METHODS OF MAKING AND USING SAME

Reference to Related Applications

[0001] This application claims priority to, and the benefit of, United States provisional patent application No. 62/347418 filed 8 June 2016, the entirety of which is hereby incorporated by reference herein.

Technical Field

[0002] Some embodiments of the present invention relate to carbon nanotube (CNT) membranes having specifically designed hydrophilic or hydrophobic surface properties, including superhydrophilic and/or superhydrophobic surface properties. Some

embodiments of the present invention relate to CNT membranes having Janus surface properties. Some embodiments of the present invention relate to methods of making or using CNT membranes having specifically designed hydrophilic or hydrophobic surface properties, including Janus surface properties.

Background

[0003] Carbon nanotubes (CNT's) have shown promising applications in catalyst support, supercapacitors, sensors, and separation materials.' ^1"51 This is due to the three-dimensional (3D) network structures and outstanding mechanical, thermal, and electrical properties of CNT's. Recently, the application of CNT membranes as a water desalination and as an oil/water separation material has attracted extensive attention because of their low density, high porosity, and hydrophobic properties.' ^6"91 However, the surface wetting properties of CNT membranes themselves may not match a desired application.

[0004] Janus materials are materials whose surfaces have two or more distinct physical properties. Janus materials have attracted tremendous attention since they were first prepared. Due to their remarkable diversity in surface structural properties, these constructs have numerous potential applications in electronics, biomedicine and biological sciences as sensors, optical probes, surfactants and imaging agents. ^{1 " J}

[0005] Among a variety of available Janus materials, two-dimensional (2D) Janus nanosheets or membranes deserve much broader attention because of their highly anisotropic shape as well as their asymmetric physical and chemical properties - inorganic Janus nanosheets can serve as solid surfactants to stabilize interfacial emulsion droplets. ^119,201 Apart from their remarkable stabilization effects, the anisotropic wettability of Janus membranes makes them potential candidates for applications in oil/water separation.

[0006] Recently, one research group fabricated a Janus CNT composite membrane through the deposition of hydrophobic and hydrophilic polymers on to the CNT membrane surface.' ²¹¹ However, these Janus CNT composite membranes require careful deposition of the CNTs onto a supporting structure in order to provide an ordered CNT surface suitable for downstream modification by self-initiated photografting and photopolymerization (SI PGP). The resultant membrane structure is thus not robust, and is used together with a secondary supporting structure (i.e. as part of a composite membrane structure) to provide sufficient strength for use of the membrane. Additionally, the polymerization process used to create this membrane results in a surface modification that can be removed under certain conditions, for example upon exposure to organic solvents. Further, the membrane is expensive to produce, and must be produced through a batch process, making it unsuitable for scale-up.

[0007] There remains a need for CNT membranes, including Janus CNT membranes, having desirable properties, and for methods of making such membranes, including Janus CNT membranes, that are scalable and cost-effective.

[0008] The foregoing examples of the related art and limitations related thereto are intended to be illustrative and not exclusive. Other limitations of the related art will become apparent to those of skill in the art upon a reading of the specification and a study of the drawings.

Summary

[0009] The following embodiments and aspects thereof are described and illustrated in conjunction with systems, tools and methods which are meant to be exemplary and illustrative, not limiting in scope. In various embodiments, one or more of the above- described problems have been reduced or eliminated, while other embodiments are directed to other improvements.

[0010] One aspect of the invention provides a multi-walled carbon nanotube (MCNT) membrane comprising at least one side that has been modified by carbon dioxide (C0 ₂ ) plasma etching. Another aspect of the invention provides a MCNT membrane wherein the at least one side has been further modified by the application of a silane coating subsequent to carbon dioxide (C0 ₂ ) plasma etching. In some embodiments, the opposite side of the membrane is optionally also treated by carbon dioxide (C0 ₂ ) plasma etching. In some embodiments, both sides of the membrane are treated by carbon dioxide (C0 ₂ ) plasma etching. In some such embodiments, both sides of the membrane are further modified by the application of a silane coating.

[0011] In some aspects, the native carbon nanotube surface is hydrophobic, the at least one side of the membrane that has been modified by carbon dioxide (C0 ₂ ) plasma etching is superhydrophilic, and the at least one side of the membrane that has been further modified by the application of a silane coating subsequent to carbon dioxide (C0 ₂ ) plasma etching is superhydrophobic.

[0012] In some aspects, a membrane having two superhydrophilic sides is used to remove water from a water-in-oil solution. In some aspects, a membrane having two

superhydrophobic sides is used to remove oil from an oil-in-water solution, including for example to clean up a spill of oil or other hydrophobic liquid in the natural environment.

[0013] In some embodiments, a membrane produced in accordance with one of the foregoing aspects is used to filter a solution containing both oil and water, including, for example, an oil-in-water or a water-in-oil emulsion.

[0014] In addition to the exemplary aspects and embodiments described above, further aspects and embodiments will become apparent by reference to the drawings and by study of the following detailed descriptions.

Brief Description of the Drawings

[0015] Exemplary embodiments are illustrated in referenced figures of the drawings. It is intended that the embodiments and figures disclosed herein are to be considered illustrative rather than restrictive.

[0016] FIG. 1 shows schematically an example strategy for obtaining multiwalled CNT (MCNT) membranes having different surface properties according to some example embodiments.

[0017] FIGs. 2A and 2B show scanning electron microscopy (SEM) images of the MCNT membranes used in some embodiments.

[0018] FIG. 3A shows relative Raman spectra vs. C0 ₂ plasma etcher time (min) for MCNT membranes being subjected to C0 ₂ plasma oxidation for a plurality of different time periods. FIG. 3B shows Raman peak Pi and P ₂ intensity ratio (i.e. the intensity ratio of I _D /IG) VS. plasma etcher time (min). FIG. 3C shows conductivity of the C0 ₂ plasma etched CNT samples after different time periods. FIG. 3D shows digital images of the surface of the CNT membranes before (left) and after C0 ₂ plasma etching (right).

