Method for producing silicon single crystals
By alternately forming reduced and expanded diameter portions with specific angle settings during the necking process, the method efficiently removes both slip and axial dislocations, addressing the limitations of conventional silicon single crystal manufacturing techniques.
Patent Information
- Application Number
- JP2021175635
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-10-27
- Publication Date
- 2025-05-16
- Estimated Expiration
- 2041-10-27
AI Technical Summary
Conventional methods for manufacturing silicon single crystals struggle to efficiently remove both slip dislocations and axial dislocations during the necking process, leading to incomplete dislocation removal.
The method involves alternately forming reduced and expanded diameter portions during the neck portion growth, maintaining a vertically upward concave solid-liquid interface shape, with specific angle settings for the reduced (80°-90°) and expanded (105°-140°) diameter portions to enhance dislocation removal.
This approach effectively removes both slip and axial dislocations by increasing thermal stress for slip dislocations and directing axial dislocations towards the crystal periphery, resulting in improved crystal quality.
Smart Images

Figure 0007678739000002 
Figure 0007678739000003 
Figure 0007678739000004
Abstract
Description
[Technical field]
[0001] The present invention relates to a method for producing a silicon single crystal by pulling a silicon single crystal by the Czochralski method (CZ method). [Background technology]
[0002] Conventionally, in the manufacture of silicon single crystals by the Czochralski method, a seed crystal is brought into contact with silicon melt and slowly pulled up while rotating to manufacture a silicon single crystal. In this process, after the seed crystal is brought into contact with the silicon melt, the diameter of the crystal is once narrowed to form a narrowed portion, a process known as seed necking. In this necking process, slip dislocations that occur at high density in the seed crystal due to thermal shock are eliminated by forming a narrowed portion in which the crystal diameter is narrowed. This type of necking is widely known as the Dash necking process.
[0003] As a method for producing a silicon single crystal by eliminating slip dislocations in the above-mentioned necking step, the following patent documents are disclosed.
[0004] For example, Patent Document 1 describes that after a seed crystal is brought into contact with a silicon melt, a crystal diameter d0 is reduced from the seed crystal to form a reduced diameter portion with a minimum diameter d1 of 5 mm or more directly below the seed crystal, and then the crystal diameter is expanded from the minimum diameter d1 to form an expanded diameter portion with a maximum diameter d2 smaller than the crystal diameter d0 of the seed crystal, and then the reduced diameter portion and the expanded diameter portion are alternately repeated in this order to form a neck portion. It also describes that the reduction rate of the reduced diameter portion is 0.6 mm / mm or more and 1.5 mm / mm or less, and the expansion rate of the expanded diameter portion is 0.1 mm / mm or more and 1.0 mm / mm or less. It also describes that the maximum diameter d2 of the expanded diameter portion is 1 mm or more and 2 mm or less larger than the minimum diameter d1 of the reduced diameter portion.
[0005] Patent Document 2 also describes a seed drawing method in which a drawn portion and an expanded diameter portion are alternately formed, and the minimum diameter of the drawn portion is set to 5 mm or more. It also describes that in forming the drawn portion and the expanded diameter portion, the amount of diameter change per unit pulling length is set to 0.5 mm / mm or more. It also describes that the ratio of the maximum diameter of the expanded diameter portion to the minimum diameter of the drawn portion is set to 2 times or more and 15 times or less.
[0006] Furthermore, the following Patent Document discloses annihilation of axial dislocations in addition to slip dislocations.
[0007] For example, Patent Document 3 describes that in order to remove slip dislocations, a necking operation is performed to narrow the neck portion to form a first neck portion and a second neck portion, and that in order to remove axial dislocations, the solid-liquid interface where the first neck portion contacts the silicon melt is temporarily changed from a vertically upward convex solid-liquid interface shape to a vertically upward concave solid-liquid interface shape.
[0008] Patent Document 4 also describes that the necking process is divided into two steps, and that the main slip dislocations are removed in the first step, and axial dislocations are removed in the second step, thereby eliminating dislocations in the neck portion. Specifically, in the first step, the solid-liquid interface shape is made vertically upward convex to reduce the crystal diameter, thereby eliminating slip dislocations. In the second step, the solid-liquid interface shape is controlled to be vertically upward concave, and the pulling conditions are controlled so that the solid-liquid interface angle is larger than the crystal diameter increase angle. It also describes that the solid-liquid interface angle is preferably 4 to 8°, and in this case, the crystal diameter increase angle is an angle not exceeding 4°.
[0009] Patent Document 5 also describes a method in which, after forming a necked portion, diameter increasing and decreasing portions are repeatedly formed in the process of forming a neck portion to increase and decrease the neck portion diameter, and the increase and decrease in the neck portion diameter are performed in the final stage of the process of forming the neck portion, thereby removing all dislocations including axial dislocations. It also describes that the increase or decrease in the neck portion diameter due to the increase or decrease in diameter is limited to within 1 mm. [Prior art documents] [Patent documents]
[0010] [Patent Document 1] JP 2011-153034 A [Patent Document 2] Japanese Patent Application Publication No. 11-199384 [Patent Document 3] JP 2011-57460 A [Patent Document 4] JP 2016-183072 A [Patent Document 5] JP 2009-263142 A Summary of the Invention [Problem to be solved by the invention]
[0011] Although the above Patent Documents 1 and 2 describe the elimination of slip dislocations, it is difficult to remove axial dislocations that propagate in a direction perpendicular to the solid-liquid interface, and there is a problem in that dislocations cannot be eliminated.
