Method for Evaluating Silicon Single Crystal Ingot

Abstract

To provide a method for evaluating a silicon single crystal ingot that, even when it is a light-dope crystal, can provide improvement in wafer yield while shortening time that is necessary for evaluating resistivity variation in a crystal length method.SOLUTION: A method for evaluating a silicon single crystal ingot, in pulling up a silicon single crystal by a Czochralski method, includes the steps of: performing counter dope of adding sub dopant during the pulling up; measuring resistivity in a crystal length direction of a lateral face with a four-point probe method after pulling up the silicon single crystal while being a single crystal ingot 1; and determining a reject range on the basis of resistivity distribution in the crystal length direction.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a method for evaluating a silicon single crystal ingot manufactured by the Czochralski method (CZ method).

Background Art

[0002] In the CZ method, a silicon single crystal ingot (hereinafter referred to as "single crystal ingot") is manufactured by bringing a seed crystal into contact with a silicon melt and slowly pulling it up while rotating. Such a CZ method is generally used as a method for manufacturing a large-diameter single crystal ingot.

[0003] On the other hand, it is known that the resistivity of a single crystal ingot manufactured by the CZ method changes along the crystal growth direction (crystal length direction). In recent years, due to the requirements for the quality of semiconductor wafers (hereinafter referred to as "wafers") manufactured from single crystal ingots, it has become important to keep the resistivity within a desired range.

[0004] For example, when measuring the resistivity of a wafer, conventionally, a single crystal ingot grown by the CZ method is peripherally ground to finish it to a predetermined dimension (diameter), the head and tail cone portions that cannot be used as products are cut off, the obtained single crystal ingot is cut at a predetermined position, and it is made into a block of a length that can be put into a slicing device such as an inner peripheral blade or a wire saw. At this time, an inspection sample wafer for resistivity and the like is cut out simultaneously. Thereby, the resistivity can be measured using the cut-out inspection sample wafer. Further, each block is sliced to a predetermined thickness to obtain a wafer. Then, the sliced wafer is extracted, and by measuring the resistivity, the resistivity in the crystal length direction can be measured.

[0005] In the growth of single-crystal ingots by the Czochralski method, when a dopant is added, a phenomenon is observed in which the resistivity changes in the crystal growth direction. This is due to the segregation of the dopant. As the silicon melt in the crucible decreases with the growth of the single crystal, the dopant concentration in the remaining liquid gradually increases, and accordingly, the resistivity of the single crystal continuously decreases. The segregation coefficient of P (phosphorus) is 0.35, but it is lower than 0.8, the segregation coefficient of B (boron) widely used as a dopant for p-type crystals. Compared with p-type crystals, the decrease in resistivity from the head to the tail is more significant. Therefore, there is a problem that the portion that can be used as a product is reduced and it is difficult to improve the yield.

[0006] Also, as a countermeasure against the problem of deviating from the desired resistivity range in the crystal length direction as described above, for example, Patent Document 1 below describes a method of measuring the resistivity in the state of a single-crystal ingot.

[0007] Specifically, the resistivity in the crystal length direction on the side surface of the single-crystal ingot is measured, the desired resistivity position indicating the desired resistivity is specified, a block of a predetermined length is cut out, and wafers are sliced from this block. As the method for measuring the resistivity in the crystal length direction, a resistivity measurement method using the four-probe method is applied. Thereby, for example, wafers with a resistivity of 1 Ωcm or less can be manufactured.

[0008] By the way, in a single-crystal ingot grown by the Czochralski method, it is known that it is difficult to make the resistivity distribution in the crystal length direction uniform due to segregation. Therefore, conventionally, a growth method in which a main dopant and a sub-dopant having a polarity opposite to that of the main dopant are added, that is, a single-crystal ingot is grown by so-called counterdoping (see Patent Document 2 below).

[0009] According to Patent Document 2 below, counter doping has "a step of doping a main dopant having, for example, an n-type (for example, phosphorus) when growing a single crystal ingot" and "while growing a single crystal ingot, according to the solidification rate represented by (pulled crystal weight) / (weight of raw material polysilicon), a sub-dopant having a p-type opposite to the n-type (for example, boron) is continuously or intermittently doped." Hereinafter, a single crystal ingot using counter doping may be referred to as a "counter-doped crystal."

