JP6304125B2 - A method for controlling resistivity in the axial direction of silicon single crystal - Google Patents

Description

  The present invention relates to an axial resistivity control method of a silicon single crystal grown by a Czochralski method (CZ method).

  An N-type crystal is mainly used in a switching device used for power such as an insulated gate field effect transistor (Insulated Gate Bipolar Transistor: IGBT). Conventionally, an epitaxial wafer (EPW) whose resistivity control is relatively easy, and a wafer (FZ-PW) manufactured by a floating zone method (FZ method) have been used.

  However, EPW is an expensive wafer because it includes an extra step (epitaxial growth step) compared to a wafer manufactured by CZ method (CZ-PW), and it is not easy to increase the crystal diameter in FZ method. CZ-PW, which is used as it is without depositing an epitaxial layer on CZ-PW, has been attracting attention. However, in a silicon single crystal manufactured by the CZ method, there is a segregation phenomenon of the dopant, and it is difficult to make the resistivity distribution in the axial direction (the pulling-up axis direction) uniform.

  As a method for solving this problem, Patent Documents 1 and 2 disclose a so-called counter-doping method in which a subdopant having a polarity opposite to that of the main dopant and having a small segregation coefficient is added. In addition, as an application of the technology (continuous charge method (Continuous Czochralski Method: CCZ method)) in which raw material silicon is added during pulling of a single crystal to control the dopant concentration in the melt, a rod-like silicon crystal containing no dopant is immersed. There is also a technique for pulling a crystal while making it happen.

  By using these methods, it is possible to improve the resistivity distribution in the axial direction of the CZ single crystal. However, the most frequently used dopant in the production of an N-type single crystal is P (phosphorus), and its segregation coefficient is about 0.35. On the other hand, when trying to improve the resistivity distribution by the counter-doping method, elements having an opposite polarity and a segregation coefficient smaller than the segregation coefficient of P include Ga, In, and Al.

  However, for example, there are reports that the electrical characteristics of the oxide film deteriorate when a heavy metal is contained in the oxide film, and it is not clear how the device characteristics are affected by the addition of these elements. For this reason, B (boron) is the mainstream as a P-type dopant, and since it is an element widely used in manufacturing devices, it is preferable to use B as an element of opposite polarity if possible. However, since the segregation coefficient of B is about 0.78, which is larger than P, the above technique cannot be used.

  The reason why the above technique cannot be applied when the segregation coefficient of the sub-dopant is larger is that the change in the concentration in the axial direction of the sub-dopant is small relative to the main dopant, and the second half of the straight body where the concentration of the main dopant is high is sufficient This is because it cannot be countered. Further, since the concentration of the sub-dopant is relatively high even in the first half of the straight body portion where the concentration of the main dopant is low, the resistivity distribution seen in the whole straight body portion is rather deteriorated as compared with the case where the counterdoping is not performed.

  In order to perform counter-doping using an element having a large segregation coefficient as the sub-dopant, the sub-dopant may be added continuously or intermittently according to the growth of the single crystal. In this way, the concentration of the sub-dopant can be made sufficiently high in the latter half of the straight body portion where the concentration of the main dopant becomes high. This technique is disclosed in, for example, Patent Document 3, and shows a calculation method of the amount of dopant added suitable for obtaining a uniform resistivity distribution.

  As a method for adding a sub-dopant to a silicon melt, a method of adding silicon crystals containing a dopant element to the melt is general. However, when a granular silicon crystal is added, the charged granular silicon crystal may reach and come into contact with the silicon single crystal grown before melting into the melt, which may cause dislocation. On the other hand, the method of inserting rod-shaped silicon crystals into the raw material melt is superior because solid silicon that has not melted does not touch the silicon single crystal that is being grown.

