JP5173370B2 - Method for manufacturing photoelectric conversion element - Google Patents

Description

The present invention relates to a method for producing a photoelectric conversion element.

In crystalline silicon solar cells, silicon nitride formed by plasma vapor chemical vapor deposition (plasma CVD method) is known as a useful material.
The reason is that the silicon nitride film plays three important roles of functioning as an antireflection film of the crystalline silicon solar cell and simultaneously functioning as a passivation film on the surface of the silicon substrate and inside.

In the function as the first antireflection film, an important parameter for taking sunlight into the silicon substrate most efficiently is the refractive index of the silicon nitride film.
Generally, the optimum refractive index when using a silicon nitride film as an antireflection film for a crystalline silicon solar cell is about 2.0. When a silicon nitride film is formed using a plasma CVD method, the refractive index can be controlled by changing the flow ratio of the mixed gas, the film forming pressure, the film forming power, and the like.

In the function as a passivation film on the surface of the second silicon substrate, important parameters for improving the function are the interface state and the fixed charge at the interface between the silicon nitride film and the silicon substrate.
That is, as described in Non-Patent Document 1, when the interface state density is low, the probability that carriers (electrons and holes) generated inside the silicon substrate are recombined at the interface is low. Increases the passivation effect.
Further, when a fixed charge is present at the interface, an energy band bend (a significant difference in carrier density) occurs in the vicinity of the silicon substrate surface, carrier recombination hardly occurs, and the passivation effect on the silicon substrate surface is enhanced.
As described above, the interface state and the fixed charge influence the passivation effect on the surface of the silicon substrate in combination by independent mechanisms.

The function as a passivation film inside the third silicon substrate is that silicon contained in a large amount in the silicon nitride film formed by the plasma CVD method is released by baking after film formation and enters the silicon substrate. This is an effect to passivate defects inside the substrate.
An important parameter for improving this function is the amount of hydrogen at the interface between the silicon nitride film and the silicon substrate and the amount of defects in the silicon substrate itself.
That is, as described in Non-Patent Document 2, when a silicon nitride film is formed by a plasma CVD method, a catalytic CVD method, a sputtering method, or the like, damage occurs on the surface of the silicon substrate, and at the same time, hydrogen accumulates in this damage. The The accumulated hydrogen is diffused inside the silicon substrate by baking after film formation, and the defects existing inside the silicon substrate are passivated. Further, Non-Patent Document 2 reports that damage formed at the interface is also rendered harmless to some extent at the same time.

As described above, the silicon nitride film plays various roles in the crystalline silicon solar cell. For this reason, in order to further improve the conversion efficiency of a crystalline silicon solar cell, there are many research report examples which devised the manufacturing method of a silicon nitride film.
However, as described above, since the function of the silicon nitride film is controlled by a plurality of parameters, it is difficult to improve these functions simultaneously.

For example, Non-Patent Document 3 reports that the effect of surface passivation is improved when the refractive index of a silicon nitride film, which is usually about 2.0, is further increased.
However, since the optimum value of the refractive index of the silicon nitride film is determined in consideration of the refractive index of the crystalline silicon substrate and air, the high refractive index reduces the function of the silicon nitride film as an antireflection film. The conversion efficiency of the crystalline silicon solar cell will be reduced.

Further, in Non-Patent Documents 4 to 6, treatment with ammonia plasma before forming a silicon nitride film on a crystalline silicon substrate improves the effective lifetime of the silicon substrate and improves the characteristics of the photoelectric conversion element. It has been reported. These improvements are reported to be the effect of a function as a passivation film on the surface of the silicon substrate, a function as a passivation film inside the silicon substrate, or a combination of these functions. These methods are excellent in that the function as a passivation film inside or on the silicon substrate can be enhanced without changing the refractive index of the silicon nitride film.
However, the technologies of Non-Patent Documents 4 to 6 focus only on one of the function as a passivation film on the surface of the silicon substrate and the function as a passivation film inside the silicon substrate, or properly separate the two effects. Not done. Therefore, it has not succeeded in controlling a parameter for improving the function of the silicon nitride film and obtaining a photoelectric conversion element having higher characteristics.

