JP2006086130A - Polymer electrolyte fuel cell, its manufacturing method, and inspection method for it - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress or remove obstruction of long stable power generation operation caused by minute short circuit between electrodes or hydrogen leak current involved in an electrolyte membrane electrode assembly (MEA) from before power generation operation starts. <P>SOLUTION: By limiting hydrogen leak current per electrode area of a cell stack in a polymer electrolyte fuel cell to 3 mA/cm<SP>2</SP>or less, generation of the factor of obstruction is suppressed and long stable power generation operation is realized. In the manufacturing method or the inspection method of the fuel cell, by removing a polymer electrolyte membrane, the MEA, or a unit cell in which initial short circuit conductivity or initial hydrogen leak current per electrode area exceeds the prescribed value, the fuel cell having high durability can efficiently be manufactured. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a fuel cell using a hydrogen ion conductive polymer electrolyte used for a portable power source, a power source for an electric vehicle, a cogeneration system, and the like, a manufacturing method thereof, and an inspection method thereof.

  A fuel cell using a hydrogen ion conductive polymer electrolyte generates electric power and heat simultaneously by electrochemically reacting a fuel gas containing hydrogen with an oxidant gas containing oxygen such as air. It is something to be made. First of all, a catalyst layer having a catalytic reaction role with carbon powder carrying a platinum-based metal catalyst as the main component is formed on both sides of a hydrogen ion conductive polymer electrolyte membrane that selectively transports hydrogen ions. To do. Next, a gas diffusion layer is formed on the outer surface of the catalyst layer by using a porous support having both fuel gas permeability and electronic conductivity, such as carbon cloth, carbon unwoven cloth or carbon paper. And the catalyst layer are combined to form an electrode. The electrode for fuel gas is called a fuel electrode or anode, and the electrode for oxidant gas is called an oxidant electrode or cathode.

  Next, a gas seal material or a gasket is disposed around the electrode with a polymer electrolyte membrane interposed so that the supplied fuel gas leaks to the outside and the fuel gas and the oxidant gas do not mix with each other. This sealing material or gasket is assembled in advance with an electrode and a polymer electrolyte membrane, and this is called an electrolyte membrane electrode assembly (MEA). As a method of assembling the MEA in advance, a method of joining the catalyst layer and the gas diffusion layer by a hot press of usually about 100 ° C. to 150 ° C. is employed. This is performed in order to improve the handling property at the time of assembling the MEA, or to improve the adhesiveness between the catalyst layer and the gas diffusion layer, thereby improving the reactivity and reducing the contact resistance.

  On the outside of the MEA, a conductive separator plate for mechanically fixing the MEA and electrically connecting adjacent MEAs to each other in series is disposed. In the portion of the separator plate that contacts the MEA, a reaction gas is supplied to the electrode surface, and a gas flow path for carrying away the generated gas and surplus gas is formed. Although the gas flow path can be provided separately from the separator plate, a system in which a gas flow channel is provided on the surface of the separator plate to form a gas flow path is common.

  In order to supply the fuel gas to the gas flow path, a piping jig for branching the piping for supplying the fuel gas to the number of separator plates to be used and directly connecting the branch destination to the separator-like groove is required. . This jig is called a manifold, and the type connected directly from the fuel gas supply pipe as described above is called an external manifold. There is a type of this manifold called an internal manifold with a simplified structure. The internal manifold is a separator plate in which a gas flow path is formed with a through-hole, through the gas flow path to the hole, and fuel gas is directly supplied from the hole.

  Since the fuel cell generates heat during operation, it is necessary to cool it with cooling water or the like in order to maintain the battery at a good temperature. Usually, a cooling part for flowing cooling water every 1 to 3 cells (unit batteries) is inserted between the separator plate and the separator plate. In many cases, a cooling section formed by joining the water flow paths to each other is used. These MEAs, separator plates, and cooling sections are alternately stacked to form a fuel cell stack by stacking about 10 to 200 cells, and then the battery stack is sandwiched between end plates through current collector plates and insulating plates. The structure of a general laminated polymer electrolyte fuel cell is fixed from both ends with fastening bolts.

The conventional fuel cell stack is fastened to about 10 to ²⁰ kgf / cm ² in order to reduce the contact resistance between the polymer electrolyte membrane, the electrode, and the separator plate, and further to ensure the gas sealing performance by the gas sealing material or gasket. It was normal to fasten the stack with pressure. Therefore, it is a general fuel cell structure that consists of an end plate made of a metal material with excellent mechanical strength, combined with a fastening bolt and a spring, and the battery stack is fixed at both ends by applying fastening pressure to the battery stack. Met.

  In addition, since the supplied humidified gas and cooling water are in contact with a part of the end plate, from the viewpoint of corrosion resistance, it is usual to use a stainless steel material having excellent corrosion resistance among metal materials. In addition, a metal material having a higher conductivity than the carbon material is used for the current collector plate, and from the viewpoint of contact resistance, it is usual to use a surface treatment in some cases. Furthermore, since the end plates at both ends are in electrical contact via the fastening bolts, an insulating insulating plate has been inserted between the current collector plate and the end plate from the viewpoint of insulation.

  The separator plate used in such a polymer electrolyte fuel cell has high conductivity and high gas tightness with respect to the fuel gas, and further has high corrosion resistance against the reaction when oxidizing / reducing hydrogen / oxygen. That is, it must have acid resistance. For this reason, the conventional separator plate is a press plate in which a gas channel is formed by cutting on the surface of a dense gas-impermeable carbon plate or a convex pattern portion corresponding to the gas channel is formed. A graphite powder was put into a mold together with a thermosetting resin, and this was hot-pressed.

  In addition, attempts have been made to use metal plates such as stainless steel instead of conventionally used carbon materials. Since the metal plate is exposed to an oxidizing atmosphere at a relatively high temperature, the metal plate is corroded or dissolved when used for a long time. When the metal plate is corroded, the electric resistance of the corroded portion increases and the output of the battery decreases. In addition, when the metal plate is dissolved, the dissolved metal ions diffuse into the polymer electrolyte membrane, which ion exchanges with the ion exchange site of the polymer electrolyte membrane, resulting in a decrease in the ionic conductivity of the polymer electrolyte membrane itself. To do. In order to avoid such deterioration, it is usual to apply noble metal plating having a certain thickness on the surface of the metal plate.

As described above, conventionally, when constructing an MEA, in order to improve the handling property when assembling the MEA into a stacked fuel cell, or by improving the adhesion between the catalyst layer and the gas diffusion layer, the reactivity can be improved. In order to reduce the contact resistance, a method of joining the catalyst layer and the gas diffusion layer by hot pressing at about 100 ° C. to 150 ° C. has been adopted. At this time, if hot pressing is performed at a high temperature of about 130 ° C. or higher and a pressure of about 20 kgf / cm ² or more, depending on the material of the gas diffusion layer to be used and the unevenness of the surface, and depending on the strength and thickness of the polymer electrolyte membrane The present inventors have found that a slight short circuit occurs.

  The slight short circuit has a problem that it is very difficult to find it in a normal fuel cell use state or power generation state as the slight short circuit. In addition, unless it is a large short circuit, the initial power generation characteristics of the fuel cell are hardly affected, and there is a problem that the discovery is further difficult. Conversely, if a large short circuit exists from the beginning, due to a decrease in open circuit voltage or due to a cross leak phenomenon in which hydrogen gas as fuel on the fuel electrode side and air as oxidant on the oxidant electrode side mix. The power generation characteristics are degraded from the beginning. Therefore, a large initial short circuit is relatively easy to find. However, it was discovered that the initial micro short circuit, which is difficult to find under normal use conditions, has a significant effect on the durability of the power generation characteristics of the fuel cell.

Even if a slight short circuit occurs, an excessive current flows in the part where the short circuit occurs, so heat is generated when the current flows, and the polymer electrolyte contained in the polymer electrolyte membrane and electrode is thermally decomposed and deteriorated due to the heat generation. It has been found that the short circuit gradually increases with the passage of time, and as a result, the cross leak phenomenon is increased and the durability is greatly affected. In addition, conventionally, when a battery stack is configured by stacking a plurality of unit batteries (cells), it is usual to fasten the battery stack with a fastening pressure of about 10 to ²⁰ kgf / cm ² , and thus the same viewpoint From the above, it was found that a minute short circuit may occur, and this minute short circuit has a great influence on durability.

  FIG. 6 shows a schematic cross-sectional view of a conceptual or ideal unit cell. In FIG. 6, the polymer electrolyte membrane 61 is sandwiched between two electrodes 64 each composed of a catalyst layer 63 and a gas diffusion layer 62 on both sides to constitute an MEA 65. The MEA 65 is sandwiched between two conductive separator plates 66 on both sides to constitute a unit cell. As can be seen from FIG. 6, in the conceptual unit cell, the polymer electrolyte membrane 61, the catalyst layer 63, and further the gas diffusion layer 62 have a uniform thickness, and a slight short circuit between the two electrodes. Such a configuration is not assumed. Accordingly, the gas flow path 67 in the conductive separator 66 also maintains the original conceptual cross section.

  However, in the unit cell that is actually manufactured, the uniformity of the thickness of each film or each layer is determined by the pressure applied by hot pressing between the catalyst layer and the gas diffusion layer or the fastening pressure when the MEA is sandwiched between the conductive separator plates. We found that there are many cases where it is greatly damaged. This will be described with reference to FIG. FIG. 7 is a schematic cross-sectional view of a unit cell in a state where such uniformity and the like are impaired and a micro short circuit is generated between both electrodes. In FIG. 7, the gas diffusion layer 72, the catalyst layer 73, and the polymer electrolyte membrane 71 are greatly distorted, and a minute short-circuit portion 78 is generated. At the same time, a part 72 a of the gas diffusion layer 72 is sunk into the gas flow path 77 in the conductive separator plate 76. The sagging gas diffusion layer blocks part of the gas flow path, hinders gas flow and impairs fuel cell characteristics.

  Also, an excessive hot press load and temperature, and an excessive fastening pressure on the battery stack not only cause a micro short circuit or a sag of the gas diffusion layer into the gas flow path, but in some cases may cause physical damage to the polymer electrolyte membrane. It has also been found that this causes a general defect and causes a cross leak. The cross-leak phenomenon, in which hydrogen, which is fuel on the fuel electrode side, and air, which is oxidant on the oxidant electrode side, is mixed, is usually caused by defects such as physical holes in the polymer electrolyte membrane. It has been considered that a cross leak has occurred through a defect.

  As a result of investigation by the inventors, when a defect occurs in the polymer electrolyte membrane, hydrogen and air that cross leak through the defect are burned, and the battery characteristics are deteriorated. The polymer electrolyte contained in the material deteriorates by thermal decomposition, and the short circuit gradually increases with time. As a result, the cross leak phenomenon is increased and the durability is greatly affected.

  As a prior art related to such a technique, a general technique related to applying a fastening pressure to a battery stack is disclosed in Patent Document 1 or 2, for example. However, those prior art documents do not teach technical problems such as a micro short circuit proposed by the present invention and solutions therefor.

Furthermore, as described above, cross-leakage, which has been considered to occur only when a physical defect occurs in the polymer electrolyte membrane, diffuses hydrogen dissolved in the polymer electrolyte membrane to the oxidant electrode side due to a concentration gradient. In addition, the present inventors have discovered that this phenomenon can also occur by reaching the oxidant electrode side. This phenomenon is unavoidable in principle in the case of a fuel cell using a perfluorosulfonic acid membrane as a polymer electrolyte, and has been considered to be a cross leak caused by a physical defect of a polymer electrolyte membrane that has been considered in the past. Need to be considered separately.
JP 2002-203578 A JP 2002-246044 A

  In other words, it is necessary to have an inspection method that eliminates cross-leakage caused by hydrogen dissolution and diffusion that is unavoidable in principle, diagnoses only cross-leakage caused by defects physically generated in the polymer electrolyte membrane, and judges whether the battery is good or bad It is. However, conventionally, it has been very difficult to diagnose only the cross leak caused by the defects physically generated in such a polymer electrolyte membrane.

