JP2006269956A - Crystalline silicon substrate, its manufacturing method and photoelectric conversion element employing it - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a crystalline silicon substrate having a higher carrier diffusion length, its manufacturing method, and a photoelectric conversion element employing it. <P>SOLUTION: In the crystalline silicon substrate, the combination mode is optimized between silicon and hydrogen is optimized and a hydrogen density profile in the silicon. In a spectrum showing a relation of desorption hydrogen speed to temperature measured by temperature rise desorption gas analysis for measuring the desorption hydrogen generated upon heating the substrate, the peak of desorption hydrogen speed is within a range of the substrate temperature higher than 600°C. Further, the maximum value of peak of the desorption hydrogen speed within the range of the substrate temperature higher than 700°C is larger than the maximum value of peak of the desorption hydrogen speed, within the range of the substrate temperature higher than 600°C and not higher than 700°C. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a crystalline silicon substrate used for a photoelectric conversion element and the like, a method for producing the same, and a photoelectric conversion element using the same.

  In order to improve the quality of the crystalline silicon substrate, hydrogen has been introduced from the surface of the crystalline silicon substrate to terminate the defective portion of the crystalline silicon substrate with hydrogen, and several attempts to introduce hydrogen into the silicon substrate have been made. The method is known. For example, as a method of introducing hydrogen into a silicon substrate via the vapor phase using a vacuum process, the polycrystalline silicon substrate is directly exposed to hydrogen plasma and irradiated with hydrogen, and the defects are inactivated by atomic hydrogen. There is a method to do (see Patent Document 1). In addition, there is a method of irradiating hydrogen into a silicon crystal by intentionally accelerating hydrogen ions extracted from hydrogen plasma (see Patent Document 2 and Patent Document 3). Furthermore, in order to avoid damage to the silicon substrate due to hydrogen ions, there is a method in which the surface of the silicon substrate is covered with an oxide film in advance, and the oxide film is removed after treatment with hydrogen plasma (see Patent Document 4).

  However, in order to enhance the effect of silicon substrate modification by hydrogen treatment, it is necessary that hydrogen forms an appropriate density profile in the silicon substrate, and that hydrogen generates effective bonds with silicon. . In the above prior art, there is no mechanism for controlling the hydrogen density profile and the coupling mode between hydrogen and silicon. Therefore, the effect of defect inactivation is small. For example, in Patent Documents 2 and 3 below, hydrogen ions extracted from hydrogen plasma are intentionally accelerated and introduced into a silicon crystal. In these methods, since hydrogen termination is performed by accelerated ions having energy of several tens eV or more, the bonding state of hydrogen and silicon (binding energy is several eV) cannot be controlled.

In Patent Document 1 below, a polycrystalline silicon substrate is directly exposed to hydrogen plasma to perform defect inactivation by atomic hydrogen. In this patent, the increase in the Si-H mode is given as an interpretation of the defect inactivation phenomenon, but the data for mode identification is not specifically presented, and it is true whether the Si-H mode is increasing or not. It is unknown. As described above, in Patent Document 1 described below, it is difficult to say that the phenomenon is described using the combined mode of hydrogen and silicon, but the combined mode is actually identified and controlled based on the data. In addition, no hydrogen density profile is presented. In Patent Document 4 below, the surface is covered with an oxide film in advance and the oxide film is removed after the hydrogen plasma treatment. However, the hydrogen density profile and the hydrogen-silicon bond are not yet determined.
JP 59-136926 A JP 58-23487 A JP 58-64035 A JP-A-3-283471

  An object of the present invention is to provide a crystalline silicon substrate having a higher carrier diffusion length than that of the prior art, a method for producing the same, and a photoelectric conversion element using the crystalline silicon substrate.

  The present invention provides a crystalline silicon substrate characterized in that the bonding mode between silicon and hydrogen is optimized, and the hydrogen density profile inside the silicon is optimized.

  Preferably, in a spectrum representing a relationship between a desorbed hydrogen velocity and a temperature measured by a temperature-programmed desorbed gas analysis that measures desorbed hydrogen generated when the substrate is heated, The peak of the desorption hydrogen rate has a peak in the range of 600 ° C. or higher, and the peak value of the desorption hydrogen rate in the range of 700 ° C. or higher of the substrate temperature is in the range of 600 ° C. or higher and lower than 700 ° C. Greater than peak maximum value of desorption hydrogen rate.

