JPH05270969A - Method for crystal growth - Google Patents

Abstract

PURPOSE:To inhibit the breakage of the quartz crucible and suppress the occurrence of crystalline defect in the case where the granules of the raw material are melted in combination with usual flake form of the material for preparation of the melt for crystallization, by improving the method for charging the raw material into the crucible in the CZ and molten zone process. CONSTITUTION:In the crystal growth process where the raw material for crystallization is melted in crucible 1b, and the melt is pulled up to effect the growth of a single crystal, the raw material in a flake form 20 is charged in the crucible 1b at the bottom, then granules of the row material 23 are charged on the flakes, and they are melted. The melting is performed, as the flakes are charged on the granules so that the flakes cover the granules, thus, granules can be prevented from flying off, thus sticking of the granules on the oven parts such as upper side walls in the crucible can preferably be prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a crystal growth method, and more particularly to a crystal growth method for growing a crystal such as a silicon single crystal used as a semiconductor material.

[0002]

2. Description of the Related Art There are various crystal growth methods, one of which is the Czochralski method (hereinafter referred to as the CZ method). FIG. 4 is a schematic cross-sectional view of a crystal growth apparatus used in the conventional CZ method, in which 1 indicates a crucible. The crucible 1 is composed of a bottomed cylindrical quartz crucible 1b and a bottomed cylindrical graphite crucible 1a fitted to the outside of the quartz crucible 1b. It is supported by a support shaft 7 which rotates at a predetermined speed and moves up and down. A resistance heating type heater 2 is arranged in a concentric cylindrical shape outside the crucible 1, and a heat insulating cylinder 3 and a heat insulating cylinder 4 made of graphite are arranged in a concentric cylindrical shape outside the heater 2, respectively. A pull-up shaft 5 such as a wire which is rotated at a predetermined speed in the same direction as the support shaft 7 or in the opposite direction is arranged on the central axis of the. A predetermined weight of the raw material for crystallization charged in the crucible 1 is heated by the heater 2 to start melting from the outer periphery in the high temperature crucible 1 and finally melt at the bottom in the relatively low temperature crucible 1. All becomes the melt 13. Then, the seed crystal 6 attached to the tip of the pulling shaft 5 is brought into contact with the surface of the melt 13 and the pulling shaft 5 is pulled up to grow the single crystal 12 formed by solidification of the melt 13. ing.

By the way, a doping impurity element is often added to the melt 13 in order to adjust the electric resistance or electric conductivity type of the single crystal 12. In this case, melt 1
Ratio C _S / C between the impurity concentration C _S in the single crystal 12 and the impurity concentration C _L in the melt 13 at the interface between the 3 and the single crystal 12.
_L (effective segregation coefficient Ke) is, for example, silicon (hereinafter, Si
0.35 for phosphorus as an impurity
(See “Si Single Crystal and Doping”; p35, Maruzen). Therefore, in the CZ method, the impurity concentration in the melt 13 increases as the single crystal 12 grows, and the impurity concentration in the single crystal 12 that is pulled up gradually increases accordingly. The impurities are segregated in the pulling direction of the crystal, so that the electric resistance of the single crystal 12 may be non-uniform.

A melt layer method is known as a method for growing crystals while suppressing the segregation of the doping impurities.
FIG. 5 is a cross-sectional view of a crystal growth apparatus used in a conventional melt layer method, in which 1 denotes a crucible. Similar to the CZ method, the crucible 1 includes a bottomed cylindrical quartz crucible 1b and a bottomed cylindrical graphite crucible 1a fitted to the outside of the quartz crucible 1b. Is supported by a support shaft 7 which rotates in a direction of an arrow in the drawing at a predetermined speed and moves up and down. A resistance heating heater 2 is concentrically arranged outside the crucible 1, and a graphite heat retaining cylinder 3 is concentrically arranged outside the heater 2. Is provided with a pull-up shaft 5 such as a wire which has the same axis as the support shaft 7 and rotates in the same direction or in the opposite direction at a predetermined speed. Only the upper part of a predetermined weight of the crystal raw material charged in the crucible 1 is melted by the heater 2 in the atmosphere of Ar gas 16 so that the melt layer 1 is formed on the upper part.
3. The solid layer 14 is formed on the lower part, and the solid layer 14 is melted from the upper side to the lower side while keeping the doping impurity concentration in the melt layer 13 constant. And the lifting shaft 5
By bringing the seed crystal 6 attached at the end of the above into contact with the surface of the melt layer 13 and pulling up the pulling shaft 5,
The single crystal 12 formed by solidification of the melt 13 is grown.

