This application claims priority under 35 U. S, C. § 1 19(e) to U.S. Provisional Application Serial No. 61 /237076 filed on August 26, 2009 entitled CATION EXCHANGE MEMBRANE and to U.S. Provisional Application Serial No. 61 /237084 filed on August 26, 2009 entitled ANION EXCHANGE

MEMBRANE both of which are incorporated by reference in its entirety for all purposes.

Embodiments of the present invention provide for ion exchange membranes and processes for their manufacture. The electrodia lysis membranes described herein combine low resistance and high permseiectivit}? ^' which make them highly effective in water desalination applications, particularly in seawater desalination. The ion exchange membranes described herein are made by polymerizing one or more monofunctionai ionogenic monomers, optionally a neutral monomer with at least one multifunctional monomer in the pores of a porous substrate.

Ion exchange membranes transport cations or anions under an electrical or chemical potential. Ion exchange membranes have either negatively or positively charged groups attached to the polymeric material making up the bulk of the membrane. The counterion of each is the transferable ion. A cation exchange membrane will have fixed negative charges and mobile positively charged cations. Similarly, anion exchange membranes will have fixed positively charged groups and mobile negatively charged anions. Ion exchange membrane properties are controlled b}/ ^* the amount, type and distribution of the fixed ionic groups. These membranes may be described as strong acid and strong base, or weak acid and weak base membranes. Strong acid cation exchange membranes usually have sulfonic acid groups as the charged group, whereas for weak acid membranes, carboxyiic acid typically makes up the fixed charged group. Quaternary and tertiary amines respectively produce the fixed positive charged groups in strong and weak base anion exchange membranes.

Among the most important applications of ion exchange membranes are desalination of water by electrodialysis (ED), as a power generating sources in reverse electrodialysis and as separators in fuels cells. Other

applications include recovery of rnetal ions in the electroplating and metal finishing industries and various applications in the food and beverage industry.

Electrodialysis desalinates water b}/ ^* transferring ions and some charged organics through paired anion- and cation selective membranes under the motive force of a direct current voltage. An ED apparatus consists of electrically conductive and substantially water impermeable anion selective and cation selective membranes arranged as opposing walls of a cell.

Adjacent cells form a cell pair. Membrane stacks consist of many, sometime hundreds of cell pairs, and an ED system may consist of many stacks. Each membrane stack has a DC (direct current) anode at one end of the stack and a DC cathode at the other end. Under a DC voltage, ions move to the electrode of opposite charge.

A cell pair consists of two types of cells, a diluting cell and a concentrating cell. As an illustrative example, consider a cell pair with a common cation transfer membrane wall and two anion transfer membrane walls forming the two cells. That is, a first anion transfer membrane and the cation transfer membrane form the dilute cell and the cation transfer membrane and a second anion membrane form the concentrating cell. In the diluting cell, cations will pass through the cation transfer membrane facing the anode, but be stopped by the paired anion transfer membrane of the concentrating cell in that direction facing the cathode. Similarly, anions pass through the anion transfer membrane of the diluting cell facing the cathode, but will be stopped by the cation transfer of membrane of the adjacent pair facing the anode. Sn this manner, salt in a diluting cell will be removed and in the adjacent concentrating cells cations will be entering from one direction and anions from the opposite direction. Flow in the stack is arranged so that the dilute and concentrated flows are kept separate and a desalinated water stream is produced from the dilute flow.

In the ED process, material commonly builds up at the membrane surface in the direction of the electric field, which can, arid usually does reduce process efficiency. To combat this effect, Electrodialysis reversal (EDR) was developed and is the primary method of use presently. In EDR, the electrodes are reversed in polarity on a regular basis, for example, every fifteen to sixty minutes. The dilute and concentrate flows are simultaneously switched as well, the concentrate becoming the dilute flow and vice versa. In this way fouling deposits are removed and flushed out.

Once the concentration in the dilution cells falls to lower than about 2000 milligrams/ liter (mg/1), electrical resistance is at a level that power demand becomes increasing expensive. To overcome this, and to be able to produce high quality water, electrodeionization (EDI), sometimes called continuous electrodeionization (CEDI) was developed. In this method the cells are filled with ion exchange media, usually ion exchange resin beads. The ion exchange media is orders of magnitude more conductive than the solution. The ions are transported by the beads to the membrane surface for transfer to the concentrate cells. EDI is capable of producing purer water then ED at less power when the feed concentration is reduced sufficiently.

ED processes for water desalination have advantages over RO. The}/ require less pretreatment which will reduce operating costs. They will have higher product water recovery and a higher brine concentration, i.e., there is less brine to dispose.

Univalent selective or monovalent selective membranes primarily transfer monovalent ions. Monovalent selective cation transfer membranes primarily transfer sodium, potassium, etc. Likewise, monovalent selective anion membranes transfer ions such as chloride, bromide, nitrate etc. Reverse osmosis (RO) dominates the production of fresh water from

seawater by membrane processes. While electrodia lysis (ED) is used for brackish water and waste water desalination, it is generally considered too expensive for seawater use. In order to be competitive with RO, ED and EDR will require membrane resistance of less than 1 ohm-cm ² , preferabfy less than 0.8 ohm-cm ² _' and most preferably less than 0.5 ohm-cm ² . Ion permselectivity of greater than 90%, more preferably greater than 95%, and most preferably greater than 98% is desired. The membrane has to have long service life, and be physically strong and chemically durable and be low cost.

Reverse electrodialysis (RED) converts the free energy generated by mixing the two aqueous solutions of different salinities into electrical power. The greater the difference in salinity, the greater the potential for power generation. For example, researchers have studied RED using Dead Sea water and fresh or seawater. Researchers in Holland have mixed river water entering the sea and seawater. RED membranes preferably will have a low electrical resistance and a high co-ion selectivity and long service life time, acceptable strength and dimensional stability and, importantly, low-cost.

The polymer electrolyte membrane (PEM) is a type of ion exchange

membrane that serves both as the electrolyte and as a separator to prevent direct physical mixing of the hydrogen from the anode and oxygen supplied to the cathode. A PEM contains negatively charged groups, usually sulfonic acid groups, attached or as part of the polymer making up the PEM .

Protons migrate through the membrane by jumping from one fixed negative charge to another to permeate the membrane.

PE M ^" s requirements include chemical, thermal and electrochemical stability, and adequate mechanical stability and strength when swollen and under mechanical stress. Other requirements include low resistance, low or preferably no methanol transport in direct methanol fuel cells (DMFC), and low cost. Development of ion exchange membranes requires an optimization of properties in order to overcome competing effects. Ion exchange

membranes for water desalination traditionally have had to meet four main characteristics to be considered successful. These are;

Low electrical resistance to reduce potential drop during operation and to increase energy efficiency

High permselectivity - that is, high permeability to counter-ions but approximately impermeable to co-ions

High chemical stability ability to withstand pH from 0 to 14 and oxidizing chemicals

Mechanical strength The membrane must be able to withstand the stresses of being handled while being manufactured into a module or other processing device. The membrane must also have good dimensional stability in operation and not swell or shrink excessively when the fluid contacting it changes concentration or temperature.

