JPH05275398A - Manufacture of porous silicon member - Google Patents

Abstract

PURPOSE:To provide a manufacturing method of a porous silicon member, which does not depend on conductivity types and resistances of silicon substrates, and which can be applied to a semiconductor field. CONSTITUTION:An even electrode 4 is provided on the rear surface of a single crystal silicon substrate 3 chiefly made of P-type and an anode potential is applied to the electrode. Meanwhile, reactive gas 9 containing halogen group elements is made to be plasma state, and ionized halogen element ion 10 is attracted toward the substance by applying a voltage to the substrate to perform fine hole etching on the surface of the silicon substrate 3. The substrate surface is irradiated with light emitted from the plasma. As a result, even on so called n-type silicon substrate which can not supply positive hole, the fine hole etching which is equal to that of p-type is performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for producing a porous silicon member using a dry process.

[0002]

BACKGROUND OF THE INVENTION Porous silicon is a structure discovered by chance during the study of electropolishing of silicon and is one of the special forms of silicon. Inner diameter is 2-100 as a general structure
Numerous ultra-fine pores of nanometer or ultra-fine pillars of the same outer diameter exist innumerably perpendicularly to the surface of the silicon substrate, and have a porosity of 10
It has a porous structure of 80% and a specific surface area of 200 to 600 m ² / cm ³ . Some researchers classify less than 20Å as microporous, 20-500Å as mesoporous, and more than 500Å as macroporous.

This porous silicon is chemically active and is expected to be applied in various ways. Among them, in recent years, much research has been conducted, and photoluminescence (PL) using this porous silicon is being actively pursued. , Development of electroluminescence (EL) light emitting device. Until the advent of this porous silicon, the observation of the light emission phenomenon from crystalline silicon alone has not been reported. The reason is that the optical bandgap of crystalline silicon is a maximum of 1.2 eV. On the other hand, in the case of porous silicon, a value of about 2.5 eV is obtained depending on the manufacturing conditions. at present,
As photoluminescence, strong visible light emission at room temperature has been confirmed by irradiation with ultraviolet rays or laser,
As electroluminescence, light emission having a center at 680 to 800 nm has been confirmed in a method using a light emitting cell.

Regarding the cause of this light emission phenomenon, excitons are confined in a quantum size silicon column, strong lattice strain in porous silicon influences the band structure, or the hydrogen termination layer on the surface. Many hypotheses have been submitted, such as luminescence by.

In addition, porous silicon is expected to have a wide range of applications such as SOI structure fabrication, heteroepitaxial substrate materials, electrode materials and gas sensors.

At present, only the so-called anodization method is used as the manufacturing method. In this anodizing method, a single crystal silicon substrate is electrochemically etched in an HF aqueous solution. Specifically, when a single crystal silicon substrate is immersed in a 5 to 50% HF aqueous solution and an anodic potential is applied, silicon atoms are etched by fluorine ions. At this time, if a current density of about 200 mA / cm ² or more is applied, electropolishing is performed on the entire substrate.
/ Cm ² , the surface silicon etching and elution is locally limited, and a sponge-like silicon layer is formed. It is said that it is necessary to form an ohmic electrode by vapor-depositing an aluminum thin film on the back surface of the substrate because the porous layer becomes non-uniform when the current density is distributed over the entire silicon substrate.

The local elution of silicon by etching proceeds only at the tip of the fine hole because the holes necessary for elution are supplied only from the tip of the hole where the electric field is concentrated.
The diameter of the micropores depends on the anode current density, and the depth depends on the time. The structure is determined by the resistivity of the substrate, that is, the impurity concentration. It is said that when the resistance is relatively low, it proceeds perpendicularly to the surface, but when the resistance is high, the elution progressing direction becomes random and a complicated network is formed. The specific resistance of the main board used is 0.001 to 10
Ωcm.

