JP2011192795A - Photoelectric conversion device, and method of manufacturing the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a photoelectric conversion device that has improved photoelectric conversion efficiency and stability by forming fine recesses, and to provide a method of manufacturing the photoelectric conversion device. <P>SOLUTION: The photoelectric conversion device includes a glass substrate 2 and a transparent conductive film 3 which covers at least a part of a principal surface of the glass substrate 2 and has an uneven shape on a surface on the opposite side from the substrate side. The photoelectric conversion device further includes a p layer 5 which covers at least a part of the uneven shape of the transparent conductive film 3 and has a first conductivity type, and an i layer 6 which covers the p layer 5. The uneven shape has projections of 50 to 1,200 nm in maximum height. The projection has on a surface, fine recesses in which intervals of local tops are 2 to 25 nm. The p layer 5 has a layer thickness larger at a part formed on the bottom of the fine recess than at a part formed on other than the bottom. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a photoelectric conversion device and a manufacturing method thereof.

  In a thin film solar cell, a transparent conductive film made of a translucent conductive material is formed on a substrate, and p-type, i-type, and n-type semiconductor layers are sequentially stacked on the transparent conductive film. In order to improve the power generation efficiency of a thin film solar cell, there are Patent Documents 1 and 2 as prior documents disclosing a thin film solar cell in which a concavo-convex shape is formed on a substrate surface or the like.

  In the solar cell substrate described in Patent Document 1, irregularities with an average step of 0.5 μm or less, preferably about 0.2 μm, are formed on the surface of an uneven glass substrate with an average step of 3 μm or more. This increases the light scattering effect. In the multi-junction thin film solar cell described in Patent Document 2, a hole having a diameter of 200 to 2000 μm and a depth of 50 to 1200 nm is provided on the surface of the transparent conductive film, and the height difference is about 10 to 300 nm (average) on the surface of the hole. 120 nm), and by forming irregularities with an interval of about 100 to 900 nm, the light confinement effect is improved.

  Further, Non-Patent Document 1 and Patent Document 3 are prior art documents disclosing the problem of reduction in the open-circuit voltage of a photovoltaic device due to the inability to uniformly form a p-type semiconductor layer on a substrate having an uneven shape.

  Further, for the purpose of improving the ohmic characteristics between the transparent conductive film and the p-type semiconductor layer, or for the purpose of reducing reduction deterioration of the transparent conductive film when forming the p-type semiconductor layer on the transparent conductive film, Patent Documents 4 to 8 and Non-Patent Document 2 are prior art documents that disclose a method for manufacturing a thin film solar cell that is subjected to plasma treatment such as hydrogen plasma.

  In the method for manufacturing a thin-film solar cell described in Patent Documents 4 to 7 and Non-Patent Document 2, the transparent conductive film is subjected to hydrogen plasma treatment. In the method of manufacturing a photovoltaic element described in Patent Document 8, noble gas plasma treatment is performed.

JP-A-10-70294 JP 2002-280590 A JP-A-4-324855 JP-A-6-92689 JP 2008-283075 A JP 63-215081 A JP-A-64-61959 JP 2001-339079 A

S. Tsuge, .et, al., Technical Digest of the International PVSEC-5, Kyoto, Japan, 1990, p.261-264 Y. Ashida, .et.al., Technical Digest of the International PVSEC-5, Kyoto, Japan, 1990, p.367-370

  In the thin film solar cell described in Patent Documents 1 and 2, the optical path length is extended by providing small irregularities on the irregular surface. However, there is room for further improvement in photoelectric conversion efficiency.

  In the method for manufacturing a thin-film solar cell described in Patent Documents 4 to 8, the surface of the transparent conductive film is modified by performing plasma treatment. However, it is also disclosed that minute recesses are formed by plasma treatment. There is no suggestion.

  The present invention has been made in view of the above-described problems, and provides a photoelectric conversion device and a method for manufacturing a photoelectric conversion device, in which a photoelectric conversion efficiency and stability are improved by forming minute recesses. With the goal.

  A photoelectric conversion device according to the present invention includes a substrate and a transparent conductive film that covers at least a part of the main surface of the substrate and has a concavo-convex shape on the surface opposite to the substrate side. In addition, the photoelectric conversion device includes a first conductive semiconductor layer having a first conductivity type that covers at least a part of the uneven shape of the transparent conductive film, and a light absorption layer that covers the first conductive semiconductor layer. . The concavo-convex shape has a convex portion having a maximum height of 50 nm to 1200 nm. The convex part has a minute concave part having a local peak sum of 2 nm or more and 25 nm or less on the surface. In the first conductivity type semiconductor layer, the layer thickness of the portion formed on the bottom of the minute recess is larger than the layer thickness of the portion formed other than on the bottom.

  By forming the minute recesses, the photoelectric conversion efficiency and stability of the photoelectric conversion device are improved.

It is sectional drawing which shows typically the structure of the photoelectric conversion apparatus which concerns on one Embodiment of this invention. It is the image which image | photographed a part of cross section of the photoelectric conversion apparatus which concerns on the same embodiment by TEM. 3 is an image obtained by binarizing the luminance value of each pixel with a threshold value of 50%. FIG. 4 is a schematic diagram of the image of FIG. 3. It is the image which image | photographed the vicinity of the boundary part of the convex part of the transparent conductive film which concerns on the same embodiment, p layer, and i layer with TEM. 6 is an image obtained by binarizing the luminance value of each pixel with the threshold value set to 50%. It is a figure which shows the image of FIG. 6 typically. It is a figure which shows the state which analyzed according to the rolling circular waviness measurement of the image of FIG. It is a schematic diagram which shows a micro recessed part and a rolling circle. It is a figure explaining the maximum depth of a micro recessed part. It is a figure explaining the maximum depth of the interface of p layer and i layer. It is a figure explaining the film thickness of p layer. It is a graph which shows the spectral transmission factor and spectral reflectance of the substrate for photoelectric conversion devices of the same embodiment, the substrate for photoelectric conversion devices not provided with the minute recesses, and the substrate for photoelectric conversion devices subjected to strong hydrogen plasma treatment. (A) is sectional drawing which shows typically the propagation state of the light in the photoelectric conversion apparatus which does not provide the micro recessed part, (B) is a schematic diagram of the propagation state of the light in the photoelectric conversion apparatus of this embodiment. FIG. It is a graph which shows the change of the characteristic value of a TCO board | substrate by hydrogen-containing plasma processing intensity. It is a figure which illustrates typically the chemical reaction in parts other than the micro recessed part of a TCO board | substrate in mixed plasma processing. It is a figure which illustrates typically the chemical reaction in the micro recessed part of a TCO board | substrate in mixed plasma processing. (A) is a schematic diagram showing an interface between a p layer and an i layer in an assumed photoelectric conversion device, and (B) is a schematic diagram showing an interface between a p layer and an i layer in a conventional photoelectric conversion device. FIG. 6C is a schematic diagram illustrating an interface between the p layer and the i layer in the photoelectric conversion device of the present embodiment. It is a figure which illustrates typically the chemical reaction in parts other than the micro recessed part of a TCO board | substrate. It is a figure which illustrates typically the chemical reaction in the micro recessed part of a TCO board | substrate.

