JP4380204B2 - Silicon single crystal and single crystal growth method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a P (phosphorus) -doped n-type silicon single crystal grown by the CZ method and a method for growing the single crystal.
[0002]
[Prior art]
A silicon single crystal used as a semiconductor material is manufactured exclusively by the CZ method. In the production of a silicon single crystal by the CZ method, a seed crystal is immersed in a raw material melt contained in a quartz crucible, and the seed crystal is pulled up while rotating the seed crystal and the crucible from this state, thereby lowering the seed crystal. A silicon single crystal is grown. Here, the resistivity of the single crystal is adjusted by the dopant added to the raw material melt. The dopant is roughly classified into n-type and p-type. P (phosphorus) is used as the n-type dopant, and B (boron) is used as the p-type dopant. It is often used because it is easy to be taken in.
[0003]
One of the problems in crystal growth by such a CZ method is a variation in resistivity in the crystal axis direction. This is a problem caused by the segregation of the dopant. Because of the segregation, the dopant is gradually concentrated in the silicon residual liquid in the crucible, and the dopant concentration in the residual liquid gradually increases, so that the resistivity of the crystal is increased. Is a phenomenon that changes continuously in the axial direction. In the case of a p-doped n-type silicon single crystal, the resistivity decreases as the P concentration increases from the top to the bottom of the crystal. Due to the change in resistivity in the crystal axis direction, the product portion where the resistivity falls within the standard range is as small as about half of the case of p-type silicon single crystal by B doping, which is why the yield does not increase sufficiently. It has become one of
[0004]
In order to solve this problem, a sub-dopant whose conductivity type is opposite to that of the main dopant that defines the conductivity type is added to the raw material melt in the crucible to cancel the resistivity change due to dopant segregation in the crystal axis direction. This is presented in Patent Document 1 and Patent Document 2.
[0005]
[Patent Document 1]
Japanese Patent No. 2550739 [0006]
[Patent Document 2]
Japanese Patent Laid-Open No. 2002-128591
As a specific example, in Patent Document 1, when a p-doped n-type single crystal is grown, B, which is a p-type dopant, is grown with the progress of crystal growth so that the resistivity is constant in the crystal axis direction. It is explained that it is continuously added to the raw material melt while increasing the amount. In Patent Document 2, when growing a p-type single crystal doped with Ga (gallium), Bi (bismuth), which is an n-type dopant, is initially added to the raw material melt, and then added several times during crystal growth. And a method of performing only several additional additions during crystal growth without performing initial addition before crystal growth. .
[0008]
[Problems to be solved by the invention]
By using a subdopant having a conductivity type different from that of the main dopant, it is possible to suppress a change in resistivity due to the segregation of the dopant in the crystal axis direction. However, in Patent Document 1, the addition of the sub-dopant is performed in a form that involves additional addition during crystal growth. The operation of adding additional sub-dopant continuously and intermittently during crystal growth is not easy in actual operation.
[0009]
On the other hand, Patent Document 2 does not disclose a method for making the resistivity distribution in the crystal axis direction uniform for a P-doped n-type single crystal. The segregation coefficient of P for silicon is about 0.35, which is less than half the segregation coefficient of B for silicon (about 0.8). For this reason, the yield of the P-doped n-type single crystal is considerably lower than that of the B-doped p-type single crystal.
[0010]
The object of the present invention relates to a p-doped n-type silicon single crystal, silicon that can make the resistivity distribution in the crystal axis direction uniform even without initial addition of a sub-dopant during crystal growth and only with initial addition before crystal growth. The object is to provide a single crystal and a method for growing the single crystal.
[0011]
[Means for Solving the Problems]
In order to achieve the above object, the present inventors, as a sub-dopant for P, which is the main dopant, are p-types having different conductivity types and sufficiently small segregation coefficients, such as Ga (gallium), In (indium) and Al ( We focused on (aluminum). The segregation coefficients for these silicon are about 0.005 for P, about 0.008 for Ga, about 0.0004 for In, and about 0.002 for Al, with a difference of two to three digits.