[Problems of the Prior Art] In addition to the resistance value of the substrate as described above, the production of the porous silicon layer has another point that it depends on the difference in the type of the silicon substrate. That is, when anodization treatment is performed using a p-type silicon substrate, the above-mentioned porosification is rapidly promoted, but n
When a silicon substrate of the type is used, the rate of porosity is significantly slower than that of the p type. Further, in the case of p type, the single crystallinity can be maintained in the porous layer, whereas in the case of n type, the single crystallinity cannot be maintained. In order to make the n-type silicon substrate porous, it has been dealt with by supplying holes by irradiating the n-type silicon substrate with light of 5 to 50 mW / cm ² , but this complicates the device configuration.

Further, as described in the above-mentioned application examples, application to the electronic material field such as formation of SOI structure is expected, but the surface state thereof is changed due to the treatment in the electrolytic solution. It is not clean enough to be used in semiconductor processes. Also, because of the porous structure, it is very difficult to clean the surface state.

[0010]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method for producing a porous silicon member which can be applied to the semiconductor field without depending on the conductivity type and resistance of the silicon substrate.

[0011]

According to the present invention, a plasma containing a halogen element is formed at a position distant from a single crystal silicon substrate, and at the same time, an anode potential is applied to the single crystal silicon substrate so that the halogen element ions are removed from the plasma. In the direction of the substrate to form fine pores on the surface of the substrate at a diameter of nm order or at intervals, which is a method for producing a porous silicon member. Further, in the above structure, by adding rare gas helium, argon, neon to the source gas, the luminous efficiency of plasma is increased,
It is characterized in that the substrate surface is irradiated with light at the same time. According to the research conducted by the present inventors, it is possible to form a porous silicon member having a clean surface state by applying a film forming or etching technique using a plasma CVD method regardless of the type of a silicon substrate. It was revealed. The device configuration and the manufacturing method will be described below.

The device used is a microwave (2.54G
Hz) or high frequency (generally 13.56 MHz) plasma CVD apparatus having a structure capable of arbitrarily applying a potential to the substrate can be widely used. In the present invention, FIG.
The quartz reaction tube 1 perpendicular to the microwave waveguide 2 as shown in
The microwave plasma CVD apparatus, which is widely used for diamond thin film production and the like, is used.

CF ₄ , He and H ₂ are supplied as reaction gases 9 from the upper part of the reaction tube of the microwave plasma CVD apparatus. CF ₄ is suitably added to He and H _{2 in} an amount of 5% or less, preferably 1%. If the concentration of CF ₄ is increased above this range, the reaction tube may be over-etched and the device may be damaged. The addition of He is
This is to increase the luminous efficiency of plasma and also to use as a diluent gas. The addition of H ₂ is for making the obtained porous silicon surface a stable hydrogen termination.

With the reaction gas stabilized, a microwave with an electric power of 300 to 600 W is applied to remove CF ₄ and He.
And H ₂ plasma is generated. By carefully adjusting the reflected wave and the like, a spherical glow discharge (spherical plasma discharge) 5 is generated at the center of the reaction tube 1 without contacting the wall of the reaction tube.

The conductivity type of the sample silicon substrate, the content (addition amount) of impurities, and the specific resistance are not particularly limited.
However, it is important to use a single crystal silicon substrate. The sample substrate is placed on a quartz substrate stage 6 which is placed on a molybdenum holder 7 and whose height can be adjusted arbitrarily. Further, the sample substrate 3 needs to be maintained at the anode potential as in the anodization method. Then, aluminum is vapor-deposited on the back surface of the substrate as an ohmic electrode 4 which is a back surface electrode of the substrate, and is connected to a DC power supply device through a substrate stage holding tube, and about +
A potential of 10 to 50 V is applied. The diameter of the micropores can be adjusted by the value of the applied potential. The substrate stage 6 is cooled by water inside the molybdenum holder 7 to protect the backside electrode.
The temperature is maintained at a low temperature of 50 to 300 ° C, preferably about 200 ° C.

