JP2002299669A - Polycrystalline solar cell and method of manufacturing polycrystalline solar cell - Google Patents

Abstract

(57) [Problem] To provide a polycrystalline solar cell having good characteristics and a method of manufacturing the polycrystalline solar cell. An active layer (12) is provided on a polycrystalline silicon substrate (11).
In a polycrystalline solar cell including the active layer 12 and the semiconductor layer 13 having the opposite polarity, the active layer 12 is formed of a polycrystal that reflects the crystal structure of the substrate 11, and the active layer 1
Since the thickness of the polycrystalline grain boundary portion 16 forming the second layer 2 is smaller than the thickness of the polycrystalline grain central portion and the grain boundary portion 16 has a groove shape, a groove is formed in the semiconductor layer 13 on the grain boundary portion 16. The electrode 14 for taking out the generated current is buried in the groove.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a polycrystalline solar cell and a method for manufacturing a polycrystalline solar cell. More specifically, the present invention relates to a polycrystalline solar cell in which thin-film polycrystalline silicon is stacked on a low-cost substrate, and a method for manufacturing a polycrystalline solar cell.

[0002]

2. Description of the Related Art Solar cells have been widely studied as power sources that are system-connected to driving energy sources of various devices and commercial power.

[0003] For a solar cell, it is desired that an element can be formed on a low-cost substrate such as a metal due to cost requirements. On the other hand, silicon is generally used as a semiconductor constituting a solar cell. Among them, single crystal silicon is the most excellent from the viewpoint of efficiency of converting light energy into electromotive force, that is, photoelectric conversion efficiency. However, amorphous silicon is considered to be advantageous from the viewpoint of increasing the area and reducing the cost.

[0004] In recent years, the use of polycrystalline silicon has been studied for the purpose of obtaining low cost comparable to amorphous silicon and high energy conversion efficiency comparable to single crystal silicon. However, in a method conventionally proposed for such a single crystal or polycrystalline silicon, a bulk crystal is sliced into a plate-like substrate.
mm was difficult. Therefore, the substrate is
The material was not thick enough to absorb the amount of light, so the material was not effectively used. That is, in order to reduce the cost, it was necessary to further reduce the thickness.

[0005] As a method of manufacturing a polycrystalline silicon substrate aiming at low cost, there has been proposed a method of forming a silicon wafer by using a spin method in which molten Si droplets are poured into a crucible. However, this method has a thickness of at least about 0.1 to 0.2 mm at the minimum, and has not been sufficiently thinned as compared with a film thickness (20 to 50 μm) necessary and sufficient for light absorption as crystalline Si. In addition, such thinning makes it difficult for the silicon wafer itself to have the strength as a substrate, which inevitably requires another inexpensive substrate for supporting the silicon wafer. Was.

Accordingly, there has been reported an attempt to form a solar cell by forming a substrate made of metal-grade silicon and then forming a silicon layer having a film thickness necessary and sufficient for light absorption on the substrate by a liquid phase growth method (see, for example, Japanese Patent Application Laid-Open No. H11-157556). TFCiszek, THWang, X.Wu, RWBurro
ws, J. Alleman, CRSchwerdtfeger and T. Bekkedahl,
"Si thin layer growth from metal solution on sing
le-crystal and cast metallurgical-grade multicryst
alline Si substrates ", 23rd IEEE Photovoltaic spec
ialists Conference, (1993) p.65).

In the above-mentioned method, a silicon layer is formed on a low-purity silicon substrate made of metal-grade silicon using copper, aluminum or tin as a metal solvent. In either case, a natural oxide film is removed at the beginning of growth. Perform etch back. When this etch-back is performed, the metal which is a solvent is mainly left behind in the crystal grain boundaries, and as a result, sufficient characteristics as a solar cell have not been obtained.

In order to solve this problem, there has been disclosed a method in which copper and an aluminum alloy are used as a solvent and etching back is not performed (THWang, TFCiszek, CRSchwerdtfeg).
er, H. Moutinho, R. Matson. "Growth of silicon thin l
ayers on cast MG-Si frommetal solutions for solar
cells ". Solar Energy Materials and Solar Cells41 / 4
2 (1996) p.19). However, this method has a problem that, when mass production is intended, the composition control of the alloy becomes complicated.

[0009] The applicant of the present invention has disclosed in
No. 8205, a method for manufacturing a solar cell including a step of forming a silicon layer on a crystalline silicon substrate by a liquid phase growth method using a metal solvent, wherein the total impurity concentration on the surface of the crystalline silicon substrate is 10 ppm or more. And
The present invention proposes a method for manufacturing a solar cell, wherein the metal solvent is indium.

[0010] Also, G. Ballhorn, KJ Weber, S. Armand, MJ
Stocks, AWBlakers. "High-efficiency multicrystall
ine silicon solar cells ". Solar Energy Material an
d Cells 52 (1998) p.61-p.68 show that when a silicon layer is formed on a polycrystalline silicon substrate by liquid phase epitaxy, growth and meltback are repeated to achieve flat growth. It is stated that it can be obtained.

However, in general, in the case of a polycrystalline solar cell, there are many defects at grain boundaries, and charge recombination there is large. In order to improve the conversion efficiency of a polycrystalline solar cell, it is necessary to suppress the recombination of charges at grain boundaries.
Conventionally, for this purpose, a thin film of amorphous silicon or microcrystalline silicon is formed on polycrystalline silicon, and the dangling bonds at the grain boundaries are terminated by H atoms contained in the film to inactivate the grain boundaries. Recombination was suppressed.
However, this method imposes great limitations on subsequent process conditions in order to prevent the escape of terminated H atoms. The effect was not enough.

In general, when power generated by a solar cell is taken out to an external circuit, it is taken out from a grid-like current collecting electrode from the light irradiation side. However, since light is not irradiated below the current collecting electrode, the region becomes a region where power generation does not occur. Therefore, an attempt is made to reduce the area of the collecting electrode as much as possible. However, if the area is too small, the series resistance increases and the conversion efficiency decreases. Therefore, there is an optimum value for the area of the collecting electrode. This value depends on the generated current and the resistance of the underlying semiconductor layer or transparent electrode layer, but is generally about 5% to 15% with respect to the area of the light incident surface of the solar cell.
Therefore, a reduction in power generation efficiency due to the presence of the collecting electrode is inevitable.

Further, in the case of a polycrystalline solar cell, a grain boundary portion exists on the light incident surface. When the polycrystalline layer as the active layer is irradiated with light, charges generated at the center of the crystal grain can be taken out to an external circuit as a photocurrent. However, the charge generated in the vicinity of the crystal grain boundary has a large proportion reaching the grain boundary. In general, the crystal grain boundaries have many defects and many recombination centers of generated charges, so that many charges recombine and disappear at these grain boundaries, and the proportion of photocurrent that can be taken out to an external circuit is reduced. I will. Therefore, when a collecting electrode is provided in a polycrystalline solar cell, it is necessary to consider the area of the grain boundary portion in addition to the area of the collecting electrode itself (the grain boundary portion is not shielded from light by the collecting electrode. However, since the rate of contribution to power generation is small, it is necessary to reduce the total area of the current collecting electrode portion and the grain boundary portion.)

