KR100445190B1 - Manufacturing method of silicon single crystal ingot - Google Patents

Abstract

The present invention improves gettering ability by adding nitrogen during the growth of silicon single crystal ingots, and uniformly grows and cools the ingots in a radial direction, thereby providing a vacancy-type. The present invention relates to a technique for reducing the area where coarse defects are distributed in the area of defects and increasing the area where microdefect areas are distributed. Method according to the invention, the method for producing a silicon single crystal ingot in the Czochralski method, adding nitrogen during ingot growth, growing the ingot with the pulling rate of the ingot to 0.55 (mm / min) or more; Controlling the ingot growth rate so that the pulling apparatus reduces the vertical temperature gradient deviation between the center and the edge of the ingot while the ingot is grown to maintain the overall vertical temperature gradient uniformly; And the pulling apparatus controls the oxidized lamination defect region to be formed in a coaxial ring shape at the edge portion in the radial direction of the ingot while the ingot is grown, and controls the microdefect region to be enlarged to be formed inside the oxidized lamination defect region. As a result, there is an effect of increasing the production yield of high-quality wafers that can be used even in highly integrated device processes having a fine line width.

Description

Manufacturing method of silicon single crystal ingot

The present invention relates to a method for producing silicon single crystal ingot by Czochralski method, and more particularly, to add getter at the time of ingot growth to improve gettering ability and to radially grow the ingot growth and cooling conditions. The present invention relates to a technique for reducing the area where coarse defects are distributed in the area of vacancy-type defects existing in the ingot and increasing the area where microdefect areas are distributed.

A silicon wafer used as a material for producing electronic components such as a semiconductor is made by thinly cutting a silicon single crystal ingot. Czochralski (CZ) method is a representative method for producing silicon single crystal ingot, which is a method of growing a crystal while immersing a single crystal seed crystal in molten silicon and slowly pulling it. S. Wolf and RN Tauber's paper, Silicon Processing for the VLSI Era, volume 1, Lattice Press (1986), Sunset Beach, CA. In the following, a general silicon single crystal ingot production process by the Czochralski method will be described.

First, after the necking process of growing thin and long single crystals from the seed crystals in order to remove the dislocations present at the lower end of the seed crystals, the shoulders are grown in the radial direction to make the target diameter. After this, single crystals having a constant diameter are grown. This process is called a body growing process, where the grown part becomes a wafer-made part. After the body is grown by a certain length, the crystal growth process is completed through a tailing process in which the diameter of the single crystal is gradually reduced and finally separated from the molten silicon. The crystal growth process is performed in a space called a hot zone, which refers to the overall environment that forms a space around the molten silicon and the ingot contact when the molten silicon is grown into a single crystal ingot in the crystal growth apparatus. do. The crystal growth apparatus is composed of various components such as a molten silicon crucible, a heating apparatus, a thermal insulation structure, and an ingot raising apparatus.

On the other hand, since the defect characteristics inside the single crystal are very sensitive to the growth and cooling conditions of the crystal, many efforts have been made to control the type and distribution of growth defects by controlling the thermal environment near the growth interface.

Growth defects are largely divided into vacancy-type and interstitial-type, and when baconic defects or interstitial defects exist above equilibrium concentrations, coagulation occurs and develops into three-dimensional defects. It is known to become. According to Voronkov (V.V. Voronkov, The Mechanism of Swirl Defects Formation in Silicon, Journal of Crystal Growth 59 (1982) 625), the formation of these defects is known to be closely related to the V / G ratio. Where V is the growth rate and G is the vertical temperature gradient in the crystal near the growth interface (hot zone). That is, if the value of V / G exceeds a certain threshold, a vacancy type and an interstitial type defect are formed under the following conditions. Therefore, when growing a crystal in a given hot zone, the kind, size, and density of defects present in the crystal are affected by the pulling speed.

1 and 2 are diagrams for explaining the characteristics of the ingot grown in a general manner.

1 is a longitudinal cross-sectional view of an ingot showing the appearance of a defect region generated along a longitudinal cross section of the ingot grown while changing the pulling speed. The ingot was first grown at high speed by drawing the upper part of the drawing, and then grown to the lower part while slowly decreasing the pulling speed.

