THEREOF

FIELD

The present disclosure relates to composite membranes, their method of fabrication and uses thereof. In particular, the present disclosure relates to composite membrane comprising a bulk phase and a binding phase for use in nanofiltration applications.

BACKGROUND

Large quantities of industrial wastewater are produced every year and discharged into the water system. Industrial wastewater contains various soluble organics that induce serious implications on the ecosystem, and removal of these from industrial effluents has attracted major interest in the past years. Accordingly, managing industrial wastewater is important for maintaining a clean and green environment, and when implemented in a corporation's social responsibility plan can further increase the goodwill to the corporation. Given the increasing complexity and need for manufactured goods, removing soluble organic compounds from industrial effluent is a fast-growing technical area with an intensive R&D focus. Many industrial effluents contain large amounts of organic solvents, which are toxic to marine animals and harmful to the environment. This situation requires the nanofiltration (NF) membrane to be also capable to reject organic solutes since the pore surface chemistry and fine structures of membrane inevitably undergo changes to certain extents with the increase in the solvent concentrations in water. Similarly, the organic solutes have different physicochemical behaviours in water-organic solvent mixture than in water. Furthermore, the removal of solutes from water-solvent mixture is a prerequisite to downstream recovery of organic solvent from water.

One of the challenges is to remove contaminants present in trace concentrations, which is usually the most energy intensive step in effluent treatment processes. The use of meso and micro porous adsorbents, such as activated carbon, clay, natural zeolite and biomass, to entrap dilute organics and the use of oxidation catalysts, typically Ti0 ₂ and supported metal nanoparticles with aid of UV-light or bubbling of ozone, to degrade the organics, represents the two-major contemporary technical approaches to tackle water-soluble organic solutes. However, such methods and devices may be technically complicated, require many additional steps and still may not provide good results.

The separation of active pharmaceutical ingredients (API) from a reaction solution after synthesis is a very important operation process. Currently, distillation is the major technique to recover organic solvents and collect API. However, it is highly energy intensive and may change the structure of product. In this regard, there is a need for another separation technique and/or separation device.

Nanofiltration (NF) shows enormous potential for the separation of small solute molecules from solvents and has particularly gained attention for applications such as wastewater treatment, desalination and water purification, and API separation. It offers higher fluxes at lower transmembrane pressures (normally 5-15 bar) as compared to reverse osmosis (RO) and higher rejections than ultrafiltration for small molecules. NF membranes can exhibit rejection to ions much smaller than its pore size owing to a combination of steric hindrance and Donnan/dielectric exclusions. Polymeric membranes, especially crosslinked polyimide (PI) matrices, have gained wide commercial acceptance for NF. However, the polymer- based NF has the limitation of advancing permeance without conceding selectivity (permeability and selectivity tradeoff). For organic solvent NF, the polymeric membranes would be intrinsically susceptible to organic solvents, particularly at elevated temperatures, which further compromise their selectivity. Alternatively, inorganic membranes such as micro and mesoporous zeolite membranes, carbon membranes and silica membranes could offer significantly improved structural stability against solvents at elevated temperatures. However, the application of these inorganic membranes in nanofiltration is rare because of its inherent problem; i.e. substantially low flux. For example, RO of water using zeolite-A inorganic membrane can only attain a permeance of about 10 ² L/m ² dvbar. One of the major challenges is the difficulty in the fabrication of a large skinned inorganic membrane that is defect free. For instance, as far as the fabrication of pristine zeolite membranes is concerned, it is generally attained through the in-situ growth of zeolite crystal grains, i.e. hydrothermal synthesis, on a porous base in an autoclave. The preparation is complex and costly, primarily due to poor reproducibility and requirement for porous ceramic support, and the resulting membranes are often too thick and dense to suit NF. Scaling-up is also quite challenging for pure zeolite membranes owing to high brittleness. Therefore, there is still a need for a selective and robust zeolite membranes and its fabrication by cost- effective protocols.

Mixed matrix membranes (MMMs), also referred to as polymer-nanocomposite membranes, where the inorganic nano-particles are embedded and dispersed in a polymer matrix, is another solution to mitigate the permeability and selectivity tradeoff. Such membranes utilize the strong aspects of both inorganic and polymeric materials, such as the ease of fabrication of polymeric membranes as well as the well-defined pore structures and thermo-chemical stability of inorganic fillers. A major class of inorganic particles that have been extensively researched in this field is zeolites, owing to their porous structure and properties such as cation exchange and molecular sieving. This is mostly attained through polymeric nanocomposite membranes in which the zeolite particles are embedded or dispersed in a polymer matrix. However, the problem with such membranes is that the filtrate can bypass the embedded zeolite particles in the membrane, discounting the role of zeolite for separation.

Accordingly, it is generally desirable to have a composite membrane that can overcome or alleviate at least one or more of the above mentioned problems.

SUMMARY

The present invention relates to composite membranes, methods of fabrication and uses. In particular, the invention relates to composite membranes comprising a bulk phase and a binding phase. In contrast to prior MMM membranes, in which the zeolite form a non- continuous solidphase inside a continuous polymer matrix and therefore cannot sufficiently execute the role of separation, the inventors have found that a high content of, for example, zeolite forming a continuous bulk phase interpenetrating with the binding carbonaceous phase is advantageous. This biphasic design realizes well-defined pores surrounded by the frame of zeolite-carbonaceous amalgam (a highly dispersed and cohesive state of the two phases) throughout the membrane. When used to separate organic molecules, probed by methylene blue (with molecular weight of about 320 g/mol), from both water and isopropyl alcohol (IPA), the composite membrane manifests high rejection of particular molecules with reasonable permeation flux. Additionally, as the liquid streams go through the composite membrane consisting of plenty micron-cells (~l pm), a low-pressure drop (0.5-2 bar) across the entire membrane is imposed.

In a first aspect, the present invention relates to a composite membrane, comprising:

a) a bulk phase comprising porous particles, the porous particles having micropores of less than about 2 nm and/or mesopores in a range of about 2 nm to about 10 nm; and

b) a binding phase comprising an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, the binding phase substantially dispersed within the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In a second aspect, the present invention relates to a composite membrane, comprising: a) a bulk phase comprising zeolite particles; and

b) a binding phase comprising an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, the binding phase substantially dispersed within the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In a third aspect, the present invention relates to a composite membrane, comprising:

a) a bulk phase comprising porous particles, the porous particles having micropores of less than about 2 nm and/or mesopores in a range of about 2 nm to about 10 nm; and

b) a binding phase comprising an at least partially carbonised polymer and an at least partially carbonised ionic liquid, the binding phase substantially dispersed within the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In a fourth aspect, the present invention relates to a composite membrane, comprising: a) a bulk phase comprising zeolite particles; and

b) a binding phase comprising an at least partially carbonised polymer and an at least partially carbonised ionic liquid, the binding phase substantially dispersed within the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In some embodiments, the binding phase further comprises glass fibers.

In some embodiments, the at least partially carbonised polymer is an at least partially carbonised acrylate polymer.

In a fifth aspect, the present invention relates to a method of fabricating a composite membrane, comprising the steps of:

a) providing a bulk phase comprising porous particles, the porous particles having micropores of less than about 2 nm and/or mesopores in a range of about 2 nm to about 10 nm; and

b) mixing the bulk phase with a binding phase comprising a polymer to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In a sixth aspect, the present invention relates to a method of fabricating a composite membrane, comprising the steps of:

a) providing a bulk phase comprising zeolite particles; and

b) mixing the bulk phase with a binding phase comprising a polymer to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In a seventh aspect, the present invention relates to a method of fabricating a composite membrane, comprising the steps of:

a) providing a bulk phase comprising porous particles, the porous particles having micropores of less than about 2 nm and/or mesopores in a range of about 2 nm to about 10 nm; and

b) mixing the bulk phase with a binding phase comprising a polymer and an ionic liquid to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer and the ionic liquid is converted into an at least partially carbonised ionic liquid;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In an eighth aspect, the present invention relates a method of fabricating a composite membrane, comprising the steps of:

a) providing a bulk phase comprising zeolite particles; and

b) mixing the bulk phase with a binding phase comprising a polymer and an ionic liquid to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer and the ionic liquid is converted into an at least partially carbonised ionic liquid;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In some embodiments, the binding phase further comprises glass fibers.

In some embodiments, prior to step (c), the paste is subjected to a pressure of about 60 MPa to about 140 MPa.

In some embodiments, the partial pyrolysis step (step C) is performed at about 300 °C under low oxygen conditions.

BRIEF DESCRIPTION OF DRAWINGS

Embodiments of the present invention will now be described, by way of non-limiting example, with reference to the drawings in which:

Figure 1 is a schematic illustration of the general protocol for the preparation of the composite membrane of the present invention;

Figure 2 illustrates scanning electron microscopy (SEM) images of the composite membrane. (Left) Structure of composite membrane with substrate partially removed leaving behind a patterned surface. (Right) Composite membrane of about 100 pm thick adjacent to a substrate (cellulose paper);

Figure 3 displays images and illustrations of the highly compact zeolite particles (~ 1 pm) bonded together by cohered partially pyrolyzed polymer of the composite membrane; Figure 4 illustrates (a) rejection-time profile of a composite membrane (NaY zeolite (82- 84 vol%) and the partially pyrolyzed acrylic polymer) when used to perform nanofiltration of a Methylene Blue (MB) aqueous solution (10 ppm); and (b) a MB rejection-time profile of a composite membrane (additionally including ionic liquid, BMIC, ca. 23.15 wt% of the acrylic polymer - based on the composition of the precursor of membrane as defined above);

Figure 5 MB rejection-time profile of composite membrane (4A-zeolite molecular sieve, acrylic polymer, BMIC and glass fiber); Figure 6 illustrates (a) rejection-permeance profile for the rejection of MB (10 ppm aqueous solution) using a composite membrane (including BMIC and short glass fibers); and (b) rejection-permeance profile for the rejection of MB (lOppm in IPA) using composite membrane (including BMIC and short glass fibers). In both profiles, three separation runs are carried out and the error percentage lies well within 10%;

