JPS5996780A - Photoelectric conversion device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a photoelectric conversion element (also simply called an element) in which a non-single-crystal semiconductor including an amorphous semiconductor having at least one P:l or PN junction is provided on a light-transmitting constant R substrate. It relates to a photoelectric conversion device that can generate high voltage by electrically connecting a plurality of (1) in series.

The present invention is characterized in that a laser scribing method is used to reduce the area required for connecting a plurality of elements to 1/10 to 1/'100K of the conventional mask alignment method.

This invention creates an opening for separating the second electrodes of the first and second elements by cutting a P- or N-type semiconductor layer of a semiconductor having a PM'i: or PIN junction under the electrode. By providing this, there will be no leakage current from the second electrode of the second element to the second electrode of the first element.
The purpose of this service is to

For this reason, the groove that separates the second electrodes of the first and second elements is provided over the semiconductor of the first element, thereby providing redundancy to improve manufacturing yield. It is characterized by

In this invention, the electrical connection at the connecting portion is
By extending the second electrode of the second element to the surface or side surface of the transparent isoelectric film (OTF) constituting the electrode, the area required for the connection part can be reduced. It is characterized by a decrease in

For this reason, the groove separating the semiconductors of the first and second elements is provided over the ↓ electrode position of the first element to provide redundancy (margin) in manufacturing.

Conventionally, in the mask matching method, the connecting part is 5 to 1 m long.
It required a width of 1/10 to 1/2 m.
100 350 to 30 μm, preferably 200 to 50 μm, in a hybrid system that requires 10 to 50 stages of this connection, the area for photovoltaic generation of all panels used as photoelectric conversion devices (effective (referred to as area or effective area) is 97-9% compared to conventional '75-50%.
0%, and is characterized by substantially improving the effective conversion efficiency by 10 to 20%.

In this invention, by using a laser beam scribing method, and by pre-memorizing the address to be scribed in the memory of a computer (microcomputer) using the alignment mark as a reference, It is characterized by eliminating at once all the causes of price increase, low stock, and reduction in manufacturing, such as mask misalignment, distortion, and decrease in manufacturing step N/C for matching accuracy required in the mask matching method. Conventionally, many examples of photoelectric conversion devices (hereinafter simply referred to as devices), that is, devices in which a plurality of elements are arranged on the same substrate and are integrated or hybridized, are known.

For example, JP-A-55-4994, JP-A-55-(2427)
4 Furthermore, the patent application No. 54-9009 filed by the present inventor
'7/9009B/90099 (Showa 54.7.16
application) is known. For example, in the patent application filed by the present inventor, the semiconductor is a Si x CI-Bi heterojunction, which is different from the case where other amorphous silicon semiconductors are simply used, and furthermore, as this semiconductor, an amorphous silicon semiconductor is used. In addition to the structure, hydrogen containing a microcrystalline structure or
It is characterized in that it is an integrated or hybridized non-single-crystal semiconductor having at least one PNt or P-G-N junction doped with a rogen element.

However, in these conventional inventions, whose vertical cross-sectional view is shown in FIG. When used, in a method of directly and selectively producing a conductive layer or a semiconductor layer, the mask that provides this selectivity may shift by KO05 to 3 mm during film formation. Furthermore, since coating components are formed on this mask, the mask becomes contaminated, the edges of the coating formed by the mask are not clearly defined, and crosstalk between adjacent electrodes (lease current ) had many drawbacks, such as being a factor in the occurrence of.

Furthermore, in conventionally known screen printing methods, conductors or semiconductors formed entirely on a substrate are independently and selectively etched away using a mask. However, in this method, the process is extremely time consuming, including the process of aligning the mask for screen printing, the process of coating the resist, the process of baking and solidifying, the process of etching the conductor or semiconductor, and the process of removing the resist. The rise was inevitable.

However, in the photoelectric conversion device of the present invention, particularly the thin film type photoelectric conversion device, each of the thin film layers, the conductive layer for the electrode and the semiconductor layer, each has a thickness of 500 to 1 μm, and is laser scribed. It has been found that by using this method, it is possible to fabricate without the need for mask matching at all.

As a result, unlike the conventional mask matching process, the present invention does not use a mask at all, so it is called a scribing process, but this scribing process is extremely simple and highly accurate due to the combined use of a microcomputer. C,
This leads to a reduction in the manufacturing cost of the device, and therefore
It is also possible to manufacture w, and by expanding the scale of production,
It is an object of the present invention to provide an extremely innovative photoelectric conversion device that is capable of converting 0 to 200 Fl/W.

A further advantage of the present invention is that in addition to using this laser scribing process, a degree of redundancy (margin) is provided in the alignment accuracy of the C scribe line.

