JPS60200578A - Manufacture of photoelectric converter - Google Patents

Abstract

PURPOSE:To save time necessary for processing and to prevent the primary material under the surface to be machined from damaging by electrically coupling between adjacent elements, forming second hole or groove in an active region to form a contact, connecting and arranging a plurality of elements in series. CONSTITUTION:A nonsingle crystal semiconductor is formed in homogeneous film thickness on the all upper surfaces of an electrode 3 and a groove 16, and the side 17 of the first electrode is exposed without damaging the surface of a film by forming a laser light. Then, the second electrode 5 is formed on the semiconductor 4 by an electron beam depositing method, and the conductive oxide of ITO is ohmically contacted with the side or upper surface 17 of the first conductive film 3. Thereafter, a laser scribing 10 is executed to form a groove at the position of the second cell side 13. The groove 13 is formed by removing the semiconductor thereunder, forming a groove, and electrically isolating between the second electrodes without the first electrode, further light transmission organic resin 28 is coated on the upper surface of the first electrode, and eventually cut at the boundary 70, and forming photoelectric converters.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to stacking photoelectric conversion elements or cells (hereinafter simply referred to as cells) on a flexible substrate having an insulating surface. Cutting (opening grooves) of the electrical film is performed by scanning linearly with a laser scribe (hereinafter referred to as LS), and the scanning speed is kept constant or approximately constant (hereinafter referred to as constant) to reduce the time required for processing. The purpose of this is to save on the amount of steel and to prevent TL scratches on the base material under the surface to be machined.

In the present invention, a first electrode is formed on a substrate having an insulating surface in a cell provided in an active region by a scribing method in which a plurality of first grooves are formed in a first conductive film by scanning a laser beam. , furthermore, a non-single crystal semiconductor that generates photovoltaic force by light irradiation is placed on this electrode, a second LS is formed on the semiconductor in parallel (including approximately parallel) to the first groove, and a second A second groove or hole is formed on top of this.
forming a conductive film, forming a third groove parallel to the first groove in the second conductive film by linearly scanning a laser beam;
It is separated into multiple electrodes. As a result, electrical connections between adjacent elements are established by providing contacts inside the active region through second openings or grooves (hereinafter simply referred to as openings), and a plurality of elements are connected in series. It is characterized by:

2. Description of the Related Art Flexible thin films having an insulating surface have been used as substrates for inexpensive, mass production of photoelectric conversion devices.

This invention provides a thin film having an insulating surface, such as an organic resin thin film (hereinafter referred to as OF), on a flexible base soil such as a stainless steel substrate, and a first 2# on the substrate having such an insulating surface.
The film consists of a thin film of chromium or a heat-resistant sublimable metal whose main component is chromium (hereinafter simply referred to as chromium), or a reflective metal such as aluminum and a translucent sublimable conductive film such as tin oxide on its upper surface. A composite conductivity test is installed, and when scanning the conductive film with the laser beam in a straight line, it is possible to scribe the first conductive film without damaging the underlying insulating V# film, such as an organic resin thin film. We have experimentally discovered that conditions that cannot be given exist, and we have attempted to utilize this fact to fabricate semiconductor devices, particularly photoelectric conversion devices.

The present invention provides that when a laser beam is irradiated to such an OF and a conductive film having a sublimating property thereon, particularly a conductive oxide film having a metal mainly composed of chromium, indium oxide, or tin oxide, this We experimentally investigated the conditions under which only the conductive thin film on the OF could be selectively removed without damaging the OF, and found that the laser beam was kept in one place for a long time (
We discovered that it is possible to selectively remove only this conductive material for electrodes by keeping the scanning speed constant or approximately constant without irradiation (for more than several tens of milliseconds). Ta.

That is, the leaves have a small thermal conductivity due to laser light irradiation (generally 1 to 7 x 10" Cal/sec/c
(at/°C/cm), therefore, if laser pulses are repeatedly applied to the same position, heat builds up within this organic resin, and this heat carbonizes and cuts the resin. However, if this is repeated once or several times, the thermal conductivity of this OF becomes conductive thin I15! ! (hereinafter referred to as CF) For example, 1/103 of a composite film of chromium or aluminum and CTF
Therefore, it has been found that conversely, only CF can be selectively removed from the area irradiated with laser light.

FIG. 1 shows a typical example of a conventional structure.

