rexresearch.com
Rahul NAIR, et al.
Graphene Oxide Filter
Graphene Oxide Filter
University of Manchester
Graphene's love affair with water
Graphene has proven itself as a wonder material with a vast range of unique properties. Among the least-known marvels of graphene is its strange love affair with water.
Graphene is hydrophobic - it repels water - but narrow capillaries made from graphene vigorously suck in water allowing its rapid permeation, if the water layer is only one atom thick - that is, as thin as graphene itself.
This bizarre property has attracted intense academic and industrial interest with intent to develop new water filtration and desalination technologies.
One-atom-wide graphene capillaries can now be made easily and cheaply by piling layers of graphene oxide - a derivative of graphene - on top of each other. The resulting multilayer stacks (laminates) have a structure similar to nacre (mother of pearl), which makes them also mechanically strong.
Two years ago, University of Manchester researchers discovered that thin membranes made from such laminates were impermeable to all gases and vapours, except for water. This means that even helium, the hardest gas to block off, cannot pass through the membranes whereas water vapour went through with no resistance.
Now the same team led by Dr Rahul Nair and Prof Andre Geim has tested how good the graphene membranes are as filters for liquid water. The results appear in the latest issue (Feb 14, 2014) of Science.
The researchers report that, if immersed in water, the laminates become slightly swollen but still allow ultrafast flow of not one but two monolayers of water.
Small salts with a size of less than nine Angstroms can flow along but larger ions or molecules are blocked. Ten Angstroms is equivalent to a billionth of a metre.
The graphene filters have an astonishingly accurate mesh that allows them to distinguish between atomic species that are only a few percent different in size.
On top of this ultraprecise separation, it is also ultrafast. Those ions that can go through do so with such a speed as if the graphene membranes were an ordinary coffee filter.
The latter effect is due to a property that the Manchester scientists call "ion sponging". Their graphene capillaries suck up small ions as powerful hoovers leading to internal concentrations that can be hundreds of times higher than in external salty solutions.
Dr Nair said: "The water filtration is as fast and as precise as one could possibly hope for such narrow capillaries. Now we want to control the graphene mesh size and reduce it below nine Angstroms to filter out even the smallest salts like in seawater. Our work shows that it is possible."
Dr Irina Grigorieva, a co-author of the study, added: "Our ultimate goal is to make a filter device that allows a glass of drinkable water made from seawater after a few minutes of hand pumping. We are not there yet but this is no longer science fiction".
This invention relates to uses of graphene oxide, and in particular graphene oxide on a porous support, and a membrane comprising these materials. This invention also relates to methods of dehydration, which include vapour phase separation and pervaporation. Pervaporation is a method of separating mixtures of liquids using a membrane. Pervaporation consists of two basic steps: permeation of the permeate through the membrane and evaporation of the permeate from the other side of the membrane. Pervaporation is a mild which can be used to separate components which would not survive the comparatively harsh conditions needed for distillation (high temp, and/or low pressure).
[0001] This invention relates to uses of graphene oxide, and in particular graphene oxide on a porous support, and a membrane comprising these materials. This invention also relates to methods of dehydration, which include vapour phase separation and pervaporation.
BACKGROUND
[0002] Separation of water from other substances is of great utility in a wide variety of contexts.
[0003] The separation of water vapour from other gases can be achieved using hydrophilic membranes, which allow water to pass through but not the gas which is being purified.
[0004] ‘Raw’ natural gas contains a number of impurities, including water vapour which must be removed before it can be used. Likewise, naturally occurring helium gas often contains water vapour which must be removed for many applications. Crude oil also frequently needs to be dehydrated as part of the refining process.
[0005] Water vapour can interfere with gas phase detector systems and pre-separation of water from the gaseous mixture can improve the sensitivity and longevity of such systems. The removal of water vapour from these and other gaseous systems may be advantageous and is an aim of the invention.
[0006] Many industries use the removal of water as a way of concentrating mixtures. Thus, the food industry will concentrate fruit juices, sugar, and milk (e.g. before cheese production) and many other industries (e.g. the galvanic industry) will concentrate their waste products in order to reduce the cost of transport and/or disposal.
[0007] Pervaporation is a method of separating mixtures of liquids using a membrane. Pervaporation consists of two basic steps: permeation of the permeate through the membrane and evaporation of the permeate from the other side of the membrane. Pervaporation is a mild process. Thus, it can be used to separate components which would not survive the comparatively harsh conditions needed for distillation (high temp, and/or low pressure). It is a low-energy alternative to distillation.
[0008] Pervaporation can be used to remove trace quantities of an impurity from a liquid. For example, hydrophilic membranes can be used to remove water from organic solvents, and can provide the solvent in a higher purity than conventional means. One particular use is to purify compounds which form azeotropic mixtures with water, such as ethanol and isopropanol.
[0009] Hydrophilic membranes can be used for the removal of water from fermentation broths or from condensation reactions such as esterification.
[0010] The currently most preferred hydrophillic membranes for pervaporation are zeolites such as those based on zeolite A. These materials have a number of drawbacks. As zeolites are polycrystalline, their formation into a stable membrane can be problematic. Zeolites dissolve in relatively weakly acidic conditions (below about pH 4) and this can limit the possible applications. It is also hard to find an effective support for such membranes.
[0011] Polymeric membranes are also widely used, but they are generally less-selective.
[0012] Graphene is believed to be impermeable to all gases and liquids. Submicrometer thick membranes made from graphene oxide are impermeable to most liquids, vapours and gases, including helium. However, an academic study has shown that, surprisingly, graphene oxide membranes which are effectively composed of graphene oxide having a thickness around 1 μm are permeable to water even though they are impermeable to helium. These graphene oxide sheets allow unimpeaded permeation of water (10<10 >times faster than He) (Nair et al. Science, 2012, 335, 442-444). The paper does not, however, disclose any workable membranes made from graphene oxide sheets nor does it disclose any practical applications of this material as a membrane.
[0013] The present invention aims to provide a means for dehydration (i.e. the separation of water or water vapour from a system) which is effective at a wider range of conditions than existing technologies. The system from which water is to be separated may be a mixture of gases or it may be a mixture of one or more liquids contaminated by water (e.g. a fuel system).
[0014] The present invention also aims to provide a means for dehydration which is as selective or more selective than existing technologies.
[0015] The present invention also aims to provide a means for dehydration which uses a material that is cheaper and/or easier to prepare in bulk than the materials of existing technologies.
[0016] The present invention also aims to provide a means for dehydration which is more efficient, in terms of yield or in terms of energy input than existing technologies.
[0017] The present invention also aims to provide a means for dehydration which is durable and has good longevity. This allows the material to be reused and/or recycled. This is particularly relevant for industrial uses.
[0018] The present invention satisfies some or all of the above aims.
[0019] We have found that sheets of graphene oxide, when combined with a porous support, can be used to selectively separate water or water vapour from fluid systems such as a gas, a liquid or mixtures of gases and liquids.
BRIEF SUMMARY OF THE DISCLOSURE
[0020] In a first aspect of the present invention is provided the use of a graphene oxide membrane for the dehydration of a product, wherein the graphene oxide membrane comprises a layer of graphene oxide on a porous support.
[0021] Dehydration comprises the separation of water from a product.
[0022] In an embodiment, the separation is by pervaporation.
[0023] In an alternative embodiment, the separation is by gas or vapour phase separation.
[0024] In an embodiment, the use is in a detector system. The species being detected may be water or it may be the dehydrated liquid or gas. In this case, it may be that, were it present, water would interfere in the detection of the dehydrated liquid or gas.
[0025] In an alternative embodiment, the use is in gas drying. In an alternative embodiment, the use is in gas humidification.
[0026] In an embodiment, the use is in a concentration step.
[0027] In an embodiment, the use is a continuous process.
[0028] In a second aspect of the present invention is provided a method of separating water from a product, the method comprising:
[0029] contacting the mixture of water and product with a first surface of a graphene oxide membrane, wherein the graphene oxide membrane comprises a layer of graphene oxide on a porous support;
[0030] removing the water from a second surface of the graphene oxide membrane.
The graphene oxide is provided on a porous support. This means that the graphene oxide is physically supported by the porous support, but does not impose any limitation on the position of the porous support relative to the position of the layer of graphene oxide. Thus, the porous support may be provided on one side of the layer of graphene oxide, this being the “top” or “bottom” side, or indeed there may be more than one porous support associated with the layer of graphene oxide, in which case the porous supports may be disposed on different sides of the layer of graphene oxide or they may be on the same side.
[0031] In an embodiment, the method of separating is a method of separating by pervaporation.
[0032] In an embodiment, the method of separating is a method of separating by gas or vapour phase separation.
[0033] In an embodiment, the method is continuous.
[0034] One effect of removing water from a product is that the product necessarily becomes more concentrated. This could be the purpose of the use or method of the invention.
[0035] Without wishing to be bound by theory, it is believed that the water passes through the layer of graphene oxide due to capillary action or other surface forces between individual water molecules and the graphene oxide. This unusual permeation property is due to the unique structure of GO membrane. The speed of water molecules inside graphene oxide membranes is high due to the low frictional graphitic structure of the membrane. It is believed that this mechanism is not available to other compounds. It is not necessary to pressurise the system i.e. the product which is to be dehydrated in order to effect dehydration as the water is able to pass selectively through the membrane. Separation may however be effected under the influence of pressure and/or the application of a vacuum. The water which is recovered from the “far” side of the membrane after the water has passed through the membrane and been released from the second surface of the membrane can be removed using a variety of methods or it can simply be left or allowed to evaporate. Active removal of the separated water is not essential to the functioning of the membrane in the dehydration process, i.e. in many cases no specific action is required.
[0036] In a third aspect of the present invention, there is provided, a composite semi-permeable membrane in the form of a laminar structure comprising at least a layer of graphene oxide and a layer of porous support.
[0037] The term “semi-permeable” refers to the fact that the composite graphene oxide membrane is impermeable to all gases and liquids but it can selectively allow permeation of water whilst being impermeable to gases, such as helium, and other vapours and liquids. Thus, in this case the membrane is exclusively pervious to water and water vapour. It also refers to the fact that composite membrane as a whole can, in some embodiments, allow selective passage of a limited number of other materials such as ethanol, ethanol and propanol, and similar small polar molecules but remains impermeable to other gases and liquids.
[0038] We have found that the nature of the porous support on to which the graphene oxide is deposited is important to the overall performance of the graphene oxide membrane composite structure and its ability to allow permeation. In other words, the porous support can modulate the ability of the overall graphene oxide membrane to allow selective permeation of liquid or vapour such as water. However, other small polar molecules may permeate through the membrane or may be prevented entirely by the membrane. In this respect the term “small polar molecules” specifically excludes water. The graphene oxide membrane according to the invention will always allow the permeation of water and the membrane materials thus govern the extent to which other small polar molecules such as C1-4 alcohols and the like will permeate or be excluded.
[0039] In an embodiment, the porous support is an inorganic material. Thus, the porous support (e.g. membrane) may comprise a ceramic. Preferably, the support is alumina, zeolite, or silica. In one embodiment, the support is alumina. Zeolite A can also be used. Ceramic pervaporation membranes have also been produced in which the active layer is amorphous titania or silica produced by a sol-gel process.
[0040] In an alternate embodiment, the support is a polymeric material. Preferably it is PTFE, PVDF or Cyclopore™ polycarbonate. In an embodiment, the porous support (e.g. membrane) may comprise a polymer. In an embodiment, the polymer may comprise a synthetic polymer. Historically, commercial pervaporatino membranes were generally based on poly(vinyl alcohol and more recently, polyimides have been used. These can be used in the invention. Alternatively, the polymer may comprise a natural polymer or modified natural polymer. Thus, the polymer may comprise a polymer based on cellulose.
[0041] In another embodiment, the porous support (e.g. membrane) may comprise a carbon monolith.
[0042] In an embodiment, the graphene oxide layer has a thickness of no more than a few tens of μm, and ideally is less than about 100 μm. Preferably, it has a thickness of 50 μm or less, more preferably of 10 μm or less, and yet more preferably is less 5 μm. In some cases it is preferably is less than about 1 μm thick. For most membrane applications, the active layer is made as thin as possible (less than 1 micrometre) to increase the flux.
[0043] In an embodiment, the porous support has a surface roughness which is the same as or smoother than as PTFE. Preferably, the surface roughness is less than that of PTFE i.e. it is smoother.
[0044] Preferably, the porous support is as smooth as Anodisc alumina. Normally, one might think of PTFE as smooth and alumina as rough. However, the reference to alumina here refers specifically to the very special case of “Anopore”™ alumina.
[0045] In an embodiment, the choice of the porous support is determined by the relative smoothness of its surface and the identity of the material which is intended to be allowed to permeate or to be stopped by the membrane structure.
[0046] We can also modify the permeability of GO and hence of the membrane as a whole by modifying not only surface roughness but also other parameters (eg. GO flake structure. For example, if flat GO is replaced by corrugated or crumpled GO it is possible to control the pore size. This will allow passage of other small molecules also through GO membranes. This provides a means to alter the selectivity of permeation of the GO coating and consequently also the membrane as a whole.
[0047] In an embodiment, the porous support layer has a thickness of no more than a few tens of μm, and ideally is less than about 100 μm. Preferably, it has a thickness of 50 μm or less, more preferably of 10 μm or less, and yet more preferably is less 5 μm. In some cases it may be less than about 1 μm thick though preferably it is more than about 1 μm.
[0048] For ceramic pervaporation membranes, the support may be rather thick (e.g., tubular structures of 7 mm internal diameter, 10 mm outside diameter, i.e. thickness 3 mm) in order to achieve sufficient mechanical strength. However, polymeric membranes are generally much thinner (e.g., for hydrophilic polyethersulfone membranes, the thicknesses are in the range 90-180 micrometres for some applications, whilst thicknesses in the range 110-150 micrometres are quoted for other applications. Also, a polymeric support may be “asymmetric”, with a large pore size substructure and smaller pores at one surface. Thin polymeric supports may also be further supported on, e.g., a fabric.
[0049] In an embodiment, the ratio of the thicknesses of the graphene oxide layer to porous support layer is in the range from 10:1 to 1:10. Usually, the thickness of the graphene oxide layer is less than the thickness of the porous support layer. The ratio is ideally from 1:1 to 1:10, and preferably is from 1:2 to 1:10. However, in some applications the ratios may be reversed, with the preference being for a thicker graphene oxide layer.
[0050] In an alternative embodiment, the porous support layer is considerably thicker i.e. from 10 to 1000 times thicker than the graphene oxide layer. In this embodiment, the above quoted ratios for the relative thicknesses of the graphene oxide layer and the porous support layer do not apply. This is because the graphene oxide layer must usually have a thickness of no more than 100 μm.
[0051] In an embodiment, the porous support is no more hydrophobic than PTFE, and preferably it is more hydrophilic than PTFE. In this respect, Anopore™ alumina is better as a support than a polymeric material such as PTFE is. The choice of porous support will thus depend on the hydrophilicity of the material used as the support in a graphene oxide membrane composite structure and the identity of the material which is intended to be allowed to permeate. The skilled person has a wide variety of more hydrophilic polymers than PTFE which are available for use in the present invention.
[0052] Hydrophilicity is a well understood term and the hydrophilicity of a material is easily determined by conventional methods. For example, hydrophilicity may be quantified in terms of, e.g., water contact angle or surface energy. For commercial materials such information is directly available from suppliers. Typical quoted values of water contact angle are: PTFE 123 deg; PVDF 111 deg; polycarbonate 70 deg., polyethersulfone 54 deg. The measurement of this angle can be performed using a Kruss drop analyser which is widely used the fields of both chemistry and materials.
[0053] The effects of smoothness and hydrophilicity can be additive in some cases, or they may act independently; this is substrate dependent. However, the above-described general principles still apply in terms of the separate, or combined, effects of increasing smoothness and increasing hydrophilicity on permeation. Without wishing to be bound by theory, we consider that the smoothness of the substrate is more important for GO coating than hydrophilicity is. This may be because graphene always tries to follow the structure of the substrate.
[0054] In some cases the effect of substrate hydrophobicity on GO coating quality can be reduced significantly in importance relative to the surface smoothness effects by using GO dispersions in other polar solvents (eg. methanol, ethanol, NMP, DMF, chloroform, etc)
[0055] We are thus able to produce a membrane whose permeation characteristics can be tailored to allow passage of water and/or to allow passage of or exclude the passage of other small polar solvents such as methanol, ethanol and propanol whilst remaining completely impermeable to other fluids such as gases and liquids. In this respect, the term “small polar molecules” specifically excludes water.
[0056] The graphene oxide membrane according to the invention will always allow the permeation of water and the membrane materials thus govern the extent to which other small polar molecules such as C1-4 alcohols and the like will permeate the membrane or be excluded by the membrane whilst allowing water to pass through.
[0057] In some embodiments, it is possible to produce a membrane which is able to separate molecules of identical weight such as different isomers of the same compounds. For example, we have found higher flux for 1-propanol than 2-propanol and thus it may be possible to use the membrane either to separate completely isomers such as these from one another, or to enrich a mixture of the isomers in one particular isomer. Thus for example, a linear isomer may be separated from a branched isomer.
[0058] In one embodiment, methanol and ethanol are small polar molecules which are of particular interest in terms of the ability to allow under certain circumstances selective transmission through the membrane. In an alternative embodiment, denial of passage through the membrane structure of small polar molecules such as these in their entirety is desirable.
[0059] We can also tailor the nature of the GO coating itself. This is a useful feature of the invention. Thus the permeability of the GO membrane can also be modified by modifying the GO coating itself to alter the spacing between the GO sheets. Permeation is purely dependent on the interlayer distance between graphene oxide sheets. We can tailor the interlayer distance by inserting (intercalating) GO with other atoms/molecules. If the interlayer distance is less than 0.7 nm (fully dried state) it completely blocks everything including water. In the wet conditions interlayer distance can be vary up to 1.3 nm depending on the hydration level.
[0060] In the fully wet conditions, molecular permeation through GO is depends on the solubility of that molecule in water and the size of the molecule. If the size of the molecule is less than the size of the capillary diameter then it permeatse with water; otherwise it will get physically blocked by the capillary.
[0061] Membranes of this type will have a number of uses in diverse applications in gas drying and liquid drying. The membranes of the third aspect can be used in the uses and methods described respectively in the first and second aspects of the invention above.
[0062] The following embodiments may also apply to any of the first second and third aspects of the invention referred to above. These individual embodiments may also be combined with one or more other embodiments in any combination.
[0063] The porous support may comprise a layer of porous membrane.
[0064] The product may comprise one or more compounds. In an embodiment, at least some of the product is in the form of a gas or a vapour. Thus, it may be that the product is substantially in the form of a gas or vapour. Or, it may be that the product is substantially in the form of a liquid. In an embodiment, at least one compound of the product is entirely in the form of a gas or a vapour. In an embodiment, the product is entirely in the form of a gas or a vapour.
[0065] In an embodiment, at least some of the water to be separated is in the form of water vapour. Thus, it may be that the water is substantially in the form of a gas or vapour. Or, it may be that the water is substantially in the form of a liquid. In an embodiment, the water is entirely in the form of a gas or a vapour.