JP-A-5-299677 describes a method using a ZnO translucent electrode as a current collecting electrode. However, in general, a light-transmitting electrode has a higher resistivity than a metal or the like and is not suitable for a large-area solar cell.

Japanese Patent Application Laid-Open No. 6-125100 discloses a method of lowering the resistance of a current collecting electrode by placing solder on top.

[0016]

SUMMARY OF THE INVENTION An object of the present invention is to provide a polycrystalline solar cell having good characteristics and a method for manufacturing the polycrystalline solar cell. That is, an object of the present invention is to provide a polycrystalline solar cell such as a thin-film polycrystalline silicon having a high conversion efficiency and a method of manufacturing the polycrystalline solar cell.

A second object of the present invention is to provide an inexpensive polycrystalline solar cell and a method for manufacturing the polycrystalline solar cell.

[0018]

The present invention solves the above-mentioned problems in the prior art, and has been completed as a result of intensive studies by the present inventors in order to achieve the above objects.

A polycrystalline solar cell according to the present invention is a polycrystalline solar cell comprising an active layer and a semiconductor layer having the opposite polarity to the active layer on a polycrystalline silicon substrate. The thickness at the grain boundary portion of the polycrystal forming the active layer is smaller than the thickness at the center portion of the polycrystal grain, and the grain boundary portion has a groove shape. A groove is formed in the semiconductor layer on the grain boundary portion, and an electrode for extracting a generated current is embedded in the groove.

Preferably, the active layer is a liquid phase grown active layer.

It is preferable that the film thickness at the grain boundary of the polycrystal forming the active layer is in the range of 0.20 to 0.85 with respect to the film thickness at the center of the polycrystal.

It is preferable that the film thickness at the grain boundary of the polycrystal forming the active layer is in the range of 0.25 to 0.75 with respect to the film thickness at the center of the polycrystal.

It is preferable that the area of the electrode portion existing per unit area is 1% or more and 20% or less.

It is preferable that the polycrystalline silicon substrate is metal-grade silicon containing 10 ppm or more of impurities.

The electrode for taking out the generated current embedded in the groove is preferably a metal paste.

The metal paste is made of Ag, Al, Cu, Cr
It is preferable that the powder is mixed with a polymer.

The metal powder is preferably a mixed metal powder of a flaky metal powder of 0.3 to 0.5 μm and a spherical metal powder of 0.1 to 0.2 μm.

The electrode for taking out the generated current embedded in the groove is composed of a metal paste, a conductive powder, and a metal or alloy or a mixture (low-melting metal) layer having a lower melting point than the conductive powder. Preferably, the conductive powder is fixed on the semiconductor layer via the metal paste, and the low melting point metal layer is formed thereon.

The conductive powder is preferably made of at least one of Ni, Ag, Cu or an alloy thereof.

The method for manufacturing a polycrystalline solar cell according to the present invention is directed to a method for manufacturing a polycrystalline solar cell comprising an active layer grown on a polycrystalline silicon substrate and a semiconductor layer having the opposite polarity to the active layer. A film thickness at a grain boundary portion of the polycrystal forming the active layer is smaller than a film thickness at a central portion of the polycrystal grain, and forming the grain boundary portion in a groove shape; Embedding an electrode for extracting a current.

It is preferable that the electrode for extracting the generated current to be embedded in the groove is a metal paste.

The electrode for taking out the generated current embedded in the groove is composed of a metal paste, a conductive powder, and a metal or alloy or a mixture (low-melting metal) layer having a lower melting point than the conductive powder. Preferably, the method further includes a step of fixing the conductive powder on the semiconductor layer via the metal paste, and a step of forming the low melting point metal layer thereon.

[0033]

The present inventor has conducted a number of experiments in order to improve the conversion efficiency of a polycrystalline silicon solar cell. As a result, it was found that the presence of the collecting electrode greatly affected the conversion efficiency. Since no light is irradiated below the collecting electrode,
No power generation occurs. If there is no current collecting electrode, power is naturally generated, so the power generation efficiency is reduced by the area ratio of the current collecting electrode. Therefore, the smaller the area ratio of the collecting electrode, the higher the conversion efficiency. However, if the area ratio of the collecting electrode is too small, the series resistance increases and the fill factor of the solar cell deteriorates,
Conversion efficiency decreases. Therefore, it was found that there was an optimum value for the area ratio of the collecting electrode. Although it depends on the material of the collecting electrode and the surface resistance of the surface semiconductor layer, it is usually about 5 to 15%. Therefore, a decrease in the conversion efficiency at this ratio cannot be avoided.

On the other hand, in a polycrystalline solar cell, it is known that the power generation efficiency in the vicinity of the grain boundary is significantly lower than that in the center of the crystal grain. This is particularly remarkable when the thickness of the polycrystalline grain boundary is large.

Therefore, as described above, when a current collecting electrode is provided in a polycrystalline solar cell, it is necessary to consider the area of the grain boundary portion in addition to the area of the current collecting electrode itself (the grain boundary portion is not Even if the upper part is not shaded by the collecting electrode, the contribution to power generation is small, so it is necessary to reduce the total area of the collecting electrode portion and the grain boundary portion.

On the other hand, if the current collecting electrode can be formed on the grain boundary, the decrease in power generation efficiency due to no light irradiation due to the presence of the current collecting electrode can be extremely reduced. However, it is extremely difficult to form a collecting electrode on a grain boundary. Since the grain boundaries are irregularly formed, there is no other method than manual operation, and industrialization is impossible.

The present inventors have conducted experiments for growing silicon crystals on a polycrystalline silicon substrate by a liquid phase method. As a result, it was found that by selecting the growth conditions, the thickness of the crystal grain boundary can be made smaller than the thickness of the crystal grain.

Further, the ratio of the thickness of the polycrystalline grain boundary to the thickness of the polycrystalline grain could be set to a desired value by controlling the growth conditions in the liquid phase growth method. That is, a groove-like structure could be formed along the grain boundaries.

Here, it is easy to embed the conductive paste only in these grooves. That is, it was found that the semiconductor surface can be rubbed with a blade made of rubber or a soft and elastic plastic.

It has also been found that a thin-film crystal solar cell having good characteristics can be formed by depositing a silicon layer on the surface of a low-purity silicon substrate such as metal-grade silicon by a liquid phase growth method using indium.

Further, it is demonstrated that a thin-film crystalline solar cell having good characteristics can be formed by depositing a silicon layer on the surface of a plate-like metal-grade silicon melted / solidified in a crucible by a liquid phase growth method using indium. I found it. Further, the surface layer of the plate-shaped metal-grade silicon melted / solidified in the crucible is further dissolved / reprecipitated with a metal solvent, and then a silicon layer is deposited thereon by a liquid phase growth method using indium. They found that a perfect solar cell could be formed, and completed the present invention.

