Fuel cell, fuel cell stack, and method for manufacturing fuel cell

Abstract

The purpose of this invention is to provide a fuel cell with which it is possible to reduce the film thickness of an electrolyte layer by suppressing the influence of a protrusion on the surface of a support substrate having through-holes. A fuel cell according to the present invention has a structure in which a porous support layer, a first electrode layer, an electrolyte layer, and a second electrode layer are stacked. The surface of the first electrode layer is flat or the film thickness of the electrolyte layer is uniform (see FIG. 2).

Description

Fuel cell, fuel cell stack, and method for manufacturing fuel cell

[0001] The present invention relates to a fuel cell.

[0002] In recent years, fuel cells have been attracting attention as a power generation system that uses fuels such as hydrogen and does not emit carbon dioxide. A fuel cell has a structure in which an electrolyte is sandwiched between two electrodes, an anode and a cathode, and generates electricity by supplying a fuel gas such as hydrogen to the anode side and an oxygen-containing gas such as air to the cathode side.

[0003] The following Patent Document 1 addresses the issue of "providing an electrochemical element or the like that has high performance and durability by providing a diffusion prevention layer that can suppress element diffusion from a metal support to an electrode layer and is less likely to inhibit gas diffusion," and describes the following technology (see Abstract): "An electrochemical element E includes a gas-permeable metal support 1, a diffusion prevention layer 7 formed on the metal support 1, an electrode layer 2 formed on the diffusion prevention layer 7, an electrolyte layer 4 formed on the electrode layer 2, and a counter electrode layer 6 formed on the electrolyte layer 4, wherein the diffusion prevention layer 7 includes a first diffusion prevention layer 7a having a first porosity and a second diffusion prevention layer 7b having a second porosity higher than the first porosity."

[0004] Japanese Patent Application Laid-Open No. 2021-163764

[0005] A thin-film SOFC (Solid Oxide Fuel Cell) has a basic structure in which an anode electrode layer, a solid electrolyte layer, and a cathode electrode layer are stacked on a support substrate. The support substrate must have through-holes to allow gas to be introduced. An anodic alumina (AAO) substrate is known as an example of such a support substrate. Furthermore, the thinner the solid electrolyte layer, the higher the output power of the SOFC.

[0006] Convex portions may be formed on the surface of a support substrate made of AAO or the like. If the height of the convex portions exceeds, for example, the thickness of the solid electrolyte layer, short circuits will occur at those locations. If the thickness of the solid electrolyte layer is increased to prevent short circuits caused by such convex portions, this will hinder the improvement of the output power of the SOFC.

[0007] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a fuel cell that can suppress the influence of protrusions on the surface of a support substrate having through holes and can reduce the film thickness of the electrolyte layer.

[0008] The fuel cell according to the present invention has a structure in which a porous support layer, a first electrode layer, an electrolyte layer, and a second electrode layer are stacked, and the surface of the first electrode layer is flat, or the thickness of the electrolyte layer is uniform.

[0009] According to the fuel cell of the present invention, the influence of the protrusions on the surface of the support substrate having the through holes can be suppressed, and the film thickness of the electrolyte layer can be reduced. Problems, configurations, and effects other than those described above will become clear from the description of the embodiments below.

[0010] 1 is a cross-sectional view showing the structure of a fuel cell 1 according to embodiment 1. FIG. 2 is an enlarged view of the fuel cell 1 shown in FIG. 1. FIG. 3 is a cross-sectional SEM image of the first electrode layer 3. FIG. 4 is a cross-sectional SEM image of the first electrode layer 3. FIG. 5 is a schematic diagram showing how the first electrode layer 3 is planarized by ion milling. FIG. 6 shows a method for deriving the arithmetic mean roughness Ra. FIG. 7 shows an example of deriving the arithmetic mean roughness Ra when the surface of the first electrode layer 3 is irregular. FIG. 8 shows an example of deriving the arithmetic mean roughness Ra when the surface of the first electrode layer 3 has holes resulting from the micro through-holes 6. FIG. 9 shows an example of deriving the arithmetic mean roughness Ra when the surface of the first electrode layer 3 includes a gentle slope. FIG. 10 shows a method for deriving the maximum height Rz of the surface shape of the first electrode layer 3. FIG. 11 is an SEM image showing protrusions on the porous support layer 2. FIG. 12 is an example of planarization of the porous support layer 2 by ion milling. FIG. 13 is a cross-sectional view showing an example of the structure of a fuel cell stack 1200. FIG. 14 is an exploded view of the layers of the fuel cell stack 1200. FIG. 15 is an enlarged cross-sectional view of a fuel cell 1 according to embodiment 2. 17 shows an enlarged cross-sectional view of a fuel cell 1 when an electrolyte layer 4 is formed on a porous support layer 2 having protrusions. This shows the results of measuring the electron leakage current when a 200 nm-thick electrolyte layer is formed on a silicon substrate without through-holes or steps, and a voltage is applied between the first electrode layer 3 and the second electrode layer 5. This shows a cross-sectional view of a fuel cell 1 according to a third embodiment. This shows the results of measuring the electron leakage current when a voltage of 1 V is applied to a first interface layer 7 of 100 nm, an electrolyte layer 4 of 200 nm, and a second interface layer 8 of 200 nm, with and without planarization treatment. This shows the yield versus the arithmetic mean roughness Ra, when electron leakage currents of 1×10^-3 A / cm^2 or more are considered defective in the measurement results of FIG. 17. This shows the results of confirming the power generation operation of a non-defective product using 3% hydrogen (nitrogen-based) as the fuel gas and air as the oxidant gas. This shows a cross-sectional view of a fuel cell 1 according to a fourth embodiment. This shows an enlarged cross-sectional view of a fuel cell 1 according to a fifth embodiment. This shows an enlarged cross-sectional view of a fuel cell 1 according to a sixth embodiment. This shows an enlarged top view of a fuel cell 1 according to the sixth embodiment.

[0011] 1 is a cross-sectional view showing the structure of a fuel cell 1 according to a first embodiment of the present invention. The fuel cell 1 is composed of a porous support layer 2, a first electrode layer 3, an electrolyte layer 4, and a second electrode layer 5. The electrolyte layer 4 is sandwiched between the first electrode layer 3 and the second electrode layer 5.