Therefore, the first electrode (lower side) between adjacent elements and the second electrode (upper electrode) of another element are connected to each other by a lead extending beyond the second electrode. By electrically connecting the sides or the surface, it was possible to provide redundancy in the position of the open grooves of the base and clei lines.

A conventional example and the structure of the present invention will be described below according to the drawings.

FIG. 1 is a vertical sectional view of a conventionally known photoelectric conversion device using a mask alignment method.

In Drawing Q, a transparent conductive film (abbreviated as OTF) constituting the first electrode is selectively formed on a transparent substrate (for example, a glass plate) (1) during the mask alignment process of Atsushi 1. let Further, a semiconductor layer (3) is selectively formed in the same manner as in the second mask alignment step.

Further, a second electrode (4) is provided in a third mask alignment step.

In Fig. 1, there is a connecting part (6) between element α and (31).
In the connection part, one side surface αQ of the CTF is covered by the semiconductor layer (3), and the other side cc! OTF to prevent the semiconductor layer (3) from covering the surface CL◆ of the TF.

The gap 01 requires a trench of 1 to 5 mm, for example 3 mm.

Furthermore, the first electrode (3η) and the second electrode (38) are electrically connected on the surface of α→, but this part is connected to the second electrode (38) of (3g).
The electrodes are also dusty, which is caused by the blurring of the mask.

1 to 5 mm, e.g. 3
A gap (6) of mm is required. Since these three masks have no self-alignment properties, the connecting portion α requires 1 to 8 mm, typically 4 mm. Furthermore, if K is 1 mm or less, the mask alignment accuracy will be extremely strict and the yield will be extremely reduced. If the gap of this connecting part 0 is 5 mm or more, for example, 20 cm x 60 cm width 15 mm (20 cm x 1
5nxm) and an end portion of 5mm, 2
Only 0-stage direct connection is possible. Furthermore, even if the gap at this connecting part is 3 mm, there will be a loss of 8 pieces in 33 stages, and a total loss of 10 cm (area of 200 cm') in the series I part, and as a result, the effective area will remain at 75 pieces when the peripheral area is taken into account. was.

The present invention eliminates such process complexity and reduces the effective area to 86.
It is an object of the present invention to provide an extremely innovative photoelectric conversion device that can be made up to 97 inches, for example, 92 inches.

The details of the embodiment are shown below according to the drawings.

FIG. 2 is a vertical sectional view showing the manufacturing process of the present invention.

In the drawing, a transparent substrate (1), for example, a glass plate (for example, thickness 1.2 mm, length in the left and right direction in the drawing) 600 m,
Using 7'Co (width 20 cm), a translucent conductive film such as TO (approx.
EInO, (200 to 400 phantom or), a luminescent specialist j1% (
-L500 to 2000 Å) was formed by vacuum evaporation ALPOVD method, plasma CvD method, or spray method.

After this, apply a multi-power output of 5 to 8 W to the lower or upper side of this board using a YAG laser processing machine (Nippon Laser Ring).
A spot diameter of 30 to 70 μm, typically 50 μm, was irradiated under the control of a microcomputer and scanned to form an opening groove α1 for the sysclub line, thereby creating a first electrode (2) between each element. .

The open groove αj formed by scribing has a width of approximately 50 mm.
Each of the first electrodes was completely cut and separated by a length of μ and a depth of 20 cm. For this reason, as is clear from the drawings, the edges of the substrate (1) were sometimes gouged out (forming recesses). Thus, the width of the first element (31) and the second element α force was set to 10 to 20 mm.

Using the above laser scribing method, the OTF (2) constituting the first electrode was cut and separated to form open grooves. After this, plasma OVD method or LPC! V
0.2 to 0.7 g of a non-single crystal semiconductor layer (3) having a PN or P-N junction, typically 0.4 to 0.
．． It was formed to a thickness of 5μ. A typical example is a P-type semiconductor (
81XO+-x! -0,250~150A) -Engineering type amol 7 asma fC is a semi-amorphous silicon semiconductor (0,4~0.5p) -N type microcrystalline (100~20A)
A non-single-crystal semiconductor having one P-N junction consisting of a semiconductor having a
O+J-N type, P type S1 semiconductor-type 5ixGel-q
Semiconductor-N type S1 semiconductor with two P-N junctions and one
Tandem type P type NP type N with two PM joints...
・P-N junction (7:-!P conductor (3 N).

Such a non-single crystal semiconductor (3) was formed to have a uniform thickness over the entire surface. Furthermore, as shown in FIG. 2(B), a second open groove α was formed on the left side of the first C open groove α→C by a second laser scribing process.

The laser may be applied either from below the glass (1) or from above this substrate.

Thus the second open groove α barrel flanks the first electrode (8).

