JP5283588B2 - Solar cell - Google Patents

Description

  The present invention relates to a solar battery, and more particularly to a multi-junction solar battery in which a plurality of solar battery cells are stacked.

  Conventionally, a single-junction solar cell composed of a single solar cell is known. However, the power generation efficiency (conversion efficiency) of a single-junction solar cell has a theoretical limit determined by the forbidden band width Eg of the semiconductor material used as a raw material. Therefore, it is known that the power generation efficiency is about 30%.

  Thus, in order to obtain higher power generation efficiency than single-junction solar cells, multi-junction solar cells formed by stacking a plurality of solar cells have been proposed. Each solar battery cell in this multi-junction solar battery has a pn junction and is composed of materials having different band gaps. In this way, the multijunction solar cell can expand the use region of the solar spectrum compared to the single junction solar cell by connecting semiconductors having different band gaps, and can realize high power generation efficiency. .

  For example, as a multi-junction solar cell, a Ge-cell / InGaAs cell / InGaP cell laminated structure three-junction solar cell, or an InGaAs cell / InGaP cell laminated structure two-junction solar cell formed on a Ge substrate, etc. Are known.

Patent Document 1 below discloses a multijunction solar cell in which lattice matching is achieved between solar cells (hereinafter also referred to as a lattice matching multijunction solar cell). In the multi-junction solar cell described in Patent Document 1, the composition x of In is in the range of 0.005 ≦ x ≦ 0.015, preferably 0 in the Ge cell (substrate) / In _x Ga _1-x As cell junction. A value in the range of .007 ≦ x ≦ 0.014. As a result, the lattice mismatch rate of the multi-junction solar cell becomes a very small value that can be regarded as almost zero. Thus, by setting the lattice constant of the InGaAs cell to a value very close to the lattice constant of the Ge cell (substrate) and lattice-matching each cell, a solar cell with few crystal defects and high power generation efficiency can be obtained.

  On the other hand, in a multi-junction solar cell in which a plurality of solar cells are connected in series, when the current matching of the output current of each solar cell is not achieved, the output current of the entire solar cell is the smallest output current There is a problem of being regulated by solar cells.

  In the multi-junction solar battery described in Patent Document 1, according to the band gap (energy conversion wavelength range of each solar battery cell) that each solar battery cell has, the relationship between the output currents of each solar battery cell is I (InGaP ), I (InGaAs) <I (Ge). Therefore, in the lattice-matched multi-junction solar cell as described in Patent Document 1, the output current I (Ge) is restricted to I (InGaP) and I (InGaAs), and the power generation efficiency is improved. It is difficult.

  On the other hand, Non-Patent Document 1 below discloses a multi-junction solar cell in which the connection between the Ge cell (substrate) / InGaAs cell is lattice-mismatched (hereinafter also referred to as a lattice-mismatched solar cell). . According to this multi-junction solar cell, by reducing the band gap of the InGaAs cell and InGaP cell, the energy conversion wavelength range of the InGaAs cell and InGaP cell is expanded, and the power generation efficiency of the entire solar cell can be improved.

JP 2004-319934 A

C. M.M. Fetzer, “1.6 / 1.1 eV metamorphic GaInP / GaInAs solar cells grown by MOVPE on Ge”, Journal of Crystal Growth 276 (2005), p. 48-56

  However, the lattice mismatched multi-junction solar cell as described in Non-Patent Document 1 described above is used for crystal growth of solar cells (for example, InGaAs cells) having different lattice constants on a substrate (for example, Ge substrate). A tensile stress is generated at the interface due to lattice mismatch between the substrate and the solar battery cell, and the crystal growth of the solar battery cell is inhibited. Due to the inhibition of the crystal growth, there is a problem that crystal defects are generated in the solar battery cell, and the power generation efficiency of the solar battery is reduced due to carrier scattering caused by the crystal defects.

  Therefore, the present invention has been made in view of the above problems, and in a lattice-mismatched multijunction solar cell, the stress caused by the difference in lattice constant is relieved to form a solar cell with few crystal defects. It is to obtain a multi-junction solar cell capable of obtaining high power generation efficiency.

  The solar cell in the first aspect of the present invention includes a first solar cell, an intermediate region formed on the first solar cell, and a second solar cell formed on the intermediate region. And the intermediate region is arranged in a distributed manner on the surface of the first solar cell, and is caused by a difference in lattice constant between the first solar cell and the second solar cell. It has a stress relaxation part that relieves stress.

  Moreover, the 2nd aspect of this invention is a solar cell in a 1st aspect, Comprising: The said stress relaxation part is arrange | positioned on the surface of a said 1st photovoltaic cell at island shape.

