WO1992010003A1 - Method for manufacturing thin silicon solar cell - Google Patents

Abstract

The surface of a silicon wafer is exposed periodically through very small openings of an insulation layer provided on the surface. In this state, angular lumps of single-crystal silicon are generated near the opening parts by a selective epitaxial growth and a lateral growth, and are grown until they are collided with each other. The insulation layer is removed through gaps left between the angular lumps, and a resin is embedded in the gaps. After electrode layers are formed on the surfaces of the angular lumps, the surfaces are fastened on a base board via a resin. Then, the angular lumps of single-crystal silicon are separated from the silicon wafer, and reverse electrodes are formed on them.

Description

 Specification

 Manufacturing method of thin silicon solar cells

 Field of the invention

 The present invention relates to a method for producing a 蘀 -type silicon solar cell having improved photoelectric conversion efficiency. Background of the Invention

 In various devices, solar cells are used as a driving energy source.

 A solar cell uses a Pn junction for a functional part, and silicon is generally used as a semiconductor constituting the pn junction. It is preferable to use single-crystal silicon in terms of the efficiency of converting light energy into electromotive force, but amorphous silicon is advantageous in terms of large area and low cost. .

 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, with the conventionally proposed method, it is difficult to reduce the thickness to 0.3 «or less, because a block-shaped polycrystal is sliced into a plate-shaped body, which is used. Therefore, the thickness was more than necessary to sufficiently absorb the light quantity, and in this respect, the effective use of the material was not sufficient. In other words, it is necessary to make it thin enough to reduce the cost.

 Therefore, attempts have been made to form a thin film of crystalline silicon using a thin film forming technique such as chemical vapor deposition (CVD), but the crystal grain size is only several hundredths of a micron at most. Energy conversion efficiency is lower than that of the bulk polycrystalline silicon slice method.

Attempts have also been made to increase the crystal grain size by irradiating the polycrystalline silicon thin film formed by the above-mentioned CVD method with laser light to melt and recrystallize it, but the cost reduction is not sufficient. In addition, stable production is difficult It is.

 Such a situation is a common problem not only in silicon but also in compound semiconductors.

 By the way, US Pat. No. 4,816,420 discloses a method for producing a thin crystalline solar cell. The method of manufacturing a solar cell disclosed in this U.S. patent is to selectively form a sheet-like crystal on a crystal substrate through a mask forest by a selective epitaxial growth method and a lateral growth method, and then separate the crystal from the substrate. It is characterized by the following.

 In this method, the opening provided in the mask material has a line shape. Then, in order to separate the sheet-like crystal grown by selective epitaxial growth and lateral growth from this line seed, the crystal is mechanically peeled off by utilizing the cleavage of the crystal. Will be · However, if the shape of the line seed is a certain size or more, the ground contact area with the substrate will increase, which may cause breakage of the sheet crystal during the peeling. Often there. No matter how narrow the line width is, especially when increasing the area of solar cells (actually,

However, if the line length is several centimeters to several centimeters or more, it becomes a fatal problem.

Also in the aforementioned US patent is shown an example of growing a silicon co down thin film by selective Ebitakisharu growth and lateral growth at a substrate temperature in 1 0 0 0 using S i 0 ₂ as the mask material but substantial stacking faults (surface to silicon co down layer side in silicon co emission layer ZS i 0 ₂ near the interface Te cowpea for reaction with silicon co emission layer and S i 0 ₂ to grow in this high temperature It is generally well known that defects are introduced. Such defects have a great adverse effect on the characteristics of the solar cell. Conversely, it has been reported that the lower the growth temperature, the lower the defect density. CH.K itajima, A.I shitani, N.Endo and K.Tano, Jpn.J, AppK P hys. 2 2> L 7 8 3 (1 9 8 3)] In this case, the growth rate drops sharply, so it is necessary to form solar cells within a practical time. It is extremely difficult to obtain film thickness

 A main object of the present invention is to eliminate the above-mentioned problems in the prior art and to provide a method capable of efficiently manufacturing a high-quality thin silicon solar cell.

 Another object of the present invention is to provide an improved solar cell manufacturing method which enables provision of an inexpensive solar cell by transferring a thin semiconductor layer formed on a silicon wafer to a substrate such as a glass substrate. Here it is.

 A further object of the present invention is to reduce the defects in the grown crystal near the interface of the grown crystal layer in the selective crystal growth through the insulating layer, thereby enabling efficient production of high quality solar cells. It is to provide a way to do this.

 The present invention has been accomplished as a result of diligent research by the present inventors to solve the above-mentioned problems in the conventional technology and to achieve the above object, and to provide a thin silicon solar cell having good characteristics. It is in a way to enable efficient manufacturing.

 The method for manufacturing a solar cell according to the present invention has the configuration described below. That is,

 (i) forming an insulating layer on a silicon wafer;

 (iii) a step of periodically providing minute openings in the insulating layer to expose the silicon surface;

 (iii) a step of performing growth until the mountain-shaped silicon single crystals generated in the opening by the selective epitaxial growth and the lateral growth collide with each other;

 (iv) removing the insulating layer through a void remaining between the single crystal bodies;

(V) a step of filling the void with a resin, (vi) after forming an electrode layer on the surface of the mountain-shaped single crystal, fixing the surface to a substrate via a resin;

 (vS) a step of separating the mountain-shaped single crystal and the silicon wafer,

 (viii) the steps of (i) to (viS) of forming a back electrode on the separated mountain-shaped 单 crystal,

And, after growing the mountain-shaped single crystal, the surface defect density in the vicinity of the interface with the insulating layer is 1 × 10 ⁴ cm ⁻¹ or less, or the ώ-type single crystal is This is a method for manufacturing a solar cell in which an impurity is added in an initial stage of growth in a growth step.

