0052] Figure 13 provides a benchmark comparison of gas separation performance for GO membranes relative to existing membrane technologies. The plot illustrates H2 permeability as a function of selectivity for various membrane materials. In particular, data pointscorresponding to crumpled GO membranes (labelled as stars - 1303) demonstrate that the disclosed membranes achieve a combination of high permeability and high selectivity that surpasses other two-dimensional (2D) material-based membranes (labelled as squares -1301) and significantly outperforms conventional polymer-based gas separation membranes (labelled as circles - 1302). The benchmark further includes performance limits defined by established upper bounds, wherein the solid line represents the Robeson 2008 upper bound for polymeric membranes, and the dashed line represents the updated 2021 upper bound for advanced gas separation membranes. As shown, the data points associated with the crumpled GO membranes (labelled as 1303) approach or exceed these upper bounds, thereby demonstrating superior separation performance relative to state-of-the-art membrane technologies.

[0053] Figure 14 illustrates the permeability performance of the crumpled GO membrane cGO4 when it was evaluated at a pressure of approximately 1 bar and at room temperature. The results demonstrate that gas permeability exhibits a dependence on the kinetic diameter of the gas molecules, wherein smaller molecules such as H2 and He exhibit significantly higher permeability compared to larger molecules such as CO2, N2, and CH4. The observed behaviour includes a distinct cutoff in permeability as the kinetic diameter increases, indicating that the crumpled GO membrane operates via a molecular sieving mechanism based on size exclusion.

[0054] In particular, the cGO4 membrane exhibits an H2 / CO2 selectivity of approximately 91 under pure gas conditions, and a separation factor of approximately 51 under equimolar mixed gas conditions, both of which exceed the corresponding Knudsen diffusion coefficient of approximately 4.69. The reduced selectivity observed under mixed gas conditions is attributed to competing diffusion effects between gas species within the nanochannels, as is characteristic of membranes comprising two-dimensional transport pathways. These results demonstrate that the crumpled GO membrane provides effective molecular discrimination while maintaining high permeability, consistent with transport through a structure comprising at least two distinct interlayer spacings and associated nanochannel configurations. The gas separation performance of cGO4 membrane in both pure gas and equimolar mixed gas permeation tests are set out in Table 2 below.Table 2

[0055] The permeability and selectivity of the crumpled GO membrane remain substantially stable across a temperature range from approximately -20°C to 80°C, as illustrated in Figures 15a and 15b. In particular, the permeability of H2 and CO2 exhibits only minor variation over the tested temperature range under both pure gas and mixed gas conditions, while the H2 / CO2 selectivity remains relatively consistent. The results further demonstrate that similar trends are observed during both increasing and decreasing temperature cycles, indicating that the membrane performance is reversible and not subject to degradation under thermal cycling. As shown in Figure 15b, at -20°C, where conventional polymer membranes typically experience phase transitions affecting their structural integrity, it can be seen that the selectivity dropped from 92 to 43 in pure gas tests and from 51 to 22 in mixed gas tests, likely due to the impact of low temperatures on the water in 2D nanochannels of the GO. Additionally, Figure 15c which illustrates a plot of the pure gas permeation tests of cGO4 under humidified environments and 1 bar pressure show that both the permeability and selectivity of the crumpled GO membrane remain stable in humidified gases with a relative humidity up to 96%.

[0056] The stability of permeability and selectivity across the temperature and relative humidity ranges as illustrated in respective plots 1601 and 1602 in Figure 16 indicate that the gas transport properties of the crumpled GO membrane are governed primarily by its intrinsic structural characteristics, including the presence of a crumpled structure comprising a plurality of strain-induced out-of-plane deformations and at least two distinct interlayer spacings. As a result, the membrane maintains reliable separation performance under varying environmental conditions. This thermal and relative humidity stability, together with consistent performance under both pure and mixed gas conditions, demonstrates the robustness of the crumpled GO membrane and supports its suitability for practical gas separation applications.

[0057] Figure 17a illustrates X-ray diffraction (XRD) analysis of the crumpled GO membrane cGO4 under varying relative humidity (RH) conditions. As shown, the XRD spectra exhibit peaks corresponding to the first interlayer spacing (d1) and the second interlayer spacing (d2), across a range of humidity levels from approximately 11% to 100% RH. Figure 17b in turn shows the interlayer spacings of the cGO4 membrane as a function of relative humidity. The results demonstrate that both the first interlayer spacing (d1) and the secondinterlayer spacing (d2) increase with increasing humidity, with a maximum increase of approximately 0.5 A observed up to 100% RH.

[0058] The limited variation in interlayer spacing for the cGO4 membrane under humid conditions indicates that the crumpled structure maintains a relatively stable interlayer architecture despite the presence of water molecules within the nanochannels. In contrast, flat GO membranes typically exhibit significant degradation in gas separation performance under humid conditions due to expansion of interlayer spacing caused by water intercalation. The comparatively small change in interlayer spacing observed for the cGO4 membrane demonstrates enhanced resistance to humidity-induced swelling, thereby enabling stable and reliable gas separation performance under high-humidity environments.

[0059] Long-term separation tests performed by separating equimolar dry H2 / CO2 mixture of cGO4 at 21 °C and 1 bar pressure under both dry and humidified environments is illustrated in Figure 18a and a long-term stability tests performed by separating equimolar humidified H2 / CO2 mixture (RH 88%) of cGO4 at 21 °C and 1 bar pressure is illustrated in Figure 18b. These results demonstrated unwavering cGO membrane performance over 100 hours, with no detectable deterioration.

[0060] Figure 19 illustrates the effect of hydrogen (H2) concentration in the feed gas on the separation performance of the crumpled GO membrane cGO4 under mixed H2 / CO2 gas conditions. As shown, the permeability of H2 increases with increasing H2 molar fraction in the feed gas, while the permeability of CO2 remains relatively low. The H2 / CO2 selectivity correspondingly increases with increasing H2 concentration and reaches a peak value of approximately 113.5 when operating with a gas mixture comprising about 90% H2and 10% CO2. This trend indicates that the crumpled GO membrane exhibits enhanced separation performance under high H2molar fraction conditions.

