JP4029155B2 - Visible-light-responsive membranous porous semiconductor photoelectrode - Google Patents

Description

  The present invention uses a photoelectrochemical cell that converts light energy such as sunlight into chemical energy such as hydrogen, a highly efficient porous-structured semiconductor photoelectrode used in the cell, and the photoelectrochemical cell. It relates to a method for decomposing water.

  Although effective use of renewable energy such as solar energy is very important, it is difficult to realize an inexpensive and high-performance optical energy conversion system. In particular, a technique for producing hydrogen by decomposing water using solar energy is a technique that is indispensable for the early commercialization and acceleration of hydrogen fuel cell vehicles. Examples of the technology for decomposing water with sunlight include the following.

First, although it is a solar cell-electrolysis system, since the power generation cost of a solar cell is very expensive, even if it combines with electrolysis, it is impossible to manufacture hydrogen cheaply.
As the next hydrogen production, there is a thermochemical decomposition method of water, but high heat of 700-800 degrees or more is required, and the cycle rate is very slow.
The photodecomposition method of water using a photocatalyst has been widely studied in recent years, but its performance is very low. A photocatalyst with a high quantum yield cannot convert solar energy because it can use only ultraviolet rays hardly contained in sunlight of 300 nm or less. Although there is a report of water splitting with a photocatalyst using ultraviolet rays up to 400 nm such as TiO ₂ , the solar energy conversion efficiency is 0.03% or less at the maximum, which is very low. Some visible light-responsive semiconductor photocatalysts are known, but most are reactions of reducing agents and oxidizing agents (so-called sacrificial reagents) such as methanol and silver nitrate, and solar energy conversion is not possible.
More recently, although the water under visible light were reports that decomposed into hydrogen and oxygen by NiOx-In1-xNixTaO _4, the quantum yield was 0.66% for 400 nm, 0 as solar energy conversion efficiency It is estimated that it is less than 0.01%. Moreover, the decomposition of water has not been successful even with non-oxide photocatalysts.
Thus, it can be said that the efficient solar energy conversion using visible light has hardly been achieved in the photocatalyst field. Also, since hydrogen and oxygen are mixed at the same time, the separation cost is high and the risk of explosion is high.

On the other hand, solar energy conversion by a semiconductor photoelectrode has also been studied for a long time (FIG. 1). When a semiconductor photoelectrode and a Pt counter electrode are combined and irradiated with light while applying a bias, holes generated in the valence band on the semiconductor oxidize water to generate oxygen, and the electrons generated on the conduction band are counter electrodes. Water can be reduced by reducing water on Pt. There is an advantage that hydrogen and oxygen can be generated separately. In such a semiconductor, an oxide semiconductor such as TiO ₂ is stable and inexpensive, but only light having a short wavelength such as ultraviolet light can be used. When a non-oxide semiconductor such as silicon is used, most of the visible light can be used, so the performance is relatively high, but the cost is high and the semiconductor is likely to deteriorate. In view of cost and stability, an oxide system is still desirable. Although a visible light responsive photoelectrode of a doped oxide semiconductor has been studied for a long time, its performance has been very low. Conventional visible light responsive semiconductor photoelectrodes have been studied mainly with electrodes having a thickness of several millimeters and a low surface area, such as single crystals and high-temperature powder sintered bodies. FIG. 1). Since charges, especially holes, are difficult to move to the surface, charge recombination progressed before the reaction occurred on the electrode surface, and the performance could not be improved.
A conventional visible light responsive photoelectrode uses pellet-type (millimeter-thickness) high-temperature-sintered semiconductor powder, the back surface is coated with an ohmic contact metal such as indium, and a conductive wire is connected. Is solidified with an adhesive or the like so that it does not come into contact with the metal or lead (J. Solid State Electrochem. 2 (1998) 176, Solar Energy Materials 21 (1991) 335, Nouv. J. Chimie 4 ( 1980) 501, Chem. Phys. Lett., 77 (1981) 6.).

