Method for searching electrode catalyst for oxygen generation and electrode catalyst for oxygen generation

Abstract

To provide a method for searching a non-noble metal electrode catalyst with high oxygen generation efficiency, a long service life, and a low cost.SOLUTION: A method for searching an electrode catalyst for oxygen generation includes the steps of: randomly selecting a chemical compound from a material candidate list composed of chemical compounds having n types of metal elements selected from an enlarged element candidate selected from a metal element, a metalloid element, a halogen, hydrogen, an oxygen group element, a nitrogen group element, and a carbon group element; synthesizing and evaluating oxides of the chemical compound; subjecting the oxides of the chemical components in the material candidate list to characteristic prediction by a regression method, extracting the oxides of unmeasured chemical compounds with a high score, and implementing material synthesis and characteristic evaluation; implementing characteristic prediction by the regression method on the basis of all the implemented characteristic evaluation results by a plurality of times; classifying the chemical components into an active material candidate and an inactive material candidate by implementing screening by a machine learning classification method; implementing characteristic prediction by the regression method with respect to the active material candidate, implementing the synthesis of chemical components with a high score and characteristic measurement by a plurality of times, and selecting a chemical compound with the most excellent characteristic.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method for searching for an oxygen generating electrode catalyst and an oxygen generating electrode catalyst. [Background technology]

[0002] In recent years, green hydrogen has been attracting attention as a clean energy source to replace fossil fuels in light of global environmental issues such as global warming. Green hydrogen is an environmentally friendly energy source that does not emit carbon dioxide during its production or use.

[0003] Since green hydrogen is produced by the electrolysis of water, an oxygen-evolving electrode catalyst is essential for producing green hydrogen. Currently, precious metals such as platinum group elements platinum (Pt), ruthenium (Ru) or iridium (Ir) are mainly used as electrode catalysts for oxygen generation, as disclosed, for example, in Patent Document 1. However, these precious metal elements are expensive and resources are limited, which is an obstacle to supplying a huge amount of green hydrogen. For this reason, there is a demand for a low-cost oxygen generating electrode catalyst that is made of a non-precious metal or has a low ratio of non-precious metal elements, has high oxygen generating efficiency, and has a long life, as well as for the establishment of a method for searching for oxygen generating electrode catalysts that are suitable for developing such catalysts. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2013-37942 A [Non-patent literature]

[0005] [Non-Patent Document 1] Xie, M. et al. An amorphous Co-carbonate-hydroxide nanowire array for efficient and durable oxygen evolution reaction in carbonate electrolytes. Nanoscale 9, 16612-16615 (2017) [Non-Patent Document 2] Xie, F. et al. Ni3N@Ni-Ci nanoarray as a highly active and durable non-noble-metal electrocatalyst for water oxidation at near-neutral pH. Journal of Catalysis 356, 165-172 (2017). [Non-Patent Document 3] Ma, M. et al. Highly efficient and durable water oxidation in a near-neutral carbonate electrolyte electrocatalyzed by a core-shell structured NiO@Ni-Ci nanosheet array. Sustainable Energy & Fuels 1, 1287-1291 (2017). [Non-Patent Document 4] Wang, N. et al. Hydration-Effect-Promoting Ni-Fe Oxyhydroxide Catalysts for Neutral Water Oxidation. Advanced Materials 32, 1906806 (2020). [Non-Patent Document 5] Chen, L., Ren, X., Teng, W. & Shi, P. Amorphous Nickel-Cobalt-Borate Nanosheet Arrays for Efficient and Durable Water Oxidation Electrocatalysis under Near-Neutral Conditions. Chemistry - A European Journal 23, 9741-9745 (2017). [Non-Patent Document 6] Yang, L. et al. A cobalt-borate nanosheet array: an efficient and durable non-noble-metal electrocatalyst for water oxidation at near neutral pH. Journal of Materials Chemistry A 5, 7305-7308 (2017). [Non-Patent Document 7] Ji, X. et al. Core-shell CoFe2O4@Co-Fe-Bi nanoarray: a surface-amorphization water oxidation catalyst operating at near-neutral pH. Nanoscale 9, 7714-7718 (2017). [Non-Patent Document 8] Dong, Y., Komarneni, S., Wang, N., Hu, W. & Huang, W. An in situ anion exchange induced high-performance oxygen evolution reaction catalyst for the pH-near-neutral potassium borate electrolyte. Journal of Materials Chemistry A 7, 6995-7005 (2019). [Non-Patent Document 9] Liu, X. et al. Nanostructured copper oxide electrodeposited from copper(II) complexes as an active catalyst for electrocatalytic oxygen evolution reaction. Electrochemistry Communications 46, 1-4 (2014).

Non-Patent Document 10

Non-Patent Document 11

Non-Patent Document 12

[0006] An object of the present invention is to provide a metal electrocatalyst which has high oxygen generation efficiency, a long life, and is also low cost, and a method for searching for such a metal electrocatalyst. [Means for solving the problem]

[0007] The configuration of the present invention for solving the problems is shown below.

