JP2004363577A - Semiconductor thin film, photoelectric conversion device and photovoltaic device using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an amorphous Si semiconductor film having high quality and excellent light stability, and to realize a photoelectric conversion device using the same and having a high efficiency and a low light deterioration rate.
SOLUTION: The ratio of the TA mode scattering peak intensity I _TA to the TO mode scattering peak intensity I _TO ( _ITA / I _TO ) containing at least silicon and hydrogen and obtained by a Raman scattering spectrum is 0.35 or less, and The semiconductor thin film has a crystallization ratio defined by the Raman scattering spectrum of 60% or less. As a result, an amorphous Si film in which the short-range ordered structure of the semiconductor thin film is significantly improved can be obtained, and the efficiency of a photoelectric conversion device using the film can be improved and the rate of photodegradation can be reduced.
[Selection diagram] Fig. 1

Description

  The present invention relates to an amorphous Si semiconductor thin film containing at least Si (silicon) and H (hydrogen), a photoelectric conversion device such as a solar cell using the semiconductor thin film, and a photovoltaic device using the same as power generation means. About.

  Devices using amorphous Si-based films typified by hydrogenated amorphous silicon (hereinafter also referred to as a-Si: H) are very diversified today. Solar cells have been particularly attracting attention as public interest has increased.

  The feature of amorphous Si in a solar cell is that the light absorption coefficient is much larger than that of crystalline Si (including microcrystalline Si film). When the film is formed by a CVD method (chemical vapor deposition method) such as a PECVD method (plasma CVD method), the productivity is much lower than when crystalline Si is used. Is to be excellent. However, it has long been known that amorphous Si has a photodegradation phenomenon called the Staebler-Wronski effect, and the efficiency of a solar cell using this film is reduced by about 10 to 15% by light irradiation. To the present day.

Although the mechanism of the photodegradation phenomenon has not yet been clarified, it is argued that it is a phenomenon related to hydrogen, and research on reducing photodegradation by at least two basic policies has been accumulated. Was. In other words, [1] the basic policy is to reduce the Si—H ₂ (die hydride) bonding state in the film (increase the proportion of the Si—H (monohydride) bonding state) and [2] reduce the hydrogen concentration in the film. There is. The basic policies of [1] and [2] have overlapping parts.

The above-mentioned [1] is a policy based on the fact that a film having a high Si—H ₂ density in the film has a large light degradation. Research on a specific method of reducing the density of Si—H _{2 in a} film has been carried out by many organizations under various hypotheses. For example, the origin of the Si—H ₂ bond state in the film may be described. It is thought that the higher silane molecules in the gas phase are taken into the film, or that the molecules become steric hindrance that disrupts the film growth reaction on the film formation surface. Methods are known to reduce the production of the relevant SiH ₂ molecules. Specifically, plasma of low electron temperature typified by VHF plasma is used in PECVD, hydrogen dilution rate is optimized (electron temperature is minimized), and substrate temperature is raised to some extent (upper limit is 350). Attempts have been made to promote the hydrogen elimination reaction from the film-forming surface. Under the conditions of a film-forming speed of 2 nm / sec, a-Si having a stabilization efficiency of 8.2% and a photo-deterioration rate of 12%: H single element characteristics are obtained (see Non-Patent Document 1). However, such stabilization efficiency and light degradation rate are not yet sufficient.

In the above [2], for example, in the film formation by the PECVD method, a method of extracting excessive hydrogen from the surface layer of the film growth while alternately repeating film deposition and hydrogen plasma treatment has been proposed (Patent Document 1). , 2, 3). However, in this method, the hydrogen concentration in the film is not sufficiently reduced as expected, and it is difficult to obtain a high film formation rate because the film deposition and the hydrogen plasma treatment are repeated many times. However, there is a problem that productivity is extremely low. Moreover, high efficiency and a low light degradation rate have not been achieved at the same time.
JP-A-5-166733 JP-A-6-120152 JP 2002-9317 A Applied Physics Vol.71 No.7 (2002) p.823

  As described above, in the above method [1], the stabilization efficiency and the light deterioration rate are not yet sufficient. In the above method [2], the hydrogen concentration in the film is not sufficiently reduced as expected, and the film deposition and hydrogen plasma treatment are repeated many times. Is difficult to obtain, and the productivity is extremely low. Also, high efficiency and low light degradation rate cannot be achieved at the same time.

  Therefore, the present invention has been made in view of the above-mentioned problems, and an object thereof is to provide an amorphous Si semiconductor film having high quality and excellent light stability and a high efficiency and low light degradation rate using the same. And to realize an excellent photoelectric conversion device and an excellent photovoltaic device.

The semiconductor thin film of the present invention 1) contains at least silicon and hydrogen, and has a ratio ( _ITA / _ITO ) of the TA mode scattering peak intensity I _TA to the TO mode scattering peak intensity I _TO obtained from the Raman scattering spectrum. 0.35 or less, and the crystallization ratio defined by the Raman scattering spectrum is 60% or less.

Also, 2) it contains at least silicon and hydrogen, the half-width of the scattering peak of the TO mode obtained by the Raman scattering spectrum is 65 cm ^-1 or less, and the crystallization ratio defined by the Raman scattering spectrum is 60% or less. It is characterized by being.

In the configuration according to 1) or 2) above, the sum (σ _S−H ) of the existing density σ _{S−H in the} Si—H bonding state and the existing density σ _{Si−H2 in} the Si—H ₂ bonding state in the semiconductor thin film. + the ratio of the density sigma _S-H of Si-H bonding state with respect to _{σ Si-H2) (σ S} -H / (σ S-H + σ Si-H2)) is equal to or is less than 0.95.

  Further, the photoelectric conversion device of the present invention includes: 4) at least one semiconductor junction unit including the semiconductor thin film according to any one of the above 1) to 3) in a photoactive portion; and an electrode for extracting power from the semiconductor junction unit. And characterized in that:

Here, particularly, 5) the semiconductor film of any of the above 1) to 3) is amorphous silicon, and in particular, the scattering peak intensity of the TA mode with respect to the scattering peak intensity I _TO of the TO mode obtained from the Raman scattering spectrum. If the ratio of _{I TA (I TA / I tO} ) is 0.175 or less, since a state in which distortion is further reduced in the film, it is suitable in order to reduce the rate of photodegradation of the photoelectric conversion device.

