JPWO2017110762A1 - Silica glass crucible strain measuring device, silicon single crystal manufacturing method, silica glass crucible strain measuring method, phase difference map, ingot and homoepitaxial wafer - Google Patents

Abstract

Accurately measure the internal residual stress of silica glass crucibles. The present invention relates to a strain for measuring the strain of a silica glass crucible comprising a cylindrical side wall, a curved bottom, and a corner provided between the side wall and the bottom and having a curvature higher than the curvature of the bottom. A measuring device, which is disposed on the side of the side wall, emits polarized light toward the side wall, an imaging unit that captures an image corresponding to the polarized light on the upper end surface of the side wall, and an imaging unit And an output unit for outputting a strain distribution of the silica glass crucible based on the captured image.

Description

  The present invention relates to a strain measuring device for a silica glass crucible, a method for producing a silicon single crystal, a strain measuring method for a silica glass crucible, a phase difference map, an ingot, and a homoepitaxial wafer.

  In one example, the method for producing a silica glass crucible includes a silica powder layer forming step of forming a silica powder layer by depositing silica powder having an average particle size of about 100 μm to 400 μm inside a rotating carbon mold using centrifugal force. And an arc melting step of forming a silica glass layer by arc melting the silica powder layer while reducing the pressure of the silica powder layer from the mold side (this method is referred to as “rotary molding method”).

  At the initial stage of the arc melting process, a so-called seal layer is formed in which the entire outermost surface of the silica powder layer is vitrified, and then the bubbles are removed by strongly depressurizing to remove a transparent silica glass layer (hereinafter referred to as “transparent layer”). Then, a bubble-containing silica glass layer (hereinafter referred to as “non-transparent layer”) in which bubbles remain is formed by reducing the reduced pressure. Thereby, for example, a two-layered silica glass crucible having a transparent layer on the inner surface side and a non-transparent layer on the outer surface side is formed.

Silica powder used for the production of crucible is natural silica powder (crystalline) produced by pulverizing 99% or more of pure silica after purifying natural quartz and then removing impurities. And synthetic silica powder (glass) manufactured by chemical synthesis. Among these, natural silica powder uses natural products as raw materials, and therefore, physical properties, shapes, and sizes tend to vary. When the physical properties, shape, and size of the silica powder change, the molten state of the silica powder changes.
That is, in the arc melting, the silica powder is first sintered, and after the volume diffusion, the temperature further rises so that the grain boundary disappears, and the vitrification and the Si—O—Si network structure are formed. The sintering speed and vitrification speed change. Specifically, for example, if the silica powder is small or has a shape with a large surface area even in the same volume, the sintering rate and the vitrification rate are increased. If the silica powder is small, the space between adjacent silica powders will also be small, and sintering and vitrification will proceed faster than the rate at which bubbles are removed under reduced pressure, so the bubbles in the produced glass crucible are small and many Become. Thus, the molecular structure of the glass after arc melting and the contained bubbles change depending on the sintering speed and vitrification speed.
Therefore, even if arc melting is performed under the same conditions, the shape of the crucible to be manufactured, the roundness of the inner diameter and the outer shape, the thickness of the transparent layer and the non-transparent layer, and the boundary surface between the transparent layer and the non-transparent layer Variations occur in roundness (vertical waviness) and the like.
In a silica glass crucible with a diameter of 32 inches (about 81 cm) used for crystal pulling for a 300 mm silicon wafer, the thickness of the transparent layer is about 1 mm to about 10 mm or less, and the thickness of the non-transparent layer is about 5 mm. It is about 50 mm or less. As the diameter of the silica glass crucible increases, the above-mentioned variation becomes more prominent.

  Moreover, the fused silica glass is solidified through a cooling process after the arc melting process. In this cooling step, the method of bonding between silicon and oxygen (for example, a 6-membered ring or an 8-membered ring) or the voids between atoms in the bonded structure of silicon and oxygen is determined by a cooling method such as cooling rate or blowing of cooling gas. The size changes. For example, as the proportion of the structure having a large member ring such as an 8-membered ring increases, the number of voids also increases. As described above, since the bonding state of the material atoms changes in a complicated manner depending on various conditions such as the melting process and the cooling process in the manufacture of the crucible, the distribution of the internal residual stress after cooling the silica glass crucible changes, and the strength of the crucible (Deformation during CZ pulling) will be affected.

  Patent Document 1 describes that different internal stresses exist in each layer of the silica glass crucible due to a difference in shrinkage amount of each layer during curing after arc melting. Patent Document 2 describes that the occurrence of defects such as cracks and peeling can be suppressed by relaxing internal stress generated when the inner layer of the crucible is densified.

JP 2013-1112597 A JP 2013-095562 A Japanese Unexamined Patent Application Publication No. 2014-094851 JP 2011-225409 A JP 2005-015288 A JP 2011-093778 A JP 2010-013306 A Japanese Patent Laid-Open No. 2005-082474 JP 2010-132498 A JP 2010-163297 A JP 2010-053015 A JP 2010-215460 A JP 2011-068531 A JP 2011-107882 A JP2011-219319A JP 2011-246341 A JP 2012-250866 A JP 2013-216505 A Japanese Patent Laying-Open No. 2015-089854 JP, 2006-018927, A Special table 2014-528643 gazette JP 2008-219002 A

Supervised by Michio Tajima, “Silicon Crystal Technology”, Japan Academic Signaling World No. 145 Committee Technology Transfer Project Editor's Committee, January 20, 2015, Second Edition, p. 88-111

  The quality of the silica glass crucible is closely related to the quality of the silicon single crystal when the silicon single crystal is pulled (for example, CZ method) using the silica glass crucible. For example, variation in the shape of the manufactured silica glass crucible leads to deterioration in the yield of silicon single crystals. Impurities in the silica glass crucible (for example, impurity metal elements in the glass) and foreign matters lead to dislocation of the silicon single crystal. Depending on the smoothness of the inner surface of the silica glass crucible (unevenness that can be seen visually), the amount and size of bubbles in the vicinity of the surface, small debris (crucible in the crucible surface) Etc.) fall off into the silicon melt. When this is mixed in the silicon ingot, the silicon ingot is dislocated. Depending on the thickness distribution and external shape of the silica glass crucible, the deformation of the silica glass crucible (side wall collapse, distortion, bottom bulge, etc.) changes during pulling, and the silicon melt liquid changes due to the change in crucible internal volume. Deviation of the surface descent speed will occur. If the minute holes on the outside of the silica glass crucible are larger than a predetermined size, or there are chips or cracks, the silica glass crucible may break during the pulling. Dispersion of the outer diameter of the silica glass crucible, variation in shape, and protrusions on the outer surface that are larger than the specified size may cause problems when the silica glass crucible is inserted into the carbon susceptor, or more than necessary with the carbon susceptor. This will cause a gap.

  Further, in a silicon single crystal pulling apparatus, polycrystalline silicon as a material is put into a silica glass crucible, and heated to 1420 ° C. or more to be melted. Then, the seed crystal is brought into contact with the silicon melt and pulled up at a predetermined speed. At this time, a heat shielding member is provided around the silicon single crystal to be pulled up above the surface of the silicon melt. The height between the liquid surface of the silicon melt and the tip of the heat shielding member is very important in controlling the temperature gradient near the solid-liquid interface of the silicon single crystal.

  Patent Documents 3 to 6 specify V / G when the growth rate of a silicon single crystal is V and the temperature gradient in the vicinity of the solid-liquid interface is G, thereby producing a silicon single crystal having excellent defect characteristics. Technology is disclosed.

  Patent Document 7 describes a method for determining a defect in a silicon single crystal wafer manufactured by a CZ method. Here, as defects, for example, COP (Crystal Originated Particle) generated depending on the pulling rate during single crystal growth and the temperature distribution in the single crystal immediately after solidification (temperature gradient in the crystal in the pulling axis direction), An OSF (Oxidation Induced Stacking Fault) that occurs depending on the thermal history of the crystal being grown is mentioned.

  Patent Document 8 discloses a method for producing a silicon single crystal that can efficiently pull up a low-defect silicon single crystal. In this manufacturing method, a pulling method (ADC: automatic diameter control) is performed in which the pulling speed and the heater temperature are changed in accordance with the deviation between the expected crystal diameter and the target crystal diameter.

  Patent Document 9 discloses a silicon single crystal manufacturing method and a manufacturing apparatus thereof that can manufacture a single crystal while suppressing variations in diameter and oxygen concentration without variation due to the characteristics of the apparatus. In this manufacturing method, the center position of the magnetic field for pulling up the single crystal is measured while applying a horizontal magnetic field, and the measured magnetic field center position is measured before and / or during the manufacturing of the single crystal. The pulling-up shaft that is the rotation axis of the single crystal is shifted in the horizontal direction within a range of 2 to 14 mm.

  In Patent Document 10, the position of the melt surface of the silicon melt is accurately controlled by accurately detecting the position of the liquid surface during crystal pulling (constant gap control), and a high level of crystal characteristics desired. A method for producing a silicon single crystal capable of producing a quality silicon single crystal is disclosed. In this manufacturing method, in order to control the V / G with high accuracy so as to obtain a defect-free region, when pulling up at a constant pulling speed, it is arranged so as to face the melt surface and cover a part thereof. The distance Δt from the heat shield member is measured.

  Patent Document 11 discloses a method for producing a single crystal by the Czochralski method capable of producing a high-quality single crystal with few crystal defects, and a single crystal produced by this production method. In this technique, the pulling speed moving average value is controlled within a preset allowable range (pulling speed control) by calculating the upper and lower limit values of the operation amount of the pulling speed of the single crystal.

  Patent Document 12 discloses a defect analysis method for a silicon single crystal that can easily analyze the distribution of crystal defects in the silicon single crystal in the pulling of the silicon single crystal by the MCZ method in which a magnetic field is applied in the horizontal direction. ing. In this technology, by using the physical property value of the silicon melt, the convection of the silicon melt is calculated by a laminar flow model, and the temperature distribution at the time of pulling up the silicon single crystal is predicted. Not crystal).

  Patent Document 13 discloses a silicon single crystal pulling method (simulation technique) in which there are few defects throughout the pulling direction and there is little variation in the defects. In this pulling method, the upper limit pulling rate and the lower limit pulling rate for pulling a silicon single crystal that does not form COP and dislocation clusters are set to vA and vB, respectively, and the pulling rate for pulling a silicon single crystal that does not form COP and dislocation clusters. When the velocity margin of (vA-vB) is taken as (vA-vB), the silicon single crystal is pulled sequentially while feeding back the actual pulling speed with the median speed margin (vA + vB) / 2 as the target pulling speed for each pulling batch of silicon single crystal. I try to raise it.

  Patent Document 14 discloses a process plan planning system, a process plan planning method, and a program for presenting a production plan to a production line having a plurality of production apparatuses that can be selected for each process in a plurality of processes in the manufacture of silicon wafers. (Simulation technology) is disclosed. This process planning system acquires quality information from a database that accumulates quality information for each process path obtained through individual production equipment for each process, and statistically analyzes the quality distribution of silicon wafers obtained by the combination of production equipment. Quality distribution estimation means for estimating the quality, and production apparatus combination determination means for determining a combination of production apparatuses satisfying a quality standard required for a silicon wafer to be manufactured based on a quality distribution obtained by the combination of the production apparatuses And a process plan determining means for determining a process route obtained through a plurality of production apparatuses that can be selected for each process based on the determined combination of production apparatuses and presenting a production plan.

  Patent Document 15 discloses a method for manufacturing a single crystal that can reduce the occurrence of crystal defects, and a method for manufacturing a semiconductor wafer that can efficiently detect and remove a site where a crystal defect has occurred from a single crystal. In this manufacturing method, in the process of pulling up a single crystal by the Czochralski method, based on the detected diameter of the single crystal and the target value of the pulling speed, the setting values for the span and heater temperature that limit the pulling speed operation are calculated. In addition to operating the pulling speed within the span and controlling the diameter of the single crystal by operating the heater temperature to the set value, the fluctuation of the moving average calculated from the actual value of the pulling speed is controlled (lifting Speed control).

  Patent Document 16 discloses a silicon single crystal pulling apparatus that accurately measures the liquid level even when a purge tube is installed inside a thermal radiation shield. This apparatus includes a crucible that supports a silicon melt in a chamber, a heater that heats the silicon melt in the crucible, a heat radiation shield disposed above the crucible, and a non-contact provided inside the heat radiation shield. A substantially cylindrical purge tube that rectifies the active gas, a CCD camera that captures a mirror image of the heat radiation shield reflected on the surface of the silicon melt, and the position of the mirror image of the heat radiation shield. A liquid level calculation unit that calculates the surface level; and a conversion table creation unit that creates a conversion table indicating a relationship between the liquid level of the silicon melt and the position of the mirror image. The liquid level calculation unit calculates the liquid level based on the conversion table (liquid level measurement).

  Patent Document 17 describes manufacturing of a high-quality semiconductor single crystal by reducing the variation in the diameter of the semiconductor single crystal, suppressing the variation in the pulling speed, which is an operation amount for controlling the diameter, and pulling up the semiconductor single crystal as set. A lifting method is disclosed. In this pulling method, a semiconductor raw material is melted by a heater, a semiconductor melt is stored in a crucible, and the semiconductor single crystal is pulled up while controlling the heater based on a preset temperature profile. Then, the past semiconductor single crystal pulling data that contributes to the setting of the heater temperature profile is stored in a database, and the semiconductor semiconductor single crystal heater temperature profile to be pulled next is specified from the past semiconductor single crystal pulling data. Evaluate based on the evaluation function. Based on this specific evaluation function, the temperature profile of the next semiconductor single crystal heater to be pulled up is corrected before pulling up, and the semiconductor single crystal is pulled up while controlling the heater based on this corrected temperature profile. That is, the past performance value of CZ furnace manufacture is applied to the next CZ breeding.

  Patent Document 18 discloses a pulling device that can control the gap between the heat shield member and the melt surface with higher accuracy. In this lifting device, a real image including at least a circular opening of a heat shield member arranged to cover a part of the silicon melt surface, and a mirror image of the heat shield member reflected on the surface of the silicon melt. The liquid surface position of the silicon melt is calculated from the interval between the real image and the mirror image obtained by imaging and the distance Δt between the heat shield member and the liquid surface position is controlled (constant gap control).