Prior Art Documents

Patent Documents

[0010]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0011] According to Patent Document 1 above, when manufacturing a wafer with a low resistivity (1 Ωcm or less), a block can be cut out so as to contain more portions having a desired resistivity, so that the yield of the wafer can be improved. Further, since the resistivity is measured in the state of a single crystal ingot, it is possible to significantly shorten the time required for resistivity evaluation as compared with the method of cutting out a test sample wafer and measuring the resistivity.

[0012] However, the method for manufacturing a wafer described in Patent Document 1 above has the following problems.

[0013] For example, in the case of a heavily doped (low resistivity) wafer with a resistivity of 1 Ωcm or less, the dopant concentration of the main dopant is sufficiently high compared to the amount of thermal donors, and the influence of thermal donors is small without heat treatment, so the resistivity can be measured even in the state of a single crystal ingot. On the other hand, in the case of a lightly doped crystal (for example, a desired resistivity of 10 Ωcm or more), since it is affected by thermal donors, a donor killer (heat treatment) is required to obtain the desired resistivity, but it is difficult to perform the donor killer in the state of a single crystal ingot. That is, in the case of a lightly doped crystal, even if the resistivity is measured in the state of a single crystal ingot, it is difficult to evaluate the true resistivity (hereinafter referred to as the "true resistivity") excluding thermal donors in the crystal length direction.

[0014] On the other hand, in the counter-doped crystal, in the vicinity of the crystal length position where the secondary dopant is introduced, there is a region where the resistivity rapidly increases and then rapidly decreases (hereinafter referred to as the "high resistivity layer"). Therefore, in the counter-doped crystal, it is necessary to detect the above high resistivity layer.

[0015] The present invention has been made in view of the above problems, and an object thereof is to provide an evaluation method capable of detecting a high resistivity layer in the crystal length direction in a lightly doped counter-doped crystal in the state of a single crystal ingot.

Means for Solving the Problems

[0016] The evaluation method of a single crystal ingot according to the present invention is an evaluation method of a single crystal ingot in which counter doping for adding a secondary dopant is performed during the pulling of a silicon single crystal by the CZ method. In the state of the single crystal ingot, the resistivity in the crystal length direction on the side surface of the single crystal ingot is measured by the four-probe method, and the reject range (non-product range) of the single crystal ingot is determined based on the resistivity distribution in the crystal length direction.

[0017] In a single crystal ingot, there exists a region where resistivity fluctuations are stable below a predetermined threshold (e.g., 1 - 3%) (hereinafter also referred to as the "stable layer"). In the counter-doped crystal according to the present invention, a rejection range is determined based on the position where a peak with a resistivity 10% or more higher than that of the stable layer on the head side exists (hereinafter also referred to as the "reference position"). That is, from the peak position (when there are multiple data of 10% or more, the position where it first becomes 10% or more is taken as the reference position), predetermined lengths (e.g., 8 - 12 mm on the head side and 10 - 20 mm on the tail side) are set as the rejection range on the head side and the tail side respectively. Also, the rejection range on the tail side may extend to the position (hereinafter referred to as the "stable position") where the fluctuation rate of the resistivity of the crystal length becomes below the threshold (e.g., 1% or less) and begins to follow segregation.

[0018] According to the above configuration, a high-resistivity layer due to counter-doping can be identified from the resistivity distribution on the side surface of the single crystal ingot, and the rejection range can be determined in the state of the single crystal ingot. Compared with the method (conventional method) of cutting out a sample wafer for inspection from the single crystal ingot and performing resistivity evaluation, the yield of the wafer can be significantly improved. Also, the time required for resistivity evaluation can be shortened compared to the conventional method.

[0019] Also, in the method for evaluating a single crystal ingot according to the present invention, it is desirable that the measurement range of resistivity includes at least the range from the position where the secondary dopant is introduced to the position where the resistivity rises, then rapidly decreases, and finally stabilizes at the desired resistivity again.

[0020] Furthermore, in the method for evaluating a single crystal ingot according to the present invention, it is desirable that the resistivity of the single crystal ingot to be grown is 10 Ωcm or more.

Effects of the Invention

[0021] According to the present invention, the time required for evaluating the resistivity in the crystal length direction can be shortened, and furthermore, the yield of the wafer can be improved.