[0066] In a specific embodiment, the product and the water are entirely in the form of a gas or a vapour.
[0067] In another specific embodiment, the product and the water are substantially in the form of a liquid.
[0068] In embodiments in which the layer of graphene oxide is supported on a layer of porous membrane, the graphene oxide membrane on a porous support is a composite material. The material will comprise one or more layers of graphene oxide and one or more layers of porous membrane in the form of a laminate or thin layer composite. It will have two surfaces, a first surface which is in contact with the product/water mixture and into which the water (and, in some embodiments, product) will enter the composite material and a second surface from which the water (and, in some embodiments, product) will exit the composite material.
[0069] Each surface may be the outer surface of a graphene oxide layer or a porous membrane layer.
[0070] In an embodiment, the first surface of the composite material is the outer surface of a graphene oxide layer and the second surface is the outer surface of a porous membrane layer. In an alternative embodiment, the first surface is the outer surface of a porous membrane layer and the second surface is the outer surface of a graphene oxide layer.
[0071] In some embodiments, the or each graphene oxide layer is sandwiched between two layers of porous membrane. Thus, in another embodiment, both surfaces of the composite material are the outer surfaces of a porous membrane layer.
[0072] In some alternative embodiments, the or each porous membrane layer is sandwiched between two layers of graphene oxide. Thus, in another embodiment, both surfaces of the composite material are the outer surfaces of graphene oxide layer.
[0073] In its simplest form, the graphene oxide membrane comprises only a layer of graphene oxide and a layer of a porous membrane. However, the laminate structure, when composed of three or more layers may be symmetrical in the case of alternative layers of graphene oxide and porous membrane as described above or it may have any arrangement of different layers of graphene oxide and porous membrane.
[0074] Throughout this specification, the term ‘layer’ when applied to graphene oxide, refers to a layer which may itself comprise any number of individual molecular layers of graphene oxide. Thus, the ‘layer’ referred to in this specification is not necessarily a single molecular layer.
[0075] Thus, in an embodiment, the or each graphene oxide layer is between 100 nm and 10 μm thick. In an embodiment, the composite material comprises a single graphene oxide layer. In another embodiment, the composite material comprises a single graphene oxide layer between 100 nm and 10 μm thick. This thickness of material represents the optimum balance between porosity and selectivity.
[0076] The porous support should be porous enough not to interfere with water transport but have small enough pores that graphene oxide platelets cannot enter the pores. Thus, the porous support must be water permeable. In an embodiment, the pore size must be less than 1 μm. In an embodiment, the support has a uniform pore-structure. Examples of porous membranes with a uniform pore structure are electrochemically manufactured alumina membranes (e.g. those with the trade names: Anopore™, Anodisc™).
[0077] In an embodiment, the composite material may comprise two or more layers of porous membrane. In this embodiment, the two or more layers may comprise different materials. ‘Different materials’ are meant to include materials with the same chemical composition but differing forms or differing structures, e.g. differing pore sizes. Of course, ‘different materials’ also includes materials with different chemical compositions. The two or more layers may comprise two layers with different pore sizes
[0078] In an alternative embodiment, the two or more layers may comprise the same material.
[0079] The porous support may be in the form of a flat sheet, a hollow fibre, a tube or a monolith structure. The graphene oxide layer may thus be presented as a coating on one or both of the surfaces of porous support.
[0080] In embodiments of the invention, the product is a mixture of chemical entities which contains, or is suspected to contain, at least one target compound.
[0081] In a more specific embodiment, the product is a gaseous mixture of chemical entities which contains, or is suspected to contain, at least one target compound.
[0082] In an embodiment, the at least one target compound is harmful. In a further embodiment, the at least one target compound is harmful to the environment. In an alternative embodiment, the at least one target compound is harmful to animals, and in particular humans. The at least one target compound may be toxic.
[0083] In an embodiment, the product is natural gas. In an alternative embodiment, the product is helium. In yet another alternative, the product is crude oil.
[0084] In an embodiment, the product is a fermentation broth or has been extracted from a fermentation broth.
[0085] In an embodiment, the fermentation broth has a pH less than about 4.
[0086] The product may comprise an alcohol, e.g. a C1-C15 alcohol and more particularly a C1-C6 alcohol. Thus, the product may comprise ethanol. The product may also comprise methanol. The product may also comprise butanol.
[0087] The product may comprise one or more alkanes, e.g. one or more C1-C15 alkanes. The product may comprise one or more C1-C10 alkanes.
[0088] The product may also comprise one or more organic molecules selected from acetone; acetaldehyde; acetic acid; formic acid; propanoic acid.
[0089] In an alternative embodiment, the product is a complex mixture of compounds. In such cases the purpose of the use or method of the invention may simply be the concentration of the product. Thus, the product may be selected from: a food stuff (e.g. fruit juice, sugar, milk) or an industrial product or waste stream, (e.g. those arising from the galvanic industry).
[0090] In an embodiment, a vacuum or sweep gas is applied to the second surface of the graphene oxide membrane to remove the water.
[0091] In an embodiment, the water is allowed to evaporate and no specific action is taken to remove the water.
BRIEF DESCRIPTION OF THE DRAWINGS
[0092] Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:
[0093] FIG. 1 shows He-leak measurements for a freestanding submicrometer-thick GO membrane and a reference PET film (normalized per square centimeter).
[0094] FIG. 2 shows weight loss for a container sealed with a GO film.
[0095] FIG. 3 shows weight loss for a container sealed with a 500 nm thick GO on Anopore membrane (20 mm diameter).
[0096] FIG. 4 shows the propanol leak rate for different support membranes with and without GO coating.
[0097] FIG. 5 shows a schematic view of the graphene oxide and graphite oxide layered structures.
DETAILED DESCRIPTION
[0098] This invention relates to the separation of water from other chemical entities. In the context of this application, ‘separation’ can be understood to mean that the proportion of the product mixture which is water is lower after the separation than it was before the separation. In some cases, water will be substantially or even entirely removed from the product. In other cases, water will be partially removed from the product. The water which has been separated will, in some cases, be substantially or entirely pure (i.e. free of the product). In other cases, the water will contain some product.
[0099] The invention relates to the separation of water from one or more products. The term ‘product’ may mean any chemical species. In an embodiment, a product may be any gas, e.g. an elemental gas. A product may be an organic molecule. Thus, a product may be any species which is a gas at standard temperatures and pressures, e.g. H2, N2, O2, methane, ethane, ethene, ethyne, ethylene oxide, propane, butane, He, Ar, Ne, CO2, CO H2S, SO2, NOx, etc. A product may be a liquid at standard temperatures and pressure such as pentane, hexane, decane, ethanol, methanol, propanol, acetone, butanol etc. The water may be separated from any single product, or a mixture of any two, three or four products or a complex mixture of many products.
[0100] In some embodiments, at least some of any one or more products are in the form of a gas or a vapour. In other words, a product may be partially in a gas or vapour form and partially in a liquid form. It may be that all of the products are partially in the gas or vapour form and partially in a liquid form. It may also be that one or more products are entirely in the gas or vapour form and one or more other products are partially in the gas or vapour form and partially in a liquid form. Thus, if at least some of the water is in the form of water vapour, then that means that the water is partially in the liquid phase and partially in the form of water vapour. It is within the scope of this invention that any product or water may be present in the gas or vapour phase, the liquid phase and the solid phase. Likewise, it is within the scope of this invention that any product may be partially present in the solid phase and partially present in the gas or vapour phase.
[0101] It is within the scope of this invention that the water and one or more products are in the liquid phase and are in contact with the porous membrane and/or the graphene oxide.
[0102] Support materials with pore size in the range 0.1 to 10 micrometers are commonly referred to as microfiltration membranes. Membrane materials with pore size in the range 0.001 to 0.1 micrometers are commonly referred to as ultrafiltration membranes. However, porous structures having a pore size in the range 100 nm (0.1 micrometers) to 500 nm are also effective in the membranes of the present invention and we have used membranes in the range of 150 nm to 300 nm e.g. 200 nm to demonstrate impermeability even at larger pore sizes (see examples).
[0103] The porous membrane may comprise a synthetic polymer. Examples of synthetic polymers include: polysulfones (e.g., PALL HT Tuffryn®); polyethersulfones (e.g., PALL Supor®, Millipore Express®, Sartorius PES); polyvinylidene difluoride (PVDF; e.g., PALL FP Vericel™, Millipore Durapore®); polypropylene (e.g., PALL GH Polypro); acrylic polymers (e.g., PALL Versapor®); polyamide (Nylon) (e.g, PALL Nylaflo™, Sartorius NY); and polytetrafluoroethylene (PTFE; e.g. Millipore Omnipore™).
[0104] The porous membrane may comprise a natural polymer or modified natural polymer. Examples of natural polymer and modified natural polymer polymers include: cellulose esters (e.g., Millipore MF-Millipore™); cellulose nitrate (e.g., Sartorius CN); cellulose acetate (e.g., Sartorius CA); and regenerated cellulose (e.g., Sartorius RC).
[0105] The porous membrane may comprise a carbon monolith. An example of a suitable monolith would be those prepared by carbonization of polymerized high internal phase emulsions (see D. Wang, N. L. Smith and P. M. Budd, Polymer Int., 2005, 54, 297-303).
[0106] The porous membrane may comprise an inorganic material. Examples of appropriate inorganic materials include: Aluminium oxide (Al2O3, Alumina; e.g., Anodisk; Anopore™); Metal oxide/ceramic (e.g., Veolia Water Solutions CeraMem®); Silicon carbide (SiC; e.g., Veolia Water Solutions CeraMem®); Zirconium oxide; Silicon dioxide; Titanium dioxide.
[0107] The graphene oxide for use in this application can be made by any means known in the art. In a preferred method, graphite oxide can be prepared from natural graphite flakes by treating them with potassium permanganate and sodium nitrate in concentrated sulphuric acid. This method is called Hummers method. Another method is the Brodie method, which involves adding potassium chlorate (KClO3) to a slurry of graphite in fuming nitric acid. For a review see, Dreyer et al. The chemistry of graphene oxide, Chem. Soc. Rev., 2010, 39, 228-240.
[0108] Individual graphene oxide (GO) sheets can then be exfoliated by dissolving graphite oxide in water with the help of ultrasound, and bulk residues can then be removed by centrifugation.
[0109] The preparation of graphene oxide supported on a porous membrane can be achieved using filtration, spray coating, casting or dip coating techniques. For large scale production of supported graphene based membranes or sheets it is preferred to use spray coating techniques. One benefit of spray coating is that spraying GO solution in water on to the porous support material at an elevated temperature produces a large uniform GO film.
[0110] Previously, graphene had been believed to be impermeable to all gases and liquids but in the present application we demonstrate that, surprisingly, a composite structure made from graphene oxide provided on a porous support can selectively allow permeation of water whilst being impermeable to gases, such as helium, and other vapours and liquids. However, we have also found that the properties of the graphene oxide composite as a whole can be modulated so as to allow selective passage of a limited number of other materials by changing the porous support.
[0111] We have found that the nature of the porous support on to which the graphene oxide is deposited is important to the overall performance of the graphene oxide membrane composite structure and its ability to allow permeation. In other words, the porous support can modulate the ability of the overall graphene oxide membrane to allow selective permeation of liquid or vapour such as water. However, other small polar molecules may permeate through the membrane or may be prevented entirely by the membrane. In this respect the term “small polar molecules” specifically excludes water.
[0112] The graphene oxide membrane according to the invention will always allow the permeation of water and the membrane materials thus govern the extent to which other small polar molecules such as C1-4 alcohols and the like will permeate or be excluded.
[0113] One factor governing the permeation is the smoothness of the surface of the porous support. A smooth porous support such as alumina is better at resisting transmission of a small polar molecule such as propanol than a rougher porous support such as PTFE or PVDF (polyvinyldifluoride) is when used with graphene oxide in a graphene oxide membrane composite. Thus the choice of the porous support will be determined in part by the relative smoothness of its surface and the identity of the material which is intended to be allowed to permeate or to be stopped by the membrane structure. Ideally, the support should have a surface roughness which is the same as or less than that of PTFE. Preferably, the surface roughness is less than that of PTFE.
[0114] Another important factor is the hydrophilicity of the porous support. A more hydrophilic support performs better in allowing selective permeation of small polar molecules such as propanol when used in combination with graphene oxide than a less hydrophilic support does. In this respect, alumina is better as a support than a polymeric material such as PTFE is if preventing permeation of small polar molecules to a higher degree or totally is the intention. The choice of porous support will thus depend on the hydrophilicity of the material used as the support in a graphene oxide membrane composite structure and the identity of the material which is intended to be allowed to permeate. Ideally, the support should be no more hydrophobic than PTFE, and preferably it is more hydrophilic than PTFE.
[0115] We are thus able to produce a membrane whose permeation characteristics can be tailored to allow the passage of water and, in some circumstances, other small polar solvents such as methanol, ethanol and propanol whilst remaining completely impermeable to other fluids such as gases and liquids.
[0116] Methanol and ethanol are small polar molecules which are of particular interest in terms of the ability to allow under certain circumstances selective transmission through the membrane. In other cases, denial of passage through the membrane structure of small polar molecules such as these in their entirety is desirable.
[0117] Membranes of this type will have a number of uses in applications for gas drying and liquid drying.
[0118] Experimental Setup
[0119] Metal containers for permeation experiments were fabricated from an aluminum alloy and sealed by using two O rings. For gravimetric measurements, the containers were specially designed to minimize their mass. The weight loss was monitored by using a computer-controlled precision balance (ADAM Equipment Ltd; accuracy 1 mg). All the gravimetric experiments were carried out in an argon atmosphere in a glove box with a negligible water pressure (<10<−3 >mbar). If the containers were sealed with submicron GO membranes, no weight loss could be detected for any liquid other than water. For the case of an open aperture, evaporation rates for other liquids were higher than for water (for example, 1.3, 6.0 and 8.3 mg/h/mm<2 >for ethanol, hexane and acetone at room temperature (T) respectively).
Example 1
[0120] This example relates to the permeation properties of GO. The studied GO membranes were prepared as follows: We employed Hummer's method to obtain graphite oxide that was dispersed in water by sonication to make a stable suspension of GO crystallites. We then used this suspension to produce laminates by spray- or spin-coating. Scanning electron microscopy and x-ray analysis reveal that such GO films have a pronounced layered structure and consist of crystals with typical sizes L of a few micrometers, which are separated by a typical distance d of ̃10 {acute over (Å)}. For the Example 1 permeation experiments, Cu foils of several centimeters in diameter were uniformly covered with the GO laminates. Then, we chemically etched Cu to produce apertures of diameter D 1 cm fully covered by freestanding GO films. Finally, a metal container was sealed by using the Cu disks. We studied membranes with thicknesses h from 0.1 to 10 μm. Even submicrometer-thick membranes were strong enough to withstand a differential pressure ΔP up to 100 mbar.
[0121] As an initial test, we filled the containers with various gases under a small overpressure (<100 mbar) and recorded its changes over a period of several days. We observed no noticeable reduction in ΔP for any tested gas including He, H2, N2, and Ar. This allowed an estimate for the upper limit on their permeation rates Pr as ≈10<−11 >g/cm<2>·s·bar, which is close to the value reported for micron-sized “balloons” made from continuous graphene monolayers.
[0122] We used mass spectrometry and found no detectable permeation of He. The accuracy was limited only by digital noise of our He spectrometer and a slightly fluctuating background, which yielded Pr<10<−12 >g/cm<2>·s·bar. Using hydrogen mass spectrometry, no permeation was found either, albeit the accuracy was three orders of magnitude lower than for He, due to a larger background. A 12-μm thick film of polyethylene terephthalate (PET) was used as a reference barrier and exhibited a He leakage rate 1000 times higher than our detection limit (FIG. 1) yielding PET's bulk permeability πHe=Pr·h≈10<−11 >mm·g/com<2>·s·bar, in agreement with literature values.
[0123] To evaluate the permeation barrier for liquid substances, we employed weight-loss measurements. FIG. 2 shows examples for evaporation rates from a metal container with an aperture covered by a 1-μm-thick GO membrane. No weight loss could be detected with accuracy of <1 mg for ethanol, hexane, acetone, decane, and propanol in the measurements lasting several days. This sets an upper limit for their π as ≈10<−11 >mm·g/cm<2>·s·bar. We observed a huge weight loss if the container was filled with water.
[0124] Moreover, the evaporation rate was practically the same as as through an open aperture i.e. in the absence of the GO film; (h≈1 μm; aperture's area≈1 cm<2>).
Example 2
[0125] This example relates to the permeation properties of GO on a permeable membrane. A 500 nm thick GO layer supported on an Anopore membrane (20 mm diameter) was prepared by vacuum filtration of graphene oxide solution in water through the alumina membrane. The pore size of the Anopore alumina membrane was 200 nm. The permeability of water, ethanol and methanol through the membrane was determined by measuring the weight loss from a metal container sealed with the GO/membrane composite. A comparative experiment was also performed to determine the permeability of water through an Anopore membrane. The measurements were carried out at room temperature and zero humidity.
[0126] FIG. 3 shows the evaporation rates. As can be seen, the evaporation rate through the GO on Anopore composite and the reference Anopore membrane are practically the same. On the other hand the GO on Anopore composite is completely impermeable to the methanol and ethanol vapours.
Example 3
[0127] This example shows 2-propanol permeation through supported GO membranes.
[0128] To enhance the mechanical stability of GO membranes, we have deposited graphene oxide on different polymer/ceramic porous support and studied their influence on the graphene oxide's membrane property. We have used anodisc alumina, Polytetrafluoroethylene (PTFE), Polyvinylidene fluoride (PVDF) and Cyclopore Polycarbonate (PC) support. We measured the vapour permeation of 2-propanol (2-propanol leak test) at room temperature and room humidity to study the influence of support membrane on property of graphene oxide membrane. Table 1 below shows the 2-propanol leak test performed for GO on different support membranes. The porosity of all the support membranes was identical and for each was 200 nm so that a direct comparison could be made. The table demonstrates that the support material has an effect on graphene oxides membrane properties. The best substrate found so far is the Anodisc alumina membrane. Graphene oxide on Anodisc alumina shows the same property as that of a free standing graphene oxide (impermeable to propanol).
TABLE 1
Propanol leak test for graphene oxide (4
μm thick) on different support membranes
Flux with-
Support porous out GO Flux with
support coating GO coating Hydrophilicity/
(200 nm pore) mg/h/cm<2> mg/h/cm<2> hydrophobicity Smoothness
PTFE 12.8 2.6 hydrophobic Rough
PVDF 13.7 2.1 hydrophilic Smooth
Cyclopore 28.8 1.0 hydrophilic Smooth
Polycarbonate (smoother
than PVDF)
Anodisc 30.6 undetectable hydrophilic Very smooth
alumina
[0129] It can also be seen that the propanol vapour barrier for GO on cyclopore Polycarbonate (PC) membranes were found better than that of PTFE and PVDF. Four micron thick GO coating on PC membranes increase the propanol barrier nearly 30 times compared to five and seven times improvement for PTFE and PVDF membranes. Table 1 also shows smoothness of different GO coating on different support obtained by microscopic or visual inception. In general very smooth hydrophilic substrates are found to be a better candidate for GO support material.