However, in recent years, several visible light-responsive porous membrane electrodes have been reported with respect to semiconductor photoelectrodes. It was found that water can be decomposed with very high efficiency using a porous semiconductor photoelectrode made of Fe ₂ O ₃ or WO ₃ [J. Phys. Chem. B, 103 (1999) 7184 (Non-Patent Document 1), J. Am. Am. Chem. Soc. 123, (2001), 10639 (Non-Patent Document 2), WO01026224A1 (Patent Document 1)]. In the porous semiconductor photoelectrode, water as a reaction substrate can penetrate into the electrode. That is, since the diffusion distance of holes generated inside the semiconductor photoelectrode is smaller than that of the conventional electrode, high efficiency can be expected (FIG. 2). However, these visible light responsive simple oxide semiconductors have a drawback of having a large conduction band level. In other words, in order to advance the reaction of reducing protons to hydrogen using electrons generated in the conduction band, it was necessary to input large electric energy from the outside. This means that the net solar energy conversion efficiency is reduced. In order to aim for practical use in such a visible light-responsive film-like porous photoelectrode, a technique for greatly improving the solar energy conversion efficiency is required. Further, since the use conditions are limited such that WO ₃ is stable only when it is acidic and Fe ₂ O ₃ is stable only when it is alkaline, development of a new stable material is desired. In addition to the above properties, the new semiconductor material must be capable of charge separation and oxygen generation efficiently.

J. et al. Phys. Chem. B, 103 (1999) 7184 J. et al. Am. Chem. Soc. 123, (2001), 10639 WO0102624A1

  The present invention has been made in view of the above circumstances, and in a photoelectrochemical system in which a specific semiconductor photoelectrode and a counter electrode are combined, visible light responsiveness is controlled by controlling the conductivity and valence level of the semiconductor. While maintaining the charge mobility, it prevents the conduction band level from becoming positively large, and by using the porous semiconductor film-like semiconductor photoelectrode, the charge diffusion distance is reduced, and the solar energy It is an object to provide a technique for greatly improving the conversion efficiency.

According to the present invention, the following film-like semiconductor photoelectrode, photoelectrochemical cell and water decomposition method are provided.
(1) A film-like semiconductor photoelectrode used in a photoelectrochemical cell that performs an energy storage type reaction by water decomposition reaction or redox reaction, and is composed of a composite metal oxide semiconductor having a porous structure that is responsive to visible light. A film-like porous semiconductor light comprising two or more metal elements, at least one of which is selected from bismuth, silver, copper, tin, lead, indium, praseodymium and nickel electrode.
(2) A film-like semiconductor photoelectrode used in a photoelectrochemical cell that performs an energy storage-type reaction by water decomposition reaction or redox reaction , and includes visible light containing one or more elements selected from nitrogen and sulfur A membranous porous semiconductor photoelectrode comprising an oxygen-containing compound semiconductor having a responsive porous structure.
(3) doped with 20 to 0.5 mol% of one or more elements selected from chromium, nickel, iron, silver, lead, copper, vanadium and bismuth, and in antimony, bismuth, vanadium, niobium and tantalum The membranous porous semiconductor photoelectrode according to the above (1), which is co-doped with one or more elements selected from:
(4) The membranous porous semiconductor photoelectrode according to any one of (1) to (3), wherein a diffusion distance of 50% or more of holes generated in the semiconductor to the semiconductor surface is within 500 nm.
(5) The membranous porous semiconductor photoelectrode according to (1), comprising a composite metal oxide based semiconductor having a visible light responsive porous structure containing both bismuth and vanadium.
(6) The membranous porous semiconductor photoelectrode according to any one of (1) to (5) , which is formed on a light-transmitting substrate.
(7) The film-like porous semiconductor photoelectrode according to any one of (1) to (6), wherein the thickness of the semiconductor is 50 μm or less.
(8) A photoelectrochemical cell using the film-like semiconductor photoelectrode according to any one of (1) to (7) as the semiconductor photoelectrode.
(9) The photoelectrochemical cell according to (8), wherein the energy storage type reaction is a water decomposition reaction.
(10) A method for decomposing water, comprising decomposing water into hydrogen and oxygen using the photoelectrochemical cell according to (8) .

  According to the present invention, a simple system that can efficiently convert visible light of sunlight into chemical energy such as hydrogen can be realized. An important point of the present invention is that it is possible to provide a film-like porous semiconductor photoelectrode having a visible light responsiveness that facilitates charge transfer or has a relatively high conduction band level (negatively large). Even in an electrode with a low quantum yield in the examples, the quantum yield is close to 100% by further optimizing the film forming method. As a result, it becomes possible to efficiently extract hydrogen from inexhaustible sunlight and water, and one step closer to realizing a hydrogen energy society.