  6) In the amorphous silicon semiconductor thin film of the above 5), the crystallization ratio forms a gradient distribution or a stepwise distribution in the film thickness direction, and goes from the light incident side (front side) to the back side. To increase the crystallization rate. This makes it possible to obtain a band profile in which the optical bandgap energy decreases from the front side to the rear side, which is more than when the optical bandgap energy is constant in the film thickness direction. Efficient photoelectric conversion becomes possible and high efficiency can be realized.

  7) In the amorphous silicon semiconductor thin film of the above 5), the crystallization ratio has a spatial periodic distribution in the film thickness direction, and the length of one cycle is 100 nm or less, preferably 10 nm. The following is assumed. By doing so, a band profile in which the optical band gap energy is periodically modulated can be obtained, and more efficient photoelectric conversion can be achieved as compared with the case where the optical band gap energy is constant in the film thickness direction. And high efficiency can be realized. In particular, when the period length is set to 10 nm or less, a quantum size effect (formation of a quantum well structure) starts to appear, and higher efficiency can be achieved. It is desirable to control the crystallization rate by controlling the temperature of the thermal catalyst.

8) If the initial ESR spin density before light stabilization in the amorphous silicon semiconductor thin film of the above 5) is 5 × 10 ¹⁵ cm ⁻³ or less, this film is used as a photoactive portion. In addition, high device characteristics can be obtained without increasing the recombination current in the same layer.

  9) If the hydrogen concentration in the amorphous silicon semiconductor thin film of the above 5) is 7 atomic% or less, it greatly contributes to improvement of the light stability of the device characteristics.

  10) It is preferable that the optical band gap energy of the amorphous silicon semiconductor thin film of the above 5) is 1.7 eV or less. In other words, when this is larger than 1.7 eV, when this is applied to a photoactive layer of a single cell or the like, the long-wavelength light cannot be sufficiently absorbed and photoelectrically converted, so that the photocurrent is reduced. This leads to a decrease in

  11) When the amorphous silicon semiconductor thin film of the above 5) is formed by a CVD method (chemical vapor deposition method), a uniform and uniform film can be formed over a large area. It is particularly suitable for manufacturing large area devices such as batteries.

  Further, the amorphous silicon semiconductor thin film is preferably formed at a substrate temperature of 100 ° C. or more and 350 ° C. or less. If the substrate temperature is lower than 100 ° C., the hydrogen extraction effect does not work effectively, and the hydrogen concentration in the film cannot be reduced effectively. On the other hand, when the substrate temperature is higher than 350 ° C., the desorption of hydrogen from the film growth surface becomes remarkable, the dangling bond density in the film increases, and a high-quality film cannot be obtained. It is.

  12) If the amorphous silicon semiconductor thin film of 5) contains at least one of Ge (germanium), Sn (tin), C (carbon), N (nitrogen) and O (oxygen), the band gap Since the energy can be adjusted (in the case of Ge and Sn, the energy can be adjusted in the direction of low energy, and in the case of C, N and O, the energy can be adjusted in the direction of high energy), the band profile of the photoelectric conversion device The degree of freedom of design can be increased, and this is suitable for higher efficiency.

  Further, the photovoltaic device of the present invention is characterized in that 13) the photoelectric conversion device is used as power generation means, and the power generated by the power generation means is supplied to a load.

According to the semiconductor thin film of the present invention, at least silicon and hydrogen are contained, the ratio of the TA mode scattering peak intensity I _TA to the TO mode scattering peak intensity I _TO obtained by the Raman scattering spectrum is 0.35 or less, and the Raman scattering spectrum Is 60% or less, an amorphous Si film in which the short-range ordered structure of the film is greatly improved can be obtained, and the efficiency of the photoelectric conversion device using the same can be improved, and The deterioration rate can be reduced.

Further, it contains at least silicon and hydrogen, the half-width of the scattering peak of the TO mode obtained by the Raman scattering spectrum is 65 cm ⁻¹ or less, and the crystallization rate defined by the Raman scattering spectrum is 60% or less. An amorphous Si film in which the short-range order structure of the film is significantly improved can be obtained, and the efficiency of a photoelectric conversion device using the film can be increased and the rate of photodegradation can be reduced.

  Further, according to the semiconductor thin film of the present invention, the film quality can be easily managed by a simple method and apparatus using Raman scattering spectroscopy.

Further, according to the semiconductor thin film of the present invention, the ratio of the existing density of the Si—H bonding state to the sum of the existing densities of the Si—H bonding state and the Si—H ₂ bonding state in the semiconductor thin film is 0.95 or more. Therefore, light degradation caused by the Si—H ₂ bond can be significantly reduced, and the efficiency of a photoelectric conversion device using the same and the rate of light degradation can be reduced.

  Further, according to the photoelectric conversion device of the present invention, since one or more semiconductor junction units including the semiconductor thin film in the photoactive portion and electrodes for extracting power from the semiconductor junction units are provided, the light degradation rate is reduced. An excellent photoelectric conversion device such as a reduced solar cell with high efficiency can be provided.

  Furthermore, according to the photovoltaic power generation device of the present invention, the photoelectric conversion device is used as a power generation means, and the power generated by the power generation means is supplied to the load, thereby providing a high efficiency photovoltaic power generation device. it can.

  Hereinafter, embodiments of a semiconductor thin film of the present invention, a photoelectric conversion device using the same, and a photovoltaic device using the same as power generation means will be described in detail.

<< First Embodiment >>
In order to obtain an amorphous silicon (hereinafter referred to as Si) film of the present invention, a Cat-PECVD method (cathode plasma CVD method with a built-in thermal catalyst; Japanese Patent Application Laid-Open No. 2001-313272; An example formed using the method described in Japanese Patent Application Laid-Open No. 2001-293031 will be described below. Note that, as described later, the amorphous Si film has a crystallization rate defined by a Raman scattering spectrum falling within a range of 0 to 60% (that is, the amorphous Si film is non-crystalline). (Included in crystalline Si film) Conversely, the crystalline Si film has a crystallization ratio in the range of 60 to 100%.