FIG. 1(A) is a first side view of the photoelectric conversion device (1) seen from the back with the transparent glass root (2) facing downward.

In the drawing, it has an active region (14) that generates a photovoltaic force upon irradiation with light, and an inactive region (15) that has a connecting portion (12) that connects each cell (11) (<13). As is clear from the longitudinal cross-sectional views taken along lines AA" and BB' in FIG. 1(A) shown in FIG. 1(B) and (C), each cell ( 11),
<13) (3) of the transparent conductive H'A (CTF) of the first electrode on the glass substrate 47m (2) are separated from each other between each cell. Further, the semiconductors (4) are connected to each other between each cell. Further, in the inactive region, the upper electrode of the cell (13) is connected to the lower electrode of the cell (11) by contacts (18) at the connecting parts (6) and (7),
Repeat this to connect 5 cells to the external electrode (8), < 9
) are connected in series.

However, in this conventional structure, since there is only one semiconductor (4), it appears that the manufacturing yield is low. However, in reality, there are three masks (the first mask for buttering the first four layers, the second mask for forming the non-active region, the second mask for buttering the second conductive film, and the third mask for buttering the second mask). mask), but
In this mask, since the first mask and the third mask are not self-aligned, mask misalignment is likely to occur.

Due to this deviation (i.e., a deviation of 0.3 to 1 mm is quite natural for a metal mask), the effective area of the cell is reduced by 10 mm.
It was found that there was a substantial reduction of ~20%.

Furthermore, even if this buttering were to be performed using laser light, it would be completely unrealistic.

That is, for example, when trying to form the first electrode (3), the scanning is not a simple linear scan, but requires several 90° turns to form one electrode. Therefore, the scanning speed becomes zero or extremely slow at the corner. This means that the Q-switched pulsed light is applied more times than necessary in this corner, and as a result, even the substrate material is damaged by Ii3. In addition, it has the double drawback of slowing down the overall machining speed. Furthermore, a mask is always required to prevent non-single crystal semiconductor from being formed in the non-active region. For this reason, it is completely impossible to form the structure shown in FIG. 1 using a laser scribing method. Therefore, the most suitable device structure for laser scribing and its manufacturing method were strongly sought after.

That is, in the present invention, a typical example is shown in FIG.
From the light irradiation surface (10), there is only an open groove (11310 to 70μ, typically 20 to 30μ, as shown in Figure 2 (20)) that is invisible to the naked eye and is used to separate the plurality of second electrodes. It is. Further, there is no non-active region such as the region (15) in FIG. 1A, and the connecting portion constitutes an iso-rayon region of each cell. In addition, since it is a completely maskless process that uses LS and simply scans the laser beam in a straight line and scans at a constant speed in the processing area, the first groove is laminated on a TV monitor. Then, a so-called computer-aided self-processing method performs optical patterning in which a laser beam is scanned in parallel or approximately parallel to a predetermined position using the open groove as a reference to form a second open groove and a third open groove. The use of one type of Lenostrei Soyonka style became i+J Noh.

Further, the contact at the connecting portion between the first electrode of the first cell and the second electrode of the second cell is located inside the semiconductor of the substrate.
(In FIG. 2, the contact is provided in the center, and the end of the opening is not cut into the semiconductor.) The position of the contact is completely different from that of the conventional contact on the outside of the semiconductor.

Furthermore, instead of creating an opening with a structure in which this contact cuts all the semiconductor between adjacent cells, the opening groove (20
~90μφ) are discontinuously provided in a circular or elliptical shape (open/111) (that is, the semiconductor constituting each cell is continuous at the end of this open groove).
As a result, this open groove can be practically seen with the naked eye.
Another feature is that the scribe line can be prevented from becoming an eyesore on the product.

The present invention has many such features, and the details thereof will be described below with reference to the drawings.

FIG. 2 shows the manufacturing process and apparatus of the photoelectric conversion device of the present invention.

In the drawing, a stainless steel thin film (2') with a thickness of 100μ is shown.
Polyimide resin or PES (polyether sulfmen film) (Sumilite manufactured by Sumitomo Heckleite Co., Ltd.) on top (continuous use temperature 150-300'C, thermal conductivity 3
~7 X 10'4Cal/sec/cnl/
°C7cm) (2' layer is formed to a thickness of 1 to 10μ,
This results in a substrate (2) with an insulating surface (for example, a thickness of 1
00μ, length in the left and right direction in the drawing) 60cm, IlJ
20 cm).