[0012] In addition, in the silicon single crystal growth method described in Patent Document 3, since the solid-liquid interface shape, which becomes a vertically upward convex shape when the first neck portion is formed, needs to be changed to a vertically upward concave shape once, it takes time for the solid-liquid interface shape to become the predetermined vertically upward concave shape. In addition, since the solid-liquid interface becomes a vertically upward convex shape when the second neck portion is formed, there is a problem that axial dislocations are pulled back to the center before reaching the crystal outer periphery and disappearing, and dislocations cannot be eliminated.
[0013] In addition, in the necking described in Patent Document 4, the crystal is pulled up so that the crystal diameter increase angle does not exceed 4°, so it takes time to expand the diameter. In addition, there is little change in the pulling speed when moving from the first step to the second step, so there is little change in the solid-liquid interface, and there is a problem that axial dislocations near the center of the neck are difficult to move and remain.
[0014] Furthermore, in the silicon single crystal growth method of Patent Document 5, when the neck diameter is reduced, the solid-liquid interface becomes a vertically upward convex shape, and as with Patent Document 3, there is a problem that axial dislocations are pulled back to the center before reaching the outer periphery of the crystal and disappearing, making it impossible to eliminate dislocations.
[0015] That is, in the conventional method for producing a silicon single crystal, both slip dislocations and axial dislocations cannot be efficiently and sufficiently removed, and there is therefore room for improvement.
[0016] The present invention has been made in consideration of the above-mentioned problems, and has an object to provide a method for producing a silicon single crystal that can efficiently remove both slip dislocations and axial dislocations in the process of forming a neck portion. [Means for solving the problem]
[0017] The method for producing a silicon single crystal according to the present invention is a method for producing a silicon single crystal in which a reduced diameter portion and an expanded diameter portion are alternately and repeatedly formed during a neck portion growth process, and is characterized in that, while maintaining a shape of a solid-liquid interface between the neck portion and the silicon melt as a vertically upward concave shape, a first angle, which is an angle between a horizontal plane and an outer peripheral surface of the reduced diameter portion, is set to be 80° or more and less than 90°, and the reduced diameter portion is formed at this first angle, and further, a second angle, which is an angle between a horizontal plane and an outer peripheral surface of the expanded diameter portion, is set to be 105° or more and less than 140°, and the expanded diameter portion is formed at the second angle which is steeper than that at which the reduced diameter portion was formed.
[0018] The formation of the reduced diameter portion increases thermal stress in the neck portion, and slip dislocations propagating toward the outer periphery of the crystal are removed by the reduced crystal diameter, thereby reducing the downward propagation of slip dislocations. The formation of the expanded diameter portion reduces thermal stress in the neck portion, making it difficult for slip dislocations propagating toward the center of the neck portion to move, thereby reducing the downward propagation of slip dislocations propagating toward the center of the reduced diameter portion. Therefore, by alternately and repeatedly forming reduced diameter portions and expanded diameter portions, slip dislocations can be eliminated.
[0019] In addition, by maintaining the solid-liquid interface shape as a vertically upward concave shape, setting the first angle between the horizontal plane and the outer peripheral surface of the reduced diameter portion to 80° or more and less than 90°, and setting the second angle between the horizontal plane and the outer peripheral surface of the expanded diameter portion to 105° or more and less than 140° (an angle steeper than that when the reduced diameter portion is formed), the height (depth) of the vertically upward concave shape can be made higher (deeper). In addition, since the axial dislocations are propagated in a direction perpendicular to the solid-liquid interface, the axial dislocations can be propagated further toward the outer peripheral side of the silicon single crystal. As a result, the axial dislocations can be easily removed toward the outer peripheral side, and therefore the axial dislocations can be eliminated.
[0020] That is, according to the present invention, in the process of forming the neck portion, the solid-liquid interface shape is always maintained as a vertically upward concave shape, while a reduced diameter portion in which the crystal diameter is reduced at a gradual angle and an expanded diameter portion in which the crystal diameter is expanded at a steep angle are alternately formed repeatedly, so that both slip dislocations and axial dislocations can be efficiently removed.
[0021] Furthermore, in the method for producing a silicon single crystal according to the present invention, it is desirable that the maximum diameter of the expanded diameter portion is at least 3 mm larger than the minimum diameter of the contracted diameter portion, and that the value of "first length / second length", which is the ratio between the first length, which is the vertical length of the contracted diameter portion, and the second length, which is the vertical length of the expanded diameter portion, is 2.5 or more.