S.W.Glunz and 14 others, 20th European photovoltaic Solar Energy Conference and Exhibition (2005), p.572 “Comparison of different dielectric passivation layers for application in industrially feasible high-efficiency crystalline silicon solar cells” Bhushan Sopori and 8 others, 4th World Conference on Photovoltaic Energy Conversion (2006), p.1028 “Damage-Layer-Mediated H Diffusion During SiN: H Processing: A Comprehensive Model”

T. Lauinger and two others, 14th European Photovoltaic Solar Energy Conference (1997), p.853 “Comparison of direct and remote PECVD silicon nitride films for low-temperature surface passivation of p-type crystalline silicon” Jin Nigo and 5 others, 21st European Photovoltaic Solar Energy Conference (2006), p.934 “Effective Interface Modification by NH3 Plasma for Silicon Surface Passivation at Very Low Temperatures”

Alexander Hauser and 3 others, Solar Energy Materials & Solar Cells 75 (2003), p.357-362 “Influence of anammonia activation prior to the PECVD SiN deposition on the solar cell performance” A.Rohatgi and 6 others, 14th International Photovoltaic Science and Engineering Conference (2004), p.635 “Record-High-EfficiencySolarCellsonMulticrystalline Materials Through Understanding and Implementation of RTP-enhanced SiNx-induced Defect Hydrogenation”

  An object of the present invention is to provide a manufacturing method for obtaining a photoelectric conversion element having higher characteristics.

  As a result of intensive studies to solve the above-described problems, the present inventor has determined that the surface of the crystalline silicon substrate on which the silicon nitride film is to be formed is formed with hydrogen gas before the silicon nitride film is formed on the surface of the crystalline silicon substrate. It has been experimentally found that the characteristics of the photoelectric conversion element can be improved by surface treatment with a plasma formed using a mixed gas containing ammonia gas and the present invention has been completed.

Thus, according to the present invention,
(A) forming an n-type semiconductor layer on the surface of the p-type crystalline silicon substrate;
(B) surface-treating the surface of the n-type semiconductor layer with plasma formed using a mixed gas containing hydrogen gas and ammonia gas;
(C) forming a silicon nitride film on the n-type semiconductor layer;
(D) including a step of forming a back electrode on the back surface of the p-type crystalline silicon substrate, and (e) a step of forming a light-receiving surface electrode on the silicon nitride film. A manufacturing method is provided .

  ADVANTAGE OF THE INVENTION According to this invention, the manufacturing method of the photoelectric conversion element of the outstanding characteristic can be provided.

The method for producing the photoelectric conversion element of the present invention is as follows.
(A) forming an n-type semiconductor layer on the surface of the p-type crystalline silicon substrate;
(B) surface-treating the surface of the n-type semiconductor layer with plasma formed using a mixed gas containing hydrogen gas and ammonia gas;
(C) forming a silicon nitride film on the n-type semiconductor layer;
(D) forming a back electrode on the back surface of the p-type crystal silicon substrate; and (e) forming a light receiving surface electrode on the silicon nitride film.
That is, the method for manufacturing a photoelectric conversion element of the present invention includes a step of forming a silicon nitride film on the main surface (surface) of a p-type crystal silicon substrate, and n on the main surface side of the p-type crystal silicon substrate. And a surface treatment of the main surface is performed by plasma formed using a mixed gas containing hydrogen gas and ammonia gas before forming the silicon nitride film. .

Although the manufacturing method of the photoelectric conversion element of this invention is demonstrated using drawing, this invention is not limited by this description.
FIG. 1A to FIG. 1F are schematic cross-sectional views showing manufacturing steps of one embodiment in the method for manufacturing a photoelectric conversion element of the present invention. Hereinafter, each step will be described in detail.

1. First, as shown in FIG. 1A, an n-type semiconductor layer 2 is formed on the surface of a p-type crystalline silicon substrate 1 to form a pn junction.

The p-type crystalline silicon substrate is not particularly limited as long as it is a crystalline silicon substrate. For example, it may be a single crystal silicon substrate or a polycrystalline silicon substrate. The resistance value and crystal orientation of the silicon substrate are not particularly limited.
Prior to the formation of the n-type semiconductor layer, the p-type crystal silicon substrate is preferably treated with a strong alkaline aqueous solution, a strong acid aqueous solution or the like for the purpose of cleaning the surface of the p-type crystalline silicon substrate or removing the surface damage layer.
In addition, before the n-type semiconductor layer is formed, incident light is subjected to multiple reflection on the light receiving surface side of the photoelectric conversion element, thereby forming a fine unevenness (texture) structure for capturing light more effectively. For the purpose, it is preferable to treat the p-type crystalline silicon substrate with an alkaline solution.

The n-type semiconductor layer 2 can be formed by a known method, for example, by doping an n-type impurity into a p-type crystal silicon substrate or by separately forming an n-type on the p-type crystal silicon substrate by a CVD method or the like. it can.
Examples of n-type impurities include Group 5 elements such as phosphorus and arsenic.
The sheet resistance of the n-type semiconductor layer is preferably in the range of 20 to 200Ω / □.

