WO2007007771A1

WO2007007771A1 - Solid electrolyte multilayer membrane, method and apparatus of producing the same, membrane electrode assembly, and fuel cell

Info

Publication number: WO2007007771A1
Application number: PCT/JP2006/313803
Authority: WO
Inventors: Naoyuki Kawanishi
Original assignee: Fujifilm Corporation
Priority date: 2005-07-07
Filing date: 2006-07-05
Publication date: 2007-01-18
Also published as: EP1911113A4; JP2007018844A; JP5093870B2; EP1911113A1; US20090169943A1

Abstract

First, second and third dopes (114, 115 and 116) containing a solid electrolyte are co-cast from a casting die (89) onto a running belt (82). The casting die (89) is provided with a feed block (119). A catalyst that promotes a redox reaction of electrodes in a fuel cell is added to the first dope (114) and the third dope (116). A casting membrane (112) having a three-layer structure is peeled from the belt (82) as a three-layered membrane (62) and sent to a tenter drier (64). In the tenter drier (64), the membrane (62) is dried in a state that both side edges thereof are held by clips, while stretched so as to have a predetermined width. The membrane (62) is then sent to a drying chamber (69) and the drying thereof is proceeded while supported by rollers.

Description

DESCRIPTI ON

SOLID ELECTROLYTE MULTILAYER MEMBRANE , METHOD AND APPARATUS OF PRODUCING THE SAME , MEMBRAlME ELECTRODE ASSEMBLY , AND FUEL CELL.

Technical Field

The present invention relates to a solid e lectrolyte multilayer membrane , a method and a.n apparatus of producing the solid electrolyt e multilayer membrane , and a membrane electrode as sembly and a fiαel cell using the solid electrolyte multilayer membrane . The present invention especially relates to a solid electrolyte multilayer membrane having excellent proton conductivity used for a fuel cell , a method and an apparatus of producing the solid electrolyte multilayer membrane , and a membrane electrode assembly and a. fuel cell using the solid electrolyte multilayer membrane .

Background Art A lithium ion battery and a .fuel cell that are used as a power source for- portable devices la.ave been actively studied in recent years . A- solid electrolyte used for the above mentioned battery or cell is also actively studied . The solid electrolyte is , for instance , a lithium ion conducting material or a proton conducting matexrial .

The protorx conducting material is generally in -the form of a membrane . The solid electrolyte i_n membrane form, vrtαich is used as a solid electrolyte layer of the fuel cell and the like , and its producing me thod have been proposed . For instance , Japanese Patent Laid-Open Publication No . 9 — 320617 discloses a. method of pnroducing a s olid electrolyte membrane by immersing a polyvinylidene fluoride resin in a liquid in which an electrolyte i

and a plasticizer are mixed. Japanese Patent Laid-Open Publ-Lcation No . 200 .1-307752 discloses a method of prroducing a proton conducting membrane by synthe sizing an inorganio compound in a solution containing an aromatic polymer compouncϋ with the sulfonic acid group , and removing a s olvent therefrom, . In this method , oxides of silicon and phosphoric acid derivative are added to tϊie solution in order to improve micropores . Japanese Patent Laid— Open Publicati-on No . 2002-231270 discloses a method of producing an ion-exahange membrane . In this method, metal oxide precursor is added to a solution containing an ion-exchange resin, and a liquid is obtained by applying hydroILysis and polyc3ondensation reaction to the metal oxide precuxrsor . The ion-exchange membrane is obtained by casting the liquid . Japanese Patent Laid-Open Publication No . 2004 - 079378 discloses a method of producing a proton concϊuσting membrane . In this method, a polymer membrane with a prot on conductivity i_s produced by a solution castiLng method . The membrane is immenrsed in an aqueous solution of an organic compound soluble to water and having a boiling point of not less than 100⁰C , and is allowed to swell to equilibrium. Water is tlien evaporated by heating . In th-is way, the prot on conducting membrane is produced. Japanese Patent Laid-Open Publication No . 2004 - 131530 discloses a method of piroducing a solid electrolyte membrane by dissolving- a compound consisting essentially of polybenzimidazole having the anionic groups into an alcohol solvent containing tetraalJcylammonium hydroxide and having a boiling poin^~t of not less than 90⁰C .

A melt -extrus ion method and the solution casting method are well known methods of forming a membrane from a. polymer . According to the melt: -extrusion method , the membrane can be formed without using a solvent . However , this method has problems in that the polymer may denature by treating , impurities in the polymer remain in the produced membarane , and the liK:e . On the O

other hand, the solution casting metliod has a problem in that its producing apparatuses become large* and complicated! since the method requires a producing apparatus of a solution „ a solvent recovery device and the like . However , this method is advantageous since a heating temperature of the membrane can be relatively low and it is possible to remove the impurities in the polymer while producing the solut ion . The solution casting method has a further advantage in that the produced membrane has better planarity and smoothness than the membrane procSuced by the melt -extrusion metliod .

When the solid electrolyte membrane produced in this way is used for the fuel cell , a catalyst layer is provided on both surfaces of the solid electrolyte membrane in order to promote redox reaction taken place on electrodes of the fuel. cell . The catalyst members and the solid electrolyte membrane have been conventionally produced separately and combined later . In addition, the electrodes for the reiox reaction are incorporated in the fuel cell . The electrodes are also produced in a separate s tep and combined, with the catalyst members and the solid electrolyte membraune . As a method of combining them, there is a press-bonding method, which is one type of lamination . The solid electrolyte membrane and the catalyst members are relatively expensi-ve , hence continuously producing them carries a risk unless stable producing conditions are established . Accordingly , it cannot be helped to make each member separately and combine them later , even though this method is inefficient .

In view of this , methods fo>r continuously producing a so-called membrane electrode assembly (MEA) having the solid electrolyte , the catalyst layers and. the electrodes arre proposed . For example . International Publication No . WO99/34466 ( corresponding to National Publication of Translated Version No . 2002 -500422 ) discloses a method in wtiiσh an electrolyte layer and two catalyst -Layers are co-extruded from a die, and electrodes sheets made fXΌIΠ carbon fiber paper are adhered thereto by pressing them between calendar rolls . The above pu/blication also discloses a method which deposits extruded ca-talyst layers between pre-formed electrolyte sheet and pre-formed electrode sheets. The at>ove publication fur^~ther discloses a method which deposits extruded solid electrolyte layer between pre-formed electrode she&ts and pre-formed two catalyst layers, and adhered together by pressing them between the calendar xrolls.

Japanese Patent Laid-Open Publication No. 2004-047489 discloses a method in which electrolyte ink for forming a first layer, catalys t layer ink for forming a second layeαr and diffusion layer ink for forming a third layeir are simultaneously injected to an applying head so as to be dϊ-sσharged in muitilayer forms on a surface of a continuously running member. In this way, a MEA is formed -

However , in the above-noted Publication No- 9-320617, the solution casting method is denied _M and there remains a problem in that the impurities contained i_n raw materials remain in the produced memb:rane. The methods disclosed in ttie above-noted Publication Nos. 2001-307752, 2002-231270, 2004-079378 and 2004-131530 are on a limited scale and not intended, to be applied in mass production. The method cLisclosed in tlie above-noted Publication No. 2001-307752 has a problem in that JLt is difficult to disperse a complex consisted of the polymer anc3L the inorganic compound. The method disclosed in the above-noted Publication No. 2002-2312^*70 has a problem in tb_at its membrane producing step is complicated. The method disclosed in trie above-noted Publication No. 2004-079378 has a. problem in that the produced membrane is not uniform in planar_L-ty and smoothness since it has micropores fourmed during the immersing in the aqueous solution. Any solution for this problem is not cited in the disclosure. Although it is cited in tlie disclosure that various solid electrolyte membranes can be produced by the solution casting method, any specific method t-herefor is not cited . The method disclosed in the above-noted Publication No . 2004 -131530 limits raw materials to be used and does not mention tlxe usage of other materials having excellent prroperties .

In orcier to produce the fuel cell efficiently, at least the solid electarolyte layer and catalyst layers should be formed at the same time . In addition, the produced fuel cell should have high and uniform quality . According to the metlxods described in International Publication No . WO99/34466 and Japanese Patent Laid-Open Publication No . 200 4- 047489 , efficiency of producing the fuel cell may be improved at some level since the fuel cell is produced integrally . However , it cannot t>e said that the methods aire capable of continuously producing fuel cells integrally "to have uniform quality without loss of the expensive catalyst and solid electrolyte . In addition, both publications do not disclose or suggest improvement of fuel cell properties . The fuel cell properties synergistically eJ-icit respective properties of the solid electrolyte and the catalyst when they are laminated . For example , the solid electrolyte layer is desired to liave high selectivity in mass transfer . That is , the. solid electrolyte is desired to carry ( transmi_t ) only protons , and to blocK fuels such as hycLzrogen or methanol - Meanwhile , the* catalyst layer is desired to have low resistance to electron- transfer, and to carry protons , fuel molecules or^* oxygen molecules with no selectivity. Thus concrete methods .for continuously- laminating the layers having opposite properties , and to assure^ uniform quality of the produced fuel cell shoiald be proposed. Without such methods , it is difficult to realize mass production. of the fuel cell having high, performance , at low cost .

It is an object of the present invention t o provide a solicl electrolyte multilayer membrane that has uniform quality an<3 excellent: ionic conductivity continuously formed from a solid electrolyte, a method ancl an apparatus of producing the solicϋ electrolyte multilayer membrane, and a membrane electrode assembly and a fuel cell vαsing the solid el_ectrolyte multilayeir membrane .

Disclosure of Invention

In order to achieve the above and ottier objects, a method of producing a solid electrolyte multilayer membrane of the present invention includes the step of casting a first dope aαd a second, dope onto a running support so as to form a casting membrane having a first layer of the first d.ope and a second layer of the second dope. The first dope contains an organic solvent and a solid electrolyte that is to be a solid electrolyte layer of a fuel cell. The second dope contains tlie solid electrolyte , the organic solvent and a catalyst that promotes a redox reaction of electrodes in the fuel cell. The methocl further includes tlxe steps of peeling the casting membrane as a wet membrane from tlxe support; performing a first drying of the wet membrane in a state that both side edges theireof are held by holding devices; arxd performing a second drying of the wet membrane supported t>y rollers to form the solid electrolyte multilayer membrane. Ttxe second <3rying step is performed after the first drying step. It is preferable that the first dope is cast from a first casting die and the seconcl dope is cast from a second casting die disposed at a downstream of the first casting die. It is preferable that wet membrane is brought into contact with a compouncl that is a poor solvent of the solid electrolyte. It is preferable that the catalyst includes at least one of Au, Ir, Pt, Rh, Ru, W, Ta, Nb, Ti Pd, Bi, Ni, Co, Fe and Hf. It is also preferable that the catalyst is an alloy of these metals. It is preferable that a thickness of a layer formed from the fiirst dope in the solid electrolyte multilayer membrane is 20 μm t o 800 μm . This layer is derived from the first layer of the cas ting film. It is preferable that a thickness of a layer formed from the second dope in the solid electrolyte multilayer membrane is 10 μm to 50O μm. This layer is derived from the second layer of the casting film .

It is preferable that a third dope containing the solid electrolyte , the organic solvent and the catalyst is cast such that tlxe first dope is interposed between, the second dope and the third oLope . When the f irst dope and the s econd dope are cast from the first casting die and second casting die , respectively , the third elope is preferably cast from a third casting die tb_at is deposecl at an upstream of the first castiixg die . It is preferable that tne catalyst in tlie second dope and the catalyst in the -third dope are different from each other. The solid electrolyte multilayer membrane of the present invention is pro duced according to the above-mentioned methoci .

An apparatus of producing a solid electrolyte multi. layer membrane of the present invention incluc.es a casting dev-Lce , a first drying device and a second drying device . The ca.sting device casts plural dopes from a casting die onto a running support so as to form a layered casting membrane and peels the casting membrane as a layered wet membrane . The plural dopes are a first dope and a second dope . The first dope contains an organic solvent and a solid electrolyte that is to be a solid electrolyte layer of a fuel cell . The second dope contains the solid electrolyte , the organic solvent and a catalyst that promotes a redox reaction of electrodes in the fuel cell . The first drying device dries the wet membrane in a state that both side edges thereof are held by hoILding devices . The second drying device dries tlxe wet membrane supported by rollers to form, the solid electirolyte o

multilayer membrane . The second drying device is disposed at a downstream of the fi_rst drying device -

A membrane electrode assembly o± the present invention includ.es the above -mentioned solid electrolyte multilayer membrane , an anode and a cathode . The anode is adherecϋ to one surface of the sol_ id electrolyte mmltilayer membrane , and generates protons from a hydrogen -containing material supplied from outside . The cathode is adhered to the other surface of the solid electrolyte multilayer membrane , and synthesizes water from the protons permeated through the solicl electrolyte multilayer membrane and gas supplied from outside .

