MANUFACTURING PROCESS FOR SINGLE CRYSTALS

Abstract

A method for producing a single crystal using the Czochralski method, comprising the following steps: a step of measuring a space between a melt surface and a lower end of a seed crystal provided above the melt; a step in which the seed crystal is lowered on the basis of the space and the seed crystal is brought into contact with the melt; and a step of growing the single crystal at the lower end of the seed crystal by pulling the seed crystal while maintaining a state of contact with the melt, wherein, in the step of measuring the space between the melt surface and the lower end of the seed crystal, an image of the seed crystal and the melt surface is taken using a camera installed obliquely above the melt surface, an approximate circle of the edge of a real image is generated by approximating a circle from an approximately arcuate edge pattern at a lower end part of a real image of the seed crystal shown in the image taken by the camera, and similarly an approximate circle of the edge of a mirror image is generated by approximating a circle from an approximately arcuate edge pattern at an upper end part of a mirror image of the seed crystal reflected at the melt surface, and the space between the melt surface and the lower end of the seed crystal is calculated based on the distance from the center coordinates of the approximation circle of the edge of the real image to the center coordinates of the approximation circle of the edge of the mirror image, and wherein in the step of measuring the space between the melt surface and the lower end of the seed crystal, the approximate circle of the edge of the real image positioned above the approximate circle of the edge of the mirror image is obtained after the approximate circle of the edge of the mirror image is obtained from the captured image, and wherein a step in which the approximation circle of the edge of the mirror image is obtained comprises the following steps: a step of determining an upper end position of the mirror image of the seed crystal from a brightness distribution in a first region defined in advance in the captured image; a step of defining a second region including the upper end position of the mirror image of the seed crystal, and detecting an edge pattern of a part of the straight body from the mirror image of the seed crystal by performing a binarization process of the second region; and a step in which a circle is approximated from the edge pattern of the part of the straight body of the mirror image, where a step in which the approximation circle of the edge of the real image is obtained comprises the following steps: a step of setting a position where a difference value of brightness exceeds a predetermined threshold as a lower end position of the real image of the seed crystal when the first area is scanned upward from the center coordinates of the approximation circle of the edge of the mirror image; a step of defining a third region including the lower end position of the real image of the seed crystal, and detecting the edge pattern of the part of the straight body from the real image of the seed crystal by performing the binarization process of the third region; and a step in which a circle is approximated from the edge pattern of the part of the straight body of the real image.

Description

FIELD OF THE INVENTION

[0001] The present invention relates to a manufacturing method for single crystals by the Czochralski (CZ) method, and more particularly, to a measuring method for a space between a seed crystal and a melt surface, and a preheating method of the seed crystal using the same.

[0002] A large number of silicon single crystals, which become substrate materials for semiconductor devices, are produced using the CZ process. In the CZ process, a polycrystalline silicon starting material is heated in a quartz crucible to form a melt. After a seed crystal is brought into contact with the melt, the seed crystal is slowly pulled while rotating the seed crystal and melt, thereby growing a large-diameter single crystal from a lower end of the seed crystal. The CZ process can increase the production yield of large-diameter single crystals.

[0003] For example, regarding the manufacturing method of the single crystals by the CZ method, Patent Literature 1 describes that before growing a single crystal, an image of a seed crystal is captured with an optical camera, the image captured with the camera is processed to detect a position of the seed crystal, a distal end of the seed crystal is stopped at a standard position provided above a melt of the raw material, a distance from the standard position to the surface of the melt of the raw material is detected, and a crucible that stores the melt of the raw material is vertically shifted according to the detected distance.

[0004] Patent Literature 2 also describes a method in which, in order to suppress the dislocation of a seed crystal due to a thermal shock when it comes into contact with a melt, the seed crystal is preheated above the melt surface to make the temperature difference between the two as small as possible, and then the seed crystal is brought into contact with the melt.It is further described that, in order to accurately measure a space between the surface of the melt of the raw material and the seed crystal, information about the position of a lower end point of a real image, which is a specific point at a lower end of the seed crystal, and information about the position of a mirror image point, which is a point corresponding to the lower end point of a real image in a mirror image of the seed crystal reflected at the melt, the space between the surface of the melt of the raw material and the lower end of the seed crystal at the point where the position of the lower end point of the real image coincides with the position of the mirror image point is defined as zero, and the space between the melt surface of the melt of the raw material and the lower end of the seed crystal is obtained.In addition, in order to suppress dislocation due to thermal shock when contacting the melt, Patent Literatures 1, 3 and 4 describe the use of a seed crystal with a tapered shape, the seed crystal having a sharp distal end.

[0005] Patent Literature 5 describes that, in order to reduce the variation of a preheating position of a seed crystal and improve the rate at which the seed crystal contacts the melt without dislocation, an image including a real image of the seed crystal and a mirror image of the seed crystal reflected on a melt surface is taken with a camera from obliquely above, a position of the lower end point of the real image, which is a point at a lower end of a part of the straight body of the real image of the seed crystal in the image, and a position of a mirror image point, which is a point at an upper end of the part of the straight body of the mirror image of the seed crystal corresponding to the lower end point of the real image, are respectively obtained, which is located between the melt surface and the lower end of the part of the straight body of the seed crystal at the point,at which the position of the lower end point of the real image on the image coincides with the position of the mirror image point is defined as zero, the space between the melt surface and the lower end of the part of the straight body of the seed crystal is obtained, and by subtracting the length of a tapered part of the seed crystal from the space between the melt surface and the lower end of the part of the straight body of the seed crystal, the space between the melt surface and the lower end of the seed crystal is obtained. RELATED PRIOR ARTPatent literature Patent Literature 1: Japanese Patent Laid-Open Publication JP 2005-170773 A Patent Literature 2: Japanese Patent Laid-Open Publication JP 2016-155729 A Patent Literature 3: Japanese Patent Laid-Open Publication JP 2000-128691 A Patent Literature 4: Japanese Patent Laid-Open Publication JP 2009-234889 A Patent Literature 5: Japanese Patent Laid-Open Publication JP 2019-214486 A

[0006] Further prior art can be found in JP 2018-100195 A. SUMMARY OF THE INVENTIONProblems to be solved by the invention

[0007] When controlling a preheating position of a seed crystal based on the position of the seed crystal in the height direction determined from an image captured with a camera, the seed crystal is advantageously brought as close as possible to a melt surface without the seed crystal coming into contact with the melt. However, when the seed crystal and the melt surface are captured from an oblique angle with the camera, and consequently, when a seed crystal with a tapered shape is used, the distal end of the tapered shape is hidden behind the seed crystal, and the distal end of the tapered shape cannot be captured with the camera. Furthermore, detecting a lower end position of the seed crystal from the mirror image of the seed crystal is also difficult because the distal end of the tapered part cannot be captured from a mirror image of the seed crystal reflected on the melt surface.In addition, there are machining variations of the tapered shape of the seed crystal and consequently, the position of the seed crystal must be controlled taking such machining variations into account.

[0008] Patent Literature 5 describes a method for obtaining the distance from the melt surface to the distal end of the seed crystal with the tapered shape. However, the accuracy with which the distance from the melt surface to the distal end of the seed crystal can be calculated is insufficient, and creativity is needed to further improve the rate at which the seed crystal contacts the melt without dislocation by calculating the distance more accurately.