The n-type impurity doping method is not particularly limited. For example, a solution containing a Group 5 compound such as phosphorus oxychloride (POCl ₃ ) is gasified in a high-temperature furnace at 700 to 1000 ° C. to form a p-type crystal. A method of diffusing into a silicon substrate (vapor phase diffusion method), a solution containing a compound of a group 5 element such as phosphorus pentoxide (P ₂ O ₅ ) (for example, an isopropyl alcohol solution of phosphorus pentoxide) is a p-type crystalline silicon substrate For example, a method of dropping the solution onto the surface and applying it uniformly with a spin coater, then putting it in a high-temperature furnace at a temperature of 700 to 1000 ° C., and diffusing the group 5 element adhering to the surface into the p-type crystalline silicon substrate, etc.
The latter method can be selectively formed only on the light-receiving surface of the n-type semiconductor layer, and an n-type semiconductor layer formed on a portion other than the light-receiving surface side (back surface and side surface) as in the vapor phase diffusion method. Since it is not necessary to include an additional step for removal, it is more suitable for mass production.

2. Plasma Surface Treatment Step Next, as shown in FIG. 1B, the surface of the n-type semiconductor layer 2 is surface-treated with plasma 3 formed using a mixed gas containing hydrogen gas and ammonia gas.
As the material and the forming method, known materials and methods can be applied, but plasma is formed using a plasma CVD apparatus, and the forming condition is an ammonia gas flow rate of 20 to 1000 sccm, preferably 50 to 500 sccm, hydrogen gas against ammonia gas. The flow rate ratio is 1 to 10 times, preferably 3 to 4 times, the forming pressure is 10 to 200 pa, preferably 50 to 150 Pa, the forming temperature is 200 to 600 ° C., preferably 400 to 500 ° C., and the forming time is 10 to 1200 seconds. Is preferred. In addition, the source power is generally about 10 to 1000 W at a frequency of 13.56 MHz), and generally no bias power is added. Under such formation conditions, excellent characteristics can be imparted to the silicon nitride film formed in the next step.
Here, the unit sccm means standard cc / min (standard: 1 atm, 0 ° C.).

3. Antireflection Film Formation Step Next, as shown in FIG. 1C, a silicon nitride film 4 is formed on the n-type semiconductor layer in order to effectively capture light such as sunlight. This silicon nitride film functions as a surface of the crystalline silicon substrate and as an internal passivation film.
As the material and the forming method, known materials and methods can be applied, but it is preferable that the silicon nitride film is formed by the plasma CVD method because it can be treated continuously from the above steps. The conditions for the plasma CVD method may be set as appropriate depending on the shape of the reaction chamber. For example, in the case of a mixed gas of monosilane 10 to 500 sccm, ammonia 10 to 1000 sccm, and nitrogen 50 to 1000 sccm, pressure 10 to 200 Pa, temperature 200 to It is desirable to be in the range of 600 ° C. In addition, the source power is generally about 10 to 1000 W at a frequency of 13.56 MHz), and generally no bias power is added. Under such formation conditions, excellent characteristics can be imparted to the silicon nitride film.
The thickness of the silicon nitride film may be appropriately set according to the refractive index of the film and the size of the surface irregularities of the p-type crystalline silicon substrate, but is usually about 60 to 100 nm.

4). Back Electrode Formation Step Next, as shown in FIG. 1D, a back electrode 5 is formed on the back surface of the p-type crystalline silicon substrate 1. The back electrode 5 is used for taking out the carriers generated in the p-type crystalline silicon substrate 1 as a current.
Known materials and methods can be applied as the material and formation method. For example, the back electrode 5 is obtained by applying (printing) a conductive paste containing aluminum powder or the like to the entire back surface of the solar cell and drying at a temperature of 100 to 400 ° C. by screen printing. The screen printing method is preferable because the cost can be reduced at the mass production level.
The thickness of the back electrode is usually about 10 to 60 μm.

5. Next, as shown in FIG. 1E, a light receiving surface electrode 6 is formed on the silicon nitride film 4.
Known materials and methods can be applied as the material and formation method. For example, a conductive paste mainly composed of silver powder, glass powder, organic vehicle and organic solvent is applied (printed) on the silicon nitride film 4 by screen printing and dried at a temperature of 100 to 400 ° C. The light receiving surface electrode 6 is obtained.
The pattern of the light-receiving surface electrode 6 is not particularly limited, and is not particularly limited as long as it is a pattern generally used for solar cells. For example, a fishbone type (comb shape) is mentioned.
The thickness of the light-receiving surface electrode is usually about 10 to 60 μm.

6). Light-receiving surface electrode fire-through and back surface electric field layer forming step After the above step, as shown in FIG. 1 (f), the obtained p-type crystalline silicon substrate 1 is subjected to a firing treatment, and the light-receiving surface electrode fire-through (It is not shown) It is preferable to form the back surface electric field layer 7 simultaneously on the interface between the p-type crystalline silicon substrate and the back electrode.
The fire-through is a phenomenon that occurs when the silicon nitride film 4 is broken by the action of the glass powder added to the light-receiving surface electrode 6 when the p-type crystalline silicon substrate 1 is baked. Thereby, the light-receiving surface electrode 6 can be brought into contact with the p-type crystalline silicon substrate 1.
The back surface electric field layer 7 is a region in which a part of aluminum contained in the back electrode 5 is diffused into the p-type crystal silicon substrate 1 when the substrate 1 is baked. It is possible to improve the collection efficiency of carriers generated in
The firing conditions are preferably a temperature range of 600 to 900 ° C. and a firing time of about 1 to 300 seconds.