[0012] When the first electrode layer 3, electrolyte layer 4, and second electrode layer 5 are made thin, e.g., 1 μm or less, for the purpose of manufacturing a high-power density cell, the porous support layer 2 serves to support the entire cell. The thickness of the porous support layer 2 is preferably 50 μm or more to ensure mechanical strength. Therefore, the actual thickness of the porous support layer 2 is 50 times or more the combined thickness of the first electrode layer 3, electrolyte layer 4, and second electrode layer 5. Furthermore, the porous support layer 2 has a porous structure with fine through-holes 6 in the Z direction, as shown in Figure 2 (described later), allowing gas to reach the first electrode layer 3. The Z direction is the direction from the bottom surface of the porous support layer 2 to the surface of the second electrode layer 5. The plane normal to the Z direction is defined as the XY plane.

[0013] The first electrode layer 3 is formed by a film formation process such as sputtering, but rather than forming the first electrode layer 3 all over the porous support layer 2, a region where the first electrode layer 3 is not formed is provided on the outer periphery of the porous support layer 2. This is because the first electrode layer 3 has a porous structure to supply gas to the electrolyte layer 4, and forming the first electrode layer 3 all over the porous support layer 2 could cause gas leakage between the top and bottom of the unit cell. Gas leakage can be prevented by leaving the outer periphery of the porous support layer 2 and covering it with the electrolyte layer 4, which has a dense structure.

[0014] When the first electrode layer 3 functions as the anode of the fuel cell, a fuel gas such as hydrogen is supplied to the first electrode layer 3. In this case, the first electrode layer 3 is made of a material such as a cermet of yttria-stabilized zirconia with an yttria composition ratio of 8% and nickel. The electrolyte layer 4 is made of an oxide having a perovskite structure, such as yttria-stabilized zirconia or lanthanum gallate. By forming the electrolyte layer 4 into a thin film of 1 μm or less using a film formation process such as sputtering, high power density power generation can be achieved. When the first electrode layer 3 functions as the anode, the second electrode layer 5 is made of a material such as silver, platinum, an alloy containing silver or platinum, a cermet of silver or platinum and gadolinium-doped ceria (GDC), or LSC ((La,Sr)CoO). The porous structure allows oxidant gas to reach the electrolyte layer 4. Alternatively, when the first electrode layer 3 is used as the cathode and the second electrode layer 5 is used as the anode, the order of forming the materials is reversed, and an oxidant gas is supplied to the first electrode layer 3 and a fuel gas is supplied to the second electrode layer 5. The following description is based on the premise that the fuel cell 1 is operated with the first electrode layer 3 as the anode and the second electrode layer 5 as the cathode.

[0015] FIG. 2 is an enlarged view of the fuel cell 1 shown in FIG. 1 . The porous support layer 2 has microscopic through-holes 6. If the microscopic through-holes 6 are circular, their diameters are preferably 300 nm or less. Even if the microscopic through-holes 6 are rectangular or have other irregular shapes, the maximum opening width (i.e., the longest part of the opening shape) is preferably 300 nm or less, more ideally 50 to 150 nm. If the opening width is too wide, short circuits are likely to occur, while if it is too narrow, the concentration of the gas flowing through the microscopic through-holes 6 decreases significantly, potentially resulting in a decrease in output power density. By achieving an ideal opening width, the thickness of the electrolyte layer 4 can be reduced to 1 μm or less, even 200 nm or less, thereby achieving a high output power density. The porous support layer 2 having such microscopic through-holes 6 can be fabricated by anodizing technology, such as anodic alumina (AAO). AAO is a porous aluminum oxide obtained by applying a voltage to aluminum immersed in an acidic solution. In the case of other materials, when a porous structure is formed by anodization, the material becomes a metal oxide, which is an insulator or semiconductor, and the electrical resistivity is generally 1 Ωcm or more.

[0016] 3A and 3B are cross-sectional SEM images of the first electrode layer 3. When the first electrode layer 3 is used as the anode, hydrogen is supplied from the back surface of the porous support layer 2 through the fine through-holes 6. The first electrode layer 3 has a porous structure so that hydrogen can reach the electrolyte layer 4. A porous electrode can be obtained by forming the layer using sputtering at a pressure of 5 to 10 Pa. When a porous structure is formed at a pressure of 5 to 10 Pa, fine, sharp-angled irregularities appear on the surface of the first electrode layer 3, as shown in FIG. 3A. If the electrolyte layer 4 is formed as a thin film in this state, the irregularities on the surface of the first electrode layer 3 may lead to short-circuit defects. Therefore, before forming the electrolyte layer 4, the first electrode layer 3 is planarized by ion milling or the like, as shown in FIG. 3B. This prevents short-circuit defects and enables high output power density, even when the electrolyte layer 4 is 1 μm or less, or even 200 nm or less.

[0017] 4 is a schematic diagram showing the process of planarizing the first electrode layer 3 by ion milling. When planarizing by ion milling, it is desirable to irradiate ions from an oblique direction. For example, if there are locally protruding portions of the electrode as shown in FIG. 4, the protruding portions can be selectively ground by irradiating ions from an angle as close to directly lateral as possible (60 degrees or more when the Z direction is set to zero degrees). In this case, to prevent ion irradiation from only a specific direction, ions are irradiated while rotating the first electrode layer 3 around the Z direction as a rotation axis.

[0018] The means for planarization is not limited to ion milling, and polishing by CMP (Chemical Mechanical Polishing) or the like may also be performed. In the case of CMP, the surface of the first electrode layer 3 can be planarized by polishing from directly above the shape shown in FIG.

[0019] FIG. 5 shows a method for deriving the arithmetic mean roughness Ra. The degree of flatness of the first electrode layer 3 can be determined by measuring the surface roughness. Among the parameters indicating surface roughness, the arithmetic mean roughness Ra and the maximum height Rz are important in the present invention. When the Z-direction coordinate of the outermost surface of the first electrode layer 3 is z = f(x), and the mean line m is defined so that the value of Ra calculated using the formula in the figure is minimized, the value of Ra is the arithmetic mean roughness. If the arithmetic mean roughness Ra is large, defects are more likely to occur when the electrolyte layer 4 and the second electrode layer 5 are formed on the first electrode layer 3. To improve calculation accuracy, the value of L is preferably equal to or greater than the thickness of the electrolyte layer 4 but not greater than 10 times the thickness of the electrolyte layer 4.

