GB2464956A

GB2464956A - Water purification method using a field separable osmotic agent

Info

Publication number: GB2464956A
Application number: GB0819952A
Authority: GB
Inventors: Barnaby Warne; Richard Buscall; Eric Mayes; Tim Oriard; Ian Norris
Original assignee: APACLARA Ltd; Cascade Designs Inc
Current assignee: APACLARA Ltd; Cascade Designs Inc
Priority date: 2008-10-31
Filing date: 2008-10-31
Publication date: 2010-05-05
Also published as: GB0819952D0

Abstract

A method of obtaining purified water from an impure water source comprises, contacting the impure source through a semi-permeable membrane with a higher osmotic potential draw solution containing a field separable osmotic agent —comprising one or more ionic salts and/or a superparamagnetic nano-particle bound to a hydrophilic polymer — maintaining contact to allow a net flow of water to into the draw solution and carrying out magnetic and/or electric field separation to remove the osmotic agent from the draw solution. The use of one or more ionic salts as field separable osmotic agents in forward osmosis processes is also disclosed and an osmotic agent comprising a superparamagnetic nano-particle bound to a hydrophilic polymer. The electric field separation is preferably carried out using capacitive deionisation wherein the electrode is made of nanoporous carbon and preferably a single ionic salt, such as Magnesium, MgCl2, or Calcium Chloride, CaCl2, is used. The superparamagnetic nano-particle is preferably iron based and may be magnetite (Fe3O4) or maghemite (γ-Fe2O3) or a mixture thereof.

Description

WATER PURIFICATION METHOD

[0001] The present invention relates to a method for obtaining purified water from an impure water source by a forward osmosis (FO) separation process and an apparatus for carrying out this method. The invention also relates to specific field separable osmotic agents and their use in the method.

BACKGROUND

[0002] Forward osmosis is a process in which a polar liquid containing impurities (e.g. contaminated water) is contacted through a semi-permeable membrane with another liquid having a higher osmotic potential (known as a "draw solution"). The liquid passes from the contaminated solution through the semi-permeable membrane into the draw solution, thereby separating the solvent from the contaminants.

[0003] Forward osmosis has a number of advantages over other separation methods employing a semi-permeable membrane. Since forward osmosis does not rely on an applied external pressure to force the polar liquid across the semi-permeable membrane, it does not necessarily require a pump or other external pressure source to operate. In addition, the absence of an external pressure source means that there is a reduced tendency for the semi-permeable membrane to become clogged by larger mechanically compacted impurities and so the need for regular cleaning or replacement of the filter and the need for pump maintenance is reduced.

[0004] Forward osmosis also has advantages over other water purification schemes when the nature of the contaminated source water is unknown. Some schemes only disinfect, some only decontaminate, yet, with the choice of an appropriately restrictive membrane, forward osmosis can disinfect, decontaminate and desalinate in one system making it a general purpose water purification technique.

[0005] Forward osmosis is therefore particularly suitable for the purification of impure or otherwise contaminated water, in order to obtain purified water such as potable water in reduced power or low maintenance situations.

[0006] In order for the draw solution to have a sufficiently high osmotic potential for water to pass from the impure water source through the semi-permeable membrane, it is necessary for the draw solution to contain an osmotic agent. Therefore, in order to obtain purified water, it is necessary for the osmotic agent and the water in the draw solution to be separated following extraction of the water from the impure solution.

[0007] A method of separating an osmotic agent and water is disclosed in Desalination 174 (2005) 1-11. In the process described by this reference, ammonium bicarbonate is dissolved in water to form a draw solution capable of extracting water from a saline feed. The ammonium carbonate is then converted to ammonia and carbon dioxide under heating. These gases then evaporate from the solution to give potable water. This method suffers from the disadvantage that it is necessary to carry out a distillation step in order to obtain a purified solution.

[0008] Alternatively, it has been proposed in US 3,670,897 to use a precipitable salt such as aluminium sulphate as the osmotic agent. Following the diffusion of water from the impure solution into the draw solution, the salt may be removed from the draw solution by precipitation using calcium hydroxide to form aluminium hydroxide and calcium sulphate. These compounds may be removed from the draw solution by conventional filtration techniques to give purified water. However, this method suffers from the disadvantage that it is difficult to precipitate substantially all of the aluminium sulphate osmotic agent without leaving an excess of calcium hydroxide in the solution.

In order to remove this excess hydroxide it is necessary to use an additional step of adding either sulphuric acid or carbon dioxide.

[0009] It would therefore be advantageous if a method could be found for obtaining purified water that does not suffer from the above disadvantages.

SUMMARY OF INVENTION

[0010] In accordance with a first aspect of the invention, there is provided a method of obtaining purified water from an impure water source comprising the steps of: (i) contacting the impure water source through a semi-permeable membrane with a higher osmotic potential draw solution containing a field separable osmotic agent, said osmotic agent comprising one or more ionic salts and/or a superparamagnetic nano-particle bound to a hydrophilic polymer; (ii) maintaining the contact for a time sufficient for a net flow of water to take place from the impure water source into the draw solution; and (iii) carrying out magnetic and/or electric field separation of the field separable osmotic agent from the draw solution to obtain purified water.

[0011] In accordance with a second aspect of the invention, there is provided a superparamagnetic osmotic agent comprising a superparamagnetic nano-particle bound to a hydrophilic polymer.

[0012] In accordance with a third aspect of the invention, there is provided the use as an osmotic agent of a superparamagnetic nano-particle bound to a hydrophilic polymer.

[0013] In accordance with a fourth aspect of the invention, there is provided the use as a field separable osmotic agent in a forward osmosis water purification process of one or more ionic salts.

[0014] In accordance with a fifth aspect of the invention, there is provided an apparatus for use in the first aspect of the invention comprising a cell for an impure water source, a cell for a draw solution containing a field separation apparatus selected from a magnetic field separation apparatus and/or a capacitive deionisation apparatus, the cell for the impure water source and the cell for the draw solution being separated by a semi-permeable membrane.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows a Transmission Electron Microscopy (TEM) image of citrate coated magnetite nanoparticles.

Figure 2 shows a Transmission Electron Microscopy (TEM) image of poly(sodium 4 styrenesulfonic acid-co-maleic acid) magnetite nanoparticles.