[0019] FIG. 4A shows X-ray photoelectron spectroscopy (XPS) results for MCNT

membranes, including an XPS survey for MCNT (solid line), MCNT after C0 ₂ plasma etcher (dash dot line) and MCNT after silane coat (dashed line); FIG. 4B shows XPS results for oxygen after C0 ₂ plasma treatment; and FIG. 4C shows XPS results for oxygen after silane coating.

[0020] FIG. 5A shows high resolution XPS spectra for carbon atoms for an example C0 ₂ plasma etched MCNT membrane sample. FIG. 5B shows high resolution XPS spectra for a C0 ₂ plasma etched MCNT membrane sample subsequent to silane coating.

[0021] FIG. 6 shows an XPS survey for three samples: MCNT (solid line), MCNT after C0 ₂ plasma etching (dashed line) and MCNT after C0 ₂ plasma etching and subsequent silane coating (dotted line).

[0022] FIG. 7 shows XPS for silicon in a sample after silane coating of the C0 ₂ plasma etched surface.

[0023] FIG. 8 shows MCNT surface wetting with water in air and the contact angle (CA) measured for the MCNT surface (first panel, top left), the C0 ₂ plasma treated MCNT surface (second and third panels moving clockwise) and silane coated MCNT surface (fourth panel moving clockwise). [0024] FIG. 9 shows the percentage increase in weight of various example MCNT membrane constructs after being dipped in water for 10 seconds.

[0025] FIGs. 10A-10C show photographs and dynamic light scattering (DLS) results for a small amount of oil-in-water emulsion before (top panel) and after (bottom panel) the filtration through an exemplary MCNT membrane construct (toluene in FIG. 10A, chloroform in FIG. 10B, hexane in FIG. 10C). FIGs. 10D-10F show photographs and DLS for a small amount of water-in-oil emulsion before (top panel) and after (bottom panel) filtration through an exemplary MCNT membrane construct (hexane in FIG. 10D, toluene in FIG. 10E, and chloroform in FIG. 10F).

[0026] FIG. 1 1 shows how the C0 ₂ plasma surface modified superhydrophilic MCNT membrane (No. 4) can be used to absorb a water/methyl blue mixture from an oil solution.

[0027] FIG. 12 shows how the superhydrophobic silane surface modified MCNT membrane (No. 5) absorbs oil from hexane/oil red mixture suspension on the surface of water.

[0028] FIGs. 13A and 13B show ¹ H NMR of oil/water absorption by the original MCNT membrane without any surface modification (FIG. 13A) and silane surface modified superhydrophobic MCNT membrane (No. 5) under the same conditions (FIG. 13B).

[0029] FIG. 14 shows ¹ H NMR of the two water signals produced when the unmodified MCNT membrane absorbs toluene from a water/toluene mixture solution (bottom line) and after addition of one drop of D ₂ 0 (upper line).

[0030] FIG. 15A shows thermogravimetric analysis of an example surface-modified CNT membrane according to one example embodiment under air oxidation; FIG. 15B shows thermogravimetric analysis of the same embodiment under nitrogen.

[0031] FIG. 16 shows a scanning electron microscope (SEM) image of an example embodiment of a surface-modified CNT membrane after 24 hours of treatment with 1 1.5 M HCI.

Description

[0032] Throughout the following description specific details are set forth in order to provide a more thorough understanding to persons skilled in the art. However, well known elements may not have been shown or described in detail to avoid unnecessarily obscuring the disclosure. Accordingly, the description and drawings are to be regarded in an illustrative, rather than a restrictive, sense.

[0033] As used herein, the term "superhydrophilic" means that water dropped onto the superhydrophilic surface forms no contact angle (i.e. has a contact angle of almost 0 degrees). As used herein, the term "superhydrophobic" means that the superhydrophobic surface is extremely difficult to wet, and the contact angle of a water droplet placed on the surface exceeds 150 degrees).

[0034] As used herein, the term "entry side" of a membrane refers to the side of a membrane contacted by a solution to be filtered. The term "exit side" of a membrane refers to the side of the membrane through which the solution to be filtered leaves the membrane, i.e. the side of the membrane that is contacted by the filtrate.

[0035] The inventors have now found that robust CNT membranes made from multiwall CNTs (MCNTs) can be modified to be superhydrophobic, superhydrophilic, or to have Janus properties, by using appropriate surface modification techniques. The inventors have found that the produced superhydrophobic membranes are able to selectively absorb a wide range of organic molecules or solvents, or hydrophilic molecules or solvents. The inventors have also found that the produced Janus membrane possesses a desirable wettability profile, with a switchable transport performance that can be controlled to effectively separate both surfactant-stabilized oil-in-water or water-in-oil emulsions. These types of robust MCNT membranes, composed of only thermal and conductive carbon nanotubes, have a great many potential applications for oil/water separation.

[0036] In one example embodiment, the surface modification technique used to render a surface of the MCNT membrane superhydrophilic is C0 ₂ plasma oxidation' ²²¹ . In some embodiments, the degree of modification of the surface of the MCNT membrane that can be achieved by C0 ₂ plasma oxidation is greater (i.e. yields a more consistent surface) than the degree of modification that can be achieved simply by depositing polymers on the surface of a MCNT membrane construct. In some embodiments, the use of C0 ₂ plasma oxidation allows for a lower quality (and therefore cheaper) MCNT starting material than, for example, polymerization techniques.

[0037] In some embodiments, the C0 ₂ plasma oxidation is carried out using a plasma etcher set at a power of around 60 Watts, a pressure of about 500 mTorr, and a C0 ₂ flow of about 135 seem. In some embodiments, the C0 ₂ plasma oxidation of the MCNT membrane is carried out for approximately 15 minutes. Those skilled in the art would be able to modify and optimize the parameters used for C0 ₂ plasma oxidation, and the stated parameters are illustrative and instructive, not limiting, in nature.

[0038] In one example embodiment, the surface modification technique used to render a surface of the MCNT membrane superhydrophobic is the application of a silane coat. In one example embodiment, the surface modification technique used to render a surface of the MCNT membrane superhydrophobic is the application of a fluorine coat or a fluorine polymer coat. In some embodiments, the surface properties of the MCNT membranes can be made switchable by incorporating stimuli-switchable polymers onto the surface of the membrane, such as those described in Patent Cooperation Treaty publication No. WO 2016/029307, which is incorporated by reference herein in its entirety.