The present invention will be described in detail below.
As the visible light responsive semiconductor used in the present invention, an oxide-based semiconductor containing an oxygen atom is preferably used in consideration of stability and economy. The visible light responsiveness in this case means not only the ability to absorb visible light but also the property that charges generated by irradiation with visible light can be used for the reaction.
In general, oxide semiconductors such as TiO ₂ , SrTiO ₃ , Ta ₂ O ₅ , and WO ₃ are formed with a 2p orbital in which a conduction band is a d-orbital of a transition metal atom and a valence band is oxygen. The formal charges of these transition metal atoms are Ti ⁴⁺ , Ta ⁵⁺ , W ⁶⁺ , and the d orbit is in a state without electrons. When irradiated with light, oxygen 2p orbital electrons are excited into transition metal vacant d orbitals. That is, a hole which is a shell of electrons is formed in the 2p orbit of oxygen. In such a semiconductor, if all the 2p orbitals of oxygen are at the same level, the conduction band level is governed by the size of the band gap. When it is made smaller, the conduction band level shifts significantly larger. There is a dilemma that as the conduction band level shifts to a large positive value, more external bias is required to generate hydrogen at the counter electrode, and the net energy conversion efficiency becomes worse. A method of eliminating this dilemma is to shift the valence band level negatively by controlling the valence band with the orbits of atoms other than oxygen. The conventional semiconductor electrode (single crystal or high-temperature sintered body) has been tried to control the valence levels by doping transition atoms, etc., or substituting atoms into the ultraviolet-responsive semiconductor electrode and to provide visible light response. The performance was very small. The reason is that holes in the valence band are less likely to move than electrons in the conduction band, and when the doping amount is small, the doping level exists spatially away, so holes move from the inside of the semiconductor bulk to the surface. Because it is difficult to do.

In order to solve this problem, the inventors have arrived at the present invention on the assumption that if the semiconductor is micronized, the distance that the electric charge moves to the surface is shortened and the oxidation reaction becomes smooth. That is, an electrode having a structure in which a semiconductor is used as a porous film and an electrolyte solution penetrates into the film is good. Such electrodes have been reported for simple oxides such as Fe ₂ O ₃ and WO ₃ , but have not been studied for complex complex oxide semiconductors controlled by valence electrons. In the present invention, for the purpose of smoothing the movement of electric charges and controlling the semiconductor band structure, (i) a visible light-responsive complex oxide semiconductor containing two or more metal elements, and (ii) an anion other than oxygen A visible light responsive oxygen-containing compound semiconductor containing a part of the reactive element (S or N), or (iii) a relatively large amount of doped oxide-based semiconductor.
Note that the composite oxide in this case is a substance that includes two or more kinds of metal elements and whose crystal structure can be defined. Doping means putting a different element into the host compound crystal within a range where the basic crystal structure of the host compound hardly changes.

  As the type of semiconductor, a visible light responsive complex metal oxide semiconductor in which the upper part of the valence level basically includes the level of an element other than an oxide is preferable. This composite metal oxide semiconductor is preferably composed of two or more kinds of metal elements, and at least one element A thereof is bismuth (Bi), silver (Ag), copper (Cu), tin (Sn), It is selected from lead (Pb), vanadium (V), indium (In), praseodymium (Pr), chromium (Cr), and nickel (Ni). In particular, bismuth, silver, tin, and nickel are preferable. The combination of two or more metal elements can indicate a combination of element B with respect to element A in addition to a combination of elements A. In this case, the element B is selected from Ti, Nb, Ta, Zr, Hf, Mo, W, Zn, Ga, In, Ge, and Sn.

In a general oxide semiconductor, the valence band is made of oxygen 2p orbitals. The upper level of this valence band determines the band gap. In order to reduce the band gap, another atomic orbital may be present at the same level as or above the upper part of the valence band (electrochemically negative direction). In addition, even if an atom having an orbit below the upper level of the valence band is present, if oxygen and the orbit of the atom hybridize, there is a possibility that a new level can be raised and used. Such orbital hybrids can move charges smoothly.
What kind of atomic compound should be used or doped can be roughly discriminated because of recent advances in computer science (such as density functional theory). A preferred semiconductor used in the present invention is specifically a visible light-responsive composite oxide semiconductor containing one or more elements having an electronic state partially filled with d orbitals, such as chromium, nickel, iron, A visible light responsive complex oxide semiconductor containing one or more elements selected from bismuth, silver, tin (preferably form II valence), lead (preferably form II valence), vanadium, indium and praseodymium. is there.

