JP2002151719A - Thin-film solar cell and manufacturing method thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a thin-film solar cell, together with its manufacturing method, wherein a micro structure in the film-thickness direction of a single doped layer is controlled to raise a photoelectric conversion efficiency. SOLUTION: A photoelectric conversion unit formed on substrate 1 constituting a thin-film solar cell comprises a part in which an n-layer 17, an i-layer 4, and a p-layer 18 are deposited in this order, with the p-layer 18 on the i-layer 4 being a silicon thin film comprising a crystal component. Here, the crystallization percentage of the p-layer 18 is higher on the near side of the i-layer 4 than on the far side of it.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a thin-film solar cell having high photoelectric conversion efficiency by controlling a microscopic structure in a thickness direction of a doped layer.

[0002]

2. Description of the Related Art Solar cells are attracting attention as an alternative energy source for fossil fuels such as petroleum, which is concerned about future supply and demand and has a problem of carbon dioxide emission causing global warming. A solar cell uses a pn junction for a photoelectric conversion layer that converts light energy into electric power, and silicon is generally most often used as a semiconductor constituting the pn junction. From the viewpoint of photoelectric conversion efficiency, it is preferable to use single-crystal silicon, but there are problems in raw material supply, increase in area, and reduction in cost.

On the other hand, a thin film solar cell using amorphous silicon as a photoelectric conversion layer has been put to practical use as a material advantageous for realizing a large area and a low cost. Inferior. Further, since a phenomenon called the Staebler-Wronski effect occurs in amorphous silicon in which the defect density in the film increases as light is irradiated, an amorphous silicon solar cell inevitably suffers from a problem of deterioration with time in photoelectric conversion efficiency.

In recent years, in order to realize a solar cell having both stable and high photoelectric conversion efficiency at the level of a single-crystal silicon solar cell and large area and low cost at the level of an amorphous silicon solar cell, a photoelectric conversion layer is required. The use of polycrystalline silicon is being considered. In particular, the same chemical vapor deposition method (C
A thin-film polycrystalline solar cell in which thin-film polycrystalline silicon is formed using a thin-film formation technique based on a VD method) has been receiving attention.

By the way, the use of a doped layer of a silicon-based thin film containing a crystalline component in thin-film solar cells using amorphous silicon or thin-film polycrystalline silicon has been performed for some time. Its advantages include higher light transmittance, lower resistance than amorphous silicon, and higher internal potential by bringing the Fermi level closer to the valence band and conduction band. In particular, Appl.Phy
As described in s. Lett., Vol. 65, 1994, p. 860,
When a thin film containing a crystal component is used for the intrinsic photoelectric conversion layer, it is more preferable to use a thin film containing a crystal component for the doped layer from the viewpoint of material similarity. However, the above documents do not consider the crystallinity of the doped layer.

[0006] In the present specification, "a silicon-based thin film containing a crystal component" means a spectrum near 520 cm ^-1 corresponding to crystalline silicon as a result of measurement by Raman scattering spectroscopy. , Or a silicon germanium thin film or a silicon carbon thin film.

As an example considering the crystallinity of the doped layer,
There is a method disclosed in JP-A-11-145498. Characteristically, by controlling the crystallization rate of the doped layer containing the crystal component serving as the underlayer, the crystal nucleation density of the intrinsic type photoelectric conversion layer is appropriately suppressed, and the crystal grain boundaries and intragranular defects are reduced. It is to obtain a good quality photoelectric conversion layer.

[0008]

In a pin-type thin-film silicon solar cell, the doping of a single film obtained by a commonly used characteristic evaluation method such as electric conductivity measurement or light transmission / reflectance measurement is carried out. The layer is determined. However, the characteristics of solar cells are often not better than expected from single-film evaluation. This is a polycrystalline silicon thin film formed by chemical vapor deposition such as plasma CVD.
This is because it often has a non-uniform microscopic structure in the thickness direction.

