JP5010691B2 - Stacked photoelectric conversion device - Google Patents

Description

  The present invention relates to an improvement in conversion efficiency of a thin film photoelectric conversion device, and more particularly to an improvement in photoelectric conversion efficiency of a thin film photoelectric conversion device in which a plurality of photoelectric conversion units are stacked.

  Note that the terms “crystalline” and “microcrystal” in the present specification include those that partially contain amorphous material. In addition, the term “pin junction” in the specification of the present application means that the stacking order on the substrate is the order of p-type layer, i-type layer, and n-type layer, and n-type layer, i-type layer, and p-type layer. All in order are included. In addition, the term “pin junction” in this specification includes both a p-type layer that is closer to the light incident side than an i-type layer and an n-type layer that is closer to the light incident side. It shall be.

  In recent years, in order to achieve both cost reduction and high efficiency of a photoelectric conversion device, a thin film photoelectric conversion device that has almost no problem in terms of resources has attracted attention and has been vigorously developed. Thin film photoelectric conversion devices are expected to be applied to various applications such as solar cells, optical sensors, and displays. An amorphous silicon photoelectric conversion device, which is one of thin film photoelectric conversion devices, can be formed on a large-area glass substrate or stainless steel substrate at a low temperature, so that cost reduction can be expected.

  A thin film photoelectric conversion device generally includes a first electrode, one or more semiconductor thin film photoelectric conversion units, and a second electrode, the surfaces of which are sequentially stacked on an insulating substrate. One thin film photoelectric conversion unit includes an i-type layer sandwiched between a p-type layer and an n-type layer.

  Most of the thickness of the thin film photoelectric conversion unit is occupied by the i-type layer which is a substantially intrinsic semiconductor layer, and the photoelectric conversion action mainly occurs in the i-type layer. Therefore, the i-type layer, which is a photoelectric conversion layer, is preferably thicker for light absorption, but if it is thicker than necessary, the deposition takes cost and time.

  On the other hand, the p-type or n-type conductive layer plays a role of generating a diffusion potential in the photoelectric conversion unit, and the magnitude of the diffusion potential causes an open end voltage, which is one of the important characteristics of the thin film photoelectric conversion device. The value depends. However, these conductive layers are inactive layers that do not contribute to photoelectric conversion, and light absorbed by impurities doped in the conductive layers does not contribute to power generation and is lost. Therefore, it is preferable that the thicknesses of the p-type and n-type conductive layers be as thin as possible within a range that generates a sufficient diffusion potential.

  Here, the photoelectric conversion unit or the thin-film solar cell has an amorphous i-type photoelectric conversion layer that occupies the main part regardless of whether the p-type and n-type conductivity type layers included therein are amorphous or crystalline. Those having a high quality are referred to as amorphous photoelectric conversion units or amorphous thin film solar cells, and those having a crystalline i-type layer are referred to as crystalline photoelectric conversion units or crystalline thin film solar cells.

  Generally, a semiconductor used for a photoelectric conversion layer has a light absorption coefficient that decreases as the wavelength increases. In particular, when the photoelectric conversion material is a thin film, sufficient light absorption does not occur in a wavelength region having a small absorption coefficient, so that the photoelectric conversion amount is limited by the film thickness of the photoelectric conversion layer. Therefore, by forming a light scattering structure that makes it difficult for light incident in the photoelectric conversion device to escape to the outside, it has been devised to increase the substantial optical path length, obtain sufficient absorption, and generate a large photocurrent. Yes. For example, when light is incident from the substrate side, a textured transparent conductive film having an uneven surface shape is used as the light incident side electrode.

  As a method for improving the conversion efficiency of a thin film photoelectric conversion device, a method of forming a stacked photoelectric conversion device in which two or more photoelectric conversion units are stacked is known. In this method, a front photoelectric conversion unit including a photoelectric conversion layer having a large band gap on the light incident side of the photoelectric conversion device (in this application, a photoelectric conversion unit relatively disposed on the light incident side is referred to as a front photoelectric conversion unit, A photoelectric conversion unit disposed on the side relatively far from the light incident side is referred to as a rear photoelectric conversion unit.) A photoelectric conversion layer having a small band gap (for example, Si-Ge alloy) is sequentially disposed behind the photoelectric conversion unit. By including the rear photoelectric conversion unit including the photoelectric conversion, it is possible to perform photoelectric conversion over a wide wavelength range of incident light, thereby improving the conversion efficiency of the entire apparatus. Among stacked thin film photoelectric conversion devices, a stack of an amorphous photoelectric conversion unit and a crystalline photoelectric conversion unit is referred to as a hybrid photoelectric conversion device. In the hybrid photoelectric conversion device, the wavelength of light that amorphous silicon can photoelectrically convert is about 800 nm on the long wavelength side, but crystalline silicon can photoelectrically convert longer light up to about 1100 nm. Therefore, it is possible to effectively photoelectrically convert a wider range of incident light.

  By the way, since each photoelectric conversion unit is connected in series in the stacked photoelectric conversion device, the short-circuit current density (Jsc) as the photoelectric conversion device is the lowest value among the current values generated in each photoelectric conversion unit. Is done. Therefore, it is preferable that the current values of the respective photoelectric conversion units are equal, and further, the conversion efficiency can be expected to increase as the absolute value of the current increases. In a stacked thin film photoelectric conversion device, a light-transmitting and light-reflective and conductive intermediate reflecting layer may be interposed between a plurality of stacked thin film photoelectric conversion units. In this case, a part of the light reaching the intermediate reflection layer is reflected, the amount of light absorption in the photoelectric conversion unit located on the light incident side of the intermediate reflection layer is increased, and the current value generated in the photoelectric conversion unit Can be increased. That is, the effective film thickness of the photoelectric conversion unit located on the light incident side with respect to the intermediate reflection layer apparently increased. For example, when an intermediate reflective layer is inserted into a hybrid photoelectric conversion device composed of an amorphous silicon photoelectric conversion unit and a crystalline silicon photoelectric conversion unit, the amorphous silicon photoelectric conversion is performed without increasing the film thickness of the amorphous silicon layer. The current generated by the unit can be increased. Alternatively, the amorphous silicon photoelectric conversion unit due to photodegradation that becomes conspicuous as the thickness of the amorphous silicon layer increases because the thickness of the amorphous silicon layer necessary to obtain the same current value can be reduced. It is possible to suppress the deterioration of characteristics.

  The intermediate reflective layer is often composed of a transparent conductive metal oxide layer such as polycrystalline ITO or ZnO, particularly ZnO. However, since ZnO is formed by a method such as sputtering or spraying, it is necessary to use equipment different from a semiconductor thin film generally formed by plasma CVD, etc., which requires equipment costs and increases production tact time. A problem occurs. Furthermore, in particular, when a sputtering method is used for forming ZnO, there is a problem that the performance may be deteriorated due to sputtering damage to the underlying semiconductor thin film.

Further, in order to suppress the influence on the series resistance of the solar cell, it is necessary to make a good ohmic contact at the interface between the transparent conductive metal oxide layer and the semiconductor thin film. For this reason, the dark conductivity of the transparent conductive metal oxide layer is 1.0 × 10 ² S / cm to 1.0 × 10 ³ S / cm by doping impurities or changing the degree of oxidation. It needs to be adjusted to a higher value.

  In particular, ZnO is generally known to be difficult to make ohmic contact at the interface with amorphous silicon or crystalline silicon. If the dark conductivity is lower than this range, the ohmic contact between the intermediate reflection layer and the front photoelectric conversion unit and between the intermediate reflection layer and the rear photoelectric conversion unit cannot be made, the contact resistance increases, and the cell fill factor ( FF) is lowered, and the characteristics of the photoelectric conversion device are deteriorated. On the other hand, if the dark conductivity is higher than this range, the transmittance of the transparent conductive metal oxide layer is lowered, the short-circuit current density (Jsc) is lowered, and the characteristics of the photoelectric conversion device are deteriorated.

  By the way, a large area thin film photoelectric conversion device is usually formed as an integrated thin film photoelectric conversion module. The integrated thin film photoelectric conversion module has a structure in which a plurality of photoelectric conversion cells, which are photoelectric conversion devices divided into small areas, are connected in series on a glass substrate. Each photoelectric conversion cell is generally formed by sequentially forming and patterning a transparent electrode layer, one or more thin film semiconductor photoelectric conversion units, and a back electrode layer on a glass substrate. .

  FIG. 19 is a cross-sectional view schematically showing an example of a conventional integrated thin film photoelectric conversion module having no intermediate reflection layer in which a plurality of stacked photoelectric conversion devices are connected in series. An integrated thin film photoelectric conversion module 101 shown in FIG. 19 includes a transparent electrode layer 103, a front photoelectric conversion unit 104a that is an amorphous silicon photoelectric unit, and a rear photoelectric conversion unit that is a crystalline silicon photoelectric conversion unit on a glass substrate 102. 104b and the back electrode layer 106 are sequentially stacked.

  As shown in FIG. 19, the integrated thin film photoelectric conversion module 101 is provided with first and second separation grooves 121 and 122 and a connection groove 123 for dividing the thin film. The first and second separation grooves 121 and 122 and the connection groove 123 are parallel to each other and extend in a direction perpendicular to the paper surface. Note that the boundary between the adjacent photoelectric conversion cells 110 is defined by the first and second separation grooves 121 and 122.