  Moreover, the 3rd aspect of this invention is a solar cell in a 1st or 2nd aspect, Comprising: The said stress relaxation part has electroconductivity.

  Moreover, the 4th aspect of this invention is a solar cell in any one of the 1st to 3rd aspect, Comprising: The said stress relaxation part has melting | fusing point higher than the material which comprises a said 2nd photovoltaic cell. Composed of materials.

  The fifth aspect of the present invention is the solar battery according to any one of the first to fourth aspects, wherein the second solar battery cell has a band gap larger than that of the first solar battery cell. Composed of materials.

  A sixth aspect of the present invention is the solar battery according to any one of the first to fifth aspects, wherein the first solar battery cell is made of Ge, and the second solar battery cell is And InGaAs.

The seventh aspect of the present invention is the solar battery according to the sixth aspect, wherein the In composition x in the In _x Ga _1-x As constituting the second solar battery cell is 0.05. <X <0.2.

  Further, an eighth aspect of the present invention is the solar battery according to any one of the first to seventh aspects, wherein the intermediate region is on the surface of the first solar battery cell and the surface of the stress relaxation part. It further has a buffer layer formed thereon.

A ninth aspect of the present invention is the solar battery according to the eighth aspect, wherein the first solar battery cell is made of Ge, and the second solar battery cell is made of InGaAs. The In composition x in In _x Ga _1-x As constituting the second solar battery cell has a value in the range of 0.05 <x <0.2, and the buffer layer is made of InGaAs. The In composition x in In _x Ga _1-x As constituting the buffer layer is a value in the range of 0.05 <x <0.2.

  The tenth aspect of the present invention is the solar battery according to the eighth or ninth aspect, wherein the stress relaxation portion has a refractive index that is the refractive index of the first solar battery cell and the refractive index of the buffer layer. It is a value between rates.

  An eleventh aspect of the present invention is the solar battery according to any one of the eighth to tenth aspects, wherein the first solar battery cell is made of Ge, and the buffer layer is made of InGaAs. The stress relaxation part is made of SiGe, Si, GaSb, or InSb.

  A twelfth aspect of the present invention is the solar battery according to any one of the first to eleventh aspects, formed on the second solar battery cell, and constituting the second solar battery cell. The battery further includes a third solar battery cell made of a material having a band gap different from that of the material.

  A thirteenth aspect of the present invention is the solar battery according to the twelfth aspect, wherein the third solar battery cell is made of a material having a larger band gap than the second solar battery cell.

  A fourteenth aspect of the present invention is the solar battery according to the thirteenth aspect, wherein the third solar battery cell is made of InGaP.

  According to the first aspect, the stress caused by the difference in lattice constant between the first solar cell and the second solar cell is relieved, and a solar cell with few crystal defects is formed even if there is lattice mismatch. And the power generation efficiency of the solar cell can be further improved.

  According to the second aspect, the stress relaxation part formed in an island shape can partially block the connection between atoms of the first solar cell and the second solar cell, resulting in a difference in lattice constant. This can reduce the stress generated at the interface between the cells.

  According to the 3rd aspect, when a stress relaxation part has electroconductivity, a charge can be moved also from a stress relaxation part, and the electric power generation efficiency of a solar cell can be improved.

  According to the fourth aspect, the stress relaxation portion is made of a material having a melting point higher than that of the material constituting the second solar cell, so that the material of the second solar cell is in the first solar cell. It is possible to suppress diffusion into

  According to the fifth aspect or the sixth aspect, the first solar cell and the second solar cell have different band gaps, so that the use area of the solar spectrum can be expanded. It is possible to improve the power generation efficiency.

  According to the 7th aspect or the 9th aspect, the output current of a 1st photovoltaic cell and a 2nd photovoltaic cell can be matched, and the electric power generation efficiency of a solar cell can be improved.

  According to the eighth aspect, it is possible to reduce crystal dislocations and defects caused by stress caused by the difference in lattice constant between the first solar cell and the second solar cell.

  According to the tenth aspect or the eleventh aspect, since the light transmitted through the second solar cell can be efficiently incident on the first solar cell, the power generation efficiency in the first solar cell is improved. Can be improved.

  According to the twelfth aspect to the fourteenth aspect, the first solar cell, the second solar cell, and the third solar cell have different band gaps, thereby expanding the use range of the solar spectrum. It is possible to improve the power generation efficiency of the solar cell.