(9) was exposed. The side surface (9) of this second open groove may be on the left side of the side surface 0Q of the first electric h (37),
That is, it is characterized in that it is provided over the first electrode position of the first element. And as an extreme example of this,
As shown in FIG. 2(B), it may enter the inside of the first electrode (31). Furthermore, as shown in the conventional example, the present invention has a frowning surface ([4 (see Fig. 1)] of the electrode of the bean paste 1.
It is not necessarily necessary to expose the laser beam, but the laser light intensity is a little too strong at 5~IOW, and the entire depth direction Y of (!TF(3η) is removed, resulting in side surface (8)K
There is no practical problem even if the second electrode (38) is in close contact with each other in FIG. 2(0). That is, it is extremely important for industrial applications of the present invention to be able to provide a margin for the intensity of the output pulse of the laser beam.

In FIG. 2, as shown in FIG. 2(0)K, a second C1ia (4) on the back surface is formed on this top surface, and a C opening groove for cutting and separation by the third laser scribing method is formed. Established.

Silver was formed on the upper surface to a thickness of 300 to 500 cm, and then aluminum or a two-layer film of aluminum and nickel was formed on the upper surface. For example, engineering To is 1
1000 pieces of 050+A1 silver, plus 150 pieces of nickel
01 3N structure. The OTO and silver are C materials in order to promote reflection of incident light 00) on the bag surface side and to effectively photoelectrically convert long wavelength light of 600 to 800 nm. Furthermore, nickel is used to improve adhesion with the external lead electrode. These methods degrade semiconductor layers using electron beam evaporation or plasma CVD.
It was formed at a temperature below 0'O.

This process also helps to prevent a decrease in reliability due to a chemical reaction between the semiconductor (3) and the back electrode (4), that is, to improve reliability.

A second laser beam is emitted upward through the back electrode like this.
This figure shows the case where the electrode is cut and separated to form an open groove (E). The C laser beam slightly excavates the semiconductor, especially the N or P type semiconductor layer that is in close contact with the top surface. This prevents the occurrence of cross leakage (leakage current) caused by metal or conductive semiconductors.

In particular, this semiconductor (3) has, for example, one P-N junction with a P-type semiconductor layer (42), an I-type semiconductor layer (43), an N-type medium conductor/@ (44), and this N-type semiconductor layer It has a microcrystalline or polycrystalline structure. Its electrical conductivity is 1~2ooCf
When the conductivity is as high as Lcm), it is extremely important to hollow out the N-type semiconductor layer of the present invention and provide a semiconductor in the recessed portion to prevent the occurrence of leak current. This punching goes beyond the #crossbody layer of the mold #C for the first electrode! T.F.K.
It is preferable not to reach this point. The present invention uses a laser beam to form grooves not only on the second electrode but also on the P'! It is industrially important that fC can also be removed from the N-type semiconductor layer and can be applied to the I-type semiconductor layer below the layer with a thickness of 0.2 μ or more.

Thus, as shown in FIG. 2(0), a plurality of elements (
31), we were able to create a photoelectric conversion device that connects α forces in series at a connecting part.

FIG. 2(D) shows the attempt to further complete the present invention as a photoelectric conversion device. In other words, a silicon nitride film H of 500-20
It was formed to have a thickness of 00λ to prevent leakage current between each element. Furthermore, connect the external lead terminal (c) to the peripheral part (5).
Established in K. These are filled with an organic resin such as polyimide polyamide, Kapton or epoxy.
0cm
Width of the connecting part is 150μ, width of the external lead electrode part is 1om.
m periphery 4mm K, actual K 5BOmmX19
40 stages are carved within 2mm, and the effective surface is 1di (192mm
x14.35mm x 40 steps 1102cm m” or 9
1.8%). If the resulting segment has a 'lo, 6% c conversion efficiency, spring #K Te9
．． It was possible to provide an output power of 11.6 W at 7% (AM'l (1 oomw/am))K.

This is 55% (
(In the case of 40 stages) This is a very significant effect compared to K. Furthermore, since no metal mask is used at all, there is no industrial problem in the manufacturing process of large-area panels, which is extremely suitable for large-area, low-cost mass production for generating large amounts of power. According to the present invention, the N or P type semiconductor layer under the second electrode was also removed at the same time when the laser scribing method was used to form the second L:D electrode. The leakage current between the electrodes could be reduced from 10 mA/cm to 10'A/Q]nK. Therefore, for general consumer use, it has become possible to omit the silicon nitride coating film (FIG. 2(D)).

Furthermore, since the electrical connection between the first electrode and the second electrode can be made on the side of the first electrode, the area required for this contact part can be reduced to less than 1/10 compared to the conventional method. Figure 3 shows the positional relationship of the most typical grooves created during the three laser scribing processes. FIG. 2 is a cross-sectional view and a plan view (end portion).

Numbers etc. are the same as in FIG. 2.

FIG. 3(A) shows the first open groove α1, the first element (31),
The second element 0]) has a connecting portion (b).