[0012]
When a raw material melt in which a sub-dopant having a sufficiently small segregation coefficient relative to the main dopant is initially added before crystal growth is used, the amount taken into the single crystal is small because the segregation coefficient is small at the initial stage of growth. . For this reason, there is almost no influence by a subdopant. However, as the growth progresses, the concentration of the sub-dopant remarkably advances as compared with the main dopant, and the influence of the sub-dopant begins to appear. This effect becomes more noticeable at an accelerated rate as the breeding progresses. Thus, even if the additional addition is not performed during the crystal growth, the effect as if the additional addition was performed is obtained.
[0013]
Based on such an idea, the present inventors conducted various investigation experiments. As a result, when a subdopant having a sufficiently small segregation coefficient with respect to the main dopant is initially added before crystal growth, the uniformity of the resistivity distribution in the crystal axis direction is the addition rate of the subdopant at the beginning of crystal growth, that is, It was determined by the initial addition rate that the setting of the initial addition rate is important to improve the uniformity. In addition, it has been found that the degree of influence of sub-dopant types on the uniformity of resistivity distribution is not significantly different among Ga, In and Al.
[0014]
FIGS. 1A to 1C are graphs showing the influence of the addition rate of the sub-dopant on the resistivity distribution in the crystal axis direction for Ga, In, and Al. The addition rate is represented by the concentration ratio of the sub-dopant and the main dopant in the top portion of the crystal body and is adjusted so that the resistivity in the top portion is the same (50 Ω · cm).
[0015]
When the initial addition rate is 0, that is, when a sub-dopant is not used, the resistivity decreases as the pulling progresses. When the secondary dopant is added, the decrease in resistivity is eased, and the influence of the secondary dopant becomes more effective in the latter half of the pulling, so that an inflection point at which the change in resistivity turns from a decrease to an increase appears. This tendency of homogenization becomes more prominent as the addition rate increases, and becomes clear from the addition rate of 30%, and the time when the decrease in resistivity starts to increase becomes earlier. Thereby, the product part in which the resistivity falls within the standard range becomes long. If the addition rate further increases, the influence of the sub-dopant becomes excessive, and the resistivity rapidly increases from the initial pulling up, and the product portion where the resistivity falls within the standard range is conversely shortened. This tendency is almost the same among Ga, In and Al.
[0016]
The present invention has been completed based on such knowledge, and the silicon single crystal is a P-doped n-type silicon single crystal grown by the CZ method, and Ga, In as a sub-dopant together with P as a main dopant. Or at least one of Al, and the concentration difference between the main dopant and the sub-dopant in the top portion of the crystal straight body portion corresponds to the target resistivity in the portion, and the concentration of the sub-dopant and the main dopant in the top portion The ratio is controlled to 30 to 70%.
[0017]
In addition, the single crystal growth method of the present invention, when growing a P-doped n-type silicon single crystal by the CZ method, uses at least one of Ga, In or Al as a sub-dopant for P which is a main dopant. Before crystal growth so that the concentration difference between the main dopant and the sub-dopant in the top portion of the trunk corresponds to the target resistivity in the portion and the concentration ratio of the sub-dopant and the main dopant in the top portion is 30 to 70%. The raw material melt initially added to is used.
[0018]
In the following description, the concentration of the main dopant taken into the top portion of the single crystal straight body portion is represented by Cs ₁ , and the concentration of the sub-dopant is represented by Cs ₂ . Accordingly, the concentration difference between the main dopant and the sub-dopant in the top portion of the crystal body portion is expressed as Cs ₁ -Cs ₂ , and the concentration ratio between the sub-dopant and the main dopant in the top portion is expressed as Cs ₂ / Cs ₁ . Further, the relationship of Formula 1 is established between the dopant concentration Co in the silicon melt before the start of pulling and the dopant concentration Cs in the crystal.
