JP2017105654A - Method of manufacturing silicon single crystal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately measure a gap width by a mirror image method at the start of lifting even when a lower end part of a heat shield body is beveled.SOLUTION: The present invention relates to a method of manufacturing silicon single crystal such that when a silicon single crystal is lifted from a silicon melt 2 by a Czochralski method, the silicon single crystal is lifted through a circular opening 17a provided at a lower end of a heat shield body 17, the method comprising: calculating a measured distance from an opening 17a of a real image of the heat shield body 17 present in an observation region when the silicon melt 2 viewed through the opening 17a is observed from obliquely above the silicon melt 2 to an opening of a mirror image 17' of the heat shield body reflected in a liquid plane of the silicon melt 2; calculating a correction distance by removing an influence of beveling performed on the lower end of the heat shield body 17 from a half value of the measured distance; and finding a gap width from the lower end of the heat shield body 17 to the liquid plane of the silicon melt 2 using the correction distance.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a method for producing a silicon single crystal by the Czochralski method (hereinafter referred to as “CZ method”), and more particularly, to a method for measuring a liquid surface position of a silicon melt.

  Many silicon single crystals used as a substrate material for semiconductor devices are manufactured by the CZ method. In the CZ method, a seed crystal is immersed in a silicon melt contained in a quartz crucible, and the seed crystal is gradually raised while rotating the seed crystal and the crucible, thereby growing a single crystal having a large diameter at the lower end of the seed crystal. Let

  It is known that the type and distribution of defects contained in a silicon single crystal grown by the CZ method depends on the ratio V / G between the pulling rate V of the single crystal and the temperature gradient G in the crystal growth direction. When V / G is large, vacancies become excessive, and microvoids (generally referred to as COP :) that form vacancies are generated. On the other hand, when V / G is small, interstitial silicon atoms become excessive, and dislocation clusters that are aggregates of interstitial silicon are generated. Therefore, in order to produce a single crystal containing neither COP nor dislocation clusters, it is necessary to strictly control V / G with respect to the radial direction and the length direction of the single crystal.

  Since the pulling speed V is constant at any position in the radial direction of the single crystal, an appropriate high temperature region (hot zone) is constructed in the chamber so that the temperature gradient G is within a predetermined range in the radial direction of the single crystal. There is a need. Further, since the temperature gradient G in the length direction of the single crystal depends on not only the hot zone structure but also the single crystal pulling speed V, the single crystal pulling speed V is adjusted so that V / G falls within a predetermined range. There is a need. At present, by strictly controlling V / G, a silicon single crystal having a diameter of 300 mm that does not include COP and dislocation clusters is mass-produced.

  However, a silicon wafer that does not contain COP and dislocation clusters pulled up by controlling V / G is never homogeneous, and includes a plurality of regions that behave differently when heat-treated. For example, there are three regions, an OSF region, a Pv region, and a Pi region, in descending order of V / G, between a region where COP occurs and a region where dislocation clusters occur.

  The OSF region contains plate-like oxygen precipitates (OSF nuclei) in an as-grown state (a state in which no heat treatment is performed after single crystal growth), and at a high temperature (generally 1000 to 1200 ° C.). This is a region where OSF (Oxidation induced Stacking Fault) occurs when thermal oxidation occurs. The Pv region includes oxygen precipitation nuclei in an as-grown state, and is a region where oxygen precipitates are easily generated when two-stage heat treatment is performed at a low temperature and a high temperature (for example, 800 ° C. and 1000 ° C.). The Pi region is a region that hardly contains oxygen precipitation nuclei in an as-grown state and hardly generates oxygen precipitates even after heat treatment. In order to grow a high-quality silicon single crystal in which the Pv region and the Pi region are separately formed, it is necessary to further strictly control V / G.

  Normally, the control of V / G is performed by adjusting the pulling speed V. In the control of V / G, the temperature gradient G in the crystal growth direction of the silicon single crystal is controlled by a heat shield provided above the silicon melt, thereby establishing an appropriate hot zone near the solid-liquid interface. can do.

  In order to control the V / G with high accuracy and form a desired defect-free region in the silicon single crystal, the distance (gap width) from the surface of the silicon melt to the heat shield must be kept constant. Desired. Variations in the gap width not only change the state of heat transfer from the silicon melt to the single crystal, but also change the flow rate of the inert gas flowing through the gap, which changes the amount of SiO evaporated from the silicon melt. The oxygen concentration in the single crystal also changes. That is, the variation in the gap width causes the interstitial oxygen concentration and the crystal defect distribution in the crystal to fluctuate, which causes the single crystal yield to deteriorate. In order to prevent deterioration of the quality of the silicon single crystal, it is necessary to set a desired gap width at the start of pulling and to control the gap width throughout the pulling process.

  However, in the manufacture of silicon single crystals, the chamber is disassembled and cleaned for each pulling batch, and the next pulling process is performed after the chamber is reassembled. In this case, the distance from the liquid surface of the silicon melt to the heat shield is increased, resulting in different batches, resulting in variations in the position of the liquid surface of the silicon melt relative to the heat shield. Therefore, it is necessary to measure the exact gap width between the heat shield and the melt surface at the start of pulling for each pulling batch.