In the present invention, since the current collecting electrode is limited to the vicinity of the grain boundary where power generation efficiency is poor, a reduction in power generation efficiency due to light not being irradiated onto the crystal grains due to the presence of the current collecting electrode is suppressed. And the conversion efficiency can be improved.

The present invention positively utilizes the structural characteristics of polycrystal, that is, the difference (= unevenness) between the thickness of the film near the grain boundary and the thickness of the central portion of the crystal grain. This is completely different from a technique that requires a flattening process.

Even when an inexpensive metal-grade silicon substrate is used as the substrate for growing the active layer in the liquid phase, the active layer with less impurities is grown by contacting the surface with indium to perform the liquid phase growth. Can be
A solar cell with high conversion efficiency can be obtained.

The point that the thickness (t2) of the film near the grain boundary is made smaller than the thickness (t1) of the central portion of the crystal grain has the following operation and effect. By setting the ratio of t2 / t1 to be 0.20 or more and 0.85 or less, more preferably 0.25 or more and 0.75 or less, the traveling distance of the electric charge near the grain boundary is shortened, and the traveling time of the electric charge is shortened. . As a result, recombination of charges in the solar cell is reduced, the current taken out to an external circuit is increased, and the conversion efficiency is improved.

In the case of 0.20 or less, the thickness of the crystal grain boundary becomes significantly smaller than the thickness of the crystal grain, and the leakage current at the crystal grain boundary increases. In particular, when the thickness of the crystal grain boundary portion is extremely small, the leakage current increases so that insulation between the upper high-concentration semiconductor layer and the current collecting electrode and the lower high-concentration semiconductor layer cannot be maintained. Conversely, when the thickness of the crystal grain portion is extremely large, recombination due to defects in the crystal grain portion increases in the course of the charge running in the crystal grain portion. In the case of 0.85 or more, the traveling distance of the charges generated in the vicinity of the crystal grain boundaries becomes longer, and recombination due to defects in the crystal grain boundaries increases.

In the present invention, the thickness (t1) of the film in the vicinity of the grain boundary is referred to as the thickness (t2) of the central portion of the crystal grain. A film thickness measuring device is simple and preferable. Such devices include devices such as Alpha Step (product name) and Tali Step (product name). The crystal grain portion and the crystal grain boundary portion are each measured over a length of about 1 mm with reference to the substrate portion on which no crystal is grown, and the average of each portion is obtained. At that time, it is preferable to measure each of the plurality of points (for example, average 10 points). Further, the thickness (t2) of the central part of the crystal grain of the present invention refers to the thickness of the substrate as shown in FIG.
The thickness of the central part of the crystal grain (part parallel to the substrate),
In the case where a part near the grain boundary is raised, the thickness of the raised part is considered.

The details of an example of a method of adjusting the thickness of the central portion of the crystal grain and the thickness of the grain boundary portion to a desired ratio are as follows, for example.

In general, when growth is performed while maintaining a thermal equilibrium state in which deposition and meltback coexist, a flat film tends to be formed. Conversely, the more the thermal equilibrium deviates from the thermal equilibrium, the more likely the difference between the thickness at the center of the crystal grain and the thickness at the grain boundary is. Therefore, the difference tends to increase as the temperature drop rate during liquid phase growth increases.
Also, the higher the degree of supersaturation before liquid phase growth, the more likely the difference will be.
In addition, for example, by circulating a metal solvent and always bringing a solvent having a high degree of supersaturation into contact with the substrate, meltback can be suppressed, and a difference between the thickness of the central portion of the crystal grain and the thickness of the grain boundary portion can be provided.

When the difference is too large, the thickness of the central portion of the crystal grains and the thickness of the grain boundary portion can be adjusted to a desired ratio by introducing a step of melt back.

In the present invention, the effect of the present invention can be obtained by controlling the area of the current collecting electrode per unit area (= the area of the grain boundary portion) to a specific range (1% or more and 20% or less). Can be further increased.

That is, the grain boundary portion has many defects as described above. Therefore, even if the current collecting electrode is embedded in the grain boundary portion, if the ratio of the current collecting electrode (= grain boundary portion) to the whole becomes larger than a certain level, the current collecting electrode is recombined at the grain boundary portion, and as a result, the photocurrent is taken out. In some cases, the current that cannot be performed increases, and the photoelectric conversion efficiency decreases.

The inventors of the present invention have studied various kinds of materials by paying attention to this point, and found that the area of the current collecting electrode (= grain boundary portion) per unit area is 20% or less, more preferably 1% or less.
It was confirmed that the effect of the present invention was even more remarkable when the content was 0% or less, optimally 8% or less. The lower limit is not particularly limited (a value of 0 is a single crystal, which is out of the scope of the present invention). However, the lower the value, the higher the cost. This is the optimum range in view of the surface and characteristics.

If the area of the current collecting electrode (= grain boundary portion) per unit area exceeds 20%, the current that can be taken out as a photocurrent is significantly reduced, so that the effects of the present invention can be sufficiently exhibited. May not be possible.

As a polycrystalline silicon substrate, 10 p
By using a metal-grade silicon substrate containing impurities of pm or more, solar cell grade (impurity concentration 10 ppm)
) Of silicon can be significantly reduced as compared with ordinary polycrystalline wafer solar cells.

In addition, by performing liquid phase growth by bringing indium into contact with the surface of the metal-grade silicon substrate, it is possible to form a silicon layer suitable for a solar cell with less impurities.

Before the liquid phase growth of indium, the substrate surface layer is dissolved with a metal solvent and then reprecipitated, so that most of the impurities can be removed from the surface layer by the segregation effect. Therefore, a higher quality silicon layer can be formed by liquid phase growth on this. At this time, the re-deposited layer contains impurities contained in the substrate,
Since boron B is likely to remain, the re-deposited silicon layer becomes p-type (p ⁺ ). This p-type (p ⁺ ) layer can be used as a BSF (Back Surface Field) layer when manufacturing a solar cell.

When a metal paste is used for an electrode for taking out the generated current embedded in the groove, the process of embedding the electrode in the groove is easy and the production cost can be reduced.

As a metal paste, Ag,
When a mixture of Al, Cu, Cr, and Ni in a polymer is used, the conductivity is high and the loss at the time of current extraction due to electrode resistance can be reduced.

When a mixed metal powder of a flaky metal powder of 0.3 to 0.5 μm and a spherical metal powder of 0.1 to 0.2 μm is used as the metal powder, the workability as a paste becomes poor. Good and high filling when dry provides high conductivity.

The electrode for taking out the generated current embedded in the groove is composed of a metal paste, a conductive powder, and a metal or alloy or a mixture (low-melting metal) layer having a lower melting point than the conductive powder. When the conductive powder is fixed on the semiconductor layer via the metal paste and the low melting point metal layer is formed thereon, the conversion efficiency of the solar cell is further improved.