 According to the present invention, a polycrystalline thin-film solar cell having high conversion efficiency formed on a wafer can be transferred to a support such as glass, and thereby a low-cost, high-quality thin solar cell can be mass-produced. Detailed description of preferred embodiments

 The technical gist of the present invention is a selective epitaxy growth and a non-nucleation surface formed on a silicon wafer and a silicon / cone seed region formed on a silicon wafer shown in FIG. Subsequent lateral growth is to form a polycrystalline silicon thin film, which is a collection of single crystals having the same crystal orientation and size (grain size).

Here, the general principle of the selective growth method is briefly described. Selective epitaxial growth refers to the oxidation formed on a silicon wafer, as shown in Fig. 2 (Α) and Fig. 2 (Β), when performing epitaxial growth using vapor phase epitaxy. Selective crystal growth in which the exposed silicon surface in the opening provided in the insulating layer is used as a seed crystal under the condition that nucleation does not occur on an insulating layer such as a film, and epitaxial growth is performed only in this opening. Is the law. If the epitaxial layer filling the opening continues to grow further, the crystal layer grows in the lateral direction along the surface of the insulating layer while continuing to grow in the vertical direction. This is called the lateral growth method (Epitaxial Lateral Overgrowth). When the growth continues in the horizontal direction, the silicon crystals grown from different closed parts eventually come into contact with each other to form a continuous film. The appearance of the vertical to lateral growth ratio / facet at this time generally depends on the growth conditions, the plane orientation of the substrate, the thickness of the insulating layer, and the shape of the opening provided on the insulating layer.

 As a result of repeated experiments, the present inventors have found that the size of the opening is a small area of several meters or less, so that the growth ratio in the vertical direction to the horizontal direction is independent of the thickness of the insulating layer. Nearly one, three-dimensional crystal growth on the insulated layer, a clear facet appears, and a mountain-shaped single crystal is obtained, and these contact to form a continuous film (Figs. 2 and 3).

 At this time, before completely forming a continuous film, that is, in a state in which a gap remains between adjacent mountain-shaped single crystals, it was found that the insulating film could be almost completely removed by etching through this gap. . As a result, it was found that a polycrystalline thin film, which is an aggregate of the above-mentioned single crystal bodies, can be easily separated from a silicon wafer.

The present inventors have further conducted experiments and found the following. That is, as shown in Fig. 4 (A), when the insulating layer on the silicon wafer was Sioz, a large number of stacking faults were found near the interface with the insulating layer around the opening. by the early growth stage by controlling the growth temperature during growth raises the temperature subsequently lowering the temperature or by changing the insulating layer from the S i 0 ₂ to S i ₃ N, 4 respectively (B) Figure and the 4 (C) As shown in the figure, it was found that it could be greatly reduced. In particular, in the case of FIG. 4 (B), a continuous thin film with few defects at the interface can be obtained without significantly lowering the effective growth rate on Si0z.

The present inventors have found out the following as a result of further experiments. That is, as shown in FIG. 4 (D), impurities are intentionally introduced into the defects shown in FIG. It has been found that a good semiconductor layer can be obtained by inactivating a high-quality region, growing a high-quality crystal layer with few defects thereon, and using this as an active region. Based on the findings, further research has led to the completion of the present invention.

 Hereinafter, an experiment performed by the present inventors will be described in detail. Experiment 1 (selective crystal growth)

As shown in FIG. 2 (A), a thermal oxide film is formed on the surface of a (1 & 0) silicon wafer 201 having a thickness of 500 m as a thermal oxide film 100A as an insulating layer 202. Etching was performed using photolithography, and square openings having a side of a in the arrangement shown in FIG. 3A were provided at intervals of b = 50 m. Here, three types of openings were provided for the value of a: Ϊ.2 m, 2 <“ΐη, and 4 m. Next, selective crystal growth was performed using a normal low-pressure CVD apparatus (LPCVD apparatus) shown in FIG. Si H ₂ C £ ₂ was used as the source gas, and H ₂ was added as the carrier gas, and HC £ was added to suppress the generation of nuclei on the oxide film of the layer 202. Table 1 shows the growth conditions at this time.

 After growth was completed, the surface of the wafer was observed with an optical microscope, and as shown in Fig. 2 (C) or Fig. 3 (B), the particle size was about 20 m for any value of a. 32 (3 03) with a mountain-shaped facet are regularly arranged on lattice points at intervals of 50 m, and the opening 3 0 1 determined in Fig. 3 (A) It was confirmed that the selective crystal growth was performed according to the pattern shown in FIG. At this time, the ratio of the crystal grown to the opening was 10% for any value of a. In addition, the proportion of the grown single crystal that clearly shows the surface facets without being deformed depends on the value of a. As shown in Table 2, the smaller the a is, the smaller the rate of destruction is.

