JP2007026694A - Separator for solid polymer type fuel cell, and solid polymer type fuel cell - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a stainless steel separator that has a low contact resistance of 10 Ωm cm<SP>2</SP>or less and maintains the low contact resistance even after a long time continuous operation of a fuel cell. <P>SOLUTION: The stainless steel separator is provided with micro pits of sizes of 0.01 to 1 μm over an entire surface, and is preferably provided with two hundred or more micro pits per 5×5 μm surface area. The micro pits are formed by the immersion of a Cr-Mo containing stainless steel in a nonoxidizing acid solution to modify the stainless steel surface to a surface mode matching irregularities in fiber surfaces of carbon paper. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a stainless steel separator incorporated in a polymer electrolyte fuel cell that can be operated at a low temperature and is easy to maintain.

  The polymer electrolyte fuel cell can operate at a low temperature of 100 ° C. or less and has an advantage of starting in a short time. In addition, since each member is a simple structure made of a solid, not only maintenance is easy, but it can also be applied to applications that are exposed to vibrations and impacts. In addition, it has advantages such as high output, suitable for downsizing, high fuel efficiency and low noise.

  In order to take out the amount of power that can be practically used from a fuel cell with a very small amount of power generation per cell, it is necessary to stack a number of cells with one unit of a solid polymer membrane sandwiched between separators. A separator that sandwiches a solid polymer film is required to have good electrical conductivity and low contact resistance, so that a graphite separator has been conventionally used. However, the graphite separator is brittle and easily breaks when excessive vibration or impact is applied. Due to its low workability, it is necessary to perform cutting work when manufacturing a product with a complicated shape, and it is unavoidable to change from a brittle material to a thick product. As a result, the graphite separator cannot sufficiently meet the demand for compactness, and the manufacturing cost increases.

Therefore, it has been studied to use stainless steel for the separator instead of graphite (Patent Documents 1 and 2). Stainless steel can be thinned because it has high strength and excellent ductility, and has the advantage that it can be formed into a target shape by an inexpensive processing method such as press molding. In addition, the steel plate surface is covered with a passive film formed from oxides and hydroxides of Cr, Mo, Fe, etc., which are constituents of stainless steel, and the corrosion of the underlying steel is suppressed by the barrier effect of the passive film. The
Japanese Patent Laid-Open No. 9-15801 JP 2000-239806 JP

The passive film is effective in improving the corrosion resistance, but exhibits semiconducting properties and is inferior in electrical conductivity compared to the base steel. Therefore, when stainless steel with a normal passive film is used for the separator, the contact resistance with the electrode is large, the electric energy generated by the cell reaction is consumed as Joule heat, and the power generation efficiency of the fuel cell decreases. .
In order to apply stainless steel to the separator while utilizing excellent corrosion resistance, it is necessary to reduce the contact resistance of the stainless steel surface, and precious metal coating, roughening of the stainless steel surface, etc. are being studied as means for reducing contact resistance.

However, the expensive precious metal coating increases the cost of the fuel cell, and restricts the spread of the fuel cell from the economical aspect. Moreover, with noble metal coating, pinholes that are the starting point of pitting corrosion are likely to be formed in the film, so strict product management is required. Although it is possible to form a noble metal film without pinholes by thick plating, a large amount of expensive noble metal is consumed, which is a bottleneck in cost reduction.
In order to reduce the contact resistance of stainless steel by roughening treatment, an alternating electrolysis method in a ferric chloride bath is partially adopted, but the equipment tends to be large because of the electrolytic treatment.

It is also known to roughen stainless steel by shot blasting, pickling and etching to reduce contact resistance. In Patent Document 3, it is said that the contact resistance is lowered when the surface roughness of the separator is adjusted to Ra 0.1 to 10 μm by shot blasting or pickling. In Patent Document 4, it is said that the contact resistance decreases when the surface roughness of the separator is adjusted to Ra 0.01 to 1.0 μm by pickling.
JP 11-297338 A JP 2003-223904 A

In Patent Document 5, cueing treatment of conductive inclusions is performed by shot blasting or pickling the surface of stainless steel in which conductive carbide metal inclusions and boride metal inclusions are deposited on stainless steel, Contact resistance is reduced. However, it is necessary to add a process for forming conductive inclusions in stainless steel, a process such as shot blasting and pickling, and the manufacturing cost increases. Further, the precipitation of chromium carbide means the consumption of Cr necessary for the development of corrosion resistance, which leads to a decrease in corrosion resistance and cannot ensure sufficient corrosion resistance in a harsh environment within the fuel cell.
Japanese Patent Laid-Open No. 2001-32056