As is clear from the drawing, the first C groove α9 slightly hollows out the substrate (1). Furthermore, the second opening α is provided across the semiconductor (3) that constitutes the first element and the first electrode (2) K, and both of these are removed. Therefore, the first electrode ring of the first element (31) and the second electrode of the second element α force are connected at the connecting portion (2) by a conductor extending from the second electrode. The side surface (8) of the electrode (2)
), and the two elements are connected in series.

Further in the drawing, a semiconductor (3) with at least one PN or P-N junction is shown, here one 5IXc, -
, L(OXl) P type 0 Lie type 5i (43) - A semiconductor having one CPIN junction such as microcrystalline N type 5i (44) is provided. Especially N of this semiconductor
The third opening groove (a) shown in the mold semiconductor (43) K that penetrates through the mold semiconductor (44) has a depth of 075≧
Thus, the second electrodes (4) of the first and second elements (31) and α are electrically cut and separated, and there is a gap (7) between the electrodes! J-riwomo1o-'A/cni (1
To improve the long-term reliability, a coating film of silicon nitride film should be added on the top surface as shown in Figure 2 (B)'. When considering that the scribing accuracy is generally ±15μ, it can be seen that this is sufficient for mass production.

FIG. 3(B) shows a plan view of the first, second and third open grooves (t) at the end (lower side in the drawing)
8), K) are provided.゛In order to further reduce leakage in this direction, the semiconductor (3) is connected to the first electrode (2).
It is characterized by having a structure that covers the 11th and 2nd electrodes to reduce short circuits.0 In this drawing, 50 to 30μ,
The width of the third open groove was 75 to 50 μm (metal is slightly wider because the laser beam conducts in the lateral direction a little better), and the width of the connecting portion was 150 to 100 μm. .

FIG. 4(A) shows a case where the second laser scribing step is performed upwardly of the substrate in the connecting portion 02, and a part of the CTF of the first electrode remains in the open groove αfl).

FIG. 4(B) shows the case where the first open groove part α→ and the second open groove part 0→ are closest to each other, and the side surface αO of the first electrode and the second open groove part
The side surface (9) of the electrode is insulated by the semiconductor (3) K, and furthermore, the width of the third groove groove is provided on the right side compared to FIG. 3 and extends over the first element side. This minimizes the loss area required for the connection part, 100μ to 60μ
It was possible to have a connecting portion with a width of .

The above YAG laser C spot layer has an output of 3 to 5 W (
30μ) 5 to 8W (50μ) is used, but by further reducing the spot diameter based on the technical concept, this connection part can be made smaller, and as a result, the C effective surface 84'x can be used as a photoelectric conversion device. It has an inventive step in that it can be improved.

FIG. 5 is an enlarged view of the external lead-out electrode portion when a photoelectric conversion device y smaller than a large panel for a calculator or the like is to be manufactured in large quantities at the same time.

Fig. 5 (A) corresponds to Fig. 2, but the external lead electrode part (5) is a conductive rubber electrode (pad 0
This pad (49) is the second C? 1 (upper electrode) (4). At this time, the pressure applied to the electrode α'Q is too strong, so that pad 09) is provided with 0 grooves to prevent a short circuit even if it passes through the electrode (2) and the semiconductor (3). Also, the outer part is mostly cut and separated with open grooves. Furthermore, FIG. 5(B) shows that the lower first electrode (2) K is connected to another / There is.

Furthermore, the pad (4 nails) has the characteristic of being able to make several photoelectric conversion devices by contacting the conductive rubber electrode (46).For example, a 20cm x 60cm panel (with a 6cm extension)
If you try to make a 1.5 cm x 1.5 cm photoelectric conversion device (for a calculator), you can make 130 calculator solar cells at once. 5) Cover everything except (45) and then cut it with a glass cutter if you want to make a small power solar cell. Furthermore, this panel, for example, 40cmX20cm or 60cmXX40cm
It is possible to provide a panel for large force according to the NEDO standard.

In addition, the panel of this (DNEDO%) is made of Seaflex, and the other glass plate of the present invention C is the photoelectric conversion device IC reflective surface side (upper side in the drawing). It is also effective to increase the mechanical strength against wind pressure, rain, etc. by placing a The side is not limited to the lower side.