  As a method for accurately measuring the position of the silicon melt at the start of pulling of the silicon single crystal, for example, Patent Document 1 discloses, for example, quartz at an end facing the melt surface of a heat shielding member covering the melt surface. A method is described in which a refractory bar made of is attached, the bar is in contact with the melt surface, and the position of the melt surface is set based on this.

  In Patent Document 2, a real image including at least a circular opening of a heat shielding member arranged so as to cover a part of the silicon melt surface and a mirror image of the heat shielding member reflected on the surface of the silicon melt are taken. The method of calculating the liquid level position of the silicon melt from the interval between the real image and the mirror image obtained in this way and controlling the distance between the heat shielding member and the liquid level position is described.

Japanese Patent Publication No. 3-31673 JP 2013-216505 A

  The method of measuring the liquid surface position by the so-called mirror image method described in Patent Document 2 is reflected in the real image of the heat shield disposed above the silicon melt visible in the observation region in the chamber and the liquid surface of the silicon melt. Based on the distance from the mirror image of the heat shield, the gap width between the liquid surface of the silicon melt and the lower end of the heat shield is calculated.

  However, when the inventors of the present application conducted thorough evaluation experiments on the measurement accuracy of the gap width by the mirror image method, it became clear that the gap width could not be measured accurately depending on the shape of the heat shield. Even if the measurement error of the gap width is only about 1 mm, the influence is very large, and therefore improvement of the measurement accuracy of the gap width is required.

  Accordingly, an object of the present invention is to provide a silicon single crystal manufacturing method capable of accurately measuring the gap width at the start of pulling by a mirror image method regardless of the shape of the heat shield.

  The inventors of the present application have made extensive studies on the cause of the gap width not being able to be measured accurately. As a result, it became clear that the cause is the chamfering of the lower end portion of the heat shield. In the single crystal pulling process, the heat shield becomes very hot, and thus is easily damaged by thermal expansion, and cracks are particularly likely to occur at the lower end of the heat shield. The chamfering process is performed to reduce the thermal stress applied to the lower end portion of the heat shield.

  However, since the mirror image method is affected by the shape of the lower end of the heat shield, if there is a chamfer at the lower end of the heat shield, just calculating the gap width as it is based on the real image and the mirror image of the heat shield The gap width cannot be measured accurately.

  The present invention is based on such technical knowledge, and the method for producing a silicon single crystal according to the present invention includes a region above a crucible containing a silicon melt and excluding the pulling path of the silicon single crystal. When the silicon single crystal is pulled up from the silicon melt by the Czochralski method, the silicon single crystal passes upward through a circular opening provided at the lower end of the heat shield. A method for producing a silicon single crystal to be pulled up, wherein the silicon melt is seen from an oblique image above the silicon melt and is observed in the observation region when the silicon melt is observed through the opening. Calculate the measured distance to the mirror image corresponding to the opening of the thermal shield reflected on the surface of the silicon melt, and apply it to the lower end of the thermal shield from the half value of the measured distance. It was calculated effect correction distance removing the chamfering, and obtains the gap width from the lower end of the heat shield to the liquid surface of the silicon melt using the correction distance. The effect of chamfering is specifically a measurement error caused by the shape and size of the chamfering applied to the lower end of the thermal shield.

  According to the present invention, even if the lower end portion of the heat shield is chamfered, the influence can be removed, thereby increasing the measurement accuracy of the position of the silicon melt relative to the heat shield. Can do.

In the method for producing a silicon single crystal according to the present invention, a distance G _c from a real image of the upper edge of the chamfered portion of the thermal shield to a mirror image corresponding to the upper edge reflected on the liquid surface of the silicon melt is the measured distance. , when the height dimension of the chamfered portion as h _1, the gap width G _a,
It is preferable to obtain by: The second term on the right side of Equation 1 is an error due to the effect of chamfering, and the gap width is corrected by subtracting the error. In the case of adopting a mirror image of the upper end edge of the chamfered portion of the heat shield as a measuring point, such a calculation method, the measurement of the gap width G _a from the lower end of the thermal shield to the liquid surface of the silicon melt Accuracy can be increased.

The method for producing a silicon single crystal according to the present invention includes a real image of an upper edge of a chamfered portion of the thermal shield to a mirror image corresponding to a lower edge of the chamfered portion of the thermal shield reflected on the liquid surface of the silicon melt. the distance G _m and the measured distance, the height h ₁ of the chamfered _portion, the width of the chamfered portion when the h _2, viewing angle of the silicon melt for pulling direction of the silicon single crystal theta, the gap width _{G a,}
It is also preferable to obtain by The second term on the right side of Equation 2 is an error due to the effect of chamfering, and the gap width is corrected by subtracting the error. In the case of adopting a mirror image of the lower edge of the chamfered portion of the heat shield as a measuring point, such a calculation method, the gap width G _a from the lower end of the thermal shield to the liquid surface of the silicon melt Measurement accuracy can be increased.

  The method for producing a silicon single crystal according to the present invention includes a step of melting a solid raw material filled in the crucible to generate a silicon melt, a step of depositing a seed crystal on the liquid surface of the silicon melt, The step of raising the seed crystal to grow a silicon single crystal, and the step of obtaining the gap width from the lower end of the thermal shield to the liquid surface of the silicon melt is after the silicon melt is generated. This is preferably performed before the seed crystal is deposited. Thus, by performing the gap width measuring method according to the present invention before starting the single crystal pulling, the initial liquid level position can be set correctly, and the crystal quality can be controlled with high accuracy.