[0130] FIG. 4 shows the propanol leak rate for different support membranes with and without a GO coating.
[0131] Supported GO membranes for Dehydration/concentration
[0132] 8 micron thick GO on Anodisc alumina membrane
[mathematical formula]
[0133] Where X is the weight fraction of components i and j in the feed and Y is the weight fraction of a component in the permeate
Initial alco-
Feed hol concen- Total flux Water flux Alcohol flux Sep.
sample tration wt % gm<−2>h<−1> gm<−2>h<−1> gm<−2>h<−1> Factor
water 0 146.64 146.64 — —
Ethanol 100 Undetect- — Undetect- —
able able
Ethanol 40 69.44 63.64 5.81 7.3
(aq)
2-propanol 40 75.71 74.96 0.75 66.9
(aq)
1-propanol 40 100.29 96.70 3.58 18.0
[0134] GO on hydrophobic substrate
GO 2-propanol
thickness Total flux Water flux flux Sep.
(microns) Feed sample gm<−2>h<−1> gm<−2>h<−1> gm<−2>h<−1> Factor
0 2-propanol 136.95 41.85 95.10 0.3
(40% aq)
1 2-propanol 54.52 27.60 26.92 0.7
(40% aq)
3 2-propanol 36.39 26.66 9.72 1.8
(40% aq)
[0135] GO on hydrophilic substrate
GO 2-propanol
thickness Total flux Water flux flux Sep.
(microns) Feed sample gm<−2>h<−1> gm<−2>h<−1> gm<−2>h<−1> Factor
0 2-propanol 41.56 21.58 19.98 0.7
(40% aq)
1 2-propanol 37.51 29.35 8.17 2.4
(40% aq)
5 2-propanol 26.93 17.89 9.04 2.5
(40% aq)
[0136] A separation factor above one corresponds to the membrane enriching the retentate in alcohol and a separation factor lower than one corresponds to the membrane enriching the retentate in water.
[0137] There are a number of uses for the composite membranes of the invention having these characteristics e.g. detector devices, the pervaporation of fermentation broths, concentration of liquids (e.g. fruit juices), liquid drying (e.g. of hydrocarbon based fuels), gas drying, gas humidification.
[0138] Even though the atomic structure and chemical composition of graphene oxide and graphite oxide membranes are same, the membrane properties are very different. For example, bulk graphite oxide membranes (Boehm et al. Journal of Chimie Physique 58, 141 (1961)) allow water and other polar solvents (eg. ethanol) to permeate but it is completely impermeable to all other gases. Graphene oxide on the other hand is completely impermeable to all gases and liquid (including polar solvents) except water. This unique property of graphene oxide membranes is due to its perfect layered structure. Permeation of polar solvents through graphite oxide membranes can be originated from the difference in their layered structure. FIG. 5 below shows schematic view of the graphene oxide and graphite oxide layered structures.
[0139] FIG. 5 shows a schematic view of the layered structure of graphene oxide (A) and graphite oxide (B) membranes
[0140] Graphite oxide membranes are consists of micrometer size thick perfectly staked graphite oxide flakes (defined by the starting graphite flakes used for oxidation, after oxidation it gets expanded due to the attached functional groups) and can be considered as a polycrystalline material. Exfoliation of graphite oxide in water into individual graphene oxide was achieved by the sonication technique followed by centrifugation at 10000 rpm to remove few layers and thick flakes. Graphene oxide membranes were formed by restacking of these single layer graphene oxides by a number of different techniques such as spin coating, spray coating, road coating and vacuum filtration.
[0141] Graphene oxide membranes according to the invention consist of overlapped layers of randomly oriented single layer graphene oxide sheets with smaller dimensions (due to sonication). These membranes can be considered as a centimetre size single crystal (grain) formed by parallel graphene oxide sheets. Due to this difference in layered structure, the atomic structure of the capillary entrance of graphene oxide and graphite oxide membranes are different. For membranes the edge functional groups are located over the functionalised regions of another graphene oxide sheet while in graphite oxide membranes mostly edges are aligned over another graphene oxide edge. These differences unexpectedly may influence the membrane properties of graphene oxide membranes as compared to those of graphite oxide...
WO2015075451
WATER PURIFICATION
WATER PURIFICATION
This invention relates to methods of purifying water using graphene oxide laminates which are formed from stacks of individual graphene oxide flakes which may be predominantly monolayer thick. The graphene oxide laminates may act as membranes which exclude large solutes i.e. with a hydration radius above about 4.5 AA or they may act as sorbents to absorb solutes having a hydration radius less than about 4.7 AA.
[0001] This invention relates to methods of purifying water using graphene oxide laminates which are formed from stacks of individual graphene oxide flakes which may be predominantly monolayer thick. The graphene oxide laminates may act as membranes or they may act as sorbents.
BACKGROUND
[0002] The removal of solutes from water finds application in many fields.
[0003] This may take the form of the purification of water for drinking or for watering crops or it may take the form of the purification of waste waters from industry to prevent environmental damage. Examples of applications for water purification include: the removal of salt from sea water for drinking water or for use in industry; the purification of brackish water; the removal of radioactive ions from water which has been involved in nuclear enrichment, nuclear power generation or nuclear clean-up (e.g. that involved in the decommissioning of former nuclear power stations or following nuclear incidents); the removal of environmentally hazardous substances (e.g. halogenated organic compounds, heavy metals, chlorates and perchlorates) from industrial waste waters before they enter the water system; and the removal of biological pathogens (e.g. viruses, bacteria, parasites, etc) from contaminated or suspect drinking water.
[0004] In many industrial contexts (e.g. the nuclear industry) it is often desirable to separate dangerous or otherwise undesired solutes from valuable (e.g. rare metals) solutes in industrial waste waters in order that the valuable solutes can be recovered and reused or sold.
[0005] Graphene is believed to be impermeable to all gases and liquids. Membranes made from graphene oxide are impermeable to most liquids, vapours and gases, including helium. However, an academic study has shown that, surprisingly, graphene oxide membranes which are effectively composed of graphene oxide having a thickness around 1 μηι are permeable to water even though they are impermeable to helium. These graphene oxide sheets allow unimpeaded permeation of water (10<10>times faster than He) (Nair et al. Science, 2012, 335, 442-444). Such GO laminates are particularly attractive as potential filtration or separation media because they are easy to fabricate, mechanically robust and offer no principal obstacles towards industrial scale production.
[0006] Zhao et al (Environ. Sci. Technol., 201 1 , 45, 10454-10462) and WO2012/170086 describe sorbents which comprise few layer graphene oxide used as a dispersion.
However, these sorbents are non selective with small ions, complex ions and organic dye molecules being removed from solution at the same time. We believe that the lack of selectivity arises because large solutes adsorb onto the surfaces of the graphene oxide flakes at the same time as small solutes absorbing into the nanocapillaries between the graphene oxide layers.
[0007] CN 101973620 describes the use of a monolayer graphene oxide sheet as an adsorbent of heavy metal ions. Again, such sheets are expected to be non-selective, with both large and small solutes being adsorbed, because the interaction is solely a surface effect.
[0008] Sun et al (Selective Ion Penetration of Graphene Oxide Membranes; ACS Nano 7, 428 (2013)) describes the selective ion penetration of graphene oxide membranes in which the graphene oxide is formed by oxidation of wormlike graphite. The membranes are freestanding in the sense that they are not associated with a support material. The resultant graphene oxide contains more oxygen functional groups than graphene oxide prepared from natural graphite and laminates formed from this material have a wrinkled surface topography. Such membranes differ from those of the present invention because they do not show fast ion permeation of small ions and also demonstrate a selectivity which is substantially unrelated to size
[0009] This study found that sodium salts permeated quickly through GO membranes, whereas heavy metal salts permeated much more slowly. Copper sulphate and organic contaminants, such as rhodamine B are blocked entirely because of their strong interactions with GO membranes. According to this study, ionic or molecular permeation through GO is mainly controlled by the interaction between ions or molecules with the functional groups present in the GO sheets. The authors comment that the selectivity of the GO membranes cannot be explained solely by ionic-radius based theories. They measured the electrical conductivities of different permeate solutions and used this value to compare the permeation rates of different salts. The potential applied to measure the conductivities can affect ion permeation through membranes.
[0010] Other publications (Y. Han, Z. Xu, C. Gao. Adv. Fund. Mater. 23, 3693 (2013); M.
Hu, B. Mi. Environ. Sci. Technol. 47, 3715 (2013); H. Huang et al. Chem. Comm. 49, 5963 (2013)) have reported filtration properties of GO laminates and, although results varied widely due to different fabrication and measurement procedures, they reported appealing characteristics including large water fluxes and notable rejection rates for certain salts.
Unfortunately, large organic molecules were also found to pass through such GO filters.
The latter observation is disappointing and would considerably limit interest in GO laminates as molecular sieves. In this respect, we note that the emphasis of these studies was on high water rates that could be comparable to or exceed the rates used for industrial desalination. Accordingly, a high water pressure was applied and the GO membranes were intentionally prepared as thin as possible, 10-50 nm thick. It may be that such thin stacks contained holes and cracks (some may appear after applying pressure), through which large organic molecules could penetrate.
BRIEF SUMMARY OF THE DISCLOSURE
[0011] In accordance with a first aspect of the invention there is provided a method of reducing the amount of one or more solutes in an aqueous mixture to produce a liquid depleted in said solutes; the method comprising:
(a) contacting a first face of a graphene oxide laminate membrane with the aqueous mixture comprising the one or more solutes;
(b) recovering the liquid from or downstream from a second face of the membrane;
wherein the graphene oxide laminate membrane has a thickness greater than about 100 nm and wherein the graphene oxide flakes of which the laminate is comprised have an average oxygen:carbon weight ratio in the range of from 0.2:1.0 to 0.5: 1.0.
[0012] Thus, in a second aspect, there is provided a fluid filtration device comprising a graphene oxide laminate membrane having a thickness greater than about 100 nm and wherein the graphene oxide flakes from which the laminate is comprised have an average oxygen:carbon weight ratio in the range of from 0.2:1.0 to 0.5: 1.0.
[0013] The term "solute" applies to both ions and counter-ions, and to uncharged molecular species present in the solution. Once dissolved in aqueous media a salt forms a solute comprising hydrated ions and counter-ions. The uncharged molecular species can be referred to as "non-ionic species". Examples of non-ionic species are small organic molecules such as aliphatic or aromatic hydrocarbons (e.g. toluene, benzene, hexane, etc), alcohols (e.g. methanol, ethanol, propanol, glycerol, etc), carbohydrates (e.g. sugars such as sucrose), and amino acids and peptides. The non-ionic species may or may not hydrogen bond with water. As will be readily apparent to the person skilled in the art, the term 'solute' does not encompass solid substances which are not dissolved in the aqueous mixture. Particulate matter will not pass through the membranes of the invention even if the particulate is comprised of ions with small radii.
[0014] The term "hydration radius" refers to the effective radius of the molecule when solvated in aqueous media.
[0015] The reduction of the amount one or more selected solutes in the solution which is treated with the GO membrane of the present invention may entail entire removal of the or each selected solute. Alternatively, the reduction may not entail complete removal of a particular solute but simply a lowering of its concentration. The reduction may result in an altered ratio of the concentration of one or more solutes relative to the concentration of one or more other solutes. The inventors have found that solutes with a hydration radius of less than about 4.5 A pass very quickly through a graphene oxide laminate whereas solutes with a hydration radius greater than about 4.7 A do not pass through at all. Thus it may be that the purified aqueous solution contains substantially no solute having a hydration radius of greater than about 4.7 A
[0016] In cases in which salt is formed from one ion having a hydration radius of larger than about 4.5 A and a counter-ion with a hydration radius of less than about 4.5 A, neither ion will pass through the membrane of the invention because of the electrostatic attraction between the ions. Thus, for example, in the case KsFe(CN)6, neither the Fe(CN)6<3">nor the K<+>pass through the membrane even though the hydration radius of K<+>is less than 4.5 A.
[0017] No electrical potential needs to be applied to the membrane. Thus, it may be that solutes of a size i.e. hydration radius of less than about 4.5 A may pass through on account of diffusion. It is also possible to achieve the same effect by the application of pressure to the solute.
[0018] The size exclusion limit of the membrane is about 4.5 A; however, this exclusion limit may vary between about 4.5 A and about 4.7 A. In the region around sizes between about 4.5 A and about 4.7 A the degree of transmission decreases by orders of magnitude and consequently the perceived value of the size exclusion limit depends on the amount of transmission of solute that is acceptable for a particular application.
[0019] Thus, the one or more solutes the amount of which are reduced in the method of the invention may have a hydration radius greater than about 4.5 A. It follows that the purified liquid may contain a reduced amount of any solutes having a hydration radius of greater than about 4.5 A. In an embodiment, the or each solute has a radius of hydration greater than 4.7 A. It may be that the or each solute has a radius of hydration greater than 4.8 A, e.g. the or each solute has a radius of hydration greater than 4.9 A, or greater than 5.0 A.
[0020] The flakes of graphene oxide which are stacked to form the laminate of the invention are usually monolayer graphene oxide. However, it is possible to use flakes of graphene oxide containing from 2 to 10 atomic layers of carbon in each flake. These multilayer flakes are frequently referred to as "few-layer" flakes. Thus the membrane may be made entirely from monolayer graphene oxide flakes, from a mixture of monolayer and few-layer flakes, or from entirely few-layer flakes. Ideally, the flakes are entirely or predominantly, i.e. more than 75%w/w, monolayer graphene oxide. [0021] The graphene oxide laminate used in the method of the invention has the overall shape of a sheet-like material through which the solute may pass when the laminate is wet with an aqueous or aqueous-based mixture optionally containing one or more additional solvents (which may be miscible or immiscible with water). The solute may only pass provided it is of sufficiently small size. Thus the aqueous solution contacts one face or side of the membrane and purified solution is recovered from the other face or side of the membrane.
[0022] In an embodiment, the method is a method of selectively reducing the amount of a first set of one or more solutes in an aqueous mixture without significantly reducing the amount of a second set of one or more solutes in the aqueous mixture to produce a liquid depleted in said first set of solutes but not depleted in said second set of solutes, wherein the or each solute of the first set has a radius of hydration greater than about 4.5 A and the or each solute of the second set has a radius of hydration less than about 4.5 A. It may be that the or each solute of the second set has a radius of hydration less than about 4.5 A, e.g. the or each solute of the second set has a radius of hydration less than about 4.4 A. It may be that the or each solute of the first set has a radius of hydration greater than about
4.7 A, e.g. the or each solute of the first set has a radius of hydration greater than about
4.8 A or optionally greater than about 5.0 A.
[0023] In an embodiment, the method is continuous. Thus, steps (a) and (b) may be carried out simultaneously or substantially simultaneously.
[0024] It may be that the aqueous mixture is permitted to pass though the membrane through diffusion and / or it may be that a pressure is applied.
[0025] Preferably, no electrical potential is applied across the membrane. In principle, an electrical potential could be applied.
[0026] The graphene oxide laminate membrane is optionally supported on a porous material. This can provide structural integrity. In other words, the graphene oxide flakes may themselves form a layer e.g. a laminate which itself is associated with a porous support such as a porous membrane to form a further laminate structure. In this embodiment, the resulting structure is a laminate of graphene flakes mounted on the porous support. In one illustrative example, the graphene oxide laminate membrane may be sandwiched between layers of a porous material.
[0027] In an embodiment, the graphene oxide flakes of which the laminate is comprised have an average oxygen:carbon weight ratio in the range of from 0.25: 1.0 to 0.45:1.0. Preferably, the flakes have an average oxygen:carbon weight ratio in the range of from 0.3:1.0 to 0.4:1.0. [0028] In an embodiment, the GO flakes which form the membrane have been prepared by the oxidation of natural graphite.
[0029] The method may involve a plurality of graphene oxide laminate membranes. Thus, the filtration device may comprise a plurality of graphene oxide laminate membranes. These may be arranged in parallel (to increase the flux capacity of the process/device) or in series (where a reduction in the amount of one or more solute is achieved by a single laminate membrane but that reduction is less than desired).
[0030] The fluid filtration device may be a filter or it may be a removable and replaceable filter unit for a filtration apparatus. The filtration device may be a filtration apparatus.
[0031] In accordance with a third aspect of the invention there is provided a method of reducing the amount of one or more solutes in an aqueous mixture to produce a liquid depleted in said solutes; the method comprising:
(a) contacting the aqueous mixture with a sorbent comprising a graphene oxide laminate membrane; and
(b) separating the sorbent from the liquid.
[0032] The invention also relates to sorbents which can be used in this method and which are described below.
[0033] Thus, in a fourth aspect, there is provided a sorbent comprising a graphene oxide laminate membrane which is optionally supported on a porous material. The porous material can provide structural integrity.
[0034] The flakes of graphene oxide which are stacked to form the laminate of the invention are usually monolayer graphene oxide. However, it is possible to use flakes of graphene oxide containing from 2 to 10 atomic layers of carbon in each flake. These multilayer flakes are frequently referred to as "few-layer" flakes. Thus the membrane may be made entirely from monolayer graphene oxide flakes, from a mixture of monolayer and few-layer flakes, or from entirely few-layer flakes. Ideally, the flakes are entirely or predominantly, i.e. more than 75%w/w, monolayer graphene oxide.
[0035] The inventors have found that solutes or ions with a radius of less than about 4.5 A absorb very quickly into a graphene oxide laminate i.e. the sorbent material to form localised (within the laminate) high concentration solutions of the solute(s) in question.
The majority of the surfaces (belonging to individual graphene oxide flakes) with which the solutes interact are internal to the graphene oxide laminate in the sense of being within the capillaries of the laminate of the graphene flakes. Migration of the solutes into the sorbent leads to a concentration within the GO laminate which can be an order of magnitude or greater higher than the concentration on the outside i.e. in the aqueous mixture or the purified liquid.
[0036] This means that substantially no solutes with a hydrated radius larger than about 4.5 A can adsorb onto the surface of the flakes. This makes the sorbents of the invention very selective in separating solutes by size. The absorption properties of the membrane are observed with solutes with a hydration radius of less than about 4.5 A; however, this exclusion limit may vary between about 4.5 A and about 4.7 A. In the region around sizes between about 4.5 A and about 4.7 A the degree of transmission decreases by orders of magnitude and consequently the perceived value of the size exclusion limit depends on the amount of transmission of solute that is acceptable for a particular application.
[0037] In an embodiment, the graphene oxide flakes of which the laminate is comprised have an oxygen:carbon weight ratio in the range 0.2 to 0.5, e.g. in the range 0.25 to 0.45. Preferably, the flakes have an oxygen:carbon weight ratio in the range 0.3 to 0.4.
[0038] It may be that the graphene oxide laminate membrane is formed from graphene oxide which has been prepared by the oxidation of natural graphite.
