Method for evaluating gettering capacity of silicon wafer and method for manufacturing silicon wafer

Abstract

To propose a method for evaluating the intrinsic gettering capacity in a silicon wafer more precisely.SOLUTION: The method for evaluating the gettering capacity in a silicon wafer uses the product R (the diameter of the sphere of an oxygen precipitated material in the silicon wafer)×δ (the strain around the oxygen precipitated material).SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method for evaluating gettering ability in a silicon wafer and a method for manufacturing a silicon wafer. [Background technology]

[0002] Semiconductor devices are manufactured using semiconductor wafers such as silicon wafers and epitaxial silicon wafers as substrate materials. The manufacturing process of semiconductor devices includes many steps in which metal impurities such as Fe, Cu, and Ni are mixed into the semiconductor wafer. If these metal impurities are present in the device formation region near the surface of the semiconductor wafer, they can cause the characteristics of the semiconductor device to deteriorate and reduce the product yield. Therefore, various gettering technologies have been developed to remove metal impurities from the device formation region.

[0003] These gettering technologies can be roughly categorized into two types. One is the intrinsic gettering (IG) method, which uses oxygen precipitates (commonly known as silicon oxide precipitates, also known as BMD: Bulk Micro Defects), which are crystal defects inside the semiconductor wafer, or dislocations as gettering sites. The other is the extrinsic gettering (EG) method, which forms gettering sites outside the silicon wafer by ion implantation on the back surface of the wafer or by forming polysilicon.

[0004] Regarding the above-mentioned IG method, it has been conventionally believed that the intrinsic gettering ability (hereinafter also simply referred to as "IG ability") of a semiconductor wafer is correlated with the size and density of oxygen precipitates, and the size and density of oxygen precipitates have also been used in the evaluation of the IG ability of a semiconductor wafer. For example, Patent Document 1, Non-Patent Document 1, and Non-Patent Document 2 describe a technique for evaluating the IG ability of a semiconductor wafer based on the diagonal length L of the oxygen precipitates and the density D of the oxygen precipitates.

[0005] Furthermore, Patent Document 2 describes a method for evaluating the internal gettering capability of a silicon substrate by calculating the concentration of contaminating heavy metals remaining inside the silicon substrate after heat treatment from the initial contamination concentration of the contaminating heavy metals, the density of oxygen precipitates, the radius of the oxygen precipitates, and the heat treatment conditions. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2000-68280 A [Patent Document 2] JP 2007-67338 A [Non-patent literature]

[0007] [Non-Patent Document 1] K. Sueoka et al., Internal Gettering for Ni Contamination in Czochralski Silicon Wafers, J. Electrochem. Soc., 147 (2000) 3074. [Non-Patent Document 2] M. Seacrist et al., Determination of minimum oxygen precipitate growth conditions for gettering of copper and nickel, The Electrochemical Society Proceeding, 2002-2 (2002) 638. Summary of the Invention [Problem to be solved by the invention]

[0008] When the present inventors evaluated the IG ability of silicon wafers by the techniques described in Patent Document 1, Non-Patent Document 1, and Non-Patent Document 2, they found that even silicon wafers evaluated to have excellent IG ability may have insufficient IG ability.

[0009] Furthermore, the technology described in Patent Document 2 merely calculates the concentration of contaminating heavy metals remaining inside the silicon substrate after heat treatment, and it is unclear whether the remaining contaminating heavy metals form silicide on the surface of the silicon substrate and affect the device characteristics, and it is unclear whether the IG ability of the silicon substrate can be correctly evaluated.

[0010] The present invention has been made in view of the above problems, and an object of the present invention is to propose a method for evaluating the intrinsic gettering ability of a silicon wafer with higher accuracy. [Means for solving the problem]

[0011] The present invention which solves the above problems is as follows. [1] A method for evaluating gettering ability in a silicon wafer, comprising the steps of: A method for evaluating a gettering ability in a silicon wafer, comprising: evaluating an intrinsic gettering ability in the silicon wafer based on a product R×δ of a sphere-equivalent radius R of an oxygen precipitate contained in the silicon wafer and a distortion δ around the oxygen precipitate.

[0012] [2] The method for evaluating a gettering ability in a silicon wafer according to [1] above, further comprising the step of evaluating an intrinsic gettering ability in the silicon wafer based on whether the product R×δ is 0.36 nm or more.