[0061] The plot further identifies operational regimes corresponding to different industrial applications. In particular, a lower H2molar fraction region corresponds to conditions relevant to carbon capture and utilization (CCU), while a higher H2molar fraction region corresponds to conditions relevant to steam methane reforming (SMR). The high selectivity achieved at elevated H2concentrations demonstrates that the crumpled GO membrane is particularly suitable for hydrogen purification processes, including SMR, where high H2purity is required. The observed performance is attributed to the molecular sieving behaviour of the membranearising from its crumpled structure comprising a plurality of strain-induced out-of-plane deformations and at least two distinct interlayer spacings.

[0062] The cost of decarbonization simulation

[0063] To assess the industrial applicability of both fGO and cGO membranes for H2 purification, their performances were analysed using experimental data on mixed H2 and CO2 permeabilities. Specifically, to evaluate the performance of fGO and cGO membranes under industrially relevant conditions, a detailed analysis of their separation capabilities for hydrogen purification was conducted. Utilizing the experimental data on mixed H2 and CO2 permeabilities, the separation process were modelled at 298 K, under feed pressures of 1.5 bar, 3 bar, and 5 bar, and a permeate pressure of 1 bar. The feed gas composition was 90 vol. % H2 and 10 vol. % CO2. The target H2 purity was set at >99.5 vol.% with the CO2 removal rate of 1 kg / h. For the simulations, a counter-current flow configuration in a flat-sheet membrane module was considered and the required membrane area, H2 permeate purity, and feed stream compression energy costs were calculated.

[0064] The flux J of gas through the membrane is calculated by:J=~ ’ (Pfeed—Pperm) U )where J is the flux (mol / (m2·s)), P is the permeability (mol·m / (m2·s·Pa)), l is the membrane thickness (m), pfeedand ppermrepresent the partial pressures of the gas on the feed and permeate sides, respectively, in Pascals (Pa).

[0065] The flux equation is stated for a solution-diffusion membrane permeation model, assuming Fick’s law applies and there is no concentration polarization. Ideal gas behaviour is assumed for both feed and permeate streams.

[0066] The separation process in each membrane was modelled as a counter-current flow system, dividing the membrane into 100 segments to iteratively calculate flux, composition, and partial pressures along the membrane length. The objective was to achieve a CO2 removal rate of 1 kg / h through the retentate stream. The total membrane area Atotai is divided equally among segments:> totalAsegment ~N(2)where N is the total number of segments. Asegmentremains fixed in each segment as we iterate Atotalto meet the desired CO2removal rate and H2purity.

[0067] For each segment / , the retentate molar flow rates of H2 and CO2 were updated as follows:^H2,ret,t+1—^H2,ret,t JH2’ -^segment (3) l-CO2,ret,t+l=^CO2,ret,t—JcO2’ -^segment (4) The partial pressures for H2 and CO2 in both feed and permeate sides of each segment were calculated using the updated mole fractions and total pressures:The fluxes JH₂and JCO₂were calculated using the partial pressure difference across the membrane:where pfeed, H₂and pfeed, CO₂were updated at each segment based on the updated compositions.

[0068] For each segment i, the permeate flow rates of H2 and CO2 are updated based on the fixed segmental area:^H2,perm,t+1=^H2,perm,t T / H2’ -^segment (11)^CO2,perm,t+l=^CO2,perm,t T JcO2’ -^segment (12)The updated permeate mole fractions for H2and CO2are calculated as:yCO2,perm,i+1 = 1 − yH2,perm,i+1 ("14)

[0069] The total membrane area Atotalis iteratively adjusted until the target CO2 removal rate in the retentate and H2 purity in the permeate are achieved.

[0070] The overall H2 purity in the permeate stream is calculated after the iteration process:The energy required to compress the feed gas from 1 bar to the specified feed pressures (1.5, 3, 5 bar) was calculated using the isentropic model:where ṅfeedis total molar flow rate of the feed gas (mol / s), R is universal gas constant with 8.314 J / (mol·K), T is temperature in Kelvin (K), y is heat capacity ratio (1.4 for diatomic gases like H2) and ηcompressoris compressor efficiency (assumed to be 0.75).

[0071] The total cost of decarbonization with CO2 removal rate of 1 kg / h (Ccost) was calculated with:Ccost= W · Celectricity+ Atotal· Cmembrane(17) where Celectricityis the electricity cost per kW·h and Cmembraneis membrane cost per square meter.

[0072] Figure 20 illustrates the performance of the crumpled GO membrane under elevated pressure conditions and its associated impact on decarbonization cost. In particular, the plot 2001 shows permeability and H2 / CO2selectivity of a mixed gas comprising approximately 0.9 H2molar fraction as a function of feed pressure. As shown, the permeability of H2increases with increasing pressure from approximately 1 bar to 5 bar, while the permeability of CO2remains significantly lower. The H2 / CO2selectivity remains at a relatively high level across the pressure range, indicating that the crumpled GO membrane maintains effective molecular sieving performance under increased pressure.

[0073] Plot 2002 illustrates the cost of decarbonization for flat GO membranes and crumpled GO membranes at different operating pressures. The results demonstrate that the crumpled GO membrane achieves significantly lower decarbonization costs compared to flatGO membranes across all tested pressures. In particular, the crumpled GO membrane achieves optimal performance at a pressure of approximately 3 bar, providing a balance between permeability, selectivity, and energy consumption. The reduced pressure requirement relative to conventional systems enables improved energy efficiency and cost reduction, thereby highlighting the suitability of the crumpled GO membrane for industrial applications, including hydrogen purification and carbon capture processes.