The obtained 单 crystals all have the same orientation, indicating that the crystal orientation is accurately inherited from the silicon wafer that is the substrate. Was

Next, under the conditions shown in Table 1, the growth time was further increased to 90 ni in, and the wafer surface was observed with an optical microscope after growth was completed. As shown in Fig. 2 (D) or Fig. 3, (C) As shown in the figure, for any value of a, adjacent single crystal bodies 203 (303) are completely in contact with each other, and have a grid-like shape when viewed from above the substrate. It was confirmed that a polycrystalline thin film composed of a collection of ordered single crystals was obtained.The height of the single crystal was about 40 from above the oxide film 202. 3 (C) The polycrystal is cleaved along the line BB 'shown in the figure, the cross section is polished, the defect is revealed by Secco etching, and the vicinity of the grain boundary is observed using a scanning electron microscope. As a result, no clear interface was observed as in the case of ordinary polycrystalline silicon having non-uniform crystal orientation. However, near the interface with the oxide film, it was observed that the crystal contained many defects as shown in Fig. 4 (A). The surface defect density at this time was about 7 × 10 ⁴ cm ⁻¹ .

 As shown in FIG. 3 (C), the obtained polycrystal has a gap 304 of about 8 m square between adjacent single crystal bodies 303, and the oxide film 3 serving as an insulating layer is formed. It was observed that 0 2 was exposed. It was confirmed by another experiment that the void 304 was completely eliminated by further increasing the growth time. Experiment 2 (Removal of insulating layer)

An attempt was made to remove the insulating layer under the polycrystalline thin film using the above-mentioned voids. The silicon wafer with the grown polycrystalline thin film with voids obtained in Experiment 1 was added to a 49% aqueous HF solution. Let it soak for hours. Thereafter, the wafer surface was washed and then dried, and the wafer surface was observed with an optical microscope and a scanning electron microscope. As a result, no oxide film was found in the voids on the polycrystalline thin film. Therefore, an epoxy resin is applied to the surface of the polycrystalline thin film, a glass plate is attached as a support, and after the epoxy is cured, When trying to separate the glass plate from the wafer, it became clear that the polycrystalline thin film was easily separated from the wafer, and the entire crystalline film was transferred directly onto the glass plate. Observation of the rear surface of the divided polycrystalline thin film with an optical microscope and a scanning electron microscope revealed that although a portion of the oxide film remained, most of the oxide film was etched. The silicon crystal was exposed. Experiment 3 (Control of defect density by growth temperature)

Defect control was attempted by controlling the growth temperature during crystal growth. Defect density at S i 0 ₂ near the interface enters the crystal grown on S i 0 ₂ as described above by connexion changes to temperature during crystal growth, as the growth temperature is low, it becomes the defect density is low H.K itajima, A.I shitani, N.Endo and K.Tanno, J pn.J.A ppl.Phys.22, L783 (1983)) . However, if the growth temperature is lowered, good quality crystals can be obtained, but on the contrary, the growth temperature drops sharply, making it difficult to grow and contact the mountain-shaped polycrystal to form a polycrystalline film.

Therefore in view of the configuration when the solar cell, by this changing the growth temperature in two steps, substantially crystalline without reducing the growth rate too much S i 0 ₂ near the interface to be柽路photocurrent An experiment was conducted to determine whether a crystal with a low defect density could be obtained in the region. Crystal growth was performed using the same substrate as in Experiment 1 and continuously changing the growth temperature to two steps under the conditions shown in Table 3, and the results are shown in Figs. 2 (D) and 3 (C). Such a continuous polycrystalline film was obtained. The polycrystal was cleaved along the line B-B 'shown in Fig. 3 (C), and after cross-section polishing, defects were revealed by Secco-notching and observed with a scanning electron microscope. In the vicinity of the interface with the oxide film, the number of defects around the opening was significantly reduced as shown in Fig. 4 (B), and was about 1 X 10 ³ cm- ¹ . Continuous polycrystal with few defects near the Sioz interface by continuously growing the crystal in two stages Experiment 4 (a defect density on Si ₃ N ₄ ) found to obtain a silicon film

Because reaction between S i 0 ₂ and S i crystals is considered to be one of the triggers of the crystal defects, the experiments using S i ₃ N ₄ instead of the S i 0 ₂ as an insulating layer lines ivy.

S i ₃ N ₄ is deposited from the thermal decomposition of S i H _Z C £ ₂ and NH ₃ on silicon co N'weha using conventional LPCVD apparatus, the film thickness was set to 1 0 0 0 A, in Experiment 1 The crystal growth was performed under the conditions shown in Table 1 in the same manner as in Example 1.The defect density in the crystal near the interface with the Si ₃ N ₄ layer was examined in the same manner as in Experiment 3. As shown, it decreased from the case of S i 0 z and was about 1 × ^{10 4} cm ⁻¹ .

Thus, it was found that the defect density was improved by replacing the insulating layer with Si ₃ N ₄ . Experiment 5 (Reduction of defects by growth temperature change + Si ₃ N)

Further, based on the results of Experiments 3 and 4 above, crystal growth was performed in the same manner as in Experiment 3 under the formation conditions in Table 3 using Si ₃ N ₄ for the insulating layer. The defect density in the crystal near the interface with the Si ₃ N ₄ layer was examined for the grown continuous polycrystalline film in the same manner as in Experiments 3 and 4.

4 (B) reduction was observed as shown in FIG, low had values were obtained of about X ¹ 0 ² cm 1. Experiment 6 (Reduction of the effect of defects due to impurity addition)

 At the initial stage of crystal growth, we tried to form an inactive region by adding impurities and use a high-quality crystal layer with few defects grown on it as the active region.

In order to minimize the influence of defects and obtain good device characteristics, impurities are added at the initial stage of crystal growth to form an interface with an oxide film with many defects. By inactivating a nearby region, then stopping the introduction of impurities and growing a crystal over the region containing the impurities, a nod region with few defects was provided, and this was used as an active layer.