  Patent Document 19 discloses a silicon single crystal manufacturing method capable of appropriately controlling the oxygen concentration of a pulled single crystal. In this manufacturing method, when a silicon single crystal is manufactured by the Czochralski method using a pulling device having a heat shield, the void ratio to the crystal diameter (the outer surface of the single crystal and the lower end opening edge of the heat shield The oxygen concentration of the crystal is controlled by adjusting the flow rate (Ar flow) of the inert gas introduced into the apparatus in accordance with the gap area / the cross-sectional area of the single crystal. .

Patent Document 20 discloses a method for manufacturing an epitaxial silicon wafer which is an epitaxial silicon wafer and has a low density of epitaxial defects and excellent gettering capability over the entire radial direction of the wafer. This manufacturing method is applied to a silicon wafer having an oxygen concentration in the range of 9 × 10 ¹⁷ atoms / cm ^{3 to} 16 × 10 ¹⁷ atoms / cm ³ , not including dislocation clusters and COP, and including an oxygen precipitation suppression region. A pre-heat treatment step for performing a heat treatment for increasing the density of oxygen precipitates, and an epitaxial layer forming step for forming an epitaxial layer on the surface of the silicon wafer after the pre-heat treatment step (for detecting COP) Evaluation method).

  In order to accurately control the temperature gradient in the pulling axis direction near the solid-liquid interface (near the silicon ingot and silicon melt interface), it is necessary to stabilize the height of the silicon melt liquid surface and the tip of the heat shield member. There is. On the other hand, it is desirable to make the temperature gradient in the radial direction of the silicon ingot uniform and to substantially eliminate defects in the silicon wafer cut out from the silicon single crystal (ingot). However, when the silica glass crucible is deformed at a high temperature during the CZ pulling, the content changes, so that the liquid surface height of the silicon melt (from the melting of the polycrystalline silicon in the silica glass crucible until the end of the pulling of the silicon single crystal) The liquid surface height of the silicon melt (height H0 in FIG. 17A) also changes. As the liquid level changes, the height between the liquid level and the tip of the heat shielding member also changes, making it difficult to accurately control the temperature gradient. The change in temperature gradient leads to crystal growth where the COP at the solid-liquid interface becomes substantially zero. In other words, insufficient control of the temperature gradient causes defects in the growth of the silicon single crystal. The silicon single crystal (silicon ingot) has a cylindrical shape, and the standard length is 2000 mm and the diameter is 300 mm to 320 mm.

  One factor affecting the quality of the silica glass crucible is the distribution of internal residual stress in the thickness direction (thickness direction) of the silica glass crucible. However, none of the patent documents discloses a technique for accurately measuring the internal residual stress in the thickness direction of the silica glass crucible. For this reason, there is a problem that a silica glass crucible that accurately grasps the relationship between the internal residual stress and the quality of the silicon single crystal when the silicon single crystal is pulled up cannot be provided.

  The present invention has been made in view of such circumstances, a strain measuring device for a silica glass crucible capable of accurately measuring the internal residual stress in the thickness direction of the silica glass crucible, a method for producing a silicon single crystal, A method for measuring strain of a silica glass crucible, a phase difference map, an ingot, and an epitaxial wafer are provided.

  A strain measuring device for a silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. An apparatus for measuring the strain of a silica glass crucible, which is disposed on the side of a side wall, and illuminates polarized light toward the side wall, and captures an image corresponding to the polarized light on the upper end surface of the side wall. An imaging unit; and an output unit that outputs a strain distribution of the silica glass crucible based on an image captured by the imaging unit.

  According to such a structure, the polarized light irradiated to the side wall part of the silica glass crucible from the light emitting part enters the silica glass crucible, and is reflected and diffused inside. At this time, birefringence occurs due to distortion based on the internal residual stress of the silica glass crucible. The birefringent light is captured by the imaging unit from the upper end surface of the side wall. And the distribution of the distortion in the thickness direction of a silica glass crucible is obtained by taking in the picture according to the polarized light of the upper end surface of a side wall part.

  A strain measuring device for a silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. A strain measuring device for measuring strain of a silica glass crucible, which has a pedestal on which a silica glass crucible to be measured is placed, and a light emission that is provided on the pedestal and irradiates at least a side wall portion of the silica glass crucible with polarized light An image capturing unit that captures an image corresponding to the polarization of at least the upper end surface of the silica glass crucible, and an output that outputs a strain distribution of the silica glass crucible based on the image captured by the image capturing unit. A section.

  According to such a configuration, polarized light is irradiated from the light emitting portion to the side wall portion of the silica glass crucible while the silica glass crucible to be measured is placed on the pedestal. The irradiated polarized light enters the silica glass crucible and is reflected and diffused inside. At this time, birefringence occurs due to distortion based on the internal residual stress of the silica glass crucible. The birefringent light is captured by the imaging unit from the upper end surface of the side wall. And the distribution of the distortion in the thickness direction of a silica glass crucible is obtained by taking in the picture according to the polarization of the upper end surface of a side wall part.

  A strain measuring device for a silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. A strain measuring device for measuring strain of a silica glass crucible, comprising a pedestal on which a silica glass crucible to be measured is placed, a light emitting unit that is provided on the pedestal and irradiates polarized light on the silica glass crucible, An imaging unit that is movably provided and captures an image corresponding to the polarized light applied to the silica glass crucible, an output unit that outputs a strain distribution of the silica glass crucible based on the image captured by the imaging unit, a pedestal, and light emission And a controller that controls the imaging unit, and the controller moves the relative position between the silica glass crucible and the imaging unit and repeats the imaging to repeat the imaging. To measure the distortion of the entire circumference worth of Rasurutsubo.

  According to such a configuration, the silica glass crucible is irradiated with polarized light from the light emitting unit in a state where the silica glass crucible to be measured is placed on the pedestal. The irradiated polarized light enters the silica glass crucible and is reflected and diffused inside. At this time, birefringence occurs due to distortion based on the internal residual stress of the silica glass crucible. The birefringent light is captured by the imaging unit. And the distribution of the distortion of the silica glass crucible is obtained by capturing an image corresponding to the polarized light. In measuring the strain of the silica glass crucible using such polarized light, the strain of the entire circumference of the silica glass crucible can be automatically measured under the control of the controller.

  A silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion that is provided between the side wall portion and the bottom portion and has a curvature higher than the curvature of the bottom portion. It was measured by. The strain measuring device is disposed on the side of the side wall, and emits polarized light toward the side wall, an image capturing unit that captures an image corresponding to the polarized light on the upper end surface of the side wall, and an image capturing unit. And an output unit for outputting the strain distribution of the silica glass crucible based on the image. The silica glass crucible has a distribution outputted from the output unit of the strain measuring device as a distribution of the first region provided from the inner surface to the middle in the thickness direction of the side wall portion, and the first region in the thickness direction of the side wall portion. A second region provided outside and having a strain distribution different from that of the first region.

  A silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. The first region is provided from the inner surface to the middle in the thickness direction and has a compressive stress as the internal residual stress, and is provided outside the first region in the thickness direction of the side wall and has the tensile stress as the internal residual stress. A second region, and irradiating toward the reference glass in which distortion is suppressed, and the order of the relative intensities of the center wavelengths of red, green, and blue in the transmitted polarized light is irradiated toward the side wall and transmitted. The order of the relative intensities of the center wavelengths of red, green, and blue in the polarized light is the same.

  According to such a configuration, since the internal residual stress along the thickness direction of the silica glass crucible is balanced, the silica glass crucible is higher in strength than the silica glass crucible having a complicated strain distribution.

  A silica glass crucible according to an embodiment is a silica glass crucible including a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. And it has the 1st field provided from the inner surface to the middle in the thickness direction in the side wall part, and the 1st field has a substantially uniform compressive stress along the inner surface.

  According to such a configuration, since the first region provided on the inner surface side in the thickness direction of the silica glass crucible has a substantially uniform compressive stress along the inner surface, the strain distribution becomes complicated. It becomes a silica glass crucible having higher strength than the silica glass crucible to be assembled.

  A method for measuring strain of a silica glass crucible according to an embodiment includes a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. A strain measuring method for measuring strain of a silica glass crucible, the step of irradiating a silica glass crucible to be measured with polarized light from a light emitting unit, and an image pickup unit according to the polarized light irradiated on the silica glass crucible The silica glass crucible is obtained by repeating the steps of taking in the image, and imaging the predetermined measurement region by the imaging unit, moving the relative position of the silica glass crucible and the imaging unit, and imaging the next measurement region. Measuring the strain of the entire circumference of the side wall portion.

  According to such a configuration, the polarized light applied to the silica glass crucible to be measured enters the silica glass crucible, and is reflected and diffused inside. At this time, birefringence occurs due to distortion based on the internal residual stress of the silica glass crucible. The birefringent light is captured by the imaging unit. And the distribution of the distortion of the silica glass crucible is obtained by capturing an image corresponding to the polarized light. By repeating the measurement of the strain of the silica glass crucible using such polarized light while moving the measurement region, the strain of the entire circumference of the silica glass crucible can be measured.

  A phase difference map according to an embodiment is a silica glass crucible including a cylindrical side wall, a curved bottom, and a corner having a curvature higher than the curvature of the bottom provided between the side wall and the bottom. FIG. 3 is a phase difference map representing a strain distribution, in which a phase difference caused by transmission of polarized light incident on a predetermined region of a silica glass crucible is displayed in gray or numerical values in association with the position of the region. According to such a configuration, the strain distribution of the silica glass crucible can be easily grasped visually.

  The silicon single crystal pulling apparatus according to the embodiment includes the silica glass crucible and a susceptor that covers the outside of the silica glass crucible. The susceptor may be made of carbon. A shielding plate that shields heat may be provided between the inner peripheral surface of the silica glass crucible and the silicon single crystal to be pulled up. According to such a configuration, since the outer side of the high-strength silica glass crucible is covered with the susceptor, a highly reliable pulling device that does not cause cracking of the silica glass crucible in pulling up the silicon single crystal is obtained. .

  The method for producing a silicon single crystal according to the embodiment includes a step of charging a silicon material into the silica glass crucible and melting the step, and a step of pulling up the silicon single crystal from a silicon melt held in the silica glass crucible. Prepare. According to such a configuration, a high-purity silicon single crystal can be pulled up by the reliability of the silica glass crucible.

  An ingot according to an embodiment includes a silica glass crucible including a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. And it is the ingot of the silicon single crystal pulled up using the silica glass crucible measured with the strain measuring device. For example, the strain measuring device is disposed on the side of the side wall, and emits polarized light toward the side wall, an imaging unit that captures an image corresponding to the polarized light on the upper end surface of the side wall, and an imaging unit And an output unit for outputting the strain distribution of the silica glass crucible based on the image captured in step (b). The ingot has a shoulder portion, a straight body portion continuing from the shoulder portion, and a tail portion continuing from the straight body portion, and the crystal defects of the straight body portion are substantially zero. According to such a configuration, it is possible to stabilize electrical characteristics and suppress deterioration of a semiconductor device manufactured using a wafer cut out from an ingot.

  The homoepitaxial wafer according to the embodiment includes a wafer substrate portion using the above-described silicon single crystal ingot and a silicon single crystal homoepitaxial layer provided on the substrate portion.

(A) And (b) is a schematic diagram which illustrates a silica glass crucible. (A) is a partial expanded sectional view of the silica glass crucible concerning this embodiment, (b) is a figure showing an example of internal residual stress. (A)-(c) is a photograph which illustrates distribution of distortion of a silica glass crucible. It is a flowchart which shows schematically the manufacturing process of a silica glass crucible. (A) And (b) is a schematic diagram for demonstrating the manufacturing method of a silica glass crucible. (A) And (b) is a schematic diagram for demonstrating the manufacturing method of a silica glass crucible. It is a schematic diagram which illustrates the distortion measuring apparatus of a silica glass crucible. (A)-(c) is a figure which shows the example of a measurement of distribution of internal residual stress. (A)-(c) is a figure which shows the example which took in the image of the polarized light which permeate | transmitted the side wall part. (A) And (b) is a schematic diagram illustrated about the detection range according to polarization | polarized-light. (A) And (b) is a schematic diagram which illustrates a robot arm type | mold distortion measurement system. (A) And (b) is a schematic diagram which illustrates the strain measuring method by a robot arm type strain measuring system. (A) And (b) is a schematic diagram explaining a measurement area | region. (A) And (b) is a schematic diagram which shows the imaging part which measures phase difference distribution. (A) And (b) is a figure which shows the example of a measurement of phase difference distribution. It is a mimetic diagram showing the whole pulling device composition concerning this embodiment. (A)-(c) is a schematic diagram explaining the manufacturing method of the silicon single crystal using the silica glass crucible which concerns on this embodiment. It is a schematic diagram which illustrates the ingot of a silicon single crystal. (A)-(c) is a schematic diagram explaining raising control. It is a figure which shows the variation | change_quantity of the internal diameter of a crucible. It is a schematic diagram explaining the condition where various defects generate | occur | produce based on the Boronkov theory. It is a schematic diagram which shows the relationship between the pulling speed at the time of single crystal growth, and defect distribution. It is a schematic cross section which illustrates an epitaxial wafer. It is a flowchart which illustrates the process from crucible manufacture to wafer manufacture.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same members are denoted by the same reference numerals, and the description of the members once described is omitted as appropriate.

<Silica glass crucible>
1A and 1B are schematic views illustrating a silica glass crucible.
FIG. 1A shows a perspective view of the silica glass crucible 11, and FIG. 1B shows a cross-sectional view of the silica glass crucible 11.
The silica glass crucible 11 to be measured includes a corner portion 11b having a relatively high curvature, a cylindrical side wall portion 11a having an edge opened on the upper surface, and a mortar shape formed of a straight line or a curve having a relatively low curvature. Bottom portion 11c.

  In the present embodiment, the corner portion 11b is a portion connecting the side wall portion 11a and the bottom portion 11c, and a common tangent line with the bottom portion 11c is formed from the point where the curved tangent line of the corner portion 11b overlaps the side wall portion 11a of the silica glass crucible 11. It means the part up to the point you have. In other words, the point at which the side wall portion 11a of the silica glass crucible 11 begins to bend is the boundary between the side wall portion 11a and the corner portion 11b. Further, a portion where the curvature of the bottom of the silica glass crucible 11 is substantially constant is the bottom portion 11c, and the point where the curvature starts to change when the distance from the center of the bottom of the silica glass crucible 11 increases is the bottom portion 11c. It is a boundary with the corner part 11b.