[0009] Accordingly, the present invention provides a manufacturing method for a single crystal that accurately measures a space between a melt surface and a lower end of a seed crystal even when a seed crystal having a tapered shape at the lower end is used, and thereby reduces the variation of a preheating position when preheating above a melt surface and enables the rate at which the seed crystal comes into contact with the melt without dislocation to be improved. Means of solving the tasks

[0010] To solve the above problems, the single crystal manufacturing method according to the present invention is a single crystal manufacturing method using the Czochralski method, and includes a step of measuring a space between a lower end of a seed crystal provided above a melt and a melt surface; a step of lowering the seed crystal based on the space and bringing the seed crystal into contact with the melt; and a step of growing the single crystal at the lower end of the seed crystal by pulling the seed crystal while maintaining a state of contact with the melt. In the step of measuring the space between the melt surface and the lower end of the seed crystal, an image of the seed crystal and the melt surface is captured using a camera installed obliquely above the melt surface.An approximate circle of the edge of a real image is generated by approximating a circle from an approximately arcuate edge pattern at the lower end part of the real image of the seed crystal shown in the image captured by the camera, and an approximate circle of the edge of a mirror image is also generated by approximating a circle from an approximately arcuate edge pattern at an upper end part of the mirror image of the seed crystal reflected at the melt surface, and the space between the lower end of the seed crystal and the melt surface is calculated based on the distance from the center coordinates of the approximate circle of the edge of the real image to the center coordinates of the approximate circle of the edge of the mirror image.

[0011] According to the present invention, the space between the melt surface and the bottom end of the seed crystal can be accurately maintained even when the bottom end of the seed crystal has a tapered shape. Therefore, when the seed crystal is preheated above the melt, the variation in the preheating position is reduced, and the rate at which the seed crystal contacts the melt without dislocation can be improved.

[0012] The single crystal manufacturing method of the present invention preferably converts the distance into real-space units by multiplying a conversion factor equal to half the pixel number from the center coordinates of the approximate circle of the edge of the real image to the center coordinates of the approximate circle of the edge of the mirror image. Accordingly, the space between the melt surface and the bottom of the seed crystal can be accurately obtained in real space.

[0013] The single crystal manufacturing method according to the present invention preferably further includes a step of obtaining the conversion factor from a change in a pixel position of the seed crystal in the captured image when the position of the seed crystal is moved by a fixed distance in the vertical direction, before measuring the space between the melt surface and the bottom of the seed crystal. Accordingly, the space between the melt surface and the bottom of the seed crystal in real space can be accurately obtained.

[0014] The single crystal manufacturing method of the present invention can calculate the space between the melt surface and the bottom of the seed crystal based on the focal length and installation angle of the camera after performing a projection transformation of the coordinates in the captured image of the approximate circle of the edge of the real image and the approximate circle of the edge of the mirror image to real space coordinates. Accordingly, the space between the melt surface and the bottom of the seed crystal can be calculated without prior calibration to obtain the conversion factor.

[0015] The single crystal manufacturing method of the present invention preferably sets a reference plane at the same height position as the bottom end of the seed crystal, performs projection transformation of the approximate circle of the edge of the real image and the approximate circle of the edge of the mirror image onto the reference plane, and calculates the distance between the melt surface and the bottom end of the seed crystal from the center coordinates of the approximate circle of the edge of the real image, the center coordinates of the approximate circle of the edge of the mirror image, and the center coordinates of a camera lens. Accordingly, by using the coordinates after the projection transformation, the space between the melt surface and the bottom end of the seed crystal can be easily calculated.

[0016] According to the invention, the step of measuring the space between the melt surface and the bottom of the seed crystal obtains the approximate circle of the edge of the real image positioned above the approximate circle of the edge of the mirror image after obtaining the approximate circle of the edge of the mirror image from the captured image. Accordingly, the positions of the real image and the mirror image of the seed crystal in the captured image can be efficiently determined.

[0017] In the present invention, a step of obtaining the approximate circle of the edge of the mirror image includes a step of determining an upper end position of the mirror image of the seed crystal from a brightness distribution in a first region defined in advance in the captured image, a step of defining a second region including the upper end position of the mirror image of the seed crystal, and detecting an edge pattern of a part of a straight body of the mirror image of the seed crystal by performing a binarization process of the second region, and a step of approximating a circle from the edge pattern of the part of the straight body of the mirror image; and a step of obtaining the approximate circle of the edge of the real image includes a step of determining a positionwhere a brightness difference value exceeds a predetermined threshold as the lower end position of the real image of the seed crystal when the first region is scanned upward from the center coordinates of the approximate circle of the edge of the mirror image; a step of defining a third region including the lower end position of the real image of the seed crystal and detecting the edge pattern of the straight body part of the real image of the seed crystal by performing the binarization process of the third region; and a step of approximating a circle from the edge pattern of the straight body part of the real image. Accordingly, the center coordinates of the approximate circle of the edge of the mirror image and the center coordinates of the approximate circle of the edge of the real image in the captured image can be efficiently obtained.

[0018] According to the present invention, the step of measuring the space between the melt surface and the bottom of the seed crystal preferably continuously captures multiple images including the real image and the mirror image of the seed crystal when the seed crystal is positioned at the same height, and calculates the space between the melt surface and the bottom of the seed crystal from the average of the values ​​obtained from each of the multiple images. Accordingly, the accuracy with which the space between the melt surface and the bottom of the seed crystal can be measured can be increased.

[0019] According to the invention, the seed crystal preferably further includes a tapered part present below the straight body part, the lower end of the seed crystal being the lower end of the tapered part, the taper angle of the tapered part being larger than the installation angle of the camera, and further includes a step of measuring the length of the tapered part in advance before measuring the space between the melt surface and the lower end of the seed crystal, and the step of measuring the space between the melt surface and the lower end of the seed crystal includes a step of subtracting the length of the tapered part from the space between the melt surface and the lower end of the straight body part of the seed crystal.If the taper angle of the seed crystal is larger than the camera installation angle, the camera cannot capture an image of the distal end of the seed crystal, and the position of the distal end of the seed crystal cannot be directly obtained from the captured image. However, according to the invention, the space between the seed crystal and the melt surface can be accurately measured without capturing the image of the distal end of the seed crystal.

[0020] The single crystal manufacturing method of the present invention preferably further includes a step of, after measuring the space after the melt surface and the lower end of the seed crystal and before the seed crystal is brought into contact with the melt, adjusting the height position of at least one crucible supporting the seed crystal and the melt so that the space becomes a target value, and a step of preheating the seed crystal by keeping it stationary at the target position. According to the present invention, when the seed crystal is preheated with the tapered portion at the distal end above the melt surface, the variation in the preheating position is reduced, and the rate at which the seed crystal comes into contact with the melt without dislocation can be improved.

[0021] In addition, the present invention is a method for measuring the space between the melt surface and the lower end of the seed crystal installed above the melt, which captures an image of the seed crystal and the melt surface using the camera installed obliquely above the melt surface, generates the approximate circle of the edge of the real image by approximating a circle from the approximately arc-shaped edge pattern at the lower end part of the real image of the seed crystal shown in the image captured by the camera, and also generates the approximate circle of the edge of the mirror image by approximating a circle from the approximately arc-shaped edge pattern at the upper end part of the mirror image of the seed crystal reflected on the melt surface,and the space between the lower end of the seed crystal and the melt surface is calculated based on the number of pixels from the center coordinates of the approximation circle of the edge of the real image to the center coordinates of the approximation circle of the edge of the mirror image.,

[0022] According to the present invention, even when the lower end portion of the seed crystal has a tapered shape, the space between the melt surface and the lower end of the seed crystal can be accurately maintained. Consequently, when the seed crystal is preheated above the melt, the variation in the preheating position can be reduced, and the rate at which the seed crystal comes into contact with the melt without dislocation can be improved.

[0023] Furthermore, a preheating method for the seed crystal according to the present invention includes a step of measuring the space between the melt surface and the lower end of the seed crystal by the above-mentioned space measuring method, a step of adjusting the height position of at least one crucible supporting the seed crystal and the melt so that the space becomes the target value, and a step of preheating the seed crystal by keeping it still at the position of the target value.

