Separator for fuel cells and method for manufacturing the same

Abstract

The present invention provides a separator for fuel cells, comprising at least a metal substrate and a carbon layer formed on the metal substrate. The carbon layer has a two-layer structure in which a granular layer in which granular crystals are randomly stacked and a columnar layer in which columnar crystals are arranged extending in a thickness direction of the carbon layer are stacked in the thickness direction of the carbon layer.

Description

Technical Field

[0001] This invention relates to a separator for fuel cells and a method for manufacturing the same. Background Technology

[0002] As a technology, a fuel cell comprises a single cell with an electrolyte membrane sandwiched between an anode electrode and a cathode electrode. Furthermore, the fuel cell is configured as a stack of multiple single cells stacked together via a partition (also called a bipolar plate) having channels for the flow of gases (hydrogen, oxygen, etc.). The output of a fuel cell can be increased by increasing the number of cells per stack.

[0003] The separator in a fuel cell also serves to transmit the generated current to adjacent cells via the surface through which cooling water flows. Therefore, the separator material must possess high conductivity. Specifically, for fuel cell separators, reducing the contact resistance of the electrode adjacent to the separator (in the case of the MEA) or reducing the contact resistance of the gas diffusion layer adjacent to the separator (in the case of the MEGA) is important. Furthermore, high conductivity refers to low contact resistance.

[0004] To meet such requirements, for example, Japanese Patent Application Publication No. 2022-045138 discloses a separator for a fuel cell comprising a metal substrate, an intermediate layer formed on the surface of the metal substrate, and a carbon layer formed on the surface of the intermediate layer, and a method for manufacturing the same. Summary of the Invention

[0005] However, the fuel cell separator shown in Japanese Patent Application Publication No. 2022-045138 sometimes comes into contact with low-pH generated water (corrosive liquid) containing chloride or fluoride ions from the electrolyte membrane during fuel cell power generation. In this case, moisture in the generated water permeates into the carbon layer, and the carbon layer may sometimes peel off. As a result, there is a possibility of increased contact resistance in the fuel cell separator.

[0006] The present invention was made in view of the following circumstances, and its object is to provide a separator for fuel cells and a method thereof capable of inhibiting the penetration of moisture into the carbon layer and preventing carbon layer stripping.

[0007] In view of the above problems, the fuel cell separator of the present invention comprises at least a metal substrate and a carbon layer formed on the metal substrate. The carbon layer has a two-layer structure in which granular layers of randomly stacked granular crystals and columnar layers of columnar crystals arranged extending along the thickness direction of the carbon layer are stacked in the thickness direction of the carbon layer.

[0008] The granular layer and the columnar layer can be sequentially formed from the surface of the metal substrate toward the separator for the fuel cell. The surface of the metal substrate may comprise titanium or a titanium alloy. In this case, an intermediate layer comprising titanium carbide can be formed between the metal substrate and the carbon layer.

[0009] The present invention relates to a method for manufacturing a separator for a fuel cell, comprising at least a metal substrate and a carbon layer formed on the metal substrate. The manufacturing method includes a step of forming the carbon layer by physical vapor deposition. The step of forming the carbon layer includes: a step of forming a granular layer in which granular crystals are randomly stacked by applying a bias voltage to the metal substrate at a predetermined voltage or without applying the bias voltage to the metal substrate; and a step of forming a columnar layer in which columnar crystals extending along the thickness direction of the carbon layer are arranged by applying a bias voltage higher than the predetermined voltage to the metal substrate.

[0010] According to the present invention, the carbon layer has a two-layer structure consisting of a granular layer and a columnar layer, thereby inhibiting water permeation near the interface between the granular layer and the columnar layer. This inhibits the penetration of water into the carbon layer and prevents carbon layer peeling. Attached Figure Description

[0011] The features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, wherein like reference numerals denote like elements, and wherein:

[0012] Figure 1A An exploded perspective view of a single cell including a separator for a fuel cell according to an embodiment of the present invention;

[0013] Figure 1B For stacking Figure 1A The diagram shows a cross-sectional view of the main components of a single-cell fuel cell.

[0014] Figure 2A for Figure 1A The diagram shows a cross-sectional view of the main components of the fuel cell separator.

[0015] Figure 2B for Figure 2A The enlarged cross-sectional view of the carbon layer in part A is shown below;

[0016] Figure 3 A photograph of the portion containing the carbon layer in a cross-section of the test piece of Example 1, taken using a transmission electron microscope (TEM);

[0017] Figure 4A To measure the spectrum of the portion containing the carbon layer using electron energy loss spectroscopy (EELS);

[0018] Figure 4B This is an image of the portion containing the carbon layer, based on the atomic bonding state, derived from EELS measurements.