JP 2002-128591 A Japanese Patent Laid-Open No. 2004-307305 JP-A-3-247585 JP 2007-290906 A JP 2008-195545 A

  Thus, when adding a subdopant by immersing a rod-shaped silicon crystal in a raw material melt, in order to obtain a crystal having a desired resistivity distribution, it is important to control the amount of dissolution of the rod-shaped silicon crystal. In particular, unlike the method of dropping granular silicon crystals into the melt, the amount of dissolution of the rod-shaped silicon crystals due to the change in the position of the molten metal surface becomes a problem.

  There are cases where the position of the molten metal surface during pulling of the single crystal is intentionally changed in order to obtain a crystal of a desired quality, and there are cases where the surface is unintentionally generated due to factors such as crucible deformation and diameter fluctuation. Among these, the latter can be solved by detecting the position of the hot water surface during operation by a method as shown in Patent Documents 4 and 5, for example, and controlling it constantly.

  However, in the case of an operation in which the molten metal surface position is changed intentionally, there is a problem that the amount of dissolution of the rod-like silicon crystal changes with the fluctuation pattern of the molten metal surface position because the control is not to keep the molten metal surface position constant. It was. Specifically, when the hot water surface position is raised during operation, the rod-shaped silicon crystals are dissolved excessively than expected, and when the hot water surface position is lowered, the amount of rod-shaped silicon crystals dissolved is less than expected. . Thus, in the case of operation that changes the melt surface position of the raw material melt in the crucible, the axial resistivity of the silicon single crystal cannot be controlled as desired (for example, uniformly). It was.

  The present invention has been made in view of the above-described problems, and controls the resistivity in the axial direction of a silicon single crystal even in an operation that changes the melt surface position of the raw material melt in the crucible. It is an object of the present invention to provide a resistivity control method capable of

In order to achieve the above object, according to the present invention, when the silicon single crystal is pulled and grown from the raw material melt by the Czochralski method while changing the surface position of the raw material melt in the crucible, A method of controlling the axial resistivity of the silicon single crystal grown by inserting a rod-like silicon crystal into the raw material melt,
Calculating a pattern for inserting the rod-like silicon crystal into the raw material melt according to the pulling length of the silicon single crystal;
When the straight body portion of the silicon single crystal is grown, the raw material melt is formed into a pattern in which the rod-like silicon crystal is inserted into the raw material melt according to a pattern that changes the position of the molten metal surface of the raw material melt. Adding a correction for changing the position of the hot water surface,
A step of inserting the rod-shaped silicon crystal during the growth of the silicon single crystal using a pattern for inserting the rod-shaped silicon crystal to which the correction has been made into the raw material melt. An axial resistivity control method is provided.

  In this way, even in the operation of changing the position of the melt surface of the raw material melt in the crucible during the growth of the silicon single crystal, the axial resistivity of the silicon single crystal can be set as desired (for example, , Uniformly).

  At this time, it is preferable to use a rod-like silicon crystal containing a sub-dopant having a conductivity type opposite to that of the main dopant having the conductivity type of the silicon single crystal.

  In this way, the change in the axial resistivity or dopant concentration of the silicon single crystal caused by the segregation of the main dopant can be offset by the sub-dopant having the conductivity type opposite to that of the main dopant. Therefore, even if the operation is to change the melt surface position of the raw material melt in the crucible, the axial resistivity of the silicon single crystal can be more reliably controlled as desired.

  At this time, the subdopant can have a segregation coefficient larger than that of the main dopant.

  Thus, even when the segregation coefficients of the main dopant and the sub-dopant are in the above relationship, the axial resistivity of the silicon single crystal can be controlled as desired.

  With the method of controlling the axial resistivity of the silicon single crystal of the present invention, even if the operation is such that the position of the melt surface of the raw material melt in the crucible is changed during the growth of the silicon single crystal, The axial resistivity can be controlled as desired. In particular, it can be controlled to be uniform.

