PHOTOELECTRIC CONVERSION DEVICE SUBSTRATE AND PHOTOELECTRIC

CONVERSION DEVICE

Technical Field

The present invention relates to a photoelectric conversion device substrate and a photoelectric conversion device which includes the photoelectric conversion device substrate and is preferably used in applications, such as solar cells and the like.

Background Art

Photoelectric conversion devices having, on a substrate, a layer structure of lower electrode (back contact electrode) , a photoelectric conversion layer that absorbs light to generate electric current, and an upper electrode (transparent electrode) are used in applications, such as solar cells and the like. Most of the conventional solar cells are Si-based cells that use bulk monocrystalline Si, polycrystalline Si, or thin film amorphous Si. Recently, however, research and development of compound semiconductor-based solar cells that do not depend on Si has been carried out. As compound semiconductor solar cells, thin film systems, such as CIS (Cu-In-Se) systems formed of group lb element, group Illb element, and group VIb element, CIGS (Cu-In-Ga-Se) systems, and the like are known to have high optical absorption and high photoelectric conversion efficiency.

In CIS or CIGS photoelectric conversion devices, it is known that diffusion of an alkali metal, preferably Na, into the photoelectric conversion layer may improve the crystallization of the photoelectric conversion layer and photoelectric conversion efficiency as described, for example, in Japanese Patent No .2922465 and Japanese Unexamined Patent Publication No. 11 (1999) -312817. Conventionally, the diffusion of Na into photoelectric conversion layer has been effected using a soda-lime glass substrate that includes Na. The use of a metal, polymer, or ceramic substrate as a solar cell substrate, however, poses a problem that the conversion efficiency is not improved as sodium is not supplied from the substrate. Consequently, in the case where a substrate that does not include sodium is used, one of the following is performed: pr ^' oviding an alkali supply layer by liquid phase method; introducing sodium by co-deposition with CIGS, or providing Mo-Na as the electrode. For example, Japanese Unexamined Patent Publication No. 2009-267332 discloses that an alkali metal silicate, more specifically, a sodium silicate is applied by liquid phase method. Japanese Unexamined Patent Publication No. 2010-232427 discloses that an anodized substrate is doped with sodium by bringing the substrate into contact with a sodium hydroxide water solution. Further, Japanese Unexamined Patent Publication No. 2004-158511 discloses that a silicon oxide film is formed on a stainless-steel substrate by sol-gel method, as well as an insulating layer using a material that includes Na.

In the mean time, in the case where a glass substrate which includes alkali is used, it is known that a diffusion layer of SiC½, A1 ₂ 0 ₃ , TiN, S13N4, Zr0 ₂ , or Ti0 ₂ is provided on the glass substrate in order to control the supply amount of alkali as described, for example, in Japanese Patent No. 4110515.

As described in the aforementioned documents, it has been a common knowledge that the power generation efficiencymaybe improved by diffusing alkali from the alkali supply layer to the photoelectric conversion layer, but it is found that the provision of alkali supply layer does not improve the power generation efficiency so much than expected. One of the reasons is that the sputter-forming of a molybdenum film of lower electrode directly on the alkali layer causes the molybdenum to react with alkali ions by the energy of sputtering and a molybdate is generated. The generation of molybdate causes a reduction in the conductivity of the molybdenum electrode and a weakening in the adhesion at the interface between the molybdenum electrode and photoelectric conversion layer or between the molybdenum electrode and alkali supply layer, thereby leading to detachment of the layers, so that it is difficult to improve the power generation efficiency.

In the case where the alkali layer is provided by a liquid phase method, a relatively large volume of water is included and the surface becomes hydrophilic, thereby causing a problem that alkali metal carbonates are likely to be deposited on the surface through reaction between moistures or carbon dioxides in the air and alkali ions included in the alkali supply layer. Such a deposit causes a problem that it becomes a starting point of detachment of the photoelectric conversion layer or electrode layer, or causes a short circuit between the upper and lower electrodes via the photoelectric conversion layer. Further, in the manufacturing process of integrated solar cells, it is necessary to wash the cells with water after the scribing step performed after the electrode forming step, which causes a problem that the sodium is eluted by the washing.

As described in Japanese Patent No. 4110515, while the liberation/diffusion of alkali metal from the alkali supply layer needs to be restricted in relation to the electrode, the alkali metal needs to be efficiently diffused into the photoelectric conversion layer provided on the electrode. Thus, with respect to alkali supply layer, conflicting functions are required for the electrode layer and photoelectric conversion layer provided on the alkali supply layer.

The present invention has been developed in view of the circumstances described above, and it is an object of the present invention to provide a photoelectric conversion device substrate capable of preventing formation of impurities on the surface of an alkali supply layer, at the interface with an electrode, and on the surface of the electrode by controlling the liberation/diffusion of alkali metal from the alkali supply layer, while efficiently diffusing the alkali metal into a photoelectric conversion layer, thereby improving the photoelectric conversion efficiency of a photoelectric conversion device. It is a further object of the present invention to provide a photoelectric conversion device having the photoelectric conversion device substrate.

Disclosure of Invention

A photoelectric conversion device substrate of the present invention includes an alkali metal silicate layer on a substrate and a diffusion control layer on the alkali metal silicate layer, wherein the diffusion control layer includes substantially no alkali metal and controls diffusion of an alkali metal. Preferably, the diffusion control layer is formed of a silicon compound, and more preferably it is formed of an organosilicon compound. Preferably, the diffusion control layer is formed by a liquid phase method. Preferably, the diffusion control layer is formed of a compound obtained by hydrolysis/condensation reaction of an organoalkoxysilane. Preferably, the diffusion control layer has a thickness of not greater than 100 run. Preferably, the alkali metal silicate layer has a thickness of not greater than 1 urn.