A substrate was fabricated in the same manner as in Experiment 1 using an Sb-doped (100) silicon wafer (P = 0.02 Ω · αη). Crystal growth was performed in two stages under the conditions shown in Table 4 to obtain a continuous polycrystalline film as shown in FIG. 2 (D) or FIG. 3 (C). The cross-sectional structure of the crystal grown at this time was as shown in Fig. 4 (D). Au was vacuum-deposited on the surface of the obtained polycrystalline film to a thickness of 200 people to form a Schottky barrier, and the reverse saturation current was measured. Table with polycrystalline film of the conditions to grow compared to those forming the shots Toki one barrier as well as re-binding current due to defects is reduced, 1 0 ² to 1 0 ³ times in the value of all the a It showed a low current value.

It was found that the influence of defects near the SiO ₂ interface can be reduced by introducing impurities into the initial stage of crystal growth by dividing the growth conditions into two stages. Experiment 7 (reduction of the impact of defects due to impurity addition + S i ₃ N ₄₎ Further, by using the S i ₃ N ₄ in absolute緣層based on the results of the above experiments 4 and 6, experiments formation conditions shown in Table 4 Crystal growth was performed in the same manner as in 6. A Schottky barrier was formed on the grown polycrystalline film in the same manner as in Experiment 6, and the reverse saturation current was measured. The characteristics obtained in Experiment 6 were 5 to 10 times higher. The current value was even lower. Experiment 8 (Formation of solar cell)

 Solar cell characteristics were evaluated using low defect crystal thin films.

In the same manner as in Experiment 3, using a Sb-doped (100) silicon wafer (p = 0.02 Ωcm), a polycrystalline silicon Is grown, and B is ion-implanted on the surface. Implantation in 2 0 K e V, 1 1 0 1 5 011- 2 conditions, at 8 0 0, formed on the surface of the polycrystal P + layer and Aniru at 3 0 min. The photovoltaic cell with the P polycrystalline silicon 3 1 0 ₂ 1 ⁺ (1 0 0) S i structure fabricated in this manner was exposed to AM I.5 (100 mW / ci) light. — When the V characteristics were measured, the open circuit voltage was 0.42 V at the cell area of 0.16 ca !, the short-circuit photocurrent was 2 2 mA / cn !. The fill factor was 0.67, and a = 4 / A conversion efficiency of 6.2% was obtained at m. It was shown that a good solar cell can be formed using the polycrystalline silicon thin film formed on the insulating layer in this way. As described above, the present invention, which has been completed based on the experimental results described above, opens a part of the insulating layer formed on the silicon wafer, and selectively forms the mountain-shaped single-crystal silicon body by selective evitaxical growth. The present invention relates to a method for manufacturing a solar cell by growing a capacitor and separating a polycrystalline thin film composed of the aggregate.

 FIG. 1 shows the configuration of a solar cell manufactured by the method of the present invention.

 A P + layer 103 is formed on the upper surface of the polycrystalline silicon composed of a collection of the single-crystal silicon 102, and the space between the adjacent single-crystal silicon 102 is formed. The voids are filled with an epoxy resin (not shown). The transparent electrode 104 also serving as an anti-reflection film and the current collecting electrode 107 are provided on the <P + layer 103. Further, the whole polycrystalline silicon is fixed by a transparent support 106 such as glass via an ethoxy resin 105 thereon. On the lower surface of the polycrystalline silicon, a back electrode 101 is formed.

 Next, a method for manufacturing a solar cell according to the present invention will be described in accordance with the process shown in FIG.

Using a Sb-doped (100) silicon wafer (P = 0.02 Ω · cm), a single-crystal silicon wafer as shown in Fig. 2 (D) was obtained in the same manner as in Experiment 1. To form a polycrystalline silicon 503. FIG. 2 (D) is a cross-sectional view taken along line B--B 'in FIG. 3 (C). However, for ease of explanation, cross-sectional views taken along line AA ′ are shown in FIGS. 5 (A) to 5 (H). FIG. 5 (A) shows a cross section in a direction in which the gap 504 is visible in FIG. 2 (D). Next, after forming a P ⁺ layer on the surface by ion implantation in the same manner as in Experiment 8, the oxide film 502, which is an insulating layer, was removed by etching with an aqueous HF solution through the void 504. FIG. 5) Epoxy resin 505 is applied to the polycrystalline surface to fill voids 504 [FIG. 5 (C)]. After the resin has hardened, only the resin on the polycrystalline surface is removed by plasma etching, and the resin is left in the voids (Fig. 5 (D)). A transparent electrode 506 and a final electrode 507 are sequentially deposited on the exposed polycrystalline surface [FIG. 5 (E)]. Further, an epoxy resin 511 is applied again from above, and the polycrystalline surface is fixed by a support 508 such as a glass plate (FIG. 5 (F)). After the resin is cured, the polycrystalline silicon layer 503 is separated from the silicon wafer 501, and the oxide film remaining on the back surface of the polycrystalline silicon layer 503 is completely removed again with an HF aqueous solution. FIG. 5 (G)], and an electrode 509 is formed on the back surface of the polycrystalline silicon layer 503 by vapor deposition to complete the process [FIG. 5 (H)].

As the insulating layer used in the solar cell of the present invention, a forest material whose nucleation density on the surface is considerably smaller than that of silicon in terms of suppressing nucleation during selective crystal growth is used. Can be For example, _Sioz , SN, etc. are used as representatives.

