JP2020132483A - Method for manufacturing cz silicon single crystal - Google Patents

Abstract

To provide a method for manufacturing a CZ silicon single crystal, capable of performing a cone step without reducing a yield by deterioration in productivity due to a long cone step time and increase in cone weight in a DFS method, reducing a variation in in-furnace heat environment due to a variation in the cone step and improving the quality of a wafer and a product yield.SOLUTION: The method for manufacturing a CZ silicon single crystal comprises: the step of attaching a seed by a DFS method; the crystal diameter stabilization step, the cone step of forming a cone part; and the straight body step of forming a straight body part. In the cone step, a cone is formed by correcting a heater output value using a correction amount obtained by measuring the crystal diameter of a crystal pulled in the crystal diameter stabilization step, calculating a difference between the measured crystal diameter and the reference diameter of a predetermined crystal and calculating the correction amount of heater output in the cone step according to the difference.SELECTED DRAWING: Figure 2

Description

The present invention relates to a method for producing a silicon single crystal by the CZ method.

In the production of a silicon single crystal by the CZ method, a seed crystal is usually deposited in a silicon melt, and a crystal diameter increasing step (cone step) for expanding the diameter to the product diameter is performed to grow a product sampling diameter portion (straight body step). ).

At this time, in order to obtain a dislocation-free single crystal used for a silicon semiconductor device, it is necessary to dislocate the crystal before the cone step, but there are mainly two types of methods. One is a method called the necking method, in which dislocations introduced by heat shock during seed crystal landing are narrowed down to a diameter of 2-5 mm and then grown to a length of about 10-20 mm to form a crystal surface ( This is a method of discharging dislocations from the side surface). This is because the dislocation plane forms an angle of 54.74 ° with respect to <100> growth and 70.53 ° with respect to <111> growth, so that the dislocation can be eliminated by making the crystal thinner.

In the necking method, it is necessary to narrow the diameter to about 2-3 mm, so temperature control during drawing is important. If the temperature is too high, there are problems such as the contact part with the melt completely melts and breaks during necking, and the strength of the thinnest part is weak and it becomes impossible to support the crystal weight that is grown after that. .. Further, when the temperature is lowered and the diameter is increased, there is a problem that dislocations cannot be completely removed and a dislocation-free single crystal cannot be obtained. Regarding the temperature control of such a necking process, Patent Documents 1 and 2 disclose techniques for controlling the necking process.

The other method is a method called Dislocation Free Seeding (hereinafter referred to as "DFS method"), in which the tip of the seed crystal is sharpened and slowly brought into contact with the melt. This is a method that enables dislocation-free growth without necking (squeezing) by avoiding the introduction of dislocations due to thermal shock and then melting the seed crystal until it reaches a predetermined diameter.

The DFS method has the advantage of eliminating the need for a necking step, but in order to succeed in dislocation-free seeding, as described in Patent Document 3, the tip of the seed crystal is sharpened and the seed crystal is shaped. Further, it is necessary to reduce the thermal shock at the time of landing by raising the melt temperature to keep the seed crystal warm.

Special Fair 6-12590 Gazette Special Fair 7-17475 Gazette JP-A-11-240793 Japanese Unexamined Patent Publication No. 2005-15287 Japanese Unexamined Patent Publication No. 4-149902

When seeding is carried out at a high melt temperature by the DFS method, during the cone process, the temperature is lowered from a high temperature to increase the diameter to the straight body, so that the cone process time becomes longer and the cone weight becomes heavier. There was a problem that it became a factor of lowering the yield.

In addition, for low-defect wafers and defect-free wafers required for recent silicon semiconductor devices that are becoming finer, the cone process (time / weight) is performed by seeding in a relatively wide seeding temperature range where the DFS method is successful. As described in Patent Document 4, the thermal environment in the furnace, such as the distance between the melt surface and the heat shield on the upper part of the melt and the relative position with the heater, changes. As a result, the variation in silicon single crystal quality causes problems such as deterioration of the yield of the final product.