That is, before hardening a metal paste such as a silver paste, a conductive powder is included in a binder of the metal paste and hardened, and a low-melting metal or alloy such as a solder material or an alloy or a mixture is selectively fixed thereon. By doing so, the series resistance can be further reduced. Therefore, the conversion efficiency of the solar cell is further improved.

High conductivity can be obtained by using, as the conductive powder, one made of at least one of Ni, Ag, Cu or an alloy thereof.

[0064]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, based on FIGS. 1 and 2 showing the configuration of a solar cell according to an embodiment of the present invention, FIG.
The present invention will be described in more detail.

FIG. 1 shows one structural example of a solar cell according to an embodiment of the present invention.

In FIG. 1, reference numeral 11 denotes a metal-grade polycrystalline silicon wafer. On top of this, a solar cell grade silicon polycrystalline layer 12 is formed reflecting the crystal structure of 11 as it is. Reference numeral 16 denotes a crystal grain boundary of the 11th metal-grade polycrystalline silicon wafer, which is inherited by 12 solar cell grade silicon polycrystalline layers.

The metal-grade polycrystalline silicon wafer of No. 11 is generally a strong p-type because it contains a large amount of B as an impurity. The resistivity is 0.1-0.001Ω · cm
The degree is a typical value. Depending on the impurities contained,
In some cases, it indicates n-type. In some cases, the conductivity type is not clear. In this case, it is preferable to use a solar cell having the structure shown in FIG.

Therefore, in order to prevent inadvertent bonding between 11 and 12, when the metal-grade polycrystalline silicon wafer 11 is p-type, the solar cell grade silicon polycrystalline layer 12 is also made p-type. It is necessary to keep. Therefore, it is necessary to dissolve a group III element slightly together with silicon in the metal solvent at the time of preparing the silicon polycrystalline layer.

The eleven metal-grade polycrystalline silicon wafers were N
In the case of the type, the twelve solar cell grade silicon polycrystalline layers also need to be N-type. Therefore 12
It is necessary to dissolve a group V element together with silicon in the metal solvent at the time of preparing the silicon polycrystalline layer.

The thickness of the solar cell grade silicon polycrystalline layer 12 is preferably 10 to 200 μm, more preferably 20 to 200 μm. Considering the absorption of sunlight, 10 μm or more is preferable, and therefore a large generated current can be obtained. On the other hand, if it is too thick, the manufacturing cost increases, which is not preferable. The preferred film thickness is in the above range.

The crystal growth method used in the method for manufacturing the twelve solar cell grade silicon polycrystalline layers includes a liquid phase growth method, an LPCVD method, a normal pressure CVD method, a plasma CVD method, a photo CVD method, and a sputtering method. A liquid phase growth method is preferably used from the viewpoint of growth rate and crystallinity.

Reference numeral 13 denotes a semiconductor layer for forming a pn junction with 12, which is a layer of the opposite type to 12. When 12 is P-type, 13 is N-type, and when 12 is N-type, 13 is P-type.

The method for forming the semiconductor layer 13 is as follows:
In this method, impurities having a polarity opposite to that of 12 are introduced into the surface of the layer by ion implantation or thermal diffusion. As impurities, P, As, Sb, etc. for n-type, and B, P-type for p-type.
Al, Ga, etc. are each suitably selected.

In forming a semiconductor junction, a semiconductor layer 13 having a conductivity type different from that of polycrystalline Si may be deposited on the surface of polycrystalline Si by a liquid phase growth method, various CVD methods, or the like. In this case, the crystal structure of the semiconductor layer 13 may or may not inherit the crystal structure of the twelve polycrystalline Si layers. If not succeeded, a film of microcrystalline silicon, amorphous silicon carbide or the like can be used.

The junction depth or the thickness of the semiconductor layer depends on the amount of impurities to be introduced.
It is preferably in the range of 1 μm, more preferably 0.1 μm.
02 to 0.5 μm. In order to form a uniform junction, a thickness of the semiconductor layer of 0.01 μm or more is required.
On the other hand, if it is too thick, light absorption in this region cannot be ignored,
Power generation efficiency decreases.

Reference numeral 14 denotes a current collecting electrode formed in the groove at the grain boundary. The collecting electrode is formed by coating a metal paste.

The collecting electrode 14 is made of a thermosetting resin such as polyester, polyimide, epoxy or phenol, or a glass frit, and is made of a conductive material such as silver, gold, copper, or Al having a diameter of about 0.1 to 1 μm. High-dispersion metal particles are monodispersed, and a conductive paste material mixed with an organic solvent such as Cellsolve is used to adjust the viscosity using rubber or a soft and elastic plastic blade.
It is buried in the groove above the semiconductor layer 13 and then cured by heating.

The thickness of the current collecting electrode is determined by the depth of the groove. The depth of the groove may be appropriately designed as long as necessary power can be taken out. When it is necessary to increase the thickness of the current collecting electrode beyond the depth of the groove, solder or the like may be provided thereon. Before the conductive paste dries, metal powder having good wettability with respect to the solder falls on the conductive paste, and the metal powder accumulated on portions other than the conductive paste is blown off by blowing air or the like. Thereafter, when the conductive paste is heated and dried, the silicon surface has poor wettability with respect to the solder. Therefore, when the sample is immersed in a solder solution, the solder is formed only on the embedded metal.

Reference numeral 15 denotes an antireflection film, which is made of TiO ₂ , Sn
A film of O ₂ , In ₂ O ₃ , ZnO or the like is formed by a sputtering method or the like.

The film thickness is determined in consideration of the refractive index of silicon and the antireflection film at a wavelength at which the photocurrent is maximized so that the antireflection effect is maximized.

FIG. 2 shows another structural example of the present invention. 2 correspond to 11 to 16 in FIG. 2
Reference numeral 7 denotes a highly doped layer introduced between 21 metal-grade polycrystalline silicon wafers and 22 solar cell grade silicon polycrystalline layers. The polarity is the same as 22, but the resistivity is low. Usually, it is 0.1 Ω · cm or less. This structure is 2
When the polarity of the first metal-grade polycrystalline silicon wafer is not clear, it is effective to make ohmic contact with 22 solar cell grade silicon polycrystalline layers.

An example of a manufacturing process of the solar cell having the structure shown in FIG. 1 will be described below. Granular metal-grade silicon was charged into a quartz crucible. The metal-grade silicon was melted by placing the crucible in a vacuum induction heating furnace and maintaining the temperature above the melting point of silicon (〜1415 ° C.) for a certain period of time. Next, an ingot of metal-grade silicon was produced by lowering the temperature and solidifying. The obtained metal-grade silicon ingot is 350 m thick.
m to produce a metal-grade silicon wafer. After cleaning the metal-grade silicon wafer with a detergent and an organic solvent, the wafer was etched with a 1.5% HE aqueous solution for 1 minute to remove an oxide film on the surface. The obtained metal-grade silicon wafer was placed in a boat made of carbon graphite or the like, and a silicon layer was deposited on the substrate by a liquid phase growth method using indium in a hydrogen atmosphere. A bond was formed on the surface of the silicon layer. A current collecting electrode pattern was formed thereon. An anti-reflection film was formed thereon, and a solar cell having the structure shown in FIG. 1 was manufactured.