Membrane developers have recognized that for a given chemistry used to make an ion exchange membrane, a thinner membrane would give a lower resistance and also allow more membrane area per unit volume of device. However, thinner membranes are more susceptible to dimensional change from environmental effects, such as changes in ionic concentration of the fluids contacting them or operating temperature changes. Moreover, to develop and produce defect-free membranes is more difficult for the case of thinner membranes because there is less margin of error during membrane production as there is for thicker membranes where the membrane thickness covers over defects that may occur in membrane formation.

US Patent 7,226.646 describes an ion conducting membrane comprising an ion conducting region and a non-ion conducting region. The ion conducting region is formed b}/ ^* a plurality of passageways extending through the membrane filled with ion conducting material. The passageways may be formed in the substrate sheet b}/ ^* a number of methods, such as drilling, chemical etching, or punching, to provide straight-through passageways. Preferred arrangements of passageways are square, rectangular, triangular or hexangular arrays, ϊonogenic polymers are deposited in the passageways to make up the ion conducting regions. In some embodiments a skin is bonded to the surface of the substrate, or coated onto one or both surfaces.

US Patent 7,649,025 describes a composite ion exchange membrane comprising a support membrane and ion exchange resin composition within the pores of the substrate. The ion exchange resin is a specific class of aromatic polyethers. In related US Patent 7,537,852, the porous membrane is a polybenzazole membrane.

US Patent 7,550,216 describes a composite solid polymer electrolyte membrane comprising a porous polymer substrate interpenetrated with a water soluble ion-conducting material. The porous polymer substrate comprises a homopolymer or copolymer of a liquid crystalline polymer such as such as a polybenzazole (PBZ) or polyaramid polymer. Preferred

polybenzazole polymers include polybenzoxazole (PBO), polybenzothiazole (PBT) and polybenzimidazole (PBI) polymers. Preferred polyaramid polymers include polypara-phenylene terephthalamide (PPTA) polymers. In other preferred embodiments, the polymer substrate comprises a thermoplastic or therrnoset aromatic polymer. The ion-conducting material comprises a water soluble homopolymer or water soluble copolymer of at least one of a sulfonated ion-conducting aromatic polymer.

W. L. Gore & Associates, Inc. (Newark, DE) describes, in US Patent

6,254,978, a integral composite membrane having a porous polymeric membrane impregnated with a perfluoro ion exchange material to make the micropores of the membrane occlusive and a surfactant having a molecular weight greater than 1 00 wherein the thickness of the composite membrane is less than 0.025 mm. US Patent 5,547,551 describes a composite membrane comprising expanded polytetrafluoroethylene membrane support impregnated with ion exchange material, having a total thickness of less than 0.025 mm. U.S. Patents 5,599,614 and 5,635,041 also describe composite membranes comprising microporous expanded PTFE substrates impregnated with Nation® (EJ. DuPont Wilmington DE). Gore-Select® membranes (W. L. Gore & Associates, Inc., Eikton, MD) are composite membranes comprising a microporous expanded PTFE membrane having an ion exchange material impregnated therein.

US Patent 6, 1 10,333 describes a composite membrane comprising an ion exchange polymer and a support of expanded polytetrafluoroethylene polymer having a porous microstructure of polymeric fibrils, said expanded polytetrafluoroethylene polymer being at least about 85% crystalline.

US 6,689,501 describes a composite membrane for use in a fuel cell membrane electrode assembly comprising a porous polymeric substrate and a cation exchange material impregnant partially filling the substrate such that the substrate comprises a first region having pores substantially filled with the impregnant, and a second substantially porous region. The cation exchange material covers one surface of the substrate in a dense surface layer, contiguous with the first region of the substrate. The substrate has greater than 10% residual porosity, and the composite membrane is substantially gas impermeable and has at least one substantially porous major surface. U.S. Pat. No. 5,985,942 describes composite membranes comprising a porous substrate and ion exchange materials comprising substituted trifluorostyrene polymers and copolymers.

McMaster University has two US patents related to composite membranes having polyelectrolytes or hydrogels bonded or crosslinked around porous support structural elements. US Patent 6,258,276 discloses charged membranes comprising a porous substrate and a cross-linked

polyelectrolyte or hydrogel located in the pores of the substrate. The patent discloses polymerization in the substrate pores of a monomer or a mixture of monomers with a cross -linking agent, the monomer or at least one of the monomer mixture being selected from those monomers which contain a functional group that provides an ion-exchange site and those which contain a group which is susceptible to a chemical reaction by which such functional groups are subsequently introduced to the in situ- formed polymer. Alternatively, an uncrosslinked polyelectrolyte or hydrogel may be formed in the pores of the substrate as described and subsequently crosslin ked.

US Patent 7,247,370 provides for an asymmetric membranes composed of a microporous substrate whose pores contain a crosslinked gel, preferably a hydrogel or a polyelectrolyte gel, located in pores of the substrate, where the density of the crosslinked gel is greater at or adjacent to one major surface of the membrane than the density at the other major surface so that there is a gradient of gel distribution from one major surface of the membrane towards the other major surface of the membrane.

US 5,510,394 discloses a process where a solid polymeric sheet which has been cast or extruded with a fixed percentage of a high boiling point ester plasticizer is then immersed or otherwise contacted with one or more monomers along with a small fraction of a crosslinking bifunctional monomer such as divin}/ ^* ! benzene. The monomers exchange with the high boiling point plasticizers and are polymerized within the interstices of the base films. The monomers may be ion containing monomers, or monomers which can be converted after polymerization into an ion exchange

membranes by for example, sulfonation of phenyl groups or amination by tertiary amines of chloromethyl groups attached to aromatic polymers.

WO2010/013861 describes an anion exchange composite membrane that is produced by impregnating a porous film with a styrene- based monomer, a vinylbenzene-based monomer, a crosslinking agent and an initiator, and after polymerization, functionalizing the resulting crosslinked polymer by adding ammonium ions.

Membranes having charged polymers are known. Charged membranes, usually negatively charged, are used for nanofiltration. Such membranes are made to have a high water permeability. Such membranes would not be suitable for ED as they would have high osmotic flow due to their high water permeability. This effect would give poor co-ion rejection. Membranes for fuel cells are designed to transport hydrogen ions and are not suitable for transferring larger ions. Figure 1 shows the relation between voles/ cell pair and total dissolved solids in the feed solution for the test described in Example 7.

Figure 2 shows the construction of the membrane test cell and the reference electrode.

Described herein are novel ion exchange membranes for electrodialysis having a beneficial combination of high energy efficiency, resulting in low operating costs, and high permselectivity. The membranes are useful for water desalination, and are suitable for seawater desalination.

The membranes are produced by a process comprising choosing a suitable porous substrate, saturating the porous regions of the substrate with a solution comprising a monofunctional ionogenic monomer, a

multifunctional monomer, and a polymerization initiator, removing excess solution from the surfaces of the substrate while leaving the porous volume saturated with solution, initiating polymerization by the application of heat, ultraviolet light, or ionizing radiation, optionally in the absence of

substantially all oxygen, to form a crosslinked ion transferring polymer substantially completely filling the pores of the substrate.