Using electron beam evaporation or magneto sputtering, chromium (JI 2000), a metal with heat resistance and sublimation properties, or aluminum (1000), a reflective metal, is coated on this OF by electron beam evaporation or magneto sputtering. A layer of SnO was formed on the top using the sppack method to a thickness of 1,100 mm, giving a composite conductivity of 1.
I used a movie. The sheet resistance was then 3.5Ω/hole.

This drawing shows a case where four cells are connected in series.

That is, the photoelectric conversion device of the present invention has a larger size than that of 100 to 2000 active regions (14) on the same substrate at the same time.
A substrate measuring cm x 60 cm was used.

In each cell, a first conductive film was formed over the entire surface of the substrate. Furthermore, this 2# membrane is shaped into a predetermined shape using a laser (here, 1.0
A first opening (16) was formed by scribing (YAG laser) at a wavelength of 6μ or 0.53μ according to a pattern marked and controlled by a microcomputer. Furthermore, in order to eliminate leakage outside the cell, isolation grooves (26) and (26') were formed. Then, the cell area (11), < 13) and the external connection electrode part (
8 ru (9) was formed.

That is, YAG laser f (emission wavelength: 0.53 μm, focusing: 11150 mm, light (¥20 μ)) was irradiated here.

The conditions are: 6Ktlz, average output (1.2W), scan speed (scanning speed, hereinafter referred to as SS).
) 30 cm/min.

Open grooves formed by scribing (16> have a width of approximately 2
5μ, length 20cm (1 cm in the drawing), and depth was completely cut W1 apart from each first electrode on the ridge.

The diameter of the first element (1 piece) and the second element (13) was 10 mm.

When examined using an electron microscope at this time, no damage or partial deterioration was observed on the leaf surface. Although this laser light is considered to be fluorous when it has a temperature of +600'C or higher, it did not impart any illumination to the leaves, which only have a low heat resistance of about 180'C or the upper limit temperature for continuous use.

That is, it was found that open grooves (16) can be selectively formed in the CFs on the ridges. Second, I was able to obtain a resistance of IMΩ or more (the inside value is Icm) between the two probes.

FIG. 3 shows the electrical resistance when the repetition frequency of the laser beam is varied and an open groove is formed.

In the drawing, YIIG (wavelength 0.53
μ) Using a laser. Then, when the frequency is lowered below 1 kHz, the curve (45) discontinuously becomes IMΩ or higher (45') below 7K II z, which means that electrical isolation can now be achieved. found.

However, this frequency is below 4 K tl z (pulse is 2
．． In addition to this CF, the underlying leaf was also damaged in its center (the area where the Gaussian energy density is the highest).

FIG. 4 shows an investigation of the necessity for constant scanning speed in order to obtain the characteristics shown in FIG. 3. That is, the first open groove shown in FIG. 2 (Δ) allows two points (46>, <47
) shows the need to scan a long distance from above to below at a constant speed.

Furthermore, an example of constant speed is scan speed 40cm/
It was possible to form open grooves up to a speed of +lJ (approximately constant scanning speed) within about 30 to 50 cm 2 /min (±25%). l! Il et al. substrate position (49) <
(46) and (47) are shown in Figure 2 (20 cm).
A) has been written off. That is, the distance between vehicles is such that the scan speed is within ±25% of the optimum speed during a constant scan.

A plan view of FIG. 2(A) and longitudinal sectional views along lines AA' and FF' are shown in (A-1)<(A-2), respectively.

Next, as shown in the plan view of FIG. 2(B), a non-single crystal semiconductor doped with hydrogen or fluorine that generates photovoltaic force when irradiated with light is placed in all of the electrodes (3 and 2 grooves (16)). A homogeneous crotch is formed on the top surface.

This semiconductor (4) is made of, for example, I) type I) of Six - Or a semiconductor whose main component is silicon to which hydrogen elements are added is 0.4 to 0.
．． The r'IN junction structure was made of a semiconductor having a thickness of 8 μm and mainly composed of N-type microcrystalline silicon or N-type 5ixC1-x (0<x<1 x to 0.9). Of course, this is P (SixC+-x x =0.7-0.8
) −1(Si)−N (μC5i ) P (Six
C1-XX=0.7 ~ 0.8)-I (SixGe
It may be a tandem structure of 111 N PIN structure such as l-X X =0.6-0.8) -N (ii&crystallization@SI or 5ixC1-y O<x<1).