A fuel cell of the present invention inclu<3.es the above- mentioned membrane electrode assembly and current collectors . One of the current collectors is provided in. contact with the anode , and the other current collector is provided in contact with the cathode . The current collector on the anode side receives and passes electrons between, the anode and outside , whereas the current collector on the cathode side recedLves and passes the electrons between the cathode and outside . According to -the present invention , it is possible to continuously produces the solid electro lyte multilayer membrane provided with the catalyst layers that promote the redox xreaction at a low cost . The produced solid electrolyte multilayer membrane has uniform quality and excellent ionic conductivity. ^"When the membrane electrode assembly using this solid electrolyte multilayer membrane is used for the ffuel cell , the fuel cell realizes an excellent electromotive force .

Brief Description of Drawings Figure 1 is a schematic diagrram illustrating a dope producing apparatus ;

Figure 2 is a schematic diagram illustrating a membrane producing apparatus ;

Figure 3 is a sectional view illustrating a simultaneous co-casting device ;

Figure 4 is a schematic diagram illustrating a sequential co-casting device ;

Figure 5 is a sectional view i-llustrating a structure of a membrane electrode assembly that uses a solid electrolyte membrane of the pre sent invention ; and

Figure 6 is an exploded sectional view illustrating a structure of a fuel cell that uses the membrane electrode assembly of the present invention .

Best Mode for Carrying Out the Invention

Embodiments o f the present invention are described below in detail . The pres ent invention, however , is not limited to the following embodiments . A solid elec trolyte multilayer membrane of the present invention is first explained and fo llowed by a producing method ttαereof .

[Material ]

In the present invention, a. polymer havin g a proton donating- group is used as a solid electrolyte , which is formed into a membrane by a producing method described later . The polymer having the proton donating -group is not particularly limited , but may be well-known proton conducting materials having an acid residue . For example , polymer compounds formed by addition polymerization having a sulfonic acid group in side chains , poly(meth) a.crylate having a phosphoric acid <jroup in side chains , sulfonat ed polyether etherketon, sulfonated polybenzimidazole , sulfonated polysulfone , sulfonated heat-resistant aromatic polymer compounds and trxe like are preferably used . As the polymer formed by addition polymerization having a sulfonic acid group in side ctiains , there are perfluorosulfonic acid, as typified by Nafioix (registered trademark) , sulfonated polystyrene, sulfonated poILyacrylonitrile styrene, sulfonated polya-crylonitrile butadiene- styrene and the like. As the sulfonated heat-resistant aromatic polymer coπvpounds , there axre sulfonated poiyimide and the like.

Substances described in, foπr example, Japanese Patent Laid-Open Publication Nos . 4-36613^*7, 6-231779 and 6-342665 are the preferable examples of the perffluorosulfonic a.cid, and the sulDStance represented by the following chemical formula 1 is especially prefer- able above all. However, in "the chemical formula 1, m is in. the range of 100 to 10000, preferably in the range of 200 to 5O00 and more prefexably in the range of 500 to 2000. In addition , n is in the range of 0.5 to 100, and especially preferably in the arange of 5 to 13.5. Moreover, x is nearly equal to m, and y is nearly equal to n.

[Chemical Formula 1]

Compounds described in, fo-ir example, Japanese Patent Laid-Open Publication Nos. 5-174856 and 6-111.834, or the substance represented by the following chemical f ormula 2 are the preferable examples of the sulfonated polystyrene, the sulfonated polyaσrylonitrila styrene and the sulfonated polyscrylonitrile butadiene- s tyrena .

[Chemical Formula 2]

Substances described in, ffor example, Japanese Patent Laid-Open Publication Nos. 6-49302, 2004-10677, 2004-345997, 2005-15541, 2002-110174, 2003-100317, 2003-55457, 9-245818, 2003-257451 and 2002-105200, and International Publication No. WO97/42253 (corresponding to National Publication of Translated Version No. 2000—510511) are the examples of tlxe sulfonated heat-resistant aromatic polymer compounds, and tine substances represented by the following chemical formulae 3 and 4 are especially preferable above all.

[Chemical Formula 3]

[Chemical Formula 4]

Sulfonat±on reaction on tlie process of obtaining the above-mentioned compounds can be performed in accordance with various synthetic methods described in the discloseσl publications . Sulfuric acid ( concentrated sulfuric acid) , fuming sulfuric acid, gaseous or liquid sulfur trioxiLde , sulfur trioscide complex, amidosulfuric acid, chlorosulfoni-c acid and the lilke are used as sulfonating agents . Hydrocarbon (benzene , toluene, nitrobenzene, chlorobenzene , dioxetane and the like ) , alkyl halide ( dictiloromethane , chloroform, di chloroethane , tetrachloromettiane and the like) and the like are used as a solvent . Reaction temperature in the sulfonation reaction is determined within the range of - 20 ⁰C to 200 ⁰C in accordance with the sulfonating agent activity . It is also possible to previously introduce a meircapto group, a disulfide group or a sulfiniσ acid group in a monomer , and synthesize the sulfonated c ompound by the oxidation reaction with an oxiσlant . In this c ase , hydrogen peroxide , nitri_c acid , bromine water , hypochlorite , hypobromite , potassium permanganate , chromic acid and the like are used as the oxidant . Wateir , acetic acid, propionic acid and the like are used as the solvent . The reaction temperature according" to this method is determined within the range of a. room temperature ( for example , 25⁰C ) to 200⁰C in accordance with the oxidant ac tivity . It is also possible to previously introduce a halogeno- alkyl group in the monomer, and synthesize tlxe sulfonated compound by the substitution reaction of a sulfite , hydrogen sulfit e and the like . In this case , water , alcohol , amide , sulfoxide , s iilfone and the o

like are used, as the solvent. The reaction temperature according to this method is determined within the range of the room temperature C for example, 25°C ) to 200⁰C . The solvent used for the above-mentioned sulfonation reactions can be a mixture of two or more substances.

In the reaction process to synthesize ttie sulfonated compound, an alkyl sulfonating agent can t>e used, and Friedel-Craf"ts reaction (Journal of Applied Polymerr Science, Vol. 36, 1753-1767, 1988) using a sulfone and AlCl₃ is a common method. When using tlie alkyl sulfonating agent for the F^*riedel-Crafts reaction, hydrocarbon (benzene, toluene, nitrobenzene, acetophenon, chlorobenzene, trichlorobenzene and the like), alkyl halide ( dichloromethane, chloroform, dJLσhloroethane, tetrachloromethane, trichloroethane, tetrachloroethane and the like) and tlie like are usecl as the solvent. The reaction temperature ±s determined in the range of the room temperature to 200 ⁰C - The solvent used for the above-mentioned Friedel-Craffts reaction can. be a mixture of two or more substances .

The solid electrolyte preferably has -the following properties . An ionic conductivity is preferably not less than 0.005 S/cm, and more prefeirably not less than 0.01 S/cm at a temperature of 25⁰C and at a relative humidity of 70%, for example. Moreover, after the solid electrolyte membrane has been soakecl in a 50% methanol aqueous solution for a day at the temperatur/e of 18⁰C , the ionic conductivity is not less than 0.003 S/cm, and more prefer-ably not less than 0.008 S/cm. At this time, it is particularIy preferable that a. percentage of reduction in the ionic conductivity of the solάd electrolyte as compared to that before the soaking is not more than 20% . Furthermore, a methanol diffusion coefficient is preferably not more than 4 xlO^"7 cm²/sec, and especially preferably not more than 2xlO^'7 cm²/sec.

As to strength, the solid electrolyte membrane preferably has elastic modulus of not less than 10 MPa, and especially preferably of not less than 20 MPa. Note that the measuring method of the elastic modulus is described in detail in paragraph [0138] in Japanese Patent Laid-Open Publication No.2005-104148. The above-noted values of the elastic modulus arre obtained by a tensile tes ter (manufactured by Toyo Baldwin Co . , Ltd. ) . In order to obtain the elastic modulvrs of the solid electrolyte by other testing methods or testers , it is preferabl_e to previously correlate the value thereof with that of the above-noted testing method and the tester.

As to durability, aftezr a test with time i_n which the solid electrolyte membrane has been soaked into the 50% methanol aqueous solution at a constant temper-ature, a percentage of change in each of weight, ion exchange capacity, and the methanol diffusion coefficient as compared to ttiat before the soaking is preferably not more ttian 20%, and especially preferably not more than 15%. Moreover, in a test with time in hydrogen peroxide, the percentage of change in each of the weight, the ion exchange capacity and the methanol diffusion coefficient as compared, to that before the soaking is preferably not mor-e than 20%, and especially preferably not more ttxan 10%. Furthermore, coefficient of volume expansion of the sol.id electrolyte membrane in the 50% methanol aqueous solution at a constant temperature is preferably not more than 10%, and especially preferably not more than 5%.

In acldition, it is preferable that the solid electrolyte has stable ratios of water absorption and water content. It is also preferable that the solid electrolyte bxas extremely low solubility in alcohol, watezr, or a mixture of alcohol and water to the extent that it is practically negligible. It is also preferable that weight reduction and shape change of the solid electrolyte membrane af ter it has been soaked in the above-mentioned liquid are also small enough to be practically negligible .

Wlαen the solid electrolyte is formed! into a membrane _^ an ion -concϋuc ting direction is preferably higher in a thickness direction of the membrane as compared to other directions thereof . The ionic conductivity bass ically depends on a ratio of the ionic conductivity to methanol transmission coefff icient . Therefore , the ion- conducting direction may be random. A ratio of the ionic conductivity to methanol diffusion coefficient is represente d as performance index . The hi_gher the index is , the higher the ionic conductivity of the solid electrolyte is . As long as the solid electrolyte has uniform performance index, ionic resistance and the mettαanol transmission of the solid electrolyte membranes; can be uniform by adjusting tlie membrane thickness . The thickness of the membrane is prefer- ably in the range of 10 μm to 300 μm . The ionic resistance is pr-oportional to the thickness , whiles the methanol transmission amount is inversely proportional to the thickness . Therefore, when the ionic conductivity and the methanol diffusion coefficient are bottα. high in the s olid electrolyte , it is especially preferable to produce the membrane with a thickness of 50 μm to 200 μm . When tlie ionic condμcti-vity and the methanol diffusion coefficient are both low in the s olid electro lyte , it is especially preferable to produce the membrane with trie thickness of 20 μm to 100 μm.

Allowable temperature limit is preferably not less than 200 ⁰C , more preferably not less than 250 ⁰C , and espeσi_ally preferably not less than 3 O0 °C . The allowable temperature Limit here means the temperature at which reduction in weight off the solid electrolyte membrane reaches 5% as it is heated at a rate of I⁰C /min . Note that ttie weight reduσti-on is calculated with the exception of evaporated contents of water and the lilce . Wlien the solid electrolyte is formed in the membrane form and usedL for the fuel ceil, the maximum power (output) density thereof is preferably not less than 10 mW/cm².

By use of the above- described so÷Lid electrolyte, it is possible to produce a solution dope preferable for the membrane production, and at the same time, it is possible to produce the solid electrolyte membrane preferable for the fuel cell- The solution, preferable for the membrane production is, for example, a solution whose viscosity is relatively low, and from which foreign matters are easily removed through, filtration. Note that the obtained solution is hereinafter referred to as the dope.

Any organic compound capable of dissolving the polymer as the solid electrolyte can be the solvent of the dope . For example , there ar/e aromatic hydrocarbon (for example, benzene, toluene and the like ) , halogenated hydrocarbon ( for example , dichloromei thane , σhlorobenzene and the .Like), alcohol (for example, mettianol, ethanol , n-propanol, n-butanol, diethylene glycol and the like) , ketone (for example, ace -tone, methyletϊiyJL ketone and the like), ester (if or example, meth.ylacetate , ethyl_acetate, propylaσetate and the like), ether C for example, tetrahydrofuran, methyl cellosolve and the like) , nitrogen compound (N-methylpyrrolidone (NMP), N,N-dimethylforrnamide (DMF), N, N' - dime thylaσet amide (DMAc) and the like) and. so forth. Note that the solvent may be a mixtuxre of a plurality of the substaixces.