[0013] [3] The density of the oxygen precipitates is 1×10 8 / cm 3 More than 1×10 11 / cm 3 A method for evaluating a gettering ability of a silicon wafer according to the above [1] or [2], which is as follows:

[0014] [4] The method for evaluating the gettering ability of a silicon wafer according to any one of [1] to [3] above, wherein the sphere-equivalent radius R is determined by a calculation simulation based on the diffusion rate control of oxygen.

[0015] [5] The distortion δ is defined as the spherical equivalent radius of the oxygen precipitate, R p , the interstitial silicon concentration at the SiO2 / Si interface is C I (R p ), the thermal equilibrium concentration of interstitial silicon in the silicon matrix is ​​C I eq , Boltzmann constant k, absolute temperature T, SiO2 molecular volume Ω(=4×10 -23 cm 3 ), f ε =7.3×10 23 eV / cm 3 , α SiO2 and α Si are the thermal expansion coefficients of SiO2 and Si (5.5×10 -7 K -1 , 5.0×10 -6 K -1 ), ΔT (=Tg-300°C, where "Tg" is the temperature during the heat treatment for growing oxygen precipitates, and "300°C" is the minimum temperature at which dislocations can move) as a cooling range, and the gettering ability is calculated from the following formula (A):

number

[0016] [6] A method for producing silicon wafers, comprising: slicing a portion of a single crystal silicon ingot grown by a prescribed method; performing wafer processing or mirror etching with a mixed solution of hydrofluoric acid and nitric acid to obtain a plurality of evaluation samples; evaluating the gettering ability of the obtained evaluation samples by the method according to any one of [1] to [5]; and obtaining silicon wafers by performing wafer processing on the single crystal silicon ingot evaluated to have a prescribed gettering ability. Effect of the Invention

[0017] According to the present invention, the intrinsic gettering ability of a silicon wafer can be evaluated with higher accuracy. [Brief description of the drawings]

[0018] [Figure 1] FIG. 1 is a diagram showing the relationship between the level of a silicon wafer and the density of shallow pits. [Diagram 2] FIG. 1 is a graph showing the relationship between the density of oxygen precipitates and R×δ and the intrinsic gettering ability for the silicon wafer of Example 1. [Diagram 3] FIG. 13 is a graph showing the relationship between the density of oxygen precipitates and R×δ and the intrinsic gettering ability for the silicon wafer of Example 2. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] (Method for evaluating gettering ability in silicon wafers) Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The method for evaluating the gettering ability of a silicon wafer according to the present invention is characterized in that the intrinsic gettering ability of the silicon wafer is evaluated based on the product R×δ of the sphere-equivalent radius R (nm) of an oxygen precipitate contained in the silicon wafer and the distortion δ around the oxygen precipitate.

[0020] The present inventors have intensively studied a method for evaluating the intrinsic gettering ability of silicon wafers with higher accuracy. Oxygen precipitates (SiO2) have a larger volume than the silicon atoms that make up the silicon wafer. Therefore, when oxygen precipitates are formed inside the silicon wafer, distortion occurs around the oxygen precipitates.

[0021] For example, if a silicon wafer is contaminated with Ni, when the silicon wafer is subjected to heat treatment, Ni diffuses into the wafer and migrates to the wafer surface and around oxygen precipitates during the subsequent cooling process to form nickel silicide. The formation of nickel silicide on the wafer surface leads to deterioration of device performance, such as leakage defects.

[0022] During the formation of the above defects, if Ni is present in the region affected by the strain around the oxygen precipitate, the Ni will be retained around the oxygen precipitate and precipitate. Therefore, the inventors of the present invention thought that the gettering ability could be estimated from the size of the oxygen precipitate and the magnitude of the strain around the oxygen precipitate. The inventors of the present invention found that the gettering ability in a silicon wafer can be evaluated with higher accuracy based on the product R×δ of the sphere-equivalent radius R (nm) of the oxygen precipitate and the strain δ around the oxygen precipitate, and thus completed the present invention. The details of the present invention will be described below.

[0023] First, the silicon wafer to be evaluated can be a wafer made by slicing a single crystal ingot grown by the Czochralski (CZ) method with a wire saw or mirror etching with a mixture of hydrofluoric acid and nitric acid. In addition, a given concentration of any dopant can be added to the wafer to obtain a so-called n + Type or p + type, or n - Type or p - The substrate may be a mold. An epitaxial silicon wafer having an epitaxial layer formed on the surface of a bulk wafer may be used. The silicon epitaxial layer can be formed under general conditions by a CVD method. The thickness of the epitaxial layer is preferably within a range of 0.1 to 10 μm, and more preferably within a range of 0.2 to 5 μm.