<Cat-PECVD method>
When the Cat-PECVD method is used, the amorphous Si film of the present invention uses a VHF band (27 MHz or more: usually about 40 to 80 MHz) as a plasma excitation frequency, and uses Ta (tantalum) provided inside the cathode. ), A temperature of a thermal catalyst made of a high melting point material such as W (tungsten) or C (carbon) is 1400 to 2000 ° C., a gas flow ratio H ₂ / SiH ₄ is 2 to 20, a substrate temperature is 100 to 350 ° C. It can be obtained under the conditions that the gas pressure is set to 13 to 665 Pa and the VHF plasma power density is set to 0.01 to 0.5 W / cm ² . Here, until H ₂ and SiH ₄ are released into the film forming space, they are led in different states through different gas introduction paths, and the H ₂ gas introduction path is part of the paths. Is arranged in the cathode, and the thermal catalyst is not arranged in the SiH ₄ gas introduction path. In this way, by activating only H ₂ with the thermal catalyst, the film is deposited and consumed in the gas introduction path until SiH ₄ is decomposed and activated by the thermal catalyst and released into the film forming space. Can be prevented, and at the same time, the effect of using a thermal catalyst described below can be obtained.

By a gas heating effects due to the use of thermal catalyst in the cat-PECVD method, the high-order silane formation reaction is exothermic (SiH ₂ insertion reaction to _Si n _{H m} molecules) is suppressed, a film The density of higher order silane molecules in the space can be kept low, and the incorporation of higher order silane molecules into the film and steric hindrance due to the same molecules on the film formation surface are reduced.

As a result, it becomes possible to form an amorphous Si film in which the existence density of the Si—H ₂ bond state in the film is greatly reduced. In addition, since the gas heating effect also has the effect of lowering the electron temperature of the plasma, the generation of SiH ₂ by the one-electron collision reaction between the source gas SiH ₄ and the electrons in the plasma space (similarly, the generation of SiH ₃ by the one-electron collision reaction). is compared require high electron temperature (collision energy) by) is suppressed, whereby it is possible to suppress the high-order silane formation reaction (SiH ₂ insertion reaction to Si _n H _m molecules) also. Furthermore, since the Cat-PECVD method can also promote the generation of hydrogen radicals, the hydrogen abstraction reaction from the film formation surface can be promoted, and the hydrogen concentration in the film can be effectively reduced. By these effects, it is possible to obtain an amorphous Si film excellent in photostability, in which short-range order (SRO: Short Range Order) described later is greatly improved.

  In the Cat-PECVD method, crystallization can be promoted by the effect of introducing a thermal catalyst even under film forming conditions that were difficult with the conventional PECVD method. Particularly preferred. Here, the origin of the crystallized region in the film is as follows: (1) those caused by the crystallization reaction on the film-forming surface, and (2) Si fine particles (crystal clusters of nanometer size) generated in the plasma space. There are at least two types, which are caused by being taken into the film.

<Crystallization rate>
The crystallization rate can be freely controlled in a range of 0 to 100% by a combination of a hydrogen dilution rate (H ₂ / SiH ₄ gas flow ratio), a thermal catalyst temperature, a VHF power density, a film forming gas pressure, and a substrate temperature. However, if the crystallization ratio is too large, the characteristics as an amorphous system will be lost. Therefore, the crystallization ratio is adjusted in the range of 0 to 60%, more preferably 0 to 30%. % (A Si film having a crystallization ratio of 60% or more usually falls within the category of a microcrystalline Si film, and the light absorption coefficient at a wavelength shorter than about 700 nm is almost the same as that of a crystalline system. To the same extent and does not meet the purpose of the present invention).

  Further, the efficiency of a photoelectric conversion element, which will be described later, can be improved by setting the crystallization ratio to have a gradient distribution or a stepwise distribution in the film thickness direction or a periodic distribution in terms of structure (not shown). That is, in the case of a slope distribution or a stepwise distribution, the crystallization rate is increased from the light incident side (front side) to the back side. By doing so, it is possible to obtain a band profile in which the optical bandgap energy decreases from the front side to the rear side, compared to the case where the optical bandgap energy is constant in the film thickness direction. More efficient photoelectric conversion (that is, higher efficiency) becomes possible. In the case of a periodic distribution, the length of one period is 100 nm or less, preferably 10 nm or less. In this way, a band profile in which the optical band gap energy is periodically modulated can be obtained, and more efficient photoelectric conversion than when the optical band gap energy is constant in the film thickness direction. (That is, higher efficiency) can be achieved. In particular, when the period length is set to 10 nm or less, a quantum size effect (formation of a quantum well structure) starts to appear, and higher efficiency can be achieved. Here, it is desirable to control the crystallization rate by controlling the temperature of the thermal catalyst. That is, since the temperature control of the thermal catalyst can be performed at very high speed and with high accuracy, it is extremely suitable for controlling the crystallization rate at high speed and with high accuracy during film formation.

In addition, the crystallization ratio referred to in this specification is defined as the crystal phase peak intensity / (crystal phase peak intensity + amorphous phase peak intensity) in the Raman scattering spectrum. a peak intensity of a peak intensity + 520 cm ^-1 at 510 cm ^-1, also an amorphous phase peak intensity shall be defined as the peak intensity at 480 cm ^-1.

<Raman scattering characteristics>
When describing the structure of an amorphous Si film, the degree of order is often handled. Particularly, short range order (SRO) has a very close relationship with film quality and photostability. It is said that there is. As a typical method for evaluating the vibration dynamics of an amorphous network in which the SRO is reflected, Raman scattering spectroscopy using a simple and simple device is known.

