JP2623390B2 - Silicon single crystal rod growth method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a manufacturing method for growing a high-purity silicon single crystal rod by the FZ method (float zone method, floating zone melting method) equipped with a high-frequency coil.
In particular, the present invention relates to a method for growing the silicon single crystal rod having a uniform electric resistivity in a cross section in the diameter direction of the single crystal rod.

[0002]

2. Description of the Related Art FIG. 8 is a diagram showing the overall configuration of a conventional silicon single crystal growing apparatus by the FZ method according to the present invention,
A polycrystalline raw material rod 1 having a predetermined diameter is held at 0, and a seed crystal 7 is held at the lower shaft 8. The raw material rod 1 and the seed crystal 7 are fused together while being melted by the high-frequency coil 2. While rotating
In this method, the silicon rod is lowered at a very low speed, the melting zone 4 is moved to the upper end of the raw material, and the single crystal rod 3 is obtained. on the other hand,
In the CZ method (Czochralski method, pulling method), a tip of a seed crystal having a desired crystal orientation is attached to a large-capacity silicon melt, pulled up while rotating a pulling shaft on which the seed crystal is mounted, and pulled to a desired diameter. This is a method for obtaining a silicon single crystal rod.

The non-uniform distribution of the electric resistivity is divided into two distributions: a growth axis direction and a cross-sectional distribution. The two growth methods will be compared with respect to the non-uniform distribution in the growth axis direction. In the CZ method, segregation of a dopant substance occurs when the silicon melt is solidified into a solid silicon single crystal, the dopant concentration of the silicon melt gradually increases, and the electrical resistivity decreases as the silicon melt grows. The non-uniform distribution of the method is large. On the other hand, in the FZ method, since the silicon melt is constantly supplied from above for a small melt volume, the dopant concentration in the direction of the growth axis becomes macroscopically more uniform than in the CZ method.

However, in the FZ method, since the volume of the melt is small, the dopant is microscopically irregularly incorporated due to fluctuations in the convection in the melt, and the distribution in the cross section becomes large.
For example, as shown in FIG. 2A, a silicon single crystal rod (rotation speed, 6 rotations per minute) having a diameter of 100 mm and a growth orientation of <111> is formed on a silicon wafer having a thickness of 300 μm. When the electrical resistivity in the direction is measured and the graph arranged and plotted in the electrical resistivity change rate A is seen, the variation in the change rate A is large.

However, when the maximum value of the measured electrical resistivity R is Rmax, the minimum value is Rmin, and the average of the electrical resistivity R in the wafer surface is Rave, the electrical resistivity change rate A is A = [( R-Rave) / Rave] × 100 (%) Further, the in-section variation rate a of the electric resistivity is defined as a = [(Rmax−Rmin) / Rmin] × 100 (%). Here, treating the value of the electrical resistivity change rate A without simply plotting the electrical resistivity R means that as the electrical resistivity R increases, the apparent change rate of the electrical resistivity increases. To avoid things.
Further, the variation of the electrical resistivity R is represented as a single numerical value according to the in-section variation rate a, whereby the electrical resistivity distribution can be compared and evaluated with each other.

FIG. 2 (a) is a distribution diagram of the electrical resistivity change rate A in the cross section of the silicon single crystal rod. It can be seen that the electrical resistivity decreases near the center of the wafer and is non-uniform.
The value of the in-section variation rate a is 22.1%. In the manufacture of individual semiconductors, the value of the in-section variation rate a is required to be as small as possible, and in severe devices, it may be required to be 3% or less. In such a case, a silicon single crystal rod is grown by the FZ method without doping with impurities, and then the single crystal rod is inserted into a nuclear reactor and irradiated with neutrons to convert ³⁰ Si into ³¹ P by a nuclear reaction. A method of doping with a doped dopant is known. However, this neutron irradiation doping method requires a nuclear reactor, and has a disadvantage that the cost of silicon wafer production is greatly increased. Therefore, without neutron irradiation, a silicon unit having a low value of the in-section variation rate a is industrially required. There is a need for a method of growing crystal rods.