[0039] The sorbents of the invention are also easier to handle than the graphene oxide dispersions of the prior art. It is difficult to separate completely a dispersed fine solid from a liquid, whereas the membranes of the invention form larger units which can be more easily filtered out.
[0040] Thus, the one or more solutes the amount of which are reduced in the method of the invention may have a hydration radius less than about 4.5 A. It follows that the purified liquid may contain a reduced amount of any solutes having a hydration radius of less than about 4.5 A. In an embodiment, the or each solute has a radius of hydration less than about 4.5 A. It may be that the or each solute has a radius of hydration less than about 4.4 A, e.g. the or each solute has a radius of hydration less than about 4.25 A.
[0041] In an embodiment, the method further comprises removing the one or more solutes from the sorbent. This may be achieved using an acid wash or alternatively by placing the sorbent in deionised water for an appropriate period of time. Where, the one or more solutes are removed by placing the sorbent in deionised water, that water may be at a temperature from about 25 °C to about 70 °C. Thus, the sorbent can be regenerated and reused and the absorbed solutes can be recovered or substantially recovered.
[0042] In an embodiment, the method is a method of selectively reducing the amount of a first set of one or more solutes in an aqueous mixture without significantly reducing the amount of a second set of one or more solutes in the aqueous mixture to produce a liquid depleted in said first set of solutes but not depleted in said second set of solutes, wherein the or each solute of the first set has a radius of hydration less than about 4.5 A and the or each solute of the second set has a radius of hydration greater than about 4.7 A. It may be that the or each solute of the first set has a radius of hydration less than about 4.4 A e.g. the or each solute of the first set has a radius of hydration less than about 4.25 A. It may be that the or each solute of the second set has a radius of hydration greater than about 4.8 A, e.g. the or each solute of the second set has a radius of hydration greater than about 5.0 A.
[0043] It may be that that the aqueous mixture is not acidic. Thus the aqueous mixture may be either neutral or alkaline. It may be that the aqueous mixture has a pH from about 5 to about 13, e.g. from about 6 to about 12 or from about 7 to about 11.
[0044] The method may comprise the step of: increasing the pH of the first aqueous mixture to obtain a modified aqueous mixture. It is this modified aqueous mixture which is modified with the sorbent in step a) described above. Increasing the pH may be achieved by adding a base, e.g. an aqueous alkaline solution, to the first aqueous mixture. Suitable alkaline solutions include: ammonium hydroxide, sodium hydroxide, potassium hydroxide, sodium carbonate, potassium carbonate, sodium bicarbonate etc.
[0045] The inventors have surprisingly found that the effectiveness of graphene oxide laminates as sorbents increases with pH.
[0046] The graphene oxide laminate membrane may is optionally supported on a porous material to provide structural integrity, e.g. a flexible porous support. The porous material may be a porous polymer support, e.g. a flexible porous polymer support.
[0047] Preferably, the graphene oxide laminate membrane has a thickness greater than about 100 nm.
[0048] In an illustrative embodiment, the sorbent comprises a composite of the graphene oxide laminate membrane on a flexible porous support in the form of a roll, a ball, a sheet or a folded sheet.
[0049] The following paragraphs may apply to the membranes of any of the first, second, third and fourth aspects of the invention.
[0050] In an embodiment, the graphene oxide laminate membrane has a thickness greater than about 100 nm, e.g. greater than about 500 nm, e.g. a thickness between about 500 nm and about 100 μηι. The graphene oxide laminate membrane may have a thickness up to about 50 μηι. The graphene oxide laminate membrane may have a thickness greater than about 1 μηι, e.g. a thickness between 1 μηι and 15 μηι. Thus, the graphene oxide laminate membrane may have a thickness of about 5 μηι. [0051] The graphene oxide laminates used in the invention may comprise a cross-linking agent.
[0052] A cross linking agent is a substance which bonds with GO flakes in the laminate. The cross linking agent may form hydrogen bonds with GO flakes or it may form covalent bonds with GO flakes. Examples include diamines (e.g. ethyl diamine, propyl diamine, phenylene diamine), polyallylamines and imidazole. Without wishing to be bound by theory, it is believed that these are examples of crosslinking agents which form hydrogen bonds with GO flakes. Other examples include borate ions and polyetherimides formed from capping the GO with polydopamine. Examples of appropriate cross linking systems can be found in Tian et al, (Adv. Mater. 2013, 25, 2980-2983), An et al (Adv. Mater. 2011 , 23, 3842-3846), Hung et al (Cross-linking with Diamine monomers to Prepare Composite Graphene Oxide- Framework Membranes with Varying d-Spacing; Chemistry of Materials, 2014) and Park et al (Graphene Oxide Sheets Chemically Cross-Linked by polyallylamine; J. Phys. Chem. C; 2009)
[0053] The GO laminate may comprise a polymer. The polymer may be interspersed throughout the membrane. It may occupy the spaces between graphene oxide flakes, thus providing interlayer crosslinking. The polymer may be PVA (see for example Li et al Adv. Mater. 2012, 24, 3426-3431). It has been found that GO laminates comprising interspersed polymer exhibit improved adhesiveness to certain substrates (e.g. metals) than GO membranes which do not comprise a polymer. Other polymers which could be used in this manner include poly(4-styrenesulfonate), Nafion, carboxymethyl cellulose, Chitosan, polyvinyl pyrrolidone, polyaniline etc. It may be that the polymer is water soluble. Where the GO laminate comprises a polymer, that polymer (e.g. PVA) may be present in an amount from about 0.1 to about 50 wt%, e.g. from about 5 to about 45 wt%. Thus, the GO laminate may comprise from about 20 to about 40 wt% polymer.
Alternatively, it may be that the polymer is not water soluble.
[0054] It may be that the GO laminate does not comprise a polymer.
[0055] The GO laminate may comprise other inorganic materials, e.g. other two dimensional materials, such as graphene, reduced graphene oxide, hBN, mica. The presence of mica, for example can slightly improve the mechanical properties of the GO laminate.
[0056] The membrane may be a graphene oxide membrane comprising only flakes of graphene oxide.
[0057] In an embodiment, the porous support is an inorganic material. Thus, the porous support (e.g. membrane) may comprise a ceramic. Preferably, the support is alumina, zeolite, or silica. In one embodiment, the support is alumina. Zeolite A can also be used. Ceramic membranes have also been produced in which the active layer is amorphous titania or silica produced by a sol-gel process.
[0058] In an alternate embodiment, the support is a polymeric material. Thus, the porous support may thus be a porous polymer support, e.g. a flexible porous polymer support. Preferably it is PTFE, PVDF or Cyclopore™ polycarbonate. In an embodiment, the porous support (e.g. membrane) may comprise a polymer. In an embodiment, the polymer may comprise a synthetic polymer. These can be used in the invention.
Alternatively, the polymer may comprise a natural polymer or modified natural polymer. Thus, the polymer may comprise a polymer based on cellulose.
[0059] In another embodiment, the porous support (e.g. membrane) may comprise a carbon monolith.
[0060] In an embodiment, the porous support layer has a thickness of no more than a few tens of μηι, and ideally is less than about 100 μηι. Preferably, it has a thickness of 50 μηι or less, more preferably of 10 μηι or less, and yet more preferably is less 5 μηι. In some cases it may be less than about 1 μηι thick though preferably it is more than about 1 μηι.
[0061] Preferably, the thickness of the entire membrane (i.e. the graphene oxide laminate and the support) is from about 1 μηι to about 200 μηι, e.g. from about 5 μηι to about 50.
[0062] The porous support should be porous enough not to interfere with water transport but have small enough pores that graphene oxide platelets cannot enter the pores. Thus, the porous support must be water permeable. In an embodiment, the pore size must be less than 1 μηι. In an embodiment, the support has a uniform pore-structure. Examples of porous membranes with a uniform pore structure are electrochemically manufactured alumina membranes (e.g. those with the trade names: Anopore™, Anodisc™).
[0063] The one or more solutes can be ions and/or they could be neutral organic species, e.g. sugars, hydrocarbons etc. Where the solutes are ions they may be cations and/or they may be anions.
[0064] In a fifth aspect of the invention is provided the use of a graphene oxide laminate membrane for the purification of water.
[0065] The laminate may be comprised in a sorbent.
[0066] The purification of water may be achieved by size exclusion filtration. [0067] The graphene oxide laminate membrane may have any of the features recited above for the first, second, fourth and fifth aspects of the invention. The use may be in any of the methods described above for the first, second, third and fourth aspects of the invention.
BRIEF DESCRIPTION OF THE DRAWINGS
[0068] Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:
[0069] Figure. 1. shows ion permeation through GO laminates: (A) Photograph of a GO membrane covering a 1 cm opening in a copper foil; (B) Schematic of the experimental setup. The membrane separates the feed and permeate containers (left and right, respectively). Magnetic stirring is used to ensure no concentration gradients; (C) Filtration through a 5 μηι thick GO membrane from the feed container with a 0.2 M solution of MgC . The inset shows permeation rates as a function of C in the feed solution. Within our experimental accuracy (variations by a factor of <40% for membranes prepared from different GO suspensions), chloride rates were found the same for MgC , KCI and CuC . Dotted lines are linear fits.
[0070] Fig. 2 shows the sieving through an atomic scale mesh. The shown permeation rates are normalized per 1 M feed solution and measured by using 5 μηι thick membranes. Some of the tested chemicals are named here; the others can be found in the Table 1 below. No permeation could be detected for the solutes shown within the grey area during measurements lasting for 10 days or longer. The thick arrows indicate our detection limit that depends on a solute. Several other large molecules including benzoic acid, DMSO and toluene were also tested and exhibited no detectable permeation. The dashed curve is a guide to the eye, showing an exponentially sharp cut-off with a semi-width of «0.1 A.
[0071] Fig. 3 shows some simulations of molecular sieving. (A) Snapshot of NaCI diffusion through a 9 A graphene slit allowing two monolayers of water. Na<+>and CI<">ions are in yellow and blue, respectively. (B) Permeation rates for NaCI, CuC , MgC , propanol, toluene and octanol for capillaries containing two monolayers of water. For octanol poorly dissolved in water, the hydrated radius is not known and we use its molecular radius. Blue marks: Permeation cutoff for an atomic cluster (pictured in the inset) for capillaries accommodating two and three monolayers of water (width of 9 A and 13 A, respectively).
[0072] Fig. 4 shows that the permeation of salts through GO membranes can be detected by using electrical measurements. The inset shows the measurement setup, and the main figure plots relative changes in resistivity of water with time in the permeate container. Changes are normalized to an initial value of measured resistance of deionized water.
[0073] Fig. 5 shows salt intake by GO laminates. (A) Relative increase in weight for 5μηι thick laminates soaked in different solutions. No intake could be detected for K3[Fe(CN)6] but it was large for small-radius salts. The dotted curves represent the best fit by the Langmuir isotherm. (B) X-ray diffraction for pristine GO, GO treated with NaCI, GO treated with KCI (all the measured at a relative humidity of «50%) and GO in water.
[0074] Fig. 6 shows a snapshot of our molecular dynamics simulations for toluene in water. All toluene molecules are trapped inside the short graphene channel and none leaves it even after 100 ns.
[0075] Fig. 7 shows a simulation of the salt-absorption effect. (A) Snapshot for the case of a 1 M NaCI solution inside the capillary and 0.1 M in the reservoirs (water molecules are removed for clarity). Despite the concentration gradient, ions move from the reservoirs into the capillary. (B) Number of ions inside a 9 A wide capillary (two layers of water) as a function of simulation time. Initial concentrations of NaCI in the two reservoirs were 0.1 , 0.5 and 1 M for the different curves. The initial NaCI concentration inside the capillary was the same for all the curves (C =1 M). (C) Comparison between graphene and GO capillaries. Evolution of the number of ions inside a capillary (1 1A wide) for initial C =1 M throughout the system.
DETAILED DESCRIPTION
[0076] The present invention involves the use of graphene oxide laminate membranes. The graphene oxide laminates and laminate membranes of the invention comprise stacks of individual graphene oxide flakes, in which the flakes are predominantly monolayer graphene oxide. Although the flakes are predominantly monolayer graphene oxide, it is within the scope of this invention that some of the graphene oxide is present as two- or few-layer graphene oxide. Thus, it may be that at least 75% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes, or it may be that at least 85% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes (e.g. at least 95 %, for example at least 99% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes) with the remainder made up of two- or few- layer graphene oxide. Without wishing to be bound by theory, it is believed that water and solutes pass through capillary-like pathways formed between the graphene oxide flakes by diffusion and that the specific structure of the graphene oxide laminate membranes leads to the remarkable selectivity observed as well as the remarkable speed at which the ions permeate the laminate structure. [0077] In one illustrative example, the graphene oxide laminate membranes are made of impermeable functionalized graphene sheets that have a typical size L «1 μηι and the interlayer separation, d, sufficient to accommodate a mobile layer of water.
[0078] The solutes to be removed from aqueous mixtures in the methods of the present invention may be defined in terms of their hydrated radius. Below are the hydrated radii of some exemplary ions and molecules.
Table 1
<img class="EMIRef" id="281513021-imgf000014_0001" />
[0079] The hydrated radii of many species are available in the literature. However, for some species the hydrated radii may not be available. The radii of many species are described in terms of their Stokes radius and typically this information will be available where the hydrated radius is not. For example, of the above species, there exist no literature values for the hydrated radius of propanol, sucrose, glycerol and PTS<4">. The hydrated radii of these species which are provided in the table above have been estimated using their Stokes/crystal radii. To this end, the hydrated radii for a selection of species in which this value was known can be plotted as a function of the Stokes radii for those species and this yields a simple linear dependence. Hydrated radii for propanol, sucrose, glycerol and PTS<4">were then estimated using the linear dependence and the known Stokes radii of those species. [0080] There are a number of methods described in the literature for the calculation of hydration radii. Examples are provided in 'Determination of the effective hydrodynamic radii of small molecules by viscometry'; Schultz and Soloman; The Journal of General Physiology; 44; 1 189-1199 (1963); and 'Phenomenological Theory of Ion Solvation'; E. R. Nightingale. J. Phys. Chem. 63, 1381 (1959).
[0081] The term 'aqueous mixture' refers to any mixture of substances which comprises at least 10% water by weight. It may comprise at least 50% water by weight and preferably comprises at least 80% water by weight, e.g. at least 90% water by weight. The mixture may be a solution, a suspension, an emulsion or a mixture thereof. Typically the aqueous mixture will be an aqueous solution in which one or more solutes are dissolved in water. This does not exclude the possibility that there might be particulate matter, droplets or micelles suspended in the solution. Of course, it is expected that the particulate matter will not pass through the membranes of the invention even if it is comprised of ions with small radii.
[0082] Particularly preferred solutes for removing from water in the first aspect of the invention include hydrocarbons and oils, biological material, dyes, organic compounds (including halogenated organic compounds) and complex ions.
[0083] Particularly preferred solutes for removing from water in the second aspect of the invention include NaCI, heavy metals, ethanol, chlorates and perchlorates, and radioactive elements.
[0084] The graphene oxide for use in this application can be made by any means known in the art. In a preferred method, graphite oxide can be prepared from graphite flakes (e.g. natural graphite flakes) by treating them with potassium permanganate and sodium nitrate in concentrated sulphuric acid. This method is called Hummers method. Another method is the Brodie method, which involves adding potassium chlorate (KCIO3) to a slurry of graphite in fuming nitric acid. For a review see, Dreyer et al. The chemistry of graphene oxide, Chem. Soc. Rev., 2010, 39, 228-240.
[0085] Individual graphene oxide (GO) sheets can then be exfoliated by dissolving graphite oxide in water or other polar solvents with the help of ultrasound, and bulk residues can then be removed by centrifugation and optionally a dialysis step to remove additional salts.
[0086] In a specific embodiment, the graphene oxide of which the graphene oxide laminate membranes of the invention are comprised is not formed from wormlike graphite. Worm-like graphite is graphite that has been treated with concentrated sulphuric acid and hydrogen peroxide at 1000C to convert graphite into an expanded "worm-like" graphite. When this worm-like graphite undergoes an oxidation reaction it exhibits a higher increase the oxidation rate and efficiency (due to a higher surface area available in expanded graphite as compared to pristine graphite) and the resultant graphene oxide contains more oxygen functional groups than graphene oxide prepared from natural graphite. Laminate membranes formed from such highly functionalized graphene oxide can be shown to have a wrinkled surface topography and lamellar structure (Sun et al,; Selective Ion Penetration of Graphene Oxide Membranes; ACS Nano 7, 428 (2013) which differs from the layered structure observed in laminate membranes formed from graphene oxide prepared from natural graphite. Such membranes do not show fast ion permeation of small ions and a selectivity which is substantially unrelated to size (being due rather to interactions between solutes and the graphene oxide functional groups) compared to laminate membranes formed from graphene oxide prepared from natural graphite.
[0087] Without wishing to be bound by theory, individual GO crystallites formed from non-worm like graphite (e.g. natural or pristine graphite) may have two types of regions: functionalized (oxidized) and pristine. The former regions may act as spacers that keep adjacent crystallites apart and the pristine graphene regions may form the capillaries which afford the membranes their unique properties.
[0088] The preparation of graphene oxide supported on a porous membrane can be achieved using filtration, spray coating, casting, dip coating techniques, road coating, inject printing, or any other thin film coating techniques
[0089] For large scale production of supported graphene based membranes or sheets it is preferred to use spray coating, road coating or inject printing techniques. One benefit of spray coating is that spraying GO solution in water on to the porous support material at an elevated temperature produces a large uniform GO film.
[0090] Graphite oxide consists of micrometer thick stacked graphite oxide flakes (defined by the starting graphite flakes used for oxidation, after oxidation it gets expanded due to the attached functional groups) and can be considered as a polycrystalline material.
Exfoliation of graphite oxide in water into individual graphene oxide flakes was achieved by the sonication technique followed by centrifugation at 10000 rpm to remove few layers and thick flakes. Graphene oxide laminates were formed by restacking of these single or few layer graphene oxides by a number of different techniques such as spin coating, spray coating, road coating and vacuum filtration.
[0091] Graphene oxide membranes according to the invention consist of overlapped layers of randomly oriented single layer graphene oxide sheets with smaller dimensions (due to sonication). These membranes can be considered as a centimetre size single crystals (grains) formed by parallel graphene oxide sheets. Due to this difference in layered structure, the atomic structure of the capillary structure of graphene oxide membranes and graphite oxide are different. For graphene oxide membranes the edge functional groups are located over the non-functionalised regions of another graphene oxide sheet while in graphite oxide mostly edges are aligned over another graphite oxide edge. These differences unexpectedly may influence the permeability properties of graphene oxide membranes as compared to those of graphite oxide.