[0022] Furthermore, in the method for producing a silicon single crystal according to the present invention, it is preferable that the first length is in the range of 9 mm to 30 mm, and the second length is in the range of 3 mm to 8 mm. Effect of the Invention
[0023] According to the present invention, it is possible to provide a method for producing a silicon single crystal that can efficiently remove both slip dislocations and axial dislocations in the step of forming a neck portion. [Brief description of the drawings]
[0024] [Figure 1] FIG. 1 is a schematic diagram showing a cross section of a single crystal pulling apparatus capable of carrying out the method for producing a silicon single crystal according to the present invention. [Diagram 2] FIG. 2 is a flowchart showing the flow of the method for producing a silicon single crystal. [Diagram 3] FIG. 3 is a schematic diagram showing an image of the neck portion. [Figure 4] FIG. 4 is a diagram showing the external appearance of the neck portion. [Diagram 5] FIG. 5 is an image diagram showing an example of a case where an axial dislocation remains. [Figure 6] FIG. 6 is an image diagram showing an example of an axial dislocation passing through to the outer periphery of a silicon single crystal. [Figure 7-1] FIG. 7-1 is a diagram showing an example of a SIRD map. [Figure 7-2] FIG. 7-2 is a diagram showing an example of a SIRD map. [Figure 7-3] FIG. 7-3 is a diagram showing an example of a SIRD map. [Figure 7-4] FIG. 7-4 is a diagram showing an example of a SIRD map. [Figure 7-5] FIG. 7-5 is a diagram showing an example of a SIRD map. [Figure 8] FIG. 8 is a diagram showing the shape of the solid-liquid interface as evaluated by vertical splitting of the neck portion using an X-ray topograph. [Figure 9] FIG. 9 is a diagram showing the observation results of the shape of the solid-liquid interface shown in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0025] Hereinafter, an embodiment of the method for producing a silicon single crystal according to the present invention will be described in detail, however, the present invention is not limited to this embodiment.
[0026] <Configuration of single crystal pulling device> FIG. 1 is a schematic diagram showing a cross section of a single crystal pulling apparatus capable of carrying out the method for producing a silicon single crystal according to the present invention.
[0027] In FIG. 1, the single crystal pulling apparatus 1 includes a furnace body 10 formed by stacking a pull chamber 10b on top of a cylindrical main chamber 10a, and includes within this furnace body 10 a carbon susceptor (or graphite susceptor) 2 that is rotatable about a vertical axis and movable up and down, and a quartz glass crucible 3 held by the carbon susceptor 2.
[0028] The quartz glass crucible 3 has a straight body portion 3a and a bottom portion 3b formed thereunder, and is rotatable about a vertical axis together with the rotation of the carbon susceptor 2. In addition, below the carbon susceptor 2, there are provided a rotation drive unit 14 such as a rotary motor that rotates the carbon susceptor 2 about a vertical axis, and an elevation drive unit 15 that moves the carbon susceptor 2 up and down. A rotation drive control unit 14a is connected to the rotation drive unit 14, and an elevation drive control unit 15a is connected to the elevation drive unit 15.
[0029] The single crystal pulling apparatus 1 also includes a side heater 4 that uses resistance heating to melt the semiconductor raw material (raw polysilicon) loaded in the quartz glass crucible 3 to produce silicon melt M, and a pulling mechanism 9 that pulls up the grown silicon single crystal C by winding up the wire 6. A seed crystal P is attached to the tip of the wire 6 of the pulling mechanism 9.
[0030] The side heater 4 is connected to a heater drive control unit 4a that controls the amount of power supplied, and the pulling mechanism 9 is connected to a rotation drive control unit 9a that controls the rotation drive thereof. In the single crystal pulling apparatus 1, a magnetic field application electromagnetic coil 8 is installed outside the furnace body 10. When a predetermined current is applied to the magnetic field application electromagnetic coil 8, a horizontal magnetic field (transverse magnetic field) of a predetermined strength is applied to the silicon melt M in the quartz glass crucible 3. The magnetic field application electromagnetic coil 8 is connected to an electromagnetic coil control unit 8a that controls its operation.
[0031] That is, in this embodiment, the MCZ method (Magnetic field applied CZ method) is carried out in which a magnetic field is applied into the silicon melt M to grow a silicon single crystal C, thereby controlling the convection of the silicon melt M and stabilizing the single crystallization.
[0032] A radiation shield 7 is disposed above the silicon melt M formed in the quartz glass crucible 3 to surround the silicon single crystal C. The radiation shield 7 is formed with openings at the top and bottom, and serves to shield the silicon single crystal C being grown from excess radiant heat from the side heaters 4 and the silicon melt M, and to straighten the gas flow in the furnace. The gap between the bottom end of the radiation shield 7 and the silicon melt M1 is controlled to maintain a predetermined distance according to the desired characteristics of the silicon single crystal C to be grown.
[0033] A cylindrical water-cooling body 12 is disposed inside the radiation shield 7. Cooling water is supplied to this water-cooling body 12 by cooling water supply means 12a, and is configured to maintain a predetermined temperature by circulating the water.
[0034] The single crystal pulling apparatus 1 further includes a controller 11 having a storage device 11a and an arithmetic and control device 11b, and the heater drive control unit 4a, the electromagnetic coil control unit 8a, the rotation drive control unit 9a, the cooling water supply means 12a, the rotation drive control unit 14a, and the lift drive control unit 15a are each connected to and controlled by the arithmetic and control device 11b.
[0035] <Method of manufacturing silicon single crystals> Fig. 2 is a flow chart showing the flow of a method for producing a silicon single crystal by the single crystal pulling apparatus 1 shown in Fig. 1. Here, it is assumed that a silicon single crystal C having a diameter of, for example, 300 mm is grown using the single crystal pulling apparatus 1 configured as described above. Then, in a state where raw polysilicon (for example, 350 kg) is loaded into the quartz glass crucible 3 and the interior of the furnace body 10 is kept in a predetermined atmosphere (mainly an inert gas such as argon gas), the controller 11 reads out and executes a program stored in the storage device 11a, thereby starting the growth of the silicon single crystal C.