  Hereinafter, an embodiment of a photoelectric conversion device and a method for manufacturing the photoelectric conversion device according to the present invention will be described with reference to the drawings. In the following description of the embodiments, the same or corresponding parts in the drawings are denoted by the same reference numerals, and the description thereof will not be repeated.

FIG. 1 is a cross-sectional view schematically showing a configuration of a photoelectric conversion apparatus according to an embodiment of the present invention. As shown in FIG. 1, in the photoelectric conversion device 1 according to the present embodiment, the transparent conductive film 3 containing SnO ₂ (tin oxide) as a main component has an uneven shape on the upper surface of a glass substrate 2 as a substrate. Is formed.

  In the present embodiment, a glass substrate is used as the substrate. However, any substrate that transmits light converted by the photoelectric conversion device 1 may be used, and a substrate that can be used for a super straight type solar cell is preferable. A super straight type solar cell is a solar cell into which light is incident from the substrate side. The glass substrate 2 and the transparent conductive film 3 constitute a photoelectric conversion device substrate 4.

  The photoelectric conversion device substrate 4 of the present invention may be one in which another layer is formed between a plate-like substrate such as the glass substrate 2 and the transparent conductive film 3. For example, a layer such as a silicon oxide film that suppresses diffusion of alkali metal impurities contained in the glass substrate 2 into the transparent conductive film 3 may be formed.

  Further, in the photoelectric conversion device substrate 4 of the present embodiment, most of one main surface of the glass substrate 2 is covered with the transparent conductive film 3, but a part of the main surface is covered with the transparent conductive film 3. There may be parts that are not. Such a substrate can be obtained, for example, by depositing the transparent conductive film 3 using a sputtering method or the like after the glass substrate 2 is housed in a frame-shaped tray. In this case, the transparent conductive film 3 is not deposited on the periphery of the main surface that is behind the tray.

  A p layer 5 made of a-SiC: H: B, which is a p-type first conductive semiconductor layer, is formed on the upper surface of the transparent conductive film 3. The p layer 5 may be composed of a-Si: H: B.

  On the upper surface of the p layer 5, an i layer 6 substantially made of a-Si: H, which is a substantially i-type semiconductor layer, is formed. Here, “substantially i-type” means not only true i-type but also weak p-type or weak n-type. Hereinafter, “substantially i-type” may be simply referred to as “i-type”. From the viewpoint of improving the interface characteristics, it is preferable to provide a so-called buffer layer in the vicinity of the interface on the conductivity type layer side of the i layer 6. As the buffer layer, for example, when a-SiC: H: B is used as the p layer 5, the composition ratio of silicon and carbon is high on the interface side with the p layer 5, and the interface side with the i layer 6. A layer whose carbon composition ratio decreases toward the bottom can be used.

  An n layer 7 made of a-Si: H: P, which is an n-type second conductive semiconductor layer, is formed on the upper surface of the i layer 6. A photoelectric conversion layer 8 is composed of the p layer 5, the i layer 6 and the n layer 7. The photoelectric conversion layer 8 is formed by a pin junction of layers made of at least one of crystalline and amorphous. Here, “crystalline” means so-called microcrystals, that is, a mixture of crystal grains and amorphous components.

  In the photoelectric conversion device of this embodiment, most of one main surface of the glass substrate 2 is covered with the semiconductor layer, but there may be a portion that is not covered with the semiconductor layer on a part of the main surface. . Such a substrate can be obtained, for example, by depositing a semiconductor layer using a plasma CVD method or the like after housing a photoelectric conversion device substrate in a frame-shaped tray. In this case, no semiconductor layer is deposited on the peripheral edge of the main surface that is behind the tray.

  The photoelectric conversion device of the present invention may have a form in which a plurality of photoelectric conversion layers are stacked. It is preferable from the viewpoint of improving the photoelectric conversion efficiency that the plurality of photoelectric conversion layers have substantially different band gaps of the i-type semiconductor layer. Specifically, for example, it is preferable to combine an i-type semiconductor layer of hydrogenated amorphous silicon and an i-type semiconductor layer of hydrogenated microcrystalline silicon. Alternatively, the composition ratio between silicon and other elements such as hydrogen, carbon, and germanium may be changed.

  Specifically, the i layer of the first photoelectric conversion layer is mainly composed of a-Si: H (hydrogenated amorphous silicon), and the i layer of the second photoelectric conversion layer is μc-Si: H (hydrogen). It is preferable that a photoelectric conversion device having a so-called tandem structure that has a stacked structure of a substrate for a photoelectric conversion device, a first photoelectric conversion layer, and a second photoelectric conversion layer.

  Moreover, the photoelectric conversion apparatus which has three photoelectric converting layers may be sufficient. In this case, each i layer of the first / second / third photoelectric conversion layers is a-Si: H / a-Si: H / .mu.c-Si: H, a-Si: H / a-SiGe: H / μc-Si: H, a-Si: H / a-Si: H / μc-SiGe: H, a-Si: H / a-SiGe: H / μc-SiGe: H, a-Si / μc- A so-called triple structure such as Si / μc-Si may be used. Furthermore, a structure in which four or more photoelectric conversion layers are stacked may be used.

When a plurality of photoelectric conversion layers are stacked, a light-transmitting conductive layer may be provided between adjacent photoelectric conversion layers. As the translucent conductive layer, for example, a transparent conductive oxide film such as ZnO or SiO _x can be used.

  On the upper surface of the n layer 7, a back transparent conductive film (not shown) mainly composed of ZnO (zinc oxide) and a back electrode layer 9 mainly composed of silver are formed. In the present embodiment, the back electrode layer 9 is made of silver, but any material having conductivity may be used, and the material is not limited to silver.

  In the photoelectric conversion device of the present embodiment, most of one main surface of the glass substrate 2 is covered with the back electrode layer, but a part of the main surface may not be covered with the back electrode layer. . Such a substrate can be obtained, for example, by depositing a back electrode layer using a sputtering method after housing a photoelectric conversion device substrate in a frame-shaped tray. In this case, the back electrode layer is not deposited on the peripheral edge of the main surface that is behind the tray.

  FIG. 2 is an image obtained by photographing a part of a cross section of the photoelectric conversion device according to the present embodiment with a TEM (Transmission Electron Microscope). In FIG. 2, the vicinity of the boundary between the photoelectric conversion device substrate 4 and the photoelectric conversion layer 8 is photographed.

  FIG. 3 is an image obtained by binarizing the luminance value of each pixel with the threshold value of 50% from the image of FIG. Here, binarization is a process of setting each pixel in the image to white when the luminance value of the pixel is equal to or greater than a threshold value and black when less than the threshold value. When the white luminance value is 255 and the black luminance value is 0, if the threshold is 50%, pixels with a luminance value of 128 or more are converted to white, and pixels with a luminance value of less than 128 are converted to black. The threshold value can be appropriately selected so that a white region and a black region are separated at the interface between the transparent conductive film 3 and the photoelectric conversion layer 8.

FIG. 4 is a schematic diagram of the image of FIG. As shown in FIG. 4, a plurality of convex portions 10 are formed on the transparent conductive film 3 formed on the upper surface of the glass substrate 2. Here, a length of 1000 nm parallel to the main surface of the glass substrate 2 is defined as a reference length (L _A ). Within the range of the reference length (L _A ), the distance in the height direction between the highest peak and the lowest valley among the peaks and valleys formed by the plurality of convex portions 10 is defined as the maximum height (H _max ).