  Preferably, the average hydrogen density of 0 to 300 μm in the thickness direction of the substrate is in the range of 1/5 to 1 times the local hydrogen density at a depth of 10 μm in the thickness direction.

  According to another situation of this invention, the photoelectric conversion element using said crystalline silicon substrate is provided.

  According to another aspect of the present invention, a step of forming an aluminum layer on the back surface of the crystalline silicon substrate, a step of heating and firing the crystalline silicon substrate on which the aluminum layer has been formed, and a step after the heating and firing A method for producing a crystalline silicon substrate, comprising: irradiating a surface of the crystalline silicon substrate with hydrogen.

  Preferably, the step of forming the aluminum layer on the back surface of the crystalline silicon substrate is one of a step of applying an aluminum paste or a step of depositing aluminum in a vacuum, and the formed aluminum layer is followed by the step. The process of removing is performed.

  Preferably, any one of a hydrogen gas exposure method, a catalytic CVD method, or a hydrogen plasma treatment method is used in the step of irradiating the crystalline silicon substrate surface with hydrogen.

  The crystalline silicon substrate of the present invention has a high carrier diffusion length. Moreover, the solar cell produced using such a board | substrate shows the outstanding characteristic (for example, conversion efficiency).

  The present invention provides a crystalline silicon substrate having a high carrier diffusion length in which a bonding mode between silicon and hydrogen and a hydrogen density profile inside the silicon are optimized, a method for producing the same, and a photoelectric conversion element using the crystalline silicon substrate. Is. These will be described in turn below.

(Crystalline silicon substrate)
The crystalline silicon substrate of the present invention is optimized for the bonding mode between silicon and hydrogen and the hydrogen density profile inside the silicon. Here, the logic of the meaning of “optimization of the bonding mode of silicon and hydrogen and the hydrogen density profile inside the silicon” and how it can be known is somewhat complicated. Will be described. Here, as a previous stage, (a) a method for measuring a bonding mode between silicon and hydrogen and (b) a method for measuring a hydrogen density profile inside silicon will be described.

(A) Method for measuring the bonding mode between hydrogen and silicon in a silicon substrate In order to obtain information on the bonding mode between hydrogen and silicon in a silicon substrate, temperature desorption gas analysis (hereinafter referred to as TDS analysis) of the silicon substrate is performed. )I do. TDS analysis examines the spectrum for the temperature of desorbed hydrogen observed when the sample is heated. FIG. 1 is a schematic diagram of an apparatus used for TDS analysis. As shown in FIG. 1, in the measurement in this embodiment, the sample such as the polycrystalline silicon substrate 11 was heated by introducing infrared rays generated by an infrared heater 13 outside the chamber through an infrared rod 12. Hydrogen generated at this time is measured by the quadrupole gas analyzer 14. The thermocouple 15 is attached to the sample fixing stage. The temperature of the sample fixing stage is used as the sample temperature. The temperature rising rate is preferably 0.1 to 1 ° C./second, and the maximum temperature is preferably 1000 ° C. or higher. The background pressure during heating is preferably 1 × 10 ⁻⁶ Pa or less.

(B) Method for Measuring Hydrogen Density Profile Inside Silicon On the other hand, the average hydrogen density in the entire substrate can be obtained by dividing the total desorbed hydrogen amount from the substrate estimated by TDS measurement by the substrate volume. By comparing the average hydrogen density of the entire substrate thus obtained with the hydrogen density in the vicinity of the surface measured by secondary ion mass spectrometry (hereinafter referred to as SIMS analysis), a hydrogen density profile is obtained to some extent. Can be estimated with accuracy.

  Here, the procedure for measuring the total amount of desorbed hydrogen from the substrate by TDS is as follows. In advance, a standard sample with a known hydrogen density is heated to desorb the gas and correspond to the number of gas molecules counted by the quadrupole gas analyzer. The amount of desorbed gas released from unknown data is estimated based on this data.

(Method for producing crystalline silicon substrate)
The method for producing a crystalline silicon substrate of the present invention comprises the steps of (a) forming an aluminum layer, (b) heating and firing the substrate, and (c) hydrogen treatment. Details of each of these steps will be described.