It is process drawing which showed an example of the resistivity control method of the axial direction of the silicon single crystal of this invention. It is the schematic which showed an example of the single crystal growth apparatus which can be used with the resistivity control method of the axial direction of the silicon single crystal of this invention. In an Example and a comparative example, when growing the straight body part of a silicon single crystal, it is the graph which showed the pattern which changes the molten metal surface position of a raw material melt. It is the graph which showed the insertion pattern to the raw material melt of the rod-shaped silicon crystal in an Example and a comparative example. It is the graph which showed the resistivity distribution of the axial direction of the silicon single crystal grown in the Example. It is the graph which showed the resistivity distribution of the axial direction of the silicon single crystal grown in the comparative example.

  Hereinafter, although an embodiment is described about the present invention, the present invention is not limited to this.

  As described above, in the case of operation that changes the melt surface position of the raw material melt in the crucible during the growth of the silicon single crystal, the amount of dissolution of the rod-shaped silicon crystal changes with the variation pattern of the melt surface position. There was a problem.

  Therefore, the present inventors have intensively studied to solve such problems. As a result, when growing the straight body portion of the silicon single crystal, the molten metal surface of the raw material melt is changed into a pattern in which the rod-like silicon crystal is inserted into the raw material melt according to the pattern that changes the molten metal surface position of the raw material melt. It has been found that a correction for changing the position is added. As a result, it has been found that the axial resistivity of the silicon single crystal can be controlled as desired even in an operation that changes the melt surface position of the raw material melt in the crucible. And the best form for implementing these was scrutinized and the present invention was completed.

  First, a single crystal growth apparatus that can be used in the axial resistivity control method for a silicon single crystal of the present invention will be described with reference to FIG.

  As shown in FIG. 2, the single crystal growing apparatus 17 has a member for accommodating and melting the raw material polycrystalline silicon, a heat insulating member for shutting off heat, and the like. Is housed inside. A pulling chamber 2 extending upward from the ceiling portion (top chamber 11) of the main chamber 1 is connected, and a mechanism (not shown) for pulling up the silicon single crystal 3 with a wire is provided on the upper portion.

  A quartz crucible 5 for containing the melted raw material melt 4 and a graphite crucible 6 for supporting the quartz crucible 5 are provided in the main chamber 1, and these crucibles 5 and 6 are rotated by a drive mechanism (not shown). It is supported by a crucible shaft so that it can move up and down. And the heater 7 for melting a raw material is arrange | positioned so that the crucibles 5 and 6 may be surrounded. A heat insulating member 8 is provided outside the heater 7 so as to surround the periphery thereof.

  A gas inlet 10 is provided at the upper part of the pulling chamber 2, and an inert gas such as argon gas is introduced and discharged from the gas outlet 9 at the lower part of the main chamber 1. Further, a heat shield member 13 is provided so as to face the raw material melt 4, so that radiation from the surface of the raw material melt 4 is cut and the surface of the raw material melt 4 is kept warm.

  Further, a gas purge cylinder 12 is provided above the raw material melt 4 so that the periphery of the single crystal rod 3 can be purged by the inert gas introduced from the gas inlet 10.

  The top chamber 11 is provided with a rod-shaped silicon crystal insertion machine 14. The rod-like silicon crystal insertion machine 14 is configured to insert the rod-like silicon crystal 15 into the raw material melt 4. Here, the rod-shaped silicon crystal insertion machine 14 can be a movable device that can move in the vertical direction as shown in FIG. 2, for example. Before the chamber is sealed (set), the rod-shaped silicon crystal 15 is attached to the tip of the rod-shaped silicon crystal insertion machine 14, and the rod-shaped silicon crystal 15 is removed from the raw material melt 4 during the growth of the silicon single crystal 3. The rod-shaped silicon crystal 15 can be immersed by extending to the surface.