In the method for producing a silicon single crystal according to the present invention, before the silicon melt is generated, the dimension of the chamfered portion at the lower end of the thermal shield is measured in advance, and the height dimension h ₁ or width of the chamfered portion is measured. it is preferable to calculate the dimension h _2, it is preferable to measure the size of the chamfered portion with a non-contact two-dimensional laser rangefinder. Previously measuring the dimensions of the chamfered portion of the thermal shield, by calculating the gap width by using the measured values of the chamfer height unit dimension h ₁ and width dimensions h _2, processing variations of the chamfered portion of the thermal shield Thus, the gap width can be accurately obtained.

In the present invention, the viewing angle theta of the silicon melt for pulling axis direction of the silicon single crystal is preferably smaller than the chamfer angle theta _c of the chamfer. In this case, the chamfering angle θ _c of the chamfered portion is preferably 45 degrees, and the viewing angle θ of the silicon melt with respect to the pulling direction of the silicon single crystal is preferably 10 degrees or more and 40 degrees or less.

  In the method for producing a silicon single crystal according to the present invention, the silicon melt seen through the opening from an obliquely upper side of the silicon melt is photographed with a camera, and the real image of the opening of the thermal shield and the liquid surface of the silicon melt It is preferable that the gap width from the lower end of the thermal shield to the liquid level of the silicon melt is determined by processing a captured image including a mirror image corresponding to the opening of the thermal shield shown in FIG. According to this, from the taking of the real image and the mirror image of the heat shield to the calculation of the gap width can be automated, and the gap width is adjusted by adjusting the height position of the quartz crucible 11 based on the measured current gap width. It can be reset to an appropriate value.

  ADVANTAGE OF THE INVENTION According to this invention, the manufacturing method of the silicon single crystal which can measure the gap width | variety at the time of a pulling start correctly by a mirror image method irrespective of the chamfering shape of the lower end part of a heat shield can be provided.

FIG. 1 is a side sectional view schematically showing a configuration of a single crystal manufacturing apparatus according to an embodiment of the present invention. FIG. 2 is a flowchart showing a manufacturing process of the silicon single crystal 3. FIG. 3 is an image of the inside of the chamber viewed from the viewing window 10e, and particularly shows the inside of the chamber before the seed crystal is deposited. FIG. 4 is a schematic diagram for explaining a method for calculating the gap width Ga from the lower end of the heat shield 17 to the liquid surface of the silicon melt 2 by _a mirror image method. FIG. 5 is a schematic diagram for explaining another method for calculating the gap width Ga from the lower end of the heat shield 17 to the liquid surface of the silicon melt 2 by _a mirror image method. 6A to 6E are diagrams for explaining an example of a method for measuring the chamfer dimension of the lower end portion 17b of the heat shield 17. FIG. FIG. 7 is a diagram for explaining an example of a method for measuring the chamfer dimension of the lower end portion 17 b of the heat shield 17. FIG. 8 is a bar graph showing the measurement results of gap width measured by three methods including the measuring method according to the present invention for each batch of single crystal manufacturing apparatuses A to E.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a side sectional view schematically showing a configuration of a single crystal manufacturing apparatus according to an embodiment of the present invention.

  As shown in FIG. 1, the single crystal manufacturing apparatus 1 includes a chamber 10, a quartz crucible 11 that supports the silicon melt 2 in the chamber 10, a graphite susceptor 12 that supports the quartz crucible 11, and a susceptor 12. The rotating shaft 13 to be supported, the shaft driving mechanism 14 for rotating and raising / lowering the rotating shaft 13, the heater 15 disposed around the susceptor 12, and the outer surface of the heater 15 along the inner surface of the chamber 10. A heat insulating material 16 disposed above the quartz crucible 11, a single crystal pulling wire 18 disposed above the quartz crucible 11 and coaxially with the rotary shaft 13, and the chamber 10. Is provided with a wire take-up mechanism 19 disposed above and a CCD camera 20 for photographing the inside of the chamber 10.Further, the single crystal manufacturing apparatus 1 controls the shaft processing mechanism 14, the heater 15, and the wire winding mechanism 19 based on an image processing unit 21 that processes an image captured by the CCD camera 20 and an output of the image processing unit 21. And a control unit 22.

  The chamber 10 includes a main chamber 10a and an elongated cylindrical pull chamber 10b connected to an upper opening of the main chamber 10a. The quartz crucible 11, the susceptor 12, the heater 15, and the heat shield 17 are composed of the main chamber. 10a. The pull chamber 10b is provided with a gas inlet 10c for introducing an inert gas (purge gas) such as argon gas into the chamber 10, and a gas for discharging the inert gas at the lower part of the main chamber 10a. A discharge port 10d is provided. Further, a viewing window 10e is provided on the upper portion of the main chamber 10a, and the pulled-up state (solid-liquid interface) of the silicon single crystal 3 can be observed from the viewing window 10e. The viewing angle θ from the viewing window 10e is preferably 10 to 40 degrees with respect to the vertical axis.