[0074] Additionally, the performance of flat GO membranes and crumpled GO membranes in achieving hydrogen (H2) purification under varying feed pressures is plotted in Figures 21a and 21b. In particular, Figure 21a shows the required membrane area for fGO and cGO membranes to achieve an H2permeate purity of at least 99.5 vol.% at a CO2removal rate of approximately 1 kg / h. The results demonstrate that the cGO membranes require substantially smaller membrane areas compared to fGO membranes across all tested pressures, indicating significantly enhanced permeability and process efficiency. Figure 21b illustrates the H2purity in the permeate stream for both fGO and cGO membranes under varying feed pressures. The data shows that the cGO membranes consistently achieve H2purity levels at or above approximately 99.5 vol.% across the tested pressure range, whereas the fGO membranes fail to reach the target purity under comparable conditions. The results demonstrate that the crumpled GO membranes enable efficient hydrogen purification with reduced membrane footprint and improved separation performance, thereby supporting their suitability for industrial applications, including hydrogen production and carbon capture processes.

[0075] The results in Figures 21 and 21 illustrate that fGO membranes require impractically large areas (730–6711 m2) and fail to achieve the target H2 purity of 99.5% under any pressure. In contrast, cGO membranes show significant improvements, with cGO4 delivering the best performance. To achieve 99.5% H2 purity a feed pressure of 3 bar is identified as the optimal operational point, balancing energy demands, membrane efficiency, and cost, with cGO4 achieving a decarbonization cost of 0.12 USD / kg CO2 for a 1-hour operation.

[0076] Furthermore, when a customized membrane module as that shown in Figure 22 is used, it was found that membrane cGO4 exhibits even higher permeability, exceeding 105Barrer for H2, while maintaining industry-relevant selectivity of ~50. These results are extremely promising, as the optimal performance is achieved at just 3 bar which is substantially lower than the typical industrial operating range of 15–30 bar which in turn offers significant energy savings and cost benefits. Given that industrial SMR processes typicallyoperate at 15–30 bar, the reduced pressure requirements of cGO membranes could present a highly efficient alternative for hydrogen purification.

[0077] It is useful to note that the customized membrane module shown in Figure 22 may comprise a plurality of stacked components configured to support and seal the crumpled graphene oxide (GO) membrane. In particular, a first component 2201 comprises an O-ring configured to provide sealing between adjacent layers of the assembly and to prevent leakage of gases during operation. A second component 2202 comprises a resin-sealed laminating pouch configured to encapsulate the membrane and provide an airtight enclosure. A third component 2203 comprises a conformal porous coating layer, for example formed from a porous polyethersulfone (PES) material, which is arranged to provide a compliant interface over the membrane surface. A fourth component 2204 comprises the crumpled GO membrane, which functions as the active separation layer. A fifth component 2205 comprises a porous support layer, for example a porous PES support, configured to provide mechanical support to the crumpled GO membrane while allowing gas transport therethrough. A sixth component 2206 comprises a laminating pouch layer arranged beneath the support layer, configured to assist in sealing and maintaining structural integrity of the membrane assembly.

[0078] In embodiments of the disclosure, it was ascertained that the gas separation performance of the crumpled GO membranes is primarily governed by the thickness of the laminated GO film in its flat state. To elucidate this relationship, the morphology of the membranes in the crumpled state was analysed. Surface roughness analysis, as illustrated in images 2301-2306 in Figure 23a, was performed using parameters including the arithmetical mean height (Sa) and the curvature–area index (CAI). Specifically, image 2301 represents an SEM image of cGO1 that was formed based on a fGO with thicknesses of 1±0.2 pm (fGO1), image 2302 represents an SEM image of cGO2 that was formed based on a fGO with thicknesses of 1.3±0.2 pm (fGO2), image 2303 represents an SEM image of cGO3 that was formed based on a fGO with thicknesses of 1.6±0.2 pm (fGO3), image 2304 represents an SEM image of cGO4 that was formed based on a fGO with thicknesses of 2.4±0.2 pm (fGO4), image 2305 represents an SEM image of cGO5 that was formed based on a fGO with thicknesses of 2.6±0.2 pm (fGO5), and image 2306 represents an SEM image of fGO1. The scale bar used in Figure 23a is 50pm for both SEM and confocal microscopy images. Based on the BET analysis, the surface area ranged from 171 m2 / g for cGO1 and 193 m2 / g for cGO5. The analysis demonstrates a correlation between membrane thickness, induced strain, and resulting surface curvature, thereby enabling estimation of the total strain within the crumpled membranes.

[0079] Figure 23b illustrates plots of the mechanical properties of flat GO membranes as a function of membrane thickness. In particular, the tensile strength and elongation at break are plotted against the thickness of the laminated GO films. The data shows that, as the thickness increases, the tensile strength decreases while the elongation increases. This indicates that thinner GO films exhibit higher strength but lower deformability, whereas thicker GO films exhibit increased flexibility but reduced mechanical strength.

[0080] Figure 23c illustrates the relationship between surface morphology parameters of the crumpled GO membranes and the thickness of the laminated GO films. In particular, the arithmetical mean height (Sa) of the surface and the curvature–area index (CAI) are plotted as a function of the flat film thickness. The CAI parameter represents a combined measure of the area of curved segments and the magnitude of curvature of the crumpled structure. The results demonstrate that both Sa and CAI increase with increasing thickness, indicating enhanced surface corrugation and curvature in thicker membranes.

[0081] The fitting behaviour of the CAI with respect to membrane thickness follows an approximate square-root dependence, i.e., proportional to √D, where D represents the thickness of the laminated GO film. This relationship indicates that the degree of curvature and surface deformation in the crumpled GO membranes is governed by thickness-dependent mechanical properties of the GO layers. Hence, these plots show that the membrane thickness serves as a controlling parameter for tuning the morphology of the crumpled structure and, consequently, the gas transport characteristics of the membrane.

[0082] Based on the images in Figure 23a, it was found that a laminated GO film having a thickness of approximately 1pm (fGO1) was shown to exhibit relatively high tensile strength, for example about 74.3 MPa, and low elongation, for example less than 1%, resulting in a substantially uniform crumpled structure with a relatively low surface roughness (Sa of approximately 6.3 pm) and a low curvature–area index. In contrast, a laminated GO film having a thickness of approximately 2.6 pm (fGO5) exhibits reduced tensile strength, for example less than 50 MPa, and increased elongation, for example greater than 1.4%, resulting in a crumpled GO membrane (cGO5) having increased surface roughness (Sa of approximately 26.9 pm) and a higher curvature–area index. However, excessive thickness leads to the formation of macroscopic defects, thereby reducing structural integrity. The formation of such defects are illustrated in Figure 24 which clearly shows the formation of macroscopic defects on the cGO5 membranes due to the overstretch of the GO flakes.

