JP7138960B2 - Solid oxide fuel cell anode material and solid oxide fuel cell using this anode material - Google Patents

Description

The present invention relates to an anode material for solid oxide fuel cells, and more particularly to an anode material for solid oxide fuel cells that significantly improves the power generation efficiency of the solid oxide fuel cell system as a whole. The invention also relates to solid oxide fuel cells using this anode material.

Solid oxide fuel cells have the highest power generation efficiency among fuel cells, and can be used to construct large-scale fuel cells. The heat efficiency can be further improved by, for example, using high-temperature exhaust heat to generate power. Since it is possible, research and development are proceeding in various fields.

One of the major goals of research and development of fuel cells is to improve power generation efficiency, and this is also a major issue for solid oxide fuel cells. In a fuel cell, the oxygen reduction reaction (reaction to create oxide ions from oxygen gas and take them into the solid electrolyte) that takes place on the cathode (oxygen electrode) side, and the hydrogen oxidation reaction that takes place on the anode (fuel electrode) side (protons are created and oxidized) It is extremely slow compared to the reaction that reacts with material ions and produces water molecules as a result), so it is important to improve the cathode to speed up this reaction to improve power generation efficiency. It has long been the consensus of those skilled in the art that it is far less important and less effective than improving cathode performance. In research on solid oxide fuel cells, improvements on the anode side have also been proposed (e.g., Non-Patent Document 1). I didn't.

In addition, the operating temperature of solid oxide fuel cells is about 1000°C, but in a relatively low temperature range of 800°C or less, the power generation efficiency drops significantly due to an increase in internal resistance. It has been thought that it should work with However, at high temperatures of 900°C or higher, the materials that can be used are limited. Especially when stacking single cells to create a stack cell for obtaining high output, lanthanum is the material for the interconnector that separates the single cells. Chromite-based oxides are mainly used. However, interconnectors using this material are very expensive, and this is one of the major factors preventing the price reduction of solid oxide fuel cells. If the operating temperature of solid oxide fuel cells can be lowered to around 650-700°C, it will be possible to use general-purpose stainless steel as the interconnect material, which is expected to significantly reduce fuel cell prices. there is

To this end, it is considered effective to reduce the thickness of the solid electrolyte to 20 microns or less to greatly reduce the internal resistance. Since it is not possible to form a free-standing membrane free of contamination, attempts have been made to form a ceramic membrane on an inexpensive anode material as a support, rather than an expensive cathode material.

However, the anode material usually consists of a mixture of Ni particles and oxide solid electrolyte particles. In this case, the Ni particles are connected and serve to transport electrons, and the oxide solid electrolyte particles (for example, yttria-stabilized zirconia particles, scandia-stabilized zirconia particles, etc. are exemplified) are connected and serve as a passage for diffusing oxide ions. have a role. When hydrogen gas comes into contact with the contact points of these two different particles, three types of hydrogen, electrons, and oxide ions meet at the same time, and the fuel cell reaction proceeds (this is called a three-phase interface, an active site in the anode). is thought to play a role).

However, in order to greatly increase the number of active sites, it is conceivable to reduce the grain size. Usually Ni oxide is used instead of Ni and is reduced at a temperature of 800° C. to form an anode layer consisting of a mixture of Ni particles and oxide solid electrolyte particles as described above.

Furthermore, when the anode layer is baked on ceramics, the baking treatment is usually performed at a temperature exceeding 1000° C., so the particle size of the Ni particles and the oxide solid electrolyte particles in the anode layer is reduced to about submicrons. However, it was difficult to dramatically improve the number of active sites compared to those in the conventional anode layer and to reflect the improvement effect in the power generation efficiency.

Due to the above-described significant drop in power generation efficiency during low-temperature operation and the difficulty of solving the problem, prospects for practical use of solid oxide fuel cells that exhibit high power generation efficiency at temperatures of 800° C. or less have not been clarified.

An object of the present invention is to greatly improve the power generation efficiency of the solid oxide fuel cell as a whole by improving the anode material.

According to one aspect of the present invention , a solid comprising zirconia or ceria, nickel and platinum, wherein platinum is present between said zirconia or ceria particles and said nickel particles in a manner not observed at a spatial resolution of 1 nm An oxide fuel cell anode material is provided.
Here, the observation may be performed using a scanning electron microscope with a spatial resolution of 1 nm.
The anode material for solid oxide fuel cells may also have a porous structure.
Also, the zirconia may be yttria-stabilized zirconia .
According to another aspect of the present invention, there is provided a solid oxide fuel cell using any of the solid oxide fuel cell anode materials described above.

According to the present invention, it is possible to remarkably improve the power generation efficiency of a solid oxide fuel cell by improving the anode material, which has been thought to be ineffective for improving the power generation efficiency of fuel cells. In addition, due to this efficiency improvement, even if the solid oxide fuel cell is operated at about 700 ° C to 800 ° C, it is possible to achieve higher power generation efficiency than before, so low temperature operation increases the choice of materials that can be used for manufacturing. It is also possible to realize cost reduction and the like.

FIG. 4 is a diagram for explaining the Pt cation-defective oxide-Ni cation cluster active site in the anode catalyst of the present invention in comparison with the three-phase surface active site of the prior art; 4 is a graph showing current density-cell voltage characteristics of solid oxide fuel cells of Examples and Comparative Examples of the present invention. The current density-cell voltage characteristics of the solid oxide fuel cell of the example of the present invention and the solid oxide fuel cell, which differ only in that NiOx in the anode layer was pre-reduced before sputtering of the PtOx thin film, are shown. , PtOx thin films with sputtering thicknesses of 10 nm, 20 nm and 30 nm. 4 is a graph showing current density at 800° C. vs. net cell voltage (IR-free) characteristics corrected for the effect of internal resistance of the solid oxide fuel cell of the example of the present invention. The figure explaining how to obtain|require IR-free by a current interruption method. Graph showing changes in IR free and cell voltage at a current density of about 250 mA cm ⁻² when the sputtering thickness of the PtOx thin film is 10 nm, 20 nm and 30 nm. The cell operating temperature is 800°C. Also shown are the results of measurements of a solid oxide fuel cell fabricated under the same conditions except that ceria nanowires were used instead of the 8YSZ powder as the material for the anode layer and the sputter thickness of the PtOx thin film was set to 10 nm. A scanning electron microscope (SEM) image of a cross-sectional surface of an example of the present invention (PtOx film thickness: 10 nm) after reduction treatment in a hydrogen atmosphere. The figure which shows the elemental mapping in the same position of the same sample as FIG. TEM bright-field image of the sample after PtOx thin film sputtering and before hydrogen reduction treatment. A STEM image of a sample after hydrogen reduction treatment at 800° C. for 1 hour. FIG. 9 shows an STEM image of the same sample as in FIG. 8 and its EDS elemental analysis result spectrum. A STEM image of the same sample as in FIG. 8 and its EDS elemental mapping image (Ni, Zr, Y). A STEM image of the same sample as in FIG. 8 and its EDS elemental mapping image (Pt, Zr, O).