[0007] In comparison, when comparing the capacity of the silicon melt between the FZ method and the CZ method, the former is about 1/100 to 1/1000 of the latter, and the FZ method is different from the CZ method in that the silicon melt is similar to the CZ method. It is said that it is difficult to artificially control the convection state in the inside. Therefore, it has been considered that the non-uniformity of the concentration distribution of the dopant in the cross section in the diameter direction in the silicon single crystal rod by the FZ method and the non-uniform distribution of the electric resistivity cannot be eliminated.

Here, considering the flow of the silicon melt in the melting zone of the FZ method, forced convection due to rotation of the seed crystal, natural convection caused by being heated by the high frequency coil, and the volume of the melt There is surface tension convection due to the surface tension induced by the melt free surface, which has a much greater ratio to the surface tension. Here, as a method of reducing the speeds of natural convection and surface tension convection, it is conceivable to cause the forced convection so as to be opposed to these flows. However, in the FZ method, the forced convection is weak because the melt volume is small. The effect of canceling out is hardly obtained. It is also conceivable to increase the forced convection by rotating the growing silicon single crystal rod at a higher speed.However, the weight of the crystal rod is supported by the first narrow portion formed at the lower end of the single crystal rod. Therefore, the growing single crystal rod cannot withstand such high-speed rotation and collapses, making this means impractical.

As means for solving such a problem, FZ
The method of applying a magnetic field to the silicon melting zone in the growth direction in parallel with the silicon melting zone is described in De Leon, etc. (N. De Leo
n, J. et al. Guldberg and J.M. Sall
ing: J. Cryst. Growth 55
(1981) 406-408), a magnetic field generating means was arranged at a position substantially coincident with the boundary surface of the melting zone, and a magnetic field was applied in a direction substantially parallel to the growth axis direction of 180 gauss or less, and a diameter of 42 mm was applied. Grow a silicon single crystal rod,
It is reported that the electric resistivity fluctuation in the cross section was reduced.

[0010]

However, the present F
The main demand for industrial wafers by the Z method is a wafer having a diameter of 75 mm or more, and a method for producing a wafer having a diameter of less than 50 mm, such as De Leon, does not satisfy the current requirements. That is, in the FZ method, the diameter of a single crystal grown by applying a magnetic field of 180 gauss or less to a melt having a grown single crystal diameter of more than 75 mm in a direction substantially parallel to the growth direction of the crystal. The change in the electric resistivity in the cross section in the direction is such that the electric resistivity in the central portion thereof is extremely low, so that the variation a in the electric resistivity cross section exceeds 20%.

In view of the drawbacks of the prior art, the present invention provides a method for growing a silicon single crystal rod having a diameter of 75 mm or more by the FZ method without using a doping step by thermal neutron irradiation. An object of the present invention is to microscopically uniform the dopant distribution in a cross section in the diameter direction of a silicon single crystal rod.

[0012]

According to the present invention, there is provided a method of manufacturing a silicon single crystal rod having a diameter of 75 mm or more by the FZ method, wherein an inner diameter of the coil is at least smaller than a diameter of a grown single crystal as a high frequency induction heating coil. Using a single-turn coil, the axial length of the melting zone is set to be smaller than the diameter of the grown single crystal, and the unmelted state at an upper position and / or a lower position in the axial direction outside the melting zone. And a magnetic field forming means surrounding the raw material polycrystalline rod or the grown single crystal rod, and applying a magnetic field to the melting zone of the silicon single crystal rod by the magnetic field forming means. In this case, the magnetic field forming means is, for example, a solenoid coil surrounding the outer periphery of the non-molten silicon single crystal rod or the raw material polycrystal rod, and is configured by supplying a direct current to the solenoid coil. It is preferable that the ripple rate of the direct current is suppressed to 8% or less and the number of rotations of the lower axis (the growing single crystal side) is set to be larger than the number of rotations of the upper axis (the growing single crystal side). 75-130 in diameter
mm, the magnetic field intensity of the magnetic field forming means is set between 150 gauss and 600 gauss and the single crystal rod is rotated while applying a magnetic field to the silicon single crystal rod. It is preferable to set the number of rotations from 1 to 8 per minute.