[0008] (Configuration 1) A method for searching for an oxygen generating electrode catalyst, the method comprising: searching for an oxygen generating electrode catalyst made of a metal oxide having n kinds of metal elements, the method comprising the steps of: A first step of selecting n+1 or more and 52 or less types of metal elements as metal element candidates of the electrode catalyst, and an element selected from the elements of the metal element candidates, metalloid elements, halogens, hydrogen, oxygen family elements, nitrogen family elements, and carbon family elements as an extended element candidate; a second step of creating a material candidate list consisting of compounds composed of elements selected from the expanded element candidates and having n kinds of metal elements selected from the metal element candidates; A third step of randomly selecting a compound from the list of candidate materials; a fourth step of synthesizing an oxide of the compound selected in the third step, and then measuring the characteristics of the synthesized oxide to obtain a characteristic evaluation result; A fifth step of performing a property prediction by a regression method for oxides of compounds in the material candidate list based on the property evaluation results, extracting oxides of compounds that have not been synthesized and measured but have the highest performance from the material candidate list, synthesizing the materials and measuring the properties, and obtaining property evaluation results; A sixth step of performing a property prediction by a regression method for oxides of compounds in the material candidate list based on all the property evaluation results performed above, extracting oxides of compounds that have not been synthesized and measured but have the highest performance from the material candidate list, and repeating material synthesis and property measurement multiple times; A seventh step of performing a classification method using machine learning based on all of the property evaluation results performed on unsynthesized and unmeasured compounds in the material candidate list, and classifying the unsynthesized and unmeasured compounds in the material candidate list into active material candidates and inactive material candidates according to a predetermined criterion; An eighth step of performing a regression method to predict properties based on all or some of the property evaluation results performed on the oxides of the compounds of the active material candidates, extracting oxides of compounds with top performance from the active material candidates in the material candidate list, synthesizing the materials and measuring their properties, and obtaining property evaluation results; a ninth step of repeating the seventh and eighth steps a plurality of sets, with the seventh and eighth steps being one set; A method for searching for an electrode catalyst for oxygen evolution, comprising a tenth step of selecting an oxide of a compound exhibiting the most excellent properties as an electrode catalyst for oxygen evolution from all of the property evaluation results obtained above. (Configuration 2) The method for searching for an oxygen evolution electrode catalyst according to configuration 1, wherein the regression method is one selected from the group consisting of Gaussian process regression, regression using a linear model, stochastic gradient descent, logistic regression, kernel ridge regression, decision tree, random forest, XGBoost, and neural network. (Configuration 3) The method for searching for an oxygen generating electrode catalyst according to configuration 1 or 2, wherein the classification method is one selected from the group consisting of random forest, support vector machine, stochastic gradient descent, local neighborhood method, decision tree, XGBoost and neural network. (Configuration 4) 4. The method for searching for an oxygen evolution electrode catalyst according to any one of claims 1 to 3, wherein the number of the expansion element candidates is 9 or more and 72 or less. (Configuration 5) The candidate elements are hydrogen (H), lithium (Li), boron (B), carbon (C), nitrogen (N), fluorine (F), sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), scandium (Sc), titanium (Ti), vanadium (V), chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu), zinc (Zn), gallium (Ga), germanium (Ge), arsenic (As), selenium (Se), bromine (Br), rubidium (Rb), strontium (Sr), yttrium (Y), zirconium (Zr), niobium (Nb), molybdenum (Mo), ruthenium (Ru), rhodium (Rh), and arsenic (Ar). h), palladium (Pd), silver (Ag), cadmium (Cd), tin (Sn), antimony (Sb), tellurium (Te), iodine (I), cesium (Cs), barium (Ba), lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), samarium (Sm), europium (Eu), gadolinium (Gd), terbium (Tb), dysprosium ( A method for searching for an electrode catalyst for oxygen evolution according to configuration 4, wherein the metal is selected from the group consisting of platinum (Dy), holmium (Ho), erbium (Er), thulium (Tm), ytterbium (Yb), lutetium (Lu), hafnium (Hf), tantalum (Ta), tungsten (W), rhenium (Re), osmium (Os), iridium (Ir), platinum (Pt), gold (Au), mercury (Hg), and lead (Pb). (Configuration 6) 6. The method for searching for an oxygen generating electrode catalyst according to any one of claims 1 to 5, wherein the materials in the material candidate list are compounds composed of the metal element candidate elements. (Configuration 7) 7. The method for searching for an oxygen generating electrode catalyst according to any one of configurations 1 to 6, wherein the evaluation index items for the characteristic evaluation are one or more selected from the group consisting of overpotential, Tafel gradient, potential dependence of change in Tafel gradient, charge transfer resistance, deterioration tendency, exchange current density, reaction current density at a specific potential, amount of oxygen generated per mass until deactivation, amount of charge per mass used for oxygen generation until deactivation, amount of oxygen generated per unit area until deactivation, amount of charge per mass used for oxygen generation until deactivation, electrochemically active surface area, pore distribution of the material, wettability of the material, amount of lattice oxygen defects introduced, increase ratio of current density between certain potentials, transfer coefficient, symmetry factor, rate-determining process, type and concentration of cations in the electrolyte, type and concentration of anion in the electrolyte, reaction intermediates in the rate-determining process and their adsorption energies, vacuum level, redox potential, band gap, Fermi level, valence, reaction rate constant, and reaction selectivity. (Configuration 8) A method for searching for an oxygen generating electrode catalyst according to any one of configurations 1 to 7, wherein the element number ratio of the metal elements in the compound containing n types of metal elements is half that of the other types of metal elements for only one type of metal element, and the element number ratios of the other types of metal elements are all equal. (Configuration 9) The method for searching for an electrode catalyst for oxygen evolution according to aspect 8, wherein n is 5. (Configuration 10) An oxygen generating electrode catalyst discovered by the method for discovering an oxygen generating electrode catalyst according to any one of configurations 1 to 9. (Configuration 11) MnFeNiZnAg 0.5 O2Cl 0.2 , MnFeNiZn 0.5 AgO2Cl 0.25 , Sc 0.5 MnFeNiAgO2Cl 0.1 , and NiFeCuMnAgO2Cl 0.3 An electrode catalyst for oxygen generation, comprising at least one compound selected from the group consisting of: Effect of the Invention

[0009] The present invention provides a metal electrocatalyst that has high oxygen generation efficiency, a long life, and is also low cost, and a method for searching for such a metal electrocatalyst. [Brief description of the drawings]