Surface roughening of stainless steel by electrolytic etching has also been studied (Patent Document 6). Contacting the fuel cell electrodes with the protrusions made of stainless steel with unevenness by electrolytic etching will increase the contact area and reduce the contact resistance. However, the unevenness produced by the electrolytic surface roughening treatment has a large pitch and is insufficient to obtain a low contact resistance comparable to that of a carbon or gold plating layer. Since it is an electrolytic treatment, it is also a problem that the equipment tends to be large.
Japanese Patent Application No. 2000-276893

  The present inventors examined various means for reducing the contact resistance of stainless steel for separators, and investigated the influence of the surface form on the contact resistance. As a result, it has been found that it is not sufficient to simply adjust the surface roughness of the separator using Ra, Ry, etc. as an index, and the microscopic surface shape has a great influence on the reduction of contact resistance. That is, if the surface of the stainless steel can be modified to a surface shape that matches the carbon fiber of the carbon paper that contacts the separator, the contact resistance can be greatly reduced and the low contact resistance can be maintained over a long period of time.

  The present invention has been completed from such a viewpoint, and by improving the contact state of the separator with respect to the carbon fiber of the carbon paper by forming a large number of micropits of substantially the same size as the carbon fiber on the stainless steel surface, the contact resistance is improved. An object of the present invention is to provide a polymer electrolyte fuel cell separator that is effective in reducing power consumption and thus improving power generation efficiency.

  In the separator for a polymer electrolyte fuel cell of the present invention, micropits having a size of 0.01 to 1 μm are formed on the entire surface, preferably Cr: 16 to 40% by mass, Mo: 1 to 5% by mass. It is produced from the stainless steel plate containing. The distribution of micropits is preferably 200 or more per 5 μm × 5 μm surface area.

  By the formation of micropits, the surface roughness measured by AFM is preferably adjusted in a surface area of 5 μm × 5 μm, and the ratio of the power spectral density of Ra ≧ 0.01 μm, 0.2 μm and 5 μm is adjusted to 0.1 or more. . Further, the surface electrode portion as a value measured by a surface roughness meter having a surface area increase rate {(surface area / projected area) −1} × 100 of the surface in contact with the electrode of 15 μm or more and a probe tip diameter of 5 μm. The surface roughness of the surface in contact with is preferably Ra ≦ 0.5 μm.

  The present inventors have made a microscopic observation of carbon paper serving as an electrode on the assumption that the contact form between the separator and the electrode is an important factor for reducing the contact resistance. The carbon paper was produced by weaving carbon fibers having a diameter of about 10 μm, and unevenness of submicron order was distributed in a gear shape on the surface of each carbon fiber (FIG. 1). The area between the protrusions on the fiber surface was also in the submicron order.

  If the separator surface is in a form that easily matches the surface form of the carbon fiber (electrode), it is considered that the separator comes into contact with the electrode surface in an optimum state, and the contact resistance can be reduced. As seen in the carbon fiber / separator contact state model diagram (FIG. 2), the smooth separator only comes in contact with the carbon paper at the carbon fiber protrusion, and the contact surface is very small (a).

  In conventional pickling and etching using an oxidizing acid, the surface roughness is increased to increase the contact point with the carbon paper and thus reduce the contact resistance (b). Roughening is the result of etching that proceeds along the grain boundaries on the surface of the stainless steel, forming irregularities with a relatively large pitch. For this reason, the separator roughened by the conventional method is theoretically limited in the contact point with the carbon paper to the convex portion on the fiber surface, and there is a limit in reducing the contact resistance. It is also a drawback that the roughened stainless steel surface is altered in a relatively short time and the contact resistance is increased.

  On the other hand, in the surface treatment using a non-oxidizing acid, fine uneven pits (hereinafter referred to as “micro pits”) can be formed in the entire surface of the stainless steel regardless of the presence of the grain boundaries. The size and pitch of the micropits can be adjusted according to the contact conditions (concentration, temperature, time, etc.) with the non-oxidizing acid. If the size is the same as the carbon fiber irregularities (0.01 to 1 μm), the carbon paper fibers And the contact resistance is sufficiently low (c).