FIG. 1 shows Raman scattering spectra of the amorphous Si films of the prior art and the present invention. Each spectrum, TO (horizontal optical vibration) band having a peak near 480cm ^-1, 385cm ^-1 LO (longitudinal optical vibrations) having a peak near zone, LA (vertical type with a peak near 305 cm ^-1 Acoustic vibration) band and TA (horizontal acoustic vibration) band energy band having a peak near 160 cm ⁻¹ . Table 1 shows the spectrum divided into the above four energy bands (modes) by a Gaussian type + Lorentzian type approximation curve, and shows the peak intensity and half width of each mode. It shows a value normalized by setting the peak intensity (TO peak intensity) of the TO mode to 1.000. Here, for the measurement of the Raman spectrum, a Renishaw Ramanscope System 1000 using a He-Ne laser (wavelength 632.8 nm) as the excitation light was used.

  The film 1 of the present invention is an amorphous Si film produced by using the Cat-PECVD method and having a crystallization rate defined by a Raman scattering spectrum of 0%, and the film 2 of the present invention is An amorphous Si film containing a small amount of a crystalline Si component, which was evaluated using the Cat-PECVD method and was evaluated to have a crystallization rate of 3%, and a conventional film which was a conventional example had a crystallization rate of 3%. It is an amorphous Si film manufactured using a conventional PECVD method evaluated as 0%.

Now, the distribution width Δθ of bond angle θ of Si, the ratio of the scattering peak intensity _{I TO} and TA band scattering peak intensity _{I TA} of TO mode _{(= I TA / I TO)} , or the half-width of the scattering peak of TO mode It is known that the film has a positive correlation with HW _TO . However, when looking at I _TA / I _TO in the film of the present invention, the conventional example takes a value of I _TA / I _TO = 0.41. On the other hand (the _ITA / _ITO ratio of the conventional amorphous Si film is at least a value greater than 0.35), the present invention film 1 has _ITA / _ITO = 0.17, and the present invention film 2 has _ITA / _ITO / _0.15 . It was found that I _TO = 0.25, which is about 2/5 to 3/5 that of the conventional case, and it was also found that although not shown in the table, it may be reduced to about 1/3. As for the half-value width _{HW TO} scattering peak of TO mode, whereas the conventional membranes take half width 65~75cm ^-1 (in Table 68cm ^-1), narrower than this with a film of the present invention Turned out to be.

  The above shows that the variation Δθ of the Si bond angle of the film of the present invention is smaller than that of the conventional film, that is, the SRO is improved. It is considered that this structural change (improvement) related to SRO is largely related to the improvement result of light stability described later.

<ESR spin density>
Before the light irradiation (initial stage), the Si films of the films 1 and 2 of the present invention have a Si dangling bond density, which is an ESR spin density measured by an electron spin resonance method, at 3.5 × 10 ¹⁵ cm ⁻³ and 2.6, respectively. × 10 ¹⁵ cm ^-3 , and a low defect density of 5.0 × 10 ¹⁵ cm ^-3 or less was also confirmed. When the dangling bond density is higher than the above value, it has been found that when applied to the active layer, the recombination current in the same layer increases and high device characteristics cannot be obtained. In addition, JES-RE3X manufactured by JEOL was used for ESR spin density measurement.

<FT-IR Characteristics (Ratio of Si-H Bonded State)>
Further, the Si films of the films 1 and 2 of the present invention have an abundance density σ _S-H of the Si—H bond state evaluated by FT-IR (Fourier transform infrared spectroscopy) analysis. the sum of the _S-H and _{Si-H 2} bonds state of the density _{σ Si-H2 (σ S-} H + σ Si-H2) for the ratio _{(σ S-H / (σ} S-H + σ Si-H2); hereinafter, Si-H bond state existing ratio) was 0.990 and 0.985, respectively. This Si—H bond state existence ratio has a large correlation with photodegradation as already described in the description of the related art, and the higher this ratio is (the more the Si—H bond is dominant, the higher the ratio). ) It is reported that light degradation is reduced. In the Si film of the present invention, the value of 0.95 or more, which is said to be significantly higher than the Si-H bonding state existence ratio of about 0.90 obtained by the conventional plasma CVD method, and that the light deterioration suppressing effect appears remarkably appears. Have been obtained. The FT-IR characteristics were measured using FTIR-8300 manufactured by Shimadzu Corporation.

<Hydrogen concentration in film>
The hydrogen concentration in the films was evaluated by FT-IR analysis to be 6.9 atomic% and 5.3 atomic% for the Si films of the films 1 and 2 of the present invention, respectively. At this time, the calculation of the hydrogen concentration in the film was performed using the absorption peak area near 630 cm ⁻¹ and A value = 1.6 × 10 ¹⁹ cm ⁻² . As described above, the film of the present invention has a low hydrogen concentration of 7 atomic% or less, which is also considered to greatly contribute to the improvement of the photostability of the device characteristics described later. If the hydrogen concentration in the film is higher than the above, the existence density of the Si—H ₂ bond state in the film is increased, and the light deterioration rate is increased.

<Optical band gap>
In terms of optical properties, the optical band gap energies Eg.opt of the films 1 and 2 of the present invention were evaluated as 1.60 and 1.55 eV, respectively, in a so-called cube root plot, and the gap was narrowed to 1.7 eV or less. It was confirmed that. When the optical bandgap energy Eg.opt is larger than 1.7 eV, when this is applied to a photoactive layer of a single cell or the like, long-wavelength light cannot be sufficiently absorbed or photoelectrically converted, so that the photocurrent is reduced. This leads to a reduction in device characteristics.

<Band gap adjusting element>
The bandgap energy Eg of the amorphous Si film can be adjusted to some extent by the hydrogen concentration in the film. However, when it is desired to adjust the Eg while keeping the hydrogen concentration in the film low, it is possible to add an Eg adjusting element. Good. Specifically, a gas containing molecular formulas such as Ge or Sn for narrow gap and C, N or O for wide gap may be introduced into the film forming space. That, _GeH 4 in case of incorporating the Ge (H contains deuterium _{D), Ge n H 2n +} 2 (n is a positive integer, the same applies hereinafter) of, _GeX 4 (X is a halogen element) such as, containing Sn _SnH 4 in the case of (H contains deuterium _{D), Sn n H 2n +} 2, SnX 4 (X is halogen), _SnR 4 (R is an alkyl group) and the like, CH ₄ in the case of incorporating the C (H includes deuterium D), C ₂ H ₂ , C _n H _{2n + 2} , C _n H _2n , CX ₄ (X is a halogen element), and N ₂ , NO _n , N _When O is contained in ₂ O, NH _3, or the like, O ₂ , CO, CO ₂ , NO _n , N ₂ O, H ₂ O, or the like may be introduced.