However, when the diameter of the silicon single crystal rod exceeds 130 mm, the magnetic field intensity of the magnetic field forming means is preferably set between 180 gauss and 200 gauss. Preferably, the speed is between 0.5 and 4 revolutions per minute.

[0014]

By setting the manufacturing method as described above, the FZ growth method of a silicon single crystal rod having a large diameter, for example, 75 mm or more, does not significantly increase the rotation speed of the lower shaft. Non-uniform distribution of the dopant can be eliminated. It has been confirmed by experiments by the inventors that increasing the rotation speed of the lower shaft causes forced convection in the opposite direction that hinders surface tension convection and natural convection. Has no effect. In addition, the rotation of the lower shaft causes the forced convection described above, causing forced agitation of the melt.However, the rotation center at the center of rotation is zero, and there is no mixing effect of the dopant due to the stirring. In addition, the center portion has a low electric resistivity due to the flat facet growth at the growth interface.

However, it surrounds the outer periphery of the raw polycrystalline rod or the grown single crystal rod in a non-molten state at a position above or below the growth axis outside the melting zone, specifically, at a position slightly distant from the melting zone. When a direct current is supplied to the solenoid coil to form a direct current magnetic field in the growth axis direction including the melting zone, the shape of the melting zone in the FZ method is affected by its own weight, so that the shape of the melting zone is outside the vertical cross section. The shape is inclined with respect to the growth axis direction, and intersects with the magnetic field lines of the solenoid coil. In particular, in the present invention, in the case of a solenoid coil disposed around the outer periphery of the raw material polycrystalline rod in the growth axis direction deviating from the melting zone, a DC magnetic field bent outward in the vicinity of the lower melting zone, and from the melting zone. In the case of a solenoid coil placed around the outer periphery of a grown single crystal rod below the deviated growth axis, a DC magnetic field that bends outward in the vicinity of the upper melting zone just causes surface tension convection and natural convection on the surface of the melting zone. It is considered that the magnetic characteristic effect is exerted so as to form a line of magnetic force perpendicular to the surface and suppress such surface tension convection and natural convection. In the prior art reported by De Leon et al., Such an effect does not occur because the magnetic field is arranged so as to surround the boundary of the melting zone.

Further, when a magnetic field is applied by supplying a direct current to the solenoid coil in the FZ method, if the direct current contains a ripple, the ripple generates an induced eddy current in the silicon melting zone, It is considered that a non-uniform distribution of the temperature distribution and the flow velocity in the cross section of the melting zone is induced, which leads to deterioration of the dopant concentration distribution in the cross section. Accordingly, the ripple rate of the DC current is suppressed to 8% or less, and the rotation rate of the lower axis (the grown single crystal side) is set to be larger than the rotation number of the upper axis (the raw material polycrystal side). The upper limit is such that the induced non-uniform distribution is suppressed to a practically permissible level. This also applies to the De L
It is not disclosed in the prior art reported by Eon et al.

Further, the present invention will be compared with the prior art reported by De Leon et al. The present invention can be said to be the same technical concept as De Leon in that a solenoid coil is surrounded around a melting zone where a single crystal is grown in the FZ method and a DC magnetic field is applied.
The method targets a grown single crystal with a diameter of 75 mm or more, uses a single-turn coil as the high-frequency induction heating coil,
Moreover, the inner diameter of the coil is at least smaller than the diameter of the grown single crystal, and the length of the melting zone is smaller than the grown single crystal diameter, and the shape for that is smaller than that of the case of a small diameter such as De Leon. The height is significantly reduced with respect to the diameter, so that even when a single crystal having a diameter of 75 mm or more is grown, a non-uniform distribution of the electric resistivity is practically recognized while preventing the deterioration of the dopant concentration distribution in the cross section. Since the FZ condition is significantly different from that in the case of De Leon or the like, the present invention cannot solve the technical problem of the present invention by merely extending the disclosed technology of De Leon or the like.