  The first separation groove 121 divides the transparent electrode layer 103 corresponding to each photoelectric conversion cell 110, has an opening at the interface between the transparent electrode layer 103 and the amorphous silicon photoelectric conversion unit 104a, and The surface of the transparent substrate 102 is the bottom surface. The first separation groove 121 is filled with an amorphous material constituting the amorphous silicon photoelectric conversion unit 104a, and electrically insulates the adjacent transparent electrode films 103 from each other.

  The second separation groove 122 is provided at a position away from the first separation groove 121. The second separation groove 122 divides the front photoelectric conversion unit 104 a, the rear photoelectric conversion unit 104 b, and the back electrode layer 106 corresponding to each photoelectric conversion cell 110, and has an opening on the upper surface of the back electrode layer 106. And the bottom surface is an interface between the transparent electrode layer 103 and the front photoelectric conversion unit. The second separation grooves 122 electrically insulate the back electrode layers 106 between the adjacent photoelectric conversion cells 110.

  The connection groove 123 is provided between the first separation groove 121 and the second separation groove 122. The connection groove 123 divides the front photoelectric conversion unit 104a and the rear photoelectric conversion unit 104b, has an opening at the interface between the rear photoelectric conversion unit 104b and the back electrode layer 106, and the transparent electrode layer 103 and the front photoelectric conversion unit. The interface of 104a is the bottom surface. The connection groove 123 is embedded with a metal material constituting the back electrode layer 106 and electrically connects one back electrode layer 106 and the other transparent electrode layer 103 of the adjacent photoelectric conversion cells 110. . That is, the connection groove 123 and the metal material filling it have a role of connecting the photoelectric conversion cells 110 juxtaposed on the glass substrate 102 in series.

  As shown in FIG. 20, when a transparent conductive metal oxide layer is simply inserted between the front photoelectric conversion unit and the rear photoelectric conversion unit as an intermediate reflection layer in the structure of FIG. Thus, the characteristics of the integrated thin film photoelectric conversion module are significantly deteriorated.

  As shown in FIG. 20, when the intermediate reflection layer 105 is provided, the connection groove 123 penetrates the front photoelectric conversion unit 104a, the intermediate reflection layer 105, and the rear photoelectric conversion unit 104b, and the back electrode layer 106 is configured in the connection groove 123. Material to be embedded. That is, the metal embedded in the connection groove 123 and the intermediate reflection layer 105 come into contact with each other.

When this intermediate reflective layer is formed of a transparent conductive metal oxide layer, it has a high dark conductivity of 1.0 × 10 ² S / cm to 1.0 × 10 ³ S / cm as described above, In the intermediate reflection layer, a current easily flows in a direction parallel to the substrate, and also serves as an electrode layer. That is, the rear photoelectric conversion unit 104b is short-circuited in the current path of the intermediate reflection layer 105, the connection groove 123, and the back electrode layer 106, and a large leak current flows. Therefore, in the structure of FIG. 20, almost no electric power generated in the rear photoelectric conversion unit can be taken out.

(Prior Example 1)
It is considered that such a problem of leakage current can be solved by adopting a structure in which a third separation groove is newly provided as described in Patent Document 1 by the present applicant shown in FIG. In Patent Document 1, the integrated thin film photoelectric conversion module 101 is provided with first to third separation grooves 121, 122, 124 and a connection groove 123 for dividing the thin film.

  The third separation groove 124 is provided between the first separation groove 121 and the connection groove 123. The third separation groove 124 divides the front photoelectric conversion unit 104a and the intermediate reflection layer 105, has an opening at the interface between the intermediate reflection layer 105 and the rear photoelectric conversion unit 104b, and covers the surface of the transparent electrode layer 103. It is the bottom. The third separation groove 124 is embedded with a crystalline material constituting the thin film photoelectric conversion unit 104b, and a portion located in the cell 110 of the intermediate reflection layer 105 is made of a conductive material such as a metal filling the connection groove 123. It is electrically insulated. Note that the third separation groove 124 may be provided such that the first separation groove 121 is located between the third separation groove 124 and the connection groove 123. However, as shown in FIG. 21, it is easier to increase the effective area for power generation by providing the third separation groove between the first separation groove 121 and the connection groove 123.

  In the module 101 of FIG. 21 described above, since the separation groove 124 is provided, a leakage current is generated between the portion of the intermediate reflection layer 105 located in the cell 110 and the metal filling the connection groove 123. Can be prevented.

  However, the structure of Patent Document 1 in FIG. 21 has one more separation groove than the structure in FIG. The first to third separation grooves or connection grooves are generally formed by patterning with a YAG laser or the like. That is, the structure of Patent Document 1 increases the number of patterning steps once, increases the number of YAG lasers or increases the tact time of patterning, and increases the device cost and manufacturing cost of the integrated thin film photoelectric conversion module. A problem occurs.

  Further, when the structure of FIG. 19 is manufactured, the front photoelectric conversion unit 104a and the rear photoelectric conversion unit 104b can be manufactured continuously in a vacuum apparatus by plasma CVD.

  However, in the case of the structure of Patent Document 1, it is necessary to produce the front photoelectric conversion unit 104a by plasma CVD and the intermediate reflective layer 105 by sputtering or the like and then take it out from the vacuum apparatus and perform patterning with a YAG laser. After that, it is necessary to place the substrate in the vacuum apparatus once more and to produce the rear photoelectric conversion unit 104b by plasma CVD. Accordingly, in the case of the structure of FIG. 21, since loading into the vacuum device, heating of the substrate, and unloading from the vacuum device increase one by one, the manufacturing time of the integrated thin film photoelectric conversion module increases and the manufacturing cost increases. It will be.

Further, since the substrate is taken out into the atmosphere once after the intermediate reflective layer 105 is formed, the intermediate reflective layer 10
In the air, impurities in the atmosphere are adsorbed on the interface between the photoelectric conversion unit 104 and the rear photoelectric conversion unit 104b, and there may be a problem in that the reliability of the integrated thin film photoelectric conversion module is deteriorated or the reliability is lowered.

  Furthermore, since the area loss of the thin film photoelectric conversion cell is increased by providing the third separation groove 124, there is a problem that the characteristics of the integrated thin film photoelectric conversion module with the intermediate reflection layer cannot be sufficiently exhibited.

  From the above, the third separation groove is required to suppress the leakage current of the integrated thin film photoelectric conversion module having the intermediate reflection layer. However, since the third separation groove is required, the patterning is increased once, and the front photoelectric conversion unit, the intermediate reflection layer, and the rear photoelectric conversion unit cannot be formed continuously. There are a number of problems such as contamination of the interface between the reflective layer and the back photoelectric conversion unit due to exposure to the atmosphere, and an increase in area loss, resulting in an increase in module cost and a decrease in characteristics, and the characteristics of the intermediate reflective layer cannot be fully exhibited.

(Prior Example 2)
Incidentally, Patent Document 2 discloses an example in which amorphous silicon oxide is used as the material of the semiconductor layer of the stacked photoelectric conversion device. In this example, on a glass substrate, a transparent electrode such as SnO ₂ , a first p-type layer of amorphous silicon carbide, a first i-type layer of amorphous silicon, a first n-type layer of amorphous silicon oxide, A second p-type layer of amorphous silicon carbide, a second n-type layer of amorphous silicon, a second i-type layer of amorphous silicon, a second n-type layer of amorphous silicon, a metal electrode such as Ag It has the structure which formed. Usually, amorphous silicon or microcrystalline silicon is used for the first n-type layer. However, Patent Document 2 reports that light absorption loss can be reduced by using amorphous silicon oxide having a wide band gap. . As a result, it is shown that the light passing through the first n-type layer and reaching the second i-type layer is increased, the short-circuit current density (Jsc) is increased, and the characteristics of the stacked photoelectric conversion device are improved.

Amorphous silicon oxide can arbitrarily adjust the oxygen concentration in the film. The higher the oxygen concentration in the film, the wider the band gap and the higher the transmittance. On the other hand, the conductivity decreases as the oxygen concentration in the amorphous silicon oxide film increases. In Patent Document 2, in order to apply to the first n-type layer, the photoconductivity, which is the conductivity when irradiated with light, is required to be 1 × 10 ⁻⁶ S / cm or more. It is reported that it is essential that x is less than 0.2 when expressed by Si _1-x O _x .

  In Patent Document 2, light reaching the second i-type layer, which is the photoelectric conversion active layer of the rear photoelectric conversion unit, increases, and Jsc increases due to an increase in the power generation current of the rear photoelectric conversion unit. It is said that the conversion efficiency of the device will be improved. However, there is no mention of improving the characteristics of the front photoelectric conversion unit. Further, there is no mention of the possibility that light reflects at the interface between the first i-type layer of amorphous silicon and the first n-type layer of amorphous silicon oxide to serve as an intermediate reflection layer. Furthermore, nothing is mentioned about the refractive index of amorphous silicon oxide. In Patent Document 2, since the oxygen concentration in the amorphous silicon oxide film is limited to less than 20%, the refractive index for light having a wavelength of 600 nm is around 3 as shown in FIG. In that case, since the difference in refractive index between amorphous silicon oxide and amorphous silicon is small, it is clear that an increase in current due to the reflection effect at the interface cannot be expected. In order to function as an intermediate reflection layer, it is necessary to increase the oxygen concentration in the amorphous silicon oxide film to decrease the refractive index. In this case, however, conversion is performed because the photoconductivity decreases and FF decreases. Efficiency will decrease. Therefore, in the configuration of Patent Document 2, amorphous silicon oxide cannot be used as the intermediate reflection layer.