It is sectional drawing which showed the structure of the multijunction solar cell which concerns on one Embodiment of this invention. It is the figure which showed the AA cross section of the multijunction solar cell of FIG. 1 schematically. It is sectional drawing which showed typically the structure of the multijunction solar cell which concerns on one Embodiment of this invention. It is the figure which showed the joining state of Ge substrate and an InGaAs layer in the multijunction type solar cell which has the stress relaxation part which concerns on one Embodiment of this invention. It is the figure which showed the joining state of Ge board | substrate and an InGaAs layer in the multijunction type solar cell which does not have a stress relaxation part. It is the figure which showed the energy conversion wavelength range of sunlight.

<Configuration of multi-junction solar cell>
Hereinafter, a multi-junction solar cell according to an embodiment of the solar cell according to the present invention will be described with reference to the drawings. Here, the multi-junction solar cell is a solar cell formed by laminating a plurality of solar cells made of materials having different light absorption wavelengths in order to expand the use region of the sunlight spectrum. Such a multi-junction solar cell has a solar cell made of a high bandgap material on the light incident side, and has a solar cell made of a low bandgap material on the light output side. A plurality of cells are connected by a tunnel junction layer.

  As shown in FIG. 1, a multi-junction solar cell 100 according to an embodiment of the present invention includes a three-junction solar cell having a stacked structure of a top cell having an InGaP layer / a middle cell having an InGaAs layer / a bottom cell having a Ge substrate. It is a battery.

  The multi-junction solar cell 100 is formed on the bottom cell 1, the intermediate region 2 formed on the bottom cell 1, the first tunnel region 3 formed on the intermediate region 2, and the first tunnel region 3. The middle cell 4, the second tunnel region 5 formed on the middle cell 4, and the top cell 6 formed on the second tunnel region 5 are provided. The multi-junction solar cell 100 includes a contact layer 7 formed on the top cell 6, a surface electrode 81 formed on the contact layer 7, a back electrode 82 formed below the bottom cell 1, a top cell. The antireflection film 9 is formed on the upper electrode 6 and the surface electrode 81.

The bottom cell 1 includes a p-type Ge substrate 11 and an n-type Ge layer 12 selectively formed on the top of the p-type Ge substrate 11. The p-type Ge substrate 11 is a substrate having a thickness of about 150 μm and doped with Ga or the like and an impurity concentration of 1 × 10 ^{17 to} 5 × 10 ¹⁸ cm ⁻³ . The n-type Ge layer 12 is formed by diffusing As or the like during film formation. The pn junction formed by the p-type Ge substrate 11 and the n-type Ge layer 12 constitutes the first solar battery cell 10.

  In the first solar battery cell 10, among the incident sunlight, photons absorbed by the p-type Ge substrate 11 generate a pair of hole-electrons, among which electrons diffuse to form a pn interface. When it reaches the depletion layer, it flows into the n-type Ge layer 12 by the internal electric field of the depletion layer and becomes a current (the same applies to the second solar cell 40 and the third solar cell 60 described later).

  The intermediate region 2 is distributed on the surface of the p-type Ge substrate 11, and the n-type formed on the surface of the first solar cell 10 and the surface of the stress relaxation portion 21. An InGaAs low temperature buffer layer 22 and an n type InGaAs buffer layer 23 formed on the n type InGaAs low temperature buffer layer 22 are provided.

The stress relaxation portion 21 is a region formed in an island shape made of SiGe, Si, GaSb, InSb, SiO ₂ , SiNx, Ti, W, or the like. As shown in FIGS. 1 and 2, the stress relaxation portion 21 in the present embodiment has a thickness of 50 to 500 mm, a circular shape in plan view with a diameter of 0.25 to 2 μm, and the surface of the bottom cell 1. It is arranged on the top at equal intervals. The stress relaxation portion 21 is for relaxing stress generated due to the difference in lattice constant (lattice mismatch) between crystal layers, and the action thereof will be described later. The stress relieving portions 21 may be other than a circle such as a quadrangle and a triangle, and may be arranged at non-uniform intervals. Moreover, the stress relaxation part 21 may be formed in a lattice shape in addition to the island shape.

  In the present embodiment, the stress relaxation portion 21 has a refractive index that is a value in a range between the refractive index of the bottom cell 1 (first solar battery cell 10) and the refractive index of the n-type InGaAs low-temperature buffer layer 22. Is preferred. In this embodiment, the p-type Ge substrate 11 and the n-type Ge layer 12 of the bottom cell 1 have a refractive index of about 4 and about 4, respectively, and the n-type InGaAs low-temperature buffer layer 22 has a refractive index of about 3.5. For example, the stress relaxation portion 21 is made of SiGe (refractive index = about 3.5 to 4), Si (refractive index = about 3.5), GaSb (refractive index = about 3.8), or InSb (refractive index = What is necessary is just to form by about 4). By doing so, the refractive index of the stress relaxation portion 21 becomes a value close to both the refractive index of the n-type InGaAs low-temperature buffer layer 22 and the refractive index of the bottom cell 1. Thereby, reflection of light incident on the stress relaxation part 21 from the n-type InGaAs low-temperature buffer layer 22 can be reduced, and reflection of light incident on the bottom cell 1 from the stress relaxation part 21 can be reduced. That is, since the light transmitted through the second solar cell 40 can be efficiently incident on the first solar cell 10 (bottom cell 1), the power generation efficiency in the first solar cell 10 is improved. Can do. The refractive index can be measured using, for example, a known ellipsometric method.