As shown in FIG. 1, interstitial defect regions 11 exist in a portion grown at a low speed, and vacancy-type defect regions 12 exist in a region formed by pulling at a high speed. Then, between the defect region of the vacancy type defect region and the defect region of the interstitial type, there is a vacancy advantage defect region 14 and an oxidative lamination defect region 13 adjacent to the vacancy advantage region.

The oxidative lamination defect causes the oxidative lamination defect to be pushed to the edge of the cross section when the pulling speed is increased to a certain level or more, resulting in the distribution of vacancy-type defects throughout the cross section.

On the contrary, if the pulling speed is decreased, the oxidative lamination area contracts to the center of the cross section and eventually disappears, and the predominantly defective area of bacon appears, and if the pulling speed is further reduced, the interstitial predominant defect area appears. Interstitial-type defect regions 11 are present throughout.

However, in the prior art, there is a problem that the cooling conditions of the vertical temperature gradient G in the radial direction of the ingot during the crystal growth are considerably uneven due to the structural problem of the hot zone. At the center of the ingot, heat is transferred to the edge of the ingot through conduction and must be radiated again, whereas at the edge of the ingot, the heat is directly radiated through radiation, resulting in a deviation of the vertical temperature gradient in the radial direction of the ingot. Since the edge of the ingot has a faster cooling rate than the center, the G value is small at the crystal center and increases in the radial direction. Therefore, even at the same pulling speed, the V / G value becomes relatively large in the center, so bacon defects become ingots. High density in the center There are many coarse bacon defects such as COP (Crystal Originated Particle) or FPD (Flow Pattern Defect) in the central area of the ingot.

FIG. 2 is a defect distribution diagram of a section cut by the II cutting line in FIG. 1. As shown in FIG. 2, in the ingot formed when pulling at the pulling speed of the portion indicated by II of FIG. 1, an oxidative-induced stacking fault (OiSF) region is located at the edge of the ingot. This figure shows a typical defect distribution of a single crystal cross section grown by the Czochralski method by adjusting the ingot pulling speed at high speed.

As shown in FIG. 2, there is a wide range of coarse bacon defect regions 12 in which coarse bacon defects exist in the center of the ingot, and an oxidative lamination defect region 13 exists outside thereof, and at the outermost portion thereof. The defect free area (Baconsea dominance) 14 is present.

As described above, when the oxidized lamination defect region 13 is placed at the edge of the ingot, coarse vacancy defects, that is, defects such as COP and FPD, exist in the center of the ingot, thereby forming a highly integrated semiconductor device having a fine line width. It cannot be used as the material of the wafer to be made. Increasing the pulling speed for ingot growth generates coarse bacon defects, which cannot be used for wafers to form fine electronic circuits, and reducing the pulling speed to reduce coarse defects reduces productivity. There was even the risk of forming a large dislocation pit (LDP) area larger than the coarse defect of the baconcitic type.

As described above, these coarse defects, which occur in some form in the radial direction of the ingot, are known to be able to be eliminated to some extent by reducing the vertical temperature gradient deviation between the center of the ingot and the outer periphery, but it is made as such. Lost wafers also cannot be a complete solution to ensure that they contain only the microscopic defects required by IC wafers above 64M DRAM.

The terminology used herein is MCLT: Minority Carrier Lifetime, COP: Crystal Originated Particle, FPD: Flow Pattern Defect, LSTD: Light Scattering Topography Defect, OiSF: Oxidation-induced Stacking Fault, DSOD: Direct Surface Oxide Defect, BMD : Bulk Micro-Defect, DZ: Denuded Zone, XRT: X-Ray Topography.

Therefore, the present invention has been proposed to solve the above-mentioned problems of the prior art, and when the nitrogen is added during the growth of the ingot by the Czochralski method, the gettering capability is improved, and the ingot is grown at the time of ingot growth. By reducing the vertical temperature gradient deviation between the center and the outer circumference to prevent coarse defect areas, the productivity and quality of the ingot are improved to increase the yield of wafers that can be used even in the high-density device process with fine line width.

