DESCRIPTION

SOLID ELECTROLYTE MULTILAYE-R MEMBRANE, METHOD AND APPARATUS FOR PRODUCING THE SAME,

MEMBRANE ELECTRODE ASSEMBLY AND FUEL CELL

Technical Field

Tfcie present invention relates to a solid electrolyte multilayer membrane, a. method and an apparatus for producing the same, membrane electrode assembly and a fuel cell using the solid electrolyte multilayer membrane, in particular, the present invention relates to a. solid electrolyte multilayer membrane having proton conductivity used for the fuel cell, a method and an apparatus for producing the same, membrane electrode as sembly and th.e fuel cell using the solid electrolyte multilayer membrane .

Background Art

Recently, active research has been directed to lithium ion batteries and fuel ceils used as power sources for mobile appliances, and to solid electrolyte memioranes constituting the above t»atteries or cells. The solid electrolyte membranes are, for instance, lithium ion conductive materials and proton conductive materials.

Generally, the proton conductive material is formed in a membrane form. The solid electrolyte in the membrane forrm used as a solid electrolyte layer of the battery or the cell such as the fuel cell, and a producing method ttiereof are suggested in the following. For instance, Japanese Patent Laid-Open. Publication No. 9-320617 suggests a me-thod in which polyvinylidene f luoricϊe resin is immersed into a liquid mixture of an electrolyte and a. plastiσizer. Japanese Patent Lai-d-Open Publication No. 2001-307752 suggests a producing method of a

proton conductive membrane by synthesizing an. inorganic compound in a solution containing aromatic polymer having sulfonic acid group , ancϋ then removing the solvent . In thi s method , oxides of silicon and phosphoric acicϋ derivative are added to improve shapes and conditions of micropores . Japanese Patent Laid-Oi?en Publication No . 2002 - 231270 suggests a method for producing an ion exchange membrane by acϊding a metal oxide precursor to a solution containing ion exchange resin , and ttien casting a liquid obtained t>y hydrolysis and polycondensation of the precursor . Japanese Patent Laid-Open Publication No . 2 004 - 79378 suggests a producing method of the proton conductive membrane . First , a polymer membrane having proton conductivity is produced by a solution casting method. To produce the proton conductive membrane , the above polyme-C ¹ membrane is immersed in a water-soluble organic compound solution who se boiling point is not less than 100° C to reaclx equilibrium swe lling , and then "the water is evaporated by heating . Japanese P atent Laid-Open Publication No . 2004 - 131530 suggests a pro ducing method of a solid electrolyte membrane ^" by dissolving a compound whose msin component is polybenzimidazole having negative ionic group in an alcoho lic solvent containing tetraalkyl ammonium hydroxiLde and whose boiling point is not less than 9 O ⁰ C .

As the membrane forming method, there are a melt extrus ion method ancl a membrane casting method as well known . In the former method, tlie membrane is produced without u sing the solvent . However , the polymer is denatured due to heat ing , and impurit ies in the polymer material remains in the membrane . On the otiier hand , the latter method requires a large sizecl facility including a producing apparatus of tfcie solution whicti is called a do;pe , a solvent recovery device and the like . However, the latter method on ly requires low heating temperatizxe , and enables to remove the impurities in the polymer material . Furthermore , in

the latter method, a membrane with superior flatness and smoothness is produced compared to the membrane produced by the former method .

In Japanese Patent Laid-Open Publication No. 9-320617, the solution casting method is rejected, but the problem of impurities contained in the raw material remaining in the membrane is not solved. The producing methods disclosed in Japanese Patent Laid-Open Publications No_ 2001-307752, 2002-231270, 2004-79378 , and 2004-131530 are for small-scale productions and not for large scale manufactures . The me-fchod disclosed in Japanese Patent Laid-Open Publication No. 2001-307752 has a problem in that dispersion of complex made of a polymer and an inorganic solvent is difficult. The metJnod disclosed in Japanese Patent Laid-Open Publication No. 2002-231270 has a problem in that the membrane production process is complicated. In a method disclosed in Japanese Patent Laid-Open Publication No. 2004-79378 has a problem in that micropores are formed on the membrane by immersing the membirane in the water. As a result , the uniform membrane is not obtained. A method, for solving the above problem is not disclosed. Furtlier , the above reference cites that the method enables to produce various kinds of solid electrolyte membranes . However, conczrete disclosure is not given . In Japanese Patent Laid-Open Publication No. 2005-146018 limits the materials to be used in the method so that otheir superior materials cannot be used.

Mone of the above references disclose a method for forming a solid electrolyte multilayer membrane which imparts desired functions to the solid electrolyte membrane.

A.n object of the present invention is to provide a solid electrolyte membrane with excellent proton conductivity in a continuous membrane form with the constant equality, a methocL and apparatus for producing the same, and the membrane electarode

assembly and the fuel cell using the solicl electrolyte membxrane .

Disclosure of invention

In order to achieve the above and other objects , in a producing method for a solid electrolyte multilayer membrane of the present invention , a layered casting membrane is formed by cast ing plural dopes firom at least one casting die onto a moving support . Each of the plural dopes contains a solid electrolyte and an organic solvent and has different composition . The casting membnrane is peeled f-trom the support as a layered memfc>rane cont aining the organic solvent . At least one of the casting membnrane and the membrrane is contacted with a liquid which is a poor solvent for ttie solid electrolyte and having a lower boiling point than that of the organic s olvent . The membrane is driecl to form a solid electrolyte mult ilayer membrane .

In the above producing method , the plural dopes are a first dope and a second dope each of which has SL different combination ratio of the solid electrolyte and the organic solvent .

It is preferable to dispose a second casting die for casting the second dope downstxream from a first casting die for casting the first dope .

A thickness of ttie solid electrolyte multilayer membrane is p-rreferably in a range of 10 μ m to 2O0 μ m .

It is preferable that the organic solvent is a mixture of a fixrst component which is a compound o f a poor solvent o f the solicl electrolyte and a second component which is a compound of a good solvent of the solid electrolyte . A weight ratio o ± the firs t component with rrespect to a sum of weights of the first component and the second component is pnreferably not less than 10% and less than 100% . It is preferable that the first component contains alcohol having one to five carbons and the second component contains dimethylsulf oxide .

It is preferable that the solid electrolyte is a hycirocarbon polymer. The hydrocarbon polymer is preferably an aromatic polymer having a sulfonic acid group. The aromatic polymer is a copolymer formed of strructural units represented by general formulae ( I) , (II) and (III ) shown in chemical formula 1.

[Chemical formula 1]

^ ^. o— z— o-)- ( in )

(X is H, Y is SO ₂ , Z has a structure represented in (E) or (II) in a chemical formu-La 2, n and m satisfy 0.1 ≤ n/(m+n) ≤ 0.5) [Chemical formula 2]

-™/ ^~ V/ ^~ V~ ( I )

Further, present invention is constituted of "the solid electrolyte membrane produced by tlxe above producing method. A producing apparatus of a solid electrolyte miiltilayer membrane is constituted of a casting- device for casti_ng plural

o

dopes each of which containing a so lid electrolyte and an organic solvent of different composition from each other frrom at least one casting die onto a moving support to form a layered casting membrane , a peeling device for peeling the casting membrane from trie support as a l ayered membrane containing an organic solvent , a drying device for drying the membrane to f orm a solid electrolyte membrane , and a membrane wetting section for contacting a liquid which is a. poor solvent o» f the solid electrolyte and having a lower boiling point than that of the oarganic solvent with at least one of the casting membrane and trie membrane .

Further , the present invention includes a membrane electrode assembly constituted of the above solid, electrolyte multilayer membrane , an anode elec "trode being adhered to one side of the solid electrolyte multilayer membrane four generating parotons from hydrogen -containing substance sixpplied from outside , and a cathode electrode b> eing adhered to tine other side of the solid electrolyte multilayer membrane for synthesizing water from the protons passed ttirough the solid electrolyte multilayer membrane and a gas sixpplied from the outside .

Further , the present invention includes a fuel cell constituted of the above membrane electrode assembly, and current collectoxrs attached to ttie electrodes of the membrane electrode assemb-Ly for transmitting electrons between the anode electrode and out side and between the cathode electrode and the outside .

According t o the present invention , the solicl electrolyte multilayer membrane having a unif orm quality and excellent ion conductivity is continuously produced . In the case the membrane electrode assembly using the solid electrolyte off the present invention is used the fuel cell , the fuel cell, exerts the excellent electromotive force .

Brief Description of Drawings

Fig . 1 is a. schematic view illustrating a dope producing apparatus ; Fig . 2 is a schematic -view illustrating a membrane producing apparatus ;

Fig . 3 is a schematic -view illustrating a membrane producing apparatus of another embodiment ;

Fig . 4 is a schematic view illustrating a membrane producing apparatus of another embodiment ;

Fig . 5 is a section view of a simultaneous co-casting apparatus ;

Fig . 6 is a section view of a sequential co-casting apparatus ; Fig . 7 is a. section view of a membrane electr-ode assembly; and

Fig . 8 is an exploded section view of a fuel cell .

Best Mode for Carrying Out the Invention Embodiment s of the present invention are de scribed below in detail . However , the present invention is not Limited to the ..following embod±-ments . First , a. solid electrolyt e membrane of the present invention is descr-ibed . Thereafter , a producing method for the solid electrolyte membrane is described . [ Material ]

In the piresent invention , a polymer having a proton cϋonating group is used as a solid electrolyte to form a membrane . A method for proclucing the membrane will be described later . The polymer having the proton donating group is not particularly limited . Any known polymer used as the proton conductive material having the acid residue is preferably used, for instance , polymer compounds f ormedL of addition polymerization having the sulfonic

acid group in si_de chains, polymethacrylate havi_ng side chains of phosphoric acid groups, sulfonated poly etheir ether ketone which is a sulfonated compound of poly ether ether ketone, ..sulfonated polytoenzimidazole, sulfonated polysizlfone which is a sulfonated compound of polysuILf one, sulfonated compound of tieat -resistant aromatic polymer? compounds and so forth. As addition polymeirization polymer having sulfonated acid in the side chains, there are perfluoro sulfonic acid polymer such as typically Naf ion (registered trad. emark) , sulfonated polystyrene, sulfonated polyacrylonitrile-styrene, sulfonated polyacrylonitrile butadiene -styrrene and the like . As sulfonated compound of the ϊieat -resistant airomatic polymer compound, there are sulfonated polyimide and the like.

As preferable examples of t-he perf luorosulffonic acid, the substances disclosed in, for instance, Japanese Patent Laid-Open .Publications No. 4-366137, 6-231779 and 6-342665 are used. Especially, the substance shown in Chemical formula 3 below is preferable. In "the Chemical formula 3, m is in a range of 100 to 10000, preferably in a range of 200 to 5000, and more preferably in a range of 500 to 2000. In addition, n is in a range from 0.5 to 100 , and especially preferable in a range of 5 to 13.5. Further , 2C is approximately equal to m, and y is approximately equal to n.

[Chemical formula 3]

Preferable examples of sulfonated polystyrene , sulfonated polyacrylonitriLle styrene and sulfonated polyacrylonitrile butadiene styrene are disclosed! in Japanese Patent Laid-Open Publications No. 5-174856 and 6- 0.11834 , and the substance shown below in Chemical formula 4.

[ ChemicaX formula 4 ]

CrHiI CcHi

Preferable examples of the sulfonated compound of the heat-resistant aromatic polymer are disclosed in , for instance, Japanese Patent Laid-Open Publications No. 6-493O2, 2004-10677, 2004-345997, 2005-15541, 2002-110174, 2003-100317, 2003-55457, 9-345818, 2003-257451 and 2002- 3.05200 , and PCT Publication No. WO/97/42253 (corresponding to Japanese Patent Publication of translated version No. 2000-510511). Among tlie above, the substances shown in the above Chemical formula 1 , and those shown in Chemical f or-mula 5 and Chemical formula 6 below are especially preferable.

[ChemzLcal formula 5]

[ChemiLcal formula 6]

Particularly, in a membrane formed of the substance shown in the chemJLcal formula 1 , a membrane expansion coefficient by water absorption is compatible -with the proton conductivity. In the case n /(m+n)<0.1, the amoiαnt of the sulfonic acid groups may be too Low for forming a path for transporting the protons, that is, the proton channel. As a result, the obtained membrane may not exer-1; the sufficient proton conductivity for a practical use. In the case n /(m+n) > 0.5, the water absorption of the membrane becomes excessively higher which result in higher membrane expansion coefficient by the water absorption. As a result, the membrane is easily degraded.

The sulfonated reaction i_n the process for obtaining the

above compounds is performed through the various synthesis methods disclosed in known references. As sulfonating agents _M sulfuric acid (concentrated sulfuric acid) , fuming sulfuric acicϋ, sulfur tr±-oxide (in a gas oar liquid), sulfur trioxide complex , amidosulfiαric acid, chlorosulf onic acid and the like are used , As the sol_ vent, hydrocarbons (benzene, toluene, nitrobenzene _λ chlorobenzene, dioxetan or the like), ha.logenated alkyls ( dichloromethane , trichloromethane , dichloroethane , tetrachloromethane, or the like) and the like are used. Reaction temperature is determined in a range of -2O ⁰ C to 200° C according to activity of the sulfonating agent. In add-±tion, it is also possible to use other meth-ods . For instance , mercapto group , disulfide group or sulfonic acid group are previously introduce*! to a monomer to synthesize the sulfonated compound by oxidatioxi with an oxidizer. As the oxidizer, hydrogen peroxide, nitric aci<l, bromine water, hypochlorous acid salt, Ixypobromite salt , potassium permanganate, chromic acid or the like are used. As the solvent, water, acetic a.cid, propionic acid or the like are used. The reaction temperature in the above method is determined in a range of room temperature (for instance, 25° C) to 200° C depending on the activity of the oxidizer. In another method , halogeno-alkyl group is previously introduced to the monomer to synthesize the sulfonated compound by substitution of sulfite salt, hydrogen sulfite salt or the like. As ttαe solvent, water , alcohol, amide, sulfoxide, sulfone or the like are used. The reaction temperature is determined in a range of the room temperature (for instance 25° C) to 200 ⁰ C. It is also possible to use a mixture of two or more solvents as ttie solvent for the above sulfonation reaction . Further, it is also possible to use alky L sulfonating agen t in the reaction process to pxoduce the sulfonated compounds. Oa e of the common methods is Friedel-Crafts Reaction (see Journa.1

of Applied Polymer Science, Vol. 36, 1753-1767, 1988) using sulfone aixd. AlCl ₃ . When ttie alkyl sulfonating agent is used to carry out the Friedel-Crafts Reaction, the .following substances are usable as the solvent: hydrocarbon (benzene, toluene _^ , , nitrobenzene, acetophenon, chlorobenzen, "trichlorobenzene ox: the like ) „ alkyl halide (dichloromethane _M trichloromethane _m dichloroethane, tetrachloromethane, trichloroethane _m tetrachlorroethane or the like) or the like. The reactiori temperaturre is determined, in a range of the room temperature to 200 ⁰ C. It is also possible to use the mixture of two or more solvents .