[0020] FIG. 6 shows an example of how to derive the arithmetic mean roughness Ra when the surface of the first electrode layer 3 is irregular. The arithmetic mean roughness Ra can be derived not only when z = f(x) is regular, but also when the surface is irregular. FIG. 6 shows such an example. Even if the line z = 0 is positioned at an arbitrary position, the total area Ap of the parts where z = f(x) is positive and the total area An of the parts where z = f(x) is negative can be extracted by image recognition or the like. When the position z = 0 is defined so that the value of Ra in the formula in the figure is minimum, the value of (Ap + An) / L ​​becomes the arithmetic mean roughness Ra.

[0021] Fig. 7 shows an example of how to derive the arithmetic mean roughness Ra when the surface of the first electrode layer 3 has holes resulting from the micro through-holes 6. The first electrode layer 3 is formed on the porous support layer 2, but if the porous support layer 2 is produced by anodization, the opening width of the micro through-holes 6 will vary, and the first electrode layer 3 may be interrupted where the opening width is wide, as shown in Fig. 7. In this case, the interrupted portions can be connected by straight lines, and Ra can be calculated in the same way as in Fig. 5 or Fig. 6.

[0022] FIG. 8 shows an example of deriving the arithmetic mean roughness Ra when the surface of the first electrode layer 3 includes a gentle slope. In anodizing technology, in addition to irregularities on the order of several tens of nanometers, relatively gentle slopes of 1 μm or more may occur. In such cases, the surface of the first electrode layer 3 also includes a gentle slope, as shown in FIG. 8. Since gentle slopes are unlikely to lead to defects, in this case the surface shape is separated into a waviness curve and a roughness curve. The waviness curve is a component with a gentle slope, while the roughness curve is a component that leads to defects. The arithmetic mean roughness Ra is calculated only for the roughness curve from which the waviness curve component has been removed, as in the case of FIG. 5 or FIG. 6 , and the value of Ra calculated from the roughness curve is the arithmetic mean roughness Ra for the case of FIG. 8 . In the example shown in FIG. 8 , since the roughness curve is irregular, the method of FIG. 6 may be used.

[0023] The process of separating the surface shape into a waviness curve and a roughness curve may be performed based on the period. For example, the Japanese Industrial Standards (JIS) defines a waviness curve as a component of 80 μm or more when the arithmetic mean roughness Ra is 20 μm or less. In the present invention, even a waviness component with a period of 80 μm or less may lead to a defect when the film thickness of the electrolyte layer 4 is 200 nm or less. Therefore, it is desirable to distinguish between a waviness curve and a roughness curve based on the film thickness of the electrolyte layer 4. Specifically, it is desirable to separate a periodic component of 10 times or more the film thickness as a waviness curve. The waviness curve is not limited to a curve, and may be, for example, a straight line that always rises to the right. In this case, the waviness curve is treated as a waviness curve with a period of L or more, and after separating the waviness curve, the Ra is calculated for the remaining roughness curve.

[0024] Depending on the state of the porous support layer 2, it may be possible to combine the cases of Figure 7 and Figure 8. In that case as well, the holes are connected by straight lines, the gently sloping components are separated as a waviness curve, and Ra is calculated using only the roughness curve.

[0025] FIG. 9 shows a method for deriving the maximum height Rz of the surface profile of the first electrode layer 3. The difference between the maximum and minimum values ​​of z within the range is the maximum height Rz. The value of the maximum height Rz does not depend on the reference position of z = 0. When the electrolyte layer 4 is formed by sputtering, it is difficult to form the electrolyte layer 4 on the side surfaces of protrusions, so if there are localized protrusions as shown in FIG. 9, defects are likely to occur regardless of the value of the arithmetic mean roughness Ra. In particular, if the value of Rz is larger than the film thickness of the electrolyte layer 4, the possibility of defects is extremely high.

[0026] Figure 10 is an SEM image showing protrusions on the porous support layer 2. In addition to foreign matter, protrusions on the porous support layer 2 can also be a factor in increasing Rz, and Figure 10 shows an example of this. When the porous support layer 2 is produced by anodization, particularly large protrusions may occur. If the first electrode layer 3 is formed in this state, the shape of the porous support layer 2 is directly reflected in the shape of the first electrode layer 3, and large protrusions remain on the surface of the first electrode layer 3. Although this problem can be solved by planarizing the surface of the first electrode layer 3, it is also effective to grind the protrusions on the porous support layer 2 before forming the first electrode layer 3. As with the planarization of the first electrode layer 3, this can be done by ion milling or polishing such as CMP.

[0027] Figure 11 shows an example of planarizing the porous support layer 2 by ion milling. The left side of Figure 11 is a cross-sectional SEM image of the porous support layer 2 before planarization. When planarization is performed until all large protrusions are removed, the entire surface of the porous support layer 2 appears as shown on the right side of Figure 11.

[0028] 12A is a cross-sectional view showing an example of the structure of a fuel cell stack 1200. The fuel cell stack 1200 is composed of a fuel cell 1, a conductive substrate 1201, a separator 1202, a lower current collector 1203, an upper current collector 1204, a cell gasket 1205, an upper electrode plate 1206, a lower gasket 1207, an upper gasket 1208, a bottom jig 1209, a top jig 1210, support columns 1211, and fastening members 1212. A unit stack 1213 is composed of, from the bottom, the separator 1202, the lower current collector 1203, the conductive substrate 1201, the fuel cell 1, and the upper current collector 1204, with the cell gasket 1205 disposed around the outer periphery of the fuel cell 1 and the upper current collector 1204. Fuel cell stack 1200 is assembled by stacking, from the bottom, bottom jig 1209, lower gasket 1207, multiple unit stacks 1213, upper electrode plate 1206, upper gasket 1208, and top jig 1210. Holes are drilled in bottom jig 1209 and top jig 1210 to pass support posts 1211 through, and both ends of support posts 1211 are machined to match the shape of fastening members 1212; for example, if fastening members 1212 are nuts, support posts 1211 are provided with threads of the same shape as fastening members 1212, and by fastening fastening members 1212, the entire fuel cell stack 1200 is fastened from above and below.