  The quartz crucible 11 is a quartz glass container having a cylindrical side wall and a curved bottom. In order to maintain the shape of the quartz crucible 11 softened by heating, the susceptor 12 supports the quartz crucible 11 in close contact with the outer surface of the quartz crucible 11. The quartz crucible 11 and the susceptor 12 constitute a double structure crucible that supports the silicon melt in the chamber 10.

  The susceptor 12 is fixed to the upper end portion of the rotary shaft 13 extending in the vertical direction. The lower end portion of the rotating shaft 13 passes through the center of the bottom portion of the chamber 10 and is connected to a shaft drive mechanism 14 provided outside the chamber 10. The susceptor 12, the rotating shaft 13 and the shaft driving mechanism 14 constitute a rotating mechanism and a lifting mechanism for the quartz crucible 11.

  The heater 15 is used to melt the silicon raw material filled in the quartz crucible 11 to generate the silicon melt 2. The heater 15 is a carbon resistance heating heater, and is provided so as to surround the quartz crucible 11 in the susceptor 12. Further, the outside of the heater 15 is surrounded by a heat insulating material 16, thereby enhancing the heat retaining property in the chamber 10.

  The heat shield 17 suppresses temperature fluctuation of the silicon melt 2 to form an appropriate hot zone near the solid-liquid interface, and prevents the silicon single crystal 3 from being heated by radiant heat from the heater 15 and the quartz crucible 11. It is provided for. The heat shield 17 is a graphite member that covers the upper region of the silicon melt 2 excluding the pulling path of the silicon single crystal 3, and has an inverted truncated cone shape whose diameter is reduced from the upper side to the lower side. ing.

  A circular opening 17a larger than the diameter of the silicon single crystal 3 is formed in the center of the lower end of the heat shield 17, and a pulling path for the silicon single crystal 3 is secured. As shown in the figure, the silicon single crystal 3 is pulled upward through the opening 17a. The diameter of the opening 17 a of the heat shield 17 is smaller than the diameter of the quartz crucible 11 and the lower end of the heat shield 17 is located inside the quartz crucible 11. The heat shield 17 does not interfere with the quartz crucible 11 even if it is raised further upward.

The heat shield 17 also functions as a gas rectifying member that rectifies the gas flow in the vicinity of the surface of the silicon melt 2. As the silicon single crystal 3 grows, the melt amount decreases and the liquid level in the quartz crucible 11 decreases, but the distance from the liquid surface of the silicon melt 2 to the lower end of the heat shield 17 (gap width G By raising the quartz crucible 11 so that _a ) becomes constant, the flow velocity of the gas flowing near the melt surface (purge gas guiding path) can be made constant. Therefore, the temperature fluctuation of the silicon melt 2 can be suppressed and the evaporation amount of the dopant from the silicon melt 2 can be controlled, and the crystal defect distribution, oxygen concentration distribution, resistivity distribution, etc. in the pulling axis direction of the single crystal can be controlled. Stability can be improved.

  Although details will be described later, the lower edge of the lower end portion 17b of the heat shield 17 is chamfered. The chamfering is, for example, C1 chamfering, the chamfering angle is 45 degrees, and the height dimension and width dimension of the chamfered part are both 1 mm. By chamfering the edge of the opening 17a at the lower end of the heat shield 17, the heat resistance of the heat shield 17 can be increased, and cracks due to thermal stress can be prevented.

  Above the quartz crucible 11, a wire 18 that is a pulling shaft of the single crystal 3 and a wire winding mechanism 19 that winds the wire 18 are provided. The wire winding mechanism 19 has a function of rotating the single crystal together with the wire 18. The wire winding mechanism 19 is disposed above the pull chamber 10b, the wire 18 extends downward from the wire winding mechanism 19 through the pull chamber 10b, and the tip of the wire 18 is located inside the main chamber 10a. The space has been reached. FIG. 1 shows a state in which the silicon single crystal 3 being grown is suspended from the wire 18. When pulling up the single crystal, the seed crystal is immersed in the silicon melt 2 and the single crystal is grown by gradually pulling up the wire 18 while rotating the quartz crucible 11 and the seed crystal, respectively.

  A gas inlet 10c for introducing an inert gas into the chamber 10 is provided at the top of the pull chamber 10b, and a gas exhaust for exhausting the inert gas in the chamber 10 is provided at the bottom of the main chamber 10a. An outlet 10d is provided. The inert gas is introduced into the chamber 10 from the gas inlet 10c, and the introduction amount is controlled by a valve. Further, since the inert gas in the sealed chamber 10 is exhausted from the gas outlet 10d to the outside of the chamber 10, the SiO gas and CO gas generated in the chamber 10 are collected to keep the inside of the chamber 10 clean. Is possible. Although not shown, a vacuum pump is connected to the gas discharge port 10d through a pipe, and the flow rate is controlled by a valve while sucking an inert gas in the chamber 10 by the vacuum pump. Is kept at a constant reduced pressure.