  The semiconductor used in the present invention may be a visible light responsive oxygen-containing compound semiconductor containing one or more elements selected from nitrogen and sulfur. For example, oxynitrides and oxysulfides. Carbons such as oxycarbides may also be included. In the semiconductor described above, oxygen defects and lattice distortion may occur as a result of doping and element substitution, which may change the valence level.

  Specific examples of combinations of metal elements in the composite metal oxide semiconductor used in the present invention include, for example, Bi / V, Ag / Nb, In / Ni / Ta, Ag / Pr / Ti, and Rb / Pb. / Nb, In / Zn, Bi / Mo, Bi / W, Ag / V, Pb / Mo / Cr, In / Zn / Cu, Na / Bi, K / Bi, and the like.

Specific examples of the composite metal oxide semiconductor include the following.
_{BiVO 4, AgNbO 3, AgPrTi 2} O 6, RbPb 2 Nb 3 O 10, In 2 O 3 - (ZnO) 3, Bi 2 MoO 6, Bi 2 WO 6, Ag 3 VO 4, In 2 -xZnxCu 2 O 5 (X = 0 to 1), ABiO ₂ (A is a monovalent metal such as Na, K, Li and Ag), ABiO ₃ (A is a monovalent metal such as Na, K, Li and Ag) and the like.

The oxygen-containing compound semiconductor containing nitrogen and / or sulfur is composed of (i) a metal, (ii) oxygen, and (iii) N and / or S elements. The metal elements in this case include Ta, Sm, Ti, Nb, Zr, Hf, Mo, W, Zn, Ga, In, Ge, Sn, Bi, V, Pb, and the like.
Specific examples of the oxygen-containing compound semiconductor include TaON, Sm ₂ Ti ₂ S ₂ O ₅ , BaNbO ₂ N, SrTaO ₂ N, LaTaON ₂ , Zr ₂ ON ₂ , Na ₂ TiOS ₂ , ZrOS, and Li _7.2 Ti. _0.8 O _1.6 N _2.4 , Ta ₅ O _1.81 N _4.79 , Ta _0.48 Zr _0.52 CaO _2.52 N _{0.48, and the} like.

In the present invention, an oxide-based semiconductor having a structure doped with metal X and co-doped with metal Y can also be preferably used.
In this case, the metal X is selected from among chromium (Cr), nickel (Ni), iron (Fe), silver (Ag), lead (Pb), copper (Cu), vanadium (V) and bismuth (Bi). At least one selected is used.
On the other hand, as the metal Y, at least one selected from antimony (Sb), bismuth (Bi), vanadium (V), niobium (Nb), and tantalum (Ta) is used. The ratio of the metal X to the metal Y is an atomic ratio [X] / [Y] and is 0.2 to 5, preferably 0.5 to 2.
Examples of the metal Z contained in the oxide semiconductor include titanium (Ti), tantalum (Ta), zirconium (Zr), hafnium (Hf), molybdenum (Mo), tungsten (W), zinc (Zn), and gallium (Ga ), Isodium (In), germanium (Te), tin (Sn), bismuth (Bi), and the like.

Specific examples of the undoped oxide semiconductor include the following.
(1) Simple _{oxides: TiO 2, Ta 2 O 5} , ZrO 2 and the like.
(2) Complex oxide (i) Ti-based: SrTiO ₃ , K ₂ La ₂ Ti ₃ O ₁₀ , Rb ₂ La ₂ Ti ₃ O ₁₀
CsLa ₂ Ti ₂ NbO ₁₀ , Na ₂ Ti ₆ O ₁₃ , BaTi ₄ O _{9 and the} like.
(Ii) Nb series: K ₄ Nb ₆ O ₁₇ , Rb ₄ Nb ₆ O ₁₇ , Sr ₂ Nb ₂ O ₇ , Na ₄
Nb ₈ P ₄ O ₃₂ , KCa ₂ Nb ₃ O _{10 and the} like.
(iii) Ta series: KTaO ₃ , NaTaO ₃ , BaTa ₂ O ₆ , Rb ₄ Ta ₆ O _{17
, K 3 Ta 3 Si 2 O} 13, Na 4 Ta 8 P 4 O 32, K 2 Ta 4 O 11, K 2 S
rTa ₂ O ₇ , Sr ₂ Ta ₂ O ₇ , RbNdTa ₂ O ₇ , LaTaO _{4 and the} like.
(Iv) In-based: CaIn ₂ O ₄ , SrIn ₂ O _{4 and the} like.
(V) Sn series: Sr ₂ SnO ₄ , Ca ₂ SnO _{4 and the} like.
(Vi) Ga-based: CaGa ₂ O ₄ , SrGa ₂ O ₄ , ZnGa ₂ O _{4 and the} like.
(vii) Ge-based, Zn ₂ GeO ₄ or the like.
(Viii) Sb-based, NaSbO ₃ , KSbO _{3 and the} like.