In order to improve the various properties of the solid electrolyte membrane, it is possible to adLd additives to the dope. As the additives, there arre antioxidants, fibers, fine particles, water absorbing agents, plasticizers and σompatibilizing agents and the like. It is preferable that a concentration of these additives is in the range of not less than 1 wt . % and 30 wt . % or less when the entire solid contents of the dope is 100 wt.% . Note, however- , that the concentration and the sorts of the additives have to be determined not to adversely affect on the ionic conductivity. Hereinafter, the additives are explained in detail.

As the antioxidants, (hindered) phenol-type compounds, monovalent or divaleixt sulfur -type compounds, trivalent phosphorus -type compounds, benzophenone-type compounds, benzotiriazole-type compounds, hinderecl amine-type compounds, cyanoaσrylate- type compounds, salicylate -type compounds, oxalic acid anilide-type compo"unds are the preferable examples . The compounds described in Japanese Patent La.άd-Open Publication Nos . 8-05363-4, , 10-101873, 11 -114430 and 2003-151346 are the specific examples thereof.

As the fibers, perrf luoroσarbon fzLbers, cellulose fibers, glass fibers, polyethylene fibers and the like are the preferable examples. The fibers ciescribed in Japanese Patent LaicS-Open Publication Nos. 10-312815, 2000-231938, 2001-3O7545, 2003-33.7748, 2004-063430 and 2004-107461 are the specific examples thereof.

A.s the fine particles, titanium oxcide, zirconium oxi_<le and the lilke are the preferable examples. The fine par-fcicles described in Japanese* Patent Laid- Open Publication Nos . 2003-1^*78777 and 2004-217931 are the specific examples thereof.

As the water absorbing agents, that is, the hydro^hilic materials, cross -linked polyacrylate salt, starch -aery late salt, poval (polyvinyl alcoti.ol) , polyacrylonitrile , carboxyrnethyl cellulose, polyvinyl pvyrrolidone, polyglycol dialkyl ether, polyglycol dialkyl ester, synthetic zeolite, titania. gel, zirconia gel and yttria. gel are the piref erable examples . The water absorbing agents described in Japanese Patent Lai<i-0pen Publication Nos . 7-1350O3, 8-020716 and θ-251857 are the specific examples thereof .

As the plasticizexrs, phosphoric a.cid ester-type corn-pound, chlorinated paraffin , alkyl naphthaleme-type compound , sulfone alkylam±de-type compound, oligoether group , aromatic nitrile group are the preferabXe examples . The plasticizers de scribed in Japanese Patent Laicl-Open Publication Nos . 2003-28S 916 and 2003- 3-L 7539 are the specific examples thereof .

A-S the compatibilizing agents , those having a boiling point or a sublimation point of not less than 250⁰C are prefr erable, and those having the same of not la ss than 300 ⁰C a.re more pref erable .

T"he dope may contain various kinds of polymer compovinds for the purpose of ( 1 ) enhancing the mechanical strength of the membrane , and ( 2 ) improving the acid concentration in the membrane .

For the purpose of ( 1 ) , a polymer having a moleculanr weight in the range of IOOOO to 1000000 or so and well compatible with ( soluble to) the sol:LcL electrolyte is preferably used . For example , the polymer such as perf luorinated polymer , polystyrene , polyethylene glycol, jpolyoxetane , pol_yether ketone , polyether sulfone , and the polymer compound having the repeating unit of at least two of these polymers are preferable . Preferatjly, the polymexr content of the membrane is in the range of 1 wt . % to 30 wt . % of the total weight . It is also possible to use the compat-±bilizing agent in order to enhance the compatibility of the polymer with the solid electrolyte . As the compatifcilizing agent , those having thte boiling point or the sublimation point of not less than 250⁰C are preferable, and those having the same of not less than 300⁰C are more preferable .

For the purpose o f ( 2 ) , proton acid segment -having polymer , and the like are preferably used. Perfluorosulfonic acid polymers such as Nafr ion ( registerecϋ trademark) , sulfonated polyetlier etherketon having a phosphoric acid group in sicϋe chains , and the sulfonated heat-resistant auromatic polymers such as sulfonated polyethex sulfone , sulfonated polysulfone , sulfonated polybenz imidazole and the like are the preferable examples thereof . Preferably, the polymer content of the membrane is in the range of 1 wt . % to 3O wt . % of the total, weight .

When the obtained solid electro lyte membrane is -used for the fuel cell , an active metal catalyst that promotes t lie redox reaction of anode fuel and cathode fuel may be added to the dope . By adding the active metal catalyst , the fuel having penetrated into the solid electrolyte from one e lectrode is well consumed inside the solid electrolyte and does not reach the other electrode , and therefore this is ef fective for preventing a crossover phenomenon . The active metal catalyst is not particularly limited as long as it functions as an e lectrode catalyst , but platinum or platinum-toased alloy is es pecially preferable .

[ Dope Production ]

In Fig . 1 , a dope producing apparatus is shown. . Note , however , that the present invention is not limited to the dope producing apparatus shown in Fig . 1 . A dope producing apparatus 10 is provided with a solvent tank 11. for storing the solvent , a hopper 12 for supplying the solid el&ctrolyte , an addi-fcive tank 15 for storing the additive , a mixing tank 17 for mixing the solvent , the solid electrolyte and the additive so as to make at mixture 16 , a heater 18 for: heating the . mixture 16 , a temperature controller 21 for controlling a temperature of the heated mixture 16 , a filtration device 22 for filter-ing the mixture 16 fed out of the temperature controller 21 , a flash device 26 for con. trolling a concentration of a dope 24 from the filtration device 22 , and a filtration device 27 for filtering the concentration-contro lled dope 24 . Tϊie dope producing apparatus 10 is further provided with a recover~y device 28 for recovering the solvent , and a refining device 29 for refining the recovered solvent. The dope producing appazratus 10 is connected to a membrane producing apparatus 33 thr-ough a stock tanlc 32. Note that the dope producing apparatus is also provided with valves 36, 37 and 38 for controlling amount of feeding, and feeding pumps 41 and 42. The number and the positi_on of the valves and feeding pumps are changed as appropriate.

First of all, the valve 37 is opened to feed "the solvent from the solvent tank 11 to the mix-ing tank 17. Successively, the solid electrolyte stored in the hopper 12 is sent to the mixing tank 17. At this time, the solid electrolyte may be continuously sent by a feed-Lng device that perfrorms measuring and sending continuously, oar may be intermitten.~tly sent by a feading device that measures a predetermined amount of the solid electrolyte first and sends the solid electrolyte, of that amount. In addition, an additive so3_ution is sent by a. necessary amount from the additive tank 15 to the mixing tanfe 17 by adjusting" the degree of opening of tlie valve 36.

In the case where the additive i_s liquid at room t eraperature, it is possible to send the additive in a liquid state to the mixing tank 17 instead of sending it as solution. Meanwhile, in the case where the additive is solid, it is possible to send the additive to the mixing tank 17 by using the hopjper and so forth . When plural kinds of additi-ves are added, the additive tank 15 may contain a solution inwhich the plural kinds of the additives aαre dissolved. Alternatively, many additive tanks may be used for irrespectively containing a solution in which one kind of the additive is dissolved. In this case, the additi^ve solutions are ^respectively sent to the mixing tank 17 throught an independent pipe.

In the above description, the solvent, the solid electrolyte andL the additive are sent to the mixing tank 17 in this order. However, this order is not exclusive. For example, the solvent of an appropriate amoun.1: may be sent after the solid electrolyte has been sent to the mixing tank 17. By the way, the additive is not necessarily contained in the mixing tank 17 beforehand . Ttie additive may be mixed in a mixture of the solid electrolyte and the solvent during a succeeding pxocess by an in-line mixing method and so forth . To mix a predetermined catalyst into the dope 24 , the catalyst may be mixed j_nto the solid electrolyte ancϋ the solvent instead of or in addition to the above additives . It is also possible to send the cataLyst from the hopper 12 along with the solid electrolyte to make the mixture 16 .

It is preferable that the mixing tank 17 is provided with a jacket for covering an outer surface thereof , a ffirst stirrer 48 rotated by a motor 47 , and a second stirrer 52 rotated by a motor 51 . A temperature of the mixing tank 17 is regulated by heat transfer medium flowing in side the jacket . A preferable temperature range of the mixing tank 17 is - 1O ⁰C to 55⁰C . The first stirrer -48 and the second s tirrer 52 are properly selected and used to swell the solid electrolyte in the solvent so that the mixture 16 is obtained . Pref erably, the first s tirrer 48 has an anchor blade and the second stirrer 52 is a decent ering stirrer of dissolver type .

Next , ttie mixture 16 is sent to the heater 1 8 by the pump 41 . It is preferable that the heater 18 is piping -with a jacket (not shown) fox letting a heat transfer medium flo-w between the piping and the jacket . It is furrther preferable thtat the heater 18 has a pressxxxe portion (not shown ) for pressurizing the mixture 16 . By using this kind of the heater 18 , solid contents of the mixture 16 are effectively and efficiently dissolved into the solvent under^* a heating condition or a pressurizing/heating condition . Hereinafter , the method of dissolving the solid contents into the solvent by heating is referrred to as a heat -dissolving method . In this case , it is preferrable that the mixture 16 is heated to have tlie temperature of 60⁰C to 250 °C .

In stead of the heat-dis solving method, it is possible to perform a cool-dissolving mettxod in order to dissolve the solid contents into the solvent . The cool -dissolving method is a method to promote trie dissolution whi le maintaining trie temperature of the mixture 1 6 or cooling the mixture 16 to have lower temperatures . In the cool-cϋissolving method, it is preferable that the mixture 16 is cooled to - 100 ⁰C to - 10 ⁰C . The above-mentioned heat -dissolving method and tlxe cool-dissolving method make it possible to sufficiently dissolve the solid electrolyte in the solvent .

After the mixture 16 has reached about a room temperature by means of the temperature controller 21 , tlie mixture 16 is filtered by the filtration devi_ce 22 to remove foreign matter like impurities or aggregations contained therein . The filtered mixture 16 i s the dope 24 . It is preferable tϊiat a filter used for the filt oration device 22 has an average poire diameter of 50 μm or less .

The dope 24 after the filtration is sent to and pooled in the stock tank 32 , and used ffor producing the membrane .

By the way , the method of swelling the so lid contents once and dissolving it to produce the solution as described above takes a longer time as a concentrati_on of the solid electrolyte in the solution incieases , and it causes a problem concerning production efficiency . In view of this , it is preferable that the dope is prepared to have a lower concentration relative to an intended concentration , and a concentration process is performed to obtain the intended concentration aft er preparing the cϋope . For example , the dope 24 filtered by the filtration device 22 is sent to the flash device 26 by the valve 38 , and the solvent of the dope 24 is partially evaporated in the flash device 26 to be concentrated. The concentrated dope 24 is extracted from th.e flash device 26 o

by the pump 42 and sent to tϊie filtration device 27. At the time^, of filtration by the filtration device 27, it is preferable that a temperature of the dope 24 is 0⁰C to 200⁰C . After removing- foreign matter by the filtration device 27, the dope 24 is sent to and pooXed in the stock tank 32, and usecl for producing the* membrane. Note that the concentrated dope 24 may contain bubbles. It is therefore preferable that a defoaming p-rocess is performeάl before sending the dope 24 "to the filtration, device 27. As thes method for removing the bubt>les, various well-known methods are applicable . For example, there is an ultrasonic irradiation* method in -which the dope 24 is irradiated with an ultrasonic.

Solvent vapor generated due to the evaporation in the flashi device 26 is condensed by the recovery device 28 having a condenserr (not shown) and becomes a liquid to be recovered. The recovered solvent is refined by the xrefining device 29 as the solvent to be reused for preparing the dope. Such recovering and reusing are advantageous in terms of production cost, and also prevent adverse effects on human bodies and the envi-ronment in a closeca. system.

By ttαe above method, the dope 24 having the solid electrolyte concentration of 2 wt.% or more and 50 wt.% o>r less is produced _ It is more preferable that the solid electrolyte concentration is 15 wt.% or more and 3O wt.% or less. Meanwhile, as to Θ. concentration of the additive, it is preferable that a range thereof is 1 wt.% or more and is 30 wt.% or Less when the entire solid contents of the dope is defined as IOO wt.%.