Figure 3 shows a static osmotic separator apparatus.

Figure 4 shows separation data from the static osmotic separator.

Figure 5 shows a dynamic osmotic separator apparatus.

DETAILED DESCRIPTION OF THE FIGURES

Figure 1 shows a Transmission Electron Microscopy (TEM) image of citrate coated magnetite nanoparticles. The scale bar is 100 nm.

Figure 2 shows a Transmission Electron Microscopy (TEM) image of poly(sodium 4 styrenesulfonic acid-co-maleic acid) magnetite nanoparticles. The scale bar is 100 nm.

Figure 3 shows a static osmotic separator apparatus. A filter unit, housing a circular piece of semi-permeable membrane (the dashed line in the diagram) is used to contact water in a dish, to the osmotic material being tested. Water is drawn across the membrane by the osmotic agent, which dilutes and increases in volume. The increase in volume is measured by the rise of the osmotic agent up the tube (designated "H").

Figure 4 shows separation data from the static osmotic separator. The osmotic agent tested was Fe304 coated with poly(sodium 4-styrenesulfonate)-co-(maleic acid) polymer (A). This osmotic agent was compared to a known superparamagnetic particle surrounded by a protein (magnetoferritin (x)).

Figure 5 shows a dynamic osmotic separator apparatus. The test solution is continuously circulated between the test solution reservoir and the semi-permeable membrane by the first peristaltic pump. Similarly, the osmotic agent is continuously circulated between the osmotic agent reservoir and the semi-permeable membrane using the second peristaltic pump. This ensures that the concentration gradient across the membrane is maintained at a maximum so that the passage of water across the membrane is also maximised.

DETAILED DESCRIPTION

Impure water source [0015] The impure water source may be any water-based composition in which the impurities in the impure water source generate an osmotic potential less than that of the draw solution. Typical impure water sources include river water, stagnant water, brackish water or salt water.

[0016] The impurities contained in the impure water source may for example be organic or inorganic. Organic impurities may be biological or non-biological. Examples of biological organic impurities include enzymes, viruses, bacteria and yeast. Examples of non-biological organic impurities include synthetic dyes, detergents, flavours, oil emulsions, paint pigments, wood resin, dyes and gross organic particulates. Examples of inorganic impurities include aqueous salts, metal ions and gross inorganic particu lates.

[0017] The impurity particles may have a size (diameter or largest dimension in the case of non-spherical particles) greater than 0.Snm or greater than mm, such as greater than lOnm or lOOnm. The impurity particles may be larger than 200nm, 400nm or 600nm for example greater than 1000nm (1pm). The impurity particle size may also be less than 6,000 nm or less than 2,000 nm, for example less than 1,000nm or less than 500nm. In some aspects, the impurity particle size may be less than 300nm, for example less than 150nm.

[0018] In cases where the impure water source contains salt, the salt concentration may be greater than 0.5 g11, for example greater than 5 g11 such as greater than 10 gl . The salt concentration may also be less than 50 g11, such as less than 30 g11, less than 20 g11 or less than 10 g11.

[0019] Particle size may be determined using any suitable technique known in the art, for example photon correlation spectroscopy (PCS). A suitable PCS instrument is the Brookhaven90Plus particle sizer. Particle size may also be determined using a Transmission Electron Microscope (TEM) for example the Philips EM43OT TEM. The TEM technique also provides a useful indication of the degree of aggregation of the particles in the sample.

Semi-iermeable membrane [0020] The semi-permeable membrane may be any semi-permeable membrane known in the art that will selectively permit the passage of water molecules from the impure water source to the draw solution whilst inhibiting the passage of impurities from the impure water source to the draw solution or the passage of the osmotic agent from the draw solution to the impure water source.

[0021] Semi-permeable membranes may be defined in terms of their effective pore sizes. Microfiltration membranes have effective pore sizes larger than 0.1 pm and are capable of filtering parasites and suspended particles from water. Ultrafiltration membranes have pore sizes between about 0.01-0.lpm and are capable of filtering bacteria and the majority of proteins and viruses from water. Nanofiltration membranes block all biological species as well as medium sized molecules such as sugars. Finally, reverse and forward osmosis membranes block all species except water and the lightest uncharged molecules such as ethanol and urea.

[0022] Since the membrane must inhibit the passage of the osmotic agent from the draw solution to the impure solution, the membrane must have an effective pore size less than the size of the osmotic agent particles. The membrane will therefore usually comprise as a component at least one ultrafiltration, nanofiltration or forward osmosis membrane type.

[0023] The membrane may be of a single type (e.g. ultrafiltration, nanofiltration or forward osmosis) or a composite comprising more than one membrane type. In some aspects, it is advantageous to use a composite membrane comprising a fine filter (e.g. a forward osmosis or nanofiltration filter) supported on a relatively coarse filter (e.g. a microfiltration or ultrafiltration filter). The supporting action of the relatively coarse filter allows the use of a thinner fine filter, thereby increasing the throughput of water into the draw solution.

[0024] Where the fine filter (e.g. a forward osmosis filter) is supported by a relatively coarse filter (e.g. a microfiltration or ultrafiltration filter), the forward osmosis membrane may be less than about 100pm thick, for example less than about 5Opm thick., [0025] The semi-permeable membrane may be inorganic e.g. ceramic or organic such as cellulose, polyamide, polycarbonate, polyester, polymer, polypropylene, polysulfone, polytetrafluoroethylene (PTFE), polyvinylidine, nitrocellulose, nylon, polyamide and polyvinyichioride. The semi-permeable membrane is preferably of hydrophilic character, for example a cellulose ester plastic.

Draw solution [0026] The draw solution comprises water and the field separable osmotic agent. The draw solution preferably does not comprise a significant proportion of any additional components although the usual minor amounts of suspended solids, minerals and dissolved gases sometimes found in water may be present. These include for example clays, calcium and magnesium ions, calcium carbonate and iron oxides. The total level of additional components in the draw solution is preferably less than lOOmgl1, for example less than 50 mgl1or less than 30 mgl*

Field Seraration

[0027] Following extraction of the water from the impure solution into the draw solution, the osmotic agent is field separated from the draw solution.

[0028] Depending on the nature of the field separable osmotic agent an electric or magnetic field may be used to separate the osmotic agent particles from the draw solution.