Preferably, the substrate is a metal substrate. Preferably, the metal substrate has an anodized aluminum film formed on a surface thereof. Preferably, the metal substrate is made of a clad material in which an aluminumplate is integrallybonded to one or both surfaces of an aluminum, stainless-steel, or steel plate. Preferably, the anodized aluminum film is a porous anodized aluminum film having a compressive stress. A photoelectric conversion device of the present invention is a device formed on the photoelectric conversion device substrate described above.

The photoelectric conversion device substrate of the present invention includes a diffusion control layer which includes substantially no alkali metal and controls diffusion of an alkali metal on an alkali metal silicate layer, so that generation of impurities through reaction between the alkali metal in the alkali metal silicate layer and molybdenum or elution of sodium through washing is prevented. It is also assumed that the liberation of alkali component included in the alkali supply layer in the form of sodium hydroxide or the like through reaction with moisture or carbon dioxide in the air, or generation of deposits such as sodium carbonate or sodium hydrogen carbonate may ,be prevented.

In particular, in the case where the diffusion control layer is formed of an organosilicon compound, the layer is an organic-inorganic hybrid compound layer of a functional group and a network structure of Si-0 bonding, having low hydrophilicity in comparison with an inorganic compound such as a silica, so that the moisture in the air is unlikely to penetrate, thereby resulting in an reduced amount of moisture reaching the alkali supply layer through the diffusion control layer, so that deposition of deposits is further prevented. In the mean time, it is assumed that the heat treatment at several hundred degrees Centigrade may provide sufficient activation energy required for the movement of alkali metal and the alkali metal is diffused in the photoelectric conversion layer via the electrode layer (Mo layer) which is an upper layer of the alkali supply layer.

Brief Description of Drawings

Figure 1 is a schematic cross-sectional view of an embodiment of the photoelectric conversion device substrate of the present invention.

Figure 2 is a schematic cross-sectional view of an embodiment of the photoelectric conversion device with the photoelectric conversion device substrate of the present invention. Best Mode for Carrying Out the Invention

Hereinafter, the photoelectric conversion device substrate will be described in detail with reference to the accompanying drawings. Figure 1 is a schematic cross-sectional view of the photoelectric conversion device substrate of the present invention. As illustrated in Figure 1, the photoelectric conversion device substrate 10 includes an alkali metal silicate layer 3 on a substrate 2 and a diffusion control layer 4 on the alkali metal silicate layer 3. The diffusion control layer 4 includes substantially no alkali metal and controls diffusion of alkali metal. The term includes substantially no alkali metal" as used herein refers to that diffusion control layer 4 does not include any alkali metal except for those unavoidablymixed therein, as impurities, from rawmaterial or manufacturing process, or a small amount of alkali metal detected as noise in composition analysis. Preferably, the diffusion control layer 4 is made of a silicon compound, and more preferably it is made of an organosilicon compound (hereinafter, referred to as the "organosilicon compound layer") .

The organosilicon compound layer is a layer of an organic compound having a carbon-silicon bond and a layer formed by liquid phase method is preferably used. The organosilicon compounds that can be used for the organosilicon compound layer may include modified silicones, silicone resins, alkoxysilane compounds, chlorosilane compounds, silazanes, and the like. From the viewpoint of securing the effect of the diffusion control layer for a long time, chlorosilane compounds and alkoxysilane compounds having high connectivity with the alkali metal silicate layer are preferably used.

Preferable chlorosilane compounds may include methyl trichlorosilane, methyl dichlorosilane, dimethyl dichlorosilane, trimethyl chlorosilane, phenyl trichlorosilane, dichloro diphenylsilane, trifluoropropyl trichlorosilane, vinyl trichlorosilane, and the like.

The alkoxysilane (starting material monomer) that can be used may include tetraalkoxysilane having four alkoxy groups, trialkoxysilane having three alkoxy groups, dialkoxysilane having two alkoxy groups, and monoalkoxysilane having one alkoxy group. There is not any specific restriction on the type of alkoxy group, but those having a relatively small number of carbon atoms (carbon number of 1 to 4) , such as methoxy group, ethoxy group, propoxy group, butoxy group and the like are advantageous from the viewpoint of reactivity. In the case where the trialkoxysilane or dialkoxysilane is used, an organic group, hydroxyl group, or the like may be bonded to silicon atoms in the alkoxysilane and the organic groupmay further include a functional group such as epoxy group, amino group, mercapto group, vinyl group, and the like. Preferable tetraalkoxysilanes may include tetra methoxysilane, tetra ethoxysilane, tetra isopropoxysilane, tetra butoxysilane, dimethoxy diethoxysilane, or the like.