[Meiαbrane Production. ]

Hereinafter, a method of producing th<e solid electrolyte multilayer? membrane is explained. In Fig. 2, the membrane producing apparatus 33 is shown. Note, however, that the presen~t invention is not limited to the membrane producing apparatus show-n in Fig.2. In the present invention, a plurality of dopes having different compositions fαrom one another is co-casted . Note that Fig . 2 shows only one dope sent from the dope producing apparatus 10 in orcler to simplify the drawing . The method of co-casting will be explained later in detail with referring to Figs . 3 and 4 .

The membrane producing apparatus 33 is provided wi_th a filtration device 61 for ^removing foreign matter contained in the dope 24 sent from the stock tank 32 ; a oasting chamber 63 for casting the dope 24 filtered by the filtration device 61 to form a solid electrolyte mu-Ltilayer membrane ( hereinafter, merely referred to as the membrane ) 62 ; a tenter drier 64 for dryin g the membrane 62 while transporting it in a state that both side edges thereof are held by clips ; a poor solvent contact device 6 5 for bringing a compound, wliich is a poor solvent of the solid electrolyte , into contact with the membiane 62 containing the solvent , for example, before feeding tlxe membrane 62 into the tenter darier 64 ; an edge slitting device 67 for cutting off both side edges of the membrane 62 ; a drying chiamber 69 for dryiixg the membrane 62 while transporting it in a state that the memtirane 62 is supported by rollers 68 ; a cooling chamber 71 for cooling the membxrane 62 ; a neutralization device 7 2 for reducing a charged voltage of the membrane 62 ; a knurling roller pair 73 for perf ormi_ng emboss processing on both s±dLe edges of the meiαlDrane 62 ; and a winding chamber 76 for wind±rxg up the membrane 62 .

Tire stock tank 32 ±s provided with a stirrer 78 rotated by a motor 77 . By the rot ation of the sti_rrer 78 , deposition or aggregat ion of the solid, contents in the dope 24 is inhib ited . The stoσlc tank 32 is connected to the filtnration device 61 thorough a pump S 0 .

A oasting die 81 ffor casting the dope 24 , and a belt 82 as a runnirxg support are parovided in the casting chamber 63 . As a material, of the casting <ϋie 81 , precipitation hardened stai-nless steel ά_s preferable and it is preferabl_e that a coefficient of thermal, expansion thereof is 2 x 10^"5( °C ^"1 J or less - It is preferable that the material has anti-cor-rosion properties , which is subs tantially equivalent with SUS316 on a compulsory corrosion examination performed in an electroHyte aqueous so lution . Further , it is preferable that the material has anti-co>rrosion properties in which pitting is not caused at a gas -liquid interface after soaked in a mixed liqiαid of dichlorome thane , methanol and water for three months . Moreover , it is preferable to make the casting die 81 by grinding a. material after at least one month has passed from foundry . In ^virtue of this , the dope 24 unifformly flows ins ide the casting die 81 and it is prrevented that st ireaks are causedL on a casting membarane 24a describecl later . As to f inishing accuracy of a dope contact surface of the casting die 81 ,. it is preferable that surface xroughness is lμm or less and strraightness is 1 μm/m or less ±_n any direction. . Slit clearance of the casting die 81 is adajpted to be automatically adjusted within the range of 0 . 5 mm to 3 . 5 mm. With respect to a corner portion of a J- ±p edge of the casting die 81 , a criamfered radius R thereof is adapted to be 50 μm or less in the entire width . Furthexrmore , it is prreferable that title casting die 81 is a coat -hanger type die .

A width of the casting die 81 is not especially Ximited . Howevex: , it is preferable that the width thereof is 1 . 3. to 2 . 0 times a width of a membrane as a final product . Moreove-ir , it is preferable that a temperature controller is attached- to the casting die 81 to maintain a predetermined temperature of the dope 24 durzLng membrane formation . Furthermore , it is prefera-ble that heat bolts for adjusting a thickness are disposed in a width direction of the casting die 81 at predetermined internals and the casting die 81 is provided with an automatic ttiickness adjusting mechanism utilizing the heat ^"bolts . In this case , the heat bolt sets a profile and forms a membrane along a prese ^~t program in accordance with a liquid amount sent by the pump 80 . In order to precisely control the sending amount of the dope 24 , the pump 80 is preferably a high-accuracy gear pump . Furthermore , feedback control may ^"be performed over the automatic thickness adjus ting mechanism . In this case , a tltiickness gauge s uch as an infra-red thickness gauge is disposed a.t the membrane producing apparatus 33 , and the feedback control is performed along an adjus tment program on the basis of a prof ±le of the thickness gauge and a detecting result from the thickness gauge . It is preferable that the casting die 81 is capable of adj usting the slit clearance of the lip edge to be -t 50 μm or less so as to regulate a thickness difference between any two points , whi_ch are located within an area excepting an edge portion , of the membrane 62 as -the final product to be 1 μm orr less .

Preferably, a hardened layer is formed on the lip edge of the casting die 81. A. method for forming the hardened, layer is not especially limited . There are ceramic coating , hard chrome-plating, nitrzLding treatment method and so for-th . When the ceramic is utilized as the hardenecϋ layer, it is pnref erable that the ceramic has grrindable properties , low porosity, strength , excellent resistance to corrosion, and no affinity, and no adhes iveness to the elope 24 . Concretely, there are tungsten carbide (WC ) , Al₂O₃, TdLN, Cr₂O₃ and so forth . Among these , the WC is especially preferable . It is possifcle to perform WC coating by a thermal spraying method.

It is preferable* that a solvent supplying device (not shown ) is at tached near the JLip edge of the casting die 81 in. order to prevent the dope from ^"being partially dried and solidified at the lip edge . It is preferable to supply a. solvent to a peripheral portion of three-phase contact lines formed by both end portions of a casting bead, both end portions off the lip edge an.d. ambient aijT. It is preferable to supply the solvent to each si±de of the end portions at a rate of 0 . 1 mL/min to L . O mL/min . Owing to this , foreign matter such as the solid contents separated out from the dope 24 , or extraneou s matter mixed into the casting bead from OU. t side can be prevented from entering" into the casting membrane 24 a . As a pump for supplying the solvent , it is prefferable to us e the one having a pulsation rate of 5% or less .

The belt 82 unόLer the casting die 81 is supported by the ro llers 85 and 86 . T-he belt 82 is continuously transported by thte rotation of at least one of these rollers 85 an<3. 86 .

A width of the belt 82 is not especially limited . However , it is preferable that "the width of the ^"belt 82 is 1 . 1 to 2 . 0 times ttLe casting width of the dope 24 . Pr-eferably, a length of the beilt 82 is 20 m to 20 0 m, and a th.icfc.ness thereof is 0 . 5 mm to 2 . 5 mm. It is prefera.t>le that the belt 82 is ground so as to have su.:rface roughness of 0 . 05 μm or less .

A material of the belt 82 is no t especially limited, but prref erably stainless - As the material of the belt 82 besides stainless , there are nonwoven plastic films such as polyethylene texephthalate (PET) film, polybutylene terephthalate ( PBT ) film, nylon 6 film, nylon 6 , 6 film, polypropylene film, polycarbonate f i_lm, polyimide film and the like . It ά_s preferable to u.se lengthy material having enough chemical stability for the used solvent and enough heat resistance to the membrane forming temperature . It is preferable that a heat transfer medium circulator 87 , wtxiσh supplies a heat medium to the rollers 85 and 8 6 so as to control surface tempeiratures thereof , is attached to the rollers 85 and 86 . For this configuration, a surface temperature of the belt 82 is kept at a predetermined value . In this embodiment , a passage (not shown ) for the heat transfer medium is formed in trie respective rollers 85 and 86. The heat transffer medium maintained at a predetermined tempenrature passes through the Δo

inside of the pas sage to keep a temperature of the respective rollers 85 and 86 at a predetermined value . The surface temperature of the belt 82 is appropriately set in accordance with a kind of the solvent , a kind of the solid contents , a concentration of the dope 24 and the like .

Instead of the rollers 85 and. 86 , and the 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 drum is capable of accurately rot ating with rotational speed unevenxiess of 0 . 2% or less . Moreover, it is preferabLe that the cast-ing drum has average surface roughness of 0 . 01 μiα or less . The sixrface of the casting drum is hard chrome platecl so as to have sufficient hardness and durr ability . Further-more , it is preferable to minimize surface ciefect of the casting drum, belt 82 , and rollers 85 and 86 . Concretely, it is preferable that there i_s no pinhole of 30 μm or more , and a number of tine pinholes of 1 0 μm or more and less than 30 μxm is at most one per square meter , and a number of the pinholes off less than 10 μm is at most two per square meter .

It is pref exrable to dispose a decompression σlxamber 90 for controlling a pressure of the cas ting bead, which is formed between the casting die 81 and the b>elt 82 , at its upstream side in the running direction of the be it 82 .

Air blowers 91 , 92 and 93 that blow air for valorizing the solvent of the casting membrane 24a , and an air shielding plate 94 that prevents the air causing uniαniformity in a shape of the casting membrane 24a from blowing oxito the casting membrane 24a are provided neaαtr the casting die 81 .

The casting chamber 63 is provided with a temperature regulator 97 for maintaining an ins ide temperature thereof at a predetermined value , and a condenser 98 for condensing and recovering solvent vapor . A recovery device 99 fo3c recovering the condensed an<3. devolatilized organic solvent is disposed at the outside of the casting chamber 63.

The poor solvent contact device 65 brings a Xiquid into contact with the membrane 62. Ttiis liquid is the poor solvent of the solid electrolyte that is combined with the catalyst in one dope. There are various ways to bring the liquid as the poor solvent into contact with the membrane 62. For example, the liquid as the poor solvent is sprayed onto the membrane 62. The membrane 62 may be fed into the atmosphere in whicfci misted or vaporized poor solvent exists. Ht is also possible to soak the membrane 62 into a bath storing the liquid as the poor solvent, or to coat the membrane 62 with the liquid as the poor solvent. Among these mettiods, the misting , the use of the vaporized poor solvent and the coating are prefeirable . The position, of the poor solvent contact device 65 is not limited to the configuration shown in Fig. 2. The poor sol-vent contact device 65 may be disposed, for example, right befoire the tenter drier 64 or between the tenter drier 64 and the drying chamber 69. However, the poor solvent contact device 65 is preferably disposed at a position where the drying of the layers containing the catalys t is not yet proceeded much.

The coating method is not particularly limiteci as long as the membrane 6Z is continuously coated with the po>or solvent. Preferably used, are extrusion coating, die coaters siαch as slide and the like, roll coaters such as forward roll coater, reverse roll coater, gravure coater and the like, rod coatenr on which a thin metal wire is wound around, and the like. These methods are described in "Modern Coating and. Drying Technology" edited by Edward Cohen and Edgar B. Gutoff (published by VCH Publishers, Inc., 1992). The rod coater, the gravure coater and a ^"blade coater, which can be stably operated even when a small amount of the poor solvent is usecϋ for the coating, are preferable among them.

When a nonflammable liquid such as water is used as the poor solvent, it is possible to adopt the soaking, the spraying and the use of the misted or the gasified poor solvent.

As the misting or the spraying method, a spraty nozzle which is utilized for air humidifiσafc ion, spray painti_ng, automatic cleaning of a tank and so fortti may be used. For example, a plurality of the spray nozzles is disposed along the width direction of tlie membrane 62 and spray the poor so J.vent onto the membrane 62 across the entire width thereof . As the spray nozzle, full cone spxray nozzles, flat spray nozzles and the like manufactured t»y H. IKEUCHI & CO _ , LTD. or Spraying Systems Co. may be used.

In order to maintain high concentration of the gasified poor solvent in the atmosphere, evaporation of the poor solvent may be enhanced by the use of an atomizer, or volatilization of the poor solvent i.n liquid form may be enhanced by he.at. Method of measuring gas concentration differs according to the type of the used poor solvent. The gas concentration may be measured by, for example, gas detecting tube, contact-combustion type gas detector, electrochemical gas detector, infrared gas detectox and the like. When flammable poor solvent is used, it is prreferable that nitrogen is pxreliminary substituted for air.

When the gasified poor solvent is brought into contact with the membrane 62 , saturated vapor concentration in "the atmosphere is preferably €0% to 95%, more preferably 60% to 9O%, and further preferably 70% to 90%.

When the membrane 62 is fed into the atmosphere in which the concentration of the gasified poor solvent is hi_gh, it is ideal to make the membrane 62 into contact with the atmosphere until the membrane 62 reaches equilibrium, in which the concentrations of the reactants and products have no net change over time. However it is impossible to proceed the impregna-tion until the membrane 62 reaches the equilibrium, since the membrane 62 is continuously transported. Therefore, the time for making the membrane 62 into contact with tϊie atmosphere is pareferably in the range of 10 sec to 300 sec, more preferably 10 to 180 sec, and most preferably 30 sec to 300 sec.