(Prior Example 3)
By the way, an amorphous silicon oxide film containing a silicon microcrystalline phase is obtained by decomposing a gas containing SiH ₄ , CO ₂ , H ₂ and limiting the flow rate ratio of CO ₂ / SiH ₄ to 1.5 or less. A method of manufacturing and applying it to a window layer of an amorphous silicon photoelectric conversion device is disclosed in Patent Document 3. In Patent Document 3, amorphous silicon oxide containing a silicon microcrystalline phase having the same photoconductivity is compared with amorphous silicon oxide having a minimum photoconductivity of 10 ⁻⁶ S / cm applicable to a window layer. Since the absorption coefficient is small, it is disclosed that the light absorption loss is reduced when applied to the window layer of the photoelectric conversion device. However, no application example to the photoelectric conversion device other than the window layer is disclosed, and any technique applied to the intermediate reflection layer of the stacked photoelectric conversion device is not disclosed. Further, there is no disclosure about the refractive index of amorphous silicon oxide containing a silicon microcrystalline phase. Patent Document 3 regarding a basic concept and structure to be applied as an intermediate reflection layer using a difference in refractive index between an amorphous silicon oxide film containing a silicon microcrystalline phase and a silicon film, which is an important effect of the present invention described later. However, nothing is disclosed.

JP 2002-261308 A JP-A-5-95126 Japanese Patent No. 3047666

  In a stacked photoelectric conversion device, when a transparent conductive metal oxide layer such as ZnO is used for the intermediate reflection layer, the manufacturing method is different from that of the photoelectric conversion unit. There is a problem that an increase cannot be avoided. Specifically, while the photoelectric conversion unit is manufactured by plasma CVD, the intermediate reflective layer of the transparent conductive metal oxide layer is manufactured by a film forming method different from the photoelectric conversion unit, such as sputtering or spraying.

  In addition, when an integrated thin film photoelectric conversion module is formed using a transparent conductive metal oxide layer such as ZnO as the intermediate reflection layer, the structure having the first separation groove, the second separation groove, and the connection groove, There is a problem that leakage current is generated and the characteristics of the integrated thin film photoelectric conversion module are deteriorated.

  This problem of leakage current can be solved by providing a third separation groove as shown in Patent Document 1, but the following new problem occurs. That is, since the patterning is increased once, there is a problem that the patterning apparatus cost and the tact time are increased. In addition, since the front photoelectric conversion unit, the intermediate reflection layer, and the rear photoelectric conversion unit cannot be formed in succession, the work of carrying the substrate into the vacuum apparatus, heating, and carrying it out increases once, resulting in an increase in tact time. There is a problem that the interface between the intermediate reflection layer and the rear photoelectric conversion unit is contaminated by exposure to the atmosphere. Furthermore, there is a problem that the area loss increases.

  Further, in the stacked photoelectric conversion device using amorphous silicon for the i-type layers of the front photoelectric conversion unit and the rear photoelectric conversion unit, amorphous silicon oxide having an oxygen concentration of less than 20% in the film is used for the first photoelectric conversion unit. When used for one n-type layer, the refractive index does not drop sufficiently, so that the reflection effect does not appear at the interface between the amorphous silicon of the first i-type layer and the first n-type layer, and it does not function as an intermediate reflective layer There is. Further, when the oxygen concentration in the amorphous silicon oxide film is increased to 20% or more and the refractive index is lowered, the photoconductivity is lowered, the FF of the stacked photoelectric conversion device is reduced, and the conversion efficiency is lowered. There is.

  According to the present invention, a stacked photoelectric conversion device including a plurality of photoelectric conversion units composed of pin junctions includes a first photoelectric conversion unit, an n-conductivity type silicon composite layer, and a second photoelectric conversion from the side close to the light incident side. The silicon composite layer includes a silicon crystal phase dispersed in an amorphous alloy matrix of silicon and oxygen. By using the n-conductivity-type silicon composite layer, it is possible to increase the oxygen concentration in the film, achieve a low refractive index, and obtain a high reflection effect at the interface. In addition, although the n-conductivity-type silicon composite layer includes a silicon crystal phase, high dark conductivity can be realized even though the oxygen concentration in the film is high. As a result, by using the silicon composite layer, it is possible to achieve both a high reflection effect and a high dark conductivity, and the power generation current of the first photoelectric conversion unit is increased to improve the characteristics of the stacked photoelectric conversion device. . Note that the same n-conductivity microcrystalline silicon layer as the silicon composite layer is preferably adjacent to the silicon composite layer.

  In addition, according to the present invention, the i-type layer of the first photoelectric conversion unit includes at least one portion sequentially configured by the first photoelectric conversion unit and the second photoelectric conversion unit from the side close to the light incident side. A stacked photoelectric conversion device having an n-type layer on the side far from the light incident side, wherein at least a part of the n-type layer is an n-type silicon composite layer, and the silicon composite layer is composed of silicon and oxygen. Including a silicon crystal phase dispersed in an amorphous alloy matrix. The silicon composite layer also serves as a part of the n-type layer of the first photoelectric conversion unit, thereby increasing the power generation current of the first photoelectric conversion unit due to the reflection effect and at the same time reducing the light absorption loss. It also becomes possible to increase the generated current of the photoelectric conversion unit. Note that the n-type layer in contact with the n-type silicon composite layer is preferably a microcrystalline silicon layer.

  In this case, the n-type layer of the first photoelectric conversion unit includes a first n-type layer that is n-type amorphous silicon or microcrystalline silicon, a second n-type layer that is an n-type silicon composite layer, and an n-type layer. The third n-type layer, which is amorphous silicon or microcrystalline silicon, can be sequentially formed. The n-type layer of the first photoelectric conversion unit is a stack of a first n-type layer that is an n-type silicon composite layer and a second n-type layer that is n-type amorphous silicon or microcrystalline silicon. It may be a structure. Furthermore, the n-type layer of the first photoelectric conversion unit may be an n-type silicon composite layer. Note that the n-type layer in contact with the n-type silicon composite layer is preferably a microcrystalline silicon layer.

  In order to obtain a sufficient reflection effect, the silicon composite layer has a refractive index with respect to light having a wavelength of 600 nm of 1.7 to 2.1, and preferably 1.8 to 2.1. In order to achieve a low refractive index, the silicon composite layer has an in-film oxygen concentration of 40 atomic% to 60 atomic%, and preferably 40 atomic% to 55 atomic%.

The silicon composite layer preferably has a dark conductivity of 10 ⁻⁸ S / cm or more and 10 ⁻¹ S / cm or less. When the dark conductivity is too low, the fill factor (FF) of the photoelectric conversion device is lowered and the conversion efficiency is lowered. On the other hand, if the dark conductivity is too high, it may cause leakage current when an integrated thin film photoelectric conversion module is formed.

  In order to achieve an optimum dark conductivity, the silicon composite layer has a peak intensity ratio of the TO mode peak of the crystalline silicon component to the TO mode peak of the amorphous silicon component measured by Raman scattering of 0.5 to 10 Preferably there is.

  The silicon composite layer preferably has a thickness of more than 20 nm and less than 130 nm, and more than 50 nm and less than 100 nm in order to ensure an optimal reflection effect.

  The substantially intrinsic i-type layer of the first photoelectric conversion unit is preferably amorphous silicon. Moreover, it is preferable that the substantially intrinsic i-type layer of the second photoelectric conversion unit is microcrystalline silicon or thin film polycrystalline silicon. Alternatively, it is preferable that the substantially intrinsic i-type layer of the second photoelectric conversion unit is amorphous silicon germanium.

According to the present invention, the stacked photoelectric conversion device includes three or more photoelectric conversion units, and has an n-type silicon composite layer between at least one of the photoelectric conversion units and the photoelectric conversion unit.

  Alternatively, according to the present invention, in the stacked photoelectric conversion device, at least one photoelectric conversion unit including three or more photoelectric conversion units and excluding the photoelectric conversion unit farthest from the light incident side is the i of the photoelectric conversion unit. An n-type layer is provided on the far side of the mold layer from the light incident side, and at least a part of the n-type layer is an n-type silicon composite layer.

In the n-type silicon composite layer, the concentration of P in the film is preferably 5 × 10 ¹⁹ cm ⁻³ or more and 2 × 10 ²² cm ⁻³ or less.

  When the substrate of the stacked photoelectric conversion device is a transparent substrate, the reflection spectrum of light incident through the transparent substrate has at least one maximum value and minimum value of reflectance in the wavelength range of 500 nm to 800 nm, respectively. The difference in reflectance between the maximum value and the minimum value is preferably 1% or more.

  According to the present invention, a stacked photoelectric conversion device having a structure including a first photoelectric conversion unit, an n-type silicon composite layer, and a second photoelectric conversion unit in order from the side closer to the light incident side is provided. As a result, light is partially reflected at the front and back interfaces of the silicon composite layer, increasing the generated current of the first photoelectric conversion unit, or reducing the thickness of the i-type layer of the first photoelectric conversion unit. An equivalent generated current can be generated.

  In addition, since the n-type silicon composite layer can be manufactured by plasma CVD in the same manner as the photoelectric conversion unit, the first photoelectric conversion unit, the n-type silicon composite layer, and the second photoelectric conversion unit are manufactured using the same apparatus. It becomes possible. This eliminates the need for another type of equipment for film formation that is necessary for the intermediate reflective layer of the conventional conductive metal oxide layer, thereby reducing the cost of the apparatus. Alternatively, the manufacturing cost can be reduced by shortening the tact time.

  Furthermore, the silicon composite layer also serves as a part of the n-type layer of the first photoelectric conversion unit, thereby increasing the power generation current of the first photoelectric conversion unit by the reflection effect and at the same time reducing the light absorption loss. The power generation current of the second photoelectric conversion unit can be increased, and the conversion efficiency of the stacked photoelectric conversion device is improved.