The niicroporous support preferably comprises microporous membranes of polypropylene, high molecular weight polyethylene, ultrahigh molecular weight polyethylene or polyvinylidene fluoride. The supports are preferably not greater than about 55 microns in thickness, more preferably not thicker than 25 microns.

Cation exchange membrane embodiments described have resistivities of no greater than about approximately 1.0 Ohm-cm ² , preferably no greater than about approximately 0.5 Ohm-cm ² . Preferred embodiments of the cation exchange membranes have perm selectivi ties of greater than about approximate!}^ 95%. more preferably greater than about approximately 99%. Preferred ionogenic monomers for the production of cation exchange membranes are 2-sulfoethylrnethacrylate (2-SEM or 2-acrylamide-2-methyi propane sulfonic acid (AMPS). Λ preferred crosslinker is

ethyleneglycolclimethacrylate.

Anion exchange membrane embodiments described have resistivities of no greater than about approximately 1.0 Ohm-cm ² , preferably no greater than about approximately 0.5 Ohm-crn- ² . Preferred embodiments of the anion exchange membranes have peraiselectivities of greater than about

approximately 90%, more preferably greater than about approximate!} ⁷ 95%. Preferred ionogenic monomers for the production of anion exchange membranes are Trimethylammoniumethylmethacrylic chloride crosslinked with ethyleneglycoldimethacrylate, or glycidyl methacrylate/ N, N- dimethylethylenediamine reaction product crosslinked with

ethyleneglycoldimethacrylate, and the crosslinked ion transferring polymer formed by polymerization of N,N,N',N',N ^'' '-pentεimethyldiethylenetriεi.mine difvinylbenzyl chloride (a quaternary salt of N,N,N',N',N"- pentamethyldiethylenetriamine and vinylbenzyl chloride) or 1 ,4- diazabicyclo[2,2,2]octane difvinylbenzyl chloride) fa quaternary salt of 1 ,4- diazabicyclo[2,2,2]octane (DABCO) and vinylbenzyl chloride) .

An embodiment of these membranes is produced by the polymerization of an one or more ionogenic monomers, a neutral monomer and a suitable crosslinker monomer. Preferred neutral monomer is hydroxyethyl acrylate and hydroxymethylmetacrylate.

The membranes described herein were developed in answer to the need for low cost electrodialysis membranes and systems for water desalination, more particularly, for low cost, energy efficient electrodialysis membranes and systems for seawater desalination. Through diligent and extensive experimentation, the inventors have found that superior cation exchange (CEM) or anion exchange membranes (AEM) are produced by the methods and procedures described herein. Faced with the problem of developing a ion exchange membranes suitable for

economically desalinating seawater, the inventors realized that a

mechanically strong membrane having low electrical resistance and high permselectivity was required. Furthermore, it was concluded that to achieve low resistance would require a thin membrane.

Good mechanical strength allows the membrane to withstand the stresses of a continuous membrane manufacturing process, and be fabricated and sealed into the final membrane-holding device or module without overt damage or hidden damage which could appear after some time of operation. In addition, mechanical strength encompasses high dimensional stability. The ED membrane must exhibit minimal variation in dimensions while working as a desalination apparatus, during cleaning, sanitizing or defouling regimes, or during shipping or while in storage. High dimensional stability to changes in ionic content or of the temperature, for example, of the fluid contacting the CEM is important because during operation variations in the distance between membrane pairs could lead to current inefficiencies. Changes in dimensions during electrodialysis could also cause stresses in the constrained ED membrane leading to membrane defects and poor performance.

Low resistance reduces the electrical energy required to desalinate and lowers operating cost. Specific membrane resistance is measured in Ohm- centimeters (Ω cm). A more convenient engineering measure is Ohm-cπi ² {Ω cm-}. Resistance may be measured by using a cell having two electrodes of known area, platinum or black graphite are typically used, with the membrane sample of known area between them in a electrotyte solution. The electrodes do not touch the membrane. Membrane resistance is estimated b}/ ^* subtracting the electrolyte resistance without the membrane from the test result with the membrane in place. The resistance may also be measured by determining a voltage vs. current curve in a cell having two well stirred chambers separated by the membrane. A calomel electrode measures the potential drop across the membrane. The slope of the potential drop vs. current curves, which may be obtained by varying voltage and measuring current. Electrochemical impedance may also be used. In this method, alternating current is applied across the membrane.

Measurement at a single frequency gives data relating to electrochemical properties of the membrane. By using frequency and amplitude variations, detailed structural information may be obtained. Herein, resistance will be defined by the methods described in the Experimental section.

Permselectivity refers to the relative transport of count erions to co-ions during electrodialysis. For an ideal cation exchange membrane only positively charged ions would pass the membrane, giving a permselectivity of 1.0. Permselectivity is found by measuring the potential across the membrane while it separates monovalent salt solutions of different concentrations. The method and calculations used herein are described in the Experimental section.

To meet these initial goals the inventors developed a type of composite ion exchange membrane in which a cross-linked polymer having charged ionic groups attached is contained in the pores of a microporous membrane substrate. The porous membrane substrate is preferably less than about approximately 155 microns thick, more preferably less than about

approximately 55 microns thick.

Substrate membranes having porosity greater than about 45% are

preferred, with those having porosities greater than about 60% more preferred. In the most preferred embodiments, the substrate membranes have porosities greater than about 70 %. Preferred substrate membranes have a rated pore size of from about approximately 0.05 microns to about approximately 1 0 microns, with a more preferred range of from about approximately 0.1 microns to about approximately 1 .0 microns. Most preferred porous substrates have a rated pore size of from about

approximately 0.1 microns to about approximately 0.2 microns. Microporous membrane supports may be manufactured from polyolefins. polyvinyl! dene fluoride, or other polymers. A class of preferred substrates comprises thin polyolefin membranes manufactured primarily for use as battery separators. A more preferred substrate class are thin battery separators manufactured from ultrahigh molecular weight polyethylene (UHMWPE).

To produce the desired ion exchange membranes, the inventors developed a method of placing the crosslinked charged polymer in the pores of the substrate by polymerizing the crosslinked polymer in these pores. The method involved saturating the porous substrate with a solution of charged monomer, multifunctional monomer, (e.g., a crosslinking agent) arid polymerization initiator. Herein we use the term ionogenic monomer to mean a monomer species having at least one charged group covalently attached. The charged group can be positively charged or negatively charged. In an embodiment, the crosslinked polymer was produced by polymerizing a multifunctional charged monomer. The Polymerization was initiated by heat or by UV light. Monofunctional monomers are monomers which have a single site for carrying forward the polymerization reaction. Multifunctional monomers have rnore than one polymerization reaction site and so can form networked or crosslinked polymers.