[0036] First, the controller 11 controls the heater drive control unit 4a to heat the side heater 4, which melts the raw material polysilicon loaded in the quartz glass crucible 3 to generate silicon melt M (step S1). Furthermore, the controller 11 controls the rotation drive control unit 14a and the elevation drive control unit 15a to rotate the quartz glass crucible 3 at a predetermined height position at a predetermined rotation speed (rpm) (step S2).
[0037] Next, the controller 11 controls the electromagnetic coil control unit 8a to pass a predetermined current through the magnetic field application electromagnetic coil 8, and applies a horizontal magnetic field in the silicon melt M at a magnetic flux density (for example, 2500 Gauss) set within a range of 1000 to 4000 Gauss (step S3). The rotation drive control unit 9a is also controlled to lower the wire 6 to bring the seed crystal P into contact with the silicon melt M, and the pulling conditions are adjusted using the power supplied to the side heater 4, the pulling speed, the magnetic field application strength, and other parameters to form a neck portion (corresponding to a reduced diameter portion and an expanded diameter portion, which will be described later) (step S4). After the neck portion is formed, the controller 11 further adjusts the power supplied to the side heater 4, the pulling speed, the magnetic field application strength, and other (pulling conditions) to form a shoulder portion (crown portion) C1 (step S5).
[0038] Then, when the crystal diameter is gradually enlarged to form a shoulder portion C1, the controller 11 further adjusts the power supplied to the side heater 4, the pulling speed, the magnetic field application strength, etc. (pulling conditions) to form a straight body portion C2 that will become a product portion (step S6). At this time, in accordance with the lowering of the silicon melt surface M1 due to the progress of the formation of the straight body portion C2 of the silicon single crystal C, the controller 11 controls the lifting drive control unit 15a to raise the carbon susceptor 2 that accommodates the quartz glass crucible 3, thereby maintaining the distance between the radiation shield 7 and the side heater 4, whose positions are fixed, and the silicon melt surface M1. In addition, natural convection of the silicon melt M is suppressed by applying a magnetic field having a magnetic flux density set in the range of 1000 to 4000 Gauss, more preferably 2000 to 3000 Gauss. If the magnetic flux density of the horizontal magnetic field is set lower than the above, for example, in the range of 800 to 1000 Gauss, the silicon melt M becomes unstable and crystal deformation becomes more likely to occur.
[0039] Then, when the body portion C2 is formed to a predetermined length, the controller 11 further adjusts the power supplied to the side heater 4, the pulling speed, the strength of the applied magnetic field, etc. (pulling conditions), and finally forms the tail portion (step S7). Here, the contact area between the bottom end of the silicon single crystal C and the silicon melt M gradually decreases, and then the silicon single crystal C and the silicon melt M are separated.
[0040] In this manner, in this embodiment, a silicon single crystal ingot is manufactured by carrying out the above steps S1 to S7.
[0041] <Neck formation> Next, the step of forming a neck portion in the method for producing a silicon single crystal according to this embodiment will be described.
[0042] Fig. 3 is a schematic diagram showing an image of the neck portion of this embodiment, (a) is a schematic external view of the neck portion, (b) is a diagram showing the process of growing the neck portion, and Fig. 4 is a diagram showing the appearance of the neck portion formed by the neck portion formation process.
[0043] In the neck portion forming step of this embodiment, the reduced diameter portion 31 and the expanded diameter portion 32 constituting the neck portion 21 are alternately formed with an appropriate angle and solid-liquid interface shape. Specifically, the reduced diameter portion 31 and the expanded diameter portion 32 are alternately and repeatedly formed in the process of growing the neck portion 21, and further, the shape of the solid-liquid interface between the silicon melt M and the neck portion 21 being grown is maintained as a vertically upward concave shape during the growth of the neck portion 21. In addition, the angle θ1 (hereinafter, may be simply referred to as "angle θ1" or "angle θ1 of the reduced diameter portion 31"), which is the angle between the horizontal plane (a plane perpendicular to the vertical direction (axial direction) in the neck portion 21) and the outer circumferential surface of the reduced diameter portion 31, is set to 80°≦θ1<90°, and the reduced diameter portion 31 is formed at the angle θ1. Furthermore, the angle θ2 between the horizontal plane and the outer circumferential surface of the enlarged diameter portion 32 (hereinafter, sometimes simply referred to as "angle θ2" or "angle θ2 of the enlarged diameter portion 32") is set to 105°≦θ2<140°, and the enlarged diameter portion 32 is formed at a steeper angle θ2 than when the reduced diameter portion 31 is formed. This makes the height (depth) of the vertically upward concave shape of the solid-liquid interface higher (deeper), making it easier for axial dislocations to escape. That is, in this embodiment, the above configuration efficiently removes slip dislocations and axial dislocations in the neck portion 21, and prevents the propagation of slip dislocations and axial dislocations to the crown portion and the straight body portion.
[0044] In detail, in growing the neck portion 21, first, the seed crystal P is brought into contact with the silicon melt M, and then the crystal diameter of the seed crystal P is reduced from the crystal diameter D0 to form a reduced diameter portion 31 having a minimum diameter D1 in the range of 3.5 mm to 6 mm directly below the seed crystal P. Here, for example, the crystal diameter is reduced at a gentle angle θ1 (80°≦θ1<90°).