The stainless steel targeted by the present invention is a ferritic stainless steel containing Cr: 15-40 mass% and Mo: 1-5 mass%.
Cr is an essential component for ensuring the corrosion resistance of stainless steel. Assuming a low pH and highly corrosive fuel cell internal environment, 15 mass% or more of Cr is required. Although an increase in Cr is effective in improving corrosion resistance, it causes a decrease in workability, so the upper limit is made 40% by mass.

Mo is a component necessary for forming micropits that improve the familiarity of carbon paper with fibers. When the Mo content is less than 1% by mass, the micropits are not formed by the dipping treatment using a non-oxidizing acid, and the entire surface is dissolved, so that it is not effective in reducing the contact resistance. However, if an excessive amount of Mo exceeding 5% by mass is included, the stainless steel becomes hard and the workability deteriorates.
In addition to Cr and Mo, stainless steel containing C, N, Si, P, S, Ni, Cu, Ti, Nb, Al, V, etc. can also be used.

  C and N have an adverse effect on the workability and low-temperature toughness of stainless steel, so they should be reduced as much as possible, and preferably both are 0.02% by mass or less. Since Si hardens stainless steel and lowers workability, it is preferably 0.5% by mass or less. P is an effective component for improving the corrosion resistance in the internal environment of the fuel cell to which the separator is exposed. However, excessive addition has an adverse effect on processability, so that when added, the content of 0.03 to 0.08 mass%. Select P content by range. Since S is a component harmful to corrosion resistance, it needs to be reduced as much as possible, and is preferably regulated to 0.005% by mass or less.

  Ni and Cu are elements that are easily eluted, so that they are not contained in large amounts, and the upper limit is preferably set to Ni: 0.5% by mass and Cu: 0.8% by mass. Among these, when the eluted Ni ions reach the catalyst layer, the catalyst is poisoned and the battery performance is deteriorated. The addition of a small amount improves the overall corrosion resistance in an acidic atmosphere and also improves the low temperature toughness of the ferritic stainless steel. Therefore, when added, Ni: 0.15 to 0.35 mass%, Cu: It is preferable to select the content of Ni and Cu in the range of 0.20 to 0.50% by mass.

  Ti and Nb have the effect of fixing C and N in the steel and improving workability, so they are both added in the range of 0.03 to 0.25% by mass as necessary. When Al is used for fixing N, the Al content is selected in the range of 0.04 to 0.25% by mass. V has the effect of improving the corrosion resistance in the internal environment of the fuel cell, and is added in the range of 0.2 to 1.0% by mass as necessary. Further, other alloy components can be added as long as the characteristics are not greatly changed.

  When Mo-containing ferritic stainless steel adjusted to a predetermined composition is immersed in a non-oxidizing acid solution, a large number of micropits are formed over the entire surface of the steel sheet. In the immersion treatment using a non-oxidizing acid solution, the acid type, concentration, temperature, and immersion time depending on the type of stainless steel so that the micropit size is 1 μm or less (preferably 0.5 μm or less). Adjust soaking conditions. For example, in 30Cr-2Mo steel, conditions of immersion in a hydrochloric acid bath having a concentration of 10 to 20% by mass and a liquid temperature of 40 to 60 ° C. for 0.5 to 10 minutes are employed. When sulfuric acid is used, it is immersed in a sulfuric acid bath having a concentration of 10 to 20% by mass and a liquid temperature of 50 to 80 ° C. for 0.5 to 20 minutes.

  When stainless steel is immersed in a non-oxidizing acid solution, dissolution of the steel sheet surface begins. In the case of high corrosion resistant ferritic stainless steel containing Mo, partial pitting corrosion occurs instead of melting the entire surface, and micropits are formed. The melting at this time is almost independent of the distribution form of the crystal grain boundaries and also occurs in the crystal grains, and micropits are distributed at a substantially uniform density on the steel plate surface. The micropits grow to the submicron order, but during that time, a large number of pits are generated in other surface areas, and as a result, the micropits are formed in a distribution covering the entire surface area. The contact resistance decreases as the number of micropits increases, and the lowest contact resistance is obtained when the micropits are formed over the entire surface of the stainless steel.

When the dipping process is further continued, the micropits are connected and integrated with each other, resulting in a surface form similar to the unevenness seen on the steel surface roughened by normal etching, and the contact resistance increases.
The micropits have a size adjusted to a range of 0.01 to 1 μm (preferably 0.05 to 1 μm, more preferably 0.1 to 0.5 μm). The size: 0.01 to 1 μm is in the same order as the uneven size of the fiber surface of the carbon paper, and improves the familiarity of the separator to the carbon paper. If the size of the pit is 0.01 μm or less, an increase in the contact area of the carbon paper with the fiber surface cannot be expected. Conversely, if the size exceeds 1 μm, it is difficult to match the carbon fiber surface and the contact area is reduced.