In the case of FIG. 3, four convex portions 10 are observed within the range of the reference length (L _A ) of 1000 nm. Of the four convex portions 10, the second peak from the left is the highest and the second valley from the right is the lowest. In the case of FIG. 3, the maximum height of the convex portion 10 is 100 nm.

In addition, the interface of the transparent conductive film 3 and the glass substrate 2 is not observed in the image of FIG. Accordingly, in order to determine the direction of the reference length (L _A ), first, during TEM observation, the horizontal direction of the image is set at such a low magnification that the interface between the main surface of the glass substrate 2 and the transparent conductive film 3 can be observed. The direction should be parallel to this interface. Thereafter, the magnification is increased to such an extent that the maximum height of the convex portion 10 can be observed as shown in FIG. 2, and the direction of the reference length (L _A ) is set parallel to the horizontal direction of the image.

  The convex portion 10 preferably has a size capable of producing a light scattering or reflection state suitable for the light absorption characteristics of the photoelectric conversion layer 8. For example, it is preferable to be able to produce a sufficient light scattering effect not only for the middle wavelength of about 450 nm to 650 nm at the center of the sunlight spectrum but also for a long wavelength of about 700 nm to 1200 nm, for example. From these things, it is preferable that the maximum height of the convex part 10 is 50 nm or more and 1200 nm or less.

  FIG. 5 is an image obtained by photographing the vicinity of the convex portion of the transparent conductive film according to the present embodiment, the p layer, and the i layer with a TEM. FIG. 6 is an image obtained by binarizing the luminance value of each pixel with the threshold value of 50% from the image of FIG. FIG. 7 is a diagram schematically showing the image of FIG.

  As shown in FIG. 7, a plurality of minute concave portions 11 are formed on the surface of the convex portion 10 of the transparent conductive film 3. In addition, an interface 12 between the p layer 5 and the i layer 6 having a buffer layer in the vicinity of the interface on the p layer 5 side extends in a gentle line shape substantially parallel to the surface of the transparent conductive film 3.

  Hereinafter, the definition of the minute recesses, the line density of the minute recesses, and the maximum depth of the minute recesses according to the present invention will be described.

  FIG. 8 is a diagram showing a state in which the image of FIG. 6 is analyzed according to rolling circle waviness measurement. A line (boundary line between the white portion and the black portion) representing the surface of the convex portion 10 of the transparent conductive film 3 shown in FIG. 8 is hereinafter referred to as a measurement cross-sectional line.

The locus of the center 14 of the rolling circle 13 when a rolling circle 13 having a diameter of 25 nm rolls following the measurement cross-sectional line is referred to as a rolling circle traced profile. In the present embodiment, the reference length (L _w ) is 80 nm. An average line of the rolling circle waviness measurement line was calculated by a least square method using a linear function (straight line) for the rolling circle waviness measurement line measured at the reference length. Hereinafter, the average line of the rolling circle waviness measurement line is referred to as a measurement average line.

  In the above-described rolling circle waviness measurement, the ratio between the radius of the rolling circle and the reference length is doubled as compared with the measurement conditions of JIS standard (JIS B0610: '01).

  FIG. 9 is a schematic diagram showing a minute recess and a rolling circle. As shown in FIG. 9, the rolling circle 13 and the minute recess 11 are in contact at at least two points. Two of the contacts that are the farthest apart are referred to as local mountain peaks 15A and 15B.

Of the lines parallel to the measurement average line, the line passing through the bottom of the minute recess 11 is referred to as a base line (L ₂ ). Of the line passing through the local peak 15A parallel to the measurement average line and the line passing through the local peak 15B parallel to the measurement average line, the line with the distance from the base line (L ₂ ) is the top line (L ₁ ).

The minute recess 11 is the depth of the minute recess 11, the distance between the base line (L ₂ ) and the top line (L ₁ ) is 1 nm or more, and the width of the minute recess 11 is 2 The distance between two local peaks 15A and 15B is a recess having a distance of 2 nm or more. Therefore, the concave portion that does not satisfy this condition is not the minute concave portion 11.

In FIG. 8, in a range of 80 nm as the reference length (L _w ), the rolling circle 13A that is in contact with the minute recess 11 that satisfies the above conditions is indicated by a solid line, and the rolling circle that is in contact with the recess that does not satisfy the condition. 13B is indicated by a broken line.

The linear density of the minute recesses 11 refers to the number of minute recesses 11 present per unit length (1 nm) within a reference length of 80 nm. Therefore, as shown in FIG. 8, when there are seven rolling circles 13A in the range of 80 nm as the reference length (L _w ), the linear density of the minute recesses 11 is 7/80 = 0.0875 nm ^−1. .

FIG. 10 is a diagram for explaining the maximum depth of the minute recesses. As shown in FIG. 10, the maximum depth (D _max ) of the minute recess 11 is defined as two straight lines (L ₄ , L ₅₎ parallel to the measurement average line and in contact with the measurement sectional line within a reference length of 80 nm. ) The distance between each other. The maximum depth of the minute recess 11 shown in FIG. 8 was 4 nm.

FIG. 11 is a diagram for explaining the maximum depth of the interface between the p layer and the i layer. As shown in FIG. 11, the interface between the p layer 5 and the i layer 6 is gently and slightly waved. The maximum depth (B _max ) of the interface between the p layer 5 and the i layer 6 is two straight lines (L) parallel to the measurement average line and in contact with the interface 12 between the p layer 5 and the i layer 6 in the measurement cross section. ₆ , L ₇ ) Distance between each other. In the region corresponding to the maximum depth of the minute recesses 11 were measured (D _max) region, the maximum of the interface of the maximum depth of the micro recesses 11 similar to the manner as (D _max), and p layer 5 and the i layer 6 The depth (B _max ) was measured.

FIG. 12 is a diagram illustrating the thickness of the p layer. As shown in FIG. 12, the film thickness of the p layer 5 refers to the film thickness in a direction orthogonal to a straight line (L ₈ ) parallel to the measurement average line. The film thickness on the bottom 26 of the minute recess 11 is the length T shown in FIG.

Hereinafter, the manufacturing method and effect | action of the photoelectric conversion apparatus 1 of this embodiment are demonstrated.
SnO ₂ is laminated on the upper surface of the glass substrate 2 using an atmospheric pressure CVD (Chemical Vapor Deposition) method to form a transparent conductive film 3 having a convex portion 10 having a maximum height of 50 nm to 1200 nm. The substrate on which the transparent conductive film 3 having the convex portions 10 is formed is hereinafter referred to as a TCO (Transparent Conductive Oxide) substrate. In this embodiment, SnO ₂ is used as the material of the transparent conductive film 3, but a transparent conductive material such as In ₂ O ₃ , ZnO, and ITO (Indium Tin Oxide) may be used.

  In the present embodiment, the transparent conductive film 3 having a texture structure is formed by using the atmospheric pressure CVD method. However, the vacuum evaporation method, the EB evaporation method, the sputtering method, the reduced pressure CVD method, the sol-gel method, the electrodeposition method, and the like. May be used.