An example of a method for manufacturing a solar cell having the structure shown in FIG. 2 according to another embodiment of the present invention is as follows. Granular metal-grade silicon was charged into a quartz crucible. The metal-grade silicon was melted by placing the crucible in a vacuum induction heating furnace and maintaining the temperature above the melting point of silicon (〜1415 ° C.) for a certain period of time. Next, an ingot of metal-grade silicon was produced by lowering the temperature and solidifying. The obtained metal-grade silicon ingot is 350 m thick.
m to produce a metal-grade silicon wafer. After cleaning the metal-grade silicon wafer with a detergent and an organic solvent, the wafer was etched with a 1.5% HE aqueous solution for 1 minute to remove an oxide film on the surface. The obtained metal-grade silicon wafer is placed in a boat made of carbon graphite or the like, and a metal solvent such as indium is brought into contact with the metal-grade silicon substrate and placed in an electric furnace, and the metal-grade silicon substrate is placed in the electric furnace. The surface layer was dissolved in the solvent. Then, the temperature was lowered to make the concentration of silicon in the solvent saturated or supersaturated, and silicon was deposited again on the surface of the metal-grade silicon substrate. At this time, the deposited silicon became p-type (p ⁺ ). The obtained metal-grade silicon substrate was placed in a boat made of carbon graphite or the like, and a silicon layer was deposited on the substrate by a liquid phase growth method using indium in a hydrogen atmosphere. A bond was formed on the surface of the silicon layer. A current collecting electrode pattern was formed thereon. An antireflection film was formed thereon, and a solar cell having the structure shown in FIG. 2 was manufactured.

Hereinafter, an example of a method for manufacturing a solar cell according to the present invention will be described in detail.

The liquid phase growth method used in the present invention is performed in a hydrogen atmosphere for the purpose of removing a natural oxide film present on the surface of a silicon substrate. The growth temperature is preferably in the range of 500 to 1100C, more preferably in the range of 700 to 1050C.

The indium used in the present invention is 9
High purity of 9.9% to 99.9999% is preferable.

As the liquid phase growth method used in the present invention, a slow cooling method or a temperature difference method is preferable, but a constant temperature method (JP-A-6-191987) by the present inventors can also be used.

As the metal-grade silicon used for the solar cell substrate of the present invention, low-grade, specifically, those containing 0.1% to 2% of an impurity element are inexpensive and easily used.
It is also possible to reduce the amount of impurities by performing a treatment with an acid such as hydrochloric acid in advance as needed before granulation or powdering and before melting.

As the material of the crucible used in the present invention, quartz or carbon graphite is used from the viewpoint of easiness of processing and cost, but a material capable of releasing the molten / solidified silicon from the mold can be applied and the melting point can be reduced. Anything higher than that of silicon can be used, and silicon carbide, silicon nitride, boron nitride, or the like can be used. In addition, the crucible is formed asymmetrically or by attaching a heat sink to control the heat flow during solidification, thereby segregating impurities contained in metal-grade silicon to one side of the sheet surface and increasing the crystal grain size. It is also possible to make it.

The release agent applied to the crucible used in the present invention is selected from those which do not react with molten silicon and have a large contact angle. Specifically, a material containing Si ₃ N ₄ as a main component is used, and SiO ₂ or the like is added as necessary. As a method of coating the release agent into the crucible, an organic solution or a silanol solution in which powdered Si ₃ N ₄ is dispersed is sprayed into the crucible, and a heat treatment at 400 ° C. or more is performed to form a coating.

As the metal soot used for dissolving / reprecipitating the surface of the metal-grade silicon substrate used in the present invention, one having a relatively low melting point and sufficiently dissolving the surface layer of the silicon substrate is selected. As such a material, for example, indium, gallium, tin and the like are preferable.

As the boat used in the present invention for the liquid phase growth with indium and the boat for dissolving / reprecipitating the surface of a metal-grade silicon wafer with a metal solvent, mainly carbon graphite is used.
Silicon carbide, silicon nitride and the like can also be used.
As a method for bringing the metal solvent into contact with the surface of the silicon wafer, a slide method or a tipping method is mainly used.

In the present invention, a vacuum induction heating furnace is preferably used as a furnace used for manufacturing a metal-grade silicon ingot from the viewpoint of controllability. -30 ° C. from the viewpoint of maintaining the crystallinity of the ingot formed
Those capable of lowering the temperature at a rate of not more than / min are preferred. The furnace used for dissolving / reprecipitating the surface of a metal-grade silicon wafer substrate with a metal solvent is an ordinary electric furnace, and the specifications thereof conform to those of the above vacuum induction heating furnace. Also in this case, an induction heating furnace may be used. FIG.
Is a cross-sectional view of only the solar cell grade silicon layer.
In the present invention, the thickness t1 of the crystal grain portion and the thickness t2 of the crystal grain boundary portion
And the ratio is defined.

Hereinafter, an experiment conducted by the present inventors to obtain the above-described solar cell will be described in detail.

(Experiment 1) In this experiment, a metal-grade silicon polycrystal substrate was placed in a boat made of carbon graphite, and a supercooling degree of 30 ° C. and a film formation starting temperature of 980 ° C. were set in a hydrogen atmosphere in a metal solvent. Was changed, and the film thickness of the crystal grain portion and the crystal grain boundary portion of the silicon layer grown in the liquid phase was measured. The descending speed is 1 ℃ / min, 2 ℃
/ Min, 3 ° C / min, 4 ° C / min, 5 ° C / mi
n, 6 ° C./min, 7 ° C./min. These samples are referred to as samples 1 to 7, respectively. As a result, the ratio (t2 / t1) of the film thickness between the crystal grain boundary and the crystal grain in each sample was as shown in Table 1.

[0096]

[Table 1]

(Experiment 2) In this experiment, a method of forming a p + silicon layer with less impurities on a metal-grade silicon wafer was examined. The metal-grade silicon wafer is placed in a carbon boat, the metal solvent of indium is brought into contact with the sheet, placed in an electric furnace, and maintained at 1050 ° C. to dissolve the surface layer of the metal-grade silicon wafer in the indium solvent. Was. After maintaining this state for a while and sufficiently saturating, the temperature was lowered by controlling the electric furnace, and silicon in the solvent was deposited again on the surface of the silicon wafer. After the deposition for a certain period of time, the indium solvent brought into contact with the wafer was removed to obtain a desired silicon deposition layer.

Table 2 shows the results of analysis of the elements contained in the silicon layer deposited on the surface of the obtained silicon wafer.

[Table 2]

From Table 2, it was confirmed that the amount of impurities in the deposited silicon layer was further reduced to 1/100 or less as compared with the metal-grade silicon.