According to one aspect of the present invention, in addition to nickel (Ni), an anode material added with platinum (Pt), which has been generally considered meaningless to use as an anode material for solid oxide fuel cells, is used. By using it, the efficiency of the solid oxide fuel cell is greatly improved.

More specifically, anode materials of the present invention include, for example, zirconia (ZrO ₂ ) (preferably yttria-stabilized zirconia (YSZ) is used) and nickel oxide (NiOx, where 1≦x≦4, hereinafter In the following, NiOx may be abbreviated as NiO) and a very thin platinum oxide (PtOx, where 0.5 ≤ x ≤ 2) is deposited on a mixture layer of a solid oxide (for example, YSZ) as an electrolyte. ) and heated in a reducing atmosphere containing hydrogen (diluted hydrogen diluted with helium or the like as in the example, or pure hydrogen may be used). A layer of a mixture of YSZ and NiOx (YSZ-NiO) on which no PtOx film is formed is heated in a reducing atmosphere to reduce NiO in the layer to Ni, thereby turning the porous anode into a solid oxide electrolyte. However, it has been recognized that the use of Pt in the anode layer of solid oxide fuel cells is not very effective. There were no examples of using it in any form to make an anode material.

In the anode material of the present invention, as will be understood from its unique microstructure, which will be described later, the amount of Pt in the total anode material is significantly reduced compared to other types of fuel cell electrode materials using Pt. Even if it is reduced, the catalytic activity is maintained at a high value. In the examples described below, the amount of Pt in the anode material was varied from 0.9% by weight to 0.3% by weight (30 nm to 10 nm as the thickness of the PtOx film formed in the example). Even at a very small amount of 0.3% by weight, no sign of deterioration in the power generation performance of the fuel cell was observed. Therefore, although the lower limit of the Pt amount specifically demonstrated in the examples is 0.3% by weight, it is believed that the anode material of the present invention exhibits sufficiently high performance even with a Pt amount less than this.

Here, the active sites formed in the anode material of the present invention will be explained in comparison with the active sites in the conventional anode material.

As described above, conventionally, the anode electrode reaction of a fuel cell is thought to occur at the site where electron-gas (hydrogen)-ion (oxide ion) meet, that is, the site called the three-phase interface. The position where the interface is formed is the contact portion between the oxide serving as the electrolyte and the electrode active material (Ni in the anode material of the conventional structure given as a comparative example in the examples). FIG. 1(a) conceptually shows the structure of a conventional anode without a PtOx thin film, that is, in which a conventional three-phase interface works as an active site. The left side of FIG. 1(a) shows a state in which a layer of a mixture of oxide electrolyte particles and NiOx particles is formed on the surface of the solid electrolyte. The right side conceptually shows the structure of the anode material after reduction treatment at 800° C. in a hydrogen atmosphere on the left side. This treatment reduces NiOx to metallic Ni. Since the volume is reduced at this time, a large number of pores are formed in the composite material of the oxide electrolyte and Ni after the reduction, and the composite material becomes porous. In this porous structure, a conventional three-phase interface is formed at the interface between Ni and the oxide electrolyte, which becomes the active site (indicated by the blank cross in the right side of FIG. 1(a)).

If the active site of the anode material of the present invention can also be explained by the above-mentioned conventional three-phase interface model, the presence of Pt is important in the anode material of the present invention. Fine particles of Pt should be present (because they are initially introduced into the anode in the form of oxide PtOx, but the subsequent reduction treatment in a hydrogen atmosphere should reduce not only NiOx but also PtOx). The Pt particles present at the three-phase interface in the electrode of the fuel cell exist in the form of fine particles with a diameter of at least several nanometers, as is well known, referring to the observation results of the electrode of the polymer electrolyte fuel cell, for example. Further, as an anode material containing Pt for a solid oxide fuel cell, for example, there is one disclosed in Non-Patent Document 1 already mentioned. It is shown that the Pt in the anode material is a particle with a diameter of about 50 nm to 100 nm. For details, refer to Non-Patent Document 2 published together with this as supplementary information of Non-Patent Document 1. B of Figure S6 of Non-Patent Document 2 shows a TEM image of the three-phase interface at the portion where SDC (samaria-doped-ceria) and Pt are in contact, and a part is visible in the upper left. It can be seen that the Pt particles have a diameter of at least 50 nm to 100 nm after being heat-treated at a temperature of about 900°C. In addition, the white and brightly shining platinum particles indicated as Pt in Figure S1 of the same document are considered to be particles after heat treatment at 900° C., referring to the caption of D in the same figure.

Therefore, even in the examples of the present invention, Pt fine particles having a diameter of at least several nanometers should be observed at the locations where the three-phase interfaces serving as active sites are formed.