[0045] Next, the crystal diameter is expanded from the minimum diameter D1 to form an expanded diameter portion 32 having a maximum diameter D2 in the range of 7 mm to 15 mm. Here, for example, the crystal diameter is expanded at a steep angle θ2 (105°≦θ2<140°) that can increase the bulge of the solid-liquid interface formed in a vertically upward concave shape. Then, the reduced diameter portion 31 and the expanded diameter portion 32 formed as described above are alternately and repeatedly formed in this order.
[0046] In growing the neck portion 21, as described above, it is preferable to form the reduced diameter portion 31 immediately below the seed crystal P. If the expanded diameter portion 32 is formed immediately below the seed crystal P, the stress in the reduced diameter portion 31 increases and dislocations tend to increase when the reduced diameter portion 31 is formed in the next stage, which is not preferable. In addition, it is preferable to repeatedly form three or more reduced diameter portions 31 and three or more expanded diameter portions 32 in the neck portion 21.
[0047] When the reduced diameter portion 31 is formed as described above, the thermal stress in the neck portion 21 increases, and slip dislocations that are generated when the seed crystal P is brought into contact with the silicon melt M tend to propagate downward (in the direction in which the silicon single crystal grows). At this time, slip dislocations that propagate toward the outer periphery of the reduced diameter portion 31 are removed by the reduction in the crystal diameter, so that the downward propagation of slip dislocations can be reduced.
[0048] On the other hand, since slip dislocations propagating toward the center of the reduced diameter portion 31 are unlikely to escape to the outer periphery, the expanded diameter portion 32 is formed after the formation of the reduced diameter portion 31. By forming the expanded diameter portion 32 as described above, the thermal stress in the neck portion 21 is reduced, so that the slip dislocations propagating toward the center of the neck portion 21 are unlikely to move. Therefore, the downward propagation of slip dislocations propagating toward the center of the reduced diameter portion 31 can be reduced.
[0049] In addition, since axial dislocations propagate in a direction perpendicular to the solid-liquid interface, by maintaining the solid-liquid interface in a vertically upward concave shape, they pass through toward the outer periphery of the silicon single crystal. In other words, if the solid-liquid interface can always be maintained in a vertically upward concave shape, the direction of axial dislocations can be directed toward the outer periphery, and if the height of the vertically upward concave shape can be increased, the direction of axial dislocations can be directed even further toward the outer periphery. This makes it possible to efficiently eliminate axial dislocations.
[0050] Fig. 5 is an image diagram showing an example of a case where an axial dislocation remains, and Fig. 6 is an image diagram showing an example of a case where an axial dislocation passes through to the outer periphery of a silicon single crystal. For example, when the diameter is suddenly reduced at an angle of θ1<80° as shown in Fig. 5, the solid-liquid interface shape cannot be maintained as a vertically upward concave shape at the reduced diameter portion 31 (because the solid-liquid interface shape changes to a vertically upward convex shape), so that the axial dislocation propagating in a direction perpendicular to the solid-liquid interface propagates toward the center along with the change in the solid-liquid interface, and as a result, the axial dislocation may not pass through to the outer periphery of the silicon single crystal.
[0051] However, when the crystal diameter is reduced at a gentle angle (80°≦θ1<90°) as in this embodiment, the solid-liquid interface shape can be maintained as a vertically upward concave shape even in the reduced diameter portion 31, so that, for example, as shown in Fig. 6, the axial dislocations are propagated in a direction perpendicular to the solid-liquid interface, and as a result, the axial dislocations pass through toward the outer periphery of the silicon single crystal. In other words, the axial dislocations can be removed.
[0052] Therefore, in the present embodiment, the angle θ1 of the reduced diameter portion 31 is set to 80°≦θ1<90° so that the solid-liquid interface shape can be maintained as a vertically upward concave shape, and the crystal diameter is reduced at a gradual angle. If the diameter is reduced suddenly at an angle of θ1<80°, it is advantageous for removing slip dislocations, but it is not preferable because the solid-liquid interface shape changes to a vertically upward convex shape, which is disadvantageous for removing axial dislocations.
[0053] On the other hand, in the present embodiment, in order to increase the bulge of the solid-liquid interface maintained in a vertically upward concave shape, the optimal angle θ2 of the expansion section 32 is set to 105°≦θ2<140°, and the crystal diameter is expanded at a steep angle. If the diameter is expanded gradually at an angle of 90°<θ2<105°, the bulge of the solid-liquid interface formed in a vertically upward concave shape cannot be increased further, and it becomes difficult to maintain the solid-liquid interface in a vertically upward concave shape when the diameter is reduced (the solid-liquid interface is likely to change to a vertically upward convex shape), which is not preferable.
[0054] By forming the neck portion as described above, the solid-liquid interface shape can be maintained as a vertically upward concave shape even in the reduced diameter portion 31 .
[0055] According to this embodiment, the solid-liquid interface shape is maintained as a vertically upward concave shape, while the reduced diameter portion 31 in which the crystal diameter is reduced at a gradual angle θ1 and the expanded diameter portion 32 in which the crystal diameter is expanded at a steep angle θ2 are alternately formed repeatedly, so that both slip dislocations and axial dislocations can be efficiently removed in a short time.
