TW202415817A - Single crystal pulling device and single crystal pulling method - Google Patents

Abstract

When growing single crystal silicon using the Czochralski method, the gap between the lower end of the radiation shield and the silicon melt is controlled with high precision to grow single crystal silicon. The single crystal pulling device comprises: a lifting drive unit for lifting and lowering the crucible; a heater for heating the crucible; a cylindrical radiation shield arranged above the silicon melt formed in the crucible and surrounding the single crystal being pulled; a light-transmitting pin member penetrating the radiation shield and arranged on a platform surface located on the front end side of the radiation shield so that the lower part is located on the silicon melt side and the upper part is located on the inner circumference side of the radiation shield; a brightness detector for detecting the brightness of the pin member; and a controller for determining that the lower end of the pin member is in contact with the silicon melt and stops the lifting of the crucible when the detected brightness of the pin member exceeds a predetermined critical value.

Description

Single crystal pulling device and single crystal pulling method

The present invention relates to a single crystal pulling device and a single crystal pulling method, and more particularly to a single crystal pulling device and a single crystal pulling method for pulling single crystal silicon using the Czochralski method.

The Czochralski method (CZ method) is used to grow single crystal silicon in the following manner: as shown in FIG. 5 , the raw material polycrystalline silicon is filled into a quartz crucible 51 disposed in a chamber 50, and the polycrystalline silicon is heated and melted into a silicon melt M by a heater 52 disposed around the aforementioned quartz crucible 51. Afterwards, the seed crystal P (seed) mounted on the seed chuck is immersed in the silicon melt M, and the seed chuck and the quartz crucible 51 are rotated in the same direction or in the opposite direction, while the seed chuck is pulled up, thereby growing a single crystal C on the inner side of the radiation shield 53.

Although the growth of single crystal silicon using this Czochralski method requires a defect-free single crystal, the conditions for producing a defect-free single crystal are extremely demanding. As an example of this production condition, there is the ratio of the pulling rate and the temperature gradient. The temperature gradient refers to the temperature change per unit length in the height direction of the single crystal near the solid-liquid interface, and the pulling rate refers to the speed of pulling the single crystal. When the pulling rate is set to V and the temperature gradient is set to G, the defects formed when producing the single crystal change sensitively according to the ratio (V/G) of the pulling rate V and the temperature gradient G. For example, if the ratio (V/G) of the pulling rate V and the temperature gradient G is too large, a vacancy type defect (void defect) is formed; on the contrary, if the ratio (V/G) of the pulling rate V and the temperature gradient G is too small, an interstitial silicon type defect (dislocation loop) is formed.

Therefore, in order to obtain a single crystal with stable quality, it is necessary to control the ratio (V/G) of the pulling speed V and the temperature gradient G with high precision. In addition, since the pulling speed V can be precisely controlled mechanically, in order to take into account both stable crystal quality and planned production, a single crystal manufacturing method is generally performed in which the pulling speed V is controlled to be fixed.

However, when only the pulling speed V is controlled to be fixed, the quality may be uneven in the direction of the crystal length. This is because the temperature gradient G is not fixed but varies during the pulling process. As a factor that has a greater impact on the control of setting the temperature gradient G to be fixed, the distance between the lower end of the radiation shield and the silicon melt surface (called the gap) is cited. It is important to control this gap to be fixed with good accuracy during the pulling of the single crystal.

In the past, there is a method for setting the gap, such as disclosed in Patent Document 1. As shown in FIG6 (a), a rod-shaped pin 60 is installed at the lower end 53a of the radiation shield 53 surrounded by the dotted circle in FIG5, and the quartz crucible 51 is raised. Then, as shown in FIG6 (b), a fusion ring 70 is detected by a CCD (Charge Coupled Device) camera or the like, and the initial gap is set based on the fusion ring 70. The fusion ring 70 is generated around the pin 60 when the pin 60 contacts the silicon melt M. [Prior Technical Document] [Patent Document]

[Patent Document 1] Japanese Patent Application Publication No. 2011-57464.