Figure 7 displays SEM image exhibiting of the zeolite composite membrane comprising approximately 85% NaY zeolite (including pores embedded), 10% carbonaceous, 5% glass fibers in volume (excluding volume of voids);

Figure 8 illustrates (a) an infrared spectrum of an example of the zeolite-carbonaceous composite membrane (ZCCM) compared to control membrane without zeolite, i.e. carbonaceous composite membrane (CCM); and (b) a thermal gravimetric analysis of the acrylic polymer (obtained from drying the emulsion) vs. the acrylic polymer with ionic liquid;

Figure 9 is an illustration of a nano-filtration cell where the round sheet membrane is sealed by a specific arrangement of rubber O-rings and porous stainless steel discs. The surrogate effluent (aqueous and organic solvent) is propelled into the cell by a liquid pump to maintain a pressure drop across the membrane;

Figure 10 compares the performances of ZCCM, a NaY zeolite powder packed bed (NPB) and CCM using an aqueous solution of MB (lOppm) as the probe effluent;

Figure 11 illustrates (a) variation of zeta potential of ZCCM and pure NaY with pH; and (b) comparison of rejection profiles of ZCCM, the packed-bed of the ground ZCCM powder and the packed-bed of NaY zeolite powder towards the aqueous solution of MB (lO ppm);

Figure 12 illustrates the effect of additives on the chemical and structural stability of the composite membranes fabricated for organic solvent systems;

Figure 13 illustrates a general protocol of forming a composite membrane on a curved surface (e.g. on a tubular substrate);

Figure 14 illustrates an exemplary setup to carry out the nano-filtration using the composite membrane;

Figure 15 presents representation of a) stainless steel-316 tube (GKN sinter metals ^® ) with an inset showing its external surface morphology; (b) carbonanceous-zeoliteY composite membrane (thickness -120 Dm) on the tubing with an inset showing its porous bulk structure; and (c) carbonanceous- kaolin composite membrane (thickness -120 Dm) on the tubing;

Figure 16 illustrates the evaluation results of the carbonaceous-kaolin composite membrane in Figure 15 with MB in water (30 ppm);

Figure 17 illustrates the evaluation results of the carbonaceous-zeolite Y composite membrane in Figure 15 with MB in water (30 ppm);

Figure 18 illustrates the evaluation results of the carbonaceous -kaolin composite membrane in Figure 15 with Brilliant Blue G (BBG) in water (30 ppm);

Figure 19 illustrates the evaluation results of the carbonaceous -kaolin composite membrane in Figure 15 with MB in water (50 ppm);

Figure 20 illustrates the evaluation results of the carbonaceous -kaolin composite membrane in Figure 15 with BBG in water (50 ppm);

Figure 21 illustrates a SEM and pictorial representation of the perfluoro-oligomer modified carbonaceous-zeolite Y composite membrane;

Figure 22 illustrates the evaluation results of the perfluoro-oligomer modified carbonaceous-zeolite composite membrane in Figure 21 with MB in isopropyl alcohol (IPA) (10 ppm); and

Figure 23 illustrates the evaluation results of the perfluoro-oligomer modified carbonaceous-zeolite composite membrane in Figure 21 with MB in isopropyl alcohol (IPA) (50 ppm).

DETAILED DESCRIPTION

The term 'nanofiltration' or NF refers to a membrane filtration-based method that uses nanometer sized through-pores that pass through the membrane. Nanofiltration membranes have pore sizes from 1-10 nanometers, smaller than that used in microfiltration and ultrafiltration, but just larger than that in reverse osmosis. NF is typically characterized by pressure drop up to 40 bars.

The term 'zeolite' refers to a microporous material that consists of three-dimensional aluminosilicate frameworks that are highly porous with varying cavity structures and negatively charged alumina tetrahedral units. These structural traits allow zeolites to possess superior cation-exchange capabilities and shape-selectivity, rendered by their rigid and regular pores.

The present invention is predicated on the understanding that while zeolite powders have relatively high specific surface areas of several hundred m ² /g, they are not good adsorbents. Instead, zeolite usually relies on its sub-nano pores to reject larger organic molecules via a size exclusion mechanism. Further zeolite lacks abundant affinity sites to act as adsorbent sites. The inventors have found that by introducing chemical functionalities to a composite membrane comprising zeolite and carbonaceous, the membrane can be an absorbent, even when it is crushed in a powder. Further, when a packed adsorption bed is packed with sub-mm adsorbent particles (for example pulverized ZCCM), a high pressure-drop across a bed is prevented.

As far as the fabrication of zeolite membranes is concerned, it is unviable to attain a zeolite membrane by the traditional ceramic sintering method because of structural vulnerability of zeolites at high temperatures. By the traditional method, a zeolite layer is grown on a porous substrate through a chemical synthesis protocol which involves depositing seeds for the growth of zeolite grain on the porous substrate followed by incubating the treated porous substrate in a specially formulated aqueous solution in order to grow zeolite grains from these seeds under hydrothermal or other heating treatment conditions. There is however a number of critical cons associated with this method. The pristine zeolite membrane is often brittle, and more significantly has a thickness of at least several microns, making it dense and resulting in substantially lower solvent permeation flux. The synthesis method is intrinsically limited to achieve a thickness of several microns without leaving defects because each zeolite grain has sizes of 1-2 micron. Besides, the approach is expensive and energy intensive.

Another known method is to embed a designed amount of zeolite into an existing polymer matrix, which is often polyimide (PI) or polybenzoimidazole (PBI) to form a composite membrane, wherein the loading of zeolite micron particles is normally below 20 wt.% (or 26 vol.%) (mixed matrix membrane). However, there are a number of uncertainties involved during moulding of such a system. The compatibility between the matrix and the filler is a concern and a poor adhesion can lead to defects. It was found that the zeolite particles are unable to form a continuous phase due to factors such as heavy aggregation and high percolation threshold and thus, the polymer matrix allows the permeation primarily. Poor interfacial compatibility between zeolite and the polymer matrix also undermine the separation performance.

Without wanting to be bound by theory, the inventors have found that the limitations of both zeolite-polymer composite and pristine zeolite membranes can be alleviated or overcome by a composite membrane in which the zeolite comprises the bulk phase, such that the zeolite particles undertakes a primary role in solvent transport and a secondary role in retention. In this composite membrane, the zeolite grains are of submicron size and are bonded/held together by a partially carbonised binding phase, where the latter accounts for a smaller volume/weight fraction (less than about 40 % of the composite membrane). This composite membrane is thus made up of closely bonded zeolite clusters, with a significant fraction of micron voids distributed among the bonded zeolite clusters as shown in Figure 3. The voids have a dendritic distribution and are not interconnected, accordingly they do not allow effluent to infiltrate through the membrane. These pores essentially collect tiny solvent streams from the bonded zeolite clusters and redistribute them. Such a porous structure largely improves permeation flux as compared to a pristine zeolite membrane. The composite membrane is also fundamentally different from a mixed matrix membrane, as the predominant zeolite phase comprising of highly compacted zeolite particles with sizes around 1 pm which aims to have the zeolites phase undertake a vast majority of the separation task (in mixed matrix membrane the polymer matrix fulfils the primary role of separation).

To this end, the inventors have found that the binding phase needs to be homogenously mixed with the bulk phase such that there is uniform contact between the binding phase (polymer) and bulk phase (zeolite particles); i.e. the binding phase is substantially dispersed and/or amalgamated within the bulk phase. For example, this can be achieved first by the mixing of zeolite powder with an emulsion of acrylic polymer, which results a homogeneous paste. After applying the paste on a porous support (by brushing or dipping) and drying to form a coating layer, the coating layer (precursor of membrane) is then facilitated by a compression step to extrude the polymer phase into the interstices of the zeolite phase, thereby substantially enhancing the interfacial contact of both (Figure 1 and 2). In addition, by converting polymer into a partially carbonised polymer via a controlled pyrolysis step, the polymer is converted into a hydrocarbon skeleton that is strengthened but still retains its functional groups such as ester, imine, carboxylate, olefin and the likes on associated polyaromatic hydrocarbons (aPAH). This has the further advantage of making the composite membrane absorbent while at the same time providing strong bonding between the zeolite particles and holds the bulk phase together (Figure 3). In addition, it also aids in retention owing to a sufficiently charged surface that holds the molecules through electrostatic association. Since, the thermal treatment is controlled, the zeolite microstructure remains intact.

It is believed that this fabrication relies on the formation of a homogeneous paste and a strong adhesion between particles (the bulk phase) and carbonaceous (the binder phase), and the micron-cell structure as described above. Hence, for other microporous particles can be used as the bulk phase in combination with an appropriate binder phase as long as the first two conditions are satisfied.

In some embodiments, the bulk phase comprises porous particles. The inventors have found that inorganic oxide particles for use in the composite membrane preferentially contain through micropores (<2 nm) or lower mesopores (2-10 nm). The selection of porous particles is dependent upon the size of solutes to be filtered. In another embodiment, the bulk phase comprises porous particles with mesoporous cage structures. In most embodiments, the bulk phase comprises zeolite particles. In some embodiments, the zeolite particles are selected from NaY, 4A or a combination thereof. In other embodiments, the zeolite particles are selected from faujasite Y, faujasite X (same pore size as faujasite Y but higher hydrophilicity), large pore zeolite such as ITG-21 (7.4 A) and titanium-silicate (8.0 A), medium pore zeolite such as ZSM-5 (5.5 A) and mordenite (6.6 A), metal organic framework (MOF) such as Zif-8 (6-8 A), or a combination thereof. In other embodiments, the zeolite particles are about 1 pm in size. In some embodiments, kaolin can be used to form the composite membrane. Kaolin, possesses the layered structure instead of the cage structure, and in this regard the formed membrane is characterized by diffused and rough pores in the amalgam of kaolin and carbonaceous. Other types of clay can also be used in place of kaolin. Other microporous particles, such as metal organic framework (MOF) or (polyhedral oligomeric silsesquioxane) POSS, can also be used.