To produce the soILid electrolyte membrane having the structure of the chemical formula 1, a dope is prepared containing a polymer whose X in the chemical- formula 1 is cation species other than a hydrrogen atom H (hereinafter referred to as a precursor) . The dope is cast onto a support and then peeled off as a membrane containing the precursor (hereinafter referred to as a precursor membrane) . By substituting the hydrogen atom H for X in the precursor membrane, tlxat is, the proton substitution, it becomes possible to produce the solid electrolyte membrane constituted of the polymer having the structure shown in the chemical formula 1.

The cation species is an atom or an atomic group whiσli generates cation(s) at the time of ionization. The ion generated from the oation species may have a valence of one or more. As the cation, alkali-metal cation, alkali eaαrth metal cation an<!l ammnonium cation are preferable in addition to proton, and calcium ion, barium ion, quaternary ammoniiim ion, lithium ion , sodium ion, potassium ion are more prefearable. The membrane obtains tlie function as the solid electrolyte even if the substitution of the hydrogen atom H for the cation species X li _¬ the chemical formula 1 is not performed. However, the protoxi

conductivity of the membrane increases as the percentage off the substitution of the hydirogen atom H for Xl (the cation species) increases. For that reason, it is espeσiaLlly preferable ttxat X is the tiydrogen atom H. It is preferable to use the solid electrolyte having the following properties. Ttte proton conductivity is preferably not less than 0.005 S/cm and more preferably aot less than 0.01 S/cm at the "temperature of, for instance, 25° C, and the relative humidity of, for instance, 70%. Further, ttie proton conductivity after immersing the membrane in 50% aqueous methanol solixtion for one day at the temperature of 18° C is preferably not less than 0.O03 S/cm, and more preferably not less than 0.008 S/cm. In particular, it is preferable that a percentage of reduction in the proton conductivity of the membrane after the immerrsion compared, to that before the immersion i_s not more than 20%. Methanol diffusivity is preferably not more than 4 x 10 ^"7 cm ² /sec, and especially preferably not more than 2 x 10 ^'7 cm ² /sec -

As the strength of the membrane, elastic modulus is preferably not less than 10 MPa, and more preferably not less than 2O MPa. Measuring methods of the elastic modulus are disclosed in a paragraph [0138] of Japanese Patent Laid- Open Publication No. 2005-104148. The above preferable values are obtained by using a tensile testing device produced by Toyo Baldwin Co. Ltd. If otheαr measuring methocL and/or other tensile testing device are used, correlation between the obtained value and the reference value obtained by using the above tensile testing device should be previously calculated.

As the durability _^ between before and after a test with time in which the membxrane is immersed in 50% methanol at a constant temperature, a percentage of a change in each of wezLght , ion exchange capacity, and methanol diff usivity is preferrably not more than 20%, and moire preferably not more than 15%. Further,

in a t est with time in hydrogen peroxicle , the percentage of the change in each of the weight , the ion exchange capacity and the methanol diffusivity is preferably not more than 20% , and more prefe-rrably not more than 10% . The volume swelling rati o of the membrane in 50% me thanol at the c onstant tempera ture is prefenrably not more than 10% and more preferably not more than 5% .

The membrane with stable water absorption ratio and. stable moisture content is preferable . It is preferable that the membrane has extreme JLy low solubility in the alcohols , water , or mixture of alcoϊiol and water to the extent thai: it is pract ically negligible . It is also prefferable that the cϋecrease of the membrane weigtit and changes in shapes and condi-tions of the membrane when the membrane is imme rsed in the above liquid is esctremely small to the extent that it is practically negligible .

The ion conductivity property of the solid electrolyte membrane is represented by an index whdLch is a ratio off the ion conductivity to the methanol transmission coefficient . The higherr the index in a certain direct ion , the higher the ion conductive property becomes in such direction . In the ttiickness direction of the solid electrolyte membrane , the ion conductivity increas es proportional to the thickness wliile the methanol permeability increases inversely proportional thereto . Accordingly, the ion conductive property of the solid electrolyte membrane is controlled by changing the thickness . In the solid electrolyte membrane used if or the fuel celL s , since the anode is provided on one side o:f the solid electrolyte membrane and the cattiode is provided on the other side thereof , it is preferable tfciat the index in the membrane ttiickness direction is larger tlian that in other cLirections . The ttiickness of the solid electrolyte membrane is preferably in a orange of

10 μm and 300 μm. If, for instance, both the ion conductivity and the methanol diffusion coefficient are high in the solid electrolyte, it is especially preferable to produce the _* membrane witti a thickness of 50μm - 200μm. If, for instance, both the ion conductivity and the methanol diffusion coefficient are low in tine solid electrolyte, it is especially preferable to produce the membrane with a thickness of 20 ^m - 100 μm.

Heat resistant temperature is preferably not less than 200° C, more preferably not less than 250 ⁰ C and especially preferably not less than 300 ⁰ C. The Ixeat resistant temperature means the temperature at which a decrease in the me.nbra.ne weight reaches 5% when the heat is increased at the measure of l°C/min. The decrease in the membrane weight cLoes not include an amount of moisture and the like evaporated from the membrane. When the solid electrolyte is formed in the membxane form and used for the fuel cell, the maximum power density thereof is preferably not less than 10 mW/crn ² .

By using the above-mentioned sol_id electrolyte, a. solution suitable for the membrane production is produced, and accordingly, the solid electrolyte membrane suitable for producing the fuel cell is produced. The solution suitable for the membrane production is, for ia stance, a solution whose viscosity is relatively low, and from, which foreign mattters are easily removed through filtration. The obtained soILution is referred to as a dope in the following descriptions .

As the solvent for the dope, an organic solvent in which the polymer, that is, the solid electrolyte is dissolved is used. For instance, aromatic hydrocarbon (for instance, benzene, toluene and the like ) , halogenated hydrocarbon (for instance, diclαlorome thane , chlorobenzene and the like), alcohol (for instance, methanol, ethanol, n-propanol, n-butanol, diethylene

glycol and the like) _w ketone (for inst ance, acetone, methyl ethyl ketone, and the like ) , ester (for inst ance, methyl acetate, ethyl acetate, propyl acetate and the Hike), ether (for instance, te-fcrahydrofuran, ethylene glycol monomethyl etlier) , and compounds containing nitrogen (N-methylpyrrolidone, N, N-<ϋimethylformamide (DMF), N, N' -dimethylacet amide (DMAc) and the like), dimethyl, surf oxide (DMSO) and the like.

As the solvent of the dope, it is also possibl_e to use a mixture in which plural substances are mixed. When the mixture is used as the solvent, it is preferable to mix the good solvent and. the poor solvent of the solid electrolyte. If "the proton substitution is caxried out in a production of the solid electrolyte membrane having the structure shown in thie Chemical fonrmula 1, it is preferable to use a good solvent and a poor solvent of the precursor of the sol-Ld electrolyte, rhe solvent and the solid electrolyte are πi-Lxed such that the solid electrolyte constitutes 5 wt . % of tlxe whole weight. Whether the solvent used is the poor solvent or tine good solvent O-f the solid electrolyte is determined by the amount of the insolubXe residues . The good solvent of the solid electrolyte in which the solid electrolyte is dissolved has a relatively high boiling point compared to the commonly used compounds. On the othexr hand, the poor solvent has a relatively low boά_ ling points compared to the commonly used compounds . Accordingly , by mixing the poor solvent to the good solvent _M the boiling point of the mixturre in which the solid electrolyte is dissolved, is lowered. As a result, efficiency and effect in removing the solvent during trie membrane production process is improved. Zn particular, the drying efficiency of the casting membrane is significantly- improved. In the mixture of the good sol_vent and the poor solvent.

larger weight ratio of the poor solvent is preferable _M concretely, not less than 10% and less than 100% is preferable, and. more preferably (weight of the good solvent) : (weight of the poor solvent) is in a range of 90:10 to 10:90. Thereby, the percentage of the component with the low boiling point increases in the total weight of the solvents. Accordingly, the drying effficiency and the drying effects are further improved during tlxe producing process of the so3_id electrolyte membrane.

As the good solvent, DMF, DMAc , DMSO and NMP are preferable. Among the above, DMSO is especially preferable in teirms of safety and relatively low boiling point. As the poor sol_vent, lower alcohol having 1 to 5 carbons, methyl acetate and acetone are preferable. Among the above, the lower alcohol having 1 to 3 carbons are more preferable. If tlie DMSO is used as the good solvent, methyl aILcohol is espec-Lally preferable in terms of excellent solubility in the DMSO.

To improve -various membrane properties off the solid electrolyte membrane, additives aare added to the elope. As the additives, antioxidant agents, fzLbers , fine particles, water absorbing agents, plasticizers, solubilizers and the like are used. A ratio of the additives is preferably in a range of 1 wt . % to 30 wt. % when the whole solid component in the dope is 100 wt.%. The additives and its mixing ratio should not adversely affect the proton conductivity. The additives will be desσxibed in the following.

As the antioxidant agent, foxr instance, compounds such as hindered phenols , monovalent or divalent sulfers , trivalent phosphates, benzophenones, bonzo-triazoles, hindsred amines, cyanoacrylates, sallicylates and. oxalic acid aixillides are preferably used. In particular, compounds disclosecl in Japanese Patent Laid-Open Publications No.8-53614, 10-101873, 11-114430 and 2003-151346 aire preferably used.

Io

As the fibers, for instance, perf luorocarbon fibers, cellulose fibers , glass fibers and polyethylene fibers are preferably used. In specific, the fibers disclosed in Japanese

Patent Laid-Open Publications No. 10-312815, 2000-231938, 2001-307545, 2003-317748, 2004-63430 and 2004-107461 are used.

As the fi_ne particles, fox: instance, titanium oxide, zirconium oxide and the like are preferably used. En specific, the fine particles disclosed in Japanese Paten ^~ t Laid-Open

Publications No. 2003-178777 and 2004-217931 are preferably used.

As the wate.tr absorbers , that i. s , the hydrophilic substances , for instance, cross -linked polyacarylate salt, starrch-acrylate salt, poval (polyvinyl alcohol), polyacrrylonitrile , carboxymethylceJLlulose , polyvinylpyrrolidone , polyglycoldialkylether, polyglycoldialkylesther, synthetic zeolite, titania gel, zirconiagel, and yttria gel are preferably used. In specif i.c, the water absorbers disclosed in Japanese Patent Laid-Open Publications No. 7-135003, 8-20716 and 9-351857 are preferably used. As the plasticizer, for ins-tance, phosphoric; acid ester compound, chlorinated paraffin, alKylnaphthalene type compound, sulfone alkylamzLde compound, oligo ether, and aromatic nitrile are preferably used. In specific, the plasticizers clisclosed in Japanese Patent Laid-Open Publications No. 2003 -288916 and 2003-317539 are preferably used.

As the soliαbilizers, substances whose boiling points or sublimation points are not less ttian 250° C are preferable, and those not less than 300° C are moire preferable.

It is also possible to add "various polymers to the dope for following objectives : (1) to eixliance the mechanical strength and (2) to increase the acid concentration in the membrane. A polymer whose molecular weight is approximately in a

1^7

range of IOOOO to 1000000 and soluble to the solid electrolyte is suitable to achieve the above objective (1). For instance, perfluoropolymer, polystyzrene, polyetlryleneglycol, polyoxetane, polyether ketone . polyether sulfone, and the polymers containing two or more structural repeating units of the above polymers are preferable. It is also possibILe to improve the solubility of the above polymer in the solid e_Lectrolyte by adding the solubilizer. As the solubilizer, the sαbstance with the boiling point or the sublimation point of not less than 250° C is preferable, and that not less than 300° C is more preferable.

A polymer having proton acid segment and the like is preferable to achieve the above objective (2) . As such polymer, for instance, perfluorosulfone acid polymer such as Nafion

(registered trademark), sulfonated polyether ether ketone having phosphoric acid in the side chain, sulfonated heat-resistant: aromatic polymex compounds such as sulfonated poly ether sulfone, sulfonated polysulfone , sulfonated polybenzimidazole and the IiKe are used. Further, it is preferable to add the above substances to the dope in a range of 1 wt. % to 30 wt .% to the wϊiole weight of the membrane.

When the obtained solid electrolyte membrane is used for the fuel cell , it is possible to add an active metal catalyst to the dope for promoting redox: reaction of the anode fuel and the cathode fuel. Since the fuel permeated into the solid electrolyte ffrom one of the electrodes is consiαmed therein without reaching the other electrode, the crossover phenomenon is prevented. Active metal cataJLyst is not particul_arly limited as long as it functions as ttie catalyst for the electrodes. However, pla. ^~ tinum or platinum based alloy is especially suitable.

[Dope Production]

Fig . 1 illustrates a dope producing apparatus 10 . The

present invention is not limited to the following method and apparatus for producing the dope . The dope prodiαcing apparatus 10 is consti-tuted of a solvent "tank 11, a hopper 12, an additive tank 15, a mixing tank 17, a tteating device 18, a temperature 5 controlling device 21, a filtrration device 22, a flash device 26, and a filtration device 27. The solvent tanlc 11 stores the solvent. The hopper 12 supplies a. solid electrolyte. The additive tank 15 stores the additive. The mixing tank 17 mixes the solvent, the solid electrolyte and the additive to form a liquid mixture 0 16. The heating device 18 heats the liquid mixture 16. The temperature controlling device 21 controls the temperature of the heated liquid mixture 16. Thereafter, the fiILtration device 22 filters the liquid mixture lβ . After the filtration, the flash device 26 controls the concentration of the dope 24. Then the 5 filtration cϋevice 27 filters "the dope 24. The dope producing apparatus IO further includes a recovery device 28 and a refining device 29. The recovery device 28 recovers tlxe solvent. The refining de~vice 29 refines tlie recovered sol~vent. The dope producing apparatus 10 is connected to a membrane producing O apparatus 33 via a stock tank 32. Valves 36 to 38 for controlling a liquid feecling amount, andpumps 41 and 42 for feeding the liquid are providecl in the dope producing apparatus 10. The positions and the number of the valves ancl the pumps are properly changed.