As for the trialkoxysilane, the following may be preferably used: methyl trimethoxysilane, propyl trimethoxysilane, hexyl trimethoxysilane, octadecyl-trimethoxysilane, phenyl trimethoxysilane, allyltrimethoxysilane, vinyltrimethoxysilane, cyano propyltrimethoxysilane, 3-bromo propyltrimethoxysilane, 3-chroro propyltrimethoxysilane, 2- (3, 4-epoxycyclohexyl) ethyltrimethoxysilane, 3-glycidyl oxypropyltrimethoxysilane, 3-iodo propyl trimethoxysilane, 3-mercapto propyl trimethoxysilane, trimethoxy [2- (7-oxabicyclo [4, 1, 0] hept-3-yl.) ethyl] silane, 1- [3- (trimethoxysilyl) propyl] urea, N-[3-

(trimethoxysilyl) propyl] aniline, trimethoxy [3-phenylaminopropyl] silane, acryloyloxy propyltrimethoxysilane, methacryloyloxy propyltrimethoxysilane, trimethoxy[2-phenylethyl] silane, trimethoxy (7-octen-l-yl) silane, trimethoxy (3, 3, 3- trifluoropropyl) silane, 3- [2- (2-aminoethylamino) ethylamino] propyltrimethoxysilane, [3- (2-aminoethylamino) propyl] trimethoxy silane, 3-glycidoxy propyltrimethoxysilane, 3-aminopropyl trimethoxy silane, 3-diethylamino propyl trimethoxysilane, bis (3- methylamino) propyl trimethoxysilane, N,N-dimethylaminopropyl trimethoxy silane, N- [3- (trimethoxysilyl) propyl] ethylenediamine, and trimethoxy (3-methylamino) propyl silane.

Preferable trialkoxysilanes further include methyltriethoxy silane, propyltriethoxysilane, hexyltriethoxysilane, octadecyltri ethoxysilane, phenyltriethoxysilane, allyltriethoxysilane, (1-naphthyl) triethoxysilane, [2- (cyclohexenyl) ethyl] triethoxysilane, 3-aminopropyltriethoxysilane, 3- [bis (2- hydroxyethyl) amino]propyltriethoxysilane, 3-chloropropyl triethoxysilane, 3-glycidyloxy propyltriethoxysilane, 3-mercapto propyl triethoxysilane, 4-chlorophenyl triethoxysilane, (bicyclo [2, 2, l]hept-5-en-2-yl) triethoxysilane, chloromethyl triethoxy silane, pentafluorophenyltriethoxysilane, 3- (triethoxysilyl) propionitrile, 3- (triethoxysilyl) propylisocyanate, bis [3- triethoxysilylpropyl] tetrasulphido, triethoxy(3- isocyanato propyl) silane, triethoxy (3-thioisocyanatopropyl) silane, and the like.

Preferable dialkoxysilanes include dimethoxydimethylsilane, diethoxydimethylsilane, diethoxy-3-glycidoxypropylmethylsilane, dimethoxydiphenylsilane, dimethoxymethylphenylsilane, and the liked.

Such alkoxysilanes may be used singly or in combination of two or more different types. Alkoxysilanes having two to four alkoxy groups described above may be used in combination with a monoalkoxysilane having one alkoxy group. Such monoalkoxysilanes include trimethylmethoxysilane, trimethylethoxysilane,

3-chloropropyl dimethylmethoxysilane, and the like.

From the viewpoint of functional group having interaction with the substrate and formation of siloxane bond, the alkoxysilane is preferable to be selected from monoalkoxysilane, dialkoxysilane, and trialkoxysilane, more preferable to be trialkoxysilane, and particularly preferable to be organotrialkoxysilane.

The organoalkoxysilane is represented by the chemical formula, Si (R ¹ ) m(OR ² ) 4- _m , in which rti is an integer of 1 to 3, R ¹ and R ² are organic groups having a carbon number of 1 or greater. Preferably, R ¹ is an organic group having a carbon number of 1 to 8 and may include different elements, such as N, 0, and S. Preferably, R ² is an organic group having a carbon number of 1 to 8. Examples of organic groups (-R ¹ ) include -C¾, -C ₂ H5,-C ₃ H7, -C ₄ H ₉ , -CHOCH-, -CH=C¾, -C ₆ H ₅ , -CF ₃ , -C2F5, -C3F ₇ , —C ₄ F ₉ , —CH ₂ CH ₂ CF ₃ , — CH2CH2C ₆ F13 , -CH2CH2C8F17 , —C ₃ H ₆ H ₂ , -C3H6NHC2H H2 , -C ₃ H ₆ OCH ₂ CHOCH2, and -CsHeOCOC (CH ₃ ) =CH ₂ and the like. Epoxy groups, amino groups, mercapto groups, and vinyl groups are more preferable.

Preferably, the alkoxy group (-OR ² ) is a methoxy group, ethoxy group, propoxy group, butoxygroup, or the like. An alkoxy group having a comparatively small number of carbon atoms (carbon number of 1 to 4) is advantageous from the viewpoint of reactivity. Note that in the case where a plurality of organic groups and a plurality of alkoxy groups are present within the same molecules, they may be different groups.

Examples of alkali metal silicates for forming the alkali metal silicate layer 3 include a lithium silicate, a potassium silicate, and a sodium silicate which may be formed singly or in combination.

As for the method of producing the sodium silicate, lithium silicate, or potassium silicate, a wet method, dry method, or the like is known. They may be formed by a method in which silicon oxide is dissolved by sodium hydroxide, lithium hydroxide, and potassium hydroxide respectively or the like. Further, various types of alkali metal silicates having different molar ratios are commercially available and these may also be used.

As for the sodium silicate, lithium silicate, and potassium silicate, various types of alkali metal silicates having different molar ratios are commercially available. As the index representing the ratio between the silicon and alkali metal, a molar ratio between S1O ₂ /A ₂ O (A: alkali metal) is often used. As for the lithium silicate, for example, Lithium Silicate 35, Lithium Silicate 45, and Lithium Silicate 75 (available from Nissan Chemical Industries Ltd.) are known. As for the potassium silicate, Potassium Silicate No.l and Potassium Silicate No.2 are commercially available.