  In the thickness direction (also referred to as the thickness direction) of the silica glass crucible 11, a transparent layer 13 is provided on the inner surface side, and a non-transparent layer 15 is provided on the outer surface side.

  The transparent layer 13 is a layer that does not substantially contain bubbles. Here, “substantially free of bubbles” means the bubble content and bubble size that do not reduce the single crystallization rate of the silicon single crystal due to the bubbles. For example, the bubble content of the transparent layer 13 is 0.1% or less, and the average diameter of the bubbles is 100 μm or less.

  The transparent layer 13 preferably contains synthetic silica glass on the inner surface side. Synthetic silica glass means silica glass produced by melting raw materials synthesized by hydrolysis of silicon alkoxide, for example. In general, synthetic silica has the characteristics that the concentration of metal impurities is lower and the concentration of OH groups is higher than natural silica. For example, the content of each metal impurity contained in the synthetic silica is less than 0.05 ppm, and the content of OH groups is 30 ppm or more. However, since synthetic silica to which metal impurities such as Al are added is also known, whether or not it is synthetic silica should not be determined based on one element, but comprehensively based on a plurality of elements It should be judged. Thus, since synthetic silica glass has fewer impurities than natural silica glass, it is possible to prevent an increase in impurities eluted from the crucible into the silicon melt, and to increase the silicon single crystallization rate.

  The non-transparent layer 15 contains a large number of bubbles. The non-transparent layer 15 is a layer that appears to be clouded by the bubbles. The non-transparent layer 15 is preferably made of natural silica glass. Natural silica glass means silica glass produced by melting natural raw materials such as natural quartz and silica. In general, natural silica has the characteristics that the concentration of metal impurities is higher and the concentration of OH groups is lower than that of synthetic silica. For example, the content of Al contained in natural silica is 1 ppm or more, the content of alkali metals (Na, K, and Li) is 0.1 ppm or more, and the content of OH groups is less than 60 ppm.

  In addition, it should not be judged based on one element whether it is natural silica, but should be judged comprehensively based on several elements. Since natural silica has a higher viscosity at high temperatures than synthetic silica, the heat resistance of the entire crucible can be increased. Natural raw materials are cheaper than synthetic silica and are advantageous in terms of cost.

FIG. 2A is an enlarged sectional view of a part of the silica glass crucible according to this embodiment, and FIG. 2B is a diagram showing an example of internal residual stress.
FIG. 2A shows an enlarged cross-sectional view of part A shown in FIG. As shown in FIG. 2, the silica glass crucible 11 according to this embodiment includes a first region R1 provided from the inner surface IS to the middle in the thickness direction of the side wall portion 11a, and a first region R1 in the thickness direction of the side wall portion 11a. And a second region R2 provided outside the first region R1. The second region R2 has a strain distribution different from that of the first region R1.

  Here, the first region R1 has a first internal residual stress, and the second region R2 has a second internal residual stress. The internal residual stress is zero at the boundary between the first region R1 and the second region R2.

  Further, the third region R3 may be provided from the outer surface to the middle in the thickness direction of the side wall portion 11a. The third region R3 includes a sintered body of silica and powder. The third region R3 is a portion that is provided on the outermost periphery of the silica glass crucible 11 and remains as a sintered body and powder without vitrification of the silica powder when the silica glass crucible 11 is manufactured. The thickness of the third region R3 is about 0.1 mm to 0.3 mm (thickness corresponding to one diameter of the raw silica powder) if it is thin, and 0.5 mm to 3 mm if it is thick. The degree (the thickness of several raw silica powders).

  FIG. 2B shows the measurement result of the internal residual stress in the thickness direction of the side wall portion 11 a of the silica glass crucible 11. In FIG. 2B, the horizontal axis represents the value of internal residual stress, “+” indicates compressive stress, and “−” indicates tensile stress. In FIG. 2B, the vertical axis indicates the thickness direction of the side wall 11a, “IN” indicates the inner surface side, and “OUT” indicates the outer surface side.

  As described above, in the first region R1 of the silica glass crucible 11, the compressive stress on the inner surface side is the highest, and the compressive stress gradually decreases (with a substantially constant inclination) from there toward the outside. Then, the second region R2 starts at a boundary where the internal residual stress becomes zero. In the second region R2, the internal residual stress becomes tensile stress. Although the tensile stress in the second region R2 changes without a large change, the tensile stress gradually weakens near the outer surface.

Then, when the internal residual stress exceeds a portion where the residual stress becomes zero, it becomes compressive stress again. There is a third region R3 outside the second region R2. The third region R3 is a portion where the silica is not vitrified when the silica glass crucible 11 is manufactured and a portion where the silica powder remains on the outer surface. It cannot be measured.
In the silica glass crucible 11 according to the present embodiment, a layered strain distribution is generated in the thickness direction. In the silica glass crucible 11, the first region R1 and the second region R2 are continuous in the circumferential direction. That is, in each of the first region R1 and the second region R2, a large stress change does not occur at least in the circumferential direction (substantially uniform stress distribution).

  Thus, the silica glass crucible 11 having two types of strain distributions that are relatively different between the inner side and the outer side in the thickness direction of the silica glass crucible 11 has higher strength than a crucible having a complicated strain distribution. It will be. The state where the strain distribution is complicated is a case where the region having the internal residual stress is not layered in the thickness direction or is not continuous in the circumferential direction. As the strain distribution is complicated as described above, cracks and peeling are likely to occur at the boundary of the strain distribution. As described above, the strain distribution has a layer structure in the thickness direction of the silica glass crucible 11 and is continuous in the circumferential direction, so that the boundary of the strain distribution in the surface direction of the silica glass crucible 11 (rapid change) ) And the occurrence of cracks and peeling is suppressed.

  In the silica glass crucible 11, the internal residual stress in the first region R1 is preferably a compressive stress, and the internal residual stress in the second region R2 is preferably a tensile stress. According to such a stress distribution, the strength of the inner surface of the silica glass crucible 11 is improved. When pulling up a silicon single crystal using the silica glass crucible 11, the silica glass crucible 11 is filled with polycrystalline silicon as a material. At this time, an impact is likely to be applied to the inner surface of the silica glass crucible 11. Since the first region R1 has a compressive stress, sufficient resistance to an impact at the time of filling the polycrystalline silicon can be obtained. Further, a bubble layer is provided on the outer surface side in the thickness direction of the silica glass crucible 11. For this reason, even if a crack occurs on the inner surface of the silica glass crucible 11, the growth of the crack is stopped by the bubbles in the bubble layer. Therefore, even if the second region R2 is a tensile stress, the elongation of the crack can be suppressed.

  In the illustrated silica glass crucible 11, an example of a two-layer structure of the first region R1 and the second region R2 is shown, but a structure of three or more layers may be used. Moreover, the layer structure of strain distribution like the 1st area | region R1 and 2nd area | region R2 should just be provided in the transparent layer 13, but it is provided in the whole thickness of the silica glass crucible 11. desirable.

3A to 3C are photographs illustrating the strain distribution of the silica glass crucible.
FIG. 3 shows the result of measuring the strain distribution of a sample obtained by slicing the silica glass crucible 11 to a thickness of about 10 mm. The strain distribution is measured by a photoelastic strain measuring device using circularly polarized light. The white portion (high luminance portion) shown in the photograph is a compressive stress region, and the black portion (low luminance portion) is a tensile stress region.

  A sample (1) shown in FIG. 3A is a silica glass crucible having an outer diameter of 32 inches. In the sample P1, a residual compressive stress layer (first region R1) is provided on the inner side, and a residual tensile stress layer (second region R2) is provided on the outer side. Thus, the intensity | strength of a silica glass crucible improves because it is a layer structure with a beautiful strain distribution.

  A sample P2 shown in FIG. 3B is a silica glass crucible having an outer diameter of 32 inches. In the sample P2, the strain distribution is uniform and does not have a layer structure. Such a silica glass crucible is lower in strength than the silica glass crucible of the sample P1.

  A sample P3 shown in FIG. 3C is a silica glass crucible having an outer diameter of 40 inches. In the sample P3, a layer of residual compressive stress (first region R1) is provided on the inner side. Outside the first region R1, a second region R2 having a strain distribution different from that of the first region R1 although strain is complicated is provided. Sample P3 is a very large silica glass crucible having an outer diameter of 40 inches. Even such an ultra-large silica glass crucible can obtain sufficient strength by having the first region R1 and the second region R2.

  In the case of a large crucible having an outer diameter of 23 inches or more or a super large crucible having a diameter of 40 inches or more, the influence of cracks, cracks, peeling, etc. due to the distribution of internal residual stress in the thickness direction of the silica glass crucible 11 is affected. large. In particular, when increasing the outer diameter of the crucible, the rate of increase in thickness is higher than the rate of increase in outer diameter. That is, the wall thickness tends to be relatively increased with an increase in the outer diameter of the crucible. For this reason, as the outer diameter of the crucible increases, the stress distribution in the thickness direction becomes more complicated, and the strength tends to be insufficient. The distribution of the internal residual stress in the thickness direction of the silica glass crucible 11 as in the present embodiment has a layered structure, which is particularly effective for improving the strength of such large and super large crucibles.

<Method for producing silica glass crucible>
FIG. 4 is a flowchart schematically showing a manufacturing process of the silica glass crucible. 5 and 6 are schematic views for explaining a method for producing a silica glass crucible.
The silica glass crucible 11 is manufactured by a rotational mold method. As shown in FIG. 4, in the rotary mold method, a silica powder layer is formed on a carbon mold (step S101), arc melting and decompression (step S102), cooling (step S103), rim cutting and edge processing (step S104). A silica glass crucible 11 is manufactured.

  First, in the formation of the silica powder layer on the carbon mold shown in step S101, a carbon mold 20 having a cavity matched to the outer shape of the silica glass crucible 11 is prepared as shown in FIG. Then, the first silica powder 201 is supplied while the carbon mold 20 is rotated, scraped off using a scraper, and molded to a predetermined thickness. Thereby, a silica powder layer is formed along the inner surface of the mold. Since the carbon mold 20 rotates at a constant speed, the supplied first silica powder 201 remains in a fixed position while being stuck to the inner surface of the mold by centrifugal force, and the shape thereof is maintained. Since the 1st silica powder 201 becomes a non-transparent layer, it is preferable that it is a natural silica powder.

  Next, as shown in FIG.5 (b), the 2nd silica powder 202 used as the raw material of the transparent layer 13 is supplied in the carbon mold 20 in which the layer of the 1st silica powder was formed, and a silica powder layer is further thickened. Form. The second silica powder 202 is supplied at a predetermined thickness onto the first silica powder 201 on the inner surface of the mold. The second silica powder 202 is preferably a synthetic silica powder, but may be a natural silica powder.

  Next, in the arc melting and depressurization shown in step S102, as shown in FIG. 6A, the arc electrode 30 is installed in the cavity of the carbon mold 20, and the carbon mold 20 is rotated from the inside while rotating. Arc discharge is performed, and the entire silica powder layer is heated to 1720 ° C. or higher and melted. At this time, a thin silica glass sealing layer is formed over the entire circumference. Simultaneously with this heating, the pressure is reduced from the carbon mold 20 side, the gas inside the silica is sucked to the outer layer side through the vent hole 21 provided in the carbon mold 20, and the voids in the silica powder layer being heated are deaerated. Remove bubbles on the inner surface of the crucible. Thereby, the transparent layer 13 substantially free of bubbles is formed.

  The carbon mold 20 is provided with a cooling means (not shown). Thereby, the silica of the part (part used as 3rd area | region R3) used as the outer surface of the silica glass crucible 11 is not vitrified. The cooling temperature by the cooling means is a temperature at which silica remains as a sintered body and powder without vitrification.

  Thereafter, the decompression for deaeration is weakened or stopped while continuing the heating, and the bubbles remain, thereby forming the non-transparent layer 15 containing a large number of minute bubbles.

Next, in the cooling shown in step S103, the power supply to the arc electrode 30 is stopped, the fused silica glass is cooled, and the shape of the silica glass crucible 11 is formed. When cooling is performed, a cooling gas is blown onto the silica glass that is the inner surface of the silica glass crucible 11. The distribution of the internal residual stress of the silica glass crucible 11 is determined by the cooling conditions such as the cooling rate and the cooling gas blowing method.
At this time, the silica glass crucible 11 having a desired strain distribution can be manufactured by adjusting the cooling conditions based on a database of strain distribution measured in advance.

Here, when the silica glass is cooled, pressure is applied to the silica glass crucible 11 due to the difference in thermal shrinkage between the silica glass crucible 11 and the carbon mold 20. For example, the linear expansion coefficient of silica glass is about 10 ⁻⁷ / K, and the silica glass crucible 11 having a total length of 0.01% at 1000 ° C., that is, a diameter of 1 m, shrinks by about 0.1 mm. On the other hand, the linear expansion coefficient of carbon is about 10 ⁻⁶ / K, and if the inner diameter is 1 m, it will shrink by about 1 mm.

  During this cooling, the sintered body and powder of silica in the third region R3 provided on the outer surface of the silica glass crucible 11 serve as a cushion. That is, when the silica glass is all vitrified, the silica glass crucible 11 directly receives the pressure due to the thermal shrinkage of the carbon mold 20 during cooling, but the outer surface of the silica glass crucible 11 in contact with the carbon mold 20. In addition, since the third region R3 (silica sintered body and powder) is present, this serves as a cushion and can relieve pressure from the carbon mold 20. By relieving the pressure from the carbon mold 20, it is possible to prevent cracks from forming in the crucible wall during cooling, and to suppress deformation of the silica glass crucible 11 when pulling up the silicon single crystal. .