[0010] [Figure 1] FIG. 2 is an explanatory diagram illustrating the framework of the search method of the present invention. [Diagram 2] FIG. 2 is a flow chart showing steps of the searching method of the present invention. [Diagram 3] FIG. 13 is a graph showing the transition of overvoltage characteristics of samples extracted by the searching method of the present invention. [Figure 4] 1 shows scanning electron microscope (SEM) images of samples prepared in the examples, including (a) MnFeNiZnAgOx, (b) MnFeNiZn0.5AgOx, (c) Ti0.5MnFeNiAgOx, (d) MnFeNiZnAg0.5Ox, and (e) Sc0.5MnFeNiAgOx. [Diagram 5] 1 shows XRD (X-Ray Diffraction) data of samples prepared in the examples. [Figure 6] These are the results of structural evaluation of the MnFeNiZn0.5AgOx sample prepared in the example. (a) is a low-magnification HR-TEM (High-Resolution Transmission Electron Microscope) image, (b) is a high-magnification HR-TEM image, (c) is an electron beam diffraction image of the area indicated by the circle, (d) is an ADF-STEM (Annular Dark-Field Scanning Transmission Electron Microscope) image, and (e) to (k) are observations of the spatial distribution of the component elements by EDS (Energy Dispersive Spectroscopy), where (e) is Mn, (f) is Fe, (g) is Ni, (h) is Zn, (i) is Ag, (j) is O, and (k) is Ti. [Figure 7]The results of XPS (X-ray photoelectron spectroscopy) measurements of the MnFeNiZn0.5AgOx sample prepared in the examples are shown here, along with the calculation results obtained by fitting the spectra corresponding to the actual measurements, (a) Mn 2p, (b) Fe 2p, (c) Ni 2p3 / 2, (d) Zn 2p3 / 2, (e) Ag 3d, and (f) O 1s. [Figure 8] FIG. 1 is a characteristic diagram showing the overvoltage at a current density of 10 mA / cm2 for various materials including MnFeNiZnAg. [Figure 9] This is a characteristic diagram evaluating the oxygen evolution reaction of the MnFeNiZn0.5AgOx electrode catalyst based on the voltage dependence of the current value, with pH as a parameter. [Figure 10] This is a characteristic diagram evaluating the oxygen evolution reaction of the MnFeNiZn0.5AgOx electrode catalyst in terms of the time dependence of potential, with pH as a parameter. [Figure 11] This is a characteristic diagram of the oxygen evolution reaction of the MnFeNiZn0.5AgOx electrode catalyst material evaluated by the potential-current characteristics (LSV curve) in an environment of 80°C using a 1M carbonate (Na-Ci) electrolyte. [Figure 12] This is a characteristic diagram comparing and evaluating the potential change over time (CP: Chronopotentiometry) of MnFeNiZn0.5AgOx and RuO2. [Figure 13] FIG. 1 is a Ragony plot showing, in map form, the relationship between current density and specific charge for various materials. [Figure 14] FIG. 2 is a configuration explanatory diagram illustrating the main configuration of the oxygen generating device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] (Embodiment 1) In the first embodiment, a method for searching for an oxygen generating electrode catalyst according to the present invention will be described with reference to FIGS.

[0012] An overview of the method for searching for electrode catalysts for oxygen evolution according to the present invention is shown in Figure 1. In the present invention, two types of AI (Artificial Intelligence) are used to execute a two-stage AI processing loop to extract materials suitable as electrode catalysts for oxygen evolution. An attempt to search for electrode materials using AI is disclosed in Non-Patent Document 16, but the configuration is different from that of the first embodiment.

[0013] 1. Overview (1) Preparation First, a list of candidate materials is created and a holder 1 storing the list is prepared. In the first step, several types of materials are randomly selected from the candidate materials, synthesized, and their properties are evaluated.

[0014] (2) First AI loop processing Then, based on the results of the property evaluation, a property prediction is performed by a regression analysis 2 using the first AI, and top-performing materials that have not been synthesized or measured are extracted from the material candidate list, and a property evaluation experiment 3 consisting of material synthesis and property measurement is performed. This material extraction loop using the first AI (first AI processing loop) is performed multiple times. Here, the regression analysis 2, which is a regression method, can be one selected from the group consisting of Gaussian process regression, regression using a linear model, stochastic gradient descent, logistic regression, kernel ridge regression, decision tree, random forest, XGBoost, and neural network. Among these, Gaussian process regression using a stochastic process is particularly preferred because it has the advantage of being able to estimate not only the predicted value but also the uncertainty of the prediction.

[0015] (3) Second AI loop processing Next, machine learning analysis 4 is performed using a second AI to classify unsynthesized and unmeasured compound oxides from the list of candidate materials stored in holder 1 based on all property evaluation results performed up to that point, and the candidate materials are classified into active material candidates and inactive material candidates. The active material candidates are stored in holder 5, which stores candidate active materials, and the inactive material candidates are stored in holder 6, which stores candidate inactive materials. Here, the classification method, which is the second AI, can be one selected from the group consisting of random forest, support vector machine, stochastic gradient descent, local neighborhood method, decision tree, XGBoost, and neural network. Among these, random forest, which belongs to ensemble learning, is particularly preferred because it has the characteristic that it is easy to obtain a model with high prediction accuracy. Thereafter, a regression process 2 using a first AI is performed on the active material candidates, and a number of unsynthesized and unmeasured materials (for example, two) with the highest scores are selected, and a property evaluation experiment 3 consisting of material synthesis and property measurement is performed to extract candidate materials for the oxygen generating electrode catalyst. This second AI loop is repeated multiple times, and the material with the best properties as an oxygen generating electrode catalyst is selected as the oxygen generating electrode catalyst from among the materials that have been experimentally evaluated so far.

[0016] In the present invention, the first AI loop process, which has a short loop, efficiently narrows down the materials, and the second AI loop process narrows down the active material candidates using an AI for classification problems different from the first AI loop process, and the first AI process is serially combined with experiments, making it possible to search for and extract materials with both high extraction accuracy and efficiency.

[0017] 2.Details Next, steps of a search method for an electrode catalyst for oxygen evolution made of a metal oxide having n kinds of metal elements will be described with reference to the flow chart of FIG.

[0018] First, n+1 to 52 types of metal elements are selected as candidate metal elements for the electrode catalyst, and elements selected from the candidate metal elements, metalloid elements, halogens, hydrogen, oxygen group elements, nitrogen group elements, and carbon group elements are selected as candidate extended elements (step S11).

[0019] Here, examples of the candidate expansion elements include hydrogen (H), lithium (Li), boron (B), carbon (C), nitrogen (N), fluorine (F), sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), scandium (Sc), titanium (Ti), vanadium (V), chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu), zinc (Zn), gallium (Ga), germanium (Ge), arsenic (As), selenium (Se), bromine (Br), rubidium (Rb), strontium (Sr), yttrium (Y), zirconium (Zr), niobium (Nb), molybdenum (Mo), ruthenium (Ru), and ruthenium (Br). Ru (Ru), Rhodium (Rh), Palladium (Pd), Silver (Ag), Cadmium (Cd), Tin (Sn), Antimony (Sb), Tellurium (Te), Iodine (I), Cesium (Cs), Barium (Ba), Lanthanum (La), Cerium (Ce), Praseodymium (Pr), Neodymium (Nd), Samarium (Sm), Europium (Eu), Gadolinium (Gd), Terbiu Examples of suitable elements include ruthenium (Tb), dysprosium (Dy), holmium (Ho), erbium (Er), thulium (Tm), ytterbium (Yb), lutetium (Lu), hafnium (Hf), tantalum (Ta), tungsten (W), rhenium (Re), osmium (Os), iridium (Ir), platinum (Pt), gold (Au), mercury (Hg) and lead (Pb). In order to reduce costs, it is preferable to keep the ratio of platinum group elements Pt, Pd, Rh, Ru, Ir and Os low even if they are used.