The poor solvent is not strictly limited as long as it is a poor solveni: of the solid electrolyte polymer "that is combined with the catalyst in one dope. The solubility of the solid electrolyte -Ln the poor solvent is preferably 1% or less. The poor solvent may be a mixture of a plurality of substances . However, substances that make -the membrane 62 extremely white or cloudy, or extremely soft are not preferable. Those described in Shinpan ϊozai Pokettobukfcu (The New Solvent Pocketbook) (published by Ohmsha, 1994) are the examples of the organic solvent to be the poor solvent . but the present .invention is not limited to them. For example, alcohol group (methanol, ethanol, n-propanol, -±sopropanol, π-butanol, isobutanoJ_, cyσlohexanol, benzyl alcoϊiol, fluorinated alcohol), keton group (acetone, methylethyl ketone, methyl zLsobutyl ketone, cyclohexanone) , ester group (methylaσetate _M ethylacetate, butylacetate) , polyalcohol gjroup (ethylene glycol, diethylene gJJLycol, propylene glycol, ethylene glycol diethyl ether), N,N-dimethylformamide, perfluorotritmtylamine, triLethylamine, dimethylformamide , dimethylsulfoxide, methyl cel-losolve, and the like.

A transfer section 101 that is disposed downstream from the casting chamber 63 is provided with an air blower 102. The edge slitting device 67 is providecl with a crusher 1O3 for shredding side edges exit from the membxrane 62.

The drying chamber 69 is provided with an absorbing device 106 to absozrb and recover solvent vapor generated due to evaporation. In Fig. 2, the cooling chamber 71 is disposed downstream from the drying chamber 69. However, a humidity-con-trolling chamber (not shown) for controlling water content of tine membrane 62 may be disposed between the drying chamber 69 ancϋ the cooling chamioer 71. The neutralization device 72 is a forced neutralization device like a neutralization bar and the like, and capable of adjusting the charged voltage of the membrane 62 within a predetermined range (for example, -3 kV to +3 kV) . Although the neutralization device 72 is disposed at the downstream side from the cooling device 71 in Fig" .2, this setting position is not exclusive, T"ϊie knurling roller pair 73 forms knurling on both side edges of tlie membrane 62 by emboss processing , The inside off the winding chainber 76 is provided with a winding roller 107 for winding the membrane 62, and a press roller 108 for controlling tension at tlxe time of winding.

Next, an embodiment of a method for producing the membrane 62 by using tlie above-described membrane producing apparatus 33 is described . The dope 24 is always uniformed, by the rotation of the stirrer 78. Various additives may be mixed in the dope 24 during the stir.

The dope 24 is sent to "the stock tank 32 by the pump 80, and deposition or aggregation of the solid contents in the dope 24 is inhibited by the stir. After that, the dope 24 is filtered. by the filtration device 61 s o as to remove tlxe foreign matter- having a size larger than a predetermined radius or foreign matter- in a gel form.

The dope 24 is then cast from the casting die 81 onto the belt 82. In order to regulate the tension of tlie belt 82 to 10^a N/m to 10⁶ N/m, a relative position of the rollers 85 and 86, anόL a rotation speed of at least one of the rolle-xs 85 and 86 are adjusted. Moreover, a relative speed difference between the belt 82 and the rollers 85 and 86 are adjusted so as to be 0.01 m/minL or less. Preferably, speed fluctuation of the belt 82 is 0.5%: or less, and meandering thereof caused in a wi-dth direction is 1.5 mm or less while the belt 82 makes one rotation. In orderr to control the meandering, it i_s preferable to provide a detectorr (not shown) Jfor detecting the positions of botli. sides of the belt 82 and a position controllear (not shown) fTor adjusting th& position of ttie belt 82 according to detection data of the detectorr, and performs feed back control of the position of the belt 82 _ With respect to a portion of the belt 82 located just under the casting die 81, it is preferable that verrtical positional fluctuation caused in association with the rotation of the roller 85 is adjusted so as to be 200 μm or less. Further, it is preferable that the temperature of the casting chamber 63 is adjusted witliin the range of -10⁰C to 57⁰C b>γ the temperature regulator 97 _ Note that the solvent vaporized inside the casting chamber 63 is reused as dope preparing soLvent after being collected by the recovery device 99.

The casting bead is formed between the casting die 81 ancii the belt 82, and the casting membrane 24a is formed on the bel-fc 82. In order to stabilize a. form of the casting bead, it is preferable ttiat an upstream-side area from the Ε>ead is controlled by the decompression chamber- 90 so as to be set to a desired pressure value. Preferably, the upstream-side area from the bead is decompressed within the range of -2500 Pa "to -10 Pa relative to its downstream-side area from the casting bead. Incidentally , it is preferrable that a jacket (not shown) ά_s attached to th«e decompression chamber 90 to maintain the inside temperature a~t a predetermined temperature. Additionally, Lt is preferable to attach a suction unit (not shown) to an edge portion of the casting die 81 and suctions both sides of the bead in order to keep a desired shape of the casting bead. A- preferable rang-e of an air amouαt for aspirating the edge is 1 L/min to 100 L/min.

After the casting membrane 24a lias possessed a. self-supporting property, tb_is casting membzrane 24a is peelecl from the bel"t 82 as the membrane 62 while supported by a peeling roller 109 . The membrane 6 2 containing the solvent is carried along the transfer section 101 while support ed by many rollears , and then fe<3 into the tenter drier 64 . In tlie transfer section 101 , it is possible to give a draw tension to the membrane 62 by increasing a. rotation speed of the downstream rrroller in comparison with that of the upstream roller . In the transfer section 1Ol , dry air of a. desired temperrature is sent near the membrane ^2 , or directly blown to the membrane 62 from tfcie air blower 102 to facilitate a drying process of the membrane 62 . At this tine , it is preferable that the temperature of the dry air is 20⁰C to 250⁰C .

The membrane 62 fed in. to the tenter drJLer 64 is dried while carried in a state that both side edges thereof are held with holding devices such as clip s 64a . At this time , pins may be u sed instead of the clips . The pins may be penetrated through -the membrane 62 to support it . It is preferable that the inside, of the tenter cirier 64 is divicled into temperature zones and drying conditions are properly adj usted in each zone . The membrane* 62 may be stretched in a width direction by using the tenter dr ier 64. It is preferable that the membrane 62 is stretched in "the casting dirreσtion and/or the width direction in the trans ±er section 101 and/or the tenter drier 64 suclx that a size of -the film 62 after the stretching becomes 100. 5% to 300% of the s ize of the same before the stretching ..

After^* the membrane 62 is dried by the t enter drier 64 un til the remaining solvent amount reaches a predetermined value, b oth edges thereof are cut off fc>y the edge slitting device 67 . The cut edges ar-e sent to the crusher 103 by a cutter blower ( not stκ>-wn) . The membrane edges are shredded by the crusher- 103 and become cb_ips . The chip is recycled for preparing the dope , and this enat> les effective u se of the raw material . The slitting process for the membrane edges may be omitted. However , it is preferable to OO

perform "the slitting process between the casting process and the membrane winding process .

Meanwhile, the membr-ane 62 of which both side edges tαave been cut off is sent to the drying chamber 69 and is further dri_ed. Although a temperature of tlie drying chambexr 69 is not especially limited, it is determined, in accordance with heat resistance properties (glass transition point Tg, heat deflection temperature under load, melting point Tm, continuous—use temperature and the like) of the solid electrolyte, and the temperature is preferably Tg or lower. In the drying chamber 69, the memb-rane 62 is carried while being bridged across the rolJ_ers 68, and the solvent gas vaporized therein is absorbed and recovered by the absorbing device 106. The air from which the solvent vapor is removed -Ls sent again into the drying chamber 69 as the dry air. Incidentally, it is preferable that the drying chamber 69 is divided into β. plurality of regions for the purpose of changing the sending aά_r temperature. Meanwhile, in a case that a preliminary drying chamber (not shown) is provided between the edge slitting device 67 and the darying chamber 69 to preliminarily dry the membrane 62, a meiαlDrane temperatures is prevented from rapidly incireasing in the drying chamber 69. Thus , in this case, it is possibie to prevent a shape of the membrrane 62 from changing.

The membrane 62 is cooled in the cooILing chamber 71 until the memt>rane temperature ^"becomes about a room temperature . A moisture control chamber (not shown) may be provided between the drying chamber 69 and the cooling chamber 71. Preferably, air having desirable humidity and temperaturre is applied to the membrane 62 in the moisture control chambex:. By doing so, it is possible to prevent the membrane 62 from curling and to prevent winding defect from occururing at the time of winding.

In the solution casting method, vari-ous steps such as the drying step , the edge s litting step and so forth are performed over tlαe membrane 62 aft er it is peeled from the support and iantil it is wound up as the final product . During or between each .step , the membrane 62 is mainly supported or transported by the rollers . Among these rollers , some are drive rollers and others are non-dr-ive rollers . The non-drive roLlers are used for determining a membrane passage , and at the same time for improving transport stability of the membrane 62.

^"While the membrane 62 is carried , the charged voXtage thereof is kept in the predetermined range. The charged voXtage is preferably at - 3 kV to +3 kV after the neutralization . Further , it is preferable that the knurling is formed on the membra:_ne 62 by the knurling roller pair 73 . Incidentally, it is preferable that asperity height of: the knurling portJLon is 1 μm to 2OO μm .

The membrane 62 is wound up by the winding roller? 107 contained in the winding chamber 76 . Αt this time, i_t is prefer-able to wind the membrane 62 in a state that a desi-βrable tension is given by the press roller 108 . Prref erably , the tension is gradually changed from the start of windi-ng to the end theareof . Owing to this , the membrane 62 is preven ted from being wound excess ively tightly. It is preferable that a width off the membrane 62 to be wound up is not less than. 100 mm. The present invent ion is applicable to a case in that a thin membrane of which thickness is 5 μm or more and 100 μm or Less is produced .

A method of producing a solid electrolyte multiILayer membrane having the catalyst layer and the solid electrolyte layer by co - casting two or more sorts of dope s is explained . The co-casting method may t>e a simultaneous co-casting method, or a sequential co-casting method . When the simultaneous co-casting is performed, a feed block may be attachecl to the casting die , or a mult i -manifold type casting die may be used.

The method of producing the solid electrolyte multilayer membrane according to the simultaneous co-casting method is explained with referring to Fig. 3. Fig. 3 shows a simulta-neous co-csasting device 111. In Fig. 3, the components identical to those shown in Fig. 2 are assigned with same numerals. The simultaneous co-casting device 111 forms a casting membrane 112 having a three-layer structure, and the obtained solid electrolyte multilayer membrane 62 is composed of three layers : a first surface layer- 112a, a second sxirface layer 112b and an inner layer 112c. The. first surface layer 112a is in contact with the ^"belt 82. The second surface layer 1 H2b is exposed to the air. The inner layer 112c is interposed between the first and the second surface layers 112a and 112b and not exposed outside.

A first dope 114 for forming the first surface layex: 112a is cast such that it contacts with the belt 82. A Seconal dope

115 forms the inner ,J_ayer 112σ, and a third dope 116 forms the second surface layer 112b. The first dope 114 and the thircl dope

116 include catalyst , which is described, later. The first, second and third dopes 114, 115 and 116 sent through dope feeding passages Ll, L2 and L3, respectively are fed to a feed block 119 attached to a casting die 89. The dopes are joJLned in the feed block 119 and simultaneously cast from the lip edge. In other worcϋs, in the feed block 119, -three dope passages are formed. The dope passage placed in the middle of the ttαree dope passages is for the second dope 115. The dope passage placed upstream from the midclle passage in the running direction of the belt 82 is for the first dope 114. The dope passage placed downstream form the mid.cl.le passage in the running direction* of the belt 82 is for the thixrd dope 116.

When the first and third dopes 114= and 116 forming the^ first and. second surface layers 112a and 112b are each made to ϊiave a viscosity lower than that of the second dope 115 forming tln& inner layer 112c, the produced membrane hardly expresses abnormal oo

characteristics sucϊi as melt fracture . When the dopes are cast af: ter adjusting the -viscosity of each, dope in this way, the second dope 115 may be sururounded by the first dope 114 and the third dope 116 in the bead , which is formed from the casting die 89 to trxe belt 82 . There are some cases that such bead is purposely formed . The first elope 114 and the third dope 116 may contain time poor solvent . In this case , poozr solvent ratio of the first dope 114 and the third dope 116 may preferably be higher than that of: the second dope 3.15 . At this time , it is preferable that the first dope 114 is cast such that the first surface layer 112a, wriich is in contact with the belt 8 2 , will have a thJLckness of 5 μm or more in a wet state . As the poor solvent , those used for ttre poor solvent contact device 65 ( see Fig . 2 ) can be used .