Then, the sample substrate is brought close to the spherical discharge region while the potential is being applied. At this time, the substrate should not be inserted inside the spherical discharge. This is because fine discharge concentration occurs at the edge portion of the fine holes in the initial state of formation, and the fine holes disappear. In order to perform the etching formation of the fine holes without generating such a fine discharge concentration, it is necessary to cause the association with the etching species in the region outside the discharge, and the role thereof is to play a role in the sample substrate 3. Application of anode potential. Fluorine ions (F ⁻ ) that control the etching rate are ions having a negative charge, and are thus attracted to the surface of the substrate as an ion flow 10 by the anode potential applied to the substrate to advance the etching. In the case of the p-type, the present inventors found that when the hole-supplying impurities are scattered, in the case of the n-type, the concentration of fluorine ions occurs in the fine defects on the surface, and the formation of fine pores occurs, and then progresses. thinking.

As described in the conventional example, in the case of the n-type in which holes cannot be supplied, the growth of fine holes tends to occur in random directions. Also in the case of p-type, the micropore formation rate remarkably differs depending on the specific resistance value, that is, the impurity concentration. In order to cope with this, in the conventional anodization method, holes are supplied by light irradiation, but the structure of the present invention has a feature that light irradiation can be performed at the same time when the fine holes are formed. That is, since holes are generated by the light emission of the spherical discharge plasma 5 above the substrate, it is possible to rapidly form fine holes perpendicular to the substrate surface regardless of the conductivity type and the specific resistance of the substrate. At this time, in order to increase the light irradiation intensity, the input power should be as high as possible within the condition range. Further, it is effective to add a rare gas such as helium in order to increase the light irradiation intensity by the plasma. Hereinafter, the present invention will be described with reference to examples.

[0018]

EXAMPLES Example 1 In Example 1, the microwave plasma CVD apparatus shown in FIG. 1 was used to obtain a p having a specific resistance value of 10 Ωcm.
This is an example in which formation of a porous layer was attempted on the surface of the single crystal silicon substrate (100). The general procedure is as described above, but will be described in detail below.

As the reaction tube 1 of the apparatus, a quartz glass tube having an inner diameter of 46 mm was used. The substrate stage 6 is also made of quartz glass with a diameter of 36 mm, and is used by covering it on a molybdenum holder 7 having a water cooling mechanism inside. However, the quartz glass substrate stage 6 and the molybdenum holder 7 are provided with holes for passing a cable for applying an anode potential to the substrate. The substrate stage 6 is cooled by the water cooling mechanism of the molybdenum holder 7 (cooling is performed by the water flow 8). The surface temperature of the substrate 3 was measured by an optical thermometer from the observation window above the reaction tube, but it was very difficult to accurately measure it in terms of focus adjustment and the like due to the structure of the device. Therefore, in this example, a thermocouple was attached to the surface of the sample to conduct a preliminary experiment, and the temperature of the substrate surface during the reaction was predicted by comparing the thermocouple temperature with the display of the optical thermometer. The surface of the sample substrate was not particularly treated except that the back surface was provided with an ohmic electrode (back surface electrode) 4 formed by aluminum vapor deposition on the back surface.

Before starting the experiment, the inside of the reaction tube 1 is evacuated to 10 ⁻⁴ Torr. Then, the reaction gas is introduced from the upper part of the reaction tube. The reaction gas is CF ₄ (1%) + He (60%) + H
₂ (39%) = 100 ccm. The reaction pressure is set within the range of 1-40 Torr by the pressure control valve of the exhaust system. In this embodiment, the set pressure is 25 Torr. If the pressure is too low, plasma spreads inside the reaction tube, sputters the inner wall, and causes generation of impurities. Conversely, if it is too high, only a very small plasma discharge can be obtained, and the effect cannot be expected.

A few minutes after starting the gas introduction (usually about 5
Microwave). After matching adjustment,
Reproducible experiments cannot be performed until the discharge stabilizes. Therefore, at this stage, it is necessary to lower the substrate from the position of the waveguide to at least 10 to 15 cm or more.
After completion of this matching adjustment, it was confirmed that a spherical discharge was stably obtained at the center of the reaction tube, and the anode potential was applied to the sample substrate. In this embodiment, it is set to + 34V. Then, with the potential applied, the height of the highest part of the substrate is adjusted to be located 5 mm below the waveguide.

In this state, the reaction is carried out for about 10 to 15 minutes. In this embodiment, it is set to 10 minutes. As described above, it is very difficult to evaluate the temperature of the substrate surface at this stage. It was estimated to be 190 to 210 ° C by measurement from the upper observation window.