The inventors have found that it is advantageous for the polymers used to form the at least partially carbonised polymer in the binding phase to be easy to thoroughly mix with bulk phase. In this regard, water-borne acrylic emulsion can be used as it wets zeolite particles entirely besides being environmental friendly. Alternatively, polymer solution, such as toluene solution of poly(ethyl methacrylate) (PEMA), may also be used because zeolite NaY can be uniformly dispersed in it with the aid of Nafion. Secondly, a polymer having a glass transition temperature (Tg) below room temperature or having its Tg to be reduced to below room temperature after entrapped a small amount of solvent is also advantageous. This allows the polymer to be extruded into interstices of zeolite particles under compression at room temperature. The acrylic polymer left behind from acrylic emulsion has Tg at about 5-10 °C, whereas PEMA entrapping said about 3-6% toluene in its matrix has Tg below room temperature. Thirdly, having the polymer can undergoing partial thermal decomposition at temperatures below but close to the controlled pyrolysis temperature (for example 300 °C) can provide another advantage. On top of the compression, the carbonisation process serves to harden the composite membrane, making it more resilient during use. Further, the carbonisation acts to further strengthen the bonding of the bulk phase. By carbonising the binding phase, the adhesive property of the binding phase is increased. In this regard, by only partially carbonising the binding phase, an improvement of the physical properties can be imparted to the composite membrane without a significant reduction of the chemical properties of the composite membrane. Additionally, the carbonisation of the polymer also introduces nanopores into the binding phase, thereby allowing the working pressure of the system to be generally lower.

In this regard, the binding phase is substantially dispersed within the bulk phase. In other embodiments, the binding phase is substantially amalgamated within the bulk phase; i.e. the binding phase is dispersed and merged with the bulk phase to form a single cohesive structure. In this sense, the binding phase is extruded into interstices of the bulk phase.

In an embodiment, the composite membrane comprises:

a) a bulk phase comprising porous particles, the porous particles having micropores of less than about 2 nm and/or mesopores in a range of about 2 nm to about 10 nm; and b) a binding phase comprising an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, the binding phase substantially amalgamated with the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In another embodiment, the composite membrane comprises:

a) a bulk phase comprising zeolite particles; and

b) a binding phase comprising an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, the binding phase substantially amalgamated with the bulk phase;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

As used herein, 'pyrolysis' and 'carbonisation' are used interchangeably. A 'partially carbonised polymer' is also referred to as 'carbonaceous'. In this regard, the term 'carbonaceous' or 'partially carbonised polymer' refers to a carbon material that has been subjected to a partial (non-complete) thermal decomposition process and contains a large amount of carbon content and at the same time also bears oxygen containing groups. This is in contrast to a complete thermal decomposition in which the oxygen containing groups are completely lost.

It will be appreciated that the at least partially carbonised polymer may have less functional groups such as COOH groups than original polymer prior to the carbonisation step. In some embodiments, the at least partially carbonised polymer comprises organic functional groups such as associated polyaromatic hydrocarbons (aPAH) bearing COOH groups, In other embodiments, the at least partially carbonised polymer comprises polycyclic aromatic hydrocarbons (PAHs), aliphatic, aromatic/olefins, phenolic or a combination thereof. The skilled person would understand that the at least partially carbonised polymer can be characterised using, for example, IR spectrum. In this regard, the composite membrane may be physically processed such that it is readable by an IR spectrometer. There are five characteristic organic structures: aliphatic hydrocarbon branches (2963, 1453 cm ¹ ); aromatic rings (1597, 1497, 889-540 cm ¹ ); olefins (1640 cm ^x ), carboxylic (1727 cm ¹ ), phenolic (3408, 1147 cm ¹ ) and tetraheral Si(Al)0 ₂ (1106, 985 cm ¹ ). An example of the infra-red (IR) values from a composite membrane of the present invention is given in Figure 8 and in Table 1 below.

Table 1. IR spectrum results of the composite membrane.

In an embodiment, the at least partially carbonised polymer comprises polycyclic aromatic hydrocarbon groups. In other embodiments, the at least partially carbonised polymer comprises polycyclic aromatic hydrocarbons (PAHs), aliphatic, aromatic/olefins, phenolic or a combination thereof. In other embodiments, the at least partially carbonised polymer has an IR absorbance of about 1714 cm ¹ and about 3289 cm ¹ . In other embodiments, the at least partially carbonised polymer has an IR absorbance of about 1714 cm ¹ , about 3289 cm ¹ and about 1641 cm ¹ . In other embodiments, the at least partially carbonised polymer has an IR absorbance of about 1714 cm ¹ , about 3289 cm ¹ , about 1641 cm ¹ and about 924-558 cm ¹ . In other embodiments, the at least partially carbonised polymer has an IR absorbance of about 1714 cm ¹ , about 3289 cm ¹ , about 1641 cm ¹ , about 924-558 cm ¹ , about 1161 cm ¹ and about 3436 cm ¹ .

The at least partially carbonised polymer can be characterised by a loss of weight compared to the original uncarbonised polymer. In this regard, the partial pyrolysis can be controlled by the reaction temperature and duration. For example, a 20 to 30% weight loss from the polymer can be obtained at 300 °C. In this way, the carbonaceous formed still keep about 80 to 70% mass. More importantly the oxygen containing groups, such as carboxylic and phenolic groups are preserved. Accordingly, 10% carbonaceous refers to its content which is removed from the original polymer, in other words, roughly, 90% of the original polymer is retained after carbonisation. This is in contrast to completely carbonising the polymer, which results in a composite membrane which is brittle, fragile and overly porous.

Accordingly, in some embodiments, the at least partially carbonised polymer of the binding phase is characterised by an at least 10% carbonisation. In other embodiments, the at least partially carbonised polymer of the binding phase is characterised by an at least 20% or at least 30% carbonisation. In some embodiments, the at least partially carbonised polymer of the binding phase is characterised by a weight loss of at least 10% compared to the original polymer. In other embodiments, the polymer of the binding phase is characterised by a weight loss of at least 20% or at least 30% compared to the original polymer. In some embodiments, the at least partially carbonised polymer of the binding phase is characterised by a weight loss of about 5% compared to the original polymer, or about 10%, about 15%, about 20%, about 25% or about 30%. Alternatively, in some embodiments, the at least partially carbonised polymer of the binding phase is characterised by a weight loss of at least 5% compared to the original precursor weight. In other embodiments, the at least partially carbonised polymer of the binding phase is characterised by a weight loss of at least 10%, at least 15%, at least 20%, at least 25% or at least 30% compared to the original precursor weight.

Different types of polymer can provide different chemical properties to the composite membrane. In some embodiments, the binding phase comprises an at least partially carbonised polymer. The polymer is preferentially a polymer with low Tg and good compatibility with the bulk phase. Accordingly, in some embodiments, the polymer is an acrylic polymer such as copolymers of methyl, ethyl, butyl, hexyl, and lauryl methacrylates. In other embodiments, the polymer is selected from poly(ethyl methacrylate), poly (propylene oxide) and its copolymers with poly (lactic acid) and poly(maleic acid).

In some embodiments, the bulk phase is about 60 vol% to about 90 vol% of the composite membrane. In other embodiments, the bulk phase is about 70 vol% to about 90 vol%. In other embodiments, the bulk phase is about 75 vol% to about 90 vol%. In other embodiments, the bulk phase is about 80 vol% to about 90 vol%. In other embodiments, the bulk phase is about 80 vol% to about 85 vol%.

In some embodiments, the binding phase is about 10 vol% to about 40 vol% of the composite membrane. In other embodiments, the binding phase, i.e. carbonaceous, is about 10 vol% to about 30 vol%. In other embodiments, the binding phase is about 10 vol% to about 25 vol%. In other embodiments, the binding phase is about 10 vol% to about 20 vol%. In other embodiments, the binding phase is about 15 vol% to about 20 vol%.

Figure 4(a) illustrates rejection-time profile of a composite membrane (NaY zeolite (82-84 vol%) and the partially pyrolyzed acrylic polymer) when used to perform nanofiltration of a Methylene Blue (MB) aqueous solution (10 ppm). A good rejection can be obtained for at least 100 minutes. Figure 16 illustrates the evaluation results of the carbonaceous-kaolin composite membrane with MB in water and Figure 17 illustrates the evaluation results of the carbonaceous-zeolite Y composite membrane with MB in water. Figure 18 further illustrates the evaluation results of the carbonaceous-kaolin composite membrane with Brilliant Blue G (BBG) in water. The higher concentrations of these two dyes were investigated as well (Figure 19 and 20).

Figure 19 and 20 compares two types of composite membrane: NaY-carbonaceous vs. kaolin-carbonaceous, which start from different compositions of precursor and are made by using different compression conditions before pyrolysis. As a result, the two composite membranes have different porous structures: the continuously cohered zeolite grains to construct a network (former) and the aggregation of kaolin grains and carbonaceous particles (latter). In some embodiments, the binding phase further comprises an additive. The additive can be glass fibers. The glass fibers may be processed such that they are of a suitable dimension as mentioned above. In other embodiments, the binding phase further comprises carbon nanotubes and/or cellulose fiber. The fibrous additive, when present, creates connections among more voids (or micron cells) within the composite membrane, which allows for a higher permeance.

It was further found that addition of ionic liquid aids in adequate compaction of the zeolite particles by shielding the charges on the particle as well as the polymeric binder. For example, when l-butyl-3-methylimidozalium chloride (BMIC) was used, the negative charges on the zeolite particles and polymer can be shielded. Also, ionic liquid (when partially carbonised) furthers the overall thermal decomposition and results in more degradation (Fig. 8b). The thermal gravimetric analysis shows that transformation of acrylic polymer to carbonaceous at 300 °C retains a major portion of the polymer. However, the real thermal treatment dwells at 300 °C for 1 h and involves the fast degradation of the ionic liquid compound that loses about 80% of mass at 300 °C, hence the degradation of ionic liquid reduces the stability of acrylic polymer, ending up with a thermally stable skeleton (carbonaceous) comprising conjugated C=C, C=N, and C(0)0 bonds as shown by infrared spectrum in Figure 8a. The addition of glass fiber is advantageous for acting as a spacer to break up connection of the zeolite-carbonaceous amalgam network (Fig. 3) and enhance the void fraction of the membrane matrix. (Fig 7). Glass fiber micron rods intermingle with the zeolite-carbonaceous phase to increase voids in order to curtail resistance to filtration. The significance of these additives in the binding phase is realized by their effect on the permeation flux and in the rejection values (Table 2).