A combination of an oxide semiconductor and a doped metal is as follows.
(1) Semiconductor: TiO ₂
Metal X: Cu
Metal Y: Sb
(2) Semiconductor: TiO ₂
Metal X: Cr
Metal Y: Bi
(3) Semiconductor: TiO ₂
Metal X: Ni
Metal Y: Sb
(4) TiO ₂ / Cr, Sb,
(5) Ta ₂ O ₅ / Cr, Sb,
(6) NaTaO ₃ / Cr, Sb

  In the oxide-based semiconductor doped with the metals X and Y, the doping amount of the metal X is 0.5 to 20 mol%, preferably 5 to 10 mol%, relative to 1 mol of the semiconductor compound. The co-doping amount of the metal Y to be co-doped is based on an amount in which the balance of the charge of the doping species matches, but may be deviated from that.

  If the doping element in the doping compound forms a valence level and the level is separated from the valence band formed by oxygen by 0.03 eV (excitation energy at room temperature) or more, the doping amount must be relatively large. For example, holes are difficult to move. Examples of the doping compound include elements having an electronic state partially filled with d orbitals such as chromium and nickel, vanadium, bismuth, silver, and tin. When a doping species having a valence different from that of the host semiconductor metal is used, it is desirable to co-dope another valence metal for neutralizing the charge. Examples of co-doped species include antimony, bismuth, vanadium, niobium, and tantalum.

  In order to shorten the movement distance of holes, it is desirable that the semiconductor is small. The shape may be spherical or rod-like, but a porous film having a diffusion distance to the semiconductor surface of 50% or more, preferably 80% or more of the holes generated in the semiconductor is 500 nm or less, preferably 10 nm or less. In order to promote charge separation and suppress charge recombination, crystallinity needs to be high, and amorphous is not preferable.

  Next, an explanatory structural diagram of the membranous porous semiconductor of the present invention is shown in FIGS.

FIG. 3 shows an example in which the pores (holes) extend vertically.
FIG. 3A shows an explanatory cross-sectional view taken along the line aa ′ in FIG. 3B, and FIG. 3B shows a plan view of the porous semiconductor film.
In FIG. 3, 1 is a substrate, 2 is a semiconductor film, 3 is a hole, 4 is a hole, 5 is a hole, and 6 is a particulate semiconductor. The arrows in FIG. 3 indicate the diffusion distance between the holes 3 and the semiconductor surface.

FIG. 4 shows an example where the semiconductor is composed of fine particles.
FIG. 4 is an explanatory cross-sectional view of the porous semiconductor film.
In FIG. 4, 1 is a substrate, 2 is a semiconductor film, 3 is a hole, and 4 is a hole. The arrows in FIG. 4 indicate the diffusion distance between the holes 3 and the semiconductor surface.

Since the semiconductor used in the present invention has a porous structure, the electrolytic solution is mostly in contact with the substrate through the semiconductor pores. If there is a possibility that leakage current will occur when the surface of the substrate comes into contact with the electrolytic solution, the performance is improved by covering the substrate surface with a dense film having almost no pores.
In order to efficiently move the electrolyte solution inside the membrane and diffuse the product such as oxygen, it is desirable that the pore size of the semiconductor membrane is relatively large, but if it is too large, the membrane strength will be weakened, It becomes difficult to move charges. A state in which pores of various sizes of 5 nm to 500 nm are combined is preferable. In order to control the pore diameter, it can be adjusted by the molecular weight or mixing amount of the organic substance to be mixed when the film is baked and formed.
In the semiconductor particles constituting the semiconductor film, large particles cause light scattering and have an effect of increasing light absorption efficiency.

  The semiconductor is usually formed on a substrate. In this case, a transparent conductor such as conductive glass or conductive plastic is the best as the semiconductor electrode substrate. Of these, heat-resistant tin oxide conductive glass is preferable. The reason why the transparent conductor is good is that the light can be irradiated from the substrate side, so that the electron moving distance is shortened and charge recombination can be suppressed. However, if the semiconductor film thickness is thin, a non-transparent substrate such as a metal or carbon plate may be used.