To solve such a problem, Patent Document 5 discloses a method of correcting the heater power based on the diameter expansion speed during the cone process, but in recent CZ single crystal growth using a large-diameter crucible, , In-core parts such as crucibles, large heat capacity of raw material melt, and longer response time for temperature control of the growing part in the center of the crucible by the heater on the outer periphery of the crucible, so the power correction in the cone is effective. There was a problem that it didn't work.

As described above, in the CZ single crystal production using the DFS method, an effective technique for controlling the variation in cone process time and weight while making the DFS method successful has not been disclosed so far.

The present invention has been made to solve the above problems, and even if the melt temperature at the time of seeding (liquidation) is raised in order to increase the success rate of the DFS method, the cone process time is lengthened. The cone process is carried out without reducing the yield due to the deterioration of productivity and the increase in cone weight, and the variation in the thermal environment in the furnace due to the variation (time, weight) in the cone process is reduced to improve the quality of low-defect and defect-free wafers. An object of the present invention is to provide a method for producing a CZ silicon single crystal, which improves the product yield.

The present invention has been made to achieve the above object, and includes a seeding step in which the tip of a seed crystal having a pointed tip is brought into contact with a silicon melt and then melted and seeded. A CZ silicon single crystal growth method including a crystal diameter stabilizing step of raising a crystal to a predetermined length to stabilize the crystal diameter, a cone step of forming a cone portion, and a straight cylinder step of forming a straight body portion. , The crystal diameter of the crystal pulled up in the crystal diameter stabilization step is measured, the amount of deviation of the measured crystal diameter with respect to a predetermined reference diameter of the crystal is calculated, and the amount of deviation in the cone step is calculated according to the deviation amount. Provided is a method for producing a CZ silicon single crystal, which calculates a correction amount of a heater output and corrects a heater output value using the correction amount in the cone step to form a cone.

According to such a CZ silicon single crystal manufacturing method, productivity deteriorates due to a long cone process time and yield decreases due to an increase in cone weight, regardless of the variation in melt temperature during seeding (liquidation) of the DFS method. It is possible to carry out the cone process without inviting, reduce the variation in the thermal environment in the furnace due to the variation in the cone process (time, weight), improve the quality of low-defect and defect-free wafers, and improve the product yield. can do.

At this time, when calculating the correction amount of the heater output, if the deviation amount is 0 or more, the correction amount is set to 0 or more, and if the deviation amount is less than 0, the correction amount is set to less than 0. It can be a CZ silicon single crystal manufacturing method.

As a result, the cone process time becomes longer when the measurement diameter with a deviation amount of less than 0 is smaller than the reference diameter, the cone weight increases, and the cone process time becomes shorter when the measurement diameter with a deviation amount of 0 or more is larger than the reference diameter. It is possible to suppress variations in the cone process due to time and reduction in cone weight, reduce variations in the thermal environment in the furnace, improve the quality of low-deficiency and defect-free wafers, and improve product yield. In addition, by setting the standard to a large (thick) diameter, it is possible to obtain the effects of improving productivity and yield by shortening the cone process time and reducing the weight of the cone while suppressing variation.

At this time, when calculating the correction amount of the heater output, the correction amount is set to 0.5 kW or more and 5.0 kW or less, more preferably 1.0 kW or more and 2.0 kW or less per 1 mm of the deviation amount. It can be a silicon single crystal production method.

In such a range, the correlation between the correction amount of the heater output and the deviation amount is higher, so that the cone process time and the variation in the cone weight can be suppressed more stably.

At this time, when calculating the correction amount of the heater output, the absolute value of the correction amount per 1 mm of the deviation amount when the deviation amount is 0 or more is used, and the deviation amount is less than 0. It is possible to use a CZ silicon single crystal manufacturing method in which the amount of deviation is less than the absolute value of the correction amount per 1 mm.