De Leon et al. Reported that a silicon single crystal having a diameter of 42 mm had a maximum rotation of 18 rotations at an upper shaft rotation speed of 7 rotations per minute and a lower rotation speed of 3.5 rotations per minute at a maximum of 18 rotations.
It was reported that a magnetic field of 0 Gauss was applied, and at 80 Gauss, the radial variation [R (peripheral) / R (center)] of electric resistivity showed the lowest value. However, from the results of the present invention, if a silicon single crystal rod having a diameter of 42 mm is grown by the method of the present invention, the optimum applied magnetic field strength is 500 Gauss or more, and the lower shaft rotation speed is 7 rotations per minute or more. It is expected that the results of the present invention will differ from the results reported by De Leon et al.

The present inventors have considered from a completely different angle to the disclosed technology of De Leon, etc., namely, the inner diameter and shape of the high-frequency induction heating coil, the position of the solenoid coil with respect to the melting zone, the direction of the magnetic field lines, and the strength of the magnetic field. The present inventors have conducted various studies on the lower shaft rotation speed, the amount of ripple, and the like, and have reached the present invention.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be illustratively described in detail below with reference to the drawings. However, unless otherwise specified, the dimensions, materials, shapes, relative arrangements, and the like of the components described in this embodiment are not intended to limit the scope of the present invention thereto, but are merely described. It is only an example.

FIGS. 1 (a) and 1 (b) are views showing the overall configuration of an apparatus for growing a silicon single crystal by the FZ method according to an embodiment of the present invention. The apparatus shown in FIG. After melting a crystal rod 1 with a high-frequency coil 2 having a single-turn inner diameter of 23 mm, a seed crystal 7 is adhered and dislocation-free by an aperture 6, and then a silicon single crystal rod 3 can be grown. The polycrystalline silicon rod 1 and the monocrystalline silicon rod 3 are rotated by an upper shaft 10 and a lower rotation shaft 8, respectively. At this time, the upper shaft 10 was rotated at a rotation speed of 0.4 rotations per minute, and the lower shaft was rotated at a rotation speed of 0.5 to 10 rotations per minute in the same direction.
For the high-frequency coil 2, a single-turn flat coil whose inner diameter is at least smaller than the diameter of the grown single crystal is used, and the axial length of the melting zone is set to be smaller than the grown single crystal diameter. 1 (a), the solenoid coil 9 is placed on the outer wall of the chamber 5 below the high-frequency coil 2, and in FIG. 1 (b), the solenoid coil 9 is placed on the chamber 5 above the high-frequency coil 2. In FIG. 1A, the solenoid coil 9 is arranged so as to surround a non-molten raw material polycrystalline rod in an axially lower position slightly deviated from a melting zone. In FIG. 1 (b), the solenoid coil 9 is arranged so as to surround the non-melted grown single crystal rod at an axially upper position slightly deviated from the melting zone. That is, specifically, the distance between the center of the high-frequency coil 2 and the center of the solenoid coil 9 is approximately 175 mm below or above the high-frequency coil 2 and concentric with the growth axis.
The dimensions of the solenoid coil 9 are 210 mm in inner diameter, 500 mm in outer diameter, and 130 mm in height.

In the apparatus shown in FIG. 1A, phosphine was flowed into the chamber 5 as a dopant to dope phosphorus, and an n-type silicon single crystal having a growth orientation of <111> was grown. In particular, here, a diameter of 75 mm, 100
mm and 125 mm silicon single crystal rods 3 were taken up. A direct current having a ripple rate of 8% or less was passed through the solenoid coil 9 so that the measured value at the center of the growth interface was changed to a magnetic force range of 0 to 1000 gauss to grow a crystal. At this time, the rotation speed of the silicon single crystal rod 3 is changed from 0.5 to 10 rotations per minute in the same direction while the rotation speed of the upper shaft 10 is constant at 0.4 rotations per minute, and forced convection in the silicon melting zone 4 is reduced. It was made to fluctuate.