The method according to the present invention for achieving the technical problem, a method for producing a single crystal silicon ingot formed in a predetermined length in the direction of the central axis with a radius of a constant length from the central axis in a predetermined pulling apparatus in the Czochralski method. The method of claim 1, further comprising the steps of: (a) adding the nitrogen while the pulling apparatus is growing the ingot, and growing the ingot with the pulling speed of the ingot being 0.55 (mm / min) or more; (b) controlling the ingot growth rate so that the pulling apparatus maintains the overall vertical temperature gradient uniformly by reducing the vertical temperature gradient deviation between the center and the edge of the ingot while the ingot is growing; And (c) the pulling apparatus controls the oxidized lamination defect region to be formed in a coaxial ring shape at the edge portion in the radial direction of the ingot while the ingot is grown, and controls the microdefect region to be formed inside the oxidized lamination defect region. Reducing the vertical temperature gradient deviation in the step (b), characterized in that the pulling device is controlled by controlling the melt gap of the predetermined heat shield provided in the pulling device.

Nitrogen added to the ingot is characterized in that 1E12 to 1E14 (atoms per cubic cm), characterized in that the micro-defective region formed in the ingot is 10% or more of the radius, the presence range of the micro-defective region formed in the ingot Is characterized in that more than 10% of the length of the ingot, the initial oxygen concentration contained in the ingot is characterized in that more than 11.5ppma. In particular, the initial oxygen concentration contained in the ingot is characterized in that more than 11.5ppma, the amount of oxygen precipitated 2.0ppma or more after a predetermined heat treatment.

1 is a longitudinal sectional view of a silicon single crystal ingot grown in a general manner;

FIG. 2 is a defect distribution diagram of a section cut by the II cutting line in FIG. 1. FIG.

3 is a schematic view showing a hot zone near an ingot growth interface in accordance with the present invention.

Figure 4 is a defect distribution diagram of the cross section cut radially ingot in accordance with the present invention.

Figure 5 is a graph showing the vertical temperature gradient in the hot zone to uniform the growth and cooling conditions in the radial direction of the ingot according to the present invention.

6 is a graph showing the pulling speed according to the position of the ingot in the longitudinal direction according to the present invention.

7 is a graph showing the nitrogen concentration according to the position in the longitudinal direction of the ingot according to the present invention.

8 is an image showing the FPD distribution and DSOD region according to the radial position observed on the wafer by cutting the ingot grown in a general manner.

9 is an LLS distribution diagram according to the position in the longitudinal direction of the ingot according to the present invention.

10 is an image of a longitudinal section according to the position in the longitudinal direction of the ingot according to the present invention.

11 is an FPD distribution diagram according to the position of the ingot in the longitudinal direction according to the present invention.

12 is a graph showing the amount of oxygen precipitated according to the initial oxygen concentration of the ingot according to the present invention.

Hereinafter, the present invention will be described with reference to the accompanying drawings.

3 is a schematic diagram illustrating a hot zone near an ingot growth interface that is added nitrogen in accordance with an embodiment of the present invention.

The method for producing nitrogen-added ingot according to the embodiment of the present invention, the temperature variation between the center portion and the edge portion, which occurs due to the fact that the cooling conditions of the vertical temperature gradient (G) in the radial direction of the ingot during the crystal growth is quite uneven. It is to reduce the coarse defects that cannot be removed even by adding nitrogen. It grows the ingot while adding nitrogen, and adjusts the amount of heat radiated from the heating element to the outer periphery of the ingot by adjusting the gap (melt gap) between the bottom of the heat shield and the silicon melt, thereby reducing the vertical temperature gradient of the ingot outer periphery, This can be solved by increasing the vertical temperature gradient at the center of the ingot by cooling the top of the top and the top of the heat shield.

That is, as shown in FIG. 3, when the ingot is grown while adding nitrogen at a pulling speed of 0.55 mm / min or more, a heat shield 34 or a crucible support 37 is used using the heat shield 34. By controlling the radiant heat from the silicon melt (32) melted in the quartz crucible (36), which is supported by the (), the cooling rate difference by radial position is reduced by reducing the cooling rate of the periphery of the ingot (33). In this case, the heat shield 34 prevents heat from being transferred from the melt 32 to the upper ingot 33 by using a material having excellent thermal insulation. In the space 35 under the heat shield (melt gap, the gap from the bottom of the heat shield to the surface of the molten silicon), a condition in which heat does not easily escape is created so that the cooling rate at the periphery of the ingot near the interface is lowered. do. In addition, the cooling conditions are controlled by adjusting the height of the melt gap to adjust the amount of radiant heat coming from the heating element and the area of the ingot to be heated.

Evaluation of the ingot manufactured as described above is based on the XRT image of the vertical end of the ingot crystal subjected to the stationary experiment in the hot zone. It was indirectly confirmed that the concentration and cooling rate were uniform in the radial direction.