An object of the present invention is to provide a polymer electrolyte fuel cell capable of performing high-performance power generation stably over a long period of time by improving the MEA disposed in the cell stack.
Another object of the present invention is that a short-circuit between two electrodes or a hydrogen leakage current found by the present inventors is a long-term power generation operation inhibiting factor included in the MEA before power generation operation in a polymer electrolyte fuel cell. By suppressing the value, it is to enable long-term power generation operation of the fuel cell.

Still another object of the present invention is to suppress the inhibition of gas flow due to the gas diffusion layer sagging into the gas flow path of the conductive separator plate in each unit cell of the polymer electrolyte fuel cell.
Still another object of the present invention is to enable high-performance power generation stably over a long period of time by introducing a new MEA quality determination method in a method for manufacturing a polymer electrolyte fuel cell or a method for inspecting the same. It is to provide a fuel cell efficiently.

In order to solve the above-described problems and achieve one of the objects of the present invention, a polymer electrolyte fuel cell according to one aspect of the present invention includes a hydrogen ion conductive polymer electrolyte membrane and both surfaces of the electrolyte membrane. An electrolyte membrane electrode assembly comprising a pair of electrodes each comprising a gas diffusion layer and a catalyst layer, and a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes. A plurality of unit cells each having a conductive separator plate in contact with the electrode and a further conductive separator plate in contact with the other electrode having a gas flow path for supplying and discharging an oxidant gas to the other electrode are stacked in the stacking direction. A polymer electrolyte fuel cell comprising a battery stack that is fastened to each other, wherein the short-circuit conductivity of each electrolyte membrane electrode assembly is 1.5 mS / cm ² or less.

Furthermore, a polymer electrolyte fuel cell according to another aspect of the present invention includes a hydrogen ion conductive polymer electrolyte membrane, and a pair of electrodes that are respectively disposed on both sides of the electrolyte membrane and each include a gas diffusion layer and a catalyst layer. An electrolyte membrane electrode assembly, a conductive separator plate having a gas flow path for supplying and discharging fuel gas to and from one of the pair of electrodes, and an oxidant gas for the other electrode A polymer electrolyte fuel cell comprising a battery stack in which a plurality of unit cells each having a gas flow path for supplying and discharging gas and having a further conductive separator plate in contact with the other electrode are stacked and fastened in the stacking direction The hydrogen leakage current of the unit battery using each electrolyte membrane electrode assembly is 3 mA / cm ² or less per electrode area.

A manufacturing method according to an aspect of the present invention includes a process of manufacturing a plurality of unit cells, a process of stacking the plurality of unit cells to manufacture a battery stack, and a process of fastening the battery stack to the stacking method. A method of manufacturing a polymer electrolyte fuel cell comprising the steps of producing each unit cell includes a pair of electrodes each comprising a gas diffusion layer and a catalyst layer on both sides of a hydrogen ion conductive polymer electrolyte membrane. A step of arranging and manufacturing an electrolyte membrane electrode assembly, a step of arranging a conductive separator plate having a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes, and the other The polymer electrolyte type fuel comprising a step of contacting a further conductive separator plate having a gas flow path for supplying and discharging the oxidant gas to and from the electrode, The method for producing a pond further includes a step of measuring a short-circuit conductivity of each electrolyte membrane electrode assembly and / or the hydrogen leakage current of the unit cell, and the short-circuit conductivity is a predetermined value of short-circuit conductivity and / or the hydrogen. When the leakage current exceeds a hydrogen leakage current predetermined value, the short-circuit conductivity predetermined value and / or the electrolyte membrane electrode assembly exceeding the hydrogen leakage current predetermined value, or a unit cell having the electrolyte membrane electrode assembly or a unit battery thereof And an inspection process including a step of removing the battery stack.

The predetermined short-circuit conductivity value is preferably 1.5 mS / cm ² , and the predetermined hydrogen leak current value is preferably 3 mA / cm ² .
The short-circuit conductivity of the electrolyte membrane electrode assembly may be converted from a steady current value or a steady voltage value obtained by applying a constant DC voltage or a constant DC current to the electrolyte membrane electrode assembly. preferable.
Here, it is more preferable that the constant voltage is 0.5 V or less per each electrolyte membrane electrode assembly.
More preferably, the constant current is 5 mA / cm ² or less per electrode area of each electrolyte membrane electrode assembly.

Furthermore, it is preferable that the measurement of the short-circuit conductivity of the electrolyte membrane electrode assembly is performed by placing a pair of the electrodes in the same air atmosphere or inert gas atmosphere.
Further, the hydrogen leak current of the unit cell supplies an inert gas to the one electrode and a fuel gas to the other electrode, and a constant DC voltage or a constant DC current is supplied to the electrolyte membrane electrode assembly. It is preferably converted by subtracting a value corresponding to the short-circuit conductivity from a value calculated from a steady-state current value or a steady-state voltage value obtained by applying.

  Further, before the inspection process stacks the unit cells to produce the battery stack, the short-circuit conductivity of the electrolyte membrane electrode assembly of the unit cells and / or the hydrogen leakage current of the unit cells is determined. A unit cell having an electrolyte membrane electrode assembly or an electrolyte membrane electrode assembly of which the short circuit conductivity exceeds the predetermined value of the short circuit conductivity and / or the hydrogen leak current exceeds the predetermined value of the hydrogen leak current It is preferable to include the process of removing.

  Further, after the inspection process has produced at least the unit battery, the short-circuit conductivity of the electrolyte membrane electrode assembly of the unit battery and / or the hydrogen leakage current of the unit battery is measured, and the short-circuit conductivity is measured. An electrolyte membrane electrode assembly, a unit cell having the electrolyte membrane electrode assembly, or a battery stack having the unit cell, wherein the short-circuit conductivity exceeds a predetermined value and / or the hydrogen leakage current exceeds the predetermined value It is preferable to include the process of removing.

A manufacturing method according to another aspect of the present invention includes a process of manufacturing a plurality of unit cells, a process of stacking the plurality of unit cells to manufacture a battery stack, and a process of fastening the battery stack in the stacking direction. A process for producing each unit cell includes a pair of electrodes each comprising a gas diffusion layer and a catalyst layer on both sides of a hydrogen ion conductive polymer electrolyte membrane. Disposing an electrolyte membrane electrode assembly, and disposing a conductive separator plate having a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes. And a step of arranging a further conductive separator plate having a gas flow path for supplying and discharging the oxidant gas to the other electrode so as to be in contact with the polymer electrolyte type fuel. Battery manufacturing method further surface of the gas diffusion layer, characterized in that they are subjected to a pre-smoothing treatment.

  An inspection method for a polymer electrolyte fuel cell according to the present invention comprises an electrolyte comprising a hydrogen ion conductive polymer electrolyte membrane and a pair of electrodes each having a gas diffusion layer and a catalyst layer disposed on both sides of the electrolyte membrane. A membrane electrode assembly, a conductive separator plate having a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes, and an oxidant gas being supplied to the other electrode A polymer electrolyte fuel cell comprising a battery stack in which a plurality of unit cells each having a gas flow path for discharging and having a further conductive separator plate in contact with the other electrode are stacked and fastened in the stacking direction The short-circuit conductivity of each electrolyte membrane electrode assembly and / or the single unit before the production of the battery stack before or after the power generation operation of the polymer electrolyte fuel cell. Measuring a hydrogen leakage current of a potential battery, and inspecting whether the short-circuit conductivity exceeds a predetermined value of the short-circuit conductivity and / or whether the hydrogen leak current exceeds a predetermined value of the hydrogen leak current It is characterized by.

In the polymer electrolyte fuel cell of the present invention, the durability of the fuel cell can be greatly improved by setting the hydrogen leakage current per unit cell electrode area to a predetermined value or less such as 3 mA / cm ^2. it can.
In the method for manufacturing or inspecting the polymer electrolyte fuel cell of the present invention, the initial short circuit conductivity exceeds a predetermined value such as 1.5 mS / cm ² , or the initial hydrogen leakage current per electrode area is 3 mA. By removing the polymer electrolyte membrane, MEA, unit cell or battery stack exceeding a predetermined value such as / cm ² , a highly durable fuel cell can be efficiently produced.

In the polymer electrolyte fuel cell according to one aspect of the present invention, the short-circuit conductivity of the MEA between the fuel electrode and the oxidant electrode of the unit cell, that is, the short-circuit conductivity due to the DC resistance component is 1.5 mS / cm. ² or less. Even if there is a micro short circuit that occurs due to the hot press of the MEA or the fastening pressure to the battery stack, if it is 1.5 mS / cm ² or less, preferably 1.0 mS / cm ² or less, the micro short circuit Has found that there is almost no effect on the durability of the fuel cell.

  The short-circuit conductivity is preferably converted from a steady current value or a steady voltage value obtained as a result of applying a constant DC voltage or a constant DC current. Here, the steady current value and the steady voltage value will be described. As will be described later with reference to FIG. 13 or FIG. 14, when a constant DC voltage or a constant current is applied between both electrodes of the MEA, immediately after the application, the current or voltage between the two electrodes due to the voltage application or current application suddenly increases. However, the change with time is almost eliminated when a certain period of time such as 1 minute or 3 minutes elapses. In this specification, the current value and the voltage value when such a change with time is almost eliminated are referred to as a steady current value and a steady voltage value.

The constant voltage is preferably 0.5 V or less per MEA, that is, per unit battery. The constant current is preferably 5 mA / cm ² or less per MEA electrode area.
In addition, the measurement of the short-circuit conductivity is performed on both the fuel electrode and the oxidant electrode of the MEA in the same gas atmosphere, and the atmosphere is an atmosphere mainly composed of an air atmosphere or an inert gas. preferable.

  Here, when there is no difference in hydrogen gas partial pressure between both electrodes of the fuel electrode and the oxidizer electrode, there is no difference in oxygen partial pressure between both electrodes, and when the supply gas and the polymer electrolyte do not contain moisture Since the ionic conductivity of the electrolyte is small, an electrochemical reaction hardly occurs. Therefore, when a constant DC voltage or a constant DC current is applied between both electrodes and a certain period of time elapses, a constant current derived from only the DC resistance component (that is, a steady current value) or a constant voltage (that is, a steady voltage value) Is observed.

Therefore, for example, the resistance value is calculated by dividing the applied voltage by the observed steady current value, and the short-circuit conductivity can be calculated by calculating the reciprocal of the resistance value per unit area of the electrode.
At this time, immediately after the DC voltage or DC current is applied, a current or voltage larger than the steady value is generated. Therefore, in order to avoid decomposition of the polymer electrolyte, the electrode catalyst, and the like due to this large current or high voltage, the applied constant DC voltage is preferably 0.5 V or less, preferably about 0.2 V.

For the same reason, a constant direct current to be applied, MEA electrode area per 5 mA / cm ² or less, preferably about 2 mA / cm ^2.
Furthermore, in order to measure the short circuit conductivity based only on the DC resistance component, it is necessary to eliminate the difference between the hydrogen partial pressure and the oxygen partial pressure between the two electrodes. For this reason, it is necessary to supply a gas having the same composition between both electrodes during measurement, and it is preferable that such a gas is an inert gas such as nitrogen. However, in order to perform the measurement simply, it is possible to expose the MEA in the air, sandwich the MEA with a conductive plate capable of collecting current, and apply a DC voltage or a DC current as it is.

In the polymer electrolyte fuel cell according to another aspect of the present invention, the hydrogen leakage current per electrode area is 3 mA / cm ² or less. It was found that if the hydrogen leakage current is 3 mA / cm ² or less, the hydrogen leakage current has little influence on the durability of the fuel cell. Further, when there is no physical defect in the polymer electrolyte membrane, the hydrogen leak current value due to only hydrogen gas dissolution and diffusion is about 1 to ² mA / cm ² , and the hydrogen leak current is ² mA / cm ² or less. Found that the influence of the hydrogen leakage current on the durability of the fuel cell was further reduced. Accordingly, it is more preferable that the hydrogen leak value is ² mA / cm ² or less.