Selective crystallization is the growth of the raw material gas S i E _Z C £ z used in the present invention, S i C, S i HC , S i H 4, S i 2 H 6, S i H 2 F _2, S i _z F silanes such as 6 and halogenated sila emissions such may be mentioned as representative.

Further, H ₂ is added as a carrier gas or in addition to the above-mentioned raw material gas to obtain a reducing atmosphere for promoting crystal growth. The ratio of the amount of the raw material gas to the amount of hydrogen is appropriately determined as desired depending on the forming method, the type of the raw material gas, the material of the insulating layer, and the forming conditions. In general, the ratio of the introduced flow rate is preferably from 1:10 to 1: 100, and more preferably from 1:20 to 1: 800.

 In the present invention, HC is used for the purpose of suppressing the generation of nuclei on the insulating layer. However, the amount of HC £ added to the source gas depends on the formation method, the type of the source gas, the material of the insulating layer, and the formation conditions. The ratio is appropriately determined as desired by the method described above, but is preferably from about 1: 0.1 to 1: 100, and more preferably from 1: 0.2 to 1:80.

The temperature and pressure at which the selective crystal growth is performed in the present invention vary depending on the formation method, the type of the source gas to be used, and the formation conditions such as the flow rate ratio between the source gas and _Hz and HC. For example, in an ordinary LPCVD method, it is appropriate to set the value to about 60 O'c or more and about 125 to below, and it is more preferable to control the value to more than 65 to about 1200. In particular, in the case of using the insulating layer Si 0 ₂ , it is preferable to set the thickness to 850 to 950 in order to suppress defects at the initial stage of crystal growth and to maintain a growth rate of a certain level or more. Similarly approximately 1 for Pressure ^{0 - 2 ~ 7 6 0 T} orr are suitable, more preferred properly ^{1 0 - 1 ~ 7 6 0} T range orr is desirable.

Also intended as a support for fixing the polycrystalline thin film used for the solar cell of the present invention have a light transmission property because it becomes a light incident surface is used, for example, _{glass, S i 0 2, S i} 3 N * Films or sheets of synthetic resin such as ceramics, polyimide, polycarbonate and boriamid are used.

In the present invention, the resin used for filling the voids between the single crystal bodies should be selected and used in terms of heat resistance, electrical insulation, adhesiveness, mechanical strength, dimensional stability, corrosion resistance, and weather resistance. Is preferred. Among such resins, the resin used in the present invention is in a liquid state, and is preferably a curable resin which produces heat curing, ultraviolet curing, electron beam curing, or the like. like that U

 Specific examples of the resin include an epoxy resin, a diglycoldialkyl-carbonate resin, a polyimide resin, a melamine resin, a phenol resin, and a urea resin.

 Hereinafter, an example of resin injection using an epoxy resin will be described, but the resin used in the present invention is not limited to the epoxy resin.

 Novolac type epoxy resin epoxy coat 15 4 [product name · made by Yuka Shell Epoxy Co., Ltd.] was used as the epoxy resin, and the polyfunctional solid acid anhydride YH-3 was used as a curing agent and a reactive diluent. After mixing 8H [trade name: Yuka Shell Epoxy Co., Ltd.] and alkyl monoglycidyl ether BGE [tradename: Yuka Shell Evoxy Hungary], perform resin casting under reduced pressure. The device shown in FIGS. & Is a device in which a resin injection nozzle 806 is provided in a processing container capable of evacuation under reduced pressure, and the resin can be cast under reduced pressure. A resin injection nozzle 806 having a structure similar to that of a commonly used injection molding machine can move on a rail 803 together with a resin tank 804. Further, a constant temperature heater 805 is provided in parallel with the nozzle 806 and the resin tank 804 so that the temperature of the resin can be kept constant. Further, the apparatus shown in the figure is provided with a squeegee moving means 8 12 for moving a squeegee (blade) 8 11 in order to spread the resin applied on the crystal evenly.

In order to cast the resin using the apparatus, first, the substrate 809 on which the single crystal 807 was grown was put into the decompression vessel 801, and this was placed on the support 808. And set the heating heater 802 disposed in the support base 808 to 1 * 0 * C. Subsequently, the constant temperature heater 800 was set to 100, and the epoxy resin ヱ Picocoat 154, the curing agent YH-308, and the reactive diluent BGE were added in a weight ratio of 100: 70. : After filling in the resin tank 804 at a ratio of 20 and leaving it to stand, then use a vacuum pump 813 to depressurize and exhaust the inside of the decompression vessel 801 to a pressure of 10 — ³ Τ ΟΤΓ . When the resin in the resin tank 804 became uniform The resin is discharged from the resin nozzle 806 onto the crystal 807 with the rollers, and the resin on the crystal 807 is distributed with a thickness of about 1 m using the squeegee moving means 812. Spread the resin. After the resin spreads, gradually open the pressure regulating valve 810 to atmospheric pressure in the vacuum vessel 801, and leave it for 5 hours while keeping the temperature of the heater 800-2 at 100 Then, the temperature of the heating heater 800 is raised to 200 and left for 6 hours. Here, when the substrate 809 is taken out of the vacuum container 801 and touched with the resin on the single crystal, it is found that the resin is strong.