[0039] In some embodiments, the application of a silane or other hydrophobic coating to a surface of the MCNT membrane is carried out after C0 ₂ plasma oxidation of that surface of the MCNT membrane, to take advantage of reactive hydroxyl and/or carbonyl groups generated on the MCNT membrane surface by C0 ₂ plasma oxidation. In some

embodiments, the covalent linkage of such groups to the MCNT surface after

functionalization using C0 ₂ plasma oxidation provides a much more stable linkage than could be achieved, for example using polymerization techniques. For example, in some embodiments, the silane coat applied as described herein may be stable to a temperature of on the order of 300°C to 400°C.

[0040] In some embodiments, the surface silane coatings are achieved using a vapor deposition technique. In some embodiments, polydimethylsiloxane (PDMS) condensed liquid is placed in a container, and is placed in an autoclave with the C0 ₂ plasma etched MCNT membrane, and the two materials are heated together for a suitable period of time, e.g. overnight. The thermal degradation of PDMS through heterolytic cleavage of the Si-O bonds leads to a mixture of volatile, low molecular weight products that form a conformal layer on the MCNT surface of the substrate through Si-O covalent bonding.

[0041] With reference to FIG. 1 , an example procedure used to prepare the

superhydrophobic, superhydrophilic, and Janus MCNT membranes is described. Five different membranes were prepared (membrane Nos. 1 through 5) having different combinations of superhydrophilic, hydrophobic and superhydrophobic surfaces.

[0042] In one example embodiment, the starting material for the MCNT membrane is a robust CNT mat assembled from multiwall CNTs (MCNTs) (hereafter referred to as a robust MCNT membrane). In one example embodiment, the robust MCNT mat that is used to provide the membrane is a non-woven MCNT mat obtained from Tortech Nano Fibers of Israel. In one example embodiment, the MCNTs are produced through a large scale and/or continuous production process.

[0043] The MCNT membrane that is used as a starting material is hydrophobic on both surfaces. In some embodiments, the MCNT membrane that is used can be provided as a large sheet or roll, and the surface modification of that MCNT membrane as described herein can be carried out in a continuous process that is suitable for efficient and cost- effective scale-up.

[0044] In order to create a superhydrophilic surface on the MCNT membrane, in one embodiment, C0 ₂ plasma etching is used. The C0 ₂ plasma etching can be selectively carried out on either surface of the MCNT membrane, or on both surfaces of the MCNT membrane. In one example embodiment, a MCNT membrane 20 is subjected to C0 ₂ plasma etching on a first surface (referred to for convenience as its upper surface, although it will be appreciated that the membrane could have any orientation in three dimensional space), as indicated by arrow 22 to produce Janus membrane No. 1 (24) having a first

(upper) surface that is superhydrophilic and a second (bottom) surface that is hydrophobic.

[0045] In some embodiments Janus membrane No. 1 (24) is coated with silane at arrow 26. In one example embodiment, the membrane is coated with silane using vapor deposition of polydimethylsiloxane (PDMS). The inventors have found that the silane reacts only with a surface of the MCNT membrane that has been subject to C0 ₂ plasma etching. Thus, as a result of this process, Janus membrane No. 2 (28) is produced that has a first (upper) surface that is superhydrophobic, and a second (bottom) surface that is hydrophobic.

[0046] In some embodiments, Janus membrane No. 2 (28) is further subjected to C0 ₂ plasma etching on its second (bottom) surface at arrow 30 to produce Janus membrane No. 3 (32) that has a first (upper) surface that is superhydrophobic, and a second (bottom) surface that is superhydrophilic. [0047] In other embodiments, MCNT membrane 20 is subjected to C0 ₂ plasma etching on both its first (upper) and second (bottom) surfaces at arrow 34 to provide membrane No. 4 (36) that has first (upper) and second (bottom) surfaces that are superhydrophilic.

[0048] In other embodiments, membrane No. 4 (36) is coated with silane at arrow 38, to yield membrane No. 5 (40) that has first (upper) and second (bottom) surfaces that are superhydrophobic.

[0049] To summarize, following this strategy, five different membranes having the following different properties can be prepared:

• Janus membrane No. 1 - one hydrophobic surface and one superhydrophilic

surface.

• Janus membrane No. 2 - one hydrophobic surface and one superhydrophobic

surface.

• Janus membrane No. 3 - one superhydrophilic and one superhydrophobic surface.

• Membrane No. 4 - two superhydrophilic surfaces.

· Membrane No. 5 - two superhydrophobic surfaces.

[0050] In one example embodiment, a representative example of membrane No. 5, i.e. a membrane having two superhydrophobic sides, is used to recover a hydrophobic substance, for example oil, gasoline, or an organic solvent, from a hydrophilic substance such as water. In one such example embodiment, a portion of membrane No. 5 is placed on top of an aqueous solution containing the hydrophobic substance to be recovered, and is used to selectively absorb the hydrophobic substance to be recovered. For example, in the case of an oil spill on a body of water, portions of membrane No. 5 could be applied to the surface of the body of water, to selectively absorb oil from the water, thereby recovering the spilled oil.

[0051] In another example embodiment, a representative example of membrane No. 4, i.e. a membrane having two superhydrophilic sides, is used to recover water or another polar substance from a hydrophobic substance, for example, oil, gasoline or an organic solvent. In one such example embodiment, a portion of membrane No. 4 is placed on top of a solution of a hydrophobic substance containing water or other polar substance, and is used to selectively absorb the water or other polar substance.

[0052] In alternative embodiments, where the density of the substance to be recovered is higher than that of the solution from which the substance is to be separated, a portion of the MCNT membrane can be passed through the bottom portion of the solution to selectively absorb the substance to be recovered.

[0053] In another example embodiment, a Janus membrane is used to filter a feed solution. In some embodiments, the Janus membrane that is used to filter a feed solution has a thickness of at least 20 μηη.

[0054] In some embodiments, the feed solution is an oil-in-water mixture. In some such embodiments, the entry side of the membrane is selected to have properties that are similar to the major component of the feed solution. In other words, where the feed solution is primarily polar, the entry side of the membrane is selected to be superhydrophilic. In some such embodiments, the exit side of the membrane is selected to also be superhydrophilic.