Brief Description of the Drawings

[0022]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Embodiments for Carrying out the Invention

[0023] Hereinafter, embodiments of the evaluation method of the single crystal ingot according to the present invention will be described in detail. Note that the present invention is not limited by this embodiment. Further, although this embodiment will be described as an n-type single crystal ingot, it is not limited thereto, and a p-type single crystal ingot may also be used.

[0024] In this embodiment, in the pulling of an n-type lightly doped (10 Ωcm or more) single crystal ingot by the CZ method, counter doping is performed by adding a p-type dopant as a sub-dopant during the pulling. For example, in a single crystal ingot grown by the CZ method, it is difficult to make the resistivity distribution uniform in the direction in which the crystal grows (crystal growth direction) due to segregation, but such a problem can be solved by counter doping with a sub-dopant having the opposite polarity to the main dopant.

[0025] For example, when silicon crystallizes, the concentration of dopants incorporated into the crystal is lower than the dopant concentration in the melt. Since the growth of the single crystal ingot is continuous, a large amount of dopants remains in the melt, and the dopant concentration in the melt gradually increases. Accordingly, the dopant concentration in the crystal also gradually increases, and the resistivity decreases. In order to prevent the resistivity from decreasing and falling below the desired resistivity range, a counter-dopant of the opposite polarity is introduced into the melt while adjusting it to an appropriate amount (counter-doping). At the crystal growth position where the counter-dopant is introduced, the resistivity increases, and then the resistivity decreases and stabilizes according to segregation.

[0026] In this embodiment, as the n-type dopant (main dopant), for example, P (phosphorus), As (arsenic), Sb (antimony), etc. can be used, and as the p-type dopant (counter-dopant), for example, B (boron), Al (aluminum), Ga (gallium), etc. can be used. However, the main dopant can also be p-type and the counter-dopant can be n-type.

[0027] In this embodiment, the counter-doped crystal is pulled up, subjected to outer peripheral grinding to finish it to a predetermined dimension (diameter), the head and tail cone portions are cut off, and then the resistivity in the crystal growth direction on the side surface is measured by the four-probe method in the state of a single crystal ingot, and the relative resistivity variation in the crystal growth direction is evaluated based on the resistivity distribution. Specifically, based on the resistivity distribution (relative resistivity distribution) in the crystal growth direction, from the peak position (reference position) where the resistivity variation is 10% or more higher than the resistivity of the stable layer on the head side to a position returned a predetermined length (for example, 8 to 12 mm) toward the head side, the range from the returned position to the stable position is detected. Alternatively, from the position returned a predetermined length (for example, 8 to 12 mm) toward the head side from the reference position, a predetermined length (for example, 10 to 20 mm) from the reference position toward the tail side is determined as the rejection range. However, the single crystal ingot is not limited to outer peripheral grinding.

[0028] FIG. 1 is a diagram showing an overview of resistivity measurement on the side surface of a single crystal ingot. After pulling up the counter-doped crystal, the outer circumference of the counter-doped crystal is ground, and the head and tail cones are cut off to obtain a single crystal ingot 1 as shown in FIG. 1. And since the crystal length position (counter-dopant injection position) where the co-dopant is injected has been recorded in advance during pulling, the resistivity of the side surface of the single crystal ingot 1 including the counter-dopant injection position is measured by the four-probe method. Specifically, as shown in FIG. 1, starting from the position 10 mm back from the counter-dopant injection position toward the head side, the resistivity of the side surface of the single crystal ingot 1 is measured at positions with a 1 mm pitch and 30 points in the crystal length direction. At this time, a sharp increase in resistivity is observed near the counter-dopant injection position, and then, after a sharp decrease, the resistivity gradually decreases according to segregation. Regarding the measurement range of the resistivity, it is sufficient that the range from the counter-dopant injection position to the position where the resistivity rises again and stabilizes at a desired resistivity is included, and it is not limited to this (the above 1 mm pitch and 30 points).