[0019] Figure 5A A coordinate graph showing the results of the deuterium concentration in the depth direction of the test pieces of Example 1 and Comparative Examples 1 and 2; and

[0020] Figure 5B To show the coordinate graphs of the contact resistance values ​​of the test pieces of Examples 2 and Comparative Examples 3 and 4 corresponding to the constant potential test time; and

[0021] Figure 6 for Figure 5B A summary table showing a portion of the test results and photographs of the test pieces. Detailed Implementation

[0022] Below, we will refer to Figures 1A to 6 The present invention relates to a separator for fuel cells. It will be noted that, in the following description, the separator for fuel cells will be referred to as a separator.

[0023] like Figure 1A , Figure 1B as well as Figure 2A and Figure 2B As shown, the fuel cell 1 in this embodiment has a structure consisting of multiple single cells 10 stacked as basic units. The fuel cell 1 is a solid polymer fuel cell that generates an electromotive force through an electrochemical reaction between an oxidant gas and a fuel gas. The oxidant gas is air. The air is compressed, for example, by a compressor. Air is supplied from the compressor to the fuel cell 1. The fuel gas is hydrogen. Hydrogen is filled in a high-pressure tank. Hydrogen is supplied from the high-pressure tank to the fuel cell 1.

[0024] like Figure 1A As shown, the single cell 10 has a power generation assembly 17 and a pair of separators 12, 12 configured to hold the power generation assembly 17. The power generation assembly 17 has an electrode-gas diffusion layer junction (MEGA) 11 and a resin frame 14 surrounding the outer periphery of the MEGA 11.

[0025] like Figure 1B As shown, MEGA11 has an electrolyte membrane 11a and a pair of catalyst layers 11b and 11c. The pair of catalyst layers 11b and 11c are the electrodes of the fuel cell. The pair of catalyst layers 11b and 11c are bonded to the respective surfaces of the electrolyte membrane 11a. The catalyst layers 11b and 11c are also bonded to gas diffusion layers 11d and 11d. MEGA11 is bonded to a resin frame 14 in the non-power generation region.

[0026] The electrolyte membrane 11a is composed of a proton-conductive ion-exchange membrane formed from a solid polymer material. Examples of electrolyte membranes 11a include thin films made of perfluorosulfonic acid polymers. The catalyst layers 11b and 11c are formed, for example, of porous carbon materials supported on catalysts such as platinum. In this embodiment, the catalyst layer 11b disposed on one side of the electrolyte membrane 11a serves as the anode catalyst layer of the single cell 10, and the catalyst layer 11c on the other side serves as the cathode catalyst layer of the single cell 10. The gas diffusion layer 11d is formed, for example, of a conductive component with gas permeability, such as a porous carbon material like carbon paper or carbon cloth, or a porous metal material like a metal mesh or foamed metal.

[0027] An opening is formed in the center of the resin frame 14, surrounding a portion of the power generation region 25 of the MEGA 11 and a portion of the surrounding non-power generation region. The resin frame 14 is fused to the periphery of one surface of the MEGA 11. The power generation region 25 of the MEGA 11 is fused to each of the partitions 12 in such a way that it is sandwiched between a pair of partitions 12, 12. A hydrogen outlet 14a, a cooling water supply port 14b, and an air supply port 14c are sequentially provided on one side of the resin frame 14 and the pair of partitions 12, 12. An air outlet 14d, a cooling water outlet 14e, and a hydrogen supply port 14f are sequentially provided on the other side of the resin frame 14 and the pair of partitions 12, 12.

[0028] The separator 12 is rectangular in top view and is a component formed from a thin sheet or foil of metal such as stainless steel, titanium, or titanium alloy. The thickness of the separator 12 is, for example, 10 μm to 200 μm. The thin sheet or foil used as the raw material for the separator 12 is formed by cold rolling or the like. The separator 12 is also formed from the aforementioned thin sheet or foil into a predetermined shape by pressing.