[0029] Specifically, where the field separable osmotic agent is superparamagnetic in nature, the field should have a magnetic component. Conversely, where the field separable osmotic agent is ionic in nature, the field should have an electric component.

Magnetic field separation method

[0030] Where the field separation process is a magnetic separation process, the separation process is preferably carried out by high gradient magnetic separation. This typically involves applying an external magnetic field to a stationary phase having a high degree of surface irregularity or curvature so that intense local fields are generated within the magnetised stationary phase, and then passing the draw solution through the magnetised stationary phase. The superparamagnetic osmotic agent particles are therefore retained within the stationary phase while the purified water passes through for subsequent recovery. Suitable materials for use as the stationary phase include steel powder (e.g. lOOmesh powder of 410-L annealed steel), steel wool or steel wire. When using the above steel powder, an external magnetic field strength of about 0.5 Tesla may be used during the magnetic separation stage.

[0031] The draw solution can then be regenerated by removing the external magnetic field and then passing water (for example a portion of the previously recovered water) through the stationary phase to form a solution of superparamagnetic osmotic agent particles.

[0032] In an aspect of the invention, the external magnetic field remains in place throughout, but is arranged in a non-uniform manner such that the flow of the draw solution through the magnetised stationary phase induces a split into purified and concentrated streams. The concentrated stream will be attracted to where the field strength is most intense. At this point, the concentrate can be continuously drawn off by a tube. Not all the draw agent may be removed in one go, in which case the purified stream passes through the magnetised stationary phase to other regions of non-uniform magnetic field where the same process of splitting into a concentrate, which is drawn off, and a purified water stream take place. High osmotic potential agents, which are harder to separate, may require several stages of separation before water of sufficient purity is produced.

[0033] In an aspect of the invention, following diffusion of the water from the impure water solution into the draw solution the speed of separation of the superparamagnetic osmotic agent from the draw solution may be increased by reducing the temperature of the draw solution below the ferromagnetic blocking temperature. This causes the formation of magnetically agglomerated particles that can be magnetically extracted more rapidly than corresponding smaller non-aggregated particles. These agglomerated particles could then be dispersed with agitation when raised back above the blocking temperature.

[0034] This may be done for example by selecting the superparamagnetic material so that the particles exhibit superparamagnetic properties under ambient conditions during daytime but exhibit ferromagnetic properties during the reduced temperatures at night.

The purified polar solvent may then be extracted from the draw solution in the morning.

[0035] During the removal or localisation of the superparamagnetic particles, it is preferred that the draw solution does not contact the impure solution through the semi-permeable membrane. This is to avoid diffusion of the water back into the impure liquid as the removal or localisation of the superparamagnetic particles reduces the osmotic potential of the draw solution.

[0036] Magnetic field separation has the advantage that no external power supply is required to drive the field separation process and so this method is particularly useful in applications where a high degree of portability is required e.g. camping.

Electric field separation Method

[0037] Where the field separation process is an electric separation process, the separation process is preferably carried out by capacitive deionisation.

[0038] Capacitive deionization utilises charged, high surface area plate capacitors to attract ionic species from solution. Dissolved ions from the draw solution are separated when they pass between two electrodes (having a typical potential difference between them of approximately 1-1.5 volts). The cations and anions in solution are attracted to the oppositely charged electrodes, and the operational cycle continues until either the electrode surfaces are saturated with adsorbed ions or all of the ions in solution are adsorbed onto the electrodes.

[0039] The draw solution can then be regenerated by either shorting the two electrodes or discharging the capacitors under controlled conditions to recover energy with the ions being released into a rinse stream.

[0040] In order to maximise the number of ions that may be separated from the draw solution (and hence the osmotic potential attainable by the draw solution when these ions are present in solution), the specific surface area of the capacitor material is preferably as high as possible. The high surface area capacitors are preferably formed from a material having a specific surface area greater then 100 m2/g, for example greater than 250 m2/g, such as greater then 500 m2/g, greater than 750 m2/g or greater than 1000 m2/g. In a preferred aspect, the high surface areas capacitors are formed from a material having a specific surface area greater then 1,250 m2/g. Due to material design constraints, the material will normally have a specific surface area of less than 3,000 m2/g, for example less than 2,000 m2/g, such as less than 1,500 m2/g. However, higher specific surface area materials can be used as capacitor materials without disadvantage.

[0041] The electrode material preferably has a low electrical resistivity. For example, the electrode material may have a resistivity of less than 10 acm, for example less than lc�=.cm or less than 100 mc.cm., In an aspect, the electrical resistivity of the electrode material is less than 70 m.cm, for example less than 40 mcm.

[0042] Suitable electrode materials include carbon nanotubes; graphenes; electrically conducting polymers having a redox storage mechanism; or nanoporous carbon.

[0043] Graphene has the form of a 2D structure of hexagonal cells with a surface area of approximately 2,630 m2 g1. Graphene exhibits good mechanical strength and flexibility and, due to its aromatic nature, has a high electrical conductivity. In an aspect, graphene films may be deposited from solution onto a substrate to form the electrode.

[0044] Suitable polymers having a redox storage mechanism include polyacenes.

Suitable nanoporous carbon materials include carbon aerogel, monolithic activated carbon material (e.g. prepared from Maxsorb 2400 (obtainable from The Kansai Coke & Chemicals Co. Ltd (Amagasaki, Japan) as described in WO 1997/029906) or mineral based carbon supercapacitors (e.g. Carbide Derived Carbon).

[0045] In order to further increase the available capacitor surface area, the capacitors may be present in a stacked or parallel configuration. For example, 2 or more, such as or more or 10 or more pairs of capacitor plates may be used in a stacked or parallel configuration. Typically, the array would contain 50 or fewer, for example 30 or fewer or or fewer pairs of capacitor plates.

[0046] To carry out capacitive deionisation separation of the ionic salt an external power source is required. This may be supplied from the mains or by a battery e.g. a lithium ion battery cell.

[0047] During the removal of the ionic salt from the draw solution, it is preferred that the draw solution does not contact the impure solution through the semi-permeable membrane. This is to avoid diffusion of the water back into the impure solution as the removal of the ionic salt reduces the osmotic potential of the draw solution. This may be achieved by placing a barrier e.g. a slideable barrier between the semi-permeable membrane and either the draw solution or the solution of the impure liquid at the beginning of the salt removal process.