The dope 24 is produced in the following method when the 5 dope producing apparatus 10 is used. First, the valve 37 is opened to feed a solvent from the solvent tank 11 to the mixing tank 17. Next, tϊie solid electrolyte in the hopper 12 is fed to the mixing tank 17. The solid electrolyte may be continuously fed to the mixing tank 17 through a supplying- device which O continuously measures and supplies the solid electrolyte, or intermittently fed to the mixing tank 17 throiαgh a supplying device whiσtα measures and supplies the solid electrolyte by a

predetermined amount. Further, the valve 36 is adjusted to feed a necessary amount of additive solution from tine additive tank 15 to the mixing tank 17.

Other -than feeding the additive in the form of solution, 5 for instance , in the case the additive is liquid at the room temperature, the additive can. be fed to the mixing tank 17 in the liquid form. Further, in tlie case the additive is solid, it is possible to use the hopper 12 to feed the addit ive to the mixing tank 17. To add several additives, it is possible to dissolve O several additives in a solution and put the solution in the additive tanJc 15. It is also pos sible to use plural additive tanks . Each of the a.dditive tanks is filled with the solvation containing a different additive. Each solution may be separately fed to the mixing tank 17 through a pipe independent from each other. 5 In tlie above descrip-fcion, the solvent , the solid electrolyte and the additive are put into the mixing tank 17 in this order; liowever, the orderr is not limited to the above. For instance, a preferable amount of the solvent is fed to the mixing tank 17 after feeding the solid electrolyte to the mixing tank O 17. Further, it is not necessary to mix the addit ive in the mixing tank 13 together with the solid electrolyte and the solvent. The additive may be mixed to the mixture of the solid electrolyte and the solvent by using an inline-mixing mathod in a later process . 5 A jacket 46, a first stirrer 48 rotated fc>y a motor 47 and a second stirrer 52 rotated by a motor 51 are preferably attached to the mixing tank 17. The jacket 46 wraps arouncl the mixing tank 17 to supply a heat transfer mecLium in a space between the mixing tank 17 and the jacket 46. The temperature of the mixing tank O 17 is controlled by the heat transfer medium flovring in the space between the -tank 17 and the jacket 46. A preferable temperature range of the mixing tank 17 is from -10 ⁰ C to 55° C. The liquid

mixture 16, in which the solid electrolyte i_s swelled in the solvent , is obtained by properly selecting and rotating the first and second stirrers 48, 52. It is preferable that the first stirrer 48 has an anchor blacle, and the second stirrer 52 has an eccentric stirrer of a dissolver type.

Next, the liquid mixture 16 is transported to the heating device 18 through the pump 41. Itis preferable that a pipe through which the liquid mixture 16 passes in the heating device 18 is provided wittx the jacket. The hteat transfer medium passes through a space between the pipe and the jacket. Further, the heating device 18 preferably has a pressurizing sectJLon (not shown) to apply pressure to the liquicl mixture 16. Tlxereby, the solid electrolyte in the liquid mixture 16 is dissolved effectively and efficiently while the liquid mixture 16 is heated and/ox pressurized. Hereinafter, the method for dissolving the solid electrolyte in the solvent by heating is referred to as a heat dissolution method. In the heat dissolution method, the liquid mixture 16 is preferably heated to reach the temperature in a range of 60° C to 250 ⁰ C. Instead! of the heat-dissolution method, a cooling- dissolution method is possibly used for dissolving the solid electrolyte in the solvent. In the cooling dissolution method, the liquid mixture 16 is preferably cooled in a range of -100° C to -10° C. It becomes possible to sufficiently cϋissolve the solid electrolyte contained in the liquid mixture 16 in the solvent by properly {selecting one of the heat-dissolving method and the cooling-dissolving method.

The temperature of the liquid mixture 16 is adjusted by the temperature control device 21 to reach the xoom temperature. Thereafter, the liquid mixture 16 is filtered through the filtration device 22 to remove the foreign maLtters such as the impurities and the agglomeration. Hereinafter the liquid mixture

o

16 is referred to as the dope 24. An average pore diameter of the filter of the filtration device 22 is preferably 50 μm or less .

After the filtration, the dope 24 is transported to the stock tank 32 through the val"ve 38 and temporrarily stored, and then used for producing the membrane.

However ¹ , a method, in which the soli_d electrolyte is swelled and then dissolved in "to the solvent, requires a longer time as the concentration of the solid elect rolyte increases, which reduces the production efficiency. In such case, it is preferable to prepare the dope with the lowerr concentration of the solid electrolyte, and tlien to carry out a concentration process to ot>tain the intendecl concentration. For instance, the dope 24 filtered through the filtration device 22 is transported to the flash device 26 throucj-h the valve 38, and a part of the solvent contained in the dope 24 is evaporated to concentrate the dope 24. The concentratetϋ dope 24 is transported from the flash device 26 to the filtration device 27 through the pump 42. At the filtration, the temperature of the dope 24 is preferably from 0 ⁰ C to 200 ⁰ C. The impurities of the dope 24 are removed through the filtration device 27. Thereafter, the dope 24 is transported "to and temporariLy stored in the stock tank 32, and then used for; the membrane production. Note that the foams may ^¬ be formed in the concentrated elope 24. It is preferable to perform. processing to remove the foams prior to transporting the concentrated, dope 24 to the fϋtration device 27. It is possible to apply known methods, for instance, an ultrasonic irradiation method in which the ultrasound is irradiated to the dope 24 for removing the foams . Further ¹ , the solvent vapor generated by the flash- evaporation in the flash device 26 is condensed to liquid and recovered by the recovery device 28 having a condenser (not

shown) . The recovered solvent is refined as the solvent to be used for the dope production -Ln the refining device 29 and reused . Such recovery and refining are advantageous to reduce production cost and also prevent adversely affecting human health and environment toy virtue of the closed system.

By using the above methods , the dope 24 wtαose concentration of the solid electrolyte or that of the precixxsor is in a range of not less than 5 wt . % and not more than 5Owt . % is produced . The concentration of the solid electrolyte or that of the precursor is more preferably in a range of not less than 10 wt. % and not more than 40 wt.%. Further, the concentration of th.e additive is preferably in a range of not less than 1 wt.% and not more than 30 wt . % when the whole solids contained in the dop e 24 is considered to be 100 wt.%. [Membrane production]

Hereinafter the method for producing the solid electrolyt e membrane is (ϋescribed. Fig. 2 is a schematic view illustratin_g a membrane producing apparatus 33. The present invention is not limited to tne following metϊiods and apparatuses for producing the solid electrolyte membrane . The membrane puroduσing apparatus 33 is provided with a filtration device 61, a casting chambevr 63, a tenter cϋevice 64, an edgre slitting device 67, a first liqui_d bath 65, a second liquid batn 66, a drying chamber 69, a cooling chamber 71, a. neutralization, device 72, a knurling roller pai_r 73 and a winding device 76. Tlie filtration device 61 removes th.e impurities from the dope 24 transported from the stock tank 32. Thereafter, from the casting _/ chamber 63, the dope 24 is cast to form a solid electrolyte membrane (hereinafter referred to a_s a membrane) 62. The tenter device 64 dries the membrane 62 whiLe holding the b>oth side edges of the membrane 62. The edge slitting device 67 cuts off the side edges of the membrane 62. Then, thte membrane 62 is immersed in the first liquid bath 65 and the second

liquid bath 66. In the drying chamber 69, "the membrane 62 is bridged across plural rollers 68 and dried while the membrane 62 is being transported by tine rollers 68. In "the cooling chamber 71, the membrane 62 is cooled. The neutralization device 72 reduces the charged voltage of the membrane 62. The nurling roller pair 73 embosses the side edges of ttie membrane 62. Tlie winding clevice 76 winds the membrane 62.

A stirrer 78 rotated t>y a motor 77 is prrovided in the stock tank 32. By using the stirrers 78, precipitation and agglomeration of the solids in the dope 24 are prevented- The stock tank 32 is connected to the filtration device 61 thrrough a pump 80. T"2ie average diameter of the filter used in the filtration device 61 is preferably lOμm or less. Thereby, impurities causing degradation of initial performance of proton conductivity a.nd degradation of proton conductivity with time are prevented from being mixed into the dope 24. Note that impurities such as insoluble contents are visually identified by emitting Ug-lit from a fluorescent lamp to a sample dope taken from the stock tank 32. The. casting chamber 63 is provided with a casting die 81 for casting the dope 24, and. a belt 82 which, is a support bei_ng transported. A precipitation hardened stainless steel is preferable for the material of the casting die 81. The mater-Lai preferably has a coefficient of thermal expansion at mo st 2 x 10 ^"5 (° C ^"1 J. Further, tlie material witli the almost saαne anti-corrosion properties as SUS316 in corrosion examination in electrolyte solution can also be used. Further, the material b_as the anti- corrosion properties which do not form pitting (holes) on the gas-liquid interface after having been dipped in a liqixid mixture of dichloromethane , methanol and water for three month-s . Further, it is preferable to manufacture the casting die 81 by grinding the material which passed more than a month after

casting. Thereby, the dope 24 flows inside -the casting die 81 uniformly. Accordingly, streaks and the Like in a casting membrane 24a are prevented, as will be descnribed later. It is preferable that the finish precision of a contacting surface of the casting die 81 to the dope 24 is lμmor Less of the surface roughnesss and the straightness is 1 μm/m or less s in any direction . Clearance of the slit of the casting die 81 is automatically controlled in a range of 0.5mm to 3.5mm. A portion of the lip end of tϊie casting die Sl contacting the dope is processed so as to have a constant chamfered radius R at 50/ii or less throughout the width off the casting die 81_ . Preferably, the casting die 81 is of a coat-hanger type.

A -width of the casting die 81 is not limited. However, the width off the casting die 81 is preferably in a range between 1.1 times and 2.0 times larger than a width of "the membrane as an end product . Further, it is preferable to install a temperature controlling device 21 to the casting die 81 for maintaining a predetermined temperature of the dope 24 dur-ing the production of the membrane. Further, the casting die 81 is preferably provided! with bolts (heat bolts) at predeterrmined intervals in the width direction of the casting die 81 for adjusting the thickness of the membrane, and an automatic thickness control mechanism which adjusts clearance of the sli"t by using the heat bolts. Zn the membrane production process, 3_t is preferable to set a profile according "to the flow volume of the pump 80 based on the previously set program. To accurately control the amount of the elope to be transported, the pump SO is preferably a high-precision gear pump. Further, in the membrane producing apparatus 33, it is also possible to carry out a feedback control based on an adjustment program according "to a profile off a thickness gauge, for instance, an infrared traickness gauge (not shown) . The casting die Sl whose slit opening of the lip end is

A I

adjust able within a range of ± 50 μ m is preferably used so as to maintain a difference in the thickness between two arbitrary positions on the membrane 62 not more than l μ m except for "the side edges of the membrane 62 as the end product . It is more preferable that lip ends of the casting die* 81 are provided with a hardened layer . Met hods for forming the hardened layer are not particularly limite d . For instance , th ere are methods such as ceramic coating , harcL chrome plating, and nitricLing treatment . If the ceramic is usecL as the hardened la/yer , the ceramic which is girindable but not friable , with a lo-wer porosity and the good corrosion resistance is preferred. "The ceramic without affinity for and adheren ce to the casting <3.ie 30 is preferable . For instance , as the ceramic , tungsten carb ide , AI ₂ O ₃ , TiN , Cr ₂ O ₃ and the like can be used, and especially tungs ten carbide (WC ) is preferable . A WC coating is performed in a thermal spraying method .

The dope discharged to the lip end of the casting die> 81 is partially dried and becomes solid. In order to prevent s uch solidification of the dope , a solvent supplying device ( n ot shown ) for supplying the solvent to the JLip end is prefera_bly disposed in the proximity of the lip encϋ . The solvent is preferably supplied in a peripheral area off a three-phase cont act line on which the lip ends contacts with, the casting bead and the outside air . It is preferable to supply- the solvent in a ra_nge from O . I mL/min to 1. 0 mL/min to each of the bead edges so> as to prevent the foreign matters such as impurities precipita_ted from the dope or those outside the dope from being mixed in the casting membrane 24a. I t is preferable to use a pump with, a pulsation of 5% or less for supplying ttie solvent . The belt 82 below the casting die 81 is bridged across the rollers 85 and 86 , and is continuously transported by the driving and rotating of at least one of the rollers 85 , 86 .

Ttie width of the b>elt 82 is not particularly limited. However- , the width is preferably in a range of 1.1 times to 2.0 times larger than the casting width of the dope 24. Further „ the length of the belt 82 is preferably 20m - 200m. The thickness of the t>elt 82 is preferably 0.5mm - 2.5mm. Further the beit 82 is preferably polished such that the surface roughness is 0. O5 μ m or less .

Material of the belt 82 is not particularly limited. However- , it is preferable to use a plastic film which is insoILuble to the organic solvent in "the dope 24. The material of the plastic film is preferably nonwoven plastic fabric; made of polyethylene terephtlialate (PET) film, polybutylene texrephthalate (PBT) film, nylon 6 film, nylon 6, 6 film, polypropylene film, polycarbonate film, polyimide film and the like. The be it 82 of a long length is preferable. It is preferable that the belt 82 has cheinical stability against the solvent used. It is also preferable that the belt 82 is heat-resistant to endure tlie temperature during the membrane production. Note that it is also possible to use a stainless support witri the long lengtrα. In order to keep surface temperaturres of the rollers 85,

86 at predetermined valixes, it is prefer ¹ able to attach a heat transfer medium circulating device 87 to each of the rollezrs 85 and 86. In this embodiment, a passage (not shown) for the heat transfer medium is formeci in each of the nrollers 85 and 86 . The temperatures of the rollers 85 and 86 are kept at the predetermined values by passing the heat transfer media ke;pt at the predetermined temperatures through the passages . The suirf ace temperature of the belt 82 is properly set according to the type of the solvent, the type of the solid component, the concentration of the dojpe and so forth.