The following laboratory method was used to investigate formulation and process effects by producing small coupons for resistivity and

permselectivity testing. Porous membrane substrate 43mm diameter coupons were die cut. Somewhat larger discs (50mm or lOOmπi diameter) of transparent potyester sheets were also die cut. A 105mm aluminum weighing boat was typically used to hold a set of coupons. The coupons were sandwiched between two polyester film discs. First, substrate coupons were thoroughly wetted with a monomer solution to make up a test sample. This was done by adding the formulated solution to the aluminum boat, and immersing a polyester film disc with a substrate coupon layered on it into the solution so that the porous support is saturated. The saturated support was then removed form the monomer solution and placed on a piece of polyester film. Air bubbles were removed from the coupon by, for example, smoothing or squeezing the coupon with a convenient tool, such as a small glass rod, or by hand. A second polyester disc was then layered on top of the first coupon and smoothed to have complete surface contact between the coupon and the lower and upper polyester film la}^ers. A second porous substrate was then layered on the upper polyester film and the saturation, smoothing and addition of a over layer of polyester film repeated to give a multilayer sandwich of two coupons and three protective polyester film layers. A t}φical experimental run would have a multi layered sandwich of 10 or more saturated substrate coupon layers. The rim of the aluminum boat can be crimped down to hold the disc/coupon assembly if required.

The boat and assembly were then placed in a scalable bag, typically a zip- lock polyethylene bag and a low positive pressure of inert gas, usually nitrogen, added before sealing the bag. The bag containing the boat and coupon assembly is placed into a oven at 80 ⁰ C for up to 30 minutes. The bag is then removed and cooled, and the now reacted cation exchange membrane coupons are placed in 0.5N NaCl solution at 40 ⁰ C-SO ⁰ C for at least 30 minutes, with NaCl soak of up to 18 hours being found satisfactory.

Monomers containing negatively charged groups useful for making cation exchange membranes of the present invention include as representative examples, without being limited by such examples; sulfonated acrylic monomers suitable to provide cation exchange capacity; e.g., 2- sulfoethylmethacrylate (2-SEM), 2-Propylacrylic acid, 2-acrylamide-2-methyl propane sulfonic acid (AMPS), sulfonated glycidylmethacrylate, 3- sulfopropj ⁷ ! methacrylate, sodium l-allyloxy-2 hydroxypropyl sulfonate and the like: other example monomers are acrylic and methacrylic acid or their salts, sodium styrene sulfonate, styrene sulfonic acid, sulfonated

vinylbenzyl chloride sodium l-allyloxy-2 hydroxypropyl sulfonate, 4- Vinvlbenzoic acid, Trichloroacrvlic acid, vinvl phosphoric acid and vinvl sulfonic acid. Preferred monomers are 2-sulfoethylmethacrylate (2-SEM), styrene sulfonic arid and its salts, and 2-acrylamtde-2-raethy[ propane sulfonic acid (AMPS).

Monomers containing positively charged groups useful for making anion exchange membranes of the present invention include at. representative examples, without being limited by such examples;

Methacrylamidoprυpyltrimethyl ammonium chloride,

trimcthylammoniumcthylmethacrylate; quarternary salts of pυlyamines and vinylaromatic halides, for example, but limited to; 1 ,4- diazabicyclo[2,2.2]octane difvinylbenzyl chloride) (a quaternary sail of 1 ,4- diazabicyclo[2,2.2]octane (DΛBCO) and piperazine divinyl chloride; or quarternary salts formed by reacting cyclic ethers, poly amines and alkyl halides; for example, but not limited to; Iodoethyldimethylethylenediamino2- hydroxylpropyl methacrylate (a quaternary ammonium salt formed by reacting glycidylmethacrylate (GMA) with N.N-dimethylethylencdiamine and ethyl iodide, and vinylbenyltrimethylammonium chloride.

Preferred monomers for anion exchange membranes are

Trimethylammoniumethylmethacrylic chloride, 3- acrylamidopropyl)trimethylammonium chloride, 1 ,4- diazabicyclo[2,2.2]octane difvinylbenzyl chloride) fa quaternary salt of 1 ,4- diazabicyclo[2,2.2]octane (DABCO) and vinylbcnzyl chloride, N,N,N',N',N"- pentamethyldiethylenetriamine difvinylbenzyl chloride (a quaternary salt of N.N,N',N',N"-pcntamethyldicthylcnetriaminc and vinylbenzyl chloride, Glycidyl methacrylate/ trimethylamine or Glycidyl methacrylate/ N, N- dimethylethylenediamine reaction product,

Multifunctional monomers suitable to provide crosslinking with monomers containing negatively or positively charged groups include as representative examples, without being limited by such examples ethyleneglycol

dimethacrylate, 1 ,3 -butanediol dimethacrylate, 1 ,3 -butanediol di aery late, 1 ,4-butanediol dimethacrylate, 1 ,4- butanediol diacrylate, 1 ,0-hexanediol diacrylate, pentaerythritol triacrylate, tetraethylene glycol dimethacrylate, divinyl benzene, trimethylolpropane triacrylate, isophorone diisocyanate, glycidylmethacrylate, trimethylolpropane trimethacrylate, ethoxylated (n) bisphenoi A difmethjacrylate (n= 1 .5. 2, 4, 6, 10, 30), ethoxylated (n) trimethylolpropanetri(rneth)Acrylate (n= 3,6,S ⁾ , 10, 15,20), propoxylated(n) trimethylolpropane triacrylate (n =3,6), vinylbenzyl chloride, glycidyl methacrylate and the like.

Multifunctional monomers containing one or rnore ionic groups may be used. Without being limited by the example, monomers such as 1 ,4- divinylbenzene-3 sulfonic acid or its salts may be used.

The degree of crosslinking may range from 2% to 60%.

Free radical initiators useful for the present invention include, but are not limited to; benzoyl peroxide (BF ⁵ O), ammonium persulfate , 2,2 ^* - azobisisobutyronitrile (AIBN), 2,2 ^* -azobis(2- methylpropionamidinejdihydroch loride , 2,2' - Azobis[2 ^■ (2 ^■ imidazolin- 2yl)propane]dihydrochloride, 2,2'-Azobis[2- {2-imidazolin-2-yl)propane] and dimethyl 2,2'-azobis(2-methylpropionate).

A person skilled in the art of membrane development and manufacturing will realize that this convenient laboratory method can be adapted to other laboratory scaled methods and may be scaled up to continuous

manufacturing.

For example, the substrate pore filling or saturation may be done at a slightly elevated temperature (>40°C) to reduce air solubility, or this step could be done after a mild vacuum treatment of the substrate sample submerged in the formulation solution. Substrate samples may be presoaked arid then placed on the polyester or similar sheet and covered with a covering sheet and smoothed out to remove air bubbles. Several presoaked pieces may be layered and then placed on the polyester or similar sheet and covered with a covering sheet and smoothed out to remove air bubbles.

Rather than heating in an oven, the saturated substrate sandwich may be placed on a heated surface at a temperature sufficient and for a time necessary to initiate and complete polymerization. Alternate methods for initiation of the polymerization reaction may be used. Ultraviolet light or ionizing radiation, such as gamma radiation or electron beam radiation may be used to initiate the pofymerization reaction.

A continuous pilot or manufacturing process may comprise a step to saturate the porous substrate, followed by a step to initiate and complete the polymerization, followed by a step to wash or leach out non-polymerized species from the now-formed membrane. The membrane may be optionally dried. Conditioning with a salt solution could be done in a continuous immersion process, such as through a tank of a salt solution, or by soaking a wound up roll of membrane, or after fabrication into a module.