[0083] This morphology aligns with elasticity theory for bi- and multilayer membranes. In the classical elasticity problems, bending rigidity is often ignored and elastic energy originates from in-plain deformations. However, this assumption holds only in specific cases, such as clamped edges. In general, wrinkles formation minimizes in-plain deformation energy, making bending energy the dominant factor. The competition between bending rigidity and external stress T governs the wrinkling and determines the corrugation properties. The corresponding energy terms are:where K is the bending rigidity, and h is variation in the interlayer distance. Under the uniform compression condition, T < 0, then the minimum of the elastic energy (equation 18) corresponds to the self-modulation of the membrane with the wave vector q* and thusto wrinkling with a characteristic wavelength =

[0084] To compare the results obtained thus far on the dependence of Sa(which is proportional to the characteristic wrinkle wavelength) on membrane thickness, the relationship between the bending rigidity (K) and membrane thickness (D) must first be established.

[0085] For classical solid membranes with D much larger than interatomic distance, the bending rigidity follows the well-known scaling relation: K~D3. This leads to a wrinkle wavelength scaling as ~D3 / 2(note, that for wrinkles induced by elongation in perpendicular direction, ~κ1 / 4~D3 / 4 52). In both cases, the wrinkling period increases super-linearly with thickness.

[0086] However, the κ~D3dependence applies only to solid membranes with strong interlayer coupling. A more detailed analysis shows that this regime is reached only at sufficiently large spatial scales Lc~(C44)−1 / 2, where C44is the shear moduli.

[0087] At smaller length scale (L < Lc), the bending rigidity is proportional to the number of layers, following the linear relation κ~D. In laminated membranes, where all shear moduli vanish due to weak interlayer coupling (allowing individual flakes to slide past each other), this linear dependence (κ~D) holds at all spatial scales. Consequently, the wavelength scale as ~κ1 / 2~D1 / 2, which appears to be consistent with experimental observations (see Figure 23c) and is further supported by previous analysis on the role of interlayer sliding.

[0088] The formation of long-lived metastable states can be naturally linked to glass-like behaviour in systems with complex energy landscape. A similar phenomenon may be observed in magnetic systems, where a spin glass state emerges due to high degeneracy along lines or surface of energy minima. This analogy suggests that the persistent crumpling of membrane may arise from an intrinsic energy frustration, where multiple, nearly equivalent, low-energy states prevent relaxation to a fully flattened configuration.

[0089] Recently, a mapping of the state of a compressed membrane onto an associated Ising model was proposed by those skilled in the art. In this model, an Ising “spin” up or down corresponds to the displacement of a given membrane point upward or downward under compression. This analogy suggests that a glassy state could emerge in such membranes, particularly given the high degree of disorder in GO membranes. The presence of multiple metastable configurations may lead to frustration in relaxation, similar to spin glass behaviour in disordered magnetic field.

[0090] Figure 25a illustrates X-ray diffraction (XRD) spectra 2501 for a crumpled GO membrane (cGO4) and spectra 2503 for a flat GO membrane (fGO1). Spectrum 2501 corresponding to the crumpled GO membrane exhibits multiple peaks, which are fitted using a plurality of components, including a first fitting curve representing a first interlayer spacing (di) and a second fitting curve representing a second interlayer spacing (d2). To recap, the first interlayer spacing (di) corresponds to a relatively smaller interlayer distance, while the second interlayer spacing (d2) corresponds to a relatively larger interlayer distance. The fitting further includes a component associated with interlayer spacing contributions arising from crumple-induced structural deformation, as well as a component associated with interlayer spacing of substantially pristine graphene oxide layers.

[0091] Figure 25b illustrates plots showing the interlayer spacing of the cGO and fGO1 membranes calculated from the XRD fitting data illustrated in Figure 25a. The XRD analysis demonstrates that the crumpled GO membrane comprises a structure having at least two distinct interlayer spacings, wherein the first interlayer spacing (di) is in a range of approximately 8.16 A to 8.61 A, and the second interlayer spacing (d2) is in a range of approximately 9.22 A to 9.89 A. Among the analysed samples, the cGO4 membrane exhibits relatively narrow values of the first interlayer spacing (di) and the second interlayer spacing (d2), for example approximately 8.16 A and 9.22 A, respectively, corresponding to a relatively compact configuration.

[0092] In contrast, the flat GO membrane (fGO1) exhibits a single interlayer spacing peak of approximately 8.7 A, indicative of a substantially uniform lamellar structure. The comparison demonstrates that deformation of the flat GO membrane into the crumpled structure results in formation of a multi-region structure comprising at least two distinct interlayer spacings, thereby providing a structural basis for enhanced gas transport and molecular sieving behaviour.

[0093] Figure 25c illustrates X-ray diffraction (XRD) spectra for crumpled GO membranes cGO1 to cGO5. Each spectrum is fitted using multiple components corresponding to distinct interlayer spacings. The d2fitting curve indicates the interlayer spacing contributed by crumpled formation, while the d1blue fitting curve represents the interlayer spacing from pristine graphene layers.

[0094] In the following section, it is proposed that corrugated membranes with spatially varying stress tend to naturally exhibit variations in the interlayer distance. To describe this behaviour, the total interlayer distance (d) is expressed as the sum of the equilibrium interlayer distance (Z) and the variation term h d = l + h.