The pn determination by the thermoelectromotive force method revealed that the deposited silicon layer was p-type (p ⁺ ).

From the results of the above-described experiments, a high-quality silicon layer can be obtained by forming a silicon layer on a crystalline silicon substrate having an impurity concentration of 10 ppm or more by a liquid phase growth method using indium as a solvent. Further, a silicon wafer is produced by melting / solidifying granular metal-grade silicon, and the surface thereof is dissolved / reprecipitated with a metal solvent to form a p-type (p ⁺ ) silicon layer, and then indium is deposited thereon. It was found that a silicon layer can be formed by a liquid phase growth method using a solvent. The thickness of the silicon layer manufactured in the present invention is preferably 1 from the viewpoint of efficient light absorption.
It is desirable that the thickness be 0 μm or more, more preferably 10 μm or more and 100 μm or less, and still more preferably 10 μm or more and 50 μm or less.

[0102]

EXAMPLES Hereinafter, the solar cell according to the present invention will be described in more detail, but the present invention is not limited to these examples.

Example 1 In this example, a silicon layer was deposited on a metal-grade silicon wafer sliced from an ingot by a liquid phase growth method using indium, and a thin-film solar cell was formed using the silicon layer as an active layer.

An ingot was prepared by casting using metal-grade silicon having a purity of 98% as a raw material, and sliced into a wafer having a thickness of 0.5 mm to prepare a metal-grade silicon wafer substrate. Elemental analysis near the surface of the produced metal-grade silicon wafer substrate was performed, and the results shown in Table 3 were obtained.

[0105]

[Table 3]

The metal-grade silicon wafer substrate has a crystal grain size of several mm to several cm and a specific resistance of 0.01 Ω · c.
The m-grade (p-type) metal-grade silicon wafer substrate was placed in a carbon boat, and high-purity silicon was dissolved in an indium solvent at 980 ° C. for 1 hour and 40 minutes before film formation. Thereafter, the wafer is placed in a solvent, and a liquid crystal growth method using indium in which high-purity silicon is dissolved is used as a solvent. In a hydrogen atmosphere, a degree of supercooling is 10 ° C., a growth start temperature is 950 ° C., and a descent rate is 4 ° C./min. To form a silicon layer until the temperature drops to 800 ° C.

When the film thickness of the formed silicon layer was measured, the center of the crystal grain was 51 μm and the grain boundary was 28 μm.
m.

Next, P was thermally diffused on the surface of the silicon layer at a temperature of 900 ° C. using POCl ₃ as a diffusion source to form an n ⁺ layer, and a junction depth of about 0.5 μm was obtained. The dead layer on the surface of the formed n ⁺ layer was removed by etching to obtain a junction depth having an appropriate surface concentration of about 0.2 μm.

A silver polymer paste material having a low specific resistance was buried in the grooves at the grain boundaries using a urethane blade. The silver polymer paste used in this example was obtained by mixing a flaky silver powder having a particle size of 0.3 to 0.5 μm and a spherical silver powder having a particle size of 0.1 to 0.2 μm with an epoxy resin using a roll mill device. is there. Here, the particle size is a value obtained by the sedimentation method.

The method of producing the powder is, for example, a method of depositing the powder on the cathode by electrolysis (electrolysis method), a method of pulverizing metal using a pulverizer or a pulverizer such as a ball mill or a crusher, or the like. (Pulverization method), a method in which a molten metal is caused to flow out through a thin nozzle, and the molten metal stream is blown out in a spray form by applying a compressed gas or a water jet to the molten metal stream, and simultaneously cooled (atomizing method). The scaly and spherical silver powders of the present example were each produced by an atomizing method.

Thereafter, the polymer paste was dried at 160 ° C. using an infrared heating and drying oven to harden the polymer paste.

Further, an antireflection film TiO ₂ is further
0 nm was formed by a sputtering method, and Al was also deposited on the back surface of the metal-grade silicon wafer to form a back electrode. (This is referred to as Sample 1.)

In the thin film crystal solar cell thus manufactured, the ratio of the area of the current collecting electrode embedded in the groove of the grain boundary was about 6% of the light receiving area of the solar cell.

Therefore, a current collecting electrode of a silver paste was formed on the n ⁺ layer by a screen printing method without considering a grain boundary for comparison. (Comparative sample 1) The area ratio of the collecting electrode was set to be about 6% of the light receiving area of the solar cell.
The width and interval of the collecting electrode were adjusted. AM1.5 (10
(0 mW / cm ² ) When the IV characteristics under light irradiation were measured, the conversion efficiency of the solar cell of Sample 1 was about 5% higher than the conversion efficiency of the solar cell of Comparative Sample 1.

(Example 2) In Example 1, after the silver paste was buried in the groove of the grain boundary to form a current collecting electrode,
Next, before the silver paste was cured by heating, Ni particles having a particle size of 3 μm to 7 μm produced by an atomizing method were sprayed on the polymer paste by a powder spray method using compressed air to be topped. At this time, the semiconductor surface was sufficiently dried. The sprayed Ni particles adhered not only on the polymer paste but also on the semiconductor surface.
In this embodiment, Ni particles are sprayed on the paste by a powder spray method using compressed air, but it goes without saying that the paste may be applied by a micro dispenser, for example.

Then, an air stream was blown to blow out and remove Ni particles other than those on the polymer paste. afterwards,
The polymer paste was dried at 160 ° C. using an infrared heating drying oven to cure the polymer paste. While heating, the polymer paste shrinked and entrained Ni particles while shrinking. The surface of the paste hardens in a form in which Ni particles having a relatively large particle diameter are wrapped in fine silver particles, and the Ni particles serve as a base for forming a solder layer.

Next, the flux 36 was applied to the entire surface on which the electrodes were formed by a spray method. As a method of applying the flux, in addition to the spray method, application by a brush, immersion in a flux solution, a jet method, a dropping method, and the like can be used.

Further, a solder layer was formed on the surface of the electrode on which the flux was applied, using a jet solder bath. Here, a molten solder flow of Sn60% Pb38% Ag2% flowing in one direction was created, and this solder flow was brought into contact with the electrode forming surface. The photovoltaic element itself is also heated before and after contact with the solder flow, so that the solder hardens and does not adhere to portions other than the electrodes. The time for passing through the solder flow was about 2 to 30 seconds. The application of the flux not only promotes the wetting of the solder with the polymer paste and the metal particles, but also can further prevent the solder from adhering to the electrode forming surface without the polymer paste.

The remaining flux was removed by washing with shower water using warm water. After the completion of the washing, the substrate was dried to complete the production of the current collecting electrode.

Further, an anti-reflection film TiO ₂ is further
0 nm was formed by a sputtering method, and Al was also deposited on the back surface of the metal-grade silicon wafer to form a back electrode. (This is referred to as Sample 2.)