  Further, the hydrogen leakage current of the unit cell is a steady current obtained as a result of supplying an inert gas to one electrode side and a fuel gas to the other electrode side and applying a constant DC voltage or a constant DC current. It is preferably converted by subtracting a value corresponding to the short-circuit conductivity from a value or a value calculated from a steady voltage value. A steady current value obtained by supplying an inert gas to one electrode side and a fuel gas to the other electrode side and applying a constant DC voltage or a constant DC current, or a total value calculated from a steady voltage value The conductivity includes both the conductivity based on the micro short circuit and the hydrogen leak current value (more precisely, the conductivity corresponding to the hydrogen leak current value). Therefore, a hydrogen leakage current value can be calculated by obtaining a value obtained by subtracting the conductivity due to a short circuit from the total conductivity and reconverting the value into a current value per unit electrode area.

The groove width of the gas flow path of the conductive separator plate is preferably 0.8 mm or more and 1.0 mm or less, and the depth of the gas flow path is preferably 0.3 mm or more and 1.0 mm or less. . In particular, when a flexible gas diffusion layer such as carbon cloth is used, a phenomenon occurs in which the gas diffusion layer sags into the flow path if the groove width of the flow path is too wide than 1.0 mm. On the other hand, if the groove width of the gas channel is narrower than 0.8 mm, the gas flow rate may be insufficient. Further, when the groove depth of the flow path is shallower than 0.3 mm, the gas diffusion layer that has sunk blocks the gas flow path, impedes the gas flow, and impairs the final fuel cell characteristics. On the other hand, if the groove depth of the gas flow path exceeds 1.0 mm, the strength of the conductive separator plate may be insufficient. Furthermore, if it is going to raise intensity | strength in the state of such a deep groove, the board thickness of an electroconductive separator board must be made thick too much.

  In the present specification, the term “groove width” means the average width of the grooves. When the groove wall is perpendicular to the surface of the conductive separator plate, that is, when the groove wall has no taper, the groove width is constant from the upper surface of the groove to the bottom surface of the groove. Therefore, for example, when the groove width is described as 1.0 mm, the groove width is 1.0 mm from the groove upper surface to the groove bottom surface. However, even when the groove width is described as 1.0 mm, when the groove wall is tapered such that the upper surface of the groove width is wider than the groove bottom surface, Under the condition that the average value is 1.0 mm, the width of the upper surface of the groove is wider than 1.0 mm, for example, 1.1 mm, while the width of the bottom surface of the groove is narrower than 1.0 mm, for example, 0.9 mm. It is also possible here.

  Moreover, it is preferable that the area of the part which the said conductive separator plate and an electrode contact is larger than the area of the part which the said conductive separator plate and an electrode do not contact among the electrode areas of said MEA. When a flexible gas diffusion layer such as carbon cloth is used, the area of the electrode contact area between the conductive separator plate and the electrode is equal to or greater than the area of the electrode contact area between the conductive separator plate and the electrode. Therefore, the gas diffusion layer can be prevented from sagging into the flow path, and the portion of the polymer electrolyte membrane in contact with the electrode can be easily and uniformly held. In addition, it has also been found that, due to such a contact area relationship, even if a conductive separator plate having lower conductivity is used, the influence on the battery characteristics due to the increase in resistance can be suppressed.

  In the method for producing a polymer electrolyte fuel cell according to another aspect of the present invention, a step of measuring the short-circuit conductivity of each electrolyte membrane electrode assembly and / or the hydrogen leakage current of the unit cell; When the short circuit conductivity is a short circuit conductivity predetermined value and / or the hydrogen leak current exceeds a hydrogen leak current predetermined value, the electrolyte membrane electrode assembly exceeding the short circuit conductivity predetermined value and / or the hydrogen leak current predetermined value or It includes an inspection process including a step of removing a unit battery having the electrolyte membrane electrode assembly or a battery stack having the unit battery. By introducing such an inspection process, a fuel cell having stable characteristics over a long period of time can be produced without requiring a power generation operation of the fuel cell.

Here, it is preferable that the predetermined value of the short circuit conductivity is 1.5 mS / cm ² and the predetermined value of the hydrogen leakage current is 3 mA / cm ² . It is preferable to use them as these predetermined value determination criteria when manufacturing a fuel cell that is stable for a long period of time.
Further, as the inspection process, it is possible to more efficiently manufacture a long-term stable fuel cell by discriminating and removing defective MEAs or defective unit cells based on the predetermined value before producing a battery stack. Become. Further, as the inspection process, after the battery stack is manufactured, the power generation characteristics must be actually measured in the past by discriminating and removing the defective MEA, the defective unit battery, or the defective battery stack based on the predetermined value. Thus, the shipment inspection process of the fuel cell or battery stack that could not be determined can be easily performed.

Hereinafter, examples of embodiments of the present invention will be described using examples, but the present invention is not limited to the configuration requirements of these examples.
Example 1
First, a method for producing an electrode having a catalyst layer will be described. An electrode catalyst comprising 25% by weight of platinum particles having an average particle diameter of about 3 nm supported on acetylene black powder was used. A dispersion solution in which perfluorocarbonsulfonic acid powder was dispersed in ethyl alcohol was mixed with a solution in which the catalyst powder was dispersed in isopropanol to form a catalyst paste.

On the other hand, carbon paper (TGP-H-120 manufactured by Toray Industries, Inc.) having an outer dimension of 9 cm × 20 cm and a thickness of 270 μm is prepared as a porous carbon support for the gas diffusion layer for electrodes. The water-repellent treatment was applied. That is, after impregnating the carbon paper with an aqueous dispersion containing fluororesin (Daikin Kogyo Co., Ltd., NEOFLON ND1), it was dried and heated at 380 ° C. for 30 minutes to give water repellency. A catalyst layer was formed on one surface of the water-repellent carbon paper by applying a catalyst paste using a screen printing method. At this time, a part of the catalyst layer is embedded in the carbon paper. Thus, carbon paper with a catalyst layer was formed as an electrode. Amount of platinum contained in the electrode after forming the 0.3 mg / cm ^2, the amount of perfluorocarbon sulfonic acid was adjusted to be 1.2 mg / cm ^2.

Next, 3 kgf at a temperature of 130 ° C. so that the catalyst layer is in contact with the side of the electrolyte membrane on both sides of the proton conductive polymer electrolyte membrane having an outer dimension of 10 cm × 26 cm. Bonding was performed by hot pressing at a pressure of / cm ² to obtain an electrolyte membrane electrode assembly (MEA). Here, a proton conductive polymer electrolyte obtained by thinning perfluorocarbon sulfonic acid to a thickness of 50 μm was used.

1A and 1B show the front and back surfaces of one conductive separator plate used in this example. Specifically, a conductive separator plate in which a grooved shape for gas circulation, that is, a gas flow path, is provided on the front and back surfaces of a carbon plate made of dense and gas-impermeable isotropic graphite. It is the typical top view which showed the shape of. FIG. 1A shows the shape of the oxidant gas flow groove formed on the surface, and FIG. 1B shows the shape of the fuel gas flow groove on the back surface thereof. The separator plate has a size of 10 cm × 26 cm, a thickness of 2 mm, and the groove portions 11a and 11b are concave portions having a width of 1.0 mm and a depth of 0.7 mm, and gas flows through these portions. Further, the rib portions 12a and 12b between the gas flow paths are convex portions having a width of 1.2 mm. Also, a manifold hole for the oxidant gas (injection port 13a, outlet 13b) and a manifold hole for the fuel gas (injection port 14a, outlet 1).
4b) and manifold holes (inlet 15a, outlet 15b) of the cooling water were formed.

  1A and 1B, when a single battery is manufactured by sandwiching an MEA with two conductive separator plates whose front and back surfaces are shown later, the surface of the first conductive separator plate on the side of FIG. 1B and The other conductive separator plate was opposed to the surface on the side of FIG. 1A. Therefore, as shown in the figure, the positions, shapes, and sizes of the corresponding elements such as the gas flow grooves on the opposing surfaces correspond to each other.

  2A and 2B are schematic plan views of the front and back surfaces of another conductive separator plate used in this example. 2A shows the shape of the oxidant gas flow groove formed on the same surface as in FIG. 1A, and FIG. 2B shows the shape of the cooling flow channel for flowing the cooling water formed on the back surface. . 2A and 2B, the position and size of the cooling water manifold holes (injection port 25a, outlet 25b) are the same as the cooling water manifold holes 15a and 15b shown in FIGS. 1A and 1B. In addition, the position and the size of each manifold hole (oxidant gas inlet 23a, outlet 23b, fuel gas inlet 24a, outlet 24b) for gas flow are also the same as the gas manifold holes in FIGS. 1A and 1B. Formed in the same position and size. Reference numeral 21 denotes a groove portion as a concave-shaped channel for water flowing in from the cooling water inlet 25a, and the depth of the concave portion is 0.5 mm. 22 is a rib part left as a convex part when the groove part 21 of cooling water is formed between cooling water passages. The cooling water flows from the inlet 25a, flows through the groove portion 21, and reaches the outlet 25b.

  When a single battery is manufactured by sandwiching an MEA between the conductive separator plate whose front and back surfaces are shown in FIGS. 1A and 1B and the conductive separator plate whose front and back surfaces are shown in FIGS. 2A and 2B as will be described later. The surface on the FIG. 1B side of the former conductive separator plate and the surface on the FIG. 2A side of the latter conductive separator plate were made to face each other. Therefore, as shown in the figure, the positions, shapes, and sizes of the corresponding elements such as the gas flow grooves on the opposing surfaces correspond to each other.

  3A and 3B are schematic plan views of the front and back surfaces of still another conductive separator plate used in this example. FIG. 3A shows the shape of the cooling channel for flowing the cooling water formed on the surface, and FIG. 3B shows the shape of the fuel gas circulation groove formed on the back surface in the same manner as FIG. 1B. . 3A and 3B, the position and size of the cooling water manifold holes (injection port 35a, outlet 35b) are the same as the cooling water manifold holes 15a and 15b shown in FIGS. 1A and 1B. In addition, the position and size of each gas distribution manifold hole (oxidant gas injection port 33a, outlet 33b, and fuel gas injection port 34a, 34b) are the same as those of the gas manifold holes of FIGS. 1A and 1B. Formed in the position and size. Reference numeral 31 denotes a concave groove serving as a flow path for water flowing in from the cooling water inlet 35a, and the depth of the concave is 0.5 mm. Reference numeral 32 denotes a rib portion left as a convex portion when the groove portion 31 of the cooling water passage is formed between the cooling water passages. The cooling water flows in from the inlet 35a, flows through the groove portion 31, and reaches the outlet 35b.

  2A and 2B, a conductive separator plate unit having a cooling part inside as described later, and a conductive separator plate having front and back surfaces shown in FIGS. 3A and 3B. When manufacturing, the surface of FIG. 2B side of the former conductive separator plate and the surface of FIG. 3A side of the latter conductive separator plate were made to face each other. Therefore, as shown in the figure, the positions, shapes, and sizes of the corresponding elements such as the cooling water flow paths on the opposing surfaces and the ribs between them are configured to correspond to each other. In addition, in the case of a conductive separator plate arranged so as to be in contact with the current collector plate at both ends of the battery stack as shown in FIG. No gas channel or cooling water channel was formed, and such a surface was left smooth.

When manufacturing each unit battery by sandwiching the MEA with a conductive separator plate, or when manufacturing a separator plate unit having a cooling water unit, the cross section is common to the six drawings from FIG. 1A to FIG. 3B. Each of the O-ring-shaped gas seal member (made by DuPont Dow Elastomer Japan Co., Ltd .: Part No. Viton) having a circular or elliptical shape surrounds each manifold hole and further surrounds each gas flow path and each cooling water flow path. It placed on the surface of the conductive separator plate. Specifically, in those six figures, numbering 16a, 16b, 17a, 17b, 18a, 18b, 26a, 26b, 27a, 27b, 28a, 28b, 36a, 36b, 37a, 37b, 38a, 38b Are circular and elliptical plan views surrounding each manifold, and a loop surrounding each gas flow passage and each cooling water flow path.