[0020] The number of the expanded element candidates is not particularly limited, but is preferably 9 to 72, more preferably 9 to 32. By making it 9 or more, the probability of extracting a material having excellent properties as an oxygen generating electrode catalyst increases, and by making it 72 or less, particularly 32 or less, the number of candidate materials in the list can be kept to 1 million candidates or less, making it possible to extract them efficiently.

[0021] Next, a material candidate list is created that is made up of compounds that are composed of elements selected from the expanded element candidates and that have n kinds of metal elements selected from the metal element candidates (step S12). Here, it is preferable that the materials in the material candidate list are compounds constituted by the metal elements in the metal element candidate list. In the examples, good results were obtained when n is 5 and the compound has n types of metal elements, with the ratio of the number of metal elements of only one type being half the ratio of the number of metal elements of the other types, and the ratios of the numbers of the other types of metal elements being all equal.

[0022] Then, a compound is randomly selected from the list of candidate materials (step S13), and an oxide of the compound selected in step S13 is synthesized. Then, the characteristics of the synthesized oxide are measured to obtain a characteristic evaluation result (step S14).

[0023] Here, examples of the evaluation indexes for the characteristic evaluation include one or more selected from the group consisting of overpotential, Tafel slope, potential dependence of change in Tafel slope, charge transfer resistance, deterioration tendency, exchange current density, reaction current density at a specific potential, amount of oxygen generated per mass until deactivation, amount of charge per mass used for oxygen generation until deactivation, amount of oxygen generated per unit area until deactivation, amount of charge per mass used for oxygen generation until deactivation, electrochemically active surface area, pore distribution of the material, wettability of the material, amount of lattice oxygen defects introduced, increase ratio of current density between certain potentials, transfer coefficient, symmetry factor, rate-determining process, type and concentration of cations in the electrolyte, type and concentration of anion in the electrolyte, reaction intermediate in the rate-determining process and its adsorption energy, vacuum level, redox potential, band gap, Fermi level, valence, reaction rate constant, and reaction selectivity. Among these, the overvoltage, the tendency to deteriorate, and the amount of charge per mass used for oxygen generation until deactivation are indicators representative of the characteristics of the oxygen generating electrode catalyst and are particularly preferably used since they are easy to measure.

[0024] Thereafter, based on the results of the property evaluation, a property prediction is performed using a regression method for the oxides of the compounds in the material candidate list, and oxides of compounds that have not been synthesized or measured but have the highest performance are extracted from the material candidate list. Materials are synthesized and properties are measured, and property evaluation results are obtained (step S15). Furthermore, based on the results of all the characteristic evaluations carried out above, characteristic predictions are performed using a regression method for the oxides of the compounds in the material candidate list, and oxides of compounds that have not been synthesized or measured but have the highest performance are extracted from the material candidate list, and material synthesis and characteristic measurement are repeated multiple times (step S16).

[0025] As mentioned above, the regression method may be one selected from the group consisting of Gaussian process regression, regression using a linear model, stochastic gradient descent, logistic regression, kernel ridge regression, decision tree, random forest, XGBoost, and neural network. Among these, Gaussian process regression using a stochastic process is particularly preferred because it has the advantage of being able to estimate not only the predicted value but also the uncertainty of the prediction.

[0026] Next, a machine learning classification method is used to select unsynthesized and unmeasured compounds from the material candidate list based on all of the property evaluation results performed above, and the unsynthesized and unmeasured compounds from the material candidate list are classified into active material candidates and inactive material candidates according to predetermined criteria (step S17). Here, as mentioned above, the classification method may be one selected from the group consisting of random forest, support vector machine, stochastic gradient descent, local neighborhood method, decision tree, XGBoost, and neural network. Among these, random forest, which belongs to ensemble learning, is particularly preferred because it has the advantage that a model with high prediction accuracy can be easily obtained.

[0027] Thereafter, a property prediction is performed using a regression method based on all or some of the property evaluation results performed on the oxides of the candidate compounds of the active material, and oxides of the compounds with the best performance are extracted from the candidate active materials in the material candidate list, and material synthesis and property measurement are performed to obtain the property evaluation results (step S18). Furthermore, step S17 and step S18 are regarded as one set, and step S17 and step S18 are repeated a plurality of sets (step S19).

[0028] Finally, from all the property evaluation results obtained above, an oxide of a compound showing the most excellent properties is selected as an electrode catalyst for oxygen evolution (step S20). Through the above steps, an electrode catalyst having properties suitable for oxygen generation, such as high activity, high durability, and high selectivity, is extracted.

[0029] The electrode catalyst extracted by the exploration method of the present invention can be composed of metal elements that are abundant and relatively low cost, and does not contain rare and costly precious metals such as platinum (Pt) and ruthenium (Ru), or if they are used, they can be used in a small proportion, thereby reducing costs. Furthermore, by using the above-mentioned material configuration, it is possible to obtain a catalyst for oxygen generation electrodes that has excellent overvoltage, stability, durability, current density characteristics, specific charge, and large current characteristics in the weakly alkaline pH range, as shown in the examples.

[0030] As shown in the examples, the oxygen evolution electrode catalyst (electrode catalyst) of the present invention can be obtained through a pyrolysis process in which a solution containing non-precious metal elements such as manganese (Mn), iron (Fe), nickel (Ni), zinc (Zn) and silver (Ag) is prepared and mixed, the mixture is applied to a conductive substrate such as titanium (Ti) by coating, and then the substrate is dried and fired. Here, as the solution containing manganese (Mn), iron (Fe), nickel (Ni), zinc (Zn) and silver (Ag), for example, a chloride solution such as MnCl2, FeCl2, ZnCl2, AgCl, or a nitrate solution such as AgNO3 can be used. In this case, it has been confirmed that the chlorine (Cl) used in the solution remains partially as an electrode catalyst, but does not adversely affect the characteristics of the electrode catalyst such as overvoltage and stability. By using a chloride solution, an oxygen generating electrode catalyst can be efficiently and easily prepared. Regarding Ag, AgNO3 is preferably used from the viewpoint of solubility. The sintering temperature is preferably 150° C. or higher and 400° C. or lower. Other methods for producing the electrode catalyst of the present invention include electrodeposition, chemical vapor deposition, hydrothermal synthesis, and solvothermal methods.