  By applying the stacked photoelectric conversion device including the n-type silicon composite layer according to the present invention to the integrated thin film photoelectric conversion module, the third separation groove is unnecessary, and the patterning device cost is reduced by reducing the patterning once. You can save time. Further, since the front photoelectric conversion unit, the n-type silicon composite layer, and the rear photoelectric conversion unit can be formed in succession, the work of carrying the substrate into the vacuum apparatus, heating, and carrying it out is only once, and the tact time is shortened. Furthermore, the interface between the n-type silicon composite layer and the rear photoelectric conversion unit is not exposed to the atmosphere, and the influence of contamination is eliminated. Further, since there is no third separation groove, the area loss can be reduced, and the conversion efficiency of the integrated thin film photoelectric conversion module is improved.

Characteristics of refractive index and dark conductivity for light with a wavelength of 600 nm of the silicon composite layer. A dark field image of a transmission electron microscope of a silicon composite layer. High resolution bright field image of transmission electron microscope of silicon composite layer. Raman scattering spectrum of silicon composite layer. Characteristics of oxygen concentration in the silicon composite layer and refractive index with respect to light having a wavelength of 600 nm. 1 is a stacked photoelectric conversion device according to a first embodiment of the present invention. Interface reflectivity considering interference when the thickness of the silicon composite layer is changed. The bright field image of the transmission electron microscope of the cross section of the laminated photoelectric conversion apparatus of this invention. The dark-field image of the transmission electron microscope of the cross section of the laminated photoelectric conversion apparatus of this invention. The reflectance of the light which injected light from the glass substrate of the laminated photoelectric conversion apparatus of this invention. A stacked photoelectric conversion device of Comparative Example 1 of the prior art. The stacked photoelectric conversion apparatus of the comparative example 2 of a prior art. 8 is a stacked photoelectric conversion device according to a second embodiment of the present invention. 7 is a stacked photoelectric conversion device according to a third embodiment of the present invention. 8 is a stacked photoelectric conversion device according to Example 4 of the present invention. The relative value of the spectral sensitivity electric current of the laminated photoelectric conversion apparatus of Example 5 of this invention. 8 is a stacked photoelectric conversion device according to Example 6 of the present invention. 8 is an integrated thin film photoelectric conversion module according to Example 7 of the present invention. The integrated thin film photoelectric conversion module of the comparative example 4 of a prior art. The integrated thin film photoelectric conversion module of the comparative example 5 of a prior art. The integrated thin film photoelectric conversion module of the comparative example 6 of a prior art.

In order to find a material having both a low refractive index and high conductivity, the present inventors have intensively studied a method of forming an alloy of silicon and oxygen by a high-frequency plasma CVD method. As a result, a structure including a silicon crystal phase dispersed in an amorphous alloy matrix of silicon and oxygen (referred to as a silicon composite layer in the present invention) has a low refractive index and a high conductivity. I found it. FIG. 1 shows the dark conductivity with respect to the refractive index measured by forming a silicon composite layer on a glass substrate. Here, the refractive index was measured with respect to light of 600 nm using spectroscopic ellipsometry. This is because in the hybrid photoelectric conversion device which is one of the stacked photoelectric conversion devices, the falling of the spectral sensitivity current of the front photoelectric conversion unit and the rising of the spectral sensitivity current of the rear photoelectric conversion unit intersect at a wavelength of about 600 nm. It is to do. It can be said that a film that reflects light in the vicinity of 600 nm well, that is, a film having a small refractive index with respect to light of 600 nm is suitable for increasing the power generation current of the front photoelectric conversion unit. The dark conductivity was measured by a current flowing in a direction parallel to the substrate with a coplanar electrode attached to the silicon composite layer. As can be seen from FIG. 1, as a result of detailed experiments, the inventors of the present invention have a low refractive index of 1.7 to 2.5 and a darkness of 10 ⁻⁸ to 10 ⁻¹ S / cm high in the n-type silicon composite layer. It has been found that conductivity can be realized simultaneously. Since the refractive index of amorphous silicon or crystalline silicon with respect to a wavelength of 600 nm is about 4, the difference in refractive index from the n-type silicon composite layer is large, and a sufficient reflection effect can be obtained. As far as the present inventors know, there is no known document regarding a film having such a low refractive index and sufficient conductivity for an alloy-based film of silicon and oxygen by plasma CVD.

  FIG. 2 is a dark field image of a transmission electron microscope (TEM) obtained by forming a silicon composite layer on a glass substrate and photographing the film from a direction perpendicular to the film surface. Since the dark field image is an image of an electron beam diffracted at a specific angle, diffraction does not occur in the amorphous portion, and only a crystal directed at a specific angle diffracts. Therefore, a dark field image brightly formed is always a crystal phase. That is, FIG. 2 shows that a crystalline phase is contained in the amorphous material. FIG. 3 is a high-resolution TEM bright-field image obtained by photographing the same silicon composite layer as in FIG. 2 from a direction perpendicular to the film surface. It can be confirmed that the regular crystal lattices are arranged, and it is clear that the crystal phase is included in the film.

FIG. 4 is a Raman scattering spectrum of the same silicon composite layer as FIG. A sharp peak of the TO mode of crystalline silicon near 520 cm ⁻¹ appears. That is, it can be seen that the crystal phase is a silicon crystal. At this time, the peak intensity ratio of the TO mode peak of the crystalline silicon component to the TO mode peak of the amorphous silicon component near 480 cm ⁻¹ is 2.5.

Such an n-type silicon composite layer that achieves both a low refractive index and a high dark conductivity uses so-called microcrystals that use SiH ₄ , CO ₂ , H ₂ , and PH ₃ as reaction gases and have a large H ₂ / SiH ₄ ratio. It has been experimentally found that it can be produced by plasma CVD under the production conditions and in the range of the CO ₂ / SiH ₄ ratio of about 2 to 10. At this time, the plasma conditions are a power coupling frequency of 10 to 100 MHz, a power density of 50 to 500 mW / cm ² , a pressure of 50 to 1000 Pa, and a substrate temperature of 150 to 250 ° C. using capacitively coupled parallel plate electrodes. When the CO ₂ / SiH ₄ ratio is increased, the oxygen concentration in the film increases monotonously. However, the carbon concentration in the film was 1 atomic% or less even when the CO ₂ / SiH ₄ ratio was changed in the range of 0 to 4, and it was found by experiment that it hardly enters the film compared with oxygen.

  FIG. 5 shows the refractive index for light having a wavelength of 600 nm of the silicon composite layer with respect to the oxygen concentration in the film of the n-type silicon composite layer. As will be described later with reference to FIG. 16, a silicon composite layer is used in the stacked photoelectric conversion device, and in order to reduce the refractive index to 2.1 or less so that a current increase of 10% or more can be observed due to the reflection effect, the oxygen concentration in the film is set to 40 It can be seen that the atomic% or more should be used.

The dark conductivity of the n-type silicon composite layer is determined by the oxygen concentration in the film, the concentration of the doping impurity (P) in the film, and the ratio of the silicon crystal phase contained in the film. In order to adjust the dark conductivity of the n-type silicon composite layer to 10 ⁻⁸ to 10 ⁻¹ S / cm and the refractive index to 1.7 to 2.1, the oxygen concentration in the film is 40 to 60 atomic%. preferable. The refractive index decreases as the oxygen concentration in the film increases, but the dark conductivity decreases, so the oxygen concentration in the film has the above-mentioned preferred upper limit.

The n-type silicon composite layer preferably has a P concentration in the film of 5 × 10 ¹⁹ cm ⁻³ or more and 2 × 10 ²² cm ⁻³ or less as a doping impurity. As the P concentration in the film increases, the dark conductivity increases. However, when the amount of P increases, the ratio of the crystal phase decreases, so the dark conductivity decreases. For this reason, it is preferable to adjust the P concentration in the film to the above range.

  Furthermore, as an index of the ratio of the silicon crystal phase of the n-type silicon composite layer, the peak intensity ratio of the TO mode peak of the crystalline silicon component to the TO mode peak of the amorphous silicon component measured by Raman scattering is 0.5 or more and 10 or less. It is preferable that As the peak intensity ratio increases, the dark conductivity increases. However, if the peak intensity ratio becomes too large, the ratio of amorphous silicon oxide in the n-type silicon composite layer decreases and the refractive index increases. For this reason, it is preferable to adjust the peak intensity ratio of Raman scattering to the above range.

  In FIG. 1, the variation in dark conductivity is large even with the same refractive index because the production conditions of the oxygen concentration in the film, the impurity concentration in the film, and the ratio of the silicon crystal phase are changed.

In the n-type silicon composite layer according to the present invention, it is considered that the silicon crystal phase dominates the electron transport path. Therefore, even if the oxygen concentration in the film is increased and the refractive index is lowered to 2.1 or less, the silicon composite phase The dark conductivity of the layer can be kept high. For this reason, even if this n-type silicon composite layer is disposed between the front photoelectric conversion unit and the rear photoelectric conversion unit of the photoelectric conversion device, the influence on the series resistance of the photoelectric conversion device is small. Accordingly, the thickness and refractive index can be designed to be optimal for optical confinement. Moreover, since the refractive index of the n-type silicon composite layer can be easily controlled simply by changing the CO ₂ / SiH ₄ gas ratio and adjusting the oxygen concentration in the film, the refractive index is changed periodically in the film thickness direction, etc. An increase in the light confinement effect can also be expected by precise optical design.