  When the silica glass is cooled, the cooling gas may be blown while measuring the temperature of the inner surface of the silica glass crucible 11 using, for example, a two-dimensional thermograph. In this case, by measuring the temperature of the inner surface of the rotating silica glass crucible 11, the temperature of the entire inner surface can be measured by repeating the temperature measurement in a predetermined region. Moreover, the temperature of a predetermined area | region may be measured and a space | interval may be calculated by simulation. Moreover, you may measure the temperature of the corner part 11b especially important for cooling of silica glass. Further, the temperature of the entire inner surface may be measured at once by a two-dimensional thermograph having the entire inner surface of the silica glass crucible 11 as a measurement range. The silica glass crucible 11 having the first region R1 and the second region R2 is configured by controlling the amount, range, time, etc. of the cooling gas while observing the temperature of the portion that becomes the inner surface of the silica glass crucible 11. .

  In the rim cutting and edge processing shown in step S104, as shown in FIG. 6B, a part of the upper end side of the side wall portion 11a of the silica glass crucible 11 taken out from the carbon mold 20 is cut to obtain the silica glass crucible 11. Adjust the height. Thereafter, the inner peripheral edge and the outer peripheral edge, which are the edges of the upper end surface TP, are chamfered to form a chamfered portion C. A vacuum adsorber used for transporting the silica glass crucible 11 is attached to the upper end surface TP of the silica glass crucible 11. Therefore, the flatness required for vacuum suction is required for the upper end surface TP. Further, by providing the chamfered portion C, burrs and the like protruding from the upper end surface TP are removed, and the flatness is improved. As a result, the suction force by vacuum suction is improved, and cracks remaining at the corners of the upper end surface TP are removed, and cracking and deformation at the time of crystal pulling are prevented.

  In rim cutting, a diamond cutter is applied at a right angle to the central axis of the silica glass crucible 11, but a thin diamond cutter tends to bend in the direction of bubbles and is not necessarily cut at a right angle. Further, the upper end surface TP may be chipped during rim cutting.

The weight of the silica glass crucible 11 is about 50 kg to 60 kg for a 32 inch type (diameter about 81.3 cm), about 80 kg to 90 kg for a 36 inch type (about 91.4 cm diameter), and 40 inch type (about 101.6 cm diameter). It becomes about 90kg-110kg. Further, when the silica glass crucible 11 is filled with polycrystalline silicon, the size is about 300 kg to 500 kg for the 32 inch type, about 400 kg to 800 kg for the 36 inch type, and about 500 kg to 1000 kg for the 40 inch type. Therefore, the flatness and flatness necessary for vacuum-sucking the silica glass crucible 11 having such a weight are required for the upper end surface TP formed in the rim cut.
Further, the thickness of the silica glass crucible 11 is 10 mm to 15 mm and is not uniform, and the thickness varies depending on the part. That is, when the silica glass crucible 11 is manufactured, the outer side of the crucible is constrained by the carbon mold, so that it can be manufactured relatively as designed. However, the inside of the crucible is not constrained by a carbon mold, and the ceramic is arc-melted at an extremely high temperature, so that it is difficult to manufacture according to the design drawing like machining. For example, the inside of the crucible often has a shape that undulates in the height direction. Therefore, it is very difficult to stabilize the manufacturing variation for each product of the silica glass crucible 11.

  In the method for producing the silica glass crucible 11, the bond between Si and O by energy is cut and the bond between Si and O by cooling is performed at the stage of melting the silica powder. Possible ways of breaking the bond between Si and O by energy include cutting by thermal energy, cutting by light energy, and cutting by radicals generated by an arc. Furthermore, the cutting method varies depending on the raw material of the silica powder. For example, if it is natural silica powder, it varies depending on the place of production, and if it is synthetic silica powder, it varies depending on the synthesis method.

  Further, the way of bonding Si and O varies depending on the material and the cooling method. For example, the bonding state of Si—O such as a 6-membered ring or an 8-membered ring changes depending on the method for cooling the fused silica powder. Further, since the bonding state of Si and O changes depending on various conditions such as a difference in materials between the first silica powder and the second silica powder and a difference in cooling rate between the inner side and the outer side, the manufactured silica glass crucible 11 is used. The distribution of internal residual stress will also change.

In the method for manufacturing the silica glass crucible, an annealing process may be performed between the cooling (step S103) and the rim cutting and edge processing (step S104).
In the annealing treatment, the silica glass crucible 11 is annealed under heating conditions corresponding to the relationship between the crucible manufacturing conditions and strain acquired in advance. That is, the strain is measured in advance for the silica glass crucible 11 manufactured under the same conditions. For example, the strain distribution of the silica glass crucible 11 is measured by a strain measuring device 100 or a robot arm type strain measuring system 200 described later and stored in a database. And about the silica glass crucible 11 in which arc melting and cooling have been completed, the annealing temperature and time are determined based on the strain distribution stored in this database, and the silica glass crucible 11 is annealed based on the determined conditions. .

Here, an example of an annealing process using an electric furnace for strain reduction will be described. The annealing treatment is appropriately set within the following conditions.
Temperature increase rate: 1 hour to several tens of hours (100 ° C./hour to 1000 ° C./hour) from room temperature to a predetermined temperature (temperature between the strain point and the annealing point of quartz glass)
Holding temperature: strain point (about 1100 ° C.) to annealing point (about 1200 ° C.)
Holding time: about 5 minutes to 2 hours Temperature decrease rate: 0.5 ° C./min to 1 ° C./min

In the above step S104, the annealing condition is changed based on the strain distribution stored in the database in such an annealing process.
For example, when the strain on the inner surface of the silica glass crucible 11 stored in the database is larger than a predetermined reference, the following annealing process is performed.
Holding temperature: 1150 ° C
Holding time: 1 hour Temperature drop rate: 1 ° C / min

Further, when the strain on the inner surface of the silica glass crucible 11 stored in the database is smaller than a predetermined reference, the following annealing process is performed.
Holding temperature: 1150 ° C
Holding time: 15 minutes Temperature decrease rate: 1 ° C / min

  Further, when the strain on the inner surface of the silica glass crucible 11 stored in the database is the same as a predetermined standard, annealing is not performed.

  That is, in this example, the holding time is lengthened as the distortion is larger than the reference. On the other hand, the holding time is shortened as the distortion is smaller than the reference. Note that the correspondence between the strain and the annealing condition is an example, and the annealing condition may be adjusted in order to obtain a desired strain.

  As described above, the silica glass crucible 11 having a desired strain distribution can be manufactured by adjusting the annealing conditions based on the strain distribution database measured in advance.

<Strain measuring device for silica glass crucible>
FIG. 7 is a schematic view illustrating a strain measuring device for a silica glass crucible.
The strain measuring device 100 is a device that measures the strain of the silica glass crucible 11 as described above in a nondestructive manner. The strain measuring apparatus 100 includes a light emitting unit 110, an imaging unit 120, and an output unit 130.

  The light emitting unit 110 is disposed outside the side wall portion 11a of the silica glass crucible 11, and irradiates polarized light toward the side wall portion 11a. The light emitting unit 110 may be disposed inside the silica glass crucible 11. The light emitting unit 110 includes a light source 111, first polarizing means 112, and second polarizing means 113. The light source 111 is a white light source that emits white light, for example. The light source 111 may be a monochromatic light source that emits monochromatic light having a predetermined wavelength such as infrared light, an LED light source that emits LED light, or a laser light source that emits laser light.

  The first polarization unit 112 extracts a linearly polarized light component from the light emitted from the light source 111. The second polarizing means 113 converts the light of the linearly polarized light component extracted via the first polarizing means 112 into light of the rotationally polarized light component. In the present embodiment, the second polarizing means 113 converts linearly polarized light into circularly polarized light. The first polarizing means 112 is arranged between the light source 111 and the side wall part 11a, and the second polarizing means 113 is arranged between the first polarizing means 112 and the side wall part 11a. Thereby, the light emitted from the light source 111 becomes linearly polarized light via the first polarizing means 112 and becomes circularly polarized light via the second polarizing means 113. The side wall portion 11a is irradiated with circularly polarized light.

  As the first polarizing means 112 and the second polarizing means 113, a transmission type or a reflection type is used. In the present embodiment, a transmissive polarizing plate is used as the first polarizing means 112. Further, a transmission type λ / 4 plate is used as the second polarizing means 113.

  The imaging unit 120 captures an image corresponding to the polarization of the upper end surface TP of the side wall part 11 a of the silica glass crucible 11. The imaging unit 120 is disposed, for example, above the side wall 11a, enters the side wall 11a from the light emitting unit 110, and takes in the polarized light reflected and diffused inside.

  The imaging unit 120 includes, for example, a light receiving unit 121, a third polarizing unit 122, and a fourth polarizing unit 123. The light receiving unit 121 includes a light receiving element such as a CCD or a CMOS. The third polarizing means 122 is an optical means that is provided between the light receiving part 121 and the upper end surface TP of the side wall part 11a and transmits more light of the rotationally polarized light component than the linearly polarized light component. The third polarizing means 122 uses the same λ / 4 plate as the second polarizing means 113.

  The fourth polarizing unit 123 is disposed between the third polarizing unit 122 and the light receiving unit 121. The fourth polarizing unit 123 is an optical unit that converts the circularly polarized light that has passed through the third polarizing unit 122 into linearly polarized light.

  The output unit 130 outputs the strain distribution of the silica glass crucible 11 based on the video captured by the imaging unit 120. The output unit 130 receives a signal based on the image captured by the light receiving unit 121 of the imaging unit 120, and performs conversion so that the distortion distribution of the silica glass crucible 11 is represented by the density and color of the image.

  In order to measure the strain of the silica glass crucible 11 by the strain measuring apparatus 100 according to the present embodiment, first, the light emitting unit 110 is disposed on the side (for example, the outside) of the side wall 11a of the silica glass crucible 11, and the side wall The imaging unit 120 is disposed above the upper end surface TP of 11a. Next, light (for example, white light) is emitted from the light source 111 of the light emitting unit 110. The light emitted from the light source 111 is converted to linearly polarized light by passing through the first polarizing means, and is converted to circularly polarized light by passing through the second polarizing means 113.

  This circularly polarized light enters the inside from the side wall portion 11a of the silica glass crucible 11, and is reflected and diffused. Here, if the silica glass crucible 11 has an internal residual stress, birefringence based on this stress occurs. Therefore, when circularly polarized light is reflected and diffused inside the silica glass crucible 11, a phase difference is generated due to birefringence. The change in the polarization state due to the phase difference is determined by the distribution of the internal residual stress of the silica glass crucible 11.

  Next, the imaging unit 120 captures light on the upper end surface TP side of the side wall 11 a of the silica glass crucible 11. This image includes light that has been reflected and diffused inside the silica glass crucible 11 to cause a change in the polarization state and has exited from the upper end surface TP. In the imaging unit 120, the circularly polarized component is extracted from this light by the third polarizing unit 122, converted into linearly polarized light by the fourth polarizing unit 123, and captured by the light receiving unit 121.

  Then, a signal based on the light captured by the light receiving unit 121 is processed by the output unit 130 and output as a distribution of internal residual stress of the silica glass crucible 11. Specifically, the amount of received light increases as the circularly polarized component in the light traveling toward the imaging unit 120 increases. That is, since the polarization state of the circularly polarized light incident from the side is changed due to the birefringence based on the distortion of the silica glass crucible 11, the circularly polarized light corresponding to the imaging region of the upper end surface TP of the silica glass crucible 11 by the imaging unit 120. By obtaining the amount, a strain distribution based on the internal residual stress along the thickness direction of the silica glass crucible 11 can be obtained.

  In the present embodiment, the image on the upper end surface TP side of the silica glass crucible 11 is captured by the imaging unit 120, and thus the distribution corresponding to the accumulation of the upper strain from the position of the light emitting unit 110 of the side wall portion 11 a of the silica glass crucible 11. Is obtained. Accordingly, the amount of accumulated distortion changes by changing the height of the light emitting unit 110. For example, when the height of the light emitting unit 110 is closer to the upper end surface TP (higher position), the accumulated amount of distortion is reduced. Conversely, when the height is closer to the bottom 11c of the light emitting unit 110 (lower position), distortion is reduced. Cumulative amount increases.

  By utilizing such characteristics, the distribution of strain at each position (each height) may be measured while moving the height of the light emitting unit 110 up and down. In addition, by using the strain distribution measured at each height, the difference between the strain distributions at two predetermined heights is obtained to obtain a cumulative strain distribution between the two heights. You can also.

8A to 8C are diagrams for explaining the distribution of internal residual stress.
FIG. 8A shows the measurement result of the strain distribution as viewed from the upper end surface side of the silica glass crucible 11. According to this measurement result, it turns out that 1st area | region R1 and 2nd area | region R2 are provided along the thickness direction of the silica glass crucible 11. FIG. That is, the first region R1 is provided from the inner surface of the side wall portion 11a of the silica glass crucible 11 to a predetermined position in the thickness direction, and the second region R2 is from a predetermined position in the thickness direction of the side wall portion 11a to the outer surface. Is provided. The thickness of each strain distribution is substantially constant, and the silica glass crucible 11 in which defects such as cracks, cracks, and peeling are suppressed is provided by being continuously provided in the circumferential direction.

  FIG. 8B illustrates the stress balance in the line SL shown in FIG. Here, “+” indicates compressive stress, and “−” indicates tensile stress. As the stress balance along the line SL in the thickness direction of the silica glass crucible 11, the first region R1 is a compressive stress and the second region R2 is a tensile stress. That is, in the thickness direction of the side wall portion 11a, the internal residual stress changes from the compressive stress to the tensile stress through the boundary region where the internal residual stress becomes zero from the first region R1 to the second region R2. The internal residual stress in the thickness direction of the side wall portion 11a is thinner in the first region R1 having compressive stress than in the second region R2 having tensile stress.

The silica glass crucible 11 has a cylindrical shape with a closed bottom, and has a transparent layer (inside the crucible) and a non-transparent layer (outside the crucible), from room temperature to high temperature (about 1500 ° C. to about 1600 ° C.). Used in the environment. In other words, it is different from glass that is flat and entirely transparent, such as automotive glass, and is used only at room temperature. Further, the silica glass crucible 11 has a bubble layer and external irregularities, and is completely different from transparent and flat glass for automobiles and the like. In such a silica glass crucible 11, it is very difficult to measure the polarization state.
As in the present embodiment, the first region R1 which is the inner side in the thickness direction of the silica glass crucible 11 is compressive stress, and the second region R2 which is the outer side through the boundary region (the region where the internal residual stress is zero) is the tensile stress. As a result, the mutual stress is relieved by the temperature of the silica glass crucible 11 at the time of pulling the silicon single crystal (about 1500 ° C. to 1600 ° C.).
Such stress relaxation occurs in the silica glass crucible 11 having a closed bottom, and deformation of the silica glass crucible 11 (falling of the side wall portion 11a, distortion, rising of the bottom portion 11c, etc.) occurs when the silicon single crystal is pulled up. It is suppressed.