When the conductive separator plates on which the gas seal members are arranged are directly or when they are combined via the MEA, the O-ring gas seal members are directly or indirectly via the MEA. These O-ring gas seal members were arranged at corresponding positions on the opposing surfaces of the combined conductive separator plates so that the gas or water was sealed against each other. This is to constitute a gasket made of a gas seal member.
Here, the term “O-ring shape” is used based on a ring shape or a loop shape corresponding to the shape of the element to be enclosed.

  For the production of the cooling water unit, the separator plate shown in FIGS. 2A and 2B and the separator plate shown in FIGS. 3A and 3B are combined so that the surface of FIG. 2B faces the surface of FIG. 3A. It was. Accordingly, the convex portions (rib portions) and the concave portions (groove portions) on the cooling water flow channel surface shown in FIG. 2B and the cooling water flow channel surface shown in FIG. The sealing members are pressed against each other, and are bonded together with a conductive adhesive as necessary, thereby obtaining a conductive separator plate unit having a cooling water flow path therein.

Next, manifold holes for circulating cooling water, fuel gas, and oxidant gas were formed in the produced proton conductive polymer electrolyte membrane of MEA. The configuration is shown in FIG. In FIG. 4, 40 is an electrode portion, 41 is a proton conductive polymer electrolyte membrane, 43a and 43b are manifold holes (inlet and outlet, respectively) for oxidant gas flow, and 44a and 44b are manifold holes for fuel gas flow. (Respectively inlet and outlet), 45a and 45b are cooling water circulation manifold holes (respectively inlet and outlet, respectively), and the positions and sizes of these holes are the same as those of the conductive separator plate shown in FIGS. 1A to 3B. Same size in position.
FIG. 4 is a view showing the surface of the MEA on the fuel electrode side, although not shown, an electrode such as an oxidizer electrode and each manifold hole formed in the MEA are visible on the back side of the MEA. .

With reference to FIG. 5, the configuration of a polymer electrolyte fuel cell having a two-cell stack battery stack manufactured in this example will be described. In FIG. 5, the lower half is a schematic front view, and the upper half is a schematic cross-sectional view taken along a plane passing through the center of the manifold hole for fuel gas injection and parallel to the length direction of the conductive separator plate. The following description based on FIG. 5 will be described mainly with reference to the schematic sectional view of the upper half.
In FIG. 5, a two-cell stack battery stack and two cooling water units arranged outside the two current collector plates 581a and 581b are sandwiched. This will be explained in order from the left side. First, reference numeral 501 denotes a conductive separator plate having a smooth surface and a configuration as shown in FIG. 2B on the back surface, that is, a cooling water flow path and manifold holes. The surface 502 has a structure as shown in FIG. 3A, that is, a cooling water flow path and manifold holes are formed on the front surface, and a structure as shown in FIG. 3B, that is, a fuel gas flow path and manifold holes are formed on the back surface. It is a separator plate.

  The two conductive separator plates are bonded together so that the O-ring gas seal members 527 and 537a disposed on both conductive separators are pressed against each other, and a cooling water flow is provided between the two conductive separator plates. A cooling water unit having a passage 511 is formed, and a part of the fuel gas manifold 554 is formed. In the figure, gas seal members 527 and 537a that seal so as to surround a part of the fuel gas manifold 554 are shown.

503 has a configuration as shown in FIG. 1A on the front surface, that is, an oxidant gas flow path and manifold holes, and a back surface as shown in FIG. 1B, that is, a fuel gas flow path and manifold holes. It is a conductive separator plate. Between the surface of the conductive separator plate 503 and the back surface of the conductive separator plate 502, a configuration as shown in FIG. 4, that is, a polymer electrolyte membrane 541 in which each manifold hole is formed, and both surfaces thereof are arranged. Further, the MEA composed of the electrodes 540a and 540b is sandwiched, the fuel gas channel 521 is in contact with the electrode 540a, and the oxidant gas channel 531a is in contact with the electrode 540b and further sandwiched by the two conductive separator plates. Thus, one unit battery (cell) is formed. In the drawing, a part of the fuel gas manifold 554 formed by the sandwiching configuration and a part of the fuel gas manifold 554 are disposed so as to surround each other, and a part of each other is interposed via the polymer electrolyte membrane 541. Gas seal members 537b and 547a that are indirectly pressed are shown.

The surface 504 has a configuration as shown in FIG. 2A, that is, an oxidant gas flow path and each manifold hole formed on the front surface, and a structure as shown in FIG. 2B, that is, a cooling water flow path and each manifold hole formed on the back surface. This is a conductive separator plate. Between the surface of the conductive separator plate 504 and the back surface of the conductive separator plate 503, a configuration as shown in FIG. 4, that is, a polymer electrolyte membrane 551 in which each manifold hole is formed, and both surfaces thereof are arranged. Further, the MEA composed of the electrodes 550a and 550b is sandwiched, the fuel gas channel 531b is in contact with the electrode 550a, and the oxidant gas channel 561 is in contact with the electrode 550b, and further sandwiched by the two conductive separator plates. Thus, one unit battery (cell) is formed. In the drawing, a part of the fuel gas manifold 554 formed by the sandwiching structure and a part of the fuel gas manifold 554 are disposed so as to surround each other, and a part of each other is interposed via the polymer electrolyte membrane 551. Gas seal members 547b and 557a that are indirectly pressed are shown.

  Reference numeral 505 denotes a conductive separator plate having a structure as shown in FIG. 3A on the front surface, that is, a cooling water flow path and manifold holes, and a smooth structure on the back surface. The conductive separator plate 504 and the conductive separator plate 505 are bonded together so that the O-ring gas seal members 557b and 567 arranged on the two conductive separators are pressed against each other. A cooling water unit having a cooling water channel 571 is formed between the conductive separator plates, and a part of the fuel gas manifold 554 is formed. In the figure, gas seal members 557b and 567 are shown for sealing so as to surround a part of the fuel gas manifold 554.

In FIG. 5, as described above, a structure in which a two-cell stacked battery stack is sandwiched between two cooling water units is sandwiched between two current collector plates 581a and 581b that are gold-plated on the surface of copper. The two insulating plates 582a and 582b are sandwiched between two end plates 583a and 583b made of SUS stainless steel. The collector plates 581a and 581b, the insulating plates 582a and 582b, and the end plate 583a are formed with manifold holes as part of the fuel gas manifold, as shown in the drawing.
The fuel gas is introduced from the fuel gas introduction pipe 554a welded to the end plate 583a, and injected into the fuel gas manifold 554 in which the fuel gas manifold portions formed as described above are connected. It passes through the gas flow path and is discharged to the outside from the fuel gas discharge pipe 554b welded to the end plate 583b.

  Although the description is omitted, as can be seen from the description centering on the upper half of FIG. 5, the lower half of FIG. 5 has a configuration similar to that of the upper half of FIG. That is, the oxidant gas is introduced from the oxidant gas introduction pipe 553a welded to the end plate 583a, and the oxidant gas manifold formed in the same manner as the fuel gas manifold part is connected to the oxidant gas manifold. It is injected, passes through an oxidant gas flow path of each conductive separator plate, and is discharged to the outside from an oxidant gas discharge pipe 553b welded to the end plate 583b.

  The laminated structure as described above is fixed by providing fastening members at the four corners of the outer surface of the end plate. That is, the above-mentioned battery is formed by providing a hole for a fastening rod in each corner of the end plate, passing the fastening rod through each hole, using a bolt and a spring (and a washer if necessary) at one end of each fastening rod, and a nut at the other end. Fix the stack by applying fastening pressure. In FIG. 5, two of these four fastening members are visible. These are bolts 592a, 595a, springs 593, 596, fastening rods 591, 594, and nuts 592b, 595b. In FIG. 5, the centers of the visible fastening members, that is, the centers of the fastening rods 591, 594 are configured to be on the cross section of the upper half of FIG. 5. In this way, a two-cell laminated polymer electrolyte fuel cell based on this example was constructed.

  By the way, in the Example group including the present Example 1, it describes about the fastening pressure per electrode area. The fastening pressure can be measured using a pressure sensor, but can also be calculated as follows. That is, the fastening load applied to the electrode portion is calculated by subtracting the fastening load of the gasket calculated from the reaction force of the gasket (gas seal member) from the total fastening load of the stacked battery stack, and this calculation is further performed. The fastening pressure applied to the electrode part is calculated by dividing the fastening load by the area of the electrode part.

  Specifically, the total fastening load of the battery stack is calculated by multiplying the spring constant by the spring shrinkage when, for example, a constant load is applied using a coil spring or a flat spring. it can. Next, regarding the calculation of the reaction force of the gasket, the tightening allowance of the gasket, that is, the shrinkage allowance, is designed so that the conductive separator plate and the electrode come into contact with each other when the gasket shrinks with a certain tightening allowance. Has been. Based on this design parameter, it is possible to calculate the gasket reaction force when the gasket shrinks with a certain tightening allowance. Alternatively, by subtracting the length between the conductive separator plates at both ends after actually fastening the battery stack from the free length between the conductive separator plates at both ends when stacked as a battery stack, Is calculated. By dividing the total length by the number of gasket layers stacked, the gasket allowance, that is, the shrinkage allowance per gasket layer is calculated.

  Furthermore, regarding the calculation of the fastening load on the electrode, the fastening load on the electrode surface can be calculated by subtracting the gasket reaction force from the total fastening load of the battery stack. Finally, the fastening load per electrode area, that is, the fastening pressure, is calculated by dividing the fastening load on the electrode surface by the electrode area. Further, by dividing the fastening load on the electrode surface by the electrode area in contact with the rib portion of the conductive separator plate, the fastening load per area of the portion where the conductive separator plate and the electrode are in contact, that is, the fastening pressure is calculated. .

A polymer electrolyte fuel cell having a two-cell stack battery stack shown in FIG. 5 was prepared, and the following experiment was conducted as Example 1.
That is, for the prepared fuel cell was varied an engagement pressure between the electrode area per ^{1kgf / cm 2 ~10kgf / cm 2} . Specifically, the power generation characteristics, short-circuit conductivity, and hydrogen leakage current are measured for each predetermined fastening pressure, and then the same measurement is performed by increasing the fastening pressure from the low fastening pressure. It was repeated by gradually changing in the direction of high fastening pressure.

First, in order to measure the power generation characteristics, the produced polymer electrolyte fuel cell of Example 1 was held at 70 ° C. and humidified and heated to a dew point of 68.5 ° C. on one electrode side. A mixed gas of% hydrogen, 20% carbon dioxide gas, and 10 ppm carbon monoxide was supplied, and air that had been humidified and heated to a dew point of 70 ° C. was supplied to the other electrode side. The open circuit voltage at this time and the battery voltage during power generation at a constant current density of 80% fuel utilization, 40% oxygen utilization, and a current density of 0.3 A / cm ² were measured. FIG. 9 shows the result converted to an average value per unit battery. In FIG. 9, V0 is a characteristic curve of open circuit voltage, and V1 is a characteristic curve of battery voltage during power generation at a constant current density of 0.3 A / cm ² current density.

As a result, it has been found that the open circuit voltage is not greatly affected by the fastening pressure, and gradually decreases at a fastening pressure of about 6 kgf / cm ² or more. However, the battery voltage during power generation significantly affected by the engagement pressure, shows an extremely low value at low engagement pressure than 2kgf / cm ^2, 2kgf / cm voltage gradually in ² or more is increased, 6 kgf / cm ² approximately As a result, it was found that it was almost saturated. From this result, it was found that the fastening pressure needs to be ² kgf / cm ² or more per electrode area.

  Next, in order to measure the short-circuit conductivity, the prepared polymer electrolyte fuel cell of Example 1 is held at 70 ° C., and humidified and heated so that a dew point of 68.5 ° C. is formed on one electrode side. Nitrogen that was humidified and heated so that the dew point of 70 ° C. was also supplied to the other electrode side. After the gas is sufficiently replaced, a DC voltage of 0.2V (0.4V in total) is applied per unit cell, and a steady-state current value is measured 3 minutes after the DC voltage is applied. The short circuit conductivity was calculated by converting by the above method.