[0024] According to the present invention, when the seed crystal is preheated with the tapered part at the distal end above the melt surface, the variation of the preheating position is reduced, and the rate at which the seed crystal comes into contact with the melt without dislocation can be improved.

[0025] Likewise, an apparatus for producing a single crystal includes a crucible supporting the melt, a heater for heating the melt, a crystal pulling mechanism that drives the seed crystal located above the melt to rise, a camera that obliquely captures the seed crystal and the melt surface from above, an image processor that processes the image captured by the camera, and a controller that controls the crystal pulling mechanism based on the processing results of the image processor. The image processor generates an approximate circle of the edge of a real image by approximating a circle from the approximately arcuate edge pattern at the lower end part of the real image of the seed crystal shown in the image captured by the camera, and also generates an approximate circle of the edge of a mirror image.by approximating a circle from the approximately arcuate edge pattern at the upper end part of the mirror image of the seed crystal reflected at the melt surface, and calculating the space between the melt surface and the lower end of the seed crystal based on the pixel number from the center coordinates of the approximate circle of the edge of the real image to the center coordinates of the approximate circle of the edge of the mirror image.

[0026] According to the present invention, even if the lower end portion of the seed crystal has a tapered shape, the space between the melt surface and the lower end of the seed crystal can be accurately maintained. Therefore, when the seed crystal is preheated above the melt, the variation in the preheating position can be reduced and the rate at which the seed crystal comes into contact with the melt without dislocation can be improved. Effect of the invention

[0027] The present invention can provide a manufacturing method for a single crystal that can accurately maintain a space between a melt surface and a lower end of a seed crystal even if the lower end part of the seed crystal has a tapered shape. BRIEF DESCRIPTION OF THE DRAWINGS [ Fig. 1] Fig. 1 is a side sectional view schematically illustrating the configuration of an apparatus for manufacturing a single crystal according to an embodiment of the present invention. [ Fig. 2] Fig. Figure 2 is a flowchart illustrating a manufacturing process for a silicon single crystal. [ Fig. 3] Fig. 3 is a cross-sectional view schematically illustrating a shape of the silicon single crystal ingot. [ Fig. 4] Fig. 4 is a side sectional view schematically illustrating a position of a seed crystal in an apparatus for producing a single crystal during a preheating process. [ Fig. 5] Fig. 5 is a view describing a method of measuring a space between a melt surface and a bottom end of the seed crystal, and is a schematic view illustrating a relationship of positions between a real image and a mirror image of the seed crystal. [ Fig. 6] Fig. 6(a) to Fig. 6(c) are views describing the relationship of positions between the seed crystal and the melt surface; Fig. 6(a) is a schematic view in which the distance between the seed crystal and the melt surface is wide, Fig. 6(b) is a schematic view in which the distance between the seed crystal and the melt surface is close, and Fig. Figure 6(c) is a graph illustrating the relationship between a height position of the seed crystal and a pixel number between the real image and the mirror image. [ Fig. 7] Fig. 7(a) to Fig. 7(c) are views describing how a pixel number N D between the real image and the mirror image. [ Fig. 8] Fig. 8 is a flowchart detailing a detection method for center coordinates of the real image P R and center coordinates of a mirror image P M for a seed crystal 5 in a recorded image. [ Fig. 9] Fig. 9 is a schematic view describing a method for projection transformation of two-dimensional coordinates of the captured image to coordinates in real space. [ Fig. 10] Fig. 10 is a schematic view describing a method for calculating the space between the seed crystal and the melt surface from the center coordinates of a circular outer edge pattern of the lower end of the part of the straight body in both the real image and the mirror image of the seed crystal. WAY TO IMPLEMENT THE INVENTION

[0028] Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

[0029] Fig. 1 is a side sectional view schematically illustrating a configuration of an apparatus for manufacturing a single crystal according to an embodiment of the present invention.

[0030] As in Fig. 1, an apparatus 1 for producing a single crystal includes a water-cooled chamber 10, a quartz crucible 11 holding silicon melt 2 in the chamber 10, a graphite crucible 12 holding the quartz crucible 11, a rotating shaft 13 supporting the graphite crucible 12, a crucible driving mechanism 14 rotating and driving the quartz crucible 11 via the rotating shaft 13 and the graphite crucible 12 to rise, a heater 15 arranged around the circumference of the graphite crucible 12, thermal insulation material 16 arranged on the outside of the heater 15 and along an inner surface of the chamber 10, a heat-shielding body 17 arranged above the quartz crucible 11, a wire 18 which is a crystal pulling axis arranged coaxially with the rotating shaft 13 and extending above the quartz crucible 11 a crystal pulling mechanism 19 arranged above the chamber 10, a camera 20,which is located outside the chamber 10 and records the interior of the chamber 10 through an observation window 10e, an image processor 21 which processes the image recorded by the camera 20, and a controller 22 which controls several components in the device 1 for producing the single crystal.

[0031] The chamber 10 is equipped with a main chamber 10a and a slender cylindrical pulling chamber 10b connected to an upper opening of the main chamber 10a. The quartz crucible 11, the graphite crucible 12, the heater 15, and the heat shield 17 are arranged inside the main chamber 10a. The pulling chamber 10b is provided with a gas inlet 10c for introducing inert gas (purge gas), such as argon gas, or doping gas, into the chamber 10, and a gas outlet 10d is provided at the bottom of the main chamber 10a for discharging atmospheric gas in the chamber 10. In addition, the observation window 10e is provided on top of the main chamber 10a, and the growth status of a silicon single crystal 3 can be observed through the observation window 10e.

[0032] The quartz crucible 11 is a container made of quartz glass with a cylindrical side wall and a curved bottom. The graphite crucible 12 adheres to an outer surface of the quartz crucible 11 and is held in place by wrapping the quartz crucible 11 to maintain the shape of the quartz crucible 11, which softens upon heating. The quartz crucible 11 and the graphite crucible 12 form a dual-structure crucible that supports the silicon melt in the chamber 10.

[0033] The graphite crucible 12 is fixed to the upper end portion of the rotating shaft 13, and the lower end portion of the rotating shaft 13 passes through the bottom of the chamber 10 to connect to the crucible drive mechanism 14 provided outside the chamber 10. The graphite crucible 12, the rotating shaft 13, and the crucible drive mechanism 14 constitute a rotation mechanism and a lifting mechanism for the quartz crucible 11. The operations of rotating and lifting the quartz crucible 11, driven by the crucible drive mechanism 14, are controlled by the controller 22.

[0034] The heater 15 is used to melt silicon raw material filled in the quartz crucible 11 to produce the silicon melt 2, as well as to maintain the molten state of the silicon melt 2. The heater 15 is a resistance heater made of carbon and is provided to surround the quartz crucible 11 inside the quartz crucible 12. Furthermore, the thermal insulation material 16 is attached to the outside of the heater 15 to surround the heater 15, thereby increasing heat retention inside the chamber 10. The power of the heater 15 is controlled by the controller 22.

[0035] The heat shielding body 17 is provided to suppress the temperature fluctuation in the silicon melt 2 and thus provide appropriate heat distribution near a crystal growth interface, and also prevents the heating of the silicon crystal 3 by radiant heat from the heater 15 and the quartz crucible 11. The heat shielding body 17 is a substantially cylindrical member made of graphite and is provided to cover the region above the silicon melt 2, excluding a pulling path of the silicon single crystal 3.

[0036] A diameter of the opening at the lower end of the heat-shielding body 17 is larger than the diameter of the silicon single crystal 3, thereby ensuring the pulling path of the silicon single crystal 3. In addition, the outer diameter of the lower end portion of the heat-shielding body 17 is smaller than an opening of the quartz crucible 11, and the lower end portion of the heat-shielding body 17 is located inside the quartz crucible 11. Thus, even if an upper end of a rim of the quartz crucible 11 is raised above the lower end of the heat-shielding body 17, the heat-shielding body 17 and the quartz crucible 11 do not interfere with each other.