[0055] In some embodiments, the feed solution is a water-in-oil mixture. In some such embodiments in which the feed solution is primarily non-polar, the entry side of the membrane is selected to be hydrophobic. In some such embodiments in which the feed solution is primarily non-polar, the entry side of the membrane is selected to be

superhydrophobic. In some such embodiments, the exit side of the membrane is selected to be hydrophobic or superhydrophobic.

[0056] In some embodiments, the properties of the MCNT membranes can be adjusted by adjusting the porosity of the MCNT membranes through densification, for example as described in PCT publication No. WO 2008/132467, which is incorporated by reference herein. In some embodiments, adjusting the porosity of the MCNT membranes allows for adjustment of the transport properties of the membrane. In some embodiments, the MCNT membranes can be prepared to have a desired density through appropriate densification during fabrication of the MCNT membrane, for example by adjusting the deposition speed of fibers during the production process.

[0057] In other embodiments, the density of the MCNT membranes can be adjusted after they have been fabricated by immersing the MCNT membranes in an inorganic solvent, for example, acetone or methanol. Subsequent to immersion, the MCNT membranes can be held under vacuum, optionally while being heated, for example in an oven, and this will cause the layers of the MCNT membrane to move closer to one another, thereby increasing the density of the MCNT membrane. The degree of densification experienced under such conditions can be varied by varying the conditions used to evaporate the inorganic solvent (for example, time and temperature at which the MCNT membranes are held in an oven during evaporation of the inorganic solvent).

[0058] Those skilled in the art will be able to determine empirically what density of MCNT membrane is optimum for a given intended function of the membrane. In some

embodiments, the porosity of the MCNT membrane is varied by densification to provide efficient separation based in the particle sizes of various components in an emulsion to be separated.

[0059] In some embodiments, the surface properties of the membranes may be adjusted having regard to the specific composition of a solution to be filtered through the membrane (for example, having regard to the presence of salts or other ions in the feed solution). For example, in some embodiments, the MCNT membranes are used to both separate an oil-in- water or water-in-oil emulsion, and to perform desalination (i.e. so that an undesired salt or ions present in the feed solution is rejected by the membrane, to yield a filtrate that does not contain the undesired salt or ions).

[0060] In some embodiments, membranes have potential application in the construction of selective gas or humidity sensors. The interaction of the Janus surface of some of the membranes in accordance with some embodiments with different gas molecules produces an interaction strength that is different, thereby generating different electron currents.

Measuring the current allows for identification of the type of gas that is interacting with the surface of the material. In some embodiments, the current generated by the interaction of gas molecules with the surface of the membrane produces an electric current that can be measured and/or detected to provide a gas sensor.

Examples

[0061] Specific embodiments of the invention are described with reference to the following examples, which are intended to be illustrative and not limiting in nature. Example 1.0 - Materials and Equipment

[0062] The CNT membranes used in this example are like a paper sheet, roughly 20 μηη thickness, assembled with multiwall CNTs (diameter, about 10-30 nm; length, about 10-30 μηη), supplied as a non-woven MCNT mat by Tortech Nanofibers (Israel). The non-woven MCNT mats supplied by Tortech Nanofibers are made using a gas-phase catalytic reaction creating a dense cloud of very long carbon nanotubes. This generates a continuous network of CNTs which is drawn out of the reactor and spun around a drum to create a non-woven mat.

[0063] The selected parameters of power (60 Watts), pressure (500 mTorr) and C0 ₂ flow (135 seem) were fixed to set up the C0 ₂ plasma etcher conditions. In alternative embodiments, other parameters of power, pressure and C0 ₂ flow could be used. These preselected conditions are believed to provide a good plasma electron density. A 3x3 cm ² cubic sample of each MCNT membrane was placed on the surface of a silicon wafer round plate and loaded into the Tegal Plasma Etcher and etched for 15 rmins.' ²²¹

[0064] The surface silane coatings of the substrates were achieved using a vapor deposition technique. In a typical experiment, a polydimethylsiloxane (PDMS) condensed liquid was placed in a glass container and the C0 ₂ plasma etched MCNT membrane was placed in a second glass container, and both containers were placed in a sealed Teflon™ autoclave and heated at 240°C overnight. In some embodiments, "overnight" is

approximately ten hours. The thermal degradation of PDMS through heterolytic cleavage of the Si-O bonds leads to a mixture of volatile, low molecular weight products that formed a conformal layer on the MCNT surface of the substrate through Si-O covalent bonding.' ²⁸¹

[0065] For oil-in-water emulsions, PEG-20 sorbitan monolaurate (Tween™ 20)

(hydrophilic/lipophilic balance (HLB) =16.7, an emulsifier of the oil-in-water type) was selected as the emulsifier. 1.1 , 1.0 and 0.7g of Tween™ 20 for toluene in water, chloroform in water, and hexane in water emulsions respectively were added into 120.0 ml of water, then stirred until completely mixed, after which 4.0 ml of oil was added. The final mixture was stirred for 3 hours. For water in oil emulsions, sorbitan monooleate (Span™ 80) (HLB = 4.3, an emulsifier of the water in oil type) was selected as the emulsifier. 0.8, 1.2 and 1.0 g of Span 80 for water in toluene, water in chloroform and water in hexane was added into 1 14.0 ml of oil. The mixed solution was completely combined, then 1.0 ml of water was also added. The mixture was stirred for 3 hours. All the prepared milk-like solution emulsions were stable for at least 48 hours, and no demulsification was observed. The separation experiments were conducted using a vacuum line driven filtration device with a 1.6 cm diameter round filtration membrane.

[0066] Instruments used for characterizations: Scanning electronic microscope: Hitachi, S- 4800 Field Emission SEM (FE-SEM); Nicolet Omega XR Raman Microscope (Raman spectroscopy); Rame-hart Model 400 Goniometer (contact angle); Kratos AXIS 165 XPS Spectrometer; Conductivity measurement: Princeton Applied Research,

Potentiostat Galvanostat Model 263A.