(A) Formation of aluminum layer The formation of the aluminum layer is obtained by vapor-depositing aluminum on a polycrystalline silicon substrate with a general-purpose vapor deposition machine. The thickness of the aluminum layer is desirably 10 to 100 μm. It is also possible to apply a paste containing fine aluminum powder to the polycrystalline silicon substrate. The thickness of the aluminum paste is desirably 100 to 1000 μm.

(B) Heating and baking of substrate For heating the substrate, it is preferable to use an infrared heating furnace as shown in FIG. The polycrystalline silicon substrate 21 is heated by the infrared heater 23 while being moved on the belt conveyor 22 made of stainless steel and moving in the direction of the arrow. The heating time is adjusted by the speed of the belt conveyor 22, and the heating temperature is adjusted by the output of the heater 23 and the speed of the belt conveyor 22. The heat treatment is performed in an inert atmosphere (for example, N ₂ gas, Ar gas, etc.) at about 600 ° C. to 1000 ° C., preferably about 700 ° C. to 850 ° C., for 10 to 100 seconds, preferably 10 to 60 seconds.

(C) Crystalline silicon substrate hydrogen treatment The crystalline silicon substrate hydrogen treatment method may be any of a hydrogen gas exposure method, a catalytic CVD method, a hydrogen plasma treatment method, and the like. In other words, it is a process in which hydrogen is exposed to the substrate in the form of molecules, atoms or ions. A case where the catalytic CVD method is used as a representative example will be described with reference to FIG.

In FIG. 3, in order to expose the surface of the polycrystalline silicon substrate 31 to hydrogen, the inside of the processing chamber is depressurized to about 10 ⁻⁵ Pa, and then hydrogen gas is allowed to flow from the gas inlet 32 at a flow rate of 100 sccm. The pressure in the chamber 34 can be adjusted to 0.1 to 10 Pa by adjusting the gas discharge amount from the outlet 33. The substrate 31 is heated to about 300 to 600 ° C., preferably about 400 to 500 ° C. by a heater (not shown). A direct current is supplied from a power source 36 to the catalyst body 35 made of tungsten wire, and the wire temperature is set to about 1500 to 2000 ° C. The hydrogen gas in contact with the wire is decomposed into atomic hydrogen and exposed to the surface of the polycrystalline silicon substrate 31. The exposure time is 10 minutes to 4 hours, preferably 30 minutes to 1 hour.

  The hydrogen treatment can be performed by other methods such as C.I. Vinckier, A .; Dumoulin, J. et al. Coroutouts and S.C. De Jaegere, J .; Chem. Soc. , Faraday Trans. 2, 84 (10), 1988, p. 1725-1740. A microwave excitation type hydrogen plasma exposure apparatus as shown in FIG. In this apparatus, the pressure in the chamber is reduced to 1 to 100 Pa, preferably 5 to 20 Pa, and the substrate is heated to about 300 to 600 ° C., preferably about 350 to 450 ° C. by a heater. The excitation power of the plasma is 20 to 100 W, preferably 40 to 60 W. The exposure of the substrate to hydrogen plasma is 30 minutes to 4 hours, preferably 90 to 210 minutes.

  Further, the hydrogen treatment may be performed by a hydrogen gas exposure method. This hydrogen gas exposure method is described in, for example, OPTOELECTRONICS— “Devices and Technologies” 9 (4), 1994, p. 523-536. 3-100% hydrogen gas is introduced into the chamber at atmospheric pressure. The substrate is heated to about 350 to 600 ° C., preferably about 400 to 500 ° C. by a heater. The hydrogen gas exposure time is 3 minutes to 2 hours, preferably 1 to 2 hours.

(Photoelectric conversion element)
A photoelectric conversion element can be manufactured using the crystalline silicon substrate of the present invention. When producing a photoelectric conversion element using the crystalline silicon substrate of the present invention, the photoelectric conversion element is produced by a known method, except that the crystalline silicon substrate of the present invention is used as the substrate. Can do. The example which manufactured the photoelectric conversion element using the crystalline silicon substrate of this invention is demonstrated taking a solar cell as an example.