The semiconductor film thickness is sufficient if it has a film thickness that can absorb light sufficiently. If it is thicker than that, problems such as cracks occur, and solution transportation and product transportation are hindered, leading to performance degradation. The semiconductor film thickness is 50 micrometers (μm) or less, preferably 20 μm or less, and more preferably 0.1 to 5 micrometers (μm).
Moreover, in the semiconductor particle which comprises the semiconductor of a porous structure, when this particle | grain is spherical shape or powder form, the average particle radius is 3-50 nm, Preferably it is 10-300 nm. On the other hand, when the particles are columnar, the average radius of the columns is 3 to 500 nm, preferably 10 to 300 nm. In the case of a hollow body with a vertical hole, the average thickness of the hollow body wall is 3 to 1000 nm. Preferably it is 10-600 nm.

As a method for preparing a semiconductor film, a metal precursor such as a sol-gel method or a complex polymerization method is dispersed in a solvent and thermally decomposed (baked) after coating, or semiconductor fine particles are prepared in advance by a solid phase method or the like. There is a method of making a paste and thermally decomposing (baking) after application. If the melting point is low, a solid phase method may be used. As a coating method, screen printing, a doctor blade method, a spin coating method, a spray method, a dip coating method, or the like can be used.
The firing temperature must basically be a temperature at which the organic matter mixed above is decomposed. However, since the substrate has heat resistance, it is desirable that the heat resistance temperature (about 600 ° C.) or less be used for the tin oxide based conductive glass. It is also effective to fire in oxygen in order to accelerate the decomposition of organic matter.

  An oxygen-containing compound semiconductor containing nitrogen, sulfur, or carbon can be synthesized by treating the oxide film with ammonia or hydrogen sulfide later. Alternatively, the precursor oxide and the nitrogen-containing compound or sulfur-containing compound may be mixed and fired. The amount of N or S is at least 0.5 atomic%, preferably 10 atomic% or more with respect to 1 atom of oxygen in the oxide semiconductor. The upper limit is usually 80 atomic%.

The prepared porous semiconductor film can improve performance by post-treatment. In the case of a Ti-based semiconductor, it is immersed in a solution containing Ti such as TiCl ₄ , in an Nb or Ta-based solution such as an alkoxide or chloride, in a Bi-based solution in a solution containing bismuth ions, and finally heat treated, Lattice defects and necking between particles are improved.
The electrode may be mixed with not only the various visible light responsive semiconductors described above but also other conductive materials. In this case, the conductive material may be an inorganic material such as SnO ₂ , TiO ₂ , WO ₃ , In ₂ O ₃ —SnO ₂ or the like, and an organic material such as a conductive polymer.

  The semiconductor film described above can be a single semiconductor or a stack of different semiconductor layers. In the case of stacking, it is necessary to consider the band potential so that charges can move.

  A conducting wire is attached to the semiconductor electrode thus prepared. In order to lower the resistance, it is desirable to attach a current collecting electrode such as a comb electrode. Indium solder is effective for reducing the contact resistance.

The photoelectrochemical cell of the present invention has a structure in which the above-mentioned visible light-responsive semiconductor electrode and a counter electrode are combined, and can advantageously carry out an energy storage type reaction.
The energy storage type chemical reaction in this case indicates a reaction in which the free energy change (ΔG) of the reaction is positive. The basic reaction is a water decomposition reaction. In addition to this, this cell can also be used, for example, for the decomposition reaction of HI and HBr and the iodine redox reaction.
The semiconductor photoelectrode of the present invention can also be used for a solar cell electrode in combination with a redox reaction such as halogen.
As for the composition of the electrolytic solution present in the photoelectrochemical cell, a stable supporting electrolyte such as sodium sulfate, sulfuric acid, perchlorate, or sodium hydroxide is used in the case of water decomposition. Basically, it may be close to a saturated solution at 0.1 mol / l or more. The pH condition is used in a region where the semiconductor electrode is stable. In the decomposition reaction of HI and HBr, HI and HBr are dissolved in these supporting electrolytes. At this time, an organic solvent may be used instead of an aqueous solvent.
The electrode of the present invention can also be used for hydrogen production from waste liquid containing substances that are easily oxidized such as organic matter and biomass. In this case, organic substances and the like are oxidized by holes on the semiconductor, and the efficiency of hydrogen production can be improved. In addition, when the semiconductor electrode of the present invention exhibits p-type characteristics, a reduction reaction such as hydrogen generation can be performed on the semiconductor electrode and an oxidation reaction can be performed on the counter electrode.