Example 2.0 - Optimization of CO? Plasma Etching Parameters

[0067] The C0 ₂ plasma etcher conditions were preset as follows: the selected parameters of power (60 Watts), pressure (500 mTorr) and C0 ₂ flow (135 seem) were fixed. A 3x3 cm ² cubic sample of each MCNT membrane was placed on the surface of a silicon wafer round plate and loaded into the Tegal Plasma Etcher and etched.' ²²¹ The time period for carrying out C0 ₂ plasma oxidation was optimized using these fixed conditions. In alternative embodiments, one skilled in the art could vary and optimize other parameters of the C0 ₂ plasma etching conditions.

[0068] The MCNT membrane was subjected to surface plasma etching under these conditions, and was monitored by Raman spectroscopy to determine the minimum etcher time. Scanning electron microscopy (SEM) images of the used MCNT membrane starting material are shown in FIGs. 2A and 2B. These images show that the MCNT membrane has a large aspect ratio with a fibre diameter in the range of 20-50 nm and a length on the order of tens of micrometers.

[0069] The correlation plot of relative Raman intensity vs. etcher time, as shown in FIG. 3A, is typical of the MCNT surface oxidation level, which is reflected in the relative change in intensity of the Raman bands (D-band and G-band) (ID/IG) as shown in FIG. 3A, in which the D-band is assigned to the presence of disorder in the graphitic material while G-band corresponds to the tangential vibration of the carbon atoms. Meanwhile, when there are few carbon impurities generated in the sample, the G-/D-band intensity ratio can be used to assess the MCNT wall defects, or oxidation disorder levels. ^{122,23,2 1} The inventors found that the intensity of the D-band (-1350 cm ^"1 ) reaches its maximum relative to that of G-band (-1590 cm ^"1 ) after being in the plasma etcher 15 minutes at the tested conditions, and that after additional time in the plasma etcher, there was almost no further obvious change. After this time, there is almost the same intensity observed for both the D-band and the G- band.

[0070] FIG. 3B shows the correlation plot of the intensity ratio of ID I _G VS. plasma etcher time. This indicated 15 minutes was a sufficient period of time to oxidize the MCNT membrane surface under the C0 ₂ plasma beam. A Gaussian peak at 1620 cm ^"1 (D'-band, visible in FIG. 3A) became separated as the plasma etcher continued to act upon the MCNT membrane surface.

[0071] CNT is a type of conductive material through conjugated π electron flow. The conductivity of the six membranes etched for varying time periods were measured and the results are given in FIG. 3C.

[0072] The membrane without plasma etching is very conductive, relative to the etched membranes. Without being bound by theory, this is believed to indicate the plasma surface- oxidation affected the conjugated structure due to surface functionalization through a σ- bond covalent bonding process, which disrupts the sp ² hybridization of the carbon atoms.' ^25,261 The disruption of the extended sp ² hybridization typically decreases the conductance of the carbon nanotubes. Interestingly, when varying the etcher time from 5 minutes to 10 minutes to 15 minutes, the conductivity of the membranes slightly decreased along with the increase of the plasma exposure time. After 20, 25 to 30 minutes, it was slightly increased and even more increased with 30 minute etcher sample of the MCNT membrane. Based on this data, 15 minutes was selected as an optimal period using the selected conditions to generate the thickest layer of the desired functionalized carbon on the MCNT membrane surface. Even with that amount of time the surface layer may become less thick due to the long plasma beam evaporating or slightly removing a defective layer on the MCNT membrane surface. This also suggested that 15 minutes was an optimal time for C0 ₂ plasma surface etching under the tested conditions, and thereafter, all the samples were exposed the surface etcher for 15 minutes at this optimized condition.

[0073] The surface change caused by C0 ₂ plasma etching can be seen from the images shown in FIG. 3D, in which the image on the left shows the surface prior to C0 ₂ plasma etching, and the image on the right shows the surface after C0 ₂ plasma etching. Prior to etching, the surface of the material is slightly shiny. After C0 ₂ plasma etching, the surface appears more black.

[0074] Functionalization of the MCNT membrane surface with the oxygen group is believed to occur via the C0 ₂ plasma beam but this is difficult to detect through FTIR and Raman spectroscopy. However, the oxygen carbon functional groups were observed using X-ray photoelectron spectroscopy (XPS), as shown in FIGs. 4A, 4B and 4C. Compared to the un- etched sample, the plasma-etched sample clearly demonstrates a large oxygen atomic peak from the MCNT membrane surface, indicating the membrane surface carbon element is being oxidized (FIG. 4A). Based on the peak fitting from an oxygen high resolution spectrum, the XPS spectrum was mainly composed of three component peaks, which corresponded to the following three functional groups, -C=0, -C-OH and -COOH (FIG. 4B). This demonstrated that the membrane surface formed different oxygen functional groups.' ²⁶¹ High resolution XPS of carbon is also in agreement with the oxygen data, providing additional evidence and confirmation (FIGs. 5A and 5B, showing the MCNT membrane after C0 ₂ plasma etching and after silane coating, respectively).

[0075] As described with reference to FIG. 1 , silane modification to the plasma etched MCNT membrane surface was carried out with a polymer of polydimethylsiloxane (PDMS) vapour deposition coat. The silane-coated MCNT membrane surface's elemental composition was characterized by XPS (see FIG. 4A, upper tracing (dashed line), and FIG. 6). In FIG. 6, the lower tracing (solid) shows the surface composition of the MCNT membrane starting material. A large peak is visible for carbon, but very little oxygen is present on the surface. After C0 ₂ plasma etching (middle tracing, dashed line), a large peak is visible for oxygen, indicating that a significant number of oxygen functional groups are present on the etched MCNT membrane surface. Subsequent to silane coating (upper tracing, dotted line), additional peaks for silicon are visible in the spectra. In addition, two peaks corresponding to Si _2s and Si _2p respectively are clearly present in the survey XPS spectrum, which showed that the silane was successfully linked to the MCNT membrane surface through covalent chemical bonds (C-O-Si) (FIG. 7). High resolution XPS of oxygen, FIG. 4C, also changed when compared to the sample surface without silane modification. All of the component peaks show small shifts to low bonding energy after silane coating, which confirms a difference in surface composition.