[0095] For a bilayer system with two identical layers, the corresponding contribution to the full elasticity Hamiltonian is given as:where κ is the bending rigidity of each layer, u is the average in-plane strain (dilatation) of the two layers, and giare corresponding elastic moduli responsible for the out-of-plane elastic energy (g1) and for the coupling of in-plane and out-of-plane deformations (g3): g3 / g1=2C13 / C33

[0096] For bilayer graphene, g1= 0.8 eV / Å2, it is estimated that the ratio g3 / g1for different graphite type is to be 0.1-0.15. Neglecting the bending energy term (κ) in equation (19), the relation: h̄ = −(g3 / 2g1)ul is obtained. This result implies that a non-uniform strain distribution in 231the layers inevitably leads to a non-uniform interlayer distance. When the bending energy is considered, the relationship becomes nonlocal but, in both cases, inhomogeneous in-plane deformation results in inhomogeneity of interlayer distances. Hence, it can be said that simple elasticity theory alone cannot explain the formation of two distinct peaks in the interlayer distance distribution observed in our XRD analysis. While elasticity theory ensures variationsin stress perpendicular to the layers, it does not account for the quantization of interlayer spacing.

[0097] To understand the origin of these two peaks, the chemical structure of the membrane and the role of intercalated water layers between GO flakes is considered. The presence of confined water layers can lead to discrete interlayer distances, where water molecules stabilize specific configurations rather than forming a continuous distribution. In the experiments, the interlayer spacing varies between 8.16 A to 9.22 A, a range previously attributed to the presence of a monolayer of water. fGO membranes exhibit a uniform interlayer distance of d = 8.7 A, whereas in cGO membranes, two distinct interlayer spacings were observed: one with a reduced spacing (di = 8.16 A) and another with an expanded spacing (d2= 9.22 A) (see Figure 25b).

[0098] In both cases, the observed shift in interlayer spacing follows a consistent trend: (d ± di,2) = ±0.5 Å. It was previously demonstrated by those skilled in the art that the transition from an ordered to disordered configuration results in an interlayer spacing change of ~0.5 A, which aligns well with the observed shifts in interlayer spacing. This suggests that strain-induced inhomogeneities in cGO contribute to this structural variation. Interestingly, an increase in interlayer spacing has been predicted as a signature of ordering, where a fully oxidized domain is surrounded by a graphene-like region. Conversely, a decrease in interlayer spacing suggests randomization, where oxidized and graphene-like regions intersect. The key question to be addressed is how H2 / CO2 selectivity is achieved in the cGO membranes described herein whereby the crumpled structure facilities intensive gas uptake at expanded sites, transport along the channels through selective filter, and release.

[0099] In the experiments, the formation of two domains with the interlayer spacings of 8.16 A and 9.22 A were observed, a range previously attributed to the presence of a monolayer of water. fGO membranes tend to exhibit a uniform interlayer distance of d = 8.7 A, whereas in cGO membranes, two distinct interlayer spacings were observed: one with a reduced spacing (d1= 8.16 Å) and another with an expanded spacing (d2= 9.22 A). In both cases, the observed shift in interlayer spacing follows a consistent trend: (d ± di,2) = ±0.5 Å. A previous study demonstrated that the transition from an ordered to disordered (random) configuration results in an interlayer spacing change of ~0.5 A. The random configuration with decreased interlayer distance where the oxidized and graphene-like regions intersect each other is shown in Figure 26a. The ordered configuration with increased interlayer spacing where there is a fully oxidized domain surrounded by a graphene-like region is shown in Figure 26b. With di =8.16 A, the effective channel size in the selective pore filter is determined by GO sheet thickness (~3.4 A) and a water monolayer (~2.5 A). This configuration leaves ‘empty’ space of ~3 A available for molecular transport. The narrow channel dimensions impose steric constraints, allowing effective H2 / CO2 separation while maintaining molecular transport efficiency. The domains with dz = 9.22 A provide an “empty” space size >3 A, forming pore openings with fast gas uptake.

[0100] It is understood that anisotropic nanochannels serve as the primary pathways for gases in membranes. However, it is essential to consider that strain-induced sliding initiates complex mechanochemical processes along the flake surfaces. A combination of X-ray photoelectron spectroscopy (XPS) was performed wherein the chemical composition of cGO was analysed using XPS data from O1s and C1s spectra and the results are plotted in Figure 27. The 0=C-0 band observed at 534 eV in O1s corresponds to carboxylic groups and C-0 detected at 533 eV in 01s, corresponds to oxygen atoms in hydroxyl (-OH) or epoxy (-C-O-C-) groups. A higher 0=C-0 / C-0 ratio indicates a greater proportion of carboxylic groups relative to hydroxyl or epoxy groups while carbon to oxygen (C / 0) mass ratio of cGO membranes calculated from XPS measurements.

[0101] The plot of the Raman spectroscopy ID1 / IG ratio for various cGO membranes is plotted in Figure 28. The D1 band in the range of 1330-1350 cm-1originates from defects and disorders in the graphitic lattice, while the G band around 1580 cm-1is associated with the inplane vibrations of sp2-hybridized carbon atoms. cGO4 was shown to exhibit the highest ID1 / IG ratio (1.45). Hence, it is shown from aberration-corrected transmission electron microscopy (TEM) and the pore size distribution (PSD) analysis that the formation of this multi-domain structure is governed by competing mechanochemical processes.

[0102] Firstly, the formation of in-plane holes: the TEM / PSD analysis reveals a dominant peak at 3 A, indicating the primary holes size where the material has the highest defect volume. Secondly, the redox-driven restructuring of the basal plane and edges: the increase in carboxylic groups, as indicated by elevated 0=C-0 (534 eV) / C-0 (533 eV) peaks for cGO4, suggests a greater presence of edges, likely due to the formation of holes. Overall, approximately 34% of the oxygen-containing functional groups (epoxy and hydroxyl) are removed from basal plane when the C / O ratio increases from 2.5 (cGO1 ) to 4.3 (cGO4) (Figure 27).