  The present invention will be described more specifically with reference to examples and comparative examples, but the present invention is not limited to the examples.

Example 1 (Evaluation Experiment 1)
In this effect demonstration experiment, interface state density and flat band voltage were evaluated by capacitance-voltage measurement (CV measurement), and the influence of plasma surface treatment on the passivation effect of the crystalline silicon substrate surface was investigated.
That is, before forming the silicon nitride film on the crystalline silicon substrate, the sample (Examples 1-1 and 1-2) in which surface treatment was performed with plasma formed using a mixed gas containing hydrogen gas and ammonia gas. ), For samples (Comparative Examples 1-2 and 1-3) subjected to surface treatment with plasma formed using ammonia gas or hydrogen gas, and for samples not subjected to surface treatment (Comparative Example 1-1) Level density and flat band voltage were measured.
Note that, in this demonstration experiment, a sample having a structure in which capacitance-voltage measurement (CV measurement) can be easily performed instead of an actual photoelectric conversion element structure was manufactured.
That is, a sample was manufactured by the manufacturing process shown in FIGS.

1-1. Step of Preparing Crystalline Silicon Substrate First, as shown in FIG. 2 (a), a single crystal p-type crystal silicon substrate 1 (plane orientation <111>, thickness 520 μm, resistivity 2.5 Ωcm) having one surface mirror-polished. It was immersed in an HF aqueous solution and then washed with pure water to remove the natural oxide film on the surface of the p-type crystalline silicon substrate 1.

1-2. Plasma surface treatment step on the surface of the crystalline silicon substrate Next, the p-type crystalline silicon substrate 1 is carried into a vacuum chamber of a plasma CVD apparatus, and as shown in FIG. The mirror-polished surface of the substrate 1 was surface treated.
The flow rate of the mixed gas for forming the plasma 3 in each sample is as shown in Table 1. Comparative Example 1-1 was not surface-treated with plasma 3.
Other plasma surface treatment conditions were source power (frequency 13.56 MHz, 500 W), treatment pressure 100 Pa, treatment temperature 450 ° C., treatment time 750 seconds. Bias power is not added.

1-3. Next, as shown in FIG. 2C, a silicon nitride film 4 having a thickness of about 90 nm was formed on the p-type crystalline silicon substrate 1 in a vacuum chamber of a plasma CVD apparatus.
The mixed gas flow rate ratio during the formation of the silicon nitride film 4 was monosilane: ammonia: nitrogen = 1: 2: 12. The other film formation conditions are source power (frequency 13.56 MHz,
250 W), a film forming pressure of 100 Pa, a film forming temperature of 450 ° C., and a film forming time of 225 seconds. Bias power is not added.
The refractive index of the silicon nitride film 4 at a wavelength of 630 nm, determined from spectroscopic ellipsometry, was 1.98.
Thereafter, the obtained p-type crystalline silicon substrate 1 was baked at 800 ° C. for 90 seconds using a near infrared furnace. Capacitance-voltage measurement is possible without firing, but the silicon nitride film 4 used in an actual solar cell includes a firing step, so firing was also performed in this demonstration experiment.

1-4. Next, as shown in FIG. 2D, 1 μm thick aluminum is deposited in a circular dot shape with a diameter of 1 mm on the silicon nitride film 4 using a vacuum evaporation apparatus and a predetermined shadow mask. Thus, the gate electrode 8 was obtained. At this time, the p-type crystalline silicon substrate 1 was not particularly heated and temperature control was not performed.

1-5. Back Electrode Formation Step Next, as shown in FIG. 2 (e), a 1 μm-thickness is formed over the entire surface of the p-type crystalline silicon substrate 1 on the opposite side to the gate electrode 8 by using a vacuum deposition apparatus in the same manner as the gate electrode 8. Aluminum was deposited to obtain a back electrode 9. At this time, the p-type crystalline silicon substrate 1 was not heated and the temperature was not controlled.

1-6. Capacitance-Voltage Measurement Next, capacitance-voltage measurement was performed on each of the produced samples.
A high frequency (0.1 MHz) voltage and a low frequency (5 to 6 Hz) voltage were respectively applied to the gate electrode 8, and the capacitance between the back electrode 9 and the gate electrode 8 at the time of voltage application was measured. The measurement start voltage was −15 V, and the measurement end voltage was +15 V. A mercury probe was brought into contact with the gate electrode 8, and the entire surface of the back electrode 9 was vacuum-adsorbed to the device-side electrode.
From the results of the obtained capacitance-voltage measurement, the interface state density and the flat band voltage were determined by the following method. The obtained results are shown in Table 1.