[The problem that the invention wants to solve]

However, as disclosed in Patent Document 1, since the fusion ring 70 formed when the front end of the pin 60 provided at the lower end of the radiation shield 53 contacts the silicon melt M is very small and is generated on the liquid surface, it is difficult to detect with good accuracy and sometimes an erroneous detection may occur. Therefore, there is a concern that the setting accuracy of the initial gap is low, and the control accuracy of the gap during the pulling of the single crystal becomes poor, and there is a problem of reducing the yield of high-quality defect-free single crystals.

The present invention was developed in view of the above-mentioned situation, and its purpose is to provide a single crystal pulling device and a single crystal pulling method, which can control the gap between the lower end of the radiation shield and the silicon melt with better precision when using the Czochralski method to pull single crystal silicon to cultivate single crystal silicon. [Means for solving the problem]

The single crystal pulling device of the present invention, which is developed to solve the above-mentioned problem, uses the Czochralski method to pull a single crystal from a silicon melt contained in a crucible in a chamber; the single crystal pulling device comprises: a lifting drive unit for lifting the crucible; a heater for heating the crucible; a cylindrical radiation shield disposed above the silicon melt formed in the crucible and surrounding the pulled single crystal; a light-transmissive pin member that penetrates the radiation shield and is disposed on a terrace surface (terrace) located on the front end side of the radiation shield; face) so that its lower portion is located on the silicon melt side and its upper portion is located on the inner peripheral side of the radiation shield; a brightness detector for detecting the brightness of the pin member; and a controller for controlling the lifting and driving portion; the controller causes the crucible to rise by means of the lifting and driving portion and causes the brightness detector to detect the brightness of the pin member, and when the detected brightness of the pin member exceeds a predetermined threshold value, it is determined that the lower end of the pin member is in contact with the silicon melt and the lifting of the crucible by the lifting and driving portion is stopped.

In addition, it is preferred that the controller, in a state where the lifting of the crucible by the lifting drive unit is stopped, lowers the crucible by the lifting drive unit to the following distance: the distance obtained by subtracting the known length from the lower end of the radiation shield in the pin member to the front end of the pin member from the expected value of the gap from the lower end of the radiation shield in the pin member to the front end of the pin member. In addition, it is preferred that the predetermined critical value of the brightness is a value obtained by adding a value of 25% or more of the standard brightness to the standard brightness when the pin member is not in contact with the silicon melt. Or, it is preferred that the predetermined critical value of the brightness is a value obtained by adding a value of 50% or more of the standard brightness to the standard brightness when the pin member is not in contact with the silicon melt.

According to this structure, the crucible is raised during the setting of the initial gap between the lower end of the radiation shield and the silicon melt surface before the single crystal is pulled; when the brightness of the pin member provided at the lower end of the radiation shield and the silicon melt changes significantly when in contact, the rise of the crucible is stopped. Among them, since the brightness is detected not on the surface of the silicon melt but on the brightness of the pin member, there is less concern about erroneous detection and a more accurate detection result is obtained. In addition, the crucible is lowered to a distance obtained by subtracting the known length of the pin member from the desired gap value, thereby enabling the initial gap to be set with better accuracy. As a result, the gap control in the subsequent single crystal pulling process can be performed with high precision, and high-quality defect-free single crystals can be pulled.

In addition, the single crystal pulling method of the present invention, which was developed to solve the above-mentioned problem, utilizes the Czochralski method to pull a single crystal from a silicon melt contained in a crucible in a cavity; the above-mentioned single crystal pulling method comprises the following steps: before pulling the single crystal, the above-mentioned crucible is raised and the brightness of a light-transmitting pin member is detected, the above-mentioned pin member extends downward from the lower end of a cylindrical radiation shield, and the above-mentioned radiation shield is arranged above the above-mentioned silicon melt formed in the above-mentioned crucible and surrounds the pulled single crystal; and when the above-mentioned brightness exceeds a predetermined critical value, it is determined that the lower end of the above-mentioned pin member is in contact with the above-mentioned silicon melt and the raising of the above-mentioned crucible is stopped.