However, as shown in the following examples, when the anode materials of the examples after completion of the reduction treatment in the hydrogen atmosphere were observed with a high-resolution SEM having a spatial resolution of 1 nm, particles with a size of about 1 nm were observed. If there is, it should be possible to observe at least something different from the surroundings, but no such signs were detected at all, just a very smooth surface was observed. Since F in Figure S1 of Non-Patent Document 2 mentioned above is also an SEM image, if there are Pt particles of about 1 nm or larger, even if the detailed shape of the particles is not known, As in Document 2, you should be able to see clearly dotted bright white masses. As a result of further research, it was found that Pt particles somewhat larger than 1 nm may be present on the surface where the oxide electrolyte particles in the anode material are exposed alone, but a three-phase interface that functions as an active site is formed. It was found that no such Pt particles were observed at the interface between the oxide electrolyte particles and the Ni particles that should have been formed. In addition, even if the Pt elemental map was observed with the same resolution, no Pt agglomerates could be found. Considering that this anode material is much more active than conventional anode materials without PtOx, so there should naturally be a large number of active sites on it, the above observations are in line with conventional three-phase It is a structure that cannot be explained by the interface model, and it suggests that there must be an active site that becomes a new three-phase interface that is much smaller than when a conventional three-phase interface structure is assumed. there is

Recently, non-patent document 3 published the results of research on a new active site structure in a Pt-supporting ceria electrode active material for a polymer electrolyte fuel cell. Non-Patent Document 3 discloses that Pt cations and cerium cations are bonded via lattice defects in ceria crystals to form clusters having extremely small sizes at the atomic level, and these clusters have strong electrochemical catalytic activity. It has been shown to exhibit As described above, in the anode material prepared according to the present invention, Pt particles with a diameter of about 1 nm or more were not observed at the interface between zirconia or ceria and nickel, and Pt did not form agglomerates on the anode material. A new three-phase cation-defective oxide cluster similar to that presented in Ref. It is reasonable to judge that it becomes an interface and is dispersed in a large amount.

This is conceptually shown in FIG. 1(b). The left side of FIG. 1(b) shows a state in which a PtOx thin film with a thickness of 10 nm is formed on the surface (lower side in the figure) of a layer with a thickness of about 40 μm formed of a mixture of an oxide electrolyte and NiOx. When this state is treated in a hydrogen atmosphere at 800° C. as in FIG. 1A, Pt cations diffuse from the PtOx thin film into the oxide electrolyte and NiOx layer. Since NiOx remains in the mixture layer during the reduction treatment in a hydrogen atmosphere,
_{PtOx +} Y2O3 stabilized ZrO2 ₊ NiO →
A reaction of Pt cation-defective oxide-Ni cation cluster formation proceeds, and as a result, at the end of the reduction treatment, as shown on the right side of FIG. (indicated by a blank asterisk ☆) is densely and uniformly formed. Since this cluster consists of a small number of atoms, its size is only a few Å. Therefore, these clusters cannot be detected at all even when observed with a spatial resolution of about 1 nm. For example, assuming a model in which a cluster of platinum cations, defective zirconia, and Ni cations exists in stabilized zirconia, the platinum cation is Pt ²⁺ and there are eight coordinated oxygen atoms around it (firefly stone ), the ionic radius of the platinum cation (Pt ²⁺ ) is considered to be about 0.7 Å (0.07 nm), so naturally it cannot be seen with a spatial resolution of about 1 nm.

It should be noted here that on the right side of FIG. 1(a), which conceptually shows the distribution of active sites (three-phase interfaces) in the prior art, the active sites are formed only at the contact portion between the oxide electrolyte and Ni. In contrast, the right side of FIG. 1(b) shows that the active sites of the present invention are distributed throughout the surface of the anode material (including the surface inside the pores of the porous body). This is explained below.

In the course of research conducted by the inventors of the present application, when the inside of the anode layer was carefully observed with an analytical TEM, it was found that the oxide component migrated to the Ni metal side, Ni cations diffused onto the oxide, and the region was found to form a cluster structure consisting of Ni and oxide components. See Non-Patent Documents 4 and 5 for details.

As shown in these non-patent documents, the width of the interface between Ni and oxide is at most 10 nm or less, but the width at which both components diffuse over the grain boundary at approximately the same distance is the width of the grain boundary. A total of 100 nm, which is about 10 times larger, was reached. In addition, it was thought that a superstructure (meaning a structure different from a matrix) was formed in the diffusion region, as expressed in non-patent literature. Based on previous simulations of defect structures by the inventors of the present application, this is also a cluster structure composed of cations and defect oxides, and has a C-type rare earth structure (the crystals that make up the matrix are stabilized zirconia or In the case of ceria, which has a fluorite crystal structure, it is known that the C-type rare earth crystal phase can thermodynamically coexist with this fluorite crystal phase, that is, they can coexist in the same phase diagram. ) were thought to have a similar arrangement. Thus, according to the conventional wisdom, the conventional three-phase interface exists only at the grain boundary between Ni and oxide, and the number of active sites is limited. On the other hand, however, there is also a cause of deteriorating the characteristics. In other words, the result contradicts the conventional technical common sense that Ni diffuses vigorously into the solid electrolyte membrane and creates an unfavorable state such as an electron conduction path in the solid electrolyte.

In the present invention, by allowing the Pt cation to participate in the formation of the cation-oxide interface in the anode layer, which conventionally played a role of deteriorating the anode characteristics, the above-mentioned long-distance migration of Ni and oxide can be achieved. It is thought that the number of new active sites (three-phase interfaces) was significantly improved only by creating a cluster structure that improves the electrode performance over a wider range, and as a result, it was possible to contribute to the improvement of power generation performance.

As described above, the inventor of the present application is not bound by the conventional stereotype that an active site is generated only in an extremely narrow place where Ni and oxide are adjacent to each other, and as a result of conducting research based on the above-mentioned experimental results, the following results are obtained. Addition of Pt, which enables the formation of new cluster structures on Ni and oxides, was conventionally thought to be meaningless from the viewpoint of improving fuel cell performance even when used in the anode layer. It was found that extremely high activity can be achieved by This finding made it possible for the first time to significantly improve the power generation performance by improving the anode layer.

Non-Patent Document 1 exists as one of the few prior art examples of using Pt as an anode material for a solid oxide fuel cell, so the difference between the present invention and the contents described in this document will be explained here.

First, it should be noted that in non-patent document 1, metal Pt is used instead of PtOx, which is an oxide of Pt, in fabricating the anode. As will be explained below, PtOx and NiOx are considered to increase the reaction activity with other particles when the oxide is reduced to metal particles, and the reduction of PtOx to platinum improves the reaction activity of Pt. effect, the effect of improving the reaction activity on the Ni surface due to the reduction of NiOx to Ni, and the effect of improving the reaction activity on the oxide particle surface due to the generation of oxygen defects on the oxide particle surface. It is thought that the new three-phase interface is formed widely and abundantly in the anode layer by combining the two-way diffusion phenomenon that occurs in a wide range between the phases. Here, if the treatment is carried out at an excessively high temperature, the bonding between Ni particles with increased surface reaction activity and the oxide particles with oxygen defects will break, resulting in the formation of new clusters over a wide area as described above. It is not preferable because it impairs the formation of phase interfaces (the novel three-phase interface in the present invention will be described in detail later).