[0005]

The crystal raw material in the above-mentioned CZ method and melt layer method is a rod-shaped raw material (hereinafter referred to as a rod) made of high-purity polycrystalline Si, and the rod is crushed to a size of 30 mm. ~ 100 mm lump material (hereinafter,
Lamps), flaky raw materials having a size of 20 mm or less (hereinafter referred to as chips), and granular raw materials having a size of several tens of μm to 2 mm (hereinafter referred to as granules).

By the way, when the large or medium-sized crystal raw materials of the rod, the lamp, and the chip are charged into the quartz crucible 1b individually or in combination, the raw materials become dense with much time and labor. Even with such filling, since the gap between the raw materials is large and the filling rate is low, the charging amount of the raw material for crystallization per time is small and the single crystal to be pulled is small. Therefore, in order to cope with an increase in the size of a single crystal, it is necessary to melt the initially charged crystal raw material and then to additionally charge another crystal raw material for melting. Further, regarding the raw material price, there is a problem that the rod and the lamp are about twice as expensive as the granules.

Therefore, in order to reduce the raw material price and increase the filling rate to efficiently perform the melting, the rod, the lamp,
A method is used in which granules are used in combination with chips to fill the inside of the quartz crucible 1b. However, in the fusion layer method, when the lamp is arranged near the bottom surface in the quartz crucible 1b, the quartz crucible 1b is used.
The phenomenon of being easily damaged occurs.

To solve this problem, a quartz crucible 1
A method of filling the lower side wall surface and the bottom surface in b with small and medium-sized chips and granules and filling the upper portion with a lamp (Japanese Patent Laid-Open No. 03-
193692 gazette) is effective, but when granules are brought into contact with the bottom surface of the quartz crucible 1b to be filled and melted, in the melt layer method, the single crystal is pulled up by about 60 to 70%. Dislocation Free (hereinafter, DF
There is a problem that the rate decreases.

Further, in the CZ method and the fused layer method, a rod, a lamp, a tip and granules are used in combination and the quartz crucible 1b is used.
When the granules appear on the uppermost surface of the filled crystal raw material, the granules burst and fly during melting and adhere to the quartz crucible 1b, etc., which disturbs the flow of the Ar gas 16, and When the crystal was pulled up, it could fall into the melt 13 and cause crystal defects (hereinafter referred to as DF breakage).

The present invention has been made in view of the above problems, and suppresses breakage of the quartz crucible 1b and occurrence of DF breakage when the granules are used as the crystallization raw material in the CZ method and the melting layer method and melted. It is an object of the present invention to provide a crystal growth method capable of suppressing the above.

[0011]

In order to achieve the above object, a crystal growth method according to the present invention is a crystal growth method in which a raw material for a crystal in a crucible is melted and the melt is pulled upward to grow a crystal. In the above, the bottom of the crucible is filled with chips, and the top of the chips is filled with granules to be dissolved.

Further, in the above-mentioned crystal growth method, chips are filled in the upper part of the granules so as to cover the granules and dissolved.

[0013]

As a result of investigations by the present inventors regarding the influence of the type of the crystal raw material and the filling method on the DF breakage in the melt bed method, when the crystal raw material was melted without using granules,
Alternatively, even when granules are used in combination with the crystal raw material, DF breakage does not occur when the crystal raw material other than the granules is brought into contact with the bottom surface in the quartz crucible to be filled and melted,
When the granules are brought into contact with the bottom surface in the quartz crucible and filled to be melted, the solid layer is melted and DF breakage often occurs when the melt reaches the bottom surface in the quartz crucible. I understood that.

Considering the above findings, since the granules have a large specific surface area and the surface is activated, they react with the quartz crucible at a relatively low temperature to form impurities such as oxides on the surface of the granules, and the molten liquid Also, it is considered that the crystals adhere to the bottom surface in the quartz crucible due to the load of the crystal raw material. Further, it is considered that the above-mentioned adhered impurities are separated from the bottom surface in the quartz crucible and float in the melt upon contact with the high temperature melt, and adhere to the single crystal being pulled to cause DF breakage.