[0019]
[Expression 1]
Cs = Co × k × (1−g) ^k−1
g: Pull rate k: Segregation coefficient for silicon
When the concentration ratio (Cs ₂ / Cs ₁ ) between the sub-dopant and the main dopant in the top portion of the crystal body is less than 30%, the effect of uniforming the resistivity distribution in the crystal axis direction is small. If this concentration ratio (Cs ₂ / Cs ₁ ) exceeds 70%, the degree of influence on the resistivity distribution in the crystal axis direction becomes excessive, so the effect of making the resistivity distribution in the crystal axis direction uniform is obtained. Disappear. In addition, it is necessary to secure a concentration difference (Cs ₁ −Cs ₂ ) corresponding to the target resistivity at the top portion of the crystal body, and the amount of dopant added is excessive. A particularly desirable concentration ratio (Cs ₂ / Cs ₁ ) is 40% or more for the lower limit and 60% or less for the upper limit.
[0021]
In a specific configuration, the single crystal growth method of the present invention adds P as a main dopant when growing a P-doped n-type silicon single crystal by the CZ method, and Ga, In or A method of growing a single crystal using a raw material melt to which at least one kind of Al is added, wherein the concentration ratio in the top portion of the crystal straight body portion is varied in a range of 32 to 61%, and the concentration ratio is 32. % Or the resistivity decreases from the top part to the bottom part of the crystal body, the target resistivity at the top part of the crystal body is set to the upper limit in the standard range, and the concentration difference is adjusted accordingly. controls, whereas, when the concentration ratio of 61% or crystal straight body resistivity toward the bottom portion from the top portion of the increases, the aim in the top part of the crystal straight body portion Set anti rate to the lower limit of the standard range, by controlling the density difference so as to correspond thereto, characterized in that during growing the single crystal not add addition of secondary dopant.
The single crystal growth method of the present invention is a growth method suitable for growing a P-doped n-type silicon single crystal having a resistivity standard range of 54 to 66 Ωcm.
In addition, the top part of the crystal straight body part is a constant-diameter part start position immediately after the shoulder part forming step for increasing the silicon single crystal diameter to a predetermined crystal diameter, which is performed by the CZ method, is completed. Say.
[0022]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below.
[0023]
When growing a P-doped n-type silicon single crystal by the CZ method, a silicon melt initially added with Ga, In or Al as a sub-dopant together with P as a main dopant is formed in a crucible by a conventional method. . Here, the sub-dopant Ga, In or Al may be added singly or in combination of two or more. Thereafter, a silicon single crystal is grown from the silicon melt by a conventional method. During the growth of single crystals, no additional sub-dopants are added in principle.
[0024]
The initial addition amount of the sub-dopant is such that the main dopant concentration in the silicon single crystal grown from the silicon melt is Cs ₁ and the sub-dopant concentration is Cs _2. the corresponding density difference _{(Cs 1 -Cs 2), 30~70} % of the concentration ratio (Cs ₂ / Cs _1), preferably such that the concentration ratio of 40~60 (Cs ₂ / Cs ₁₎ is secured, This is determined based on Equation 1 described above.
[0025]
Here, the target resistivity at the top part of the single crystal straight body part will be described.When the resistivity of the single crystal decreases from the top part to the bottom part, the target resistivity is set to the vicinity of the upper limit value in the standard range of the resistivity. It is good to do. When the resistivity of the single crystal increases from the top portion to the bottom portion, the resistivity should be set near the lower limit in the resistivity specification range. When the resistivity of the single crystal decreases from the top part to the bottom part and then starts to increase, the difference between the upper limit value and the lower limit value is set so that the minimum value of the resistivity matches the lower limit value in the resistivity specification range. The target resistivity should be set.
[0026]
This is the target resistivity setting method when the first half of the growth from the top part of the single crystal straight body part to the middle part is commercialized, but it is also possible to commercialize the middle part of the single crystal straight body part In this case, the target resistivity at the top portion is shifted from the above-described target resistivity, and it goes without saying that the target resistivity is out of the specified range. As described above, the target resistivity in the top portion may be out of the standard range of the resistivity, and does not necessarily mean the standard range.