[0048] Following removal of the ionic salt osmotic agent from the draw solution, the purified water may be recovered for use.

Field Serarable osmotic agent

[0049] The field separable osmotic agent may comprise or consist of an ionic salt osmotic agent and/or a superparamagnetic osmotic agent comprising a superparamagnetic nano-particle bound to a hydrophilic polymer. The field separable osmotic agent may comprise a mixture of the ionic salt and superparamagnetic osmotic agents. However, in most embodiments, either an ionic salt osmotic agent or a superparamagnetic osmotic agent is used depending on the application. This simplifies the subsequent separation process since only an electric field separation process or a magnetic separation process is then required.

Ionic salt osmotic agent [0050] The ionic salt osmotic agent comprises one or more water soluble ionic salts.

The anion of the salt may be selected, for example, from chloride, bromide, iodide, acetate, nitrate sulphate. The cation of the salt may be selected, for example, from sodium, lithium, potassium, magnesium, calcium or ammonium.

[0051] In a preferred aspect, the ionic salt osmotic agent has a higher osmotic potential per given mass when dissolved in water than NaCI. This is advantageous in applications where the impure solution exhibits a high osmotic potential (e.g. sea water) since the use of the high osmotic potential salt minimises the amount of the osmotic agent required to obtain a given level of purification. This in turn reduces the required capacity of the electrode to adsorb the ions from the draw solution during the separation stage. Preferred salts include chloride salts of divalent metals.

[0052] In a more preferred aspect of the invention, the salt has a higher osmotic potential than sodium chloride and is also biologically benign at low concentrations.

This is advantageous in embodiments in which the impure solution (e.g. sea water) is to be purified for human or animal consumption. Preferred salts for use in accordance with the invention are therefore MgCI2 or CaCI2.

[0053] In a preferred aspect, the ionic salt osmotic agent is a single salt (e.g. MgCI2 or CaCI2). This gives better adsorption onto the electrode than in the case where more than one salt is used (in which case, competitive adsorption onto the electrolyte can impair the adsorption of one or more of the salts).

[0054] The initial salt concentration used in the draw solution will depend on the osmotic potential required to obtain a net flow of water from the impure water solution to the draw solution. This in turn will depend on the purification application. For example, a lower salt concentration will be required in the draw solution for use in the purification of fresh water (e.g. from a river) than would be required for desalination of brackish or sea water.

[0055] As water flows from the impure solution into the draw solution, the concentration of salt in the draw solution and hence the osmotic potential associated with the draw solution will be reduced. Similarly, as water flows out of the impure solution, its osmotic potential will rise. Therefore, the initial concentration of the salt osmotic agent in the draw solution should be high enough that a difference in osmotic potential between the draw solution and the impure liquid is maintained as the water flows from the impure liquid into the draw solution in order to allow the efficient recovery of water from the impure solution. A higher initial salt concentration will typically be used when the separation process is a batch separation process (in which all of the osmotic agent is added at the start of the process) as opposed to a continuous separation process (in which osmotic agent is added to the draw solution periodically to maintain an osmotic potential difference between the impure solution and the draw solution). In general, the maximum value of the initial salt concentration in the draw solution will be limited by the solubility of the salt in the draw solution. The maximum initial salt concentration will therefore typically be less than 6Owt%, for example less than SOwt% or less than 4Owt%.

[0056] In aspects in which the impure liquid contains a salt of a monovalent metal (e.g. where the impure solution is brackish water or seawater), the initial salt concentration in the draw solution may be selected to be at least lwt% higher than in the impure liquid, for example at least 2wt% higher, such as at least Swt% higher, at least lOwt% higher, at least 2Owt% higher, at least 3Owt% higher or at least 4Owt% higher in order to obtain a good initial net rate of flow from the impure liquid to the draw solution. This is particularly advantageous where the salt osmotic agent is the salt of a multivalent metal (e.g. divalent chloride salts such as CaCI2 and MgCI2) since osmotic potential increases more rapidly with increasing concentration for these salts than the osmotic potential of salts of monovalent metals (e.g. NaCI).

[0057] Similarly, the difference between the salt concentration in the draw solution and the salt concentration in the impure liquid may be maintained by the addition of salt osmotic agent to the draw solution. For example, the salt osmotic agent may be maintained at a concentration of at least lwt% greater than that of the impure liquid, for example at least 2wt% greater, such as at least 4wt% greater, at least lOwt% greater or at least 2Owt% greater.

[0058] In general, the initial salt concentration selected will depend on the application.

For example, in the purification of fresh water, the initial salt concentration may be selected to be greater than 0.Swt%, for example greater than lwt%. In this aspect, the initial salt concentration may be selected to be less than 3wt%, for example less than 2wt%.

[0059] For the purification of brackish water, the initial salt concentration may be selected to be greater than 3wt%, for example greater than 4wt%. In this aspect, the initial salt concentration may be selected to be less than 5wt%, for example less than 4wt%.

[0060] For the purification of sea water the initial salt concentration may be selected to be greater than 7wt%, for example greater than 8wt%. In this aspect, the salt concentration may be selected to be less than 4Owt%, for example less than 3Owt%, such as less than 2Owt% or less than lOwt%.

[0061] In these aspects, the initial salt concentration in the draw solution will reduce as water flows from the impure liquid to the draw solution. These concentration values can be maintained if desired by the addition of further salt to the draw solution during the separation process.

Superparamagnetic osmotic agent [0062] The superparamagnetic osmotic agent particles each comprise a superparamagnetic nano-particle bound to a hydrophilic polymer.

[0063] The superparamagnetic osmotic agent particles must be superparamagnetic at the temperature of the draw solution during the separation process. If the liquid temperature is too low, the osmotic agent particles will become ferromagnetic and aggregate, thereby forming a composition of agglomerated particles with a greatly reduced osmotic potential.

[0064] However, in normal use some degree of aggregation of the superparamagnetic osmotic agent particles may be tolerated while still retaining a composition exhibiting a useful osmotic potential. For example, in a composition in which aggregation of individual superparamagnetic osmotic agent particles has taken place, the number average particle diameter (or largest dimension in the case of non-spherical particles) should be less than 20 individual particle diameters, for example less than 10 individual particle diameters, or less than 5 individual particle diameters, such as less than 3 or 2 individual particle diameters.