  Moreover, it is preferable that the stress relaxation part 21 uses the material which has electroconductivity. As a result, the charge can be transferred from the stress relieving portion 21 and the power generation efficiency of the multi-junction solar cell 100 can be further improved.

The n-type InGaAs low-temperature buffer layer 22 is a layer having a thickness of about 0.05 to 0.1 μm and doped with Si or the like and an impurity concentration of 5 × 10 ¹⁷ to 2 × 10 ¹⁸ cm ⁻³ . The n-type InGaAs buffer layer 23 has a thickness of about 1 to 3 μm and is doped with Si or the like and has an impurity concentration of 5 × 10 ¹⁷ to 2 × 10 ¹⁸ cm ⁻³ .

  The n-type InGaAs buffer layers 22 and 23 are provided to reduce dislocations and defects caused by stress caused by lattice mismatch at the Ge / InGaAs heterointerface. However, since the n-type InGaAs buffer layers 22 and 23 alone cannot sufficiently reduce dislocations and defects, the stress relaxation portion 21 is used for the purpose of relaxing stress caused by lattice mismatch occurring at the Ge / InGaAs heterointerface. Is provided.

The first tunnel region 3 includes an n ⁺⁺ type InGaAs tunnel junction layer 31 formed on the intermediate region 2 and a p ⁺⁺ type InGaAs tunnel junction formed on the n ⁺⁺ type InGaAs tunnel junction layer 31. Layer 32.

The n ⁺⁺ type InGaAs tunnel junction layer 31 is a layer having a thickness of about 0.01 to 0.05 μm and doped with Si, Te or the like and having an impurity concentration of 1 × 10 ¹⁸ to 1 × 10 ¹⁹ cm ⁻³ . It is. The p ⁺⁺ type InGaAs tunnel junction layer 32 is a layer having a thickness of about 0.01 to 0.05 μm and doped with Zn, C or the like and having an impurity concentration of 1 × 10 ¹⁹ to 1 × 10 ²⁰ cm ⁻³ . It is.

  This first tunnel region 3 has a pn junction made of a thin InGaAs layer in which a dopant is implanted at a high concentration, and in order to improve the connection between solar cells using a tunnel current. (The same applies to the second tunnel region 5 described later).

  The middle cell 4 includes a p-type InGaP back surface field layer 41 formed on the first tunnel region 3, a p-type InGaAs base layer 42 formed on the p-type InGaP back surface field layer, and a p-type InGaAs base layer 42. An n-type InGaAs emitter layer 43 formed thereon and an n-type AlInP window layer 44 formed on the n-type InGaAs emitter layer 43 are provided. The pn junction formed by the p-type InGaAs base layer 42 and the n-type InGaAs emitter layer 43 constitutes the second solar cell 40.

The p-type InGaP back surface electric field layer 41 is a layer having a thickness of about 0.05 to 0.3 μm and doped with Zn, C or the like and having an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ^−3. . The p-type InGaAs base layer 42 is a layer having a thickness of about 2 to 4 μm and doped with Zn or the like and having an impurity concentration of 0.8 × 10 ^{17 to} 5 × 10 ¹⁷ cm ⁻³ . The n-type InGaAs emitter layer 43 is a layer having a thickness of about 0.05 to 0.3 μm and doped with Si or the like and an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ⁻³ . The n-type AlInP window layer 44 is a layer having a thickness of about 0.025 to 0.1 μm and doped with Si or the like and an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ⁻³ . Further, the n-type AlInP window layer 44 may be configured using InGaP.

  The p-type InGaP back surface field layer 41 reduces back surface recombination loss caused by diffusion of electrons generated in the p-type InGaAs base layer 42 through the p-type InGaAs base layer 42 and into the bottom cell 1 immediately below. It is provided for. Therefore, the p-type InGaP back surface field layer 41 acts as a band barrier against the electrons as described above (the same applies to the p-type AlInP back surface field layer 61 described later).