  A viewing window 10e for observing the inside is provided at the upper part of the main chamber 10a, and the CCD camera 20 is installed outside the viewing window 10e. The captured image of the CCD camera 20 may be grayscale or color. During the single crystal pulling process, the CCD camera 20 takes an image of the boundary portion between the silicon single crystal 3 and the silicon melt 2 that can be seen from the viewing window 10 e through the opening 17 a of the heat shield 17.

  The CCD camera 20 takes an image of the silicon melt obliquely from above. The CCD camera 20 is connected to an image processing unit 21, the captured image is processed by the image processing unit 21, and the processing result is used by the control unit 22 for controlling the lifting conditions.

  FIG. 2 is a flowchart showing a manufacturing process of the silicon single crystal 3.

As shown in FIG. 2, in the manufacture of the silicon single crystal 3, the silicon raw material 3 such as polycrystalline silicon previously filled in the quartz crucible 11 is heated by the heater 15 to generate the silicon melt 2 (step S <b> 11). Next, the liquid level position (gap width G _a ) of the silicon melt 2 viewed from the heat shield 17 is measured (step S12). Thereafter, the seed crystal attached to the tip end portion of the wire 18 is lowered and deposited on the silicon melt 2 (step S13). The amount of seed crystal drop at this time is determined based on a gap width measured in advance.

  Next, a single crystal pulling process is started for growing the single crystal by gradually pulling the seed crystal while maintaining the contact state with the silicon melt. In the single crystal pulling step, first, seed drawing (necking, step S14) is performed by the dash neck method in order to make the single crystal dislocation-free. Next, in order to obtain a single crystal having a required diameter, a shoulder portion having a gradually widened diameter is grown (step S15), and when the single crystal has a desired diameter, a body portion whose diameter is maintained constant is grown. (Step S16). After the body portion is grown to a predetermined length, tail restriction (tail portion growth, step S17) is performed to separate the single crystal from the silicon melt 2 in a dislocation-free state.

During the pulling process of the single crystal, the diameter of the silicon single crystal 3 and the liquid surface position of the silicon melt 2 are controlled. The controller 22 controls pulling conditions such as the pulling speed of the wire 18 and the power of the heater 15 so that the diameter of the single crystal 3 becomes the target diameter. The control unit 22 controls the vertical position of the quartz crucible 11 so that the liquid level position (gap width G _a ) is constant.

  FIG. 3 is an image of the inside of the chamber viewed from the viewing window 10e, and particularly shows the inside of the chamber before the seed crystal is deposited.

As shown in FIG. 3, the silicon melt 2 in the quartz crucible 11 can be seen through the opening 17a of the heat shield 17 from the viewing window 10e. The inside of the edge 17ae of the opening 17a of the heat shield 17 is all the liquid surface of the silicon melt 2. Further, a mirror image 17 ′ of the heat shield 17 is reflected on the liquid surface of the silicon melt 2, and the edge 17′ae of the opening of the mirror image 17 ′ of the heat shield 17 can be recognized. That is, the edge 17 ′ ae is a mirror image corresponding to the edge 17 ae of the opening 17 a of the heat shield 17 reflected on the liquid surface of the silicon melt 2. Edge when the gap width G _a from the lower end of the thermal shield 17 to the liquid surface of the silicon melt 2 is relatively wide mirror 17 of the thermal shield 17 from the edge 17ae of the opening 17a of the real image of the thermal shield 17 '17 'distance D to ae becomes long, the distance D is shortened when the gap width G _a is relatively narrow.

The liquid level position of the silicon melt 2 with respect to the heat shield 17 before starting the pulling of the silicon single crystal can be measured by a so-called mirror image method using a mirror image 17 ′ of the heat shield 17. Next, _a method for measuring the initial gap width Ga by the mirror image method will be described.

FIG. 4 is a schematic diagram for explaining a method for calculating the gap width Ga from the lower end of the heat shield 17 to the liquid surface of the silicon melt 2 by _a mirror image method.

As shown in FIG. 4, in the lower corner of the lower end portion 17b of the thermal shield 17 is chamfered processing is given, and substantially vertical side end surfaces S _1, a horizontal plane positioned below the side end surface S ₁ or It has a lower end surface S ₂ composed of a substantially inclined surface, and a chamfered portion (square surface) S ₃ positioned between the side end surface S ₁ and the lower end surface S ₂ . Chamfer S ₃ is formed by C1 chamfered example the lower corner defined by the side edge surface S ₁ and the lower end surface S _2, the inclination angle (bevel angle) theta _c of the chamfer S ₃ with respect to the vertical plane The height dimension h ₁ and the width dimension h ₂ are both 1 mm at 45 degrees. Angle of chamfer S ₃ is not limited to 45 degrees with respect to the vertical plane may be smaller than the angle of the lower end surface S _2. The present invention is a method for calculating the gap width G _a from the lower end of the heat shield 17 to the liquid surface of the silicon melt 2 by the mirror image method.

In the calculation of the gap width G _a by the mirror image method, the opening of the mirror image 17 ′ of the thermal shield 17 ′ from the edge 17ae of the real image aperture 17a of the thermal shield 17 existing in the observation region viewed from the observation window 10e of the chamber 10 is calculated. A distance G _c (measured distance) to the edge 17′ae is calculated. The distance G _c is equivalent to the distance D in FIG. 3. And unless chamfering processing is given to the edge of the opening 17a of the thermal shield 17, the half value G _{c /} 2 of the distance G _c is the gap width G _a.