Instead of the rollers 85, 86 and tlie belt 82, it is also possible to use a casting drum (not shown) as the support . In

this case, it is preferable that the casting rotates with a. high precision such that the variation in the rotation speed is 0.2% or less _ It is preferable that the polishing is made suclx that a surface roughness is 0.01 μm or less. It is preferable that the surface of the casting is hard chrome-plated which offers sufficient corrosion resistance and hardness . It is preferable to minimize the surface defect of the casting drum, the bslt 82 and the rotation rollers 85, 86. Concretely, the number of pin holes wnose diameter is 30 μm or more j_s preferably zero . The number of pinholes whose diameter is not less than 10 μm ancϊ less than 30 ^m is preferably 1 or less per Im ² - The number of pinholes whose diameter is less than 10 μm is prreferably 2 or less s per Im ² .

Further , a decomprression chamber 9O is preferably provided in the proximity of the casting die 81 forr adjusting the pressure in the upstream area from the casting bead in the support moving direction. The casting bead is formed between the casting die 81 and -the belt 82.

In the proximity of the belt 82, air blowers 91-93 and an air shielding plate 94 are provided. The air blowers 91-93 blow air onto the casting membrane 24a to evaporate the solvent. The air shielding plate 94 prevents the air which may damacje the surface of the casting membrane 24a from blowing onto the casting membrane 24a. In the casting chamber 63 , a temperature controlling device

97 and a. condenser 98 are provided. The temperature controlling device Sl keeps the temperature inside the casting chamt>er 63 at the predetermined value. The condenser 98 condenses and recovers the solvent vapor. A recovery device 99 is provided outside the casting chamber 63. The recovery device 99 recovers the condensed and liquefied organic solvent.

An area indicated by a short dashed line in Fig. 2 _r that

is , the area from the downstream from the casting chamber? 63 to the cooling chamber 71 i s a drying area 1 OO for drying the membrane 62 . A -transporting section 101 is provided downstream fr-om the casting chamber 63 . In the transporting section , an air ^" blower 102 is provided for bl owing dry air onto the membrane 6 2 . The membrane 62 passed thorough the transporting section ILOl is transported to the tenter device 64 in which the membrane 62 is stretched in the width direction while both side edges arre held by memlDrane holding members such as cl_ips 64a or pins . T"he dry air is introduced to t he tenter devic e 64 to dry the membrane 62 . No te that it is preferable to separate inside the "tenter device 64 into different temperature zones to adjust the drying condit ions . After passing the tenter device 64 , the membrane 62 is transported to the edge slitting device 67 . In the edge slitti-ng device 67 , a crusher 103 is provided for crushing the side edges cut off from the membrane 62 into chips .

The membrane 62 whose side edges are cut off and removed is transported to the first and secon d liquid baths 65 and 66 through guide rollers 65b , 65σ , 66b and 66c . A first liquid 65a is put in the first liquid bath 65 . As the first liquid 65a , a liquid soluble to and having a lower boiling point than the solvent of the dope 24 and in which the polymer is insolixble is used . It is possible to use a liquid mixture containing l iquids other than the above JLiquid . It is preferable that the temperature of the first liquid 65a in the first liquid t>ath 65 is adj ustable in a range of 10° C-150* C .

A second liquid 66a is put in tlxe second liquid ba_th 66 . As the second liquid 66a , a similar li_quid to the first .liquid 65a is used. It is als o possible to use the liquid mixt ure in the same manner as the first liquid 65a . Zn addition , as the second liquid 66a , it is preferable to use the Liquid whose boiling point is lower than that of the first liquid S 5a . It is preferabl_e that

the temperature of tlie second liquid 66a in the second liquid bath 66 is adjustable in a range of .LO ⁰ C -150 ⁰ C. Positions of the first and second, liquid baths 65, 66 are not particularly limited. In the case -where the first and the second liquid baths 65 anc3.66 are disposed at positions otluer than those immediately before the drying chamber 69 of a drying area 100, it is possible to dispose a drying chamber or the lifce for drying the membrane 62 at a position, for instance, immediately after tlie second liquid bath 66. In the drying chamber 69, an absorbing device 106 is provided. The absorbing device 106 absorbs and recovers the solvent vapor evaporated from the membrane 62. In Fig. 2, a cooling chamber 71 is provided downstream from the dryin g chamber 69. It is also possible to provide a hu.midifiσation chamber (not shown. ) between the drying chamber 69 and the cooling chamber 71. The h/umidification cliamber adjusts ttxe moisture content in the membrane 62. The neutralization device 72 is a neutralization bar oxr the like which controls the charged voltage of the membrane 62 in. a predetermined range (for instance from -3kV to +3kV) . The neutralization cϋevice 72 is disposed downstream from the cooling device 71 as an example. The position of the neutralization device 72 is not limited to that illustrated in Fig. 2. The knurling xoller pair 73 embosses the both side edges of the membrane 62. Inside the windingr device 76, a winding roll 107 and a press rol-Ler 108 are provicϋed. The winding roll 107 winds up the membrane 62. The press roller 108 contrrols the tension of the membrane 62 at the time of winding.

Next, an example of a method for producing the membrane 62 using the membrane producing apparatus 33 is described in the following. The dope 24 is kept unifoarm by the rotation of the stirrer 78. It is possible to add various additives to the dope 24 while the dope 24 is being stirred.

The dope 24 is transported to t lie stock tank 32 . Until the casting of the dope 24 , the precipitation and the aggl omeration of the solids are prevented by stirrring the dope 24 . Then , the dope 24 is filtered through the filtration device 63_ so as to remove the foreign matters whose panrticle size is lirger than a predetermined siz e and those in a gel-form .

Then the dope 24 is cast onto tϊie belt 82 from ttxe casting d-Le 81 . It is preferrable that the rollers 85 and 86 are driven so as to adjust the "tension of the belt 82 in a range of 10 ³ N/m and 10 ⁶ N/m. It is pxreferable to adjxist the relative positions of: the rollers 85 and 86 or the rotation speed of at least one off the rollers 85 and 86 . Moreover , a relative speed difference between the belt 82 and the rollers 85 and 86 are ad justed to be 0 . 01 m/min or less . Preferably, speed fluctuation o> ± the belt 82 is 0 . 5% or less , and meandering thereof caused in a width direction while the ^" belt 82 makes one rotation is 1 . 5mrn or less . In. order to control, the meandering, it is more pref erable to prrovide a detector ( not shown ) and a position contro ller ( not stiown ) to perform feedback control of the position off the belt 82 . The detector det ects the positions of both sides o-f the belt 82 . The position controller adjusts "the position of ttie belt 82 according to a measurement value of the detector . With respect to a portion of the belt 82 located directly below tlxe casting diLe 81 , it is preferable that vertical positional fluctuation caused in association with the rotation of the belt 82 is adjusted to be 200 Mm or les s . Further , it i s preferable that the temperature in the casting chamber 63 is adjusted with-in a range of: - 1O ⁰ C to 57° C by the temperature controlling device 97 . The solvent vapor in th.e casting chamber 63 is collecte όl by the recovery device 99 and is recycled and reused as the dope for prreparing the solvent .

The casting bead is formed between the casting αlie 81 and

the belt 82 , and the casting membrane 24a is formed on the belt 82 . In order to stabilize the casti_ng bead , it is preferable that the upstream area from the casting bead in the transporting direction of the casting die 81 is decompressed by the decompression chamber 90 to achieve a predetermined pressure value . Preferably , the upstream area, from the casting bead is decompressed within a range of - 2500E ^» a to - lOPa relat ive to the downstream area farom the casting bead. Moreover , it is preferable that a jacket ( not shown) is attached to the decompression chamber 90 to maintain the inside temperature at a predetermined value . Further , it is preferable to attach a suction unit ( not shown ) to an edge of the casting die 81 in order to keep a desired shape of the cas ting bead . A preferable range of air volume aspirated in the edge portion is 1 X/min to 100 L/min. After the casting membrane 24a has possessed a self -supporting property, the casting membrane 24a is peeled off as the membrane 5 2 from the belt 82 while being supp orted by a peel roller 109 . After that , the membrane 62 containing the solvent is carried, along the transporting section 101 supported t>y the many rollerrs to the tenter device 64 . In the transporting section 101 , it is possible to give a d.:raw tension to tlxe membrane 62 by increasing- a rotation speed of the downstream roller relative to that of the upstream ro ller. In the transporting section 101 , dry air of a desired temperature is sen t from the air blower 102 to the proximity of or directly to the membrane 62 to promote drying of the membrane 62 . At this t-±me , it is preferable that the temperature of tine dry air is in a range of 20 ⁰ C to 250 ⁰ C .

The membrane 62 transported to tϊie tenter device C 4 is dried while carried in a state that both side edges thereo f are held with clips 64a or the like . Instead of the clips 64a , it is also possible to put pins through the membrane 62 for holding the

membrane 62. It is preferable to separate inside the tenter device 64 into different temperatuure zones to adjust the drying conditions. In ttie tenter device 54, it is possible "to stretch the membrane 62 in the width direction. Through at least one of the transporting section 101 and. the tenter device 64, the membrane 62 is preferably stretched in at least one of the casting direction and the width direction by 100.5%-300% with respect to the size of the membrane 62 before the stretching.

The membrane 62 is dried by tlie tenter device 64 until the remaining solvent reaches a predetermined value. Both side edges of the membrane 62 are cut off by "the edge slitting device 67. The cut edges are sent to the crusb_er 103 by a cutter blower not shown. The side edges of the membrane 62 are shredded into chips by the crusher IO 3. Since the chips are recycled for ^* preparing the dope, the materials are efficiently utilized. Tlxe slitting process for the membrane side edges may be omitted. However, it is preferable to perform the εslitting process between the casting process and the membrane winding process.

The membrane 62 whose side edges are cut off and removed is transported to the drying chamfcer 69 for further ¹ drying. A temperature inside the drying chamber 69 is not particularly limited. However, it is preferable to determine the temperature according to the txeat resistance (glass transition point Tg, heat deflection tempenrature under load, melting point Tm, continuous working temperature and the like) of the solid electrrolyte, and the temperature is preferably not more than the Tg. It is preferable to dry the membrane 62 such that the remaining solvent of the membrane 62 is less than 5 wt . % . In the drying chamber 69, the membrane 62 is transported while being bridged across the rollers 68. The solvent vapor generated by cirying the membrane 62 in ttie drying chamber 69 is adsorbed anόl recovered by the absorbing device 106. The air from which the solvent is

removed is supplied to the drying chamber 69 as thte dry air. The drying chamber 69 is preferably divided into plural sections so as to change the temperature of the dry air in each section. It is also preferable to provide a predrying chamber (not shown) between the edge slitting device 67 and the drying chamber 69 to predry the membrane 62. Ther-eby, in the drying chamber 69, an abrupt increase of the membrane temperature is prevented so that changes in shapes and contϋitions of the membrane 62 are prevented. After the drying in the dirying chamber 69, the membrane 62 is cooled to the room temper-ature in a cooling chamber 71. When the humid-Lfication chamber is provided between the drying chamber 69 and the cooling chamber 71, it is prefexable to spray air whose humidity and temperature are adjusted to desired values to the membrane 62. Thereby, cur-ling and winding defects in the membrane 62 are prevented.

In the solution casting method, various processes such as the drying process and the edge slitting and removing process are performed between the membrane peeling process in which the membrane (solicl electrolyte membrane) 62 is peeled off from the support, and the membrane winding process in whiσli the membrane 62 is wound. During each process or between the above processes, the membrane 62 is mostly suppoirted or transported by using the rollers. There are driven rollers and non-driven rollers. The non-driven rollers determine ttie transporting passage of the membrane 62 ancL improving the transportation stability during the transportation of the membarane 62.

The charged voltage during the transportation of the membrane 62 is controlled by the neutralization device 72 at a desired value. The charged voltage after the neutralization is preferably in a. range of -3JcV to +3kV. Furthers, knurling is preferably provided to the memfc>rane 62 by using the knurling

roller pair 73. Note that the height of each of projections and depressions of the knurling is preferably in a range of lμm to

The membrane 62 is woundby a winding roll 107 of the winding device 76. It is preferable to apply the tension of the desired value to the meiαbrane 62 by the press roller 108 durJLng the winding of the membrane 62. It is preferable to gradually change the tension appliecl to the membrane 62 from the start to the end of the winding. Ttiereby, excessive tightening during the winding is prevented. A. width of the meiαbrane 62 is prefeacably 100mm or more. The present invention is also applicable to the production of thin membranes with the thickness in a range of 5μm to 300 μm. The present invention is especially effective in producing the membrane with -the thickness in a range of 10μm to 200 μm. Next, refferring to Fig. 1, a membrane producing apparatus 233 which is a second embodiment of the present invention is described. The membrane producing apparatus 233 uses a plastic membrane (herej±nafter referred to as a web) 111 wrapped around a casting drum 110 as the support instead of the belt 82 used in the membrane producing apparatus 33. The web 111 is loaded in a web feeding device 112 in a roll form. From tlie web feeding device 112, the web 111 is fed into the casting darum 110. Above the casting dr-um 110, the casting die 81 is disposed in the proximity of trie web 111. The dope is cast from tine casting die 81 onto the wet> 111 to form the casting membrane 24 on the web 111 while the web 111 is being "transported. Note that the dope 24 and the casting die 81 are the same as those used in the first embodiment so that the descrip-tion thereof is omitted.

Along the passage of the web 111, a casting membrane drying device 113 is provided. The casting membrane drying device 113 is constituted of a drying section 114. The drying section 114 is constituted of a duct 116 having an outlet 116a and an exhaust

port 116b, an air blower, a treating device, a.n opening to introduce outside air and so forth. Dry air 117 is blown from the outlet 116a to the casting membrane 24a in the direction of and parallel to the transporting direction of the web 111. Thus, the evaporation of the solvent is promoted. When the casting membrane drying device 113 uses tlie dry air for drying the casting membrane 24a as above, an air shield plate 118 is necessary between the casting die 81 and the outlet L 16a. Surface fluctuations of the casting memiorane 24a caused toy the dry air 117 is prevented by the air shield plate 118 so that the membrane with low uneveαness in the thickness is obtained.. The casting die 81, the casting drum 110, tlie outlet 116a and the exhaust port 116b of the casting membrane drying device 113 are provided in a casting ctiamber 119. In the case the predetermined amount of solvent is evaporating from the casting membrane 62, in orier to control the amount of solvent vapor in the casting chamber 119, the gases other than the solvent vapor in the casting chamber 119 should be recovered ancl kept at a predetermined amount. Instead of the above dry air supplying method, JLt is also possible to put a cover in the transportation passage of the support membrane 111 from the casting die 81 to the first liquid bath 120 or to adjust a distance between the casting and the first liquid bath 120. Moreover, it is also possible to adjust pressure of the solvent vapor, the pressure of gases o-fcher than the sol-vent vapor, temperature, aix supplying velocity, air discharge velocity and the like in this ambience. As "the drying method , it is also possible to use infrared rays, decompression, far infrared rays and microwaves for drying instead or in addition to the above dry air.