[0029] And in the present embodiment, measurement is performed starting from the stable layer on the head side of the crystal length, and the peak position where the peak of the resistivity is 10% or more higher than the resistivity of the stable layer is used as the reference position. Further, the resistivity fluctuation on the tail side of the crystal length becomes stable, and the position where the fluctuation rate of the resistivity becomes equal to or less than a threshold value (for example, 1% to 3%) and starts to follow segregation is defined as the stable position, and the range from the position 10 mm from the reference position toward the head side to the stable position is defined as the rejection range. Thereby, compared with the method (conventional method) of cutting out an inspection sample wafer from a single crystal ingot and measuring the resistivity, the processing loss in the wafer processing step and the time required for evaluating the resistivity fluctuation can be significantly reduced. Note that the rejection range on the tail side from the peak position is not limited to the range up to the above stable position, and may be set to a predetermined length. For example, the rejection range on the tail side may be set to a range of 10 mm or more and 20 mm or less from the reference position. Also, the rejection range on the head side from the reference position may be set to a range of 8 mm or more and 12 mm or less from the peak position (reference position).

[0030] In addition, in the n-type counter-doped crystal grown by the CZ method as described above, a certain amount of oxygen has dissolved due to its manufacturing method, and a part of it becomes a thermal donor and acts as an n-type dopant, thereby reducing the resistivity. Therefore, the counter-doped crystal grown by the CZ method is subjected to heat treatment (donor killer) to obtain the original resistivity of the crystal determined by the main dopant. In addition, thermal donors are generated in large numbers when the history around 450 °C during crystal growth is long and the oxygen concentration is high.

[0031] In addition, the peak formed by the resistivity variation on one side of the single crystal ingot 1 changes depending on the dopant concentration, the amount of thermal donors, and the variation amount of thermal donors at the measurement position. However, in this embodiment, regardless of the level of the oxygen concentration, when the resistivity is 10 Ωcm or more, a peak with a resistivity variation of 10% or more is formed near the crystal length position where the sub-dopant is introduced.

[0032] The resistivity variation due to thermal donors on one side of the single crystal ingot 1 can be expressed by the following ratio formula. Resistivity variation due to thermal donors = (Resistivity corresponding to the variation amount of thermal donors) / (Resistivity of the single crystal ingot + Resistivity of thermal donors)

[0033] For example, when the oxygen concentration in the single crystal ingot 1 is high, the resistivity of the side surface is almost determined by the amount of thermal donors. Since the denominator becomes smaller due to the influence of thermal donors, the ratio (resistivity variation) becomes larger. On the other hand, when the oxygen concentration is low, the variation amount of thermal donors becomes smaller (the resistivity of thermal donors increases), so the numerator becomes larger. In this way, by the change in the oxygen concentration causing the denominator or the numerator to vary, a peak with a resistivity 10% or more higher than the stable layer is formed, and the reference position (the position of increased resistivity due to counter-doping) can be detected.

[0034] In addition, in this embodiment, the resistivity is measured by the four-probe method, but it is not limited to this. Any method may be used as long as the resistivity in the crystal length direction on the side surface can be measured in the state of the single crystal ingot.

[0035] <Effect> Thus, in the method for evaluating a single crystal ingot of this embodiment, the time required for resistivity evaluation can be shortened.

Example

[0036] Next, an example of the method for evaluating a silicon single crystal ingot according to the present invention will be described. Note that the present invention is not limited by the following examples.

[0037] <Example 1> A counter-doped crystal of n-type light doping was pulled up by counter-doping with phosphorus as the main dopant and boron as the sub-dopant. At this time, during the growth of the counter-doped crystal under the conditions of crystal rotation of 10 rpm, crucible rotation of 1 rpm, crystal pulling of 1 mm / min, and magnetic field strength of 2000 G, boron as the sub-dopant was introduced at the position of crystal length 500 mm and the position of crystal length 800 mm. In addition, in Example 1, the desired resistivity was 50 Ωcm (phosphorus concentration: about 8.64E13 / cm 3 ) and the desired oxygen concentration was 0.55E18 / cm 3 .

[0038] The counter-doped crystal was pulled up with the above settings, ground on the outer periphery to finish at a diameter of 300 mm, and the head and tail cone portions were cut off to produce a single crystal ingot as shown in FIG. 2.

[0039] Then, the resistivity in the crystal length direction on the side surface of the single crystal ingot was measured by the four-probe method in the state of the single crystal ingot. Specifically, starting from the position 10 mm back to the head side from the position of the secondary dopant input, that is, starting from the positions of 490 mm and 790 mm of the crystal length from the head side, the resistivity of the side surface of the single crystal ingot was measured at 30 positions with a 1 mm pitch in the crystal length direction (see Fig. 2).