1-7. Interface state density The interface state density is a high frequency and low frequency described in EHNicollian and JRBrews, "MOS (Metal Oxide Semiconductor) Physics and Technology", 1982, published by WILEY-INTERSCIENCE, p.331. It was determined by the hi-lo method combining the capacitance-voltage characteristics.

Specifically, the interface state density _Dit was determined based on the following equation.
D _it = (1 / q ₀ ) × [[(1 / C _LF ) − (1 / C _SIN )] ⁻¹ − [(1 / C _HF ) − (1 / C _SIN )] ⁻¹ ]
q ₀ : Elementary quantity of electricity C _SIN : Capacity of silicon nitride film 9 C _LF : Capacity measured at low frequency C _HF : Capacity measured at high frequency From this equation, the range corresponding to the energy gap of silicon, that is, valence electrons The continuous D _it can be obtained in the range from the band to the conduction band, and the minimum value of the continuous D _it is shown in Table 1.
Note that the lower the interface state density, the lower the probability that carriers are trapped on the surface, which means that the passivation effect on the surface of the crystalline silicon substrate is higher.

1-8. Flat Band Voltage The flat band voltage was determined by the method described in EHNicollian and JRBrews, “MOS (Metal Oxide Semiconductor) Physics and Technology”, 1982, published by WILEY-INTERSCIENCE, p.487.

Specifically, the flat band capacitance C _FB was obtained based on the following equation.
C _FB = C _FBS C _SIN / (C _SIN + C _FBS )
C _SIN : capacitance of the silicon nitride film 4 (for convenience, the capacitance measured when a voltage of −15 V is applied to the gate electrode 8 is used. Although the measured value of C _SIN is slightly different for each sample, It was about 500 to 600 [pF].)
C _FBS : Flat band capacity of semiconductor (p-type crystalline silicon substrate 1) (determined by the following formula)
C _FBS : ε _s / λp
ε _s : Dielectric constant of semiconductor (product of dielectric constant ε _{0 in} vacuum and relative dielectric constant ε _Si of semiconductor)
λp: Debye length, which is expressed by the following equation.
λ _p = {(kTε ₀ ε _Si ) / (q ₀ ² N _d )} ^1/2
k: Boltzmann constant T: Temperature of semiconductor (T is slightly different for each sample, but is about 25 ° C.)
q ₀ : Elementary quantity of electricity N _d : Semiconductor carrier concentration (N _d is slightly different for each sample, but is about 5 × 10 ¹⁵ [cm ⁻³ ])
The gate voltage when the C _FB thus obtained matches the measured value of the capacitance when a high frequency voltage is applied in the actual capacitance-voltage curve is obtained, and the value is defined as the flat band voltage V _FB . .

The flat band voltage and the fixed charge density can be related by the following equation. The fixed charge density means the density of charges fixed inside the silicon nitride film 4 or in the vicinity of the interface between the silicon nitride film 4 and the silicon substrate 1.
Q = C _SiN (Φ _MS −V _FB ) / q ₀ A
Q: fixed charge density C _SIN : capacitance of the silicon nitride film 4 Φ _MS : work function difference between the material of the gate electrode 8 and the material of the silicon substrate 1 V _FB : flat band voltage q ₀ is the elementary charge A is the gate electrode 8 Area of

  According to the above equation, the flat band voltage increases in the negative direction as the fixed charge density increases in the positive direction. Therefore, in Table 1, a sample in which the flat band voltage has a large negative value has a positive value in which the fixed charge density is large.

According to the results of Table 1, it can be seen that Examples 1-1 and 1-2 are superior to Comparative Examples 1-1 to 1-3. Therefore, when a silicon nitride film having a large flat band voltage and a negative value is formed on the n-type semiconductor layer 2, a large energy band is bent on the surface of the n-type semiconductor layer, and a storage layer is formed. The hole concentration on the surface of the n-type semiconductor is remarkably reduced, the recombination is substantially suppressed, and the surface passivation effect is enhanced.
Although the flat band voltage of the silicon nitride film of Comparative Example 1-1 is large, it can be seen that the interface state density is also large and the passivation effect is low. That is, since the passivation effect also depends on the interface state density and the amount of surface hydrogen, even if the flat band voltage is a large negative value, the surface passivation effect is not necessarily high.

  In consideration of the prior art described in Non-Patent Documents 4 to 6 in which ammonia plasma treatment is effective for improving the surface passivation effect, Comparative Example 1-2 of ammonia plasma treatment is Comparative Example 1-1 without treatment. The plasma processing Examples 1-1 and 1-2 formed using a mixed gas containing hydrogen gas and ammonia gas are more advantageous than Comparative Example 1-2. Can understand.