The single crystal constituting the polycrystalline thin film formed by the method of the present invention can form a junction by doping with an impurity element during or after crystal growth. As the impurity element used, an element of the HI group A of the periodic table, for example, an element of B, A, G a, In, or the like is preferably cited as an impurity of the type. Elements of Group A of Table V, for example, P> As, Sb, Bi and the like are preferable, but B, Ga, P, Sb and the like are particularly preferable. The amount of the impurity to be doped is appropriately determined according to the desired electrical characteristics. As a substance containing such an impurity element as a component (impurity introduction substance), it is preferable to select a compound which is in a gaseous state at normal temperature and normal pressure or a compound which can be easily vaporized by a suitable vaporizer. Such compounds include Ρ Η ₃ , P ₂ H 4, PF a, PF ₅ , PC £ _{3> As} H a, As F 3, As F 5, As C, S b H ₃ , _{S b F 5, BF a,} BC, BB ra, B Z H 6, B 4 H, 0, B 5 H,, B 5 H n, B 6 H,. , B ₆ H ₁₂ > A £ C £ ₃ and so on. The compounds containing impurity elements may be used alone or in combination of two or more.

When performing the selective crystal growth method used in the method of manufacturing the solar cell of the present invention, the shape of the opening provided in the layer is not particularly limited, but a square, a circle, and the like are typical examples. . As for the size of the opening, the facet of the growing mountain-shaped single crystal was as shown in Experiment 1. The opening tends to collapse as the size of the opening increases, that is, the crystallinity tends to deteriorate, and is preferably several meters or less in order to suppress collapse of the facets and facilitate separation. In practice, since it depends on the pattern accuracy of photolithography, a is appropriate when the shape is a square, from 1 to 5 m. In addition, the interval b where the opening is provided is such that the unevenness of the surface of the polycrystalline silicon is 2 m or more and 40 or less in consideration of the light reflectance and the ease of electrode formation in the above-mentioned mountain facet. It is appropriate to set the distance between 10 m and 200 m so that

 The layer structure of the solar cell according to the method of the present invention is not particularly limited, and can be applied to any structure such as a Schottky type, a MIS type, a pn junction type, a pin junction type, a hetero type, and a dandem type.

 Hereinafter, the formation of a desired solar cell by the method of the present invention will be described in more detail with reference to examples, but the present invention is not limited to these examples. Example 1

 As described above, a pn-type solar cell consisting of a collection of single-crystal silicon with a uniform crystal orientation as shown in Fig. 1 or Fig. 5 (H) was fabricated in the same manner as in Experiment 3. did. FIGS. 5A to 5H show the manufacturing process. The substrate was a P-doped (100) silicon wafer (ρ = 0.05 Ω · αιι). PSG (2000 A) is deposited instead of the thermal oxide film and used as an insulating layer 502, the opening is square, and one side is a = 1.2 m, 2 um, 4 111ゎ = 50 m intervals, selective crystal growth was performed using a normal LPCVD apparatus under the conditions shown in Table 3, and single crystal silicon 5 with the same crystal orientation as shown in Fig. 3 (C) A polycrystalline silicon thin film consisting of a collection of 03 was formed on a silicon wafer [Fig. 5 (iv)].

At this time, Ρ is simultaneously diffused from the PSG, and the backside of the polycrystalline silicon An n + layer is formed on the side (not shown).

Implantation more 2 0 Ke V, 1 X 1 0 15 cm_ 2 conditions To polycrystalline sheet re co down surface b on-implanted B in the same manner as in Experiment 8, 8 0 0 and Aniru at 3 0 min Thus, a P + layer 5 10 was formed on the surface of the polycrystal. When ion implantation is performed, the substrate is implanted at an angle several degrees off from the direction perpendicular to the substrate so that the ion does not reach the peripheral portion of the crystal in contact with the insulating layer in the gap formed between the single crystals. (This avoids leakage between the back surface electrode and the joint when forming the back electrode.)

Next, the wafer is immersed in a 49% strength HF aqueous solution for 24 hours, and the PSG 502 is removed by etching through a gap 504 of about 8 / _m square, and the wafer is washed with running water and dried. Evoxy resin 505 was applied to the surface of the polycrystal to fill voids 504. Novolac Epoxy Resin Ebiko-To 15 4 [trade name: manufactured by Yuka Shirekiboshi] was used as the epoxy resin, and this was used as a curing agent and reactive diluent. Product YH-308H [trade name: Oil-based Shellepoxy Co., Ltd.] and Alkyl monodalicydyl ether BGE [trade name: Yuka Shell-Evoxy Co., Ltd.] in a weight ratio of 100: 70: 20. After mixing, the resin was cast under reduced pressure. The resin was cast using a device as shown in Fig. 8, and the resin was discharged from the resin injection nozzle 806 onto the polycrystalline surface 807, and the resin was spread evenly by the squeegee 811. After curing the resin in 1 0 0 5 hours, 0 ₂ plasma etch (0 ₂ flow: 5 0 0 sccm / min, pressure: 1 Torr, high-frequency discharge power over: 2 0 0 W) by an air gap 5 0 4 The polycrystalline surface was completely exposed, leaving only the portion of the resin, and then the resin was cured at a temperature of 250 for 3 hours (FIGS. 5 (B) to 5 (D)). Electron beam evaporation of a transparent conductive film IT 0506 was performed on the entire surface of the polycrystalline surface, and Cr was vacuum-deposited thereon as a current collecting electrode 507 by 1 m [Fig. 5 (E)]. Apply epoxy resin 511 again to the electrode surface in the same manner as above, and After the entire polycrystalline silicon film is fixed by the support 508 and the resin is cured, the polycrystalline silicon, which is an aggregate of the single crystal 503 from the silicon wafer 501, is formed. The layers were separated, and a portion of the PSG remaining on the back surface was etched with a 10% aqueous HF solution. Finally, a back electrode 509 was formed on the back surface of the separated polycrystalline silicon layer by attaching 300 Cr by vacuum evaporation (Figs. 5 (F) to 5 (H)). . The IV characteristics of the obtained Pn-type solar cell were measured under AM 1.5 light irradiation. The cell area at this time was 0.16 αί. Table 5 shows the results of comparing the characteristics with those of a solar cell fabricated in the same manner using a polycrystalline silicon film obtained when the growth time was set to 90 rain under the conditions shown in Table 1.