However, the C1 chamfered lower corners of the opening 17a of the thermal shield 17 is applied as, the visibility angle theta when observing the silicon melt 2 obliquely from above is smaller than the chamfer angle theta _c (e.g. 30 If degrees) can not only see the upper edge E ₁ of the chamfered portion S _3, it can not be seen the lower edge E ₂ of the chamfered portion S _3. That is, so that the edge of the opening 17a of the real image of the thermal shield 17 visible from the viewing window 10e of the chamber 10 17ae (see FIG. 3) coincides with the lower end of the thermal shield 17 when there is no chamfered portion S ₃ , the edge of the opening 17a of the thermal shield 17 17Ae (see FIG. 3) is not coincident with the lower end of the thermal shield 17 when there is a chamfer S _3, the lower end of the thermal shield 17 is the back side of the chamfer S ₃ It will be hidden and hidden from view. When calculating the gap width G _a in such a state, the measurement error always occurs in the gap width G _a due to the chamfered portion S _3, it is impossible to accurately determine the liquid level position of the silicon melt 2 .

The present embodiment, by correcting the measured distance from the measured distance based on the position 17ae of the opening 17a of the edge of the thermal shield 17 by removing the influence of the chamfered portion S _3, to accurately calculate the gap width G _a Is. Specifically, the height dimension h ₁ of the chamfered portion S ₃ from the half value G _c / 2 of the distance G _c between the real image of the edge 17 ae of the opening 17 a of the heat shield 17 and the edge 17 ′ ae that is a mirror image thereof. by subtracting, it can be determined gap width G _a. In other words, the gap width G _a can be obtained from the following equation (1).

The method of calculating the gap width G _a as described above may be performed by processing an image by the CCD camera 20 is captured by the image processing unit 21, the result of measurement by visual while observing the worker viewing window 10e May be used. After adjusting the height position of the quartz crucible 11 based on the current gap width G _a thus measured and resetting the gap width G _a to an appropriate value, the step of pulling up the silicon single crystal is started.

FIG. 5 is a schematic diagram for explaining another method for calculating the gap width Ga from the lower end of the heat shield 17 to the liquid surface of the silicon melt 2 by _a mirror image method.

As described above, the lower edge E ₂ of the chamfered portion S ₃ cannot be visually recognized with respect to the real image of the heat shield 17. However, the mirror image 17 of the thermal shield 'for the upper edge E ₁ of the chamfered portion S _3' it is possible to visually recognize the lower edge E ₂ as well _'. Mirror image 17 of the thermal shield 'when determining visually the opening edge 17'Ae (see FIG. 3) of the upper edge E ₁ of the chamfered portion S _3' viewing is near the bottom side edge E _{2 'than} It may be easy. Therefore, as shown in FIG. 5, in this embodiment, a distance G from the upper edge E _{1 at} the lower end of the real image of the heat shield 17 to the lower edge E ₂ ′ at the lower end of the mirror image 17 ′ of the heat shield. _m is calculated, and the gap width G _a is calculated from the distance G _m . In this case, the gap width G _a can be obtained from the following equation (Expression 2) using the height dimension h ₁ and the width dimension h ₂ of the chamfered portion S ₃ .

The position of the quartz crucible 11 in the height direction is adjusted based on the current gap width G _a measured in this way to reset the gap width G _a to an appropriate value, and then the silicon single crystal pulling process is started. .

The chamfer dimension of the edge of the opening 17 a of the heat shield 17 may be measured in advance before using the heat shield 17. Since the chamfer dimension there are some processing variation for each thermal shield 17, it is possible to improve the measurement accuracy of the gap width G _a by using actually measured chamfer dimension. Because chamfer dimension be used repeatedly heat shield 17 is hardly changed, chamfer dimension may if measured prior to first use of the thermal shield 17, thereby taking into account the effect of the chamfered portion S ₃ of precise calculation of the gap width G _a. The chamfer dimension of the heat shield 17 is preferably measured in a single state, not in the state of being installed in the chamber 10.

  FIGS. 6A to 6E and FIG. 7 are diagrams for explaining an example of a method for measuring the chamfer dimension of the lower end portion 17b of the heat shield 17.

  As shown in FIG. 6A, a non-contact two-dimensional laser rangefinder 23 can be used for measuring the chamfer dimension of the lower end portion 17 b of the heat shield 17. The measurement data of the contour shape of the corner measured using the non-contact two-dimensional laser distance meter 23 is displayed on two-dimensional coordinates.

Next, as shown in FIG. 6 (b), sets a first approximate straight line LN ₁ and the second approximate straight line LN ₂ respectively on the side end surface and the lower end surface of the corner portion. Next, as shown in FIG. 6 (c), the first approximate straight line LN ₁ so that the angle theta _x formed by the first approximate straight line LN ₁ and the second approximate straight line LN ₂ there is an angle of the design of the corner portion The angle of the second approximate straight line LN ₂ is adjusted.