From the above consideration, when the granules are melted together with the rod, the lamp and the tip in the melting layer method to prevent the damage of the quartz crucible 1b and the crystal defects, the tip should be the quartz crucible. It has been found that the impurities do not stick to the bottom of the quartz crucible 1b when the bottom of the quartz crucible 1b is filled by contacting the bottom of the quartz crucible 1b and then filling the top of the chip with granules, rods or lamps and melting.

Further, as a result of confirming by filling a chip in the bottom of the container and a granule in the upper part of the chip using a transparent container, for example, in the case of a chip having a size of about 10 mm, it is about 50 mm. It was found that when the particles were filled to a thickness, or, in the case of chips having a size of about 5 mm, to a thickness of about 30 mm, the granules were trapped in the gaps between the chips and did not reach the bottom of the container.

Also, since the granules contain hydrogen and chlorine in the manufacturing process, even if the discharge treatment of these elements by heating is performed prior to melting, the above-mentioned contained elements remain in a considerable amount, and the content is close to the melting temperature. It was considered that when heated, the remaining gaseous hydrogen and chlorine rapidly expanded, causing the granules to burst and scatter. On the other hand, since the chip does not contain the above-mentioned elements, it was considered that the chips would not burst and scatter even if heated near the melting temperature.

Then, when the chip was covered with a thickness of about 10 mm or more and heated so that the granules were completely hidden above the granules, it was found that the granules burst but did not scatter to the outside.

According to the crystal growth method of the present invention, the bottom portion of the crucible is filled with the chips, and the upper portion of the chips is filled with the granules in combination with the lamp to melt the granules. Impurities on the surface are suppressed from sticking to the bottom of the crucible, and therefore the occurrence of DF breakage is suppressed. Further, since the lamp is not filled near the bottom of the crucible, damage to the crucible is prevented.

Further, when the chips are filled in the uppermost part of the crystal raw material so as to cover the granules and melted, the scattering of the granules is suppressed, and the granules adhere to the parts in the furnace such as the upper side wall in the crucible. Is suppressed. Therefore, the above-mentioned turbulence in the flow of argon is suppressed, and the deposits are prevented from falling into the molten liquid during the pulling of the crystal, and the occurrence of DF breakage is suppressed.

[0021]

EXAMPLES Examples and comparative examples of the crystal growth method according to the present invention will be described below with reference to the drawings.

FIG. 1 is a cross-sectional view schematically showing a method of filling a silicon polycrystal raw material according to an embodiment, in which 1b shows a quartz crucible. The bottom of the quartz crucible 1b is filled with 11 kg of a chip 20 having a size of approximately 3 to 20 mm, and four rods 22 having a diameter of approximately 120 mm and a length of approximately 200 mm are provided at the top of the chip 20 (total weight: 17. 2k
g) and then the size around 30 around the rod 22
10 kg of 100 mm lamp 21 is filled. further,
22.8 kg of granules 23 are charged and filled in the gap between the rod 22, the lamp 21, and the tip 20, and the tip 20 (4 kg) having a size of about 3 to 20 mm is filled so as to cover the granules 23 at the uppermost portion. ..

FIG. 6 is a cross-sectional view schematically showing a conventional method of filling a raw material for crystal as a comparative example, in which 1b shows a quartz crucible. At the bottom of the quartz crucible 1b, 15 kg of chips 20 having a size of approximately 3 to 20 mm and granules 2
11 kg of the mixed raw material 24 in which 22.8 kg of 3 was mixed, and four rods 22 having a diameter of approximately 120 mm and a length of approximately 200 mm were provided above the mixed raw material 24 (total weight: 17.2 k).
g) and then the size around the rod 22 is about 30 ~
10 kg of 100 mm lamp 21 was filled. further,
The remaining 26.8 kg of the mixed raw material 24 is charged into the rod 22.
And the gap and the upper part of the lamp 21 were filled.

In the following, a quartz crucible filled with the crystal raw material by the above method is used, and phosphorus having an effective segregation coefficient of 0.35 with respect to silicon is used as a doping impurity.
The result of pulling a silicon single crystal having a length of 1100 mm under the conditions shown in Table 1 below will be described.