As the web 111 , the nonwoven plastic membirane such as polyethylene terephthalate ( PET ) film, polybutyJLene

terephthalate CPBT) film, nylon. 6 film, nylon 6, 6 film, polypropylene ffilm, polycarbonate film, polyimi.de film or the like is used. It is preferable that the web 111 has chemical stability against the solvent. Zt is also prefexable that the web 111 is heat resistant to wi-thstand the membrane forming temperature. In this embodiment, the PET film is used as the web 111.

A feeding speed of the wet> 111, that is, a. casting speed is preferably ά_n a range of 0.5m/min. to lOOm/m-Ln. The surface temperature of the web 111 is properly determined according to the material tliereof . The surface temperature is preferably adjusted in a .range of -2O ⁰ C to 100 ⁰ C. To adjust the surface temperature of the web 111, the passage for the heat transfer medium (not shown) is formed inside the casting ixum 110 to feed the heat transfer medium whose temperature is kept at the predetermined value. During the rotation of the casting drum 110, the position fluctuations in the vertical direction thereof due to displacements of its rotation center is preferably adjusted to be not more than 0.2mm. Defects on the surface of the web 111 should be minimized. ParticularILy, the number of j?in holes whose diameter is 30 jum or more is zero, the number of pinholes whose diameter is 10 /*m or more and less than 30μm is 1 or less per Im ² , and the number of pinholes wϊiose diameter is less than 1O ₁ Um is 2 or less per Im ² . The first and second liquid baths 120, 121 are constituted in the same manner as those in the first embodiment . The first liquid 65a is stored in the first liquid bath L20. The second liquid 66a is stored in the second liquid bath 121. After the immersion in the first liquid bath 120, the casting membrane 24a on the web 111 has a self-supporting property. The casting membrane 24a is peeled off from tlie web 111 by using a peel roller 123. Hereinafter the peeled membzrrane is referred to as a membrane

124. The membrane 124 is immersed in the second liquid bath 121 by using guide rollers 121b. The solvent containecϋ in the casting membrane 24a is reduced by substituting the first liquid 65a for the solvent contained in the casting membrane 24a in the first liquid bath 12O to promote the dirying of the membxane 124 in the next drying process in the drying chamber 69. En addition, by substituting the second liquid 66a for the first liquid 65a and a mixture of tlie first liquid 65a and the solvent contained in the casting membrane 24a in the second liquid bath 121 , the drying of the membrane 124 is further promoted. Hexreinafter, the substitution of the first and the second liquids 65a and 66a for the organic solvent contained in. the casting membrane 24a and/or the membrane 62 described above is referred to as a solvent substitution. After the immersion in the second liquid bath 121 , the membrane 124 is dried in the drying chamber 69 and wound by the winding device 76 in the roll form.

The remaining solvent in -the membrane 124 at the time of peeling off from the web 111 is preferably in a range of 200 wt . % to 950 wt. % to the total solid components in the membrane 124. As a predrying process before the immersion in the second liquid bath 121, it is possible to use the tenter device 64 of the first embodiment shown in Fig. 2 and/or the drying criamber 69 using the rollers. For instance, the casting membrane 24a is dried in the tenter device 64 together witϊi the web 111 in trie tenter device 64, and then -the casting membrane 24a is dried in the drying chamber 69 using the rollers. The order is not particularly limited in the present invention. Further, the tenter device 64 is properly installed, for instance, between the membrane peeling process and the membrane winding process. After tlxe membrane 124 is peeled off, the web 111 is wound by a web winding device 125 in a roll form. To supply the web 111 continuously, it is preferable that both the web feeding

device 112 and the web winding de-vice 125 have a turret mechanism. In the second embodiment, instead of the web feeding device 112 and the web -winding device 125, it is also possible to provide guide rollers only. In this case, the web 111 is circulated in an endless loop. A surface detecting device is pnrovided between the guide rollers to detect the surface roughness of the web 111. When the numt>er and the size of pά_n holes exceed tlxe predetermined value, a new web 111 is supplied . To supply the new web 111, the old web 111 is cut and the new web 111 is adhered thereto. When the new web 111 is rotated by a. round, the old web 111 is cut off and removed so as to bond the ends of the new web 111 to form the endless loop. Further, to prevent the membranes 124 from sticking tog/ether and to protect the surface theαceof , it is also possible to "Wind the web 111 together with the casting membrane 24a. In this case, the casting membrane 24a is peeled off from the web 111 at the time of procϋucing the fuel cell as will be described later.

Next, a third embodiment which is the most preferable embodiment in the present invention is described. Fig. 4 is a schematic vά_ew of a membrane producing apparatus 333 of the third embodiment. The membrane producing apparatus 333 is provided with the wet> 111 instead of the belt 82 shown ±n Fig. 2. In a producing method using the membirane producing apparatus 333 , the casting membrane 24a is peeled off as a membrane 410 after the casting membrane 24a formed on. the web 111 is immersed in the first liquicl 65a. Note that the same numerals are assigned to the devices and members constituting the membrane producing apparatus 333 similar to those in the first and second embodiments shown in Figs.2 and 3, and descriptions thereof are omitted.

The web 111 is loaded in the web feeding device 112 in the roll form in the same manner as that in the second embodiment.

The casting chamber 63 is provided with a belt 400 for supporting the web 1 11 . The belt 400 zLs bridged across drums 401a , 401b to form a passage through whi ch the web 111 pas ses in the casting chamber 6 3 . The web 111 is fed by the web fe&ding device 112 to the belt 400 and transported along the passage in the casting chamber 6 3 . Thereafter, the web 111 is transported out of trxe casting chamber 63 . Instead of using the beILt 400 , it is also possible "to use the above-mentioned casting cϋrum 110 to supporrt the web 1 11 . In the proximity of the transportati on passage in tlαe casting chamber 63 , the casting die 81 is disposed . The dope 24 is cast fxrom the casting die 81 to the web 111 "to form the casting membrane 24a while the web 111 is being transported . In thie proximity of the passage in the downstream frrom the casting die 81 , the air blowers 91-93 and the air shieXding plate 94 arre provided in the same manner as those in the first embodiment . The casting membrane 24a is dried by the dr^y air from the air blowers 9 1 - 93 while the wet> 111 is being transported along the passage . The casting membrane 24a is dried until the remaining solvent reaches the predetermined value . Thereafter , the casting membrane 24a together witti the web 111 is transported outsitϋe the casting chamber 63. Zn the downstream, from the casting chamber 6 3 , the web winding device 125 is disposed . The web 111 is transported by the guide rollers and wound by the membrane winding clevice 125 .

Guide rollers 405b are provided between "the casting chamber 63 and trie web winding chamber 125 . A first liquid bath 405 , a first wat er remover 415 an<3L the peel roller X23 are disposed in this order . The casting membrane 24a is transported from the casting chamber 63 by the -web winding device 125 and the guicξle rollers 4 05a . The casting membrane 24a comes in contact with ttie

first liquid 65a. while being supporrted by the web 111. "Thereafter, the casting membrane 24a supported, by the web 111 is transported from the first liquid bath 405 to the first water remover 415.

Water on tlie casting membrarxe 24a supported by "the web 111 is removed by trie water remover 415. As the water remover 415, for instance, blades, an air knife, rolls or the like are used.

Among the above, the air kni-fe is most preferable for the water remover 415 since the air knife removes water most efficiently. By adjusting air flow volume and air pressure of the air blown onto the casting membαrane 24a, the air kniLfe removes the remaining moisture content on the surface of the casting membrane 24a almost completely. However, if the air fflow volume is too large, flutter or tilt may occur in the casting membrane 24a which adversely affect the transporting stability. For that reason, the airflow volume is preferably in a range of 10m/s to 500m/s, more preferably, 20m/s to 300m/s, and most preferably, 30m/s to 200m/s . Note that the above air flow volume is not particularly limited. The air flow volume is properly determined depending on the moisture content on the surface of the casting membrane 24a before using the water? remover 415 , the transporting speed or the like.

To uniformly remove the moisture content on the surface of the casting membrane 24a, a variation range in the airflow velocity distribution in the wiclth direction of the casting membrane 24a is preferably set at 10% or less, and more preferably 5% or less by adjusting the outlet of the air knife or the air supplying method of the air knife. The closer the clearance between the surface of the casting membrane 24a and the outlet of the air knife, the more moisture content on the surface of the casting memt>rane 24a is removed. However, at the same time, the surface of the casting membrrane 24a is more likely to be damaged by the outlet of the air knife. Accordingly, trαe air knife

is ins "tailed such that tine clearance between the surface of the casting membrane 24a and the outlet of the sir knife is in a range of 10 ju m to 10cm, more preferably 100 /u m to 5cm, and most preferably 500 μ m to lcm . It is preferable to install the air knife and a backup roll on opposite sides of the transportation passage of the casting membrane 24 . The backup roll supports the casting membrane 24a so as to stabilize "the clearance setting and recϋuce the flutters , -wrinkles and def ormations of the casting membrane 24a . The casting membrane 24a which pas sed through the fzLrst water remover 415 is transported to the peel roller 123 . The ipeel rollezr 123 peels off the casting membrane 24a from the web 111 as the membrane 410 and transport the membrane 410 to the tenter device 64 . In the tenteir device 64 , the membrane 410 is daried until the remaining solvent reaches the predetermined valLue . Theresf ter , the membrane 410 is transported to the edge slitting device 67 .

In a second liquid, bath 420 in which, the second liquid. 66a is stored , guide roller's 420b are provicϋed . The membrane 410 whose side edges have been cut off and removed by the edge slitting device 67 is transported, to the second L iquid bath 420 by the guide rollers 420 , immersed into the s econd liquid 66a and transported out of the s econd liquid bath 420 . Thus , the second liquici 66a substitutes ffor the first liquid 65a and the mixture of the first liquid 65a and the solvent contained in the membzirane 62 by contacting the membrane 410 with tϊie second liquid € 6a . Thereafter , the membrane 410 is transported to a second water remover 425 . The second water remover 425 lias the same structure as the first water remover 415 and is used for removing the water from the surface of the membrane 410. The membrane 410 wlhich passecl the first water remover 415 is transported to the drying chamber 69 . In the drying chamber 69 , the dry air is blown onto

the membrane 410 for drying while th.e membrane 410 i s being transported . As descritied above , the time for drying the membrane 410 in the tenter devi ce 64 and the drying chamber 69 , that is , the time for removing the organic solvent contained in the membrane 410 is shortened by contacting the membrane 4 10 with the first and second liquids 65a and 66a . The time for- drying the membrane 410 is furrther shortened b>γ removing the water from the surface of the membrane 410 with "the use of the fíLrst and second water removers 415 and 425 . In the present invention , as described above , the casting membirane or the membrane is dried before contacting with, the poor solvent of the solid electrolyte . Sucϊi drying process reduces the ^remaining solvent in the casting membrane or the membrane so ttiat the formation of pores in the casting membrane oar in the membrane is prevented - Thus , it becomes possible to obtain the solid, electrolyte memt>rane with very little defects .

In chemical formula 1 , if X is a polymer which is a. cation species other than the hydrogen atom H , that is , the precursor of the solid electrolyte , it is possible to perform acid processing during the above-mentioned, producing methocl of the solid electrolyte memt>rane . In the acid processing, p αroton subs titution is perfoirmed by contacting the precursor membrane with, the solution con-taining acid whi_ch is a proton- donating subs tituent . Thereby , the solid elect acolyte is generat ed from the precursor in the precursor membrane . Thus , the so lid electrolyte membrane ά_s produced from the precursor membrane by the proton substitution . Note that the proton substitution in the present invention is to substitute the hydrogen atom for the cation species X other than the hydrogen, atom( s ) H in the p> olymer . To perform the proton substitut ion in the preσuarsor membrane with a high degree of efficiency, the remaining solvent in tlαe precursor membrane is preferably not less than 1 wt . % and

not more than 100 wt . % ( dry measure ) - If the drying is continued until the remaining solvent achieves less than 1 wt . % , the drying time becomes too long which is not preferable . If tine acid processing is perfor-med to the precursor membrane containing the remaining solvent exceeding 100 wt . % ( dry measure) , a p ercentage of voids becomes to o large which is not preferable .

After the prot on substitution, it is preferable to perform a washing process to remove the solution containing a.cid which has not been used for substituting lnydrogen atom( s ) ff or cation species from the membrane . Thereby , it becomes possible to prevent the polymer constituting the solid electrolyte membrane from being contaminated by the remaining acid .

As a method for washing the membrane after tlie proton substitution , it is preferable to Linmerse the membrane in the water . However , the method is not particularly limit ed to the above as long as the method is capable of removing ttxe acid by contacting the membrane to the watenr . For instance , JLt is also pos sible to coat or spray the water onto the surface of the solid electrolyte membrane . Such methods are applicable -while the membrane is being transported σont±rLuously without recϋucing the productivity thereof .

As the method for applying tha water to the membrane , for ins -tance , a method vising an extrusion coater or various coating heads such as a fountain coater or a. frog mouth coater: is used . As "the method for spraying the water onto the membirane , for ins tance , a method using a spray nozzle which is commonly used for humidification of air , painting, automatic washing of a tank and. so forth is used - The coating mettiods disclosed in "All about coa ^~ ting" , edited by Masayoshi Araki , published by Kafeo Gijutsu KenJcyukai ( Converting Technical Institute) , 1999 , are also applicable to the present invention . Further , in the case the spray nozzle is usecϋ , it becomes possible to spray tne washing

liq;-uid across the entire width of the membrane by arranging a plurality of conical or sector spuray nozzles manufactured by Ikexichi Co. , Ltd. or Spraying Systems Co. , Ltd. along the width dir-eσtion of the membrane. The higher ttαe velocity of sp -raying water, the higher the turrbulent mixing effect is obtained. However, such, turbulent mixing effect may cause the reduction in the transportation stability of the membrane. For that reason, it is preferable to spr-ay the washing liquid at a spraying velocity of 50 cm/s to 1OO 0 σm/s, more preferably 100 csm/s to 700 cm/s , and most preferably 100 cm/s to 500 cm/s.