Example 2 (Effect Demonstration Experiment 2)
In this demonstration experiment, the amount of hydrogen desorption was measured by thermal desorption spectroscopy (TDS), and the influence of plasma surface treatment on the passivation effect inside the crystalline silicon substrate was investigated.

That is, before the silicon nitride film was formed on the crystalline silicon substrate, the samples (Examples 2-1 and 2-2) in which surface treatment was performed with plasma formed using a mixed gas containing hydrogen gas and ammonia gas. ), Samples heated by plasma formed using ammonia gas or hydrogen gas (Comparative Examples 2-2 and 2-3) and samples not subjected to surface treatment (Comparative Example 2-1) were heated. The amount of hydrogen that was occasionally desorbed was measured.
In this demonstration experiment, a sample having a structure that facilitates measurement of the amount of hydrogen desorption from the surface (temperature-programmed desorption analysis), not the actual structure of the photoelectric conversion element, was prepared.
That is, a sample was manufactured and evaluated by the manufacturing process shown in FIGS.

2-1. Step of Preparing Crystalline Silicon Substrate First, as shown in FIG. 2A, a single-crystal p-type crystal silicon substrate 1 (plane orientation <111>, thickness 400 μm, resistivity 1000 to 4000 Ωcm) whose one side is mirror-polished is prepared. After cutting to a size of 1 cm × 1 cm, it was immersed in an HF aqueous solution and then washed with pure water to remove the natural oxide film on the surface of the p-type crystalline silicon substrate 1.

2-2. Plasma surface treatment step on the surface of the crystalline silicon substrate Next, the p-type crystalline silicon substrate 1 is carried into a vacuum chamber of a plasma CVD apparatus, and as shown in FIG. Each sample was obtained by subjecting the mirror-polished surface of the substrate 1 to surface treatment.
The flow rate of the mixed gas and other plasma surface treatment conditions for forming the plasma 3 in each sample were the same as in Example 1.

2-3. Temperature-programmed desorption analysis Temperature-programmed desorption analysis is a technique in which a sample installed in a vacuum chamber is heated with a heater, and components desorbed from the sample at that time are detected by a mass spectrometer.
First, the sample was carried into the vacuum chamber of the temperature programmed desorption analyzer and kept at room temperature until the degree of vacuum was 2.5 × 10 ⁻⁷ Pa or less.
The sample in the vacuum chamber was then heated to a temperature of 200 ° C. and held for 15 minutes.
Thereafter, the sample is heated to 1000 ° C. at a heating rate of 1 ° C./second, and hydrogen desorbed from the sample using a quadrupole mass spectrometer (Pfeiffer Vacuum, model: QMG422) The thermal desorption spectrum) was measured over time.

Next, the sample was naturally cooled to room temperature in the vacuum chamber, the sample was heated with the same thermal history as above, and the temperature programmed desorption spectrum was measured in the same manner as described above.
The obtained temperature-programmed desorption spectrum was used as the background, and the background was subtracted from the temperature-programmed desorption spectrum measured previously to obtain the final temperature-programmed desorption spectrum. The obtained results are shown in FIG. However, FIG. 3 shows a sample in which surface treatment was performed with plasma formed using a mixed gas containing hydrogen gas and ammonia gas before forming a silicon nitride film on a crystalline silicon substrate (Example 2- Only the results of 1) and a sample (Comparative Example 2-2) subjected to surface treatment with plasma formed using ammonia gas are shown.

According to the results of Table 1 and FIG.
(1) The sample (Comparative Example 1-2) subjected to the surface treatment with ammonia gas plasma has a significantly reduced interface state density compared to the sample (Comparative Example 1-1) which was not subjected to the surface treatment. It can be seen that the surface passivation effect is improved.
However, the absolute value of the flat band voltage is smaller in the former than in the latter, and the size of the storage layer on the surface of the n-type semiconductor layer 2 is reduced accordingly, so the passivation effect due to the bending of the energy band is slightly weakened. It is thought that.
Non-Patent Documents 4 to 6 describe that the surface treatment with ammonia gas plasma improves the characteristics of the photoelectric conversion element. This improvement is considered to be mainly caused by the reduction of the interface state density. .

(2) Samples (Examples 1-1 and 1-2) that were surface-treated with a mixed gas plasma containing hydrogen gas and ammonia gas (Examples 1-1 and 1-2) were samples that were surface-treated with ammonia gas plasma (Comparative Example 1). -2), the interface state density is further reduced, and the absolute value of the flat band voltage is increased.
This means that the former is more excellent in surface passivation effect than the latter.