As a result, when the measured diameter with a deviation amount of 0 or more is larger than the reference diameter, the effect of shortening the cone process time and reducing the cone weight further improves the productivity, and further improves the yield by reducing the cone weight. Obtainable.

At this time, the CZ silicon single crystal manufacturing method can be used in which the measurement of the crystal diameter is performed when the length of the crystal to be pulled up in the crystal diameter stabilizing step is 10 mm or more and 100 mm or less.

As a result, the crystal diameter can be measured more accurately, so that the accuracy of the correction value can be improved and the decrease in productivity can be further suppressed.

At this time, the correction of the heater output value is performed within 60 minutes from the start of the cone process or 60 minutes before the start of the straight body process, whichever is longer. It can be a crystal production method.

As described above, the correction of the heater output value may be performed 60 minutes before the start of the straight body process, but if it is performed within 60 minutes from the start of the cone process, the cone process time is unclear. Can be corrected without any problem. As a result, the productivity and yield improvement effect can be obtained more reliably. Further, even if the time required to reach the product sampling diameter portion (straight body process) is unknown, appropriate correction can be performed more reliably.

As described above, according to the CZ silicon single crystal production method of the present invention, the cone process time is long regardless of the variation in the melt temperature at the time of seeding (landing) of the dislocation-free seeding method (DFS method). It is possible to carry out the cone process without deterioration of productivity due to conversion and decrease in yield due to increase in cone weight, and in the production of CZ silicon single crystal using the DFS method, variation in the thermal environment in the furnace due to variation in the cone process (time, weight). It is possible to improve the quality of low-defect / defect-free wafers and improve the product yield.

An example of a silicon single crystal manufacturing apparatus is shown. The relationship between the measured crystal diameter deviation (horizontal axis) and the cone process time (vertical axis) with respect to a predetermined crystal reference diameter is shown. The comparison result of the cone weight reduction effect is shown. The comparison result of the cone weight variation is shown. The defect evaluation result for PW of the comparative example is shown.

Hereinafter, the present invention will be described in detail, but the present invention is not limited thereto.

As described above, even if the melt temperature at the time of seeding (liquidation) is raised in order to increase the success rate of the DFS method, the productivity deteriorates due to the lengthening of the cone process time and the yield decreases due to the increase in cone weight. CZ silicon single crystal that implements the cone process without causing it, reduces the variation in the thermal environment in the furnace due to the variation (time, weight) in the cone process, improves the quality of low-defect / defect-free wafers, and improves the product yield. A manufacturing method was required.

As a result of diligent studies on the above-mentioned problems, the present inventors have brought the tip of a seed crystal having a pointed tip into contact with a silicon melt, and then melted and seeded the seed crystal. A CZ silicon single crystal growth method including a crystal diameter stabilizing step of raising a predetermined length to stabilize the crystal diameter, a cone step of forming a cone portion, and a straight cylinder step of forming a straight body portion. The crystal diameter of the crystal pulled up in the crystal diameter stabilizing step is measured, the amount of deviation of the measured crystal diameter with respect to a predetermined crystal reference diameter is calculated, and the heater in the cone step is calculated according to the deviation amount. When the DFS method is seeded (liquid landing) by the CZ silicon single crystal manufacturing method in which the output correction amount is calculated and the heater output value is corrected using the correction amount in the cone step to form a cone. Regardless of the variation in melt temperature, it is possible to carry out the cone process without causing deterioration in productivity due to the lengthening of the cone process time and reduction in yield due to an increase in cone weight, and variation in the cone process (time, weight). ), The variation in the thermal environment in the furnace can be reduced, the quality of low-defect / defect-free wafers can be improved, and the product yield can be improved, and the present invention has been completed.

Hereinafter, description will be made with reference to the drawings.

FIG. 1 shows a single crystal manufacturing apparatus 100 used for manufacturing a CZ silicon single crystal according to the present invention.