In addition, it is preferred that after the step of determining that the lower end of the pin member is in contact with the silicon melt when the brightness exceeds a predetermined threshold value and stopping the ascent of the crucible, the following step is provided: lowering the crucible to the following distance: the distance obtained by subtracting the known length from the lower end of the radiation shield of the pin member to the front end of the pin member from the expected value of the gap from the lower end of the radiation shield of the pin member to the front end of the pin member. In addition, it is preferred that the predetermined threshold value of the brightness is the value obtained by adding a value of 25% or more of the standard brightness to the reference brightness when the pin member is not in contact with the silicon melt. Alternatively, preferably, the predetermined threshold value of the brightness is a value obtained by adding a value greater than 50% of the reference brightness to the reference brightness when the pin member is not in contact with the silicon melt.

According to this structure, the crucible is raised during the setting of the initial gap between the lower end of the radiation shield and the silicon melt surface before the single crystal is pulled; when the brightness of the pin member provided at the lower end of the radiation shield and the silicon melt changes significantly when the pin member is in contact with the silicon melt, the rise of the crucible is stopped. Among them, since the brightness is detected not on the surface of the silicon melt but on the brightness of the pin member, there is less concern about erroneous detection and a more accurate detection result is obtained. In addition, the crucible is lowered to a distance obtained by subtracting the known length of the pin member from the desired gap value, thereby enabling the initial gap to be set with better accuracy. As a result, the gap control in the subsequent single crystal pulling process can be performed with high precision, and a high-quality, defect-free single crystal can be pulled. [Effect of the invention]

According to the present invention, when using the Czochralski method to grow single crystal silicon, the gap between the lower end of the radiation shield and the silicon melt can be controlled with better precision to grow single crystal silicon.

Hereinafter, the single crystal pulling device and the single crystal pulling method of the present invention will be described using drawings. However, this embodiment is described as an example of the present invention, and the present invention is not limited thereto.

FIG1 is a cross-sectional view showing an example of a single crystal pulling device of the present invention. The single crystal pulling device 1 comprises a furnace body 10 formed by stacking a pull chamber 10b on a cylindrical main chamber 10a; the furnace body 10 comprises: a carbon crucible (or graphite crucible) 2 which is arranged to be rotatable around a lead straight axis and to be able to be raised and lowered; and a quartz glass crucible 3 (hereinafter referred to as crucible 3) which is held by the carbon crucible 2. The crucible 3 can rotate around the lead straight axis as the carbon crucible 2 rotates.

In addition, below the carbon crucible 20, there are provided: a rotation drive unit 14, which is a rotation motor, etc., for rotating the carbon crucible 2 around the lead straight axis; and a lifting drive unit 15, which is for lifting and moving the carbon crucible 2. In addition, the rotation drive control unit 14a is connected to the rotation drive unit 14, and the lifting drive control unit 15a is connected to the lifting drive unit 15.

In addition, the single crystal pulling device 1 is equipped with a heater 4 of a resistance heating type or a high-frequency induction heating type, which is used to heat the semiconductor raw material (raw material polycrystalline silicon) loaded on the crucible 3 to melt and form a silicon melt M. In addition, the single crystal pulling device 1 is equipped with a pulling mechanism 9, which winds up the wire 6 and pulls the cultivated single crystal C. A seed crystal P is installed at the front end of the wire 6 of the pulling mechanism 9. A rotation drive control unit 9a for controlling the rotation drive of the pulling mechanism 9 is connected to the pulling mechanism 9.

In addition, a radiation shield 7 surrounding the single crystal C is arranged above the silicon melt M formed in the crucible 3. The radiation shield 7 has openings formed at the upper and lower parts, and is used to shield the excess radiation heat from the heater 4, the silicon melt M, etc. to the single crystal C being grown and to rectify the gas flow in the furnace. The shape of the radiation shield 7 is described in detail. As shown in FIG. 2, the radiation shield 7 has: a tapered portion 7b that tapers from the upper opening 7a toward the silicon melt M (downward); and a platform portion 7c that extends horizontally inward from the lower end of the tapered portion 7b, and the platform portion 7c is formed in a ring shape. The single crystal is pulled through the lower opening 7d and the upper opening 7a formed on the inner edge of the terrace portion 7c.