In this way, tlie first , second and third dopes 11 3 , 115 and 116 share the feed ^"block 119 to be simultaneously co— cast from trie casting die 89 having one castJLng opening . Instead of the feed block 119 and the casting die 89 , it is also possible to use a casting die having three casting openings . When such casting die is used, the fiϋrst , second and third dopes 114 , 13.5 and 116 aπre cast from different openings . Three openings of this kind of casting die are arranged along the running direction of the belt 82 .

Thickness of each layer ll_ 2a , 112b or 112 c is not particularly restricted, however the first , second and third dopes 114 , 115 and 116 are preferably cast such that the first and second surface ILayers 112a and 11.2b , that is , catalyst layers will each have the thickness of 10 μm to 500 μm.

Each dope 114 , 115 or 116 may tiave the viscosity different farom each other. However , it is preferable that the solid electrolyte in the .first dope 114 and the third dope 1 16 is same as or compatible with that in the second dope 115.

Each dope 114 , 115 or 1.16 may contain the additives different firoiπ each other. Specifically, tlxe types or the concentration of the additives sixch as the above -described antioxidants , fibers , fine particles , water absorbing agents , plasticizers , compatibilizing agents and the liJse may be varied -f rom dope to dope . For example , the antioxidants and fine particX.es (matting agents ) may be ad<3.ed more to the ffirst and third dopes 114 and 11 6 forming the surface layers as compared to the second dope 115 forming the inner layer . Alternatively, the antioxidants and fine particles may be added only to the first and thijtrd dopes 114 and 116 . MeanwhilLe , the water abs orbing agents , pLasticizers , compatibilizing agents may be addad more to the second dope 115 forming the inner layer as comparecϋ to the first and third dopes 11_ 4 and 116 forming the surface layers . Alternatively , the water absorbing agents , plasticizers , compatibilizing agents may be added only to the second dope 115. There is another configuration ttiat the antioxidants having a low volatility are contained in tlxe surface layers 112a and 112b , while the plasticizers having an excellent plasticity and the water absorbing agents having a high water-absorbing property are contained in the inner layer 112c . There is further another con-figuration that peeling agents aire added only to the first dope 114 forming the first surface layer 112a being in contact with "the belt 82 . Thus each layer can independently Ihave desirable f unctions by adjusting the types O2z concentration of the additives . . Moreover , the elopes of the pxresent invention, are capable o=£ forming different sort of function layers ( f or example , catalyst layer , antioxidant layer , antistatic layer , lubricating layei: and the like) siπuαltaneously. In order to give rubricating property to the produced membrane , fine parotides are preferably contained in. the surface layers . Note that at least one odE the surface layers 112a and I X 2b should contain the fine parrticles so that the produced membrane comes to have lubricity . Apparent specific gravity of the fine particle is preferably 70 g/liter or- more , more preferably 90 g/L ±ter to 200 g/li-ter , and further preferably 100 g/liter to 200 g/liter . The produced dispersion li<juid can have higher concentration of the fine particles as the apparent specific gravity of the fine pairticle is larger . When silicon dioxide is used as the fine particles , average dά-ameter of an initial particle is preferably 20 nm or less and the apparent specific gravity is preferably 7O g/liter or more . Such silicon dioxide fine part icles can be obtained by, for example , burning a. mixture of vapoarized silicon tetrachloride and hydrogen in the air at a temperature of 1000⁰C to 1200⁰C . Beside the silicon dioxide fine particles obtained t>y the above method , AEROSIL ® 200V or AEROSIL ® R972V (manuf actured by NIPPON AEROSIL CO . , LTD . ) may be used.

The method of producing the solid electrolyte multilayer membrane according to the sequential co-castin-g method is explained with referring to Fig . 4 . Fig . 4 shows a sequential co-casting device 121 . The sequential co-casting device 121 is provided with three casting dies 122 , 123 and 124. "These casting dies 122 , 123 anca. 124 are sequentially disposed along the belt 82. The casting die 122 casts tϊie first dope 114 , the casting die 123 casts the second dope 115 and the casting άlie 124 casts the third dope 116 .

When the first , second and third dopes 114 , 1 15 and 116 of the same composoLtion are sequentially co-cast , the membrane production speed can be improved as compared to tha.t in a single layer casting. I n this case , the positions of the s econd and the third casting dies 123 and 124 aαre determined according to the drying speed and the like of the preceding layer . For example , it is preferable t o dispose the second casting die 123 at a position ^■where a ratio of the distance between the most upstream casting die 122 and the second casting die 123 to the distance between fclie most upstream casting die 122 and the position at which the casting membrane is peeled is in. a range of 30% to 60% .

Besides the above methods , following method is also available as an example of the co -casting method. A first dope is cast from a fi_rst casing die onto a support to form a membrane , and the membrane is peeled off . With transporting the peeled membrane while supporting it by .rollers , a second dope is cast from a second casting die onto tlxe peeled surface* of the peeled membrane to form a double- layer membrane .

Regardless of the single layer casting method or the co-casting method , there are various methods for casting the dope . For example , a method to uniformly extrude the dope from the pressurizing die , a doctor blade method, a revers e roll coating method and the like . In the doctox¹ blade method, tlie dope is cast on the support and smoothed by -the blade so as to adjust the membrane thickness . In the reverse roll coating me-thod, a casting amount of the dope is adjusted by smoothing the surface of the <ϋope by using rollers rotating reversely to one a.nother . Above all , the method using the pressurizing die is preferable . As the pressurizing die , there are a coat -hanger type die , T- type die and so forth . Any type of the pressurizing die is prref erably used.

Instead off the above -descriioed method for f ojrming the solid electrolyte int o a membrane , it is possible to infiltrate the solid electrolyte into micropores of a so-called porous substrate in order to produce different type of the solid electrolyte membrane . As such method of producing the solid electrolyte membrane , there are a method in which a sol-gel reaction liquid containing the solid electrolyte is applied to the porous substrate so that the sol- gel reaction liquid is infiltrated into the micropores thereof , a method in which such porous substrate is dipped in the sol-gel reaction liquid containing the solid electrolyte to thereby fill trie micropores with the solid electrolyte, and the like . Preferred examples of the porous substrate are porous polypropylene , porous polytetrafluo-coethylene , porous cross -linked heat-resistant polyethylene , porous polyimide, and the like. Additionally, it is also possible to process the solid electrolyte into a fiber form and fill spaces therein with other polymer compounds , and forms this fiber into a membrane to produce the so lid electrolyte membrane . In this case , for example , those used as the additives in the present invention may be used as the polymer compounds to fill the spaces .

The solicϊ electrolyte membrane of the present invention is appropriately used for the fuel cell , especially as a proton conducting meπibrane for a direct methanol fuel cell . Besides that , the solid electrolyte membrane of the present invention is used as a solid electrolyte membrane interposed b etween the two electrodes of the fuel cell . Moreover, the solid electrolyte membrane of tine present invention is used as an electrolyte for various cells ( redox flow cell , lithium cell , and the like) , a display element , an electrochemical censor, a signal transfer medium, a condenser, an electrocϋalysis , an electrolyte membrane for electrolysis , a gel actuator , a salt electrolyte membrane , a proton -exchange resin , and the like .

( Fuel Cell )

Hereinaffter, an example of using the solid electrolyte membrane in a Membrane Electrode Assembly (hereinafter, MEA) and an example of using this MEA in a fuel cell are explained. Note, however, that forms of the MEA and the fuel cell, described here are just an example and the present invention is not limited to them. In Fig. 5, a MEA 131 has the membrane 62 and an anode 132 and a cathode 133 opposing each other. The membrane 62 is interposed between the anode 132 and the cathode 133.

The anocie 132 has a porous conductive sheet 132a and a catalyst .Layer 132b contacting the membrane 62, whereas the cathode 133 has a porous conductive sheet 133a and. a catalyst layerr 133b contacting the membrane 62.. As the porous conductive sheets 132a and 133a, there are a carbon sheet and the l^ke . The catalyst layers 132TD and 133b are made of a dispersed siibstance in which catalyst metal-supporting cartoon particles are dispersed in the proton conducting material. -As the catalyst metal, there are platinum and the like. As the carbon particlas, there are, fo.tr example, ketjenblack, acetylene black, carbon nanotube (CNT) ancϋ the like. As the proton conducting material- , there are, fox: example, Mafion (registered trademark) and ttie like.

As a method of producing the MEA 131, tltie following fouαr methods axre preferable.

(1) Proton conducting material coating method: A catalyst paste (ink ) that has an active metal-supporting carbon, a protoxi conducting material and a solvent is directly applied onto botli surfaces of the membrane 62, ancL the porous condxictive sheets 132.a and 133a aire (thermally) adhered under pressure thereto to form a five-layered MEA.

(2) Porous conductive sheet coating method: A liquid containing the materials of tlie catalyst layers 132b and 133b , that is, for example the catalyst paste is applzLed onto the poroαs conductive sheets 132a and 133a to form the catalyst layers 132TD and 133b thereon, and the membrane ,62 is adhered thereto undei pressure to form a five-layered MEA.

(3) Decal method: The catalyst paste is applied onto polytetrafrluoroethylene (PTFE) to form the catalyst layers 132TD and 133b tliereon, and the catalyst layers 132b and 133b alone arre transferred to the membrane 62 to form a three-layer structure.. The porous conductive sheets 132a and 133a arre adhered thereto under pressure to form a five-layered MEA.

(4) Catalyst post-attacrnnent method: Ink prepared by mixiixg a carbon material not supporting platinmm and the proton conducting material is applied onto the membi-cane 62, the poroTis conductive sheet 132a and 133a or the PTFE to form a membrane. After that, the membrane is impregnated with liquid containing platinum ions, and platinum particles are precipitated in tϊie membrane through reduction to thereby form the catalyst layeαrs 132b and 133b. After the catalyst layers 132b and 133b are formed, the MEA 131 is formed according to one of the above-described methods (1) to (3).

Note that the method of producing the MEA is not limited to the aloove-described metϊiods, but various well-known methods are applicable. Besides tϊie methods (1) to (4), there is, for example, the following method. A coating liquid containing the materials of the catalyst layers 132b and 133b is previously prepared . The coating liqvxid is applied onto supports and driad. The supports having the catalyst layers 152b and 133b forined thereon are adhered so as to contact with both surfaces of the membrane 62 under pressure. After peeling the supports therefrrom, the memb-rane 62 having the catalyst layers 132b and 133b on both surfaces is interposed by the porous conductive sheets 132a and 133a. Tune porous conductive sheets 132a and L 33a and the catalyst layers IL32b and 133b are tightly adhered to form a MEA 131.

In. Fig. 6, a fuel cell 141 has the MEA 131, a pair of separators 142, 143 holding the MEA 131 therebetween, current collectors 146 made of a stainless net attached to the separators 142, 143, and gaskets 147. The fuel cell 141 is illustrated in exploded! fashion in Fig. 6 for the sake of convenience of explanation, however, eaαϊi element of the fuel cell 141 are adhered to each other to be used as a fuel call. The anode-si_de separator 142 has an anode-side opening 151 formed through ±_t; and the cathode-side separator 143 has a cathode-side opening 1_52 formed through it. Vapox" fuel such as h/ydrogen or alcohiol (methanol and the like) or liquid fuel such as aqueous alcoliol solution is fed to the cell via the anode— side opening 151; .and an oxidizing gas such as oxygen gas or air is fed thereto via "the cathode — side opening 152.

For the anode 132 and the cathode 133, for example _M a catalyst that supports active metal particles of platinum or "the like on a carbon material may be used. The particle size of "the active metal particles that are generally "used in the art is from 2 nm to 10 nm. Active metal particles having a smaller particle size may have a larger surface area per the unit weight thereof, and are -therefore more advantageous since t-tieir activity is hicylier . If too small, however, the particles are difficult to disperse with no aggregation, and it is said that -the lowermost limit of the particle size will be 2 nm or so.