  Next, the TCO substrate is subjected to a hydrogen-containing plasma treatment in a capacitively coupled parallel plate high-frequency plasma reaction chamber. This hydrogen-containing plasma treatment is performed such that the main reduction reaction species that reach the TCO surface are hydrogen radicals. As a result of etching using the hydrogen radical as a reactive species, a minute recess 11 is formed on the surface of the protrusion 10 of the transparent conductive film 3. The reason why the minute recesses 11 are formed without being uniformly etched over the entire surface layer is considered to be because the etching rate at the defects existing on the surface of the TCO substrate is higher than the etching rate at other portions.

In this embodiment, since only hydrogen gas is used as the gas species introduced into the plasma reaction chamber, the production equipment can be simplified. The flow rate of H ₂ introduced into the plasma reaction chamber was 70 SLM. However, a mixed gas of a gas such as nitrogen, argon, or xenon and hydrogen gas may be used as a gas species introduced into the plasma reaction chamber. The “hydrogen-containing plasma” in the present invention is not limited to one using only hydrogen gas as long as the main reactive species is a hydrogen radical, and also means one using such a mixed gas. To do.

  In the present embodiment, the pressure in the plasma reaction chamber is 600 Pa. Since the density of hydrogen radicals increases as the pressure in the plasma reaction chamber increases, the maximum depth of the minute recesses 11 formed on the surface of the TCO substrate increases. On the other hand, as the pressure in the plasma reaction chamber is lowered, the ion damage received on the surface of the TCO substrate increases, so that the linear density of the minute recesses 11 formed on the surface of the TCO substrate increases. Therefore, in order to increase the linear density of the minute recesses 11 while increasing the depth of the minute recesses 11, the pressure in the plasma reaction chamber is preferably 100 Pa or more and 1500 Pa or less.

In the present embodiment, the power density input into the plasma reaction chamber is 40 mW / cm ² . The higher the power density during the hydrogen plasma treatment, the higher the density of hydrogen radicals generated in the plasma reaction chamber. Therefore, by applying a high power density and performing hydrogen plasma treatment, the maximum depth of the minute recesses 11 formed on the surface of the TCO substrate is increased, and the linear density of the minute recesses 11 is increased.

However, if the power density is too _high , the reduction amount of SnO ₂ constituting the transparent conductive film 3 in the TCO substrate is remarkably increased, and Sn is deposited on the surface of the transparent conductive film 3 to reduce the transmittance of the transparent conductive film 3. To do. Thus, when the transmittance | permeability of a TCO board | substrate becomes low, since it will cause the fall of the power generation efficiency in the photoelectric conversion apparatus 1, it is necessary not to make power density too large. Therefore, the power density input into the plasma reaction chamber is preferably 10 mW / cm ² or more and 200 mW / cm ² or less.

  The processing time in the plasma reaction chamber was 15 seconds. The longer the processing time, the deeper the maximum depth of the minute recesses 11 formed on the surface of the TCO substrate, and the greater the linear density of the minute recesses 11.

However, if the treatment time is too long, the reduction amount of SnO ₂ constituting the transparent conductive film 3 on the TCO substrate is remarkably increased, and Sn is deposited near the surface of the transparent conductive film 3 so that the transmittance of the transparent conductive film 3 is increased. descend. Thus, when the transmittance | permeability of a TCO board | substrate becomes low, since the fall of the power generation efficiency in the photoelectric conversion apparatus 1 will be caused, it is necessary not to make processing time excessively long. Therefore, the treatment time in the plasma reaction chamber is preferably 5 seconds or more and 200 seconds or less.

  When pulse plasma is used in the hydrogen-containing plasma processing, the effective processing time may be controlled by changing the duty ratio between the On time and the Off time. The frequency of ON / OFF in a pulsed manner is desirably a sufficiently low frequency (100 Hz to 10 kHz) with respect to the plasma excitation frequency (5 MHz to 80 MHz). The duty ratio can be 0.05 to 0.5. By reducing the duty ratio and increasing the input power, the ion damage to the substrate due to the processing can be increased while controlling the processing speed.

The processing intensity in the hydrogen-containing plasma processing is considered to be proportional to the total energy that the substrate receives from the plasma in the processing step. The total energy that the substrate receives from the plasma is proportional to the plasma input power density, the processing time, and the pulse duty ratio. The total energy per unit area (= the power density (mW / cm ²⁾ × processing time (sec) × duty ratio) is preferably ^{20mJ / cm 2 ~300mJ / cm 2} .

  In the present embodiment, the substrate temperature during the hydrogen-containing plasma treatment is 190 ° C. The higher the substrate temperature during processing, the smaller the difference in the reduction reaction rate between the defective portion on the surface of the convex portion 10 of the TCO substrate and the other portion. Therefore, since etching occurs over the entire surface of the convex portion 10 of the TCO substrate, the maximum depth of the fine concave portion 11 formed on the surface of the TCO substrate becomes shallow, and the linear density of the fine concave portion 11 becomes small. Therefore, the substrate temperature during the hydrogen-containing plasma treatment is preferably room temperature (20 ° C.) or higher and 200 ° C. or lower.

  Preferably, a treatment for introducing defects into the surface of the TCO substrate is performed before the hydrogen-containing plasma treatment. This defect means the defect which becomes the micro recessed part 11 by hydrogen-containing plasma processing. For example, the TCO substrate is plasma-treated for a very short time in a state where a power density significantly higher than the conditions of the hydrogen-containing plasma treatment step is applied. By performing this treatment, defects that may be caused by ion damage are introduced into the surface of the TCO substrate. As the power density increases, the linear density of defects generated increases, and the linear density of the minute recesses 11 generated by the hydrogen-containing plasma treatment in the subsequent process can be increased.

  When plasma is used in the defect introduction step, ion damage extends from the surface of the TCO substrate to a deeper position as the plasma potential is increased. Therefore, deep defects are generated, and the maximum depth of the minute recesses 11 generated by the subsequent hydrogen-containing plasma treatment is increased. In order to increase the plasma potential, for example, a short-time pulse plasma in which the on-time of one pulse is 1 msec or less is used, the power density input for generating the plasma is increased, the gas for plasma generation, etc. Can be used. In addition, a gas species having a relatively large atomic weight such as argon or xenon can be used.

  When performing the hydrogen-containing plasma treatment by increasing the power density input for generating the plasma as the defect introduction process, the hydrogen-containing plasma is reduced by reducing the power density without turning off the plasma after the defect introduction process. It is possible to shift to a processing step. Therefore, both shortening of the manufacturing time and simplification of the manufacturing apparatus can be achieved.

  FIG. 13 is a graph showing the spectral transmittance and the spectral reflectance of the photoelectric conversion device substrate of the present embodiment, the photoelectric conversion device substrate not provided with the minute recesses, and the photoelectric conversion device substrate subjected to strong hydrogen plasma treatment. It is. In FIG. 13, the vertical axis represents the spectral transmittance and the spectral reflectance, and the horizontal axis represents the wavelength of the incident light. The data of the photoelectric conversion device substrate 4 of the present embodiment is a solid line, the data of the photoelectric conversion device substrate not provided with the minute recesses 11 is a dotted line, and the data of the photoelectric conversion device substrate subjected to strong hydrogen plasma treatment is a one-point difference line Is shown.

  Since the substrate for a photoelectric conversion device not provided with the minute recesses is not subjected to the hydrogen plasma treatment process, the minute recesses 11 are not formed on the surface of the TCO substrate. Other configurations of the photoelectric conversion device substrate not provided with the minute recesses are the same as those of the photoelectric conversion device substrate 4 of the present embodiment.