Further, from the results of FIG. 3, the sample (Example 2-1) subjected to surface treatment with plasma of a mixed gas containing hydrogen gas and ammonia gas is a sample (Comparative Example) subjected to surface treatment with plasma of ammonia gas. Compared with 1-2), it can be seen that the amount of desorbed hydrogen is large in the region of 600 to 1000 ° C., that is, the amount of accumulated hydrogen is large.
This means that the former has a better passivation effect on the inside of the crystalline silicon substrate than the latter.

(3) The sample subjected to surface treatment with hydrogen gas plasma (Comparative Example 1-3) has a significantly increased interface state density compared to the sample not subjected to surface treatment (Comparative Example 1-1). It can be seen that the surface passivation effect is deteriorated.
This shows that it is essential to use a mixed gas of both hydrogen and ammonia in order to improve the passivation effect by plasma treatment.

  In summary, the samples (Examples 1-1 and 1-2) that were surface-treated with plasma of a mixed gas containing hydrogen gas and ammonia gas were samples that were surface-treated with plasma of ammonia gas. (Comparative Example 1-2), compared to a sample subjected to surface treatment with plasma of hydrogen gas (Comparative Example 1-3) and a sample not subjected to surface treatment (Comparative Example 1-1), from the viewpoint of surface passivation Is also excellent from the viewpoint of the passivation effect inside the silicon substrate

The reason why such a result is obtained is not necessarily clear, but considering that the silicon nitride film 4 is formed under the same conditions in all samples, the mixed gas containing hydrogen gas and ammonia gas is used. It is considered that the surface treatment by the plasma 3 changes the state in the vicinity of the interface between the silicon nitride film 4 and the silicon substrate 1, resulting in termination of unbonded silicon by hydrogen or a change in the fixed charge amount.
In this embodiment, the interface state density and the flat band voltage are measured on the p-type crystalline silicon substrate. The interface state density is determined by the fact that the silicon nitride film is not bonded to the silicon of the silicon substrate. Since this effect is considered to be almost independent of a small amount of impurity atoms (dopants), the relative comparison of each sample is This is considered to apply to an n-type crystalline silicon substrate.
Further, since the magnitude of the flat band voltage changes depending on the fixed charge of the silicon nitride film, it is also applicable to the n-type crystalline silicon substrate in making a relative comparison for each sample. Conceivable.

Example 3
In this example, it was examined that the effects demonstrated in Examples 1 and 2 contribute to the improvement of the characteristics of the actual photoelectric conversion element.
In this example, sheet silicon (a p-type polycrystalline silicon substrate 1 having an area of 15 cm × 15 cm and a thickness of 350 μm) produced by the method described in Japanese Patent No. 2001-223172 is used, and FIG. The photoelectric conversion element was produced by the manufacturing process shown by 1 (a)-(f).

3-1. Step of Preparing Crystalline Silicon Substrate First, the p-type polycrystalline silicon substrate 1 is treated with a HNO ₃ / HF mixed solution and etched to about 10 μm on both sides to simultaneously clean the surface and remove the damaged layer. (Not shown). Next, anisotropic etching with an aqueous NaOH solution was performed to form a fine concavo-convex structure called texture on the surface of the p-type polycrystalline silicon substrate 1 (not shown). Here, the texture is a structure for taking in light more effectively by multiple reflection of incident light on the light receiving surface side of the photoelectric conversion element.

3-2. Step of forming n-type semiconductor layer Next, an isopropyl alcohol solution of phosphorus pentoxide (concentration 15 g / L) was dropped onto the p-type polycrystalline silicon substrate 1 and uniformly applied by a spin coater. Thereafter, the p-type polycrystalline silicon substrate 1 was placed in a high-temperature furnace at 900 ° C. for 15 minutes, and as shown in FIG. 1A, an n-type semiconductor layer 2 was formed on the light receiving surface side of the photoelectric conversion element.

3-3. Plasma surface treatment step on the surface of the crystalline silicon substrate Next, the p-type crystalline silicon substrate 1 is carried into a vacuum chamber of a plasma CVD apparatus, and as shown in FIG. The surface of the n-type semiconductor layer of the substrate 1 was surface treated.
The flow rate of the mixed gas for forming the plasma 3 in each sample is as shown in Table 2. Comparative Example 3-1 was not surface-treated with plasma 3.
Other plasma surface treatment conditions were source power (frequency 13.56 MHz, 500 W), treatment pressure 100 Pa, treatment temperature 450 ° C., treatment time 750 seconds. Bias power is not added.

3-4. Next, as shown in FIG. 1C, a silicon nitride film 4 having a film thickness of about 80 nm was formed on the p-type crystalline silicon substrate 1 in a vacuum chamber of a plasma CVD apparatus.
The mixed gas flow rate ratio during the formation of the silicon nitride film 4 was monosilane: ammonia: nitrogen = 1: 2: 12. The other film formation conditions are source power (frequency 13.56 MHz,
250 W), a film forming pressure of 100 Pa, a film forming temperature of 450 ° C., and a film forming time of 225 seconds. Bias power is not added.
The refractive index of the silicon nitride film 4 at a wavelength of 630 nm, determined from spectroscopic ellipsometry, was 1.98.