As for the sodium silicate, Sodium Orthosilicate, Sodium Metasilicate, Sodium Silicate No.l, Sodium Silicate No.2, Sodium Silicate No.3, and Sodium Silicate No.4 are known and a high-molar sodium silicate in which the molar ratio of silicon is increased as high as a few dozens is also commercially available.

Each of the sodium silicate, lithium silicate, and potassium silicate may be mixed with water at a desired ratio to obtain a solution of desired concentration. The viscosity of each application liquid may be controlled by controlling the addition of water and appropriate coating conditions may be determined.

Note that the supply sources of the lithium silicate, potassium silicate, and sodium silicate of the alkali metal silicate layer at the time of preparation are not necessarily lithium silicate, potassium silicate, and sodium silicate respectively. For example, in the case where the alkali metal silicate layer includes lithium silicate and sodium silicate, an alkali metal silicate layer that includes the lithium silicate and sodium silicate may be produced by mixing the lithium silicate and sodium hydroxide or lithium hydroxide and sodium silicate with water at a desired ratio . Further, in the case where the alkali metal silicate layer includes the potassium silicate and sodium silicate, an alkali metal silicate layer that includes the potassium silicate and sodium silicate may be produced by mixing potassium hydroxide and sodium silicate or potassium silicate and sodium hydroxide with water at a desired ratio. Further, lithium salt, potassium salt, and sodium salt may be added to the respective supply sources. For example, nitrate salt, sulfate salt, acetate salt, phosphoric salt, chloride, bromide, iodide, and the like are used.

An application liquid of the alkali metal silicate other than the lithium silicate, potassium silicate, and sodium silicate may be obtained easily by adding a desired alkali metal nitrate salt, sulfate salt, acetate salt, phosphoric salt, chloride, bromide, or iodide to a sodium silicate solution.

Further, a boron-containing compound or a phosphorus-containing compound may be added to an alkali metal silicate water solution. The addition of these compounds may improve the Mo film formability andpower generation efficiency. The detailed mechanism for this phenomenon is still unclear, but it is assumed that the addition of the boron or phosphorus to the alkali metal silicate causes a change in microstructure of glass and the stability of alkali ions in the glass is increased. As a result, liberation of the alkali ions is prevented, thereby improving the Mo film formability and power generation efficiency.

A preferable example of boron source is a borate salt, such as boric acid or sodium tetraborate. Examples of phosphorus sources include phosphoric acid, peroxophosphoric acid, phosphonic acid, phosphinic acid, diphosphoric acid, triphosphoric acid, polyphosphoric acid, cyclo-triphospheric acid, cyclo-tetraphospheric acid, diphosphonic acid, and salts of these acids. Preferable examples include, for example, lithium phosphate, sodium phosphate, potassium phosphate, lithium hydrogen phosphate, ammonium phosphate, sodium hydrogen phosphate, calcium hydrogen phosphate, ammonium hydrogen phosphate, lithium dihydrogen phosphate, sodium dihydrogen phosphate, calcium dihydrogen phosphate, ammonium dihydrogen phosphate, sodium pyrophosphate, sodium triphosphate, and the like.

As for the substrate 2, ceramics substrates (alkali-free glass, silica glass, alumina, etc.), metal substrates (stainless steel, titanium foil, silicon, etc.), polymer substrates (polyimide and the like) may be used. From the viewpoint of thermal resistance and light weight, metal substrates are preferable. In particular, a material that causes a metal oxide film formed on a metal substrate by anodxzation to be an insulating bodymaybe used. More specifically, a substrate that includes at least one metal selected from aluminum (Al) , iron (Fe) , zirconium (Zr) , titanium (Ti) , magnesium (Mg) , cupper (Cu) , niobium (Nb) , and tantalum (Ta) , or a substrate made of an alloy of the aforementioned metals is preferable. In particular, a clad material in which an aluminum plate is integrally bonded to one or both surfaces of an aluminum, stainless steel, or steel plate is more preferable from the viewpoint of ease of anodization and high durability. A clad material with an aluminum plate on both surfaces is more preferable as warpage of the substrate due to difference in thermal expansion coefficient between the aluminum and oxide film (AI2O3) and hence detachment of the film due to the warpage may be prevented.

Preferably, the substrate is subjected to washing and polishing/smoothing treatments, such as a degreasing process for removing attached rolling oil, desmutting process for removing smut from the surface of the aluminum plate, and roughening process for roughening the surface of the aluminum plate, as required, before being used.

The anodized film is an insulating film having a plurality of pores formed by anodization, thereby ensuring high insulating performance. The anodization is performed by immersing the substrate 2, as anode, with a cathode in an electrolyte and applying a voltage between the anode and cathode. As for the cathode, carbon, aluminum, or the like is used.

There is not any specific restriction on anodization conditions, which are dependent on the type of electrolyte used. Appropriate anodization conditions include an electrolyte concentration of 0.1 to 2 mol/L, a solution temperature of 5 to 80°C, a current density of 0.005 to 0.60A/cm ² , a voltage of 1 to 200V, and an electrolyzing time of 3 to 500 minutes. There is not any specific restriction on the electrolyte and an acidic electrolyte that includes one or more acids, such as sulfuric acid, phosphoric acid, chromic acid, oxalic acid, malonic acid, sulfamic acid, benzenesulfonic acid, amidosulphonic acid, and the like is preferably used. In the case where one of such electrolytes is used, an electrolyte concentration of 0.2 to 1 mol/L, a solution temperature of 10 to 80°C, a current density of 0.05 to 0.30A/cm ² , and a voltage of 30 to 150V are preferable.