The micropits are preferably distributed on the separator surface at a density of 200 or more (preferably 500 or more) per surface area of 5 μm × 5 μm. When the distribution density is less than 200, the smooth surface occupies a large proportion, so that a sufficient contact resistance reduction effect cannot be obtained, and the contact resistance in the environment inside the fuel cell increases.
In addition, the average roughness in the surface area of 5 μm × 5 μm measured with a tapping mode atomic force microscope (AFM) is Ra ≧ 0.01 μm (preferably Ra ≧ 0.02 μm), and calculation was performed using AFM data. In the two-dimensional isotropic power spectral density (2D isotropic PSD), the power spectral density ratio (0.2 / 5: PSD) between the wavelength of 0.2 μm and the wavelength of 5 μm is 0.1 or more (preferably 0.5 or more). is there.

  The two-dimensional isotropic power spectral density is suitable for expressing the pitch of the projections and depressions such as the interval between the peaks and depressions of the projections and depressions instead of the roughness index such as Ra and Ry when evaluating the surface morphology. A preferable surface form of the separator is that there are unevenness of submicron order matching the unevenness of the carbon fiber surface. In a separator exhibiting low contact resistance, the ratio of the power spectral density in the submicron order to the power spectral density in the micron order (submicron order power spectral density / micron order power spectral density) needs to be increased. For example, when a power spectral density of a wavelength of 0.2 μm is taken as a submicron order and a power spectral density of a wavelength of 5 μm is taken as a micron order, the ratio (0.2 / 5: PSD) is 0.1 or more (preferably, 0 It is possible to reduce the contact resistance.

The distribution density of micropits existing on the separator surface can also be specified by the amount of increase in surface area. Although the effective surface area of the separator is increased by micropits, the increase rate of the effective surface area can be quantified as a ratio to the projected area of the steel sheet surface. That is, assuming that the effective surface area obtained by AFM measurement is S ₁ and the projected area is S ₀ , the surface area increase rate ΔS (%) is calculated as {(S ₁ / S ₀ ) −1} × 100.
A large surface area increase rate ΔS means that there are a large number of micropits on the separator surface, and ΔS ≧ 15% (preferably ΔS ≧ 20%) improves the contact state of the separator to the fiber surface of the carbon paper, and the contact resistance The reduction effect becomes remarkable.

  In addition to the surface area increase rate ΔS ≧ 15%, the surface roughness measured with a probe having a tip diameter of 5 μm is preferably Ra ≦ 0.5 μm. The probe with a tip diameter of 5 μm is subject to measurement of micron-order irregularities, and a measured value of Ra ≦ 0.5 μm means that there are few irregularities that do not contribute to contact with the carbon fiber. On the contrary, if Ra> 0.5 μm, not only the increased surface does not work for contact with the carbon fiber, but also the amount of metal ions eluted from the metal separator as the surface area increases unnecessarily. This degrades the performance of the fuel cell.

A total of four types of ferritic stainless steels of 30Cr-2Mo, 22Cr-1.2Mo, 18Cr-2Mo, and 18Cr were used as separator materials, and the influence of acid immersion on the surface morphology of stainless steel was investigated. Prior to acid immersion, all stainless steels were subjected to a degreasing treatment in which they were immersed in a sodium orthosilicate solution having a concentration of 5% by mass and a liquid temperature of 60 ° C. for 10 seconds.
The stainless steel after the degreasing treatment was immersed using a hydrochloric acid solution having a concentration of 10% by mass and a liquid temperature of 50 ° C. Immediately after the dipping treatment, it was washed with water and dried with a dryer. In the immersion treatment, the immersion time was varied in order to investigate the effect of the surface morphology on the contact resistance. For comparison, an aqua regia in which 60% by mass nitric acid and 36% by mass hydrochloric acid were mixed at a volume ratio of 1: 3 and a mixed acid of 2.5% by mass hydrofluoric acid and 8% by mass nitric acid were used as the acid solution, and the same immersion treatment was performed. did.

  A test piece was cut out from the immersion-treated stainless steel, and the initial contact resistance and the contact resistance after the wet test were measured. The initial contact resistance was measured by measuring the contact resistance of stainless steel / carbon paper after leaving the test piece in the room for 72 hours and then contacting the test piece with carbon paper at a load of 1 MPa. The contact resistance after the wet test is the same as the test piece left for 72 hours in a wet environment of temperature: 70 ° C. and humidity: 98% RH simulating the internal environment of the fuel cell. Contact resistance was measured.