  The n-type AlInP window layer 44 is formed on the p-type InGaAs emitter layer 43 in order to prevent carriers generated by sunlight from disappearing due to recombination with surface defects of the n-type InGaAs emitter layer 43 directly below. (The same applies to an n-type AlInP window layer 64 described later).

The second tunnel region 5 includes an n ⁺⁺ type InGaP tunnel layer 51 formed on the middle cell 4 and a p ⁺⁺ type InGaP tunnel junction layer 52 formed on the n ⁺⁺ type InGaP tunnel junction layer 51. With.

The n ⁺⁺ type InGaP tunnel junction layer 51 is a layer having a thickness of about 0.01 to 0.05 μm and doped with Si, Te or the like and having an impurity concentration of 1 × 10 ¹⁸ to 1 × 10 ¹⁹ cm ⁻³ . It is. The p ⁺⁺ type InGaP tunnel junction layer 52 is a layer having a thickness of about 0.01 to 0.05 μm and doped with Zn, C or the like and having an impurity concentration of 1 × 10 ¹⁹ to 1 × 10 ²⁰ cm ⁻³ . It is. The p ⁺⁺ type InGaP tunnel junction layer 52 may be formed using AlInGaAs.

  The top cell 6 includes a p-type AlInP back surface field layer 61 formed on the second tunnel region 5, a p-type InGaP base layer 62 and a p-type InGaP base layer formed on the p-type AlInP back surface field layer 61. An n-type InGaP emitter layer 63 formed on 62 and an n-type AlInP window layer 64 formed on the n-type InGaP emitter layer 63 are provided. The pn junction formed by the p-type InGaP base layer 62 and the n-type InGaP emitter layer 63 constitutes the third solar battery cell 60.

The p-type AlInP back surface electric field layer 61 is a layer having a thickness of about 0.025 to 0.2 μm and doped with Zn or the like and having an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ⁻³ . The p-type InGaP base layer 62 is a layer having a thickness of about 0.5 to 1.5 μm and doped with Zn or the like and having an impurity concentration of 0.8 × 10 ^{17 to} 5 × 10 ¹⁷ cm ⁻³ . The n-type InGaP emitter layer 63 is a layer having a thickness of about 0.05 to 0.3 μm and doped with Si or the like and having an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ⁻³ . The n-type AlInP window layer 44 is a layer having a thickness of about 0.025 to 0.1 μm and doped with Si or the like and an impurity concentration of 1 × 10 ¹⁸ to 4 × 10 ¹⁸ cm ⁻³ . The p-type AlInP back surface field layer 61 may be configured using AlInGaAs.

The n-type InGaAs contact layer 7 is a layer for easily connecting to an external circuit, has a thickness of about 0.2 to 0.5 μm, is doped with Si and Te, and has an impurity concentration of 2 × 10. It is a layer of ¹⁸ to 1 × 10 ¹⁹ cm ⁻³ .

The front electrode 81 is formed of, for example, an AuGe / Ni / Au multilayer film, and the back electrode 82 is formed of Au or Ag. The antireflection film 9 is made of, for example, SiNx, SiO ₂ , TiO ₂ / SiO ₂ , MgF ₂ / ZnS, or the like.

  In the multi-junction solar cell 100 as shown in FIG. 1, the relationship between the band gaps of the solar cells is as follows: the band gap Eg1 of the third solar cell 60> the band gap Eg2 of the second solar cell 40> the first. The band gap Eg3 of one solar battery cell 10 is obtained. That is, the top cell 6 absorbs photons having energy larger than Eg1. The middle cell 4 absorbs photons having energy between Eg2 and Eg1 out of the light that has passed through the top cell 6. Further, the bottom cell 1 absorbs photons having energy between Eg3 and Eg2 out of the light passing through the top cell 6 and the middle cell 4. The band gap is obtained by, for example, using a PL (Photoluminescence) measuring device, irradiating a solar cell that measures laser light (for example, laser light having a wavelength of 532 nm), and examining the wavelength spectrum emitted from the solar cell. Can be measured.

<Action of stress relaxation part>
Next, the effect | action of the stress relaxation part 21 is demonstrated.

  FIG. 3 is a cross-sectional view schematically showing the structure of a multi-junction solar cell having the stress relaxation portion 21. The stress relaxation portion 21 will be described below using a multi-junction solar cell having a stacked structure of Ge substrate / InGaAs layer / InGaP layer shown in FIG.

  In order to relieve the stress caused by the difference in lattice constant between the layers (lattice mismatch), the stress relaxation portion 21 is distributed in the form of islands dispersed at the interface between the layers causing the lattice mismatch. It is an area. In the case of the multi-junction solar cell shown in FIG. 3, a plurality of stress relaxation portions 21 are arranged dispersed on the surface of the Ge substrate (interface between the Ge substrate and the InGaAs layer).