[0037] Although the amount of melt in the quartz crucible 11 decreases as the silicon single crystal 3 grows, the quartz crucible 11 is raised so that the space (gap) between a melt surface 2a and the heat shielding body 17 is constant, thereby controlling the temperature fluctuation of the silicon melt 2 and controlling the evaporation of dopant from the silicon melt 2 by keeping the flow rate of gas flowing near the melt surface 2a constant. Such gap control allows for the improvement of the stability of the crystal defect distribution, the oxygen concentration distribution, the resistivity distribution, and the like in the pulling axis direction of the silicon single crystal 3.

[0038] The wire 18, which is the pulling axis of the silicon single crystal 3, and the crystal pulling mechanism 19, which pulls the silicon single crystal 3 by winding the wire 18, are provided above the quartz crucible 11. The crystal pulling mechanism 19 functions to rotate the wire 18 and the silicon single crystal 3. The crystal pulling mechanism 19 is controlled by the controller 22. The crystal pulling mechanism 19 is arranged above the pulling chamber 10b, and the wire 18 extends downward from the crystal pulling mechanism 19 through the pulling chamber 10b, and the distal end of the wire 18 reaches the interior of the main chamber 10a. Fig. 1 shows a state in which the silicon single crystal 3 is suspended from the wire 18 in the middle of its growth. When the silicon single crystal 3 is pulled, the silicon single crystal 3 grows by gradually pulling the wire 18 while rotating both the quartz crucible 11 and the silicon single crystal 3. The crystal pulling speed is controlled by the controller 22.

[0039] The camera 20 is arranged outside the chamber 10. The camera 20 is, for example, a CCD camera and captures an image of the interior of the chamber 10 through the observation window 10e formed in the chamber 10. The installation angle of the camera 20 forms a predetermined angle (preferably 20 to 30°) with a vertical direction, and the camera 20 has an optical axis inclined with respect to the pulling axis of the silicon single crystal 3. In other words, the camera 20 captures an area of ​​the upper surface of the quartz crucible 11, including the circular opening of the heat-shielding body 17 and the melt surface of the silicon melt 2, from an oblique top view.

[0040] The camera 20 is connected to the image processor 21, and the image processor 21 is connected to the controller 22. The image processor 21 calculates the crystal diameter near the solid-liquid interface from the edge pattern of the single crystal shown in the image captured by the camera 20. In addition, a gap, which is the distance from the lower end of the heat-shielding body 17 to the melt surface 2a, is calculated from the position of the mirror image of the heat-shielding body 17 reflected on the melt surface in the captured image.The method for calculating the gap is not particularly limited, but for example, a conversion formula obtained by performing a linear approximation of the relationship between the gap and the position of the mirror image of the heat-shielding body 17 is generated in advance, and the gap can be determined by substituting the position of the mirror image of the heat-shielding body during a crystal pulling process into the conversion formula. Furthermore, the gap can be geometrically calculated from the relationship of the positions between the real image and the mirror image of the heat-shielding body 17 shown in the captured image.

[0041] The controller 22 controls the crystal diameter by controlling the crystal pulling speed based on crystal diameter data obtained from the image captured by the camera 20. Specifically, when a measured value of the crystal diameter is larger than a target diameter, the crystal pulling speed is increased, and when the measured value of the crystal diameter is smaller than the target diameter, the pulling speed is decreased. In addition, the controller 22 controls the amount of movement (crucible lifting speed) of the quartz crucible 11 based on the crystal length data of the silicon single crystal 3 obtained from a sensor of the crystal pulling mechanism 19 and the crystal diameter data obtained from the image captured by the camera 20.

[0042] Fig. 2 is a flowchart illustrating a manufacturing process for the silicon single crystal 3. Fig. Figure 3 is a cross-sectional view schematically illustrating the shape of a silicon single crystal ingot.

[0043] As in Fig. 2, the manufacturing step of the silicon single crystal 3 according to the present embodiment includes a raw material melting step S11 that heats the silicon raw material in the quartz crucible 11 with the heater 15 and produces the silicon melt 2, a preheating step S12 that preheats the seed crystal attached to the distal end of the wire 18 before bringing the seed crystal into contact with the melt, a melt contacting step S13 that lowers the seed crystal to bring the seed crystal into contact with the silicon melt 2, and crystal pulling steps (S14 to S17) that grow the single crystal by gradually pulling the seed crystal while maintaining the state of contact with the silicon melt 2.

[0044] In the crystal pulling steps (S14 to S17), a neck formation step S14 forming a neck 3a where the crystal diameter is thinly narrowed so as not to retain a dislocation; a shoulder growth step S15 forming a shoulder 3b where the crystal diameter gradually increases as the crystal grows; a body growth step S16 forming a body 3c maintained at a certain crystal diameter; and a tail growth step S17 forming a tail 3d where the crystal diameter is gradually reduced as the crystal grows are performed in sequence.

[0045] Thereafter, a cooling step S18 is performed, which separates the silicon single crystal 3 from the melt surface and promotes cooling. The above steps complete a silicon single crystal ingot 3I having the neck 3a, the shoulder 3b, the body 3c, and the tail 3d, as shown in Fig. 3 is shown.

[0046] Fig. 4 is a side sectional view schematically illustrating a position of the seed crystal in the single crystal manufacturing apparatus 1 during the preheating step S12.

[0047] As in Fig. As illustrated in Figure 4, the preheating step S12 is the step in which the seed crystal 5 is heated by keeping the seed crystal 5 still above the melt surface 2a for a fixed period of time. When the seed crystal 5, which has a large temperature difference from the silicon melt 2, is brought into contact with the silicon melt 2, slip dislocation occurs due to thermal shock; however, if the preheating step S12 is performed, the dislocation is suppressed, and the rate at which the seed crystal 5 comes into contact with the melt without dislocation can be increased. If the seed crystal 5 is separated from the melt surface 2a, heat is not sufficiently supplied in the preheating step S12, and the temperature difference between the seed crystal 5 and the silicon melt 2 cannot be reduced.Consequently, the space between the lower end of the seed crystal 5 and the melt surface 2a must be controlled to be 5 mm or less, and the space between the lower end of the seed crystal 5 and the melt surface 2a must be accurately measured to accomplish this.

[0048] Fig. 5 is a view describing a method for measuring the space between the lower end of the seed crystal 5 and the melt surface 2a, and is a schematic view illustrating the relationship of positions between a real image 5R and a mirror image 5M of the seed crystal 5.

[0049] As in Fig. As illustrated in Fig. 5, the seed crystal 5 according to the present embodiment has a straight body part 5a with a constant thickness and a tapered part 5b extending downward from the lower end of the straight body part 5a. The diameter of the straight body part 5a is preferably 6 to 12 mm, and the length of the straight body part 5a is preferably 30 to 80 mm. In addition, the tapered part 5b is preferably 5 to 100 mm long. The seed crystal 5 of this type is manufactured by cutting a cylindrical silicon single crystal rod into a predetermined length, then performing a grinding process to sharpen the distal end, and further performing an etching process to remove damage from processing. The tapered shape may be a conical shape or a truncated cone shape with a flat distal end.