3-5. Back Electrode Formation Step Next, as shown in FIG. 1 (d), an aluminum paste is printed on the back surface of the p-type crystalline silicon substrate 1 by a screen printing method and sufficiently dried at a temperature of 150 ° C. An aluminum back electrode 5 was obtained.

3-6. Next, as shown in FIG. 1E, silver powder, glass powder, organic vehicle, and organic solvent as main components are formed on the silicon nitride film 4 of the p-type crystalline silicon substrate 1 by screen printing as shown in FIG. The aluminum paste to be printed was printed in a fishbone pattern and sufficiently dried at a temperature of 150 ° C. to obtain an aluminum light-receiving surface electrode 6 having a thickness of 30 μm. In addition, the fish-bone type pattern used was formed of bus bar electrodes (main grid) and finger electrodes (sub grid), and the finger electrodes were arranged perpendicular to the two bus bar electrodes.

3-7. Next, the p-type crystalline silicon substrate 1 is baked at a temperature of 850 ° C. for 120 seconds using a near-infrared furnace, and as shown in FIG. A back surface electric field layer (not shown) was formed at the interface between the fire-through (not shown) and the p-type crystalline silicon substrate 1 and the back electrode 5 to obtain a photoelectric conversion element (solar cell).

3-8. Evaluation of photoelectric conversion element Using a solar simulator, the obtained photoelectric conversion element was irradiated with pseudo light of AM 1.5 and 100 mW / cm ² (temperature condition: 25 ° C.), short-circuit current density (mA / cm ² ), The open circuit voltage (V), curvature factor and conversion efficiency (%) were measured. The obtained results are shown in Table 2.

  From the results of Table 2, it can be seen that Examples 3-1 and 3-2 have characteristics of photoelectric conversion elements that exceed those of Comparative Examples 3-1 to 3-3 according to the prior art. As can be seen from the results of Examples 1 and 2, considering that the silicon nitride films are formed under the same conditions in all the samples, the plasma treatment according to the present invention is performed with the silicon nitride film and the silicon substrate. It is considered that the characteristics of the photoelectric conversion element were improved by changing the state in the vicinity of the interface.

  From the results of the above embodiments, before forming the silicon nitride film on the surface of the crystalline silicon substrate, the surface of the crystalline silicon substrate on which the silicon nitride film is formed is plasma of a mixed gas containing hydrogen gas and ammonia gas. It can be seen that the characteristics of the photoelectric conversion element can be improved by performing the surface treatment in advance.

It is a schematic sectional drawing which shows the manufacturing process of one Embodiment in the manufacturing method of the photoelectric conversion element of this invention. It is a schematic sectional drawing which shows the manufacturing process of the sample in the effect verification experiment (capacitance-voltage measurement) of this invention. It is a graph which shows the result of the desorption amount of hydrogen obtained by the effect verification experiment (temperature-programmed desorption analysis) of this invention.

Explanation of symbols

1 p-type crystalline silicon substrate 2 n-type semiconductor layer 3 plasma 4 silicon nitride film 5, 9 back electrode 6 light-receiving surface electrode 7 back surface field layer 8 gate electrode

Claims (4)

(A) forming an n-type semiconductor layer on the surface of the p-type crystalline silicon substrate;
(B) surface-treating the surface of the n-type semiconductor layer with plasma formed using a mixed gas containing hydrogen gas and ammonia gas;
(C) forming a silicon nitride film on the n-type semiconductor layer;
(D) including a step of forming a back electrode on the back surface of the p-type crystalline silicon substrate, and (e) a step of forming a light-receiving surface electrode on the silicon nitride film. Production method.   The plasma in the step (b) is formed using a plasma CVD apparatus, and the formation conditions are a hydrogen gas flow rate ratio of 1 to 10 times to ammonia gas, a processing pressure of 10 to 200 pa, a processing temperature of 200 to 600 ° C., and a processing time of 10 to 10. It is 1200 second, The manufacturing method of the photoelectric conversion element of Claim 1.   The method for manufacturing a photoelectric conversion element according to claim 1, wherein the silicon nitride film in the step (c) is formed by a plasma CVD method.   After the step (e), the obtained p-type crystalline silicon substrate is subjected to a baking treatment, and a back surface electric field layer is simultaneously formed at the interface between the fire-through of the light-receiving surface electrode and the p-type crystalline silicon substrate and the back electrode. The manufacturing method of the photoelectric conversion element as described in any one of Claims 1-3 formed.

Images