[0029] like Figure 1B As shown, the separator 12 has an isosceles trapezoidal wave shape. One side of the separator 12 is in non-adhesive contact with the gas diffusion layer 11d of the MEGA 11. The other side of the separator 12 is in non-adhesive contact with one side of an adjacent separator 12. Each separator 12 has a plurality of protrusions 15f, 15f... formed on the MEGA 11 side. The protrusions 15f are arranged along the long side of the power generation region 25 of the MEGA 11. Therefore, gas flow paths 15a, 15b are formed for hydrogen or air to flow between the protrusions 15f, 15f. A cooling water flow path 15c is formed on the back side (opposite side) of the protrusions 15f of each separator 12 for cooling water to flow.

[0030] When hydrogen is supplied from supply port 14f to gas flow path 15a and air is supplied from supply port 14c to gas flow path 15b, an electrochemical reaction occurs within the single cell 10 to generate an electromotive force. In this electrochemical reaction, protons (H+) are generated in the catalyst layer 11b on the anode side. + Water permeates through the electrolyte membrane 11a in a hydrated state. As a result, generated water is produced in the catalyst layer 11c on the cathode side. The generated water becomes an acidic aqueous solution with a low pH. The generated water may also contain fluoride ions from the electrolyte membrane 11a. Such generated water may damage the surface of the separator 12. In view of these points, the following separator 12 is used in this embodiment.

[0031] like Figure 2A As shown, the separator 12 has at least a metal substrate 21 and a carbon layer 23 formed on the metal substrate 21. In this embodiment, an intermediate layer 22 comprising titanium carbide is formed between the metal substrate 21 and the carbon layer 23.

[0032] The metal substrate 21 has a stainless steel substrate 21A made of stainless steel and a titanium layer 21B made of titanium or a titanium alloy. There are no particular limitations on the type of stainless steel used; for example, austenitic, ferritic, or austenitic-ferritic dual-phase systems can be included. The thickness of the metal substrate 21 is, for example, 0.05 to 1 mm.

[0033] The titanium layer 21B is a layer made of titanium or a titanium alloy. By providing the titanium layer 21B, corrosion of the stainless steel substrate 21A can be suppressed. Examples of titanium alloys include Ti-Al, Ti-Nb, Ti-Ta, Ti-6Al-4V, or Ti-Pd. There are no particular limitations on the thickness of the titanium layer 21B, for example, it is 0.1 to 1000 nm, preferably 10 to 500 nm, and more preferably 30 to 300 nm. Furthermore, the metal substrate 21 can also be made of titanium or a titanium alloy.

[0034] The intermediate layer 22 is made of titanium carbide. By providing the intermediate layer 22, the adhesion between the titanium layer 21B and the carbon layer 23 can be ensured. The thickness of the intermediate layer 22 is, for example, 1 to 10 nm, preferably 1 to 5 nm. Furthermore, if the adhesion between the metal substrate 21 and the carbon layer 23 can be ensured, the intermediate layer 22 can be omitted.

[0035] The carbon layer 23 has a two-layer structure in which a granular layer 23A and a columnar layer 23B are stacked in the thickness direction of the carbon layer 23. The granular layer 23A is a layer of numerous randomly stacked granular crystals 23a. The granular crystals 23a are crystals composed of carbon and have a particle size of about 0.1 to 1.0 nm. The crystallinity of the granular crystals 23a is lower than that of the columnar crystals 23b, which will be described later.

[0036] The columnar layer 23B is a layer of columnar crystals 23b arranged along the thickness direction of the carbon layer 23. The columnar crystals 23b are carbon crystals with an outer diameter of approximately 0.1–1.0 nm. Compared to the granular crystals 23a, the columnar crystals 23b are clusters of carbon crystals and exhibit higher crystallinity. The shape and crystallinity of the granular crystals 23a and columnar crystals 23b can be seen from photographs of the carbon layer 23 taken using a transmission electron microscope (TEM) as described later (see reference). Figure 3 ) and images of the spectrum and mapping analysis of carbon layer 23 using electron energy loss spectroscopy (EELS) (see reference). Figure 4A , Figure 4B (To confirm)

[0037] According to this embodiment, by stacking granular layer 23A and columnar layer 23B into two layers, an interface 23C is formed between granular layer 23A and columnar layer 23B. Within carbon layer 23, the interface 23C, with its different crystalline shapes and crystallinities, functions as a barrier against liquid permeation within the carbon layer 23. Therefore, water permeation into the carbon layer 23 can be suppressed.