The silicon single crystal manufacturing apparatus 100 shown in FIG. 1 is composed of a main chamber 9a and a pull chamber 9b communicating with the main chamber 9a. Inside the main chamber 9a, a graphite crucible 1b and a quartz crucible 1a are installed on the support shaft 6. A heater 2 is provided so as to surround the graphite crucible 1b and the quartz crucible 1a, and the heater 2 melts the raw material silicon polycrystal contained in the quartz crucible 1a to obtain the raw material melt 3. Further, a heat insulating cylinder 8a and a heat insulating plate 8b are provided to prevent the influence of the radiant heat from the heater 2 on the main chamber 9a and the like.

On the melt surface of the raw material melt 3, heat shield members 11 are arranged to face the melt surface at predetermined intervals to block radiant heat from the melt surface of the raw material melt 3.

Further, when the single crystal is grown, an inert gas such as argon gas is introduced as a purge gas from a gas introduction port (not shown) in the upper pull chamber 9b, and is between the silicon single crystal 4 being pulled up and the gas rectifying cylinder 10. After passing through, it passes between the heat shield member 11 and the melt surface of the raw material melt 3, and is discharged from the gas outlet (not shown) at the lower part of the main chamber 9a. By controlling the flow rate of the gas to be introduced and the amount of gas discharged by the pump or valve, the pressure in the chamber being pulled up is controlled.

When growing a crystal by the CZ method, a magnetic field may be applied by a magnetic field application device (not shown). Further, the single crystal manufacturing apparatus 100 may be controlled by a control unit (not shown).

The silicon single crystal is produced by the CZ method as follows using the above-mentioned single crystal production apparatus 100. After the raw material is made into the raw material melt 3 in the crucible 1, the tip of the seed crystal 7 installed on the pulling shaft 5 is landed in the raw material melt 3 in the crucible 1a, and the strength is such that the weight of the pulled crystal is maintained. As described above, the tip of the seed crystal 7 is submerged for a predetermined length and immersed to melt and seed the seed crystal (hereinafter referred to as "seed step"). At this time, the length of the seed crystal 7 to be submerged is preferably, for example, about 20 mm or more and 40 mm or less. Within such a range, the desired strength can be stably obtained. After that, the seed crystal 7 is pulled up to a predetermined length to stabilize the crystal diameter (crystal diameter) (hereinafter, referred to as "crystal diameter stabilizing step"). At this time, the predetermined length for pulling up the seed crystal 7 can be, for example, 10 mm or more and 100 mm or less. Within such a range, the crystal diameter can be stabilized more stably.

After the crystal diameter becomes stable, a process of expanding the diameter to the product diameter is performed (hereinafter referred to as "cone process"). Then, after the crystal to be pulled up reaches the product diameter, the straight body portion of the ingot is grown while maintaining the crystal diameter (hereinafter referred to as "straight body process").

As described above, the rod-shaped silicon single crystal 4 is pulled up from the raw material melt 3. The crucibles 1a and 1b can be moved up and down in the crystal growth axis direction, and the crucibles 1a and 1b are grown during the growth so as to compensate for the decrease in the liquid level of the raw material melt 3 as the growth of the silicon single crystal 4 progresses. The height of the melt surface of the raw material melt 3 is kept substantially constant by increasing the amount.

Next, the feature points in the CZ silicon single crystal manufacturing method according to the present invention will be described.

First, the measurement of the crystal diameter of the crystal pulled up in the crystal diameter stabilizing step will be described. As described above, after the seeding step, a crystal diameter stabilizing step is performed in which the seed crystal is pulled up to a predetermined length to stabilize the crystal diameter. The crystal diameter of the crystal pulled up in the crystal diameter stabilizing step is measured when it is pulled up to a predetermined length, for example, 10 mm or more, preferably 30 mm or more. Within such a range, the crystal diameter is stable, so that more accurate measurement can be performed. The upper limit is not particularly limited, but may be 100 mm or less, preferably 70 mm or less. Even if it is pulled up longer than necessary, only a lot of waste will be generated, and the loss of raw materials and time can be suppressed more effectively.