  FIG. 4 is a diagram showing each step in manufacturing a photoelectric conversion element using the crystalline silicon substrate of the present invention. First, the polycrystalline silicon substrate 41 having a thickness of 300 μm was cleaned. Next, as shown in (a), the texture 42 of the polycrystalline silicon substrate 41 is etched at a liquid temperature of about 90 ° C. using a mixed solution of NaOH aqueous solution and isopropyl alcohol, and the surface of the silicon substrate 41 is highly etched. A micro pyramid of several μ was formed.

Subsequently, as shown in (b), a phosphorus glass liquid (hereinafter referred to as PSG liquid) is applied to the surface of the polycrystalline silicon substrate 41 by a spin coating method and baked at 850 ° C. An n-type silicon layer 43 having a thickness of about 1.0 μm and an impurity concentration of 1.2 × 10 ²⁰ cm ⁻³ was formed.

  Next, the steps shown in (c) to (f) are performed. These processes are processes that characterize the present invention. As shown in (c), an aluminum layer 44 is deposited on the back surface of the polycrystalline silicon substrate 41 to a thickness of 10 μm, and baked at 750 ° C. for 30 seconds as shown in (d). An aluminum silicon alloy layer 45 was formed on the contact surface. At this time, the aforementioned infrared heating furnace was used. Furthermore, as shown to (e), it immersed for 30 minutes using 100% hydrochloric acid, and the aluminum of the back surface was peeled.

  Next, as shown in (f), hydrogen atoms are exposed to the surface of the crystalline silicon substrate 41. At this time, the above-described catalytic CVD apparatus can be used. The substrate temperature is 450 ° C., the pressure is 0.4 Pa, and the processing time is 1800 sec. Hydrogen atoms may be exposed on the back side.

  Next, as shown in FIGS. 4G to 4I, electrodes are formed. First, as shown in (g), a patterning mask is placed at a place where the light receiving surface electrode is deposited, and an antireflection titanium oxide film 46 is formed at 400 ° C. by a normal pressure CVD method at a place other than the light receiving surface electrode. Formed. On the other hand, as shown in (h), a mask having a pattern opposite to that described above is installed on the portion where the light-receiving surface electrode is deposited, and the surface silver electrode 47 is deposited to a thickness of 10 μm on the fishbone pattern. did. Then, as shown in (h), an aluminum paste was screen-printed on the back surface of the polycrystalline silicon substrate 41 and then baked at 300 ° C. as shown in (i) to form a back electrode 48. Usually, the aluminum paste is fired at 750 ° C. instead of 300 ° C. This is because the aluminum silicon alloy layer 45 is formed on the contact surface between the back surface aluminum electrode and the silicon substrate. In the process of the present embodiment, since the aluminum silicon alloy layer 45 has already been formed in FIG. 4D, the process in FIG. 4I only needs to dry the aluminum paste serving as the back electrode. Therefore, low-temperature firing at about 300 ° C. is sufficient.

  The solar cell manufactured through the above steps has a conversion efficiency of about 0.5% and high efficiency as compared with the solar cell manufactured by omitting the steps (c) to (f) based on the present invention. It has become clear.

  The crystalline silicon substrate of the present invention will be further described. As described in the above embodiment, in order to further describe the “substrate in which the bonding mode of silicon and hydrogen and the hydrogen density profile inside the silicon are optimized”, this embodiment will be described in the following order.

  (I) The example of the manufacturing method of the polycrystalline silicon substrate of this invention is shown, and carrier diffusion length is shown. It is clearly shown that the electrical characteristics are improved when the treatment according to the present invention is performed.

  (Ii) The measurement result of the bonding mode between silicon and hydrogen of the substrate subjected to the processing according to the present invention will be described. It shows that the bonding mode of silicon and hydrogen is actually optimized as described in claim 2 in the substrate that has been processed according to the present invention and has improved electrical characteristics.

  (Iii) The measurement result of the hydrogen density profile inside the silicon of the substrate that has been processed according to the present invention will be described. It shows that the hydrogen density profile inside the silicon is actually optimized as described in claim 3 in the substrate that has been processed according to the present invention and has improved electrical characteristics.