[0092] We have studied GO laminates that were prepared from GO suspensions by using vacuum filtration as described in Example 1. The resulting membranes were checked for their continuity by using a helium leak detector before and after filtration experiments, which proved that the membranes were vacuum-tight in the dry state. Figure 1 shows schematics of our experiments. The feed and permeate compartments were initially filled with different liquids (same or different height) including water, glycerol, toluene, ethanol, benzene and dimethyl sulfoxide (DMSO). No permeation could be detected over a period of many weeks by monitoring liquid levels and using chemical analysis. The situation principally changed if both compartments were filled with water solutions. In this case, permeation through the same vacuum-tight membrane can readily be observed as rapid changes in liquid levels (several mm per day). The direction of flow is given by osmotic pressure. For example, a level of a one molar (1 M) sucrose solution in the feed compartment rises whereas it falls in the permeate compartment filled with deionized water. For a membrane with a thickness h of 1 μηι, we find osmotic water flow rates of «0.2 L nr<2>h<"1>, and the speed increases with increasing the molar concentration C. Because a 1 M sucrose solution corresponds to an osmotic pressure of «25 bar at room temperature (van't Hoff factor is 1 in this case), the flow rates agree with the evaporation rates of «10 L nr<2>h<"1>reported for similar GO membranes (Nair et al. Science, 2012, 335, 442-444), in which case the permeation was driven by a capillary pressure of the order of 1 ,000 bars. Note that hydrostatic pressures in these experiments never exceeded 10<"2>bar and, therefore, could be neglected.
[0093] After establishing that GO membranes connect the feed and permeate containers with respect to transport of water molecules, we have investigated the possibility that dissolved ions and molecules can simultaneously diffuse through capillaries. To this end, we have filled the feed container with various solutions and studied if any of the solutes appears on the other side of GO membranes, that is, in the permeate container filled with deionized water (Fig. 1 B). As a quick test, ion transport can be probed by monitoring electrical conductivity of water in the permeate container (Fig. 4). We have found that for some salts (for example, KCI) the conductivity increases with time but remains unaffected for others (for example, K3[Fe(CN)6]) over many days of measurements. This suggests that only certain ions may diffuse through GO laminates. Note that ions are not dragged by the osmosis-driven water flow but move in the opposite direction.
[0094] To quantify permeation rates for diffusing solutes and test those that do not lead to an increase in conductivity (sucrose, glycerol and so on), we have employed various analytical techniques. Depending on a solute, we have used ion chromatography, inductively coupled plasma optical emission spectrometry, total organic carbon analysis and optical absorption spectroscopy. As an example, Figure 1C shows our results for MgC which were obtained by using ion chromatography and inductively coupled plasma optical emission spectrometry for Mg<2+>and CI<">, respectively. One can see that
concentrations of Mg<2+>and CI<">in the permeate container increase linearly with time, as expected. Slopes of such curves yield permeation rates. The inset of Fig. 1 C illustrates that the observed rates depend linearly on C in the feed container. Note that cations and anions move through membranes in stoichiometric amounts so that charge neutrality within each of the containers is preserved. Otherwise, an electric field would build up across the membrane, slowing fast ions until the neutrality is reached. In Fig. 1 C, permeation of one Mg<2+>ion is accompanied by two ions of chloride, and the neutrality condition is satisfied.
[0095] Figure 2 summarizes our results obtained for different ionic and molecular solutions. The small species permeate with approximately the same speed whereas large ions and organic molecules exhibit no detectable permeation. The effective volume occupied by an ion in water is characterized by its hydrated radius. If plotted as a function of this parameter, our data are well described by a single-valued function with a sharp cutoff at «4.5Α (Fig. 2). Species larger than this are sieved out. This behavior corresponds to a physical size of the mesh of «9Α. Fig. 2 also shows that permeation rates do not exhibit any notable dependence on ion charge and triply charged ions such as As04<3">permeate with approximately the same rate as singly-charged Na<+>or CI<">. Finally, to prove the essential role of water for ion permeation through GO laminates, we dissolved KCI and CuS04in DMSO, the polar nature of which allows solubility of these salts. No permeation has been detected, proving that the special affinity of GO laminates to water is important.
[0096] To explain the observed sieving properties, it is possible to employ the model previously suggested to account for unimpeded evaporation of water through GO membranes (Nair ef al. Science, 2012, 335, 442-444). Individual GO crystallites may have two types of regions: functionalized (oxidized) and pristine. The former regions may act as spacers that keep adjacent crystallites apart. It may be that, in a hydrated state, the spacers help water to intercalate between GO sheets, whereas the pristine regions provide a network of capillaries that allow nearly frictionless flow of a layer of correlated water. The earlier experiments using GO laminates in air with a typical d «10 A have been explained by assuming one monolayer of moving water. For GO laminates soaked in water, d increases to «13+1 A, which allows two or three monolayers. Taking into account the effective thickness of graphene of 3.4 A (interlayer distance in graphite), this yields a pore size of «9-10 A, in agreement with the mesh size found experimentally.
[0097] To support this model, molecular dynamics simulations (MDS) can be used. The setup is shown in Fig. 3A where a graphene capillary separates feed and permeate reservoirs, and its width is varied between 7 and 13 A to account for the possibility of one, two or three monolayers of water. It is found that the narrowest MDS capillaries become filled with a monolayer of ice as described previously and do not allow inside even such small ions as Na<+>and CI<">. However, for two and three monolayers expected in the fully hydrated state, ions enter the capillaries and diffuse into the permeate reservoir. Their permeation rates are found approximately the same for all small ions and show little dependence on ionic charge (Fig. 3B). Larger species (toluene and octanol) cannot permeate even through capillaries containing three monolayers of water (Fig. 6). Large solutes have been modelled as atomic clusters of different size and it is found that the capillaries accommodating 2 and 3 water monolayers rejects clusters with the radius larger than «4.7 and 5.8 A, respectively. This probably indicates that the ion permeation through GO laminates is limited by regions containing two monolayers of water. The experimental and theory results in Figs 2 & 3B show good agreement.
[0098] Regarding the absolute value of ion permeation rates found experimentally, it is possible to estimate that, for laminates with h «5 μηι and L «1 μηι, the effective length of graphene capillaries is Lxh/d «5 mm and they occupy d/L «0.1 % of the surface area of the GO membrane. For a typical diffusion coefficient of ions in water («10<-5>cm<2>/s), the expected diffusion rate for a 1 M solution through GO membrane is «10<"3>mg/h/cm<2>, that is, thousands of times smaller than the rates observed experimentally. Moreover, this estimate neglects the fact that functionalized regions narrow the effective water column. To appreciate how fast the observed permeation is, we have used the standard coffee filter paper and found the same diffusion rates for the paper of 1 mm in thickness (the diffusion barrier is equivalent to a couple of mm of pure water). Such fast transport of small ions cannot be explained by the confinement, which increases the diffusion coefficient by 50%, reflecting the change from bulk to two-dimensional water. Furthermore, functionalized regions (modeled as graphene with randomly attached epoxy groups) do not enhance diffusion but rather suppress it as expected because of the broken translational symmetry.
[0099] To understand the ultrafast ion permeation, it should be recalled that graphene and GO powders exhibit a high adsorption efficiency with respect to many salts. Despite being very densely stacked, GO laminates are surprisingly found to retain this property for salts with small hydrated radii. Experiments show that permeating salts are adsorbed in amounts reaching as much as 25% of membranes' initial weight (Fig. 5). The large intake implies highly concentrated solutions inside graphene capillaries (close to the saturation). MDS simulations confirm that small ions prefer to reside inside capillaries (Fig. 7). The affinity of salts to graphene capillaries indicates an energy gain with respect to the bulk water, and this translates into a capillary-like pressure that acts on ions within a water medium, rather than on water molecules in the standard capillary physics. Therefore, in addition to the normal diffusion, there is a large capillary force, sucking small ions inside the membranes and facilitating their permeation. Our MDS provide an estimate for this ionic pressure as «50 bars. The membranes would therefore be expected to form efficient sorbents for appropriate solutes.
Example 1. Fabrication and characterization of GO membranes and Experimental Set-up
[00100] Graphite oxide was prepared by exposing millimeter size flakes of natural graphite to concentrated sulfuric acid, sodium nitrate and potassium permanganate (Hummers' method). Then, graphite oxide was exfoliated into monolayer flakes by sonication in water, which was followed by centrifugation at 10,000 rpm to remove remaining few-layer crystals. GO membranes were prepared by vacuum filtration of the resulting GO suspension through Anodisc alumina membranes with a pore size of 0.2 μηι. By changing the volume of the filtered GO solution, it was possible to accurately control the thickness h of the resulting membranes, making them from 1 to more than 10 μηι thick. For consistency, all the membranes described in this report were chosen to be 5 μηι in thickness, unless a dependence on 7 was specifically investigated.
[00101] GO laminates were usually left on top of the Anodiscs that served as a support to improve mechanical stability. In addition, influence of this porous support on permeation properties of GO was checked and they were found to be similar to those of free standing membranes.
[00102] The permeation experiments were performed using a U-shaped device shown in Fig. 1 of the main text. It consisted of two tubular compartments fabricated either from glass or copper tubes (inner diameters of 25 mm), which were separated by the studied GO membranes. The membranes were glued to a Cu foil with an opening of 1 cm in diameter (see Fig. 1 of the main text). The copper foil was clamped between two O-rings, which provided a vacuum-tight seal between the two compartments. In a typical experiment, one of the compartments was filled (referred to as feed) with a salt or molecular solution up to a height of approximately 20 cm (0.1 L volume). The other (permeate) compartment was filled with deionized water to the same level. Note that the hydrostatic pressure due to level changes played no role in these experiments where the permeation was driven by large concentration gradients. Magnetic stirring was used in both feed and permeate compartments to avoid possible concentration gradients near the membranes (concentration polarization effect).
[00103] The GO membranes including their entire assembly with the O-rings were thoroughly tested for any possible cracks and holes. In the first control experiment, GO membranes were substituted with a thin Cu foil glued to the Cu foil with all the other steps remaining the same. Using a highly concentrated salt solution in the feed compartment, we could not detect any permeation. In the second experiment, we used reduced GO, which makes the GO membrane water impermeable. Again, no salt permeation could be detected, which proves the absence of holes in the original GO membrane. Finally and most conclusively, we used a helium-leak detector. No holes could be detected in our GO membranes both before and after permeation measurements
[00104] Although graphite oxide is known to be soluble in water, the vacuum-filtered GO laminates were found to be highly stable in water, and it was practically impossible to re- disperse them without extensive sonication. No degradation or damage of membranes was noticed in these filtration experiments lasting for many weeks. To quantify the solubility of GO laminates, we accurately measured their weight and thickness before and after immersing in water for two weeks. No weight or thickness loss could be detected within our accuracy of <0.5%.
[00105] Membranes were thoroughly tested for any possible cracks or holes by using a helium-leak detector as described in Nair et al. Science, 2012, 335, 442-444. To check the laminar structure of our GO membranes, we performed X-ray diffraction measurements, which yielded the interlayer separation d of 9-10 A at a relative humidity of 50±10%.
[00106] PVA-GO laminate samples were prepared by blending water solutions of GO and PVA using a magnetic stirrer. The concentrations were chosen such that a weight percentage of GO in the final laminates of 60-80% was achieved, after water was removed by evaporation. We used vacuum filtration, drop casting and rod coating techniques to produce free standing PVA-GO membranes and PVA-GO coated substrates.
Example 2. Monitoring ion diffusion by electrical measurements
[00107] For a quick qualitative test of ion permeation through GO membranes, the setup shown in Fig. 4 was used. The feed and permeate compartments were separated by GO membranes. We used the same assembly as described above but instead of Cu foil GO were glued to a glass slide with 2 mm hole and the liquid cell was small and made entirely from Teflon. The feed compartment was initially filled with a few ml_ of a concentrated salt solution, and the permeate compartment contained a similar volume of deionized water. The typical feed solution was approximately a million times more electrically conducting than deionized water at room temperature. Therefore, if ions diffuse through the membrane, this results in an increase in conductivity of water at the permeate side.
Permeation of salts in concentrations at a sub-μΜ level can be detected in this manner. Resistance of permeate solution was monitored by using a Keithley source meter and platinum wires as electrodes.
[00108] Figure 4 shows examples of our measurements for the case of NaCI and potassium ferricyanide K3[Fe(CN)6] . The observed decreasing resistivity as a function of time indicates that NaCI permeates through the membrane. Similar behavior is observed for CuS04, KCI and other tested salts with small ions (see the main text). On the other hand, no noticeable changes in conductivity of deionized water can be detected for a potassium ferricyanide solution during measurements lasting for many days (Fig. 4).
Example 3. Quantitative analysis of ion and molecular permeation
[00109] The above electrical measurements qualitatively show that small ions can permeate through our GO membranes whereas large ions such as [Fe(CN)6]<3">cannot. The technique is not applicable for molecular solutes because they exhibit little electrical conductivity. To gain quantitative information about the exact amount of permeating ions as well as to probe permeation of molecular solutes, chemical analysis of water at the permeate side was carried out. Samples were taken at regular intervals from a few hours to a few days and, in some cases, after several weeks. Due to different solubility of different solutes, different feed concentrations were used. They varied from 0.01 to 2 M, depending on a solute. For each salt, measurements were performed at several different feed concentrations to ensure that we worked in the linear response regime where the permeation rate was proportional to the feed concentration (Fig. 1C) and there was no sign of the concentration polarization effect.
[00110] The ion chromatography (IC) and the inductively coupled plasma optical emission spectrometry (ICP-OES) are the standard techniques used to analyze the presence of chemical species in solutions. The IC for anionic species was employed, and the ICP-OES for cations. The measurement techniques provided us with values for ion concentrations in the permeate water. Using the known volume of the permeate (-0.1 L) the number of ions diffused into the permeate compartments were calculated. For certain salts (those with low solubility), the obtained permeate solutions were first concentrated by evaporation to improve the measurement accuracy. Furthermore, the results of the chemical analysis were crosschecked by weighing a dry material left after evaporation of water in the permeate compartment. This also allowed the calculation of the amounts of the salt permeated through the GO membranes. The weight and chemical analyses were found in good quantitative agreement.
[00111] To detect organic solutes such as glycerol, sucrose and propanol, the total organic carbon (TOC) analysis was employed. No traces of glycerol and sucrose could be found in the permeate samples after several weeks, but propanol could permeate, although at a rate much lower than small ions as shown in Fig. 2. The detection limit of the TOC was about 50 μg/L, and this put an upper limit on permeation of the solutes that could not be detected. The corresponding limiting values are shown by arrows in Fig. 2. The above techniques were calibrated using several known concentrations of the studied solutes, and the detection limits were identified by decreasing the concentration of the standard solution until the measured signal became five times the baseline noise.
[00112] The optical absorption spectroscopy is widely used to detect solutes with absorption lines in the visible spectrum. This technique was employed for large ions such as [Fe(CN)6]<3">, [Ru(bipy)3]<2+>of Tris(bipyridine)ruthenium(ll) dichloride ([Ru(bipy)3]Cl2) and PTS<4">of pyrenetetrasulfonic acid tetrasodium salt (Na4PTS). It was not possible to detect any signatures of [Fe(CN)6]<3">, [Ru(bipy)3]<2+>and PTS<4">on the permeate side, even after many weeks of running the analysis. The absorption spectra were taken with air as a background reference. The detection limit was estimated by measuring a reference solution and gradually decreasing its concentration by a factor of 2-3 until the optical absorption peaks completely disappeared. The penultimate concentration was chosen as the corresponding detection limits in Fig. 2.
[00113] An experiment was performed in which a mixture of 0.5M NaCI and 0.01 M tris(bipyridine)ruthenium(ll) dichloride ([Ru(bipy)3]Cl2) was tested. It was found that only sodium chloride diffused through the membrane and [Ru(bipy)3]Cl2 was blocked by the membrane. This indicates that the presence of small ions don't open up the channels enough to allow larger ions to permeate. However, the presence of [Ru(bipy)3]Cl2 decreases the NaCI permeation rate through the membrane by a factor of ten.
Example 5. Salt absorption by GO laminate membranes
[00114] To test the adsorbing efficiency of GO laminates with respect to salts, the following experiments were carried out. GO laminate membranes were accurately weighed and placed in a solution with salt's concentration C (we used MgC , KCI and K3[Fe(CN)6]). After several hours, the laminates were taken out, rinsed with deionized water and dried. An intake of the salts was then measured. Fig. 5 shows that for salt that cannot permeate through GO laminates, there is no increase in weight. On the other hand, for the salts that fit inside GO capillaries, a massive intake was observed that reaches up to 25% in weight (Fig. 5). The intake rapidly saturates at relatively small C of «0.1 M. Further analysis shows that more than a half of this intake is a dry salt with the rest being additional bound water.
[00115] The mass intake in Figure 5A can be due to both dry salt and extra water accumulated inside the capillaries in the presence of the salt. It was possible to separate the contributions by using the following two approaches. For the first one, a GO laminate was dried out in zero humidity of a glove box for one week and accurately weighted. Then, the laminate sample was exposed to a 1 M MgC solution and dried again in zero humidity for a week. A mass intake of 13±2% was found, that is, smaller than that in Fig. 5A. The latter result was confirmed by using a chemical approach. A sample of the same GO laminate exposed to a 1 M MgC1⁄2 solution was dissolved in a mixture of nitric, sulphuric and perchloric acids at 300°C, which effectively turned carbon into CO2 (chemical burning of graphene). After this, ICP-OES (iCap 6300) was employed to measure the amount of Mg in the resulting solution. We found 3-3.2% of Mg in weight, which translates into «13% mass intake of MgC from a 1 M solution, in agreement with the above result based on weighing. This percentage means that more than a half of the intake in Fig. 5A was due to the dry salt with the rest being additional bound water.
[00116] This experiment was repeated at high and low pH. pH of the salt solution was adjusted by adding either dilute hydrochloric acid or ammonium hydroxide solutions as appropriate. The amount of salt absorbed into the GO membranes were measured after treating (immersing) the GO membranes in one molar salt solution with different pH for 24hour. The weight uptake was measured by using the gravimetric methods.
[00117] The table below shows the results for MgC .
pJH Mass intake (wt%)
2 7.1 %
6 12.6%
11 18.2%
[00118] As can be seen, the sorption capability of the membrane increases with pH. A similar trend was observed for NaCI. [00119] The large salt intake proves that the permeating solutes accumulate inside GO capillaries, leading to highly concentrated internal solutions. Using the measured amount of adsorbed salts and the known amount of water in fully hydrated GO laminates, we estimate that concentrations of the internal solutions reach several molar, that is, can exceed external C by a factor of 10 or more. This "salt sponge" effect is in qualitative agreement with the strong adsorption properties reported previously for graphene and GO powders.
[00120] The accumulation of salts inside GO capillaries means that there is a significant energy gain when ions move inside capillaries from the bulk solution. Our MDS confirm this effect and indicate that the energy gain is mostly due to interaction of ions with graphene walls. The ion sponging is reminiscent of the standard capillary effects where molecules can gain energy by moving inside confined regions. In this case, water plays a role of a continuous medium in which the capillary-like pressure acts on ions, sucking them inside capillaries from the bulk water.