  On the TCO substrate of the photoelectric conversion device subjected to strong hydrogen plasma treatment, the minute recesses 11 are formed by strong hydrogen plasma treatment with a treatment time of 300 seconds in the hydrogen plasma treatment step. Other configurations of the photoelectric conversion device substrate subjected to the strong hydrogen plasma treatment are the same as those of the photoelectric conversion device substrate 4 of the present embodiment.

  As shown in FIG. 13, the spectral transmittance is remarkably lowered in the substrate for a photoelectric conversion device subjected to strong hydrogen plasma treatment. This indicates that the amount of reduction of the transparent conductive film 3 is large and the transmittance of the transparent conductive film 3 is reduced as described above.

  When the wavelength of the incident light is in the range from about 400 nm to about 700 nm, the photoelectric conversion device substrate 4 of this embodiment has a spectral transmittance and a spectral transmittance that are smaller than those of the photoelectric conversion device substrate in which the minute recesses 11 are not provided. The amplitude of increase / decrease in reflectance is small. This is probably because the photoelectric conversion device substrate 4 of the present embodiment is less susceptible to light interference in the transparent conductive film 3 than the photoelectric conversion device substrate in which the minute recesses 11 are not provided. The reason is considered to be light scattering caused by the minute recess 11.

  FIG. 14A is a cross-sectional view schematically showing a light propagation state in a photoelectric conversion device not provided with a minute recess, and FIG. 14B is a light propagation state in the photoelectric conversion device of the present embodiment. It is sectional drawing which shows this typically.

  As shown in FIG. 14A, a part of the light 18 incident on the transparent conductive film 16 in which no minute recess is formed is transmitted at the interface between the transparent conductive film 16 and the p layer 5 and transmitted light 19. And part of the light is reflected to become reflected light 20.

  As shown in FIG. 14B, a part of the light 18 incident on the transparent conductive film 3 of the present embodiment is transmitted and becomes transmitted light 21 at the interface between the transparent conductive film 3 and the p layer 5. The part is reflected to become reflected light 22. It is considered that the transmitted light 21 and the reflected light 22 are dispersed in a wide range by the minute recesses 11.

  For this reason, the photoelectric conversion device substrate 4 of the present embodiment is less likely to cause light interference in the wavelength range of about 400 nm to about 700 nm, as compared to the photoelectric conversion device substrate in which the minute recesses 11 are not provided. It is thought that.

  Further, as shown in FIG. 14B, the photoelectric conversion device 1 of the present embodiment has a convex portion 10 in which a part of the reflected light 22 reflected at the interface between the transparent conductive film 3 and the photoelectric conversion layer 8 is reflected. It is considered that the light propagates to the minute concave portion 11 located on the slope opposite to the slope. A part of the reflected light 22 is further dispersed by the minute recesses 11, and a part of the reflected light 22 is considered to propagate into the photoelectric conversion layer 8 as transmitted light 23. Therefore, it is considered that the amount of light incident on the photoelectric conversion layer 8 increases and the light contributing to power generation can be increased.

  Here, although the details of the reason why the very small minute recesses 11 having a width of 2 nm or more and 25 nm or less scatter light are unknown, the width of the minute recesses 11 is the wavelength of sunlight incident on the photoelectric conversion device substrate 4, In particular, since it is less than one tenth of the wavelength of about 350 nm to about 1100 nm where the energy intensity is large, it is likely that each minute concave portion 11 is performing Rayleigh scattering of incident light. Guessed. Rayleigh scattering is only relevant when the size of the scatterer is very small compared to the wavelength of the light. The shorter the wavelength of light, the greater the degree of scattering, and the larger the scatterer, the greater the degree of scattering.

  From this, when considering the uneven surface of the transparent conductive film 3, the minute recess 11 having a width of 2 nm or more and 25 nm or less functions as a scatterer of incident light, and the degree of scattering increases as the wavelength of the incident light decreases. Become. Therefore, in a photoelectric conversion device having a plurality of photoelectric conversion layers, it is considered that the optical path length in the first photoelectric conversion layer located in the immediate vicinity of the transparent conductive film 3 is increased. The main component of the i layer of the first photoelectric conversion layer is made of hydrogenated amorphous silicon or hydrogenated amorphous silicon carbon, for example. If such short wavelength light is subjected to Rayleigh scattering, it is considered that the reflected light from the minute recess 11 is also strongly scattered. As a result, the probability that the scattered light reflected on one of the two inclined surfaces facing one convex portion 10 is incident again on the other inclined surface increases. In general, even when the i-layer thickness of the first photoelectric conversion layer is relatively thin, short wavelength light can be sufficiently absorbed. As a result, when the i layer is mainly composed of an amorphous silicon-based material, it is possible to reduce the influence of the i layer being photodegraded.

Among short-wavelength light, the range of about 350 nm to about 550 nm where the energy intensity of sunlight is relatively large should improve the absorption rate in the photoelectric conversion layer 8. It is particularly preferably 25 nm or less and 20 nm or less. Moreover, since it is considered that it is higher scattering effect of the more minute recesses 11 are more light, it is preferable that the linear density of the micro recesses 11 is 0.05 nm ^-1 or more, it is 0.07 nm ^-1 or more Particularly preferred. However, if the linear density is too high, the distance between the local peaks of each minute concave portion 11 becomes too small, and the effect of light scattering is reduced. Therefore, the linear density is preferably 0.3 nm ⁻¹ or less. Particularly preferably, it is 0.1 nm ⁻¹ or less. It is preferable that the minute concave portions are formed in substantially all the portions of the convex surface of the transparent conductive film.

  Incident light with a relatively long wavelength has a small degree of Rayleigh scattering, and light scattering based on a geometrical optical or Mie scattering mechanism by the convex part 10 whose maximum height is mainly 50 nm or more and 1200 nm or less seems to be dominant. It is. In FIG. 11, when the wavelength of incident light is about 800 nm or longer, the spectral reflectances of the photoelectric conversion device substrate 4 of the present embodiment and the photoelectric conversion device substrate having no minute recesses are substantially the same. It is thought that this proves that the degree of Rayleigh scattering will be small. Therefore, it is considered that the optical path length in the first photoelectric conversion layer located in the vicinity of the transparent conductive film 3 is relatively small.

  Therefore, the second photoelectric conversion layer is laminated at a position that covers the first photoelectric conversion layer by the amount of absorption of long-wavelength light in the first photoelectric conversion layer that is small in increment due to Rayleigh scattering. The amount of incident light is considered to increase. The main component of the i layer of the second photoelectric conversion layer is made of, for example, hydrogenated amorphous silicon germanium, hydrogenated microcrystalline silicon, or hydrogenated microcrystalline silicon germanium. As a result, the amount of light absorption in the i layer of the second photoelectric conversion layer can be improved. When the i-layer thickness of the first photoelectric conversion layer is relatively thin as described above, the light absorption amount in the i-layer of the second photoelectric conversion layer can be further improved.

  From the above considerations, it can be understood that the substrate for a photoelectric conversion device of the present invention is preferably applied to a so-called super straight type stacked photoelectric conversion device. The stacked photoelectric conversion device has a structure in which a plurality of photoelectric conversion layers are stacked so as to cover most of the main surface of the substrate. The band gap of the i layer of the photoelectric conversion layer closer to the light incident side is increased to mainly absorb short wavelength light, and the band gap of the i layer of the photoelectric conversion layer farther from the incident side is decreased to By absorbing the long wavelength light, it is possible to widen the usable range of the wavelength of the incident light and improve the photoelectric conversion efficiency.