[0027]
As a general rule, this operation is not performed for the additional addition of the sub-dopant during single crystal growth, but this operation is not completely excluded. Even if an additional addition of a subdopant is performed during single crystal growth, the present invention is effective in that the addition amount and the number of times can be reduced.
[0028]
【Example】
Next, the effectiveness of the present invention will be described with reference to examples.
[0029]
When 140 kg of silicon was melted to grow a P-doped n-type silicon single crystal having a diameter of 8 inches and a resistivity standard range of 54 to 66 Ωcm, Ga was initially added as a sub-dopant to the silicon melt. . The addition amount of the subdopant was adjusted so that the concentration ratio (Cs ₂ / Cs ₁ ) between the subdopant and the main dopant in the top portion of the crystal body was 32% and 61%.
[0030]
FIG. 2 shows a dopant concentration distribution, a concentration difference (Cs ₁ -Cs ₂ ) distribution, and a resistivity distribution in the crystal axis direction when no subdopant is added. FIG. 3 shows the dopant concentration distribution, the concentration difference (Cs ₁ -Cs ₂ ) distribution, and the resistivity distribution in the crystal axis direction when the sub-dopant is added at a concentration ratio (Cs ₂ / Cs ₁ ) of 32%. FIG. 4 shows the dopant concentration distribution, the concentration difference (Cs ₁ -Cs ₂ ) distribution, and the resistivity distribution in the crystal axis direction when 61% of the sub-dopant is added at a concentration ratio (Cs ₂ / Cs ₁ ).
[0031]
In the case where no subdopant is added, the resistivity decreases monotonously from the top portion to the bottom portion of the crystal body. The target resistivity at the top portion of the crystal body is the upper limit (66 Ωcm) of the standard range, and in order to ensure this, the P concentration at the top portion was about 7.6 × 10 ¹³ atoms / cc. When the pulling rate was about 0.32, the resistivity was out of the standard range. The product part that satisfies the standard range is only 28% of the total length of the straight body part.
[0032]
When the initial addition of the sub-dopant is performed so that the concentration ratio (Cs ₂ / Cs ₁ ) in the top part of the crystal body is 32%, the resistivity decreases from the top part to the bottom part of the crystal body However, the degree is milder than when no additive is added. The target resistivity at the top portion of the crystal body is set to the upper limit value (66 Ωcm) of the standard range, and in order to ensure this, the concentration difference (Cs ₁ −Cs ₂ ) at the top portion is about 7.6 × 10. ¹³ atoms / cc. In order to secure a concentration difference (Cs ₁ -Cs ₂ ) after initially adding the sub-dopant, the amount of the main dopant is increased. Despite the addition of no additional dopant, the pulling rate at which the resistivity deviated from the standard range increased by about 0.1, and the product portion satisfying the standard range was 36% of the total length of the straight body.
[0033]
When the initial addition of the sub-dopant is performed so that the concentration ratio (Cs ₂ / Cs ₁ ) in the top part of the crystal body is 61%, the resistivity increases from the top part to the bottom part of the crystal body I tend to. The target resistivity at the top portion of the crystal body is set to the lower limit (54 Ωcm) of the standard range, and in order to ensure this, the concentration difference (Cs ₁ −Cs ₂ ) at the top portion is about 9.2 × 10. ¹³ atoms / cc. In order to secure the concentration difference (Cs ₁ -Cs ₂ ) after increasing the initial amount of sub-dopant, the amount of main dopant is further increased. Despite the addition of no additional dopant, the pulling rate at which the resistivity deviated from the standard range was further increased, and the product portion that satisfied the standard range reached 55% of the total length of the straight body. The yield is almost double compared to the case of no addition.