[Brief explanation of drawings]

Figure 1 is a vertical cross-sectional view of a conventional photoelectric conversion device C0 Figure 2 is a vertical cross-sectional view showing the manufacturing process of the photoelectric conversion device of the present invention Figure 5 is a photoelectric conversion device of the present invention FIG. 3 is an enlarged view of a portion of a C-board I# side view and a plan view. 4th shi 1 to 5th shi! is an enlarged vertical sectional view of another light converting device C according to the present invention. 1980 Patent Application No. 206809 (filed on November 24, 1982) 2. Name of the invention Photoelectric conversion device 3. Relationship with the person making the amendment Patent applicant's specification and Drawing 5, Description of amendments 1, Title of the invention Photoelectric conversion device 2, Claim 1, A first electrode and at least one PN or PIN junction on the electrode on a substrate having an insulating surface. In the photoelectric conversion device, a plurality of photoelectric conversion elements each having a non-single crystal semiconductor and a second electrode on the semiconductor are electrically connected in series and disposed on the insulating substrate. and an opening groove that separates the semiconductor of the second element and the first electrode thereunder is provided over the first electrode position of the first element, and from the second electrode of the second element. A photoelectric conversion device characterized in that the extended connecting portion is provided in close contact with a side surface of the first electrode of the first element. 2. In claim 1, the first element A photoelectric conversion device characterized in that the opening for separating the semiconductors of the first and second elements provided over the first electrode position is provided inside the first electrode. 3. Invention DETAILED DESCRIPTION OF THE INVENTION The present invention provides electrical conversion of a plurality of photoelectric conversion elements (also simply referred to as elements) in which a non-single crystal semiconductor including an amorphous semiconductor having at least one PIN or PN junction is provided on a transparent insulating substrate. This invention relates to a photoelectric conversion device that is connected in series to a photoelectric conversion device that can generate a high voltage.The present invention uses a laser The present invention is characterized in that a scribing method is used.The present invention is characterized in that the second groove for electrically connecting the first and second elements is formed between a non-single crystal semiconductor having a PN or PIN junction and a second groove underneath. 1 electrode was removed, and the 1st electrode was removed.
The second electrode of the second element is connected to the side surface of the first electrode of the element, and in this case, the second groove is provided to extend inside the second electrode of the first element. The purpose of this is to make the first element and the first isolation more perfect. For this reason, the groove separating the second electrodes of the first and second elements is provided over the semiconductor of the first element, thereby providing redundancy for improving manufacturing yield. In this invention, the electrical connection at the connecting portion is
By extending the second electrode of the second element to the side surface of the transparent conductive film (CTF) constituting the electrode, and using it in close contact with the side surface, the area required for the connection part is reduced. It is characterized by For this reason, an opening groove separating the semiconductors of the first and second elements is provided over the first electrode position of the first element, and L
It is characterized by providing manufacturing redundancy (margin) due to fluctuations in the scanning of 'S. Conventionally, in the mask alignment method, the number of connecting parts is 5 to 1.
It required a width of mm, but this is 1/10 to 1
/100 350-30μ, preferably 200-50μ
By setting this connecting part to 10 to 50 stages (10 to 50 steps), the area for photovoltaic generation (effective area or effective area) of all panels used as a photoelectric conversion device can be reduced. and l, ) u) power (, (ashiki) 75-50
% to 97-90%, and the effective conversion efficiency is 1.
It is characterized by a substantial improvement of about 6% by 0 to 20%. In this invention, by using a laser beam scribing method, the address to be scribed is prepared in advance in a memo on a computer (microcomputer) using the alignment mark as a reference.
It is characterized by eliminating all the causes of price increases and yield decreases in manufacturing at once, such as mask misalignment, warping, and decreases in manufacturing yield due to blue discoloration required by conventional mask alignment methods. shall be. Conventionally, many photoelectric conversion devices (hereinafter referred to as type G devices), that is, devices in which a plurality of elements are arranged on the same substrate and are integrated or hybridized, are well known. For example, JP-A-55-4994, JP-A-55-12427
4 Furthermore, patent application No. 54-90097 filed by the present inventor
/90098/90099 (Showa 54.7.16
) is known. For example, a patent application filed by the inventor, a semiconductor application, etc.
This is a heterojunction of i x Cl-x S i, which is different from the case where a single G amorphous silicon semiconductor is used. In addition, it is said that a non-single-crystal semiconductor having at least one PN or PIN junction doped with a halogen element is integrated into an integrated single-crystal semiconductor/S phyllide (which has the following characteristics. However, these conventional Regarding the invention, FIG.