[00121] This data shows that GO laminates can be effective as sorbents, particularly at high pH. The physical and chemical structures of the laminates, and particularly their capillary structures, are not altered when supported on, for example, a porous support. The size of the capillaries and the distribution of chemical functionalisation across the flakes and through the capillary network remains substantially the same. It is expected therefore that supported laminates are effective sorbents. For most practical applications, it is expected that the laminates would be supported on a porous support as indicated elsewhere in this specification. However, for some applications it may be advantageous to have an unsupported laminate.
Example 6. X-ray analysis of GO laminates
[00122] The GO laminates were also examined using X-ray diffraction. Figure 5B shows the diffraction peak for GO at a relative humidity of «50%. It corresponds to the interlayer separation d «8Α. For GO laminates immersed in water, the peak is shifted to d «13Α. It is interesting to note that the peak in water has not become notably broader. This means that the layered structure of GO laminates is preserved in the fully immersed state, and the additional water is adsorbed as an extra layer with a rather uniform thickness of «5Α.
Taking into account that d for reduced GO is «4A, a free space of 9±1A exists between graphene sheets and it is available for transport of water. The latter value is in good agreement with the permeation cut-off observed experimentally and in the MD simulations. We have also measured the volume of taken-in water by weighing. GO laminates exposed to nearly 100% humidity exhibited a water intake that was equal to approximately hxS where S is the laminate area and h its thickness. This is the volume of graphene capillaries, which is used to estimate diffusion through the equivalent water column (see the main text).
[00123] When the GO laminates were soaked in NaCI or KCI solutions and then dried out at the same 50±10% humidity as above, d increased from 8 to 9 A (see Fig. 5B). This increase in the interlayer spacing is consistent with the fact that a significant amount of salts is trapped within graphene capillaries. On the other hand, d for GO inside a concentrated salt solution was found to be similar to that for GO in clean water. The latter observation is attributed to the fact that d is determined by the number of intercalating layers of water whereas ions are moving inside this water layer, in agreement with the MD simulations.
Example 7. Molecular dynamics simulations
[00124] Our basic modeling setup consisted of two equal water reservoirs connected by a capillary formed by parallel graphene sheets as shown in Fig. 3A. Sizes of the reservoirs and capillaries varied in different modeling experiments. To analyze the salt-sponge effect and study ion diffusion in the confined geometry, we used reservoirs with a height of 51.2 A, a length of 50 A and a depth of 49.2 A, which were connected by a 30 A long capillary. A slightly smaller setup was used to assess sieving properties of graphene capillaries. It consisted of the reservoirs with a height of 23.6 A, a length of 50 A and a depth of 30.1 A, which were connected by a 20 A long capillary. For both setups, we varied the capillary width d from 7 to 13 A (d is the distance between the centers of the graphene sheets). When the same property was modeled, both setups yielded similar behavior. Periodic boundary conditions were applied in the Z direction, that is, along the capillary depth. Ions or molecules were added until the desired molar concentrations were reached. Water was modeled by using the simple point charge model. Sodium and chlorine ions were modeled by using the parameters from E. S. David, X. D. Liem. J. Chem. Phys. 100, 3757 (1994) and S. Chowdhuri, A. Chandra. J. Chem. Phys. 115, 3732 (2001); magnesium and copper anions with the OPLS-AA parameters. Intermolecular interactions were described by the 12-6 Lennard-Jones (LJ) potential together with a Coulomb potential. Parameters for water/graphene interactions were reported in C. Ailan, W. A. Steele. J. Chem. Phys. 92, 3858 (1990) and T. Werder, J. H. Walther, R. L. Jaffe, T. Halicioglu, P. Koumoutsakos. J. Phys. Chem. B 107, 1345 (2003).
[00125] The system was initially equilibrated at 300 K with a coupling time of 0.1 ps<_1>for 500 ps . In the modeling of sieving properties, our typical simulation runs were 100 ns long and obtained in the isobaric ensemble at the atmospheric pressure where the simulation box was allowed to change only in the X and Y direction with a pressure coupling time of 1 ps<-1>and a compressibility of 4.5x10<"5>bar<1>. The cutoff distance for nonbonding interactions was set up at 10 A, and the particle mesh Ewald summations method was used to model the system's electrostatics. During simulations, all the graphene atoms were held in fixed positions whereas other bonds were treated as flexible. A time step of 1 fs was employed.
[00126] To model sieving properties of graphene, the GROMACS software was used. At the beginning of each simulation run, water molecules rapidly filled the graphene capillary forming one, two or three layer structures, depending on d. Then after a certain period of time, which depended on a solute in the feed reservoir, ions/molecules started enter the capillary and eventually reached the pure water reservoir for all the modeled solutes, except for toluene and octanol. The found permeation rates are shown in Fig. 3B. We have also noticed that cations and anions move through the capillary together and without noticeably changing their hydration shells.
Example 8. Theoretical analysis of permeation for large molecules
[00127] In the case of organic molecules (for example, propanol) simulations showed that they entered the graphene capillary but then rapidly formed clusters that resided inside the capillary for a long time. The cluster formation is probably due to confinement. On the other hand, the long residence times can be attributed to van der Waals forces between the alcohol molecules and graphene. Toluene molecules exhibited even stronger interaction with graphene (due to π-π staking). In simulations, toluene molecules entered the channel but never left it being adsorbed to graphene walls (Fig. 6). This adsorption is likely to be responsible for the experimentally undetectable level of toluene permeation. Therefore, despite the experimental data suggesting a rather simple sieving behavior that can be explained just by the physical size effect, we believe that van der Waals interactions between solutes and graphene may also play a role in limiting permeation for those molecules and ions that have sizes close to the cutoff radius.
[00128] To better understand the observed sieving effect with its sharp physical cutoff, the following analysis was performed. An artificial cluster was modeled as a truncated icosahedron and placed in the middle of the capillary as shown in the inset of Fig. 3B. The size of the cluster was varied by changing the distance between the constituent 60 atoms, and the interaction energy between the cluster and the graphene capillary was calculated. The energy was computed as the sum of interactions between all the atoms involved which were modeled with a 12-6 LJ potential. Positive and negative values of the calculated energy indicate whether the presence of the cluster in the capillary is energetically favorable or not, respectively. The minimum radius for which the spherical cluster was allowed into the graphene capillary obviously depended on the capillary size. For capillaries that allowed two monolayers of waters (d = 9A) this radius was found to be 4.7 A. For wider capillaries containing three water monolayers (d = 13A), the radius was 5.8A. These values are shown in Fig. 3B as the blue bars.
Example 9. Simulations of the ion sponge effect
[00129] In this case, a relatively long capillary (482 A) was employed such that its volume was comparable to that of the reservoirs (see Fig. 7A). The capillary width was varied between 9 and 11 A, which corresponds to 2 and 3 monolayers of water. MDS were carried out in a canonical ensemble using LAMMPS. The temperature was set at 300 K by using the Nose-Hoover thermostat. The equations of motion were integrated using a velocity- Verlet scheme with a time step of 1.0 fs. The snapshots obtained in these simulations (an example is shown in Fig. 7A) were processed by Atomeye.
[00130] During the simulations, we counted the number of ions inside the capillary as a function of time (Fig. 7B). If the initial concentration C of NaCI was taken constant over the entire system (for example, C =1 M for the black curve in Fig. 7B), we found that the salt moved from the reservoirs into the capillary, that is, ions were attracted to the confined region. Then, we used smaller initial C inside the two reservoirs (0.1 and 0.5 M) while keeping the same C =1 M inside the capillary. Despite the large concentration gradient, the salt still moved into the capillary rather than exited it (see Fig. 7B).
[00131] In the next MDS experiment, a low concentration of NaCI was kept in the two reservoirs (C =0.1 M) and gradually increased C inside the graphene capillary up to 3 M. For C =2 M inside it, an influx of NaCI from the reservoirs was still observed. The salt started leaving the capillary only if C inside approached ~3 M. This allows an estimate of the equilibrium concentration of NaCI inside the graphene capillary as 2-3M, in good agreement with the experiments discussed in section #5. The concentration gradient corresponds to a capillary-like pressure of «50 bars, which acts on salt ions against the osmotic pressure.
[00132] We have also assessed whether functionalized GO regions can play any major role in the salt sponge effect and, more generally, in molecular permeation through GO laminates. To this end, we used the same MDS setups as described above but added hydroxyl and epoxy groups to both walls of graphene capillaries. The epoxy group was modeled by binding an oxygen atom to two carbon atoms of graphene and the hydroxyl group (OH) by its oxygen bonded to a carbon atom. For simplicity, oxygen atoms were fixed in their positions whereas the O-H bond was treated as flexible. Fig. 4C shows an example of the latter simulations. Both ion and water dynamics inside GO capillaries is found to be extremely slow. Accordingly, we expect that the sponge effect should be weaker for functionalized capillary regions compared to pristine ones. In addition, the case where only regions near the entrances of the graphene capillary were covered with hydroxyl and epoxy groups was simulated (Fig. 7C). These simulations again showed slow water and ion dynamics, similar to the case of GO capillaries.
[00133] Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of them mean "including but not limited to", and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.
[00134] Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.
[00135] The reader's attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.
OSMOSIS
WO2015075453
WO2015075453
This invention relates to methods of purifying water using forward osmosis, with a graphene oxide laminate acting as a semi-permeable membrane. The laminate is formed from stacks of individual graphene oxide flakes which may be predominantly monolayer thick. The methods of the invention find particular application in the desalination of salt water.
[0001] This invention relates to methods of purifying water using forward osmosis, with a graphene oxide laminate acting as a semi-permeable membrane. The laminate is formed from stacks of individual graphene oxide flakes which may be predominantly monolayer thick. The methods of the invention find particular application in the desalination of salt water.
BACKGROUND
[0002] The removal of solutes from water finds application in many fields.
[0003] This may take the form of the purification of water for drinking or for watering crops or it may take the form of the purification of waste waters from industry to prevent environmental damage. Examples of applications for water purification include: the removal of salt from sea water for drinking water or for use in industry; the purification of brackish water; the removal of radioactive ions from water which has been involved in nuclear enrichment, nuclear power generation or nuclear clean-up (e.g. that involved in the decommissioning of former nuclear power stations or following nuclear incidents); the removal of environmentally hazardous substances (e.g. halogenated organic compounds, heavy metals, chlorates and perchlorates) from industrial waste waters before they enter the water system; and the removal of biological pathogens (e.g. viruses, bacteria, parasites, etc) from contaminated or suspect drinking water.
[0004] In many industrial contexts (e.g. the nuclear industry) it is often desirable to separate dangerous or otherwise undesired solutes from valuable (e.g. rare metal) solutes in industrial waste waters in order that the valuable solutes can be recovered and reused or sold.
[0005] Forward osmosis (FO) is an emerging membrane technologies which has recently found use in low energy desalination process and in brackish water filtration. In FO impure water and a highly concentrated solution of a salt (known as the draw solution (DS)) are separated by a semi-permeable membrane, water moves from the saline water to the concentrated DS due to the osmotic gradient. Hence in FO the driving force is the differential osmotic pressure between feed and DS rather than applied hydraulic pressure as in reverse osmosis. One of the key challenges remaining in this technology is developing a suitable DS that can generate a high osmotic pressure to produce higher water flux while being easy to re-concentrate and recover at a lower energy cost.
[0006] Graphene is believed to be impermeable to all gases and liquids. Membranes made from graphene oxide (GO) are impermeable to most liquids, vapours and gases, including helium. However, an academic study has shown that, surprisingly, graphene oxide membranes which are composed of graphene oxide having a thickness around 1 μηι supported on porous alumina are permeable to water even though they are impermeable to helium. These graphene oxide sheets allow unimpeaded permeation of water (10<10>times faster than He) (Nair et al. Science, 2012, 335, 442-444). Such GO laminates are particularly attractive as potential filtration or separation media because they are easy to fabricate, mechanically robust and offer no principal obstacles towards industrial scale production.
BRIEF SUMMARY OF THE DISCLOSURE
[0007] In accordance with a first aspect of the invention there is provided a method of reducing the amount of one or more solutes in an aqueous mixture to produce a liquid depleted in said solutes; the method comprising the steps of:
a) contacting the aqueous mixture with a first face of a membrane which comprises graphene oxide; and
b) contacting a second face of the membrane with at least one draw solute.
[0008] The membrane may be a graphene oxide membrane comprising only flakes of graphene oxide which may be bound together due to van der Waals forces or the like, or it may comprise graphene oxide flakes which are bound together by chemical or physical means such as with a polymer or adhesive. Alternatively, the membrane may comprise flakes of graphene oxide which are supported on a porous material to provide structural integrity. The flakes may be bound to one another and to the support due to van der Waals forces or the like, or by physical or chemical means.
[0009] In the membrane which comprises graphene oxide, the graphene oxide itself is preferably in the form of a laminate membrane. This is the case irrespective of whether or not a porous material is present to provide additional support.
[0010] The term "solute" applies to both ions and counter-ions, and to uncharged molecular species present in the solution. Once dissolved in aqueous media a salt forms a solute comprising hydrated ions and counter-ions. The uncharged molecular species can be referred to as "non-ionic species". Examples of non-ionic species are small organic molecules such as aliphatic or aromatic hydrocarbons (eg toluene, benzene, hexane, etc), alcohols (eg methanol, ethanol, propanol, glycerol, etc), carbohydrates (eg sugars such as sucrose), and amino acids and peptides. The non-ionic species may or may not hydrogen bond with water. As will be readily apparent to the person skilled in the art, the term 'solute' does not encompass solid substances which are not dissolved in the aqueous mixture. Particulate matter will not pass through the membranes of the invention even if the particulate is comprised of ions with small radii.
[0011] The term "draw solute" refers to ionic or non-ionic species which are readily soluble in water. The draw solute may be in the form of an aqueous solution with a concentration which is sufficient to exert an osmotic effect on an aqueous mixture present on the other side of the membrane of the invention. Alternatively, the draw solute may be in the form of a solid which rapidly forms an aqueous solution during the practising of the method of the invention, thus generating an aqueous solution with a concentration which is sufficient to exert an osmotic effect on an aqueous mixture present on the other side of the membrane of the invention. The osmotic effect results in the transport of water through the membrane from the aqueous mixture into the draw solute.
[0012] Solutes i.e. ionic and non-ionic species present in the aqueous mixture having a hydration radius of greater than about 4.7 A are not transported through the membrane. Solutes having a hydration radius smaller than about 4.5 A may pass through the membrane but only to a limited extent. In this way, the concentration of such solutes of less than about 4.5 A hydration radius may be reduced in the resulting draw solute solution relative to the concentration of the same solutes in the original aqueous mixture. The reduction is typically in the range of about 10-90%, e.g. the range of about 30-80% or the range of about 50-70%.
[0013] The term "hydration radius" refers to the effective radius of the molecule when solvated in aqueous media.
[0014] The reduction of the amount of one or more selected solutes in the solution which is treated with the GO membranes used in the methods of present invention may entail entire removal of the or each selected solute. Alternatively, the reduction may not entail complete removal of a particular solute but simply a lowering of its concentration. The reduction may result in an altered ratio of the concentration of one or more solutes relative to the concentration of one or more other solutes. The inventors have found that solutes with a hydration radius of less than about 4.5 A pass very quickly through a graphene oxide laminate whereas solutes with a hydration radius greater than about 4.7 A do not pass through at all. The inventors have found that under forward osmosis conditions even the concentrations of the solutes with a hydration radius of less than about 4.5 A are lower in the product aqueous mixture, i.e. the 'purified' liquid, than they were in the original aqueous mixture which contained those solutes. It is thought that this is due to the osmotic effect of the draw solute.
[0015] In cases in which a salt is formed from one ion having a hydration radius of larger than about 4.7 A and a counter-ion with a hydration radius of less than about 4.7 A, neither ion will pass through the membrane of the invention because of the electrostatic attraction between the ions. Thus, for example, in the case KsFe(CN)6, neither the Fe(CN)6<3">nor the K<+>pass through the membrane even though the hydration radius of K<+>is less than 4.7 A.
[0016] The size exclusion limit of the membrane is about 4.7 A; however, this exclusion limit may vary between about 4.5 A and about 4.7 A. In the region around sizes between about 4.5 A and about 4.7 A the degree of transmission decreases by orders of magnitude and consequently the perceived value of the size exclusion limit depends on the amount of transmission of solute that is acceptable for a particular application.
[0017] The flakes of graphene oxide which are stacked to form the laminates which may be used in the methods of the invention are usually monolayer graphene oxide. However, it is possible to use flakes of graphene oxide containing from 2 to 10 atomic layers of carbon in each flake. These multilayer flakes are frequently referred to as "few-layer" flakes. Thus the membrane may be made entirely from monolayer graphene oxide flakes, from a mixture of monolayer and few-layer flakes, or from entirely few-layer flakes. Ideally, the flakes are entirely or predominantly, i.e. more than 75% w/w, monolayer graphene oxide.
[0018] The method may further comprise the step (c) recovering the purified aqueous liquid from or downstream from the second face of the membrane. That purified liquid will typically be a solution of the draw solute, but will typically contain substantially no other solute having a hydration radius of greater than about 4.7 A. The purified aqueous mixture may also contain a reduced amount of one or more solutes with a hydration radius less than about 4.5 A relative to the original aqueous mixture.
[0019] In one embodiment, the draw solute may have a hydration radius greater than 4.7 A. Thus the draw solute may be one or more carbohydrate, e.g. sucrose, fructose, glucose or a mixture thereof. A draw solute having a lower hydration radius than 4.7 A may also be used provided that the osmotic pressure in the draw solute is sufficient to ensure forward osmosis occurs and to prevent any unwanted escape of draw solute through the membrane.
[0020] The method may comprise the step (d) separating the draw solute from the purified aqueous liquid, for example, by the evaporation/condensation of water.
Alternatively, the purified aqueous solution comprising the draw solute may be the desired product.
[0021] The step of separating the draw solute from the purified aqueous liquid may comprise (e) contacting a first face of a size exclusion (e,g. a second graphene oxide laminate) membrane with the purified aqueous liquid containing the draw solute;
(f) recovering the purified aqueous liquid containing a substantially reduced amount of the (e.g. substantially no) draw solute, from or downstream from a second face of the membrane.
[0022] It may be that the draw solute includes one or more consumable carbohydrates (e.g. sucrose, glucose, fructose) and the method of the invention is a method of producing drinking water. In this case, the purified aqueous mixture comprising the draw solute will be drinkable as a sugary solution.
[0023] It may be that the method of the invention comprises the iterative repetition of steps (a) and (b) (and optionally steps (c) and/or (d)). This may be needed in the case where a single iteration of steps (a) and (b) only provides a reduction in the concentration of a solute with a hydration radius less than about 4.5 A, but a greater reduction is required. The method may be repeated until the concentration of the solute is reduced to the required level. This may be the case in the desalination of water for drinking, where a reduced concentration of salt is acceptable.
[0024] The method may also be part of a larger separation process involving other conventional separation steps (before and / or after the graphene oxide separation step(s)) designed to remove other contaminants.
[0025] The method may involve a plurality of graphene oxide laminate membranes. Said plurality of membranes may be used in parallel (to increase the total water flux of the process) or in series (to provide an iterative purification process).
[0026] In a preferred embodiment, the method is a method of desalination. Thus, the solutes the concentrations of which are reduced in the methods of the invention may include NaCI.