[0025]

[Table 1]

As a result of melting the crystallization raw material by the fusion layer method under the above conditions, in the method of the comparative example, a large amount of the molten liquid was scattered at the initial stage of melting, and then the mixed raw material filled in the bottom of the quartz crucible was Impurities surfaced as they began to melt. On the other hand, in the method according to the example, it was observed that the melt was scattered and the impurities were floated up relatively.

FIG. 2 is a graph showing the frequency of the single crystal length when the single crystal pulling was performed 30 times by each of the crystal growth methods of the example and the comparative example and DF breakage occurred. As is clear from FIG. 2, in the comparative crystal growth method, the pulled crystal has a length of 700 mm to 800 m.
At the position of m, that is, when the crystal pulling rate is approximately 60 to 70%, D
About 70% of F breakage occurred, but in the crystal growth method of the example, about 90% of the crystals had no DF breakage at a total length of 1100 mm, and the occurrence rate of DF breakage was only about 10%.

As another example, as a result of melting the raw material for crystallization under the above conditions by the CZ method, scattering of the molten liquid and DF breakage due to floating of impurities in the granules were suppressed in the initial stage of melting.

As is clear from the above results, in the crystal growth method according to the embodiment, the DF of the crystal is larger than that in the conventional method.
Cutting is suppressed. By thus filling the bottom of the crucible with chips and filling the upper part of the chips with the granules and melting them, it is possible to prevent impurities from sticking to the bottom of the crucible, and to obtain crystals with less DF breakage. It was confirmed.

Further, by filling a chip on the upper part of the granules so as to cover the granules and melting the chips, the scattering of the granules is suppressed and the falling of the granules into the molten liquid during the pulling up of the crystals is suppressed. It was confirmed that crystals with few DF breaks were obtained.

[0031]

As described above in detail, in the crystal growth method according to the present invention, the bottom of the crucible is filled with chips, and the top of the chips is filled with granules and melted. It is possible to suppress the generation of impurities generated by the reaction between the granules and the quartz crucible without contacting the inner bottom part, to prevent the occurrence of DF breakage, and to prevent the crucible from being damaged.

When chips are filled and melted at the uppermost part of the charging raw material so as to cover the granules, the granules may be ruptured and scattered to adhere to internal parts such as a quartz crucible. Since it can be suppressed, it is possible to prevent the occurrence of DF breakage.

Further, since a large amount of granules can be used in combination, the filling rate is improved, the work of charging the raw material for crystallization is facilitated, and the raw material price can be reduced.

[Brief description of drawings]

FIG. 1 is a cross-sectional view schematically showing a method of filling a raw material for Si polycrystal according to an example of the present invention.

FIG. 2 is a graph showing D in the crystal growth methods of Examples and Comparative Examples.
It is a graph which showed the frequency of the single crystal length when F breakage occurred.

FIG. 3 is a cross-sectional view schematically showing a method for filling a Si polycrystal raw material according to another embodiment.

FIG. 4 is a schematic sectional view of a crystal growth apparatus used in a conventional CZ method.

FIG. 5 is a schematic sectional view of a crystal growth apparatus used in a conventional melt layer method.

FIG. 6 is a cross-sectional view schematically showing a conventional method of filling a raw material for crystal as a comparative example.

[Explanation of symbols]

 1b Quartz crucible 20 Tip 21 Lamp 22 Rod 23 Granule

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Takayuki Kubo 4-53-3 Kitahama, Chuo-ku, Osaka City, Osaka Prefecture Sumitomo Metal Industries, Ltd. (72) Inventor Shuichi Inami 4-chome, Kitahama, Chuo-ku, Osaka City, Osaka Prefecture No. 5 33 Sumitomo Metal Industries, Ltd.

Claims (2)

[Claims] 1. A crystal growth method in which a raw material for a crystal in a crucible is melted, and the molten liquid is pulled upward to grow a crystal, in which the bottom of the crucible is filled with a flaky raw material,
A crystal growth method, characterized in that a granular raw material is filled in the upper portion of the flaky raw material and dissolved. 2. The granular raw material is filled and dissolved in an upper portion of the granular raw material so as to cover the granular raw material.
The described crystal growth method.