4 is a defect distribution diagram of a cross section cut radially in the nitrogen-added ingot according to an embodiment of the present invention.

When a single crystal ingot grown in a thermal environment is minimized to minimize the thermal environment difference in the radial direction of the ingot and the entire range of the ingot exists as a vacancy region, the wafer is fabricated and the defect distribution is examined, as shown in FIG. Similarly, coarse bacon defects, such as COP (Crystal Originated Particle) or FPD (Flow Pattern Defect), are present at low density in the central portion 41, followed by the formation of microdefect regions 42, followed by oxidative lamination defect regions ( 43) and a bacon sea predominance defect-free region 44 is narrowly formed. The baconic dominant defect free area 44 is formed with a width of 10% or less of the radius.

The remarkable difference with the cross section of the ingot grown by the general method shown in Fig. 2 is that coarse defects such as COP and FPD are limited in the center, and microdefect regions 42 are distributed outside.

Here, the coarse defect region refers to an area in which FPD defects are distributed, and the microdefect region refers to an area without FPD defects and has a DSOD (Direct Surface Oxide Defect, Reference: JG Park, JM Park, KC Cho, GS Lee and HK Chung, Electrochemical Society). Proceedings 97-22 (1997) 173). In addition, COP, LSTD, and the like are found in the coarse defect region, and such coarse defects are not allowed to be used as wafers for IC production of 64M DRAM or more.

With reference to the drawings of the accompanying embodiment will be described in more detail the ingot production method is nitrogen-added according to an embodiment of the present invention.

Experimental conditions when producing a nitrogen-added ingot according to an embodiment of the present invention are shown in Figs. In addition, the initial oxygen concentration included in the ingot was adjusted to about 11.0 ppm to 12.5 ppm by controlling the flow of the atmospheric gas and the rotational speed of the quartz crucible.

5 is a graph showing a vertical temperature gradient in a hot zone in which the growth and cooling conditions in the radial direction of the nitrogen-added ingot are uniform according to an embodiment of the present invention.

In order to uniformize the growth and cooling conditions of the crystal in the radial direction, as described above, by adjusting the distance between the bottom of the heat shield and the silicon melt, the amount of heat radiated from the heating element to the outer circumference of the ingot is reduced, thereby reducing the vertical temperature gradient of the outer circumference of the ingot, By increasing the vertical temperature gradient in the center of the ingot by cooling the upper part of the ingot and the top of the heat shield, the radial Gr / Gc (Gr is the vertical temperature gradient along the radial distance, Gc is the vertical temperature gradient on the central axis) is shown in FIG. 5. Ingots were grown to give a curved graph as shown in.

As shown in the graph of Fig. 5, the vertical temperature gradient 132 from the ingot center to the outer circumferential portion is more uniform than the conventional vertical temperature gradient 131 under the atmospheric conditions of the embodiment of the present invention. In particular, the difference (G) of the temperature gradient between the central part and the outer circumferential part was 16.49 K / cm in the prior art, so that it was 2.87 K / cm in this embodiment, which became very uniformly 3 K / cm or less. In addition, the average value of the vertical temperature gradient between 1120 and 1070, which is known as the temperature section where COP is formed, is much larger than the prior art, with 32.31 K / cm at the center and 43.55 K / cm at the edge, and is known as the section where OiSF nuclei are formed. Since the average value of the vertical temperature gradient between 1070 and 800 is also much larger than in the prior art, at 23.81 K / cm at the center and 26.14 K / cm at the edges, these defects pass very quickly through the temperature range where these defects occur. This occurred less.

6 is a graph showing the pulling speed according to the position in the longitudinal direction of the ingot to which nitrogen is added according to the embodiment of the present invention.

As described above, it was found that the pulling speed was more than 0.55 mm / min under the temperature gradient distribution, which was tested at the pulling speed of about 0.6 mm / min as shown in FIG. 6.

7 is a graph showing the nitrogen addition concentration according to the position in the longitudinal direction of the ingot to be nitrogen added according to an embodiment of the present invention.

As shown in FIG. 7, in the present experiment, the nitrogen addition level was 2.4E13 (atoms per cubic cm) at the start of the crystal growth, and thus the calculated value according to the longitudinal position of the ingot was shown for the grown ingot. The graph is shown as 7. Here, as for the amount of nitrogen added to the single crystal silicon ingot, it was found that 1E12 to 1E14 (atoms per cubic cm) are suitable as shown in FIG.