[0103] Finaly, TEM / PSD analysis, cross-sectional TEM / STEM and XPS depth profiling enable both the in-plane and out-of-plane structure and chemistry of the crumpled membraneto be investigated. First, the in-plane scanning proves that spatial resolved regions with holes measuring 3 A width and 5-8 A width may be distinguished as shown in plots 2901-2905 of Figure 29 (which shows the pore size distribution of the various membranes) and images 3001-3003 of Figure 30 (where image 3001 shows a cross-sectional TEM image of cGO membranes on silicon wafer prepared by focused ion beam (FIB), and images 3002, 3003 show a cross-sectional STEM image of cGO membranes on silicon wafer prepared by FIB). The images in Figure 30 indicate the distance profiles between each layer whereby the cross-sectional samples were prepared by sonicating an aqueous suspension of free-standing cGO membranes for 2 hours, drop-casting the dispersion onto a silicon wafer, and subsequently processing it with FIB. Due to the inherently low crystallinity of graphene oxide and the presence of significant wrinkling and turbostatic stacking in the crumpled membrane structure, the resulting images had limited contrast and resolution. These structural features complicated direct measurement of interlayer distances at atomic resolution or detailed lattice fringe analysis. Nevertheless, qualitative differences were consistently observed in interlayer spacing across regions with varying membrane thicknesses, suggesting that the crumpling process induces spatial heterogeneity in the interlayer structure.

[0104] The 3 A defects are comparable to the kinetic diameter of H2 (2.89 A) while blocking the CO2 (3.30 A), allowing selective molecular sieving. Meanwhile, 5–8 Å defects introduce more effective diffusion pathways, thereby enhancing permeability. Second, XPS depth profiling, performed with 100 nm etching steps, was used to analyse the out-of-plane chemical composition of the membranes. This technique combines XPS surface analysis with ion sputtering (etching) to progressively remove layers, allowing for composition analysis at different depths. Notably, only in the cGO4, the 0=C-0 (534 eV) / C-0 (533 eV) ratio exhibits multiple peaks with a maximum amplitude of difference throughout the depth profile, indicating out-of-plane chemical anisotropy.

[0105] Due to this multi-domain anisotropic configuration, the diffusivity of H2 and CO2 through cGO membranes as shown in Figure 31 remains largely unaffected by membrane thickness, as intrinsic diffusion pathways are preserved within the few nanometer-thick compact domains. In particular, Figure 31 illustrates plots of diffusivity of H2 and CO2 in cGO membranes across different thickness at 21 °C and 1 bar pressure.

[0106] Furthermore, the solubility of both H2 and CO2 as shown in Figure 32 peaks for cGO4, driven by the larger gas volume available in expanded domains. Specifically, Figure 32 illustrates plots of solubility of H2 and CO2 in cGO membranes across different thickness at 21 °C and 1 bar pressure.

[0107] In conclusion, it can be said that the transition from flat to crumpled topology in 2D lamellar membranes significantly enhances gas permeability keeping high selectivity. cGO membranes overcome the gas barrier limitations of fGO membranes, achieving an impressive H2 permeability of 2.1 *104Barrer with H2 / CO2 selectivity of 91. Crumpling creates anisotropic gas permeation paths, improving both gas flow and molecular sieving. The tightened GO layers form highly selective nano-filters, while expanded domains provide fast access to nanofilters for gas molecules, boosting permeability. The multidomain 2D structure demonstrates strong thermal and humidity stability when tested in harsh conditions (-20°C, 96% relative humidity) and enables the cGO membranes to outperform both 2D material-based and polymer-based membranes. The scalability of the membranes ensures that the thickness of the fGO film plays a key role in controlling crumpling. Additionally, the efficient and scalable fabrication process, coupled with the use of sustainable materials, aligns with the principles of the circular economy, promoting resource efficiency and minimizing environmental impact.

[0108] Exemplary Embodiment

[0109] The following section describes a non-limiting exemplary embodiment for forming the cGO membranes described in this disclosure.

[0110] Materials

[0111] Graphene oxide (GO) dispersion (30 mg / ml in water, lateral dimensions of flakes are approx. 1 pm, Guangzhou Coral Industrial Co. Ltd.) was used to fabricate the membranes. Polystyrene (PS) sheets (Grafix KSF50-C) were utilized to serve as the template for casting. Poly (vinyl alcohol) (PVA) powder (Mw 89,000-98,000, 99% hydrolyzed, Sigma-Aldrich) was dissolved with deionized (DI) water to produce 20 mg / ml solution. Ethyl acetate (>99.5% ACS, VWR Chemicals BDH®) was employed to dissolve the PS substrate.

[0112] Assembly of fGO and cGO membranes

[0113] First, the fGO films were made by casting the GO dispersion (30 mg / ml in DI water) on the plasma-treated PS template with the automatic film applicator. The thickness of the membrane is controlled by adjusting the gap between the casting blade and the substrate. The thickness of fGO varied between 1 pm to 2.6pm. After the membrane was dried in a fume hood for 24 h, it was heated to 140°C for 10 min in convection oven to induce the thermal shrinkage of PS template, which crumpled the GO film due to the mismatch of the thermal coefficients. The membrane retracted to one-third of its original size after the thermal treatment.Subsequently, a layer of PVA was spin coated on the membrane to protect the crumpled structure. Finaly, ethyl acetate was utilized to dissolve the template to produce free-standing membranes. To assemble a control fGO membrane, we used the same procedures except the thermal treatment.