  At this time, it is desirable that the rod-like silicon crystal 15 can be immersed in a portion close to the quartz crucible 5. In such a case, it is possible to prevent the rod-shaped silicon crystal 15 from being inserted in the immediate vicinity of the growing silicon single crystal 3. For example, the rod-shaped silicon crystal 15 includes a sub-dopant. In some cases, the melt containing a high concentration of dopant reaches the silicon single crystal 3 before being stirred and the in-plane and axial resistivity distribution becomes unintentionally non-uniform, or the melt near the silicon single crystal is melted. Since liquid temperature is low temperature, it can suppress that solidification generate | occur | produces in an immersion part and leads to the dislocation | rearrangement of the silicon single crystal 3.

  Further, the magnetic field applying device 16 may be further installed outside the main chamber 1 in the horizontal direction. Thus, a single crystal growth apparatus based on the so-called MCZ method that applies a magnetic field in the horizontal direction or the vertical direction to the raw material melt 4 to suppress the convection of the raw material melt 4 and achieve stable growth of the silicon single crystal 3 is obtained. You can also.

  Next, an axial resistivity control method of the silicon single crystal of the present invention will be described with reference to FIGS. Here, an example in which a uniform resistivity distribution is obtained in the axial direction will be described. However, the present invention is not limited to this example, and can be appropriately adjusted so as to obtain a desired resistivity distribution.

(Insertion pattern calculation process: SP1 in FIG. 1)
First, a pattern for inserting the rod-like silicon crystal 15 into the raw material melt 4 is calculated according to the pulling length of the silicon single crystal 3. Thus, first, the insertion pattern of the rod-like silicon crystal 15 corresponding to the pulled length of the silicon single crystal 3 is obtained by calculation without taking into account the change in the molten metal surface position of the raw material melt 4.

  At this time, as an insertion pattern of the rod-like silicon crystal 15 that can make the resistivity in the axial direction of the silicon single crystal 3 uniform, for example, it can be obtained by calculation based on a calculation formula disclosed in Patent Document 3, for example.

(Insertion pattern correction step: SP2 in FIG. 1)
Next, when the straight body portion of the silicon single crystal 3 is grown, the rod-like silicon crystal 15 obtained in the above SP1 is inserted into the raw material melt 4 according to the pattern for changing the position of the molten metal surface of the raw material melt 4. A correction corresponding to the change of the molten metal surface position of the raw material melt 4 is added to the pattern to be performed.

  The amount of change that changes the position of the melt surface of the raw material melt 4 when the straight body portion of the silicon single crystal 3 is grown is relative to the position of the melt surface when the insertion of the rod-shaped silicon crystal 15 into the material melt 4 is started. It is desirable that the range be within ± 10 mm. The change of the molten metal surface position during the growth of the straight body part is mainly used for flattening the crystal defect distribution over the entire straight body part. Although it depends on the HZ (hot zone) to be used and the desired defect region, it is usually preferable to change the molten metal surface position within a range of ± 10 mm or less because the influence on various qualities of the crystal can be suppressed.

  For example, it is assumed that a pattern in which the molten metal surface position of the raw material melt 4 is lowered by 2 mm while the straight body portion of the silicon single crystal 3 is grown 10 cm. In this case, the time corresponding to growing the straight body portion of the silicon single crystal 3 by 10 cm is calculated from the pulling set speed. Then, the insertion speed of the rod-shaped silicon crystal 15 by the rod-shaped silicon crystal insertion machine 14 may be corrected so that the rod-shaped silicon crystal 15 advances by an extra 2 mm during that time.

  Assuming that the pulling rate of the silicon single crystal in the above section is 0.5 mm / min, it takes 200 minutes to pull the silicon single crystal 3 by 10 cm. During this time, the rod-shaped silicon single crystal 15 only needs to be advanced by 2 mm, so the insertion speed of the rod-shaped silicon crystal 15 by the rod-shaped silicon crystal insertion machine 14 should be corrected by +0.01 mm / min = 0.6 mm / h. It will be.