[0029] This structure prevents the fuel gas and oxidant gas used in the operation of the fuel cell 1 from leaking outside the fuel cell stack 1200. The conductive substrate 1201 and the separator 1202 are joined by welding or the like to prevent gas leakage to the outside and electrically connect the separator 1202 to the first electrode layer 3 of the fuel cell 1. By shaping the separator 1202 so that a recess is formed on the inside, leaving the edges intact, a gas flow path is formed that allows gas to be supplied to the through-holes in the conductive substrate 1201. Inserting the lower current collector 1203 into this gas flow path reduces the resistance of the electrical connection between the conductive substrate 1201 and the separator 1202. Furthermore, the second electrode layer 5 is electrically connected to the separator of the unit stack 1213 one level above via the upper current collector 1204, and multiple unit stacks 1213 are connected in series. The second electrode layer 5 of the topmost unit stack 1213 is electrically connected to the upper electrode plate 1206 via the upper current collector 1204, and the generated power can be extracted to the outside. The output current does not depend on the number of unit stacks 1213, but the output voltage is proportional to the number of unit stacks 1213, so the output power is determined by the product of the output current and the output voltage.

[0030] 12B is an exploded view of each layer of the fuel cell stack 1200. For simplicity, only one unit stack 1213 is shown, but the same applies when there are multiple unit stacks 1213, except that the unit stacks 1213 are inserted below the upper electrode plate 1206.

[0031] The conductive substrate 1201 has a first gas inlet 1201 a, a first gas outlet 1201 b, a second gas inlet 1201 c, and a second gas outlet 1201 d. The first gas is a fuel gas or an oxidizing gas, and the second gas is an oxidizing gas when the first gas is a fuel gas, or is a fuel gas when the first gas is an oxidizing gas.

[0032] Similar to the conductive substrate 1201, (a) the separator 1202 has a first gas inlet 1202a, a first gas outlet 1202b, a second gas inlet 1201c, and a second gas outlet 1202d, (b) the cell portion gasket 1205 has a first gas inlet 1205a, a first gas outlet 1205b, a second gas inlet 1205c, and a second gas outlet 1205d, and (c) the upper electrode plate 1206 has a first gas inlet 1206a, a first gas outlet 1206b, a second gas inlet 1206c, and a second gas outlet 1206d.

[0033] By providing a first gas inlet 1207a or 1208a, a first gas outlet 1207b or 1208b, a second gas inlet 1207c or 1208c, and a second gas outlet 1207d or 1208d in at least one of the lower gasket 1207 and the upper gasket 1208, the first gas and the second gas can be supplied to the fuel cell 1 from the bottom jig 1209 or the top jig 1210. Gases can also be supplied from both the bottom jig 1209 and the top jig 1210. The example in FIG. 12B shows a configuration example in which all layers have inlets and outlets on four sides. Below, a case will be described as an example in which the first gas is supplied from the first gas inlet 1207a of the lower gasket and the second gas is supplied from the second gas inlet 1208c of the upper gasket.

[0034] The first gas supplied through the first gas inlet 1207a of the lower gasket flows into the first gas inlet 1202a of the separator. Because the first gas inlet 1202a of the separator has a notch, the first gas passes over the separator 1202 and heads toward the first gas outlet 1202b. The lower current collector 1203 is located at the point where the first gas passes. If the lower current collector 1203 has a gas-permeable structure, such as a mesh structure, the gas flows through the lower current collector 1203 and is supplied to the back surface of the fuel cell 1 via the through-holes in the conductive substrate 1201. The first gas supplied to the back surface of the fuel cell 1 is supplied to the first electrode layer 3 via the fine through-holes 6, and further, if the first electrode layer 3 has a porous structure, it is supplied to the interface between the first electrode layer 3 and the electrolyte layer 4.

[0035] The first gas that reaches first gas outlet 1202b of the separator flows toward first gas outlet 1207b of the lower gasket and is discharged to the outside of fuel cell stack 1200 via bottom jig 1209. Since there are actually multiple unit stacks 1213, the first gas flows upward, passing through first gas inlet 1201a of the conductive substrate and first gas inlet 1205a of the cell part gasket in that order.

[0036] The second gas supplied from the second gas inlet 1208c of the upper gasket flows into the second gas inlet 1206c of the upper electrode plate 1206. The second gas then flows toward the second gas inlet 1205c of the cell gasket. Because the second gas inlet 1205c of the cell gasket has a notch, the second gas passes over the fuel cell 1 on the conductive substrate 1201 and flows toward the second gas outlet 1205d of the cell gasket. The upper current collector 1204 is located on the fuel cell 1, and the second gas flows through the upper current collector 1204 in the same way as the lower current collector 1203, allowing the second gas to be supplied to the second electrode layer 5 on the surface of the fuel cell 1. Furthermore, by making the second electrode layer 5 porous, the second gas can be supplied all the way to the interface between the second electrode layer 5 and the electrolyte layer 4.

[0037] The second gas that reaches the second gas outlet 1205d of the cell section gasket passes through the second gas outlet 1206d of the upper electrode plate 1206 to the second gas outlet 1208d of the upper gasket, and is discharged to the outside of the fuel cell stack 1200 via the top surface jig 1210.

[0038] 13 is an enlarged cross-sectional view of a fuel cell 1 according to a second embodiment of the present invention. In the second embodiment, the porous support layer 2 has specific protrusions, which affect the first electrode layer 3, resulting in a shape with convex portions.

[0039] FIG. 14 is an enlarged cross-sectional view of a fuel cell 1 when an electrolyte layer 4 is formed on a porous support layer 2 having protrusions. When the porous support layer 2 has protrusions as shown in FIG. 13 and the electrolyte layer 4 is formed by a method with poor step coverage, such as sputtering, the thickness tb of the electrolyte layer 4 in the direction perpendicular to the tangent to the surface of the protrusion at point B on the slope of the protrusion is thinner than the thickness ta at other locations, resulting in localized thin electrolyte layer 4 thickness. Even at point B, the thickness in the Z-axis direction is equal to ta, but the current flows through a path with low resistance, so it flows in the shortest direction from point B to the surface of the first electrode layer 3. When viewed as a whole, current tends to concentrate in areas with locally thin thickness. Because the electrolyte layer 4 does not exhibit ionic conductivity at room temperature, only electron leakage current flows. Even in this case, if there are areas where the electrolyte layer 4 is thin, the electron leakage current increases at those locations, compromising long-term reliability.