[0050] The taper angle θ1 of the seed crystal 5 is larger than the installation angle θ0 of the camera 20. If the distal end of the seed crystal 5 has a tapered shape of this type, the distal end P0 of the seed crystal 5 cannot be photographed by the camera 20 because the distal end P0 of the seed crystal 5 (the lower end of the tapered part 5b) is located in a blind spot of the camera 20. With this in mind, a length L T of the tapered part 5b is precisely measured in advance and a space L P1 between the melt surface 2a and a lower end P1 of the part 5a of the straight body of the seed crystal 5, which can be photographed with the camera 20, is calculated, and then by subtracting the length L T of the tapered part 5b of the space L P1 a room L S = L P1 - L T between the distal end P0 of the seed crystal 5 and the melt surface 2a.

[0051] In this example, the room L P1 between the melt surface 2a and the lower end P1 of the part 5a of the straight body of the seed crystal 5, half a distance (distance between the real image and the mirror image) L D from the lower end P1 of the part 5a of the straight body of the real image 5R of the seed crystal 5 to the mirror image P2 of the lower end P1. Accordingly, the space L S between the distal end P0 of the seed crystal 5 and the melt surface 2a by determining the distance L D between the real image and the mirror image.

[0052] It is preferred that the length L Tof the tapered part 5b is accurately measured, for example, using a non-contact measuring method. When measuring the length of the tapered part 5b in advance, it is preferable to measure the length of the tapered part 5b at multiple positions in a circumferential direction and use the average of the multiple measured values. This is because the length of the tapered part 5b may vary depending on the machining accuracy of the tapered part 5b based on the position of the tapered part 5b in the circumferential direction.

[0053] As described in detail below, the distance L D between the real image and the mirror image in real space can be obtained by multiplying a predetermined conversion factor α by a number of pixels (number of pixels N D between the real image and the mirror image) in the recorded image, which corresponds to the distance L D between the real image and the mirror image. The number of pixels N Dbetween the real image and the mirror image can be defined as the number of pixels from the center coordinates (center coordinates of the real image P R (X R , Y R )) of a circular outer edge of the lower end of the part 5a of the straight body of the real image 5R of the seed crystal 5 to the center coordinates (center coordinates of the mirror image P M (X M , Y M )) of a circular outer edge of the upper end of the part 5a of the straight body of the mirror image 5M of the seed crystal 5.

[0054] The conversion factor α can be obtained by a calibration operation that establishes a relational expression between a relative displacement amount (millimeters) in real space and a relative displacement amount (number of pixels) on the captured image when the seed crystal 5 is moved a fixed distance in the vertical direction. In calibration, first, while the seed crystal 5 is moved in the vertical direction, an image including the seed crystal 5 and the melt surface 2a is captured with the camera 20 at multiple height positions of the seed crystal 5. The multiple height positions of the seed crystal 5 should be relative height positions rather than absolute height positions, so that the multiple height positions of the seed crystal 5 with respect to the melt surface 2a can be defined by fixing the height position of the seed crystal 5 and changing the height position of the melt surface 2a (height position of the crucible).

[0055] Fig. 6(a) to Fig. 6(c) are views describing the positional relationship between the seed crystal and the melt surface; Fig. 6(a) is a schematic view in which the distance between the seed crystal and the melt surface is wide, Fig. 6(b) is a schematic view in which the distance between the seed crystal and the melt surface is close and Fig. Figure 6(c) is a graph illustrating a relationship between a height position of the seed crystal and the pixel number between the real image and the mirror image.

[0056] As in Fig. 6(a) and Fig. As illustrated in Fig. 6(b), the real image 5R and the mirror image 5M of the seed crystal 5 have a symmetrical positional relationship across the melt surface 2a, and this relationship does not change even when the seed crystal 5 is moved in the vertical direction. For example, a lowering amount (feed amount of the wire) when the seed crystal 5 is lowered from a first height position h1 to a second height position h2 is defined as Δh = h1 - h2 (mm), and in this connection, when the pixel number between the real image and the mirror image differs from N D1 to N D2 (N D1 >N D2 ), the ratio of a magnitude of change in the number of pixels relative to the feed amount Δh of the wire as ΔN D / Δh = (N D1 -N D2 ) / (h1 - h2) and using ΔN D / Δh as the conversion factor α, the magnitude of the change in the height position of the seed crystal 5 in the recorded image can be converted into the magnitude of the change in the height position of the seed crystal 5 in the real space. As in Fig. 6(c), the conversion factor α can be obtained as the slope of a first regression line (y = αx + β) showing the relationship between the plurality of height positions h of the seed crystal 5 and the pixel number N D between the real image and the corresponding mirror image.

[0057] It is preferred as the distance L Dbetween the real image and the mirror image, the average of the multiple values ​​obtained from the multiple images of the seed crystal 5 continuously acquired at the same height position is used. The continuous acquisition cycle here is several hundred ms. Accordingly, the measurement accuracy can be increased by reducing measurement errors caused by fluctuations in the measurement environment.

[0058] Fig. 7(a) to Fig. 7(c) are views describing how the pixel number N D between the real image and the mirror image can be obtained.

[0059] As in Fig. 7(a) to Fig. 7(c), the distance L D between the real image and the mirror image from the number of pixels N D between the real image and the mirror image, and the number of pixels N Dbetween the real image and the mirror image can be defined as the number of pixels from the center coordinates P R of the real image of the seed crystal 5 in the recorded image to the center coordinates of the mirror image P M By performing a binarization process on the captured image and detecting an approximately arcuate edge pattern of the part 5a of the straight body of the real image 5R or the mirror image 5M of the seed crystal 5, the center coordinates P R of the real image and the center coordinates P M of the mirror image of the seed crystal 5 as the center coordinates of an approximation circle when a circle is approximated from the edge pattern.

[0060] As another method to determine the distance L Dbetween the real image and the mirror image from a surface at the lower end of the part 5a of the straight body of the real image 5R of the seed crystal 5 to a surface at the upper end of the part 5a of the straight body of the mirror image 5M of the seed crystal 5, the distance L Dbetween the real image and the mirror image is defined as the space from the lowest end of the edge pattern of the part 5a of the straight body of the real image 5R of the seed crystal 5 to the lowest end of the edge pattern of the part 5a of the straight body of the mirror image 5M of the seed crystal 5 in the captured image. However, with this method, the measurement error is likely to increase due to a change in the brightness distribution or effects of eccentric rotation caused by the installation position and the deviation of the inclination of the seed crystal 5. On the other hand, if, as in the present embodiment, the corresponding center coordinates of the approximate circle of the edge pattern of the part 5a of the straight body of the real image 5R and the mirror image 5M obtained from the captured image are calculated and the distance between the two center coordinates is defined as the distance L Dbetween the real image and the mirror image, it is unlikely that the eccentric rotation of the seed crystal and the change in the brightness distribution have an effect and the measurement error of the distance L D between the real image and the mirror image can be reduced.

[0061] Fig. 8 is a flowchart showing a detection method for the center coordinates P R of the real image and the center coordinates P M of the mirror image of the seed crystal 5 in the recorded image in detail.

[0062] As in Fig. 8, in the present embodiment, after cutting out a first region including the real image 5R and the mirror image 5M of the seed crystal 5 from the image taken with the camera 20 (step S20), the center coordinates P M (X M , Y M) of the mirror image are first obtained (steps S21 to S24) and then the center coordinates P R (X R , Y R ) of the real image (steps S25 to S28). The reason for cutting out the first area from the image captured by the camera 20 is to restrict a target area of ​​the image processing, because the image captured by the camera 20 is an image of a wide area of ​​the interior of the furnace, and the area thereof in which the seed crystal 5 is captured is only a small part of the captured image. In addition, the reason is that the center coordinates P M of the mirror image first, that the brightness of the game image 5M of the seed crystal 5 in the recorded image is greater than the brightness of the real image 5R and its position is easier to identify.