If the monomer solution is formulated with a solvent which wets out the substrate, the process would start by feeding substrate from a roll into and through a tank of the monomer formulation and wiping off excess solution. The saturated substrate then could be sandwiched between two layers of plastic sheeting fed from rolls and nipped between two rolls to remove air and produce a smooth multilayered assembly. A preferred sheeting material is polyethylene terepthalate film. This assembly would then be processed through an oven, or over a heated roll, to initiate and complete

polymerization. An alternative would be to run the saturated sheet through an oven blanketed with inert gas. This alternative would be suitable for use with solvents having a high boiling point.

UV light initiation with suitable initiators could be used. The three layer sandwich described would be run through a tunnel or other process equipment having an inlet and outlet for the substrate web with UV light sources on one or both sides of the web. With a high boiling formulation, this may be done in an inert gas atmosphere.

If the three layer sandwich is used, the covering sheets are removed after polymerization and the now-formed membrane is washed and optionally dried.

In the experiments shown in the Examples section, an organic solvent. N- methyl pyrrolidone was used as the reactant carrier. A preferred class of solvents is dipolar aprotic solvents. Without being limited, some examples of suitable solvents include dimethyl acetamicle. dimethyl formamide, dimethyl sulfoxide, hexaniethylphosphoramide or triaraide, acetone acetonitrile, and acetone. This class of solvents has the advantage of solvating ionic group containing monomers and monomers that are not water soluble.

Other solvents found useful are N-propariol arid dipropylene glycol. Similar hydroxy containing solvents, such as alcohols, for example isopropanol, butanol: diols such as various glycols, or polyols, such as glycerine may be useful in sorne cases. These are given as examples, not to be limiting to a practitioner.

The solvents discussed may be used alone or in combination. In some cases, they may be used with water to increase solubility of ionic containing organics.

This allows the membrane developer to have a broad range of monomer mixtures from which to tailor a crosslinked copolymer in order to obtain the optimum balance of properties. By combining a water soluble and/or swellable ion containing monomer with a non-water swelling comonomer, the developer can obtain a copolymer with a high degree of ionic groups and reduced swelling in water desalination use. Such copolymers will have better physical strength in water and suffer less dimensional change in use due to changes in water ionic content or temperature changes.

Earlier attempts to make ion transporting membranes with a porous substrate relied on imbibing a ion-conducting polymer into the pores of the substrate. US Patent 7,550,216 uses a water soluble ion-conducting polymer. In water such polymers are expanded, due to charge repulsion from the similar charges on the monomer units. This will hinder diffusion into the pores of the substrate, and reduce the amount of charge that can be permanently placed into the substrate. The 7,550,21 6 patent also shows that water insoluble (i.e., organic solvent solution borne) ion-conducting polymers do not produce satisfactory fuel cell membranes. The process use to make these membranes requires a very long repetitive drying/ soaking protocol, which would increase manufacturing costs. Furthermore, the dry/ soak steps are needed to fully densify the membrane, which indicates that polymer diffusion does not completely fill the porous regions.

The present inventors have gone beyond the standard requirements of an electrodialysis membrane in the present embodiments. While a

mechanically strong membrane having low electrical resistance and high permselectivity are fundamental requirements for a seawater electrodialysis membrane, other characteristics are important in order to obtain the optimum balance of properties.

The inventors of the several embodiments of the invention described herein have also developed ion exchange membranes that have reduced water permeation. Permeation of the dilute flow through membrane defects under the driving force of the osmotic pressure difference between the dilute and concentrated streams is deleterious to high efficiency. Water permeation reduces current efficiencies and purified water productivity by removing pure water. Water loss is particularly severe in seawater electrodialysis with thin membranes because the high concentration difference between the concentrate (brine) side of the membranes and the pure water side of the membrane increases the osmotic driving force. Membrane defects will be particularly detrimental to operation as the high osmotic pressure will force pure water through such defects and increase water loss and increase power consumption.

To reduce water permeation, the inventors developed several methods that 111337- be applied to the membrane production process in order to produce a membrane with minimal defects, preferably essentially no defects. In the Examples, particularly Examples 1 and 2, an apparent optimum soaking time was shown to produce the best combination of membrane properties. In Example 1 , soaking times between 6 and 24 hours gave the best combination of low resistivity and high permselectivity. In Example 2, using a different monomer formulation, soaking times of between 1 and 3 hours gave the best combination of these properties. The inventors also found that a procedure involving a double pore filling process reduced apparent defects without greatl}' affecting resistivity and but improving apparent permselectivity. One or both of these findings may be used in a scaled up process

Moreover, the crossliriked polymer used to fill the porous substrate was developed to have a structure that allowed high permeability of cations and low osmotic flow. Apparent permselectivity is the ratio of transport rate of counter-ions (cations) to co-ions (anions) but does not indicate the rate of counter-ion removal.

Cation permeability is controlled by the structure of the ion (molecular size and total charge) and by the effect of membrane microstructure. The membrane microstructure can retard counter-ion permeability if it can be thought that the membrane has pores that are "small." This qualitative term means that the counter-ions encounter high resistance from

interactions with the membrane material in traversing the membrane, as if they were traversing a tunnel slightly larger than their apparent diameter. The membrane may have a relatively low water content which will reduce the pathways for counter-ion permeability. By balancing the content of hydrophilic monomer to increase counter-ion permeability, with the amount and nature of crosslinkiiig monomer, the water content and effective pore size of the membrane can be optimized. The crosslinking monomer is usually a hydrophobic chemical, but may be chosen to be hydrophilic.

The results of Experiments 1 , 2, and 4 illustrate how differences in

formulation can affect membrane properties, in the table below some results are abstracted to show this. The formulation is given as weight percent of the reactants. Resistivity and perm selectivity are given at the soaking time which gave the best combination of properties, (Sample number in parentheses. As the %2-SEM is decreased, with a constant % of crosslinker (EGDM), the resistivity increases and permselectivity decreases, somewhat more pronounced with the APorous substrate. Note the very low resistivity and high permselectivity of Sample 24.

It is evident from the results presented here that achieving an ED

membrane for seawater use is a difficult problem. The inventors have found that certain combinations of membrane-making variables will produce membranes with resistivities of less than 1 .0 Ohra-cm2, and in some cases, less than 0.5 Ohra-cm2, and permselectivities of greater than 90%.

In Example 1. a long soaking time, 6-24 hours produces excellent

membranes with resistivities of from approximately 0.2 to approximately 1.0 and permselectivities greater than 99%. When the formulation was changed to a higher level of solvent in Example 2, the soak time to produce 99% pcrmselectivity was 0.5 - 1.0 hours, although resistivity was between 1 and 2 Ohm-crri2.