<Substrate temperature>
Next, the substrate temperature is set in the range of 100 ° C to 350 ° C. This is because if the substrate temperature is lower than 100 ° C., the hydrogen extraction effect does not work effectively and the hydrogen concentration in the film cannot be reduced effectively.If the substrate temperature is higher than 350 ° C., the film growth This is because the desorption of hydrogen from the surface becomes remarkable, the dangling bond density in the film increases, and a high-quality film cannot be obtained.

<Photoelectric conversion device>
Next, an example in which the amorphous Si film of the present invention is applied to a photoactive layer of a photoelectric conversion device such as a solar cell will be described in detail. That is, a description will be given of a photoelectric conversion device including one or more semiconductor bonding units including the above-described semiconductor thin film in the photoactive portion, and electrodes for extracting power from these semiconductor bonding units.

  FIG. 2 shows an example in which the amorphous Si film of the present invention is applied to a photoactive layer for a super-straight tandem-type thin-film Si (silicon) solar cell element, which is one of the photoelectric conversion devices. It is. In the figure, 1 is a light-transmitting substrate, 2 is a first electrode which is a front electrode, 3 is a semiconductor multilayer film, 31 is a first semiconductor bonding layer, 32 is a second semiconductor bonding layer, and 31a is a first semiconductor bonding layer. A p-type semiconductor layer in the first semiconductor bonding layer 31, 31b is a photoactive layer in the first semiconductor bonding layer 31, 31c is an n-type semiconductor layer in the first semiconductor bonding layer 31, and 32a is a second semiconductor. A p-type semiconductor layer in the bonding layer 32, 32b is a photoactive layer in the second semiconductor bonding layer, 32c is an n-type semiconductor layer in the second semiconductor bonding layer 32, and 4 is a second electrode which is a back electrode. It is. Further, thick arrows in the drawing indicate the incident direction of light (hν). Note that an intermediate layer made of a transparent conductive film, a thin metal layer, or silicide may be inserted between the n-type semiconductor layer 31c and the p-type semiconductor layer 32a (not shown).

  In order to manufacture such a photoelectric conversion device, first, the translucent substrate 1 is prepared. Here, as the translucent substrate 1, a plate material or a film material made of glass, plastic, resin, or the like can be used.

Next, the first electrode 2 serving as a front electrode is formed on the translucent substrate 1. As the first electrode 2, a known oxide transparent conductive film can be used. Specifically, materials such as SnO ₂ which is a tin oxide, ITO which is an indium-tin oxide, and ZnO which is a zinc oxide can be used. The thickness of the transparent conductive film is adjusted in the range of about 60 to 600 nm in consideration of the antireflection effect and the reduction in resistance. As a guide for lowering the resistance, it is desirable to set the sheet resistance to about 10Ω / □ or less. Note that this transparent conductive film is exposed to an active hydrogen gas atmosphere due to the use of SiH ₄ and H ₂ when a Si film is subsequently formed thereon, and thus has excellent reduction resistance. It is desirable to form a ZnO film at least as a final surface. The thickness of this ZnO film is in the range of 10 to 100 nm. As a method of forming the transparent conductive film, known techniques such as a thermal CVD method, an evaporation method, an ion plating method, a sputtering method, a spray pyrolysis method, and a sol-gel method can be used.

  Note that, before forming the transparent electrode, an uneven structure is formed on the surface of the glass substrate by a method such as RIE (reactive ion etching) or blasting, and then the transparent electrode is formed on the uneven structure, thereby increasing the light scattering effect. The light confinement effect is promoted, which is suitable for improving efficiency.

  Next, the semiconductor multilayer film 3 is formed. The semiconductor multilayer film 3 includes a first semiconductor bonding layer 31 and a second semiconductor bonding layer 32. As a production method, in addition to the known PECVD method and Cat-CVD method, the present inventors have already disclosed Cat-type compounds disclosed in Japanese Patent Application Nos. 2000-130858, 2001-293031 and 2002-38686. A PECVD method can be used.

  First, a first semiconductor bonding layer 31 is formed on the first electrode 2. The semiconductor junction layer 31 has a pin junction in which a p-type semiconductor layer 31a, a photoactive layer 31b (substantially i-type layer), and an n-type semiconductor layer 31c are sequentially stacked.

Here, as the p-type semiconductor layer 31a, a conventional amorphous Si film, a crystalline Si film containing a crystalline Si phase and having a crystallization ratio of 60% or more, or an amorphous Si film of the present invention is used. it can. The film thickness is adjusted in the range of about 2 to 100 nm depending on the material. The concentration of the doping element (for example, B (boron)) is about 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , and is substantially a p ⁺ type. In addition, if an appropriate amount of a gas containing C (carbon) such as CH ₄ is mixed in addition to SiH ₄ , H ₂ and a gas such as B ₂ H ₆ which is a doping gas used at the time of forming the film, the Si _x C _1-x film is formed. And is very effective in forming a window layer with small light absorption loss, and also effective in reducing dark current components for improving open circuit voltage. A similar effect can be obtained by mixing an appropriate amount of a gas containing O (oxygen) or a gas containing N (nitrogen) in addition to C.

For the substantially i-type semiconductor layer, which is the photoactive layer 31b, the amorphous Si film of the present invention is used, and the film thickness is adjusted in the range of about 0.1 to 0.5 μm. The film formation conditions at this time were as follows: a plasma excitation frequency of 60 MHz, a catalyst temperature of 1600 to 2000 ° C., a H ₂ / SiH ₄ flow ratio of 5 to 10, a substrate temperature of 200 to 250 ° C., and a gas pressure of 133 to 200 Pa. The VHF power density is set to 0.1 to 0.3 W / cm ² . Under such film forming conditions, the amorphous Si film of the present invention can be formed at a high film forming rate of 2 nm / sec. Here, since the non-doped film usually actually shows a slight n-type characteristic, in this case, the p-type doping element (for example, B (boron)) is slightly contained and substantially i-type. It can be adjusted to be a mold. In some cases, for the purpose of fine adjustment of the internal electric field intensity distribution, an n ^- type or a p ^- type may be used.