(I) Example of manufacturing method of polycrystalline silicon substrate based on the present invention Each step of the method for manufacturing a crystalline silicon substrate of the present invention will be described as follows. First, the carrier diffusion length of a polycrystalline silicon substrate having a thickness of 300 μm is measured in advance by the SPV method (Surface Photo voltage method). Second, an aluminum paste is printed on the back surface of the substrate and heated at 750 ° C. for 10 to 120 seconds. Further, the aluminum is peeled off by immersion in hydrochloric acid. Third, hydrogen irradiation is performed. At this time, hydrogen may be a molecule, an atom, or an ion. In the experiment illustrated in this example, hydrogen atom irradiation is performed using a catalytic hydrogen treatment apparatus. The processing conditions are a substrate temperature of 450 ° C., a pressure of 0.4 Pa, and a processing time of 1800 sec. Fourth, the carrier diffusion length is measured again by the SPV method. From the above experiment, the change in carrier diffusion length before and after the basic process described in the above embodiment can be calculated.

  FIG. 5 shows the results of examining how the carrier diffusion length due to the subsequent hydrogen treatment changes when the firing time is changed in aluminum firing. The vertical axis represents the carrier diffusion length change (ΔDL) in the substrate due to aluminum coating firing and hydrogen treatment, and the horizontal axis represents the aluminum firing time (seconds).

  The following can be seen from FIG. That is, (1) the carrier diffusion length is improved in the range where the aluminum firing time is less than 60 seconds. When the aluminum firing time is set to about 30 seconds, ΔDL becomes the largest. (2) ΔDL of aluminum firing time of 10 seconds or more and less than 60 seconds is larger than ΔDL when hydrogen treatment is performed without firing aluminum (data of firing time 0). (3) On the other hand, at 60 seconds or more, ΔDL decreases rapidly, and rather becomes smaller than when hydrogen treatment itself is not performed.

  As described above, it can be seen that ΔDL becomes larger than that obtained when hydrogen treatment is simply performed by performing aluminum baking prior to hydrogen treatment and appropriately selecting the treatment time.

(Ii) Measurement result of bonding mode of silicon and hydrogen As described above, the improvement degree of the carrier diffusion length varies depending on the aluminum baking time prior to the hydrogen treatment. In order to clarify what causes the difference in the improvement degree of the carrier diffusion length, the TDS is plotted for 10 samples having ΔDL in the range of −70 to +40 μm plotted in FIG. Analysis was carried out.

  FIG. 6 is a hydrogen desorption spectrum showing a relationship between a typical substrate temperature and a hydrogen desorption rate. (A) shows the case of ΔDL = + 40 μm whose carrier diffusion length has been improved by hydrogen treatment, and (b) shows the case of ΔDL = −70 μm sample whose carrier diffusion length has deteriorated.

  Two substrates were prepared for each of (a) and (b). One substrate was subjected to hydrogen treatment, and the other substrate was used as a reference substrate not subjected to hydrogen treatment. The solid line shows the hydrogen desorption spectrum of the former (with hydrogen treatment), and the dotted line shows that of the latter (without hydrogen treatment). The difference between the two is hydrogen introduced into the substrate by hydrogen treatment.

  Observe data with hydrogen treatment. Although several peaks are observed, peaks below 550 ° C. have been investigated in detail in single crystal silicon, and it is known that their origin is hydrogen adsorbed on the substrate surface. On the other hand, the peak produced by the desorbed hydrogen at 600 ° C. or higher is considered to originate from hydrogen that was present inside the substrate, not the silicon surface. Since the quality of the substrate depends not on the surface but on the inside of the substrate, the following can be understood by comparing (a) and (b) for the peak of 600 ° C. or higher.

  (1) The peak position of the high temperature peak in (a) is near 800 ° C., whereas that in (b) is 700 ° C. or less. Thus, it can be seen that in the sample in which the carrier diffusion length is improved by the hydrogen treatment, the desorption of hydrogen occurs to a higher temperature.

  (2) In detail, in (a) and (b), the peak in the region of 550 ° C. or higher is a superposition of two peaks centered at 650 ° C. and 730 ° C. (The presence of two peaks is particularly clearly observed in (b).) As described above, two types of bonds having different bond forms are present inside the crystal grains.