Based on this idea, when platinum particles are sputtered from the beginning, there is no such reduction process from oxides to metal particles, but grain growth and sintering of platinum particles due to heat treatment at a high temperature of 800 ° C. Therefore, the reaction activity with other particles is not sufficiently increased, and only a moderate effect can be expected. Therefore, in order to increase the anode performance improvement effect to a measurable range, a large amount of platinum needs to be sputtered, and platinum, which is a rare metal, is used extensively, creating a state of low industrial utility value. put away.

In addition, to further explain the use of Pt in Non-Patent Document 1, in Non-Patent Document 1, metal Pt is applied by sputtering, but the film thickness is 10 nm to 30 nm in the examples of the present application described later (this film is Note that it is PtOx, not metallic Pt. When converted to metallic Pt, it is thinner), the thickness is 150 nm to 200 nm, which is about 10 times (non-patent document page 160, left column, second column in the "Method" section). paragraph). That is, in Non-Patent Document 1, it is confirmed that the electrode performance is improved by using a much larger amount of Pt than in the present invention. On the other hand, in the examples of the present application, although only a small amount of Pt, which is about 1/10, was used, a large improvement effect was observed even in the battery performance. It can be understood that the efficiency is high.

If the three-phase interface is explained here, the three-phase interface consists of electrons (a phase responsible for charge transfer, conventionally a path of Ni particles), a gas phase (hydrogen in the anode layer), and ions ( It refers to the interface where the phase responsible for the diffusion of oxide ions (conventionally, the path of oxide particles) is concentrated in one place. In the case of the present application, in addition to the conventional three-phase interface, platinum nanoparticles, which are considered to be below the detection limit of high-resolution SEM, are formed by reduction of platinum oxide. Together with the two-way diffusion phenomenon that occurs between the oxide particles with defects and the Ni particles whose surface reaction activity is increased by reduction, the three-phase interface that appears as the new highly active sites described above can be expanded over a wide area. could be formed. As a result, it is considered that the fuel cell power generation performance can be improved.

Here, the two-way diffusion described above and the three-phase interface formed at points other than the contact points between the Ni particles and the oxide particles will be described.

Non-Patent Document 6 discloses a Pt—CeO ₂ /C electrode material in which ceria nanowires impregnated with Pt are mixed with carbon black. Figure 3 shows an example of the result of TEM observation of the diffusion phenomenon from oxide particles (here, ceria) onto metal platinum, and Figure 12 shows an illustration of this diffusion phenomenon. It is shown. Although Pt is used as the metal in Non-Patent Document 6, it is considered that the same phenomenon occurs with Ni. As can be seen from these figures, the oxide particle component diffuses from the oxide particles onto the metal particles in the form of a film with a thickness of several nanometers. This diffusion proceeds while creating an interface structure (cluster structure). On the other hand, regarding the diffusion of the metal to the oxide particle side, in the Pt—CeOx nanowire electrode, the metal component diffuses to a slightly deeper range (considered to be about 10 nm) from the oxide particle surface. In addition, as described above with reference to Non-Patent Documents 4 and 5, equidistant diffusion is observed from both sides in the anode layer for a solid oxide fuel cell, so that oxidation Diffusion in the thin film state of the substance, and on the oxide side, the metal component diffuses from the surface to the inside (depth direction) and along the surface. Diffusion proceeds while forming a structure. It is believed that this diffusion range increases when platinum is added. Therefore, in both Ni particles and oxide particles, Pt cation-defective oxide-Ni cation clusters are formed everywhere as the diffusion progresses.

Also, although the metallic Ni particles are connected to each other, the surfaces thereof are partially covered with an oxide component thin film. Since this oxide thin film naturally has an oxygen defect structure, it can serve as a path for transporting oxide ions. As for electrons, they naturally move in Ni with low resistance and can encounter thin oxide films at various portions of the surface of the entire Ni particle. In addition, since this thin film should take the form of nonuniformly (in other words, partially) covering the Ni metal particle surface, the number of three-phase interfaces encountered by the gas, electrons, and ions on the Ni metal particle surface is This greatly increases compared to the case where only the contact points between the Ni particles and the oxide particles become the three-phase interfaces.

In addition to this, considering the electron conduction path on the oxide particles, the metal component diffuses on the oxide particle side as described above. This diffusion is considered to be of two types, that is, diffusion from the surface to the inside and diffusion that spreads along the surface. In either case, the metal cations form clusters. That is, the Pt ²⁺ site and the Ni ²⁺ site are Zr ⁴⁺ and Y ³⁺ in the crystal lattice of the oxide (Since Y ₂ O ₃ is a solid solution in the ZrO ₂ lattice of the stabilized zirconia, tetravalent and trivalent cations coexist) or Ce ⁴⁺ , Ce ³⁺ (ceria usually consists of tetravalent and trivalent cerium). In this way, divalent cations, trivalent cations, and tetravalent cations are constituent components of newly formed clusters, and it is thought that these different valences coexist in the cluster (cluster The overall electric charge is, of course, compensated by the amount of oxygen defects and the like so that the overall is neutral). It is generally known that charge transfer is promoted on the surface of materials with different valences. , the oxide ions also diffuse in the immediate vicinity, so that both the electrons and the oxide ions can reach very close locations of the clusters. Furthermore, if the anode layer has good porosity, gas (hydrogen) will also contact many of the places electrons and oxide ions can reach. Since this state occurs everywhere near the surface, it is thought that a new three-phase interface state is formed there, and as a result, even on the oxide surface, the number of three-phase interfaces encountered by gases, electrons and ions is naturally greatly increased. To increase.