In a pin-type thin-film polycrystalline silicon solar cell, the thickness of the doped layer is usually a few in order to reduce light absorption in the doped layer that does not contribute to photoelectric conversion and to reduce the series resistance component of the solar cell. It is formed so as to be within 10 nm. However, generally, the formation conditions of the doped layer are determined by evaluating the characteristics of a single film having a thickness of several hundreds to 1,000 nm. The thickness of about several hundreds to several thousand nm is a thickness necessary to ensure the reliability of the measured values in a characteristic evaluation device using various optical or electrical methods. Since the value is only information averaged over the entire film, the state of the doped layer having a non-uniform microstructure in the film thickness direction cannot be accurately predicted. Therefore, a solar cell expected from the characteristics of a single film cannot be obtained.

In particular, in a solar cell having a structure in which a current flows in the film thickness direction, the electrical characteristics in the film thickness direction are important, and in a film containing a crystal component, the crystallinity greatly influences the electrical characteristics. I do. Therefore, it is important to design an optimal crystal structure of the doped layer in the thickness direction by using an evaluation method capable of correctly grasping a change in the microscopic structure in the thickness direction of the doped layer.

Also, in the method disclosed in JP-A-11-145498, the crystallization ratio of the doped layer is not evaluated at all with respect to the microstructure in the thickness direction.
Therefore, the microstructure is not properly controlled and improved, and a sufficient crystal structure as a doped layer of the thin-film solar cell has not been obtained.

The present invention has been made in view of the above problems, and has as its object to provide a thin-film solar cell having a high photoelectric conversion efficiency by controlling a microscopic structure in a thickness direction of a single doped layer. Also,
An object of the present invention is to provide a method for manufacturing the thin-film solar cell.

[0013]

In order to achieve the above object, a thin film solar cell according to the present invention includes at least one photoelectric conversion unit formed on a substrate, wherein the photoelectric conversion units are p, i, n layers, or n, i,
A solar cell including a portion deposited in the order of a p-layer, wherein the p- or n-layer on the i-layer is a silicon-based thin film containing a crystalline component, and the crystallization rate of the p- or n-layer on the i-layer is: It is configured such that the side closer to the i-layer is larger than the side farther from the i-layer.

As another structure, the crystallization ratio of the p or n layer on the i layer is gradually reduced from the side closer to the i layer to the side farther from the i layer.

Preferably, the i-layer is a silicon-based thin film containing a crystal component.

More preferably, light is incident from the side of the p or n layer on the i layer.

The p-type or n-type layer on the i-layer may have a conductivity type determining impurity density of 10 ¹⁷ cm ⁻³ or more and 10 ²¹ cm ⁻³ or less.

Further, the p or n layer on the i layer may have a specific resistance of 0.01 Ω · cm or more and 10 Ω · cm or less.

In the method of manufacturing the thin-film solar cell, the p or n layer on the i-layer can be formed by a plasma CVD method using a VHF frequency band larger than 13.56 MHz and smaller than 300 MHz.

Further, in the method of manufacturing the thin-film solar cell, the crystallization ratio can be controlled by forming the p or n layer on the i-layer while changing the dilution ratio of the dilution gas.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, a description will be given of an evaluation method capable of measuring a microscopic structure of a doped layer in a film thickness direction. Journal
Ge _x S formed on a single crystal silicon wafer described in a paper by SJ Chang et al. in Applied Physics, Vol. 64, 1988.
A method using Raman scattering spectroscopy used for stress evaluation of the i _1-x film (hereinafter referred to as oblique polishing Raman scattering spectroscopy) was applied to the evaluation method of the present invention.

FIG. 1 is a sectional view of a sample used for measurement by oblique polishing Raman scattering spectroscopy. The silicon-based thin film 9 including a crystal component formed on the glass substrate 8 is mechanically polished so as to form a small inclination angle α, and the inclined surface 10 is formed between the glass substrate 8 and the silicon-based thin film 9 including the crystal component. To have. Then, the smoothness of the inclined surface 10 is measured by a surface roughness meter, and a boundary 11 between the glass substrate 8 and the silicon-based thin film 9 containing a crystal component formed on the inclined surface 10 is obtained. Here, the boundary 11 is formed due to a difference in hardness between the silicon-based thin film 9 containing a crystal component and the glass substrate 8.
Is obtained from the fact that the inclination angle is different.