  (3) The peak at 650 ° C. is dominant in (b), and the peak at 750 ° C. is dominant in (a). That is, in the sample in which the effect of improving the carrier diffusion length is seen, a bond that generates a peak near 750 ° C. is selectively generated compared to a bond that generates a peak near 650 ° C. by hydrogen treatment. Recognize.

  (4) The fact that hydrogen desorption is performed at a higher temperature means that the bond between hydrogen and silicon is strong. In the sample in which the carrier diffusion length improving effect is seen, it is considered that the hydrogen silicon bond formed in the substrate by the hydrogen treatment is stronger than the sample in which the improvement is not seen.

  Although the above is the result of only two examples, in order to make the reliability of the above result more reliable, the same check is performed for more samples. FIG. 7 shows the ratio of the amount of hydrogen desorbed at 750 to 800 ° C. and 800 to 850 ° C. with respect to the total amount of desorbed hydrogen and the change in carrier diffusion length for 10 samples having a ΔDL in the range of −70 to +40 μm. It is a figure showing the relationship between these using a graph.

  In FIG. 7, the vertical axis represents the ratio of the amount of desorbed hydrogen at 750 to 800 ° C. and 800 to 850 ° C. with respect to the total amount of desorbed hydrogen, and the horizontal axis represents the change in carrier diffusion length (ΔDL) (μm). . It can be seen that the greater the carrier diffusion length improvement, the greater the proportion of hydrogen desorbed at high temperatures of 750-800 ° C. and 800-850 ° C., respectively. FIG. 7 is a simple one that mechanically divides the temperature range and estimates the amount of hydrogen desorption existing in the range, and is not conscious of the two peaks centering on the above-mentioned 650 ° C. and 730 ° C., Even with such a rough analysis, there is a certain correlation between the improvement of the carrier diffusion length and the shape of the desorption hydrogen spectrum. Thus, it was experimentally confirmed that the desorption spectrum of hydrogen above 700 ° C. has a correlation with ΔDL.

(Iii) About the measurement result of the hydrogen density profile inside silicon The above result was the measurement result about the bonding state of hydrogen and silicon. Next, the hydrogen density profile was examined. Five typical samples were extracted from the series of samples described in FIG. 5, SIMS analysis was performed on them, and a hydrogen density profile up to a depth of 10 μm in the thickness direction in the substrate was measured. On the other hand, the average hydrogen density in the whole substrate (thickness 300 μm) was determined by TDS measurement.

  FIG. 8 shows local hydrogen densities at depths of 0.1 μm and 10 μm obtained by SIMS measurement, substrate average hydrogen density (0 to 300 μm) derived from TDS measurement, and carrier diffusion length change (ΔDL). It is the figure which represented this relationship using the graph. The following can be understood.

  (1) As ΔDL increases, the hydrogen density at a depth of 10 μm decreases, and conversely, the average hydrogen density of the entire substrate increases. In the sample where ΔDL is less than 0 μm, the ratio value to the former is 1/10 or less, whereas in the sample where ΔDL is +40 μm, the ratio value to the latter is increased to ½ or more.

  (2) This result means the following regarding the hydrogen density profile. In other words, in samples with ΔDL less than 0 μm, the hydrogen density decreases sharply with respect to the depth of the substrate, whereas in samples with ΔDL of 0 to 40 μm, the hydrogen density changes almost with respect to the substrate depth. do not do.

  (3) That is, a substrate showing ΔDL of 0 to 40 μm has a substantially flat hydrogen density profile in the thickness direction of the substrate, and hydrogen is well diffused in the substrate.

  Thus, the hydrogen density profile correlates with the carrier diffusion length change due to the hydrogen treatment, and it was experimentally confirmed that the effect of improving the carrier diffusion length increases as hydrogen penetrates deeper into the substrate.

  As described above, it was found that the hydrogen silicon bond or hydrogen density profile of crystalline silicon with a large improvement in diffusion length by hydrogen treatment was different from that of crystalline silicon with a small improvement in diffusion length. In other words, a polycrystalline silicon substrate having a high carrier diffusion length can be obtained by controlling and optimizing the bonding mode of silicon and hydrogen and the hydrogen density profile inside the silicon.