[0040] Furthermore, depending on the degree of the protrusion, tb in Fig. 14 may become extremely thin, or the first electrode layer 3 and the second electrode layer 5 may short-circuit. If a short-circuit occurs, the fuel cell will not function and will be defective. If the thickness becomes extremely thin without shorting, not only will long-term reliability be reduced due to the aforementioned electron leakage current, but there is also the possibility of defects due to insulation breakdown.

[0041] Figure 15 shows the results of measuring the electron leakage current when a 200 nm-thick electrolyte layer was formed on a silicon substrate without through-holes or steps, and a voltage was applied between the first electrode layer 3 and the second electrode layer 5. The vertical axis represents the current value divided by the area of ​​the second electrode layer 5. As the voltage increases, the electron leakage current also increases, eventually resulting in breakdown at a voltage of around 10 V. During operation of the fuel cell 1, a voltage of up to 1.2 V or more may be applied to the electrolyte layer 4, so breakdown will occur if there are any locations where the film thickness is 24 nm or less. Even if the thickness does not result in breakdown, the electron leakage current will increase, reducing long-term reliability.

[0042] To prevent this, as shown in Figure 13, regardless of where point B is placed on the slope resulting from the convex portion, the thickness tb between the first electrode layer 3 and the second electrode layer 5 at the shortest distance is equal to the thickness ta at the average location, thereby achieving a uniform thickness throughout the fuel cell 1. While achieving a completely uniform thickness is difficult due to the small irregularities on the surface of the first electrode layer 3 resulting from its porous structure for gas permeation, it is desirable to achieve a thickness that does not cause breakdown even at the thinnest point when 1.2 V is applied, and more ideally, does not result in an electron leakage current of 1 x 10^-3 A / cm^2 or more when 1.2 V is applied. In the results shown in Figure 15, the worst sample exhibited a value of 1 x 10^-3 A / cm^2 at 2.0 V. Since the thickness of the electrolyte layer 4 was 200 nm, the electric field strength was 0.01 V / nm. Therefore, even when the thickness of the electrolyte layer 4 is 200 nm or less, it is ideal for the thinnest point to be 120 nm or more.

[0043] While such a film formation is difficult using the sputtering method described in Embodiment 1, it is possible to form an electrolyte layer 4 with sufficient thickness in the vertical direction from the slope by using a method with good step coverage, such as atomic layer deposition (ALD). When the present invention is applied, protrusions remain on the surface of the electrolyte layer 4, and protrusions also form on the surface of the second electrode layer 5 formed thereon. However, leaving the protrusions on the surface of the second electrode layer 5 or flattening it as shown in FIG. 13 does not affect the defect rate. Protrusions on the surface of the first electrode layer 3 can be caused by protrusions on the porous support layer 2 as shown in FIG. 13 as well as foreign matter, but the present invention is not limited to slopes. For example, even if a surface that becomes a sidewall parallel to the Z axis occurs, no defects will occur as long as the film thickness in the vertical direction from the side is equivalent to that of other average locations and is maintained uniform throughout the fuel cell 1.

[0044] 16 is a cross-sectional view of a fuel cell 1 according to a third embodiment of the present invention. In the first and second embodiments, the fuel cell 1 is constructed by forming a first electrode layer 3, an electrolyte layer 4, and a second electrode layer 5 on a porous support layer 2. In the third embodiment, a first interface layer 7 is inserted between the first electrode layer 3 and the electrolyte layer 4, and a second interface layer 8 is inserted between the electrolyte layer 4 and the second electrode layer 5.

[0045] It is known that the higher the ionic conductivity of a material in contact with the electrode, the higher the output density. Therefore, by using a material with higher ionic conductivity than the electrolyte layer 4 for the first interface layer 7 and the second interface layer 8 (hereinafter, the two may be referred to as interface layers without distinction), the output power density can be improved. For example, GDC (gadolinium-doped ceria) has an ionic conductivity 10 times or more that of stabilized zirconia, making it suitable as a material for the interface layer. However, since GDC is reduced in a high-temperature hydrogen atmosphere and exhibits electronic conductivity, it is often used in a laminate with stabilized zirconia, which is not susceptible to reduction. At least one of the first interface layer 7 and the second interface layer 8 is inserted into the interface between the electrode layer and the electrolyte layer 4 using GDC. Alternatively, since GDC is also an electrolyte, the laminate of stabilized zirconia and GDC can be treated as an electrolyte layer. Even when an interface layer is inserted, if the surface of the first electrode layer 3 is made flat as in embodiment 1, defects can be prevented when the electrolyte layer 4 is made thin, such as 1 μm or less or 200 nm or less, and both an improvement in output power density and a low defect rate can be achieved. When the first electrode layer 3 is used as the anode, the first interface layer 7 is in a hydrogen atmosphere and is reduced depending on the operating temperature, resulting in electronic conductivity. Therefore, it is desirable that the surface of the first interface layer 7 is also flat. This is easy if the first electrode layer 3 is flat.

[0046] The insertion of the interface layer in this embodiment can be used in combination with Embodiment 2. In Embodiment 2, the surface of the first electrode layer 3 has a convex portion. However, when an electrolyte material is inserted as the first interface layer 7 or the second interface layer 8, one or both of the first interface layer 7 and the second interface layer 8 are included in the electrolyte film thickness of Embodiment 2, as long as they are not reduced in the operating environment and do not exhibit electronic conductivity. For example, when the first interface layer 7 is reduced in the operating environment, the total film thickness of the electrolyte layer 4 and the second interface layer 8 only needs to be 120 nm or more at its thinnest point. When neither the first interface layer 7 nor the second interface layer 8 is reduced in the operating environment, the total film thickness of the first interface layer 7, the electrolyte layer 4, and the second interface layer 8 only needs to be 120 nm or more at its thinnest point.