In order to suppress the influence on the series resistance of the photoelectric conversion device, the dark conductivity of the intermediate reflective layer of the transparent conductive metal oxide layer such as ZnO needs to have a high value of 10 ² to 10 ³ S / cm. . In general, it is known that it is difficult to make ohmic contact at the interface between ZnO and amorphous silicon or crystalline silicon. In particular, it is difficult to make ohmic contact at the interface between ZnO and p-type amorphous silicon or p-type crystalline silicon. However, if the silicon composite layer has a dark conductivity of 10 ⁻⁸ to 10 ⁻¹ S / cm, it is detailed that a good ohmic contact can be realized between the photoelectric conversion unit of amorphous silicon and crystalline silicon. It became clear by examination. One reason for this is that the n-type silicon composite layer is a semiconductor thin film mainly composed of silicon, similarly to amorphous silicon and crystalline silicon.

  Furthermore, as described below, it can be considered that there is a current path through the silicon crystal phase in the cross-sectional direction of the silicon composite layer as a reason why a good ohmic contact can be obtained. The dark conductivity in FIG. 1 is a value obtained from a current flowing in a direction parallel to the film. In the case of a photoelectric conversion device, a current flows in the cross-sectional direction of the film. In the dark field image of the TEM in FIG. 2, since a brightly visible crystal phase is visible in some places, it can be said that the silicon crystal phase penetrating the film thickness of the silicon composite layer has a structure dispersed in a planar shape. In other words, it is considered that the silicon composite layer applied to the stacked photoelectric conversion device is connected to the cross-sectional direction of the film by the silicon crystal phase. Therefore, even if the silicon composite layer has a low dark conductivity in the direction parallel to the film surface, a current flows in the cross-sectional direction mainly through the silicon crystal phase, and an increase in series resistance of the photoelectric conversion device can be suppressed. it is conceivable that.

  Even if the dark conductivity of the n-type silicon composite layer is several to ten orders of magnitude lower than that of the transparent conductive metal oxide, a good ohmic contact can be obtained, so that the structure of the integrated thin film photoelectric conversion module can be simplified. The apparatus cost can be reduced and the conversion efficiency of the module can be improved. Although a detailed description will be given later, in the integrated thin film photoelectric conversion module, the problem of leakage current does not occur even in the structure without the third separation groove as shown in FIG. Therefore, in the integrated thin film photoelectric conversion module, patterning is reduced once, and the patterning apparatus cost and tact time can be shortened. In addition, since the front photoelectric conversion unit, the n-type silicon composite layer, and the rear photoelectric conversion unit can be formed continuously, the work of carrying the substrate into the vacuum apparatus, heating, and carrying it out is only once, and the tact time is shortened. The interface between the n-type silicon composite layer and the rear photoelectric conversion unit is not exposed to the atmosphere, and the influence of contamination is eliminated. Further, since there is no third separation groove, the area loss is reduced and the conversion efficiency of the integrated thin film photoelectric conversion module is improved.

  FIG. 6 is a cross-sectional view of a stacked photoelectric conversion device according to an embodiment of the present invention. Hereinafter, the present invention will be described in detail with reference to FIG. On the glass substrate 1 which is a transparent substrate, the transparent electrode layer 2, the amorphous photoelectric conversion unit 3 as the first photoelectric conversion unit 3, the n-type silicon composite layer 4, the crystalline as the second photoelectric conversion unit 5 A silicon photoelectric conversion unit and a back electrode layer 6 are formed.

  In addition to the glass of FIG. 6, a transparent resin film or the like is used for the transparent substrate used in the present invention, but it should be as transparent as possible in order to transmit and absorb more sunlight into the photoelectric conversion layer. preferable. From the same intention, if non-reflective coating is performed in order to reduce the light reflection loss on the substrate surface on which sunlight is incident, high efficiency can be achieved.

A transparent conductive oxide (TCO) is used as the transparent electrode layer 2 that is an electrode on the light incident side, and materials constituting the TCO include tin oxide (SnO ₂ ), indium tin oxide (ITO), Zinc oxide (ZnO) or the like can be used, but SnO ₂ is particularly preferable. Moreover, it is preferable that the unevenness | corrugation which has a pitch of 200-900 nm is formed in the interface by the side of the photoelectric conversion unit of the transparent electrode layer 2, The constituent material forms the transparent electrode layer 2 with a particle size of 200-900 nm for this purpose (However, “having a pitch” as used herein does not only indicate that the pitch repeats regularly with a specific pitch, but also includes those that change randomly while changing the value.) )

  Although two photoelectric conversion units may be used as illustrated, three or more photoelectric conversion units may be stacked. In addition, when three or more photoelectric conversion units are stacked, the n-type silicon composite layer 4 may be formed between the photoelectric conversion units, but may be one layer (any number of layers as long as it is one or more layers). Is possible).

  As the photoelectric conversion unit, one composed of one conductivity type layer, an i type layer which is a substantially intrinsic photoelectric conversion layer, and a reverse conductivity type layer can be used. The one conductivity type layer may be a p-type layer or an n-type layer, and the opposite conductivity type layer correspondingly becomes an n-type layer or a p-type layer. However, since a p-type layer is usually arranged on the light incident side in a photoelectric conversion device, generally in the structure of FIG. 6, the one conductivity type layers 31 and 51 are p-type layers, and the opposite conductivity type layers 33 and 53 are It is an n-type layer. The p-type layer and the n-type conductive layer serve to generate a diffusion potential in the photoelectric conversion unit, and the open-circuit voltage (Voc), which is one of the characteristics of the thin film photoelectric conversion device, depending on the magnitude of the diffusion potential. Is affected. However, these conductive layers are inactive layers that do not contribute to photoelectric conversion, and light absorbed by impurities doped in the conductive layers does not contribute to power generation. Therefore, it is preferable to make the p-type layer and the n-type conductive layer as thin as possible within a range in which a sufficient diffusion potential is generated. Since the i-type layers 32 and 52 play a role of absorbing light and performing photoelectric conversion, the plurality of photoelectric conversion units to be combined are combinations in which the band gaps of the intrinsic photoelectric conversion layers are different, that is, combinations of materials having different light absorption wavelength regions. It is preferable that it has absorption in the main wavelength range (400-1200 nm) of sunlight as a whole. For example, a combination of an amorphous silicon thin film and an amorphous silicon germanium thin film, a combination of an amorphous silicon thin film and a crystalline silicon thin film, or the like can be given.

  In the case where an amorphous silicon thin film photoelectric conversion unit is formed as the front photoelectric conversion unit of FIG. 6, that is, the first photoelectric conversion unit 3, it is possible to stack them in the order of nip, but as shown in FIG. It is preferable that the respective semiconductor layers are stacked by the plasma CVD method in this order because the conversion efficiency becomes higher. In this case, for example, a p-type amorphous silicon carbide layer 31 doped with 0.01 atomic% or more of boron, which is a conductivity-determining impurity atom, an i-type amorphous silicon layer 32 serving as a photoelectric conversion layer, and a conductivity-type determination An n-type microcrystalline silicon layer 33 doped with 0.01 atomic% or more of phosphorus, which is an impurity atom, may be deposited in this order. However, these layers are not limited to the above. For example, the p-type layer may be made of amorphous silicon, microcrystalline silicon, or amorphous silicon nitride. Further, amorphous silicon may be used for the n-type layer. Note that the film thickness of the conductive type (p-type, n-type) layer is preferably 3 nm to 100 nm, and more preferably 5 nm to 50 nm.

The n-type silicon composite layer 4, which is a feature of the present invention, reflects a part of the light reaching the silicon composite layer 4 to the front photoelectric conversion unit 3 located on the light incident side of the silicon composite layer 4, and the rest The light is transmitted to the rear photoelectric conversion unit 5. In the case where a silicon-based material is used for the photoelectric conversion layer, the refractive index of the photoelectric conversion layer with respect to light of 600 nm is about 4, and therefore the refractive index of the silicon composite layer 4 is preferably in the range of 1.7 to 2.1. . Further, since current flows through the silicon composite layer 4, the silicon composite layer 4 preferably has a dark conductivity of 10 ⁻⁸ S / cm or more and 10 ⁻¹ S / cm or less.

  FIG. 7 shows the reflectance of light at 600 nm in consideration of the interference at the interface before and after the silicon composite layer 4 when the thickness of the n-type silicon composite layer 4 is changed. At this time, the refractive index of the n-type silicon composite layer with respect to light of 600 nm is 2. From FIG. 7, it is preferable to make the film thickness of the n-type silicon composite layer 4 larger than 20 nm and smaller than 130 nm in order to ensure that the light reflected to the front photoelectric conversion unit 3 side is 10% or more. Recognize. Furthermore, it can be seen that the thickness of the n-type silicon composite layer 4 is preferably 50 nm or more and 100 nm or less in order to make the light reflected to the front photoelectric conversion unit 3 side 30% or more.

  The front photoelectric conversion unit 3, the n-type silicon composite layer 4, and the rear photoelectric conversion unit 5 are preferably formed continuously without being taken out into the atmosphere. Here, “without taking out into the atmosphere” means to maintain the environment in which the contamination of the surface can be prevented, and various methods are possible if this can be achieved.