 In any case, the polycrystalline solar cell formed by the two-step growth of the present invention has higher characteristics than the polycrystalline solar cell formed by a single substrate temperature of 130 ° C. I have. The reason that the characteristics tend to deteriorate as a becomes larger is thought to be due to the collapse of the facet during crystal growth and, consequently, the increase in defect levels resulting from the decrease in crystallinity. Example 2

 In the same manner as in Example 1, an amorphous silicon carbide Z-polycrystalline silicon hetero solar cell was manufactured. Using a P-doped (100) silicon wafer (P-0.05 Ω · α) as the substrate, a PSG film with a thickness of 300 was deposited by atmospheric pressure CVD. The opening was a-1.2 m in size. The polycrystalline silicon thin film is composed of a collection of single-crystal silicon with a uniform crystal orientation, grown by selective LP under the conditions shown in Table 6 by the ordinary LPCVD method. At this time, the size of the voids remaining between the single crystals was about 6;

FIGS. 6 (A) to 6 (H) show the process of the fabricated hetero-type solar cell (almost the same as the case of FIG. 5 shown in Example 1). Thus, in (B), a p-type amorphous silicon carbide 610 is formed on the polycrystalline silicon instead of the p + layer 510).

The p-type amorphous silicon force-byte layer 610 was deposited on a polycrystalline silicon surface at 100 A using a conventional plasma CVD apparatus under the conditions shown in Table 7. Dark conductivity of Amorufu § mortal Li co Nkabai preparative film at this time is ~ 1 0 - ² S 'is cm- ^1, the composition ratio in the film of C and S i is

2: 3.

 The transparent conductive film 606 was formed by electron beam evaporation of about 100 OA of ITO, and the collector electrode 607 was formed by vacuum evaporation of Cr by 1 m.

 The I-V characteristics of the amorphous silicon hetero-byte solar cell thus obtained under AM 1.5 light irradiation were measured (cell area 0.16 erf ), The open-circuit voltage was 0.52 V, and the short-circuit photocurrent was 2 O m AZdU. The fill factor was 0.56, and a high conversion efficiency of 5.8% was obtained. This is a result comparable to that of the conventional amorphous silicon carbide Z 单 crystalline silicon hetero-type solar cell. Example 3

In the same manner as in Examples 1 and 2, a pn-type solar cell in which impurities were added to the region near the interface as shown in FIG. 9 (H) was produced. FIGS. 9A to 9H show the manufacturing process. A P-doped (100) silicon wafer (p = 0.05 Ω.cm) was used for the substrate, and a thermal oxide film (100 A) was used for the upper layer 902. The parts were a = 1.2 ^ m and b = 50 m. Selective crystal growth was performed by the ordinary LPCVD apparatus shown in Fig. 7 under the conditions shown in Table 8, and single crystal silicon 903 having the same crystal orientation as shown in Fig. 3 (C) was obtained. A polycrystalline silicon thin film consisting of aggregates was formed on a silicon wafer [Fig. 9 (A)]. Hereinafter, a pn junction is formed by the same process as in Example 1. A release solar cell was fabricated.

 The I-V characteristics of the obtained n-type solar cell under AM 1.5 light irradiation were measured.The cell area was 0.16 αί, the open-circuit voltage was 0.47 V, and the short-circuit photocurrent was 24 m. The conversion efficiency was 0.71, and the conversion efficiency was 8%. The characteristics were further improved as compared with the case where no impurity was added in the initial stage of crystal growth (Example 1). Example 4

 In the same manner as in Examples 1, 2 and 3, a pin-type polycrystalline solar cell as shown in FIG. 10 (Η) was produced. The solar cell shown in FIG. 10 is one in which a pin junction is formed in a single crystal by adding a small amount of impurities to a source gas during the growth of the selective yarn ± i crystal. The important point here is that since the end face of the P / i junction is on the insulating layer 1002 at the portion where a gap is formed, the insulating layer is etched away to separate the obtained polycrystal from the wafer. If this happens, the joint end face will be exposed,

(Such as electrical isolation due to the formation of the back electrode) becomes severe. Therefore, as shown in Fig. 10 (A), an insulating layer is previously formed on the substrate 100 1 / S i 0 ₂ 1 0 0 2 / S i ₃ N ₄ 1 0 0 2 ' After that, a minute opening was formed by a reactive ion etching method, and a polycrystalline film was obtained by a selective crystal growth method.

_{S i 0 2, S i 3} N 4 are each normal pressure CVD method, and sedimentary by LPCVD, and the thickness of the film each 1 0 0 0 people was 3 0 0 0 A. The size of the openings was a-1.2 m and the spacing was b = 50 m. During the growth of the single crystal by the selective crystal growth method, the type and amount of impurities were changed, and the junction was formed by changing the conductivity type with niP in order from the substrate side. Table 9 shows the growth conditions.