Next, as shown in FIG. 6D, the position of the intersection point P ₀ between the first approximate line LN ₁ and the second approximate line LN ₂ is aligned with the origin of the two-dimensional coordinates, and the first approximate line LN _{1 with} respect to the X axis is set. angle and rotating the corners of the contour as the angle and the second approximate straight line LN ₂ equals about the origin of the two-dimensional coordinates. When the angle of the corner portion before chamfering is 90 degrees, the angles of the two approximate lines are both 45 degrees.

Next, as shown in FIG. 6 (e), it sets the third approximate straight line LN ₃ is a straight line connecting the beveled start point _{P 1} and the chamfered end point _{P 2.} Here, the chamfered start point _{P 1} is a point of the first on approximate straight line LN ₁ a deviation between the measured value is 0.1mm for example, the chamfered end point _{P 2} has a deviation between the measured value is, for example, 0.1mm It becomes a single point of a second on the approximate straight line LN _2. Then, the intersection point P ₀ and the distance D ₁ of the between the chamfered start point P ₁ can be obtained as the height h ₁ of the chamfered portion, the intersection P ₀ and chamfer distance D ₂ between the chamfered end point P ₂ it can be determined as a width dimension h _2.

As shown in FIG. 7, the lower end surface S ₂ of the thermal shield 17 is slightly inclined rather than horizontal, when the angle of the lower corner consisting of the side end face S ₁ and the lower end surface S ₂ is smaller than 90 degrees, If the height dimension h ₁ of the chamfered portion S ₃ is directly calculated by the method shown in FIG. 6, ΔD ₁ = D ₂ sin θ _d (θ _d is an inclination angle of the lower end surface S _{2 with} respect to the horizontal plane) is different from the actual one. Value. In this case, the correct height dimension h ₁ of the chamfered portion S ₃ can be obtained by subtracting ΔD ₁ from the distance D ₁ between the intersection point P ₀ and the chamfering start point P ₁ . Further, the correct width h ₂ of the chamfered portion S ₃ can be obtained from h ₂ = D ₂ sin θ _d .

As described above, the silicon single crystal manufacturing method according to the present embodiment is between the real image of the heat shield 17 existing in the observation region and the mirror image 17 ′ of the heat shield reflected on the surface of the silicon melt. calculating the measured distance, by removing the influence of the chamfered portion S ₃ of the lower end of the thermal shield 17 from the measured distance, the gap width G _a from the lower end of the thermal shield 17 to the liquid surface of the silicon melt 2 Therefore, the measurement accuracy of the liquid surface position of the silicon melt 2 with respect to the thermal shield 17 can be increased. Further, the dimensions of the chamfered portion S ₃ of the thermal shield 17 measured in advance, so calculating the gap width G _a using measured value of the dimension of the chamfered portion S _3, the chamfered portion S ₃ of the thermal shield 17 by suppressing the influence of processing variations can be determined accurately gap width G _a.

  The preferred embodiments of the present invention have been described above, but the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention. Needless to say, it is included in the range.

  For example, although the case where C chamfering was given to the lower end part of the heat shield in the above-mentioned embodiment was mentioned as an example, the case where R chamfering is given may be used. Moreover, although the calculation method of the initial liquid level position was demonstrated in the said embodiment, it is also possible to apply to the calculation method of the liquid level position in a single crystal pulling process.

  An evaluation test was performed on the gap width measurement method according to the present invention. For the evaluation test, five single crystal production apparatuses A to E were used, and the heat shields installed in the chambers of the respective apparatuses were chamfered at the lower end opening. Although the design chamfer dimension is C1 chamfer, the actual chamfer dimension varies depending on the apparatus due to the influence of processing variations. The evaluation tests for measuring the gap width G between the lower end of the heat shield and the surface of the silicon melt were performed four times for the devices A and D and twice for the devices B, C, and E, respectively.

  In each evaluation experiment, after melting a raw material in a quartz crucible installed in a chamber to generate a silicon melt, a conventional method that does not consider the chamfered shape of the heat shield, a method using a quartz pin, and a heat shield The gap width of the silicon melt was simultaneously measured by three methods according to the present invention in consideration of the chamfered shape. As described in the prior art, the quartz pin is attached to the end facing the melt surface of the heat shielding member covering the melt surface, and the position where the quartz pin is in contact with the melt surface is used as a reference. In this way, the gap width is calculated. Since this method has high measurement accuracy, it can be referred to as the true value of the gap width, and it can be said that the measurement accuracy closer to this value is higher.

  FIG. 8 is a bar graph showing the measurement results of the gap width measured by the above three methods for each batch of the single crystal manufacturing apparatuses A to E. The vertical axis of the bar graph is the relative value of the gap width, and one scale on the vertical axis is 0.5 mm. The origin (0 mm position) on the vertical axis indicates a reference point that is a fixed distance away from the absolute origin, and the longer the bar graph, the greater the gap width.

  As shown in FIG. 8, the gap width value measured by the conventional method is larger than the gap width measured by the method using the quartz pin in any batch regardless of the single crystal manufacturing apparatuses A to E. Therefore, the error for the true value was large. On the other hand, the gap width value measured by the method of the present invention was almost the same as the gap width value measured by the method using a quartz pin. That is, it became clear that the gap width measurement method according to the present invention has higher measurement accuracy than the conventional method. The gap value measurement results tended to differ from device to device, but this was due to the fact that the measurement conditions such as the variation in the chamfered shape of the heat shield and the mounting state of the heat shield differ slightly from device to device. It is guessed.