FIG. 4 is a diagram showing a bonding state between the Ge substrate and the InGaAs layer in the multi-junction solar cell having the stress relaxation portion 21. FIG. 5 is a view showing a bonding state between a Ge substrate and an InGaAs layer in a multi-junction solar cell that does not have the stress relaxation portion 21, for comparison with FIG. Here, in FIGS. 4 and 5, the InGaAs layer has a lattice constant different from that of the Ge substrate, and the junction with the Ge substrate becomes lattice mismatch. For example, as described later, the In composition x in the In _x Ga _1-x As layer is set to a value in the range of 0.05 <x <0.2.

  As shown in FIG. 5, since the InGaAs layer has a larger lattice constant than that of the Ge substrate, stress caused by the difference in the lattice constant is generated at the interface between the two. This stress is caused by the bond between atoms of Ge and InGaAs. This stress causes a crystal defect in the InGaAs layer, and causes a decrease in power generation efficiency of the solar cell due to carrier scattering caused by the crystal defect.

  On the other hand, as shown in FIG. 4, when a plurality of stress relaxation portions 21 are dispersed between the Ge substrate and the InGaAs layer, the bonds between the atoms of Ge and InGaAs are partially blocked. Become. Thereby, the stress caused by the difference in lattice constant between the Ge substrate and the InGaAs layer is reduced, and crystal defects caused by the stress are reduced.

  As described above, by dispersing the stress relaxation portions between the Ge substrate and the InGaAs layer, the stress due to the difference in lattice constant between the Ge substrate and the InGaAs layer is relaxed, and the lattice mismatch is caused. Also, an InGaAs layer with few crystal defects can be obtained.

<Manufacturing method of multi-junction solar cell>
Next, a method for manufacturing the multijunction solar cell 100 according to the present embodiment will be described below.

First, the stress relaxation part 21 is formed on the p-type Ge substrate 11. First, a film of SiGe, Si, GaSb, InSb, SiO ₂ , SiNx, Ti, W or the like is formed on the p-type Ge substrate 11 using a sputtering method, a plasma CVD method, or a vapor deposition method. After film formation, a resist is applied, exposed and developed to form a resist pattern in an island shape. Thereafter, the film is etched by a dry etching method or the like, the resist is peeled off, and the stress relaxation portion 21 having an island pattern is formed. The shape of the island pattern that becomes the stress relieving portion 21 is a circle having a diameter of 0.25 μm to 2 μm.

  The stress relaxation portion 21 has a temperature higher than the film formation temperature of each semiconductor layer so that the stress relaxation portion 21 does not dissolve when forming an InGaAs low-temperature buffer layer 22 described later and each semiconductor layer formed subsequently. It is preferable to form with a material having a high melting point.

Moreover, since the stress relaxation part 21 also plays the role which suppresses the spreading | diffusion of As to the p-type Ge board | substrate 11 so that it may mention later, it is preferable to form with the material whose diffusion coefficient is smaller than the p-type Ge board | substrate 11. FIG. In the present embodiment, the diffusion coefficient of As in Ge forming the p-type Ge substrate 11 is, for example, about 3 × 10 ⁻¹⁴ cm ² / sec at 600 ° C., for example, Si (at 600 ° C. The diffusion coefficient of As in Si = about 4 × 10 ⁻²³ cm ² / sec) or the stress relaxation portion 21 is preferably formed of SiGe. This diffusion coefficient can be obtained, for example, by measuring an impurity concentration profile using a known secondary ion mass spectrometry.

Next, a film is formed on the p-type Ge substrate 11 on which the stress relaxation portion 21 is formed by the MOCVD method or the like. When forming an InGaAs film, TMG (trimethylgallium) or TMI (trimethylindium) is used as a group III atom source gas, and AsH ₃ (arsine) is used as a group V atom source gas. In the case of depositing InGaP, TMG (trimethylgallium) or TMI (trimethylindium) is used as a group III atom source gas, and PH ₃ (phosphine) is used as a group V atom source gas. When p-type dopant control is performed, DMZ (dimethyl zinc) or CBr ₄ (carbon tetrabromide) is used as an impurity doping source. When n-type dopant control is performed, SiH ₄ (silane) or DETe (diethyl tellurium) is used as an impurity doping source.

First, the n-type InGaAs low-temperature buffer layer 22 is formed on the p-type Ge substrate 11 on which the stress relaxation portion 21 is formed. SiH ₄ (silane) is used as an impurity doping source. Film formation is performed at a relatively low temperature of 500 to 600 degrees.