[0040] Fig. 3 is a diagram showing the measurement results of the resistivity (relative value) on the side surface of the single crystal ingot. Fig. 3(a) shows the transition of the relative resistivity in the crystal length direction starting from the position of 490 mm of the crystal length from the head side, and Fig. 3(b) shows the transition of the relative resistivity in the crystal length direction starting from the position of 790 mm of the crystal length from the head side. As a result of measuring the resistivity (the above 1 mm pitch, 30 points) on the side surface of the single crystal ingot by the four-probe method, the relative resistivity increased rapidly near the position of the secondary dopant input, and the maximum peak height was 42% at the position of Fig. 3(a) and 21% at the position of Fig. 3(b). Thereafter, the relative resistivity decreased rapidly and stabilized at the desired resistivity.

[0041] And in the vicinity of the crystal length of 500 mm, taking the peak position of the resistivity fluctuation in the crystal length direction of 42% as the reference position, further, the resistivity fluctuation stabilized, and the position where the fluctuation rate of the relative resistivity in the crystal length direction became 1% or less and began to follow segregation was defined as the stable position. As shown in Fig. 4, the range from the position 10 mm to the head side from the reference position to the stable position was defined as the rejection range (16 mm). Also, in the vicinity of the crystal length of 800 mm, taking the peak position of the resistivity fluctuation in the crystal length direction of 21% as the reference position, further, the resistivity fluctuation stabilized, and the position where the fluctuation rate of the relative resistivity in the crystal length direction became 1% or less and began to follow segregation was defined as the stable position. As shown in Fig. 4, the range from the position 10 mm to the head side from the reference position to the stable position was defined as the rejection range (15 mm).

[0042] (Verification 1) Next, the resistivity of the single crystal ingot side surface was measured as described above, the blocks in each of the above reject ranges were processed into wafers at a 1 mm pitch, and wafers within the above reject ranges were obtained. Then, the oxygen concentration and the thermal donor amount of each obtained wafer were calculated, and the influence on the evaluation of the resistivity variation of the single crystal ingot side surface was verified. The oxygen concentration at the resistivity measurement site was 0.55E18 / cm 3 and the thermal donor amount determined from the resistivity before and after the donor killer was approximately 3.9E13 / cm 3 and the variation amount of the thermal donor in the crystal length direction was 3.18E12 / cm 3 It was thus. From such results, it was confirmed that this thermal donor amount is an amount capable of evaluating the peak of the resistivity of the single crystal ingot side surface shown in FIG. 2.

[0043] (Verification 2) Next, from among all the wafers processed in the above Verification 1, wafers in the range from the position of crystal length 490 mm to the position of crystal length 519 mm and wafers in the range from the position of crystal length 790 mm to the position of crystal length 819 mm were obtained, and the resistivity of the center and the outer periphery of each obtained wafer was measured by the four-probe method.

[0044] FIG. 5 is a diagram showing the measurement results of the resistivity (relative value) at the center and outer periphery of the wafer. FIG. 5(a) shows the transition of the relative resistivity by 30 wafers obtained from the range from the position of crystal length 490 mm to the position of crystal length 519 mm, and FIG. 5(b) shows the transition of the relative resistivity by 30 wafers obtained from the range from the position of crystal length 790 mm to the position of crystal length 819 mm. As shown in FIG. 5, in both (a) and (b), at the position where the secondary dopant was introduced, an increase in the relative resistivity (increase in resistivity) occurred at the center of the wafer. Also, in both (a) and (b), on the tail side of several millimeters from the peak position at the center of the wafer, an increase in the relative resistivity (increase in resistivity) occurred at the outer periphery of the wafer. From these results, it was confirmed that the resistivity variation on the side surface of the single crystal ingot and the resistivity variation at the outer periphery of the wafer generally agreed. Also, as shown in FIG. 5, since the peak position at the outer periphery of the wafer is shifted 4 to 6 mm to the tail side from the peak position at the center of the wafer, from this verification, it was confirmed that it is necessary to reject from the above-mentioned "position 10 mm from the reference position to the head side" determined in Example 1.

[0045] (Verification 3) Next, the resistivity in the radial direction of each processed wafer was measured.