Next, the calculation of the measured deviation amount of the crystal diameter with respect to the predetermined reference diameter of the crystal will be described. In the present invention, the reference diameter of the crystal is determined in advance. The reference diameter of the crystal can be appropriately set according to the diameter and the maximum weight of the crystal to be manufactured. For example, in the case of manufacturing a crystal having a diameter of 300 mm, the crystal diameter is selected from the range of 3 mm or more and 20 mm or less. Is preferable. The difference between the reference diameter set in this way and the measured crystal diameter (“measured crystal diameter”-“predetermined reference diameter of crystal”) is calculated to obtain the amount of deviation.

Next, the calculation of the correction amount of the heater output in the cone process will be described. It is preferable to select a value for correcting the heater output with respect to the amount of deviation so that the variation in cone process time and weight becomes uniform. In particular, it is more effective to set the correction amount to 0 or more when the deviation amount is 0 or more, and to set the correction amount to less than 0 when the deviation amount is less than 0. As a result of investigation by the present inventor, the correction amount of the heater output is 0.5 kW or more, 5.0 kW or less, preferably 1.0 kW or more, per 1 mm of the deviation amount of the crystal diameter with respect to the predetermined reference diameter. It was found that it is preferable to set it to 0 kW or less. Within such a range, the shortening of time becomes more reliable, and the variation in cone process time and weight can be suppressed more reliably. At this time, the absolute value of the correction amount per 1 mm of the deviation amount when the deviation amount is 0 or more may be less than the absolute value of the correction amount per 1 mm of the deviation amount when the deviation amount is less than 0. It is possible to obtain the effect of further improving productivity by shortening the time and further improving the yield by reducing the weight of the cone.

In the cone process, the heater output is corrected as follows, based on the heater output correction amount calculated from the deviation amount obtained by measuring and calculating the crystal diameter.

The effect of the present invention is exhibited regardless of the range of the deviation amount. For example, when the reference diameter is set to 10 mm, the deviation amount is the reference diameter minus 7 mm or more and the reference diameter plus 10 mm or less. In some cases, it is preferable to correct the heater output. By targeting crystals with a crystal diameter in such a range, even if an abnormal value such as a detection abnormality or a communication data error occurs, it is possible to avoid a trouble that excessive heater output correction is applied. It is possible to stably produce a single crystal.

The timing, number of times, and the like are not particularly limited as long as the correction of the heater output is performed during the cone process. The timing of correcting the heater output may be 60 minutes before the start of the straight body process, but if it is performed within 60 minutes from the start of the cone process, the correction effect can be obtained more reliably. it can. Further, if the correction is performed within 60 minutes from the start of the cone process, the correction can be reliably performed even if the time until the start of the straight body process is unknown.

Further, the heater output may be corrected once during the cone process, but the correction value for each correction may be set low, and the correction may be performed intermittently in a plurality of times or continuously. It is also possible to carry out by gradually changing the correction value. By reducing the change in the output of the heater due to the correction, the surface condition of the melt and the cone shape can be maintained more stably, so that a crystal with good quality can be produced more reliably.

Hereinafter, the present invention will be described in detail with reference to examples, but this does not limit the present invention.