  In the process of forming the InGaAs low temperature buffer layer 22, As diffuses toward the p-type Ge substrate 11. Specifically, since the stress relaxation portions 21 are discretely arranged on the upper surface of the p-type Ge substrate 11, As is mainly in a region where the stress relaxation portions 21 on the upper surface of the p-type Ge substrate 11 are not formed. Spread. In this way, As is partially diffused into the p-type Ge substrate 11, the n-type Ge layer 12 is formed. Note that a part of As diffused in the p-type Ge substrate 11 wraps around the lower surface of the stress relaxation part 21, so that the n-type Ge layer 12 extends to a part of the lower surface of the stress relaxation part 21. The p-type Ge substrate 11 and the n-type Ge layer 12 function as the first solar battery cell 10.

  The stress relaxation part 21 also has a function of adjusting the amount of As diffused into the p-type Ge substrate 11. By providing the stress relaxation portion 21, excessive diffusion of As to the p-type Ge substrate 11 can be suppressed, and thereby the n-type Ge layer 12 can be formed thin. Thereby, since absorption of the light in the n-type Ge layer 12 can be suppressed, the 1st photovoltaic cell 10 with sufficient photoelectric conversion efficiency can be formed.

  Next, an InGaAs buffer layer 23 is formed. The InGaAs buffer layer 23 is formed at a temperature of 600 to 700 degrees, which is about 100 degrees higher than the film formation temperature of the InGaAs low temperature buffer layer 22.

  Next, each layer is formed using a predetermined source gas and a doping source. The film formation temperature is 600 to 700 degrees excluding the tunnel junction layer. The film formation temperature of the p-type tunnel junction layer is 500 to 600 degrees, and the film formation temperature of the n-type tunnel junction layer is 700 to 750 degrees.

  Next, after film formation by MOCVD or the like, an AuGe / Ni / Au multilayer film is formed as a surface electrode 81 by lift-off. First, a resist is applied on the n-type InGaAs contact layer 7, and exposure and development are performed to form a predetermined resist pattern. Thereafter, the AuGe / Ni / Au multilayer film is deposited by vacuum deposition, and then the resist is removed. Thereby, the electrode pattern of the surface electrode 81 as shown in FIG. 1 is formed.

Next, the n-type InGaAs contact layer 7 is etched using the electrode pattern of the surface electrode 81 as a mask. As the etching, wet etching using an H ₂ SO ₄ —H ₂ O ₂ —H ₂ O-based etching solution is performed.

After the n-type InGaAs contact layer 7 is etched, an antireflection film 9 is formed by lift-off. First, a resist is applied on the n-type AlInP window layer 64 and the surface electrode 81, and exposure and development are performed to form a predetermined resist pattern. Thereafter, MgF ₂ / ZnS is deposited by sputtering or vacuum deposition, and then the resist is removed. Thereby, the antireflection film 9 is formed. Finally, a back electrode 82 made of Au or Ag is formed under the p-type Ge substrate 11 by vapor deposition or the like.

<Power generation efficiency of multi-junction solar cells>
Next, the power generation efficiency of the multijunction solar cell 100 in one embodiment of the present invention will be described.

  As described above, by forming the stress relaxation portion 21, it is possible to obtain the multi-junction solar cell 100 that solves the problem of stress caused by lattice mismatch between cells. In the following, conditions for the multi-junction solar cell 100 including the lattice mismatch system to have high power generation efficiency will be examined.

  FIG. 6 is a diagram showing the energy conversion wavelength range of sunlight. The band gap shown in FIG. 6 indicates a value in a lattice-matched multijunction solar cell. From the relationship between the conversion wavelength region and the band gap shown in FIG. 6, the relationship between the output currents of the solar cells in the lattice-matched multijunction solar cell is I (InGaP), I (InGaAs) <I (Ge ). That is, since the current matching of the output current of each solar battery cell is not achieved, the output current I (Ge) is regulated by I (InGaP) and I (InGaAs), and energy loss occurs.

  Therefore, in order to improve the power generation efficiency of the multi-junction solar cell 100, the energy loss caused by the regulation of the matching of the output current between the cells is reduced, and the current matching (I ( InGaP) ≈I (InGaAs) ≈I (Ge)) needs to be realized.

  For this purpose, it is required to reduce the band gap of the InGaAs cell and InGaP cell and to widen the energy conversion wavelength region of the InGaAs cell and InGaP cell.

When the relationship between the lattice constants of each cell in a multi-junction solar cell in which this is realized is examined, the In composition x in In _x Ga _1-x As is a value in the range of 0.05 <x <0.2. It becomes.