Furthermore, a second aperture (15) is formed using a laser beam,
Vertical cross-sectional views along lines BB' and CC' in FIG. 2(B) are shown corresponding to (B-1) and (B-2).

Thus, the second opening (15) exposed the side surface (17) of the first electrode without causing any damage to the surface of the OF itself.

At this time, as a result of exposing the upper end of the conductive thin film within 0 to 5 μm, the side and top surfaces of the conductive thin film (3) formed contacts of the connecting portion.

Furthermore, if the output is reduced, only silicon can be removed because the wavelength of 0.53 μ of the laser beam has a quotient absorption coefficient of silicon. In this case, the contact can be made on the top surface.

The conditions for forming the second opening (15) are the same as those for forming the first groove, except that pulses of laser light are discontinuously applied only to the position (15). It was formed by scanning in parallel using the groove as a base plate.

The characteristics shown in FIGS. 3 and 4 could be used.

Next, the pattern shown in FIG. 2(C) was formed. Figure 2 (C
)'s D-Ll', E-E', and GG' longitudinal cross-sectional views according to J (C-2>, <C-3), (C-1)
It is shown in

That is, a second electrode was formed on the semiconductor (4) by electron beam evaporation to a thickness of 400 to 1000, for example 700.

Then, the conductive oxide of ITO came into contact with the side surface or top surface (17) of the first conductive film (3) at the opening (15), making ohmic contact possible.

In particular, chromium was preferred for laser machining because it has low contact resistance and excellent laser machinability.

Furthermore, since the second electrode was made only of fTO, the laser light was transmitted through it, and the third groove substantially scribed the ITO and the semiconductor underneath.

After this, the laser star live (19
) was carried out. This is a YAG laser (wavelength 0.53μ)
While monitoring the first groove on a television monitor, an open groove was created at a position 50 to 200 μm further into the second cell side (13). The average output of the laser beam is 0.1
~0.3 W, a beam diameter of 10 to 3 μΦ, and a beam scanning speed of 0.1 to 1 m/min, generally 0.3 m/min.

A Q-switched laser beam having a wavelength of 0.6 μ or less, that is, 0.53 μ, which has an extremely large light irradiation coefficient of the semiconductor, was particularly effective for this third open groove.

Thus, since Si (semiconductor with a thickness of 0.5μ or more to which hydrogen or halogen elements are added) is inherently sublimable and has a large absorption coefficient at a wavelength of 0.53μ,
The temperature is rapidly raised by laser light, and the semiconductor and its
It was possible to vaporize and remove TO at the same time.

Thus, in the connecting part (12), the first electrode (23') of the cell (13) and the second electrode (23') of the cell (11)
5) is in ohmic contact with an oxide contact through the second open groove (18). In particular, the contact l-(17) in the connecting part (]2) is connected to the side surface or side face of the first electrode made by the second aperture (15) and the I of 0 to 5μ.
This is achieved with the upper end surface of the first electrode of 'lJ, forming a so-called side contact structure. That is, the site contact of the second opening of only lO~70μφ is sufficient for two cells. The material constituting the second 'd1 pole is brought into close contact with this part to electrically connect it in series. (C
-1>, <C-2) As is clear from the vertical Wi plane view, the second electrode (5) is simply formed on the semiconductor (4).

This third trench (20) is created by removing the semiconductor underneath, and creating an electrical isolation between the second electrodes of each element without hollowing out the first electrode. was completed.

Furthermore, in Fig. 2 (C), a translucent self-produced resin (28) is applied to the upper surface of these using a 2P process (particularly, using a liquid that is cured by ultraviolet light without heating +3+-) to remove the translucent insulation. It is completed by coating using the +) process.

As a result, instead of making a single 1 cm x 5 cm photoelectric conversion device on a thin metal layer with an insulating surface,
20cm x 20cm or 20cm x 60cm
Or, it has become possible to create a large number of photoelectric conversion devices at once on a large grass root measuring 40 cm x 40 cm.

Finally, these were cut at the boundary (70) by a cutting method to produce each photoelectric conversion device. For this purpose,
Instead of creating an active region and a non-active region as in conventional photoelectric conversion devices, the entire active region is made into a substantially active region, and grooves formed by laser light are created from end to end to increase the scanning speed of the laser light. mf is always kept constant over a large OF. Otherwise, the leaves will be damaged in the parts where the SS is slow.