As the n-type semiconductor layer 31c, a conventional amorphous Si film, a crystalline Si film containing a crystalline Si phase and having a crystallization rate of 60% or more, or the amorphous Si film of the present invention can be used. . The thickness is adjusted in the range of about 2 to 100 nm depending on the material. The concentration of the doping element (for example, P (phosphorus)) is set to about 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , and is substantially an n ⁺ type. Note _SiH 4 used at the time of film formation, _{H 2,} and in addition to gas, such as PH ₃ is a doping gas, as described above, when an appropriate mixture of gas containing C, such as CH _{4 Si x C 1-x} A film can be obtained, and a film with little light absorption loss can be formed, and it is also effective in reducing a dark current component for improving an open circuit voltage. A similar effect can be obtained by mixing an appropriate amount of a gas containing O or a gas containing N in addition to C.

In order to further improve the junction characteristics, a substantially i-type non-single-crystal Si layer is provided between the p-type semiconductor layer 31a and the photoactive layer 31b or between the photoactive layer 31b and the n-type semiconductor layer 31c. the or non-single-crystal _{Si x C 1-x} layer may be inserted as a buffer layer (not shown). At this time, the thickness of the insertion layer is about 0.5 to 50 nm.

  Here, in order to realize good recombination characteristics at the junction between the first semiconductor bonding layer 31 and the second semiconductor bonding layer 32 described below (that is, to realize ohmic contact-like electrical connection characteristics). For this reason, in the n-type semiconductor layer 31c included in the first semiconductor bonding layer 31 and the p-type semiconductor layer 32a included in the later-described second semiconductor bonding layer 32, at least a portion where they are in contact with each other has a lower crystallization ratio. It is desirable to keep it high. At this time, if the crystallization ratio is set to 60% or more, tunnel junction characteristics can be obtained, and good reverse junction recombination characteristics can be realized. Note that the crystalline reverse junction is formed by forming a microcrystalline Si layer of the same conductivity type about 5 to 30 nm following the n-type semiconductor layer 31 c and subsequently forming a microcrystalline Si layer of the opposite conductivity type about 5 to 30 nm. Then, the n-type semiconductor layer 31c and the p-type semiconductor layer 32a may be formed independently by inserting a dedicated crystalline Si layer such that a p-type semiconductor layer 32a is formed (not shown). Illustrated).

In order to realize ohmic contact electrical connection characteristics between the first semiconductor bonding layer 31 and the second semiconductor bonding layer 32, a transparent conductive film, a thin metal layer, or a thin silicide layer (silicon and metal (An unalloyed layer) as a conductive intermediate layer (not shown). Here, as the transparent conductive film, SnO ₂ which is a tin oxide, ITO which is an indium-tin oxide, ZnO which is a zinc oxide, or the like can be used. Here, when a transparent conductive film is used, an optical effect (reflection and transmission characteristics) can be introduced by the presence of the transparent conductive film, and therefore, the use of a transparent conductive film is extremely excellent in terms of high efficiency. That is, by adjusting the thickness of the transparent conductive film, the short-wavelength light is reflected by the transparent conductive film and re-enters the first semiconductor bonding layer 31 preferentially, and the long-wavelength light has the same antireflection effect. According to the principle, the second semiconductor bonding layer 32 can be preferentially confined, and more efficient photoelectric conversion of light energy becomes possible. In the case where a thin metal layer or a silicide layer is used, an effect can be expected that an ohmic contact between the first semiconductor bonding layer 31 and the second semiconductor bonding layer 32 can be more reliably and simply realized with a high yield.

  Next, a second semiconductor bonding layer 32 is formed on the first semiconductor bonding layer 31. The second semiconductor junction layer 32 is formed of a pin junction in which a p-type semiconductor layer 32a, a photoactive layer (substantially i-type semiconductor layer) 32b, and an n-type semiconductor layer 32c are sequentially stacked.

Here, as the p-type semiconductor layer 32a, a conventional amorphous Si film, a crystalline Si film containing a crystalline Si phase and having a crystallization rate of 60% or more, or an amorphous Si film of the present invention is used. it can. The thickness is adjusted in the range of about 2 to 100 nm depending on the material. The concentration of the doping element (for example, B (boron)) is about 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , and is substantially a p ⁺ type. Incidentally, it used during film _SiH 4, _{H 2,} and when an appropriate mixture of gas containing C (carbon), such as CH ₄ in addition to the gas such as _B 2 _{H 6} is a doping gas _{Si x C 1-x} A film is obtained, which is very effective for forming a window layer with small light absorption loss, and is also effective for reducing a dark current component for improving an open circuit voltage. A similar effect can be obtained by mixing an appropriate amount of a gas containing O (oxygen) or a gas containing N (nitrogen) in addition to C.

When the amorphous Si film of the present invention is used, the thickness of the substantially i-type semiconductor layer, which is the photoactive layer 32b, is adjusted in the range of about 0.5 to 1 μm, and When a typical crystalline Si film is used, the thickness is adjusted in a range of about 1 to 3 μm. At this time, when the amorphous Si film of the present invention is used, the film forming conditions are as follows: a plasma excitation frequency of 60 MHz, a catalyst temperature of 1600 to 2000 ° C., an H ₂ / SiH ₄ flow ratio of 5 to 10, a substrate When the temperature is 200 to 250 ° C., the gas pressure is 133 to 200 Pa, the VHF power density is 0.1 to 0.3 W / cm ² , and a crystalline Si film is used, the flow rate ratio of H ₂ / SiH ₄ to the above conditions is Is 5 to 20, the gas pressure is 200 to 665 Pa, and the VHF power density is 0.2 to 0.5 W / cm ² . Under such film formation conditions, the amorphous Si film or the crystalline Si film of the present invention can be formed at a high film formation rate of 2 nm / sec. Here, since the non-doped film usually shows a slight n-type characteristic in this case, in this case, the p-type doping element (for example, B (boron)) is slightly contained, and the i-type film is substantially formed. It can be adjusted so that In some cases, for the purpose of fine adjustment of the internal electric field intensity distribution, an n ^- type or a p ^- type may be used. In the case where the crystalline Si film is used for the photoactive layer 32b, the film structure is an aggregate of (110) -oriented columnar crystal grains, and the surface shape after film formation is a spontaneous suitable for optical confinement. It is desirable to have an uneven structure.