Further, the glass substrate 8
Assuming that x is an arbitrary position on the interface 12 between the silicon-based thin film 9 and the silicon-based thin film 9 containing the crystal component, and d is the thickness of the silicon-based thin film 9 containing the crystal component at that position, d = xtan 留. By measuring at a plurality of points on the inclined surface 10 by Raman scattering spectroscopy, it is possible to evaluate a microscopic structure with respect to a continuous change in film thickness.

For evaluation of the microscopic structure, the ratio between the crystalline component and the amorphous component, that is, the crystallization ratio is used. The peak height Ic of the Raman scattering spectrum of about 520 cm ^-1 corresponding to crystalline silicon and the peak height I of the Raman scattering spectrum of about 480 cm ^-1 corresponding to amorphous silicon
By superimposing Gaussian distribution with a, the ratio Ic / (Ic + Ia) of crystalline silicon in the doped layer can be used as an evaluation index of the crystallization ratio. This value does not match the volume fraction.

Next, the reliability of the measured value of the oblique polishing Raman scattering spectroscopy will be described. FIG. 2 shows an obliquely polished sample 13.
The evaluation result of the crystallization ratio of the sample 14 and the unprocessed sample 14 by Raman scattering spectroscopy is shown. The obliquely polished sample 13 is a polycrystalline silicon thin film sample having a thickness of about 1000 nm formed on a glass substrate using a general parallel plate electrode type plasma CVD apparatus, and has a crystallization rate by obliquely polished Raman scattering spectroscopy. Measured. On the other hand, the unprocessed sample 14 is the obliquely polished sample 1
These are several types of flat samples in which the thickness of polycrystalline silicon is variously changed under the same conditions as in No. 3, and the crystallization ratio is measured by ordinary Raman scattering spectroscopy. In FIG. 2, since the polished sample 13 and the non-processed sample 14 show good agreement, the damage of the polished sample 13 due to mechanical polishing is so small as not to affect the measurement, and the oblique polishing Raman scattering spectroscopy is a sufficiently reliable method. It means there is.

As described above, since the microstructure in the film thickness direction can be evaluated, the influence of the microstructure of the doped layer on the characteristics of the solar cell can be seen.

Next, the design of the thin-film solar cell of the present invention will be described. In order to increase the photoelectric conversion efficiency in a thin-film solar cell, it is necessary to design an appropriate doped layer. here,
An appropriate doped layer has a thin film for reducing light absorption that does not contribute to power generation and a reduction in activation energy for obtaining a sufficient built-in potential, that is, a low resistance.

Then, the covering property of a film important for reducing the thickness of the doped layer will be described. As a method of evaluating the coatability of the film, about 500 nm of ZnO is deposited on a substrate such as glass by sputtering, and further, doped layers having various crystallization rates are deposited, and the above-described method is performed by Auger electron spectroscopy (AES). When the Auger electrons derived from ZnO are no longer detected, it is considered that the film of the doped layer has covered the entire surface, and the coating film thickness is determined. Therefore, a film having a smaller coating thickness has a better covering property, and can be made thinner.

FIG. 3 shows the measurement results of the film thickness of the doped layer with respect to the crystallization ratio of the doped layer. As the crystallization rate increases, the coating film thickness for covering the entire surface increases. This is thought to be because the increase in the crystallization ratio makes it easier for the film to grow in an island shape, and as a result, the coating film thickness increases.
Therefore, in order to reduce the thickness of the doped layer, it is appropriate to use a film having a small crystallization ratio and good coverage.

Next, the specific resistance of the film, which is important for reducing the resistance of the doped layer, will be described. FIG. 4 shows the measurement results of the resistivity of the doped layer with respect to the crystallization ratio of the doped layer.
Here, the film thickness of the sample used for the measurement was set to 80 nm, which becomes the coating film thickness at an arbitrary crystallization rate from the results of FIG. FIG. 4 shows that the specific resistance decreases as the crystallization rate increases. This is because the crystallization proceeds, the activation rate of the dopant increases, the Fermi level approaches the conduction band and the valence band more, and the activation energy decreases.
It is considered that the resistance of the doped layer is reduced.