Table 2: Comparison of composite membrane and the effect of additives on the final performance to separate MB from water (10 ppm)

Assessment time 60 min; ^b The mass ratio of IL/ Acrylic.

Figure 4(b) illustrates a MB rejection-time profile of a composite membrane (additionally including ionic liquid, BMIC, ca. 23.15 wt% of the acrylic polymer). A rejection rate of at least 97% is obtained.

In some embodiments, the at least partially carbonised ionic liquid is selected from an ionic liquid comprising an imidazolium type ionic liquid or a cationic ionic liquid. In other embodiments, the at least partially carbonised ionic liquid is selected from an ionic liquid such as N-butyl-3-methyl imidazolium chloride (BMIC), N-butyl- 3 -methyl imidazolium nitrate, N-butyl- 3 -methyl imidazolium acetate, N-alkyl-pyridinium, tetraalkyl phosphonium, tetraalkyl ammonium chloride, tetraalkyl ammonium nitrate and tetraalkyl ammonium acetate.

In some embodiments, the binding phase further comprises glass fibers (Figure 7). The glass fibers have a diameter of about 0.5 pm to about 0.9 pm. When in use, the glass fibers are grounded such that their aspect ratio is about 10:1 to about 2:1 and thoroughly dispersed. The additive can be about 1% v/v of the composite membrane, or about 2%, about 3%, about 4% or about 5%.

Figure 5 illustrates the rejection profile of MB using composite membranes comprising zeolite-4A, ionic liquid and glass fibers as control. Although, as demonstrated, a rejection rate of at least 98% is obtainable, the use of 4A bring about a much lower permeance than the composite membrane where only zeolite-Y is used construct the bulk phase. Figure 6 further illustrates the rejection and permeance profiles of MB using the composite membranes comprising ionic liquid and glass fibers, in which water and IPA are solvent, respectively. As shown in Figures 6(b) and 20, using a same feed and compared with the organic solvent nanofiltration (OSN) using a non-fluorinated membrane with an average permeance of 7.5 (L/m ² -h-bar), the fluorine modified membrane displays a clearly higher average permeance of 75 (L/m ² -h-bar).

The composite membrane can be used for organic solvent nanofiltration. An example of a set-up for use in organic solvent nanofiltration is shown in Figure 9. To allow the composite membrane to be usable with an organic solvent, i.e. IPA (instead of water), other additives can be added to the binding phase. In this regard, a fluorinated polymer can be added. For example, a perfluoro sulfonated ionomer resin, Nafion (5wt% in alcohol/water) can be added to the binding phase while the bulk phase remains the same (Figure 21). This will facilitate the permeation of the organic solvent, to sustain NF of the IPA solution of MB because MB is a neutral molecule in IPA rather than ions in water. The results of separating MB from IPA are shown in Table 3 and Figure 22. This Figure presents the same assessment as Fig. 6b except for a longer duration. In addition, the membrane also demonstrates enough capability to separate a higher concentration feed (50 ppm) (Fig. 23). It can be seen that composite membrane when used in water or organic solvent both manifest strong efficacies to remove soluble organic species from their respective solvents through NF mechanism.

Table 3: Effect of additives on the chemical and structural stability of the composite membranes fabricated for organic solvent systems

Assessment by running IPA through the membrane for 4 hours In an embodiment, the binding phase further comprises an at least partially carbonised fluorinated polymer. In another embodiment, the binding phase comprises Nafion and poly(ethyl methacrylate). In another embodiment, the binding phase comprises Nafion, poly(ethyl methacrylate) and polycyanoacrylate.

It is believed that the extension of hydraulic compression duration to enhance amalgamation between fluorinated polymer and zeolite through squeezing extensive polymer flow into interstices of zeolite grains and their surface defects. The use of a right adhesive to clip the amalgamation attained is believed to be responsible for the above additional improvement.

The composite membrane can be used with any aqueous solvents. "Aqueous solvent" as used herein, refers to a water based solvent or solvent system, and which comprises mainly water. Such solvents can be either polar or non-polar, and/or either protic or aprotic. Solvent systems refer to combinations of solvents which resulting in a final single phase. Both 'solvents' and 'solvent systems' can include, and is not limited to, pentane, cyclopentane, hexane, cyclohexane, benzene, toluene, dioxane, chloroform, diethylether, dichloromethane, tetrahydrofuran, ethyl acetate, acetone, dimethylformamide, acetonitrile, dimethyl sulfoxide, nitromethane, propylene carbonate, formic acid, butanol, isopropanol, propanol, ethanol, methanol, acetic acid, ethylene glycol, diethylene glycol or water. Water based solvent or solvent systems can also include dissolved ions, salts and molecules such as amino acids, proteins, sugars and phospholipids. Such salts may be, but not limited to, sodium chloride, potassium chloride, ammonium acetate, magnesium acetate, magnesium chloride, magnesium sulfate, potassium acetate, potassium chloride, sodium acetate, sodium citrate, zinc chloride, HEPES sodium, calcium chloride, ferric nitrate, sodium bicarbonate, potassium phosphate and sodium phosphate. As such, biological fluids, physiological solutions and culture medium also falls within this definition.

The composite membranes can be used with any organic solvents. "Organic solvent" as used herein, refers to an organic based solvent or solvent system, and which comprises mainly organic solvent. Organic based solvents can be any carbon-based solvents. Such solvents can be either polar or non-polar, and/or either protic or aprotic. Solvent systems refer to combinations of solvents which resulting in a final single phase. Both 'solvents' and 'solvent systems' can include, and is not limited to, pentane, cyclopentane, hexane, cyclohexane, benzene, toluene, dioxane, chloroform, diethylether, dichloromethane, tetrahydrofuran, ethyl acetate, acetone, dimethylformamide, acetonitrile, dimethyl sulfoxide, nitromethane, propylene carbonate, formic acid, butanol, isopropanol, propanol, ethanol, methanol, acetic acid, ethylene glycol, diethylene glycol or water. Organic based solvents or solvent systems can include, but not limited to, any non-polar liquid which can be hydrophobic and/or lipophilic. As such, oils such as animal oil, vegetable oil, petrochemical oil, and other synthetic oils are also included within this definition. In some embodiments, the solvent is selected from isopropyl alcohol (IPA), ethyl acetate, toluene or dimethylformamide. In other embodiments, the solvent is selected from IPA, ethanol, acetone, toluene, cyclohexane and dimethylformamide.

Figure 12 illustrates the effect of additives on the chemical and structural stability of the composite membranes fabricated for organic solvent systems. For example, the addition of an additive, while providing more voids, also improves chemical stability of the membrane. The degree of pyrolysis and/or the amount of binding phase can affect the degree of delamination. The inventors have further found that adding an adhesive in the binding phase also help to improve the adhesion to a surface. Alternatively, the adhesive can be applied directly to the surface. Accordingly, in an embodiment, the composite membrane further comprises an adhesive. The adhesive can for example be an acrylic polymer and/or polycyanoacrylate.

Without wanting to be bound by theory, it is believed that the bonded zeolite-carbon clusters offer two pathways for the solvent flow: the nano pores of zeolite and the pores of carbonaceous. Zeolite offers enhanced water permeation. Water flows through the zeolite pores as well as the porous carbonaceous phase, into the micron voids (enclosed by the zeolite-carbonaceous clusters), and then out of the membrane. Similarly, in the case of organic solvent systems, the solvent travels through both the zeolite pores and the carbonaceous. The enhanced hydrophobicity between the solvent and the carbonaceous favours the passage of solvent through the membrane owing to selective repulsion of carbonaceous against the solute over solvent.

The composite membrane, excluding the substrate such as the carbonized filtration paper (CFP), is about 100 pm thick (Figure 2) and has a tap density of about 0.4-0.5 g/cm ³ . Accordingly, in some embodiments, the composite membrane has a thickness of about 100 pm. In other embodiments, the composite membrane has a thickness of about 90 pm to about 130 pm, or about 80 pm to about 140 pm, 70 pm to about 150 pm, 60 pm to about 160 pm, 50 pm to about 170 pm or 50 pm to about 200 pm.

As shown in Figure 3, voids (<l pm) are distributed throughout the entire structure and these are essentially the results of the interstices present in the packing of the zeolite grains-carbonaceous amalgam. The zeolite-carbonaceous phase surrounding the voids constitute a lot of tiny filters and hence significantly reduce the pressure required to drive NF. As a result, the composite membrane requires low transmembrane pressures in the range of 0.5-8 bars to carry out NF. The pore channels are also extremely tortuous in nature and such a dendritic channel network virtually works to gather thin filtrate streams from various bonded zeolite clusters. Such a network allows a low trans-membrane pressure, thus sustaining an average water permeation flux of around 88 L/m ^2- h while offering rejection rates > 95% for MB (Fig. 6a).

The membrane made by using CFP as substrate possesses adequate mechanical strength and can withstand pressure drop up to about 3 bars. The membranes made by using stainless steel disc (SSD) as substrate can withstand transmembrane pressure up to about 20 bars or more.

Figure 11(b) provides a comparison of rejection profiles of ZCCM, powder packed-bed of the pulverized ZCCM and NaY zeolite powder packed bed (NaY) towards the aqueous solution of MB (10 ppm). The composite membrane is able to perform better than the composite powder after compaction.