  As the counter electrode, an appropriate material suitable for the reaction is used. For hydrogen generation, Pt or carbon having a low hydrogen generation overvoltage is effective, but an inexpensive Co—Mo electrode can also be used.

The bias potential is controlled by a potentiostat or a source meter. The light irradiation is finally sunlight, but an artificial light source such as a xenon lamp is used when it is desired to check the performance of the electrode or to perform a special chemical reaction.
Further, instead of applying a bias by a power source, a bias can be applied by connecting a solar cell or using another semiconductor electrode as a counter electrode.

  Next, the present invention will be described in detail by examples.

Example 1
(1) Preparation method of BiVO ₄ electrode Bismuth 2-ethylhexanoate (Bi (hex) ₃ )) and vanadium oxyacetylacetonate (VO (acac) ₂ ) were dissolved and mixed in acetylacetone at a stoichiometric ratio of 1: 1. . After stirring for 1 hour, the mixture was concentrated with an evaporator, and 50 vol% of polyethylene glycol was added. The obtained solution was applied to conductive glass (F-SnO ₂ , 10 ohm / sq) by a doctor blade method, and air baked at 500 degrees for 1 hour. This was repeated three times. The film thickness was about 0.3 micrometers. By SEM observation, it was found that the large pores were films having pores of 100 to 200 nm. XRD confirmed that monoclinic BiVO ₄ was made.

(2) Evaluation of electrode This electrode was connected to a potentiostat. Ag / AgCl was used for the reference electrode, and Pt was used for the counter electrode. Water decomposition reaction was carried out with 0.1 mol / l Na ₂ SO ₄ aqueous solution. Monochromatic light of various wavelengths was irradiated while changing the bias voltage. The band gap was about 2.4 eV, and the open circuit voltage was +0.2 V (vs. Ag / AgCl). Compared to the band gap of the WO ₃ (about 2.7 eV) and open circuit voltage (+ 0.25 V), the conduction band level is negative is more of BiVO ₄ than WO _3, it can be seen that BiVO ₄ is preferred. The reason why the 6s orbital of Bi constitutes the upper part of the valence band seems to be the reason for the superiority. At 400 nm, the quantum yield (QE) was 33% at an open circuit voltage of 0.6 V, and the QE was 64% at 0.8 V, which was very high efficiency. It can be seen that the solar energy conversion efficiency is about 0.36%, which is more effective than a photocatalytic system that cannot decompose water.

Example 2
(1) Preparation of AgNbO ₃ electrode AgNO ₃ dissolved in methanol and Nb ethoxide dissolved in ethanol were mixed at a stoichiometric ratio of 1: 1. After stirring, 50 vol% of polyethylene glycol was added. The obtained solution was applied to conductive glass (F-SnO ₂ , 10 ohm / sq) by a doctor blade method and air baked at 550 ° C. for 1 hour. This was repeated 5 times. The film thickness was about 0.2 micrometers. It was confirmed by XRD that AgNbO ₃ was formed.

(2) Electrode Evaluation The electrode evaluation method is the same as in Example 1. The open circuit voltage was difficult to observe because it was hidden behind the silver redox peak, but it was 0.2 V or less. In the band calculation using the density functional method (CASTEP), the conduction band uses the d orbit of Nb. Therefore, it can be said that the conduction band is negative compared to WO ₃ and Fe ₂ O ₃ . As a result of the measurement, at 400 nm, the quantum yield (QE) was 1.4% at 0.5 V with respect to the open circuit voltage. At 420 nm with no absorption in undoped TiO ₂ , the quantum yield (QE) was 0.4%.

Example 3
(1) Preparation of TiO ₂ electrode doped with Cr and Sb Titanium isopropoxide, chromium nitrate (Cr: 2.3 mol%), and antimony chloride (Sb: 3.5 mol%) were mixed in an ethanol solvent. The obtained solution was applied to conductive glass (F-SnO ₂ , 10 ohm / sq) by a doctor blade method, and air baked at 500 degrees for 1 hour. Thereafter, it was immersed in an aqueous TiCl ₄ solution (0.2 mol / l) for 18 hours, and air baked at 500 ° C. for 1 hour. This was repeated three times. The film thickness was about 1 micrometer.