[0063] When detecting the center coordinates P Mof the mirror image of the seed crystal 5 (steps S21 to S24), a brightness level of the brightness peak value in the first region, for example, 99.9%, is defined as a threshold value for the detection of the mirror image, and this threshold value is used to identify the upper end position of the mirror image 5M at the position of the topmost white pixel obtained when the binarization process is performed on the first region (step S21).

[0064] Next, using the upper end position of the pixels identified as the mirror image 5M as a standard, an image within a set range (e.g., 50 pixels long × 100 pixels wide) below it is defined as the detailed detection range (second range) of the mirror image 5M (step S22), and a brightness level, which is, for example, 99% of the brightness peak value in the second range, is defined as the threshold for edge detection. Then, this threshold is used to binarize the image in the second range and obtain a position of the black-and-white boundary of the binary image as the edge of the part 5a of the straight body of the mirror image 5M of the seed crystal 5 (step S23). After that, a circle is approximated from the approximately arcuate edge pattern using the least squares method, and the center coordinates P Mof the mirror image, which are the center coordinates of the approximation circle of the edge of the mirror image, are obtained (step S24).

[0065] Next, when detecting the center coordinates P R of the real image of the seed crystal 5, the detailed detection range of the real image 5R (third range) is obtained from the first range, which is the general detection range of the real image 5R and the mirror image 5M of the seed crystal 5. Since the real image 5R of the seed crystal 5 is known to exist above the mirror image 5M, using the center coordinates P Mof the mirror image of the seed crystal 5 as a standard, the first area is scanned upward from these coordinates, and a position where the difference value of the brightness level exceeds the predetermined threshold is set as the lower end position of the real image 5R (step S25). Using the thus detected lower end position of the real image 5R as a standard, an image within a set range above this position (e.g., 50 pixels long × 100 pixels wide) is defined as the third area (step S26), and a brightness level that is, for example, 99% of the brightness peak value in the third area is defined as a threshold for edge detection.Then, this threshold is used to binarize the image in the third region and obtain the position of the black-and-white boundary of the binary image as the edge of the straight body part 5a of the real image 5R of the seed crystal 5 (step S27). After that, a circle is approximated from the approximately arcuate edge pattern using the least squares method, and the center coordinates P are determined. R of the real image, which are the center coordinates of the approximation circle of the edge of the real image, are obtained (step S28).

[0066] When the lower end of the real image 5R of the seed crystal 5 overlaps with the upper end of the mirror image 5M because the lower end of the seed crystal 5 is close to the melt surface 2a, the mirror image 5M of the seed crystal 5 becomes noise for the real image 5R, and it is difficult to separate the real image 5R from the mirror image 5M. In such a case, by performing an open process (expansion and contraction process) on the binary image, the mirror image 5M can be separated from the real image 5R, and the edge of the real image 5R can be fixed.

[0067] After the center coordinates P R of the real image and the center coordinates P M of the mirror image of the seed crystal 5 were each obtained in this way, the distance L D between the real image and the mirror image of the seed crystal 5 from the pixel number N D between the two.

[0068] In the preheating step S12 of the seed crystal 5, after measuring the space L S between the seed crystal 5 and the melt surface 2a, using the above-mentioned method, the seed crystal 5 is moved in a direction in which the difference between the measured value and the target value becomes smaller. In other words, the controller 22 adjusts the height of the seed crystal 5 by operating the crystal pulling mechanism 19, or adjusts the height of the quartz crucible 11 by operating the crucible driving mechanism 14 to compensate for the difference between the measured value and the target value. Accordingly, the actual space L S between the melt surface 2a and the seed crystal 5 can be defined as the target space.

[0069] The space (target value) between the lower end of the seed crystal 5 and the melt surface 2a in the preheating step S12 of the seed crystal 5 is preferably 5 mm or less, and more preferably 3 mm or less. Accordingly, the temperature difference between the seed crystal and the melt can be sufficiently reduced, and the rate at which the seed crystal comes into contact with the melt without dislocation can be increased. On the other hand, the initial space between the melt surface 2a and the seed crystal 5 before measuring the space between the melt surface 2a and the lower end of the seed crystal 5 is preferably 5 mm or larger. This is because there is a risk that the seed crystal 5 may come into contact with the melt if the seed crystal 5 is brought too close to the melt surface 2a in a state where the space between the melt surface 2a and the seed crystal 5 is not accurately known.

[0070] In the preheating step S12 of the seed crystal 5, the seed crystal 5 and the melt surface 2a are maintained in a neighboring state for, for example, several minutes to several hours. Accordingly, the seed crystal 5 is heated by radiant heat from the silicon melt 2. By preheating the seed crystal 5 each time at a predetermined position above the melt surface 2a, the reproducibility of the temperature of the seed crystal 5 after preheating can be increased. In addition, by setting the target space sufficiently small, the temperature difference between the two can be sufficiently small, and the thermal shock when the seed crystal 5 is in contact with the melt can be reduced. Therefore, the introduction of dislocations into the seed crystal 5 can be suppressed.

[0071] Furthermore, in the conventional method in which an operator visually adjusts the space between the melt surface 2a and the bottom end of the seed crystal when preheating the seed crystal 5, the seed crystal 5 may inadvertently be in contact with the silicon melt 2 if the target space is set to a very small value, for example, 3 mm. However, by accurately measuring the space between the melt surface 2a and the bottom end of the seed crystal 5 according to the present embodiment, the contact between the seed crystal 5 and the melt surface 2a can be avoided.

[0072] After preheating the seed crystal 5 in this way, the seed crystal 5 is lowered and brought into contact with the silicon melt 2, and then the seed crystal 5 is lifted to grow the silicon single crystal 3.

[0073] As described above, the manufacturing method for the single crystal according to the present embodiment approximates a circle from the edge pattern of the real image and mirror image of the seed crystal 5 and obtains the distance L D between the real image and the mirror image of the seed crystal 5 based on the space of the center coordinates P R of the approximation circle of the edge of the real image to the center coordinates P M of the approximation circle of the edge of the mirror image. Therefore, the space between the melt surface 2a and the lower end of the seed crystal 5 can be accurately obtained even if the seed crystal 5 has the tapered portion 5b.

[0074] Next, a description will be given of another method for measuring the space between the melt surface 2a and the lower end of the seed crystal 5. The above-mentioned method for measuring the space between the seed crystal and the melt surface needs to use calibration to convert the space between the real image and the mirror image of the seed crystal in the image into the space between the real image and the mirror image of the seed crystal in real space, and calibration to obtain the conversion factor α is necessary to do this.

[0075] Against this background, the present embodiment directly calculates the space between the seed crystal and the melt surface without calibration by performing projection transformation of coordinate points in an extracted image to real-space coordinates. Specifically, the image of the melt surface 2a inside the furnace is captured, the coordinates of the mirror image of the seed crystal 5 reflected on the melt surface 2a and the real image of the seed crystal 5 in the captured image are detected, the projection transformation of the respective coordinates of the real image and the mirror image of the seed crystal 5 to real-space coordinates is performed, and a space ΔG between the seed crystal and the melt surface is calculated from the real-space coordinates of the real image and the mirror image of the seed crystal, respectively.The following describes in detail the procedure of projection transformation of two-dimensional coordinates of the captured image to coordinates in real space.

[0076] Fig. Figure 9 is a schematic view describing the method of projection transformation from two-dimensional coordinates of the captured image to coordinates in real space.

[0077] As on the left side of the Fig. As illustrated in Figure 9, the camera 20 records the interior of the chamber obliquely from above, and consequently, the outer contour of the circle at the lower end of the straight body portion of the seed crystal 5 is distorted in the recorded image. In order to accurately calculate dimensions of both the real image and the mirror image of the seed crystal 5, correction of the image distortion is necessary. In this regard, the distortion is corrected by performing projection transformation of the coordinates of the image recorded with the camera 20 to coordinates on a reference plane at the same height position as the lower end of the straight body portion of the seed crystal 5.