The thin-film crystal solar cell (sample 2) manufactured in this manner had an AM of 1.5 (100 mW / cm ² ).
When the IV characteristics under light irradiation were measured,
The conversion efficiency of the solar cell of Sample 2 was about 8% higher than the conversion efficiency of the solar cell of Comparative Sample 1.

(Example 3) In Example 1, after forming a silicon layer under the conditions of Experiment 1, a silver paste was embedded in each groove under the same conditions as in Example 1, and then under the same conditions as in Example 1. A solar cell was formed. Each sample was transferred to AM1.5 (1
(00 mW / cm ² ) When the IV characteristics under light irradiation were measured, as shown in FIG. 4, the conversion efficiency varied depending on the depth of the groove.

The reason why the conversion efficiency is low at the shallow groove is that the resistance of the current collecting electrode has increased. The reason why the conversion efficiency is low in the deep part of the groove is that the leak current at the grain boundary part has increased.

Example 4 A metal-grade silicon wafer was placed in a carbon boat, and a metal solvent of indium was brought into contact with the sheet and placed in an electric furnace. The layer was dissolved in the In solvent. In this state, when several hours have passed and saturation is sufficient, the electric furnace is controlled to lower the temperature at a rate of 0.5 ° C./min, and silicon in the solvent is deposited again on the silicon wafer surface to 700 ° C. Was. After deposition for 2 hours, the boat was slid to remove the indium solvent on the wafer, and a desired silicon deposition layer was obtained.

Table 4 shows the results of analysis of elements contained in the silicon layer deposited on the surface of the metal-grade silicon wafer.

[0126]

[Table 4]

From Table 4, it was confirmed that the amount of impurities in the deposited silicon layer was significantly reduced as compared with the granular metal-grade silicon as the raw material. Further, the thickness of the obtained reprecipitated silicon layer was about 15 μm from a cross-sectional observation by SBM / EDX.

The pn judgment by the thermoelectromotive force method revealed that the deposited silicon layer was p-type (p ⁺ ).

On the surface of the silicon wafer, that is, on the deposited silicon layer, liquid phase growth was performed using indium other than the indium metal solvent used in the above-described dissolution / reprecipitation step as a solvent. Before film formation, high-purity silicon was dissolved in an indium solvent at 980 ° C. for 1 hour and 40 minutes. Thereafter, the wafer is placed in a solvent, and in a hydrogen atmosphere, the degree of supercooling is 30 ° C., and the growth start temperature is 950 ° C.
The silicon layer was formed at a temperature lowering rate of 4 ° C./min.
The thickness of the obtained silicon layer is 53 μm at the center of the crystal grain.
m, 30 μm at the grain boundary.

The surface of the silicon layer was subjected to thermal diffusion of P at a temperature of 900 ° C. using POCl ₃ as a diffusion source to form an n ⁺ layer, and a junction depth of about 0.5 μm was obtained. Formed n ⁺
The dead layer on the layer surface is removed by etching,
A junction depth with an appropriate surface concentration of μm was obtained.

A silver polymer paste material having a low specific resistance value was formed on the n ⁺ layer by using a urethane blade.
The grain boundary was embedded only in the groove. The silver polymer paste used in the present example was prepared by using a roll mill device to obtain a particle size of 0.3.
A mixture of epoxy resin and flaky silver powder of about 0.5 μm and spherical silver powder of 0.1 to 0.2 μm. here,
The particle size is a value determined by the sedimentation method.

The powder may be produced by, for example, a method of depositing the powder on a cathode by electrolysis (electrolysis method), or a method of pulverizing a metal by using a pulverizer or a crusher such as a ball mill or a crusher. (Pulverization method), a method of producing a molten metal by flowing it out through a fine nozzle, blowing the molten metal stream against a compressed gas or a water jet to blow it out in a spray form, and cooling it at the same time (atomizing method). The scaly and spherical silver powders of the present example were each produced by an atomizing method.

Thereafter, the polymer paste was dried at 160 ° C. using an infrared heating and drying oven to harden the polymer paste.

Further, an antireflection film TiO ₂ is further
0 nm was formed by a sputtering method, and Al was also deposited on the back surface of the metal-grade silicon wafer to form a back electrode.

In the thin film crystal solar cell manufactured in this manner, the ratio of the area of the current collecting electrode embedded in the groove of the grain boundary was about 6% of the light receiving area of the solar cell. (This is referred to as Sample 3.)

Therefore, a current collecting electrode of a silver paste was formed on the n ⁺ layer by a screen printing method without considering a grain boundary for comparison. The width and interval of the collecting electrodes were adjusted so that the area ratio of the collecting electrodes was about 6% of the light receiving area of the solar cell. (Comparative sample 2) AM1.5 (10
(0 mW / cm ² ) When the IV characteristics under light irradiation were measured, the conversion efficiency of the solar cell of Sample 3 was about 6% higher than the conversion efficiency of the solar cell of Comparative Sample 2.

(Example 5) In Example 1, after the silver paste was buried in the groove of the grain boundary to form a current collecting electrode,
Next, before the silver paste was cured by heating, Ni particles having a particle size of 3 μm to 7 μm produced by an atomizing method were sprayed on the polymer paste by a powder spray method using compressed air to be topped.

At this time, the semiconductor surface was sufficiently dried. The sprayed Ni particles adhered not only on the polymer paste but also on the semiconductor surface. In this embodiment, Ni particles are sprayed on the paste by a powder spray method using compressed air, but it goes without saying that the paste may be applied by a micro dispenser, for example.

Then, an air stream was blown to blow and remove Ni particles other than those on the polymer paste. afterwards,
The polymer paste was dried at 160 ° C. using an infrared heating drying oven to cure the polymer paste. While heating, the polymer paste shrinked and entrained Ni particles while shrinking. The surface of the paste hardens in a form in which Ni particles having a relatively large particle diameter are wrapped in fine silver particles, and the Ni particles serve as a base for forming a solder layer.

Next, a flux 36 was applied to the entire surface on which the electrodes were formed by a spray method. As a method of applying the flux, in addition to the spray method, application by a brush, immersion in a flux solution, a jet method, a dropping method, and the like can be used.

Further, a solder layer was formed on the electrode forming surface to which the flux was applied, using a jet solder bath. Here, a molten solder flow of Sn60% Pb38% Ag2% flowing in one direction was created, and this solder flow was brought into contact with the electrode forming surface. The photovoltaic element itself is also heated before and after contact with the solder flow, so that the solder hardens and does not adhere to portions other than the electrodes. The time for passing through the solder flow was about 2 to 30 seconds. The application of the flux not only promotes the wetting of the solder with the polymer paste and the metal particles, but also can further prevent the solder from adhering to the electrode forming surface without the polymer paste.

The remaining flux was removed by washing with hot water using a shower. After the completion of the washing, the substrate was dried to complete the production of the current collecting electrode.

Further, an anti-reflection film TiO ₂ is further
0 nm was formed by a sputtering method, and Al was also deposited on the back surface of the metal-grade silicon wafer to form a back electrode. (This is referred to as Sample 4.)