[0076] The MCNT membrane surface was originally hydrophobic with a water wetting contact angle of 108° (first panel of FIG. 8), but it became superhydrophilic after being plasma etched. As shown on the right-hand side of FIG. 8, the water disappeared within 1 second after a water droplet touched the C0 ₂ plasma etched membrane surface. It was surprising that it eventually changed to superhydrophobic after the silane coating was applied to the plasma etched MCNT surface, with a water wetting contact angle of 163° as shown in the fourth panel of FIG. 8.

[0077] The silane coating could not be adhered to the MCNT surface without first treating the surface with the C0 ₂ plasma etcher. Without being bound by theory, this observation is believed to indicate that the silane coat application was most likely successfully completed through the linkage of oxygen functional groups to the MCNT membrane surface.

Example 3.0 - Examination of Surface Wetting Properties of Membrane Constructs

[0078] The surface wetting properties of the different MCNT surfaces were measured under the same conditions for the purpose of comparison. Almost identical sizes of four MCNT membranes were dipped in water for 10 seconds, and the weight increase percentages of the four membranes were determined. Results are shown in FIG. 9. These results demonstrate that applying the C0 ₂ -plasma etching to the MCNT membrane surface dramatically increases the surface wetting and the membrane absorption of water by almost six times as compared with the unmodified MCNT membrane surface (3425% and 513% respectively). In addition, the silane coated MCNT membranes do not absorb any water, and experience no weight change due to the evenly superhydrophobic surface properties of the modified surface that does not absorb water.

[0079] For comparison, the surface wetting for the C0 ₂ plasma etched MCNT surface and the C0 ₂ plasma/silane coat treated MCNT surface were observed visually. For the C0 ₂ plasma etched MCNT surface, water would almost not stay on the surface (i.e. was rapidly absorbed by the surface), while subsequent silane coating of the surface resulted in a situation where water would not be absorbed into the surface. Thus, the superhydrophilic C0 ₂ plasma etched surface is superwetting with water, as confirmed by a significant increase in weight after contact with water (FIG. 9). However, subsequent to silane coating to provide a superhydrophobic surface on the MCNT membrane, the membrane becomes non-wetting with water, so that no water is absorbed by the MCNT membrane.

Example 4.0 - Filtration of Emulsions Through Different Membrane Constructs

[0080] As shown in FIG. 1 , five different examples of MCNT membranes having different surface properties were generated through a design strategy using a combination of the unmodified MCNT surface, C0 ₂ plasma etching, and silane coating: three Janus membranes (No. 1 , No. 2 and No. 3), a fourth membrane with a superhydrophilic coating on both sides (No. 4), and a fifth membrane with a superhydrophobic coating on both sides (No. 5) (FIG. 1) were produced.

[0081] The characteristics of these membranes were assessed under the same conditions to show how the surface modification of the MCNT membranes could be designed to improve the membrane performance for a given application. Flux of two thicknesses (10 and 20 μηη) of MCNT membranes using only water or only oil (hexane, chloroform, or toluene) was measured. The data are listed in Table 1. The flux is an important parameter when used to evaluate or compare the function of different membranes.

Table 1. Flux parameters for tested MCNT membrane constructs.

8 One side plasma 10 hexane 3 min 3 s 40.98 Native MCNT (Janus No. 1 )

9 Two sides plasma 10 hexane 2 min 52 s 43.60 Plasma

(Membrane No. 4)

10 No modification 10 chloroform 3 min 2 s 41.21 Native MCNT

11 One side plasma 10 chloroform 3 min 30 s 35.71 Native MCNT (Janus No. 1 )

12 Two sides plasma 10 chloroform 3 min 30 s 35.71 Plasma

(Membrane No. 4)

13 No modification 20 hexane 2 min 15 s 55.56 Native MCNT

14 One side plasma 20 hexane 4 min 19 s 28.96 Native MCNT (Janus No. 1 )

15 One side plasma, 20 hexane 55 s 136.36 Silane

one side silane

(Janus No. 3)

16 No modification 20 chloroform 3 min 41 s 33.94 Native MCNT

17 One side plasma 20 chloroform 5 min 30 s 22.73 Native MCNT (Janus No. 1 )

18 One side plasma, 20 chloroform 3 min 2 s 41.21 Silane

one side silane

(Janus No. 3)

19 No modification 20 toluene 3 min 53 s 32.19 Native MCNT

20 One side plasma 20 toluene 4 min 52 s 25.68 Native MCNT (Janus No. 1 )

21 One side plasma, 20 toluene 2 min 25 s 51.72 Silane

one side silane

(Janus No. 3)

[0082] Some conclusions that can be drawn from the above results are as follows. First, the surface property of the side of the membrane the liquid enters is a significant factor that will affect the liquid flux. Meanwhile, varying the surface property on the exit side of the membrane produces much less of a change in flux. For example, water/10 μηη (test 1-3), if the entry side of the membrane is changed to superhydrophilic from hydrophobic, the flux is doubled, while with two sides (both entry and exit side) changed to superhydrophilic, flux only increased by 16.8%. The second flux can be quite different with different pass liquids, for instance, for an organic liquid/membrane thickness 20 μηη, the flux of hexane is the highest, and toluene is in the middle, slightly faster than chloroform (test 13-21). Third, the change of surface property on the exit side that is slightly affected by the flux depends on the match of the liquid property to that surface property, such as if the exit side is superhydrophilic, which matches water, a slight increase in the flux is observed (test 2-3).

[0083] Meanwhile, for water if the exit side is superhydrophobic, this does not match the property of water as the pass liquid, and therefore decreases the water pass flux (test 5-6). This is understandable because a superhydrophobic surface will not allow water to pass through and the water flux will decrease. Thus, the filtration membrane should be selected based on both the surface property of both sides of the membrane and the pass liquid. For example, for filtration of water, the No. 4 membrane (both sides superhydrophilic) worked best under the tested conditions. In contrast, for organic solvents, Janus membrane No. 3 (one side superhydrophilic, one side superhydrophobic) was the better choice under the tested conditions.