As the n-type semiconductor layer 32c, a conventional amorphous Si film, a crystalline Si film containing a crystalline Si phase and having a crystallization rate of 60% or more, or an amorphous Si film of the present invention can be used. . The thickness is adjusted in the range of about 2 to 100 nm depending on the material. The concentration of the doping element (for example, P (phosphorus)) is set to about 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , and is substantially n ⁺ type. In addition, if an appropriate amount of a gas containing C (carbon) such as CH ₄ is mixed in addition to SiH ₄ and H ₂ used for film formation and a gas such as PH ₃ as a doping gas, a Si _x C _1-x film is obtained. As a result, it is possible to form a film with less light absorption loss, and it is also effective in reducing the dark current component for improving the open-circuit voltage. A similar effect can be obtained by mixing an appropriate amount of a gas containing O (oxygen) or a gas containing N (nitrogen) in addition to C.

  In order to further improve the junction characteristics, a substantially i-type non-single-crystal Si layer is buffered between the p-type semiconductor layer 32a and the photoactive layer 32b or between the photoactive layer 32b and the n-type semiconductor layer 32c. It may be inserted as a layer (not shown). At this time, the thickness of the insertion layer is about 0.5 to 50 nm.

  Next, a metal film is formed as the second electrode 4 serving as the back electrode. As this metal film material, it is desirable to use Ag (silver) which is excellent in the conductive property and the light reflection property in the Si medium. It is also possible to use. Of course, a stacked body of Ag and Al or an alloy material containing Ag and Al may be used. Further, the second electrode has a two-layer structure, the first layer on the semiconductor layer side is the metal film containing Ag or Al, and the second layer laminated on the first layer is less expensive than Ag. It may be a metal material layer. As the film forming method, known techniques such as a vapor deposition method, a sputtering method, an ion plating method, a screen printing method, and a plating method can be used. At this time, the film thickness is about 0.1 μm or more.

It is more preferable that the second electrode 4 has a structure in which a transparent conductive film / metal film is stacked in this order from the surface contacting the semiconductor layer. By inserting the transparent conductive film between the semiconductor layer and the metal film in this manner, it is possible to suppress a phenomenon that the metal film component is diffused into the semiconductor layer and deteriorates device characteristics. In addition, if the surface on which the transparent conductive film is formed has an appropriate uneven structure, light can be effectively scattered, so that the light confinement effect that is effective for improving the efficiency of the solar cell can be enhanced. Here, as the transparent conductive film material, SnO ₂ , ITO, ZnO or the like can be used as described above, and as a film forming method, a CVD method, an evaporation method, an ion plating method, a sputtering method, a spray method, A known technique such as a sol-gel method can be used.

<Device characteristics>
Table 2 shows the characteristics of the device manufactured by the above method in comparison with the conventional example. The devices shown in Table 2 are those in which the photoactive layer 32b of the second semiconductor bonding layer 32 is formed of crystalline Si.

As can be seen from Table 2, the element applied with the film 1 of the present invention and the element applied with the film 2 of the present invention have a high initial conversion efficiency despite the formation of the photoactive layer 31b at a high film forming rate, and have a high photo-degradation efficiency. It was also found that the rate was extremely lower than that of the conventional film application element. From this, it was proved from the element characteristics that the amorphous Si film of the present invention had high quality and high light stability. In addition, the characteristics after light irradiation shown in Table 2 were measured after simulated sunlight of AM 1.5 and 100 mW / cm ² was continuously irradiated at a substrate temperature of 48 ° C. for 200 hours. The initial state before light irradiation is realized according to the annealing method shown in International Standard IEC1646.

<< Embodiment 2 >>
Next, among the film forming conditions described in the first embodiment, several kinds of amorphous Si semiconductor films having a crystallization ratio in the range of 0 to 65% were prepared using the VHF power density and the film forming gas pressure as parameters. Table 3 shows the evaluation results obtained by converting the device into an element according to the procedure described in the first embodiment.

As can be seen from Table 3, a film having a crystallization rate of 60% or less has a small photodegradation rate and a relatively high conversion efficiency, but a film having a crystallization rate of 30% or less has a particularly high efficiency. Characteristics (initial efficiency of about 11% or more) are obtained. The reason why the crystallization ratio is 0% and the light deterioration ratio is high is that the TA / TO peak intensity ratio and the TO band half width are large. As the crystallization rate becomes higher, the efficiency decreases due to a decrease in Jsc (short-circuit current density) and a decrease in Voc (open-circuit voltage) based on a decrease in the light absorption rate. It was found that the efficiency was greatly reduced when the pressure exceeded. Further, from the same table, it is found that a film having a TA / TO peak intensity ratio of 0.35 or less or a TO band half-width of 65 cm ^-1 or less can obtain a much lower light degradation rate than the conventional film (the conventional 10 to 15). % To about 8% or less).

<< Embodiment 3 >>
Next, of the film forming conditions described in the first embodiment, the catalyst body temperature in the Cat-PECVD method was used as a parameter, and the Si—H bond was formed using the same method and apparatus described in the first embodiment. Table 4 shows the measurement results of the state existence ratio and the light deterioration ratio. SiH ₄ / H ₂ in the table is a gas flow ratio of SiH ₄ gas to H ₂ gas.

  According to Table 4, the existence ratio of the Si—H bond state in the film is 0.95 or more (in this example, there are only two cases of 0.967 of the film 3 of the present invention and 0.997 of the film 4 of the present invention, but 0.934 of the conventional film 2 and 0.934 of the present film). It is considered that there is an inflection point in the middle of 0.967 (0.951) of the film 3.) It can be seen that the light deterioration rate is significantly reduced. The table also shows that the Cat-PECVD method is a very suitable method for this control.