Table 8 of Example 3 illustrates the use of two different formulations for a two pore filling process. The average values and standard deviation for resistivity are 1.31 0hm-cm2 (range 0.38 - 1.72) and 0.481 and for pcrmselectivity, average 97.02% (range 95.77% - 98.08%) and standard deviation 0.918. The spread, as calculated from the ratio of standard deviation to average, is larger for resistivity than for the permselectivity, 0.395 vs. 0.0094. This indicates that low values of resistivity combined with high permselectivity are possible (See sample MB) but that it is difficult to control resistivity. Example 4 is made with a formulation having the highest level of non- reactive solvent (-52% NMP). Here the effect of soak time is not apparent, which indicates that there is a optimal combination solvent content and soak time. Since in these examples the crosslink percentage was constant, this is the same as finding an optimum 2-SEM level, for these particular processing conditions.

Example 5 shows examples of anionic exchange membranes. Table 1 1 shows examples where supports and crosslinkers were varied. The results show that the type of support plays an important role in the amount of polymer formed in the support (%weight gain). The Celgard(/C) and

Durapore(/F) supports have the highest first weight gain for each

crosslinker in all cases but one, followed by Ticoria(/T). However, the Ticoria supported membranes have very high resistivity. (Rows 6, 13, and 18} The PETA crosslinked membranes have high weight gain with the Celgard support, which strangely also had an increased resistivity after the second pore filling. Permselectivity generally increased a few percent after the second pore filling. Weight pick-up arid resistivity vary so considerably that no clear understanding of the relation between crosslinker type arid support properties was evident. These results show that considerable experimentation would be needed.

The examples in Table 12 give the results of further combinations of supports and formulations. Several have resistivities of less than 0.5 Ohm- cm2 and permselectivities greater than 90%. (several samples of row 2, and rows 3, 4,) and resistivities of less than 1.0 0hm-cm2 and permselectivities greater than 90%. (Rows 13, 32). The PMDETA/ VBC and DABCO/ VBC adduct based membranes had the highest permselectivities seen in this work for AEM's. Overall, the combination of the TMAEMC/ EDGM

formulation and the APorous support seemed to have the best consistency and a good combination of properties.

Example 6 gives results from a continuous machine run using formulations developed from previous experiments discussed above. Table 1 5 gives the formulations and substrate used. Membrane development had advanced to the stage where a 10-cell prototype module could be used for testing.

Results in standard tests are shown in Table 16.

The voltage required for these membranes at various salt concentrations is shown in Figure 1. It can be seen that the example membranes require lower voltage than the commercial membranes at the same current density, i.e., are more energy efficient. This is particularly evident at salt

concentrations below SOOOppm. The low resistivity at comparable apparent permselectivities is strong indication of a lack of random defects.

In related example, an experimental production run of one embodiment of the anion membrane of the present invention, using thinner substrate, generated eighteen samples to determine the presence of defects. Apparent permselectivity was used to measure defects in the membrane samples. The data below show a low standard deviation in apparent permselectivity, indicating the absence of random defects.

Mean apparent permselectivity = 93.75%

Standard deviation - 0.0031

Formulation: TMAEMC 58.1%, EGDM 18.1 %, DPG 16.4%, 1 - Propanol 6.7%, AlBN 0.7%; Substrate: Teklon, 35 μm, 48% porosity

The results discussed for Example 6 are riot limiting to the formulations and other process conditions described, but are related here to show one example of exceptional elect rodia lysis membranes having been developed as a result of prolonged and difficult experimental research.

Further improvement was surprisingly found (data in Table 17) by adding a non-charged neutral monomer to the polymerization formulation.

Hydroxy ethyl aery late, hydroxyethylmethacrylate and vinylbenzyl chloride are preferred neutral monomers. The inventors believe that these

monomers and similar monomers moderate phase separation arid produce a more homogeneous crosslinked polymer. Other example monomers are, for example, but not limited by these examples, methyl methacrylate, ethyl methacrylate, and hydroxypropylacrylate. Table 17 of Example 7 gives results for experiments done with several formulations using two ionogenic monomers (VBTMΛC and TMΛEMC) and hydroxyethylacryate as the neutral monomer, with divinylbenzene used as crossliriker monomer. The results show consistent values of resistivity below I and down to approximately 0.5 Ω-cm ² . Permselectivities were in the range of 92% to 95%. The preferred support was APorous H6Λ.

The results further show the importance of choosing the correct support membrane as the APorous support membrane consistently produces membranes with resistivity values less than 1.0, whereas both the Tcklon 3 and Soluporl 6P10A supports give values above 1 .

The inventors control membrane properties by balancing the amount of Crosslin king formed in the polymerized membrane. In general, a high level of crosslinking gives high perm selectivity with concurrent high resistivity. Conversely, too low a crosslinking content will give low resistivity, but low permselectivity. Without desiring to be limited by the following, the inventors also believe that several properties of the polymerization

formulation affect the properties of the resultant membrane. The use of a mϊcroporous substrate is believed to improve membrane properties by reducing the microheterogeneity formed polymer. This is believed due to the constrained crosslinked polymer formation in the porous volume, possibly enhanced by the surface effects of the hydrophobic pore walls. In the polymerization of an ionogenic monomer (hydrophilic) and a hydrophobic crosslinker. the tendency of the hydrophobic and hydrophilic regions will he to separate and aggregate. Even with f^drophilic crosslinkers, such as ethylenegfycoldimethacrylate, the inventors see evidence of phase separation when a film of solution is applied to a dense substrate film, for example, a polyethyleneterepthalate film, as the applied solution becomes cloudy as solvent is removed. The inventors believe that the improved combination of low resistivity and high apparent permselectivity is in part due to the retarded motion engendered by the microporous dimensions of the pores and possibly aided by the hydrophobic attraction of the pore walls.

The conclusion shown by these examples is that membranes with goal properties of less than 1.0 Ohm-crri2 or more preferably 0.5 0hm-cm2 and greater ihari 90% permselectivity can be made. For CEM's, permselectivities greater ihari 99% have been shown with similar goal resistivities. The experimental work shows that proper choice of support is important, but that different formulations combine differently with supports so that a general choice cannot be made. Process variables such as soaking time are also important. Λ practitioner skilled in the art of membrane development and particularly ED membrane development will be able to use these examples and the teachings herein to make membranes having goal properties with the use of standard experimental procedures and without undue experimentation

The following examples are meant to illustrate the extent of effort expended in developing the subject membranes. The finding resulted in showing that ion exchange membranes having the desired properties could be made and that improvements are possible with further experimentation. These results are meant to be illustrative and to indicate developmental directions to those skilled in the art of membrane development and associated arts and not to be limiting as to the extent of the matter disclosed herein. Properties and suppliers of the supports used are given in Table 1 below.

Λ treatment solution was formulated from an Ionic sulfonated methacrylate - 2-sulfoethyl methacrylate (2-8BM; CAS# 10595-80-9), Ethylene

glycoldimethacrylatc (EGDM; CAS# 97-90-5, with a non-reactive solvent N- methyl pyrrolidonc (NMP; CAS# 873-50-4) and polymerization free radical iniiiator 2.2-azυbisisobutyrnitrile (ΛIBN; CAS # 78-67- 1) VΛZO-64 DuPont. The formulation was made up to;

Membrane coupons were made up by the aluminum boat method described in the Detailed Description section. Soaking time, the time that the substrate was in the formulation before polymerization was initiated, was varied. Resistivity and permselectivity results are given in Table 4.