[0031] (Embodiment 2) In the second embodiment, an oxygen generating device 101 using the oxygen generating electrode catalyst explained in the first embodiment and an anode electrode (oxygen electrode) 12 used therein will be described with reference to FIG.

[0032] 14, the oxygen generating device 101 is equipped with an anode electrode (oxygen electrode) 12 and a cathode electrode (hydrogen electrode) 13, and its main components are an electrolytic cell 11 filled with water 14 containing an electrolyte, an oxygen gas separation tower 16, a dehumidifier 17, a water tank 18, a hydrogen gas separation tower 19, and a dehumidifier 20. A power source 15 is connected to the anode electrode 12 so that a positive voltage is applied to the cathode electrode 13, and a negative voltage is applied to the cathode electrode 13.

[0033] Water 14 is electrolyzed by applying a voltage to the anode electrode 12 and the cathode electrode 13 , and oxygen gas and hydrogen gas are generated within the electrolytic cell 11 . The generated oxygen gas is transported together with water vapor to the oxygen gas separation tower 16, where the moisture and oxygen gas are separated. The separated water is then poured into a water tank 18, and the oxygen gas is recovered as oxygen (O2) via a dehumidifier 17. Here, the separation of oxygen gas or oxygen gas from water can be performed using a density difference separation method that separates water, which is a high-density liquid, from gas, which is low-density, by the action of gravity or collision force, and / or a gas-liquid separation method that utilizes the surface tension of the liquid.

[0034] The generated hydrogen gas is transported together with water vapor to the hydrogen gas separation tower 19, where the water and hydrogen gas are separated. The separated water is then poured into a water tank 18, and the hydrogen gas is recovered as hydrogen (H2) via a dehumidifier 20. The water supplied via piping 21 and the water poured from the oxygen gas separation tower 16 and the hydrogen gas separation tower 19 are returned to the electrolytic cell via the water tank 18.

[0035] Here, the anode electrode 12 is an electrode using an electrode catalyst made of a non-precious metal material made of an oxide of a mixture containing manganese (Mn), iron (Fe), nickel (Ni), zinc (Zn), and silver (Ag) as described in embodiment 1, or an oxide of a mixture made of manganese (Mn), iron (Fe), nickel (Ni), zinc (Zn), and silver (Ag). That is, as the anode electrode 12, an electrode in which the material described in embodiment 1 is deposited on a collector electrode substrate such as titanium, or an electrode made of the material described in embodiment 1 is used.

[0036] The oxygen generating device 101 equipped with the anode electrode 12 using the oxygen generating electrode catalyst described in embodiment 1 is an oxygen generating device 101 that has excellent characteristics such as overvoltage, stability, durability, current density characteristics, specific charge, and large current characteristics in the weakly alkaline pH range, has high oxygen generating efficiency, a long life, and is also low in cost. EXAMPLES

[0037] The present invention will be described below with reference to examples, but the present invention is not limited to these examples.

[0038] Example 1 In Example 1, an example of searching for an electrode catalyst for oxygen evolution will be described. In Example 1, five types of metal elements were used to search for oxygen evolution electrode catalysts (QEE: Quinary-element electrocatalysts). Here, 11 kinds of metal element candidates were selected, namely, scandinium (Sc), titanium (Ti), zirconium (Zr), tantalum (Ta), manganese (Mn), iron (Fe), nickel (Ni), copper (Cu), zinc (Zn), tin (Sn) and silver (Ag). Five kinds of metal elements were selected from them, and the composition ratio, in terms of element number ratio, was set to 0.5 for one kind of metal element and 1 for the remaining four kinds. Therefore, the number of QEE candidates was about 3000, and from these, the ones considered to be optimal as electrode catalysts for oxygen evolution were extracted by the method (AI optimization algorithm) described in the first embodiment.

[0039] First, 10 compounds were randomly selected from this list of candidate materials, and the materials were synthesized using the method shown in Example 2 to evaluate their overvoltage characteristics (see Figure 3). During the preparation of the materials, these compounds were exposed to the air and turned into oxides. The 10 compounds were selected so that all 11 elements were always included in at least one compound.

[0040] Next, for the oxide compounds in the material candidate list, the properties were predicted using a regression method using Bayesian optimization based on the property evaluation results, and the top eight unsynthesized and unmeasured compound oxides were extracted, and materials were synthesized and properties were measured. This first loop was performed twice, and overvoltage data was collected for a total of 16 compound oxides. The results are shown in AI0-16 near the center of Figure 3. As a trend, it can be seen that compound oxides with smaller overvoltage properties are extracted as the sample evaluation progresses.

[0041] Then, machine learning using random forest was performed based on all the property evaluation results performed up to that point on the unsynthesized and unmeasured compound oxides in the material candidate list, and the material candidates were divided into active material candidates and inactive material candidates. Bayesian optimization was performed on the active material candidates, and the top two were selected from them, and materials were synthesized and properties were evaluated. This second loop was repeated four times, and properties were evaluated for a total of 16 samples (AI17 to 32 in Figure 3). Here, the active material candidates were samples with an overvoltage of less than 2.0 V, and the inactive material candidates were samples with an overvoltage of 2.0 V or more. For reference, FIG. 3 also shows the results of randomly selecting materials from the candidate materials, synthesizing the materials, and evaluating their properties (Random 1 to 32), as well as the results when RuO2 was used as a reference.

[0042] As a result, the QEE by the AI ​​of the present invention conforming to the first embodiment was 10 mA / cm 2 The overpotential defined by the current density of 1000 s was clearly lower than that of random extraction, and in a total of 32 material extraction trials (48 materials synthesized, including the data used to create the classification model), a material comparable to the overpotential when using RuO2 was obtained. Furthermore, MnFeNiZn 0.5 AgO x was extracted.

[0043] Example 2 In Example 2, the MnFeNiZn 0.5 AgO x The sample shows excellent overvoltage characteristics. x , MnFeNiZnAg 0.5 O x , Sc 0.5 MnFeNiAgO x and Ti 0.5 MnFeNiZnAgO x The samples were also partially evaluated and their properties were investigated in detail.