The alignment accuracy was insufficient, and Yoshibe G required a large area for the series. For example, when a metal mask is used, a method of directly producing a selective conductive layer or a semiconductor layer is used. It may be stored away. Furthermore, since a film component is formed on the mask here, the mask force (Nη) is stained, and the peripheral edge of the film formed by the mask is not clear.
This method has many drawbacks, such as causing crosstalk (leakage current) between adjacent electrodes. Furthermore, conventionally known screen printing methods and the like are methods in which conductors or semiconductors formed entirely on a substrate are independently and selectively etched away using a mask. However, in this method, the process is extremely time consuming, including the process of aligning the mask for screen printing, the process of coating the resist, the process of beta solidification, the process of etching the conductor or semiconductor, and the process of removing the resist. The rise was inevitable. However, in the photoelectric conversion device of the present invention, particularly the thin film type photoelectric conversion device, each of the thin film layers, the conductive layer for electrodes and the semiconductor layer, each has a thickness of 500 to 1 μm, and by using a laser scribing method, , it has become possible to fabricate without the need for mask alignment at all. As a result, the conventional mask alignment process has been replaced by a female scribing process, which is extremely simple and highly accurate due to the combined use of a microcomputer, resulting in a reduction in the manufacturing cost of the device. Therefore, it is possible to manufacture a photoelectric conversion device at a price of 500 yen/W, and by expanding the manufacturing scale, it becomes possible to produce a photoelectric conversion device at a price of 100 to 200 yen/W. Furthermore, in the present invention, in addition to using this laser scribing process, it is important to provide a degree of redundancy (margin) in the alignment accuracy of the scribe lines. Therefore, the first electrode (lower side) between adjacent elements and the second electrode (upper electrode) of another element are electrically connected to the first electrode and its side surface by a lead extending from the second electrode. By connecting it to the scribe line, it was possible to increase redundancy at the position of the open groove of the scribe line. A conventional example and the structure of the present invention will be described below according to the drawings. FIG. 1 is a longitudinal sectional view of a conventionally known photoelectric conversion device using a mask alignment method. In the drawings, a transparent conductor 11i (abbreviated as CTF) constituting a first electrode is selectively formed on a transparent substrate (for example, a glass plate) (1) by a first mask alignment step. Further, a semiconductor layer (3) is similarly selectively formed by a second mask alignment step. Further, a second electrode (4) is provided by a third mask alignment step. In FIG. 1, there is a connecting part (12) between the elements (11) and (31), and in the connecting part, the semiconductor layer (3) covers one side surface (16) of the CTF, and the semiconductor layer (3) covers the other side surface (16) of the CTF. In order to prevent the semiconductor layer (3) from covering the surface (14),
A gap of 1 to 5 mm, for example 3 mm, is required between the CTFs (13). Furthermore, the first electrode (37) and the second electrode (
Since the second electrode of 38) must not be short-circuited, including the spread caused by mask blur, a gap (6) of 1 to 5 mm, for example 3 mm, is required. These three masks do not have any self-alignment properties, so the connection part (
12) requires 1 to 8 mm, typically 4 mm. Furthermore, if this is set to 1 mm or less, the mask alignment accuracy will be extremely strict, and the yield will be extremely reduced. If the gap between this connecting part (12) is 5 mm or more, for example, 20 cm x 60 ctrl and 15 mm
mm (20cm x 15mm) element, edge 5mm
If you try to make one, you will only be able to connect 20 stages directly. Furthermore, even if the gap at this joint is 3 mm, there are 33 stages, resulting in a total loss of 10 cm (area of 200 cA) at the joint, and as a result, the effective area remains at 75% when the peripheral area is taken into account. . The present invention eliminates such process complexity and reduces the effective area to 86.
The object of the present invention is to provide an extremely innovative photoelectric conversion device that can increase the photoelectric conversion efficiency by 97%, for example, 92%. The details of the embodiment will be shown below according to the drawings. FIG. 2 is a longitudinal sectional view showing the manufacturing process of the present invention. In the drawings, a substrate with an insulating surface, such as a transparent substrate (
1) That is, a glass plate (for example, 1.2 inches thick, 60 cm long in the horizontal direction in the drawing, 20 cm in the middle) was used. Furthermore, a light-transmitting conductive film, such as ITO (approx. 1,500) +5nOz (200 to 40
A light-transmitting conductive film (1,500 to 2,000 layers) containing tin oxide as a main component or containing a halogen element was formed by vacuum evaporation, LP CVD, plasma CVD, or spraying. After that, a YAG laser processing machine (manufactured by Nippon Laser) applies an output of 5 to 8 W from the bottom or top of this substrate, and the spot diameter is 30 to 70μφ.
Typically, 5oμφ was irradiated by controlling a microconbeater, and by scanning it, an open groove (13) for a scribe line was formed, and a first electrode (2) was produced between each element. The open groove (13) formed by scribing has a width of approximately 5