Therefore, in the anode material of the present invention, the surface of either Ni or oxide can have three-phase interfaces functioning as active sites everywhere, and thus achieves extremely high activity compared to conventional anode materials. It becomes possible to

Here, before forming the PtOx thin film, NiO was completely reduced to metallic Ni by heating in a hydrogen atmosphere. Although it depends on the sputtering thickness, when the power generation performance of a solid oxide fuel cell single cell using this is measured, it is at most the same level of power generation performance as compared to the case where the PtOx thin film is sputtered before reduction. Or, in some cases, the performance was extremely degraded. When considering the conventional three-phase interface model, if the PtOx thin film is formed after the porous structure is completed by reducing NiOx, PtOx will enter every corner of the already completed porous structure from the beginning. should form more three-phase interfaces, resulting in a more active anode material. However, contrary to this prediction, experimental results were obtained in which no improvement in performance was observed. can be done.

The phenomenon that the power generation performance does not improve even if the NiOx in the anode layer is pre-reduced as described above, and in some cases, the power generation performance is extremely lowered is explained as follows.

Even when PtOx is sputtered on the anode layer, if the anode is reduced before this sputtering treatment, the above-mentioned reaction activity improvement on the Ni surface in the process of reducing NiO to Ni is improved. effect is reduced. Therefore, even if the subsequently formed PtOx film is subjected to a reduction treatment, a sufficient effect on the formation of a new active three-phase interface cannot be obtained. The power generation performance improvement effect obtained from the reduced anode layer is reduced.

This negative effect appears in the form that it becomes more difficult to form new composite cluster active sites where high activity can be expected in regions where the thickness of PtOx is thinner, that is, the amount of platinum sputtered is smaller, resulting in a significant decrease in the effect. In conclusion, it is preferable to sputter PtOx onto NiO-oxide in order to minimize the use of expensive and rare metals and to have the greatest effect on fuel cell power generation.

As shown in the reaction formula shown below in FIG. 1(b), a platinum cation-defective oxide-Ni cation cluster structure is formed by a similar reaction even if ceria is used instead of zirconia as the oxide electrolyte. resulting in a highly active anode. That is, a solid oxide fuel cell was fabricated under the same conditions except that ceria nanowires were used instead of 8YSZ, and the cell voltage and IR free at a current density of 250 mAcm ⁻² were measured. As can be seen, it was confirmed that the power generation performance is very close to that of using 8YSZ as an anode component.

In addition, the oxide electrolyte is usually used in the form of powder, and the fine particles that constitute this powder may be in the form of spheres or other low aspect ratios as in the case of ordinary powders, or may be in the form of particles as described above. , may be of high aspect ratio, such as nanowires. Therefore, it should be noted that the term "powder" in the present application does not exclude cases where some or all of the fine particles that constitute it are of high aspect ratio, such as nanowires.

[Preparation, Measurement and Evaluation of Anode Material of the Present Invention and Solid Oxide Fuel Cell Using the Same]
In the examples described below, comparative examples and examples were prepared by forming a conventional cathode material and an anode material of the example of the present invention on two surfaces of solid oxide fuel cell solid electrolyte pellets. The characteristics of a solid oxide fuel cell single cell using

Specifically, the single cell of the comparative example was produced as follows. First, an anode (NiO-8YSZ) slurry was applied by screen printing to one side of a zirconia (8YSZ) sintered body pellet (thickness: 500 μm, diameter: 10 mm) in which 8 mol % yttria was dissolved, and dried. . The composition of the solid content in this anode (NiO-8YSZ) slurry was NiO:8YSZ=4:1 (weight ratio). After that, baking treatment was performed in the air at 1200° C. for 1 hour, and slowly cooled to room temperature. Next, a slurry of cathode ((La _0.80 Sr _0.20 ) _0.95 MnO _3-x : manufactured by Fuel Cell Materials, product name LSM20-I) was applied to the other surface, dried, and dried. Baking treatment was carried out in air for 1 hour at a temperature of °C. After baking, the NiO in the anode layer was reduced using 4% hydrogen (helium diluted) gas at a temperature of 800°C. Since large voids were generated in the anode layer when NiO was reduced to Ni, a solid oxide fuel cell single cell having a porous anode layer was obtained.

On the other hand, in producing the examples of the present invention, first, as in the comparative example, NiO-8YSZ (NiO:8YSZ = 4:1 (weight ratio)) anode was baked into pellets of the same type and thickness. A cathode ((La _0.80 Sr _0.20 ) _0.95 MnO _3-x : manufactured by Fuel cell materials, product name LSM20-I) slurry was applied to one surface, dried, and then heated to 1100°C. Baking treatment was performed in the air for 1 hour. After that, a platinum oxide (PtOx) film was formed on the NiO-8YSZ anode. A PtOx film was formed on the anode by magnetron sputtering using a platinum target (purity: 99.95%) with a diameter of about 5 cm. Here, a sample was placed at a position 90 mm away from the platinum target, and DC sputtering was performed at room temperature under an oxygen partial pressure of 4×10 ⁻¹ Pa using a 10 W DC power supply. The deposition of PtOx on the anode was performed at a rate of about 2 nm per minute. After forming a PtOx film with a thickness of 10 nm on the anode layer, NiO was reduced using 4% hydrogen (helium diluted) gas at a temperature of 800° C. to make the anode layer porous as in the comparative example. . In this embodiment, a direct-current sputtering apparatus is used for sputtering, but any film-forming means such as an alternating-current sputtering apparatus can be used as long as a required PtOx film can be formed.

As can be seen from the above description, all the PtOx film formation on the anode side was performed after baking on the cathode side. The reason for this is that further heat treatment at 1100° C. for 1 hour in the air after forming the PtOx film causes grain growth of the PtOx itself, which causes grain growth in the PtOx reduction process described below. This is because the diffusion of Pt cations and the formation of Pt cation-defective oxide-Ni cation clusters may be adversely affected.

Here, the particle sizes of NiO and 8YSZ to be used have a preferred range. If NiOx is unnecessarily fine, it will sinter itself and the connection between NiOx will be broken. Therefore, when producing this kind of anode, particles of micron order are usually used. If the 8YSZ is too large, there is a possibility that the NiOx will not be in contact with the joints between the 8YSZs, so powder with a particle size on the order of submicron to micron is normally used here as well. In the present examples and comparative examples, TZ-8YS grade powder manufactured by Tosoh Corporation was used as 8YSZ, and NIO-F grade powder manufactured by Fuel Cell Materials was used as NiO.

According to Tosoh Corporation, the BET specific surface area of TZ-8YS grade powder, which is 8YSZ, is about 7 m ² g ⁻¹ . Normally, in the case of this BET specific surface area, the average particle size is 2 to 3 μm, so the average particle size of 8YSZ used here is also considered to be about 2 to 3 μm (the average particle size is in the submicron, the BET specific surface area becomes double digits).