  The super straight type means a type in which light is incident from the other main surface of the translucent substrate using a substrate having a transparent conductive film at a position covering one main surface of the translucent substrate such as glass. To do. In the super straight type stacked photoelectric conversion device using the photoelectric conversion device substrate 4 having the minute recesses 11, the short-wavelength light is greatly scattered to improve the absorption amount in the first photoelectric conversion layer close to the incident side. It is possible to scatter long-wavelength light relatively small and improve the amount of incident light on the second photoelectric conversion layer far from the incident side, thereby realizing improvement in photoelectric conversion efficiency.

  On the other hand, if the size of the minute recess 11 is too small, the surface is substantially the same as the uneven surface without the minute recess 11, and Rayleigh scattering does not occur. From this viewpoint, the interval between the local peaks of the minute recesses 11 is preferably 5 nm or more.

As described above, by providing the minute recesses 11, light is scattered by the minute recesses 11 and enters the photoelectric conversion layer 8, so that the optical path length of the light propagating in the photoelectric conversion layer 8 is increased. As a result, the photoelectric conversion efficiency of light in the photoelectric conversion layer 8 is improved, and the short-circuit current density (J _sc ) of the photoelectric conversion device 1 is increased.

  The photoelectric conversion device 1 of the present embodiment has a larger adhesive force between the photoelectric conversion layer 8 and the TCO substrate than a photoelectric conversion device that does not have the minute recesses 11. In the case where the photoelectric conversion layer 8 is laminated on the TCO substrate 17 not provided with the minute recesses 11 as shown in FIG. 14A, the adhesive force between the surface of the TCO substrate 17 and the photoelectric conversion layer 8 is relatively weak. . When the photoelectric conversion layer 8 is stacked on a TCO substrate provided with the minute recesses 11 as shown in FIG. 14B, the adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8 is relatively strong.

  As a factor of the difference in adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8, the first is considered to be a difference in contact area between the surface of the TCO substrate and the photoelectric conversion layer 8. By forming the minute recesses 11, the contact area between the surface of the TCO substrate and the photoelectric conversion layer 8 increases dramatically. It is considered that the adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8 increases due to the increase in the contact area.

  As a second factor of the difference in adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8, the difference in the size of the minute recesses 11 can be considered. As the minute concave portion 11 of the present embodiment, a narrow minute concave portion 11 in which the interval between the local peaks is 2 nm or more and 25 nm or less is provided. Therefore, at the interface between the surface of the TCO substrate and the photoelectric conversion layer 8, the photoelectric conversion layer 8 enters the narrow concave portion 11 having a small width.

In this way, at the complicated intricate interface, the anchor effect works between the surface of the TCO substrate and the photoelectric conversion layer 8, and the adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8 is remarkably increased. It is thought that there is. Therefore, the maximum depth of the minute recess 11 is preferably 2 nm or more. In addition, since the minute recesses 11 are formed on the surface of the protrusions 10 at a predetermined ratio or more, the anchor effect can be stabilized. Therefore, the linear density of the minute recesses 11 is preferably 0.05 nm ⁻¹ or more.

As a third factor of the difference in adhesive force between the surface of the TCO substrate and the photoelectric conversion layer 8, the third is considered to be a difference in activity on the surface of the TCO substrate. In the present embodiment, the minute recess 11 is formed by hydrogen plasma treatment. The surface of the TCO substrate 17 in which the minute recesses 11 are not formed as shown in FIG. 14A is an inert surface that is in a chemically relatively stable state and the number density of dangling bonds is relatively small. is there. For this reason, the TCO substrate 17 and the photoelectric conversion layer 8 are considered to have relatively few portions chemically bonded to each other. Therefore, in particular, in the case of a substrate having a large area of 1 m ² or more, the photoelectric conversion layer may be peeled off unless the surface is treated with hydrogen-containing plasma.

  When the hydrogen-containing plasma treatment is performed, it is considered that hydrogen radicals desorb oxygen contained in the TCO substrate on the surface of the TCO substrate, and dangling bonds are formed on the surface of the TCO substrate. In light of the presumed that the minute recesses 11 are formed as a result of hydrogen radicals preferentially etching the defects on the surface of the transparent conductive film, the dangling bonds are relatively smooth without the minute recesses 11 formed. It is presumed that there are relatively more areas around the valleys of the minute recesses 11 than there are.

Since these dangling bonds are reactive, when the photoelectric conversion layer 8 made of silicon is formed, SiH ₃ radicals and the like are firmly bonded to dangling bonds formed on the surface of the TCO substrate. Therefore, the TCO substrate and the photoelectric conversion layer 8 have relatively many portions that are chemically bonded to each other. As a result, it is considered that the adhesive force between the TCO substrate and the photoelectric conversion layer 8 increases.

FIG. 15 is a graph showing changes in the characteristic value of the TCO substrate due to the hydrogen-containing plasma treatment intensity. In FIG. 15, the vertical axis represents the transmittance of the TCO substrate, the ratio of Sn / Sn ^{x + on} the surface of the TCO substrate, and the horizontal axis represents the hydrogen-containing plasma treatment intensity. In FIG. 15, the hydrogen-containing plasma processing conditions of this embodiment are indicated by dotted lines.

As shown in FIG. 15, under the hydrogen plasma processing conditions of the present embodiment, the hydrogen plasma processing is performed with a hydrogen plasma processing intensity at which metal Sn starts to slightly deposit on the surface of the TCO substrate. Metal Sn is deposited by reducing SnO ₂ constituting the transparent conductive film 3 by hydrogen radicals.

  As described above, in the hydrogen-containing plasma processing step, the surface of the TCO substrate is reduced and etched by hydrogen radicals. As a result, the concentration of the reduced metal atoms is slightly present on the surface of the TCO substrate as compared with the bulk portion. A high surface layer is formed. A minute recess 11 is formed in this surface layer.

In the modification of the present embodiment, hydrogen plasma treatment (hereinafter referred to as mixed plasma) containing a small amount of carbon was performed after the hydrogen-containing plasma treatment step. During mixing the plasma, for example, CH _3, such as radicals, believed that there is a radical having a carbon atom. Hereinafter, the mixed plasma processing will be described by taking CH ₃ radicals as an example.

  FIG. 16 is a diagram schematically illustrating a chemical reaction in a portion other than the minute concave portion of the TCO substrate in the mixed plasma processing. FIG. 17 is a diagram schematically illustrating a chemical reaction in a minute recess of the TCO substrate in the mixed plasma processing. 16 and 17, hydrogen atoms 24 are indicated by white circles, and carbon atoms 25 are indicated by black circles.

In the mixed plasma treatment, as shown in FIG. 16A, CH ₃ radicals and hydrogen radicals are physically adsorbed on the surface of the TCO substrate at portions other than the minute recesses of the TCO substrate. Thereafter, as shown in FIG. 16B, CH ₃ radicals and hydrogen radicals thermally diffuse on the surface of the TCO substrate. Then, as shown in FIG. 16C, the CH ₃ radical and the hydrogen radical are desorbed from the surface of the TCO substrate by recombination.