[0047] When this embodiment is applied, the electrolyte film thickness is substantially increased by the amount of the interface layer, and the interface layer is made of a material with higher ionic conductivity than the electrolyte layer 4, so defects can be prevented without impairing the output power density. Furthermore, the voltage that would be applied only to the electrolyte layer 4 if the interface layer were not present is distributed to both the interface layer and the electrolyte layer 4, thereby reducing electronic leakage current and improving long-term reliability. For example, at 450°C, the ionic conductivity of GDC is approximately 30 times that of stabilized zirconia. Therefore, if the stabilized zirconia film thickness is 190 nm and the GDC film thickness is 300 nm, the total film thickness can be 490 nm while maintaining the resistance to ionic conduction equivalent to that of stabilized zirconia 200 nm.

[0048] Figure 17 shows the results of measuring the electron leakage current when a voltage of 1 V was applied to samples with and without planarization, with the first interface layer 7 being 100 nm, the electrolyte layer 4 being 200 nm, and the second interface layer 8 being 200 nm. Twenty-four samples of each type were measured. When no planarization was performed, all samples had a leakage current of 1 x 10^-4 A / cm^2 or higher, with 30% experiencing short-circuit defects. When only the first electrode layer 3 was planarized (solid open squares and dotted open squares) and when only the porous support layer 4 was planarized (solid filled triangles and solid filled circles), approximately half of the samples had an electron leakage current of 1 x 10^-4 A / cm^2 or lower, and the number of defects was also reduced. The reduction in leakage current is due to the effect of planarization increasing the minimum film thickness in locally thin areas of the electrolyte layer 4. Planarizing the first electrode layer 3 has the effect of reducing the irregularities that accompany the formation of the porous electrode, as shown in Figure 3A, and planarizing the porous support layer 4 has the effect of reducing the irregularities of the underlying surface during the formation of the first electrode layer 3. Regarding short-circuit defects, the cause is that the electrolyte layer 4 is locally extremely thin or not formed locally, so planarizing the first electrode layer 3 and the porous support layer 2 reduced the frequency of these defects and therefore the defect rate. Furthermore, if the first electrode layer 3 is planarized by ion milling from an oblique direction, as shown in Figure 4, it may be possible to block any current paths that could cause short-circuit defects between the first electrode layer 3 and the second electrode layer 5.

[0049] When both the first electrode layer 3 and the porous support layer 2 were flattened (solid line with open triangles and circles, dotted line with open triangles and circles), the electron leakage current was reduced to a minimum of 1×10 A / cm, resulting in zero defects. It is believed that the combined effects described above prevented the occurrence of extremely thin areas in the electrolyte layer 4.

[0050] Fig. 18A shows the yield versus the arithmetic mean roughness Ra when defects are defined as those with an electron leakage current of 1 x 10^-3 A / cm^2 or more in the measurement results of Fig. 17. It can be seen that the larger the value of Ra, the lower the yield, and that the unevenness of the surface of the first electrode layer 3 is the cause of the defects.

[0051] Figure 18B shows the results of power generation testing of a non-defective battery using 3% hydrogen (nitrogen-based) as the fuel gas and air as the oxidant gas. The vertical axis represents an index called OCV (Open Circuit Voltage), which is the voltage when the current output is set to zero (open circuit state). Large Ra values ​​result in a decrease in OCV, which is caused by gas leakage. In the fuel cell 1, the fuel gas and oxidant gas are separated by the electrolyte layer 4. However, if the surface of the first electrode layer 3 is highly uneven, voids may remain in the electrolyte layer 4 formed on top of it. This results in direct contact between the fuel gas and oxidant gas, preventing them from contributing to the cell reaction and resulting in a decrease in OCV. Thus, if the surface of the first electrode layer 3 is highly uneven, the fuel cell 1 may malfunction even without a short circuit. By applying the present invention to flatten the surface of the first electrode layer 3, a high output power density can be achieved without defects, even when the electrolyte layer 4 is made thin.

[0052] <Fourth Embodiment> Figure 19 is a cross-sectional view of a fuel cell 1 according to a fourth embodiment of the present invention. While the first to third embodiments use aluminum oxide formed by anodization as the porous support layer 2, the fourth embodiment uses a silicon wafer as the porous support layer 2. Silicon wafers have superior surface flatness compared to those using anodization, allowing the electrolyte layer 4 to be made even thinner. In addition, the diameter can be easily increased, enabling a significant reduction in manufacturing costs.

[0053] The micro through holes 6 can be formed by forming openings 9 on the back surface of the silicon wafer by wet etching using an alkaline solution or the like, and then using a combination of photolithography and dry etching on a portion of the silicon wafer where the thickness is 10 μm or less. This separates the porous support layer 2 into a porous region 10 where the micro through holes 6 are lined up, and a support region 11 where the micro through holes 6 are not formed.

[0054] In this embodiment, a stress relief layer 12 is inserted between the porous support layer 2 and the first electrode layer 3. This is to prevent defects such as cracking of the first electrode layer 3 and the electrolyte layer 4 due to the difference in thermal expansion between the materials. The linear thermal expansion coefficient of silicon is approximately 4×10^-6 / K, roughly half that of aluminum oxide. The linear thermal expansion coefficient of the first electrode layer 3, a cermet of nickel and yttria-stabilized zirconia, is 10 to 17×10^-6 / K. This difference generates stress, which may lead to cracks in the thin film structure and result in destruction. Therefore, by inserting a stress relief layer 12 whose linear thermal expansion coefficient is between that of the porous support layer 2 and the first electrode layer 3, stress can be relieved. For example, if the linear thermal expansion coefficient of the first electrode layer 3 is 15×10^-6 / K, the stress relief layer 12 should preferably be made of a material with a linear thermal expansion coefficient of approximately 9.5×10^-6 / K, such as ferritic stainless steel. Electrical conductivity is also required when using other materials. The stress relaxation layer 12 has a porous structure for gas supply.

[0055] Similarly, thermal stress can be relieved by providing a second stress relief layer 13 on the second electrode layer 5. The fuel cell 1 is assembled into a stack in the manner shown in Figure 6, and the second electrode layer 5 contacts the separator 1202 or the upper electrode plate 1206 via the upper current collector 1204. For example, if the upper current collector 1204 is silver, the linear thermal expansion coefficient is approximately 19 x 10^-6 / K. Therefore, if the linear thermal expansion coefficient of the second electrode layer 5 is 10 x 10^-6 / K, the second stress relief layer 13 should preferably be made of a material with a linear thermal expansion coefficient of approximately 14.5 x 10^-6 / K. An example of such a material is a cermet material made by mixing silver and GDC.