[0114] Instrumental characterizations

[0115] Basic plasma cleaner (PDC-32G-2) was used to treat the PS template to enhance the adhesion between GO and PS surface. Scanning electron microscope (SEM, Zeiss Supra 40) was used to depict the crumpled structure of GO membrane and image the cross-section of the membrane for thickness measurements. Confocal microscope (Zeiss LSM710) was utilized to map the topography of the membrane surface. The interlayer spacing of the membrane was measured with the X-ray diffractometer (Bruker D8 Advance) with Cu Ka radiation wavelength (A) of 1.5418 A and quantified with Bragg’s law (nA = 2dsin(9)). Attenuated total reflection - Fourier transform infrared spectroscopy (ATR-FTIR, IRTracer-100, Shimadzu; 4 cm-1resolution) and Raman spectroscopy (WITEC ALPHA300R, 532 nm laser) were operated to investigate the functional groups of the membrane. Brunauer-Emmett-Teller (BET) analysis was conducted using Micromeritics ASAP 2460 surface area and porosity analyser with CO2 serving as the sorption gas at 273K. The pore size distribution (PSD) was acquired by analysing the gas sorption with density function theory (DFT). Before each measurement, the samples (around 150 mg) were degassed at 110°C for 68 h. The mechanical tensile test was conducted by using dynamic mechanical analyser (DMA 850, TA Instruments) with rate control strain ramping of 0.25 % / min. The water contact angle was measured by Drop Shape Analyzer (DSA25E, KRUSS GmbH) with a water drop of 2 pL and analysed by Ellipse (TangentOI) fitting method. XPS analysis (Kratos AXIS Supra) was undertaken under pressure below 5*1 O'9Torr using Al-Ka ( / ?u=1486.6 eV) radiation at 75W (15kV). High resolution scans were implemented with a 40-eV pass energy and 0.1 eV step size. Wide-scan survey was acquired with 160 eV pass energy, 100ms dwell time and 1 eV step size. The angle between the sample surface and detector was 90°. For the XPS depth profiling, the etching conditions were: 10 keV Ar1000+ (cluster size of 1000), raster size of 1.5 mm, etching time of 12s for cGO1, 24s for the rest. Aberration-corrected TEM images were acquired by using JEOL JEM-ARM200F at 60 kV.

[0116] Curvature analysis

[0117] For the curvature analysis of the cGO membranes, 3D surface profiles were collected using a digital microscope VHX-7000 (Keyence, Japan). The samples exhibitedvarying degrees of crumpling and roughness. The microscope provided high-resolution images at a magnification of 150x with a pixel size of 0.69 pm and Z resolution of 1 pm. The scanned area for each sample was 1500 pm x 2000 pm. The collected 3D surface maps were imported into MeshLab as point clouds. Normals of the vertices were computed to create the surface mesh using the screened Poisson surface reconstruction method with a reconstruction depth of 8, interpolation weight of 6, and number threads set to 20. Scale dependent quadric fitting with a curvature scale of 1% was used to compute principal directions of curvature and the mean curvature values for each surface. A mean curvature map provided the spatial distribution of curved and flat segments. Areas with curvature values greater than one standard deviation from the mean were classified as curved, while those within one standard deviation were considered flat. The parameters such as the area of curved segments in pm2, along with curvature magnitude in pm-1for each segment were derived to compare the intensity of curved features across samples.

[0118] Gas permeation set-up and operating conditions

[0119] In brief, the membranes were put in the membrane modules and the permeated gas was carried to the bench-top mass spectrometric residual gas analyzer (Hidden Analytical, HPR-20 TMS) with argon as the sweeping gas for analyzing the composition. The total feed flow rate of all the measurements in this work was set at 150 mLnmin-1and the argon sweeping flow rate was 50 mLnmin-1under the controlled feed pressure of 1 bar with the help of the mass flow controller (MFC). The temperature-controlled chamber enabled the membrane to be tested in both high and low temperatures from -20 to 80°C. For hydrated gas permeation test, the feed gas was mixed with water vapor with the assistance of the controlled evaporator mixer (CEM), and the relative humidity (RH) of the feed gas was further confirmed with the hydrometer.

[0120] Determination of permeability and selectivity

[0121] The calculation of the permeability is set out below. The permeate flow rate, QPii, was calculated from the sweep flow rate, QAr, according to the relationship of:QPH,i = QAr -^ (20)PArwhere pp iand pArwere discretely the normalized partial pressure of permeate and argon in the sweep / permeate mixed stream and were obtained by processing the signal in the mass spectrometer. The permeance, Fhand the permeability, P of the tested gas were yielded by:A and I were respectively membrane active area and thickness, while pptwas the partial pressure of tested gas in the feed stream:Pf,i = Xf Pf(23)where xtsymbolized the volume fraction of the tested gas in the feed stream, and p was the total feed pressure which was 1.01 bar in all tests. The ideal gas selectivity was the ratio of the permeance or the permeability:

[0122] Determination of diffusivity and solubility

[0123] The time-lag method was exploited to calculate the diffusivity and solubility of the gases in cGO membranes. The total permeate volume, Vp i, was obtained by integrating the discrete flow rate data:

[0124] The permeate flow rate was converted from MS standard signal (partial pressure of the permeate gases) by equation (20). Fitting of the cumulative volume data in the time period of steady permeation state yielded the total time lag (ttotc ii) as the intercept with the horizontal axis. The total time lag was contributed by the residence time of the permeating gas / in the system (ts i, instrumental time lag) and in the membrane (tmemi, membrane time lag):total.i ~ s,i T ^mem,i (26)

[0125] The instrumental time lag can be considered as constant once operating in the same testing conditions such as flow rate, temperature and pressure. Normally, the membrane time lag is proportional and inversely proportional to square of the membrane thickness, I2, and diffusivity, D, respectively:

[0126] Considering that the slope of the line with as the y-axis and I2as the x-axis is k, the diffusivity can be obtained by:D = — (28)

[0127] The solubility, S, can then be calculated by:where P is the permeability.

[0128] A process for forming a crumpled graphene oxide (GO) membrane in accordance with embodiments of the disclosure is illustrated in Figure 33. Process 3300 begins at step 3301 with process 300 depositing a GO dispersion onto a shrinkable substrate to form a laminated GO film comprising a plurality of stacked GO layers. At step 3302, process 3300 then drys the laminated GO film to form a planar GO membrane. Process 3300 then heats the shrinkable substrate to induce lateral shrinkage of the substrate, such that a compressive strain is applied to the GO membrane during shrinkage of the substrate. This takes place at step 3303. The GO membrane is then deformed at step 3304, in response to the application of the compressive strain, to form a crumpled structure comprising a plurality of strain-induced out-of-plane deformations, wherein the crumpled structure exhibits at least two distinct interlayer spacings between adjacent GO layers.