In hydro gen -oxygen fuel cells, the active polarization of cathode . namely air electrode is higher tha.n that of anode, namely hydrogen electrode. Th.±s is because the cathode reaction, namely oxygen reduction is slow as compared with trie anode reaction. For enhancing the oxygen electrode activity, usable are vari-ous platinum-based binary alloys such as Pt-Crr, Pt-Ni, Pt-Co, Pt-Cu, Pt-Fe. In a direct methanol fuel cell in -which aqueous methanol is used for the anode fuel, usable are platinum-based binary alloys such as Pt-Ru, Pt-Fe, Pt-Ni, Pt-Co, Pt-Mo, and platinum- based ternary alloys such as Pt-Ru-Mo, Pt-Ru-W, Pt-Rm-Co, Pt-Ru-Fe, Pt-Ru-Ni, Pt-Ru-Cu, Pt-Ru-Sn, Pt-Ru-Au in orderr to inhibit the catalyst Poisoning with CO that is formed durring methanol oxidation. For the carbon matenrial that supports the active metal thereon, preferred are acetylene black, Vulcan XG- 72 , ketjent>lack, carbon nanohorn (CNH) and CNT.

The function of ttie catalyst layers H32b, 133b includes (1) transporting fuel to active metal , ( 2 ) providing the reaction site for oxidation of fuel (anode) or for reduction of fuel (cathode) , (3) transmitting the electrons releasetϋ in the redox reaction to the current collector 146, and (4) transporting the protons generated in the reaction to the solid, electrolyte, namely the membrane 62. For (1) , the catalyst layers 132b, 133b must be porous so that liquid and vapor fuel may penetrate into the depth thereof. The catalyst supporting active metal particles on a carbon material works for (2); and the carbon material woxrks for (3) . For attaining tlxe function of (4) , the catalyst layers 132b, 133b contain a proton conducting material added thereto. The proton conducting material to be in the catalyst layers 13213, 133b is not specifically defined as long as it is a solid tha"t has a proton-donating group. The proton conducting material may pref exably be acid res ±due-having polymer compounds that axe used for the membrane 62 such as perf luoro sulfonic acids, as typified by Nafion (registered, trademark); po.Lγ(meth)acrylate having a phosphoric acid group in side chains; sulfonated heat-resistant aromatic polymers such as sulfonated polyether etherketoxies and sulfonated polybenz imidazoles. When the solid electrolyte for the membrane 62 is used for the catalyst layers 132b, 133b, the membr-ane 62 and the catalyst layers 132b, 133b are formed of a material of the same* type. As a resτilt, the electroctiemical adhesiveness between the solid electrolyte and catalyst layer becomes high. Accordingly, this is advantageous in terms of the ionics conductivity. The amount of the active metal to t>e used hereiLn is preferably from 0.03 mg/cm² to 10 mg/cm² in viewy of the cell output and economic efficiency. The amount of the carbon material that supports the active metal is preferably from 1 to 10 times the weight off the active metal- The amount of the proton conducting material is preferably from O .1 to 0.7 times the weight of tine active metal- supporting carbon .

The anode 132 and the cathode 133 act as current collectors (power collectors) and also act to prevent water from staying 4

thezrein to worsen vapor permeation. Zn general, carbon paper or cartoon cloth may be used. If desired, the carbon paper or the cart>on cloth may be processed with PTFE so as to be rapellent to water.

The MEA has a value of area .resistance prefeαrably at 3 Ω cm² or less, moire preferably at 2. Q cm² or less, and most preferably at 0.5 Ω cm² or less according to alternating -current (AC) impedance mettiod in a state that the MEA is incorrporated in a cell and the cell i.s filled with fuel - The area resistance value is calculated by a product of the measured resistance value and a sample area..

Fuel for fuel cells is described. For anode fuiel, usable are hydrogen, alcotiols (methanol, is opropanol , ethylene glycol and the like), ethers (dimethyl ether, dimetho>xyme thane, trimethoxyme thane and the like), formic acid, boronhydride complexes, ascorbic acid, and so fortti. For cathode fuel, usable are oxygen (including oxygen in air) „ hydrogen peroxide, and so forth.

In direct metlianol fuel cells, the anode fuel may be aqueous methanol having a methanol concentration of 3 wt.% to 64 wt.%. As in the anode reaction formula (CH₃OH + H₂O → CO₂ + 6H⁺ + 6e^~), 1 iriol of methanol requires 1 mol of water, and tlxe methanol concentration at this time corresponds to 64 wt.%. A higher metlnanol concentration in fuel is more effective for reducing the weight and the voIL-uine of the cell including a fuel tank of the same energy capacity. However, if ttxe methanol concentration is too high, much methanol may penetrate through the solid electrolyte to reach the cathode on -which it reacts with oxygen to l_ower the voltage . This is so-called the crossover phenomenon . When the methanol concentration is too high, the crossover phenomenon is remaarkable and the cell output tends to> lower. In vievr of this, the optimum concentration of methanol shall be determined depending on the methanol perviousness through the solid electrolyte used. The cathode reaction formula in direct methanol fuel cells is (3/2) O₂ + 6H⁺ + 6e^~ → H₂O , and oxygen (generally, oxygen in air) is use<3 for the fuel in the cells.

For supplying the anode fuel, and the cathode* fuel to the respective catalyst layers 132b and 133b, there are two applicable methods: (1) a method of forcedly sending the fuel toy the use of an auxiliary device such as pump (active method) , ancϋ (2) a method not using such an auxiliary device, in which liquid fuel is supplied through capillarity or by spontaneously dropping it, and vapor fuel is supplied by exposing the catalyst layer to air (passive method) . It is also possi-^"ble to combine th.e methods (1) and (2) . In the method (1) , high- concent rat ion methanol is usable as fuel, and air supply enables h.άgh output from the cells by extracting water formed in the cathode area. These are the advantages of the method (1). However, this me-thod has the disadvantage in that the necessary fuel supply unit will make it difficult to downsize the cells. On the other hand, -the advantage of the method (2) is capability of downsizing the cells, but the disadvantage thereof is that the fuel supply rate is readily limited and high output from the cells is often clifficult.

Unit cell voltage of fuel cells is generally at most 1 V. Ttierefore, the unit cells are stacked up in series depending on tine necessary voltage for load. For cell stacking, employable methods are a method of "plane sta.cking" that arranges the unit cells on a plane, and a method of " bipolar stacking " that stacks Uj? the unit cells via a separator "with a fuel pathway formed on both sides thereof. In the plane stacking, the cathode (air electrode) is on the surface of the stacked structure and therefore it readily takes air ther^*einto. In addition, since the s izacked structure may be thinned, it is more ffavorable for small-sized fuel cells. Besides the above -de scribed methods, MEMS technology may be employed „ in which a silicon wafer is processed to form a micropatterrn and fuel cel-Ls are stacked thereon.

Fuel cells may have many applications fox: automobiles, electric and electronic appliances for household use, mobile devices, portable devices, and ttαe like. In part icular, direct methanol fuel cells can be downsized, the weight thereof can be reduced and do no"t require charging • Having such many advantages , they are expected to be used for various energy souxces for mobile appliances and portable appliances . For example , mobile appliances in wlxich fuel cells are favorably used include mobile phones, mobile notebook-size personal computers, electronic still cameras, PDA, video cameras , mobile game machines , mobile servers, wearable personal computers , mobile displays and the like . Portable appliances in which fuel cells are favorably used include portable generators, outdoor lighting devices, pocket lamps, electrically-powered (or assisted) bicyσlets and the like. In addition, fu^l cells are also favorable for power sources for robots for industrial and household use and for other toys. Moreover, they are further usable as power sources for charging secondary batteiries that are mounted on these appliances.

[Example 1 ]

Hereinafter, examples of the present invention are explained. In tlie following description. Experiment 1 of Example 1 and Experiment 1 of Example 2 are explained in detail. With respect to Experiments 2 to 7 of Example 1 and Experiments 2 to 6 of Example 2, conditions different from each Experiment 1 of Examples 1 and 2 are only explained. Note that Experiments 2 to

6 of Example 1 and. Experiments 2 to 5 of Example 2 azre the examples of the embodiments of the present invention. Experiments 1 and

7 of Example 1, and Experiments 1 and 6 of Example 2 are the comparative experiments of the embodiments o:f the present invention.

[Experimen.t 1 ]

{Production of First, Second and Third dopes 114, 115 and 116}

5 A material A was condensed t>y the flash device 26 and dried. Solid contents containing the dried material A was dissolved in the solvent according to the following compositioia., and the dopes having the solid contents of 30 wt.% were produceci. The solvent was perfluorohexane . Note that catalyst fine particles did not O dissolve in, bu"t dispersed in the solvent. Adclitive rate of dichloromethane to the dope was varied in each Experiment 1 to 7 as shown in Tafcϊle 1. The dichloromethane was the poor solvent of the dried material A. The di_chloromethane was added to the first dope 114 and the third dope 116, but was not added to the S second dope 115. Each Experiment 1 to 7 was perforiried with varying the additive rate of dichloromethane that was ttαe poor solvent of the dried material A. The first to third dopes 114 to 116 in Experiments 1 to 7 all had 3O wt.% of the solid contents concentration. Note that the material A was 20% Nafion O (registered trademark) Dispersion Solution DE202O (manufactured by US Dupont) . First dope 114:

Dried material A 80 pts.wt

Pt catalyst ffine particles TEC10E50E 20 pts.wt 5 (manufactured by Tanaka Kikinzoku Kogyo K.K.) Second dope 115 s

Dried material A Third dope 116:

Dried material A 80 pts.wt 0 Pt-Ru catalyst fine particles TEC61E54 20 pts.wt (manufactured by Tanaka KikdLnzoku Kogyo K. K.)

{Production of Solid Electrolyte Multilayer Membrane 62} The solid electrolyte multilayer- membrane having three-layer structure was produced by the simuiltaneous co-casting device 11 1 according to the following method - After the drying , the solid electrolyte multilayer membrane 62 was made to have the total thickness of 140 μm in which the first surface layer , the second surface layer and the inner layer wesre made to have the thickness of 20 μm, 20 μm and 100 μm, respectively. The casting width was 380 mm, and the flow amount of eac-h dope was adjusted during tfcne co-casting . The casting die 89 "was provided with a jacket ( not shown) in which, a heat transfer medium was supplied . A temperature of the heat transfer medium wa.s regulated at 4 O ⁰C so as to maintain the temperature of each first to third dope 114 to 116 at 40⁰C .

The temperatures of the casting die 89 _r the feed block 2L 19 , and the dope feeding passages Ll to L3 for the first to third dopes 114 to 11 6 were all maintained at 40 °C . The casting die 89 was the coat - lianger type and had the width of 0 . -4 m. The heat bolts provided to the casting die 89 for adjusting the membrrane thickness were disposed at the pitch of 20 mm. The casting die 89 had the automatic thickness adjusting mechanism for adjust ing the slit clearance thereof . The profile of the heat bolt could be set corresponding to the flow amounts of the first to tlαird dopes 114 to 116 by the accuracy gear pump , on the basis of the preset program. Thus the fr eed back control could be made by the control program on the bas is of the profile of an infrared ray thickness meter (not shown ) disposed in the membrane producing apparatus 33 . The slit clearance of the li;p edge was adjus ted such that , with exception of both side edge porrtions ( specif iσaully, 20 mm eacti in the widthwise direction of the produced membraixe ) , the difference of the membrane thickness between any two po-Lnts which were 50 mm apart from each other might ^"be at most 1 μm , and the largest difference between the minimal values of the membirane __Q

thickness in the widtlrwise direction migh.t be at most 3 μjn/m. Moreover, the slit clearance of the lip eclge was adjusted such that the average thickness accuracy of eacln surface layer might be at most ±2%, that O-f the inner layer might be at most → 1%, and the average membrane thickness might be at most ±1.5%.

In order to prevent the dope from partially drying- and solidifying at the lip edge of the casting die 89, a liquid used as the solvent of the elope was supplied to three-phase contact lines formed by both erxd portions of the casting bead, botrx end portioixs of the lip edge and ambient air at a rate of 0.5 ml/'min. The pulse rate of a pump for supplying the liquid was at most 5%.

Tlie material of the belt 82 was SUS316 having enough corrosi-on resistance and strength. The beLt 82 was polished such that the surface roughness might be at most 0 - 05 μm. The thickness of the ^"belt 82 was 1.5 mm and the thickness unevenness thereof was at most 0.5%. The fc>elt 82 was moved by rotating the rollers 85 and 86, and the relative speed between the rollers 85, 86 and the bell; 82 was at most 0.01 m/min. The speed fluctuation of the belt 82 was at most 0.5% . The positions of ^"both sides of the belt 82 were detected so as to control the position of the belt 82. The position of the belt 82 was controlled such that the meandering thereof: in the width ddLxection might be at most 1.5 mm while the belt 82 makes one rotation. The distance fluctuation between the lip edg-e and the belt 82 was regulated to fc>e at most 200 μm - In the casting chamber 63 , a wind pressure fluctuation controller (not sh.own) for control,ling the wind pressure fluctuation inside of the casting chamber- 63 was provided.