[0046] FIG. 6 shows, as an example, the measurement results of the resistivity (relative value) in the radial direction of the wafers at positions a, b, c, and d shown in FIG. 2. FIG. 6(a) shows the change in the relative resistivity in the radial direction at position a, FIG. 6(b) shows the change in the relative resistivity in the radial direction at position b, FIG. 6(c) shows the change in the relative resistivity in the radial direction at position c, and FIG. 6(d) shows the change in the relative resistivity in the radial direction at position d. In Verification 3, the resistivity was measured for each wafer at a pitch of 5 mm in the diameter direction. As a result, for example, the wafer at position a before the co-dopant was incorporated into the crystal had a substantially constant relative resistivity (see FIG. 6(a)) and a substantially constant in-plane resistivity variation RRG (Radial Resistivity Gradient) of 5% or less. Note that RRG is a value obtained by expressing, as a percentage, the difference between the maximum value and the minimum value in a group of resistivity measurements taken at an arbitrary position within a single silicon single crystal substrate plane, divided by the minimum value. Also, for example, the wafer at position b where the co-dopant was introduced had an increase in relative resistivity at its center (see FIG. 6(b)) and a deteriorated RRG (RRG exceeded 5%). Also, for example, the wafer at position c after the co-dopant was introduced had an increase in relative resistivity at its outer periphery (see FIG. 6(c)) and a state where the RRG exceeded 5% continued. Also, for example, the wafer at position d (further on the tail side) after the co-dopant was introduced had no increase in resistivity in the radial direction (see FIG. 6(d)), the relative resistivity was stable, and the RRG was also improved to 5% or less.

[0047] By measuring the resistivity in the radial direction of the processed wafers, it was possible to identify the position where, in the crystal growth direction, after the RRG deteriorated beyond 5%, it became stable again within 5%. That is, from this verification, it was confirmed that it was necessary to reject up to the above-mentioned "stable position" determined in Example 1. Also, by this verification, it was confirmed that for wafers outside the above rejection range, all had an RRG within 5%.

[0048] <Example 2> An n-type lightly doped counter-doped crystal was pulled under the same conditions as in Example 1, except that the desired oxygen concentration was set to 1.20E18 / cm 3

[0049] After that, it was finish-ground on the outer circumference to a diameter of 300 mm, the head and tail cone portions were cut off, and a single crystal ingot as shown in Fig. 2 was produced. Then, in the same manner as in Example 1, the resistivity in the crystal length direction on the side surface was measured by the four-probe method in the state of the single crystal ingot (see Fig. 2).

[0050] As a result, also in the single crystal ingot of Example 2, as in Example 1, the relative resistivity rapidly increased near the positions of the secondary dopant injection at the crystal length of 500 mm and the crystal length of 800 mm, and then the relative resistivity rapidly decreased and stabilized at the desired resistivity.

[0051] Also in Example 2, the range from the position 10 mm from the reference position (the position where the increase rate of the relative resistivity in the crystal length direction is 10% or more) to the head side to the stable position (the position where the fluctuation rate of the relative resistivity in the crystal length direction becomes 1% or less and segregation starts) was defined as the rejection range.

[0052] The single crystal ingot of Example 2 had a desired oxygen concentration of 1.20E18 / cm 3 and the variation amount of the thermal donor was 7.21E12 / cm 3 Therefore, a resistivity variation of 12% or more and the reference position (the crystal length position where the secondary dopant was injected) were detected.

[0053] <Comparative Example> In the comparative example, an n-type lightly doped counter-doped crystal was pulled up under the same conditions as in Example 1, and a single crystal ingot similar to that in Example 1 was produced.

[0054] Then, in the comparative example, the produced single crystal ingot was cut into blocks, and resistivity evaluation (evaluation of whether it is within RRG 5%) was performed using the inspection sample wafer cut out during block production (conventional method).

[0055] However, in this method, since it is not clear at what position from the block end the resistivity increases (whether the RRG exceeds 5%), the cutting positions of the inspection sample wafers were obtained from a crystal length of 490 mm and a crystal length of 510 mm with respect to the secondary dopant input position of 500 mm. After processing each inspection sample wafer and performing resistivity evaluation, the RRG was 3.4% at a crystal length of 490 mm and 7.0% at a crystal length of 510 mm. The RRG results of the cut samples are shown in Table 1.