The radial length (width) of the platform portion 7c is formed to be, for example, more than 10 mm and less than 150 mm, so as to prevent the reflection or stray light of the silicon melt M from reaching the vicinity of the lower end of the conical portion 7b. As shown in the figure, a light-transmissive quartz pin 8 (pin member) extending downward is provided on this platform portion 7c. That is, this quartz pin 8 penetrates the radiation shield 7 and is arranged on the platform portion 7c (platform surface) located on the front end side of the radiation shield 7, so that the lower part is located on the silicon melt M side and the upper part is located on the inner circumference of the radiation shield 7. This quartz pin 8 is composed of, for example, a straight tube 8a and a disk-shaped head 8b having a larger diameter than the straight tube 8a, and is designed by inserting the straight tube 8a from the top to the bottom to form a hole 7c1 in the platform 7c that matches the diameter of the straight tube 8a and locking the head 8b to the platform 7c. In this quartz pin 8, the length L from the lower end of the radiation shield 7 to the front end 8a1 of the quartz pin 8 is formed to be, for example, 15 mm.

In addition, the single crystal pulling device 1 is equipped with an optical diameter measuring sensor (diameter measuring device) 16, which is a CCD camera, etc., for measuring the diameter of the single crystal being grown. A small window 10a1 for observation is provided on the upper surface of the main chamber 10a, and the position change of the crystallization end (the position shown by the dotted arrow) of the solid-liquid interface is detected from the outside of the small window 10a1.

In addition, the single crystal pulling device 1 is provided with a brightness detector 17, which is a CCD camera or the like, for measuring the brightness of the head 8b of the quartz pin 8 provided at the lower end of the radiation shield 7. As shown in FIG2 , the brightness detector 17 is configured to measure the brightness of the head 8b of the quartz pin 8 provided on the upper surface of the platform portion 7c of the radiation shield 7. A small window 10a2 different from the small window 10a1 is provided on the upper surface of the main chamber 10a, and the brightness change of the head 8b of the quartz pin 8 is detected from the outside of the small window 10a2.

In this embodiment, the brightness change of the head 8b of the quartz pin 8 is detected in this way in order to set the initial gap between the lower end of the radiation shield 7 and the silicon melt surface M1 with high precision. The setting method of this initial gap is described in detail below. When the front end 8a1 of the quartz pin 8 contacts the silicon melt M, light passes through the quartz pin 8, and the head 8b of the quartz pin 8 emits light with higher brightness. The brightness detector 17 is used to detect this brightness change. In addition, the head 8b of the quartz pin 8 emits light with high brightness based on the following principle. The silicon melt M is orange and brighter than the surrounding environment in the furnace. When the transparent quartz pin 8 contacts the silicon melt M, a ring is formed around the quartz pin 8 due to surface tension, and receives more visible light emitted by the silicon melt M than before the contact. Since the quartz pin 8 is transparent, the visible light is hardly reflected or absorbed and is transmitted to the head 8b of the quartz pin 8, thereby emitting high brightness.

In addition, the single crystal pulling device 1 is equipped with: a controller 11 having a memory device 11a and an operation control device 11b; and a rotation drive control unit 14a, a lifting drive control unit 15a, a rotation drive control unit 9a, a diameter measuring sensor 16, and a brightness detector 17 are respectively connected to the operation control device 11b.

In the single crystal pulling device 1 thus constructed, if, for example, a single crystal C with a diameter of 300 mm is to be grown, the pulling is performed in the following manner. That is, firstly, the raw material polycrystalline silicon (for example, 460 kg) is loaded on the crucible 3, and the crystal growth process is started according to the program stored in the memory device 11a of the controller 11.

First, a predetermined atmosphere (mainly inert gas such as argon) is formed in the furnace body 10. For example, a furnace atmosphere with a furnace internal pressure of 65 torr and an argon flow rate of 90 l/min is formed. Then, while the crucible 3 rotates in a predetermined direction at a predetermined speed (rpm), the raw material polycrystalline silicon loaded in the crucible 3 is melted by heating of the heater 4 and becomes silicon melt M (step S1 in Figure 3).