Preferably, the anodized film is constituted by a barrier layer portion and a porous layer portion in which the porous layer portion has a compressive strain at room temperature. Generally, it is known that the barrier layer has a compressive stress while the porous layer has a tensile stress, and the entire anodized film has a tensile stress if the film thickness is not less than several micrometers. In the mean time, a porous layer having a compressive stress may be produced by the used of the clad material described above and heat treatment to be described later. Consequently, it is possible to cause the entire anodized film to have a compressive stress even when the film thickness is not less than several micrometers, whereby cracking due to a difference in thermal expansion at the time of film forming is prevented and the filmbecomes an insulating film excellent in long term reliability near room temperature .

In this case, it is preferable that the magnitude of the compressive strain is not less than 0.01%, more preferably not less than 0.05%, and further preferably not less than 0.10%. It is also preferable that the magnitude of the compressive strain is not greater than 0.25%. A compressive strain of less than 0.01% is too small to have an anti-cracking effect. Consequently, the anodized film formed as the insulating layer may have cracking, leading to reduced insulating performance if subjected to bending strains, long lasting temperature cycles, external shocks, or stresses in the final product .

On the other hand, an excessive compressive strain may cause detachment of the anodized film or a large compressive strain on the anodized film, thereby causing cracking in the film, lost of flatness due to surface rising, leading to reduced insulating performance. Thus, it is preferable that the compressive strain is not greater than 0.25%. It is known that the Yong' s modulus of the anodized film is about 50 to 150 GPa, so that it is preferable that the magnitude of the compressive stress is 5 to 300 MPa.

Heat treatment may be performed after anodization. The heat treatment may cause the anodized film to have a compressive stress and anti-cracking property is improved. Thus, thermal resistance and insulation reliability are improved so that it is preferably used as a metal substrate with an insulation layer. Preferably, the heat treatment temperature is 150°C or higher. In the case where the clad material is used, heat treatment at a temperature of 300°C or higher is preferable. The preheat treatment may reduce the amount of moisture included in the porous anodized film, whereby the insulating performance may be improved.

In a conventional substrate made of only aluminum, heat treatment at a temperature of 300°C or higher causes a problem that the aluminum substrate loses the function of substrate due to softening of the aluminum or insulating property due to cracking in the anodized film arising from the difference in thermal expansion coefficient between the aluminum and anodized film. But the use of a clad material of aluminum and a dissimilar metal allows heating at a temperature of 300°C or higher.

The anodized film is an oxide layer formed in a water solution, and it is known that moisture is held inside of the solid as described, for example, in T. Iijima et al., "Structure of Duplex Oxide Layer in Porous Alumina Studiedby ⁷ A1 MAS andMQMAS NMR", ChemistryLetters, Vol. 34, No.9, pp.1286-1287, 2005. A solid NMR measurement of an anodized film identical to that described in the aforementioned document shows that the amount of moisture (OH group) inside of the solid of the anodized film is reduced when heat-treated at a temperature of 100°C or higher, and the moisture reduction is significant when heat-treated at a temperature of 200°C or higher. It is assumed that the heating may change the bonding state of Al-O and Al-OH, thereby inducing stress relaxation (annealing effect) .

It is clear from the dehydration measurement of the anodized film conducted by the present inventors that most of the dehydration occurs from the room temperature to a temperature of about 300°C. In the case where the anodized film is used as the insulating film, heat treatment at a temperature of 300°C or higher is extremely effective in order to improve the insulating performance since the greater the amount of moisture the poorer the insulating performance . The combination of the used of clad material of aluminum and a dissimilar metal and heat treatment at a temperature of 300°C or higher may effectively cause annealing effect, whereby a high compressive strain and a less amount of moisture content yet to have been achieved may be realized. This allows a photoelectric conversion device substrate having a further high insulation reliability to be provided.

Preferably, the anodized film has a thickness of 3 to 50 urn from the viewpoint of electrical insulation. A film thickness of 3 in or greater allows the film to have insulating property and compressive stress at room temperature, whereby both the thermal resistance at the time of film forming and long term reliability maybe obtained. Preferably, the film thickness is in the range from 5 urn to 30 um, and more preferably in the range from 5 urn to 20 urn.

In the case where the film thickness is extremely small, the electrical insulation may not be secured and damage due to mechanical impact at the time of handling may not be prevented. Further, insulating performance and thermal resistance are reduced rapidly with increased temporal degradation. This is due to the fact that a small film thickness causes relative increase in the influence of uneven surface of the anodized film, whereby cracking is likely to occur with the uneven surface as the starting point, reduced insulating performance due to metal deposits, intermetallic compounds, and metal oxides in the anodized film resulted from metal impurities included in the aluminum, in addition to relative increase in the influence of the voids, and increased damage due to external impact or stress whereby cracking is likely to occur. As such, an anodized film with a thickness of less than 3 urn is not suitable for applications of flexible thermal resistance substrates and roll-to-roll manufacturing due to reduced insulating performance.