[0034]
The same comparative test was performed by changing the sub-dopant from Ga to In. The results are shown in FIGS. As in the case of Ga, the resistivity distribution in the direction of the crystal axis is made uniform even though no sub-dopant is added, and the product portion that satisfies the standard range is 37%, 61% when 32% is added. In the case of% addition, it increased to 53%.
[0035]
The same comparative test was performed with the subdopant changed to Al. The results are shown in FIGS. As in the case of Ga and In, the resistivity distribution in the crystal axis direction is uniformed even though no additional dopant is added, and the product portion satisfying the standard range is 37% when 32% is added. In the case of 61% addition, it increased to 54%.
[0036]
【The invention's effect】
As described above, the silicon single crystal of the present invention is a P-doped n-type silicon single crystal grown by the CZ method, and at least one of Ga, In, or Al as a sub-dopant, By containing so that the concentration ratio of the sub-dopant and the main dopant in the top portion of 30 to 70%, the resistivity distribution in the crystal axis direction can be made uniform without adding an additional sub-dopant during crystal growth, Since products that fall within the resistivity range can be collected with a high yield, it is very economical.
[0037]
The silicon wafer manufactured from the silicon single crystal of the present invention can be used not only as a bulk wafer but also as an annealing or epitaxial wafer and a support substrate for an SOI wafer.
[0038]
In addition, the single crystal growth method of the present invention, when growing a P-doped n-type silicon single crystal by the CZ method, uses at least one of Ga, In or Al as a sub-dopant for P which is a main dopant. By using the raw material melt initially added before crystal growth so that the concentration ratio of the sub-dopant to the main dopant in the top part of the trunk is 30 to 70%, no additional sub-dopant is added during crystal growth. In both cases, the resistivity distribution in the crystal axis direction can be made uniform, and the yield at the time of product collection can be greatly improved.
[Brief description of the drawings]
FIG. 1 is a graph showing the influence of the addition rate of a sub-dopant on resistivity distribution in the crystal axis direction for Ga, In, and Al.
FIG. 2 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when Ga is not added.
FIG. 3 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when Ga is added at 32%.
FIG. 4 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when 61% Ga is added.
FIG. 5 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when In is not added.
FIG. 6 is a graph showing a dopant concentration distribution, a concentration difference distribution, and a resistivity distribution in the crystal axis direction when 32% of In is added.
7 is a graph showing a dopant concentration distribution, a concentration difference distribution, and a resistivity distribution in the crystal axis direction when 61% of In is added. FIG.
FIG. 8 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when Al is not added.
FIG. 9 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when 32% Al is added.
FIG. 10 is a graph showing dopant concentration distribution, concentration difference distribution, and resistivity distribution in the crystal axis direction when 61% Al is added.

Claims (2)

When growing a P-doped n-type silicon single crystal by the CZ method, P as a main dopant is added, and a raw material melt to which at least one of Ga, In, or Al is added as a sub-dopant is added. A method for growing crystals,
The main dopant concentration is Cs1, the subdopant concentration is Cs2, and the concentration ratio (Cs2 / Cs1) in the top portion of the crystal straight body portion is varied in the range of 32 to 61%.
When the concentration ratio (Cs2 / Cs1) is 32% or the resistivity decreases from the top portion to the bottom portion of the crystal straight body portion, the target resistivity at the top portion of the crystal straight body portion is the upper limit value in the standard range. And the density difference (Cs1-Cs2) is controlled to correspond to it,
On the other hand, when the concentration ratio (Cs2 / Cs1) is 61% or the resistivity increases from the top portion to the bottom portion of the crystal body portion, the target resistivity at the top portion of the crystal body portion is within the standard range. Set the lower limit value and control the density difference (Cs1-Cs2) to correspond to it,
A method for growing a single crystal, wherein no additional addition of a sub-dopant is performed during the growth of the single crystal. The method for growing a single crystal according to claim 1, wherein a P-doped n-type silicon single crystal having a resistivity range of 54 to 66 Ωcm is grown .

Images