The composite membrane demonstrates advantageous capabilities to remove trace amount of small molecules from effluents. When tested with methylene blue in proxy effluent, the composite membrane shows a rejection of more than about 90%. In other embodiments, the rejection is more than about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99% or about 99.5%. It is believed that the rejection is primarily a combination of size exclusion (steric hindrance) and coulombic interactions, particularly for the aqueous system. Physical adsorption does not play an important role since the individual components of the composite membrane (acrylic polymer binder and IL) or pristine NaY powder in a fixed- bed shows trivial capability to remove MB (10 ppm aqueous solution of MB in all the control experiments) (Fig 10). Further, a control sample of a powder formed by crushing a few pieces of composite membrane and assessing the separation potential in a packed bed performed in a similar way as that for the NaY powder (Fig l lb). The control test exhibits a rejection drop of 28% at 6 hours whereas composite membrane displays a stable rejection until 10 hours. Pore blocking initiated by membrane fouling (due to accumulation of MB) may also play a role in solute exclusion thus contributing rejections over long periods. For the NF of the IPA solution of MB, a range of complex interactions is believed to take place, where stronger hydrophobic rejection between MB and perfluorocarbon-modified carbonaceous pores of the membrane (owing to weaker dispersion force and larger interaction cross-section) primarily impacts the overall membrane performance (solute rejection and permeance). It is noteworthy that MB does not dissociate in IPA to form ions like what happens in water, charge interactions are no longer significant in solute rejection.

The molecular weight cut-off (MWCO) refers to the lowest molecular weight solute (in daltons) in which 90% of the solute or molecule is retained by the membrane. In some embodiments, the molecular weight cut-off (MWCO) is about 200 Da, or about 250 Da, about 300 Da, about 350 Da, about 400 Da or about 500 Da.

The composite membranes can be used for removing dyes from effluents in textile industry or large polar organic compounds, which are mainly phenolic, carboxylic and polyaromatic types (e.g. > 400 au.), from waste or produced water in petroleum refinery.

The composite membrane can be used in purification processes in the pharmaceutical industries. In this regard, in designing the composite membrane, the inventors have considered Lipinski's rule of five (Pfizer's rule of five or R05) as a rule of thumb to evaluate druglikeness or determine if a chemical compound with a certain pharmacological or biological activity has chemical properties and physical properties that would make it a likely orally active drug in humans. This is based on the observation that most orally administered drugs are relatively small and moderately lipophilic molecules. The rule describes molecular properties important for a drug's pharmacokinetics in the human body, including their absorption, distribution, metabolism, and excretion. Lipinski's rule states that, in general, an orally active drug has no more than one violation of the following criteria:

1) No more than 5 hydrogen bond donors (the total number of nitrogen-hydrogen and oxygen-hydrogen bonds);

2) No more than 10 hydrogen bond acceptors (all nitrogen or oxygen atoms);

3) A molecular mass less than 500 daltons; and

4) An octanol- water partition coefficient log P not greater than 5.

In this regard, the inventors have studied the electrostatic attraction, charge exclusion, interfacial p-effect, and dragging of counterion clouds of interactions between the membrane and molecules/compounds.

The inventors have also considered privileged scaffolds for designing drug compounds and have found that the composite membrane can be useful for purifying these compounds. Such privileged compounds are for example:

The composite membranes of the present invention are suitable for separating solutes from a solution and/or insoluble particles from a mixture. For example, solutes such as vitamin E and tannic acid can be separated from solvents such as isopropyl alcohol and ethyl acetate. Depending on the additive used, the composite membrane can also be used to separate charged solutes in for example an aqueous solvent.

To realize this composite structure, a specific low-Tg acrylic polymer in the form of emulsion or solution was blended with the zeolite grains to form a paste. Following the coating and drying process, the precursor layer was subjected to hydraulic compression. The compressed precursor was then subjected to controlled partial pyrolysis under low- oxygen conditions to convert the acrylic to carbonaceous. The controlled pyrolysis conducted aims at mediating the extent of carbonization so that pendant carboxylic (COOH) groups are preserved. Further the formation of a slightly flexible aPAH frame is also obtained.

The composite membrane of the present invention can be supported on a substrate. For example, the substrate can be a porous surface. This porous surface can be a filter paper or a porous metallic surface, such as a stainless steel 316 tube or a metal mesh fabric. As is discussed below, the composite material for forming the composite membrane is first applied to the substrate and allowed to dry. Several layers can be applied in this manner to control the thickness of the resultant membrane. The dried material is subsequently compressed to form the precursor of composite membrane on the substrate. This further allows for the amalgamation of the bulk phase and the binding phase. Advantageously, due to the porous nature of the substrate/surface, some of the composite material enters the pores of the substrate, thereby increasing the adhesion of the composite membrane to a substrate to provide a more rigid and durable composite membrane for use. Because the substrate is porous, there is no need to remove it and the composite membrane and substrate can be used as formed. Further, as only physical interaction such as Van der Waals forces is employed and which is shown to be strong enough, there is no need for further chemical processing.

Accordingly, in another aspect, the present invention provides a method of fabricating a composite membrane. Figure 1 illustrates the general protocol, which includes coating, drying, hydraulic compression and thermal treatment to conduct partial pyrolysis. Figure 13 illustrates a general protocol of forming a composite membrane on a curved surface (e.g. a tubular metal or ceramic substrate). Figure 14 illustrates an exemplary setup to carry out the nano-filtration using the composite membrane and Figure 15 presents representation carbonanceous-zeoliteY composite membrane (thickness -120 Dm) on stainless steel-316 tube (GKN sinter metals®) and carbonanceous-kaolin composite membrane (thickness -120 Dm) on the tubing.

In a method of fabricating a composite membrane, the following steps are taken:

a) providing a bulk phase comprising zeolite particles; and

b) mixing the bulk phase with a binding phase comprising a polymer to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups, wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In another method of fabricating a composite membrane, the following steps are taken: a) providing a bulk phase comprising zeolite particles; and

b) mixing the bulk phase with a binding phase comprising a polymer and an ionic liquid to form a paste, such that the binding phase is substantially dispersed within the bulk phase; and

c) partially pyrolysing the paste such that the polymer is converted to an at least partially carbonised polymer and the ionic liquid is converted into an at least partially carbonised ionic liquid;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In this method, the ionic liquid (IL) can be added to the formulation of the paste to form the binding phase. The IL can function as dispersion aid as well as pyrolysis promoter; i.e. inclusion of IL into the acrylic emulsion enhances the homogenization of zeolite with the acrylic polymer.

In some embodiments, the binding phase further comprises glass fibers. The glass fibers may be processed such that they are of a suitable dimension as mentioned above to introduce more connections among cluster in the ZCCM.

To transform the paste into a composite membrane, the paste can be subjected to a drying step. The mixing the bulk phase with a binding phase initially comes together in the form of an emulsion. The emulsion can be in an aqueous medium or organic solvent. To be able to apply this emulsion consistently to a surface, the emulsion is formed into a paste; i.e. thickening of the emulsion via the removal of the solvent. In this regard, the binding phase is substantially blended with the bulk phase. The paste can thus be easily applied to a surface (such as a porous metallic or ceramic support) to form a coating layer, without the paste displaying substantial flow rheology.

The paste may further be subjected to a drying step. The drying step is performed after the paste is applied to a porous substrate or surface. The drying step allows for coalescence of polymer latex left behind from the emulsion coating as described above and therefore better compaction of the membrane. In an embodiment, the drying step is performed under ambient conditions. In another embodiment, the drying step is performed at about 30 °C, about 40 °C, about 50 °C, about 60 °C, about 70 °C or about 80 °C. In another embodiment, the drying step is performed for more than 1 h, 1.5 h, 2 h, 3 h, 4 h, 5 h, 6 h, 8 h, 10 h or 12 h.

To transform the paste into a composite membrane, the paste can be further subjected to a compression force. The compression forces the binding phase to flow and not only fill interstices and crevices in the zeolite (for example) but also cover the surfaces of the zeolite. In this regard, the binding phase is thoroughly amalgamated with the bulk phase due to the effect of pressurizing. This allows formation of unique frame and micron-cell membrane structure and robust binding between membrane and substrate after carbonisation. This compression can be a hydraulic compression or can be an isostatic compression. For example, the pressure for hydraulic compression (0.8 T/cm ² or 78.4 MPa) can be chosen since it is approximately the median pressure for extruding polymer melts. Hence, the polymer in highly rubbery state can be extruded to move into tiny seams among zeolite grains and to wrap their surface as well. Such in-depth compositing can be used to achieve an amalgam of zeolite and carbonaceous, which can in turn affect membrane porosity, permeance and selectivity. Accordingly, in some embodiments, prior to step (c), the paste is subjected to a pressure of about 60 MPa to about 100 MPa. The pressure may further be adjusted to suit the application needs. For example, a higher pressure about 80 MPa or higher is suitable if a lower permeance is acceptable with limited improvement on rejection. In an embodiment, the compression pressure is from about 60 MPa to about 150 MPa, from about 60 MPa to about 140 MPa, from about 65 MPa to about 140 MPa or from about 65 MPa to about 100 MPa. In some embodiments, the hydraulic pressure is of about 60 MPa to about 100 MPa or the isostatic pressure is of about 67 to 140 MPa.

In an embodiment, the compression force is exerted for a period of about 10 min, about 15 min, about 20 min, about 25 min or about 30 min. In another embodiment, the compression force is exerted for more than 10 min, 15 min, 20 min or 30 min. The duration of compression has the effect on the penetration of polymer into interstices of the bulk phase within the dry paste coating, i.e. the precursor.

The paste may further be subjected to a compression force at temperatures above ambient temperature to facilitate extrusion of the polymer precursor.

This final thickness of composite membrane is the result of the paste coating and subsequent compression as described above. It has been found that the compression will in most cases reduces the membrane thickness by 15 pm when formed. The inventors believe that a thicker membrane is unnecessary because this will increase rejection at the cost of permeance.