An excessively thick film thickness is undesirable as flexibility is reduced and anodization becomes costly and time consuming. Further, bending resistance and thermal distortion resistance are reduced. The reason for the reduction in bending resistance is assumed that, when the anodized film is bent, a stress distribution in the cross-sectional direction becomes large due to difference in tensile stress between the surface and interface with the aluminum, whereby a local stress concentration is likely to occur. The reason for the reduction in thermal distortion resistance is assumed that, when a tensile stress is exerted on the anodized film due to thermal expansion of the base material, a larger stress is exerted on the interface with the aluminum and the stress distributionbecomes large in the cross-sectional direction, whereby a local stress concentration is likely to occur. As such, an anodized film with a thickness of greater than 50 urn is not suitable for applications of flexible thermal resistance substrates and roll-to-roll manufacturing due to reduced bending resistance and thermal distortion resistance. Further, the insulation reliability is reduced.

. A manufacturing method of the photoelectric conversion device substrate of the present invention will now be described. First, an alkali metal silicate layer is formed on a substrate. The alkali metal silicate layer may be formed by applying an alkali metal silicate solution on the substrate and heat treating the substrate. Each of the sodium silicate, lithium silicate, and potassium silicate described above may be mixed with water at a desired ratio to obtain a solution of desired concentration. The viscosity of each application liquid may be controlled by controlling the addition of water and appropriate application conditions may be determined. There is not any specific restriction on the application method and, for example, doctor blade method, wire bar method, gravure method, spraying method, dip coating method, spin coating method, capillary coating method, and the like may be used.

Heat treatment is performed after the application liquid is applied on the substrate. Dehydration temperature measurements performed by the present inventors using the thermogravimetric analysis and thermal desorption spectroscopy show that the dehydration occurs at a temperature of 200°C to 300°C. A temperature below 200°C is not desirable since the application liquid may not be dried sufficiently and a high water resistant alkali metal silicate layer may not be formed. Further, a heat treatment at a temperature below 300°C causes problems of a large amount of residual moisture which may react with carbon dioxide or the like in the air to form an impurity, such as carbonate or the like, on the surface, and formation of sodium molybdate or the like when Mo electrode is sputtered. Therefore, it is preferable that the heat treatment is performed at a temperature of not less than 200°C, more preferably at a temperature not less than 300°C, and further preferably at a temperature not less than 400°C.

As the heat treatment is performed at such a high temperature, the substrate used in the present invention is preferable to be a clad substrate formed of aluminum and a dissimilar metal combined together with an anodized film formed on a surface of the aluminum. It is known that clad substrates have a high thermal resistance such that cracking does not occur at a temperature of 400°C or higher. It is also known that the heat treatment at a temperature not less than 300°C performed in advance may cause the anodized film to have a compressive strain, which further improves the thermal resistance property and ensures long term insulation reliability. The heat treatment after the application of the alkali metal silicate layer may serve the heat treatment required for dehydration of the alkali metal silicate layer and the heat treatment required for causing the anodized film to have compressive strain at the same time. On the other hand, a temperature exceeding 600°C is not desirable because such temperature exceeds the glass transition temperature of the alkali metal silicate.

Preferably, the thickness of the alkali metal silicate layer after the heat treatment is not greater than 1 um, more preferably in the range from 0.01 to 1 um, and further preferably in the range from 0.1 to 1 um. A thickness greater than 1 um is not desirable because if the thickness of the alkali metal silicate layer is greater than 1 um, the amount of contraction of the alkali metal silicate layer becomes large at the time of heat treatment and cracking is likely to occur.

The diffusion control layer may be formed by application. Hereinafter, description will be made of a case taking alkoxysilane as an example of organosilicon compound. The application liquid is prepared by mixing alkoxysilane with a solvent. /As for the solvent, for example, water, ethanol, or methanol may be used. Further, a combined solvent prepared by mixing isopropyl alcohol, methyl ethyl ketone, or the like into the aforementioned solvent may also be used.

The application liquid may further include other components, including but not limited to various acids (e.g. , hydrochloric acid, acetic acid, sulfuric acid, nitric acid, phosphoric acid, and the like), various bases (e.g., ammonia, sodium hydroxide, sodium hydrogen carbonate and the like), a hardener (e.g., metal chelate compound and the like) , a viscositymodifier (e.g., polyvinyl alcohol, polyvinylpyrrolidone, and the like) .

The application liquid prepared in the manner described above is applied on the alkali metal silicate layer formed in the manner as described above to form a coated layer. There is not any specific restriction on the method of applying the application liquid on the alkali metal silicate layer, and for example, doctor blade method, wire bar method, gravure method, spraying method, dip coating method, spin coating method, capillary coating method, and the like may be used, as in the alkali metal silicate layer.

Then the coated layer is heated to cause hydrolysis/condensation reaction of the alkoxysilane in the coated layer. As the hydrolysis/condensation reaction of the alkoxysilane progresses by sol-gel reaction, an alkoxysilane condensate gradually turns to have a high molecular weight. Preferably, the heating temperature is in the range from 50°C to 200°C with a reaction time in the range from 5 minutes to 1 hour. A temperature exceeding 200°C causes the alkoxysilane condensate to have pores. Preferably, the thickness of the diffusion control layer after formation is not greater than 100 nm, more preferably not greater than 50 nm, and particularly preferably not greater than 20 nm. If the layer has a thickness greater than 100 nm, cracking is likely to occur and influence the characteristics of the upper layers of Mo electrode and photoelectric conversion layer. In the case where an organosilicon compound is used for the diffusion control layer, if the thickness of the layer is greater than 100 nm, the amount of degassing at a high temperature becomes large, thereby impacting on vacuum film forming in the subsequent steps. On the other hand, if the organosilicon compound layer is too thin, the diffusion control function is deteriorated, so that the thickness is preferable to be not less than 1 nm and more preferable to be not less than 3 nm.