As described above, the doped layer having a high crystallization rate is
Although it is advantageous for lowering the resistance, the film has poor coatability and is not suitable for thinning. Further, a doped layer having a low crystallization rate has a good film covering property and is advantageous for thinning, but has a high resistance. Therefore, in order to obtain a thin-film solar cell having a high photoelectric conversion efficiency, in the doped layer on the i-layer, the crystallization ratio is increased on the side in contact with the i-layer, thereby reducing the activation energy of the doped layer and increasing the built-in energy It is preferable to obtain a potential and further reduce the crystallization rate in the film thickness direction, thereby improving the coatability of the film and reducing the film thickness. By controlling the crystallization ratio of the doped layer in this way, a thin-film solar cell with reduced light absorption and increased short-circuit current due to thinning of the doped layer, and a large open-circuit voltage due to a high built-in potential can be manufactured.

Hereinafter, embodiments of the thin-film solar cell will be described.
FIG. 5 is a schematic side sectional view illustrating a thin-film solar cell according to an example of the embodiment. The substrate 1 supports and reinforces the entire thin-film solar cell 15 and may be made of a material having heat resistance, such as a glass substrate, a heat-resistant polymer film, or a SUS substrate.

On the substrate 1, a back electrode 2, which is a single-layer or multiple-layer metal film, is formed with a thickness of several hundred to 1,000 nm by a sputtering method or a vacuum evaporation method. As a material of the back electrode 2, Ag, Cu, Au, Ni, Cr, W, Ti, Pt,
Fe, Mo, etc. can be used. The back electrode 2 and the substrate 1
, A transparent conductive film (not shown) such as SnO ₂ , InO ₂ , ZnO, or ITO may be interposed.

On the back electrode 2, a first doped layer 3, which is a silicon-based thin film containing a crystal component, is deposited to a thickness of several tens nm by a plasma CVD method. First doped layer 3
Is deposited by using a silicon-based gas such as SiH ₄ , Si ₂ H ₆ or SiF ₄ and a p-type dopant gas such as B ₂ H ₆ or PH.
A rare gas such as H ₂ or Ar is used as a diluent gas in a mixed gas with an n-type dopant gas such as ₃ .

On the first doped layer 3, an i-layer 4 is deposited to a thickness of about 1000 nm by a plasma CVD method. i layer 4
As a material of the above, a silicon-based thin film containing a crystal component is preferable. SiH ₄ , Si ₂ H ₆ , SiF ₄
The silicon-based gas and the like, H ₂ or A as the diluent gas
A rare gas such as r is used.

On the i-layer 4, as a doped layer, a second doped layer 5, which is a silicon-based thin film containing a crystal component of a conductivity type opposite to that of the first doped layer 3, is formed by several tens of n by a plasma CVD method.
m. The deposition of the second doped layer 5 includes:
A silicon-based gas such as SiH ₄ , Si ₂ H ₆ , SiF ₄ ,
A rare gas such as H ₂ or Ar is used as a diluent gas in a mixed gas with an n-type dopant gas such as PH ₃ or a p-type dopant gas such as B ₂ H ₆ . Then, by gradually decreasing the flow rate of the dilution gas, the second doped layer 5 is deposited while gradually decreasing the crystallization ratio.

On the second doped layer 5, a transparent conductive film 6 is deposited to a thickness of about 100 nm by sputtering or vacuum evaporation. As a material of the transparent conductive film 6, SnO ₂ , In
O ₂ , ZnO, ITO and the like can be used.

On the transparent conductive film 6, a comb-shaped metal electrode 7 serving as a collecting electrode is formed for several hundred n by sputtering or vacuum evaporation.
m. The material of the metal electrode 7 is A
g, Al, Cu, Au, Ni, Cr, W, Ti, Pt,
Fe, Mo, etc. can be used.