[0044] 1. Electrocatalyst Preparation An electrode catalyst made of a mixture of oxides of manganese (Mn), iron (Fe), nickel (Ni), zinc (Zn), and silver (Ag) was prepared by the following process. In addition, titanium (Ti) was used instead of Mn. 0.5 FeNiZnAg oxide, Sc using scandium (Sc) instead of Zn 0.5 The oxide of MnFeNiAg was also prepared.

[0045] First, 0.05M solutions of MnCl2, FeCl2, NiCl2, ZnCl2 and AgNO3 were prepared. Also, 0.05M solutions of TiCl2 and ScCl2 were prepared. Next, the solutions were mixed according to the ratio. 0.5 AgO x In preparing the mixture, the components were mixed in the following ratio: MnCl2:FeCl2:NiCl2:ZnCl2:AgNO3 = 1:1:1:0.5:1. The mixed solution was then drop-cast onto a Ti substrate measuring approximately 0.5cm x 2.0cm x 0.01cm in size, dried at room temperature (approximately 25°C), and pre-sintered in an oven at 350°C for 10 minutes. This step of drop-casting the mixed solution and pre-sintering was repeated nine times, and then sintered at 450°C for one hour to produce an electrode coated with an electrode catalyst containing five types of metals on a Ti substrate. Here, the pre-sintering and sintering were performed in air. Thus, oxidation was performed by natural oxidation, pre-sintering, and sintering.

[0046] For comparison, a RuO2 metal oxide catalyst was prepared. For the RuO2 metal oxide catalyst, the above mixture of five metal solutions was replaced with 0.05 M RuCl3, and the rest of the procedure was followed to prepare the RuO2 catalyst-coated electrodes. The mass loading of the electrode coated with the electrocatalyst on the Ti substrate was approximately 0.13 mg for all samples.

[0047] 2. Characterization 2.1 Material characterization Scanning electron microscope (SEM) photographs of the prepared electrode samples are shown in Figure 4. Figure 4(a) shows the MnFeNiZnAgO x , (b) MnFeNiZn 0.5 AgO x , (c) Ti 0.5 (d) MnFeNiZnAg 0.5 O x and (e) Sc 0.5 MnFeNiAgO x Nanoparticles were observed in both cases, and the surface had irregularities. Here, the SEM used was SU8000 (manufactured by Hitachi, Ltd.). XRD (X-ray Diffraction) measurements of each sample revealed that the crystal structure of the prepared samples consisted of cubic silver chloride with space group Fm-3m and hexagonal nickel oxide with space group R-3m, as shown in Figure 5. A MiniFlex X-ray diffractometer (manufactured by Rigaku) ​​was used for the measurements. MnFeNiZn 0.5 AgO x The results of cross-sectional observation of the sample using a high-resolution transmission electron microscope (HR-TEM) are shown in Figure 6(a). 0.5 AgO x The average thickness of the nanoparticles was 978.0±139.8 nm. As the black spots were observed, it was found that the nanoparticles were composed of nanoparticles. The HR-TEM used was JEM-2100F1 (manufactured by JEOL).

[0048] MnFeNiZn 0.5 AgO xThe results of detailed observation of the sample by Selected Area Electron Diffraction (SAED) are shown in Figures 6(b) and (c), which indicate that the materials other than the AgCl crystals are mainly amorphous or low-crystalline materials. The main SAED pattern, as shown in Figure 6(c), corresponds to the three-dimensional structure of AgCl, and this result is in good agreement with the XRD measurement result shown in Figure 5. MnFeNiZn 0.5 AgO x The results of observation with an annular dark field-scanning transmission electron microscope (ADF-STEM) are shown in Figure 6(d), and the results of element mapping using energy dispersive spectroscopy (EDS) are shown in Figures 6(e)-(k). Figures 6(e), (f), (g), (h), (i), (j), and (k) show the distribution of Mn, Fe, Ni, Zn, Ag, O, and Ti elements, respectively. These results indicate that the elements Mn, Fe, Ni, Zn and O are uniformly distributed throughout the electrocatalyst. On the other hand, most of the Ag element is in the form of nanoparticles, with only a small amount forming metal oxides with other metal species and dispersed therein. The ADF-STEM used was JEM-ARM200F-B (manufactured by JEOL).

[0049] Next, high-resolution X-ray photoelectron spectroscopy (XPS) was used to characterize the MnFeNiZn 0.5 AgO x The chemical composition and oxidation state of each element were investigated. The results are shown in Figure 7. Here, a Quantera SXM (Ulvac-PHI) was used for XPS. As a result, MnFeNiZn 0.5 AgO xThe XPS measurement results confirmed that the sample contains seven elements: Mn, Fe, Ni, Zn, Ag, Cl, and O. Figure 7 shows the spectra of the six elements: Mn, Fe, Ni, Zn, Ag, and O. As shown in FIG. 7(a), Mn 3+ 641.9 eV (Mn 2p 3 / 2 ) and 653.6 eV (Mn 2p 1 / 2 ), and Mn 4+ 646.7 eV (Mn 2p 3 / 2 ) and 658.4 eV (Mn 2p 1 / 2 ) and Fe has two oxidation states, Fe as shown in Fig. 7(b). 2+ 711.0 eV (Fe 2p 3 / 2 ) and 724.1 eV (Fe 2p 1 / 2 ), and Fe 3+ 713.0 eV (Fe 2p 3 / 2 ) and 726.1 eV (Fe 2p 1 / 2 ) It consists of two oxidation states. Ni 2p 3 / 2 As shown in Fig. 7(c), 2+ 856.2 eV and Ni 3+ It consists of two states at 858.9 eV, Zn 2p 3 / 2 As shown in Fig. 7(d), Zn 2+ 1022.0 eV and Zn 0 It consists of two states at 1023.6 eV. As shown in Figure 7(e), Ag 2+ 367.3 eV (Ag 3d 5 / 2 ) and 373.3 eV (Ag 3d 3 / 2 ), and Ag + 368.1 eV (Ag 3d 5 / 2 ) and 374.1 eV (Ag 3d 3 / 2 ) It consists of two oxidation states.

[0050] From the above results, MnFeNiZn 0.5 AgO x The sample contains five metal elements, Mn, Fe, Ni, Zn, and Ag, each of which has multiple valence states.4+ Or Ag 2+ It was shown that the valence states of MnFeNiZn and other species are mainly distributed in the valence state. This forms a highly active material, 0.5 AgO x It is presumed that this is the basis for the high oxygen evolution reaction.