[0056] Current collector wiring 15 is provided on the back surface of the porous support layer 2. The electrical resistivity of silicon is generally 1 to 10 Ωcm, and when the porous region 10 is 10 μm thick, the sheet resistance is 1 kΩ or more, resulting in significant loss in a thin-film structure with a current output of 1 A / cm^2 or more. The electrical resistance in the Z-axis direction in the porous region 10 is 1.25 mΩ per 1 cm^2 area when the electrical resistivity is 1 Ωcm and the porosity is 20%, allowing for low loss extraction to the back surface. Current flows in the X-axis and Y-axis directions via the current collector wiring 15, and is electrically connected to the separator 1202 via the convex portions of the support region 11. The current collector wiring 15 preferably has an electrical resistivity of 100 μΩcm or less and a thickness of 10 μm or more. It is desirable for the difference in linear thermal expansion coefficient between the current collector wiring 15 and the porous support layer 2 to be small. For example, an alloy of iron, nickel, and cobalt can be made to have a thermal expansion coefficient similar to that of silicon. Similarly, the conductive substrate 1201 in contact with the current collecting wire 15 and the separator 1202 in contact with the conductive substrate 1201 can be made of a material with a linear thermal expansion coefficient similar to that of silicon, thereby suppressing thermal stress between the respective components. In this case, the thermal stress between the separator 1202 and the upper current collector 1204 increases, but because both are 0.1 mm or more thick and the upper current collector 1204 has a mesh structure, damage caused by thermal stress is unlikely to occur. Furthermore, the current collecting wire 15 needs to be porous to allow gas to be supplied to the first electrode layer 3.

[0057] The methods for forming the first electrode layer 3, electrolyte layer 4, and second electrode layer 5 are the same as in embodiment 1, but stress relief slits 14 can be provided during formation by masking or other methods. This divides the first electrode layer 3, electrolyte layer 4, and second electrode layer 5, reducing their respective areas and suppressing thermal stress. In this embodiment, too, if the first electrode layer 3 is formed as a porous film, unevenness as shown in FIG. 3A will occur on the surface. By applying the present invention to flatten the surface, defects can be prevented even when the electrolyte layer 4 is a thin film with a thickness of 1 μm or less or 200 nm or less. Furthermore, a first interface layer 7 or a second interface layer 8 can be inserted in combination with embodiment 3.

[0058] Fifth Embodiment FIG. 20 is an enlarged cross-sectional view of a fuel cell 1 according to a fifth embodiment of the present invention. The porous support layer 2 is an insulator or a semiconductor. In the case of a semiconductor, the electrical resistivity is 1 Ωcm or more. Therefore, the front and back surfaces of the porous support layer 2 are not electrically connected or have high resistance. In this embodiment, therefore, an intra-through-hole conductive film 16 is formed on the inner walls of the micro through-holes 6 to achieve low-resistance electrical connection between the front and back surfaces of the porous support layer 2. The intra-through-hole conductive film 16 can be formed over the entire inner walls of the micro through-holes 6 of the porous support layer 2 by a method with good step coverage, such as ALD (Atomic Layer Deposition). In this case, the intra-through-hole conductive film 16 is also formed on the front and back surfaces of the porous support layer 2, as shown in FIG. 20 . The front surface side is more easily electrically connected to the first electrode layer 3, and the back surface side is more easily electrically connected to the conductive substrate 1201.

[0059] The thickness of the conductive film 16 in the through-holes must be less than half the width of the through-holes 6 so as not to fill the micro-through-holes 6, and specifically, a thickness of 100 nm or less is desirable. Furthermore, a minimum thickness of 20 nm or more is desirable to ensure reliable electrical connection. Examples of materials for the conductive film 16 in the through-holes include metals such as nickel and platinum, and conductive oxides such as zinc oxide and ruthenium oxide. However, when the first electrode layer 3 is used as a cathode, an oxidant gas is supplied, and therefore the conductive film must not lose its conductivity in the operating environment; for example, nickel is unsuitable.

[0060] After the formation of the through-hole conductive film 16, the first electrode layer 3 is formed in the same manner as in the first embodiment, and then planarized by ion milling or the like, thereby preventing defects that may occur when the electrolyte layer 4 is thinned. Furthermore, by applying this embodiment, the high output power density achieved by the thin electrolyte layer 4 can be extracted to the outside through a low-resistance path, thereby reducing power loss.

[0061] 21A is an enlarged cross-sectional view of a fuel cell 1 according to a sixth embodiment of the present invention. In this embodiment, the first electrode layer 3 is not formed on the protrusions of the porous support layer 2.

[0062] 21B is an enlarged top view of the fuel cell 1 in embodiment 6. Since the first electrode layer 3 is not formed on the protrusions of the porous support layer 2, the protrusions are exposed in those areas.

[0063] 21A and 21B, defects can be prevented even when the electrolyte layer 4 is formed to a thickness of 1 μm or less or 200 nm or less on the first electrode layer 3. Such a structure can be fabricated by (a) masking using photolithography when forming the first electrode layer 3, or (b) forming the entire layer once and then grinding it by ion milling from an oblique direction as shown in FIG.

[0064] 21A and 21B, the first electrode layer 3 is interrupted midway, but the arithmetic mean roughness Ra is calculated by connecting the ends of the interrupted portion with a straight line, as shown in Fig. 7. Fig. 21A is a cross-sectional view taken along dashed line A in Fig. 21B. Even if the first electrode layer 3 is interrupted at a specific cross-section, such as cross-section AA, the first electrode layer 3 is connected at other locations, so no problems arise with electrical connection.

[0065] Seventh Embodiment In a seventh embodiment of the present invention, hydrogen generation using a fuel cell 1 will be described. When the electrolyte layer 4 is made of yttria-stabilized zirconia, hydrogen generation using the fuel cell 1 is also possible. Specifically, when water vapor is supplied to the first electrode layer 3 or the second electrode layer 5, a voltage is applied so that the potential of the electrode layer to which water vapor is not supplied is increased, thereby electrolyzing the water vapor and generating hydrogen. In this case, the thinner the electrolyte layer 4, the lower the power consumption relative to the amount of hydrogen generated, and the greater the amount of hydrogen generated for a given amount of surplus power. Therefore, by using the present invention to make the electrolyte layer 4 1 μm or less or 200 nm or less, highly efficient hydrogen generation is possible.