Example 5.0 - Filtration of Surfactant-Stabilized Emulsions

[0084] The most difficult task for a filtration membrane to accomplish is to break down the microcells of a surfactant stabilized emulsion, i.e. a water-in-oil or an oil-in-water emulsion solution. Three different organic solvents: hexane, toluene and chloroform, with water and six emulsion solutions were prepared and tested with the Janus No. 3 membrane (one side superhydrophilic, one side superhydrophobic). For oil-in-water emulsions, the entry side of the membrane used was the superhydrophilic surface treated with C0 ₂ plasma oxidation. For water-in-oil emulsions, the hydrophobic native MCNT membrane surface or the superhydrophobic silane coated surface of the membrane was the entry side of the membrane.

[0085] The separation tests using the 10 μηη thick sample of Janus No. 3 membrane to separate an oil/water emulsion solution were not successful - the solutions were still turbid after multiple filtrations. However separation tests with the 20 μηη thick sample of Janus No. 3 membrane were all successful. The superhydrophilic surface as the entry side could be used to separate the small amount of the oil in the oil-in-water emulsion. Meanwhile the reversal to use the superhydrophobic side as the entry side allowed the Janus No. 3 membrane to separate a small amount of water from a water-in-oil emulsion. The filtration solution was very clean and DLS (dynamic light scattering) was not able to locate the microcell emulsion in the filtered solution (FIGS. 10A-10F).

[0086] The characterization results for the 20 μηη thick sample of Janus No. 3 membrane are given in FIGs. 10A-10F (for oil-in-water emulsions of toluene in FIG. 10A, chloroform in FIG. 10B, hexane in FIG. 10C; and for water-in-oil emulsions of hexane in FIG. 10D, toluene in FIG. 10E, and chloroform in FIG. 10F). The upper panels of each figure show the measured particle sizes for the prepared emulsions, as well as a photograph of the solution in which the particles present in the emulsion are visible. The lower panels of each figure show the measured particle sizes for the filtered emulsion solutions, as well as a photograph showing the clarity of the filtered solution.

Example 6.0 - Evaluation of Rejection Ability of Membrane

[0087] The rejection of the membrane is the main parameter used to assess its separation performance, and the oil rejection of the Janus No. 3 MCNT membrane was tested with an n-hexadecane in water emulsion solution. After filtration using the superhydrophilic side of the membrane as the entry side, the remaining oil liquid was purified and more than 95.0 % of the n-hexadecane was recovered by the membrane filtration process.

Example 7.0 - Recovery of Water from Water-in-Oil Mixture

[0088] The two sides of the C0 ₂ plasma etched membrane (No. 4) are superhydrophilic, providing a superwetting with water membrane (see the 3425% weight increase when dipped in water as shown in FIG. 9). The No. 4 membrane can be used to absorb water from a water-in-oil mixture (FIG. 1 1). In this example, an aqueous solution of water containing methyl blue dye was applied on top of an oil solution, in this case hexane. A portion of No. 4 membrane was applied on top of the solution, and rapidly absorbed the water and methyl blue dye, leaving a clear oil solution (far right panel).

Example 8.0 - Recovery of Oil from Oil-in-Water Mixture

[0089] The silane coated membrane (Membrane No. 5) is superhydrophobic on both sides and has no wetting to water (FIG. 9), which demonstrates it only absorbs oil and can be used to recover oil from an oil in water mixture. A red colored oil was suspended on top of water and then the MCNT membrane (No. 5) was laid down. It contacted the suspended oil liquid for a few seconds and then the membrane was lifted out. The suspension of oil was fully absorbed by the MCNT membrane, as shown by the clear solution in the right-most panel of FIG. 12. Example 9.0 - Evaluation of Solvent Selectivity of Membranes

[0090] As a control experiment, ¹ H NMR was used to compared the solvent selectivity of the original and silane modified MCNT membranes, with water contact angles of 108° and 163° respectively (FIG. 8). The original MCNT membrane was found to absorb both toluene and water, whereas the surface modified superhydrophobic MCNT membrane absorbed only toluene, confirming the important role of the superwetting surface in selective absorption (FIGs. 13A and 13B). The inventors found two water signals, which shifted from 2.84 ppm, mimicking the chemical shift of water in acetone (FIG. 14). ^[27] These two signals are caused by presence of both HOD and H ₂ 0 in solution when the exchange rate between HOD and H ₂ 0 is slow on the NMR timescale. ^[27] Without being bound by theory, the chemical shift to a higher ppm is probably due to the mixture being composed of acetone and toluene, rather than pure acetone. To ensure the water signals were accurate and to ascertain the cause of the chemical shift, one additional drop of D ₂ 0 was placed in the solution, which resulted in an additional ¹ H NMR spectrum being recorded. The results observed were that the ratio of the two water signal peaks changed and the chemical shift moved to a higher ppm. This indicates it is, in fact, water signals and the chemical shift was caused by the water exchange in the mixture of the two solvents (FIG. 14).

[0091] The inventors assigned the largest intensity signal as H ₂ 0 and the small intensity signal as HOD. After adding of one drop of D ₂ 0 to increase the molar ratio of HOD inside the solution, the relative intensity of the H ₂ 0 signal increases. This confirms that the signal is water. The chemical shift of water in pure acetone is reported to be 2.84 ppm, but the inventors observed it at around 3.1 ppm in the sample of water in mixture of toluene and acetone, which is slightly shifted in the higher ppm direction. Further chemical shift change to higher frequency after addition of D ₂ 0 indicates this discrepancy is caused by two species of H ₂ 0 and HOD exchange in the solution. Still, two distinct separation peaks can be identified, showing the exchange rate is slow on the NMR timescale.

Example 10.0 - Stability of Surface-Modified CNT Membranes

[0092] The inventors evaluated the stability of the surface-modified CNT membranes having Janus properties. Thermogravimetric analysis (TGA) shows that the membranes are thermally stable up to 490 °C (FIGs. 15A and 15B). Treatment with 1 1.5 M HCI for 24 hours showed that the membranes are also chemically stable (FIG. 16).