In order to switch the impurities, PH ₃ is introduced until the growing single crystal becomes about 10 m to form an n-layer, then PH ₃ is stopped, and the single crystals contact each other. When the gap is about 6 m In the future, B ₂ H ₆ was introduced to form a P layer on the crystal surface at about 200 OA.

A solar cell having a P in junction was manufactured from the polycrystalline thin film obtained in this manner by a process as shown in FIGS. 10 (A) to 10 (H). As described above, the insulating layer has a two-layer structure. First, Si ₃ N ₄ 10 2 ′ is etched by reactive ion etching through the voids 10 4, and then a 49% concentration aqueous HF solution is used. Then, the Si 02 layer 1002 was removed. In this way, the end surface of the pin junction is protected by the Si ₃ N ₄ film, and even after separation, the n-layer 1003 ′ has a foot, so that it can be easily connected to the back surface electrode. All other processes were performed in the same manner as in Examples 1 and 2. Investigation of the IV characteristics of the P-in junction type polycrystalline solar cell fabricated through the above process under AM I.5 irradiation revealed that the cell area was 0.16 crf, the open voltage was 0.51 V, and the short-circuit photocurrent was The fill factor was 0.75, and a high conversion efficiency of 8.9% was obtained.

 As described above, according to the method for manufacturing a polycrystalline solar cell of the present invention, defects in a grown crystal near an interface between a grown crystal and an insulating layer in selective crystal growth are reduced, and a high-quality thin solar cell can be manufactured. It came to be.

 In addition, it is possible to separate from the wafer and fix it to a support such as glass using the voids remaining between the single crystals that are constituents of the polycrystalline silicon thin film. It is shown that mass production of inexpensive solar cells is possible.

1

 Gas flow ratio Substrate temperature Pressure Growth time

(1 / min) ("C) (Torr vmin)

SiH _z Cl ₂ / HCl / H ₂

1030 80 20 = 0.53 / 2/100 a (μm) 1.2 24 Facet rate (%) 96.4 93.1 84.6

 Four

a ί μm) 1.2 2 4 Conversion efficiency (%)

 7.6 7.3 7.1 950-1060

 Conversion efficiency (%)

 4.5 4.1 4.0 1030

Gas flow ratio Substrate temperature Pressure Growth time (l / min) (in) (Torr) (min;

 0.53 / 2.0 / 100 950 100 20

 1 i 1

0.53 / 1.6 / 100 950 100 40

I i 1 i

0.53 / 2.0 / 100 1060 100 80 Gas flow ratio Substrate temperature Pressure Discharge power

Sif / Cl-

0.5cc 8W at 0.8cc / 0.2cc 350

B ₂ H ₆ / SiH ₄ =

1.5 x 10- ²

図面の簡単な説明

BRIEF DESCRIPTION OF THE FIGURES

 FIG. 1 is a cross-sectional view of a Pn-type solar cell manufactured by the method of the present invention.

 FIG. 2 is a diagram illustrating a selective crystal growth method. FIG. 3 is a view for explaining a process in which a mountain-shaped crystal obtained by the method of the present invention grows three-dimensionally.

Fig. 4 shows how defects entering the crystal are reduced by the method of the present invention. FIG.

 FIG. 5 is an explanatory diagram of a manufacturing process of the Pn-type solar cell shown in FIG.

 FIG. 6 is an explanatory diagram of a manufacturing process of a hetero solar cell manufactured by the method of the present invention.

 FIG. 7 is a diagram schematically showing a general LPCVD apparatus.

 FIG. 8 is a diagram schematically showing an apparatus for performing resin casting.

 FIG. 9 and FIG. 10 are diagrams illustrating the steps of manufacturing a Pϋ type or pin type solar cell manufactured by the method of the present invention, respectively.

Claims

PCT / JP91 / 01657 25 Model of Thunder Blue 1. In a method of manufacturing a solar cell obtained by separating an epitaxial layer grown on a silicon wafer,  5 (i) a process of forming an insulating layer on a silicon wafer,  (iii) a step of providing minute openings periodically in the insulating layer to expose a silicon surface;  (iii) 10 steps of performing growth until the mountain-shaped silicon single crystals generated in the opening by the selective epitaxial growth and the lateral growth collide with each other;  (ίν) removing the insulating layer through voids remaining between the single crystal bodies,  (V) a step of filling the void with a resin,  (vi) after forming an electrode layer on the surface of the mountain-shaped single crystal, fixing the surface to the substrate via 15 resin;  (vii) a step of separating the mountain-shaped single crystal and the silicon wafer,  (vi) A method for manufacturing a thin silicon solar cell, comprising the steps of (i) to (viii) of forming a back electrode on the separated mountain-shaped single crystal. 20 2. The mountain-shaped single crystal body has, after growth, a region having a surface defect density of 1 × 10 cm ^-1 or less in the vicinity of an interface with an insulating layer at least in the vicinity of the minute opening. 2. The method for producing a thin silicon solar cell according to 1.  3. The method of manufacturing a thin silicon solar cell according to claim 1, wherein the selective epitaxial growth and the lateral growth are performed by a CVD method in which a substrate temperature is changed stepwise. 4. The method for manufacturing a thin silicon solar cell according to claim 1, wherein the insulating layer is Si ₃ N ₄ . 5. In the step of growing the mountain-shaped single crystal, an initial stage of growth 5. The method for producing a thin silicon solar cell according to claim 1, wherein impurities are added to the thin silicon solar cell.