Each of the first electrodes was completely cut and separated at a length of 0μ and a depth of 20cm. For this reason, as is clear from the drawings, a portion of the substrate (1) may be pinched (a recessed portion may be formed). Thus, the first element (31) and the second element (11
) was set to have a width of 10 to 20 mm. Using the above laser scribing method, the CTF (2) constituting the first electrode is cut and separated to form an open groove)
accomplished. After this, put plasma cvoa or LP on this top surface.
A non-single crystal semiconductor layer (3) having a PN or PIN junction is formed by CVD to a thickness of 0.2 to 1.0μ, typically 0.4
It was formed to a thickness of ~0.5μ. Typical examples are P-type semiconductors (SixC: I-x
0 to 200), or a P-type semiconductor (Si
, x C+-x) -■ type, N type, P type Si semiconductor -■
Since it has two PIN junctions and one PN junction made of a type 5ixGe I-x semiconductor and an N-type Si semiconductor, it is a tandem type PJNPI'N...PIN junction semiconductor (3). Such a non-single crystal semiconductor (3) was applied over the entire surface. The film was formed to have a uniform thickness. Furthermore, as shown in FIG. 2(B), there is a second open groove (18) on the left side of the first open groove (13).
was formed by a second laser scribing step over a distance of 100 to 500 μ over a width of 50 μ. The laser may be applied either from below the glass (1) or from above the substrate. The second open groove (18) thus forms a side surface (8) of the first electrode.
, <9> was exposed. The presence of the right side surface (9) of the first electrode formed by this second groove is due to the presence of the first electrode (37).
It is characterized by being provided over the electrode positions. Then, as shown in FIG. 2(B), by entering the inside of the first electrode (31), the side surface of the first electrode (8) is exposed (<9). As a result, a portion of the first electrode (37) of the first element remains on the right side of the second groove. If there is no such remaining area, the high heat (~2000°C) of the laser beam will cause
Since it is much easier to process than TF (2), the semiconductor filled in the first groove (13) will be blown away. Therefore, isolation between the first electrodes of the first and second elements becomes impossible. From this, Figure 2 (B)
It is very important that the second electrode is provided inside the first electrode, as shown in FIG. This (9)
The CTF remaining in the portion had a width of 50 to 500μ. Furthermore, the present invention is not intended to expose the surface of the first electrode (14><see Fig. 1) as shown in the conventional example, but because the laser beam is somewhat too strong at 5 to IOW, this CTF
(37) in the depth direction, and as a result, there is no practical problem even if the second electrode (38) is brought into close contact with the side surface (8) as shown in FIG. 2(C). That is, it is extremely important for the industrial application of the present invention to be able to provide a margin for the intensity of the output pulse of the laser beam. In FIG. 2, as shown in FIG. 2(C), a second electrode (4) on the back surface is further formed on this top surface, and a third groove for cutting and separation in the third laser scribing method is formed. (
2o) was provided. This second electrode (4) has a transparent conductive film of 700 to 140
It is made of ITO (indium tin oxide) to a thickness of 0.000 mm, and a reflective metal of silver is coated on the top surface with a thickness of 300 to 300 mm.
It was formed to a thickness of 0.00 people. Furthermore, a 2N film of aluminum or aluminum and disol was formed on the upper surface thereof. For example, 1050 people hold ITO, 1000 people hold silver,
Furthermore, a three-layer structure with 1,500 nickel members was used. This I
TO and silver promote reflection of incident light (10) on the back side, and 6
This is for effectively photoelectrically converting long wavelength light of 00 to 800 nm. Furthermore, nickel is used for external extraction electrodes (2
This is to improve the adhesion with 3). These were formed using an electron beam evaporation method or a plasma CVD method at a temperature of 300° C. or lower, which does not cause deterioration of the semiconductor layer. This ITO also helps to prevent a decrease in reliability due to a chemical reaction between the semiconductor (3) and the back electrode (4), that is, to improve reliability. The back electrode is irradiated with a laser beam from above to cut and separate the second electrode, and the third electrode is opened! (20) (middle school 50
The figure shows the case where μ) is formed. This laser beam slightly gouges out the semiconductor, especially the N or P type semiconductor layer that is in close contact with the top surface. Alternatively, the occurrence of crosstalk (leakage current) due to conductive semiconductors was prevented. type semiconductor layer (43), N type semiconductor layer (44) and, for example, 1
This N-type semiconductor layer has a microcrystalline or polycrystalline structure. Its electrical conductivity is 1-200 (
When the conductivity is as high as Ωc+11)-', the N of the present invention
It was extremely important to hollow out and remove the mold semiconductor layer and provide semiconductor in the recessed portions to prevent leakage current generation. It is preferable that this extrusion does not exceed the ■-type semiconductor layer and does not reach the CTP for the first electrode. The present invention uses laser light to form grooves not only on the second electrode, but also in the process of forming the groove portion (20) with a margin equal to the thickness of the I-type semiconductor layer that is 0.2μ or more below the second electrode. It is industrially important to make it durable. Thus, as shown in FIG. 2(C), a plurality of elements (
We were able to create a photoelectric conversion device in which 31) and <11) were connected in series at a connecting part. FIG. 2(D) shows the attempt to further complete the present invention as a photoelectric conversion device, that is, a silicon nitride film (21) with a thickness of 500 to