In addition, the initial power supplied to the heater 4, the pulling speed, etc. are used as parameters to adjust the pulling conditions, and the seed crystal P starts to rotate around the axis at a predetermined speed. The rotation direction is opposite to the rotation direction of the crucible 3. When the liquid level of the silicon melt M is stable, under the control of the controller 11, the lifting drive control unit 15a drives the lifting drive unit 15 to increase the height of the crucible 3 by an inching action at intervals of, for example, 0.05 mm (step S2 of Figure 3). In addition, during this period, the controller 11 uses the brightness detector 17 to monitor the brightness change of the head 8b of the quartz pin 8 (step S3 of Figure 3). In addition, in this embodiment, the controller 11 pre-memorizes the maximum brightness of the quartz pin 8 after contacting the silicon melt M as 100%. In addition, the brightness of the quartz pin 8 before it contacts the silicon melt M (set as the reference brightness) is, for example, 40%.

When the silicon melt surface M1 of the silicon melt M contacts the front end 8a1 of the quartz pin 8 at the lower end of the radiation shield 7 due to the rise of the crucible 3, the brightness of the head 8b of the quartz pin 8 rises to a predetermined threshold value or more (step S4 in FIG. 3 ). The controller 11 detects the brightness rise by the brightness detector 17 and controls the lifting drive 15 to stop the rising action of the crucible 3 (step S5 in FIG. 3 ). Specifically, the controller 11 determines that the quartz pin 8 at the lower end of the radiation shield 7 contacts the silicon melt M when the brightness change detected by the brightness detector 17 exceeds the predetermined threshold value. For example, the predetermined threshold value of the brightness refers to the value obtained by adding, for example, 25% or more (or 50% or more) of the reference brightness (here 40%) when the quartz pin 8 is not in contact with the silicon melt M when the maximum brightness after the quartz pin 8 is in contact with the silicon melt M is 100%.

The controller 11 lowers the crucible 3 by the lifting drive 15 to a distance obtained by subtracting the length L of the quartz pin 8 from the desired gap value (step S6 in FIG. 3 ). In this way, the initial gap value is set with high precision. In addition, since the height of the fixed radiation shield 7 is known, the height position (initial liquid level) of the silicon melt liquid surface M1 after setting the gap can be easily calculated. The controller 11 memorizes this initial liquid level in advance (step S7 in FIG. 3 ).

Next, the wire 6 is lowered and the seed crystal P is brought into contact with the silicon melt M, and after the front end of the seed crystal P is dissolved, necking is performed to form a neck (step S8 in FIG. 3 ). Then, the crystal diameter gradually expands to form a shoulder C1 (step S9 in FIG. 3 ). In addition, the controller 11 drives and controls the lifting drive unit 15 to fix the pulling speed at, for example, 0.55 mm/min through the lifting drive control unit 15a, and moves to the process of forming the straight tube portion C2 that becomes the product part (step S10 in FIG. 3 ).

During the formation of this straight tube portion C2, the controller 11 uses the measurement result of the diameter measuring sensor 16 to obtain the solidification rate of the single crystal silicon, and calculates the height position and gap of the silicon melt surface M1 at this time point based on this solidification rate and the initial height of the silicon melt surface M1 (step S11 in Figure 3). Then, it is determined whether the change in the measured dimension of the gap is within ±0.1mm (step S12 in Figure 3). If it is not satisfied, the lifting drive control unit 15a is controlled to adjust the height position of the crucible 3 by the lifting drive unit 15 to meet the condition (step S13 in Figure 3). During the pulling process, the control is to meet the condition of step S12, and the single crystal pulling process is performed.

After the straight tube portion C2 is formed to a predetermined length (step S14 in FIG. 3 ), the process proceeds to the final tailing process (step S15 in FIG. 3 ). In this tailing process, the contact area between the lower end of the crystal and the silicon melt M is gradually reduced, and the single crystal C is separated from the silicon melt M to produce single crystal silicon.