[0051] In addition, as shown in Fig. 7(f), the O 1s spectrum was determined by the lattice oxygen (O L ), and oxygen vacancy (O V ) and 533.7 eV surface adsorbed oxygen / interlayer water (O W ) and deconvolution analysis is performed, 0.5 AgO x The oxygen (O) in the sample is oxygen vacancy (O V The area ratio of oxygen vacancies in the MnFeNiZn alloy was as high as 48 atomic %, indicating that the sample had a large number of oxygen vacancies. 0.5 AgO x It is presumed that this is one of the reasons why the oxygen evolution reaction is so high.

[0052] 2.2 Electrochemical characterization MnFeNiZnAgO x , MnFeNiZn 0.5 AgO x and MnFeNiZnAg 0.5 O x MnFeNiZnAgO x Materials based on Ti 0.5 FeNiZnAgO x and Sc 0.5 MnFeNiAgO x Regarding the material, its 10mA / cm 2 The overpotential at 1000 kV was evaluated. The measurement was carried out using a three-electrode beaker cell consisting of a working electrode with the above-mentioned material placed thereon, a reference electrode, and a counter electrode, and 0.1 M KOH as the electrolyte. The reference electrode was a silver-silver chloride electrode, and the counter electrode was platinum. As is clear from the results shown in Figure 8, MnFeNiZnAgO xThe MnFeNiZn system has low overvoltage and excellent properties as an electrode catalyst for oxygen evolution. 0.5 AgO x had the lowest overvoltage of 0.42 V and the most excellent overvoltage characteristics.

[0053] Next, MnFeNiZn 0.5 AgO x The electrochemical stability of the samples was evaluated by chronopotentiometry. One is the current (j) versus applied voltage (E-iR), and the other is 50mA / cm 2 The time dependence of the voltage E at which the hydrogen peroxide is obtained was evaluated by controlling the pH of the electrolyte with a buffer solution. Here, the voltage E is the voltage applied to a reversible hydrogen electrode (RHE). The results are shown in Figures 9 and 10, respectively. It can be seen from FIG. 9 that a large current can be obtained by setting the pH at 13 or higher. Here, as shown in FIG. 10, in an electrolyte environment controlled at pH ≦ 8 with phosphate buffer, the voltage E increased rapidly within 1 hour. This increase was due to the MnFeNiZn 0.5 AgO x It has been confirmed that the electrode made of the sample deteriorates rapidly during the oxygen evolution reaction, causing Mn, Fe and Ni to dissolve and flow into the electrolyte. On the other hand, in an electrolyte environment controlled to an alkaline pH of 9.2 or higher using carbonate or KOH buffers, the current is 50 mA / cm 2 The high durability and stability of the sample have been ensured, allowing it to maintain this temperature for more than three hours. It has been confirmed that the distribution and composition of all the elements in this sample remain almost unchanged before and after the experiment.

[0054] The electrochemical stability can be significantly enhanced by controlling the electrolyte environment at 80° C. and pH 9.2 with 1.0 M Na-Ci buffer. Figure 11 shows the MnFeNiZn 0.5 AgO xThe results of LSV (Linear Sweep Voltammetry) measurements performed on the sample under these conditions are shown below. The voltage E was 2.6 V relative to RHE, and the current was 1000 mA / cm. 2 was obtained.

[0055] Figure 12 shows the results of the MnFeNiZn 0.5 AgO x The sample was evaluated by chronopotentiometry, where the current was 400 mA / cm 2 The temperature was set to 0.05°C, and the electrolyte was added at the timing indicated by the arrow. Figure 12 also shows the results of RuO2 measured under similar conditions as a comparative example. As is clear from FIG. 0.5 AgO x The sample was 400mA / cm 2 The voltage E of the RuO2 sample was stable for more than 600 hours. On the other hand, the voltage E of the MnFeNiZn sample changed rapidly after 7 hours. 0.5 AgO x The superiority of the sample was confirmed.

[0056] Figure 13 is a visualization of the current density and specific charge of various oxygen evolution electrode catalysts as a map. Since oxygen evolution electrode catalysts are desired to have high current density and specific charge, the materials in the upper right corner are desirable. From Figure 13, MnFeNiZn 0.5 AgO x It can be seen that this material has excellent current density and specific charge characteristics as an electrode catalyst for oxygen evolution. The electrolyte environment and citations for each characteristic evaluation are summarized in Table 1.

[0057] [Table 1]

[0058] From the above examples, it is clear that the oxide of a mixture of manganese, iron, nickel, zinc, and silver, particularly MnFeNiZn 0.5 AgO xIt was confirmed that this material has excellent properties as an electrode catalyst for oxygen evolution.

[0059] Chlorine (Cl) is used in the material synthesis process. Therefore, the composition of each sample with good properties, including the chlorine content, was investigated using X-ray photoelectron spectroscopy. As a result, the aforementioned MnFeNiZnAg 0.5 O x MnFeNiZnAg 0.5 O2Cl 0.2 , MnFeNiZn 0.5 AgO x MnFeNiZn 0.5 AgO2Cl 0.25 , and Sc 0.5 MnFeNiAgO x Sc 0.5 MnFeNiAgO2Cl 0.1 It was. In addition, NiFeCuMnAgO2Cl extracted by the above search method 0.3 The catalyst had excellent activity, durability and corrosion resistance, and was particularly good in durability at high temperatures and when a corrosive electrolyte was used, and had excellent properties as an electrode catalyst for oxygen evolution. [Industrial Applicability]

[0060] Metal catalysts for oxygen generation are used in equipment that produces green hydrogen, which is seen as the next-generation energy source, and are also applied to fuel cells, FCVs, etc. The present invention provides a metal catalyst with high oxygen generation efficiency without using platinum group precious metal elements, which are rare, expensive, and limited in resource. It is expected that this will support next-generation industry and society, regardless of whether it is for consumer or industrial use, and greatly contribute to their development. [Explanation of symbols]