[0129] In embodiments of the disclosure, process 3300 may deposit a protective polymer layer onto the crumpled structure of the deformed GO membrane. Process 3300 may then subsequently remove the shrinkable substrate to obtain a free-standing crumpled GO membrane.

[0130] In embodiments of the disclosure, the protective polymer layer may comprise poly(vinyl alcohol) (PVA), and the step of removing the shrinkable substrate comprises process 3300 dissolving the substrate using an organic solvent comprising ethyl acetate.

[0131] In embodiments, the laminated GO film has a thickness in a range between 1pm and 3pm. In further embodiments, when process 3300 deposits the GO dispersion onto the shrinkable substrate, process 3300 may utilize a blade casting technique to disperse the GO dispersion onto the shrinkable substrate.

[0132] In embodiments, process 3300 may heat the shrinkable substrate at a temperature range between 120°C and 160°C. In embodiments, the at least two distinct interlayer spacings comprise a first interlayer spacing having a spacing between 8.0 A and 8.7 A, and a second interlayer spacing having a spacing between 9.0 A and 10.0 A.

[0133] In embodiments, the crumpled structure comprises in-plane pores having a size in a range between 2.5 A and 3.5 A, and the shrinkable substrates is shrunk by at least 50%.

[0134] In embodiments, the first interlayer spacing defines a selective region configured for molecular sieving; and the second interlayer spacing defines a transport region configured to facilitate gas diffusion. Further, the GO dispersion comprises GO flakes dispersed in an aqueous solution, the flakes having a lateral size between 0.5 pm and 2.0 pm, and the dispersion having a concentration between 10 mg / ml and 50 mg / ml.

[0135] Numerous other changes, substitutions, variations, and modifications may be ascertained by the skilled in the art and it is intended that the present application encompass all such changes, substitutions, variations, and modifications as falling within the scope of the appended claims.

Claims

CLAIMS:

1. A method for forming a crumpled graphene oxide (GO) membrane, the method comprising:depositing a GO dispersion onto a shrinkable substrate to form a laminated GO film comprising a plurality of stacked GO layers;drying the laminated GO film to form a planar GO membrane;heating the shrinkable substrate to induce lateral shrinkage of the substrate, such that a compressive strain is applied to the GO membrane during shrinkage of the substrate; anddeforming the GO membrane, in response to the application of the compressive strain, to form a crumpled structure comprising a plurality of strain-induced out-of-plane deformations, wherein the crumpled structure exhibits at least two distinct interlayer spacings between adjacent GO layers.

2. The method according to claim 1 further comprising the step of:depositing a protective polymer layer onto the crumpled structure of the deformed GO membrane.

3. The method according to claim 2, further comprising the step of:removing the shrinkable substrate to obtain a free-standing crumpled GO membrane.

4. The method according to claim 3, wherein the protective polymer layer comprises poly(vinyl alcohol) (PVA), and whereby the step of removing the shrinkable substrate comprises the step of:dissolving the substrate using an organic solvent comprising ethyl acetate.

5. The method according to any one of claims 1 to 4, wherein the laminated GO film has a thickness in a range between 1pm and 3pm.

6. The method according to any one of claims 1 to 5, wherein the step of depositing the GO dispersion onto the shrinkable substrate comprises blade casting the GO dispersion onto the shrinkable substrate.

7. The method according to any one of claims 1 to 6, wherein the shrinkable substrate comprises polystyrene.

8. The method according to any one of claims 1 to 7, wherein the shrinkable substrate is plasma-treated prior to the step of depositing the GO dispersion onto the shrinkable substrate.

9. The method according to any one of claims 1 to 8, wherein the step of heating the shrinkable substrate is performed at a temperature range between 120°C and 160°C.

10. The method according to any one of claims 1 to 9, wherein the at least two distinct interlayer spacings comprise a first interlayer spacing having a spacing between 8.0 A and 8.7 A, and a second interlayer spacing having a spacing between 9.0 A and 10.0 A.

11. The method according to any one of claims 1 to 10, wherein the crumpled structure comprises in-plane pores having a size in a range between 2.5 A and 3.5 A.

12. The method according to any one of claims 1 to 11, wherein the shrinkable substrate is shrunk by at least 50%.

13. The method according to claim 10, wherein the first interlayer spacing defines a selective region configured for molecular sieving; and the second interlayer spacing defines a transport region configured to facilitate gas diffusion.

14. The method according to any one of claims 1 to 13, wherein the GO dispersion comprises GO flakes dispersed in an aqueous solution, the flakes having a lateral size between 0.5 pm and 2.0 pm, and the dispersion having a concentration between 10 mg / ml and 50 mg / ml.

15. A crumpled graphene oxide (GO) membrane comprising:a plurality of stacked GO layers arranged in a laminated structure that is deformed into a crumpled structure having a plurality of strain-induced out-of-plane deformations, wherein the crumpled structure exhibits at least two distinct interlayer spacings between adjacent GO layers.

16. The crumpled GO membrane according to claim 15 further comprising:a protective polymer layer disposed onto the crumpled structure.

17. The crumpled GO membrane according to claim 16, wherein the protective polymer layer comprises poly(vinyl alcohol) (PVA).

18. The crumpled GO membrane according to any one of claims 15 to 17, wherein the laminated structure has a thickness in a range between 1 pm and 3pm.

19. The crumpled GO membrane according to any one of claims 15 to 18, wherein the plurality of stacked GO layers arranged in the laminated structure is obtainable by depositing a GO dispersion onto a shrinkable substrate using blade casting.

20. The crumpled GO membrane according to any one of claims 15 to 19, wherein the at least two distinct interlayer spacings comprise a first interlayer spacing having a spacing between 8.0 A and 8.7 A, and a second interlayer spacing having a spacing between 9.0 A and 10.0 A.

21. The crumpled GO membrane to any one of claims 15 to 20, wherein the crumpled structure comprises in-plane pores having a size in a range between 2.5 A and 3.5 A.

22. The crumpled GO membrane to claim 20, wherein the first interlayer spacing defines a selective region configured for molecular sieving, and the second interlayer spacing defines a transport region configured to facilitate gas diffusion.