As described above, according to the present embodiment, in setting the initial gap between the lower end of the radiation shield 7 and the silicon melt surface M1 before the single crystal is pulled, the crucible 3 is raised by a micro-motion; when the brightness of the quartz pin 8 provided at the lower end of the radiation shield 7 and the silicon melt M changes significantly, the rise of the crucible 3 is stopped. Among them, since the brightness is detected by detecting the brightness of the quartz pin 8 instead of the brightness on the silicon melt surface M1, there is less concern about erroneous detection, and a more accurate detection result is obtained. Then, the crucible 3 is lowered to a distance obtained by subtracting the known length L of the quartz pin 8 from the desired gap value, thereby setting the initial gap with better accuracy. As a result, the gap control in the subsequent single crystal pulling process can be performed with high precision, and high-quality single crystals can be pulled.

In addition, in the above-mentioned embodiment, in the gap setting during the single crystal pulling period, the height position detection on the silicon melt surface M1 is calculated based on the solidification rate of the single crystal, but the present invention is not limited to this. For example, as disclosed in Japanese Patent Publication No. 2008-189522, a CCD camera can be used to detect the reflection of the laser light irradiated to the silicon melt surface M1, and the detected image signal can be processed to obtain the liquid level height of the silicon melt M. In addition, in the above-mentioned embodiment, the pulling speed is set to be fixed, but the pulling speed may not be set to be fixed and may vary.

In addition, in the above-mentioned embodiment, a quartz pin 8 is provided at the lower end of the radiation shield 7 as an example. However, in addition to the quartz pin 8, a quartz pin shorter than the quartz pin 8 can also be provided at the lower end of the radiation shield 7. At this time, in the control during the single crystal pulling period, although the crucible 3 is controlled to rise as the silicon melt M decreases, when the shorter pin contacts the silicon melt M, the crucible 3 can also be stopped from rising or the crucible 3 can be controlled to fall, so as to prevent the main body of the radiation shield 7 from contacting the silicon melt M.

In addition, in the above-mentioned embodiment, the quartz pin 8 is composed of a straight tube portion 8a with a fixed diameter and a disc-shaped head portion 8b with a larger diameter than the straight tube portion 8a, but the present invention is not limited to this form. For example, the diameter of the lower portion can be smaller than the diameter of the upper portion (the diameter of the upper portion is larger than the diameter of the lower portion). That is, in the quartz pin 8, when the surface area per length of the head portion 8b above the radiation shield 7 (the second area above the radiation shield 7) is larger than that of the straight tube portion 8a below the radiation shield 7 (the first area below the radiation shield 7), it is easier to detect the brightness using the brightness detector 17 (if the surface area of the head portion 8b is too small, it is more susceptible to disturbances, etc., resulting in reduced measurement accuracy). [Example]

The single crystal pulling device and the single crystal pulling method of the present invention are further described according to the embodiments.

[Experiment 1] In Experiment 1, 460 kg of silicon raw material was filled into a quartz crucible with a diameter of 32 inches in the single crystal pulling device shown in Figure 1 to form a silicon melt. After the silicon melt was melted in the crucible, the crucible was raised to verify whether the brightness changed when the silicon melt came into contact with the quartz pin at the lower end of the radiation shield. The crucible was raised in micro-movements of 0.05 mm and the brightness at that time was detected.

The results of this experiment 1 are shown in Figure 4. In the graph of Figure 4, the left vertical axis is the crucible displacement (mm) relative to the reference value (0mm), and the right vertical axis is the brightness (%). In addition, the brightness (%) on the vertical axis is the maximum brightness after the quartz pin contacts the silicon melt as 100%. As shown in the graph of Figure 4, the brightness increases significantly at the position where the crucible is raised by +0.4mm. By visual inspection, this position is the position where the silicon melt contacts the quartz pin located at the lower end of the radiation shield. That is, it is confirmed that the brightness increases significantly when the silicon melt contacts the quartz pin located at the lower end of the radiation shield. In addition, the graph in FIG4 shows that the critical value used by the controller to judge whether the quartz pin is in contact with the silicon melt is a brightness of 50% or more (the value obtained by adding 25% of the reference brightness (40%) to the reference brightness before the quartz pin is in contact with the silicon melt), which is almost impossible to cause an erroneous judgment due to the pickup disturbance. More specifically, it is preferably 60% or more (the value obtained by adding 50% of the reference brightness (40%) to the reference brightness before the quartz pin is in contact with the silicon melt).