With such a layer configuration, the second layer on the i-layer 4
The crystallization ratio of the doped layer 5 is large on the side in contact with the i-layer 4 and is gradually reduced in the film thickness direction. The solar cell 15 can be obtained. In particular, the effect is increased by making the i-layer 4 a silicon-based thin film containing a crystal component. This is because band gap discontinuity does not occur at the junction between the second doped layer 5 and the i-layer 4, and the resistance of the second doped layer 5 and the activation energy are reduced, so that p
This is because the built-in potential of the in or nip junction is increased, and among the characteristics of the thin-film solar cell 15, the short-circuit current density and the open-circuit voltage are improved.

Other methods for changing the crystallization ratio of the second doped layer 5 include a method for changing the power of the plasma generator and a method for changing the substrate temperature.

Further, since light is incident from the second doped layer 5 side, light absorption that does not contribute to power generation due to the thinning of the second doped layer 5 can be reduced.
Is more efficient.

As another embodiment of the thin-film solar cell, a thin-film solar cell in which light is incident from the substrate side, or a plurality of photoelectric conversion units including at least one second doped layer of the above-described embodiment are stacked. Tandem thin film solar cells are also possible.

Hereinafter, examples will be described together with comparative examples. Example 1 FIG. 6 shows the results of oblique polishing Raman scattering spectroscopy of n layers deposited at different frequencies. In the plasma CVD apparatus of a parallel plate electrode type, the PH ₃ is used as a dopant gas, 13.56 MHz and 81.36MH
An n-layer of a silicon-based thin film containing a crystal component is deposited on a substrate at different frequencies of z. The film forming conditions other than the frequency are fixed.

Here, the n-layer deposited at 13.56 MHz has no crystal component up to about 10 nm from the start of the film formation, and thereafter the crystal component increases as the film thickness increases. In the n-layer deposited at .36 MHz, a crystal component is present immediately after the start of film formation.
The saturation value is reached at a film thickness of m. Thus, the n-layer is 81.
This indicates that a film sufficiently containing a crystal component can be formed at 36 MHz.

Example 2 FIG. 7 shows the results of oblique polishing Raman scattering spectroscopy of p layers deposited at different frequencies. In a parallel plate electrode type plasma CVD apparatus, B ₂ H ₆ is used as a dopant gas, and p-layers of silicon-based thin films containing crystal components are deposited on the substrate at different frequencies of 13.56 MHz and 81.36 MHz, respectively. .
The film forming conditions other than the frequency are fixed. here,
The p-layer deposited at 13.56 MHz has no crystal component over the entire thickness, whereas the p-layer deposited at 81.36 MHz has a crystal component immediately after the start of film formation, and has a thickness of about 10 nm. To reach the saturation value. in this way,
The p-layer shows that a film sufficiently containing a crystal component can be formed at 81.36 MHz.

Further, from the results of Example 1 and Example 2, it is understood that the p layer is generally less likely to be crystallized than the n layer. This means that in the p-layer film formation, radicals such as BH ₃ generated from B ₂ H ₆ extract hydrogen atoms terminating the film surface, and thus the surface diffusion distance of SiH ₃ radicals of the deposition precursor. It is thought that this is to shorten crystallization and prevent crystallization. However, even if the p-layer is formed,
By using the VHF frequency band of 1.36 MHz, it is possible to sufficiently crystallize from the initial stage of film formation. As described above, it is understood that the use of a VHF frequency band exceeding 13.56 MHz without limitation to the n-type or p-type as the doped layer is advantageous for controlling the crystallization ratio of the doped layer. However, if a frequency of 300 MHz or more is used, power cannot be effectively supplied to the electrode surface, and film formation becomes difficult.

Example 3 FIG. 8 is a schematic side sectional view showing a thin-film solar cell of Example 3. Ag was sputtered on a substrate 1 made of glass having a thickness of 3 mm by 400 n.
The back electrode 2 is formed by depositing
The transparent conductive film 16 is formed by depositing 100 nm of nO by sputtering. Then, on the transparent conductive film 16,
13.56 of a silicon-based thin film containing a crystal component
Gas flow rate of SiH by RF plasma CVD method of MHz
_{4 5sccm, PH 3 0.01sccm, H} 2 100sc
40 nm under a condition of 200 cm in temperature, a substrate temperature of 200 ° C., and an internal pressure of 0.25 Tool.