[0056]

Table 1

[0057] Therefore, resistivity evaluation was performed using inspection sample wafers cut to a certain thickness with a wire saw. After processing the inspection sample obtained at the position of a crystal length of 511 mm and performing resistivity evaluation, since the RRG was 5.2%, again, the inspection sample obtained at the position of a crystal length of 512 mm was processed and resistivity evaluation was performed. As a result, the RRG was 3.9%. The loss of crystal length due to the block and processing was 22 mm.

[0058]

Table 2

[0059] On the other hand, in Example 1 and Example 2, as described above, the high-resistivity layer can be specified from the resistivity distribution on the side surface of the single-crystal ingot, and the rejection range can be determined immediately. Furthermore, after rejection, wafer processing can be performed using only silicon blocks with an RRG within 5%. As a result, compared with the method of cutting inspection sample wafers from a single-crystal ingot and performing resistivity evaluation (conventional method), the yield of wafers was significantly improved (it is possible to improve by about 30% on average). Also, the time required for resistivity evaluation could be significantly shortened compared with the conventional method (it is possible to shorten by about 12 hours on average).

[0060] Incidentally, as another comparative example, there is also a method of pulling up an n-type lightly doped counter-doped crystal under the same conditions as in Example 1 and performing block cutting at the position of the crystal length where the co-dopant is introduced from the log data. However, the position of the co-dopant introduction becomes unclear due to the outer circumference grinding process of the single crystal ingot and the process of cutting out the sample wafer, and there has been a problem that block cutting cannot be performed near the co-dopant introduction position.

[0061] As described above, the present invention is not limited to the above-described embodiments and the above-described examples. The above-described embodiments and the above-described examples are illustrative, and any configuration that has substantially the same configuration as the technical idea described in the claims of the present invention and exhibits the same operational effects is included in the technical scope of the present invention.

Explanation of Reference Numerals

[0062] 1 Single crystal ingot

Claims

1. An evaluation method for a silicon single crystal ingot in which counter doping is performed by adding a secondary dopant during pulling in the pulling of a silicon single crystal by the Czochralski method, comprising: measuring the resistivity in the crystal length direction on the side surface by the four-probe method in the state of the silicon single crystal ingot; determining a rejection range based on the resistivity distribution in the crystal length direction; taking, as a reference position, a position where a sharp increase in resistivity is observed near the position of the secondary dopant input recorded in advance during pulling; setting the rejection range to a predetermined length from the reference position to the head side and the tail side respectively; An evaluation method for a silicon single crystal ingot, characterized by the above.

2. An evaluation method for a silicon single crystal ingot in which counter doping is performed by adding a secondary dopant during pulling in the pulling of a silicon single crystal by the Czochralski method, comprising: measuring the resistivity in the crystal length direction on the side surface by the four-probe method in the state of the silicon single crystal ingot; determining a rejection range based on the resistivity distribution in the crystal length direction; taking, as a reference position, a position where a sharp increase in resistivity is observed near the position of the secondary dopant input recorded in advance during pulling, and further, taking, as a stable position, a position where the resistivity variation is stable on the tail side from the reference position, the variation rate of the resistivity becomes equal to or less than a threshold value, and segregation starts to follow; setting the rejection range to a range from a position at a predetermined length from the reference position to the head side to the stable position; An evaluation method for a silicon single crystal ingot, characterized by the above.

3. The reference position is a peak position where the resistivity is 10% or more higher than the resistivity of the stable layer on the head side where the resistivity variation is stable. The evaluation method for a silicon single crystal ingot according to claim 1 or 2, characterized by the above.

4. The measurement range of the resistivity includes at least the range from the position where the co-dopant is introduced to the position where the resistivity increases and then stabilizes again at the desired resistivity. The method for evaluating a silicon single crystal ingot according to claim 1 or 2, characterized in that.

5. The resistivity of the silicon single crystal to be grown is 10 Ωcm or more. The method for evaluating a silicon single crystal ingot according to claim 1 or 2, characterized in that.

6. The rejection range is 8 mm or more and 12 mm or less on the head side from the reference position, and 10 mm or more and 20 mm or less on the tail side from the reference position. The method for evaluating a silicon single crystal ingot according to claim 1 or 2, characterized in that.

Images