  Further, while maintaining the same fastening pressure as the fastening pressure measured for the short-circuit conductivity, hydrogen which has been humidified and heated so as to have a dew point of 68.5 ° C. on one electrode side is also 70 on the other electrode side. Nitrogen that has been humidified and heated to a dew point of ° C. is supplied, and after the gas has been sufficiently replaced, a DC voltage of 0.2 V (0.4 V in total) is applied per unit cell, and after applying the DC voltage, 3 By measuring the steady-state current value at the time of minutes and subtracting the short-circuit conductivity obtained earlier from the conductivity obtained by converting this current value by the above-mentioned method, and further converting this by the above-mentioned method. Hydrogen leak current was calculated.

The result is shown in FIG. In FIG. 10, Cs shows a characteristic curve of short circuit conductivity, and Ih shows a characteristic curve of hydrogen leakage current. From this result, it was found that when a fastening pressure exceeding 4 kgf / cm ² per electrode area was applied, the short-circuit conductivity started to increase, and a micro short circuit occurred at a fastening pressure exceeding 4 kgf / cm ² . However, it has been found that the hydrogen leakage current does not depend on the fastening pressure within the range of this embodiment and is substantially constant. Furthermore, from this result, it was discovered that the fastening pressure needs to be 4 kgf / cm ² or less per electrode area in order to suppress a minute short circuit.

  In the above experiment, the current was measured at 3 minutes after the DC voltage was applied. The time point of 3 minutes was selected, as shown in the experimental results shown in FIGS. 13 and 14 to be described later, when about 1 minute has passed after the DC voltage application, the current becomes a steady current value. This is because the current value was almost steady at 3 minutes.

Next, six fuel cells of the same two-cell stack as prepared above are prepared, one of which is applied with a fastening pressure of 1 kgf / cm ² , and the remaining five are 2 kgf / cm ² to in the range of 10 kgf / cm ^2, it was varied engagement pressure for each 2 kgf / cm ^2. The power generation characteristics of these fuel cells were measured under the same conditions as described above. That is, each fuel cell is maintained at 70 ° C., and a mixed gas of 80% hydrogen, 20% carbon dioxide gas, and 10 ppm carbon monoxide, which is humidified and heated so that a dew point of 68.5 ° C. is supplied to one electrode side. Then, air humidified and heated so as to have a dew point of 70 ° C. was supplied to the other electrode side. The open circuit voltage at this time and the battery voltage during power generation (durability test) at a constant current density of 80% fuel utilization, 40% oxygen utilization, and a current density of 0.3 A / cm ² were measured.

As a result, the performance of the fuel cell having a fastening pressure of 1 kgf / cm ² was extremely poor from the initial stage. Next, the fuel cells with fastening pressures of ² kgf / cm ² and 4 kgf / cm ² have good durability with an average deterioration rate of 1 mV / 1000 hours and 1.5 mV / 1000 hours, respectively, for 10,000 hours of continuous power generation time. The characteristics are shown. In addition, short-circuit conductivity and hydrogen leakage current were measured after 10,000 hours, but there was no change from the initial state. However, in the case of a fastening pressure of 6 kgf / cm ² , the deterioration was an average deterioration rate of 7 mV / 1000 hours up to 5000 hours, but after 6000 hours, the performance deteriorated rapidly and continuous power generation became impossible.

Further, in the case of the fastening pressure of 8 kgf / cm ^2, the average degradation was 16 mV / 1000 hours until 2000 hours later, but the performance was lowered to a voltage where continuous power generation was impossible after 3000 hours. Further, in the case of a fastening pressure of 10 kgf / cm ² , the average deterioration rate was 120 mV / 1000 hours up to 800 hours, but after 1000 hours, the performance suddenly decreased to a voltage where continuous power generation was impossible. Further, when the short-circuit conductivity and the hydrogen leakage current after the completion of the fuel cell test at the fastening pressure of 6 kgf / cm ^{2 to} 10 kgf / cm ² were measured, it was found that both increased by about two digits.

Example 2
A polymer electrolyte fuel cell having a two-cell stacked battery stack having the configuration shown in FIG. 5 was manufactured as a fuel cell in Example 2 by using the same method as that described in Example 1, except that a part thereof was different. . The difference was that instead of the carbon paper used in Example 1, a carbon cloth (manufactured by Nippon Carbon Co., Ltd.) having the same outer dimensions and thickness of 300 μm was used, and a conductive separator plate. The dimensions of the groove portions (11a, 11b) are 0.8 mm in width and 1.0 mm in depth. The following experiment was performed using the fuel cell of Example 2 manufactured in this way.

That is, the fuel cell of the second embodiment, the fastening pressure was varied between 2kgf / cm ² ~20kgf / cm ² per area of the portion where the conductive separator plate and the electrode is in contact. Specifically, the power generation characteristics, short-circuit conductivity, and hydrogen leakage current are measured for each predetermined fastening pressure, and then the same measurement is performed by increasing the fastening pressure from the low fastening pressure. It was repeated by gradually changing in the direction of high fastening pressure.

First, in order to measure the power generation characteristics, the prepared polymer electrolyte fuel cell of Example 2 was held at 70 ° C., and humidified and heated to a dew point of 68.5 ° C. on one electrode side. A mixed gas of% hydrogen, 20% carbon dioxide gas, and 10 ppm carbon monoxide was supplied, and air that had been humidified and heated to a dew point of 70 ° C. was supplied to the other electrode side. The battery voltage at the time of power generation was measured with the open circuit voltage at this time, a fuel utilization rate of 80%, an oxygen utilization rate of 40%, and a constant current density of 0.3 A / cm ² . FIG. 11 shows the result converted to an average value per unit battery. In FIG. 11, V0 is a characteristic curve of the open circuit voltage, and V1 is a characteristic curve of the battery voltage during power generation at a constant current density of 0.3 A / cm ² .

As a result, it has been found that the open circuit voltage is not greatly affected by the fastening pressure, and gradually decreases at a fastening pressure of about 12 kgf / cm ² or more per area where the conductive separator plate and the electrode are in contact. However, the battery voltage during power generation is greatly affected by the fastening pressure, and shows an extremely low value at a fastening pressure lower than 4 kgf / cm ² per area where the conductive separator plate and the electrode are in contact, and gradually increases at 4 kgf / cm ² or more. It has been found that the voltage rises and is almost saturated at about 12 kgf / cm ² or more. From this result, it has been found that the fastening pressure is required to be 4 kgf / cm ² or more per area of the portion where the conductive separator plate and the electrode are in contact.

  Next, in order to measure short-circuit conductivity, the prepared polymer electrolyte fuel cell of Example 2 is held at 70 ° C., and humidified and heated so that a dew point of 68.5 ° C. is provided on one electrode side. Nitrogen that was humidified and heated so that the dew point of 70 ° C. was also supplied to the other electrode side. After the gas is sufficiently replaced, a DC voltage of 0.2V (0.4V in total) is applied per unit cell, a steady current value is measured at 3 minutes after the DC voltage is applied, and this current value is described above. The short circuit conductivity was calculated by converting by the above method.

  Further, while maintaining the same fastening pressure as the fastening pressure measured for the short-circuit conductivity, hydrogen which has been humidified and heated so as to have a dew point of 68.5 ° C. on one electrode side is also 70 on the other electrode side. Nitrogen that has been humidified and heated to a dew point of ° C. is supplied, and after the gas has been sufficiently replaced, a DC voltage of 0.2 V (0.4 V in total) is applied per unit cell, and after applying the DC voltage, 3 Measure the steady-state current value at the time of the minute, subtract the short-circuit conductivity obtained earlier from the conductivity obtained by converting this current value by the above method, and further convert this to obtain the hydrogen leakage current. Calculated.

The result is shown in FIG. In FIG. 12, Cs shows a characteristic curve of short circuit conductivity, and Ih shows a characteristic curve of hydrogen leakage current. From this result, when a fastening pressure exceeding 8 kgf / cm ² per area where the conductive separator plate and the electrode are in contact is applied, the short circuit conductivity starts to increase, and a micro short circuit occurs at a fastening pressure exceeding 8 kgf / cm ^2. Turned out to be. However, it has been found that the hydrogen leakage current does not depend on the fastening pressure in the range of the present Example 2, and is almost constant. Furthermore, from this result, it was discovered that the fastening pressure must be 8 kgf / cm ² or less per area of the contact portion between the conductive separator plate and the electrode in order to suppress a slight short circuit.

Next, six fuel cells of the same two-cell stack as prepared above are prepared, and one of them is applied with a fastening pressure of ² kgf / cm ² per area where the conductive separator plate and the electrode are in contact, the five remaining in the range of ^{4kgf / cm 2 ~20kgf / cm 2} , was varied engagement pressure for each 4 kgf / cm ^2. The power generation characteristics of these fuel cells were measured under the same conditions as described above. That is, each fuel cell is maintained at 70 ° C., and a mixed gas of 80% hydrogen, 20% carbon dioxide gas, and 10 ppm carbon monoxide, which is humidified and heated so that a dew point of 68.5 ° C. is supplied to one electrode side. Then, air humidified and heated so as to have a dew point of 70 ° C. was supplied to the other electrode side. The open circuit voltage at this time and the battery voltage during power generation (durability test) at a constant current density of 80% fuel utilization, 40% oxygen utilization, and a current density of 0.3 A / cm ² were measured.

As a result, first, the performance of the fuel cell having a fastening pressure of ² kgf / cm ² was extremely poor from the initial stage. Next, the fuel cells with a fastening pressure of 4 kgf / cm ² and 8 kgf / cm ² have good durability with an average deterioration rate of 1 mV / 1000 hours and 1.5 mV / 1000 hours, respectively, for 10,000 hours of continuous power generation time. The characteristics are shown. In addition, short-circuit conductivity and hydrogen leakage current were measured after 10,000 hours, but there was no change from the initial state.

However, in the case of the fastening pressure of 12 kgf / cm ^2, the average deterioration rate was 7 mV / 1000 hours until 5000 hours. However, after 6000 hours, the performance deteriorated rapidly and continuous power generation became impossible. Further, in the case of a fastening pressure of 16 kgf / cm ^2, the average degradation was 16 mV / 1000 hours until 2000 hours later, but after 3000 hours, the performance was lowered to a voltage where continuous power generation was impossible. Furthermore, in the case of a fastening pressure of 20 kgf / cm ² , the average deterioration rate was 120 mV / 1000 hours up to 800 hours, but after 1000 hours, the performance suddenly decreased to a voltage where continuous power generation was impossible. The measured short conductivity and hydrogen leak current of 12kgf / cm ² ~20kgf / cm ² after the end of fuel cell test of the engagement pressure, it was found that one is also increased by about 2 orders of magnitude.

Example 3
When measuring only the conductivity due to the DC resistance component, that is, the short-circuit conductivity, in other words, when not measuring the hydrogen leakage current or the power generation characteristics, as a simple measurement method, Measurement of short-circuit conductivity is possible. In Example 3, an example of the measurement method will be described.
First, in the same manner as in the production of the fuel cell in Example 2, a carbon cloth was used, and a groove was measured using a conductive separator plate having a width of 0.8 mm and a depth of 1.0 mm. Thus, a unit cell of Example 3 was produced.

  First, as shown by 502 in FIG. 5, a conductive surface having a structure as shown in FIG. 3A, that is, a cooling water flow path, etc. formed on the front surface and a structure as shown in FIG. A separator plate was prepared. Next, a configuration as shown in FIG. 2A, that is, an oxidant gas flow path or the like is formed on the front surface, which is indicated by reference numeral 504 in FIG. 5, and a configuration as shown in FIG. A conductive separator plate was prepared. An MEA composed of a polymer electrolyte membrane indicated by 541 in FIG. 4 or FIG. 5 and electrodes 540a and 540b disposed on both sides thereof was sandwiched between the two conductive separator plates. In this way, a single battery in Example 3 was produced.

  On both end faces of this unit cell, that is, on the surface of each of the two conductive separator plates on the side of the cooling water flow path, two rigid insulating plates with both main surfaces being smooth are disposed, and this insulating plate and the unit cell are laminated. The body was sandwiched between the two pressure surfaces of a press having two smooth and opposed pressure surfaces so that the two pressure surfaces were in contact with the respective insulating plates of the laminate. Further, two electric leads for applying a voltage between the two conductive separator plates of the unit cell were connected to each conductive separator plate.