The BET specific surface area of the NIO-F grade powder used as the NiOx powder was about half, about 3.1 m ² g ⁻¹ (as stated on the purchased bottle). The average grain size is estimated to be about twice as large as that of 8YSZ.

When powders other than those used in the examples are used, the BET specific surface areas of 8YSZ and NiO fall within the ranges of about 7 m ² g ^-1 ±20% and about 3.1 m ² g ^-1 ±20%, respectively. is considered suitable. If the BET specific surface area is too small (that is, the average particle diameter is too large), the surface area in which active sites can exist decreases, and naturally the activity as an electrode catalyst decreases. Conversely, if the BET specific surface area is too large (i.e., the average grain size is too small), a high temperature treatment exceeding 1000°C is performed during the baking process, so grain growth proceeds, and the connection between NiOx and 8YSZ is broken. On the contrary, the charge transfer path and the oxide ion diffusion path in the anode layer are cut off, which lowers the ratio of three-phase interfaces in the anode layer, which is not preferable.

The solid oxide fuel cell single cells of Examples and Comparative Examples thus produced were operated as solid oxide fuel cells, and the relationship between the current density and the cell voltage was measured. Pure hydrogen (room temperature, saturated with water) was supplied at a flow rate of 80 sccm as the anode gas, and pure oxygen was supplied at a flow rate of 80 sccm as the cathode gas. After that, the supplied gas was switched, and power generation and measurement were performed). Moreover, the operating temperature of the solid oxide fuel cell single cell during this power generation test was set to 700°C and 800°C. In obtaining measurement data, a retention time of about 10 minutes was taken for measurement at one point, and the measured value at the measurement point was taken after waiting for the measured value to stabilize. A graph of this measurement result is shown in FIG.

Normally, the power generation test temperature for a solid oxide fuel cell is 1000.degree. Therefore, as can be seen from the graph in FIG. 2, in the comparative example, even at 800° C. (indicated by a blank circle), the cell voltage drops to 0.5 V when the current density of the output current is about 40 mA cm ⁻² . In addition, when the operating temperature is lowered to 700°C (indicated by a blank triangle), almost no output current can be obtained, and it practically functions as a fuel cell. confirmed not to. On the other hand, as can be seen at a glance, the example showed a fuel cell output performance that was greatly improved compared to the comparative example even at such a low temperature. Specifically, when the operating temperature was 800° C., the cell voltage was 1.02 V at a fuel cell current density of 50 mA cm ⁻² , which was much higher than that of the comparative example.

In general, the fuel cell reaction is characterized by the extraction of electricity using the reverse reaction of the electrolysis reaction of water. At that time, as in thermal power generation, the chemical energy of the fuel is converted into thermal energy by burning it, and the thermal energy is used to turn the turbine of the generator, and this kinetic energy is used to generate electricity. It is a power generation method that can generate electricity directly when synthesizing water from hydrogen and oxygen without a large loss of power generation efficiency that occurs when converting to energy.

Therefore, the potential energy required to produce water from hydrogen and oxygen is theoretically 1.48 V. Therefore, by dividing the cell voltage at a given fuel cell current density by this theoretical value, the fuel cell Power generation efficiency can be estimated. In the above example, the fact that the current density is 50 mA cm ⁻² and the cell voltage is 1.02 V means that the power generation efficiency = (1.02/1.48) × 100 = 68.9% was obtained at this point. Become.

Also, even when the operating temperature is 700 ° C., the cell voltage can be confirmed to be about 0.85 V at a current density of 50 mA cm ^-2 , so power generation efficiency = (0.85 / 1.48) × 100 = 57.4% is obtained. In the comparative example, no cell voltage was observed at a current density of 50 mAcm ⁻² , so the effect is clear. A large expansion was confirmed.

As already mentioned, it is generally recognized by those skilled in the art that, among the reactions occurring in the fuel cell system, the cathode performance in which the oxygen reduction reaction, which has a slow reaction rate, takes place has a great effect on the power generation performance of the fuel cell. Even if the performance on the anode side, where the hydrogen oxidation reaction takes place with a much higher reaction rate than the oxygen reduction reaction, is improved, even if the improvement in the anode performance is recognized, the high power generation efficiency peculiar to the fuel cell maintained, it did not contribute significantly to the improvement of the performance of the fuel cell. The measurement results shown above completely overturn this common sense, and are the first report showing that the improvement of the anode significantly improved the overall performance of the fuel cell, which exhibits high power generation efficiency.

This is a big problem in the world of oxide fuel cell development, and it is difficult to improve the performance of anode-supported solid electrolyte membrane type oxide fuel cells. It can be said that it is an invention that brings

Further, instead of reducing NiO in the anode layer and PtOx formed thereon at the same time, the anode layer before forming the PtOx film is first reduced under the same conditions as the reduction treatment in the hydrogen atmosphere described above, Thereafter, PtOx films were formed with sputtering thicknesses of 10 nm, 20 nm, and 30 nm, and the current density-cell voltage characteristics of samples obtained by reduction under the same conditions were measured (operating temperature 800° C.). The measurement results are shown in FIG. In the graph of FIG. 3, the term "pre-reduction" means that the anode layer before the PtOx film formation is first reduced under the same conditions as the reduction treatment in the hydrogen atmosphere described above. As can be seen from this graph, the thickness of the PtOx thin film is 10 nm, although there is an advantageous condition that PtOx can sufficiently penetrate deep into the porous anode layer by pre-reduction. In the case of , the performance of the prior reduction treatment (Δ) was extremely lower than that of the example (▴) in which the reduction treatment was performed together with NiOx, and a conventional anode that did not use Pt at all was used. It was confirmed that the power generation performance was not much different from that of the case (■). The expected performance improvement was not observed when the thickness of the PtOx thin film was 20 nm and 30 nm (○ and ◇, respectively). (● and ♦) gave no significant change. In the graph, the marker (●) for plotting the data of the example when the thickness of the Pt thin film is 20 nm is the marker (●) for plotting the data when the Pt thin film having a thickness of 30 nm is formed after pre-reduction. The former is hard to see because it is almost hidden behind the ◇), but the data of both are almost the same.