(Example 1)
A 32 inch (800 mm) diameter acupuncture point is filled with 360 kg of raw material and melted to form a raw material melt. A seed crystal with a sharp tip is landed in the raw material melt, and the tip of the seed crystal is submerged and melted by 30 mm. After that, the crystal diameter was measured when the crystal diameter was stabilized by pulling up by 50 mm. With the reference diameter as 11 mm, the amount of deviation between the reference diameter and the measured crystal diameter (hereinafter, may be simply referred to as the "deviation amount") is calculated, and the cone process is started after measuring the crystal diameter according to the amount of deviation. The heater output of 1.4 kW per 1 mm of deviation was corrected within 5 minutes. If the amount of deviation is 0 or more (when the measured crystal diameter is greater than or equal to the reference diameter), increase the heater output, and when the amount of deviation is less than 0 (when the measured crystal diameter is less than the reference diameter), increase the heater output. The correction was made to lower it. In this example, the heater output was continuously corrected within 5 minutes after the start of the cone process. While performing such control, the production of a silicon single crystal having a diameter of 300 mm was repeated many times. The range for performing correction was set from a deviation amount of -8 mm to +9 mm.

(Example 2)
The heater output is corrected in the same manner as in Example 1 except that the correction value of the heater output is set to 0 when the deviation amount is 0 mm or more (when the measured crystal diameter is equal to or larger than the reference diameter). The production of a silicon single crystal having a diameter of 300 mm was repeated many times.

(Comparison example)
The production of a silicon single crystal having a diameter of 300 mm was repeated many times in the same manner as in Example 1 except that the heater output was not corrected. In the comparative example, the crystal diameter was measured and the amount of deviation was calculated for comparison with Examples 1 and 2.

First, with respect to Examples 1 and 2 and Comparative Example, the results of investigating the time required for the cone process and the weight of the cone portion will be described.

FIG. 2 shows the relationship between the measured crystal diameter deviation (horizontal axis, unit mm) and the cone process time (vertical axis) with respect to the reference diameter for Example 1 and Comparative Example. The cone process time is expressed as a relative value when the average value of the cone process times of the comparative example is set to 1 as a reference. The amount of deviation calculated in the produced crystal was in the range of −5.0 mm to +2.0 mm. Further, FIG. 3 shows the comparison result of the cone weight reduction effect, and FIG. 4 shows the comparison result of the cone weight variation.

As shown in FIG. 2, in the comparative example, it was confirmed that the smaller the measured crystal diameter (thin crystal) and the larger the absolute value in the minus (minus) direction, the longer the cone process time tends to be. .. In addition, the cone process time varied in the range of the average value of -14% to the average value of + 12%. It was confirmed that the cone weight has the same relationship, that is, the smaller the measured crystal diameter (thinr crystal), the larger the cone weight.

On the other hand, as is clear from FIG. 2, in Example 1, it can be seen that the cone process time is stable regardless of the amount of deviation of the measured crystal diameter with respect to the reference diameter. The cone process time of Example 1 was shortened by 9% on average as compared with Comparative Example. In particular, when the amount of deviation was in the range of −4.5 mm to −5.0 mm, the cone process time could be shortened by 21%. In addition, the cone process time was in the range of an average value of -2% to an average value of + 3%, and variation could be suppressed. The cone process time of Example 2 was shortened by 10% on average.

As shown in FIG. 3, the weight of the cone was reduced by 8.5% on average in Example 1 and 9.8% on average in Example 2 as compared with the comparative example. Further, as shown in FIG. 4, the variation σ of the cone weight was also reduced to 1.3% in Example 1 and 2.4% in Example 2 compared to 4.8% in Comparative Example.

In particular, in Example 2, since the correction value of the heater output when the deviation amount is 0 mm or more is set to 0, the cone process time shortening effect and the cone weight reduction effect when the deviation amount is 0 mm or more are enhanced. Conceivable.