  The film thickness and shape of the silicon composite layer 4 applied to such a stacked photoelectric conversion device can be measured with a cross-sectional transmission electron microscope (cross-section TEM). FIG. 8 is a cross-sectional TEM image of the vicinity of the interface before and after the silicon composite layer in a stacked photoelectric conversion device having a structure of glass substrate / amorphous photoelectric conversion unit / silicon composite layer / crystalline photoelectric conversion unit / back electrode layer. It is a bright field image. In the bright-field image of FIG. 8, the silicon composite layer portion appears whitish reflecting that the density of the silicon composite layer is lower than that of amorphous silicon and crystalline silicon before and after that. FIG. 9 is a dark field image obtained by photographing the same place as the bright field image of FIG. It can be seen that the silicon composite layer has a partly bright place and the silicon composite layer contains a crystalline phase.

  Further, the oxygen concentration and the P concentration in the silicon composite layer of the stacked photoelectric conversion device can be detected by a known analysis method. For example, the composition can be analyzed by SIMS, ESCA, EPMA, Auger electron spectroscopy or the like while changing the depth detected by wet etching, plasma etching, ion sputtering, or the like.

  Furthermore, the refractive index of the silicon composite layer of the stacked photoelectric conversion device is measured by ellipsometry by removing the back electrode layer with an acid such as HCl and exposing the silicon composite layer by wet etching, plasma etching, ion sputtering, etc. Can be detected. In addition, the determination of the presence or absence of the silicon composite layer can be easily detected even by the difference in reflectance of light incident from the glass substrate. FIG. 10 shows a stacked photoelectric conversion device A having a structure of glass substrate / amorphous photoelectric conversion unit / silicon composite layer / crystalline photoelectric conversion unit / back electrode layer, and glass substrate / amorphous photoelectric conversion unit / crystalline. It is the reflection spectrum which injected from the glass substrate into the laminated photoelectric conversion apparatus of the structure of a photoelectric conversion unit / back surface electrode layer, and the laminated photoelectric conversion apparatus without a silicon composite layer. In the case where there is a silicon composite layer, light is reciprocally reflected in the amorphous photoelectric conversion unit to cause interference, and a maximum value and a minimum value having a reflectance difference of 1% or more appear at wavelengths of 500 nm to 800 nm. On the other hand, when there is no silicon composite layer, clear maximum and minimum values do not appear in this wavelength region.

Even when a crystalline silicon photoelectric conversion unit, for example, is formed as the second photoelectric conversion unit 5 on the silicon composite layer 4, it may be formed at a low temperature of 400 ° C. or lower by a plasma CVD method in the order of the pin type. preferable. It is preferable that the crystalline silicon photoelectric conversion layer which is the intrinsic photoelectric conversion layer 52 is formed at a low temperature so as to include a large number of hydrogen atoms that terminate and inactivate defects in the crystal grain boundaries and grains. Specifically, the hydrogen content of the photoelectric conversion layer 52 is preferably in the range of 1 to 30 atomic%. Further, this layer is preferably formed as a thin film that is substantially an intrinsic semiconductor having a conductivity type determining impurity atom density of 1 × 10 ¹⁸ cm ⁻³ or less. Further, most of the crystal grains contained in the intrinsic crystalline silicon layer are grown in a columnar shape from the front electrode side, and preferably have a (110) preferential orientation plane parallel to the film surface. This is because the crystalline silicon thin film having such a crystal orientation has a surface texture structure in which the surface of the photoelectric conversion unit deposited thereon has fine irregularities even when the surface of the transparent electrode 2 is substantially flat. Indicates. Therefore, in the case where the surface of the transparent electrode 2 has a surface texture structure including irregularities, the surface of the photoelectric conversion unit has a texture structure with small irregularities compared to the surface of the transparent electrode 2, and thus has a wide wavelength range. This is because the structure has a large light confinement effect suitable for reflecting light. The film thickness of the intrinsic crystalline silicon layer is preferably 0.1 μm or more and 10 μm or less. However, since the thin film photoelectric conversion unit preferably has absorption in the main wavelength region (400 to 1200 nm) of sunlight, an amorphous silicon germanium layer (for example, an alloy material) is used instead of the intrinsic crystalline silicon layer (for example, An amorphous silicon germanium layer made of amorphous silicon containing 30 atomic% or less of germanium) or crystalline silicon germanium may be formed.

  By the way, the film thickness of the p-type crystalline silicon of the crystalline silicon photoelectric conversion unit is preferably in the range of 3 nm to 25 nm. When the film thickness of the p-type crystalline silicon is smaller than 3 nm, an internal electric field sufficient to take out carriers generated inside the crystalline i-type silicon photoelectric conversion layer by light irradiation, i.e., a function as a p layer, is provided. It cannot be generated. If it is larger than 25 nm, the light absorption loss of the p layer itself is increased. The film thickness of the n-type crystalline silicon is preferably in the range of 3 nm to 20 nm as in the case of p-type crystalline silicon.

As the back electrode layer 6, it is preferable to form at least one metal layer made of at least one material selected from Al, Ag, Au, Cu, Pt and Cr by sputtering or vapor deposition. Between the photoelectric conversion unit and the metal electrode, ITO, may be formed a layer made of SnO _2, conductive oxides such as ZnO (not shown).

  For example, the back electrode 6 is preferably a multilayer film in which ZnO having a thickness of 10 nm to 150 nm and a silver film having a thickness of 30 nm to 500 nm are formed in this order. When ZnO is thinner than 10 nm, the adhesion between the crystalline silicon photoelectric conversion unit and the silver film is deteriorated, and conversely, when it is thicker than 150 nm, the light absorption of ZnO itself is increased, which is a factor of deteriorating the characteristics of the photoelectric conversion device. The silver film has a function of reflecting light on a long wavelength side that is difficult to be absorbed by the crystalline silicon photoelectric conversion unit and making it incident on the crystalline silicon photoelectric conversion unit again. When the film thickness of the silver film is 30 nm or less, the effect as the reflective layer is drastically reduced, and when it is 500 nm or more, the manufacturing cost is increased.

  In the example of FIG. 6, an embodiment using a transparent substrate is shown. However, a laminated type in which a back electrode layer, a rear photoelectric conversion unit, a silicon composite layer, a front photoelectric conversion unit, and a transparent electrode layer are sequentially stacked on an opaque substrate. Similarly, in the photoelectric conversion device, the generation efficiency of the front photoelectric conversion unit can be increased to improve the conversion efficiency. However, in this case, it is preferable to laminate the rear photoelectric conversion unit and the front photoelectric conversion unit in the order of the nip type layer.

  Hereinafter, examples according to the present invention and comparative examples according to the prior art will be described in detail. In the drawings, the same members are denoted by the same reference numerals, and redundant description is omitted. Moreover, this invention is not limited to a following example, unless the meaning is exceeded.

  First, a two-stage stacked photoelectric conversion device will be described while comparing Comparative Examples 1 and 2 according to the prior art and Examples 1 to 4 according to the present invention. The characteristics of the stacked photoelectric conversion devices of Comparative Examples 1 and 2 and Examples 1 to 4 are summarized in Table 1. At this time, the values obtained by normalizing the power generation currents of the front photoelectric conversion unit and the rear photoelectric conversion unit measured by the spectral sensitivity measuring apparatus with the respective values of the front photoelectric conversion unit and the rear photoelectric conversion unit of Comparative Example 1 are also described. Further, the total output current obtained by summing the output currents of the front photoelectric conversion unit and the rear photoelectric conversion unit was also normalized and described with the value of Comparative Example 1.

(Comparative Example 1)
As Comparative Example 1, a stacked photoelectric conversion device as shown in FIG. A pyramidal SnO ₂ film having a thickness of 800 nm was formed as a transparent electrode layer 2 on a glass substrate 1 having a thickness of 1.1 mm and a 127 mm square by a thermal CVD method. The sheet resistance of the obtained transparent electrode layer 2 was about 9Ω / □. Further, the haze ratio measured with a C light source was 12%, and the unevenness depth d was about 100 nm. On this transparent electrode layer 2, a p-type amorphous silicon carbide layer 31 having a thickness of 15 nm, an i-type amorphous silicon layer 32 having a thickness of 0.3 μm, and an n-type having a thickness of 30 nm are formed using plasma CVD. The front photoelectric conversion unit 3 composed of the type microcrystalline silicon layer 33 is formed, followed by the p-type microcrystalline silicon layer 51 having a thickness of 15 nm, the i-type crystalline silicon layer 52 having a thickness of 2.5 μm, and the layer having a thickness of 15 nm. The rear photoelectric conversion units 5 composed of the n-type microcrystalline silicon layer 53 were sequentially formed. Thereafter, Al-doped ZnO having a thickness of 90 nm and Ag having a thickness of 300 nm were sequentially formed as the back electrode layer 6 by sputtering.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., the open circuit voltage (Voc ) Was 1.353 V, the short circuit current density (Jsc) was 11.61 mA / cm ² , the fill factor (FF) was 0.734, and the conversion efficiency (Eff) was 11.53%.

(Comparative Example 2)
As Comparative Example 2, a stacked photoelectric conversion device as shown in FIG. This has a structure in which the n-type microcrystalline silicon layer 33 of Comparative Example 1 is replaced with an n-type amorphous silicon oxide 39 having a thickness of 30 nm. This has a structure similar to the preceding example 2 except that the rear photoelectric conversion unit is a crystalline photoelectric conversion unit. The flow rate of the gas when forming the n-type amorphous silicon oxide 39 is SiH ₄ / CO ₂ / PH ₃ / H ₂ = 5 / 2.5 / 0.1 / 100 sccm. The film was formed at a power frequency of 13.56 MHz, a power density of 20 mW / cm ² , a pressure of 100 Pa, and a substrate temperature of 200 ° C. At this time, the amorphous silicon oxide 39 has an oxygen concentration in the film of 18 atomic%, a refractive index for light of 600 nm is 3.0, and the TO of the crystalline silicon component with respect to the TO mode peak of the amorphous silicon component measured by Raman scattering. The peak intensity ratio of the mode peak was 0, there was no crystal phase, and the dark conductivity was 1.2 × 10 ⁻⁶ S / cm. Other than that, it was formed by the same manufacturing method as Comparative Example 1.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., Voc was 1. 354 V, Jsc of 11.64 mA / cm ² , FF of 0.730, and Eff of 11.51%. When the spectral sensitivity current was measured, the front photoelectric conversion unit was 0.99, the rear photoelectric conversion unit was 1.01, and the total was 1.00, compared with Comparative Example 1.