DESCRIPTION OF SYMBOLS 1 Single crystal manufacturing apparatus 2 Silicon melt 3 Silicon single crystal 10 Chamber 10a Main chamber 10b Pull chamber 10c Gas inlet 10d Gas outlet 10e Viewing window 11 Quartz crucible 12 Susceptor 13 Rotating shaft 14 Shaft drive mechanism 15 Heater 16 Heat insulating material 17 Thermal shield (real image)
17 ′ Mirror image of heat shield 17′ae Mirror image of opening edge of heat shield 17a Heat shield opening (real image)
17ae Edge of thermal shield opening (real image)
17b Lower end 18 of heat shield 18 Wire 19 Wire winding mechanism 20 CCD camera 21 Image processing unit 22 Control unit 23 Non-contact two-dimensional laser distance meter D Distance between real image and mirror image of heat shield D From _{one surface} capture start point to intersection Distance D ₂ Distance from the chamfering end point to the intersection E ₁ Upper edge of the chamfered portion of the real image of the heat shield E ₂ Lower edge E ₁ ′ of the chamfered portion of the real image of the heat shield Upper edge E ₂ ′ Lower edge G _a Gap width G _c of mirror image chamfered portion of heat shield Distance from upper edge E ₁ of real image of heat shield to upper edge E ₁ ′ of mirror image G _m width LN ₁ first approximate straight line LN ₂ second approximate straight line LN ₃ third approximate straight line of height _{h 2} chamfer distance _{h 1} chamfer from the upper edge _{E 1} of the real image to the bottom edge _{E 2} 'mirror image P ₀ first approximate straight line And the second approximate straight line P ₁ chamfering start point P ₂ chamfering end point S ₁ side end surface S ₂ lower end surface S ₃ chamfered portion (corner surface)
θ Viewing angle of silicon melt θ _c Chamfer angle

Claims (10)

When placing a heat shield above the crucible containing the silicon melt and covering the region excluding the silicon single crystal pulling path, when pulling the silicon single crystal from the silicon melt by the Czochralski method, A method for producing a silicon single crystal, wherein the silicon single crystal is pulled upward through a circular opening provided at a lower end of the thermal shield,
The thermal shield reflected on the liquid surface of the silicon melt from a real image of the opening of the thermal shield existing in an observation region when the silicon melt seen through the opening is viewed from obliquely above the silicon melt. Calculate the measured distance to the mirror image corresponding to the aperture of
From the half value of the measured distance, calculate a correction distance that removes the effect of chamfering applied to the lower end of the thermal shield,
A method for producing a silicon single crystal, wherein a gap width from a lower end of the thermal shield to a liquid surface of the silicon melt is obtained using the correction distance. The distance G _c from the real image of the upper end edge of the chamfer of the heat shield to the mirror image corresponding to the upper edge appearing on the liquid surface of the silicon melt and the actual distance, the height of the chamfered portion h ₁ And when
The gap width _{G a,}
The method for producing a silicon single crystal according to claim 1, obtained by: The distance G _m to mirror image corresponding to the lower end edge of the chamfered portion of the heat shield from the real image of the upper end edge of the chamfer of the thermal shield reflected on the liquid surface of the silicon melt and the actual distance, the chamfered When the height dimension of the part is h ₁ , the width dimension of the chamfered part is h ₂ , and the viewing angle θ of the silicon melt with respect to the pulling direction of the silicon single crystal,
The gap width _{G a,}
The method for producing a silicon single crystal according to claim 1, obtained by: Melting the solid material filled in the crucible to produce a silicon melt;
Depositing seed crystals on the surface of the silicon melt;
And raising the seed crystal to grow a silicon single crystal,
The step of obtaining the gap width from the lower end of the thermal shield to the surface of the silicon melt is performed after the silicon melt is generated and before the seed crystal is deposited. The manufacturing method of the silicon single crystal as described in any one of these.   The method for producing a silicon single crystal according to any one of claims 1 to 4, wherein a dimension of a chamfered portion at a lower end portion of the thermal shield is measured in advance before the silicon melt is generated.   The method for producing a silicon single crystal according to claim 5, wherein the dimension of the chamfered portion is measured using a non-contact two-dimensional laser distance meter. The method for producing a silicon single crystal according to claim 1, wherein a viewing angle θ of the silicon melt with respect to a pulling axis direction of the silicon single crystal is smaller than a chamfering angle θ _{c of the} chamfered portion. . The method for producing a silicon single crystal according to claim 7, wherein a chamfering angle θ _c of the chamfered portion is 45 degrees.   The method for producing a silicon single crystal according to claim 7 or 8, wherein a viewing angle θ of the silicon melt with respect to a pulling direction of the silicon single crystal is 10 degrees or more and 40 degrees or less.   The silicon melt that can be seen through the opening from obliquely above the silicon melt is photographed with a camera, and corresponds to the real image of the opening of the thermal shield and the opening of the thermal shield that is reflected on the surface of the silicon melt. 10. The silicon single crystal according to claim 1, wherein the gap width from the lower end of the thermal shield to the liquid surface of the silicon melt is determined by processing a captured image including a mirror image. Manufacturing method.