Next, on the n-layer, a silicon-based thin film 4 containing a crystal component was coated with a 13.56 MHz VHF band plasma CVD method at a gas flow rate of 5 sccm for SiH ₄ and H ₂ 90
1000 nm is deposited under the conditions of sccm, a substrate temperature of 200 ° C., and an internal pressure of 0.3 Tool.

Further, on the i-layer 4, a silicon containing a crystal component is formed.
The p-layer 18 of the capacitor-based thin film is
Gas flow rate of SiH by plasma CVD_Four4sccm,
B_TwoH ₆0.15 sccm, H_Two90 sccm flow rate
From 20 to 30 sccm
You. These sequentially stacked nip junction photoelectric conversion units
100 nm of ITO was deposited on the substrate by sputtering.
Thus, the transparent conductive film 6 is formed. On the transparent conductive film 6,
Ag is formed into a comb shape by vacuum evaporation to extract current
The metal electrode 7 is formed by depositing 300 nm.

Here, the crystallization ratio of the p layer 18 was measured.
FIG. 9 shows the result of measuring the change in the crystallization ratio in the thickness direction of the p layer 18 by oblique polishing Raman scattering spectroscopy.
When the film thickness is small, that is, on the i-layer 4 side, the crystallization ratio is large, and as the film thickness increases, the crystallization ratio decreases. Thus, it is understood that the p-layer 18 which is a thin doped layer having a high built-in potential is obtained.

Comparative Examples 1 and 2 Comparative Examples 1 and 2
In the third embodiment, the H ₂ flow rate was set to 80 under the conditions for forming the p-layer 18.
It is constant at sccm, and can be completely covered in Comparative Example 1.
In Comparative Example 2, the thickness was set to 20 nm as in Example 3.
nm. Other conditions were the same as in Example 3 to produce a thin-film solar cell. The crystallization ratio in the thickness direction of the p-layer deposited on the i-layer in Comparative Example 1 and Comparative Example 2 becomes smaller on the side in contact with the i-layer, similarly to the 81.36 MHz film shown in FIG. I have.

Comparative Example 3 In Comparative Example 3, p
Under the conditions for forming the layer 18, the H ₂ flow rate was kept constant at 30 sccm, and the film thickness was set to 20 nm as in Example 3. Other conditions were the same as in Example 3 to produce a thin-film solar cell. The crystallization ratio in the thickness direction of the p-layer deposited on the i-layer in Comparative Example 2 is smaller on the side in contact with the i-layer, as in Comparative Examples 1 and 2.

Table 1 summarizes the characteristics of the thin-film solar cells of Example 3 and Comparative Examples 1 to 3. The characteristic evaluation is standardized based on the results of Comparative Example 1. The conversion efficiency was measured by AM
Simulated sunlight of 1.5 and 100 mW / cm ² was irradiated.

[0054]

[Table 1]

As shown in Table 1, Example 3 is thinner than Comparative Example 1, and the short-circuit current density is improved. Further, in Example 3, since the crystallization ratio is large and the activation energy is small on the side where the p layer is in contact with the i layer, the built-in potential can be increased, and the conversion efficiency is improved without lowering the open circuit voltage. . On the other hand, in Comparative Example 2, since the crystallization ratio of the p-layer, which is the doped layer, is high, the layer cannot be completely covered with a film thickness of 20 nm, and the conversion efficiency is reduced due to the leak current. In Comparative Example 3, the p-layer, which is a doped layer, has a low crystallization rate and a large activation energy, so that the open-circuit voltage and the short-circuit current density decrease due to the decrease in the built-in potential, and the conversion efficiency decreases. are doing.

Embodiment 4 In Embodiment 4, i of Embodiment 3
The flow rate of H ₂ was changed under the conditions for forming the layer 4 so that the i-layer became amorphous. Other conditions were the same as in Example 3 to produce a thin-film solar cell.

Comparative Example 4 In Comparative Example 4, i of Comparative Example 1
The H ₂ flow rate was changed depending on the layer formation conditions so that the i-layer became amorphous. Other conditions were the same as in Comparative Example 1 to produce a thin-film solar cell. That is, in Comparative Example 4, a film having a higher crystallization rate of the p-layer, which is the doped layer on the i-layer, was deposited to a thickness of 60 nm as compared with Example 4.