On the other hand, in the minute recess 11 of the TCO substrate, as shown in FIG. 17A, CH ₃ radicals and hydrogen radicals thermally diffuse after physical adsorption on the surface of the TCO substrate. Next, as shown in FIG. 17B, since the dangling bonds are present in the minute recess 11 and the reaction activity is high, the CH ₃ radical chemically bonds with the dangling bonds in the minute recess. Then, as shown in FIG. 17C, the CH ₃ group bonded at the minute recess and the hydrogen radical collide, and the hydrogen molecule is desorbed due to the strong bonding force between the TCO substrate and the carbon atom. In the vicinity, a carbon atom having an unbonded hand is bonded.

  Note that the hydrogen-containing plasma treatment step and the mixed plasma treatment step can be continuously performed while the plasma is turned on. Specifically, for example, a solid carbon source that reacts with hydrogen plasma to generate carbon radicals is introduced into the plasma reaction chamber. By turning on the hydrogen plasma in this state, at the start of the hydrogen plasma treatment, a treatment with substantially pure hydrogen plasma is performed. With the passage of time, the amount of carbon-based radicals contained in the plasma increases and the mixed plasma treatment is performed. The substrate for the photoelectric conversion device is manufactured through the steps so far.

Note that the defect introduction step and the mixed plasma treatment step are not necessarily performed.
After the mixed plasma processing step, the p layer 5 is laminated using a plasma CVD apparatus. The precursor, such as SiH ₃ radical, is thermally diffused on the surface of the TCO substrate, and then a part thereof is chemically bonded at a chemically active site. Micro recesses 11 are dangling bonds are often active sites, also, because when subjected to mixed plasma treatment of carbon atoms having dangling bonds are abundant in the vicinity of the minute recess 11, SiH ₃ radicals micro recesses 11 It is easy to chemically bond with the TCO substrate. Therefore, the p layer 5 and the TCO substrate are firmly bonded in the minute recess 11. Considering that the thickness of the p layer 5 is preferably in the range of 5 nm or more and 15 nm or less, the maximum height of the minute recesses 11 so that the p layer 5 can appropriately cover the minute recesses 11 of the transparent conductive film 3. Is preferably 10 nm or less.

  Thereafter, the i layer 6 and the n layer 7 are sequentially stacked using a plasma CVD apparatus, thereby forming the photoelectric conversion layer 8. Finally, the back electrode layer 9 is formed by laminating zinc oxide and silver in this order so as to cover the n layer 7 using a sputtering apparatus. The photoelectric conversion device 1 of this embodiment is manufactured by the above process. As a method for forming the photoelectric conversion layer 8 and the back electrode layer 9, a method used for a general thin film solar cell or the like can be used.

  Hereinafter, the p layer 5 in the photoelectric conversion device 1 in the present embodiment will be described. FIG. 18A is a schematic diagram showing an interface between the p layer and the i layer in the assumed photoelectric conversion device, and FIG. 18B is a schematic diagram showing an interface between the p layer and the i layer in the conventional photoelectric conversion device. It is a figure and (C) is a schematic diagram which shows the interface of the p layer and i layer in the photoelectric conversion apparatus of this embodiment. In FIG. 18, defects present at the interface between the p layer and the i layer are indicated by black marks. This defect has an interface state that functions as a carrier recombination center.

  As shown in FIG. 18A, in the hypothetical photoelectric conversion device, the shape of the minute concave portion on the surface of the TCO substrate is copied at the interface between the p layer 28 and the i layer. Therefore, the area of the interface between the p layer 28 and the i layer is relatively large, and the number of defects 27 is also relatively large.

  As shown in FIG. 18B, in the conventional photoelectric conversion device, the minute recess 11 is not provided, and the shape of the surface of the TCO substrate is copied at the interface between the p layer 29 and the i layer. . Therefore, the area of the interface between the p layer 29 and the i layer is relatively small, and the number of defects 27 is relatively small.

  As shown in FIG. 18C, in the photoelectric conversion device of this embodiment, the shape of the minute recesses on the surface of the TCO substrate is not copied on the interface between the p layer 5 and the i layer 6. Therefore, the area of the interface between the p layer 5 and the i layer 6 is relatively small, and the number of defects 27 is relatively small.

  In the photoelectric conversion device 1 of the present embodiment, although the minute recesses 11 are provided, the number of defects 27 present at the interface between the p layer 5 and the i layer 6 is relatively small, and therefore, carrier recombination occurs. The frequency can be suppressed to the same level as that of a conventional photoelectric conversion device.

  In addition, as compared with the assumed photoelectric conversion device, the photoelectric conversion device 1 of the present embodiment can suppress diffusion of boron and carbon contained in the p layer 5 to the i layer 6. Therefore, when the i layer 6 is made of a silicon hydride-based material, the progress of photodegradation caused by the diffusion of impurities toward the i layer 6 can be suppressed to the same level as that of a conventional photoelectric conversion device. .

  Hereinafter, in the photoelectric conversion device 1 of the present embodiment, the reason why the shape of the minute recesses on the surface of the TCO substrate is not copied on the interface between the p layer 5 and the i layer 6 will be described. FIG. 19 is a diagram schematically illustrating a chemical reaction in a portion other than the minute concave portion of the TCO substrate. FIG. 20 is a diagram schematically illustrating a chemical reaction in a minute recess of the TCO substrate. 19 and 20, hydrogen atoms 24 are indicated by white circles, and silicon atoms 28 are indicated by black circles.

When deposition of the p-layer 5 is started on the TCO substrate, SiH ₃ radicals and hydrogen radicals are physically adsorbed on the surface of the TCO substrate at a position away from the minute recesses of the TCO substrate, as shown in FIG. The Thereafter, as shown in FIG. 19B, SiH ₃ radicals and hydrogen radicals thermally diffuse on the surface of the TCO substrate. As shown in FIG. 19C, since the density of active sites such as unbonded hands is low at a position away from the minute recess, the SiH ₃ radical and the hydrogen radical are recombined with each other on the TCO substrate. High probability of desorption from the surface.

On the other hand, when the deposition of the p layer 5 is started on the TCO substrate, SiH ₃ radicals and hydrogen radicals are physically adsorbed on the surface of the TCO substrate as shown in FIG. After that, thermal diffusion.

Next, as shown in FIG. 20B, since the dangling bonds are present in the minute recesses 11 and the reaction activity is high, the SiH ₃ radical is chemically bonded to the dangling bonds in the minute recesses. Then, as shown in FIG. 20C, the SiH ₃ group bonded to the minute recess and the hydrogen radical collide with each other, and since the bonding force between the TCO substrate and the carbon atom is strong, the hydrogen molecule is desorbed, In the vicinity, a SiH ₂ group having an unbonded hand is bonded. As shown in FIG. 20D, SiH ₃ radicals are chemically bonded to SiH ₂ groups having dangling bonds, whereby the bonding between the precursor and the TCO substrate and the deposition of the p layer proceed in the vicinity of the minute recess 11. To do.

Thus, in the vicinity of the minute recess 11, the deposition rate of the p layer 5 is faster than the position away from the minute recess 11. Therefore, the layer thickness of the P layer 5 formed on the bottom portion 26 of the minute recess 11 is larger than the layer thickness of the portion formed other than on the bottom portion 26. In addition, the maximum depth (B _max ) of the interface between the p layer 5 and the i layer 6 is smaller than the maximum depth (D _max ) of the minute recess 11. As a result, the interface between the p layer 5 and the i layer 6 has a gentle line shape in the measurement cross section.