[0024] The silicon wafer used in the evaluation of the present invention is a silicon wafer in which oxygen precipitates are precipitated as gettering sites. The oxygen precipitates can be precipitated by performing an oxygen precipitate nucleation heat treatment at 600 to 800°C, followed by a nucleation growth heat treatment at 1000°C or higher. For example, heat treatments can be performed in succession in a nitrogen gas atmosphere at 780°C for 3 hours and at 1000°C for 16 hours.

[0025] The density of the oxygen precipitates is 1×10 8 / cm 3 More than 1×10 11 / cm 3 It is preferable that:

[0026] In the present invention, the sphere-equivalent radius R of the oxygen precipitate is used to evaluate the IG ability. The sphere-equivalent radius R of the oxygen precipitate can be determined, for example, by a calculation simulation. Specifically, it can be determined, for example, by the method described in Reference 1 (International Publication No. 2005 / 071144). The method described in Reference 1 will be described below.

[0027] First, the calculation for determining the density and amount of oxygen precipitates will be described. The calculation comprises the following steps 1) to 3). 1) A process for determining the concentration of thermal donors generated in a silicon single crystal from the thermal history between 400℃ and 550℃ that a silicon single crystal ingot undergoes during the cooling process during crystal growth and the initial oxygen concentration. 2) A process of dividing the heat treatment process into multiple time intervals, calculating the nucleation rate for each time interval using a nucleation rate equation that uses the initial oxygen concentration, thermal donor concentration, and temperature, and calculating the density of nuclei generated in each time interval. 3) A process to determine the growth and annihilation of nuclei occurring in each time interval Here, we assume that the oxygen precipitates are silicon-oxygen compounds (SiO2) and are spherical.

[0028] [Regarding process 1)] First, the process of calculating the thermal donor concentration in 1) will be described. The thermal donor concentration of silicon single crystals is strongly correlated with the time required to pass between 450°C and 500°C during the cooling process during crystal growth and with the oxygen concentration, and can be calculated using an empirical formula. However, it is preferable to calculate it using the following calculation method.

[0029] The basic formula used is that proposed by Reference 2 (K. Wada, N. Inoue, Semiconductor Silicon, ed. HR Huff et al., (Electrochem. Soc., Pennington NJ, 1986) p. 778). This formula is expressed as the following formula (1). The following formula (1) shows the relationship between the heat treatment time under constant temperature conditions and the thermal donor concentration, and indicates that the thermal donor concentration TD is expressed as a function of the oxygen concentration C and temperature (absolute temperature) T.

[0030] In the following formula (1), TD: thermal donor concentration, TD eq : Thermal equilibrium concentration of thermal donor, a: coefficient (= 9.2 × 10 -50 ), k: Boltzmann constant, T: absolute temperature, D: oxygen diffusion coefficient, C: oxygen concentration, and t: time.

number

number

[0031] Since the above formula (1) is valid under the condition of a constant temperature, the following device is used to apply this formula under the condition of changing temperature, such as the cooling process during crystal growth.

[0032] When thermal donors are generated with temperature changes, the time t in the above formula (1) must be converted to an effective time for each temperature. Here, consider a case where heat treatment is performed at temperature T1 for time t1 (seconds), and after the thermal donor concentration becomes TD1, heat treatment is performed at temperature T2 for time t2 (seconds). In such a case, consider the case where the amount (concentration) of thermal donors generated at temperature T1, TD1, is generated at temperature T2, and the time t 12 From the above formula (1), time t 12 is expressed as the following equation (3).

number

[0033] Then, when a heat treatment is added at a temperature T2 for a time t (seconds), the amount (concentration) of thermal donors generated TD is considered to be expressed by the following equation (4).

number

[0034] The temperature of the silicon single crystal continues to drop during crystal growth. For this reason, the above calculation must be repeated. n Thermal donor generation (concentration) up to TD n at temperature T n The equivalent time t n Calculate the equivalent time t n The thermal donor generation is most active between 450°C and 500°C, but it is preferable to use a wider temperature range for the calculation, 400°C to 550°C.

[0035] By using the above calculation method, the thermal donor concentration can be predicted with high accuracy.