[0065] Whether or not a particle exhibits superparamagnetism depends on the volume, temperature, and anisotropy of the particle. Therefore, for particles of a given size, whether or not those particles will be superparamagnetic at a given temperature depends on the material chosen. In cases where the volume of a particle is substantially fixed, it is possible to tune the temperature at which the particle changes from being ferromagnetic to superparamagnetic (the blocking temperature) by modifying the particle material. Correspondingly, for a particle of a given material, the blocking temperature can be modified by adjusting the particle size.

[0066] Via energy considerations, one can derive an equation relating these quantities. The volume (m3) at which a particle becomes superparamagnetic (Vp) is given by: Vp=25 kTIK, where k is Boltzman's constant (JK1), T the temperature of the particle in degrees Kelvin, and K the anisotropy constant of the material (Jm3). Using this formula, it is possible to determine the temperature at which a particle becomes superparamagnetic (the "blocking temperature") for a given material at a fixed volume.

[0067] Alternatively, for a given particle material and temperature of operation, it allows the calculation of the particle size required in order to obtain particles exhibiting superparamagnetic properties.

[0068] Since the draw solution is water based, the superparamagnetic nanoparticles preferably retain their superparamagnetic properties at a temperature in the range at which water is liquid under conditions of standard temperature and pressure i.e. 0- 100°C. The superparamagnetic particles may, for example, retain their superparamagnetic properties at temperatures of greater than 60°C, for example greater than 40°C, or greater than 20°C. In an aspect of the invention, the particles are superparamagnetic above about 0°C.

[0069] In general, the osmotic potential associated with a composition of osmotic agent particles of constant total volume increases as the size of the particles decreases. This is because the osmotic potential depends to some extent on the total surface area of the hydrophilic polymer in contact with the water in the draw solution, which increases as the particle size decreases. However, as the size of the osmotic agent particles decreases, subsequent magnetic separation becomes more difficult.

Therefore, the osmotic agent particle size selected represents a balance between maximising osmotic potential and maximising subsequent ease of separation following passage of water into the draw solution.

[0070] The diameter of the superparamagnetic part of the osmotic agent particles (or largest dimension in the case of non-spherical particles) depends on the configuration of the osmotic agent particles.

[0071] In cases where the superparamagnetic part of the osmotic agent particle forms a core particle that has the hydrophilic polymer component bound to at least a portion of its surface, the diameter of the superparamagnetic part of the osmotic agent particles (or largest dimension in the case of non-spherical particles) is preferably greater than 3nm, for example greater than 6 or greater than 8nm. In addition, the diameter of the superparamagnetic part of the osmotic agent particles (or largest diameter in the case of non-spherical particles) is preferably less than 3Onm, for example less than 2Onm, or less than l2nm.

[0072] The total diameter of the superparamagnetic osmotic agent particle will be larger than the diameter of the superparamagnetic component due to the presence of the hydrophilic polymer outer component. The diameter (or largest dimension in the case of non-spherical particles) of the superparamagnetic osmotic agent particles (including both the superparamagnetic particles and the hydrophilic polymer component) is preferably greater than 4nm, for example greater than 8nm or greater than lOnm. In addition, the diameter of the superparamagnetic part of the osmotic agent particles (or largest diameter in the case of non-spherical particles) is preferably less than 4Onm, for example less than 3Onm, or less than 22nm such as less than l5nm.

[0073] The superparamagnetic core particle size may be measured using TEM (since this technique will not detect the hydrophilic polymer). The osmotic agent particle size (including both the superparamagnetic component and hydrophilic polymer) may be measured using PCS.

[0074] The apparent particle size may vary due to variation in the hydrophilic component under different conditions of pH, ionic strength and temperature. In cases where such variation occurs, the particle size should be measured in distilled water at pH 7 under conditions of standard temperature and pressure.

[0075] In cases where the hydrophilic polymer forms a core particle with one or more superparamagnetic particles bound to its outer surface, the diameter of the superparamagnetic part of the osmotic agent particles (or largest dimension in the case of non-spherical particles) is preferably greater than mm, for example greater than 3nm or greater than 5nm. In addition, the diameter of the superparamagnetic part of the osmotic agent particles (or largest diameter in the case of non-spherical particles) is preferably less than 2Onm or less than lOnm, for example less than 8nm, or less than 6nm.

[0076] In this case, the diameter of the hydrophilic polymer particle (or largest dimension in the case of non-spherical particles) is preferably greater than 3nm, for example greater than 6nm or greater than 8nm. In addition, the diameter of the hydrophilic polymer particle (or largest diameter in the case of non-spherical particles) is preferably less than 3Onm, for example less than 2Onm, or less than l2nm.

[0077] The total osmotic agent particle size in this embodiment (or largest dimension in the case of non-spherical particles) is preferably greater than Snm, for example greater than 8nm or greater than lOnm. In addition, the total osmotic agent particle size (or largest diameter in the case of non-spherical particles) is preferably less than 32nm, for example less than 22nm, or less than l4nm.

[0078] The superparamagnetic particle size may be measured using TEM (since this technique will not detect the hydrophilic polymer). The osmotic agent particle size (including both the superparamagnetic component and hydrophilic component) may be measured using PCS.

[0079] The apparent particle size may vary due to variation in the hydrophilic component under different conditions of pH, ionic strength and temperature. In cases where such variation occurs, the particle size should be measured in distilled water at pH 7 under conditions of standard temperature and pressure.

[0080] The superparamagnetic particle material may be any material exhibiting superparamagnetic properties at the selected particle size and operating temperature.

A preferred superparamagnetic particle material is one that is iron based, for example magnetite (Fe304), maghemite (y-Fe203) or a mixture thereof. This is particularly preferred in cases where the purified water is for human consumption, due to the relatively low toxicity of iron-based materials. Iron based materials are also preferable due to their comparatively low cost.

[0081] Alternatively, for example in obtaining purified water not for human consumption, the superparamagnetic particle material may be a metal or alloy comprising a metal selected from aluminium, barium, bismuth, cerium, chromium, cobalt, copper, iron, manganese, molybdenum, neodymium, nickel, niobium, platinum, praseodymium, samarium, strontium, titanium, vanadium, ytterbium, and yttrium.