  As described above, by not increasing the defects at the interface between the p layer 5 and the i layer 6 due to the formation of the minute recesses 11, it is possible to suppress a decrease in photoelectric conversion efficiency due to an increase in carrier recombination.

  The thickness of the p layer 5 is preferably 5 nm or more and 15 nm or less. When the thickness of the p layer 5 is 5 nm or more, the internal electric field of the photoelectric conversion device 1 can be sufficiently increased. When the thickness of the p layer 5 is 15 nm or less, in a photoelectric conversion device in which light absorption in the p layer 5 does not contribute to power generation, such as an amorphous silicon solar cell, loss due to light absorption in the p layer 5 is reduced. can do.

In addition, in the structure of the photoelectric conversion apparatus of this embodiment, you may employ | adopt the following structures. As the light-transmitting substrate, one having heat resistance that can withstand the plasma CVD process and having light-transmitting properties can be used, and glass, polyimide, and the like can be used. For example, alkali-free glass may be used. As the transparent conductive layer, ITO, ZnO, or the like can be used. For example, a SnO ₂ film may be used.

  As the photoelectric conversion layer, a pin single junction, a top layer pin / bottom layer pin two junction, a top layer pin / middle layer pin / bottom layer pin three junction, or a four-junction device structure higher than that can be used. A junction tandem structure may be employed.

  When a two-junction tandem structure is employed, an amorphous silicon carbide (a-SiC: H: B) layer, amorphous silicon ( An a-Si: H: B) layer and an amorphous silicon nitride (a-SiN: H: B) layer, or a laminate of these layers can be used. An a-Si: H layer can be used as the i layer (thickness is preferably 100 nm to 400 nm) constituting the top layer. As an n layer (thickness is preferably 5 nm to 40 nm) constituting the top layer, an a-Si: H: P layer and a microcrystalline silicon (μc-Si: H: P) layer, or these are used. Stacked ones can be used.

  As the p layer (thickness is preferably 5 nm to 30 nm) constituting the bottom layer, a μc-Si layer, a microcrystalline silicon nitride (μc-SiN: H: B) layer, and a microcrystalline silicon carbide ( A μc-SiC: H: B) layer or a laminate of these can be used. As the i layer (thickness is preferably 1000 nm to 3000 nm) constituting the bottom layer, a μc-Si: H layer can be used. As an n layer (thickness is preferably 5 nm to 40 nm) constituting the bottom layer, an a-Si: H: P layer and a microcrystalline silicon (μc-Si: H: P) layer, or these are used. Stacked ones can be used.

  As the back electrode layer, a laminate of a ZnO film (thickness is preferably 20 nm to 150 nm) and an Ag film (thickness is preferably 50 nm to 500 nm) can be used. Further, ITO may be used instead of ZnO. Al may be used instead of Ag.

According to the present embodiment, the optical path length of incident light is extended by exposing the transparent conductive film to hydrogen-containing plasma and providing the fine conductive layer 11 with a local peak sum of 2 nm to 25 nm in the transparent conductive film. In addition, the photoelectric conversion efficiency can be improved. Furthermore, the stability of the photoelectric conversion device 1 can be improved by increasing the bonding strength between the photoelectric conversion device substrate 4 and the photoelectric conversion layer 8. In order to obtain these effects, the linear density of the minute recesses 11 is preferably 0.05 nm ⁻¹ or more. Moreover, it is preferable that the maximum depth of the micro recessed part 11 is 2 nm or more and 10 nm or less.

  In addition, the said embodiment disclosed this time is an illustration in all the points, Comprising: It does not become a basis of limited interpretation. Therefore, the technical scope of the present invention is not interpreted only by the above-described embodiments, but is defined based on the description of the scope of claims. Further, all modifications within the meaning and scope equivalent to the scope of the claims are included.

  DESCRIPTION OF SYMBOLS 1 Photoelectric conversion apparatus, 2 Glass substrate, 3,16 Transparent conductive film, 4 Substrate for photoelectric conversion apparatuses, 5, 28, 29 p layer, 6 i layer, 7 n layer, 8 Photoelectric conversion layer, 9 Back electrode layer, 10 Convex part, 11 Minute concave part, 12 Interface, 13, 13A, 13B Rolling circle, 14 Rolling circle center, 15A, 15B Local peak, 17 TCO substrate, 18 light, 19, 21, 23 Transmitted light, 20, 22 Reflected light, 24 hydrogen atoms, 25 carbon atoms, 26 bottom, 27 defects, 28 silicon atoms.

Claims (9)

A substrate,
A transparent conductive film that covers at least a part of the main surface of the substrate and has a concavo-convex shape on the surface opposite to the substrate side;
A first conductivity type semiconductor layer covering at least a part of the irregular shape of the transparent conductive film and having a first conductivity type;
A light absorption layer covering the first conductivity type semiconductor layer,
The uneven shape has a convex portion having a maximum height of 50 nm or more and 1200 nm or less,
The convex part has a minute concave part on the surface where the distance between the local peaks is 2 nm or more and 25 nm or less,
The first conductivity type semiconductor layer is a photoelectric conversion device, wherein a layer thickness of a portion formed on a bottom portion of the minute recess is larger than a layer thickness of a portion formed on a portion other than the bottom portion.   The photoelectric conversion device according to claim 1, wherein the maximum depth of the minute recess is 2 nm or more and 10 nm or less. The photoelectric conversion device according to claim 1, wherein a linear density of the minute recesses is 0.05 nm ⁻¹ or more.   4. The photoelectric conversion device according to claim 1, wherein a thickness of the first conductivity type semiconductor layer is 5 nm or more and 15 nm or less. Forming a transparent conductive film having a concavo-convex shape in which the maximum height of the convex part is 50 nm or more and 1200 nm or less so as to cover at least a part of the main surface of the substrate;
Exposing the transparent conductive film to a hydrogen-containing plasma to form micro-recesses on the surface of the protrusions with a local peak sum of 2 nm or more and 25 nm or less;
Forming a first conductive semiconductor layer so as to cover at least a part of the transparent conductive film;
Forming a light absorption layer so as to cover the first conductivity type semiconductor layer,
In the step of forming the first conductivity type semiconductor layer, the layer thickness of the portion where the first conductivity type semiconductor layer is formed on the bottom of the minute recess is formed on the portion other than on the bottom. A method for manufacturing a photoelectric conversion device, which is formed to be larger.   The method for manufacturing a photoelectric conversion device according to claim 5, wherein the minute recesses are formed so that a maximum depth of the minute recesses is 2 nm or more and 10 nm or less. The method for manufacturing a photoelectric conversion device according to claim 5, wherein the minute recesses are formed so that a linear density of the minute recesses is 0.05 nm ⁻¹ or more.   The photoelectric conversion device according to claim 5, further comprising a step of introducing a defect into a surface of the transparent conductive film between the step of forming the transparent conductive film and the step of forming the minute recesses. Manufacturing method.   The method for manufacturing a photoelectric conversion device according to claim 5, wherein hydrogen plasma containing carbon is used in the step of forming the minute recesses.