The silicon single crystal rod 3 was taken out of the chamber 5 after growth, and a silicon wafer having a thickness of 300 μm was cut out from a predetermined position with a diamond saw to obtain a sample for electric resistivity measurement. After measuring the electric resistivity R of the cut wafer by a four-probe measuring method, the average value of the electric resistivity R in the wafer surface is Rave, and the maximum value is Rm.
Assuming that amax and the minimum value are Rmin, the rate of change A in electrical resistivity is A = [(R−Rave) / Rave] × 100 (%), and the rate of change a in the cross section of electrical resistivity is a = [(Rmax −Rmin) / Rmin] × 100 (%).

FIG. 2 is a graph in which the value of the electric resistivity change rate A is plotted with respect to the distance from the center of the wafer. FIG. 2A shows the case where a magnetic field is applied to a melting zone in which the silicon single crystal is grown. Indicates when not to do so. FIG. 2 (b)
The case where a magnetic field strength of 250 Gauss is applied to the melting zone is shown. Here, a measurement was performed on a wafer cut out from a silicon single crystal rod (growth orientation <111>, phosphorus-doped n-type crystal) having a diameter of 100 mm generated by rotating the lower rotation shaft 8 at a rotation speed of 6 rotations / minute. Separately, when the variation a in the cross section of the electrical resistivity is obtained from the measured value, it is 22.1% and 9.7%, respectively, which indicates that a uniform distribution in the cross section was obtained by applying a magnetic field. .

FIG. 3 is a table showing the in-section variation rate a of the electrical resistivity for the sample wafers having diameters of 75 mm, 100 mm and 125 mm. However, each of them rotates in an appropriate range around a suitable lower shaft rotation speed, and the intensity of the magnetic field by a DC current having a ripple rate of 3% is 0 to 1.
This is a measurement and arrangement of a sample wafer cut from a silicon single crystal rod applied and grown over a range of 000 gauss.

The lower shaft rotation speed and the strength of the magnetic field, which are preferable growth conditions, are shown in Table 3 above.
Is read at a small value, the rotation speed is 1 to 8 rotations / minute, the strength of the magnetic field is 190 to 600 gauss, and more preferably, the optimum condition that the in-section variation rate a is the minimum value is
In the case of a silicon single crystal rod having a diameter of 75 mm, the intensity of the magnetic field applied at a lower shaft rotation speed of 7 rotations / minute is 500 gauss,
It can be seen that at a diameter of 100 mm, 250 gauss at 6 revolutions / min and at a diameter of 125 mm, 220 gauss at 2 revolutions / min. The reason why the lower shaft rotation speed is reduced as the diameter of the wafer is increased and the magnetic field strength is not reduced to obtain a good result unless the lower shaft rotation in the silicon melting zone is described as described above. It is considered that the uneven distribution of the thickness of the boundary diffusion layer was improved on the delicate balance between the centrifugal force and the force acting on the magnetic field.

FIG. 4 shows the variation a in the cross section of the electrical resistivity of the single crystal rods of various diameters when the lip ratio included in the DC current under the above-mentioned optimum conditions was changed to 3 to 15%. FIG. It can be seen from FIG. 4 that the range of the ripple rate in which the value of the variation rate a in the cross section is a value small enough to be allowed is 8% or less.

Next, the diameter of the silicon single crystal rod is 130 m.
The second preferred embodiment in the case where m is exceeded is further fried.
The same silicon single crystal growth apparatus by the FZ method as in FIG. 1 was used. Further, the same solenoid coil was used, and the spatial positional relationship with the high-frequency coil 2 was the same. Dopants include:
Phosphine is flowed into the chamber 5 to dope phosphorus, and the growth direction is <111>. An n-type silicon single crystal rod having a diameter of 150 mm was grown. DC currents with ripple rates of 3%, 8% and 15% were passed through the solenoid coil 9 and the measured value at the center of the growth interface was changed from 0 gauss to 2 gauss.
It varied within the range of 50 Gauss. At this time, the upper shaft 10
The rotation speed of the silicon single crystal rod 3 was changed from 0.5 rotations per minute to 4 rotations per minute in the same direction, with the rotation speed kept constant at 0.4 rotations per minute.