Contact resistance stainless steel which is not immersed was 550mΩ · cm ² at 40m · cm ² in 30Cr-2Mo, 210mΩ · cm ² in 22Cr-1.2Mo, 18Cr-2Mo by 50mΩ · cm ^2, 18Cr However, in all cases, the contact resistance was greatly reduced by acid immersion. Moreover, the immersion time for obtaining a low contact resistance varied depending on the steel type.

As a result of investigating the surface morphology of each immersed stainless steel by SEM, the surface of the stainless steel immersed in hydrochloric acid was clearly different from the stainless steel immersed in aqua regia or mixed acid. Micropits were counted from the surface image photographed by FE-SEM, and the distribution density of micropits was determined as the number per 5 μm × 5 μm surface area.
When the number of micropits is 200 or more per 5 μm × 5 μm surface area, as shown in Table 1, low contact resistance is exhibited by acid immersion, and the low contact resistance is maintained even after 72 hours of the wet test. On the other hand, with a distribution density of less than 200, the contact resistance was not sufficiently reduced by acid immersion, and the contact resistance was significantly increased even after the wet test.

Next, taking 30Cr-2Mo as an example, the immersion time was changed by immersion treatment using 10% by mass hydrochloric acid (50 ° C.), and the influence of the immersion time on contact resistance and surface roughness was investigated. The surface roughness was measured with a stylus type surface roughness meter having a tip diameter of 5 μm.
As can be seen from the measurement results in Table 2, the surface roughness was in the range of Ra: 0.16 to 0.17 μm regardless of before and after the immersion treatment and the immersion time, but the contact resistance was greatly reduced by the immersion treatment. Was. The treatment time of 5 minutes showed the same Ra as that of the untreated, but the surface area increase rate was 45.6% which was very large. The results in Table 2 mean that the conventional methods (Patent Documents 3 and 4) in which the contact resistance is reduced by adjusting the surface roughness Ra cannot be ignored.

  Since the reduction in contact resistance cannot be explained by a simple surface roughness Ra, a scanning probe microscope (Nanoscope IIIa, D3100: Digital Instrument) was used to investigate the surface morphology of the immersed stainless steel in more detail. The surface of the stainless steel was observed using a tapping mode atomic force microscope (AFM) with a scan size of 5 μm × 5 μm. Table 3 shows the surface roughness of the micro area measured by AFM.

The surface of the carbon paper fiber surface was also subjected to AFM analysis in advance. The carbon paper had irregularities on the order of submicron as seen in the cross-sectional profile of FIG. 3, and the pitch (wavelength) of the irregularities was approximately 0.2 μm.
Next, the surface shape of the stainless steel was evaluated by two-dimensional isotropic power spectral density using the AFM data obtained by the measurement. The evaluation of the surface morphology based on the two-dimensional isotropic power spectral density is effective when there is a difference in the unevenness scale on the stainless steel surface even if Ra is substantially the same.

  As can be seen from FIG. 4 showing the calculation result of the two-dimensional isotropic power spectral density, the spectral density of the stainless steel immersed in hydrochloric acid is increased in both short wavelength and long wavelength compared with the untreated stainless steel. The wavelength of the surface of the fiber of carbon paper is about 0.2 μm, and a corresponding surface form with a large ratio of the wavelength of 0.2 μm can be obtained by the immersion treatment for 5 minutes, and the wavelength of 0.2 μm is obtained by the immersion treatment for 60 minutes. The proportion occupied was decreasing. Corresponding to the ratio of the wavelength of 0.2 μm, a significantly lower contact resistance was obtained in 5 minutes of immersion treatment compared to 60 minutes of immersion treatment. This result shows that a surface form with a larger ratio of short wavelengths is effective in reducing the contact resistance, not just the surface roughness Ra.

An SEM image (FIG. 5) observing the surface of stainless steel immersed in hydrochloric acid for 5 minutes shows that micropits are formed over the entire surface. Even if this surface is measured with a stylus type surface roughness meter with a tip diameter of 5 μm, Ra is 0.16 μm, which is not much different from the conventional pickling material, but the cross-sectional profile obtained by cross-sectional analysis of the AFM image shows that the carbon paper It can be seen that irregularities having substantially the same pitch as the irregularities on the fiber surface are formed. As a result, it is considered that the stainless steel was in contact with the carbon paper with the good matching property described in FIG. 2 (c), and the contact resistance was as low as 4 mΩ · cm ² .