[0082] It is preferred that the bond strength between the polymer and the superparamagnetic particle is sufficiently strong that only a small proportion of the polymer detaches from the superparamagnetic particle during the operating life of the osmotic agent (e.g. less than 5% by weight of the polymer over more than 100 cycles of purification, for example more than 300 or 1,000 cycles of purification). The bond strength between the hydrophilic polymer and the superparamagnetic particle (or where more than one bond is present between the polymer and the superparamagnetic particle, the sum of the bond strengths) may therefore be greater than 50 KJ mo11, for example greater than 150 KJ mol-1, such as greater than 200 KJ mol-1. The bond strengths between the hydrophilic polymer and the superparamagnetic particle is also usually less than 300 KJ mol-1.

[0083] The hydrophilic polymer may for example be attached to the superparamagnetic core particle by adsorption onto the surface of the superparamagnetic particle e.g. by chemisorption. In some aspects, the hydrophilic polymer may be attached to the superparamagnetic particle via a bidentate bond between the superparamagnetic particle and one or more carboxylate groups on the polymer.

[0084] The hydrophilic polymer may be a natural or synthetic polymer, for example synthetic.

[0085] In some aspects, the hydrophilic polymer may comprise one or more carboxylic acid groups or an alkali metal salt thereof, capable of dissociation in solution to form a carboxylate group. These groups may either be directly bound to the polymer backbone, or bound via an intermediate group (e.g. a group comprising a carbon chain of length 1-3 carbon atoms). For example, the carboxylic acid groups may be present as maleic acid groups in the hydrophilic polymer.

[0086] In order to maximise the osmotic potential, the polymer preferably comprises polar groups, optionally capable of dissociation in solution to leave residual charged groups. Suitable groups include hydroxyl (-OH), aldehyde (-CHO), carboxyl (-COOH), amines, amides, imines, phosphonates, pyridiniums, pyrrolidones and sulfonic acid (- 503H), including benzene sulfonic acid.

[0087] In order to maximise the osmotic potential associated with the use of a given polymer, it is desirable that the maximum length of the hydrophilic polymer be available to interact with the water in the draw solution. This available length may be reduced where groups capable of binding with the superparamagnetic particle (e.g. carboxylate or amine groups) in preference to other polar or charged groups on the polymer are distributed along the entire length of the polymer since the presence of these preferential binding groups can result in the polymer being bound to the superparamagnetic particle at many points along its length with a resultant reduction in the ability of the remainder of the polymer to interact with the water in the draw solution.

[0088] To overcome this difficulty, in an aspect, the hydrophilic polymer preferably comprises a binding part containing groups (e.g. amine, carboxylate or carboxylic acid groups) which bind, or are capable of dissociation in solution to form groups which bind to the superparamagnetic particle in preference to the remaining groups on the polymer and an osmotic part in which the preferential binding groups are absent but other more weakly binding polar or charged groups are present (e.g. styrene sulfonate groups).

The hydrophilic polymer is therefore predominantly bound to the superparamagnetic particle via the preferential binding groups in the binding part (e.g. amine or carboxylic acid/carboxylate groups), with the osmotic part free to interact with the draw solution, thereby nlaximising the osmotic potential associated with each superparamagnetic osmotic agent particle.

[0089] The preferential binding groups present on the binding part of the polymer may for example be amine, carboxylic acid or carboxylate groups e.g. carboxylic acid or carboxylate groups present in a maleic acid group.

[0090] The osmotic part of the hydrophilic polmer may for example comprise sulphonate groups, for example styrene sulfonate groups.

[0091] In a preferred embodiment, the hydrophilic polymer is a polyelectrolyte. The polyelectrolyte may for example be PoIy(acrylic acid), PoIy(acrylamide), PoIy(acrylamide-co-acrylic acid), PoIy(acrylic acid-co-maleic acid), PoIy(sodium 4- styrenesulfonate), Poly(diallyldimethylammonium chloride), Poly(sodium 4-styrenesulfonic acid-co-maleic acid). In a preferred aspect, the polymer is one in which the polymer comprises a binding part comprising maleic acid groups and an osmotic part comprising styrene sulfonic acid groups, for example poly(sodium 4-styrenesulfonic acid-co-maleic acid) polymer.

[0092] The hydrophilic polymer may be a copolymer in which the binding part and osmotic parts make up the polymer backbone. Alternatively, the hydrophilic polymer may be a graft copolymer in which the binding part or osmotic part is grafted onto the other part. In this aspect, the binding part may have a length of greater than 5% of the total polymer length. The binding part may also have a length of less than 25% of the total polymer length.

[0093] The hydrophilic polymer preferably binds to the superparamagnetic particle through the preferential binding groups on the binding part so that at least 70%, more preferably 80%, even more preferably 90% and most preferably about 100% of the binding between the hydrophilic polymer and the superparamagnetic particle is between preferential binding groups on the binding part on the hydrophilic polymer and the superparamagnetic particle.

[0094] Similarly, preferably at least 50%, more preferably 70%, even more preferably 90% and most preferably about 100% of the surface of the superparamagnetic particle is covered by the hydrophilic polymer, preferably contacting the superparamagnetic particle via the binding part of the polymer.

[0095] This enables the maximum number of hydrophilic polymers to bond to the particle and also reduces bonding between the groups on the osmotic part and the superparamagnetic particle, which can reduce the contribution of the hydrophilic polymer to the osmotic potential of the particle.

[0096] The molecular weight of the hydrophilic polymer selected depends on the application. In general, for a given hydrophilic polymer, as the molecular weight of the polymer increases, the osmotic potential associated with each particle also increases.

However, the increase in polymer weight also causes an increase in viscosity of the draw solution. Therefore, the molecular weight of the polymer is selected to give a balance between obtaining a high osmotic potential draw solution and keeping viscosity of the solution at acceptable levels.