[0038] From this perspective, if the carbon layer 23 has a two-layer structure consisting of a granular layer 23A and a columnar layer 23B, the order in which the granular layer 23A and the columnar layer 23B are stacked relative to the metal substrate 21 is not particularly limited. However, a more preferred approach is that, in this embodiment, the granular layer 23A and the columnar layer 23B are formed sequentially from the metal substrate 21 side toward the surface 12f of the separator 12. Specifically, in this embodiment, the granular layer 23A is formed on the surface of the intermediate layer 22. The columnar layer 23B is formed on the surface of the granular layer 23A. As described above, the columnar crystals 23b of the columnar layer 23B have a higher crystallinity than the granular crystals 23a of the granular layer 23A. By placing such a columnar layer 23B closer to the surface of the carbon layer 23 than the granular layer 23A, the contact resistance of the separator 12 can be reduced.

[0039] There is no particular limitation on the thickness of the carbon layer 23, but it is preferably 10 nm to 500 nm, and more preferably 15 to 200 nm. If the above-mentioned interface 23C can be formed, there is no particular limitation on the thickness of the granular layer 23A and the columnar layer 23B, but it is preferably 5 to 100 nm. Similarly, if the above-mentioned interface 23C can be formed, there is no particular limitation that the ratio of the thickness of the columnar layer 23B to the thickness of the granular layer 23A is preferably in the range of 0.1 to 10 times.

[0040] The separator 12 can be manufactured as follows: First, a metal foil equivalent to the metal substrate 21 is prepared. Second, the metal foil is formed into the shape of the separator 12 by pressing. Next, an intermediate layer 22 and a carbon layer 23 are formed on the surface of the metal substrate 21.

[0041] In the method for forming the intermediate layer 22 and the carbon layer 23 (film formation method), the intermediate layer 22 and the carbon layer 23 are formed sequentially by physical vapor deposition (PVD). Here, a bias voltage is applied to the metal substrate 21. If the aforementioned carbon layers with different crystal structures can be formed, the specific method of physical vapor deposition (PVD) is not particularly limited. In this embodiment, sputtering or ion plating can be cited as examples of this method. According to these methods, carbon layers with low hydrogen content can be formed. As a result, the bonding between carbon atoms (sp...) can be increased. 2 The proportion of hybrid carbon can achieve excellent electrical conductivity.

[0042] Sputtering methods include magnetron sputtering, unbalanced magnetron sputtering (UBMS), dual magnetron sputtering or ECR sputtering, and high-power pulsed magnetron sputtering (HiPIMS).

[0043] In the process of forming carbon layer 23, granular layer 23A and columnar layer 23B can be formed by the following steps. In the process of forming granular layer 23A, granular crystals 23a are randomly stacked to form granular layer 23A by applying a bias voltage to metal substrate 21 at a predetermined voltage. Alternatively, in the process of forming granular layer 23A, granular crystals 23a can be randomly stacked to form granular layer 23A without applying a bias voltage to metal substrate 21.

[0044] The specified voltage mentioned here refers to the voltage applied between the carbon target and the metal substrate 21. The specified voltage is preferably, for example, less than 100V. Not applying a bias voltage to the metal substrate 21 means not applying a voltage between the carbon target and the metal substrate 21. That is, the applied bias voltage is 0V. By not applying a voltage to the metal substrate 21, granular crystals with a much lower crystallinity than the columnar crystals 23b of the columnar layer 23B can be obtained. Therefore, in the portion of the interface 23C containing the carbon layer 23, the barrier properties of the carbon layer 23 to liquid permeation can be improved.

[0045] On the other hand, in the process of forming the columnar layer 23B, a columnar layer 23B in which columnar crystals 23b extending along the thickness direction of the carbon layer 23 are arranged by applying a bias voltage higher than the aforementioned predetermined voltage to the metal substrate 21. The bias voltage applied here is preferably, for example, 100V or more, and preferably in the range of 300V or less. By applying a bias voltage in this range to the metal substrate 21, columnar crystals 23b with higher crystallinity than the granular crystals 23a of the granular layer 23A can be obtained.

[0046] In this embodiment, by setting the bias voltage to the voltage relationship described above, the granular layer 23A and the columnar layer 23B can be formed continuously using the same film-forming apparatus. There are no particular restrictions on the order in which the granular layer 23A and the columnar layer 23B are formed. However, in this embodiment, it is preferable to form the columnar layer 23B on the surface of the granular layer 23A after the granular layer 23A has been formed. This allows the columnar layer 23B to be formed closer to the surface of the carbon layer 23 than the granular layer 23A. As a result, the contact resistance of the separator 12 can be reduced.

[0047] The following examples illustrate this implementation method.