Opening of lecture (20), <27>, <2 in Figure 2 (C)
7') is scanned in a straight line from end to end.
This is important when considering mass production. In particular, in areas where the scanning speed at the beginning and end of laser beam scanning changes, the energy density of Lj in the irradiation area will increase.For this reason, it is necessary to perform LS using only a constant speed area. Therefore, opening formation using a linear LS is an extremely important process.Of course, these opening grooves are not visible at all from the entrance side, so they do not hinder high commercial value.

Furthermore, as is clear from FIG. 2(C), the effective area of the cell is 92%, except for a very small part of the connecting part (12) of 10 to 300μ. The structure of the present invention was significantly superior to 80% of the conventional example.

Furthermore, this conversion efficiency is 300Lx
6.3%, and an open circuit voltage of 1.9V could be achieved.

Taking these things into consideration, it has been found that the present invention has the following major features.

That is, the present invention adopts the method of [1] creating a large number of photoelectric conversion devices at the same time on a dog-sized substrate with an insulating surface, and dividing the devices into 11 parts to create one photoelectric conversion device on each root. became Kuchi J Noh. Therefore, it is possible to manufacture it at 1/3 to 115 times the conventional price.

[2] Since the first opening, second opening, and third opening are self-registration methods controlled by a computer, the self-J area of the cell is large, and in between - each photoelectron made of Hanochi. There is little variation between conversion devices [3
Since this is a maskless process using a linear scanning LS, the manufacturing yield is high.

[4] The IC of the scribe line of the first and third open grooves of each nacelle separation is extremely small, 10 to 70μ, and
The openings are also extremely small, 10 to 50 μψ, and the first opening grooves are not visible at all from the light irradiation surface.
It was possible to give the value of the product.

In Fig. 2, only one second flowering (15) is present in the semiconductor R1 (particularly near the center). Even if a circular structure is created between the first and third open grooves in the Y direction,
In addition, a first groove (16) is formed inside the semiconductor (3) in a comb shape.
) may be formed along the lines.

The above description is not limited to the pattern of FIG. 2 of the present invention. The number and size of cells are determined by the design specifications. Further, for the semiconductor, a plasma Cv1) method, a low pressure CVD method, a photo CVD method, or a photo plasma CVD method was used.

Applications are possible not only to gate-N junctions, heterojunctions, and tandem junctions, but also to many structures that are mainly composed of non-single-crystal silicon.

In addition, the present invention has a structure formed on an organic resin. However, in the present invention, an insulating film of silicon nitride, silicon oxide, or chromium oxide is formed on a thin metal layer to a thickness of 300 to 3000 mm, or the substrate itself has an insulating surface (non-transparent). Needless to say, a conductive thin film for the first electrode may be formed thereon.

In addition, in the drawings, a method may be used in which the substrate is made translucent, for example, made of glass or a translucent plastic such as PES, the first electrode on the lower side is made of CTF, and light is irradiated from the side. Needless to say.

[Brief explanation of drawings]

FIG. 1 is a longitudinal sectional view of a conventional photoelectric conversion device. FIG. 2 is a cross-sectional view and a vertical cross-sectional view of the photoelectric conversion device of the present invention, taken along the manufacturing process. FIG. 3 shows the change in electrical resistance due to laser scribing when the conductive thin film on the organic resin of the stainless steel substrate of the present invention is laser scribed. FIG. 4 shows the situation when the scanning speed of the present invention is kept constant. Patent issued person〈9 zu・no (/?I-himee-Ck Hz)
f-, 5(@L-BooeηiLfCc,su() 4ta

Claims (1)

[Claims] 1. A step of forming a plurality of first electrodes by linearly scanning and scribing a first conductive film formed on a substrate having an insulating surface with a laser beam; forming a non-single crystal semiconductor that generates a photovoltaic force by light irradiation on a second electrode and a first groove between the electrodes, and forming the linear first groove in the non-single crystal semiconductor; forming a second groove or groove in the non-single crystal semiconductor by scanning a laser beam parallel to Form a conductive film-
(A plurality of third electrodes are formed and integrated by scanning the laser beam in a straight line parallel to the first opening and scribing the second conductive gland that constitutes the connecting part. 2. In claim 1, when scribing is performed by linearly scanning a laser beam, the scanning speed is constant or approximately constant. A method for manufacturing a photoelectric conversion device characterized by the following.