The thin-film crystal solar cell (sample 4) manufactured in this manner had an AM of 1.5 (100 mW / cm ² ).
When the IV characteristics under light irradiation were measured,
The conversion efficiency of the solar cell of Sample 4 was about 9% higher than the conversion efficiency of the solar cell of Comparative Sample 2.

Example 5 In this example, a current-collecting electrode was produced in the same manner as in Example 2, except that silver having a particle size of 3 to 10 μm was used as the conductive powder. Other conditions are the same as those in Example 2.
When a solar cell was produced in the same manner as in Example 1, excellent characteristics similar to those of Example 2 were obtained.

(Example 6) In this example, a current collecting electrode was produced in the same manner as in Example 2 except that copper having a particle size of 3 to 6 µm was used as the conductive powder. The other conditions were the same as in Example 2, and a solar cell was produced. As a result, the same excellent characteristics as in Example 2 were obtained.

Example 7 In this example, a current-collecting electrode was produced in the same manner as in Example 4, except that silver having a particle size of 3 to 10 μm was used as the conductive powder. The other conditions were the same as in Example 4.
When a solar cell was manufactured in the same manner as in Example 1, excellent characteristics similar to those of Example 4 were obtained.

Example 8 In this example, a current collecting electrode was produced in the same manner as in Example 4, except that copper having a particle size of 3 to 6 μm was used as the conductive powder. Other conditions were the same as in Example 4, and a solar cell was fabricated. As a result, excellent characteristics similar to those of Example 4 were obtained.

[0149]

As described above, according to the present invention, since the current collecting electrode is limited to the vicinity of the grain boundary where power generation efficiency is poor, light caused by the presence of the current collecting electrode causes light on the crystal grains. Not irradiating can reduce the decline in night power generation efficiency,
Conversion efficiency can be improved.

Further, before the metal paste is hardened, the conductive powder is contained in a binder of the metal paste and hardened, and a low-melting metal or alloy or mixture such as a solder material is selectively fixed thereon. The series resistance can be further reduced.

Further, the amount of silicon of the solar cell grade to be used can be significantly reduced as compared with a normal polycrystalline wafer solar cell.

In addition, by performing liquid phase growth by bringing indium into contact with the surface of the metal-grade silicon substrate, it is possible to form a silicon layer suitable for a solar cell with less impurities.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram showing a configuration example of a solar cell of the present invention.

FIG. 2 is a conceptual diagram showing another configuration example of the solar cell of the present invention.

FIG. 3 is a schematic view of a cross section of a polycrystalline silicon layer of the present invention.

FIG. 4 is a diagram showing the relationship between the conversion efficiency and the ratio of the film thickness between the crystal grain boundary and the crystal grain in the present invention.

[Explanation of symbols]

11,21 metal grade polycrystalline silicon wafer 12,22 solar cell grade polycrystalline silicon layer 13,23 12 and 22 silicon layers and p, respectively
Semiconductor layer for forming n-junction 14, 25 Current collecting electrode 15, 25 Antireflection film 16, 26 Crystal grain boundary 27 Highly doped layer

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Akishi Nishida 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc. (72) Inventor Hiroshi Sato 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon F term (reference) 5F051 AA03 CB04 DA03 DA20 FA14 FA30 GA04 GA20 HA03

Claims (16)

[Claims] 1. A polycrystalline solar cell comprising an active layer and a semiconductor layer having the opposite polarity to an active layer on a polycrystalline silicon substrate, wherein the active layer is grown by reflecting the crystal structure of the substrate. The thickness at the grain boundary of the polycrystal forming the active layer is smaller than the thickness at the center of the polycrystal grain, and the grain boundary becomes a groove, so that the grain boundary is formed on the grain boundary. A polycrystalline solar cell, wherein a groove is formed in a semiconductor layer, and an electrode for extracting a generated current is embedded in the groove. 2. The polycrystalline solar cell according to claim 1, wherein said active layer is an active layer grown by liquid phase. 3. A film according to claim 1, wherein the thickness of the polycrystalline grain forming the active layer at the grain boundary portion is in the range of 0.20 to 0.85 with respect to the thickness of the central portion of the polycrystalline grain. A polycrystalline solar cell, characterized in that: 4. A film according to claim 1, wherein the film thickness of the polycrystal forming the active layer at the grain boundary is 0.25 or more and 0.75 or less with respect to the film thickness at the center of the polycrystal. A polycrystalline solar cell, characterized in that: 5. The polycrystalline solar cell according to claim 1, wherein the area of the electrode portion per unit area is 1% or more and 20% or less. 6. A polycrystalline solar cell according to claim 1, wherein said polycrystalline silicon substrate is metal-grade silicon containing 10 ppm or more of impurities. 7. The polycrystalline solar cell according to claim 1, wherein the electrode for extracting a generated current embedded in the groove is a metal paste. 8. The polycrystalline solar cell according to claim 7, wherein the metal paste is a mixture of Ag, Al, Cu, Cr, and Ni powder in a polymer. 9. The method according to claim 8, wherein the metal powder contains 0.1.
3 ~ 0.5μm scaly metal powder and 0.1 ~ 0.2μm
A polycrystalline solar cell characterized by being a mixed metal powder with a spherical metal powder. 10. The electrode according to claim 1, wherein the electrode for extracting the generated current embedded in the groove is formed of a metal paste, a conductive powder, and a metal or alloy or a mixture (melting point) lower than the conductive powder. A low melting point metal) layer, wherein the conductive powder is fixed on the semiconductor layer via the metal paste, and the low melting point metal layer is formed thereon. Solar cells. 11. The polycrystalline solar cell according to claim 10, wherein the conductive powder is made of at least one of Ni, Ag, Cu, and an alloy thereof. 12. A method for manufacturing a polycrystalline solar cell comprising an active layer grown in a liquid phase on a polycrystalline silicon substrate and a semiconductor layer having the opposite polarity to the active layer, wherein the active layer is formed. A step of forming the grain boundary portion in a groove shape in which the thickness at the grain boundary portion of the polycrystal to be formed is smaller than the thickness of the central portion of the polycrystal grain; A method for producing a polycrystalline solar cell, comprising: 13. The method for manufacturing a polycrystalline solar cell according to claim 12, wherein the electrode for extracting a generated current embedded in the groove is a metal paste. 14. An electrode according to claim 11, wherein the electrode for extracting a generated current embedded in the groove is made of a metal paste, a conductive powder, and a metal, an alloy or a mixture having a lower melting point than the conductive powder. And a step of fixing the conductive powder on the semiconductor layer via the metal paste, and a step of forming the low melting point metal layer thereon. Solar cell manufacturing method. 15. The method for manufacturing a polycrystalline solar cell according to claim 12, wherein the polycrystalline silicon substrate is a metal-grade silicon substrate containing 10 ppm or more of impurities. 16. A method for manufacturing a polycrystalline solar cell according to claim 15, wherein indium is brought into contact with the surface of the metal-grade silicon substrate to perform liquid phase growth.