  Comparative Example 2 exhibited almost the same characteristics as Comparative Example 1, and no significant change was observed in the increase in Jsc or the spectral sensitivity current of the front photoelectric conversion unit. That is, it can be said that the n-type amorphous silicon oxide 39 has no effect of reflecting light toward the front photoelectric conversion unit. Since the n-type amorphous silicon oxide 39 has a high refractive index of 3.0, the difference in refractive index from amorphous silicon or crystalline silicon is small, so that it can be said that almost no reflection effect is obtained. In order to reduce the refractive index of the amorphous silicon oxide, it is necessary to increase the oxygen concentration in the film as compared with the comparative example 2. In this case, the dark conductivity of the amorphous silicon oxide is lowered and the FF Since the reduction in Eff due to the reduction is inevitable, the refractive index cannot be lowered.

Example 1
As Example 1, a stacked photoelectric conversion device as shown in FIG. The difference from Comparative Example 1 is that an n-type silicon composite layer 4 having a thickness of 30 nm is provided between the front photoelectric conversion unit 3 and the rear photoelectric conversion unit 5. Other than that, it produced similarly to the comparative example 1.

The gas flow rate when forming the n-type silicon composite layer 4 is SiH ₄ / CO ₂ / PH ₃ / H ₂ = 5/10 / 0.1 / 1000 sccm. Power frequency is 13.56 MHz, power density is 100
The film was formed at mW / cm ² , a pressure of 100 Pa, and a substrate temperature of 200 ° C. At this time, the n-type silicon composite layer 4 has an oxygen concentration in the film of 42 atomic%, a refractive index for light of 600 nm is 2.0, and the TO of the crystalline silicon component with respect to the TO mode peak of the amorphous silicon component measured by Raman scattering. The peak intensity ratio of the mode peak was 2.0, and the dark conductivity was 5 × 10 ⁻⁶ S / cm.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light quantity of 100 mW / cm ² and measured for output characteristics at 25 ° C., Voc was 1.338V. , Jsc was 12.71 mA / cm ² , FF was 0.701, and Eff was 11.92%. When the spectral sensitivity current was measured, the front photoelectric conversion unit was 1.09, the rear photoelectric conversion unit was 1.06, and the total was 1.08, compared with Comparative Example 1.

In Example 1, although FF is slightly lowered as compared with Comparative Example 1, Jsc is increased by 1 mA / cm ² or more and Eff is improved. Further, the spectral sensitivity current of the front photoelectric conversion unit is increased by 9%, and it can be seen that the silicon composite layer 4 effectively reflects incident light to the front photoelectric conversion unit side. Further, the spectral sensitivity current of the rear photoelectric conversion unit is also increased by 6%, and it is considered that light scattering occurs in the silicon composite layer and the optical path length of the rear photoelectric conversion unit is increased.

(Example 2)
As Example 2, a stacked photoelectric conversion device as shown in FIG. The difference from Example 1 is that the n-type silicon composite layer 4 having a thickness of 30 nm is used for the n-type layer of the front photoelectric conversion unit 3 and the intermediate reflection layer and the n-type layer are combined. Other than that, the manufacturing method was the same as in Example 1, and the same film characteristics of the silicon composite layer 4 were used.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., Voc was 1. 340 V, Jsc was 13.29 mA / cm ² , FF was 0.692, and Eff was 12.32%. When the spectral sensitivity current was measured, the front photoelectric conversion unit was 1.14, the rear photoelectric conversion unit was 1.15, and the total was 1.14 as compared with Comparative Example 1.

  In Example 2, Jsc is further increased and Eff is improved compared to Example 1. Further, the spectral sensitivity current is increased in both the front photoelectric conversion unit and the rear photoelectric conversion unit as compared with the first embodiment. This is because the silicon composite layer 4 also serves as the n-type layer of the front photoelectric conversion unit, so that both the light reflected to the front photoelectric conversion unit side and the light transmitted to the rear conversion unit side are n for 30 nm in thickness. This is because it is no longer necessary to pass through the type microcrystalline silicon, and the absorption loss is reduced accordingly. However, FF is lower than that of Comparative Example 1 and Example 1, and it is considered that the contact resistance is increased at the interface of i-type amorphous silicon layer 32 / silicon composite layer 4.

(Example 3)
As Example 3, a stacked photoelectric conversion device as shown in FIG. The difference from the first embodiment is that the n-type layer of the front photoelectric conversion unit 3 is an n-type silicon composite layer 34 having a thickness of 30 nm which is a first n-type layer and an n-type layer having a thickness of 5 nm which is a second n-type layer. That is, the type microcrystalline silicon layer 35 is laminated. Other than that, the manufacturing method was the same as in Example 1, and the same film characteristics of the silicon composite layer were used.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., Voc was 1. 346 V, Jsc 13.03 mA / cm ² , FF 0.721, and Eff 12.65%. When the spectral sensitivity current was measured, the front photoelectric conversion unit was 1.12, the rear photoelectric conversion unit was 1.08, and the total was 1.12.

  In Example 3, Jsc is slightly reduced compared to Example 2, but FF is improved and Eff is improved. Although the spectral sensitivity current of the front photoelectric conversion unit is slightly lower than that of the second embodiment, it is higher than those of the first comparative example and the first embodiment. By inserting the n-type microcrystalline silicon layer 35 between the n-type silicon composite layer 34 and the p-type microcrystalline silicon 51, it is considered that the contact resistance at the n / p interface is reduced and the FF is improved. Since the thickness of the inserted n-type microcrystalline silicon layer 35 is as thin as 5 nm, the decrease in Jsc is not so large as compared to Example 2.

Example 4
As Example 4, a stacked photoelectric conversion device as shown in FIG. The difference from the first embodiment is that the n-type layer of the front photoelectric conversion unit 3 is an n-type microcrystalline silicon layer 36 having a thickness of 10 nm that is a first n-type layer and a 60-nm thickness that is a second n-type layer. That is, an n-type silicon composite layer 37 and a third n-type n-type microcrystalline silicon layer 38 having a thickness of 5 nm are stacked. Other than that, the manufacturing method was the same as in Example 1, and the same film characteristics of the silicon composite layer were used.

When the laminated photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., Voc was 1. 350 V, Jsc was 12.96 mA / cm ² , FF was 0.732, and Eff was 12.80%. When the spectral sensitivity current was measured, the front photoelectric conversion unit was 1.07, the rear photoelectric conversion unit was 1.11 and the total was 1.10.

  In Example 4, Jsc is slightly reduced compared to Example 3, but FF is further improved than Example 3 and Eff is improved. By inserting the n-type microcrystalline silicon layer 36 between the i-type amorphous silicon 32 and the n-type silicon composite layer 37, the contact resistance at the i / n interface is also reduced, and the FF is further improved than in the third embodiment. It is thought that it improved. The n-type microcrystalline silicon of the front photoelectric conversion unit has a total thickness of 15 nm of the first n-type layer and the third n-type layer, and is thinner than 30 nm of Example 1, so that the absorption loss is reduced. It is increased from 1.

(Example 5)
FIG. 16 shows the relative value of the spectral sensitivity current when the refractive index of the silicon composite layer is changed in the stacked photoelectric conversion device having the structure of Example 2 as Example 5 of the present invention. The silicon composite layer was produced in the same manner as in Example 1 except that the CO ₂ / SiH ₄ ratio was varied from 1 to 15. In FIG. 16, the horizontal axis represents the refractive index of the silicon composite layer with respect to light having a wavelength of 600 nm, and the vertical axis represents the relative value to the spectral sensitivity current of the stacked solar cell having the structure of Comparative Example 1 having no silicon composite layer. The spectral sensitivity current of the front photoelectric conversion unit increases as the refractive index decreases, and decreases when the refractive index is less than about 1.8. As the refractive index decreases, the light reflected to the front photoelectric conversion unit increases and the spectral sensitivity current increases. However, when the refractive index is less than about 1.8, the dark conductivity of the silicon composite layer decreases, and silicon decreases. It is considered that the current decreases because the influence of the increase in the resistance of the composite layer and the contact resistance at the interface cannot be ignored.

  The spectral sensitivity current of the rear photoelectric conversion unit increases as the refractive index decreases, and decreases when the refractive index is less than about 2. As the refractive index decreases, the transmittance of the silicon composite layer increases, so that the light reaching the rear photoelectric conversion unit increases and the current increases. If the refractive index is less than about 2, the amount of light reflected toward the front photoelectric conversion unit increases, and the influence of the decrease in the light reaching the rear photoelectric conversion unit cannot be ignored and the current decreases.

The total spectral sensitivity current obtained by adding the front photoelectric conversion unit and the rear photoelectric conversion unit also has a characteristic having a maximum value with respect to the refractive index. When the refractive index is 1.7 or more and less than 2.5, the entire spectral sensitivity current is increased as compared with Comparative Example 1. Further, in order to increase the entire spectral sensitivity current by 10% or more from Comparative Example 1, the refractive index needs to be 1.8 to 2.1.