In order to accurately control the temperature gradient in the pulling axis direction near the solid-liquid interface (near the silicon ingot and silicon melt interface) when pulling up the silicon single crystal (silicon ingot), the liquid surface and heat of the silicon melt are controlled. It is necessary to stabilize the height from the tip of the shielding member. On the other hand, it is desirable to make the temperature gradient in the radial direction of the silicon ingot uniform and to substantially eliminate defects in the silicon wafer cut out from the silicon single crystal (ingot).
At the time of pulling up the silicon single crystal, the interface between the silicon melt at the center of the silica glass crucible 11 and the silicon single crystal is determined from the temperature (about 1500 ° C.) of the peripheral edge of the silicon melt in the silica glass crucible 11. It is necessary to reliably heat the silicon melt to a temperature required for pulling up the silicon single crystal up to a temperature (hereinafter referred to as “solid-liquid interface”) (about 1420 ° C.).
Further, since the distance between the heater and the center of the silica glass crucible 11 increases as the diameter of the silica glass crucible 11 becomes larger than 32 inches, a stronger heater is used, and the temperature of the silica glass crucible 11 (from about 1500 ° C.). About 1600 ° C.).
Therefore, as the diameter of the silica glass crucible 11 increases, the deformation resistance characteristic of the silica glass crucible necessary for suppressing deformation at high temperatures becomes severer as described later.

  On the other hand, when the silica glass crucible 11 is filled with polycrystalline silicon, the corner of the polycrystalline silicon lump hits the inner surface of the silica glass crucible 11 at normal temperature. If the strength of the silica glass crucible 11 at room temperature is insufficient, cracking or chipping will occur when filling polycrystalline silicon.

  As in the present embodiment, the first region R1 which is the inner side in the thickness direction of the silica glass crucible 11 is compressive stress, and the second region R2 which is the outer side through the boundary region (the region where the internal residual stress is zero) is the tensile stress. As a result, the strength at normal temperature can be ensured, and cracks and chips during filling of polycrystalline silicon can be suppressed.

  As described above, the silica glass crucible 11 needs to have both the properties required at normal temperature and the properties required at a high temperature (about 1500 ° C. to 1600 ° C.) when pulling up the silicon single crystal (silicon ingot).

  Since the silica glass crucible 11 according to the present embodiment has both the characteristics required at normal temperature and the characteristics required at high temperatures (about 1500 ° C. to 1600 ° C.), the first on the line SL It is desirable that the sum of the compressive stress in the first region R1 and the tensile stress in the second region R2 is zero. Furthermore, the line SL is moved in the vertical direction of the silica glass crucible 11, and the compressive stress in the first region R1 at each height position of the side wall portion 11a, the corner portion 11b, and the bottom portion 11c, and the second region R2 It is more preferable that the sum total with the tensile stress is zero. Thus, since the sum of the compressive stress and the tensile stress is zero, the strength of the silica glass crucible 11 at room temperature can be secured, and cracks and chips during filling of polycrystalline silicon can be suppressed. The balance is maintained so that the compression and the tension in the internal residual stress in the thickness direction when heated can cancel each other, and the collapse of the shape of the silica glass crucible 11 can be suppressed.

  FIG. 8C shows an example of an image of the upper end surface TP of the silica glass crucible 11 captured by the strain measuring device 100. By using polarized light, the strain distribution in the thickness direction of the side wall portion 11a of the silica glass crucible 11 can be accurately grasped by the density of the image. In the image shown in FIG. 8C, a state where the first region R1 having compressive stress is provided on the inner side and the second region R2 having tensile stress is provided on the outer side clearly appears. Further, the third region R3 outside the second region R2 can also be grasped from this image.

Here, when the sum of the compressive stress and the tensile stress along the thickness direction (thickness direction) in the side wall portion 11a is zero, the polarized light that is irradiated and transmitted toward the reference glass (blank) in which the strain is suppressed. The order of the relative intensities of the center wavelengths of red, green, and blue is the same as the order of the relative intensities of the center wavelengths of red, green, and blue in the polarized light that is irradiated and transmitted toward the side wall 11a. Including.
If there is no polarized light of any of red, green, and blue among the polarized light to be irradiated, that wavelength is excluded from the order.

As a reference glass, a sample piece (for example, 10 cm × 10 cm) of a side wall portion of a silica glass crucible manufactured under the same conditions as the silica glass crucible 11 to be measured is prepared, and annealed at 1200 ° C. for about 24 hours, for example. It was confirmed that the distortion was suppressed (substantially no distortion) by cross-sectional observation or the like. Then, the reference glass is irradiated with polarized light, and the order of the relative intensities of the center wavelengths of red, green, and blue when transmitted is measured.
Relative intensities are measured with I1 (R, G, B) as red, green, and blue intensities of light immediately before passing through the polarizing filter from the light source and entering from the outside (or inside) of the silica glass crucible 11. When the intensity of each of the red, green, and blue polarized light transmitted through the object is I2 (R, G, B), it is expressed as I2 (R, G, B) / I1 (R, G, B). The

  Then, the polarized light irradiated toward the side wall portion 11a is divided into red, green, and blue components, and the order of the relative intensities of the respective center wavelengths is obtained, and the red, green, and blue colors after passing through the reference glass are obtained. If the relative intensity order does not change, the sum is assumed to be zero.

  For example, if the relative intensity is high in the order of red, green, and blue as the polarized light applied to the side wall portion 11a, the light is yellow as a whole. When this polarized light is applied to the side wall portion 11a and the transmitted polarized light is viewed, it is the same yellow if the relative intensity does not change, and there is little color change if the relative intensity order does not change. If it is this grade, it will become the high intensity | strength silica glass crucible 11, assuming that the sum total of the compressive stress and the tensile stress along the thickness direction of the side wall part 11a is zero.

FIGS. 9A to 9C are diagrams illustrating an example in which an image of polarized light transmitted through the side wall portion is captured.
In any of the diagrams shown in FIGS. 9A to 9C, polarized light whose relative intensities at the center wavelengths of red, green, and blue are in a specific order is incident on the side wall 11a.

  In the example shown in FIGS. 9A and 9B, the order of the relative intensities of the transmitted polarized red, green, and blue is the same as the order of irradiation. In these examples, the sum of the compressive stress and the tensile stress along the thickness direction of the side wall 11a is zero in such a state. In addition, it can be seen that the color is almost constant in the measurement region, and the total stress is zero over the entire measurement region.

  On the other hand, in the example shown in FIG. 9C, the order of the relative intensities of the transmitted polarized red, green, and blue is different from the order when the reference glass is transmitted. In such a state, it cannot be said that the sum of the compressive stress and the tensile stress along the thickness direction of the side wall portion 11a is zero. In addition, it can be seen that a change in color occurs in the measurement region, and a change occurs in the stress sum in the plane of the measurement region.

  In the strain measurement apparatus 100, when an image corresponding to the polarization of the upper end surface is captured by the imaging unit 120, a mechanism is provided that enables adjustment of the incident angle of light irradiated from the light emitting unit 110 to the side wall portion 11a of the silica glass crucible 11. Also good. That is, in the reflection (diffusion) of the light incident on the side wall portion 11a from the light emitting unit 110, the side wall portion 11a is made to have a Brewster angle with respect to the surface of the side wall portion 11a. The reflectance of p-polarized light at the surface of the film can be made zero. As a result, a large amount of linearly polarized light component can be incident on the side wall portion 11a, and an image of the internal residual stress distribution of the side wall portion 11a can be clearly obtained while suppressing the incidence of unnecessary polarized light components.

  As described above, according to the strain measuring apparatus 100 according to the present embodiment, it is possible to accurately measure the strain distribution in the thickness direction of the silica glass crucible 11. Further, the strain measuring apparatus 100 can measure the strain in the thickness direction without destroying the silica glass crucible 11, and can inspect the distortion of the product itself.

  In this embodiment, optical means for converting linearly polarized light into circularly polarized light is used as the second polarizing means 113 and the third polarizing means 122, but optical means for converting linearly polarized light into elliptically polarized light may be used. By using elliptically polarized light, the light detection range in the light receiving unit 121 can be set.

FIGS. 10A and 10B are schematic views illustrating the detection range corresponding to the polarized light.
FIG. 10A shows a detection range when circularly polarized light is used, and FIG. 10B shows a detection range when elliptically polarized light is used.
In the example shown in FIGS. 9A and 9B, the internal residual stress of the silica glass crucible 11 is measured with the amount of light (corresponding to the length of the arrow in FIG. 10) using the polarizing plate in the linear direction of the arrow in the figure. .
In this case, the difference in the magnitude of the internal residual stress cannot be distinguished because the amount of light is the same for the phase difference (π / 16 = −0.19) and the phase difference +0.19, which are elliptically polarized light close to linearly polarized light. Therefore, when the circularly polarized light (π / 4 = 0.78) shown in FIG. 10A is used, the detection range is a range from the phase difference +0.78 to the phase difference 0. That is, the light incident on the silica glass crucible 11 without any internal residual stress is not polarized (circularly polarized), but is polarized by the internal residual stress to become linearly polarized (the phase difference is 0 in FIG. 10A). Can be measured up to the state of linearly polarized light).

  On the other hand, when the elliptically polarized light shown in FIG. 10B is used (in FIG. 10B, for example, the phase difference ((7/16) π = 1.37)), the detection range is the phase difference +1.37 to about The range is up to zero phase difference. That is, it is possible to measure from a state where the incident light having no internal residual stress to the silica glass crucible 11 is not polarized (still elliptically polarized) to a state where it is polarized by the internal residual stress and becomes linearly polarized light. Thereby, when the elliptically polarized light is used, the measurement range of the internal residual stress becomes wider than when the circularly polarized light is used.

Furthermore, the following effects can be obtained by using elliptically polarized light.
When circularly polarized light is irradiated from the air to the inner surface of the silica glass crucible 11, the reflectance of s-polarized light and p-polarized light may differ depending on the incident angle to the surface of the silica glass crucible 11. If the reflectances of s-polarized light and p-polarized light are different, one of the s-polarized component and the p-polarized component is reflected more on the surface of the silica glass crucible 11 than the other, and attenuates when entering the silica glass crucible 11. As a result, the circularly polarized state collapses and becomes elliptically polarized light, and enters the silica glass crucible 11. If measurement is performed as it is, there is a possibility that even a portion without distortion may be observed as if there is distortion. For this reason, the silica glass crucible 11 is irradiated with elliptically polarized light in which one of the s-polarized component and the p-polarized component is larger than the other in consideration of the attenuation due to reflection in advance. As a result, the polarization state after entering the silica glass crucible 11 becomes circularly polarized light, and it becomes possible to accurately measure the original internal residual stress.
If the magnitude of the internal residual stress of the silica glass crucible 11 can be measured, the deformation of the silica glass crucible 11 can be correctly predicted.

<Robot arm type strain measurement system>
FIGS. 11A and 11B are schematic views illustrating a robot arm type strain measurement system.
As shown in FIG. 11A, a robot arm type strain measurement system 200 is provided with an articulated robot arm 210, an imaging unit 120 attached to the robot arm 210, and a silica glass crucible 11 to be measured. A gantry 220, a light emitting unit 110, a controller 250, and an output unit 130 are provided.

  The gantry 220 includes a horizontal gantry 221 and a vertical gantry 222 attached substantially perpendicular to the horizontal gantry 221. The horizontal platform 221 is provided with a lower irradiation unit 110A of the light emitting unit 110, and the vertical platform 222 is provided with a side irradiation unit 110B of the light emitting unit 110. A light source 111, a first polarizing unit 112, and a second polarizing unit 113 are provided in each of the lower irradiation unit 110A and the side irradiation unit 110B. Between the light source 111 and the first polarizing means 112, a diffusion plate 115 that diffuses the light emitted from the light source 111 may be provided. 110 A of lower irradiation parts have the length more than the outer diameter of the silica glass crucible 11 of a measuring object, and the side irradiation parts 110B have the length more than the height of the silica glass crucible 11 of a measuring object.

  A slide rail 223 may be provided on the horizontal base 221 of the base 220. The slide rail 223 is provided with a pedestal 224 that can move horizontally along the slide rail 223. The silica glass crucible 11 to be measured is placed on the pedestal 224 and moved to the measurement position along the slide rail 223.

  The controller 250 controls the operation of the robot arm 210. The controller 250 controls the position of the robot arm 210 using the design data (CAD data or the like) of the silica glass crucible to be measured, and controls the imaging area by the imaging unit 120. Here, as the CAD data, the outer diameter, the inner diameter, and the height of the crucible (the height from the bottom portion 11c of the crucible to the upper end surface TP, the height of the side wall portion 11a), the wall thickness, and the curvature (from the bottom portion 11c to the corner portion 11b). 3 dimensional coordinate data (crucible outer surface, inner surface, rim end face, finite element mesh, polygon data, etc.). By using the design data, the positional relationship between the imaging unit 120 and the measurement location of the silica glass crucible can be accurately set. The controller 250 may be provided with the output unit 130.

<Strain measurement method>
12A and 12B are schematic views illustrating a strain measurement method using a robot arm strain measurement system.
FIG. 12A shows an example in which the imaging unit 120 is disposed on the upper end surface TP side of the silica glass crucible 11 to perform strain measurement. First, the controller 250 controls the position of the pedestal 224 to place the silica glass crucible 11 at the measurement reference position. Next, the controller 250 controls the robot arm 210 to place the imaging unit 120 above the upper end surface TP of the silica glass crucible 11. Then, the imaging direction by the imaging unit 120 is set downward.

  In this state, light is irradiated from the side irradiation unit 110 </ b> B toward the side wall part 11 a of the silica glass crucible 11, and an image of the upper end surface TP of the silica glass crucible 11 is captured by the imaging unit 120. A signal based on the video captured by the imaging unit 120 is sent to the output unit 130. The output unit 130 outputs a distortion distribution based on the signal sent from the imaging unit 120. Thereby, for example, a strain distribution as shown in FIG. 8 can be obtained.