  In this way, the amount of correction to be added to the pattern for inserting the rod-like silicon crystal 15 into the raw material melt 4 according to the fluctuation pattern of the molten metal surface position of the raw material melt 4 is obtained, and a new insertion pattern is obtained. .

(Stick silicon crystal insertion step: SP3 in FIG. 1)
The rod-shaped silicon crystal 15 is inserted during the growth of the silicon single crystal 3 by using the pattern in which the rod-shaped silicon crystal 15 corrected as described above is inserted into the raw material melt 4.

  In this way, the axial resistivity of the silicon single crystal can be uniformly controlled even in an operation that changes the melt surface position of the raw material melt in the crucible.

  Depending on the conditions, the insertion speed of the rod-shaped silicon crystal 15 by the rod-shaped silicon crystal inserter 14 is very slow, around 1 mm / h. For this reason, the method of continuously inserting the rod-shaped silicon crystal 15 may be difficult due to the accuracy of the rod-shaped silicon crystal insertion machine 14. In such a case, it is preferable that the rod-shaped silicon crystal 15 is inserted intermittently so that the rod-shaped silicon crystal 15 is immersed only once every 10 minutes, for example, so as to follow the fluctuation of the molten metal surface position.

  The present invention is particularly effective when growing an N-type single crystal used for power devices such as IGBTs, specifically, a relatively high resistance product having a resistivity of about 30 Ωcm to 3000 Ωcm.

  In the case of a silicon single crystal having a relatively high resistivity in such a range, the volume of the sub-dopant necessary for making the resistivity distribution uniform is small. In such a case, since the influence on the resistivity due to the change in the melting amount of the sub-dopant caused by the change in the molten metal surface position is large, the present invention considering the correction for changing the molten metal surface position is particularly preferably applied. Can do.

  At this time, it is preferable to use a rod-like silicon crystal 15 containing a sub-dopant having a conductivity type opposite to that of the main dopant of the silicon single crystal 3.

  In this way, the change in the axial resistivity or dopant concentration of the silicon single crystal 3 caused by segregation of the main dopant can be offset by the sub-dopant having the conductivity type opposite to that of the main dopant. The axial resistivity of the crystal 3 can be controlled as desired. In particular, the resistivity distribution in the axial direction of the silicon single crystal 3 can be controlled uniformly.

  At this time, it is preferable that the subdopant has a segregation coefficient larger than that of the main dopant.

  Thus, even if it is a case where the segregation coefficient of a main dopant and a subdopant has a relationship as mentioned above, this invention can be applied suitably. Even in the latter half of the straight body portion, the main dopant can be sufficiently canceled as necessary.

  Specifically, for example, P (phosphorus) can be used as the main dopant, and B (boron) can be used as the sub-dopant. If it is an N-type silicon single crystal in which the main dopant is P and the sub-dopant is B, both elements are widely used for device manufacturing, and there is little fear of unexpected effects.

Further, the B concentration in the rod-like silicon crystal 15 used for the addition of the subdopant is preferably about 6 × 10 ¹⁹ atoms / cm ³ or less although it depends on the target resistivity of the silicon single crystal 3.

  In the case of the rod-shaped silicon crystal 15 having such a B concentration, even if the amount of dissolution of the rod-shaped silicon crystal 15 changes due to, for example, an error in the position accuracy of the crucibles 5 and 6, the effect is suppressed to some extent. Therefore, the targeted resistivity can be obtained.

  The upper limit of the concentration of the sub-dopant in the rod-shaped silicon crystal 15 is determined as appropriate depending on many parameters such as the target resistivity of the silicon single crystal 3, the volume of the raw material melt 4, and the size of the rod-shaped silicon crystal 15. It is preferable to change.