Table 2 summarizes the characteristics of the thin film solar cells of Example 4 and Comparative Example 4. The characteristic evaluation is standardized based on the results of Comparative Example 4. The conversion efficiency was measured according to AM1.
5, simulated sunlight of 100 mW / cm ² was irradiated.

[0059]

[Table 2]

From Table 2, it can be seen that when the i-layer is amorphous, the effect of the present invention is small since the material consistency cannot be obtained as compared with the case where a silicon-based thin film containing a crystal component is used for the i-layer.

Embodiments 5 to 7 Embodiments 5 and 6
In the third embodiment, the conductivity type determining impurity density, that is, the impurity concentration of the doped layer is changed by changing the flow rate of the dopant gas B ₂ H _{6 under} the conditions for forming the p-layer 18 in the third embodiment. Thus, a p-layer, which is a doped layer, is obtained in which the dopant gas concentration is increased as far as the properties can be obtained and the specific resistance is minimized. Other conditions were the same as in Example 3 to produce a thin-film solar cell.

[Comparative Example 5] In Comparative Example 5, p
The flow rate of the dopant gas B ₂ H ₆ was reduced under the conditions for forming the layer to reduce the impurity density of the doped layer. Other conditions were the same as in Comparative Example 1 to produce a thin-film solar cell. That is,
Comparative Example 5 is different from Example 5 in that a film having a high crystallization rate of a p-layer, which is a doped layer on the i-layer, is deposited to a thickness of 60 nm.

Table 3 summarizes the characteristics of the thin film solar cells of Examples 3, 5 to 7 and Comparative Examples 1 and 5. The characteristic evaluation is standardized based on the results of Comparative Example 1. The conversion efficiency was measured by irradiating simulated sunlight of AM 1.5 and 100 mW / cm ² .

[0064]

[Table 3]

From Table 3, it can be seen that Comparative Example 5 has a lower doped layer impurity density than Comparative Example 1, and therefore has a large specific resistance and a small short-circuit current density. Further, since the activation energy is also increased, the built-in potential is reduced, the open-circuit voltage is reduced, and the conversion efficiency is reduced. Although the conversion efficiency of Example 5 is higher than that of Comparative Example 5, the conversion efficiency is lower than that of Comparative Example 1 because the impurity density of the doped layer is low, which is not desirable as solar cell characteristics.

In Example 6, although the impurity concentration of the doped layer is lower and the specific resistance is higher than that of Comparative Example 1, the conversion efficiency is improved. Therefore, in order to obtain sufficient solar cell characteristics, the impurity density of the doped layer of Example 6 is 1 × 10 ¹⁷ cm ⁻³ and the specific resistance is 10 Ω · cm. Moreover, Example 7 has a higher conversion efficiency than Comparative Example 1, and has sufficient solar cell characteristics. In order to obtain a doped layer having sufficient crystallinity at the device level, the impurity density of the doped layer of Example 7 is 10 ²¹ cm ⁻³ as the maximum value, and the specific resistance is 0.01 Ω · cm as the minimum value.

In the above Examples and Comparative Examples, the doped layer on the i-layer was the p-layer and was on the light incident side.
In addition to the structure of the thin-film solar cell, the same effect can be obtained even if the doped layer on the i-layer is an n-layer. The effect can be obtained even when the light incident direction is reversed, but the effect of reducing the light absorption by reducing the thickness of the doped layer becomes small. Furthermore, it goes without saying that the same effect can be obtained even in the case of a stacked solar cell in which a plurality of pin or nip junction photoelectric conversion units are stacked.

[0068]

According to the present invention, the microscopic structure in the thickness direction of a single layer of the doped layer is analyzed by oblique polishing Raman scattering spectroscopy, and the optimal structure of the doped layer using a silicon-based thin film containing a crystal component is analyzed. As a crystal structure, a doped layer that achieves both thinning and low resistance is designed, and the activation energy is reduced by increasing the crystallization rate of the doped layer on the side in contact with the i-layer, resulting in a high built-in solar cell device. The potential is obtained, the crystallization ratio of the doped layer is reduced in the film thickness direction, thereby improving the film coverage and reducing the film thickness, obtaining a large short-circuit current density, and producing a thin-film solar cell with high photoelectric conversion efficiency. .