CEM (2 -SEM //EGDM) Teklon 0,25 1.345 102.4 1

CEM (2-SEM/ EGDM) Teklon 0.5 1.1 18 98.55

CEM (2-SEM/EGDM) Teklon 1 1.163 96.09* CEM (2-SEM/EGDM) Teklon 3 1.532 100.04 CEM [2 -SEM /EGDMj Teklon 48 1. 195 100.08

CEM (2-8EM/ /EGDM) APorous H6A 0.25 2.029 97.41 CEM (2-SEM/EGDM) APorous H6A0.5 1.422 100.69

CEM (2 ^■ SEM/ EGDM) APorous H6A 1 1.063 97.24**

CEM (2-SEM/ EGDM) APorous H6A3 1 .646 97.24

CEM (2-SEM/EGDM) APorous H6A48 1.169 99. 10

(all single pore -filling)

*: some white spots; **: A large white area.

A treatment solution was formulated from an ionic sulfonated methacrylate - 2-sulfoethyl methacrylate (2-SEM; CAS# 10595-80-9), Ethylene

glycoldimethacrylate (EGDM; CAS# 97-90-5, with a higher percent of non- reactive solvent N-methyl pyrrolidone (NMP; CAS# 873-50-4) and polymerization free radical initiator 2,2-azobisisobutyrnitrile (AIBN; CAS # 78-67- 1) VAZO-64 DuPont. The formulation was made up to (weight percentages);

2-SEM (gram EGDM(grams NMP(grams) VAZO

64(grams)

12.53 6.32 1 1 .94 I 0.1895

Table 6 Substrates used in example 2

Rated I Thickne: Porosity

Type ! Manufactur Material

pore siz 1 (microns (%)

Membrane coupons were made up by the aluminum boat method described in the Detailed Description section. Soaking time, the time that the substrate was in the formulation before polymerization was initiated, was varied. Resistivity and permselectivity results are given in Table 2.

1%;

CEM (2-SEM// EGDM) Teklon 0.25 1.1 15 95.19

CEM (2-SEM/ EGDM) Teklon 1 2.141 96.16

CEM (2-SEM/EGDM) Teklon 3 1.475 99.05

CEM (2-SEM/EGDM) Teklon 6 1.600 95.52

CEM (2-SEM/EGDM) Teklon 30 1.658 95.02*

CEM (2-SEM/EGDM) Teklon 48 1.102 94.70

CEM {2-SEM/EGDM) Teklon 48 1.900 93.54**

CEM (2-SEM/ /EGDM) APorous H6A 0.25 0.724 92.22

CEM (2-SEM/EGDM) APorous H6A 1 1.451 97.66

CEM (2-SEM/EGDM) APorous H6A3 1.054 96.60

CEM (2-SEM/EGDM) APorous H6A 6 1.801 95.0:

CEM (2-SEM/EGDM) APorous H6A30 0.579 90.54 * _*

CEM (2-SEM/EGDM) APorous H6A48 2.412 96.67 ***

CEM (2-SEMZEGDMj APorous H6A48 2.047 93.54

*: One white spot ; **: many white spots; ***: resin sticked on surface

This batch of 2-SEM/ EGDM/ NMP@40%NP formulation showed behavior consistent with solution instability which appeared after overnight soaking. 2-SEM (gram EGDMfgrams NMP(grams) AJ prams)

7.30 2.72 2.87 0. 104

Membr 'pe Substra te Pore fi RfC a2| App.

CEM (2-SEM/EGDM) CelgardEZ2590 1 1.294 97.87

CEM (2-SEM/EGDM) AP H6A I 1.456 99.17

CEM (2-SEM/EGDM) Solupori 6Pl OA 2 1.941 102.4

CEM (2-SEM/EGDM) Solupor 16P5A 2 2.159 102.0

A sample of Teklon (25" x25") was saturated with 186-76-2 formulation for 1.5 hours. Then a second formulation was made (186-77-2) and the saturated Teklon (25 ^" x 25") was soaked in 186-77-2 for another hour.

Then it was placed on a large piece of PET film, smoothed out to remove air bubbles, and the soaked Teklon folded over itself twice into a narrower strip. Air bubbles were again removed by smoothing the folded Teklon. The four - folded Teklon strip was then covered with another piece of PET, and any air bubbles smoothed out. Rolled the Teklon/PET sandwich up into a 4" diameter cylinder and cured in an 80° C oven for 1 hour. The 25" x 25" CEM thus made was conditioned overnight in 0.5 N NaCl.

Table 8 shows resistivity and permselectiviiy measurements of the 9 coupons die cut from the 25" X 25" CEM.

Results: Resistivity (or .2) App.

Permselectiviiy (%)

Right Top (RT): 1. .7244 97.75

Right Middle(RM): 1. .6989 95.77

Right Bottora(RB): I . .6159 9676

Middle Top(MT): 1. .3933 9808

Middle MkWIe[MM): 0 9453 96.27

Middle Bottom(MB): 0.3768 95.77

Left Top (LT): 1.5462 97.91

Left Middle(LM): 1. ,7167 97.75

Left Bottom(LB): 0, .8558 9709

Λve 1.31

97.02

Std Dev 0.481

0.918

StdDcv/Ave 0.395

0.0094

This set of samples was made with the formulation below in the usual manner. Results are shown in Table 9.

CEM (AMPS/ /EGDM) Teklon 0.25 0.6774 93.71

CEM (AMPS/ EGDM) Teklon 0.5 1.0454 89.91

CEM (AMPS/ EGDM) Teklon ] 1.1092 94.86

CEM (AMPS/ EGDM) Teklon 3 1.3873 96.92

CEM (AMPS/ EGDM) Teklon 6 1 .2506 95.19

CEM (AMPS/ EGDM) Teklon 24 1.0877 94.61

CEM (AMPS/ EGDM) Teklon 48 1.8887 94.12

CEM (AMPS/ 2- SEM /

EGDM/HOEMA*) Teklon 1 1.48 100

CEM(AMPS/ 2-SEM/

EGDM/HOEMA*) AP H6A 1 2.09 96.05

CEM (AMPS /EGDM) APorous H6A 0.25 1.4805 93.71

CEM (AMPS/EGDM) APorous H6A 0.5 0.9379 91.73

CEM (AMPS/EGDM) APorous H6A 1 1 .0487 92.22

CEM (AMPS/EGDM) APorous H6A 3 1.1669 91.15

CEM (AMPS/EGDM) APorous H6A 6 2.3513 91.73

CEM (AMPS/EGDM) APorous H6A 24 0.5015 91.31

CEM (AMPS/EGDM) APorous H6A 48 1.6426 92.78 A related experinment was conducted with a neutral monomer HOEMA (Hydroxyethylmethacrylate). The solution composition is given in Table 9A. Membrane results are given in Table IC)A.

CEM (AMPS/ 2■ SEM /

EGDM /HOEMA*) Teklon 3 .48 100

CEM(AMPS/2-SEM/

EGDM/HOEMA*) AP H6A 2.09 96.05

Λ series of experiments were conducted varying the crosslinkcr type and support used to make anionic exchange membranes. Results are shown in Table 1 1 . All formulations were 16.1% TMAEMC. 6.8%DPG, 3.6%N- propanol, 18.8% cross- linker, and 0.6% VAZO. The formulation in the Coating Formulation/ Support column is named for the cross-linker used.