[Experiment 2] [Example 1] In Experiment 2, ten single crystal silicons were pulled. In Example 1, 460 kg of silicon raw material was filled into a quartz crucible with a diameter of 32 inches and a silicon melt was formed as in Experiment 1. According to this embodiment, a quartz pin was set at the lower end of the radiation shield and the initial gap between the radiation shield and the silicon melt surface was set. This gap was set to 50 mm. In addition, the furnace internal pressure was set to 50 torr, and argon gas was flowed in at a flow rate of 100 L/min to create a furnace internal environment. Then, the crucible speed was set to 1rpm, the crystallization speed was set to 7rpm (in the opposite direction of the crucible rotation), and the single crystal was grown at a pulling speed of 0.6mm/min with a crystal diameter of 305mm to 310mm as the target. In addition, the gap between the radiation shield and the silicon melt surface was controlled to 50mm±0.1mm during the pulling process.

In this embodiment 1, ten single crystals were pulled to be defect-free. As an evaluation of the defect-free area, the V-rich side where the void defect exists is confirmed to be a cluster of fifty overlay images of LPD (Light Point Defect) evaluation to determine whether there is COP (Crystal Originated Particle).

The I-rich side is judged by the Cu (copper) decoration method to determine whether there is a B-band region, which is a region where oxygen precipitates are easily generated. As a defect-free crystal, it is defined as a crystal without COP and B-band region. The results of Example 1 are shown in Table 1. As shown in Table 1, the results of evaluating ten single crystals confirmed that all ten single crystals were defect-free crystals throughout the entire length.

[Comparative Example 1] In Comparative Example 1, a gap combination method using a mirror gap was used to attempt to pull a defect-free crystal. The mirror gap uses the same CCD camera as the one that detects the brightness of the quartz pin to detect the edge of the shield of the real image of the radiation shield and the mirror image reflected on the silicon melt surface, and the gap is calculated from the difference. The results of Comparative Example 1 are shown in Table 1 together with the results of Example 1. As a result of pulling ten single crystals, two single crystals were detected to have inter-lattice defects (I-defect; Inter-Grid Type Defect) throughout the entire length by wafer evaluation using the Cu decoration method. In addition, cluster patterns such as DSOD (Direct Surface Oxide Defect; direct surface oxide defect) or LPD were observed in part of the two crystals.

[Table 1] Number of crystal roots Gap Unevenness(σ) Defect-free crystal yield (average) Embodiment 1 10 0.03 100% Comparison Example 1 10 0.60 72%

The results of Example 1 confirmed that the gap unevenness became smaller and the defect-free crystal yield was higher according to the present invention.

1: Single crystal pulling device 2: Carbon crucible (graphite crucible) 3: Quartz glass crucible (crucible) 4,52: Heater 6: Wire 7,53: Radiation shield 7a: Upper opening 7b: Cone 7c: Platform 7c1: Hole 7d: Lower opening 8: Quartz pin 8a,C2: Straight tube 8a1: Front end 8b: Head 9: Pulling mechanism 9a: Rotation drive control unit 10: Furnace body 10a: Main chamber 10a1,10a2: Small window 10b: Pulling chamber 11: Controller 11a: Memory device 11b: Operation control device 14: Rotation drive unit 14a: Rotation drive control unit 15: Lifting drive unit 15a: Lifting drive control unit 16: Diameter measuring sensor (diameter measuring device) 17: Brightness detector 50: Cavity 51: Quartz crucible 53a: Lower end 60: Pin 70: Fusion ring C: Single crystal C1: Shoulder L: Length M: Silicon melt M1: Silicon melt surface P: Seed crystal S1~S15: Steps

[FIG. 1] is a cross-sectional view showing an example of a single crystal pulling device of the present invention. [FIG. 2] is a schematic cross-sectional view showing a portion of the single crystal pulling device of FIG. 1 in an enlarged manner. [FIG. 3] is a flow chart showing an example of a single crystal pulling method of the present invention. [FIG. 4] is a graph showing the results of an embodiment. [FIG. 5] is a cross-sectional view of a conventional single crystal pulling device. [FIG. 6], (a) and (b) are partial enlarged views of FIG. 5, used to illustrate the fusion ring generated when the pin provided at the lower end of the radiation shield contacts the silicon melt.