This result is due to the fact that interdiffusion, in which substances move through grain boundaries, is fairly rapid, and therefore it is not effective to spread PtOx deep into the pores by preparing a porous structure in advance. This is thought to be due to inhibition of the formation of Pt cation-defective oxide-Ni cation clusters because there is little or, on the contrary, some of the metal platinum particles are sputtered instead of PtOx.

Next, the current density-IR free (net cell voltage that compensates for the effect of the voltage drop due to the internal resistance of the fuel cell) characteristic (operating temperature 800 ° C.) was obtained for the single cell of the above example. It is shown in FIG. As can be seen from FIG. 4, the cell voltage of about 1.1 V in the no-load state drops only to about 0.78 V even at a high current density of 500 mA cm ⁻² , confirming that it has a large power generation capacity. was done.

When the same comparison is made with the data in the example, the current density is 50 mAcm ⁻² and the cell voltage is about 1.05 V, and the power generation efficiency at that time = (1.05/1.48) × 100 = 70.9% .
Furthermore, at a current density of 150 mAcm ⁻² , a cell voltage of about 0.96 V is observed,
It can be estimated that the power generation efficiency at that time=(0.96/1.48)×100=64.8%.

This means that by minimizing loss due to ohmic drops present in the solid electrolyte membrane, seals, etc., a fuel cell that exhibits a power generation performance of 150 mAcm ⁻² while maintaining a power generation efficiency of 64.8% can be achieved according to the present invention. It suggests that it is possible by using an anode layer.

It is clear from a comparison of FIGS. 2 and 4 that, in the conventional anode layer, if the solid electrolyte membrane of the same type and thickness is used, the effect cannot be expected.

The method for obtaining IR-free is well known to those skilled in the art, and FIG. 5 shows an outline of the method for obtaining it by the current interruption method. As the name suggests, the output current from the fuel cell is changed as steeply as possible from a state in which a predetermined current i is flowing to a current of 0, as shown in the upper graph of FIG. This can be realized by inserting an oscilloscope in the circuit that extracts current from the fuel cell, setting this steep current change, and observing the resulting voltage change waveform. IR-free measurement is the net measurement of the fuel cell, which compensates for the ohmic voltage drop IR due to the internal resistance R of the fuel cell system with respect to the cell voltage V of the fuel cell when a given current I is output. The cell voltage Vo is obtained. When the output current of a fuel cell is suddenly cut off, the cell voltage immediately rises by an ohmic voltage drop in a simple model of a series circuit consisting of a current source with an internal resistance of 0 and an internal resistance R. , the cell voltage does not rise vertically immediately after cutting, as shown in the lower graph of FIG. In the lower graph of FIG. 5, it converges to a constant value while rising slightly). Therefore, in the graph on the lower side of FIG. 5, the cell voltage coordinates at the intersection of the vertical line extending in the vertical direction from the time when the cell voltage started to change due to the current interruption and the extension line of the cell voltage change immediately after the transition to the steady state. is the ohmic (IR) voltage drop Vo. IR free = V + Vo
is obtained as

Next, the thickness of the PtOx thin film was changed to 20 nm and 30 nm. IR-free and normal (ie including ohmic voltage drop) cell voltages at current densities of about 250 mA cm ⁻² were measured for . The results are shown in FIG. The thickness of the anode layer of the NiO _- 8YSZ mixture before reduction in a hydrogen atmosphere was about 40 μm. Assuming uniform diffusion in the anode layer, the concentration of Pt in the anode layer is 0.3 wt%, 0.5 wt% and 0.9 wt% when the PtOx thin film has a sputtering thickness of 10 nm, 20 nm and 30 nm, respectively. was calculated.

FIG. 6 shows the measurement results of IR-free and cell voltage for three examples in which the PtOx thin film sputtered thicknesses are 10 nm, 20 nm, and 30 nm. It is shown as a graph taking voltage values for IR-free and cell voltage. Furthermore, ceria (CeOx, 1.5≤x≤2) nanowires were used instead of the 8YSZ powder as the raw material for the anode layer (the composition ratio (weight ratio) was NiOx:CeOx nanowires = 3:2), and the PtOx thin film Additional examples of solid oxide fuel cells were produced under the same conditions except that the sputtering pressure was one type of 10 nm, and the results of measurements under the same conditions are also shown. To distinguish between these two types of examples on the graph, the IR-free and cell voltage measurements for the example using 8YSZ powder as the source of the oxide electrolyte are indicated by open circles (○) and triangles (Δ). , and IR-free and cell voltage measurements of examples using ceria nanowires are plotted with filled circles (●) and triangles (▴), respectively.

Note that the weight ratio (NiOx: ceria nanowires = 3:2) when ceria nanowires were used as the oxide electrolyte was different from the weight ratio when 8YSZ was used (NiOx: 8YSZ = 4:1). The reason is explained below. Since 8YSZ is an easily sinterable powder, sintering proceeds by heat treatment at an electrode baking temperature of 1200° C. for 1 hour. Therefore, when the amount of NiOx is small, 8YSZ bonds with each other by sintering to form large grains, so-called grain growth progresses too much, and the connection between 8YSZ and between NiOx may be easily broken. To avoid this problem, the weight ratio of NiOx to 8YSZ is 4:1. On the other hand, CeOx nanowires originally have ceria linked one-dimensionally and are bulkier than fine particles, and grain growth like 8YSZ is not observed experimentally at a temperature of about 1200°C. Therefore, by setting the weight ratio of the NiOx--CeOx nanowires to 3:2, sufficient charge transfer paths and oxide ion diffusion paths are ensured in the anode layer.

As can be seen from FIG. 6, when the thickness of the PtOx thin film is in the range of 10 to 30 nm, even if this thickness is changed (that is, the Pt concentration in the anode layer is changed in the range of 0.3 to 0.9% by weight). At current densities of about 250 mA cm ⁻² , the IR-free and cell voltage became nearly constant (as mentioned above, the exact value of IR-free for 10 nm thickness has not been determined). From this, even if the Pt concentration in the anode layer is lowered to 0.3% by weight, the lower limit of the preferable Pt concentration has not yet been reached. And it is suggested that the cell voltage does not decrease so much. However, in the film thickness range (concentration range) tested, the IR-free and cell voltage were almost constant, or at least assumed to be so, and the trend of these changes could not be read.