The amount of: water to be useαl in washing should be larger than the calculatecl amount based on a theoretical dilution ratio defined by an equation (1) below. Thee theoretical dilution ratio is defined on the assumption that the whole amount o_f water for washing contributes to dilution of the contact solution coiitaining the acid. Actually, since the whole amount of water does not contribute to form a mixtixre, a larger amount of water than that derived from the theoretical dilution ratio is used in practice. The amount of water varies depending upon the acid concentration of the solution used, additives, and type of a sol. vent; however, water is used in an amount providing- a dilution ratio of at least 100 to 1000 times, preferably 50O to 10,000 times, more preferably 1,000 to 100,000 times. Note "that in the eqxxation (1) below, volumes each of water and aqueous acid soJ_ution is a liquid amount contacting with the membrane per a unά-t area.

Theoretical cϊilution ratio = < amount of washingr liquid 121 [cc/m ² ]) í (amount of aqueous acid solution 110[cc/rn ² ]) ''-(I) When a predetermined amount of water is used for washing, it is preferable to divide the prede-termined amount of water into several portions to contact the membrane with the water several

t imes rather than to use the whole amount of water at one time , that is , a so-called batch washing . The predetermined amount of water is divided and supplied to plural washing means disposed along the transportation passage . Appropriate distance is preferably provided between the adjacent washing devices so as to diffuse the water on the membrane to promote diliation . Further, it is preferable to incline the so lid electrolyte membrane being transported sucti that the water f lows over the membrane surface so as to diffuse the water and dilute the soluti on containing acid . The dilut ion efficiency is improved by providing water remover which removes the water- on the membrane between the washing devices . As the washing device , those s imilar to the af orementioned ff irst and second water removers 415 _^ 425 are used .

It is advantageous to dispose the large number of the washing devices as much as possible along the transportation passage . However , in terms of in stallation spacer and cost , it is preferable to dispose the washing devices in 2- 10 positions , more preferably 2 - 5 positions along the transportation passage .

The above acid processing and washing are performed between a process after ¹ the formation o f the casting membrane and a process for obtaining the membrane product . For ins tance , a first tank and a second tank are provided between the casting chamber and the tenter device . The solution containing the acid is stored in the first tank: . Water is stored in the second tanlc . The casting membrane which has been dried is transported to the first tank for the acid processing and then to the second tank for washing . The casting memt>rane is transported to each tanlk while being supported by the support , or after" the casting membrane is peeled off from the support . After the washing , it is preferable to remove the water from the surface of the casting membrane or the membrane by using the water remover . The water remover is not particularly limited, and it is possible to use aiE orementioned

water removers .

In the above embodiment , two liquid baths are provided . However, one or more liquid baths may be used.. In the above embodiment, two liquid baths are installed in tandem. However, it is also possible to provide another process such as the drying process between the first and second liquid baths . To contact the membrane with the solution. , in addition to immersing the membrane in the liquid baths, spraying, coating and other methods may be used. A single solvent or a mixture of two o>r" more solvents is used for preparing an optimxim solution for contacting the membrane depending on the organic solvent used for the membrane production. The organic solvent is more securely removed from the membrane fc>y performing the solution contact process for several times . In the paresent invention, two or more sorts of dopes are co-cast to produce a solid electrolyte multilayer membrane. As the co-casting method, there are a simultaneous co-casting method in which two or more sorts of the dopes are simultaneously cast, or a sequentially casting method in which two or more sorts of the dopes are sequentially cast. When the co-casting is performed, the casting die with a. feed block or a rnulti-manifold type casting cLie can be used.

The simultaneous co-casti-ng method is described. Fig. 5 is a schematic view illustrating the simultaneous co-casting apparatus. In Fig. 5, the same numerals are assigned to the members similar to those in Fig. 2. A simultaneous co-cast apparatus 111 forms a three— layer casting membrane 112. Accordingly, a three layer solid electrolyte membrane is obtained. However, according to prescriptions of dopes 114 to 116, boundaries of the layers may become indistinguishable. In such case, the solid electrolyte membrane may appear to have only one layer or two layers. When tlxe solid electrolyte multilayer

membrane has three layers, a layer contacting a belt 86 is referred to as a first surface layer 112a, a layer exposed to the air is referred to as a second surface layer 112b, and a layer interposed toy the first and second surface layers 112a and 112b is referred to as an inner layer 112c.

The first, second and thixrd dopes 114, 115 and 116 are cast onto the belt 82 to form the firrst surface layer- 112a, the inner layer 112c and the second surface layer 112b respectively. The first to third dopes 114-116 are fed to a feed block 119 attached to the casting die 81 through individual liquid passages Ll - L3. In the freed block 119, the liquid passages X1-L3 are joined together to simultaneously cast the first to thi_rd dopes 114-116 from the lip end of the feed block 119. In other words, in the feed block 119, three liquid passages which join into one liquid passage are formed. The liquid passage placed in the middle of the three liquid passages is fox the second dope 115. The liquid passage placed upstream from the middle liquid passage in the transport direction of the belt 86 is for the first dope 114. The liquid passage downstream from the middle passage in the transport direction of the belt 86 is for the third dope 116.

The first and third dopes 114, 116 forming the first and. second surface layers 112a, 112b have lower viscosity than that of the second dope 115 forming the inner layer? 112c. Thereby, it becomes possible to form -the membrane in which abnormal properties such as melt fracture is prevented- In some cases, it is preferrable to intentionally form a bead between the casting die 81 and the belt 86 such that the higher viscosity layer (the inner layer- 112c) is sandwiched by the lower viscosity layers (the first and second surface layers 112a and 112b) as above when the first, second and third dopes 114, 115 and 116 are cast after each viscosity is properly adjusted. In some cases, it is preferable to add the poor solvent only to the first and third.

5U

dopes 114 and 116. In some cases, it is prefexable to add tϊi& poor solvent to all of the first to third dopes 114-116 but to> make the ratio of the poor solvent in the first and third dopes 114 and 116 higher than that in the second dope 3_15. In this case, it is prefezrable to cast the dope such that the thickness of the* first surface layer 112a contacting the belt 8S is not less than. 5//In in a wet state.

Thus, the casting die 81 has one outlet for casting the dope. By tlxe combined use off the feed block IIS and the casting die 81, the first to third dopes 114-116 are simultaneously cast from the outlet. Instead of using the feed block 119 and the casting die 81, it is also possible to use a casting die whicrx casts the first to third dopes 114-116 separately from three individual outlets. In this case, the three outlets formed in the casting die are aligned in the transport direction of the belt 86.

In a single layer casting, it is often necessary to cast a solid electrolyte solution of higher concentration and higtα. viscosity so as to form the membrane with the desired thickness- In such case, dope often becomes unstable and solids may be generated. The generated solids are mixed into the membrane as foreign matters, compromising the planarity off the membrane. Bousing the above co- casting method, not only the planarity is improved, tout also the second dope with the high concentration is cast without trouble. Accordingly, drying efficiency is improved to enhance the production speed.

The thickness of each of the layers IL12a-112c is not particularly limited. However, the sum of the thickness of the first and second surface layers 112a, 112b preferably occupies l%-50%, moi-e preferably 2% -.30% of the total thickness of the solid electrolyte membrane as the product after drying.

It is possible to use the dopes 114-116 with the different

concentrations . It is also possible to add different additives to each elope . In particular , it is also possible to add the aforementioned antioxidants , fibers , fine particles , water: absorbing agents , plasticizers and solubilizers of different types and concentrations to each dope . For instance , amounts of the antioxidants and fine particles (matt ing agents ) added to the first and second surf ace layers 112a and 112b may be largex: than those added to the inner layer 112c , or the antioxidants and fine particles may t>e added only to the first and seconcϋ surface Xayers 112a and 112b . Amounts of the water absorbing agents , solubilizers and. plasticizers added to the inner dope 112c may t>e larger than tho se added to the f irrst and second surface layers I-L 2a and 112b , or such additives may be added only to the inner doipe 112c . It is also possible to add the low volatile antioxidant to the surface layers 112a and 112b , and add the plasticizer and the water absorbing agent to the inner layer 112c It is als o possible to add the peeling agent only to the first dope 114 forming the fiirst surface layer 112a contacting the casting ciie 86 . Thus , by adding appropriate additives to eacln dope , the desirable funct ions are added to each layer . Further , it is also possible to cast the solid electrolyte solution of the present invention simultaneously witti the other function layers , for instance , a catalyst layer, an antioxidant layer , antistatic layer , a lubricating layer anal the like . It is preferable to add the fine pairticles to the first and secon-d surface layers 112a , 112b so as to improve lubricating property of the membrane . To improve the .lubricating property of the membrane , it is effective to add tlie fine particles to at least one of the first and second surfaoe layers 112a, 112b _ An apparent specific gravity of the fine pairticles is preferably not less than 70 g/liter , more preferably 9O g/liter - 200g/litexr , and most preferably 10Og/ liter - 200g/litear . The fine particles

with the higher apparent specific gravity enable to prepaare a dispersion liquid witn higher concentration which prevents agglomer-ation. If silicon dioxide is used as the fine particH.es, an average diameter of a. primary particle ±_s preferably not more than 20nm, and the apparent specific gravity is preferably not less than 70 g/liter. Such silicon dioxicϋe is obtained by, for instance, mixing silicon tetrachloride and hydrogen, and "then burning "the mixture in the air. Instead of such silicon dioxide, it is also possible to use Aerosil 200V or i^erosil R972V prodxaced by Nippon Aerosil Co. Ltd.

Fig. 6 is a schematic view of the sequential co-cas ^~ ting apparatus. In Fig. 6, an example in which, first to third dopes 114-116 are co-cast is described. In a sequential co-casting device 121, three casting dies 122-124 are disposed in this older along tϊie transport passage of the belt S 6. The first dope 114 is cast from the casting die 122. The second dope 115 is cast from the casting die 123. The third dope 116 is cast from, the casting die 124.

If: the first to th.dL.rd dopes 114-116 of the same composition are sequentially co-cast, the membrane production speecl is enhancecl compared to that in the single Layer casting. In this case, the positions of the second and third casting dies 123, 124 are determined according to the drying speed and the like of the preceding layer. For instance, it is; preferable to dispose the second casting die 123 at a position, where a ratio off the distance between the most upstream casting- die 122 and the second casting die 123 to the distance between the most upstream casting die 122 and the position to peel the cas ting membrane is in a range of 30%-60%. In addition to the above methods _r following metho»d is available as an example . A first dope is cast from a first casting die to form a membrane , and the membrane* is peeled off. T?hen,

a second! dope is cast from a second casting die onto the p eeled surface of the peeled membrane to form the double -layer membrane .

Regardless of -the single layer casting method or co-casting method , there are various methods for casting the dope , for instance , a method ±n which the dope is uniformly ext ruded from a pressure die , a cioctor blade method in which a thickness of the casting membrane on the support is adjusted by us ing a blade , and a reverse roll coating method Ln which a casting amount of the elope is adjusted! by leveling the surface of the dope by using rollers rotating in reverse directions . Among the above , the mettiod using the pressure die is pref erable . As the pre ssure die , there are a coat rianger type , a T- die type and so ff orth . Any type of the pressure die is prefer ably used .

Instead of the above method , it is also possible to prroduce a different type of the solid electrolyte membrane by pixtting the solid electrolyte in the micropores of a so-called porous substrate in which a plurality of micropores are forma d . As examples of such methods , there are a me thod in which the solid electro lyte is put in the micropores by applying a sol -gel solution containing the solid electrolyte onto the porous substrate , a method in -which the solid eiectrolyte is filled in the micropores by immer sing the porous s iibstrate in the sol-gel reaction liquid and the like . As the porous substrate , porous polypropylene , porous polytetraf lvαoroethylene , porous cross-linked heat-resistant polyethylen.e , porous polyimicϋe and the like are preferably used . It is also possible to produce the membrane by processing the solid electrolyte into a fiberr-form and f HX. the voids in the fibers with other polymers , and forming the fibers into the membrane . As the polymer for filling the voids , it is possible to use the additives described in this specification .

The solid electrolyte membrane off the present invention

is sui "t ably used for the fuel cell, in particular, for the proton conductive membrane in a direct methanol full cell. In addition, the solid electrolyte membrane is used a.s a component of the fuel cell interposed between the two electrodes of the fuel cell. Further, the solid electrolyte membrane of the present in-vention is used for the electrolyte in various batteries or ceILls such as a redox flow battery and the lithium battery , a display element , an electrochemical sensor, a signal transmission meclium, a capacitor, electrodialysis , elect xolyte membrane for electrolysis , a gel actuator, salt electrolyte membrane and proton exchange membrane. (Fuel Cell)

In the following _r an example of using the solid electrolyte membrane in a membrane electrode assembly, hereinafter ref erred to as MEA) , and an example of using the MEA in the fuel cell are described. The MEA and, the fuel cell described in the following are examples of the present invention, tout the present irtvention is not limited to the following examples . Fig. 7 is a section view illustrating a configuration of the MEA. An MEA. 131 is constituted of the membrane 62, and an anode 132 and catfciode 133 placed opposite to each other. The membrane 62 is interposed between the anode 132 and the cathode 133.

The anode 132 is constituted of a. porous conductive sheet 132a and a catalyst layer 132b contacting the membrane 62. The cathocϋe 133 is constituted of a porous conductive sheet 133a and a catalyst layer 133b contacting the membrane 62. As the porous conductive sheets 132a . 133a, carbon pa;per and the like a:re used. The catalyst layers 132b, 133B are form&d. of a dispersion in which carbon particles are dispersed into the proton conductive material. The carbon particles support a catalyst metal thereon such as platinum. As the carbon particles, there are ket jen black, acetylene black, carbon nanotubes. As the proton conductive

material , for instance, Nafion and ttie like are used .

The following me -fchods are preferably applied for p producing the MEA 131:

(1) Proton conductive material coating method: a catalyst paste (ink) including an active metal- supported carbon, a proton conductive material and a solvent is clirectly applied on both surfaces of the membrane 62, and a porous conductive sheets 132a, 133a are thermally adlxered under pressure thereto to construct a 5 -layered MEA. (2) Porous contϋuctive sheet coating method: α liquid containing the material for the catalyst layer 132b, l_33b, for instance, the catalyst paste is applied onto the porous conductive sheet 132a and 133a to form a catalyst layer "thereon, and a solid electrolytic membrane 62 is adhered there to under pressure to construct a 5 -layered MEA..

(3) Decal method : The catalyst paste is applied onto PTFE to form catalyst layenrs 132b, 133b tlαereon, and the catalyst layears 132b, 133b alone are transferred, to a solid eleσ "trolytic membxrane 62 to construct a 3 -layered structure. A porous conductive sheet is adiiered thereto uncϋer pressure to construct a 5 -layered MEA.