1: Single crystal pulling device

2: Carbon crucible (graphite crucible)

3: Quartz glass crucible (crucible)

4: Heater

6: Line

7: Radiation shielding

7b: Cone

7c: Platform Department

8: Quartz pins

9: Lifting mechanism

9a: Rotation drive control unit

10: Furnace body

10a: Main cavity

10a1,10a2: small window

10b: Pulling cavity

11: Controller

11a: Memory device

11b: Computational control device

14: Rotary drive unit

14a: Rotation drive control unit

15: Lifting drive unit

15a: Lifting drive control unit

16: Diameter measuring sensor (diameter measuring device)

17: Brightness tester

C: Single crystal

C1: Shoulder

C2: Straight section

M: Silicon melt

M1: Silicon melt surface

P: Seed crystal

Claims (8)

A single crystal pulling device uses the Czochralski method to pull a single crystal from a silicon melt contained in a crucible in a chamber; The single crystal pulling device comprises: A lifting drive unit for lifting and lowering the crucible; A heater for heating the crucible; A cylindrical radiation shield is arranged above the silicon melt formed in the crucible and surrounds the pulled single crystal; A light-transmissive pin member is passed through the radiation shield and arranged on a platform surface located on the front end side of the radiation shield so that the lower part is located on the silicon melt side and the upper part is located on the inner circumference of the radiation shield; A brightness detector for detecting the brightness of the pin member; and A controller for controlling the lifting drive unit; The controller causes the crucible to rise by the lifting drive unit and causes the brightness detector to detect the brightness of the pin member. When the detected brightness of the pin member exceeds a predetermined threshold value, it is determined that the lower end of the pin member contacts the silicon melt and the lifting of the crucible by the lifting drive unit is stopped. A single crystal pulling device as described in claim 1, wherein the controller, in a state where the lifting of the crucible by the lifting drive unit is stopped, causes the crucible to be lowered to the following distance by the lifting drive unit: a distance obtained by subtracting a known length from the lower end of the radiation shield in the pin member to the front end of the pin member from the expected value of the gap from the lower end of the radiation shield in the pin member. A single crystal pulling device as recited in claim 1, wherein the predetermined threshold value of the brightness is a value obtained by adding a value of 25% or more of the baseline brightness to a baseline brightness when the pin member is not in contact with the silicon melt. A single crystal pulling device as recited in claim 1, wherein the predetermined threshold value of the brightness is a value obtained by adding a value of 50% or more of the baseline brightness to a baseline brightness when the pin member is not in contact with the silicon melt. A single crystal pulling method is to pull a single crystal from a silicon melt contained in a crucible in a chamber by using the Czochralski method; The single crystal pulling method comprises the following steps: Before pulling the single crystal, the crucible is raised and the brightness of a light-transmitting pin member is detected, the pin member extends downward from the lower end of a cylindrical radiation shield, and the radiation shield is arranged above the silicon melt formed in the crucible and surrounds the pulled single crystal; and When the brightness exceeds a predetermined threshold value, it is determined that the lower end of the pin member contacts the silicon melt and the lifting of the crucible is stopped. The single crystal pulling method described in claim 5, wherein after the step of judging that the lower end of the pin member contacts the silicon melt when the brightness exceeds a predetermined threshold value and stopping the ascent of the crucible, the following step is included: The crucible is lowered to the following distance: the distance obtained by subtracting the known length from the lower end of the radiation shield of the pin member to the front end of the pin member from the expected value of the gap from the lower end of the radiation shield to the silicon melt. A single crystal pulling method as described in claim 5, wherein the predetermined threshold value of the brightness is a value obtained by adding a value of 25% or more of the baseline brightness to a value when the pin member is not in contact with the silicon melt. A single crystal pulling method as described in claim 5, wherein the predetermined threshold value of the brightness is a value obtained by adding a value of 50% or more of the baseline brightness to a value when the pin member is not in contact with the silicon melt.