FIG. 5 is a graph in which the value of the electric resistivity change rate A in the second embodiment is plotted with respect to the distance from the center of the wafer similarly to FIG. When the rotation speed of the crystal rod is (a) 0 gauss,
Two revolutions per minute, and (b) 185 gauss, two revolutions per minute are shown.

FIG. 6 shows that the applied magnetic force is from 0 gauss to 250 gauss, and the rotation speed of the silicon single crystal rod is 0.5 g / min.
It is the table | surface figure which changed from the rotation to 4 rotations per minute, and measured and calculated the variation rate a in a cross section of an electrical resistivity. In comparison with the case where the diameter of the silicon single crystal rod is 75 mm to 125 mm in FIG. 3, the preferable range is reduced and moved when the diameter is 150 mm, and the rotation speed of the silicon single crystal rod is 0.5 mm / min. From rotation to 4 revolutions per minute, the applied magnetic field strength is 180
It turns out that it becomes 200 Gauss from Gauss. The diameter is 75
mm to 125 mm or more, 150 m in diameter
In addition to m, an attempt was made for a diameter of 140 mm, but the optimum range was the same.

FIG. 7 is a table showing the influence of the ripple rate contained in the DC current on the in-section variation rate a of the electrical resistivity. Regarding the ripple rate, as the diameter of the silicon single crystal rod increases, the allowable upper limit value seems to decrease. However, for a ripple rate of 8%, the in-section variation rate a of the electrical resistivity is about 16% and the magnetic field is increased. Compared to those when no voltage is applied, it is significantly improved.

Although the above experiment was not carried out with respect to FIG. 1 (b), the growth apparatus uses the grown single crystal in which the solenoid coil 9 is placed above the high-frequency coil 2 in other words, in the non-melted bear above the melting zone. Except for the point that it is disposed around the outer periphery of
Even in this growth apparatus, since the magnetic field similarly includes the melting zone and the crystal growth area, effects similar to those of the growth apparatus are easily expected. Also, the solenoid coil 9 is connected to the high-frequency coil 2
If the magnetic fields formed by the two solenoid coils are in the same direction when they are simultaneously arranged above and below the magnetic field, the strength of the magnetic field formed by each of the solenoid coils is approximately half that of the single solenoid coil, The same effect as that of the growth apparatus described above is achieved. Further, when the magnetic field formed by the two solenoid coils is in the opposite direction, when the difference between the strengths of the magnetic fields becomes substantially equal to the strength of the magnetic field in the single solenoid coil, the same as the growth apparatus in the above embodiment is performed. The effect can be expected to be obtained.

[0033]

As described above, according to the present invention, FZ
In a manufacturing method for growing a silicon single crystal rod having a diameter of 75 mm or more according to the method, the dopant distribution in the cross section in the diameter direction of the silicon single crystal rod can be microscopically uniform. Further, according to the present invention, since the step of doping by thermal neutron irradiation is not included in the step of the FZ method, the silicon single crystal rod can be grown at a desired cost. And so on.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of a silicon single crystal growth apparatus according to the present invention. FIG. 1 (a) is an overall configuration diagram in a case where a solenoid coil is disposed below a melting zone, and FIG. 1 (b) is a solenoid. FIG. 4 is an overall configuration diagram when a coil is disposed at a position above a melting zone.

FIG. 2 is a distribution diagram of an electric resistivity change rate A in a cross section of the silicon single crystal rod according to the first embodiment of the present invention, and FIG. 2 (a) is a distribution diagram when no magnetic field is applied; (B)
8 is a distribution diagram when a magnetic field is applied.