[0066] This hydrogen generation operation is possible with a fuel cell 1 having the same configuration as in the first embodiment shown in Fig. 1. When assembling into a stack in the same manner as in Fig. 5, hydrogen generation is possible by increasing the potential of the electrode layer to which water vapor is not supplied while water vapor is being supplied to the first electrode layer 3 or the second electrode layer 5. In other words, one fuel cell 1 can perform both power generation and hydrogen generation operations.

[0067] The power generation operation and hydrogen generation operation can be switched freely. For example, when the power demand is high, fuel gas and oxidant gas are supplied to perform the power generation operation, and when the power demand is low, water vapor is supplied and the surplus power is used to perform the hydrogen generation operation. The water vapor used for hydrogen generation can be vaporized using the exhaust heat of fuel cell operation, which is a high temperature of 450°C or higher, or the water vapor generated when fuel gas is consumed during power generation operation can be used.

[0068] The surplus power can be generated by a solar power generation facility or the like, or it can be divided into two parts: one for power generation by simultaneously operating multiple fuel cells 1, and the other for hydrogen generation. For example, if the power demand is sufficient without operating all of the multiple fuel cells 1, the remaining fuel cells 1 can be operated to generate hydrogen, and the surplus power can be used to generate hydrogen according to the power demand. Fuel cells 1 using yttria-stabilized zirconia have a high operating temperature and require time to wait for temperature changes before restarting after a complete shutdown. Therefore, when power demand decreases, they can be operated efficiently by switching to hydrogen generation without changing the temperature. According to the present invention, defects that occur when the electrolyte layer 4 is 1 μm or less or 200 nm or less can be prevented. During power generation, the amount of power generated by hydrogen consumption increases, and during hydrogen generation, the amount of hydrogen produced relative to power consumption increases, resulting in improved energy utilization efficiency.

[0069] <Regarding Modifications of the Present Invention> The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0070] In the above embodiment, the arithmetic mean roughness of the surface of the first electrode layer 3 is preferably 100 nm or less, and more preferably 20 nm or less. Similarly, the arithmetic mean roughness of the surface of the porous support layer 2 is preferably 20 nm or less.

[0071] 1: Fuel cell 2: Porous support layer 3: First electrode layer 4: Electrolyte layer 5: Second electrode layer 6: Micro through-holes

Claims

1. A fuel cell comprising: a porous support layer having through holes and made of an insulator or semiconductor; a first electrode layer disposed on said porous support layer; an electrolyte layer disposed on said first electrode layer; and a second electrode layer disposed on said electrolyte layer, wherein the surface of said first electrode layer is flat.

2. A fuel cell comprising: a porous support layer having through holes and formed from an insulator or semiconductor; a first electrode layer disposed on the porous support layer; an electrolyte layer disposed on the first electrode layer; and a second electrode layer disposed on the electrolyte layer, wherein the electrolyte layer has a uniform thickness.

3. The fuel cell according to claim 1, wherein the arithmetic mean roughness of the surface of the first electrode layer is 100 nm or less.

4. The fuel cell according to claim 1, wherein the arithmetic mean roughness of the surface of said first electrode layer is 20 nm or less.

5. The fuel cell according to claim 1 or 2, wherein the electrolyte layer has a thickness of 1 μm or less.

6. The fuel cell according to claim 1 or 2, wherein the longest part of the opening shape of said through-holes has a length of 300 nm or less.

7. The fuel cell according to claim 1 or 2, wherein the surface of the porous support layer is flat.

8. The fuel cell according to claim 7, wherein the arithmetic mean roughness of the surface of said porous support layer is 20 nm or less.

9. The fuel cell according to claim 1 or 2, characterized in that the porous support layer is formed from a metal oxide.

10. The fuel cell according to claim 9, wherein the porous support layer is formed of aluminum oxide.

11. A fuel cell according to claim 1 or 2, characterized in that the electrolyte layer is formed from zirconium oxide containing yttrium oxide or an oxide having a perovskite structure.

12. A fuel cell according to claim 1 or 2, characterized in that an interfacial layer having a higher ionic conductivity than the electrolyte layer is disposed at least either between the first electrode layer and the electrolyte layer or between the first electrolyte layer and the second electrode layer.

13. A fuel cell as described in claim 1 or 2, characterized in that the porous support layer has an opening on the surface opposite to the first electrode layer, the porous support layer has a porous region in which the through holes are arranged and a support region in which the through holes are not arranged, the porous region is arranged between the opening and the first electrode layer, and the support region is arranged between the part in which the openings are not formed and the first electrode layer, and the fuel cell further comprises wiring that electrically connects the surface of the porous region opposite to the first electrode layer to the surface of the support region.

14. A fuel cell according to claim 1 or 2, characterized in that the fuel cell further comprises a stress relaxation layer for relieving thermal stress, at least either between the porous support layer and the first electrode layer, or on the surface of the second electrode layer opposite the electrolyte layer.

15. The fuel cell according to claim 1 or 2, further comprising a slit penetrating through and dividing the first electrode layer, the electrolyte layer, and the second electrode layer.

16. The fuel cell according to claim 1 or 2, characterized in that a conductive film is formed on the surface of the inner wall of at least one of the through holes.

17. The fuel cell according to claim 2, wherein the electrolyte layer has a thickness of 120 nm or more at its thinnest point.

18. A fuel cell according to claim 1 or 2, characterized in that the fuel cell is configured to generate hydrogen when water vapor is supplied to one of the first electrode layer or the second electrode layer and a voltage is applied to the other electrode layer so that the other electrode layer has a higher potential than the first electrode layer or the second electrode layer.

19. A fuel cell stack comprising a plurality of fuel cells according to claim 1 or 2 stacked and connected to each other.

20. A method for manufacturing a fuel cell, comprising the steps of: forming a first electrode layer on a porous support layer having through holes; planarizing the surface of the first electrode layer; forming an electrolyte layer on the first electrode layer planarized by the planarizing step; and forming a second electrode layer on the electrolyte layer.

Images