It was formed to a thickness of 2,000 mm to prevent leakage current between each element. Furthermore, an external lead-out terminal (23) was provided at the peripheral portion (5). These were filled with an organic resin (22) such as polyimide, polyamide, Kapton or epoxy. Thus, for the irradiation light (10), each element on a substrate (60 cm x 20 cm) as in this example was
．． The effective area (192 m
m x 14.35 mm x 40 stages ILO2c (ie, 91.8%) could be obtained. As a result, if the segment has a conversion efficiency of 10.6%, at 9.7% (AMI (100mW/c)) at the panel, 1
It was possible to provide an output power of 1.6H. This is 55% (
This is a very significant effect compared to the case of 40 stages). Furthermore, since no metal mask is used at all, there is no industrial problem in the manufacturing process of large-area panels, making it extremely suitable for large-area, low-cost mass production for generating large amounts of power. As is clear from the above examples, the N or P type semiconductor layer under the second electrode was also removed at the same time by cutting and separating using the laser scribing method to form the second electrode according to the present invention. The leakage current between the two electrodes could be lowered from 10-''mA/cm to 10''L/crn. Therefore, for general consumer use, it has become possible to omit the silicon nitride film coating (21) in FIG. 2(D). Furthermore, regarding the connection part, since the first electrode is connected to the second electrode of the adjacent element on the side surface, the area required for this connection part (contact part) is reduced to 1/10 or less compared to the conventional method. As a result, the effective area of the panel can be improved. The above YAG laser spot layer has an output of 3 to 5 W (
30μ), when used at 5 to 8W (50μ),
Furthermore, by reducing the spot diameter based on the technical concept, this connecting portion can be made smaller, and as a result, the effective area of the photoelectric conversion device can be further improved, which is an inventive step. FIG. 3 is an enlarged view of the externally drawn electrode portion when photoelectric conversion devices smaller than a large panel for a calculator or the like are to be manufactured in large quantities at the same time. FIG. 3 (Δ) corresponds to FIG. 2, but the external extraction electrode part (5) has a pad (49) that contacts the conductive rubber electrode (47), and this pad (49) It is connected to the second electrode (upper electrode) (4). At this time, the electrode (47)
An opening groove (13') is provided to prevent a short circuit even if the pressure is too strong and the nodule (49) penetrates the first electrode (2,) and semiconductor (3) below. . Outside again (
1! Part j is cut and separated by openings i (18') and (20').Furthermore, FIG. 3(B) shows the lower first electrode (20').
) is connected to the conductive rubber electrode (46) by a second electrode material. It is connected to the outside with an electric curved plate.Here too, there is an open groove (1
8'), (18''), (20''), (20) The pad (48) is electrically isolated from other photoelectric conversion devices, and the glass cutting between these devices is performed in a post-process. It has the feature that a large number of photoelectric conversion devices can be made with one panel without using any alignment masks.For example, 20cm x 60cm.
If you try to make a 6cm x 1.5cm photoelectric conversion device (for a calculator) using a panel of m, it turns out that you can make 130 solar cells for a calculator at once. In other words, the photoelectric conversion device is an organic resin mold (22) with an electrode part (5),
<45) is covered, and if you want to make a small power solar cell after this, you can easily cut it with a glass cutter. Furthermore, this panel, for example, 40cm x 20cm or 60cm
Packaged by combining 6 or 4 cm x 20 cm in an aluminum sash frame, 120
It is possible to provide a NEDO standard high power panel measuring 40 cm x 40 cm. In addition, this NEDO standard panel is made by laminating another glass plate using Seaflex on the reflective surface side (upper side in the drawing) of the photoelectric conversion device of the present invention, and the photoelectric conversion device is placed between them. It is also effective to increase mechanical strength. In FIGS. 2 and 3, light was incident from the lower force glass plate. However, the present invention does not limit the light incident side to the lower side. 4. Brief Description of the Drawings FIG. 1 is a longitudinal sectional view of a conventional photoelectric conversion device. FIG. 2 is a longitudinal sectional view showing the manufacturing process of the photoelectric conversion device of the present invention. FIG. 3 is an enlarged vertical cross-sectional view of an electrode portion of an external bow 1 of another photoelectric conversion device of the present invention. patent applicant

Claims (1)

[Claims] 1. On an insulating substrate, a first electrode of a transparent conductive film, a non-single crystal semiconductor having at least one PN or P-N junction on the electrode, and a second electrode on the semiconductor. In a photoelectric conversion device in which a plurality of photoelectric conversion elements are electrically connected to each other in series and arranged on the insulating substrate, the opening between the semiconductors constituting the first and second edge button elements is located below the opening. A photoelectric conversion device 02 characterized in that a side surface or surface of the first electrode that is in close contact with the semiconductor is exposed; ~ that the separating groove is not provided over the first electrode position of the first element;
Features: Photoelectric conversion device.