To apply this to the multi-junction solar cell 100 shown in FIG. 1, an n-type InGaAs low-temperature buffer layer 22, an n-type InGaAs buffer layer 23, an n ⁺⁺ type InGaAs tunnel junction layer 31, a p ⁺⁺ type InGaAs. In the tunnel junction layer 32, the p-type InGaAs base layer 42, and the n-type InGaAs emitter layer 43, the In composition x in In _x Ga _1-x As is set to a value in the range of 0.05 <x <0.2. . The second solar cell 40 at this time is a lattice mismatch system, and for example, the lattice constant becomes larger and the band gap becomes smaller than when the second solar cell 40 is a lattice matching system.

That is, in the multi-junction solar cell 100, current matching (I (InGaP)) is achieved by setting the In composition x in In _x Ga _1-x As to a value in the range of 0.05 <x <0.2. ≈I (InGaAs) ≈I (Ge)), and the power generation efficiency of the solar cell can be improved.

  As described above, according to the present embodiment, the stress relaxation portion is disposed between the first solar battery cell 10 and the second solar battery cell 40 of the multi-junction solar battery 100, so that the lattice between the cells can be obtained. Since stress due to the difference in constant is relieved, a solar battery cell with few crystal defects can be formed even if lattice mismatch occurs, and power generation efficiency can be further increased.

DESCRIPTION OF SYMBOLS 1 Bottom cell 2 Middle area | region 3 1st tunnel area | region 4 Middle cell 5 2nd tunnel area | region 6 Top cell 7 Contact layer 9 Antireflection film 10 1st photovoltaic cell 11 p-type Ge substrate 12 n-type Ge layer 21 Stress relaxation part 22 n-type InGaAs low-temperature buffer layer 23 n-type InGaAs buffer layer 31 n ^++- type InGaAs tunnel junction layer 32 p ^++- type InGaAs tunnel junction layer 40 second solar cell 41 p-type InGap back surface electric field layer 42 p-type InGaAs base Layer 43 n-type InGaAs emitter layer 44 n-type AlInP window layer 44 contact layer 51 n ^++- type InGaP tunnel junction layer 52 p ^++- type InGaP tunnel junction layer 60 third solar cell 61 p-type AlInP back surface field layer 62 p Type InGaP base layer 63 n-type InGaP emitter 64 n-type AlInP window layer 81 surface electrode 82 back-surface electrode

Claims (14)

A first solar cell;
An intermediate region formed on the first solar cell;
A second solar cell formed on the intermediate region,
The intermediate region is arranged in a distributed manner on the surface of the first solar cell, and relieves stress caused by a difference in lattice constant between the first solar cell and the second solar cell. A solar cell having a stress relaxation portion.   The solar cell according to claim 1, wherein the stress relaxation portion is arranged in an island shape on a surface of the first solar cell.   The solar cell according to claim 1, wherein the stress relaxation part has conductivity.   The said stress relaxation part is a solar cell in any one of Claim 1 to 3 comprised with the material whose melting | fusing point is higher than the material which comprises the said 2nd photovoltaic cell.   The solar cell according to any one of claims 1 to 4, wherein the second solar cell is made of a material having a band gap larger than that of the first solar cell. The first solar cell is made of Ge,
The solar cell according to claim 1, wherein the second solar cell is made of InGaAs. The solar cell according to claim 6, wherein a composition x of In in In _x Ga _1-x As constituting the second solar battery cell is a value in a range of 0.05 <x <0.2.   The solar cell according to claim 1, wherein the intermediate region further includes a buffer layer formed on a surface of the first solar battery cell and a surface of the stress relaxation portion. The first solar cell is made of Ge,
The second solar battery cell is made of InGaAs, and the In composition x in In _x Ga _1-x As constituting the second solar battery cell is in a range of 0.05 <x <0.2. Value of
The buffer layer is made of InGaAs, and a composition x of In in In _x Ga _1-x As constituting the buffer layer is a value in a range of 0.05 <x <0.2. The solar cell as described in.   The solar cell according to claim 8 or 9, wherein a refractive index of the stress relaxation part is a value between a refractive index of the first solar battery cell and a refractive index of the buffer layer. The first solar cell is made of Ge,
The buffer layer is made of InGaAs,
The solar cell according to any one of claims 8 to 10, wherein the stress relaxation portion is made of SiGe, Si, GaSb, or InSb.   The solar cell according to claim 1, further comprising a third solar cell formed on the second solar cell and made of a material having a band gap different from that of the material constituting the second solar cell. The solar cell in any one.   The solar cell according to claim 12, wherein the third solar cell is made of a material having a larger band gap than the second solar cell.   The solar cell according to claim 13, wherein the third solar cell is made of InGaP.

Images