  The controller 250 rotates the pedestal 224 by a certain amount and rotates the silica glass crucible 11 by a predetermined angle after the strain measurement at one place is completed. In this state, the image of the upper end surface TP is taken in by the imaging unit 120 and the distortion distribution is output by the output unit 130 in the same manner as before. By repeating this operation, a strain distribution for one round of the upper end surface TP of the silica glass crucible 11 can be automatically obtained.

  The controller 250 may irradiate all of the light sources 111 arranged in the vertical direction of the side irradiation unit 110B with the same light amount, or the light amount of any one of the light sources 111 may be more than the light amount of the other light source 111. You may make it increase. Adjusting the light amounts of the plurality of light sources 111 arranged in the vertical direction individually is equivalent to adjusting the vertical position of the light source 111 in a pseudo manner.

  FIG. 12B shows an example in which the imaging unit 120 is arranged on the inner surface IS side of the silica glass crucible 11 to perform strain measurement. First, the controller 250 controls the position of the pedestal 224 to place the silica glass crucible 11 at the measurement reference position. Next, the controller 250 controls the robot arm 210 to place the imaging unit 120 at a predetermined height inside the silica glass crucible 11. Then, the imaging direction by the imaging unit 120 is adjusted so as to face the inner surface IS.

  In this state, light is irradiated from the lower irradiation unit 110 </ b> A and the side irradiation unit 110 </ b> B toward the silica glass crucible 11, and an image of light transmitted through the silica glass crucible 11 is captured by the imaging unit 120. A signal based on the video captured by the imaging unit 120 is sent to the output unit 130. The output unit 130 outputs a distortion distribution based on the signal sent from the imaging unit 120. Thereby, the strain distribution seen from the inner surface IS side of the silica glass crucible 11 can be obtained.

  After the distortion measurement at one location is completed, the controller 250 controls the robot arm 210 to adjust the imaging area of the imaging unit 120 to be an area adjacent to the lower side of the previous imaging area. In this state, the image of the upper end surface TP is captured by the imaging unit 120 and the distortion distribution is output by the output unit 130 in the same manner as described above. By repeating this operation, as shown in FIG. 13A, it is possible to obtain a strain distribution for one column of the inner surface IS of the silica glass crucible 11.

  As an example, when the size of one measurement region MR is 100 mm × 100 mm, if the length from the upper end surface TP of the side wall portion 11a to the center of the bottom portion 11c is 790 mm, the length of one column is about eight. A measurement area MR is assigned. In other words, one column can be measured with 8 images.

  After measuring the strain distribution for one vertical column, the controller 250 rotates the pedestal 224 by a certain amount to rotate the silica glass crucible 11 by a predetermined angle. In this state, a distortion distribution for one vertical column adjacent to the inner surface IS is obtained by the imaging unit 120 as before. By repeating this operation, as shown in FIG. 13B, a strain distribution for the entire circumference of the inner surface IS of the silica glass crucible 11 can be automatically obtained. By obtaining the strain distribution for the entire circumference of the inner surface IS of the silica glass crucible 11, it is also possible to grasp the local strain distribution change in the silica glass crucible 11.

  As an example, when the circumference of the inner surface IS of the silica glass crucible 11 is 2450 mm, about 25 measurement regions MR are assigned in the circumferential direction. That is, since about 200 measurement regions MR are allocated on the entire circumference of the inner surface IS, the distribution of distortion for the entire circumference can be measured with about 200 images. For example, if it takes 10 seconds to acquire one image, the entire distortion of the inner surface IS of the silica glass crucible 11 can be automatically acquired in about 34 minutes.

  In the strain measurement by the robot arm strain measurement system 200 as described above, the robot arm 210 and the imaging unit 120 are controlled by controlling the positions of the robot arm 210 and the imaging unit 120 using the CAD data of the silica glass crucible 11 and the like. The positional relationship between the imaging unit 120 and the measurement region can be set accurately and quickly while avoiding interference between the unit 120 and the silica glass crucible 11.

<Phase difference distribution measurement>
14A and 14B are schematic diagrams illustrating an example of an imaging unit that measures the phase difference distribution.
The imaging unit 120 illustrated in FIG. 14 includes a polarization element 122B having a direction for each pixel of the light receiving unit 121. As illustrated in FIG. 14B, the polarization element 122 </ b> B has a polarization direction corresponding to each pixel of the light receiving unit 121. In this example, the polarization directions corresponding to two pixels that are adjacent vertically or horizontally are deviated from each other by 45 degrees. Accordingly, the polarizing element 122B has four polarization directions that differ by 45 degrees corresponding to 2 × 2 pixels in the vertical and horizontal directions.

  By using this imaging unit 120, the phase difference for each pixel can be measured, and a phase difference distribution with 2 × 2 pixels in the vertical and horizontal directions as one unit can be obtained.

FIGS. 15A and 15B are diagrams showing measurement examples of the phase difference distribution.
FIG. 15A shows a phase difference distribution image received by the light receiving unit 121 and the polarizing element 122B corresponding to 256 pixels × 256 pixels. By expressing the phase difference by shading or color, the phase difference distribution can be visually recognized.

  FIG. 15B shows phase difference values corresponding to the phase difference distribution image shown in FIG. Based on the intensity (luminance information) of the signal captured by each pixel of the light receiving unit 121, the phase difference from the polarization direction of the polarizing element 122B corresponding to the pixel is obtained. That is, the numerical value of the phase difference can be obtained by obtaining the signal strength. By obtaining the phase difference numerically, the strain distribution can be quantitatively analyzed. For example, statistical processing such as standard deviation of distortion and total distortion can be easily performed. The display of the shade of the phase difference shown in FIG. 15A and the numerical value of the phase difference shown in FIG. 15B are examples of the phase difference map. The phase difference map is a distribution map of distortion in a predetermined area. In the example shown in FIG. 15A, the difference in the magnitude of distortion in the predetermined area is represented by a shading pattern.

  Further, by applying the configuration of the imaging unit 120 shown in FIG. 14 to the robot arm type strain measurement system 200 shown in FIG. 11, a vertical row of the inner surface IS of the silica glass crucible 11 as shown in FIG. The distortion distribution of the minute can be obtained as a phase difference map. Further, a strain distribution for the entire circumference of the silica glass crucible 11 as shown in FIG. 13B can be obtained as a phase difference map. That is, the robot arm type strain measurement system 200 can obtain two-dimensional and three-dimensional phase difference maps.

<Pulling device>
FIG. 16 is a schematic diagram showing an overall configuration of a pulling apparatus which is a silicon single crystal manufacturing apparatus according to the present embodiment.
A crucible CR that accommodates the silicon melt 23 is provided inside the chamber 510 that forms the appearance of the pulling device 500, and a carbon susceptor 520 is provided so as to cover the outside of the crucible CR. The crucible CR used in the pulling device 500 is the silica glass crucible 11 according to this embodiment. The carbon susceptor 520 is fixed to the upper end of the support shaft 530 parallel to the vertical direction. The crucible CR fitted to the carbon susceptor 520 is rotated in a predetermined direction by the support shaft 530 together with the carbon susceptor 520, and the liquid level of the silicon melt can be controlled to a constant height with respect to the heater 540 in the furnace. (So that the temperature gradient is constant), it can move up and down.

  The outer peripheral surfaces of the crucible CR and the carbon susceptor 520 are surrounded by a heater 540. The heater 540 is further surrounded by a heat insulating cylinder 550. In the raw material melting process in the growth of the silicon single crystal, the high-purity polycrystalline silicon raw material filled in the crucible CR by heating the heater 540 is heated and melted to form the silicon melt 23.

  A lifting means 560 is provided at the upper end of the chamber 510 of the lifting device 500. The pulling means 560 is provided with a wire cable 561 hanging down toward the rotation center of the crucible CR, and a pulling motor (not shown) for winding or feeding the wire cable 561 is provided. The seed crystal 24 is attached to the lower end of the wire cable 561. During the pulling, the seed crystal 24 rotates, and the silicon single crystal 25 (ingot) also rotates with the growth.

  A cylindrical heat shield member 570 is provided between the silicon single crystal 25 and the heat insulating cylinder 550 so as to surround the silicon single crystal 25 being grown. The heat shielding member 570 includes a cone portion 571 and a flange portion 572. By attaching the flange portion 572 to the heat insulating cylinder 550, the heat shielding member 570 is disposed at a predetermined position (hot zone). The diameter of the body part of the silicon single crystal 25 to be grown is about 465 mm at the maximum including the cutting allowance when it is pulled up so that it can be set to 450 mm after surface cutting, for example. At this time, when the silica glass crucible of the present invention is used, deformation of the silica glass crucible at the time of pulling up can be prevented, so that the dimensions and mounting clearance of the carbon susceptor 520 and the heat shielding member 570 are widened.

  In the pulling device 500, the crucible CR is heated by the heater 540 in a state where the periphery of the crucible CR is covered with the carbon susceptor 520. In recent years, the diameter of the crucible CR has increased to 32 inches or more and is heated to about 1500 ° C. to 1600 ° C. in order to dissolve the polycrystalline silicon raw material. At this time, it expands due to the crucible CR heating, but cannot be expanded outward because the periphery is covered with the carbon susceptor 520, and expands upward to the open side.

  In the silica glass crucible 11 according to the present embodiment as the crucible CR, since the region of substantially uniform strain continues in the vertical direction of the side wall portion 11a in both the first region R1 and the second region R2, the crucible CR is heated. A stable shape can be maintained without causing cracks, cracks, delamination, inward collapse, and the like.

  Here, the gap D between the crucible CR that is pulling up the silicon single crystal and the cone portion 571 that is a hot zone needs to be as narrow as possible. That is, in order to efficiently reach the center of the crucible CR from the heater 540 and to heat the solid-liquid interface to about 1420 ° C., for example, a crucible CR having a large diameter of 32 inches or more has a gap D of, for example, 30 mm to It needs to be as narrow as about 40 mm. Further, as the silicon single crystal pulling progresses, the silicon melt in the crucible CR decreases. Therefore, when the crucible CR is raised in order to make the silicon melt level constant with respect to the heater 540 in the furnace, the gap D between the cone portion 571 and the crucible CR becomes narrower. In such a situation, if the crucible CR falls inside, the crucible CR comes into contact with the heat shielding member 570 (cone portion 571).

Since the crucible CR is rotating while the silicon single crystal 25 is being pulled up, if the inward collapse of the crucible CR occurs and the straight body portion rotates and contacts the heat shielding member 570, the heat shielding member 570 or The crucible CR will be damaged. When the heat shielding member 570 is broken, the pulling of the silicon single crystal 25 must be stopped, and when the crucible CR is broken, it leads to leakage of the silicon melt, and the pulling device is broken. Repair is required.
The control of the height H between the silicon melt surface and the cone portion 571 is very important in controlling the temperature gradient in the vicinity of the solid-liquid interface of the silicon single crystal 25, and is controlled in units of 0.1 mm. (See Non-Patent Document 1).

  When deformation of the crucible CR occurs, the position of the liquid surface changes due to the change in the volume of the crucible CR, and the height H changes, leading to a decrease in crystal quality (crystal diameter, defects in the crystal, etc.) Yield is worse.

  Further, as shown by an arrow F in the figure, the cooling gas flows outside through the gap D through the portion indicated by the height H from between the silicon single crystal 25 and the cone portion 571. Therefore, when the gap D or the height H changes, the flow rate of the cooling gas changes, and the set temperature gradient changes, leading to a decrease in crystal quality.

  By suppressing the crucible CR from falling or deforming, contact between the crucible CR and the heat shielding member 570 can be avoided, and the silicon can be pulled up by the set temperature gradient, and the silicon single crystal having excellent crystal quality can be obtained. Crystal 25 can be manufactured. Therefore, the production yield of the silicon single crystal can be improved.

<Method for producing silicon single crystal>
FIGS. 17A to 17C are schematic views illustrating a method for producing a silicon single crystal using the silica glass crucible according to the present embodiment.
As shown in FIG. 17A, when pulling up the silicon single crystal, the silica glass crucible 11 is filled with polycrystalline silicon, and in this state, the polycrystalline silicon is heated by a heater disposed around the silica glass crucible 11. And melt. Thereby, the silicon melt 23 is obtained. At this time, by using the silica glass crucible of the present invention, the crucible during filling can be prevented from being damaged.

  The volume of the silicon melt 23 is determined by the mass of polycrystalline silicon. Therefore, the initial height position H 0 of the liquid surface 23 a of the silicon melt 23 is determined by the mass of the polycrystalline silicon and the three-dimensional shape of the inner surface of the silica glass crucible 11. That is, when the three-dimensional shape of the inner surface of the silica glass crucible 11 is determined, the volume up to an arbitrary height position of the silica glass crucible 11 is specified, whereby the initial height of the liquid surface 23a of the silicon melt 23 is determined. The position H0 is determined.

  After the initial height position H0 of the liquid surface 23a of the silicon melt 23 is determined, the tip of the seed crystal 24 is lowered to the height position H0 and brought into contact with the silicon melt 23. Then, the silicon single crystal 25 is grown by slowly pulling up the wire cable 561 while rotating it. At this time, the silica glass crucible 11 is rotated opposite to the rotation of the wire cable 561.

  As shown in FIG. 17B, when the liquid surface 23 a is located on the side wall portion 11 a of the silica glass crucible 11 when the straight body portion (the portion having a constant diameter) of the silicon single crystal 25 is pulled up. In this case, if the liquid surface 23a is pulled up at a constant speed, the descending speed Vm of the liquid surface 23a becomes almost constant, so that the control of the pulling up is easy.

  However, as shown in FIG. 17 (c), when the liquid level 23a reaches the corner 11b of the silica glass crucible 11, the area rapidly decreases as the liquid level 23a descends. The speed Vm increases rapidly. The descending speed Vm depends on the inner surface shape of the corner portion 11b.

  By accurately measuring the three-dimensional shape of the inner surface of the silica glass crucible 11, the inner surface shape of the corner portion 11b can be known, and therefore, how the descent speed Vm changes can be accurately predicted. it can. Based on this prediction, pulling conditions such as the pulling speed of the silicon single crystal 25 are determined. At this time, since the silica glass crucible 11 of the present embodiment is used, the predicted accuracy of the descending speed Vm is further improved because the silica glass crucible 11 is hardly deformed from the predicted shape. Thereby, it is possible to prevent the transition from occurring in the corner portion 11b and to automate the pulling-up.