Herein, when nitrogen is added during silicon single crystal ingot growth, aggregation of atomic vacancy in silicon is suppressed, as described in T.Abe, H. Take no, Mat, Res. Soc. Symp. Proc. Vol. 262, 3,1992 '. This effect is because the aggregation process of atomic vacancies shifts from homogeneous nucleation to heterogeneous nucleation. When nitrogen is added to grow single crystal silicon ingot by Czochralski method, the size of void defects, which are aggregates of atomic vacancies, is increased. It is possible to make very small, and such small crystal defects are easily extinguished by heat treatment.

Furthermore, in the case where nitrogen is added during silicon single crystal ingot growth, the oxygen precipitate density increases with the effect of promoting the precipitation of oxygen, 'F. Shimura, R. S. Hawkock, Apply. Phys. Lett. 48, 224, 1986 'and the like. Therefore, when the wafer made by cutting the ingot is heat-treated, the bulk region has a high density of oxygen precipitates, and the oxygen precipitates on the wafer surface are extinguished due to the diffusion of nitrogen added to the wafer, thus forming a wafer having excellent gettering capability. To make it possible.

Experimental results of the nitrogen added ingot according to an embodiment of the present invention is shown in Figures 8 to 13.

FIG. 8 is an image showing an FPD distribution and a DSOD region according to a radial position of a wafer in which the ingot grown by the general method is observed. However, this is a diagram for explaining the distribution of the oxidized lamination defect region and the micro defect region, and shows an example in which nitrogen is not added.

As shown in FIG. 8, the characteristic is that there is a micro-defective region between the dotted line and the oxidatively-laminated defect region in the bacony region. 8 is a graph showing a distribution of FPD defects in which a wafer of a cross section is inspected by a chemical etching method. A bacon predominant defect free zone (see reference numeral 44 in FIG. 4) is present at a width of 10% or less in the direction of the central axis from the wafer edge, followed by the oxidized stacked defect zone. Here, it can be seen that the DSOD region in which microdefects exist is located from the center to the oxidative lamination defect region, and the FPD region in which coarse defects exist exists only in the center, and as a result, the microdefect region actually exists.

By the way, the microdefect region shown in FIG. 8 can be expanded to some extent by increasing the thermal thermal uniformity in the radial direction, but the wafers thus produced also contain only the microdefects of the level required by the IC production wafer of 64M DRAM or more. Since this is not a complete solution, the addition of nitrogen during the growth of silicon single crystal ingots allows the elimination of cohesion of atomic vacancy, ultimately eliminating coarse defect areas. Therefore, all of the inside of the oxidized lamination region can be a non-defective region. Also, if the vertical temperature uniformity and the amount of nitrogen addition are properly adjusted, no COP or FPD defects are found and the central axis of the ingot including the oxidative lamination region is not found at all. It is possible to grow ingots that can produce wafers with only minute defects.

FIG. 9 compares the addition of nitrogen with and without addition of LLS (Localized Light Scatterer) distribution according to the longitudinal position of the ingot according to the embodiment of the present invention. Most of the LLS here is known to be COP originating from a vacancy cluster.

As shown in Figure 9, the LLS distribution according to the position of the longitudinal direction of the ingot can be seen to be significantly reduced compared to the case where the nitrogen is not added, when the nitrogen is added, in particular, the micro LLS of less than 0.14㎛ 50% or more Decreased.

FIG. 10 is an image of a longitudinal cross section according to a longitudinal position of a nitrogen-added ingot according to an embodiment of the present invention, and FIG. 11 is added to a FPD distribution chart according to a longitudinal position of an ingot according to an embodiment of the present invention. It is a comparison with the one that is not added with nitrogen.

As shown in FIG. 10, as the length of the ingot varies (corresponding to the positions 130, 280, 430, 580, and 730 mm from the top of the ingot from top to bottom, respectively), the oxygen lamination defect region is almost unchanged at the edge portion. It can be seen that it is maintained, and the observation of the FPD is rarely seen, as shown in the distribution diagram of the FPD shown in Figure 11, when nitrogen is added, compared to the case where the nitrogen is not added, more than 75% FPD A decrease distribution is shown.