[0097] It is possible to reduce the viscosity exhibited by a hydrophilic polymer of a given length by crosslinking the polymer to adjacent polymers. This makes it possible to achieve a higher osmotic potential at a given viscosity than in the case where the hydrophilic polymers are not crosslinked. A cross linked polymer may be obtained by including a proportion of a trifunctional monomer in the reaction mixture during the polymerisation stage. Any suitable trifunctional monomer may be used, for example pentaerythritol allyl ether. The degree of crosslinking may be such that at least 0.lwt%, for example at least 0.Swt% or at least lwt% of the polymers are cross linked to an adjacent polymer. The degree of cross-linking between polymers may be measured using NMR. The molecular weight of the hydrophilic polymer is typically greater than 5 kDa, such as greater than 40 kDa, greater than 100 kDa, greater than 200 kDa, greater than 300 kDa or greater than 400 kDa. The molecular weight of the hydrophilic polymer is typically less than 1,000 kDa, such as less than 800 kDa, less than 700 kDa, less than 600 kDa, or less than 500 kDa.

[0098] In cases where the hydrophilic polymer contains an appreciable degree of crosslinking, the molecular weight of the hydrophilic polymer may be greater than 5kDa. The molecular weight of the hydrophilic polymer may also be less than 1,000kDa.

[0099] In cases where the hydrophilic polymer is not crosslinked to adjacent polymers, the molecular weight of the hydrophilic polymer may be greater than 5 kDa.

The molecular weight of the hydrophilic polymer may also be less than 300 kDa.

Molecular weight may be measured using any suitable technique, e.g. by PCS.

EXAMPLES

Example I -Synthesis of I litre of Fe304 coated by citrate [00100] Magnetite particles coated by citrate were prepared as follows.

Starting materials: 47g FeCI3 and 27g FeCI2 imi conc. HCI 30g citric acid lOOmI conc. NH3 [00101] The following process steps were used: (i) Acidify Fe2/Fe3 with HCI while dissolving in 200m1 water; (ii) Filter dissolved mixture through 0.1.tm or 300K MWCO filter; (iii) Purge solutions with N2 for 15 mins; (iv) Heat to 8000; (v) Add NH3 rapidly (by pouring); (vi) Stir vigorously (300rpm overhead); (vii) Maintain temperature/stirring for 1 hour; (viii) Heat to 950; (ix) Add 30g citric acid solution (in -50m1 water); (x) Continue heating/stirring for another hour; (xi) Allow to cool; (xii) Separate aggregated material with a string permanent magnet; (xiii) Decant supernatant; (xiv) Repeat the above process steps twice.

[00102] A Transmission electron microscopy (TEM) image of citrate coated magnetite nanoparticles is shown in Figure 1.

Example 2 -Synthesis of I litre of Fe304 coated by poly(styrene sulfonic acid)-co-(maleic acid)polymer [00103] Magnetite particles coated by poly(styrene sulfonic acid)-co-(maleic acid)polymer were prepared as follows.

Starting materials: 47g FeCI3 27g FeCI2 imI conc. HCI 30g poly(styrenesu Ifonate-co-maleic acid) lOOmI conc. NH3 [00104] The following process steps were used: (i) Acidify Fe2/Fe3 with HCI while dissolving in 200ml water; (ii) Filter dissolved mixture through 0.1 im or 300K MWCO filter; (iii) Purge solutions with N2 for 15 mins; (iv) Heat to 80°C; (v) Add polymer solution or powder; (vi) Add NH3 rapidly (by pouring); (vii) Stir vigorously (300rpm overhead); (viii) Heat to 95°C after mixing; (ix) Continue heating/stirring for 1 hour.

[00105] A transmission electron microscopy (TEM) image of poly(sodium 4-styrenesulfonic acid-co-maleic acid) coated magnetite nanoparticles is shown in Figure 2. It may be seen that the co-polymer coated particles seem to be better stabilized than the citric acid coated particles as seen in Figure 1. Without wishing to be bound by theory, this may be due to greater steric stabilization i.e. the polymer coating is thicker than the citric acid coating and physically prevents the particles from approaching each other too closely.

Example 3 -Osmotic separation (static) [00106] A simple apparatus for assessing the osmotic behavior of a material is illustrated in Figure 3. A filter unit, housing a circular piece of semi-permeable membrane (the dashed line in the diagram) is used to contact water in a dish, to the osmotic material being tested. Water is drawn across the membrane by the osmotic agent, which dilutes and increases in volume. The increase in volume is measured by the rise of the osmotic agent up the tube (designated "H" in the diagram).

[00107] The results obtained from magnetoferritin and the poly(sodium 4-styrenesulfonic acid-co-maleic acid) polymer coated material are shown graphically in Figure 4 below.

[00108] The performance of the poly(sodium 4-styrenesulfonic acid-co-maleic acid) polymer coated material can be seen to be about 40% better than magnetoferritin in respect of the water drawn in one day.

Example 4 -Osmotic separation (dynamic) [00109] In order to keep the concentration gradient across the membrane at a maximum so that the passage of water across the membrane is also maximised, the solution at both sides of the membrane was constantly refreshed using two peristaltic pumps using an apparatus as shown in Figure 5.

[00110] The use of a peristaltic pump to constantly refresh the solutions at both sides of the semi-permeable membrane maintains the osmotic potential at its maximum possible value.

[00111] A test solution of 250 ml of water was placed in the apparatus and contacted through a semipermeable membrane with a solution of the osmotic agent (65ml 2Owt% Fe304 (average diameter 9-lOnm) coated with 4-styrene sulfonic acid-co maleic acid - 3:1 ratio, molecular weight 20K). Over a period of operation of 10 hours, nearly all the 250ml of water was drawn through the membrane by the osmotic agent.

Example 5 -Data for various osmotic agents A selection of hydrophilic polymers were dissolved in water and tested for their ability to draw water across a semi-permeable membrane in a static cell. The volume drawn across the membrane in 1 hour was measured in each case and then normalised to a concentration equivalent to 2Owt%. The results are shown below.