For example, specifically, consider the B concentration of the rod-shaped silicon crystal 15 when the target resistivity of the silicon single crystal 3 is N-type 100 Ωcm and the positional accuracy error of the crucibles 5 and 6 is 0.1 mm. Since the P concentration in the N-type 100 Ωcm silicon single crystal is about 4.2 × 10 ¹³ atoms / cm ^{3 in the} case where no counter-doping is performed, the concentration error of B has a relatively large effect on the order of 1 × 10 ^12. It can be said from about atoms / cm ³ . When the contact area of the rod-shaped silicon crystal 15 with the molten metal surface is 25 mm ² and the weight of the raw material melt 4 is 380 kg, the B concentration in the raw material melt 4 is 1 × 10 ¹² atoms / min with an error of the crucible position 0.1 mm. The change of cm ³ or more occurs when the B concentration of the rod-like silicon crystal 15 is about 6 × 10 ¹⁹ atoms / cm ³ or more. Therefore, it is preferable to set the concentration below this level.

Of course, the above is only an example because it changes if the target resistivity changes.
On the other hand, if the B concentration of the rod-shaped silicon crystal 15 is too low, the volume of the rod-shaped silicon crystal 15 necessary for keeping the resistivity constant increases, which causes problems due to difficulty in handling. However, the lower limit of the concentration of the rod-shaped silicon crystal 15 should be determined by the pulling weight of the silicon single crystal 3 in addition to the target resistivity of the silicon single crystal 3, the volume of the raw material melt 4, and the size of the rod-shaped silicon crystal 15. Because it changes, it is difficult to limit specifically. It can be determined appropriately according to each condition.

  As a guideline, for example, if the lower limit of the concentration is set such that the length of the rod-shaped silicon crystal 15 required through the pulling of the silicon single crystal 3 is 30 cm or less, preparation and handling of the rod-shaped silicon crystal 15 can be facilitated. This is preferable because it is possible.

  EXAMPLES Hereinafter, the present invention will be described more specifically with reference to examples and comparative examples of the present invention, but the present invention is not limited to these.

(Example)

  The single crystal growth apparatus 17 as shown in FIG. 2 was used to pull up the N-type phosphorus-doped silicon single crystal 3 by the MCZ method.

  As conditions for growing the silicon single crystal 3, the charge amount was 380 kg, the diameter of the silicon single crystal 3 to be pulled up was 306 mm, and the target resistivity after counter-doping was 65 Ωcm (resistivity standard 65 ± 10%).

The rod-like silicon crystal 15 for counter dope has a immersible portion having a size of 5 mm × 5 mm × 100 mm and a B concentration of 5 × 10 ¹⁸ atoms / cm ³ . Further, in order to control the defect distribution, the molten metal surface position is changed as shown in FIG. 3 by correcting the positions of the crucibles 5 and 6 during the growth of the straight body portion. A positive value indicates correction for lowering the hot water surface position relative to the current hot water surface position, and a negative value indicates correction for increasing the hot water surface position.

  First, a polycrystalline silicon and phosphorus dopant was introduced into the quartz crucible 5 and the raw material was heated in an argon atmosphere while rotating the crucibles 5 and 6 to form a silicon melt (raw material melt 4).

  Next, a horizontal magnetic field having a central magnetic field strength of 4000 G was applied to the raw material melt 4 from the magnetic field application device 16. When the temperature of the raw material melt 4 was stabilized, the seed crystal was immersed in the raw material melt 4, and the silicon single crystal 3 was started to grow by gradually pulling up while rotating the seed crystal. On the way, when the pulling length ratio of the straight body part (the pulling length of the straight body part / the final pulling length of the straight body part) is 26%, the rod-like silicon crystal 15 is brought into contact with the melt surface, and the counter dope is applied. Started.

  The insertion pattern of the rod-like silicon crystal 15 at the time of counterdoping was obtained in advance. First, it calculated | required by calculation from the target resistivity and the pulling speed of the silicon single crystal 3 in the insertion pattern calculation process. As described above, this insertion pattern calculation step does not take into account the change in the melt surface position of the raw material melt 4.