On the other hand, the stainless steel immersed in aqua regia for 5 minutes (FIG. 6a) had micron-order irregularities, but the proportion of submicron-order irregularities was significantly less. The surface morphology of the stainless steel immersed in the mixed acid for 5 minutes (FIG. 6b) hardly changed. These contact resistances were large values of 9.8 mΩ · cm ² and 19.5 mΩ · cm ² , respectively.
Further, a material immersed in hydrochloric acid for 5 minutes was press-molded into a separator shape. SEM observation of the surface of the part of the press-molded material that contacts the electrode showed almost no change from that before press molding. The contact resistance was 4.6 mΩ · cm ² , and there was no substantial increase in contact resistance even by press molding.

Next, an acid-immersed stainless steel separator was assembled on the fuel electrode side and the oxidation electrode side of the fuel cell, and the fuel cell was continuously operated for 100 hours, and an increase in contact resistance and an output change were investigated.
As can be seen from the results of the investigation in Table 4, the fuel cell incorporating the stainless steel separators No. 1 to 3 has no decrease in output even after 100 hours of continuous operation, and the increase in contact resistance is 5 mΩ · cm ^2. It was suppressed to the following. When the change in the contact resistance is organized by the surface area increase rate, the separator adjusted to a surface form with a surface area increase rate of 15% or more maintained a low contact resistance even after 100 hours of continuous operation.

On the other hand, in the 18Cr stainless steel separator in which the micropits are not formed, the increase in the contact resistance is large and the output is reduced. Moreover, in the stainless steel separator whose surface area increase rate does not reach 15%, the contact resistance significantly increased after continuous operation, and the output also decreased accordingly.
From this comparison, it can be seen that by forming micropits on the surface of stainless steel, low contact resistance and high output are maintained over a long period of time, and characteristics suitable for fuel cells are imparted. In addition, by suppressing the surface roughness Ra to 0.5 μm or less with a surface form with a surface area increase rate of 15% or more, a stainless steel separator that is also effective in maintaining low contact resistance and preventing fuel cell output from decreasing is obtained.

  As explained above, the stainless steel separator in which micropits are formed on the steel surface by dipping treatment using a non-oxidizing acid solution has a high matching with the unevenness on the fiber surface of the carbon paper, and the gap between the carbon electrode and the carbon electrode. In good contact. Therefore, a low contact resistance is achieved, and an increase in the low contact resistance is suppressed even after long-time continuous operation of the fuel cell, so that it is used as a fuel cell fuel cell separator suitable for improving power generation efficiency.

SEM image of carbon fiber that forms carbon paper Schematic diagram for explaining the familiarity of separators with carbon fibers Graph showing the distribution of unevenness on the carbon fiber surface Graph showing that the surface morphology changes according to the treatment time of acid immersion using non-oxidizing acid solution Chart showing surface morphology and contact resistance of 30Cr-2Mo stainless steel immersed in hydrochloric acid SEM image showing surface morphology of 30Cr-2Mo stainless steel immersed in aqua regia (a) and mixed acid (b)

Claims (6)

  A separator for a polymer electrolyte fuel cell, comprising a stainless steel plate in which micropits having a size of 0.01 to 1 μm are formed on the entire surface.   2. The polymer electrolyte fuel cell separator according to claim 1, wherein 200 or more micropits are formed per surface area of 5 μm × 5 μm.   2. The polymer electrolyte fuel cell according to claim 1, wherein the surface roughness measured by AFM in the surface area of 5 μm × 5 μm is Ra ≧ 0.01 μm, and the ratio of the power spectral density of 0.2 μm to 5 μm is 0.1 or more. Separator for use.   Surface increase rate {(surface area / projected area) -1} × 100 with respect to the projected area of the surface in contact with the electrode is 15% or more, and the surface electrode portion is in contact with the surface measured by a surface roughness meter having a probe tip diameter of 5 μm. The separator for a polymer electrolyte fuel cell according to claim 1, wherein the surface roughness of the surface to be satisfied is Ra ≦ 0.5 μm.   5. The polymer electrolyte fuel cell separator according to claim 1, wherein the stainless steel contains Cr: 16 to 40 mass% and Mo: 1 to 5 mass%.   A separator for a polymer electrolyte fuel cell incorporating the separator according to any one of claims 1 to 5.