Polymer Wt % Mw (kD) lhr Adjusted for polymer Draw Concentration ___________________________________________ __________________ ______________ (ml) (ml) Sucrose (comparative 10 0.342 0.95 1.90 example) ___________ _________ _________ ________________ Sodium Chloride 10 0.058 3.50 7.00 Ferritin (comparative 15 450 0.10 0.13 example) ___________ _________ _________ ________________ Poly(acrylic acid) 10 450 0.10 0.20 Poly(acrylamide) 50 10 0.90 0.36 Poly(acrylamide-co-acrylic 20 200 0.70 0.70 acid) ___________ _________ _________ ________________ Poly(acrylicacid-co-maleic 20 50 1.10 1.10 acid) ___________ _________ _________ ________________ Poly(sodium 4-20 70 1.60 1.60 styrenesulfonate) _____________ __________ __________ _________________ Poly(diallyldimethylammonium 20 150 1.00 1.00 chloride) ___________ _________ _________ ________________ PoIy(acrylic acid-co-maleic 50 3 1.00 0.40 acid) ___________ _________ _________ ________________ PoIy(sodium 4-styrenesulfonic 20 20 1.60 1.60 acid-co-maleic acid) ___________ _________ _________ ________________ The two polymers found to give the best results were Poly (sodium 4-styrenesulfonate) and Poly(sodium 4-styrenesulfonic acid co-maleic acid). These polymers are comparable in efficacy to sucrose (a commonly used osmotic agent).

Claims

CLAIMS1. A method of obtaining purified water from an impure water source comprising the steps of: (i) contacting the impure water source through a semi-permeable membrane with a higher osmotic potential draw solution containing a field separable osmotic agent, said osmotic agent comprising one or more ionic salts and/or a superparamagnetic nano-particle bound to a hydrophilic polymer; (ii) maintaining the contact for a time sufficient for a net flow of water to take place from the impure water source into the draw solution; and (iii) carrying out magnetic and/or electric field separation of the field separable osmotic agent from the draw solution to obtain purified water.
2. The method of claim 1, wherein the field separation is a magnetic field separation.
3. The method of claim 1, wherein the field separation is an electric field separation.
4. The method of claim 3, wherein the electric field separation is carried out using capacitive deionisation.
5. The method of claim 4, wherein the capacitive deionisation is carried out using electrodes having a specific surface area of greater than 1 00m2/g.
6. The method of claim 5, wherein the capacitive deionisation technique is carried out using electrodes having an electrical resistivity of less than 40 mO.cm.
7. The method of any one of claims 5 or claim 6, wherein the electrode is made of nanoporous carbon.
8. A superparamagnetic osmotic agent comprising a superparamagnetic nano-particle bound to a hydrophilic polymer.
9. The superparamagnetic osmotic agent of claim 8, wherein the superparamagnetic nano-particle is iron based.
10. The superparamagnetic osmotic agent of claim 9, wherein the superparamagnetic nano-particle is magnetite (Fe304), maghemite (y-Fe203) or a mixture thereof.
11. The superparamagnetic osmotic agent of claim 8, wherein the superparamagnetic part of the superparamagnetic osmotic agent forms a core particle with the hydrophilic polymer bound to its surface.
12. The superparamagnetic osmotic agent of claim 11, wherein the diameter (or largest dimension in the case of non-spherical particles) of the superparamagnetic osmotic agent particles is in the range from 4nm to 4Onm.
13. The superparamagnetic osmotic agent of claim 8, wherein the hydrophilic polymer forms a core particle with one or more superparamagnetic particles bound to its surface.
14. The superparamagnetic osmotic agent of claim 13, wherein the diameter (or largest dimension in the case of non-spherical particles) of the superparamagnetic osmotic agent particles is in the range from 5nm to 32nm.
15. The superparamagnetic osmotic agent of claim 8, wherein the bond strength between the hydrophilic polymer and the superparamagnetic particle is in the range from 50 KJ mo11 to 300 KJ mo11.
16. The superparamagnetic osmotic agent of claim 8, wherein the hydrophilic polymer comprises polar groups.
17. The superparamagnetic osmotic agent of claim 16, wherein the polar groups are selected from hydroxyl (-OH), aldehyde (-CHO), carboxyl (-COOH), amine, amide, imine, phosphonates, pyridinium, pyrrolidons and sulfonic acid (-SO3H).
18. The superparamagnetic osmotic agent of claim 8, wherein the hydrophilic polymer comprises a binding part containing groups which bind to the superparamagnetic particle in preference to the remaining groups on the polymer and an osmotic part in which the preferential binding groups are substantially absent but other more weakly binding polar or charged groups are present.
19. The superparamagnetic osmotic agent of claim 18 wherein the preferential binding groups present on the binding part of the polymer comprise carboxylate and/or amine groups.
20. The superparamagnetic osmotic agent of claim 18 wherein the osmotic part of the hydrophilic polmer comprises sulphonate groups.
21. The superparamagnetic osmotic agent of claim 20 wherein the sulphonate groups are present as styrene sulfonate groups.
22. The superparamagnetic osmotic agent of claim 8, wherein the hydrophilic polymer is a polyelectrolyte.
23. The superparamagnetic osmotic agent of claim 22, wherein the polyelectrolyte is selected from poly(acrylic acid), poly(acrylamide), poly(acrylamide-co-acrylic acid), poly(acrylic acid-co-maleic acid), poly(sodium 4-styrenesulfonate), poly(diallyldimethylammonium chloride), poly(acrylic acid-co-maleic acid), poly(sodium 4-styrenesulfonic acid-co-maleic acid).
24. The superparamagnetic osmotic agent of claim 23, wherein the polyelectrolyte is poly(sodium 4-styrenesulfonic acid-co-maleic acid) and the superparamagnetic particle is magnetite (Fe304).
25. The superparamagnetic osmotic agent of claim 18, wherein the hydrophilic polymer is a graft copolymer in which the binding part is grafted onto the osmotic part or the osmotic part is grafted onto the binding part.
26. The superparamagnetic osmotic agent of claim 18, wherein the hydrophilic polymer molecules are crosslinked to adjacent polymers.
27. The use as an osmotic agent of a superparamagnetic nano-particle bound to a hydrophilic polymer.
28. The use as an electric field separable osmotic agent in a forward osmosis water purification process of one or more ionic salts.
29. The use of claim 28, wherein a single salt is used.
30. The use of claim 28 or 29, wherein the salt is selected from MgCI2 or CaCI2.
31. An apparatus for use in a method as claimed in claim 1 comprising a cell for an impure water source, a cell for a draw solution containing a field separation apparatus selected from a magnetic field separation apparatus and/or a capacitive deionisation apparatus, the cell for the impure water source and the cell for the draw solution being separated by a semi-permeable membrane.
32. The apparatus of claim 31, wherein the field separation apparatus is a magneticfield separation apparatus.
33. The apparatus of claim 31, wherein the electric field separation apparatus is a capacitive deionisation apparatus.