[0027] In an embodiment, the method is continuous.
[0028] In accordance with a second aspect of the invention is provided the use of a graphene oxide laminate membrane in the purification of water by forward osmosis.
[0029] In accordance with a third aspect of the invention there is provided a forward osmosis membrane comprising graphene oxide.
[0030] The membrane may be a graphene oxide membrane comprising only flakes graphene oxide which may be bound together due to van der Waals forces or the like, or it may comprise graphene oxide flakes which are bound together by chemical or physical means such as with a polymer or adhesive. Alternatively, the membrane may comprise flakes of graphene oxide which are supported on a porous material to provide structural integrity. The flakes may be bound to one another and to the support due to van der Waals forces or the like, or by physical or chemical means.
[0031] In the membrane which comprises graphene oxide, the graphene oxide itself may in one embodiment be in the form of a laminate. This is the case irrespective of whether or not a porous material is present to provide additional support.
[0032] The graphene oxide membrane may be in the form of a container which is able to retain a draw solute or it may form part of an interchangeable element which itself is part of a container for draw solute.
[0033] The following embodiments can apply to the first, second or third aspects of the invention.
[0034] The graphene oxide laminates used in the invention may comprise a cross-linking agent.
[0035] A cross linking agent is a substance which bonds with GO flakes in the laminate. The cross linking agent may form hydrogen bonds with GO flakes or it may form covalent bonds with GO flakes. Examples include diamines (e.g. ethyl diamine, propyl diamine, phenylene diamine), polyallylamines and imidazole. Without wishing to be bound by theory, it is believed that these are examples of crosslinking agents which form hydrogen bonds with GO flakes. Other examples include borate ions and polyetherimides formed from capping the GO with polydopamine. Examples of appropriate cross linking systems can be found in Tian et al, {Adv. Mater. 2013, 25, 2980-2983), An et al {Adv. Mater. 2011 , 23, 3842-3846), Hung et al (Cross-linking with Diamine monomers to Prepare Composite Graphene Oxide- Framework Membranes with Varying d-Spacing; Chemistry of Materials, 2014) and Park et al (Graphene Oxide Sheets Chemically Cross-Linked by polyallylamine; J. Phys. Chem. C; 2009)
[0036] The GO laminate may comprise a polymer. The polymer may be interspersed throughout the membrane. It may occupy the spaces between graphene oxide flakes, thus providing interlayer crosslinking. The polymer may be PVA (see for example Li et al Adv. Mater. 2012, 24, 3426-3431). It has been found that GO laminates comprising interspersed polymer exhibit improved adhesiveness to certain substrates (e.g. metals) than GO membranes which do not comprise a polymer. Other polymers which could be used in this manner include poly(4-styrenesulfonate), Nafion, carboxymethyl cellulose, Chitosan, polyvinyl pyrrolidone, polyaniline etc. It may be that the polymer is water soluble. Where the GO laminate comprises a polymer, that polymer (e.g. PVA) may be present in an amount from about 0.1 to about 50 wt%, e.g. from about 5 to about 45 wt%. Thus, the GO laminate may comprise from about 20 to about 40 wt% polymer.
Alternatively, it may be that the polymer is not water soluble.
[0037] It may be that the GO laminate does not comprise a polymer.
[0038] The GO laminate may comprise other inorganic materials, e.g. other two dimensional materials, such as graphene, reduced graphene oxide, hBN, mica. The presence of mica, for example can slightly improve the mechanical properties of the GO laminate.
[0039] The membrane may be a graphene oxide membrane comprising only flakes of graphene oxide.
[0040] Preferably, the graphene oxide laminate membrane is supported on a porous material. This can improve structural integrity. In other words, the graphene oxide flakes may themselves form a layer e.g. a laminate which itself is associated with a porous support such as a porous membrane to form a further laminate structure. In this embodiment, the resulting structure is a laminate of graphene flakes mounted on the porous support. In a further illustrative example, the graphene oxide laminate membrane may be sandwiched between layers of a porous material.
[0041] Thus, the graphene oxide laminate membrane may be comprised in a composite with a porous support, e.g. a flexible porous support.
[0042] In an embodiment, the graphene oxide laminate membrane has a thickness greater than about 100 nm, e.g. greater than about 500 nm, e.g. a thickness between about 500 nm and about 100 μηι. The graphene oxide laminate membrane may have a thickness up to about 50 μηι. The graphene oxide laminate membrane may have a thickness greater than about 1 μηι, e.g. a thickness between 1 μηι and 15 μηι. Thus, the graphene oxide laminate membrane may have a thickness of about 5 μηι.
[0043] In an embodiment, the graphene oxide flakes of which the membrane is comprised have an average oxygen:carbon weight ratio in the range 0.2: 1.0 to 0.5: 1.0, e.g. in the range 0.25: 1.0 to 0.45: 1.0. Preferably, the flakes have an average oxygen:carbon weight ratio in the range 0.3:1.0 to 0.4:1.0.
[0044] It may be that the graphene oxide laminate membrane is formed from graphene oxide which has been prepared by the oxidation of natural graphite.
[0045] In an embodiment, the porous support is an inorganic material. Thus, the porous support (e.g. membrane) may comprise a ceramic. Preferably, the support is alumina, zeolite, or silica. In one embodiment, the support is alumina. Zeolite A can also be used. Ceramic membranes have also been produced in which the active layer is amorphous titania or silica produced by a sol-gel process.
[0046] In an alternate embodiment, the support is a polymeric material. Thus, the porous support may thus be a porous polymer support, e.g. a flexible porous polymer Preferably it is PTFE, PVDF or Cyclopore™ polycarbonate. In an embodiment, the porous support (e.g. membrane) may comprise a polymer. In an embodiment, the polymer may comprise a synthetic polymer. These can be used in the invention. Alternatively, the polymer may comprise a natural polymer or modified natural polymer. Thus, the polymer may comprise a polymer based on cellulose.
[0047] In another embodiment, the porous support (e.g. membrane) may comprise a carbon monolith.
[0048] In an embodiment, the porous support layer has a thickness of no more than a few tens of μηι, and ideally is less than about 100 μηι. Preferably, it has a thickness of 50 μηι or less, more preferably of 10 μηι or less, and yet more preferably is less 5 μηι. In some cases it may be less than about 1 μηι thick though preferably it is more than about 1 μηι.
[0049] Preferably, the thickness of the entire membrane (i.e. the graphene oxide laminate and the support) is from about 1 μηι to about 200 μηι, e.g. from about 5 μηι to about 50.
[0050] The porous support should be porous enough not to interfere with water transport but have small enough pores that graphene oxide platelets cannot enter the pores. Thus, the porous support must be water permeable. In an embodiment, the pore size must be less than 1 μηι. In an embodiment, the support has a uniform pore- structure. Examples of porous membranes with a uniform pore structure are electrochemically manufactured alumina membranes (e.g. those with the trade names: Anopore™, Anodisc™).
BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:
[0052] Figure. 1. shows ion permeation through GO laminates: (A) Photograph of a GO membrane covering a 1 cm opening in a copper foil; (B) Schematic of the experimental setup. The membrane separates the feed and permeate containers (left and right, respectively). Magnetic stirring is used to ensure no concentration gradients; (C) Filtration through a 5 μηι thick GO membrane from the feed container with a 0.2 M solution of MgC . The inset shows permeation rates as a function of C in the feed solution. Within our experimental accuracy (variations by a factor of <40% for membranes prepared from different GO suspensions), chloride rates were found the same for MgC , KCI and CuC . Dotted lines are linear fits.
[0053] Fig. 2 shows the sieving through an atomic scale mesh. The shown permeation rates are normalized per 1 M feed solution and measured by using 5 μηι thick membranes. Some of the tested chemicals are named here; the others can be found in the Table 1 below. No permeation could be detected for the solutes shown within the grey area during measurements lasting for 10 days or longer. The thick arrows indicate our detection limit that depends on a solute. Several other large molecules including benzoic acid, DMSO and toluene were also tested and exhibited no detectable permeation. The dashed curve is a guide to the eye, showing an exponentially sharp cut-off with a semi-width of «0.1Α.
[0054] Fig. 3 shows some simulations of molecular sieving. (A) Snapshot of NaCI diffusion through a 9 A graphene slit allowing two monolayers of water. Na<+>and CI<">ions are in yellow and blue, respectively. (B) Permeation rates for NaCI, CuC , MgC , propanol, toluene and octanol for capillaries containing two monolayers of water. For octanol poorly dissolved in water, the hydrated radius is not known and we use its molecular radius. Blue marks: Permeation cutoff for an atomic cluster (pictured in the inset) for capillaries accommodating two and three monolayers of water (width of 9 A and 13 A, respectively).
[0055] Fig. 4 shows that the permeation of salts through GO membranes can be detected by using electrical measurements. The inset shows the measurement setup, and the main figure plots relative changes in resistivity of water with time in the permeate container. Changes are normalized to an initial value of measured resistance of deionized water.
[0056] Figure 5 shows the dependence of water flux rate through GO membrane on thickness of the membrane (differential osmotic pressure is ~ 100 atm)
[0057] Figure 6 shows the dependence of water flux rate through a five micron thick GO membrane on concentration gradient between feed and draw solution.
DETAILED DESCRIPTION
[0058] The present invention involves the use of a graphene oxide laminate membrane. Typically, these are made of impermeable functionalized graphene sheets that have a typical size L «1 μηι and the interlayer separation, d, sufficient to accommodate a mobile layer of water. The graphene oxide laminates and laminate membranes of the invention comprise stacks of individual graphene oxide flakes, in which the flakes are predominantly monolayer graphene oxide. Although the flakes are predominantly monolayer graphene oxide, it is within the scope of this invention that some of the graphene oxide is present as two- or few-layer graphene oxide. Thus, it may be that at least 75% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes, or it may be that at least 85% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes (e.g. at least 95 %, for example at least 99% by weight of the graphene oxide is in the form of monolayer graphene oxide flakes) with the remainder made up of two- or few- layer graphene oxide. Without wishing to be bound by theory, it is believed that water and solutes pass through pathways formed between the graphene oxide flakes by capillary action and that the specific structure of the graphene oxide laminate membranes leads to the remarkable selectivity observed as well as the remarkable speed at which the ions permeate the laminate structure.
[0059] The solutes to be removed from aqueous mixtures in the methods of the present invention may be defined in terms of their hydrated radius. Likewise, the draw solutes used in the methods of the present invention may be defined in terms of their hydrated radius. Below are the hydrated radii of some exemplary solutes.
[0060] Table 1
<img class="EMIRef" id="281513025-imgf000011_0001" />
[0061] The hydrated radii of many species are available in the literature. However, for some species the hydrated radii may not be available. The radii of many species are described in terms of their Stokes radius and typically this information will be available where the hydrated radius is not. For example, of the above species, there exist no literature values for the hydrated radius of propanol, sucrose, glycerol and PTS<4">. The hydrated radii of these species which are provided in the table above have been estimated using their Stokes/crystal radii. To this end, the hydrated radii for a selection of species in which this value was known can be plotted as a function of the Stokes radii for those species and this yields a simple linear dependence. Hydrated radii for propanol, sucrose, glycerol and PTS<4">were then estimated using the linear dependence and the known Stokes radii of those species.
[0062] There are a number of methods described in the literature for the calculation of hydration radii. Examples are provided in 'Determination of the effective hydrodynamic radii of small molecules by viscometry'; Schultz and Soloman; The Journal of General Physiology; 44; 1 189-1199 (1963); and 'Phenomenological Theory of Ion Solvation'; E. R. Nightingale. J. Phys. Chem. 63, 1381 (1959).
[0063] The term 'aqueous mixture' refers to any mixture of substances which comprises at least 10% water by weight. It may comprise at least 50% water by weight and preferably comprises at least 80% water by weight, e.g. at least 90% water by weight. The mixture may be a solution, a suspension, an emulsion or a mixture thereof. Typically the aqueous mixture will be an aqueous solution in which one or more solutes are dissolved in water. This does not exclude the possibility that there might be particulate matter, droplets or micelles suspended in the solution. Of course, it is expected that the particulate matter will not pass through the membranes of the invention even if it is comprised of ions with small radii.
[0064] The graphene oxide for use in this application can be made by any means known in the art. In a preferred method, graphite oxide can be prepared from graphite flakes (e.g. natural graphite flakes) by treating them with potassium permanganate and sodium nitrate in concentrated sulphuric acid. This method is called Hummers method. Another method is the Brodie method, which involves adding potassium chlorate (KCIO3) to a slurry of graphite in fuming nitric acid. For a review see, Dreyer et al. The chemistry of graphene oxide, Chem. Soc. Rev., 2010, 39, 228-240.
[0065] Individual graphene oxide (GO) sheets can then be exfoliated by dissolving graphite oxide in water or other polar solvents with the help of ultrasound, and bulk residues can then be removed by centrifugation and optionally a dialysis step to remove additional salts.
[0066] In a specific embodiment, the graphene oxide of which the graphene oxide laminate membranes of the invention are comprised is not formed from wormlike graphite. Worm-like graphite is graphite that has been treated with concentrated sulphuric acid and hydrogen peroxide at 1000C to convert graphite into an expanded "worm-like" graphite. When this worm-like graphite undergoes an oxidation reaction it exhibits a higher increase the oxidation rate and efficiency (due to a higher surface area available in expanded graphite as compared to pristine graphite) and the resultant graphene oxide contains more oxygen functional groups than graphene oxide prepared from natural graphite. Laminate membranes formed from such highly functionalized graphene oxide can be shown to have a wrinkled surface topography and lamellar structure (Sun et al,; Selective Ion Penetration of Graphene Oxide Membranes; ACS Nano 7, 428 (2013) which differs from the layered structure observed in laminate membranes formed from graphene oxide prepared from natural graphite. Such membranes do not show fast ion permeation of small ions and a selectivity which is substantially unrelated to size (being due rather to interactions between solutes and the graphene oxide functional groups) compared to laminate membranes formed from graphene oxide prepared from natural graphite.
[0067] Without wishing to be bound by theory, individual GO crystallites formed from non-worm like graphite (e.g. natural or pristine graphite) may have two types of regions: functionalized (oxidized) and pristine. The former regions may act as spacers that keep adjacent crystallites apart and the pristine graphene regions may form the capillaries which afford the membranes their unique properties.
[0068] The preparation of graphene oxide supported on a porous membrane can be achieved using filtration, spray coating, casting, dip coating techniques, road coating, inject printing, or any other thin film coating techniques
[0069] For large scale production of supported graphene based membranes or sheets it is preferred to use spray coating, road coating or inject printing techniques. One benefit of spray coating is that spraying GO solution in water on to the porous support material at an elevated temperature produces a large uniform GO film.
[0070] Graphite oxide consists of micrometer thick stacked graphite oxide flakes (defined by the starting graphite flakes used for oxidation, after oxidation it gets expanded due to the attached functional groups) and can be considered as a polycrystalline material.
Exfoliation of graphite oxide in water into individual graphene oxide flakes was achieved by the sonication technique followed by centrifugation at 10000 rpm to remove few layers and thick flakes. Graphene oxide laminates were formed by restacking of these single or few layer graphene oxides by a number of different techniques such as spin coating, spray coating, road coating and vacuum filtration.
[0071] Graphene oxide membranes according to the invention consist of overlapped layers of randomly oriented single layer graphene oxide sheets with smaller dimensions (due to sonication). These membranes can be considered as a centimetre size single crystals (grains) formed by parallel graphene oxide sheets. Due to this difference in layered structure, the atomic structure of the capillary structure of graphene oxide membranes and graphite oxide are different. It is believed that for graphene oxide membranes the edge functional groups are located over the non-functionalised regions of another graphene oxide sheet while in graphite oxide mostly edges are aligned over another graphite oxide edge. These differences unexpectedly may influence the permeability properties of graphene oxide membranes as compared to those of graphite oxide.
[0072] We have studied GO laminates that were prepared from GO suspensions by using vacuum filtration as described in Example 1. The resulting membranes were checked for their continuity by using a helium leak detector before and after filtration experiments, which proved that the membranes were vacuum-tight in the dry state. Figure 1 shows schematics of our experiments. The feed and permeate compartments were initially filled with different liquids (same or different height) including water, glycerol, toluene, ethanol, benzene and dimethyl sulfoxide (DMSO). No permeation could be detected over a period of many weeks by monitoring liquid levels and using chemical analysis. The situation principally changed if both compartments were filled with water solutions. In this case, permeation through the same vacuum-tight membrane can readily be observed as rapid changes in liquid levels (several mm per day). The direction of flow is given by osmotic pressure. For example, a level of a one molar (1 M) sucrose solution in the feed compartment rises whereas it falls in the permeate compartment filled with deionized water. For a membrane with a thickness h of 1 μηι, we find osmotic water flow rates of «0.2 L nr<2>h<"1>, and the speed increases with increasing the molar concentration C. Because a 1 M sucrose solution corresponds to an osmotic pressure of «25 bar at room temperature (van't Hoff factor is 1 in this case), the flow rates agree with the evaporation rates of «10 L nr<2>h<"1>reported for similar GO membranes (Nair et al. Science, 2012, 335, 442-444), in which case the permeation was driven by a capillary pressure of the order of 1 ,000 bars. Note that hydrostatic pressures in these experiments never exceeded 10<"2>bar and, therefore, could be neglected.
[0073] After establishing that GO membranes connect the feed and permeate containers with respect to transport of water molecules, we have investigated the possibility that dissolved ions and molecules can simultaneously diffuse through capillaries. To this end, we have filled the feed container with various solutions and studied if any of the solutes appears on the other side of GO membranes, that is, in the permeate container filled with deionized water (Fig. 1 B). As a quick test, ion transport can be probed by monitoring electrical conductivity of water in the permeate container (Fig. 4). We have found that for some salts (for example, KCI) the conductivity increases with time but remains unaffected for others (for example, K3[Fe(CN)6]) over many days of measurements. This suggests that only certain ions may diffuse through GO laminates. Note that ions are not dragged by the osmosis-driven water flow but move in the opposite direction.
[0074] To quantify permeation rates for diffusing solutes and test those that do not lead to an increase in conductivity (sucrose, glycerol and so on), we have employed various analytical techniques. Depending on a solute, we have used ion chromatography, inductively coupled plasma optical emission spectrometry, total organic carbon analysis and optical absorption spectroscopy. As an example, Figure 1 C shows our results for MgC which were obtained by using ion chromatography and inductively coupled plasma optical emission spectrometry for Mg<2+>and CI<">, respectively. One can see that concentrations of Mg<2+>and CI<">in the permeate container increase linearly with time, as expected. Slopes of such curves yield permeation rates. The inset of Fig. 1 C illustrates that the observed rates depend linearly on C in the feed container. Note that cations and anions move through membranes in stoichiometric amounts so that charge neutrality within each of the containers is preserved. Otherwise, an electric field would build up across the membrane, slowing fast ions until the neutrality is reached. In Fig. 1 C, permeation of one Mg<2+>ion is accompanied by two ions of chloride, and the neutrality condition is satisfied.