With respect to the unit cell of this example configured as described above, a pressure of 6 kgf / cm ² per area of the portion where the conductive separator plate and the electrode are in contact with each other by the press machine, that is, a fastening pressure, was applied to the unit cell with DC 0.2 V A constant voltage was applied. FIG. 13 shows changes with time in current density and short-circuit conductivity when short-circuit conductivity is measured in air at room temperature. In FIG. 13, Id is a current density characteristic curve, and Cs is a short circuit conductivity characteristic curve.

  In order to measure the short-circuit conductivity, it is necessary to separate the ionic conductivity of the polymer electrolyte membrane from the DC resistance component of electronic conduction and extract only the DC resistance component based on electronic conduction. The separation can be performed by applying a DC voltage or a DC current to the polymer electrolyte membrane for a sufficient time. The reason is that the ionic conductor is polarized by the application of the DC voltage or DC current, so that ideally infinite time, practically enough time, DC voltage or DC current is applied for ionic conduction. This is because the resulting current or voltage is so small that it can be ignored, and as a result, the DC resistance component resulting from ionic conduction is so small that it can be ignored. FIG. 13 shows such a phenomenon.

  Therefore, when a constant voltage is applied, if the voltage is divided by the current value measured after a sufficient time from the voltage application, the resistance value caused by only the direct current resistance component can be calculated, and the reciprocal thereof. The short-circuit conductivity can be calculated by taking At this time, in order to reduce the influence of the ionic conductivity of the polymer electrolyte membrane, it is desirable to perform measurement under conditions where the polymer electrolyte membrane is less likely to exhibit ionic conductivity, that is, in a dry state. Therefore, in Example 3, the measurement was performed in a constant temperature and humidity chamber, and the humidity was kept at a low humidity of 24%.

  For the unit cell of this example, the same measurement was repeated 20 times, and when the current value and the short-circuit conductivity when DC 0.2V was applied to the unit cell were measured, the same results were obtained with good reproducibility. Obtained. Since the decay curve of the current value and the short-circuit conductivity with respect to time decreases logarithmically, the result after a long time passes shows the original short-circuit conductivity. However, from the viewpoint of efficient measurement, when the reproducibility of the short-circuit conductivity calculated from the slope of the current value decay curve and the current value at that time is examined, the slope of the current value decay curve is about -0.0025. It was found that when the short-circuit conductivity was calculated from the current value after attenuation to the following, a reproducible result was obtained. In Example 3, the slope of the current value attenuation curve was about −0.0025 or less after about 1 minute from the voltage application.

Example 4
First, 11 fuel cells having the same two-cell stack as prepared in Example 2 were prepared. For these 11 fuel cells, a fastening load is equally applied between 100 kg and 1 ton to the four fastening parts of each fuel cell, and the total load (absolute value) is varied between 400 kg and 4 ton. As a result of measuring the short-circuit conductivity and the hydrogen leakage current for each predetermined fastening pressure, the short-circuit conductivity measured by the same method as in Example 2 is 0.2 mS / cm ² , 0.4 mS / cm ² , 0 ^{.6mS / cm 2, 0.8mS / cm} 2, 1.0mS / cm 2, 1.2mS / cm 2, 1.4mS / cm 2, 1.5mS / cm 2, 1.6mS / cm 2, 1. The fastening pressure was adjusted to 8 mS / cm ² and 2.0 mS / cm ² . These 11 fuel cells were subjected to a durability test with a constant current density of 80% fuel utilization, 40% oxygen utilization, and 0.3 A / cm ^{2 under} the same measurement conditions as described in Example 2. .

As a result, the fuel cell having an initial short-circuit conductivity of 0.2 mS / cm ^{2 to} 0.8 mS / cm ² has a good average deterioration rate of about 1 mV / 1000 hours with respect to 10,000 hours of continuous power generation time. The short-circuit conductivity after 10,000 hours and the hydrogen leakage current were unchanged. The fuel cell having an initial short-circuit conductivity of 1.0 mS / cm ² exhibited good durability characteristics of 1.5 mV / 1000 hours for 10,000 hours of continuous power generation. In addition, short-circuit conductivity and hydrogen leakage current were measured after 10,000 hours, but there was no change from the initial state. The fuel cell initial short circuit conductivity of 1.2 mS / cm ^2, was 1.4 mS / cm ² and 1.5 mS / cm ^2, compared continuous power generation of 10,000 hours, 2.5 mV / 1000 hours First of all, it showed good durability characteristics. Further, when the short circuit conductivity and the hydrogen leakage current were measured after 10,000 hours, the hydrogen leakage current did not change at all from the initial value, but the short circuit conductivity increased by about 20%.

However, the fuel cell with an initial short-circuit conductivity of 1.6 mS / cm ² had an average deterioration rate of 7 mV / 1000 hours until 5000 hours, but after 6000 hours the performance deteriorated rapidly and continuous power generation was impossible. Reached. In addition, the fuel cell with an initial short-circuit conductivity of 1.8 mS / cm ² was deteriorated by an average of 16 mV / 1000 hours until 2000 hours passed, but after 3000 hours, the performance was lowered to a voltage where continuous power generation was impossible. Further, the fuel cell with an initial short-circuit conductivity of 2.0 mS / cm ² had an average deterioration rate of 120 mV / 1000 hours up to 800 hours, but after 1000 hours, the performance suddenly decreased to a voltage that could not be continuously generated. . Moreover, when the short-circuit conductivity and the hydrogen leakage current after the fuel cell test of the fuel cell having an initial short-circuit conductivity of 1.6 mS / cm ^{2 to} 2.0 mS / cm ² were measured, both increased by about two digits. Turned out to be.

Example 5
First, nine fuel cells having the same two-cell stack as prepared in Example 2 were prepared. For these nine fuel cells, the fastening load is equally applied to the four fastening parts of each fuel cell between 100 kg and 1 ton, and the total load (absolute value) is changed between 400 kg and 4 ton. from the result of measuring the short-circuit conductivity and hydrogen leak current in a predetermined fastening圧毎hydrogen leakage current measured in the same manner as in example 2, 1.2 mA / cm ^2, 1.6 mA / cm ^2, 2 0.0 mA / cm ² , 2.4 mA / cm ² , 2.8 mA / cm ² , 3.0 mA / cm ² , 3.2 mA / cm ² , 3.6 mA / cm ² , 4.0 mA / cm ² The fastening pressure was adjusted. These nine fuel cells were subjected to an endurance test at a constant current density of 80% fuel utilization, 40% oxygen utilization, and 0.3 A / cm ² .

As a result, the fuel cell having an initial hydrogen leakage current of 1.2 mA / cm ² to 2.0 mA / cm ² has a good average deterioration rate of about 1 mV / 1000 hours with respect to 10,000 hours of continuous power generation time. The short-circuit conductivity after 10,000 hours and the hydrogen leakage current were unchanged. The fuel cell having an initial hydrogen leakage current of 2.4 mA / cm ² exhibited good durability characteristics of 1.5 mV / 1000 hours with respect to 10,000 hours of continuous power generation. In addition, short-circuit conductivity and hydrogen leakage current were measured after 10,000 hours, but there was no change from the initial state. In addition, a fuel cell having an initial hydrogen leakage current of 2.8 mA / cm ² and 3.0 mA / cm ² has a good durability characteristic of 2.5 mV / 1000 hours for 10,000 hours of continuous power generation. Indicated. Further, when the short circuit conductivity and the hydrogen leakage current were measured after 10,000 hours, the hydrogen leakage current did not change at all from the initial value, but the short circuit conductivity increased by about 20%.

However, the fuel cell with an initial hydrogen leakage current of 3.2 mA / cm ² had an average deterioration rate of 7 mV / 1000 hours until 5000 hours, but after 6000 hours the performance deteriorated rapidly and continuous power generation was impossible. Reached. Further, the fuel cell with an initial hydrogen leakage current of 3.6 mA / cm ² deteriorated by an average of 16 mV / 1000 hours until 2000 hours passed, but after 3000 hours, the performance decreased to a voltage where continuous power generation was impossible. Further, the fuel cell with an initial hydrogen leakage current of 4.0 mA / cm ² had an average deterioration rate of 120 mV / 1000 hours up to 800 hours, but after 1000 hours, the performance suddenly decreased to a voltage that could not be continuously generated. . Moreover, when the short-circuit conductivity and the hydrogen leak current after the fuel cell test of the fuel cell having an initial hydrogen leak current of 3.2 mA / cm ^{2 to} 4.0 mA / cm ² were measured, both increased by about two digits. Turned out to be.

  In addition to the measurement method used in this example, the hydrogen leak current can also be measured using cyclic voltammetry (CV). Specifically, for example, hydrogen that has been humidified and heated to a dew point of 68.5 ° C. is supplied to the fuel electrode side of the MEA, and humidified and heated to a dew point of 70 ° C. to the oxidant electrode side. Nitrogen is supplied, whereby the atmosphere gas in each electrode is sufficiently replaced by hydrogen and nitrogen, and then the fuel electrode supplied with hydrogen is used as a reference electrode, and the oxidant electrode potential supplied with nitrogen is set. This is a measurement method in which CV measurement is performed by sweeping from an equilibrium potential to +1.0 V by a single sweep at a sweep speed of 50 mV / s. The hydrogen leak current value per unit area of the electrode can be obtained by dividing the leak current value thus measured by the electrode area.

Example 6
A plurality of two-cell stacked fuel cells were produced in the same manner as in Example 2, except that each of the conductive separator plates in which the groove width and depth of each gas flow path were changed was used.
That is, corresponding to the conductive separator plate group used in Example 2, the gas channel rib width is fixed to 1.0 mm, the gas channel groove width is changed between 0.6 mm and 2.0 mm, By changing the groove depth between 0.2 mm and 1.1 mm as shown in Table 1 below, there are eight kinds of combinations of the groove width and the groove depth, and which conductive separator Eight types of conductive separator plate groups were produced in which the flow groove depth and the number of flow grooves were changed according to the flow groove width so that the gas flow rates flowing through the gas flow grooves were substantially the same.
The combinations of gas channel grooves are shown in the table below. Using these eight types of conductive separator plate groups, eight fuel cells were produced in the same manner as in Example 2, and a test was performed using these eight fuel cells.

At this time, the fastening pressure was varied between 2kgf / cm ² ~20kgf / cm ² per area of the portion where the conductive separator plate and the electrode is in contact. Specifically, for each predetermined fastening pressure, the power generation characteristics, short-circuit conductivity, and hydrogen leakage current are measured in the same manner as in Example 2, and then the fastening pressure is increased and the same measurement is performed. This operation was repeated by gradually changing the fastening pressure from the low fastening pressure to the high fastening pressure.

As a result, when a fastening pressure exceeding 8 kgf / cm ² per area where the conductive separator plate and the electrode are in contact is applied depending on the channel groove width, the short-circuit conductivity starts to increase, and the fastening pressure exceeding 8 kgf / cm ^2. I discovered that a short circuit occurred. However, it has been found that the hydrogen leak current does not depend on the fastening pressure in the range of the present embodiment, and is substantially the same hydrogen leak current. From this result, it was discovered that the fastening pressure needs to be 8 kgf / cm ² or less per area where the conductive separator plate and the electrode are in contact with each other in order to suppress a slight short circuit.

Next, eight similarly prepared two-cell stacked fuel cells were prepared, and fastened with a load of 6 kgf / cm ² per area of the portion where the separator plate and the electrode were in contact. With these fuel cells, the power generation characteristics of these fuel cells were measured under the same conditions as in Example 2 under a constant current density of 80% fuel utilization, 40% oxygen utilization, and 0.3 A / cm ^2. The initial characteristics were measured.
As a result, in all the cells of cell numbers 1 to 8, the open circuit voltage was 0.980 V to 0.985 V, indicating a good result.