A photoelectric conversion device having the photoelectric conversion device substrate of the present invention will now be described. Figure 2 is a schematic cross-sectional view of an embodiment of the photoelectric conversion device. Each component in Figure 2 is not necessarily written to scale for facilitating visual recognition. As illustrated in Figure 2, a photoelectric conversion device 1 has a structure in which following are layered on a photoelectric conversion device substrate 10 of the present invention in the order listed below: a lower electrode 40, a photoelectric conversion semiconductor layer 50, a buffer layer 60, _

a translucent conductive layer (transparent electrode) 70, and an upper electrode (grid electrode) 80.

There is not any restriction on the component of the lower electrode (back contact electrode) 40, but Mo, Cr, W, and a combination thereof are preferable, in which Mo or the like is particularly preferable. There is not any specific restriction on the thickness of the lower electrode (back contact electrode) 40 but a thickness of about 200 to about 1000 nm is preferable.

The photoelectric conversion semiconductor layer 50 is a compound semiconductor system photoelectric conversion semiconductor layer, and there is not any specific restriction on the major component (component with a content of 20% by mass or more) thereof, but from the viewpoint of high photoelectric conversion efficiency, a chalcogen compound semiconductor, a compound semiconductor having a chalcopyrite type structure, or a compound semiconductor having a defect stannite type structure is preferably used.

As for chalcogen compounds (compounds containing S, Se, Te) , the following are preferable:

II-VI compounds: ZnS, ZnSe, ZnTe, CdS, CdSe, CdTe and the like; I-III-VI2 compounds: CuInSe2, CuGaSe2, Cu(In, Ga)Se2, CuInS2, CuGaS2,

Cu(in, Ga) (S, Se) ₂ , and the like; and

I-III3-VI ₅ compounds: CuIn ₃ Ses, CuGa3Se ₅ , Cu(In, Ga) ₃ Se ₅ , and the like.

As for the compounds having a chalcopyrite type structure or defect stannite type structure, the following are preferable:

I-III-VI2 compounds: CuInSe ₂ , CuGaSe ₂ , Cu(In, Ga)Se ₂ , CuInS2, CuGaS ₂ ,

Cu(In, Ga) (S, Se) ₂ , and the like; and

I-III ₃ -VI5 compounds: CuIn ₃ Ses, CuGa ₃ Ses, Cu(In, Ga) ₃ Ses, and the like. Here, (In, Ga) and (S, Se) represent (Ini- _x Ga _x ) and (Si-. _y Se _y ) (where, x=0 to 1 and y=0 to 1) .

There is not any specific restriction on the method of forming the photoelectric conversion layer. For example, in the film forming of a CI (G) S photoelectric conversion semiconductor layer which includes Cu, In, (Ga) , and S, selenization method or multi source deposition may be used. There is not any specific restriction on the thickness of the photoelectric conversion semiconductor layer 50, but a thickness in the range from 1.0 to 3.0 urn is preferable and a thickness in the range from 1.5 to 2.0 urn is particularly preferable .

There is not any specific restriction on the buffer layer 60, but it is preferable that the buffer layer 60 includes a metallic sulfide that includes at least one metal element selected from the group consisting of Cd, Zn, Sn, and In, such as Cds, ZnS, Zn(S, O), and/or Zn(S, 0, OH), SnS, Sn(S, 0) and/or Sn(S, 0, OH), InS, In(S,0), and/or In(S, 0, OH) , and the like. Preferably, the thickness of the buffer layer 60 is in the range from 10 nm to 2 urn and more preferably in the range from 15 to 200 nm.

The translucent conductive layer (transparent electrode) 70 is a layer through which light is introduced into the device and functions as an electrode through which electric current generated in the photoelectric conversion layer 50 flows with the lower electrode 40 as a pair. There is not any specific restriction on the composition of the translucent conductive layer 70, but n-ZnO such as ZnO:Al or the like is preferable. There is not any specific restriction on the thickness of the translucent conductive layer 70, but a thickness in the range from 50 nm to 2 urn is preferable. There is not any specific restriction on the upper electrode (grid electrode) 80, but Al or the like may be preferably used. There is not any specific restriction on the thickness of the upper electrode 80, but a thickness in the range from 0.1 to 3 urn is preferable.

The photoelectric conversion device substrate of the present invention may preferably be used in solar cells and the like. A solar cell may be provided by attaching a cover glass, a protection film, and the like to the photoelectric conversion device 1. Hereinafter, the photoelectric conversion device substrate of the present invention will be described in more detail by way of Examples. [Examples]

(Preparation of Substrate)

A 3x3 cm alkali-free glass substrate, a SUS430 substrate, and ah anodized aluminum substrate were provided. The anodized aluminum substrate was produced in the following manner. A clad material of a 30 um thick aluminum and a 100 um thick SUS430 was anodized in an oxalic acid electrolyte with a constant voltage of 40V to produce a substrate with a 10 um thick anodized aluminum on the surface. (Preparation of Application)

Thirty grams of sodium silicate No.3 is mixed with ten grams of pure water to prepare an application liquid "A" for an alkali metal silicate layer. Further, an application liquid "B" for a diffusion control layer and an application liquid "C" for a diffusion control layer were prepared according to the prescriptions in Tables 1 and 2 below respectively.