[0061] 1: Material candidate list (holder) 2: First regression analysis (Bayesian estimation analysis) 3: Characterization experiments (material synthesis and characterization) 4: Machine learning analysis (random forest analysis) 5: Active material candidates (holders) 6: Inactive material candidate (holder) 11: Electric field cell 12: Anode electrode (oxygen electrode) 13: Cathode electrode (hydrogen electrode, RHE) 14: Water (H2O) 15: Power supply 16: Oxygen gas separation tower 17:Dehumidifier 18: Water tank 19: Hydrogen gas separation tower 20:Dehumidifier 21: Piping

Claims

1. A method for searching for an oxygen generating electrode catalyst, the method comprising: searching for an oxygen generating electrode catalyst made of a metal oxide having n kinds of metal elements, the method comprising the steps of: A first step of selecting n+1 or more and 52 or less types of metal elements as metal element candidates of the electrode catalyst, and an element selected from the elements of the metal element candidates, metalloid elements, halogens, hydrogen, oxygen family elements, nitrogen family elements, and carbon family elements as an extended element candidate; a second step of creating a material candidate list including compounds composed of elements selected from the expanded element candidates and having n kinds of metal elements selected from the metal element candidates; a third step of randomly selecting a compound from the list of candidate materials; a fourth step of synthesizing an oxide of the compound selected in the third step, and then measuring the characteristics of the synthesized oxide to obtain a characteristic evaluation result; a fifth step of performing a property prediction by a regression method for oxides of compounds in the material candidate list based on the property evaluation results, extracting oxides of compounds that have not been synthesized and measured but have the highest performance from the material candidate list, synthesizing the materials and measuring the properties, and obtaining property evaluation results; A sixth step of performing a property prediction by a regression method for oxides of compounds in the material candidate list based on all the property evaluation results performed above, extracting oxides of compounds that have not been synthesized and measured but have a high performance from the material candidate list, and repeating material synthesis and property measurement multiple times; A seventh step of performing a classification method using machine learning based on all the property evaluation results performed on unsynthesized and unmeasured compounds in the material candidate list, and classifying the unsynthesized and unmeasured compounds in the material candidate list into active material candidates and inactive material candidates according to a predetermined criterion; An eighth step of performing a characteristic prediction by a regression method based on all or some of the characteristic evaluation results performed on the oxides of the compounds of the active material candidates, extracting oxides of compounds with top performance from the active material candidates in the material candidate list, synthesizing the materials and measuring their characteristics, and obtaining characteristic evaluation results; a ninth step of repeating the seventh and eighth steps a plurality of sets, with the seventh and eighth steps being one set; A method for searching for an electrode catalyst for oxygen evolution, comprising a tenth step of selecting an oxide of a compound exhibiting the most excellent properties from all the property evaluation results obtained above as an electrode catalyst for oxygen evolution.

2. The method for searching for an oxygen evolution electrode catalyst according to claim 1, wherein the regression method is one selected from the group consisting of Gaussian process regression, regression using a linear model, stochastic gradient descent, logistic regression, kernel ridge regression, decision tree, random forest, XGBoost, and neural network.

3. 3. The method for searching for an oxygen generating electrode catalyst according to claim 1 or 2, wherein the classification method is one selected from the group consisting of random forest, support vector machine, stochastic gradient descent, local neighborhood method, decision tree, XGBoost and neural network.

4. The method for searching for an oxygen evolution electrode catalyst according to claim 1 , wherein the number of the expansion element candidates is 9 or more and 72 or less.

5. The candidate elements are hydrogen (H), lithium (Li), boron (B), carbon (C), nitrogen (N), fluorine (F), sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), scandium (Sc), titanium (Ti), vanadium (V), chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu), zinc (Zn), gallium (Ga), germanium (Ge), arsenic (As), selenium (Se), bromine (Br), rubidium (Rb), strontium (Sr), yttrium (Y), zirconium (Zr), niobium (Nb), molybdenum (Mo), ruthenium (Ru), rhodium (Rh), and arsenic (Ar). ), palladium (Pd), silver (Ag), cadmium (Cd), tin (Sn), antimony (Sb), tellurium (Te), iodine (I), cesium (Cs), barium (Ba), lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), samarium (Sm), europium (Eu), gadolinium (Gd), terbium (Tb), dysprosium (Dy), holmium (Ho), erbium (Er), thulium (Tm), ytterbium (Yb), lutetium (Lu), hafnium (Hf), tantalum (Ta), tungsten (W), rhenium (Re), osmium (Os), iridium (Ir), platinum (Pt), gold (Au), mercury (Hg), and lead (Pb). The method for searching for an electrode catalyst for oxygen evolution according to claim 4.

6. The method for searching for an oxygen evolution electrode catalyst according to claim 1 , wherein the materials in the material candidate list are compounds composed of the metal element candidates.

7. 7. The method for searching for an oxygen generating electrode catalyst according to claim 1, wherein the evaluation index items for the characteristic evaluation are one or more selected from the group consisting of overpotential, Tafel gradient, potential dependence of change in Tafel gradient, charge transfer resistance, deterioration tendency, exchange current density, reaction current density at a specific potential, amount of oxygen generated per mass until deactivation, amount of charge per mass used for oxygen generation until deactivation, amount of oxygen generated per unit area until deactivation, amount of charge per mass used for oxygen generation until deactivation, electrochemically active surface area, pore distribution of the material, wettability of the material, amount of lattice oxygen defects introduced, increase ratio of current density between certain potentials, transfer coefficient, symmetry factor, rate-determining process, type and concentration of cations in the electrolyte, type and concentration of anion in the electrolyte, reaction intermediates in the rate-determining process and their adsorption energies, vacuum level, redox potential, band gap, Fermi level, valence, reaction rate constant, and reaction selectivity.

8. 8. A method for searching for an oxygen generating electrode catalyst according to claim 1, wherein the elemental ratio of the metal elements in the compound containing the n types of metal elements is half that of the other types of metal elements for only one type of metal element, and the elemental ratios of the other types of metal elements are all equal.

9. The method for searching for an oxygen evolution electrode catalyst according to claim 8, wherein n is 5.

10. An oxygen generating electrode catalyst discovered by the method for discovering an oxygen generating electrode catalyst according to any one of claims 1 to 9.

11. MnFeNiZnAg ０．５ O ２ C ０．２ , MnFeNiZn ０．５ AgO ２ C ０．２５ , Sc ０．５ MnFeNiAgO ２ C ０．１ , and NiFeCuMnAgO ２ C ０．３ An electrode catalyst for oxygen generation, comprising at least one compound selected from the group consisting of:

Images