[0048] Example 1

[0049] Test pieces of a metal substrate acting as a separator were fabricated. First, a 0.2 mm thick stainless steel substrate (SUS304) was prepared. Titanium layers were formed on both sides of the stainless steel substrate using a PVD film deposition apparatus (Hautzer FC1200). Specifically, after placing the stainless steel substrate inside the reaction vessel of the apparatus, the reaction vessel was evacuated and heated using an internal heater. Next, a pure Ti cathode target for sputtering was etched (cleaned) with plasma-treated Ar gas. Additionally, to remove passivation states present on the surface of the stainless steel substrate, it was etched with plasma-treated Ar gas. Next, using the aforementioned pure Ti cathode target, a 220 nm thick titanium layer was formed by unbalanced magnetron sputtering (UBMS). Similarly, an intermediate layer of 2 nm thick titanium carbide was formed using a titanium carbide target.

[0050] Secondly, using high-purity ta-C as the carbon target, a carbon layer with a thickness of 30 nm was formed by arc ion plating (AIP). Specifically, the bias voltage applied to the metal substrate was set to 0V, i.e., no bias voltage was applied to the metal substrate, thereby forming a granular layer with a thickness of 15 nm. Next, the bias voltage was changed to 150V, thereby forming a columnar layer with a thickness of 15 nm. The test specimen corresponding to Example 1 was thus prepared.

[0051] (Comparative Example 1)

[0052] The sample was prepared in the same manner as in Example 1. The difference between Comparative Example 1 and Example 1 is that the bias voltage was set to 0V, thereby forming a carbon layer with a thickness of 35nm.

[0053] (Comparative Example 2)

[0054] The sample was prepared in the same manner as in Example 1. The difference between Comparative Example 2 and Example 1 is that the bias voltage was set to 150V, thereby forming a carbon layer with a thickness of 35nm.

[0055] Analysis of carbon layers

[0056] The carbon layer containing the test specimen in Example 1 and Comparative Examples 1 and 2 was photographed using a transmission electron microscope (TEM). Figure 3 A TEM image of Example 1 is shown. Furthermore, the atomic bonding state of a portion of the carbon layer containing the test specimen of Example 1 was determined by electron energy loss spectroscopy (EELS). Figure 4A It measures the spectrum of the portion containing the carbon layer. Figure 4B It is an image based on EELS measurements, which is a mapping analysis of the portion containing the carbon layer based on the atomic bonding state.

[0057] from Figure 3 The results show that in Example 1, a granular layer 23A with randomly stacked granular crystals and a columnar layer 23B with columnar crystals arranged along the thickness direction of the carbon layer 23 were formed. Furthermore, the carbon layer in Comparative Example 1 consisted only of granular layers. The carbon layer in Comparative Example 2 consisted only of columnar layers. Figure 4A and Figure 4B As shown, carbon crystals with different crystallinity were found to have formed in the portions forming granular layer 23A and columnar layer 23B. Figure 4A The intensity (count) of the spectra at energies of 290 eV and 295 eV indicates that the carbon clusters constituting the columnar crystals of columnar layer 23B are aggregated. Therefore, it can be concluded that the columnar crystals of columnar layer 23B have higher crystallinity than the granular crystals of granular layer 23A.

[0058] (Corrosion resistance test)

[0059] For the test specimens of Example 1 and Comparative Examples 1 and 2, corrosion resistance tests (potential constant corrosion tests) were performed according to the Japanese Industrial Standard for Metallic Materials Electrochemical High-Temperature Corrosion Test Method (JIS Z2294). In an open-atmosphere system, an impregnation solution containing 3 ppm fluoride ions was prepared in a sulfuric acid aqueous solution (300 mL, pH 3) adjusted to 80°C with temperature-controlled water. The test specimens were immersed in this impregnation solution for 8 hours. Under these conditions, a potential difference of 0.9 V was generated between the counter electrode (sample electrode) and the test specimen by electrically connecting a platinum plate counter electrode, thereby causing corrosion of the test specimen. Furthermore, the potential of the test specimen was kept constant by a reference electrode.

[0060] (Determination of deuterium concentration)

[0061] For the test specimens of Example 1 and Comparative Examples 1 and 2, the deuterium (heavy hydrogen) concentration in the thickness direction of the carbon layer was determined by secondary ion mass spectrometry (SIMS). The deuterium concentration corresponds to the amount of water that penetrates the carbon layer during the corrosion resistance test. The results are shown below. Figure 5A . Figure 5AThis is a coordinate graph showing the results of deuterium concentration along the depth direction of the test subjects in Example 1 and Comparative Examples 1 and 2.