[0056] In this embodiment, it is preferable that the maximum diameter of the enlarged diameter portion 32 is 3 mm or more larger than the minimum diameter of the reduced diameter portion 31. When the maximum diameter of the enlarged diameter portion 32 is 3 mm or more larger than the minimum diameter of the reduced diameter portion 31, the thermal stress during enlargement is reduced and dislocations are more likely to stop, which is advantageous for removing slip dislocations. In addition, it is possible to form the reduced diameter portion 31 long, and the bulge of the solid-liquid interface maintained in a vertically upward concave shape is larger, which is advantageous for removing axial dislocations. On the other hand, if the above size is 3 mm or less, it is difficult to obtain the effect of forming the enlarged diameter portion 32, which is not preferable.
[0057] The length L1 of the narrowed portion 31 in the vertical direction and the length L2 of the expanded portion 32 in the vertical direction are appropriately set according to the crystal diameter D0 of the seed crystal P used, the minimum diameter D1 of the narrowed portion 31, the maximum diameter D2 of the expanded portion 32, the angle θ1 of the narrowed portion 31, and the angle θ2 of the expanded portion 32. The length L1 of the narrowed portion 31 is preferably in the range of 9 mm or more and 30 mm or less. The length L2 of the expanded portion 32 is preferably in the range of 3 mm or more and 8 mm or less. The ratio of the length L1 of the narrowed portion 31 to the length L2 of the expanded portion 32 is preferably 2.5 or more, and more preferably 3 or more. If the value of L1 / L2 is less than 2.5, the proportion of the length L1 of the narrowed portion 31, which is the portion of the neck portion 21 where axial dislocations are likely to escape, becomes relatively small, which is undesirable because it may be difficult for axial dislocations to escape to the outer periphery of the silicon single crystal.
[0058] It is not preferable that the minimum diameter D1 of the reduced diameter portion 31 is less than 3.5 mm, since the strength decreases when producing a large-diameter (for example, a diameter of 12 inches or more) silicon single crystal, and the neck portion 21 may break. On the other hand, if the minimum diameter D1 of the reduced diameter portion 31 exceeds 6 mm, there is no problem in terms of strength, but it is difficult to eliminate dislocations in the neck portion 21 and produce a dislocation-free silicon single crystal. Therefore, it is preferable that the minimum diameter D1 of the reduced diameter portion 31 is in the range of 3.5 mm to 6 mm.
[0059] Furthermore, if the maximum diameter D2 of the enlarged diameter portion 32 is less than 7 mm, the proportion of the length L1 of the reduced diameter portion 31, where axial dislocations are likely to escape, in the neck portion 21 tends to be relatively small, which is undesirable from the viewpoint of manufacturing. Furthermore, if the maximum diameter D2 of the enlarged diameter portion 32 exceeds 15 mm, the stress of the reduced diameter portion 31 increases and dislocations tend to multiply when the reduced diameter portion 31 is formed in the next stage, which is undesirable. Therefore, the maximum diameter D2 of the enlarged diameter portion 32 is preferably in the range of 7 mm to 15 mm.
[0060] <Effects> As described above, in the method for producing a silicon single crystal according to the present embodiment, the reduced diameter portion 31 and the expanded diameter portion 32 are alternately and repeatedly formed during the neck portion growth process, and further, the shape of the solid-liquid interface between the silicon melt M and the neck portion 21 being grown is maintained as a vertically upward concave shape during the growth of the neck portion 21. Also, the angle θ1 of the reduced diameter portion 31 is set to 80°≦θ1<90°, and the reduced diameter portion 31 is formed at the angle θ1. Furthermore, the angle θ2 of the expanded diameter portion 32 is set to 105°≦θ2<140°, and the expanded diameter portion 32 is formed at an angle θ2 that is steeper than that when the reduced diameter portion 31 is formed.
[0061] The formation of the reduced diameter portion 31 increases the thermal stress in the neck portion 21, and the slip dislocations propagating to the outer periphery of the crystal are removed by the reduction in the crystal diameter, so that the downward propagation of the slip dislocations can be reduced. The formation of the expanded diameter portion 32 reduces the thermal stress in the neck portion 21, and the slip dislocations propagating to the center of the neck portion 21 are less likely to move, so that the downward propagation of the slip dislocations propagating to the center of the reduced diameter portion 31 can be reduced. Therefore, by alternately and repeatedly forming the reduced diameter portion 31 and the expanded diameter portion 32, the slip dislocations can be eliminated. Furthermore, by maintaining the solid-liquid interface shape as a vertically upward concave shape, the axial dislocations are propagated in a direction perpendicular to the solid-liquid interface and are removed toward the outer periphery of the silicon single crystal, so that the axial dislocations can be eliminated. That is, according to this embodiment, in the process of forming the neck portion, the solid-liquid interface shape is always maintained as a vertically upward concave shape, while the reduced diameter portion 31, in which the crystal diameter is reduced at a gradual angle θ1, and the expanded diameter portion 32, in which the crystal diameter is expanded at a steep angle θ2, are alternately and repeatedly formed, so that both slip dislocations and axial dislocations can be efficiently removed.