  In the method for producing a silicon single crystal according to the present embodiment, since the silica glass crucible 11 is prevented from being deformed by the heating of the silica glass crucible 11 when the silicon single crystal 25 is pulled up (such as falling of the side wall portion 11a, distortion, rising of the bottom portion 11c). The deviation of the descending speed Vm of the liquid surface 23a obtained from the three-dimensional shape of the inner surface of the glass crucible 11 is suppressed, and the silicon single crystal 25 having a high crystallization rate can be manufactured with a high yield. The silicon single crystal is pulled up in an argon atmosphere and under reduced pressure (about 660 Pa to 13 kPa).

<Silicon single crystal ingot>
FIG. 18 is a schematic view illustrating a silicon single crystal ingot.
The silicon single crystal ingot 600 is manufactured by setting the silica glass crucible 11 of the present invention in the pulling device 500 and pulling it up by the above-described silicon single crystal manufacturing method. The silica glass crucible 11 is measured by, for example, the strain measuring apparatus 100 according to the present embodiment.

  The ingot 600 includes a shoulder 610 on the seed crystal 24 side, a straight body 620 continuous from the shoulder 610, and a tail 630 continuous from the straight body 620. In some cases, the seed crystal 24 is removed from the ingot 600. The diameter of the shoulder portion 610 gradually increases from the seed crystal 24 side to the straight body portion 620. The diameter of the straight body 620 is substantially constant. The diameter of the tail 630 gradually decreases as the distance from the straight body 620 increases.

  The quality of the ingot 600 is closely related to the quality of the silica glass crucible 11 to be pulled up. For example, the contamination of the silica glass crucible 11 (for example, an impurity metal element in the glass) or foreign matter leads to dislocation of the silicon single crystal in the ingot 600. Further, depending on the smoothness of the inner surface of the silica glass crucible 11 (unevenness that can be seen by appearance), the amount and size of bubbles near the surface, minute cracks in the silicon due to chipping of the crucible surface, cracking of the bubbles, and crushing. When debris (particles peeled off from the crucible) falls into the silicon melt, this mixes in the ingot and leads to dislocation.

In addition, the quality of the ingot 600 greatly depends on the pulling control in manufacturing the ingot 600. Below, the specific example of the relationship between the quality of the ingot 600 and raising control is demonstrated.
FIGS. 19A to 19C are schematic diagrams for explaining the pull-up control.
As shown in FIG. 19A, when the growth rate of the silicon single crystal is Vg, the pulling rate of the silicon single crystal is V, the lowering rate of the silicon melt is Vm, and the rising rate of the crucible is C. The following relationship holds.
Vg = V + Vm-C

  Among these, the liquid level lowering speed Vm is determined by a function f of the crucible inner volume and the silicon single crystal growth speed Vg (see FIG. 19B). In the conventional technique, the liquid level lowering speed Vm is obtained by calculation using this function f. Further, since the pulling speed V and the crucible rising speed C are known as the conditions of the pulling apparatus, the silicon single crystal growth speed Vg = V + Vm−C is obtained thereby.

  However, in actual pulling up, the inner shape of the crucible is deformed due to exposure to high temperature, and the internal volume is also changed (see FIG. 19C). In the pulling device, the silica glass crucible is inserted into the carbon susceptor. Therefore, the outer peripheral surface of the silica glass crucible is in a state of being fitted to the carbon susceptor. For this reason, the silica glass crucible is not deformed outward but deformed only inward. If the internal volume of the crucible changes, the calculation of the liquid level lowering speed Vm becomes inaccurate, and the silicon single crystal growth speed Vg cannot be determined accurately. This growth rate Vg is an important factor in the generation of crystal defects. Therefore, if the growth rate Vg cannot be accurately controlled, the quality of the ingot 600 is greatly affected.

If the inner radius of the crucible at the position of the silicon melt liquid surface is R, the diameter of the silicon single crystal (ingot) is r, the density of the silicon melt is ρL, and the density of the silicon single crystal is ρs, the liquid surface is in the crucible straight body. In some cases, the following equation holds:
Vg = ρL / ρs · (R / r) ² · Vm
Vg / Vm = ρL / ρs · (R / r) ² = k

When the variation rate of the inner radius of the crucible is α, the following equation is established.
Vg = ρL / ρs · (αR / r) ² · Vm
Vg = α ² · {ρL / ρs · (αR / r) ² · Vm}

  From this, the square of α contributes to the deviation of Vg. Therefore, if R varies by 1%, Vg varies by about 2%.

Assuming that R = 0.797 m, r = 0.3 m, ρL = 2570 kg / m ³ , and ρs = 2300 kg / m ³ , k = 7.95 and 1 / k = 0.126.
For example, when a silicon single crystal (ingot) corresponding to a thickness of 1 mm of a silicon wafer is manufactured, the drop in the silicon melt level is 0.126 mm. Considering the cutting width (width of a blade or the like) when cutting a silicon wafer from an ingot and polishing after cutting, the thickness of the silicon wafer is about 700 μm to 800 μm. In order to make the COP substantially zero no matter where the ingot is cut, it is necessary to make the COP substantially zero in the entire area of the straight body portion of the ingot. In addition, when the structure portion falls within a range of 1/10 to 1/100 or less of the thickness of the silicon wafer, such as a semiconductor device having a three-dimensional structure to be described later, in the pulling of the silicon single crystal, the thickness of the silicon wafer is 1 A pulling control of / 10 to 1/100 or less (pulling control for making COP substantially zero) is necessary. In this case, in order to control the decrease in the liquid level of the silicon melt, it is necessary to control the accuracy of 0.01 mm or less.

Thus, if the inner diameter of the silica glass crucible 11 varies by 1%, the growth rate Vg of the silicon single crystal varies by 2%. Moreover, the rate of decrease Vm of the silicon melt at the corner portion 11b of the silica glass crucible 11 is higher than the rate of decrease of the level of the silicon melt at the straight body portion of the silica glass crucible 11. Therefore, the influence of the fluctuation of the inner diameter of the crucible on the fluctuation of the liquid level is larger in the corner portion 11b than in the straight body portion of the crucible.

  In this embodiment, since the internal residual stress in the thickness direction of the silica glass crucible 11 actually used for pulling can be accurately measured, the relationship between the internal residual stress and the change in the inner diameter of the crucible after use (in terms of operation results) Based on the simulation of the fluctuation amount of the inner diameter of the crucible based on this, the inner diameter fluctuation amount of the crucible in use can be estimated at the stage of the silica glass crucible 11 before use (before the silicon single crystal is pulled up). This makes it possible to reduce the deviation from the target value of the growth rate Vg of the silicon single crystal compared to the case where the deformation of the crucible is not considered at all as in the conventional technique, and the entire length of the straight body portion 620 of the ingot 600 can be reduced. Defects can be suppressed (substantially zero).

FIG. 20 is a diagram showing a variation amount of the inner diameter of the crucible.
In FIG. 20, the horizontal axis indicates the amount of variation in the inner diameter of the crucible, and the vertical axis indicates the height from the bottom of the crucible.
The plot of FIG. 20 is a measured value. Moreover, the line L connects the average of the measured value in each height.
As shown by line L, it can be seen that fluctuations in the inner diameter of the crucible (that is, fluctuations in the crucible internal volume) occur on average. As in this embodiment, if the rising speed A of the silicon single crystal is changed based on the shape of the inner surface of the crucible, it is possible to control the growth rate Vg of the silicon single crystal so that the entire length of the silicon single crystal is within a defect-free range. become.
On the other hand, in the prior art, feedback control during CZ single crystal growth is performed only by a combination of ADC (automatic diameter control) and liquid level control. That is, in the prior art, the shape of the crucible in actual use is not taken into consideration at all, and the shape change of the crucible cannot be accurately grasped, so that the growth rate Vg is accurately controlled in pulling up the silicon single crystal. I can't. In other words, the conventional technology does not correspond to the Vg control corresponding to the accuracy of the liquid level lowering velocity Vm of 0.01 mm or less as described above, and the performance of the semiconductor device, particularly the device of the three-dimensional structure is sufficient. It is not a silica glass crucible that can produce a silicon single crystal (ingot) to be drawn out.

Here, it is possible to estimate the behavior of the crucible from the production history, inspection results, and use results of the crucible so far by simulation technology (example of crucible behavior). From this, the following can be understood about the deformation of the crucible.
(1) The amount of fluctuation is large in the thin part.
(2) The amount of deformation increases as the weight of the crucible increases.
(3) The amount of deformation of the inner surface increases as the crucible has a smaller outer diameter.
(4) The amount of deformation in the eccentric part is large.
(5) The crucible is likely to be deformed at a non-symmetrical portion of the carbon susceptor.
(6) The silica glass crucible is also ceramic, and the inner peripheral surface of the crucible is not a perfect circle.

  As described above, in order to control the growth rate Vg of the silicon single crystal by Vg = V + Vm−C, it is necessary to accurately grasp the crucible information. Therefore, it is desirable to record all crucible information from the past in association with each other so that they can be searched.

  It is also important to regulate the relationship between the growth rate (Vg) of the silicon single crystal and the temperature gradient (G) in the pulling axis direction near the solid-liquid interface in order to suppress the occurrence of crystal defects in the ingot 600. Become. Here, the temperature gradient (G) in the pulling axis direction is higher on the melt side than on the solid side (in other words, lower on the solid side than on the melt side). In addition, the temperature gradient in the plane (in the radial direction) perpendicular to the pulling axis (in the radial plane) is constant.

  The silica glass crucible 11 of the present invention can stabilize the height H between the surface of the silicon melt and the tip of the heat shielding member, since deformation and collapse of the silicon single crystal are suppressed. In the ingot 600 obtained by pulling up the silicon single crystal using such a silica glass crucible 11, the crystal defects in the straight body 620 are substantially zero. For example, the COP (Crystal Originated Particle) in the straight body 620 is substantially zero. COP is one of crystal defects and is a fine defect in which silicon atoms are not present at lattice points of a single crystal (holes are collected). The presence of the COP causes deterioration of electrical characteristics (leakage current, resistance value distribution, carrier mobility, etc.) of the semiconductor device.

Here, generation of COP will be described.
FIG. 21 is a schematic diagram for explaining a situation in which various defects occur based on the Boronkov theory.
As shown in FIG. 21, in the Boronkov theory, when the pulling speed is V (mm / min) and the temperature gradient in the pulling axis direction in the vicinity of the solid-liquid interface of the ingot (silicon single crystal) is G (° C./mm), The relationship between V / G and point defect concentration is schematically shown with the ratio V / G being the horizontal axis and the concentration of vacancy type point defects and the concentration of interstitial silicon type point defects being the same vertical axis. expressing. It is shown that there is a critical point that becomes a boundary between a region where a vacancy type point defect occurs and a region where an interstitial silicon type point defect occurs.

  When V / G falls below the critical point, a single crystal having a dominant interstitial silicon point defect concentration is grown. In a range where V / G is less than the critical point (V / G) I, the interstitial silicon type point defects are dominant in the single crystal, and the region where the aggregate of interstitial silicon point defects exists [I ] Appears.

  On the other hand, when V / G exceeds the critical point, a single crystal having a dominant vacancy point defect concentration is grown. In a range where V / G is greater than the critical point (V / G) v, a region where vacancy type point defects are dominant in the single crystal and agglomerates of vacancy type point defects exist [V] Appears and COP occurs.

FIG. 22 is a schematic diagram showing the relationship between the pulling rate and the defect distribution during single crystal growth.
The defect distribution shown in FIG. 22 is that a silicon single crystal is grown while gradually lowering the pulling speed V, and the grown single crystal is cut along a central axis (pulling axis) to form a plate-shaped specimen. It shows the occurrence of defects. The defect distribution is a result of evaluating the occurrence of defects by decorating Cu on the surface of the plate-shaped specimen and performing heat treatment, then observing the plate-shaped specimen by the X-ray topograph method.

As shown in FIG. 22, when the growth is performed at a high pulling speed, a region [VOPs of vacancy-type point defects (COP) exist over the entire in-plane region perpendicular to the pulling axis direction of the single crystal [V ] Occurs. When the pulling speed is decreased, the OSF region appears in a ring shape from the outer peripheral portion of the single crystal. The diameter of the OSF region gradually decreases as the pulling speed decreases, and disappears when the pulling speed becomes V1. Accordingly, a defect-free region [P] (region [PV]) appears instead of the OSF region, and the entire in-plane area of the single crystal is occupied by the defect-free region [P]. When the pulling rate is reduced to V ₂ , a region [I] where interstitial silicon-type point defect aggregates (LD) are present appears, and finally appears in a defect-free region [P] (region [PI]). Instead, the entire in-plane area of the single crystal is occupied by the region [I].

  In the present embodiment, the above-mentioned COP is substantially zero means that the number of detected COPs is substantially zero. COP is detected by a particle counter. In the particle counter, when the number of particles of 0.020 μm or more is detected only 30 or less on the wafer surface (semiconductor device forming surface), the number is substantially zero. In this specification, “0.020 μm COP” means, for example, 0.020 μm when measured with the SP series manufactured by Tencor or the particle counter device for semiconductors and silicon wafers having the same performance as this device. A COP detected as a particle size.

  As described above, the ingot 600 in which the COP of the straight body portion 620 is substantially zero is sliced into, for example, a diameter of 300 mm and a thickness of about 1 mm to form a silicon wafer. In a semiconductor device manufactured using a silicon wafer cut out from the ingot 600, electrical characteristics can be stabilized and deterioration can be suppressed.

  The method for detecting COP may be other than the particle counter. For example, a method using a surface defect inspection apparatus, after forming an oxide film of a predetermined thickness on the surface of a wafer, applying an external voltage to destroy the oxide film at the defective portion of the wafer surface and deposit copper, Examples include a method of detecting defects (COP) by observing the deposited copper with the naked eye, a transmission electron microscope (TEM), a scanning electron microscope (SEM), and the like. In the straight body 620 of the ingot 600, COP is not detected by such a detection method (substantially becomes zero).