[0062] from Figure 5A The results showed that the deuterium concentration of the carbon layers in Comparative Examples 1 and 2 was about two orders of magnitude higher than that in Example 1. This result confirms that water permeation into the carbon layer was achieved even when the carbon layer was only a granular layer, as in Comparative Example 1, or even when the carbon layer was only a columnar layer, as in Comparative Example 2. On the other hand, in the case of Example 1, it was believed that by forming the carbon layer into a two-layer structure of granular and columnar layers, water permeation could be suppressed in the portion including the interface.

[0063] (Example 2)

[0064] The test specimen was prepared in the same manner as in Example 1. Example 2 differs from Example 1 in that no intermediate layer is formed; instead, a carbon layer is formed on the surface of the metal substrate. Furthermore, Example 2 differs from Example 1 in that a 35 nm carbon layer is formed, and the thicknesses of the granular layer and the columnar layer are each set to 17.5 nm.

[0065] (Comparative Example 3)

[0066] The test specimens were prepared in the same manner as in Example 2. The difference between Comparative Example 3 and Example 2 is that the bias voltage was set to 250V to form a carbon layer (columnar layer) with a thickness of 100nm.

[0067] (Comparative Example 4)

[0068] The test specimens were prepared in the same manner as in Example 2. The difference between Comparative Example 4 and Example 2 is that the bias voltage was set to 150V to form a carbon layer (columnar layer) with a thickness of 35nm.

[0069] Contact resistance measurement test

[0070] For the obtained separator, the initial contact resistance was measured. The contact resistance was measured as follows: carbon cloth was placed on one side of the test piece (splitter), and the laminate was clamped between two copper electrodes and subjected to a load of 98 N (10 kgf). Under this condition, a DC power supply was applied, and the voltage applied between the carbon cloth and the substrate was measured using a voltmeter. The contact resistance value was then calculated. The results are presented below. Figure 5B and Figure 6 . Figure 5B This is a coordinate graph showing the contact resistance values ​​corresponding to the constant potential test time of the test pieces in Example 2 and Comparative Examples 3 and 4. Figure 6 yes Figure 5B A summary table showing a portion of the test results and photographs of the test pieces.

[0071] like Figure 5B and Figure 6 As shown, even with increased test time, the contact resistance value of the test specimen in Example 2 differed from that in Comparative Examples 3 and 4, showing almost no increase. In the case of Comparative Example 3, the contact resistance value increased with increasing test time, and carbon layer peeling was confirmed at a test time of 240 hours. The contact resistance value of the test specimen in Comparative Example 4 increased significantly with increasing test time compared to the sample in Example 1. Furthermore, the test specimen in Comparative Example 4 turned brown at a test time of 360 hours. As a result, the test specimen in Example 2 was able to suppress water penetration into the carbon layer in the same way as in Example 1. Furthermore, the test specimen in Example 2 was also able to suppress the increase in contact resistance.

[0072] The embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above embodiments, and various design changes can be made without departing from the spirit of the present invention as set forth in the claims.

Claims

1. A separator for a fuel cell, characterized by It has at least a metal substrate and a carbon layer formed on the metal substrate, wherein the carbon layer has a two-layer structure in which a granular layer of randomly stacked granular crystals and a columnar layer of columnar crystals arranged extending along the thickness direction of the carbon layer are stacked in the thickness direction of the carbon layer.

2. The separator for fuel cells according to claim 1, characterized by The granular layer and the columnar layer are sequentially formed from the metal substrate side toward the surface of the fuel cell separator.

3. The separator for fuel cells according to claim 1, characterized by The surface of the metal substrate comprises titanium or a titanium alloy, and an intermediate layer comprising titanium carbide is formed between the metal substrate and the carbon layer.

4. A method for manufacturing a separator for a fuel cell, wherein the separator for a fuel cell comprises at least a metal substrate and a carbon layer formed on the metal substrate, characterized in that, The manufacturing method includes a step of forming the carbon layer by physical vapor deposition. The process of forming the carbon layer includes: A process of forming a granular layer in which granular crystals are randomly stacked by applying a bias voltage to the metal substrate with a specified voltage or without applying the bias voltage to the metal substrate; and The process of forming a columnar layer by applying a bias voltage higher than the specified voltage to the metal substrate, thereby arranging columnar crystals that extend along the thickness direction of the carbon layer.

Images