(4)Catalyst post -attachment method: Ink prepared by mixing a carbon material not supporting platinum powder and a proton conductive material is applied or cast onto a membrane 63, porous conductive sheet 132a, 133a ox PTFE to form a membrane. Thereafter, the membrane 62 is immersed into a liquid containing platinum ion so as to reduce and pxrecipitate the platinum particles in the membrane 62 to form tlie catalyst lay&xs 132b, 133b- After the catalyst layers 132b, 133b are formed, the MEA 131 is produced by one of the above methods (l)-(3) .

The method for puroducing the MEA 131 is not limited to the above and other known methods are also> used. For insta_nce, the

OD

following method is used instead of the above methods (I)- (4) . A coating liquid containing materials of the catalyst la/yers 132b, 133b is previously prepared. The coating liquid is applied onto tne support (or the web) and dried. T?he supports (or the webs) on which the catalyst Xayers 132b, 133b are formed are thermally aclhered to both surfaces of the membrane 62 such that the catalyst layers 132b, 133b contact the membrane 62. After peeling the supports (or the wet>s), the membrane 62 interposed by the catalyst layers 132b, 133b is sandwiched between trae porous conductive sheets 132s, 133a. Thus, the catalyst layers 132b, 133b are airtightly aclhered to the membrane to produce the MEA 131.

Fig. 8 is an exploded section view illust orating a configuration of the fuel cell. The fuel cell 141 is constituted off the MEA 131, a paiαr of separators 142, 143 for sandwiching ttie MEA 131 , current collectors 142 which are formed of stainless nets attached to the separators 142, 143, and gaskets 147. The anode-side separator 142 has an anocle-side opening 1-43 formed tnere through; and tlie cathode -side separator 142 has a cathode- side opening 152 formed their e through. Vapor fuel such as hydrogen or alcohoJL (e.g., methanol) or liquid fuel such as aςjueous alcohol solution is fed to ttie cell via the anode- side opening 151; and an O-scidizing gas such as oxygen gas or air is fed thereto via the csathode-side opening 152. For the anode 132 and the cathode 133, for example, a catalyst that supports active metal particles of platinum or the like on a carbon materrial may be used.. The particle sj_ze of the active metal particles generally used, is in a range o>f 2 nm to IO nm. Active metal panrticles having a smaller particle size may have a large surface area per the αnit wt. thereof, and are therefore advantageous since their activity is highezr. If too small, however, the particles are difficult to disperεse with no

aggregation, and it is said that ttie lowermost limit of the particle size is 2 nm or so.

In hydrogen- oxygen fuel cells, the active polarization of cathode (air electrode) is higher tlxan that of anode (hydrogen electrode) . This is because the cathode reaction (oxygen reduction) is slow as compared witti the anode reaction. For enhancing the oxygen electrode activity, usable are various platinum-basedbinary alloys such as Pt-Cr, Pt-Ni, Pt-Co, Pt-Cu, P-t-Fe. In a direct methanol fuel cell, in which aqueous methanol is used for the anode fuel, it is important that then catalyst poisoning with CO "that is formed duri_ng methanol oxidation must be inhibited. For this purpose, usable are platinum-based binary alloys such as Pt-Ru, Pt-Fe, Pt-Ni, Pt-Co, Pt-Mo, and platinum-based ternary alloys sucsh as Pt-Ru-Mo, Pt-Ru-W, P-t-Ru-Co, Pt-Ru-Fe, Pt-Ru-Ni, Pt-Ru-Cu, Pt--Ru--Sn, Pt-Ru-Au. As the carbon mater-ial for supporting "the active metal, acetylene black, Vulcan XC -"72, ketjen black, carbon nanohorn C CNH), and carbon nanotube (CNT) are preferabl_y used.

The catalyst layer 132b, 133b have following ff unctions: ( 1) transporting fuel to active metal „ (2) providing the reaction site for oxidation of fuel (anode) or for reduction thereof ( cathode) , (3) transmitting the electrons released by the redox reaction to the σuαrrent collector 146, and (4) transporting the protons generated in the reaction to solid electrolytic; membrane 62. For (1), the catalyst layers 132b, 133b must be porous so tliat liquid and vapor fuel may penetrate the pores. T7he active metal catalyst supported by the cart>on material works for (2); and the carbon material also works for (3). For attaining the ff "unction of (4), a proton conductive material is mixed into the catalyst layers 132b, 133b. The proton conductive mate:r:ial mixed in the catalyst layers 132, 133b is not particularly limited provided that it is a solid that has a proton -donati_ng group.

Oo

Polymer compounds having acid- residue used for the membrane 62, for instance, perf luorosulf onic ac-ids such as typically Nafion; phosphoric acid- ^" branched poly(iαeth)acrylates; sulfonated, heat-resistant aromatic polymers such as sulfonated polyether-ether ketones, sulfonated polybenzimidazoles are pireferably used. IEf the material of the membrane 62, that is, the solid electroXyte is used for "the material of the catalyst layers 132b, 133fc>, the catalyst layers 132b, 133b and the membrane 62 are made of the material of the same type. For that reason, the electrochemical contact between tlxe solid electrolytic membrane and the cataLyst layer becomes high, which is more advantageous in view of the proton conduction. Title amount of the active metal is preferably rfrom 0.03 mg/cm ² to 20 mg/cm ² in view of cell output and the cost efficiency. The amount of trie carbon material that supports tlie active metal is prref erably furom 1 to 10 times the mass of the active metal. The amount of the solid electrolyte is preferably from 0.1 to 0.7 times the mass of the active metal -supporti-ng carbon.

The anode 3.32 and the cathode 133 serve to prevent interference in current collection and gas permeation due to water accumulation. The carbon papers and carbon fi-bers are commonly used for the anode 132 an<3. the cathode 133. It is also possible to perfonnn polytetrafluoaroethylene (PTFE) processing to the carbon paper and carbon fibers for repelling ttie water. The MEA is preferably incorporated in the batteary. Sheet resistivity measurred by an AC impedance method when tlxe fuel is loaded is preferably 3 ωcm ² or less, more preferably 1 ω cm ² and most preferably 0.5 ωcm ² or less. The sheet resistivity is obtained by a product of an actuall_y measured value ancl a sample aarea .

Fuel for the fuel cell is described. For anode fuel, hydrogen, alcohols (e.g., methanol, isopropanol, ethylene

glycol), ethers (e.g., dimethyl ether, dimettxoxymethane , trimethoxyraethane) , formic acid , boron hydride complexes, ascorbic acid and the like are used. For cathode JEuel, oxygen ( including oxygen in air) , hydrogjen peroxide and the like are used. For cathode fuel, oxygen (including oxygen in air), Ixydrogen peroxide and the like aure used.

In direct methanol fuel cells , methanol aqueous solution txaving a methanol concentration o± 3 wt . % to 64 wt. % is used as the anode fuel. As in the anode xreaσtion formula (CH ₃ OH + H ₂ O — > CO ₂ + 6H ⁺ + 6e ^" ), 1 mol of methanol requires 1 mol of water, and the methanol, concentration in this case corresponds to 64 wt. % . A higher methanol concentration in fuel is more effective jE or reducing the weight and the voILume of the cell including the fuel tank of the same energy capacity. However, the higher methanol concentration tends to reduce the cell output due to the so-called crossover phenomenon in which methanol penetrates through the solid electrolyte and reacts with oxygen at the cathode to reduce the voltage. Wh^n the methanol concentration J_s too high, the crossover phenomenon is remarkable which reduces the cell output. To prevent the* above problem, the optimum concentration of methanol is cletermined depentϋing on the methanol permeatoility through the solid electrolytic membrane tαsed. The cathoόle reaction formula, in direct methanol fuel cells As (3/2) O ₂ + 6H ⁺ + 6e ^~ → H ₂ O, and oxygen (generally, oxygen in air) is used as the fuel in the cells.

To supply the anode fuel and the cathode fuel to the corresponding catalyst layers 132fc>, 133b, there are two methods: <1) a method of forcedly circulating the fuel by the use of an auxiliary device such as pump, that is, an active method, and (2) a method not using such auxil.La.ry device, that is, a passive method, for example, the liqui-d fuel is suppLied through capillarity or by free-fall, an-d vapor fuel is supplied by

exposing the catalyst layer to air- . It is also possible to combine these methods. The method (1) has some advantages in that water formed in the cathode area is circulated, and high— concentration methanol is usable as fuel, and that air supply enables high output from the cells , while it is difficult to downsize the cell because a fuel supply unit is necessary. On the other hand, the method (2) enat>les to downsize tne cells, while tlαe fuel supply ratio is readiXy limited and high, output from the cells is often difficult . The unit cell voltage of fuel cells is generally at most I V. It is desirable to stack up the unit cells in series, depending on the necessary voltage for load. A.s methods for stacking, a plane stacking in which unit cells aare arranged on a plane, and a bipolar stacking in which unit cells are stacked up via a separator with a fuel passage formed on bottα sides thereof are used. In tne plane stacking, the cathode (air- electrode) is on the surface of the stacked structure so that it is easy to take in air and realizes a thin structure. Accordingly, the plane stacking is sixitable for small-sized fuel cells - In addition, it is also possible to apply MEMS technology, in which a silicon wafer is processed to form a miσropattern thereon and fuel cells are stacked on the processed silicon wafer.

Fuel cells are used in varrious appliances , for example, for automobi-Les , electric ancϋ electronic appliances for household use, mobile devices and the like. In particular, direct methanol fuel cells enable downsizing, lightweight and do not require charging. Having such acLvantages , they are expected to be used for various energy sources for mobile appliances and portable appliances. For example, mobile appliances in which fuel cells are favorably used include mobile phones , mobile notebook-size personal computers , electronic still cameras, PDA, video cameras , mobile game drivers , mobile servers , wearable

personal computers, mobile displays and so fortli. The portable appliances in which fuel cells are favorably used include portable generators , outdoor .Lighting devices., pocket lamps, electrically-powered (or assisted) bicycles and so forth. In addition, fuel, cells are also dEavorable for power sources for robots for industrial and household use and ffor other toys. Moreover, they are further usable as power sources for charging secondary batteries that are mounted on these appliances. [ Example 1 ] Next, examples of the present invention arre described. In the following examples, the example 1 is described in detail. With regard to examples 2-8, only the conditions which differ from those of the example 1 are described. The examples 1, 2, 5-8 are examples of the embodiments in the present invention. The most preferable examples are the examples 7 and 8. Further, the examples 3 and 4 are the comparison experiments of the examples 1 and 2.

[Production of dope]

As the solid electrolyte, sulfonated polyacrylonitrile butadiene- s tyirene with the sulf onation degree of 35% which is referred to as a material A is used. First, the material A is prepared by ttie following synthesis.

(1) Synthesis of 4- (4-(4-pentylcyclohexyl) pherxoxymethyl) styrene A substance of the following composition is reacted for

7 hours at the temperature of 100 ⁰ C. Thereafteir, the obtained liquid is cooled to the room temperature. Water J_s added thereto to crystallize 4- (4- (4-pen-fcylcyclohexyl) phenoxymethyl ) styrene. The liquid containing 4- (4- (4-pentylcyclohexyl) phenoxymethyl) styrene is filtered. Thereafter _^ the crystal is washed by an aqueous solution containing water and acetonitrile at a ratio of IL to 1. Thereafter", the 4- (4- (4-pentylcyclohexyl)

phenoxymethyl) styrene is dried by the air.

4- (4-pentylcyclohexyl) phenol 14 pts. wt.

4-chloromethylstyrene 9 pts. wt.

5 Potassium carbonate 11 pts. wt.

N,N-dimethylformamide 66 pts. wt

(2)Synthesis of graft copolymer

A substance of the following composition is heated up to 60 ⁰ C. O Polybutadienlatex 100 pts. wt.

Potassimm rosinate 0.83 pts. wt.

Dextrose 0.50 pts. wt.

Sodium pyrophosphate 0.17 pts. wt.

Ferrous sulfate 0.08 pts. wt. 5 Water 250 pts. wt.

Polymerization is performed by dropping a mixture of the following composition onto the mixture of the above composition for 60 minutes.

Acrylonitrile 21 pts. wt. O 4- (4- (4—pentylcyclohexyl ) phenoxymethyH. ) styrene

62 pts. wt. t-dodecylthiol 0.5 pts. wt. cumene hydroperoxide 3.0 pts. wt.

After tlxe dropping, 0.2 pts. wt . of cumene hydroperoxide 5 is added to tine above, and cooled for one hour?. Thereby, latex is obtained. The obtained latex is put into 1% sulfuric acid at the temperature of 60 ⁰ C, and ttxen, the tempera-fcure is increased to 90 ⁰ C to coagulate. The coagulated latex is properly washed and dried to obtain the graft copolymer. O (3) synthesis of the material A by sulfonating the graft copolymer

The grafft copolymer (100 pts. wt . ) obtained by the process

bo

(2) is dissolved in methylene chloride (1300 pts. wt.)- Whiles the temperatuxre of the obtained liquid is kept at 0° C or below, concentrated sulfuric acid is slowly added thereto. Thereafter, the liquid is stirred for 6 hours to form precipitates. The precipitates are dried after" the solvent is rremoved to obtairϊ. sulfonated poJLyacrylonitorile butadiene styrene. The percentage of introduction of the sulfonic acid group is 35% measured by- titration. A όlope A is formed of the material A which is the solicl electrolyte, and the solvent which is N, N- dimethylf ormamide > [Produc-tion of solid eILectrolyte multilayer membrane 621

The solid electrolyte multilayer membrane 62 having three layers is procluced by using a simultaneous co- casting apparatus 111 by the following method. A second dope 115 is the dope A- A first dope 114 and a third, dope 116 are mixture of the dope A and N, N-dimethylf ormamide in which the concentration of the solid electrolyte is 18.0 wt .%. To be more specific, there are three liquid passages to be connected to the casting die 89 off the dope A. A static mixer is provided to one of the three liquicl passages. N, N -dime thy If ormamide is fed to tha dope immediately before the static mixer so as to mix the N, N-cϋimethylf ormamide and the dope A. by inline-mixing to form the dope 114. The statics mixer has 60 mixing elements - Another static mixer is provided!! to one of the remaining liquid passages to form the third dope 116 in the same manner as the first dope 114. In the simultaneous co — casting of the dopes 114-116, the casting width*, is set at 380mm and the flow volume of the each dope is adjusted so as to form the thickness of the first, seconcϋ and third surface layers of ttie dried solid electrolyte membrane 62 to be 5/<m, 5μm and 40μin, that is, 50μm in total. To kee∑? the temperatυixe of each of tlie dopes 114-116 at 40 ⁰ C, a jacket (not shown) is provided in th.e casting die 89 to supply the heat transfer medium at 40 ⁰ C.