However, while cell numbers 3 to 5 showed a voltage of 0.73 V to 0.74 V at a current density of 0.3 A / cm ² , cell numbers 1, 2, 6, 7, and 8 A voltage of 0.70 V or less was exhibited.
In cell numbers 1 and 2, since the gas flow channel groove width was too small compared to the rib width, the total area of the electrode portion corresponding to the gas flow channel groove was small, and sufficient performance could not be obtained. In cell numbers 7 and 8, since the contact area between the gas flow path rib and the electrode was small, the contact resistance was large and sufficient performance could not be obtained. In cell number 6, since the gas channel groove depth was too shallow, gas supply to the gas channel groove was hindered, and sufficient performance could not be obtained.

Example 7
100 electrolyte membrane electrode assemblies (MEA) were produced in the same manner as in Example 2. The short-circuit conductivity of one piece of this MEA was measured by the following method. That is, as shown in FIG. 8, MEA 85 is sandwiched between two current collector plates 80 each plated with gold on a copper plate having a thickness of 5 mm, and no particular fastening pressure is applied. The MEA was held with a load that was only the weight of one plate. This sandwich was placed in a constant temperature and humidity chamber at 25 ° C. and humidity 30%, DC 0.2V was applied between the current collector plates, the change in current value over time was measured, and the short-circuit conductivity was calculated from the current value. Calculated. The result is shown in FIG. In FIG. 14, Id is a current density characteristic curve, and Cs is a short circuit conductivity characteristic curve. At this time, since the slope of the attenuation curve of the current value (current density) was −0.0022 after 1 minute, the short-circuit conductivity was calculated from the current value after 1 minute had elapsed after the voltage application.

Next, 100 unit cells were produced in the same manner as in Example 3 using the 100 MEAs described above. That is, each unit cell has a configuration in which each MEA is sandwiched between a conductive separator plate indicated by 502 and a conductive separator plate indicated by 504 in FIG.
For each of these 100 single cells, the short-circuit conductivity was measured in the same manner as in Example 3, using the same two insulating plates and press as those used in Example 3, and two electrical leads. .

As a result, the short-circuit conductivity of each unit cell, that is, each MEA, is the short-circuit conductivity of 70 sheets of MEA of 0.5 mS / cm ² or less, and 15 sheets of more than 0.5 to 1.0 mS. For 10 sheets / cm ² or less, 10 sheets exceeded 1.0 to 1.5 mS / cm ² or less, and 5 sheets exceeded 1.5 mS / cm ² . By the method described below, a battery stack was manufactured in which 100 ME cells were stacked such that these MEAs were stacked in ascending order of short-circuit conductivity, and a cooling water unit was arranged for each cell. Here, a cell using MEA having the smallest short-circuit conductivity is No. No. 1 cell, a cell using MEA having the highest short-circuit conductivity. 100 cells were used.

  The manufacturing method of the battery stack of the 100 cell stack is as follows. First, a conductive separator plate having a smooth surface on the front surface (left side in the cross-sectional view) indicated by 501 in FIG. 5 and a cooling water channel on the back surface was prepared as a leftmost conductive separator plate. The left side of the leftmost conductive separator plate, It arrange | positioned so that the surface of one electroconductive separator board of 1 cell, ie, the surface which has a cooling water flow path of the structure of FIG. 3A may contact | connect. Next, no. The surface of the other conductive separator plate of one cell, that is, the surface having the cooling water flow path having the configuration of FIG. It arrange | positioned so that the surface of one electroconductive separator plate of 2 cells, ie, the surface which has a cooling water flow path of the structure of FIG. 3A may contact | connect. Such an arrangement is referred to as “No. 3 cells to No. The same was repeated up to 100 cells.

  Finally, a conductive separator plate indicated by 505 in FIG. 5 having a smooth back surface (right side in the sectional view) and a cooling water channel on the surface was prepared as a right-end conductive separator plate. And No. The surface of the right-side conductive separator plate was joined to the surface of the conductive separator plate outside the 100 cells, that is, the surface having the cooling water flow path having the configuration shown in FIG. 2B. In addition, as shown by 527, 537, 547 and 557 in FIG. 5, adjacent O-ring gas seal members are arranged on adjacent conductive separator plates, and these O-ring gasses are arranged. The seal members are configured to be pressed against each other directly or indirectly via the MEA, thereby sealing the oxidant gas, the fuel gas, and the cooling water. In this way, a battery stack having a stack of 100 cells in which both ends were sandwiched between cooling water units and the cooling water unit was arranged for each cell was produced.

Two current collector plates were arranged on both end faces of the battery stack thus produced, and two insulating plates were further arranged thereon to produce a fuel cell test unit. The fuel cell test unit was sandwiched between the two pressure surfaces of a press having two pressure surfaces so that the two pressure surfaces were in contact with the respective insulating plates. A pressure of 6 kgf / cm ² , that is, a fastening pressure was applied to the area of the battery stack where the conductive separator plate and the electrode are in contact with each other by the pressing force of the press.

The thus produced polymer electrolyte fuel cell test unit of this example was maintained at 70 ° C., and steam reforming of methane that was humidified and heated to a dew point of 68.5 ° C. on one electrode side. The gas was supplied to the other electrode side with air that was humidified and heated to a dew point of 70 ° C. As a result, 97.4V battery open voltage (average 0.974V) was obtained at no load when no current was output to the outside.
This fuel cell test unit was subjected to a continuous power generation test under the conditions of a fuel utilization rate of 80%, an oxygen utilization rate of 40%, and a current density of 0.3 A / cm ² , and the change over time in output characteristics was measured. At this time, the initial average generated voltage of the 100-cell stacked fuel cell was 0.728V.

As a result of the durability test, the initial short-circuit conductivity exceeded 1.5 mS / cm ² . 96 cells to No. With respect to 5 cells of 100 cells, the battery voltage in the power generation state began to drop sharply when about 5500 hours had elapsed after the start of power generation operation. Since the battery voltage of these 5 cells dropped to an average of 0.4 V or less in about 200 hours after the voltage began to drop, the operation of the 100-cell stacked fuel cell test unit was temporarily stopped. The open circuit voltage at this time was reduced to 96.8V (average 0.968V). Therefore, once the fastening of the fuel cell test unit is released, 96 cells to No. Remove 5 cells of 100 cells, 96-No. When 5 cells of 100 cells were disassembled and observed, it was confirmed that pores were formed in the polymer electrolyte membrane in all the cells.

The remaining 95 cells were fastened again and the durability test was continued. As a result of the continuous durability test, the initial short-circuit conductivity was 1.0 to 1.5 mS / cm ² . 86-No. With respect to 10 cells of 95 cells, the battery voltage in the power generation state suddenly decreased when about 18000 hours had elapsed after the start of the power generation operation. No. 86-No. The battery voltage of 10 cells of 95 cells decreased to an average of 0.4 V or less in about 200 hours after the voltage began to drop, so the operation of the fuel cell test unit of the 95 cell stack was stopped. The open circuit voltage at this time was reduced to 91.4V (average 0.962V). Therefore, once the fastening of the fuel cell is released, the upper No. 86-No. Ten cells of 95 cells were removed. No. 86-No. When 10 cells of 95 cells were disassembled and observed, it was confirmed that pores were formed in the polymer electrolyte membrane in all the cells.

The remaining 85 cells were fastened again and the durability test was continued. As a result of the continuous durability test, the initial short-circuit conductivity was 0.5 to 1.0 mS / cm ² . 71-No. For 15 cells of 85 cells, the battery voltage in the power generation state suddenly decreased when about 23,000 hours had elapsed after the start of the power generation operation. No. 71-No. Since the battery voltage of 15 cells of 85 cells dropped to an average of 0.4 V or less in about 200 hours after the voltage began to drop, the operation of the fuel cell test unit of 85 cells was stopped. The open circuit voltage at this time was lowered to 80.9V (average 0.952V). Therefore, once the battery stack is closed, the upper No. 71-No. 15 cells of 85 cells were removed. At this time, the no. 71-No. When 15 cells of 85 cells were disassembled and observed, it was confirmed that pores were formed in the polymer electrolyte membrane in all the cells.

The remaining 70 cells were fastened again and the durability test was continued. As a result of the continuous durability test, the average short-circuit conductivity of 0.5 mS / cm ² or less for all the cells was 0.7V on average even during power generation after about 30000 hours after the start of power generation operation. It was confirmed that the generated voltage was above. The open circuit voltage at this time was maintained at 67.9V (average 0.970V).

As described above, in this example, it was discovered that the life of the fuel cell can be predicted by measuring the short-circuit conductivity of the MEA before assembling the fuel cell. That is, it was confirmed that the manufacturing method or the inspection method using MEA having a short-circuit conductivity of 1.5 mS / cm ² or less is a method capable of confirming the life of the fuel cell for 10,000 hours or more. More preferably, by selecting only MEAs having a short-circuit conductivity of 0.5 mS / cm ² or less, it was confirmed that the manufacturing method or the inspection method can predict the life of the fuel cell of about 30000 hours.

Example 8
In the same manner as in Examples 1 and 2, but in this example, the carbon paper used as the gas diffusion layer substrate in Example 1 was replaced with a carbon nonwoven fabric, and each gas diffusion layer substrate was pre-smoothed. The MEA was produced after performing the chemical treatment.
As the preliminary smoothing process, the following method was used. That is, when a carbon non-woven fabric was used as the gas diffusion layer base material, the fine carbon fibers fluffed on the surface of the carbon non-woven fabric were removed by using a burner in advance.
Further, when carbon cloth was used as the gas diffusion layer base material, preliminary smoothing treatment was performed by the following three methods.

The first method is a method of performing preliminary smoothing processing by simply applying a load by pressing. In this method, a preliminary smoothing treatment was performed by applying a pressure of 30 kgf / cm ² for 5 minutes.
The second method is a method of performing a pre-smoothing process by hot pressing. In this method, a pre-smoothing treatment was performed by applying a pressure of 150 ° C. and 30 kgf / cm ² for 5 minutes. At this time, the heating temperature was examined in a range of 120 ° C. to 180 ° C., but as a result of surface observation, a remarkable difference was not observed in the range of 120 ° C. to 180 ° C.
The third method is a method of scoring fluff on the surface of the carbon cloth with a burner after hot pressing.

For each of these one type of carbon non-woven fabric and three types of carbon cloth as the gas diffusion layer base material, 25 sheets were prepared, and using them, 25 sheets were used to make a total of 100 MEAs.
Each MEA produced in this manner is sandwiched between two current collector plates plated with gold on a copper plate having a thickness of 5 mm as shown in FIG. 8, and no particular load (pressure) is applied, only the weight of the current collector plate. Each MEA was pinched with a load of. These were placed in a constant temperature and humidity chamber at 25 ° C. and a humidity of 30%, DC 0.2V was applied between the current collector plates, the change in current value over time was measured, and the short-circuit conductivity was calculated from the current value. .

When the short-circuit conductivity of the 100 MEAs was measured, all MEAs had a short-circuit conductivity of 0.5 mS / cm ² or less. Among them, all the carbon non-woven fabrics that were only treated with a burner had a short-circuit conductivity of 0.2 mS / cm ² or less. Moreover, the thing which performed only the press processing to the carbon cloth was all the short circuit conductivity of 0.5 mS / cm < ² > or less. In addition, all the carbon cloth subjected to hot press treatment had a short-circuit conductivity of 0.3 mS / cm ² or less. Further, all the carbon cloth subjected to the hot press treatment and then subjected to the treatment with a burner had a short-circuit conductivity of 0.2 mS / cm ² or less.

  Also, when providing a water-repellent layer on a carbon non-woven fabric or carbon cloth, in order to confirm whether or not the pre-smoothing process is effective, when pre-smoothing is performed before providing the water-repellent layer, Both the case of performing the pre-smoothing process after providing the water repellent layer were conducted. As a result, it was confirmed that the pre-smoothing treatment was effective regardless of whether the pre-smoothing treatment was performed before or after the water-repellent layer was provided. However, from the viewpoint of the water repellency of the water repellent layer, it was also confirmed that the pre-smoothing treatment should be performed before the water repellent layer is provided.