Next, the results of producing a PW (polished wafer) from the silicon single crystal ingots obtained in Examples 1 and 2 and Comparative Examples and evaluating the defects in the wafer surface will be described. For the evaluation of defects, a laser scattering type defect inspection device SP3 manufactured by KLA-Tencor was used to evaluate defects detected as 26 nm or more in the Oblique mode. FIG. 5 shows the defect evaluation results for PW of the comparative example. In the comparative example, it was confirmed that the defect quality of PW deteriorated as the amount of deviation of the measured crystal diameter with respect to the predetermined reference diameter of the crystal increased to 0, -2, -4 mm. As shown in FIG. 2, in the comparative example, since the cone weight increases as the deviation amount shifts to the minus side, the melt surface and the upper heat of the melt when reaching the straight body product portion during silicon single crystal production. It is considered that the distance from the shield becomes wider, which reflects the effect of changes in the thermal environment during crystal growth. On the other hand, in Examples 1 and 2, since the variation in cone weight is suppressed as described above, such variation in the defect quality of PW does not occur, and a defect-free wafer is manufactured with a good yield. Was made.

As described above, according to the CZ silicon single crystal production method according to the present invention, the cone process time is not affected by the variation in the melt temperature at the time of seeding (landing) of the dislocation-free seeding method (DFS method). It is possible to carry out the cone process without deterioration of productivity due to long time and decrease in yield due to increase in cone weight, and heat in the furnace due to variation in cone process (time, weight) in CZ silicon single crystal production using the DFS method. It is possible to reduce environmental variation, improve the quality of low-defect / defect-free wafers, and improve product yield.

The present invention is not limited to the above embodiment. The above-described embodiment is an example, and any object having substantially the same configuration as the technical idea described in the claims of the present invention and exhibiting the same action and effect is the present invention. Is included in the technical scope of.

1a ... Quartz crucible, 1b ... Graphite crucible, 2 ... Heater, 3 ... Raw material melt,
4 ... Silicon single crystal, 5 ... Pull-up shaft, 6 ... Support shaft, 7 ... Seed crystal,
8a ... Insulation cylinder, 8b ... Insulation plate, 9a ... Main chamber,
9b ... Pull chamber, 10 ... Gas rectifier cylinder, 11 ... Heat shield member,
100 ... Single crystal manufacturing apparatus.

Claims (6)

A seeding step in which the tip of a seed crystal having a pointed tip is brought into contact with a silicon melt and then melted for seeding, and a crystal diameter stabilization in which the seed crystal is pulled up to a predetermined length to stabilize the crystal diameter. A CZ silicon single crystal growth method including a step, a cone step of forming a cone portion, and a straight body step of forming a straight body portion.
The crystal diameter of the crystal pulled up in the crystal diameter stabilization step is measured, and
The amount of deviation of the measured crystal diameter with respect to a predetermined crystal reference diameter is calculated.
The correction amount of the heater output in the cone process is calculated according to the deviation amount.
A method for producing a CZ silicon single crystal, which comprises correcting a heater output value using the correction amount in the cone step to form a cone. When calculating the correction amount of the heater output,
When the deviation amount is 0 or more, the correction amount is set to 0 or more.
The CZ silicon single crystal production method according to claim 1, wherein when the deviation amount is less than 0, the correction amount is set to less than 0. When calculating the correction amount of the heater output,
The method for producing a CZ silicon single crystal according to claim 1 or 2, wherein the correction amount is 0.5 kW or more and 5.0 kW or less per 1 mm of the deviation amount. When calculating the correction amount of the heater output,
The absolute value of the correction amount per 1 mm of the deviation amount when the deviation amount is 0 or more shall be less than the absolute value of the correction amount per 1 mm of the deviation amount when the deviation amount is less than 0. The method for producing a CZ silicon single crystal according to any one of claims 1 to 3, wherein the CZ silicon single crystal is produced. The CZ silicon single according to any one of claims 1 to 4, wherein the measurement of the crystal diameter is performed when the length of the crystal to be pulled up in the crystal diameter stabilizing step is 10 mm or more and 100 mm or less. Crystal production method. The claim is characterized in that the correction of the heater output value is performed within 60 minutes from the start of the cone process or 60 minutes before the start of the straight body process, whichever is longer. Item 5. The method for producing a CZ silicon single crystal according to any one of Items 1 to 5.