(Example 6)
FIG. 17 shows a three-layer stacked photoelectric conversion device according to Example 6 of the present invention. 17 from the glass substrate 1 side to the second-stage photoelectric conversion unit 5a except that the film thickness of the i-type layer is different from the glass substrate 1 of Example 1 in FIG. 6 to the rear photoelectric conversion unit 5. It was produced by the same method as before. The thickness of the amorphous silicon that is the i-type layer of the first-stage photoelectric conversion unit is 100 nm, and the thickness of the crystalline silicon that is the i-type layer of the second-stage photoelectric conversion unit is 1.2 μm. On the second-stage photoelectric conversion unit 5a, a second silicon composite layer 7 having a thickness of 30 nm, a p-type microcrystalline silicon layer 81 having a thickness of 15 nm, an i-type crystalline silicon layer 82 having a thickness of 2.0 μm, A third-stage photoelectric conversion unit 8 composed of an n-type microcrystalline silicon layer 83 having a thickness of 15 nm was sequentially manufactured. Thereafter, Al-doped ZnO with a thickness of 90 nm and Ag with a thickness of 300 nm were sequentially formed as the back electrode 6 by sputtering. The first silicon composite layer 4a and the second silicon composite layer 7 have the same film characteristics as the silicon composite layer 4 of Example 1.

When the three-stage stacked photoelectric conversion device (light-receiving area 1 cm ² ) obtained as described above was irradiated with AM1.5 light at a light amount of 100 mW / cm ² and the output characteristics were measured at 25 ° C., Voc. Was 1.905 V, Jsc was 10.07 mA / cm ² , FF was 0.745, and Eff was 14.29%.

As Comparative Example 3, a three-stage stacked photoelectric conversion device having the same structure as that of Example 6 except that the first silicon composite layer 4 and the second silicon composite layer 7 were not provided. When the output characteristics of Comparative Example 3 were measured, Voc was 1.910 V, Jsc was 9.50 mA / cm ² , FF was 0.749, and Eff was 13.59%.

  Also in the three-stage stacked photoelectric conversion device, Jsc increased due to the reflection effect of the silicon composite layer, and Eff improved.

(Example 7)
FIG. 18 shows an integrated thin film photoelectric conversion module according to Example 7 of the present invention configured to include a stacked photoelectric conversion device. The structure of FIG. 18 is the same as that of FIG. 20 except that the ZnO intermediate reflection layer 105 of FIG. The thickness and manufacturing method of each layer were manufactured in the same manner as in Example 1. The size of the module was 910 mm × 455 mm, and 100 stages of photoelectric conversion cells were connected in series by dividing by patterning. The silicon composite layer 107 having the same film characteristics as the silicon composite layer 4 of Example 1 was used.

  As Comparative Example 4, an integrated thin film photoelectric conversion module having no intermediate reflection layer shown in FIG. 19 was produced. As Comparative Example 5, a structure having no third separation groove was produced by using ZnO having a thickness of 30 nm produced by sputtering as shown in FIG. 20 for the intermediate reflection layer. As Comparative Example 6, an integrated thin film photoelectric conversion module having a structure in which ZnO was used for the intermediate reflective layer and a third separation groove was provided as shown in FIG.

Table 2 summarizes output characteristics measured at 25 ° C. by irradiating the integrated thin film photoelectric conversion modules of Example 7 and Comparative Examples 4 to 6 with light of AM1.5 at a light amount of 100 mW / cm ² . Compared to Comparative Example 4 without the intermediate reflective layer, Comparative Example 5 with the ZnO intermediate reflective layer inserted significantly reduced Voc and FF, significantly reducing maximum power (Pmax) and conversion efficiency (Eff). Yes. This is because a leakage current is generated in the current path of the ZnO intermediate reflection layer 105, the connection groove 123, and the back electrode layer 106.

  In Comparative Example 6 provided with the third separation groove, the leakage current is suppressed, the short-circuit current (Isc) is increased, and Pmax is improved by about 3 W.

  In Example 7 using the silicon composite layer, Jsc was further increased as compared with Comparative Example 6, and Pmax was improved by about 10 W compared with Comparative Example 4. This is considered to be an improvement in Isc because the area loss corresponding to the need for the third separation groove is eliminated. In addition, since the front photoelectric conversion unit 104a, the silicon composite layer 107, and the rear photoelectric conversion unit 104b can be continuously formed by plasma CVD, there is no influence of air pollution at the interface between the silicon composite layer 107 and the rear photoelectric conversion unit 104b. It is thought that FF improved.

Furthermore, compared to Comparative Example 6, Example 7 eliminates the need for the third separation groove, so that the number of patterning operations can be reduced and one YAG laser can be reduced, thereby reducing the apparatus cost. Moreover, since it can be produced with the same equipment as the photoelectric conversion unit by using plasma CVD just by adding a CO ₂ gas line, the film forming equipment dedicated to the intermediate reflection layer such as sputtering required for ZnO is not necessary. The equipment cost was greatly reduced. Further, in comparison with Comparative Example 6, Example 7 has reduced the work of carrying in, heating, and carrying out the substrate to and from the plasma CVD apparatus, which is a vacuum device, and the patterning time has been reduced by one time. The time has been greatly reduced and the manufacturing cost has been reduced.

1 glass substrate, 2 transparent electrode layer, 3 front photoelectric conversion unit, 3a first photoelectric conversion unit, 31 p-type amorphous silicon carbide layer, 32 i-type amorphous silicon layer, 33 n-type microcrystalline silicon layer, 34 An amorphous silicon oxide alloy layer containing an n-type silicon crystal phase as a first n-type layer, 35 An n-type microcrystalline silicon layer as a second n-type layer, 36 An n-type microcrystal as a first n-type layer Silicon layer, 37 Amorphous silicon oxide alloy layer containing n-type silicon crystal phase as second n-type layer, 38 N-type microcrystalline silicon layer as third n-type layer, 39 n-type amorphous silicon oxide Layer, 4 amorphous silicon oxide alloy layer containing n-type silicon crystal phase, 5 backward photoelectric conversion unit, 5a second photoelectric conversion unit, 51 p-type microcrystalline silicon layer, 52 i-type crystalline silicon layer, 53 n Type microcrystalline silicon layer, 6 back electrode 7 Amorphous silicon oxide alloy layer containing the second n-type silicon crystal phase, 8 Third photoelectric conversion unit, 81 p-type microcrystalline silicon layer, 82 i-type crystalline silicon layer, 83 n-type microcrystalline silicon Layer, 101 integrated thin film photoelectric conversion module, 102 glass substrate, 103 transparent electrode layer, 104a front photoelectric conversion unit, 104b rear photoelectric conversion unit, 105 ZnO intermediate reflection layer, 106 back electrode layer, 107 n-type silicon crystal phase Amorphous silicon oxide alloy layer containing. 121 first separation groove, 122 second separation groove, 123 connection groove, 124 third separation groove.

Claims (7)

A stacked photoelectric conversion device including a plurality of photoelectric conversion units composed of pin junctions,
Including at least one portion sequentially formed by the first photoelectric conversion unit, the n-type silicon composite layer, and the second photoelectric conversion unit from the side close to the light incident side,
The silicon composite layer includes a silicon crystal phase dispersed in an amorphous alloy matrix of silicon and oxygen, contains an oxygen concentration in the film of 40 atomic% to 60 atomic%, and emits light having a wavelength of 600 nm. In contrast, the stacked photoelectric conversion device has a refractive index of 1.7 or more and 2.1 or less and a thickness of 50 nm or more and less than 130 nm.
  2. The stacked photoelectric conversion device according to claim 1, wherein a microcrystalline silicon layer having the same conductivity type as that of the silicon composite layer is adjacent to the silicon composite layer. A stacked photoelectric conversion device including a plurality of photoelectric conversion units composed of pin junctions,
Including at least one part composed of the first photoelectric conversion unit and the second photoelectric conversion unit sequentially from the side closer to the light incident side,
The i-type layer of the first photoelectric conversion unit includes an n-type layer on a side far from the light incident side, and at least a part of the n-type layer is an n-type silicon composite layer,
The silicon composite layer includes a silicon crystal phase dispersed in an amorphous alloy matrix of silicon and oxygen, contains an oxygen concentration in the film of 40 atomic% or more and 60 atomic% or less, and with respect to light having a wavelength of 600 nm. A multilayer photoelectric conversion device having a refractive index of 1.7 to 2.1 and a thickness of 50 nm to 130 nm.
4. The stacked photoelectric conversion device according to claim 3, wherein a portion of the n-type layer other than the silicon composite layer is a microcrystalline silicon layer. 5. The stacked photoelectric conversion device according to claim 1, wherein the silicon composite layer has a dark conductivity of 10 ⁻⁸ S / cm or more and 10 ⁻¹ S / cm or less.   6. The peak intensity ratio of the TO mode peak of the crystalline silicon component to the peak derived from the amorphous measured by Raman scattering in the silicon composite layer is 0.5 or more and 10 or less. A stacked photoelectric conversion device according to any one of the above.   The stacked photoelectric conversion device is stacked on a transparent substrate, and the reflection spectrum of light incident through the transparent substrate has at least one maximum value and minimum value of reflectance in the wavelength range of 500 nm to 800 nm. The stacked photoelectric conversion device according to claim 1, wherein a difference in reflectance between the maximum value and the minimum value is 1% or more.

Images