[0036] [Regarding process 2)] Next, the process of dividing the heat treatment process in 2) into a number of time intervals of a given interval, for example, 5 seconds, calculating the nucleation rate for each time interval using a nucleation rate equation using the initial oxygen concentration, thermal donor concentration, and temperature, and calculating the density of nuclei generated in each time interval will be described. The nucleation rate at each temperature is expressed as a function of the oxygen concentration and thermal donor concentration, as shown in the following equation (5).

[0037] In the following formula (5), I(T,C,TD): nucleation rate (cm -3 S -1 ), T: temperature, a(T): coefficient determined by temperature, C: oxygen concentration (×10 17 cm -3 ), TD: thermal donor concentration (×10 15 cm -3 ) It is.

number

[0038] The thermal donor concentration TD used in the above formula (5) is the as-grown concentration that occurs during the crystal growth process. Although their electrical properties disappear during the temperature rise in heat treatment, their physical effects remain, so the as-grown concentration is used. Furthermore, when a heat treatment process is performed at 400°C to 550°C, the thermal donors that occur during the heat treatment are also included. Thus, the nucleation rate of oxygen precipitates is expressed by the above formula (5), and the effect of thermal donors is applied to the computational prediction.

[0039] The density N(t') of nuclei generated during the time Δt from time t' is given by the following equation (6).

number

[0040] The radius of the generated nucleus is the critical nucleus radius R, given by the following equation (7): cri In the following formula (7), R cri: Critical nucleus radius, V: SiO2 molecular volume, σ: Si-SiO2 interface energy, C: oxygen concentration, C eq is the thermal equilibrium concentration of oxygen.

number

[0041] Data on nuclei generated for each time interval at the density and size shown in equations (6) and (7) above is recorded in memory, and the growth and annihilation of each nucleus is calculated.

[0042] [Regarding process 3)] Next, the process of calculating the growth and annihilation of nuclei generated in each time interval of 3) will be explained. The growth rate of nuclei generated from time t' until Δt has elapsed at time t is expressed by the following formula (8). In the following formula (8), R(t', t) is the radius of the nuclei generated from time t' until Δt has elapsed at time t, and Ci is the equilibrium oxygen concentration at the interface of a spherical particle of radius R.

number

number

[0043] R(t',t) can be found by calculation using the above formula (8). The change in oxygen concentration in a silicon single crystal is shown by the following formula (10), so the amount of oxygen precipitates can be found by calculation. Since the oxygen concentration decreases by the amount of precipitated oxygen, the remaining oxygen concentration can be found as the difference between the initial oxygen concentration and the amount of oxygen precipitates found using the following formula (10). The calculation process of finding the remaining oxygen concentration for each time interval, finding the nucleation rate of oxygen precipitates, and finding the growth of oxygen precipitates is repeated.

number

[0044] By applying the above calculations to the heat treatment process, the density and size distribution (i.e., spherical equivalent diameter distribution) of oxygen precipitates and the amount of oxygen precipitates can be obtained.

[0045] The strain δ around the oxygen precipitate can be calculated from the following formula (11) described in Reference 3 (WJ Taylor, U. Goesele and TY Tan, J. Appl. Phys., 72 (1992) 2192.):

number

[0046] In the present invention, the IG ability of a silicon wafer is evaluated based on the product R×δ of the sphere-equivalent radius R of the oxygen precipitate and the distortion δ around the oxygen precipitate. The larger the value of the product R×δ, the higher the IG ability can be evaluated. Since the required IG ability varies depending on the specifications of the silicon wafer, the threshold value of R×δ used to evaluate whether or not the desired IG ability is present can be appropriately set according to the specifications. However, as shown in the examples described later, as a result of the study by the present inventors, when R×δ≧0.36 nm is satisfied, excellent gettering ability can be exhibited even in the case of severe heavy metal contamination. Therefore, it is preferable to evaluate the IG ability based on whether R×δ is 0.36 nm or more.

[0047] (Silicon wafer manufacturing method) The method for producing a silicon wafer according to the present invention is characterized in that a portion of a single crystal silicon ingot grown by a prescribed method is sliced, and a plurality of evaluation samples are obtained by performing wafer processing or mirror etching with a mixed solution of hydrofluoric acid and nitric acid, and the gettering ability of each of the evaluation samples is evaluated by the above-mentioned method for evaluating the gettering ability in a silicon wafer according to the present invention, and the single crystal silicon ingot evaluated to have the prescribed gettering ability is subjected to wafer processing to obtain a silicon wafer.