4

The casting die 89, a feed block 119, and the liquid passages of the first to the third dopes 114 -116 are insulated at 40 ⁰ C. The casting die 89 is of the coat lxanger type and ttie width thereof is 0.4m. The casting die 89 is provided with heat bolts witli 20mm pitches for adjusting the. thickness of tlie membrane. The casting die 89 has an automatic.: thickness control mechanism which adjusts clearance of the IfLp end by using tϊie heat bolts. The heat bolts set a profile according to the total flow volixme of the first to third dopes 114-116 through a high-precision pump based on the previously set program. To set the profile, a feedback control is carried out based on an adjustment program according to a profile measured by an infrared thickness gauge (not shown) installed in the membrane producing apparatus 33. The clearance of the lip end is adjusted such that a difference in the thickness between two arbitrary points 50mm away from each other is lμm or less, and those Im away from each other in the width direction is 3 _/ /mor less in a portion of t2ie membrane 62 whose side edges have been cut off, to be more specif ic , in the ponrtion 20mm inward from each side ecϋge of the membrane 62. Further, the clearance of the lip end is adjusted such tb_at the thickness accuracy of -the surface layexrs is ±2% or less, that of tlie inner layer is ±1% or less and. that of the total thickness is ± 1.5% or less (with respect to the above-mentioned values of the dried membrane 62.) The dope discharged to the lip end of the casting die 81 is partially dried and becomes solid. In orcler to prevent sαch solidification of the dope _m the liquid useci as the solvent of the dope i_s supplied to botlα side edges of a three-phase contact line formed by both ends off the casting beacL, those of the -Lip ends and the outside air. A. pump with a pulsation of 5% or lass is used for supplying the solvent.

The belt 82 is formecL of SUS 316 having sufficient

anti-corrosion and strength. The belt is pol_ished such that the surface roughness is 0.O5μm or less. The tlxiσkness of the t>elt is 1.5mm. The unevennes s in the thickness i-S 0.5% or less. The belt 82 is driven by two rollers 85 and 86. A relative speed difference between the belt 82 and the roJ-lers 85 and 86 are adjustecl so as to be 0. Olm/min or less. Prreferably, speed fluctuations of the be_l_t 82 is 0.5% or less, and meanderi_ng thereof caused in a width direction while ttie belt 82 makes one rotation is 1.5mm or less. In order to contarol the meandering, the position of the belt 82 is controlled b>γ detecting the end positions thereof . Distance fluctuations between the lip end. and the casting belt 82 are controlled to be 20Oμm or less. In the casting chamber 63, a clevice for preventing air pressure fluctuations (not shown) which prevents tine air pressure fluctuations inside of the casting chambe-cr 63 is installed.

Trie first to thizrd dopes 114-116 arre cast to form the casting membrane 112. Tine dry air in a tempeirature range of 80° C -120 ⁰ C is blown onto the casting membrane 1_12 by using the air blowers 91-93 until the remaining solvent reaches 30 wt. % with respect to the weight of the solid component , that is, the solid electrolyte contained in the material A. At "the time the casting membrane 112 obtains the self-supporting prroperty, the casting membrane 112 is peeled off as the membrane <52 from the belt 82. The membrane 62 is transported to the tenterr device 64, and then through the tenter device 64 while the side edges of the memburane 62 are held by the clips 64a. In the tenter device 64 , the membnrane 62 is daried by the dry air at 140° C until tine remaining solvent reaches 15 wt. % with respect to the weight of the solid component . At the exit of the tenter device 64, the cJ-ips 64a release the side edges of the membrane 62. Then, the side edges of the membirane 62 are cut off by the edge slitting device 67 disposed in the downstream from the tenter device 64. Therreafter, the sol-vent

subst itution is performed by immersing the membrane 62 in a liquJLd. mixture of methanol and water at the ratio of 1 to 1 - The liquid, mixture is insulated at 60 ⁰ C . Af-fcer the immersion ,, the membirane 62 is transported to the drying chamber 69 and dried

5 in a range of 140° C-16O° C while being transported by the rol_lers 68. Thus , the solid eiectrolyte multilayer membrane 62 which contains the solvent ILess than 3 wt . % i_s obtained .

The following e-valuation is perf ormed to the obtained membirane . The result of the evaluation is shown in table I - The

LO numerals of the evaluation items in tabILe 1 correspond to the numeirals assigned to "the following evaluation items .

1 . Thickness

The membrane thickness is continixously measured at the velocity of 600mm/min by using the electronic micrometer

L5 produced by Anritsu Corporation Ltd . The data obtained trjr the measurement is recorded on a chart shee i: with a 1/20 scal_e at the velocity of 3 Oπun/min . The data curves are measured by αsing a ruler . According to the measured value „ average thickness and variations in thickness with respect to the average thickcness

10 are obtained . In table 1 , ( a ) is averages thickness (unit : // m) , ( b) is the variation in thickness (unit : μ m) with respecst to ( a) .

2 . Number of de .Erects

Light is emitted onto an area of t lhe whole width x l_m on .5 the membrane 62 and reflected . By the reflected light , deffects such as deformation and the like are detected by visual inspection . Thereafter , the detected defects are checked with a polarization microscope and the number of the defects per lmm ² is counted. Deformations after the sampling , for instance, ff laws 30 are not included in ttie counts .

3 . Measurement of Proton Conductivity

Ten measurement points are selected at intervals of Im

along a lengthwise direction of the obtained solid electrolyte memburane . Each of ten measurement points is punched out as a disc — shaped sample wi-th a diameter o>f 13mm. Each sample is intearposed by two stainless steel plates. The proton conductivity thereof JLs measured by AC impedance methocϋ using 1470 and 1255B produced by Solartron Co., Ltd. The measurement is performed at 8O ⁰ C, and relative humidity of 95%. The proton conductivity is indicated by AC impeόLance value (units S/cm) shown in table 1. 4. Power density of the fuel cell 141

A fuel cell 141 is produced by using the membrane, and the output thereof is measured. The producing method and the measurement method of the power density of the fuel cell 141 are desσiribed in the following. (1) Formation of catalyst sheet A used as the catalyst layers 132b and 133b

2 g of platinum- supporting carbon is mixed with IL 5 g of the solid electrolyte (5% DMF solution), and dispersed for 30 minutes by using an ultrasonic disperser. The average particle diameter of the resulting dispersion is about 500 nm. The dispersion is applied onto the carbon paper with a thick.ness of 350 μ,m. and dried. Thereafter, a disc- shape with a diameter of 9 mm is punched out of the carbon paper. Thus, the catalyst sheet A is formed. Note that the above platinum- supported σairbon is Vulcan XC72 with 50 wt _ % platinum. The solid electrolyte is the same as that in the membrane product-Lon. (2) Formation of MEA 131

The catalyst sheet A is attached, to both surfaces of the solici electrolyte membrane 62 such tha.t the coated surfaces of the catalyst sheet A are attached the solid electrolyte membrane and heat -pressed at 80 ^β C under 3MPa for- 2 minutes. Thus, the MEA 131 is formed.

bo

(3) Power density of fuel cell 141

The MEA 131 obtained in the aloove process (2) rLs set in the fuel cell shown, in Fig. 8. An aqueous 15 wt. % methanol solution is put into the fuel cell thorough an anode-side opening 5 151 > At this time, a cathode-side opening 152 is kep-t open to the air. The anode IL 32 and the cathode 133 are connected via a multi-channel battery test system produced by Solartron Co. , Ltd. to measure the poweir density (W/cm ² ) .

[Example 2]

0 As the solid electrolyte, sulfopropyl polyethersulf one with the sulfonation degree of 35% which is referrecϋ to as a material B is usecl. First, the material B is synthesized according to a methuod disclosed in Japanese Patent Iiaid-Open Publication No.2002-110174. The dried material B and tlie solvent L5 are mixed by the following composition to dissolve the* material B in the solvent. Tlius , the solid electrolyte dope 24 with 20 wt. % is formed. Other conditions arre the same as those in the example 1. Hereinafter the dope 24 is referred to as a dope B.

Material B 100 pts. wt .

-0 Solvent: N-Methyl-2-pyrrolidone 400 pts. wt .

[Production of solid electrolyte membrane 62] The dope B is used instead of -the dope A. The temperature of the air from the air blowers 91-93 is set at 80° C -l_40°C. The side edges of the membrane 62 are removed by the edge slitting 25 device 67. Thereafter, the membrane 62 is immersed in the liquid mixture of methanol and water at the ratio of 1 to 1 insulated at 60° C to substitute the liquid, mixture for the solvent contained in the meitilDrane 62. The membrane is dried in the drying chamber in a range of 160 ⁰ C -18O ⁰ C. "Thus, the solid eLectrolyte 0 menYbrane 62 with the remaining solveixt of less than 3% is obtained. The evaluation results of the obtained membrane 62 are shown in Table 1.

by

[Example 3]

[Production of solid electroJLyte membrane] Only the dope A is cast withoiat using the first dope 114 and the third dope 116 to form a single layer membrane with the thickness of 50 μm. Other conditions are the same as those in the example 1. The evaluation results of the obtained membrane are shown in Table 1. [Example 4]

[Production ode solid electro JLy te membrane] Only the dope B is cast without using the first dope 114 and the third dope L 16 to form a single layer membrane with the thickness of 50 μm. Other conditions are the same as those in the example 1. The evaluation results of the obtained membrane are shown in Table 1. [Example 5]

The compound shown in chemical formula 1 is used as the solid electrolyte. The proton substitution, that is, the acid processing to obtain the compound of the chemical formula 1 is performed in the membrane production process as described below instead of prior to the dope production. A substance preceding the proton substitution, that is, the precursor of the solid electrolyte is referred to as a ma-terial C. The mateαcial C is dissolved in the solvent to form the dope for casting. The dope is formed in the sajne manner as the dope 24 in the example 1. The solvent is a mixture of the solvent component 1 and tlxe solvent component 2. The solvent component 1 is a good solvent of the material C, and the solvent component 2 is a poor solvesnt of the material C. In the example 5, the chemical formula 1 with the following compositi-on is used: X JLs Na, Y is SO ₂ , Z is (I) of the chemical formula.2, n is 0.33, m is 0.67, the number average molecular weight Mn is 61000, and the weight average molecular weight Mw is 159000.

Material C XOO pts . wt .

Solvent component 1 ( dimethyl sulfoxide ) 256 pts . wt .

Solvent component 2 (methanol) 171 pts . wt .

The dope is cast onto the beILt 82 and peeled off from the belt 82 . The peeled membrane is referred to as trie precursor membrane since the membrane is formed of the material C . The precursor membrane is subject to t lie same processes as those in trie first example , and side edges of the precursor membrane are ciαt off by the edge slitting device 67 . The proton substitution is performed to the precursor membrane through acid processing .

Trien , the membrane is washed . The acid processing is to make the precursor membrane contact to the aqueous acid solut ion . Through trie acid processing , the structure of the precursor is changed to that shown in the chemical formula 1 , that is , the solid eXectrolyte . To contact the membrane to the aqueous a.cid solution , trie aqueous acid solution is continuously supplied, to the liquid bath and the membrane formed of the solid electrolyte is immersed in the aqueous acJLd solution . The -water is used for- washing the membrane after trie acid processing . After the washing , the membrane is transported to the dryi_ng chamber 69 . The evaluation results of the ob tained membrane are shown in table 1 .

[ Example 6 ]

A compound shown in the chemical formula 1 which is different from that shown in the example 5 is used as the solid eXectrolyte . The proton substituti_on for obtaining -the compound of the chemical formula 1 is performed in the membrane production process instead of prior to the dope production . Tlie precursor used as the component of the dope is referred to a s a material

D - The solvent is a mixture of the solvent compon&nt 1 and the solvent component 2 as shown below. The solvent component 1 is a good solvent of the material D , and the solvent component 2 is a poor solvent o f the material D . In the example 6 , the chemical

formula 1 with tlαe following composition is used: X is Na, Y is SO ₂ , Z is (I) ancl (II) of the chemical formula 2, n is 0.33, m is 0.67, the nunVber average molecular weight Mn is 68000, and the weight average molecular weight Mw is 200000. In the chemical formula 2, (I) is 0.7 mol%, (II) JLs 0.3 mol%. Othezr conditions are the same as those in the example 5.

Material D ^r 100 pts . vet.

Solvent 1 ( dimethylsulfoxicle ) 200 pts. wt.

Solvent 2 (methanol) 135 pts. wt . [Example 7 ]

In the example 7, after the -washing, the casting membrane 24a is peeled off from the PET ffilm to obtain t_tie precursor membrane. After the washing, the membrane is immersed in the water at 30° C foir five minutes ancl the water on the membrane is removed after tlxe immersion. Other conditions ara the same as those in the example 5.

[Example 8 ]

The membrane is produced under the same condit ions as those in the example 7 except that the material C is clxanged to the material D.

[Table 1]

Evaluation Item

1 (μm) 2 3 4

(a) (b) (defeσts/m ² ) (S/cm) (W/cm ² )

Example 1 51 ±1.5 0.4 0.08-0.1 0 0.45-0.52

Example 2 52 ±1.5 0.3 0.10-0.1 1 0.50-0.53

Example 3 50 ±3.7 12.4 0.04-0.0 6 0.26-0.30

Example 4 51 ±3.2 9-1 0.07-0.0» 9 0.35-0.43

Example 5 51 ±1.5 0-4 0.11-0.1 3 0.51-0.54

Example 6 52 ±1.5 0-3 0.11-0.1 3 0.52-0.55

Example 7 51 ±1.4 0-3 0.12-0.1 4 0.52-0.55

Example 8 51 ±1.3 0-2 0.12-0. L 5 0.53-0.56

According to the results of the above examples, the present invention enables to continuously produce the soli_d electrolyte multilayermembrane with excellent planarity and reduced defects . The obtained solid electrolyte ππ_xltilayer membrane is suitably used for the solid electrolyte layer of the fueJL cell.

Industrial applicability The solid electrolyte multi.layer membrane, the method and the apparatus; for producing the same, the membrane electrode assembly and tlie fuel cell using tlxe solid electrolyte multilayer membrane of tlie present invention are applicable to the power sources for -various mobile appJLiances and various portable appliances.