[0062] In the neck portion forming step in the method for producing a silicon single crystal according to the present embodiment, when forming an enlarged portion and a reduced portion, the method of adjusting the diameter reduction and the diameter expansion by changing the pulling speed is generally used, but is not limited to this. For example, it is also possible to adjust the diameter reduction and the diameter expansion by combining changes in pulling conditions such as the pulling speed, the gap width between the minimum diameter D1 and the maximum diameter D2 of the neck portion 21, the heater power, and the rotation speed of the crucible. EXAMPLES
[0063] Next, an embodiment of the method for producing a silicon single crystal according to the present invention will be described.
[0064] (Test 1) In Test 1, in a single crystal pulling apparatus 1 having the configuration shown in the above-described embodiment, 300 kg to 400 kg of raw material polysilicon and a dopant of boron (B) were placed in a quartz glass crucible 3, and the raw materials were heated in an inert gas (argon gas) atmosphere while rotating the quartz glass crucible 3 to generate silicon melt M.
[0065] A cylindrical silicon seed crystal P was brought into contact with silicon melt M to form a neck portion and a shoulder portion (crown portion) C1 as shown in FIG. 3, and then a boron-heavy doped silicon single crystal ingot (resistivity 20 mΩcm or less) having a straight body portion C2 with a diameter of 300 mm to 320 mm and a length of 1600 mm to 1800 mm was pulled up.
[0066] In order to suppress natural convection of the silicon melt M, the magnetic flux density of the horizontal magnetic field (transverse magnetic field) applied during pulling was set to 2500 Gauss. The pressure inside the furnace was set to 20 Torr to 30 Torr. The rotation speed of the crucible was set to 1 rpm, and the rotation speed of the silicon single crystal was set to 9 rpm (the rotation directions were opposite to each other).
[0067] Furthermore, to form the neck portion, a horizontal magnetic field was applied in the same manner as described above, while controlling the pulling speed and the temperature of the side heater 4. The angle θ1 of the narrowing portion and the angle θ2 of the expanding portion were set for each embodiment and comparative example shown in Table 1 (setting the narrowing / expanding conditions), and the other conditions were kept constant or within a specified range to form the narrowing portion and the expanding portion. The pulling of the silicon single crystal was repeated several times under each condition (while changing the minimum diameter D1 and maximum diameter D2 within the specified range).
[0068] Other conditions are as follows: Minimum diameter of the tapered part D1: 4.5mm~5.5mm Maximum diameter of expansion part D2: 7.5mm~10.0mm Number of repeated forming of narrow and wide sections: 6 times
[0069] Thereafter, the silicon single crystal ingots obtained for each of the examples and comparative examples were used to evaluate slip dislocations based on the presence or absence of dislocations in the shoulder (crown) C1 of the silicon single crystal. In addition, the presence or absence of axial dislocations was evaluated by strain determination using a SIRD (Scanning Infrared Depolarization) device. Figure 7 (Figures 7-1 to 7-5) is a diagram showing an example of a SIRD map for the presence or absence of axial dislocations, where Figure 7-1 shows a case where there is no axial dislocation, and Figures 7-2 to 7-5 show a case where there is axial dislocation (enlarged view). The positions of the circles in Figures 7-2 to 7-5 represent axial dislocations. In strain determination using the SIRD device, the shoulder (crown) C1 is cut into wafers, and if the strain continues at the same position in each wafer, it is evaluated that there is axial dislocation: "present." For example, in Figures 7-2 to 7-5, as a result of observing the raw images (DEP) and the averaged images (GBA) of two wafers (GFMF01 and GFMF02) of the same silicon single crystal, the distortion at the same position is identified as an axial dislocation and is represented by a circle. Note that for the two wafers (LBLK01 and LBLK02) shown in Figure 7-1, no distortion is observed at the same position in the four DEP and GBA images. In other words, the axial dislocation can be evaluated as "none." Table 1 shows the list of tests and the evaluation results in this study.
[0070] [Table 1]
[0071] As shown in Table 1, for the neck portions of Examples 1 to 6 formed under diameter contraction / expansion conditions within the ranges of 80°≦θ1<90° and 105°≦θ2<140°, even when the minimum diameter D1 and maximum diameter D2 were changed within the above-mentioned ranges, it was evaluated that both slip dislocations and axial dislocations were removed (slip dislocations "absent", axial dislocations "absent"). On the other hand, for the neck portions of Comparative Examples 1 to 6 formed under diameter contraction / expansion conditions outside the above-mentioned ranges, it was found that slip dislocations were removed by repeated diameter contraction and expansion, but axial dislocations remained.
[0072] (Test 2) The neck portions of the silicon single crystal ingots produced in Example 1 (constriction / expansion conditions: θ1 = 85°, θ2 = 115°) and Example 2 (constriction / expansion conditions: θ1 = 85°, θ2 = 110°) of Test 1 and the silicon single crystal ingot produced in Comparative Example 1 (constriction / expansion conditions: θ1 = 80°, θ2 = 95°) of Test 1 were cut, and the solid-liquid interface shape of the neck portion was evaluated using an X-ray topograph.