A series of further experiments were conducted on severe.! supports utilizing formulations of varying monomer/ crosslinker combinations. Formulation details are given in Table 12.

Formulations for the examples of Table 12 are given below. Table 14 has the abbreviations used.

Further evidence of essentially defect-free membranes is shown in the comparison of this Example. Membranes made according to the description described herein were compared to a thicker ( 120- 160 μm) commercial electrodialysis membranes. The formulation used for these membranes is given in Table 14,

These membranes were tested in a ten ceil eiectrodialysis module at a current density 88 A/m ² . Results shown below illustrate the lower resistivity at comparable apparent permselectivities. The feed solution made up with NaCl, which has been found to be a good laboratory stand-in for seawater.

Cation, mean resistance = ^■ 0.9163Ω-cm' ^; , Standard deviation=O.42 Ω- cm ² ; average apparent permsclectivity = 104%

Anion: mean resistance = 0.813211-Cm ⁹ , Standard deviation = 0.35 Ω- cm-: average apparent permselectivity = 93.5%

Cation: resistance - -2.8 Ω-cm ² , apparent permselectivity ^~ 04% Anion: resistance ^~ -2.8 Ω-cm ⁴ , apparent perraselectivity = 104%

Further improvement was surprisingly found by adding a non-charged neutral monomer to the polymerization formulation. Hydroxyethyl acrylate and vinylbenzyl chloride are preferred monomers. The inventors believe that these monomers, and similar monomers moderate phase separation and produce a more homogeneous crosslinked polymer. Other examples of neutral monomers are, for example, but not limited by these examples, glycidylmethacrylate, methyl methacrylate. ethyl raethacrylate,

hyd roxyethy 1 m ethacrylate . hyd roxypropylmet h aery 1 ate , hydroxypropylacrylale and vinylbenzyl chloride and the like.

The membrane resistance and counter ion transport number

(permselectivity) can be measured using an electrochemical cell. This bench top experiment provides us with a very effective and quick experiment using a small piece of sample. The equipment and procedure are described here.

(1) Solartron 1280 electrochemical measurement unit

The Solartron 1280 electrochemical measurement unit enables us to apply a current between the two platinum electrodes on the two sides of the cell and to measure the voltage drop across membrane. It has 4 connectors: work electrode (WEj, counter electrode (CE), Reference electrodes (RE l and RE2). The CE and WE are used to apply current and REl and RE2 to measure the voltage drop.

(2) Reference electrodes Reference electrodes (see the insert in figure 1 ) used for membrane characterization are made in R&D lab. Vi" glass tubing is softened, then bent and drawn to the form shown. A porous plug is inserted in the tip to control solution flow to a low rate.

Silver/ silver chloride wire is freshly made for each day's testing. Λ current of 2-3 mΛ was supplied and controlled by a power supplier and an ampere meter to a platinum wire cathode and silver wire anode immersed in a 0.1 N HCl solution. After several minutes, the sliver wire starts to turn black, indicating the formation of ΛgCl layer on the surface. The solution used inside the reference electrode tubing is 1.0M KCl solution. Since the solution will leak through the porous tip, constant addition of KCl is a necessary (-every 20 min) during experiment.

(3) Membrane test cell

Figure 2 shows the detailed electrochemical testing cell 10 construction used for the experiment to measure resistance and counter ion

permselectivity of the membrane. The membranes are cut into disc using a die cutter. The reference electrodes are used to monitor the voltage drop across the testing membrane 14 and the two platinum discs 16 are used to provide a current through the membrane. The cylindrical path of the cell 10 has a cross section area of approximately 7.0 cm-

(4) Solutions

All the solutions need to be prepared with quantitative level as indicated by their significant figures. These includes 0.500N NaCl. 1.0N HCl and 1.0N NaOH (caustic, using plastic container or volumetric flask). The 0.5N Na2SO4 is used to feed the electrode compartments without evolution of chlorine gas.

(1) Resistance measurement Resistance here refers to area resistance Ω-cm ² . The measurement contains 3 steps.

(a) Set up electrode positions: Prior to a measurement, the reference

electrode horizontal positions are set. To set reference electrode position, a rigid plastic disc is used as a stand-in for the membrane. Each reference electrode is adjusted to just touch the plastic disc and their position fixed by two set screws.

(b) Mg ^a _. surg _.. ihg solution _:.. conductiyitγ: The plastic disc was then removed and the two reference electrodes moved to 1.0 cm apart b}/ ^* removing the two 0.50 mm plastic blocks. The voltage drop between the two reference electrodes is recorded at an applied a current (~ 10-50 mA} by the Solartron 1280. The distance of the 2 reference electrodes! 1.00 cm here), the current density ( 10.00 mA) and voltage (to 0.1 mV precision) used to obtain the conductivity of the solution (0.50 N NaCl typically.

(c) MgJ!^§I10ilgJILg™ _L bj^ne_i^sistaj]^: The membrane sample is then

placed by the sample slider and the voltage and current measured again. The resistance of membrane is the total resistance less the solution resistance measured in procedure (b)

(2) Counter ion Permselectivity (Transport number)

The measurement procedures are:

(a) Reference electrode position is set as described by part(a) of

Resistance measurement. The reference electrodes position may be approximate since the voltage measured in this test is theoretically

independent of the distance, but it is recommended that the position be located as reproducibly as possible.

(b) Solutions: After emplacing the sample membrane with the slider, pour 0.500N NaCl solution in the right part of the cell separated by the testing membrane and 0.250NNaCl on the left side of the cell. ) Measuring the voltage: the voltage was measured (without applying current) using a voltage meter attached to the platinum electrodes and data were entered the spreadsheet to obtain counter ion permselectivϊty.

C = conductivit}? ^' (siemens/cm)

p = resistance (ohms/ cm)

R = resistivity (ohm-cm- ² or Ω- cm ² )

A = area (cm ² )

U, V = measured voltage (mV)

I = measured current (mA)

L = distance between reference electrodes

( 1) Conductivity of the 0.500 M NaCl at 10.00 mA current and 33.1 mV measured for a reference electrode distance of 1.00 cm, the conductivity of solution:

C - ¹ - ^1' - ^L - ^{} Mcm} - 0 043^ ^/ ™

p R ^] J 33. ImK _^ ,, ₂

h \ A x 7.00cm

I lO.OmA

Arjga_resistance of the membrane calculation needs to subtract the solution resistance. For a CMX membrane with a measured potential of 38.OmV, the area resistance is:

Perniselectiviiy (transport number) of cation{-+-) or anion(-) membrane T ₊ is obtained by:

(21 ll ^ln^

Vhich rearranges to;

a

I T ₁ - 1 + VP _/ 'RT i in--M

a,

Where V is measured voltage by the reference electrodes, R is gas constant (8.314 Joule»K ¹ »mole ^) T is Kelvin temperature of solution, F is Faraday constant (96480 coulomb/ mole) and a _R and a, are concentration (activity format) of the solution on the two sides of the membrane in the cell.