[0078] The figure on the right in Fig. 9 shows a coordinate system when the image is corrected. In this coordinate system, the reference plane is referred to as the xy plane. In addition, the origin point C0 of the XY coordinates is located at an intersection point of the reference plane and a straight line (dotted line) drawn from a center position C of an imaging device 20a of the camera 20 so as to pass through a center position F (0, y f , e.g. f ) of a lens 20b of the camera 20. This straight line is the optical axis of the camera 20.

[0079] Furthermore, the pulling direction of the silicon single crystal 3 is in a positive direction of a z-axis and the center position C (0, y c , e.g. c ) of the imaging device 20a and the center position F (0, y f , e.g. f ) of the lens 20b lie in a yz-plane. The coordinates (u, v) in the diagram on the left side are Fig. 9 are represented by pixels of the imaging device 20a and correspond to an arbitrary point P (x p , y p , e.g. p ) on the imaging device 20a, as illustrated in equation (1) below. [Equation 1] xp=−αuuyp=yc−αvv cos θczp=zc+αvv sin θc}

[0080] In this example, α u and α v a pixel size in the vertical and transverse directions of the imaging device 20a and y c , e.g. c are the y-coordinate and z-coordinate of the center position C of the imaging device 20a. Furthermore, as shown on the right side of the Fig. 9 is illustrated, θ c an angle formed by the z-axis and the optical axis of the camera 20 and is the installation angle of the camera 20.

[0081] If the distance from a coordinate origin point C0 on the reference plane to the center position C (0, yc , e.g. c ) of the imaging device 20a as L c is defined, y c and z c each expressed as equation (2) below. [Equation 2] yc=Lc sin θczc=Lc cos θc}

[0082] If the distance from the coordinate origin point C0 to the center position F of the lens 20b of the camera 20 is defined as a and the distance from the center position F of the lens 20b to the center position C of the imaging device 20a is defined as b, the distance L C from the coordinate origin point C0 to the center position C of the imaging device 20a is expressed as equation (3) below. [Equation 3] Lc=L1+L2

[0083] Similarly, based on the lens imaging formula, the focal length f1 is expressed as equation (4) below using the distances a and b. [Equation 4] 1fl=1L1+1L2

[0084] Equation (5) below shows that if the distance L2 is eliminated from equations (3) and (4) and L C with a distance L1 and the focal length f1. [Equation 5] Lc=L1+L1flL1−fl

[0085] The value of the distance L1 from the coordinate origin point C0 to the center position F of the lens 20b of the camera 20 can be expressed as equation (6) below, where the center position F (0, y f , e.g. f ) of the lens 20b of the camera 20. [Equation 6] L1=yf2+zf2

[0086] Therefore, the equation (2) given above can be expressed as the equation (7) below, where the focal length f1 of the lens 20b and the center position F (0, y f , e.g. f ) of the lens 20b of the camera 20. [Equation 7] yc=yf2+zf2[1+fl / (yf2+zf2−fl)]sin θczc=yf2+zf2[1+fl / (yf2+zf2−fl)]cos θc}

[0087] If the lens 20b is considered as a pin hole, an arbitrary point P (x p , y p , e.g. p ) on the imaging device 20a by F (0, y f , e.g. f ) is projected onto the reference plane, and the projected point P' (X, Y, 0) is expressed as equation (8) below. [Equation 8] X=−xpzf / (zp−zf)Y=(yfzp−ypzf) / (zp−zf)}

[0088] Consequently, the coordinates of the real image and the mirror image of the seed crystal undergoing the projection transformation to the reference plane can be obtained using equations (1), (7) and (8).

[0089] Next, a method for calculating the radius and center coordinates of the circular outer edge pattern of the lower end of the straight body part of the seed crystal is described. The least squares method can be used as the method for calculating the radius r and the center coordinates (x0, y0) of the seed crystal from the coordinates of the real image and the mirror image projected onto the reference plane. The outer edge of the lower end of the straight body part of the seed crystal is shaped into a circle, and the outer edge image satisfies the circle equation shown in Equation (9) below. [Equation 9] (x−x0)2+(y−y0)2=r2

[0090] In this example, the least squares method is used to calculate r and (x0, y0) in equation (9). To simplify the least squares calculation, the transformation shown in equation (10) below is performed. [Equation 10] z=a+bx+cyz=x2+y2a=r2−x02−y02b=2x0c=2y0}

[0091] The variables a, b, and c in equation (10) are determined by the least squares method. This allows us to obtain a condition where the sum of the roots of the difference between equation (10) and the measured point is at its minimum, which is obtained by solving the partial differential equation shown in equation (11) below. [Equation 11] ∂∂a,b,c∑i(a+bxi+cyi−zi)2=0

[0092] The solution of this equation 11 can then be calculated by the simultaneous equation shown in equation (12) below. [Equation 12] (∑izi∑izixi∑iziyi)=(n∑ixi∑iyi∑ixi∑ixi2∑ixiyi∑iziyi∑ixiyi∑ixi2)(abc)

[0093] Using the least squares method in this way, it is possible to calculate the radius and center coordinates of the circular outer edge of the lower end of the part of the straight body from both the real image 5R and the mirror image 5M of the seed crystal projected onto the reference plane.

[0094] Fig. 10 is a schematic view describing a method for calculating the space ΔG between the seed crystal and the melt surface from the center coordinates of the circular outer edge pattern of the lower end of the part of the straight body from both the real image 5R and the mirror image 5M of the seed crystal 5.

[0095] As in Fig. 10, the center coordinates C m of the lower end of the part of the straight body of the mirror image 5M of the seed crystal 5 naturally on the opposite side of the melt surface 2a from the center coordinates C h (X hc , Y hc , 0) of the lower end of the part 5a of the straight body of the real image 5R of the seed crystal 5, and the straight line connecting the two points passes through C h (X hc , Y hc , 0) and is parallel to the Z-axis.

[0096] On the other hand, the center coordinates C m ' (X mc , Y mc , 0) of the lower end of the part of the straight body of the mirror image 5M of the seed crystal 5 on the reference plane, the coordinates of the center coordinates C h (X hc , Y hc, 0) of the lower end of the part of the straight body of the real image 5R of the seed crystal 5, projected onto the reference plane, and consequently the center coordinates (X mc , Y mc , Z gap ) of the lower end of the part of the straight body of the mirror image 5M of the seed crystal 5 on the straight line passing through the center position F (0, y f , e.g. f ) of the lens 20b and the center coordinates (X mc , Y mc , 0) of the lower end of the part of the straight body of the mirror image 5M of the seed crystal 5 on the reference plane. Consequently, the space ΔG to be calculated between the melt surface and the seed crystal is half the value Z gap , which can be calculated from equation (13) shown below. [Equation 13] −2ΔG=Zgap=Zf−Zf(Ymc−Yf) / (Yhc−Yf)

[0097] In this way, the space ΔG between the melt surface and the seed crystal can be calculated from the center coordinates C h (X hc , Y hc , 0) of the real image of the seed crystal 5 on the reference plane, the center coordinates C m ' (X mc , Y mc , 0) of the mirror image of the seed crystal 5 on the reference plane and the center coordinates F (0, y f , e.g. f ) of the lens 20b of the camera 20. In the present embodiment, the reference plane for projection transformation is set at the same height position as the lower end of the straight body portion 5a of the seed crystal 5; however, it may be set at other height positions. Thus, for example, the reference plane may be set at the same height position as an opening edge of the lower end of the heat-shielding body 17.

[0098] A preferred embodiment of the present invention has been described above, however, the present invention is not limited to the above-mentioned embodiment and various modifications are possible without departing from the scope of the present invention, and such modifications are of course covered by the scope of the present invention.