[0073] Fig. 8 is a diagram showing the solid-liquid interface shape by vertical cut evaluation of the neck portion using an X-ray topograph, where (a) shows the expanded diameter portion and the contracted diameter portion formed under the diameter contraction / expansion conditions of Example 1, (b) shows the expanded diameter portion and the contracted diameter portion formed under the diameter contraction / expansion conditions of Example 2, and (c) shows the expanded diameter portion and the contracted diameter portion formed under the diameter contraction / expansion conditions of Comparative Example 1. Fig. 9 is a diagram showing the observation results of the solid-liquid interface shape by vertical cut evaluation of each neck portion shown in Fig. 8, where (a) shows the solid-liquid interface height relative to the neck length (observation of the solid-liquid interface height at 2 mm intervals), and (b) shows the average solid-liquid interface height by diameter contraction / expansion conditions. In FIG. 9( a ), the straight lines connected by △ (triangles) represent the solid-liquid interface height of the neck portion formed under the diameter contraction / expansion conditions of Example 1, the straight lines connected by ◯ (circles) represent the solid-liquid interface height of the neck portion formed under the diameter contraction / expansion conditions of Example 2, and the straight lines connected by ◇ (diamonds) represent the solid-liquid interface height of the neck portion formed under the diameter contraction / expansion conditions of Comparative Example 1.
[0074] As a result of observing the above-mentioned Figs. 8 and 9, it was found that the solid-liquid interface of a vertically upward concave shape was maintained in all three neck parts, but for the neck part of Comparative Example 1, which was evaluated as having axial dislocations "present" in Test 1, both the peak value of the solid-liquid interface height shown in Fig. 9(a) and the average solid-liquid interface height shown in Fig. 9(b) were lower than those of the neck parts of Examples 1 and 2, which were evaluated as having axial dislocations "absent" in Test 1. That is, for the neck parts formed under the diameter contraction / expansion conditions of Examples 1 and 2, the bulge of the solid-liquid interface maintained in a vertically upward concave shape can be further increased in the expanded diameter part, while, on the other hand, for the neck part formed under the diameter contraction / expansion conditions of Comparative Example 1, the bulge of the solid-liquid interface maintained in a vertically upward concave shape cannot be further increased in the expanded diameter part. In particular, Example 1 is expanded at a larger angle than Example 2, and as a result, the average solid-liquid interface height is higher than that of Example 2. In addition, in Example 1, the ratio of length L1 to length L2 of the expanded diameter portion is set to a value of L1 / L2 of 3 or more to lengthen the contracted diameter portion, thereby making it easier for axial dislocations to escape. It is also preferable to set the value of L1 / L2 to 2.5 or more as in Example 2, and it is even more preferable to set the value of L1 / L2 to 3 or more as in Example 1.
[0075] Furthermore, according to FIG. 8, for the neck portion formed under the diameter contraction / expansion conditions of Comparative Example 1, the value of L1 / L2, which is the ratio of the length L1 of the contracted portion to the length L2 of the expanded portion, was "15 mm / 12 mm," which was not 2.5 or more.
[0076] In this embodiment, the test was performed on a silicon single crystal ingot heavily doped with boron, but the test is not limited to this, and the same effect (evaluation) can be obtained with heavy doping of arsenic, red phosphorus, etc. Also, the test is not limited to heavy doping, and can be applied to light doping as well, but it is known that the dislocation movement speed is faster in light doping than in heavy doping, and axial dislocations are less likely to remain. [Explanation of symbols]
[0077] 1. Single crystal pulling device 2 Carbon susceptor (graphite susceptor) 3. Quartz glass crucible 3a Straight body part 3b bottom 4 Side heater 4a Heater drive control section 6 Wire 7 Radiation Shield 8. Magnetic field application electromagnetic coil 8a Electromagnetic coil control section 9 Lifting mechanism 9a Rotation drive control section 10 Furnace body 10a Main Chamber 10b Pull chamber 11 Controller 11a Storage device 11b Arithmetic and control device 12 Water cooling body 12a Cooling water supply means 14 Rotation drive unit 14a Rotation drive control section 15 Lifting drive unit 15a Lifting drive control section 21 Neck 31 Reduced diameter part 32 Expanded diameter part
Claims
1. A method for producing a silicon single crystal in which a reduced diameter portion and an expanded diameter portion are alternately and repeatedly formed in a neck portion growing process, comprising: a first angle between a horizontal plane and an outer circumferential surface of the tapered portion that is equal to or greater than 80° and less than 90°, while maintaining a shape of a solid-liquid interface between the neck portion and the silicon melt in a vertically upward concave shape; Furthermore, a second angle between a horizontal plane and an outer circumferential surface of the expanded diameter portion is set to 105° or more and less than 140°, and the expanded diameter portion is formed at the second angle which is steeper than that at the time of forming the reduced diameter portion.
2. A method for producing a silicon single crystal comprising the steps of:
2. The maximum diameter of the enlarged diameter portion is 3 mm or more larger than the minimum diameter of the reduced diameter portion. The method for producing a silicon single crystal according to claim 1 .
3. The ratio of a first length, which is the vertical length of the reduced diameter portion, to a second length, which is the vertical length of the expanded diameter portion, is set to 2.5 or more. The method for producing a silicon single crystal according to claim 1 or 2.
4. The first length is in the range of 9 mm to 30 mm, and the second length is in the range of 3 mm to 8 mm. The method for producing a silicon single crystal according to claim 3 .
Citation Information
Patent Citations
Production of oxide single crystal by lifting up method
JP1996225391A
Growth of single crystal
JP1997235180A
Growth of silicon single crystal
JP1999199384A
Method for growing silicon single crystal
JP2009263142A
Method for growing silicon single crystal
JP2011057460A