After the reaction time had elapsed, the substrate stage was moved downward, the discharge and gas supply were stopped, and the anode potential application was stopped to terminate the reaction. The sample is cooled immediately by leaving it to stand. At this time, H ₂ may be flown to promote the formation of hydrogen termination, or He may be flowed to promote the cooling of the substrate. After completion of the reaction, vacuum is pulled to 10 ^-4 Torr as before starting.

The obtained sample was evaluated as follows. i Evaluation of Porous Structure Evaluation of the structure of the porous layer, especially its cross section, is direct and important. However, since the produced porous layer has angstrom-level fine pores and columnar structures, it often exceeds the resolution of a scanning electron microscope (hereinafter referred to as SEM), which is easy to handle. Also in this example, following the conventional example, the cross-section was observed with a transmission electron microscope (hereinafter referred to as TEM). The sample was made from the sample of the above experiment. A cross-sectional view prepared from the observation result photograph is shown in FIG. Although exhibiting some refraction, the micropores 21 are smoothly grown in a generally downward direction. No branching of the fine holes 21 was observed at all. Pore diameter is about 20 nm, pitch is 2
It seems to be 0 to 30 nm. SEM that preceded TEM observation
From the observation, it was found that the thickness of the porous layer was about 50 μm.

Ii Evaluation of Crystallinity The extent to which the crystallinity of the silicon substrate is hindered by the porosity was evaluated using X-ray diffraction (hereinafter referred to as XRD). Here, attention is paid to the (400) plane peak of the diffraction chart, and the half width thereof is compared between the substrate before the porosification treatment and the substrate after the porosification treatment of this embodiment. The full width at half maximum of the (400) plane peak of the substrate before the porosification treatment was 5 (deg), which is very close to 7 (deg) in the porous layer obtained in this example. It was confirmed that the single crystal property of was well maintained.

Iii Evaluation of Photoluminescence (PL) The results of evaluation of PL of porous silicon are shown below. On the surface of the porous layer of the sample obtained in Example 1, He-
When excited by irradiation with Cd laser light (325 nm), red emission having a peak at 700 nm was observed as shown in FIG.

Example 2 In this example, the effect was observed by changing the value of the anode potential. The experimental conditions and experimental procedure are the same as in Example 1, but the anode potential applied before moving the substrate stage was set to 50V. The evaluation results are described below.

I Evaluation of Porous Structure TEM observation result In the photograph, as in Example 1, the micropores were steadily grown in the general downward direction although showing some refraction, and branching of the micropores was observed at all. There wasn't. However, the pore size was about 40 nm and the pitch was 15 to 20 nm, and there was a tendency for the pores to become larger overall, and it was found from SEM observation that the thickness of the porous layer was as thick as about 65 μm. It is considered that fluorine ions flying to the surface of the substrate are increased due to the influence of the increase in the anode potential.

Ii Evaluation of Crystallinity Below, the results of evaluating the crystallinity of the porous layer and the silicon substrate using XRD as in Example 1 are shown. Here, the full width at half maximum of the (400) plane peak by the XRD of the silicon substrate before the porosification treatment of this example and the half width at the (400) plane peak by the XRD of the silicon substrate after the porosification treatment were compared. It is a thing. As a result of the measurement, (4
It was confirmed that the half width of the (00) plane peak was 5 (deg) and that of the porous layer was 9 (deg), which was wider than that of Example 1, and that the crystallinity of silicon was inhibited. ..