  The reason why the carrier diffusion length improvement differs depending on the manufacturing method of the polycrystalline silicon substrate based on the present invention is still in an imaginary range, but it is considered that depletion existing in the silicon crystal is related. That is, depletion is injected into the silicon substrate by baking aluminum before hydrogen treatment. Due to the difference in the aluminum firing conditions, some difference is given to the density and form of the depletion. On the other hand, it is known that hydrogen atoms enter the substrate via depletion. Any difference in depletion injected into the silicon substrate will change the hydrogen density profile in the silicon substrate, or the hydrogen-silicon bonding state, caused by subsequent hydrogen treatment. This gives a difference in the degree of defect inactivation by hydrogen treatment.

It is the schematic of the apparatus used for a temperature-programmed desorption gas (TDS) analysis. It is the schematic of the infrared heating furnace used for the heating of a board | substrate. It is the schematic of the catalytic CVD apparatus used for hydrogen irradiation in this invention. It is a figure which shows each process at the time of manufacturing the photoelectric conversion element using the crystalline silicon substrate of this invention. It is a figure showing the relationship between the baking time of aluminum, and the carrier diffusion length by subsequent hydrogen treatment using a graph. It is a figure showing the relationship between the substrate temperature at the time of hydrogen desorption, and a hydrogen desorption rate using a graph. It is a figure showing the relationship between the ratio of the amount of hydrogen desorbed at 750 to 800 ° C. and 800 to 850 ° C. and the change in carrier diffusion length with respect to the total amount of desorbed hydrogen. It is the figure which represented the relationship between the local hydrogen density in the depth of 0.1 micrometer and 10 micrometers, the average hydrogen density of a board | substrate, and carrier diffusion length change using a graph.

Explanation of symbols

  11 Polycrystalline silicon substrate, 12 Infrared rod, 13 Infrared heater, 14 Quadrupole gas analyzer, 15 Thermocouple, 21 Polycrystalline silicon substrate, 22 Belt conveyor, 23 Infrared heater, 31 Polycrystalline silicon substrate, 32 Gas inlet 33 Gas exhaust port, 34 chamber, 35 catalyst body, 36 power source, 41 polycrystalline silicon substrate, 42 texture, 43 n-type silicon layer, 44 back surface aluminum layer, 45 aluminum silicon alloy layer, 46 antireflection titanium oxide film, 47 Front silver electrode, 48 Back electrode.

Claims (7)

  A crystalline silicon substrate characterized in that a bonding mode between silicon and hydrogen is optimized, and a hydrogen density profile inside silicon is optimized.   In a spectrum representing a relationship between a temperature and a desorbed hydrogen velocity measured by a temperature-programmed desorption gas analysis that measures a desorbed hydrogen generated when the substrate is heated, the substrate temperature is 600 The peak of the desorbed hydrogen velocity has a peak in the range of ≧ 600 ° C., and the maximum peak value of the desorbed hydrogen velocity in the range of the substrate temperature of not less than 700 ° C. is in the range of not less than 600 ° C. and less than 700 ° C. The crystalline silicon substrate according to claim 1, wherein the crystalline silicon substrate is larger than a peak maximum value of velocity.   The average hydrogen density of 0 to 300 μm in the thickness direction of the substrate is in the range of 1/5 to 1 times the local hydrogen density at a depth of 10 μm in the thickness direction. A crystalline silicon substrate as described in 1.   A photoelectric conversion element using the crystalline silicon substrate according to claim 1. Forming an aluminum layer on the back surface of the crystalline silicon substrate;
Heating and baking the crystalline silicon substrate on which the aluminum layer is formed;
Irradiating the crystalline silicon substrate surface after the heating and baking with hydrogen;
The manufacturing method of the crystalline silicon substrate in any one of Claims 1-3 including this.   The step of forming the aluminum layer on the back surface of the crystalline silicon substrate is either a step of applying an aluminum paste or a step of vapor-depositing aluminum in vacuum, and after that step, the formed aluminum layer is removed. The method for producing a crystalline silicon substrate according to claim 5, wherein a process is performed.   7. The crystalline silicon substrate according to claim 5, wherein any one of a hydrogen gas exposure method, a catalytic CVD method, and a hydrogen plasma processing method is used in the step of irradiating the crystalline silicon substrate surface with hydrogen. Manufacturing method.

Images