[0099] For example, in the embodiment described above, the edge of the real image is obtained after the edge of the mirror image is obtained; however, it is also possible to obtain the edge of the mirror image after the edge of the real image is obtained.

[0100] In addition, in the above-described embodiment, a case is exemplified where the distal end portion of the seed crystal 5 has a tapered shape. However, the present invention can also be applied when the seed crystal 5 consists only of the straight body portion 5a, and the space L S between the melt surface 2a and the lower end of the seed crystal 5 can be calculated based on the number of pixels N D from the center coordinates P R of the real image to the center coordinates P M of the mirror image in the captured image can be accurately measured.

[0101] Furthermore, in the above-described embodiment, a manufacturing method for a silicon single crystal is given as an example. However, the present invention can also be applied to a manufacturing method for various single crystals that can be grown by the CZ method, such as germanium and sapphire. DESCRIPTION OF REFERENCE SYMBOLS 1 Device for producing a single crystal 2 silicon melt 2a Melt surface 3 Silicon single crystal 3I silicon single crystal ingot 3a Neck 3b shoulder 3c Body 3D cock 5 seed crystals 5M mirror image 5R real image 5a Part of the straight body 5b tapered part 10 chambers 10a Main Chamber 10b Drawing chamber 10c Gas inlet 10d Gas outlet 10th observation window 11 quartz crucibles 12 graphite crucibles 13 Rotation shaft 14 Crucible drive mechanism 15 Heating 16 thermal insulation material 17 Heat shielding body 18 wire 19 Crystal pulling mechanism 20 Camera 21 Image processor 22 Control L D Distance between real image and mirror image L P1 Space between melt surface and lower end of the part of the straight body of the seed crystal L S Space between melt surface and lower end of the seed crystal N D Number of pixels between real image and mirror image P0 distal end of the seed crystal P1 lower end of the part of the straight body of the seed crystal P2 Mirror image of the lower end of the part of the straight body of the seed crystal P MCenter coordinates of the mirror image P R Center coordinates of the real image S11 Step of melting the starting material S12 Preheating step S13 Step of contacting with the melt S14 Step of neck formation S15 Step of shoulder growth S16 Step of growth of the body S17 Step of tail growth S18 Cooling step S20 Step of cutting out a first area S21 Step of detecting the position of the mirror image S22 Step of cutting out the second area S23 Binarization step of the second area S24 Step of calculating the center coordinates of the mirror image S25 Step of detecting the position of the real image S26 Step of cutting out the third area S27 Binarization step of the third area S28 Step of calculating the center coordinates of the real image h, h1, h1 Height position of the seed crystal Δh supply amount of seed crystal α conversion factor θ, θ0 Installation angle of the camera θ1 taper angle

Claims

[1] Manufacturing process for a single crystal using the Czochralski method, which includes the following steps: a step of measuring a space between a melt surface and a lower end of a seed crystal provided above the melt; a step in which the seed crystal is lowered on the basis of the space and the seed crystal is brought into contact with the melt; and a step of growing the single crystal at the lower end of the seed crystal by pulling the seed crystal while maintaining a state of contact with the melt, wherein, in the step of measuring the space between the melt surface and the lower end of the seed crystal, an image of the seed crystal and the melt surface is taken using a camera installed obliquely above the melt surface, an approximate circle of the edge of a real image is generated by approximating a circle from an approximately arcuate edge pattern at a lower end part of a real image of the seed crystal shown in the image taken by the camera, and similarly an approximate circle of the edge of a mirror image is generated by approximating a circle from an approximately arcuate edge pattern at an upper end part of a mirror image of the seed crystal reflected at the melt surface, and the space between the melt surface and the lower end of the seed crystal is calculated based on the distance from the center coordinates of the approximation circle of the edge of the real image to the center coordinates of the approximation circle of the edge of the mirror image, and wherein in the step of measuring the space between the melt surface and the lower end of the seed crystal, the approximate circle of the edge of the real image positioned above the approximate circle of the edge of the mirror image is obtained after the approximate circle of the edge of the mirror image is obtained from the captured image, and where a step in which the approximation circle of the edge of the mirror image is obtained comprises the following steps: a step of determining an upper end position of the mirror image of the seed crystal from a brightness distribution in a first region defined in advance in the captured image; a step of defining a second region including the upper end position of the mirror image of the seed crystal, and detecting an edge pattern of a part of the straight body from the mirror image of the seed crystal by performing a binarization process of the second region; and a step in which a circle is approximated from the edge pattern of the part of the straight body of the mirror image, where a step in which the approximation circle of the edge of the real image is obtained comprises the following steps: a step of setting a position where a difference value of brightness exceeds a predetermined threshold as a lower end position of the real image of the seed crystal when the first area is scanned upward from the center coordinates of the approximation circle of the edge of the mirror image; a step of defining a third region including the lower end position of the real image of the seed crystal, and detecting the edge pattern of the part of the straight body from the real image of the seed crystal by performing the binarization process of the third region; and a step in which a circle is approximated from the edge pattern of the part of the straight body of the real image. [2] The manufacturing method for the single crystal according to claim 1, further comprising converting into units of distance in a real space by multiplying a conversion factor by a half value of a pixel number from the center coordinates of the approximate circle of the edge of the real image to the center coordinates of the approximate circle of the edge of the mirror image. [3] The manufacturing method for the single crystal according to claim 2, further comprising a step, before measuring the space between the melt surface and the lower end of the seed crystal, of obtaining the conversion factor from a quantity of change in a pixel position of the seed crystal in the captured image when the position of the seed crystal is moved in the vertical direction by a fixed distance. [4] The manufacturing method for the single crystal according to claim 1, wherein the space between the melt surface and the lower end of the seed crystal is calculated based on a focal length and an installation angle of the camera after performing a projection transformation of the coordinates in the captured image of the approximate circle of the edge of the real image and the approximate circle of the edge of the mirror image to coordinates in the real space. [5] The manufacturing method for the single crystal according to claim 4, wherein a reference plane is set at the same height position as the lower end of the seed crystal, the projection transformation of the approximate circle of the edge of the real image and the approximate circle of the edge of the mirror image onto the reference plane is performed, and the distance between the melt surface and the lower end of the seed crystal is calculated from the center coordinates of the approximate circle of the edge of the real image, the center coordinates of the approximate circle of the edge of the mirror image, and the center coordinates of a lens of the camera. [6] The manufacturing method for the single crystal according to any one of claims 1 to 5, wherein in the step of measuring the space between the melt surface and the bottom end of the seed crystal, a plurality of images including the real image and the mirror image of the seed crystal when the seed crystal is positioned at the same height are continuously taken, and the space between the melt surface and the bottom end of the seed crystal is calculated from an average of the values obtained from each of the plurality of images. [7] A manufacturing method for the single crystal according to any one of claims 1 to 6, wherein the seed crystal further comprises a tapered part present below the part of the straight body, wherein the lower end of the seed crystal is the lower end of the tapered part and the taper angle of the tapered part is greater than the installation angle of the camera, and the method further comprises a step of measuring a length of the tapered part in advance before measuring the space between the melt surface and the lower end of the seed crystal, wherein the step of measuring the space between the melt surface and the lower end of the seed crystal includes a step of subtracting the length of the tapered part from the space between the melt surface and the lower end of the straight body part of the seed crystal. [8] A manufacturing method for the single crystal according to any one of claims 1 to 7, further comprising the following steps: a step in which, after measuring the space between the melt surface and the lower end of the seed crystal and before the seed crystal is brought into contact with the melt, the height position of at least one crucible supporting the seed crystal and the melt is adjusted so that the space becomes a target value; and a step in which the seed crystal is preheated by holding it still at the target position.

Images