  Then, in the insertion pattern correction step, according to the pattern in which the molten metal surface position of the raw material melt 4 is changed, the rod-like silicon crystal 15 obtained in the above-mentioned insertion pattern calculation step is inserted into the raw material melt 4 in a pattern. Correction for changing the position of the hot water surface of the liquid 4 was added, and the example of the graph shown in FIG. 4 was made. As a result, in the example, the corrected insertion pattern obtained in the insertion pattern correction step was corrected for changes in the molten metal surface position by a total of 5.5 mm.

  After the operation, the amount of dissolution of the rod-like silicon crystal 15 was measured and found to be 53.2 mm. A test sample is cut out from each cut end face of the pulled silicon single crystal 3, and the result of measuring the resistivity is shown in FIG. As shown in FIG. 5, it was possible to obtain a resistivity distribution according to the target resistivity over the entire straight body.

(Comparative example)
According to the pulling length of the silicon single crystal, as shown in the comparative example of FIG. 4, except that the rod-shaped silicon crystal was inserted into the raw material melt in accordance with the pattern of inserting the rod-shaped silicon crystal into the raw material melt. A silicon single crystal was manufactured under the same conditions as those described above. Note that the pattern of this comparative example is the same as the pattern before correction in the embodiment, that is, the pattern obtained in the insertion pattern calculation step, and does not take into consideration the influence due to the change in the molten metal surface position.

  It was 58.7 mm when the melt | dissolution amount of the rod-shaped silicon crystal was measured after the operation.

  Further, the resistivity distribution in the axial direction of the silicon single crystal grown in the comparative example was measured and shown in FIG. As a result, as shown in FIG. 6, it can be seen that the resistivity of the silicon single crystal grown in the comparative example increases in the second half. And in the end part, it has deviated from the standard of resistivity.

  The present invention is not limited to the above embodiment. The above-described embodiment is an exemplification, and the present invention has any configuration that has substantially the same configuration as the technical idea described in the claims of the present invention and that exhibits the same effects. Are included in the technical scope.

1 ... main chamber, 2 ... pulling chamber, 3 ... silicon single crystal,
4 ... Raw material melt, 5 ... Quartz crucible, 6 ... Graphite crucible, 7 ... Heater,
8 ... Insulating member, 9 ... Gas outlet, 10 ... Gas inlet, 11 ... Top chamber,
12 ... Gas purge cylinder, 13 ... Heat shield member, 14 ... Rod-shaped silicon crystal insertion machine,
DESCRIPTION OF SYMBOLS 15 ... Rod-shaped silicon crystal, 16 ... Magnetic field application apparatus, 17 ... Single crystal growth apparatus.

Claims (3)

While growing the silicon single crystal from the raw material melt by the Czochralski method while changing the molten metal surface position of the raw material melt in the crucible, by inserting a rod-like silicon crystal into the raw material melt, A method of controlling the axial resistivity of the grown silicon single crystal,
Calculating a pattern for inserting the rod-like silicon crystal into the raw material melt according to the pulling length of the silicon single crystal;
When the straight body portion of the silicon single crystal is grown, the raw material melt is formed into a pattern in which the rod-like silicon crystal is inserted into the raw material melt according to a pattern that changes the position of the molten metal surface of the raw material melt. Adding a correction for changing the position of the hot water surface,
A step of inserting the rod-shaped silicon crystal during the growth of the silicon single crystal using a pattern for inserting the rod-shaped silicon crystal to which the correction has been made into the raw material melt. Resistivity control method in the axial direction.   2. The silicon single crystal axial resistance according to claim 1, wherein the rod-like silicon crystal includes a sub-dopant having a conductivity type opposite to a main dopant having a conductivity type of the silicon single crystal. Rate control method.   The method for controlling the axial resistivity of a silicon single crystal according to claim 2, wherein the subdopant has a segregation coefficient larger than that of the main dopant.

Images