TABLE 1

(Examples and Comparative Example)

The application liquid "A" was applied to each substrate shown in Table 3 by spin coating method. The application thickness was adjusted such that the thickness of the layer becomes 0.5um. Thereafter, heat treatment was performed for 30 minutes with a temperature of 450°C. Then, the application liquid "B" or "C" shown in Table 3 was applied by spin coating method. The application thickness was adjusted such that the thickness of the layer becomes 0.5um. Thereafter, the substrates were subjected to heat treatment for 30 minutes with a temperature of 450°C. Note that the Comparative Example was not provided with a diffusion control layer.

(Washing Acceptability Evaluation-Residual Na Ratio)

The substrates of Examples and Comparative Example were immersed in pure water for three minutes and an amount of Na before and after the immersion was measured by measuring intensity ratio of NaKa peaks near 1041 eV by XRF before and after the immersion. That is, the amount of Na was measured by measuring the amount of NaKa radiation by an XRF measuring system. When the amount of Na before the immersion is taken as 1, the ratio of the amount of Na after the three-minute immersion was determined as the residual Na ratio. As the penetration depth of the incident X-rays is about 10 to 20 urn, the entire amount of Na included in the porous anodized aluminum film may be evaluated.

(Mo Film Formability Evaluation)

Mo was formed by DC sputtering on the substrate of each of Examples and Comparative Example with a thickness of 800 nm. Impurities on the surface of Mo formed on the alkali metal silicate layer were observed by optical microscope, and following evaluations were made as shown in Table 3 according to the number of impurities per 1 mm square: "A" if no impurity was found, "B" if 1 to less than 10 impurities were found, and "C" if 10000 or more impurities were found.

(Solar Cell Manufacturing)

ACIGS solar cell was formed on the Mo electrode. In the present embodiment, granular raw materials of high purity cupper and indium (purity of 99.9999%), high purity Ga (purity of 99.999%), and high purity Se (purity of 99.999%) were used as the deposition sources. A chromel-alumel thermocouple was used for monitoring the substrate temperature. The main vacuum chamber was evacuated to 10 ^"6 Torr (1.3xl0 ^~3 Pa) and then a CIGS thin film was formed by controlling deposition rate of each deposition source with a thickness of about 1.8 um under the film forming condition of maximum substrate temperature of 530°C. Then, as the buffer layer, a CdS thin film was deposited about 90 nm by solution-growth technique followed by the formation of a 0.6 urn thick transparent conductive film of ZnO:Al by DC sputtering. Finally, as the upper electrode, an Al grid electrode was formed by deposition to complete the manufacture of the solar cell.

(Power Generation Efficiency Measurement)

Pseudo solar light with air mass (AM) = 1.5, lOOmW/cm ² was directed to each manufactured solar cell (area of 0.5 cm ² ) to measure the energy conversion efficiency. Eight photoelectric conversion device samples were produced for each Example and Comparative Example. With respect to each photoelectric conversion device, photoelectric conversion efficiency was measured under the condition described above. Among the measured values, a maximum value in each Example and a maximum value in Comparative Example were determined to be the conversion efficiency of the respective groups. Further, a coefficient of variation (value obtained by dividing the standard deviation of eight cells by the average value) was evaluated as the efficiency variation of the cells.

(Sodium Concentration Measurement)

With respect to each of the photoelectric conversion devices of Examples and Comparative Example, sodium concentration of the photoelectric conversion layer (CIGS layer) was measured. The measurement was performed using SIMS (secondary ion mass spectrometer) . The primary ion species for the measurement was Cs ⁺ with an acceleration voltage of 5.0 kV. The sodium concentration in the photoelectric conversion layer (CIGS layer) is distributed in the thickness direction, so that the distribution was integrated to obtain an average value for use in the evaluation of sodium concentration.

Table 3 shows measurement results of washing acceptability evaluation, Mo film formability evaluation, sodium concentration, power generation efficiency, and coefficient of variation, as well as the type of substrate and prescriptions of application liquids for the alkali metal silicate layer and diffusion control layer of each Example and Comparative Example.

As shown in Table 3, substantially no foreign substance was generated in each Example provided with a diffusion control layer, indicating that the generation of impurities through reaction between the alkali metal ih the alkali metal silicate layer and molybdenum is prevented. Further, from the comparison of the residual Na ratio between each Example and Comparative Example, it is known that the sodium elution due to washing is prevented. Further, from the Na concentrations of Examples and Comparative Example, it is assumed that the liberation of alkali component included in the alkali supply layer in the formof sodiumhydroxide or the like through reaction with moisture or carbon dioxide in the air, or generation of deposits such as sodium carbonate or sodium hydrogen carbonate is prevented.

In the mean time, comparison of power generation efficiency between each Example and Comparative Example shows that the Na supply to the CIGS is sufficient and high power generation efficiency is obtained in each Example. This shows that the diffusion control layer mayprevent formation of impurities on the surface of an alkali supply layer, at the interface with Mo, and on the surface of the Mo by controlling the liberation/diffusion of alkali metal from the alkali supply layer, while efficiently diffusing the alkali metal into the photoelectric conversion layer. Note that each Example has a small coefficient of variation and all manufactured solar cells in each Example have high power generation efficiency, indicating that power generation efficiency of all manufactured solar cells has been improved.