<< Photovoltaic device >>
A photovoltaic device can be provided in which the above-described photoelectric conversion device is used as power generation means, and power generated from the power generation means is supplied to a load. That is, one or more of the above-described photoelectric conversion devices (in the case of a plurality of photoelectric conversion devices, these are connected in series, in parallel, or in series-parallel) are used as power generation means (in the case of a plurality of photoelectric conversion devices, these are connected in series, What is connected in parallel or series-parallel is used as the power generating means), and the generated power may be supplied directly to the DC load.

  In addition, the DC power generated by the above-described power generation means is converted into appropriate AC power by power conversion means such as an inverter, and the converted power is converted into an AC load such as a commercial power supply system or various electric devices. A photovoltaic device that can be supplied to

  Further, such a photovoltaic device can be used as a photovoltaic device such as a photovoltaic power generation system of various modes by installing the photovoltaic device on a roof or a wall of a sunny building.

<< Generalization >>
In the above description of the embodiment, an example is described in which the amorphous Si film of the present invention is applied to a tandem solar cell having two semiconductor junctions in a semiconductor multilayer film. Amorphous Si films include single-junction solar cells with one semiconductor junction (not shown), triple-junction solar cells with three semiconductor junctions (not shown), and more. The present invention can be applied to a multi-junction solar cell (not shown) having a semiconductor junction, and the same effect can be obtained.

  Although the solar cell in which the pin junction of the semiconductor bonding layer is formed in the order of pin from the light receiving surface side has been described, the same effect can be obtained with a solar cell formed in the order of nip from the light receiving surface side.

  In addition, although the superstrate type solar cell in which light is incident from the substrate side has been described, the same effect can be obtained for a substrate type solar cell (not shown) in which light is incident from the semiconductor film side. In the case of a substrate type, the substrate is not limited to a light-transmitting substrate, but may be a non-light-transmitting substrate such as stainless steel. The electrode may be a translucent material.

  Further, in the above description of the embodiment, the case where the amorphous Si film of the present invention is formed by using the Cat-PECVD method, which is a kind of the PECVD method, is described, but the manufacturing method is not limited to this. However, this does not exclude the possibility of forming by the conventional PECVD method by adjusting the film forming conditions. In addition, the possibility of forming by a CVD method other than the PECVD method typified by the Cat-CVD method (catalytic CVD method) by adjusting the film forming conditions is not excluded.

  Further, the amorphous Si film of the present invention can be applied to photoelectric conversion devices such as photodiodes, phototransistors, and optical sensors in addition to solar cells.

<< Summary >>
Thus, according to the semiconductor thin film of the present invention, the semiconductor thin film contains at least silicon and hydrogen, and the ratio of the scattering peak intensity I _TA of the TA mode to the scattering peak intensity I _TO of the TO mode obtained by the Raman scattering spectrum is 0.35 or less. Since the crystallization ratio defined by the scattering spectrum is 60% or less, an amorphous Si film in which the short-range ordered structure of the film is significantly improved can be obtained, and the efficiency of the photoelectric conversion device using the same can be improved. In addition, the light deterioration rate can be reduced. In addition, the film quality can be easily controlled by a simple method and apparatus using Raman scattering spectroscopy.

Further, it contains at least silicon and hydrogen, the half-width of the scattering peak of the TO mode obtained by the Raman scattering spectrum is 65 cm ⁻¹ or less, and the crystallization rate defined by the Raman scattering spectrum is 60% or less. As a result, an amorphous Si film in which the short-range ordered structure of the film is significantly improved can be obtained, and the efficiency of a photoelectric conversion device using the film can be increased and the rate of photodegradation can be reduced. In addition, the film quality can be easily controlled by a simple method and apparatus using Raman scattering spectroscopy.

Further, by the ratio of the density of Si-H bonding state to the sum of the density of the density and Si-H ₂ bonds state of Si-H bonding state of the semiconductor thin film is 0.95 or more, Si-H ₂ bonds Light degradation caused by the above can be greatly reduced, and the efficiency of a photoelectric conversion device using the same can be increased, and the light degradation rate can be reduced.

  Further, according to the photoelectric conversion device of the present invention, it is possible to provide a high-efficiency and excellent photoelectric conversion device such as a solar cell with a reduced light degradation rate.

  Further, according to the photovoltaic device of the present invention, the photoelectric conversion device of the present invention is used as power generating means, and the power generated by the power generating means is supplied to the load, thereby providing a high-efficiency photovoltaic device. can do.

FIG. 3 is a diagram showing a Raman scattering spectrum of an amorphous silicon film. FIG. 3 is a cross-sectional view schematically illustrating an example in which the semiconductor thin film of the present invention is applied to a photoelectric conversion device.

Explanation of reference numerals

1: substrate 2: first electrode 3: semiconductor multilayer film
31: first semiconductor bonding layer
31a: p-type semiconductor layer
31b: Photoactive layer
31c: n-type semiconductor layer
32: second semiconductor bonding layer
32a: p-type semiconductor layer
32b: Photoactive layer
32c: n-type semiconductor layer 4: second electrode

Claims (5)

It contains at least silicon and hydrogen, the ratio of the scattering peak intensity of the TA mode to the scattering peak intensity of the TO mode obtained by the Raman scattering spectrum is 0.35 or less, and the crystallization ratio defined by the Raman scattering spectrum is A semiconductor thin film characterized by being 60% or less. It contains at least silicon and hydrogen, the half-width of the scattering peak of the TO mode obtained by the Raman scattering spectrum is 65 cm ⁻¹ or less, and the crystallization rate defined by the Raman scattering spectrum is 60% or less. Characteristic semiconductor thin film. The ratio of the existing density of the Si—H bonding state to the sum of the existing density of the Si—H bonding state and the existing density of the Si—H ₂ bonding state in the semiconductor thin film is 0.95 or more. The semiconductor thin film according to claim 1 or 2. A photoelectric conversion device comprising: a semiconductor junction unit including the semiconductor thin film according to claim 1 in a photoactive portion; and an electrode electrically connected to the semiconductor junction unit. . 5. A photovoltaic device, wherein the photoelectric conversion device according to claim 4 is used as power generation means, and the power generated by the power generation means is supplied to a load.

Images