Example 1 1.0 - Regeneration of Surface-Modified CNT Membrane

[0093] The successful regeneration of a membrane having both sides coated with silane (i.e. both sides superhydrophobic, membrane No. 5) after recovery of oil from an oil-in-water solution was demonstrated. An emulsion was prepared using 0.2g of n-hexadecane in 5g of water. The No. 5 membrane with two superhydrophobic sides was dipped five times in the solution for 30 seconds each time, and was washed with 5 mL hexane (hydrophobic solvent) after each trial. The results are shown in Table 2. After each trial, the membrane was washed with hexane. The wash solution from all of the trials was combined together, and hexane was evaporated to measure the amount of oil (i.e. n-hexadecane) recovered. 0.19 g (95%) of oil was recovered. The membrane was dried, and was recovered as having its original weight of 0.9g. The dried membrane retained its original performance in absorbing oil from an oil-in-water solution. This example demonstrates that the membrane has a high recovery capacity, and further can be regenerated for subsequent use.

Table 2. Results of Experiment to Regenerate Surface-Modified CNT Membrane.

[0094] The foregoing examples demonstrate that the utility of MCNT membranes can be extended through a facile surface modification. The examples demonstrate the construction of MCNT membranes with five different surface properties through surface C0 ₂ plasma treatment in combination with a surface silane coat. The membranes produced have application as filters, and in the selective absorption of oil and organic solvents from a water mixture. Some of these membranes can separate stable oil-water emulsions and can withstand high temperature and pressure conditions with excellent chemical tolerance, thereby providing an economical solution for environmental remediation such as oil spill clean ups, waste water treatment, and oil recovery. Additionally, the membranes have potential application in the construction of selective gas or humidity sensors with different surface properties.

[0095] While a number of exemplary aspects and embodiments have been discussed above, those of skill in the art will recognize certain modifications, permutations, additions and sub-combinations thereof. It is therefore intended that the following appended claims and claims hereafter introduced are interpreted to include all such modifications, permutations, additions and sub-combinations as are consistent with the broadest interpretation of the specification as a whole.

References

[0096] The following references are of interest with respect to the subject matter described herein. Each of the following references is incorporated by reference herein in its entirety.

[1] M. F. L. De Voider, S. H. Tawfick, R. H. Baughman, A. J. Hart, Science 2013, 339, 535.

[2] Q. Cao, J. A. Rogers, Adv. Mater. 2009, 21, 29.

[3] L. B. Hu, D. S. Hecht, G. Gruner, Chem. Rev. 2010, 110, 5790.

[4] J. M. Schnorr, T. M. Swager, Chem. Mater. 2011 , 23, 646.

[5] S. J. Gao, Z. Shi, W. B. Zhang, F. Zhang, J. Jin, ACS Nano, 2014, 8, 6344.

[6] H. Y. Yang, J. H. Zhao, S. F. Yu, L. P. Kin, O. Kostya, K. Rohit, Nat. Commu. 2013, 4, 2220 (DOI: 10.1038/ncomms3220).

[7] Z. Shi, X. J. Chen, X. W. Wang, I. Zhang, J. Jin, Adv. Fund. Mater. 2011 , 21, 4358.

[8] S. J. Gao, Z. Shi, W. B. Zhang, F. Zhang, J. Jin, ACS Nano, 2014, 8, 6344.

[9] Z. Shi, W. Zhang, F. Zhang, X. Liu, D. Wang, J. Jin, L. Jiang, Adv. Mater. 2013, 25, [10] J. Hu, S. X. Zhou, Y. Y. Sun, X. S. Fang, L. M. Wu, Chem. Soc. Rev. 2012, 41, 4356.

[1 1] S. Jiang, S. Granick, Janus Particles Synthesis, Self-Assembly, and Applications RSC: London, England, 2012.

[12] M. Lattuada, T. A. Hatton, Nano Today, 2011 , 6, 286.

[13] P. G. De Gennes, Angew. Chem., Int. Ed. 1992, 31, 842.

[14] F. X. Liang, C. L. Zhang, Z. Z. Yang, Adv. Mater. 2014, 26, 6944.

[15] A. Walther, M. Hoffmann, A. H. E. Muller, Angew. Chem. Int. Ed. 2008, 47, 71 1.

[16] F. V. Leonardo, Y. G. Tao, N. S. Zacharia, V. Kitaev, F. Galembeck, R. Kapral, G. A. Ozin, Small, 2010, 6(4): 565.

[17] S. H. Hu, X. H. Gao, J. Am. Chem. Soc. 2010, 132 (21): 7234.

[18] Y. J. Liu, F. X. Liang, Q. Wang, X. Z. Qu, Z. Z. Yang, Chem. Commun. 2015, 51, 3562.

[19] T. Nisisako, T. Torii, T. Takahashi, Y. Takizawa, Adv. Mater. 2006, 18 (9): 1 152.

[20] H. Yang, F. Liang, X. Wang, Y Chen, C. Zhang, Q. Wang, X. Qu, J. Li, D. Wu, Z. Z. Yang, Macromolecules, 2013, 46, 2754.

[21] J. C. Gu, P. Xiao, J. Chen, J. W. Zhang, Y. J. Huang, T. Chen, ACS Appl. Mater.

Interfaces, 2014, 6, 16204.

[22] D. J. Babu, S. Yadav, T. Heinlein, G. Cherkashinin, J. J. Schneider, J. Phys. Chem. C 2014, 118, 12028.

[23] S. Osswald, M. Havel, Y. Gogotsi, J. Raman Spectrosc. 2007, 38, 728.

[24] A. I. Lopez-Lorente, B. M. Simonet, M. Valcarcel, Analyst, 2014, 139, 290.

[25] N. Karousis, N. Tagmatarchis, Chem. Rev. 2010, 110, 5366.

[26] W. Xia, C. Jin, S. Kundu, M. Muhler, Carbon, 2009, 47(3), 919.

[27] H. E. Gottlieb, V. Kotyar, A. Nudelman, J. Org. Chem. 1997, 62, 7512.

[28] J. K. Yuan, X. G. Liu, O. Akbulut, J. Q. Hu, S. L. Suib, J. Kong, F. Stellacci, Nature Nanotechnology, 2008, 3, 332.

[29] WO 2008/132467

[30] WO 2009/153576

[31] WO 201 1/138595

[32] WO 2015/082936

[33] WO 2016/087857

[34] S.A. Abraham, G. Ma and CD. Montemagno, Adv. Mater. Interfaces, 2016, 3,

1600445.

24