Iii Evaluation of PL Similar to the sample obtained in Example 1, when the surface of the porous layer was irradiated with He—Cd laser light (325 nm) for excitation, the emission peak shifted to 760 nm. I found out that

[Embodiment 3] In this embodiment, the substrate type is p
An attempt was made to produce porous silicon by changing the mold to n-type.
In the case of n-type, the holes necessary for the reaction cannot be supplied from the substrate, and since the hole-supplying impurities that are the starting points for forming the micropores do not exist, it is difficult to form the porous layer and light irradiation is required. It has been said that there is. On the other hand, according to the present invention, it is possible to produce porous silicon that is comparable to p-type by the light from the plasma discharge space. In this example, the substrate was changed to the n-type silicon substrate (100) surface having a specific resistance value of 2 Ωcm, and the effect was observed. The other experimental conditions and experimental procedures are the same as in Example 1, but in order to give a starting point for starting the formation of fine pores on the surface of the n-type silicon substrate, a liquid in which 300-mesh alumina powder is dispersed in ethanol is used. Ultrasonic damage treatment was performed. As a result of SEM observation after the scratching treatment, a scratch of a recognizable size could not be confirmed (magnification: 7).
Up to 10,000 times). The evaluation results are described below.

I Evaluation of Porous Structure As shown in the schematic diagram of FIG. 4, the micropores 41 in the photograph of the TEM observation result show some refraction as shown in the schematic view of FIG. However, no branching of fine pores was observed at all. Further, the hole diameter and the pitch were found to vary within a range of about 10 to 15 nm. FIG. 4 shows a schematic diagram prepared from the TEM observation result. According to SEM observation, the thickness of the porous layer was about 45 μm. It can be said that the formation of porous silicon comparable to the p-type substrate was successful due to the effect of light irradiation by plasma light.

Ii Evaluation of Crystallinity As in Examples 1 and 2, using XRD, the half-angle value of the (400) plane peak in the sample substrate before the porous treatment of this example and the (400) in the sample substrate after the porous treatment were measured. ) The peak full width at half maximum of the
The half width of the 0) plane peak is 5 (deg), but that of the porous layer is about 12 (deg), which is wider than in Examples 1 and 2, and the crystallinity of silicon is obstructed. I understand. Also, the peak position was slightly shifted to the higher angle side, and internal stress (compression) was generated. This shift was also detected in Examples 1 and 2 although the amount was extremely small.

Iii Evaluation of PL Similar to the samples obtained in Examples 1 and 2, when the surface of the porous layer was irradiated with He—Cd laser light to be excited, the emission peak was as shown in FIG. It was shifted to 610 nm, and the emission color was closer to orange than red. It has been reported that the emission peak tends to shift to the blue side as the diameter of the fine pores becomes smaller, and the results of this example are in good agreement with the report.

[0035]

As described above, by using the method of the present invention, porous silicon can be produced regardless of the type of substrate. Although a microwave plasma CVD apparatus using a quartz reaction tube is used in the present invention, the same effect can be obtained by incorporating a high output high frequency power source and a bias potential applying mechanism in a parallel plate type high frequency plasma CVD apparatus. right. Even with the same microwave CVD apparatus, it is possible to increase the area by using a magnetic field type apparatus.

[Brief description of drawings]

FIG. 1 is a microwave plasma CVD used in the present invention.
apparatus

FIG. 2 is a schematic diagram of cross-sectional TEM observation of the sample obtained in Example 1.

FIG. 3 is a PL spectrum of the sample obtained in Example 1.

FIG. 4 is a schematic diagram of cross-sectional TEM observation of the sample obtained in Example 3.

FIG. 5: PL spectrum of the sample obtained in Example 3

[Explanation of symbols]

 1 Quartz Reaction Tube 2 Microwave Waveguide 3 Silicon Substrate 4 Substrate Backside Electrode 5 Spherical Discharge Plasma 6 Quartz Substrate Stage 7 Molybdenum Stage Holder 8 Cooling Water Flow 9 Reactive Gas Flow 10 Ion Flow

Claims (3)

[Claims] 1. A plasma containing a halogen element is formed at a position distant from a single crystal silicon substrate, and at the same time, an anode potential is applied to the single crystal silicon substrate to attract halogen element ions from the plasma toward the substrate.
A method for producing a porous silicon member, characterized in that fine holes are formed on the surface of the substrate at diameters or intervals of nm order. 2. A method for producing a porous silicon member, wherein the halogen-based element ion of claim 1 is fluorine. 3. The method for producing a porous silicon member according to claim 1, wherein the rare gas is used as a source gas to enhance the luminous efficiency of plasma and irradiate the substrate surface with light.