In some embodiments, the partial pyrolysis step is performed at about 300 °C. In other embodiments, the partial pyrolysis step (step C) is performed at about 300 °C under low oxygen conditions. It was found that a higher temperature of 350 °C adversely impacts the rejection whereas a lower temperature (around 280 °C) results in much lower rejection and solvent flux. The fundamental need for conducting pyrolysis is to achieve a carbonaceous binding phase as it possesses firstly a stable porous structure against temperature and solvent, secondly pendant functional groups to sustain the electrostatic actions, strong adhesion to zeolite, and thirdly a better ductility than the usual carbon matrix to ensure freestanding membrane structure. Therefore, control over the pyrolysis temperature is needed as adoption of too high or too low pyrolysis temperature will erode the above properties.

In an embodiment, the partial pyrolysis is performed for a period of about 1 h, about 1.5 h, about 2 h, about 2.5 h or about 3 h. In another embodiment, the partial pyrolysis is performed for more than 1 h, 1.5 h, 2 h or 3 h.

To further retain the functionality of the binding phase and/or to partially carbonise the polymer/ionic liquid, the partial pyrolysis can be performed under low oxygen conditions. In this regard, the oxygen content in the pyrolysis chamber is controlled to be less than atmospheric conditions; i.e. less than 21%. In other embodiments, the oxygen content is less than 18%, less than 15%, less than 12%, less than 10%, less than 5% or less than 3%.

In some embodiments, in the method of fabricating a composite membrane, the following steps are taken:

a) providing a bulk phase comprising zeolite particles;

b) mixing the bulk phase with a binding phase comprising a polymer to form a paste, such that the binding phase is substantially dispersed within the bulk phase;

c) applying the paste to a surface and allowing the paste to dry to form a coating layer; d) compressing the coating layer; and

e) partially pyrolysing the coating layer such that the polymer is converted to an at least partially carbonised polymer, the at least partially carbonised polymer comprises COOH groups,

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

In another method of fabricating a composite membrane, the following steps are taken: a) providing a bulk phase comprising zeolite particles; and

b) mixing the bulk phase with a binding phase comprising a polymer and an ionic liquid to form a paste, such that the binding phase is substantially dispersed within the bulk phase; c) applying the paste to a surface and allowing the paste to dry to form a coating layer; d) compressing the coating layer; and

e) partially pyrolysing the coating layer such that the polymer is converted to an at least partially carbonised polymer and the ionic liquid is converted into an at least partially carbonised ionic liquid;

wherein the bulk phase is about 60 vol% to about 90 vol% of the composite membrane, and

wherein the binding phase is about 10 vol% to about 40 vol% of the composite membrane.

The present invention demonstrates that a highly compact zeolite particle phase bound by a carbonaceous phase with abundant micron pores or voids is viable and effective in performing NF. While the invention is demonstrated as a flat sheet membrane, the skilled person would understand that the composite membrane structure can be realized in a tubular or flat interlayer forms by applying extrusion, roll milling or isostatic compression techniques for large scale industrial applications. For example, the composite membrane can be applied on alumina or stainless steel by applying cold or hot isostatic pressing. It is envisioned that the present invention is applicable in water and organic solvent treatment industries aiming to remove dilute soluble organics from industrial effluents. The composite membrane structure can also be used to concentrate organic solutions by permeating solvents for isolating solutes that may include relatively larger reactant and product molecules as well as homogeneous catalysts.

Those skilled in the art will appreciate that the invention described herein in susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications which fall within the spirit and scope. The invention also includes all of the steps, features, compositions and compounds referred to or indicated in this specification, individually or collectively, and any and all combinations of any two or more of said steps or features.

Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

The reference in this specification to any prior publication (or information derived from it), or to any matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of endeavour to which this specification relates.

Certain embodiments of the invention will now be described with reference to the following examples which are intended for the purpose of illustration only and are not intended to limit the scope of the generality hereinbefore described.

EXAMPLES

Materials

Sodium Y zeolite (Sigma-Aldrich, p = 1.92 g/cm ³ ), waterborne acrylic emulsion (W-l, Cement strengthener, Warrior Pte. Ltd Singapore, 48 wt.% acrylic resin, p = 1.12 g/cm ³ ), ionic liquid (IL), 1 -butyl-3 -methylimidazolium chloride (Sigma-Aldrich, > 98.0%), glass fiber pre-filter (Sigma-Aldrich, Millipore type, 5*4), cellulose filtration paper (Advantec, diameter=l50mm, thickness= 0.2lmm), Methylene Blue hydrate (Sigma-Aldrich, Bioreagent), Rhodamine B (Sigma-Aldrich, >=95% HPLC), Methyl Orange (Sigma- Aldrich, ACS reagent, dye content 85%), are used as received without further purification. The porous stainless-steel round disc (PSD, 2.5 cm _ 0.1 cm with a mean pore size of 0.25 mm) is obtained from Mott Corp. Farmington, CT, USA. General Protocol: Preparation of the precursor (zeolite-acrylic composite)

The precursor used to fabricate the membrane varies in accordance with the end application (typically, the identity of the solvent). The variation is introduced in the form of binder system and additives. The primary constituent (NaY Zeolite) as well as the steps for fabrication remain the same.

Example 1: For the separation of aqueous MB solution (lOppm)

A given amount of NaY zeolite powder (1.6 g) was first blended in a mortar and pestle with the W-l waterborne acrylic emulsion (1.8 g,) accompanying grinding. To this dispersion, ionic liquid (IL) [BMIMjCl (0.2 g), a piece of glass filtration paper (ca. 0.14 g) and 3 ml water were added. The mixture was homogenized well to ensure there is no aggregation of the zeolite submicron-particles and until the precursor had the consistency of a paste. The formulated precursor contains about 67.3% NaY, 23.3% acrylic latex resin, 5.4% IL and 3.73% GF by weight. Subsequently, about a quarter of this precursor paste was carefully applied using a brush to one face of a cellulose filtration paper (2pm pore size, used as a porous substrate) of around 2.5 cm diameter, ensuring the precursor covers the entire area evenly. A thin metal mesh fabric (for example, -300 mesh and 50pm thick) may be placed in between to reinforce the membrane. The coated layer of precursor was then dried overnight to allow complete evaporation of the solvent (mainly water). Thereafter, two such pieces of the substrate were sandwiched with the coated face facing each other and the sandwiched green body was then subjected to compression (at ca. 0.8 T/cm ² or 78-79 MPa) in a hydraulic press for 10 minutes. Thereafter, the compressed membrane green body was subjected to moderate pyrolysis at 300 °C for 1 h under a weight (around 0.02-0.04 MPa), which converts the organic binding phase to a carbonaceous binding phase that strongly coheres the zeolite particles. The obtained membrane round sheet is free-standing that displays a thickness of about 100 pm (excluding the filtration paper). Analysis shows that the membrane contains about 0.25 g NaY and 0.08 g carbonaceous binding phase. The cellulose filtration paper was basically intact throughout the controlled pyrolysis at 300 °C because it undergoes decomposition at 390 °C. Example 2: For the separation of MB (10 ppm) in IPA

The precursor is prepared by blending NaY zeolite (l.6g, 0.4 g/cm ³ ) in a transparent solution of PEMA (Poly(ethyl)methacrylate) in Toluene (l.3g of PEMA in lOml of Toluene) ensuring the particles are well dispersed with no visible agglomeration. To this, Nafion (5 wt% in alcohol/water, 0.27g), commercial acrylic binder emulsion (0.2gm) and polycyanoacrylate (0.05gm) are added. The precursor is then let to disperse overnight until homogeneity is obtained. The precursor is then applied equally on one face of the cellulose filtration paper (d = 2.5 cm) and one face of stainless steel disc using a brush to obtain a uniform coating and allowed to dry for 1-2 hr ensuring all the solvent has evaporated. It is then processed as in Example 1. PS: The same fabrication steps are carried out for the membrane pertaining to organic solvent systems. Additionally, one of the substrates used is a stainless-steel disc, which enhances the mechanical strength of the membrane, allowing the membrane to withstand transmembrane pressures of more than about 30 bar.

Example 3: Assessing the nanofiltration efficacy of composite membrane

The composite membrane was evaluated in a cylindrical cell as shown in Fig. 9, where the membrane is sealed in a specific fashion using two rubber O-rings and porous stainless- steel discs, the latter with an average pore size of 2 pm. The feed is propelled into the cell using a liquid pump to maintain a trans -membrane pressure in the range of 0.5 to 2 bars. Three aqueous dye solutions, methylene blue (MB) (10 mg/L), Rhodamine B (RB) (5 mg/L) and methyl orange (MO) (10 mg/L), were used as feed to examine the membrane. These batch solutions were prepared by adding an accurate amount of the respective dye in Millipore water. The solution was stirred at around 250 rpm and the temperature was controlled at 25 °C. The separation for each of these dyes was carried out for 10 h on an average and samples were collected at definite intervals of time. Three control samples were evaluated as well: the first control sample was a carbonaceous composite membrane (CCM) prepared by following the same fabrication steps to make the composite membrane (see Example 1 and 2), although exclusive of NaY and GF, the second control sample was a packed-bed of NaY powder (NPB) (around 0.5 g), which was made by compressing the NaY powder between two porous stainless-steel discs in the same separation module; and the third control sample was made by crushing a few pieces of composite membrane together in a mortar pestle to obtain a fine powder, whose adsorption capability in a packed-bed was also assessed in a similar way as that for NPB. The water flux (L/lrm ² ) was calculated by the formula, J _w =V/At, where V is the volume of permeate in liter, A is the effective membrane area in m ² and t is the time in hr.

The effective membrane area was measured to be 7.85 xl0 ^_5 m ² . For each membrane and dye type, measurements were carried out three times where the deviation of uncertainty falls within 10%.

Example 4: Characterizations of the structure/properties of composite membrane and its NF performance

The concentration of dye in a permeate stream was determined by a double beam UV- Visible spectrometer (Shimadzu UV-1800), in the range from 200 to 800 nm. The permeate liquid was collected at regular intervals and the dye removal efficiency (rejection, R %) was calculated according to the Beer-Lambert’s law, R=( 1 C _P /C /)x 100 , where C _f (mg/L) is the concentration of feed (initial dye concentration) and C _p (mg/L) the concentration of the permeate.