(2) Electrode Evaluation The electrode evaluation method is the same as in Example 1. The open circuit voltage was 0 V (vs. Ag / AgCl) which was almost the same as that of undoped TiO ₂ . In the band calculation using the density functional method (CASTEP), since the valence level is the d orbital of Cr and the conduction band is the d orbital of Ti, the conduction band is WO ₃ or Fe ₂ O ₃ . It can be said that it is negative in comparison.
At 420 nm with no absorption in undoped TiO ₂ , the quantum yield (QE) was 1% at 0.6 V with respect to the open circuit voltage.

Example 4
(1) Preparation of Bi ₂ WO ₆ electrode Bismuth nitrate dissolved in an aqueous nitric acid solution and ammonium tungstate were mixed at a stoichiometric ratio of 1: 1. After stirring, the mixture was allowed to stand for 5 days to form a sol. The obtained sol solution was dispersed with ultrasonic waves, and then applied to conductive glass (F-SnO ₂ , 10 ohm / sq) by a doctor blade method, followed by air baking at 550 ° C. for 1 hour. This was repeated 5 times. It was confirmed that Bi ₂ WO ₆ was mainly made by XRD.

(2) Evaluation of electrode The evaluation of the electrode is the same as in Example 1. The open circuit voltage was -0.12 V, which was the most negative in the present example.
At 400 nm, the quantum yield (QE) was 1.8% at 0.9 V with respect to the open circuit voltage.

(Example 5)
It shows the preparation of N-doped TiO ₂ electrode. Ammonia was added to TiCl ₃ to cause precipitation, which was calcined at 450 degrees. The color was yellow and the absorption extended to 470 nm. This was mixed with acetylacetone and Triton X to form a paste, and applied to conductive glass by the doctor blade method. Further, TiCl ₄ treatment was performed. In the case of TiO _{2 having} no visible light response at 440 nm (P-25, Nippon Aerosil), QE was 0.2% or less, but the N-doped electrode showed 0.3% or more under the same conditions. It was. It was found that the relative quantum yield based on 400 nm shows that the N-doped electrode exhibits a performance 6 times or more higher than that of undoped at 440 nm.

It is explanatory drawing of the conventional semiconductor photoelectrode. It is explanatory drawing of the manufacture principle of hydrogen by the photoelectrode using a porous semiconductor film. An explanatory structural view of an example of a porous semiconductor film is shown. The explanatory sectional view about other examples of a porous semiconductor film is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Substrate 2 Semiconductor film 3 Hole 4 Hole 5 Hole 6 Particle semiconductor

Claims (10)

A film-like semiconductor photoelectrode used in a photoelectrochemical cell that performs an energy storage type reaction by water decomposition reaction or redox reaction, and is composed of a composite metal oxide semiconductor having a porous structure with visible light response. A film-like porous semiconductor photoelectrode comprising the above metal elements, wherein at least one of the metal elements is selected from bismuth, silver, copper, tin, lead, indium, praseodymium and nickel. A film-like semiconductor photoelectrode used in a photoelectrochemical cell that performs an energy storage type reaction by water decomposition reaction or redox reaction , and having visible light responsiveness containing one or more elements selected from nitrogen and sulfur A film-like porous semiconductor photoelectrode comprising an oxygen-containing compound semiconductor having a porous structure.   One or more elements selected from nickel, iron, silver, lead, copper and bismuth are doped with 20 to 0.5 mol%, and one or more elements selected from antimony, bismuth, niobium and tantalum are added. The membranous porous semiconductor photoelectrode of claim 1 co-doped.   The membranous porous semiconductor photoelectrode according to any one of claims 1 to 3, wherein a diffusion distance of 50% or more of holes generated in the semiconductor to the semiconductor surface is within 500 nm.   The film-like porous semiconductor photoelectrode according to claim 1, comprising a composite metal oxide based semiconductor having a porous structure with visible light response and containing both bismuth and vanadium. The membranous porous semiconductor photoelectrode according to any one of claims 1 to 5 , which is formed on a light-transmitting substrate.   The film-like porous semiconductor photoelectrode according to any one of claims 1 to 6, wherein the thickness of the semiconductor is 50 µm or less.   A photoelectrochemical cell using the film-like semiconductor photoelectrode according to claim 1.   The photoelectrochemical cell according to claim 8, wherein the energy storage type reaction is a decomposition reaction of water. A method for decomposing water, comprising decomposing water into hydrogen and oxygen using the photoelectrochemical cell according to claim 8 .