In the above, the effect of this invention was demonstrated using the Example group. In some of the examples in the above example group, the short-circuit conductivity associated with the micro-short-circuit is obtained by converting from a steady-state current value measured by applying a constant voltage of 0.5 V or less to the MEA. In addition, the hydrogen leakage current value was calculated, but the short-circuit conductivity and the hydrogen leakage current value determined by conversion using the steady voltage value observed when a constant current of 5 mA / cm ² or less was applied. In addition, it was confirmed that an equivalent result was obtained with good reproducibility.

Moreover, it is preferable to apply a voltage or a small current as low as possible to prevent oxidation of the electrode catalyst layer material as a constant voltage or a constant current applied to the MEA. However, if the constant voltage or constant current to be applied is too small, the steady current or steady voltage becomes small. Considering the resolution of existing measuring instruments, the constant voltage is about 0.2 V, or constant is about ² mA / cm ^2. Although it was confirmed that a highly reproducible result was obtained by applying a current, it should be within a range that can be measured with a measuring instrument with good reproducibility at a constant voltage of 0.5 V or less and a constant current of 5 mA / cm ² or less. It was also confirmed that there was no effect on MEA.

In the polymer electrolyte fuel cell of the present invention, the fastening pressure of the fuel cell is set to 2.0 to 4.0 kgf / cm ² per electrode area, or 4 per area where the conductive separator plate and the electrode are in contact with each other. 0.0 to 8.0 kgf / cm ² , or the short-circuit conductivity due to the DC resistance component between the fuel electrode and the oxidant electrode of the unit cell is set to a predetermined value or less such as 1.5 mS / cm ^2. Alternatively, by setting the hydrogen leakage current per electrode area to a predetermined value or less such as 3 mA / cm ² , the durability of the fuel cell can be greatly improved.

In the method for manufacturing or inspecting the polymer electrolyte fuel cell of the present invention, the initial short circuit conductivity exceeds a predetermined value such as 1.5 mS / cm ² , or the initial hydrogen leakage current per electrode area is 3 mA. By removing the polymer electrolyte membrane, MEA, unit cell or battery stack exceeding a predetermined value such as / cm ² , a highly durable fuel cell can be efficiently produced.

It is a typical top view which shows the structure of the surface of the electroconductive separator plate used with the fuel cell of the Example of this invention. FIG. 1B is a schematic plan view showing the configuration of the back surface of the conductive separator plate of FIG. 1A. It is a typical top view which shows the structure of the surface of the other kind of electroconductive separator board used with the fuel cell of the Example of this invention. It is a typical top view which shows the structure of the back surface of the electroconductive separator board of FIG. 2A. It is a typical top view which shows the structure of the surface of another kind of electroconductive separator plate used with the fuel cell of the Example of this invention. FIG. 3B is a schematic plan view showing the configuration of the back surface of the conductive separator plate of FIG. 3A. It is a typical top view which shows the structure of the surface of MEA produced with the fuel cell of the Example of this invention. It is a typical partial front sectional view showing the composition of the fuel cell of the 2 cell lamination produced in the example of the present invention. It is typical sectional drawing which shows the structure of the cross section of a conceptual unit battery. It is typical sectional drawing which shows the structure of the cross section of a realistic unit battery. It is typical sectional drawing which shows the structure which clamped MEA produced in the Example of this invention with the two current collecting plates. It is a characteristic view which shows the relationship between the fastening pressure per electrode area of the fuel cell produced in the Example of this invention, and battery voltage. It is a characteristic view which shows the relationship between the fastening pressure per electrode area, a short circuit conductivity, and a hydrogen leak current in the fuel cell produced in the Example of this invention. It is a characteristic view which shows the relationship between the fastening pressure per area which an electroconductive separator plate and an electrode contact | abut, and the battery voltage in the fuel cell produced in the Example of this invention. It is a characteristic view which shows the relationship between the fastening pressure per area where an electroconductive separator plate and an electrode contact | connect in the fuel cell produced in the Example of this invention, a short circuit conductivity, and a hydrogen leak current. It is a characteristic view which shows a time-dependent change of a current density and a short circuit conductivity in the fuel cell produced in the Example of this invention. FIG. 9 is a characteristic diagram showing changes with time in current density and short-circuit conductivity in a configuration in which the MEA is sandwiched between two current collector plates shown in FIG. 8.

Explanation of symbols

11a, 11b, 21, 31 Groove parts 12a, 12b, 22, 32 Rib parts 13a, 13b, 23a, 23b Manifold holes for oxidizing gas 14a, 14b, 24a, 24b Manifold holes for fuel gas 15a, 15b, 25a, 25b Cooling Water manifold holes 33a, 33b, 43a, 43b Oxidant gas manifold holes 34a, 34b, 44a, 44b Fuel gas manifold holes 35a, 35b, 45a, 45b Cooling water manifold holes 501, 502, 503, 504, 505 Conductive separator plate 71,541,551 Polymer electrolyte membrane 540a, 540b, 550a, 550b Electrode 72 Gas diffusion layer 73 Catalyst layer 78 Slightly short-circuited portion

Claims (16)

An electrolyte membrane electrode assembly comprising a hydrogen ion conductive polymer electrolyte membrane, and a pair of electrodes each comprising a gas diffusion layer and a catalyst layer disposed on both surfaces of the electrolyte membrane, and a fuel on one of the pair of electrodes A conductive separator plate having a gas flow path for supplying and discharging gas and contacting the one electrode, and a gas flow path for supplying and discharging oxidant gas to the other electrode and contacting the other electrode A polymer electrolyte fuel cell comprising a battery stack in which a plurality of unit cells each comprising a further conductive separator plate are stacked and fastened in the stacking direction, and the short-circuit conductivity of each electrolyte membrane electrode assembly is 1. A polymer electrolyte fuel cell characterized by being 5 mS / cm ² or less. An electrolyte membrane electrode assembly comprising a hydrogen ion conductive polymer electrolyte membrane, a pair of electrodes each comprising a gas diffusion layer and a catalyst layer, disposed on both surfaces of the electrolyte membrane, and a fuel on one of the pair of electrodes A conductive separator plate having a gas flow path for supplying and discharging gas and contacting the one electrode, and a gas flow path for supplying and discharging oxidant gas to the other electrode and contacting the other electrode A polymer electrolyte fuel cell comprising a battery stack in which a plurality of unit cells each comprising a further conductive separator plate are stacked and fastened in the stacking direction, wherein the unit cell using each electrolyte membrane electrode assembly A polymer electrolyte fuel cell having a hydrogen leakage current of 3 mA / cm ² or less per electrode area.   A process for producing a polymer electrolyte fuel cell comprising: a process for producing a plurality of unit cells; a process for producing a battery stack by laminating the plurality of unit cells; and a process for fastening the battery stack to a lamination method. In the process for producing each unit battery, a pair of electrodes each comprising a gas diffusion layer and a catalyst layer is arranged on both sides of a hydrogen ion conductive polymer electrolyte membrane to produce an electrolyte membrane electrode assembly. A step of disposing a conductive separator plate having a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes, and supplying and discharging oxidant gas to the other electrode A step of arranging a further conductive separator plate having a gas flow path in contact therewith, and the method for producing the polymer electrolyte fuel cell further includes the electrolysis A step of measuring the short-circuit conductivity of the membrane electrode assembly and / or the hydrogen leakage current of the unit cell, and when the short-circuit conductivity exceeds a predetermined value of the short-circuit conductivity and / or the hydrogen leak current exceeds a predetermined value of the hydrogen leak current And a step of removing the electrolyte membrane electrode assembly exceeding the predetermined value of the short-circuit conductivity and / or the hydrogen leak current predetermined value, or the unit cell having the electrolyte membrane electrode assembly or the battery stack having the unit cell. A process for producing a polymer electrolyte fuel cell comprising a process. The method for producing a polymer electrolyte fuel cell according to claim 3, wherein the predetermined short-circuit conductivity value is 1.5 mS / cm ² and the predetermined hydrogen leak current value is 3 mA / cm ² .   The short-circuit conductivity of the electrolyte membrane electrode assembly is converted from a steady current value or a steady voltage value obtained by applying a constant DC voltage or a constant DC current to the electrolyte membrane electrode assembly. 3. A method for producing a polymer electrolyte fuel cell according to 3.   6. The method for producing a polymer electrolyte fuel cell according to claim 5, wherein the constant voltage is 0.5 V or less per each electrolyte membrane electrode assembly. The method for producing a polymer electrolyte fuel cell according to claim 5, wherein the constant current is 5 mA / cm ² or less per electrode area of each electrolyte membrane electrode assembly.   The method for producing a polymer electrolyte fuel cell according to claim 3, wherein the measurement of the short-circuit conductivity of the electrolyte membrane electrode assembly is performed in the same air atmosphere or inert gas atmosphere of the pair of electrodes.   The hydrogen leak current of the unit cell supplies an inert gas to one of the electrodes, a fuel gas to the other electrode, and applies a constant DC voltage or a constant DC current to the electrolyte membrane electrode assembly. 4. The method for producing a polymer electrolyte fuel cell according to claim 3, wherein the value is obtained by subtracting a value corresponding to the short-circuit conductivity from a value calculated from a steady-state current value or a steady-state voltage value obtained by performing the calculation. .   The method for producing a polymer electrolyte fuel cell according to claim 3, wherein the inspection process is performed before the step of stacking the unit cells to produce the battery stack.   The method for producing a polymer electrolyte fuel cell according to claim 3, wherein the inspection process is performed after the step of stacking the unit cells to produce the battery stack. A process for producing a polymer electrolyte fuel cell comprising: a process for producing a plurality of unit cells; a process for producing a battery stack by laminating the plurality of unit cells; and a process for fastening the battery stack to a lamination method. In the process for producing each unit battery, a pair of electrodes each comprising a gas diffusion layer and a catalyst layer is arranged on both sides of a hydrogen ion conductive polymer electrolyte membrane to produce an electrolyte membrane electrode assembly. A step of disposing a conductive separator plate having a gas flow path for supplying and discharging fuel gas to one of the pair of electrodes, and supplying and discharging oxidant gas to the other electrode The process for producing the unit cell includes the steps of arranging a further conductive separator plate having a gas flow path in contact with each of the electrolyte membrane electrodes. Before the step of producing a combined, on the surface of the gas diffusion layer, the manufacturing method of the polymer electrolyte fuel cell, characterized in that the pre-smoothing process is performed.   An electrolyte membrane electrode assembly comprising a hydrogen ion conductive polymer electrolyte membrane, and a pair of electrodes each comprising a gas diffusion layer and a catalyst layer disposed on both surfaces of the electrolyte membrane, and a fuel on one of the pair of electrodes A conductive separator plate having a gas flow path for supplying and discharging gas and contacting the one electrode, and a gas flow path for supplying and discharging oxidant gas to the other electrode and contacting the other electrode An inspection method for a polymer electrolyte fuel cell comprising a battery stack in which a plurality of unit cells each comprising a further conductive separator plate are stacked and fastened in the stacking direction, the power generation of the polymer electrolyte fuel cell Measuring the short-circuit conductivity of each electrolyte membrane electrode assembly and / or the hydrogen leakage current of the unit cell before or after the battery stack is manufactured before operation; A test method for a polymer electrolyte fuel cell comprising: checking whether a conductivity exceeds a predetermined value of short circuit conductivity and / or whether the hydrogen leak current exceeds a predetermined value of hydrogen leak current. The short conductivity predetermined value is 1.5 mS / cm ^2, the test method for hydrogen leakage current polymer electrolyte fuel cell having a predetermined value according to claim 13, wherein a 3mA / cm ^2. The short-circuit conductivity of the electrolyte membrane electrode assembly is obtained by applying a constant voltage of 0.5 V or less or a constant current of 5 mA / cm ² or less per electrode area to the electrolyte membrane electrode assembly. The inspection method for a polymer electrolyte fuel cell according to claim 13, which is converted from a current value or a steady current value.   The hydrogen leak current of the unit cell supplies an inert gas to one of the electrodes, a fuel gas to the other electrode, and applies a constant DC voltage or a constant DC current to the electrolyte membrane electrode assembly. The polymer electrolyte fuel cell according to claim 13, obtained by subtracting a value corresponding to the short-circuit conductivity from a value calculated from a steady-state current value or a steady-state voltage value obtained by performing the calculation. Inspection method.

Images