The sandwich structure and the morphology of the NaY-carbonaceous bulk were determined using a field emission scanning electron microscopy (FESEM, JEOL, JSM- 6700F). The thermal degradation profiles of the acrylic resin as well as the blend of the ionic liquid and the acrylic resin were obtained from the thermogravimetric analysis (TGA, Q50 Shimadzu) using a ramp of 10 °C min ¹ up to 400 °C under a purge of N ₂ (50 mL min ' ). The organic functionality of CCM and composite membrane were identified using infrared spectroscopy (Bio-Rad Excalibur FTS-3500 FT-IR spectrometer). The zeta potential-pH dependencies of CCM, the pristine NaY zeolite and composite membrane were measured using a particle size analyzer (LS 13320 Laser Diffraction Particle Size Analyzer). The surface area and pore distribution of composite membrane and NaY were determined by the BET method on a physical adsorption instrument (ASAP 2020 Plus Physisorption, Micromeritics). Similarly, the mean pore radius and the pore size distribution of the composite membrane and CCM samples were measured by mercury porosimetry (Micrometries Auto Pore III, Norcross). The crystalline structure of NaY after soaking in HC1 (0.1 M, pH ~ 1) was examined using X-ray diffractometer (Bruker D8 Advance, Cu Ka radiation l = 1.54 A).

The rejection-time profiles and corresponding average permeation flux readings are reported in Fig 4-6 and Table 2.

Example 5: Performance of composite membrane with NaY zeolite and composite membrane with 4A zeolite

The composite membrane of NaY (Si:Al = 1.5 to 3, pore diameter: 7.4 A) with the composite membrane of 4A zeolite (Si:Al = 1, pore diameter: 4.2 A) as shown in Fig. 5b were fabricated as described above. The composition of these composite membranes are similar, except for the zeolite used; i.e. about 82-84 vol% zeolite and about 16-18 vol% acrylic polymer of which about 23.15 wt% is BMIC and about 16 vol% is glass fiber. The performance of these composite membranes were tested using an aqueous solution of MB as the model effluent. The 4A-composite membrane shows a rejection of around 8% lower than that of composite membrane at the beginning of the separation, however, the difference contracts to around 3% after 60 minutes. 4A-composite membrane offered a reasonable average rejection of 95% towards MB. The average permeance for 4A- composite membrane was found to be around 16 L/h-m ^2- bar over 100 mins at a constant pressure drop of 1.25 bar, which is around 3 times lower than that of composite membrane. The reduced permeance is believed to be due to the relatively smaller pore cavity of 4A zeolite. Thus, identifying a proper zeolite phase is significant as it influences the separation through its pore size and cage volume.

A composite membrane made of 4A zeolite instead of NaY was assessed using the same MB feed. Compared with NaY (7.4 A), 4A-zeolite has a smaller aperture diameter (3.8 A) but a higher cation exchange capacity by around 51%. The 4A-ZCCM showed a rejection of around 8% lower than that of ZCCM at the beginning of the separation, however, the difference contracts to around 3% after 60 minutes. Although 4A-ZCCM offers a reasonable average rejection of 95% to MB, it shows a large reduction in the permeance. The average permeance is around 16 L/lrm ² bar, which is 5 times lower than that of MB. Clearly, this reduction in permeance is caused due to the relatively smaller pore sizes and cavity of 4A zeolite. Identifying a proper zeolite bulk phase is therefore significant as it influences the separation performance through its pore size, cage volume and possibly the interfacial compatibility with the carbonaceous phase. Incorporating IL and GF as additives into the acrylic binder while fabricating the precursor aims to improve the permeation flux. The corresponding separation assessments show the noticeable enhancement of permeation flux upon the incorporation of IL and GF in the precursor (Table 2). As proposed, formation of IPN comprising ZCF and GF introduces additional voids enabling substantial increase in the solvent flux through the membrane at low pressure-drop of around 0.5-2 bars. Additionally, ZCF also intercepts pore flows in many locations throughout the membrane to assure nanofiltration.

Example 6: Properties of Example 1

The rejection is primarily a combination of size exclusion (steric hindrance) and electrostatic interactions, and physical adsorption does not play an important role. This has been verified by two control experiments; using pure carbonaceous membranes (excluding the NaY and glass fiber) and NaY powder packed bed to evaluate rejection of MB (10 ppm) in aqueous solution (Fig. 10). As can be seen from the figure, both the solid media own limited capability to sustain continuous MB rejection. The IR spectrum of the membrane exhibits its characteristic structure (Fig 8a). The strong peaks of the pendant carboxylic groups as well as phenolic groups (oxygenate) are responsible for the negative surface charge of the membrane. The zeta ^-potential of pristine NaY zeolite at neutral conditions is around -40.5 mV whereas the z-potential of composite membrane is only slightly more negative at the same pH (Fig. 11a). However, the z-potential gap between the two profiles expands quickly with an increase in pH value because the composite membrane becomes far more negative owing to the formation of more pendant anionic Ar- C0 ₂ and Ar-O moieties. This plays a role in retention of the cationic MB molecules through electrostatic attraction. NaY particles on the other hand offer substantial hindrance to the passage of the probe solute. MB being a larger molecule is unable to access the pore cavity of the nanoparticles and thus, is retained. The solute is thus retained on the membrane surface of feed side as well as along the pore walls on account of electrostatic attraction with the membrane, which leads to substantial membrane pore blocking. This in turn leads to further solute exclusion sustaining higher rejections over longer periods. Further analysis of the surface morphologies of the NaY grains in composite membrane and in the pristine form exhibits a carbonaceous covering on NaY grains, which as proposed above, originates from the extrusion of the polymer phase. This carbonaceous covering has a major impact on the rejection of organic solutes because it provides the first layer of screening against the solute. Furthermore, the BET analysis of composite membrane and pristine NaY powder supports the above-mentioned morphology. The amalgamation of NaY and carbonaceous brings about a drastic loss in the BET surface area of NaY from 599 (NaY) to 2 (composite membrane) m ² /g. This significant loss of surface area originates from the compression of the precursor paste. The pressure applied at ~79 MPa drives the acrylic polymer to undergo an extensive flow because of its low Tg (-10 °C) inside the packed NaY grains. Such polymer flow would not only fill interstices and crevices in NaY grains but also cover the surface of zeolites grains, as showed by the t-plot external surface area of NaY (17 m ² /g). It is important to note that the BET average pore width and volume of the pristine NaY are 23 A and 0.34 cm ³ /g, respectively. This data takes into account the impacts of micro- to macro-pores present in the interstices and crevices of the NaY grains. After pyrolysis of the compressed sandwich, the carbonaceous joints and stains formed in the structural defects and on the external surface of NaY grains adversely affect the N ₂ adsorption because of significant shortage of micro- and meso- pores in the carbonaceous phase. On the other hand, composite membrane shows an apparently larger average BET pore size of around 4653 A when compared to NaY, reflecting the occurrence of zeolite-carbonaceous frames (ZCF) that surrounds submicron voids and has a rather dense frame, i.e. very low surface area. In addition to BET measurements, the pore-size distribution profile of composite membrane was also determined by Hg-porosimetry for pores larger than macro-pores (> 0.05 pm). It was observed that the pore size close to 0.5 pm contributes to the main specific pore volume. This outcome is consistent with the void sizes in ZCF. Moreover, the pore-size distribution curve of composite membrane also extends beyond 1 micron that can be attributed to the presence of glass fibers (GF), which, despite being present in a smaller volume fraction (3 vol. %), is able to constitute a network that interpenetrates with the ZCF. Both the networks contain voids, which may merge or separate, and are bonded by carbonaceous at various locations within the membrane matrix as shown in Fig. 7. This resulting interpenetrating network (IPN) enhances the permeation flux, without compromising MB rejection (Fig. 6a). The contribution of carbonaceous in permeation as well as rejection is a function of its pore size and surface charge. The porosimetric analysis of membrane without zeolite shows a trace of pore volume corresponding to the pore size of around 200 nm. The evaluation of membrane without zeolite for the separation of MB from water in the following section will support the above estimation.

ZCF provides two passages for water; the regular sub-nano pores of NaY grains, both wrapped and unwrapped by carbonaceous, and the random sub-micron pores of carbonaceous. The NaY cage has a smaller aperture size as compared to the probe dye molecules, which is further narrowed by the carbonaceous spread. Hence, these zeolite cages allow the passage of only water molecules. Although, sub-micron pores are present in the carbonaceous phase of ZCF, the permeation of dye molecules through these pores is restricted by Coulombic interaction between the probe solutes and the ZCF. Thus, the carbonaceous phase offers the first shielding layer as discussed above. Another factor in the context of rejection is the optimal sub-micron porosity in the carbonaceous phase. The extent of pyrolysis affects sub-micron porosity and pore sizes of the carbonaceous phase and in turn affects dye rejection. A balance is required among the magnitude of surface charge in contact with liquid, the volumetric flow rate, and the size of solvated solute. A given amount of surface charge can screen the dyes only when the volumetric flow rate falls in a certain range.

Example 7 : Properties of Example 2

The rejection mechanism can be considered to be a combined solute-solvent-membrane interaction phenomenon. However, solvent-membrane affinity majorly impacts the overall membrane performance. Nafion is a sulfonated perfluoro polymer, which has excellent thermal stability and distinct amphiphilicity. The structure starts to break down only after 320 °C, which is beyond the temperature at which the membrane is thermally treated and thus, the backbone is more or less kept intact. Nafion is able to partially invade and infiltrate the surface of carbonaceous, thus making it hydrophobic in nature due to the outward orientation of the perfluoro-segments on carbonaceous. The carbonaceous surface doped by perfluorocarbon segments would therefore possess the three sites: the embedded perfluorocarbon chain site, the polyaromatic hydrocarbons (PAH) site, and the pendant polar group site (e.g. carboxylic and hydroxyl). This increase in surface hydrophobicity differentiates the interaction of the carbonaceous phase with IPA from the interaction with MB. The latter interaction is weak due to larger interaction cross-section and weak dispersion force. The average NF rejection of MB through the membrane is around 96% over 15 hours Fig.6b). Finally, the hydrophobic modification promotes solvent-membrane interaction vs. solute-membrane interaction. This promotes higher IPA recovery as well as MB rejection. The average IPA permeance through the membrane over 15 hours is around 8.22 L/hrm ² bar. As aforementioned, the magnitude of electrostatic interactions is largely challenged in the presence of IPA (as the solvent) as MB does not bear net charge in IPA. Hence, charge effects are no longer significant in solute rejection.