As described above, the results as shown in FIGS. 9 to 11 show that the aggregation of atomic vacancy is suppressed by the nitrogen addition effect made during the growth of the silicon single crystal ingot, and various defect densities such as COP and FPD Is reduced.

12 is a graph comparing the amount of precipitated oxygen (Oi) according to the initial oxygen concentration of the ingot according to the embodiment of the present invention with and without the addition of nitrogen. Here, the precipitation of oxygen appears after a predetermined heat treatment performed under an atmosphere such as hydrogen and oxygen.

As shown in Fig. 12, the amount of oxygen precipitation according to the initial oxygen concentration is proportional in all cases, but in the case of adding nitrogen, the amount of precipitation uniformly at the radial position of the ingot is 2 ppm or more when the initial oxygen concentration is 11.5 ppm or more. This is an effect of uniformly improving the overall gettering capability of the wafer cut in the nitrogen-added ingot according to the embodiment of the present invention, and serves to reduce various defect densities such as COP and FPD.

As described above, in the method for producing a silicon single crystal ingot which is nitrogen-added according to the embodiment of the present invention, when nitrogen is added when the silicon single crystal ingot is grown by the Czochralski method, the size of void defects, which are aggregates of atomic vacancy, etc. It is possible to make very small, such small crystal defects are easily extinguished by heat treatment, and with the effect of promoting oxygen precipitation, when the oxygen precipitate density becomes high and heat treated, high density oxygen precipitates in the bulk region Oxygen precipitate on the surface of the wafer is a gettering effect that is extinguished due to the diffusion of nitrogen added to the wafer, and has a constant length from the central axis and a constant length in the direction of the central axis in a given pulling device. Single crystal silicon ingot formed by the Czochralski method In the method, while the ingot is grown in nitrogen while the ingot is grown, and the ingot is grown with the pulling speed of the ingot to be 0.55 (mm / min) or more, the center of the ingot and the edge portion of the ingot are vertical. The ingot growth rate is controlled to maintain the overall vertical temperature gradient uniformly by reducing the temperature gradient variation, so that the oxidative lamination region is formed in the coaxial ring shape at the edge portion in the radial direction of the ingot while the ingot is growing. The microdefective region was formed inside the region.

As described above, the silicon single crystal ingot according to the present invention has a very low crystal defect size and density compared to the conventional ingot and can increase the pulling speed, thereby increasing the productivity of the ingot and increasing the quality of the ingot without increasing the manufacturing price. As a result, it is possible to improve the yield of high-quality wafers that can be used even in highly integrated device processes having a fine line width.

Claims (7)

In the method for producing a silicon single crystal ingot having a constant radius from the central axis and a predetermined length in the direction of the central axis in a predetermined pulling apparatus in a Czochralski method, (a) doping nitrogen into the silicon melt in the pulling apparatus and growing an ingot from the silicon melt, wherein the ingot is grown with a pulling rate of the ingot at least 0.55 (mm / min); (b) controlling the ingot growth rate so that the pulling device maintains the overall vertical temperature gradient uniformly by reducing the vertical temperature gradient deviation between the center and the edge of the ingot while the ingot is growing; And (c) the pulling apparatus controls the oxidized lamination defect region to be formed in a coaxial ring shape at the edge portion in the radial direction of the ingot while the ingot is grown, and controls the microdefect region to be formed inside the oxidized lamination defect region. Silicon single crystal ingot manufacturing method comprising a. The method of claim 1, Reducing vertical temperature gradient deviation in the step (b), And the pulling device controls the melt gap of a predetermined heat shield provided in the pulling device to be adjusted. The method of claim 1, Nitrogen added to the ingot is 1E12 to 1E14 (atoms per cubic cm) Method for producing a silicon single crystal ingot, characterized in that. The method of claim 1, The micro-defective region formed in the ingot is at least 10% of the ingot radius silicon single crystal ingot manufacturing method. The method of claim 1, Existing range of microdefect areas formed in the ingot is at least 10% of the ingot length Method for producing a silicon single crystal ingot, characterized in that. The method according to any one of claims 1 to 5, Silicon initial crystal ingot manufacturing method characterized in that the initial oxygen concentration contained in the ingot is at least 11.5ppma. The method according to any one of claims 1 to 5, The initial oxygen concentration contained in the ingot is at least 11.5ppma, silicon single crystal ingot manufacturing method characterized in that the amount of oxygen precipitates at least 2.0ppma after a predetermined heat treatment.