[0048] As described above, in the method for evaluating the gettering ability of a silicon wafer according to the present invention, the IG ability of the silicon wafer is evaluated based on the product R×δ of the sphere-equivalent radius R of the oxygen precipitate and the distortion δ around the oxygen precipitate. As a result, the IG ability of the silicon wafer can be evaluated with higher accuracy than in the past.

[0049] Therefore, in the method for producing a silicon wafer according to the present invention, a single crystal silicon ingot grown by a predetermined method is subjected to wafer processing, and the gettering capabilities of the obtained silicon wafers are evaluated by the above-mentioned method for evaluating the gettering capability of a silicon wafer according to the present invention, and silicon wafers evaluated to have a predetermined gettering capability are obtained. This makes it possible to produce silicon wafers with high IG ability.

[0050] The predetermined method for growing a single crystal silicon ingot can be the CZ method. EXAMPLES

[0051] Examples of the present invention will be described below, but the present invention is not limited to these examples.

[0052] Example 1 First, we prepared a total of 29 silicon wafers of the following two types. (1) Nitrogen-doped epitaxial silicon wafer (N-doped EW) Silicon wafers Diameter: 300mm Oxygen concentration: 11.0 x 10 17 ~12.5×10 17 atoms / cm 3 (ASTM F121-1979) Dopant: Boron Resistivity: 8~12Ω cm Silicon epitaxial layer Film thickness: 4μm (2) Polished wafer (PW) Diameter: 300mm Oxygen concentration: 17.0 x 10 17 atoms / cm 3 (ASTM F121-1979) Dopant: Boron Resistivity: 8~12Ω cm

[0053] Wafer (1) was subjected to an oxygen precipitation nucleation heat treatment in a nitrogen gas atmosphere at (A) 900°C for 90 minutes or 5 hours, or (B) 1000°C for 1 hour, 6 hours, or 16 hours, or (C) was not subjected to an oxygen precipitation nucleation heat treatment.

[0054] In addition, for wafer (2), the precipitation nuclei were once annihilated by heat treatment in an argon gas atmosphere at 1150°C for 1 minute, and then the wafer was subjected to oxygen precipitation nuclei generation heat treatments in a nitrogen gas atmosphere, namely (D) 3 hours at 650°C and 90 minutes or 8 hours at 800°C, or (E) 4 hours at 650°C and 90 minutes or 8 hours at 800°C.

[0055] Next, the wafers (1) and (2) were subjected to a heat treatment for oxygen precipitate nucleus growth at 1000° C. for 16 hours.

[0056] After that, the wafers (1) and (2) were subjected to light etching for 2 minutes, and the cross sections of the wafers were observed under an optical microscope to measure the BMD density. In addition, after the above-mentioned oxygen precipitate nucleation heat treatment was performed on the wafers (1) and (2), the above-mentioned oxygen precipitate nucleation growth heat treatment was not performed, but instead, the wafers (1) and (2) were subjected to a heat treatment at 900°C for 5 minutes, which is performed following the forced contamination treatment in the evaluation of IG ability described later, and the sphere-equivalent radius R of the BMD in the wafers (1) and (2) was calculated based on the method described in Reference 1 above. Furthermore, the distortion generated around one oxygen precipitate was calculated based on the above formula (11) described in Reference 3. The density of the oxygen precipitates, the sphere-equivalent radius R of the oxygen precipitates, the distortion δ around the oxygen precipitates, and the value of R×δ measured for each wafer, as well as the heat treatment conditions for the wafers, are shown in Table 1.

[0057] [Table 1]

[0058] On the other hand, the IG ability was evaluated as follows. First, the surface after the precipitation heat treatment was made hydrophilic by using the technique described in JP-A-2-156636, and a solution prepared by diluting a specified Ni atomic absorption standard solution with a solvent was applied to the surface and spin-dried, whereby the surface contamination concentration was reduced to 7.4 × 10 11 ~1.5×10 12 atoms / cm 2 The wafer was then heat-treated at 900°C for 5 minutes to diffuse Ni into the wafer, after which it was lightly etched for 2 minutes. The wafer surface after etching was then observed with an optical microscope to measure the density of shallow pits in the Ni silicide. The shallow pit density was 1×10 3 atoms / cm 2 Wafers with a shallow pit density of 1×10 or less were evaluated as being getterable. 3 atoms / cm 2 Wafers with a level above this were evaluated as not capable of gettering. The evaluation results and the measured shallow pit densities are shown in Table 1. The relationship between the wafer level and the measured shallow pit density is shown in Figure 1. In Figure 1, the shallow pit densities for levels 12 to 29 are in the range of 1×10 3 atoms / cm 2 It shows that:

[0059] Figure 2 shows the relationship between the density of oxygen precipitates and R×δ and the intrinsic gettering ability for the above 29 levels of silicon wafers. In Figure 2, the dashed line indicates the boundary of R×δ=0.36 nm, and all silicon wafers of a level evaluated as capable of gettering satisfied R×δ≧0.36 nm, while all silicon wafers of a level evaluated as not capable of gettering did not satisfy R×δ≧0.36 nm.

[0060] Example 2 The intrinsic gettering ability was evaluated when the contamination concentration was higher than that in Example 1. However, a total of 16 wafers (1) and (2) were prepared. In addition, Ni contamination treatment was performed when the surface contamination concentration was 1.1 to 2.0 × 10 13 atoms / cm 2 The intrinsic gettering ability was evaluated in the same manner as in Example 1, with the shallow pit density set to 1×10 3 atoms / cm 2 This was based on whether or not:

[0061] 3 shows the relationship between the density of oxygen precipitates, R×δ, and the intrinsic gettering ability for the silicon wafer of Example 2. As is clear from FIG. 3, similar to Example 1, all of the silicon wafers that were evaluated as being capable of gettering satisfied R×δ≧0.36 nm, while all of the silicon wafers that were evaluated as being unable to be gettered did not satisfy R×δ≧0.36 nm. Thus, it was confirmed that the gettering ability could be evaluated without problems even when there was Ni contamination at a high concentration of the 13th power level. [Industrial Applicability]

[0062] INDUSTRIAL APPLICABILITY The present invention is useful in the semiconductor industry since it enables the intrinsic gettering ability of a silicon wafer to be evaluated with higher accuracy.

Claims

1. A method for evaluating a gettering ability in a silicon wafer, comprising the steps of: A method for evaluating a gettering ability in a silicon wafer, comprising: evaluating an intrinsic gettering ability in the silicon wafer based on a product R×δ of a sphere-equivalent radius R of an oxygen precipitate contained in the silicon wafer and a distortion δ around the oxygen precipitate.

2. 2. The method for evaluating the gettering ability of a silicon wafer according to claim 1, further comprising the step of evaluating the intrinsic gettering ability of the silicon wafer based on whether the product R×δ is 0.36 nm or more.

3. The density of the oxygen precipitates is 1×10 ８ / cm ３ Above 1 x 10 １１ / cm ３ 3. The method for evaluating a gettering ability of a silicon wafer according to claim 1, wherein the gettering ability is:

4. 3. The method for evaluating the gettering ability of a silicon wafer according to claim 1, wherein the sphere-equivalent radius R is determined by a calculation simulation based on the diffusion rate limit of oxygen.

5. The distortion δ is defined as the radius of the oxygen precipitate equivalent to a sphere R ｐ , SiO ２ The interstitial silicon concentration at the / Si interface is C Ｉ (R ｐ ), the thermal equilibrium concentration of interstitial silicon in the silicon matrix is ​​C Ｉ ｅｑ , Boltzmann constant k, absolute temperature T, SiO ２ The molecular volume is Ω (= 4 × 10 －２３ cm ３ ), f ε = 7.3 x 10 ２３ eV / cm ３ , α ＳｉＯ２ and α Ｓｉ Each of them is SiO ２ , the thermal expansion coefficient of Si (5.5×10 －７ K －１ , 5.0×10 －６ K －１ 3. The method for evaluating the gettering ability of a silicon wafer according to claim 1, wherein the gettering ability is determined from the following formula (A) with a cooling range of ΔT (=Tg-300°C, where "Tg" is the temperature during the heat treatment for growing oxygen precipitates, and "300°C" is the minimum temperature at which dislocations can move). [0010]

6. 3. A method for producing silicon wafers, comprising: slicing a portion of a single crystal silicon ingot grown by a predetermined method; performing wafer processing or mirror etching with a mixed solution of hydrofluoric acid and nitric acid to obtain a plurality of evaluation samples; evaluating the gettering ability of the obtained evaluation samples by the method according to claim 1 or 2; and obtaining silicon wafers by performing wafer processing on the single crystal silicon ingot evaluated to have a predetermined gettering ability.

Images