Further, according to the present invention, by using a silicon-based thin film containing a crystal component in the i-layer, discontinuity of the band gap does not occur at the junction between the doped layer and the i-layer, and the thin-film solar cell having higher photoelectric conversion efficiency Can be produced.

Further, according to the present invention, light is incident from the side of the doped layer on the i-layer, so that the thinned doped layer allows
Since light absorption that does not contribute to power generation can be reduced, the photoelectric conversion efficiency of the thin-film solar cell can be higher.

[Brief description of the drawings]

FIG. 1 is a cross-sectional view of a sample of the present invention used for measurement by oblique polishing Raman scattering spectroscopy.

FIG. 2 shows the results of evaluation of the crystallization ratio of the obliquely polished sample and the unprocessed sample by Raman scattering spectroscopy.

FIG. 3 is a measurement result of a coating thickness of a doped layer with respect to a crystallization ratio of the doped layer of the present invention.

FIG. 4 is a measurement result of the specific resistivity of the doped layer with respect to the crystallization ratio of the doped layer of the present invention.

FIG. 5 is a schematic side sectional view showing a thin-film solar cell according to an example of an embodiment of the present invention.

FIG. 6 is a measurement result of oblique polishing Raman scattering spectroscopy of the n-layer of the present invention deposited at different frequencies.

FIG. 7 shows the results of oblique polishing Raman scattering spectroscopy of p-layers of the present invention deposited at different frequencies.

FIG. 8 is a schematic side sectional view showing a thin-film solar cell of Example 3 of the present invention.

FIG. 9 shows the results of measurement of the change in the crystallization ratio in the thickness direction of the p-layer according to the present invention by oblique polishing Raman scattering spectroscopy.

[Explanation of symbols]

 Reference Signs List 1 substrate 4 i layer 9 silicon-based thin film containing crystal component 15 thin film solar cell 17 n layer 18 p layer

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 4K030 AA06 BA29 BB12 CA02 CA06 CA07 FA03 HA03 HA04 JA06 JA18 LA16 5F045 AA08 AB04 AC01 AC19 AF07 AF10 CA13 DA59 DA61 EE14 5F051 AA03 AA04 CB12 CB29 CB30 DA04

Claims (8)

[Claims] And a photoelectric conversion unit including at least one photoelectric conversion unit formed on a substrate.
a solar cell including a portion deposited in the order of an i, n layer, or an n, i, p layer, wherein the p or n layer on the i layer is a silicon-based thin film containing a crystalline component; A thin-film solar cell, wherein the crystallization ratio of the n-layer is larger on the side closer to the i-layer than on the side farther from the i-layer. 2. The thin-film solar cell according to claim 1, wherein the crystallization ratio of the p-layer or the n-layer on the i-layer gradually decreases from a side closer to the i-layer to a side farther from the i-layer. . 3. The thin-film solar cell according to claim 1, wherein the i-layer is a silicon-based thin film containing a crystalline component. 4. The thin-film solar cell according to claim 1, wherein light is incident from the side of the p-layer or the n-layer on the i-layer. 5. The p-type or n-type layer on the i-layer has a conductivity type determining impurity density of 10 ¹⁷ cm ⁻³ or more and 10 ²¹ cm ⁻³ or less. Thin-film solar cells. 6. The thin-film solar cell according to claim 1, wherein a specific resistance of the p or n layer on the i-layer is 0.01 Ω · cm or more and 10 Ω · cm or less. 7. The method of manufacturing a thin film solar cell according to claim 1, wherein the p or n layer on the i layer is formed by a plasma CVD method using a VHF frequency band greater than 13.56 MHz and less than 300 MHz. Method. 8. A thin-film solar cell according to claim 1, wherein the p-layer or the n-layer on the i-layer is formed while changing the dilution ratio of the dilution gas to control the crystallization ratio. Method.