  A more preferable form of the ingot 600 according to the present invention is that there is no region where point defects (vacancies) called vacancy are aggregated (V-Rich region where COP exists) in all the straight body portions 620, and OSF (Oxidation Induced Stacking). Fault) is not detected, and there is no region (I-Rich region) where interstitial point defects called interstitial exist, that is, all of the straight body portion 620 is a neutral region. Here, the neutral region includes not only a region having no defects, but also a region that does not exist as an agglomerated defect or is so small that it cannot be detected even if a slight vacancy or interstitial is included.

  As described above, since the crystal defects in the straight body portion 620 are zero, the electrical characteristics of the semiconductor device manufactured using the wafer cut out from the ingot 600 can be stabilized and the deterioration can be suppressed.

<Homoepitaxial wafer>
Further, a homoepitaxial wafer (hereinafter also simply referred to as “epitaxial wafer”) using this wafer as a substrate portion may be configured. FIG. 23 is a schematic cross-sectional view illustrating an epitaxial wafer. The epitaxial wafer 700 includes a wafer substrate portion 710 cut out from the ingot 600, and a silicon single crystal epitaxial layer 720 provided on the substrate portion 710. In this embodiment, the epitaxial layer 720 is a silicon homoepitaxial layer. The thickness of the epitaxial layer 720 is about 0.5 μm to 20 μm.

An example of the manufacturing method of the epitaxial wafer 700 is shown. First, the substrate unit 710 is heated to about 1200 ° C. in an epitaxial furnace. Next, vaporized silicon tetrachloride (SiCl ₄ ) and trichlorosilane (trichlorosilane, SiHCl ₃ ) are flowed into the furnace. Thus, a silicon single crystal film is vapor-phase grown (epitaxial growth) on the surface of the substrate portion 710, and an epitaxial layer 720 is formed.

  By forming the epitaxial wafer 700 using a wafer cut out from the ingot 600 having substantially zero crystal defects, the epitaxial layer 720 having substantially zero crystal defects can be formed.

In recent years, miniaturization of semiconductor integrated circuits has progressed, and the limits of conventional planar transistors are approaching. Accordingly, a transistor called a Fin-type FET (fin-type field effect transistor) structure has been proposed (see, for example, Patent Documents 21 and 22).
In the conventional planar type, a MOSFET (Metal Oxide Semiconductor Field Effect Transistor) structure is formed inside the silicon wafer surface.
In the planar type, the source and drain are two-dimensionally configured. However, the Fin-type FET has a channel region called FIN in the upper layer of the silicon surface and is in contact with the silicon wafer to form a three-dimensional MOSFET.
The planar type has been miniaturized by the gate length, but in the Fin type FET, the fin width is managed as the minimum dimension. There is also a Fin type FET having a fin width of about 20 nm, that is, about the same as COP.
Therefore, it is required to reduce the size of the COP to the limit as the surface quality of the silicon wafer directly under the fin.
Such a three-dimensional structure is adopted not only in a Fin type FET but also in a three-dimensional NAND type flash memory.
In order to manufacture such a semiconductor device, a homoepitaxial wafer with improved quality is desired.
When forming a homoepitaxial layer using a silicon wafer, the size of the COP of the silicon wafer needs to be smaller and smaller. Although there is a heat treatment method to suppress COP on the silicon wafer, it is important to control the silicon melt at the time of pulling in order to make COP substantially zero at the ingot stage of the silicon single crystal. is there. The inventors of the present application have found that the silicon melt can be controlled by paying attention to the relationship between the liquid level fluctuation of the silicon melt and the silica glass crucible.

  The epitaxial layer 720 may be formed on the entire surface of the substrate portion 710 or may be partially formed. As a result, it is possible to provide a high-quality epitaxial wafer 700 that can be used when crystal integrity is required or when a multilayer structure with different resistivity is required.

<Process from crucible manufacturing to silicon single crystal product manufacturing>
FIG. 24 is a flowchart illustrating the steps from crucible manufacturing to wafer manufacturing.
Steps S201 to S206 shown in FIG. 24 are crucible manufacturing processes, steps S207 to S214 are ingot manufacturing processes, steps S215 to S221 are silicon wafer manufacturing processes, and steps S222 to S227 are the same. It is a manufacturing process of an epitaxial wafer.

A series of processes from crucible production to ingot production shown in steps S201 to S214 will be referred to as crucible-ingot production processes.
A series of processes from crucible production to silicon wafer production shown in steps S201 to S221 will be referred to as a crucible-silicon wafer production process.
A series of processes from crucible manufacturing to epitaxial wafer manufacturing shown in steps S201 to S227 will be referred to as a crucible-epiwafer manufacturing process.

  In this embodiment, in order to perform consistent manufacturing condition control and quality control in each of the crucible-ingot manufacturing process, the crucible-silicon wafer manufacturing process, and the crucible-epi wafer manufacturing process, an integrated control system that collectively manages each process is provided in this embodiment. Used.

  In this embodiment, an integrated control system is used for production management that assumes the quality of silicon single crystal products (ingots, silicon wafers, epitaxial wafers) due to crucible manufacturing.

  Conventionally, for example, when an ingot is manufactured by pulling up a silicon single crystal, the diameter of the straight body portion is controlled to be constant by ADC (automatic diameter control). The time required for pulling up the straight body having a diameter of about 300 mm to a total length of 2000 mm is about 4000 minutes as 0.5 mm / min. In addition, as a whole in the production of silicon ingots, (1) an operation of carefully loading the silica glass crucible so that the silica glass crucible does not break when filling the silica glass crucible, (2) melting of the polycrystalline silicon, (3) Dash necking (dislocation removal) process, (4) forming the shoulder of the silicon ingot, (5) lifting the total length of the straight body part to 2000 mm, (6) tail tailing to prevent dislocation from entering the silicon ingot, (7) The furnace is cooled and the silicon ingot is recovered. In order to manufacture a single silicon ingot having a diameter of 300 mm and a total length of 2000 mm by performing such a series of processes, it takes about 7 days.

  The control during this time is mainly based on the relationship between the pulling speed and the weight, and aims at COP-free pulling over the entire length of the straight body with a constant diameter. The height H between the surface of the silicon melt important for pulling and the cone portion 571 is high when the pulling speed is high, and is low when the pulling speed is slow. Conventionally, the height H is controlled based on the individual difference for each lifting device and the experience of the operator.

  In the present embodiment, the height H at the time of pulling up can be controlled more uniformly by predicting the amount of inner surface deformation of the crucible. That is, in the pulling device, the crucible is housed in the carbon susceptor 520, and becomes a weight of, for example, 500 kg due to the filling of polycrystalline silicon. In addition, the crucible being pulled becomes a high temperature of about 1600 ° C. and is pushed outward by the silicon melt, and the gap with the carbon susceptor 520 disappears. Since the carbon susceptor 520 is not deformed, as a result, the crucible is easily deformed inward by a reaction force from the carbon susceptor 520.

  The integrated control system of the present embodiment accumulates information such as the manufacturing history of the crucible used so far, the measurement result of the internal residual stress before use, the shape change after use, etc. Calculate the behavior and deformation of the crucible when it is pulled up before use. Thereby, the deformation | transformation of the crucible internal volume can be known from the deformation | transformation of the crucible during raising to be estimated, and the height H during raising can be strictly controlled. Therefore, it is possible to perform consistent control to manufacture an ingot in which crystal defects are substantially zero, manufacture a silicon wafer from the ingot, and manufacture an epitaxial wafer using the silicon wafer.

  As described above, according to the embodiment, the internal residual stress of the silica glass crucible 11 can be accurately measured.

  Although the present embodiment has been described above, the present invention is not limited to these examples. For example, those in which the person skilled in the art appropriately added, deleted, and changed the design of each of the above-described embodiments, and combinations of the features of each embodiment as appropriate, also have the gist of the present invention. As long as they are within the scope of the present invention.

DESCRIPTION OF SYMBOLS 11 ... Silica glass crucible 11a ... Side wall part 11b ... Corner part 11c ... Bottom part 13 ... Transparent layer 15 ... Non-transparent layer 20 ... Carbon mold 21 ... Vent 23 ... Silicon melt 23a ... Liquid surface 24 ... Seed crystal 25 ... Single silicon Crystal 30 ... Arc electrode 100 ... Strain measuring device 110 ... Light emitting unit 110A ... Downward irradiation unit 110B ... Side irradiation unit 111 ... Light source 112 ... First polarizing means 113 ... Second polarizing means 115 ... Diffuser 120 ... Imaging unit 121 ... Light receiving unit 122 ... third polarizing means 122B ... polarizing element 123 ... fourth polarizing means 130 ... output unit 200 ... robot arm type strain measuring system 201 ... first silica powder 202 ... second silica powder 210 ... robot arm 220 ... stand 221 ... Horizontal base 222 ... Vertical base 223 ... Slide rail 224 ... Base 250 ... Controller 500 ... Lifting device 510 ... 520 ... Carbon susceptor 530 ... Support shaft 540 ... Heater 550 ... Insulating cylinder 560 ... Pulling means 561 ... Wire cable 570 ... Heat shielding member 571 ... Cone part 572 ... Flange part 600 ... Ingot 610 ... Shoulder part 620 ... Straight body part 630 ... tail 700 ... epitaxial wafer 710 ... substrate part 720 ... epitaxial layer CR ... crucible H0 ... height position IS ... inner surface MR ... measurement region R1 ... first region R2 ... second region TP ... upper end surface V ... lifting speed Vg ... Growth speed Vm ... Descent speed

Claims (14)

Strain measurement for measuring strain of a silica glass crucible comprising a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. A device,
A light emitting part that is arranged on the side of the side wall part and irradiates polarized light toward the side wall part;
An imaging unit that captures an image corresponding to polarized light on the upper end surface of the side wall;
An output unit that outputs a distribution of strain of the silica glass crucible based on the image captured by the imaging unit;
A strain measuring device for a silica glass crucible provided with The light emitting unit
A light source;
First polarizing means for extracting a linearly polarized light component from the light emitted from the light source;
Second polarizing means for converting light of the linearly polarized light component extracted through the first polarizing means into light of a rotationally polarized light component;
The apparatus for measuring strain of a silica glass crucible according to claim 1, comprising: The imaging unit
A light receiver;
A third polarizing means provided between the light receiving portion and the upper end surface of the side wall portion, and transmits more light of a rotationally polarized component than a linearly polarized component;
The strain measuring apparatus for a silica glass crucible according to claim 2, comprising: It further comprises a gantry having a pedestal on which the silica glass crucible to be measured is placed,
The light emitting unit is provided on the gantry,
The imaging unit is provided movably,
The strain measuring apparatus for a silica glass crucible according to any one of claims 1 to 3. An arm that movably supports the imaging unit;
A controller for controlling the position of the arm using design data of the silica glass crucible, and controlling an imaging region by the imaging unit;
The apparatus for measuring strain of a silica glass crucible according to claim 4, further comprising: A controller for controlling the pedestal, the light emitting unit, and the imaging unit;
5. The silica according to claim 4, wherein the controller measures distortion of the entire circumference of the silica glass crucible by repeating imaging by moving a relative position between the silica glass crucible and the imaging unit. Glass crucible strain measuring device.   The strain measuring apparatus for a silica glass crucible according to claim 6, wherein the controller performs control to measure strain for at least one round of the upper end surface of the silica glass crucible.   The apparatus for measuring strain of a silica glass crucible according to claim 6, wherein the controller performs control for measuring strain of the entire inner surface of the silica glass crucible. An arm that movably supports the imaging unit;
The strain measurement apparatus for a silica glass crucible according to claim 6, wherein the controller controls the position of the arm using design data of the silica glass crucible to control an imaging region by the imaging unit. A silica glass crucible comprising a cylindrical side wall part, a curved bottom part, and a corner part provided between the side wall part and the bottom part and having a curvature higher than the curvature of the bottom part, and a strain measuring device A method for producing a silicon single crystal using the silica glass crucible measured in
The strain measuring device includes:
A light emitting part that is arranged on the side of the side wall part and irradiates polarized light toward the side wall part;
An imaging unit that captures an image corresponding to polarized light on the upper end surface of the side wall;
An output unit that outputs a strain distribution of the silica glass crucible based on the image captured by the imaging unit; and
The silica glass crucible is
As the distribution output from the output unit of the strain measuring device,
A first region provided from the inner surface to the middle in the thickness direction of the side wall,
A second region provided outside the first region in the thickness direction of the side wall and having a strain distribution different from the first region;
The method for producing the silicon single crystal is:
Adding a silicon material into the silica glass crucible and melting it;
And a step of pulling up the silicon single crystal from the silicon melt held in the silica glass crucible. Strain measurement for measuring strain of a silica glass crucible comprising a cylindrical side wall portion, a curved bottom portion, and a corner portion provided between the side wall portion and the bottom portion and having a curvature higher than the curvature of the bottom portion. A method,
Irradiating the silica glass crucible to be measured with polarized light from the light emitting part;
Capturing an image corresponding to the polarized light applied to the silica glass crucible with an imaging unit;
By imaging the predetermined measurement area by the imaging unit, moving the relative position of the silica glass crucible and the imaging unit and repeating the imaging of the next measurement area, the silica glass crucible Measuring the distortion of the entire circumference of the side wall,
Method for measuring strain of silica glass crucible provided with A position representing the distribution of strain of a silica glass crucible comprising a cylindrical side wall, a curved bottom, and a corner provided between the side wall and the bottom and having a curvature higher than the curvature of the bottom. A phase difference map,
A phase difference map in which a phase difference caused by transmission of polarized light incident on a predetermined region of the silica glass crucible is displayed in a shade or numerical value in association with the position of the region. A silica glass crucible comprising a cylindrical side wall part, a curved bottom part, and a corner part provided between the side wall part and the bottom part and having a curvature higher than the curvature of the bottom part, and a strain measuring device A silicon single crystal ingot pulled up using the silica glass crucible measured in
The ingot is
Shoulder and
A straight body continuous from the shoulder;
A tail portion continuous from the straight body portion,
An ingot in which the crystal defects in the straight body portion are substantially zero. A substrate portion by a wafer using the silicon single crystal ingot according to claim 13;
A homoepitaxial wafer comprising a silicon single crystal homoepitaxial layer provided on the substrate part.