[0075] Figure 2 summarizes our results obtained for different ionic and molecular solutions. The small species permeate with approximately the same speed whereas large ions and organic molecules exhibit no detectable permeation. The effective volume occupied by an ion in water is characterized by its hydrated radius. If plotted as a function of this parameter, our data are well described by a single-valued function with a sharp cutoff at «4.5Α (Fig. 2). Species larger than this are sieved out. This behavior corresponds to a physical size of the mesh of «9Α. Fig. 2 also shows that permeation rates do not exhibit any notable dependence on ion charge and triply charged ions such as As04<3">permeate with approximately the same rate as singly-charged Na<+>or CI<">. Finally, to prove the essential role of water for ion permeation through GO laminates, we dissolved KCI and CuS04in DMSO, the polar nature of which allows solubility of these salts. No permeation has been detected, proving that the special affinity of GO laminates to water is important. [0076] To explain the observed sieving properties, it is possible to employ the model previously suggested to account for unimpeded evaporation of water through GO membranes (Nair ef al. Science, 2012, 335, 442-444). Individual GO crystallites may have two types of regions: functionalized (oxidized) and pristine. The former regions may act as spacers that keep adjacent crystallites apart. It may be that, in a hydrated state, the spacers help water to intercalate between GO sheets, whereas the pristine regions provide a network of capillaries that allow nearly frictionless flow of a layer of correlated water. The earlier experiments using GO laminates in air with a typical d «10 A have been explained by assuming one monolayer of moving water. For GO laminates soaked in water, d increases to «13+1 A, which allows two or three monolayers. Taking into account the effective thickness of graphene of 3.4 A (interlayer distance in graphite), this yields a pore size of «9-10 A, in agreement with the mesh size found experimentally.
[0077] To support this model, molecular dynamics simulations (MDS) can be used. The setup is shown in Fig. 3A where a graphene capillary separates feed and permeate reservoirs, and its width is varied between 7 and 13 A to account for the possibility of one, two or three monolayers of water. It is found that the narrowest MDS capillaries become filled with a monolayer of ice as described previously and do not allow inside even such small ions as Na<+>and CI<">. However, for two and three monolayers expected in the fully hydrated state, ions enter the capillaries and diffuse into the permeate reservoir. Their permeation rates are found approximately the same for all small ions and show little dependence on ionic charge (Fig. 3B). Larger species (toluene and octanol) cannot permeate even through capillaries containing three monolayers of water (Fig. 6). Large solutes have been modelled as atomic clusters of different size and it is found that the capillaries accommodating 2 and 3 water monolayers rejects clusters with the radius larger than «4.7 and 5.8 A, respectively. This probably indicates that the ion permeation through GO laminates is limited by regions containing two monolayers of water. The experimental and theory results in Figs 2 & 3B show good agreement.
[0078] Regarding the absolute value of ion permeation rates found experimentally, it is possible to estimate that, for laminates with h «5 μηι and L «1 μηι, the effective length of graphene capillaries is Lxh/d «5 mm and they occupy d/L «0.1 % of the surface area of the GO membrane. For a typical diffusion coefficient of ions in water («10<-5>cm<2>/s), the expected diffusion rate for a 1 M solution through GO membrane is «10<"3>mg/h/cm<2>, that is, thousands of times smaller than the rates observed experimentally. Moreover, this estimate neglects the fact that functionalized regions narrow the effective water column. To appreciate how fast the observed permeation is, we have used the standard coffee filter paper and found the same diffusion rates for the paper of 1 mm in thickness (the diffusion barrier is equivalent to a couple of mm of pure water). Such fast transport of small ions cannot be explained by the confinement, which increases the diffusion coefficient by 50%, reflecting the change from bulk to two-dimensional water. Furthermore, functionalized regions (modeled as graphene with randomly attached epoxy groups) do not enhance diffusion but rather suppress it as expected because of the broken translational symmetry.
[0079] To understand the ultrafast ion permeation, it should be recalled that graphene and GO powders exhibit a high adsorption efficiency with respect to many salts. Despite being very densely stacked, GO laminates are surprisingly found to retain this property for salts with small hydrated radii. Experiments show that permeating salts are adsorbed in amounts reaching as much as 25% of membranes' initial weight (Fig. 5). The large intake implies highly concentrated solutions inside graphene capillaries (close to the saturation). MDS simulations confirm that small ions prefer to reside inside capillaries (Fig. 7). The affinity of salts to graphene capillaries indicates an energy gain with respect to the bulk water, and this translates into a capillary-like pressure that acts on ions within a water medium, rather than on water molecules in the standard capillary physics. Therefore, in addition to the normal diffusion, there is a large capillary force, sucking small ions inside the membranes and facilitating their permeation. Our MDS provide an estimate for this ionic pressure as «50 bars. The membranes would therefore be expected to form efficient sorbents for appropriate solutes.
Example 1. Fabrication and characterization of GO membranes and Experimental Set-up
[0080] Graphite oxide was prepared by exposing millimeter size flakes of natural graphite to concentrated sulfuric acid, sodium nitrate and potassium permanganate (Hummers' method). Then, graphite oxide was exfoliated into monolayer flakes by sonication in water, which was followed by centrifugation at 10,000 rpm to remove remaining few-layer crystals. GO membranes were prepared by vacuum filtration of the resulting GO suspension through Anodisc alumina membranes with a pore size of 0.2 μηι. By changing the volume of the filtered GO solution, it was possible to accurately control the thickness h of the resulting membranes, making them from 1 to more than 10 μηι thick. For consistency, all the membranes described in this report were chosen to be 5 μηι in thickness, unless a dependence on 7 was specifically investigated.
[0081] GO laminates were usually left on top of the Anodiscs that served as a support to improve mechanical stability. In addition, influence of this porous support on permeation properties of GO was checked and they were found to be similar to those of free standing membranes. [0082] The permeation experiments were performed using a U-shaped device shown in Fig. 1 of the main text. It consisted of two tubular compartments fabricated either from glass or copper tubes (inner diameters of 25 mm), which were separated by the studied GO membranes. The membranes were glued to a Cu foil with an opening of 1 cm in diameter (see Fig. 1 of the main text). The copper foil was clamped between two O-rings, which provided a vacuum-tight seal between the two compartments. In a typical experiment, one of the compartments was filled (referred to as feed) with a salt or molecular solution up to a height of approximately 20 cm (0.1 L volume). The other (permeate) compartment was filled with deionized water to the same level. Note that the hydrostatic pressure due to level changes played no role in these experiments where the permeation was driven by large concentration gradients. Magnetic stirring was used in both feed and permeate compartments to avoid possible concentration gradients near the membranes (concentration polarization effect).
[0083] The GO membranes including their entire assembly with the O-rings were thoroughly tested for any possible cracks and holes. In the first control experiment, GO membranes were substituted with a thin Cu foil glued to the Cu foil with all the other steps remaining the same. Using a highly concentrated salt solution in the feed compartment, we could not detect any permeation. In the second experiment, we used reduced GO, which makes the GO membrane water impermeable. Again, no salt permeation could be detected, which proves the absence of holes in the original GO membrane. Finally and most conclusively, we used a helium-leak detector. No holes could be detected in our GO membranes both before and after permeation measurements
[0084] Although graphite oxide is known to be soluble in water, the vacuum-filtered GO laminates were found to be highly stable in water, and it was practically impossible to re- disperse them without extensive sonication. No degradation or damage of membranes was noticed in these filtration experiments lasting for many weeks. To quantify the solubility of GO laminates, we accurately measured their weight and thickness before and after immersing in water for two weeks. No weight or thickness loss could be detected within our accuracy of <0.5%.
[0085] Membranes were thoroughly tested for any possible cracks or holes by using a helium-leak detector as described in Nair et al. Science, 2012, 335, 442-444. To check the laminar structure of our GO membranes, we performed X-ray diffraction measurements, which yielded the interlayer separation d of 9-10 A at a relative humidity of 50±10%.
[0086] PVA-GO laminate samples were prepared by blending water solutions of GO and PVA using a magnetic stirrer. The concentrations were chosen such that a weight percentage of GO in the final laminates of 60-80% was achieved, after water was removed by evaporation. We used vacuum filtration, drop casting and rod coating techniques to produce free standing PVA-GO membranes and PVA-GO coated substrates.
Example 2. Monitoring ion diffusion by electrical measurements
[0087] For a quick qualitative test of ion permeation through GO membranes, the setup shown in Fig. 4 was used. The feed and permeate compartments were separated by GO membranes. We used the same assembly as described above but instead of Cu foil GO were glued to a glass slide with 2 mm hole and the liquid cell was small and made entirely from Teflon. The feed compartment was initially filled with a few mL of a concentrated salt solution, and the permeate compartment contained a similar volume of deionized water. The typical feed solution was approximately a million times more electrically conducting than deionized water at room temperature. Therefore, if ions diffuse through the membrane, this results in an increase in conductivity of water at the permeate side.
Permeation of salts in concentrations at a sub-μΜ level can be detected in this manner. Resistance of permeate solution was monitored by using a Keithley source meter and platinum wires as electrodes.
[0088] Figure 4 shows examples of our measurements for the case of NaCI and potassium ferricyanide K3[Fe(CN)6] . The observed decreasing resistivity as a function of time indicates that NaCI permeates through the membrane. Similar behavior is observed for CuS04, KCI and other tested salts with small ions (see the main text). On the other hand, no noticeable changes in conductivity of deionized water can be detected for a potassium ferricyanide solution during measurements lasting for many days (Fig. 4).
Example 3. Quantitative analysis of ion and molecular permeation
[0089] The above electrical measurements qualitatively show that small ions can permeate through our GO membranes whereas large ions such as [Fe(CN)6]<3">cannot. The technique is not applicable for molecular solutes because they exhibit little electrical conductivity. To gain quantitative information about the exact amount of permeating ions as well as to probe permeation of molecular solutes, chemical analysis of water at the permeate side was carried out. Samples were taken at regular intervals from a few hours to a few days and, in some cases, after several weeks. Due to different solubility of different solutes, different feed concentrations were used. They varied from 0.01 to 2 M, depending on a solute. For each salt, measurements were performed at several different feed concentrations to ensure that we worked in the linear response regime where the permeation rate was proportional to the feed concentration (Fig. 1C) and there was no sign of the concentration polarization effect. [0090] The ion chromatography (IC) and the inductively coupled plasma optical emission spectrometry (ICP-OES) are the standard techniques used to analyze the presence of chemical species in solutions. The IC for anionic species was employed, and the ICP-OES for cations. The measurement techniques provided us with values for ion concentrations in the permeate water. Using the known volume of the permeate (-0.1 L) the number of ions diffused into the permeate compartments were calculated. For certain salts (those with low solubility), the obtained permeate solutions were first concentrated by evaporation to improve the measurement accuracy. Furthermore, the results of the chemical analysis were crosschecked by weighing a dry material left after evaporation of water in the permeate compartment. This also allowed the calculation of the amounts of the salt permeated through the GO membranes. The weight and chemical analyses were found in good quantitative agreement.
[0091] To detect organic solutes such as glycerol, sucrose and propanol, the total organic carbon (TOC) analysis was employed. No traces of glycerol and sucrose could be found in the permeate samples after several weeks, but propanol could permeate, although at a rate much lower than small ions as shown in Fig. 2. The detection limit of the TOC was about 50 μg/L, and this put an upper limit on permeation of the solutes that could not be detected. The corresponding limiting values are shown by arrows in Fig. 2. The above techniques were calibrated using several known concentrations of the studied solutes, and the detection limits were identified by decreasing the concentration of the standard solution until the measured signal became five times the baseline noise.
[0092] The optical absorption spectroscopy is widely used to detect solutes with absorption lines in the visible spectrum. This technique was employed for large ions such as [Fe(CN)6]<3">, [Ru(bipy)3]<2+>of Tris(bipyridine)ruthenium(ll) dichloride ([Ru(bipy)3]Cl2) and PTS<4">of pyrenetetrasulfonic acid tetrasodium salt (Na4PTS). It was not possible to detect any signatures of [Fe(CN)6]<3">, [Ru(bipy)3]<2+>and PTS<4">on the permeate side, even after many weeks of running the analysis. The absorption spectra were taken with air as a background reference. The detection limit was estimated by measuring a reference solution and gradually decreasing its concentration by a factor of 2-3 until the optical absorption peaks completely disappeared. The penultimate concentration was chosen as the corresponding detection limits in Fig. 2.
[0093] An experiment was performed in which a mixture of 0.5M NaCI and 0.01 M tris(bipyridine)ruthenium(ll) dichloride ([Ru(bipy)3]Cl2) was tested. It was found that only sodium chloride diffused through the membrane and [Ru(bipy)3]Cl2 was blocked by the membrane. This indicates that the presence of small ions don't open up the channels enough to allow larger ions to permeate. However, the presence of [Ru(bipy)3]Cl2 decreases the NaCI permeation rate through the membrane by a factor of ten.
Example 4. Molecular dynamics simulations
[0094] Our basic modeling setup consisted of two equal water reservoirs connected by a capillary formed by parallel graphene sheets as shown in Fig. 3A. Sizes of the reservoirs and capillaries varied in different modeling experiments. To analyze the salt-sponge effect and study ion diffusion in the confined geometry, we used reservoirs with a height of 51.2 A, a length of 50 A and a depth of 49.2 A, which were connected by a 30 A long capillary. A slightly smaller setup was used to assess sieving properties of graphene capillaries. It consisted of the reservoirs with a height of 23.6 A, a length of 50 A and a depth of 30.1 A, which were connected by a 20 A long capillary. For both setups, we varied the capillary width d from 7 to 13 A (d is the distance between the centers of the graphene sheets). When the same property was modeled, both setups yielded similar behavior. Periodic boundary conditions were applied in the Z direction, that is, along the capillary depth. Ions or molecules were added until the desired molar concentrations were reached. Water was modeled by using the simple point charge model. Sodium and chlorine ions were modeled by using the parameters from E. S. David, X. D. Liem. J. Chem. Phys. 100, 3757 (1994) and S. Chowdhuri, A. Chandra. J. Chem. Phys. 115, 3732 (2001); magnesium and copper anions with the OPLS-AA parameters. Intermolecular interactions were described by the 12-6 Lennard-Jones (LJ) potential together with a Coulomb potential. Parameters for water/graphene interactions were reported in C. Ailan, W. A. Steele. J. Chem. Phys. 92, 3858 (1990) and T. Werder, J. H. Walther, R. L. Jaffe, T. Halicioglu, P. Koumoutsakos. J. Phys. Chem. B 107, 1345 (2003).
[0095] The system was initially equilibrated at 300 K with a coupling time of 0.1 ps<"1>for 500 ps . In the modeling of sieving properties, our typical simulation runs were 100 ns long and obtained in the isobaric ensemble at the atmospheric pressure where the simulation box was allowed to change only in the X and Y direction with a pressure coupling time of 1 ps<-1>and a compressibility of 4.5* 10<"5>bar<1>. The cutoff distance for nonbonding interactions was set up at 10 A, and the particle mesh Ewald summations method was used to model the system's electrostatics. During simulations, all the graphene atoms were held in fixed positions whereas other bonds were treated as flexible. A time step of 1 fs was employed.
[0096] To model sieving properties of graphene, the GROMACS software was used. At the beginning of each simulation run, water molecules rapidly filled the graphene capillary forming one, two or three layer structures, depending on d. Then after a certain period of time, which depended on a solute in the feed reservoir, ions/molecules started enter the capillary and eventually reached the pure water reservoir for all the modeled solutes, except for toluene and octanol. The found permeation rates are shown in Fig. 3B. We have also noticed that cations and anions move through the capillary together and without noticeably changing their hydration shells.
Example 5. Theoretical analysis of permeation for large molecules
[0097] In the case of organic molecules (for example, propanol) simulations showed that they entered the graphene capillary but then rapidly formed clusters that resided inside the capillary for a long time. The cluster formation is probably due to confinement. On the other hand, the long residence times can be attributed to van der Waals forces between the alcohol molecules and graphene. Toluene molecules exhibited even stronger interaction with graphene (due to π-π staking). In simulations, toluene molecules entered the channel but never left it being adsorbed to graphene walls. This adsorption is likely to be responsible for the experimentally undetectable level of toluene permeation. Therefore, despite the experimental data suggesting a rather simple sieving behavior that can be explained just by the physical size effect, we believe that van der Waals interactions between solutes and graphene may also play a role in limiting permeation for those molecules and ions that have sizes close to the cutoff radius.
[0098] To better understand the observed sieving effect with its sharp physical cutoff, the following analysis was performed. An artificial cluster was modeled as a truncated icosahedron and placed in the middle of the capillary as shown in the inset of Fig. 3B. The size of the cluster was varied by changing the distance between the constituent 60 atoms, and the interaction energy between the cluster and the graphene capillary was calculated. The energy was computed as the sum of interactions between all the atoms involved which were modeled with a 12-6 LJ potential. Positive and negative values of the calculated energy indicate whether the presence of the cluster in the capillary is energetically favorable or not, respectively. The minimum radius for which the spherical cluster was allowed into the graphene capillary obviously depended on the capillary size. For capillaries that allowed two monolayers of waters (d = 9A) this radius was found to be 4.7 A. For wider capillaries containing three water monolayers (d = 13A), the radius was 5.8A. These values are shown in Fig. 3B as the blue bars.
Example 6. Forward Osmosis
[0099] One of the advantages of using GO membrane for forward osmosis (FO) is that we can use any molecule or salt higher than 4.7 A radius as a draw solute. This techniques was used to estimate the salt rejection capability of GO membranes and also to establish the feasibility of using GO for FO applications.
[00100] In our FO experiment one side of a tube was filled with concentrated solution of large molecules such as glycerol or sucrose (DS) while other side was filed with dilute solution of NaCI. In such conditions, glycerol and sucrose being larger in size are completely impermeable to the membrane demonstrated an ideal situation of forward osmosis. Because of the differential osmotic pressure between glucose and NaCI water from the NaCI side flowed towards glucose side. The water flux rate was measured by measuring the increase in height of the liquid column. Some NaCI also diffuses with water and we estimated the amount of NaCI on the other side by ion chromatography. Salt rejection was calculated using the equation 1-Cp/Cf where Cp is the concentration of NaCI in transmitted water and Cf is the concentration of NaCI in feed side. This analysis yields 62% salt rejection for the GO membrane.
[00101] The dependency of the water flux rate in these FO conditions with the thickness of the membrane has also been studied. Figure 5 shows the dependence of water flux rate on thickness of the membrane for a differential osmotic pressure of ~ 100 atm. Our typical micron thick membranes yield 1 L/h-m2 water flux. The observed water flux is comparable to the conventional FO membranes.
[00102] Water flux rate through GO membrane in FO condition with different concentration gradients between feed and drain solution has also been studied. Fig. 6 shows water flux through a five micron thick membrane for different concentration gradients. This study shows that water flux rate increases initially with increasing concentration gradient and for very high (~ > 7 M) concentration gradients the water flux decreases. Even at high concentrations, however, there is still a reasonable water flux.
[00103] Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of them mean "including but not limited to", and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.