7 and 8 show an SEM image and elemental mapping at the same position of an example of the present invention (PtOx film thickness: 10 nm). The analytical SEM used here is a JEOL SU-8000 with a spatial resolution of 1 nm. Upon inspection of the SEM image of FIG. 7, no protrusions or other features that could be attributed to Pt particles could be found. If Pt particles with a diameter of about 1 nm were present in the cross section of the anode of the example, the existence of the particles should have been recognizable from the SEM image, regardless of the detailed shape. Therefore, it can be said that there were no Pt nanoparticles with a diameter of about 1 nm or more in the cross section of the anode of the example. The elemental mapping shown in FIG. 8 also suggests that Ni is unevenly distributed and forms agglomerates in some places. It was confirmed that they were distributed with high uniformity throughout.

[Measurement and analysis of the anode material of the present invention]
Further measurements and analyzes were carried out on the form in which Pt is present in the anode material of the present invention. The samples of the anode material to be measured and analyzed here were prepared using the same material and manufacturing method as the anode material of the example of the present invention prepared above.

FIG. 9 is a TEM bright-field image of the sample immediately after forming a PtOx thin film on the NiO-8YSZ mixture anode layer by room temperature sputtering (that is, before hydrogen reduction treatment). A large number of black-looking Pt fine particles were observed in the PtOx sputtered layer formed on the surface of the anode particles. The particle size of the Pt fine particles was about 2 nm.

Observation of the result of hydrogen reduction treatment of this sample revealed that the Pt fine particles were mainly present on the surface of the yttria-stabilized zirconia YSZ particles, as shown in the STEM image of FIG. In FIG. 10, the YSZ particles are vertically elongated rod-shaped particles visible in the center of the STEM image shown on the left side. It was confirmed from the element mapping of Zr, Y, and O shown in (c) and (d) of FIG. 12 and (d) of FIG. 13 that this is YSZ. White-appearing Pt particles were observed on the upper flat-appearing surface of the YSZ particles. It was confirmed from the elemental mapping image of Pt shown in FIG. 13(b) that these particles were Pt. Note that the lower part of the STEM image where the YSZ grains seemed to bite was Ni, as can be seen from the elemental mapping image of Ni shown in FIG. 12(b). That is, in this sample, YSZ grains were present on large Ni grains.

FIG. 11 is a full spectrum of information from the entire viewing area shown in FIGS. The spectrum confirmed that the amount of Pt present in this area was very small.

In the left STEM image, the right side of FIG. 10 shows a high-resolution image of the upper surface of the YSZ particle near the starting point of the arrow drawn there. As can be seen from this high-resolution image, despite the reduction treatment in hydrogen at 800° C. for 1 hour, the particle size of these Pt particles was about 2 nm, the same as during sputtering. near the tip of the arrow, etc.). In addition, some Pt particles grew to a size of about 10 nm to 20 nm at maximum.

As can be seen from the elemental mapping image of Pt shown in FIG. 13(b), Pt was widely distributed not only on the upper surface of the YSZ grain but also on this grain. However, although the STEM image shown here also shows the interface between YSZ and Ni, even if the inventors of the present application repeatedly make full use of the means available for observation, clear Pt particles existence could not be confirmed. In other words, no Pt particles with a particle size of 1 nm or more were present at this interface.

It is well known to those skilled in the art that, when only fine particles of Pt are fired at 800° C. for 1 hour in a reducing atmosphere, grains grow to a size of several μm to several tens of μm. . What happened contrary to the technical common sense as explained above is that the surface of YSZ and the fine Pt cause a strong interaction, and the fine particles move through the grain boundary between YSZ and Ni (so-called interdiffusion). , it is considered that the fine Pt observed on the surface of the YSZ particles dissolved in YSZ or Ni. It is reasonable to interpret that this made it difficult to observe the existence of Pt.

In this way, the strong interaction between fine Pt and YSZ, which suppresses the grain growth of Pt fine particles on the YSZ surface, is used as a medium to form solid oxide particles at many locations on the grain boundaries between YSZ and Ni in the anode layer. It is concluded that an interface was formed which is conducive to improving the anode performance of the shaped fuel cell.

As described in detail above, according to the present invention, the solid oxide solid oxide can exhibit extremely high power generation performance and can extend the operating temperature range toward low temperatures by containing only a small amount of Pt in the anode layer. Since it is possible to provide a type fuel cell, it is expected that the present invention will greatly contribute to the practical use and spread of solid oxide fuel cells.

William C. Chueh, Yong Hao, WooChul Jung and Sossiana M. Haile, Nature Mater., 11, 155-161, (2012). http://www.nature.com/nmat/journal/v11/n2/extref/nmat3184-s1.pdf Keisuke Fugane, Toshiyuki Mori, Pengfe Yan, Takaya Masuda, Sunya Ymamoto, Fei Ye, Hideki Yoshikawa, Graeme Auchterlonie, and John Drennan, ASC Appl. Materi. Interfaces 7, 2698-2707 (2015). Zhi-Peng Li, Toshiyuki Mori, Graeme John Auchterlonie, Jin Zou and John Drennan, Phys. Chem. Chem. Phys. Vol.13(20), 9685-9690 (2011). Zhi-Peng Li, Toshiyuki Mori, Graeme John Auchterlonie, Yanan Guo, Jin Zou, John Drennan, and Masaru Miyayama, J. Phys. Chem. C, Vol.115(14), 6877-6885 (2011). Ding Rong Ou, Toshiyuki Mori, Hirotaka Togasaki, Motoi Takahashi, Fei Ye, and John Drennan, Langmuir, 27, 3859-3866 (2011).

Claims (4)

containing zirconia or ceria, nickel and platinum,
The platinum is dissolved in the zirconia or ceria particles or the nickel particles,
Containing 0.3 to 0.9% by weight of the platinum ,
The dissolved platinum has the form of Pt cation-defective oxide-Ni cation cluster
Anode material for solid oxide fuel cells. The solid oxide fuel cell anode material according to claim 1 , which has a porous structure. 3. The solid oxide fuel cell anode material according to claim 1, wherein said zirconia is yttria-stabilized zirconia. A solid oxide fuel cell using the anode material for a solid oxide fuel cell according to any one of claims 1 to 3 .

Images