Example 8: Performance with different small molecules

Physical adsorption of MB and electrostatic attraction are the primary driving forces for the separation in both the above control systems. However, both the solid media own limited separation capability to sustain continuous MB rejection. It turns out that zeolite alone in packed bed (NPB) shows just a transient MB separation capability, whereas CCM exhibits a much weaker rejection capacity than composite membrane. These outcomes justify the essential of the integration of NaY and carbonaceous phase in submicron scale. The unique aspect of the carbonaceous phase in composite membrane versus in CCM can be perceived through the IR spectra of these, which show rather different carbon skeletons. CCM comprises higher content of aliphatic and smaller aromatic rings, as labeled in its IR spectrum. This supports the theory that the organic phase underwent deeper pyrolysis when present in a thorough amalgamation with NaY than when it existed alone. Consequently, the short-lived separation performance of CCM stems from the lack of essential organic functionalities, i.e. the carboxylic groups on PAHs (based on the relative peak intensities of both), and the eminently extended carbonaceous phase to sustain the negative surface charge and structural stability. The NF efficacy of a blank sample by stacking two pieces of the pyrolyzed cellulose filtration paper shows nil separation capability. To clarify if ZCF adsorption capability dominates MB rejection, separation of MB (10 ppm) was evaluated using a packed-bed of composite membrane powder. The rejection exhibits a drop of around 28 % by the end of 6 hours whereas composite membrane can maintain high rejection values until 10 h. The separation mechanism of MB by composite membrane thus includes both adsorption due to Coulombic attraction and resistance to pore diffusion. To further verify this conclusion, methyl orange (MO) solution was used instead of MB and found nil rejection through the composite membrane powder packed- bed. Since MO bears negative sulfonate at neutral pH, it encounters Coulombic repulsion, which fails to facilitate rejection through the packed-bed where the interstices possess larger widths as compared to MO. On the contrary, composite membrane demonstrates MO percolation.

MB ionizes into a cationic moiety and CT anion in aqueous solutions at neutral conditions. Composite membrane bears a negative z-potential at this pH value and thus arrests the cationic dye through electrostatic association. It is important to note that the carbonaceous binding phase, possessing obviously smaller space occupancy in ZCF, is extensively spread out over the surface as well as in the seams of NaY grains. Such a highly spread-out carbonaceous phase is drastically more effective to block MB molecules as it could be envisioned to possess higher PAH moieties on the surface than those present in CCM. Thus, besides Coulombic interactions, there is a higher z-affinity between MB molecules and the carbonaceous, which further supplements the rejection. This rejection mechanism results in deposition of MB molecules on various ZCF locations that deters the flow of solvent as the separation progresses. In other words, the filtration process in ZCF involves advection and diffusion of the permeate stream inside the carbonaceous phase. Moreover, the permeance-dwindling profile is an outcome of using a one-end filtration cell set-up and the absence of a retentate stream results in a higher accumulation of the solute on the membrane feed side. The permeance starts off at a reasonable value of 168 at lh but drops to 91 in the next 2h.

Table 3 illustrates the average rejection and permeance of three dyes tested. Table 3: Average rejection and permeance of the three dyes over 10 hours at neutral pH and the corresponding maximum UV-Vis absorption wavelengths

Feed (ppm) JV-Vis Average Permeance Average rejection

Example 9: Effect of pH on dye rejection

The point of zero charge (Pzc) of composite membrane is identified to be around 1.6 as ZCF carries zero surface charge at pH = 1.6 according to the z-rH profile of ZCF. For aqueous separation systems where H ⁺ /OH are mobile charge carriers, the membrane is negatively charged at pH > Pzc and is positively charged at pH < Pzc. As for dye molecules, the effect of pH on their rejection is a tad bit more complex owing to the variation of the charges of both solute and membrane with pH. Hence, the separation of solute proceeds on account of either electrostatic repulsion (if both the membrane and solute carry the same type of charge) or attraction (if both carry opposite charges) depending on the pH of the system. The separation performance of composite membrane to remove three probe dyes from water at pH = 1 and 7, respectively was studied. Composite membrane demonstrates excellent MB separation performance at the neutral pH, which relies on a combination of 3 factors: the electrostatic association of opposite charges between the carbonaceous pores and MB molecules, size exclusion by NaY cage, and p-effects along the pores. However, composite membrane unveils negligible rejection towards MB at pH = 1 that is below the pKa (= 3.8) of MB. Under this acidic condition, the ionization of MB is restricted and hence MB exists as a close ion-pair. Since the pH is very close to Pzc, the membrane carries almost nil net charges. Consequently, the weak van der Waals forces and p-effects are not strong enough to halt MB molecules and allow them to slip over the carbonaceous barrier and appear in the permeate stream. As for the structural stability of NaY in the acidic feed, the XRD diffraction verifies the acidic refractory of NaY at pH =1. Similarly, it may be worthy to note that the spent ZCCM after lOhrs of NF for the MB solution presents the same XDR pattern as the above two. Although being cationic in nature, rhodamine B (RB) possesses a zwitterionic structure at pH > pKa (3.4) by losing both H ⁺ and Cl ions simultaneously. Our study verifies that RB could be separated by composite membrane at both pH = 1 and 7. Composite membrane exhibits a stable rejection towards RB at pH = 7 with an average rejection of 95% over a period of 10 h. The effect of pH on rejection is fundamentally realized owing to the change in charge distribution around the RB molecule. Hence, it is proposed that RB interacts with the negatively charged ZCF through its positively charged end to form the first entrapping layer. Such a molecular alignment is then replicated with the progression of separation, resulting in buildup of solute at the membrane surface. The effect of such a layer-by-layer buildup is more severe in case of RB because of its bulky and unsymmetrical shape (15 x 9.8 x 4.3 A) which impedes the solvent permeation flux. This is supported by a decreasing trend of permeance over the entire assessment period, averaging to value of around 24 L/h-m ² bar.

Alternatively, at pH = 1, RB possesses a positive core due to protonation of the tertiary amine and the ether but this positive charge density is largely shielded by the surrounding of counter ions CT, which is partially from the dissociation of the quaternary ammonium group and partially from hydrochloric acid used to adjust the pH. The membrane on the other hand possesses a positively charged external (pzc > pH), which exerts a drag force on the entity comprising the protonated RB core and the attached counter-ion cloud in a direction opposite to the pore flow, resulting in high rejection rates of 94% over a 2h assessment period. An average permeance of 40 over 2h is observed at pH=l.

MO being an anionic dye (pKa =3.1) displays an opposite interaction pattern with ZCCM when compared to MB at neutral pH conditions. Besides, there are no other organic groups sterically hindering the sulfonate group of MO to interact with the negatively charged carbonaceous surface. Although, the basic amine group of MO is to be slightly protonated at neutral pH to become the cationic site, its anionic trait is nevertheless dominant. Therefore, pore flows in the carbonaceous phase with diameters larger than the linear and symmetric MO (with the length of 10.5 A) would favor a higher permeance over rejection. Moreover, the molecules near the pore surface are effectively excluded but clearly there is a gradual drop of such interfacial interaction while moving to the bulk of flow. To overcome the unmatched dimensions between pores and the dye molecules, narrowing down the average pore width was attempted. The duration of the hydraulic compression of the sandwiched membrane green body was prolonged to about 20 min from 10 min. The resulting composite membrane manifests a higher rejection of around 92% over the 10 h assessment period. The permeance experienced a continuous drop from 36 to 9 in the first 2 h, which can be interpreted by the possibility of MO molecules accumulating at certain torturous throats owing to favorable interfacial associations. A steady value of permeance is attained after 4h of operation.

A high rejection towards MO (average of around 95.5%) was also achieved at pH =1. At this pH, both the tertiary amine and azo groups of MO are protonated while the sulfonic acid group does not undergo ionization. Thus, MO bears a positive moiety at one end, which is surrounded by a cloud of counter-ions (e.g. Cl ions, from HC1 used to adjust pH), and the sulfonic acid group on the other end of molecule. The molecule experiences a similar effect as RB in acidic medium; surface charge drag with the counter-ion cloud resulting in high rejection rates. An average permeance of 37 L/hmr bar over 120 min has been obtained. For the operation under acidic medium, there was no need to extend the duration for hydraulic compression.

Example 10: Fabrication of composite membrane on a curved surface

The composite membrane can be applied to a curved surface, such as a surface of a stainless steel-316 (SS-316) tube. Briefly, this can be achieved through forming a slurry of the composite comprising a water-borne emulsion and kaolin, applying the composite as a coating on the surface, drying and isostatic pressing the coating, drying the coating layer, which is then subjected to isostatic compression under typically 67-140 MPa for a given duration (e.g. 30-40 min) and then subjected to partially pyrolysising under the anaerobic condition (Figure 13). Figure 15 shows pictographs of a) stainless steel-316 tube (GKN sinter metals ^® ); b) and c) that show the two coatings (membrane) whose thickness is ~ 120 mm.

The evaluation results of these composite membranes are as shown in Figure 16-20. These composite membrane (on curved SS-316 tube) demonstrates the possibility of scaling up the present invention. The composite membranes also do not need a dual surfactant system to modify its pore channels for the separation of 30 ppm dye (MB) aqueous solution. This tubular composite membrane has a different carbonaceous- zeolite(clay) interface because of the use of a different compression approach.