FIG. 3 shows the growth of silicon single crystal rods of various diameters according to the first embodiment of the present invention when the lower shaft rotation speed and the intensity of the applied magnetic field are appropriately changed within a range. FIG. 4 is a table showing the electric resistivity variation rate a in the cross section of the rod.

FIG. 4 is a cross-sectional view of a silicon single crystal rod of various diameters according to a first embodiment of the present invention when the ripple rate included in a direct current forming a magnetic field to be applied is changed. FIG. 4 is a table showing the electrical resistivity fluctuation rate a in FIG.

FIG. 5 is a distribution diagram in which values of a change rate of electric resistivity A according to a second embodiment of the present invention are plotted with respect to a distance from the center of the wafer, and FIG. FIG. 5B is a distribution diagram when a magnetic field is applied.

FIG. 6 is a cross-sectional view of a silicon single crystal rod according to a second embodiment of the present invention, when the lower shaft rotation speed and the intensity of an applied magnetic field are appropriately changed within the range. FIG. 4 is a table showing the electrical resistivity fluctuation rate a in FIG.

FIG. 7 is a graph showing the relationship between the electric current in the cross section of a single-crystal silicon rod and the variation of the ripple included in the DC current that forms the applied magnetic field during the growth of the silicon single-crystal rod according to the second embodiment of the present invention. FIG. 4 is a table showing a resistivity variation rate a.

FIG. 8 is an overall configuration diagram of a conventional silicon single crystal growth apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Silicon polycrystal rod 2 High frequency coil 3 Silicon single crystal rod 4 Melting zone 5 Chamber 6 Restrictor 7 Seed crystal 8 Lower rotation axis 9 Solenoid coil

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification number Agency reference number FI Technical indication location H01L 21/208 H01L 21/208 M H05B 6/32 H05B 6/32 (56) Reference JOURNAL OF CRYSTA L GROWTH, 55 (1981) p. 406 -408

Claims (5)

(57) [Claims] 1. A method for producing a silicon single crystal rod having a diameter of 75 mm or more by the FZ method, wherein a single-turn coil having an inner diameter of a coil smaller than at least the diameter of a grown single crystal is used as a high-frequency induction heating coil. The axial length of the melting zone is set to be smaller than the diameter of the grown single crystal, and the raw material polycrystalline rod or the grown material in the non-molten state at an upper position and / or a lower position in the axial direction outside the melting zone. A method for growing a silicon single crystal rod, wherein a magnetic field forming means is arranged so as to surround the single crystal rod, and a magnetic field is applied to the melting zone of the silicon single crystal rod by the magnetic field forming means. 2. The method according to claim 1, wherein the magnetic field forming means is a solenoid coil surrounding an outer periphery of the silicon single crystal rod or the raw material polycrystal rod in the non-molten state, and supplies a direct current to the solenoid coil. The method for growing a silicon single crystal rod according to claim 1. 3. The method according to claim 1, wherein the ripple rate of the direct current is suppressed to 8% or less, and the number of rotations of the lower axis (growth single crystal side) is set to be larger than the number of rotations of the upper axis (polycrystalline material side). 3. The method for growing a silicon single crystal rod according to claim 2, wherein 4. A method for producing a silicon single crystal rod having a diameter of 75 to 130 mm by an FZ method, wherein a magnetic field intensity of said magnetic field forming means is from 150 gauss to 600 gauss.
The method is characterized in that the single crystal rod is rotated while applying a magnetic field to the melting zone of the silicon single crystal rod while being set to be between Gauss, and the number of rotations is set to 1 to 8 rotations per minute. Item 3. The method for growing a silicon single crystal rod according to Item 3. 5. A method for producing a silicon single crystal rod having a diameter of 130 mm or more, wherein the magnetic field intensity of said magnetic field forming means is from 180 gauss to 200 gauss.
4. The method for growing a silicon single crystal rod according to claim 3, wherein the silicon single crystal rod is set between Gauss and the rotation speed of the silicon single crystal rod is set from 0.5 rotations per minute to 4 rotations per minute.