TVhe first, second and third dopes 114 „ 115 and 116 were cast so as to form the casting membrane 112. Tlie dry air of 50⁰C to 70⁰C was applied to the casting membrane 112 by the air blowers 91, 92 and 93 so as to dry the casting membrane 112 until the solvent content thereof reached 30 wt . % with respect to the solid contents OO

of the material A. , namely the solid electrolyte . Avfter the casting membrane 1 12 had possessed a self -supporting property , tine casting membrane 112 was peeled from the belt 82 as the membrane 62 . The membrane 62 was fed into the tenter drieir 64 and transported therein in a state that tooth side edges theareof were held with the clips 64a . In the tenter drier 64 , the membrane 62 was dried until, the solvent content thereof reachecϋ 15 wt . % with respect to the solid contents by the dry air of 14O °C . The membrane 62 was ttien released from "the clips 64a at an exit of tine tenter drier 6 4 , and both edges of the membrane 62 were cut off by the edge slitting device 67 disposed downstream, from the tenter drier 64. The membrane 62 of which both side edges had been cut off was s ent to the drying chamber 69 and was further clαried at the temperature of 160⁰C to 180⁰C while transported by tine rollers 68 . In this way, the so lid electrolyte membrane 62 having a solvent content rate of le ss than 1% was obt ained. A tliickness of the obtained membrane 62 was 80 μm.

The obtained membrane 62 was evaluated in each of the following items . Evaluation results are shown in Tables 1 . Note tliat the number of the evaluation items in Table 1 correspond to tlie number assigned to each of the following items . 1 . Thickness

Thickness of: the membrane 62 was continuously measured at a. speed of 600 mm/min . by the use of an electronic micrometer manufactured by Anritsu Electric Co . , Ltd . Data obtained by the measurement was recorded on a chart on a scale of 1/20 , at a chart speed of 30 mm/min. . After obtaining measurements of data curve bry a ruler, an average thickness value of the membrane 62 and tliickness unevenness relative to the average thickness value were obtained based on the obtained measurements . In Table 1 , ( a) represents the average thickness value (unit : μm) and (b ) .represents the thickness unevenness ( unit : μm) relative to ( a) . 2. Ionic Conductivity Coefficient

On the obtained solid electrolyte multilayer- membrane 62, ten measurement points each of whi-ch is Im apart fream one another were selected along a longitudina.,1 direction of the membrane 62. These ten measurement points were cut out into cizrcular sample having a diameterr of 13 mm. Each sample was interposed by a pair of stainless plates, and the ionic conductivity coefficient of the sample was measured in accordance with the AC impedance method by the use of a Multichannel Battery Test System L 470 and 1255B manufactured by Solartron Co., Ltd. The meatsurement was performed under the condition of a temperature at 25⁰C and a relative humidity of 100% . The ionic conductivity i_s represented by a value of the AC impedance (unit: S/cm) as shown in Table 1.

3. Output Density of Fuel Cell 141

The fuel cell 141 using ttie membrane 62 was formed, and output thereof was measured. According to the following methods, the fuel cell 141 was formed, and the output density thereof was measured.

(1) Formation of MEA 131

A carbon paper having a thickness of 350 μiα was attached to both surfaces of the solid electrolyte memt>rane 62, and thermally adhered for 2 minutes a.t a temperature of 8O⁰C under a pressure of 3 MPa. In this way, a MEA 131 was formed.

(2)Output Density of Fuel Cell 141

The MEA fabricated in (1) was set in a fuel cell as shown in Fig. 6, and an aqueous 15 wt.% methanol solution was fed into the cell via tine anode-side opening 151. At ttiis time, the cathode-side opening 152 was kept open to air. The anode 132 and the cathode 133 were connected to the Multichannel. Battery Test System (Solartron 1470) , and the output density (un.it: W/cm²) was measured. Table -L

Evaluation Item

Example L 1 (μm) 2 3

(a) Cb) (xlO^"2 S/σm) (mW/cm²)

Experiment 1 33.7 ± 2.0 7.9 228

Experiment 2 33.7 2.0 8.1 331

Experiment 3 33.8 2.0 8.3 338

Experiment 4 33.9 2.0 8.4 375

Experiment 5 34.0 ± 2.0 8.8 401

Experiment 6 34.2 ±: 2.0 8.3 441

Experiment 7 34.2 2.0 8.0 329

According to the result s of Example 1 , the value of a simple cell according to the AC impedance method and ttie output density of the fuel cell as the unit cell are both higher in Experiments 2 to 6 as compared to Experiment 1 which is a prior art and Experiment 7 which is the comparative example - In Experiments 2 to 6 , an appropriate amount of the poor solvent of the solid electrolyte was added to the first and the third dopes 114 and 116 for the catalyst layer 132b and 133b . Accordingly, it will be understood that the solid electrolyte multiILayer membrane of the present invention is suitably used for tϊxe fuel cell .

[ Example 2 ]

Solid contents containing a dried material B was dissolved in the solvent according to the following coiϊijposition, and the first, seconcl and third dopes 114, 115 and 116 having the solid contents of: 30 wt.% werre produced. T"lie solvent was N-methylpyrrolidone . Note tihat catalyst fine particles did not dissolve in, ^"but dispersed in the solvent. Note that the material B was sulfonated polyacrylonitrile styrene. First dope 114:

Dried material B 10 pts.wt Pt catalyst fine particles TEC10E50E 20 pts.wt

(manufactured by Tanalca Kikinzoku Kogyo K. K.) Second dope 115:

Dried material B Third dope 116:

Dried material B 10 pts.wt

Pt-Ru catalyst fine panrticles TEC61E54 20 pts.wt

(manufactured by Tanaka Kikinzoku Kogyo K. K. )

{Production of Solid Electrolyte Multilayer Membrane 62} Instead of the first to third dopes IH 4 to 116 of Example 1, the above-noted first to third dopes 114 to 116 were used. The temperatmres of the dry air from the air blLowers 91, 92 ancL 93 were regulated to be 10O⁰C to 120⁰C . A thickness of each membrrane produced in this Example 2 was 35 μm. In Experiment 2, water was sprayed on "to the just peeled membrane 62 f e<3. out of the casting chamber 63. The spraying was performed by ttie use of an atomi_zer manufacture ed by H. IKEUCHI & CO., LTD. Note that water was the poor solvent of the material B. In Experiment 3, the spraying was performed at the exit of the tenter driezr 64. In Experiment 4, water was added to tine just peeled membrane 62 by vapor humidifica-tion. In Experiment 5, water was added to the membrrane 62 at the exit of the tenteir drier 64 by the vapor humidif icatiLon . In Experiment 6 , water was added to the dry membrane before wound up by the -vapor humidif iσ at ion. In .Experiment 1, water was not added at all. Other conditions were same as Example 1. Evaluation results of the obtained membrane 62 are shown in Table 2. Table 2

Evaluation Item

Example 2 1 (μm) 2 3

(a) (b) (XlO^'2 S/cm) (mW/cm² )

Experiment 1 33.7 ±2.0 7 .9 228

Experiment 2 33.7 ±2.0 8 .1 331

Experiment 3 33.8 ±2.0 8 .3 348

Experiment 4 33.9 ±2.0 8 .4 385

Experiment 5 34.O ±2.0 8 .8 351

Exper imen t 6 34.2 ±2.0 8 .0 229

According to the ztresults of Example 2 , the value of a simple cell according to the AC impedance methocL and the output density of the fuel cell as the unit cell are botϊi higher in Experiments 2 to 5 as compared to Experiment 1 wh-Lch is a prior arrt and Experiment 6 which is the comparative essample. In Experiments 2 to 5, an appropriate amount of the poor solvent of the solid electrolyte was applied to the surfaces of^" the catalyst laye or 132b, 133b beffore fully dried. Accordingly, it will be understood, that the solid electrolyte multilayer meiαlDrane of the present invention is suitably used for the fuel. cell.

Fa-Tom the results of the above-mentioned examples, i-fc will be understood that it is possible to continuously produce the solid el_ectrolyte multilayer membrane having excellent planarity and reduced defects according to the present invention. I^~t will be also understood that "the obtained solid, electrolyte multilayer membrane can be appropriately used as the solid electrolyte layer for the fuel cell.

Industrial Applicability

Tlie solid electrolyte multilayer membrane, the method and the apparatus of producing the same, the membrane electrode assembly and the fuel cell using the solid electrolyte multilayer membrane of the present invention are applicable t o the power sources for various mobile appliances and varioxαs portable appliances .

Claims

CLArMS

1. A method of producing a. solid electrol_yte multilayer membrane, comprising the steps of: casting a first dope and a second dope onto a .running support so as to form a casting membrane having a first layer of said first dope and a second layer of saicϊ second dope, said first dope containing an organic solvent and a solid electrolyte being a solid electrolyte layer of a fuel cell, sai_d second dope containing said solid electrolyte, said organic solvent and a catalyst promoting a redox reaction of electrodes in said fuel cell; peeling said casting membrane as a wet membrane from said support ; performing a first drying of said wet membxane in a state that both side edges thereof are held by holding devices; and performing a second drying of said wet membrane supported by rollers to form said solid electrolyte multiILayer membrane, said second drying step being performed after sa.±d first drying step.

2. A methocϋ described in ciaim 1, wherein said first dope is cast from a fά_rst casting die eand said second dope is cast from a second casting die disposed at a downstream of saixd first casting die .

3. A method described in claim 1 , wherein said wet membrane is brought into contact with a compound that is a poor solvent of said solid elLeσtrolyte .

4. A method described in claim 1, wherein said catalyst includes at least one of Au, Ir, Pt, Rh, Ru, W, Ta, Nb, Ti Pd, Bi , Ni , Co , Fe and Hf .

5 . A method described in claim 1 , wrαerein a thicknes s of a layer- formed from said first dope in said solid electrolyte multilayer membrane is 20 μm to 800 μm, saά_d layer being derrived from said first layer of said casting membrane .

6 . A method described in claim 1 , wtierein a thickness s of a layer- formed from said second dope in said solid electrolyte multilayer membrane is 10 μm to 500 μm, sai_d layer being derrived from said second layer of said casting membrane .

7 . A method described in claim 1 , therein a third dope containing said solid electrolyte , said organic solvent and said catalys t is cast such that said first dope is interposed between said second dope and saicϋ third dope .

8 - A method described in claim 2 , -wherein a third dope containing said solid electrolyte , said organic solvent and said catalyst is cast from a third casting die disposed at an upstream of said first casting die .

9 . A method described in claim 7 , wherein said catalyst in said second dope and said catalyst in said tlxird dope are different from each other .

IO . An apparatus of pzroducing a solid electrolyte multilayer membrane , comprising : a casting device fox¹ casting plural dopes from a casting die onto a running support so as to form a layered casting membrane and peeling said casting membrane as a layered wet membrane; a first drying device for drying said wet membrane in a state that both side edges thereof are held by holding cϋevices; and a second durying device for dryiixg said wet membrane supported by rollers to form said solid electrolyte multilayer membrane, said second drying device being disposed at a downstream of said first drying device, wherein said plural dopes are a first dope and a second dope, said first dope containing an organic sol.vent and a solid electrolyte being a solid electrolyte layer of &. fuel cell, and said second dope containing said solid electrolyte, said organic; solvent and a catalyst promoting a redox reaction of electrodes in said fuel cell.

11. A sol±.d electrolyte multilayer membrane piroduced by a method described in claim 1.

12 . A membrane electrode assembly, comprising ; a solid electrolyte multilayerr membrane descriLbed in claim 11 ; an anode adhered to one surface of said solid electrolyte multilayer membrane , said anode generating protons from a hydrogen-containing material suppl ied from outside ; and a cathode adhered to the other surface of said solid electrolyte multilayer membrane, said cathode synthe sizing water from said protons permeated throiαgh said solid electrolyte multilayer membrane and gas supplied from outside -

13. A fuel cell , comprising : a membrane electrode assembly described in claim 12 ; current collectors one of whzLch provided in contact with said anode and the other of which parovided in contact with said cathode , said current collector on said anode side receiving and passing electrons between said anoc3.e and outside , whereas said current collector on said cathode side receiving and passing said electrrons between said cathode and outsicle .