JP6645545B1 - Method for evaluating carbon concentration of silicon sample, method for evaluating silicon wafer manufacturing process, method for manufacturing silicon wafer, and method for manufacturing silicon single crystal ingot - Google Patents

Abstract

A new method for evaluating the carbon concentration of a silicon sample is provided. Introducing a hydrogen atom into a silicon sample to be evaluated, subjecting the silicon sample to which the hydrogen atom has been introduced to evaluation by an evaluation method for evaluating a trap level in a band gap of silicon, and Among the evaluation results obtained by the above evaluation, based on the evaluation result regarding the density of at least one trap level selected from the group consisting of Ec-0.10 eV, Ec-0.13 eV and Ec-0.15 eV, Including the evaluation of the carbon concentration of the silicon sample to be evaluated, the introduction of the hydrogen atoms is performed by bringing the silicon sample to be evaluated into contact with a solution, and the solution has a molar ratio (HNO3 / (HNO3 + HF)) of 0.72. Method for evaluating carbon concentration of silicon sample which is hydrofluoric nitric acid in the range of not less than 0.82 or not more than 0.87 and not more than 0.91[Selection diagram] None

Description

  The present invention relates to a method for evaluating a carbon concentration of a silicon sample, a method for evaluating a silicon wafer manufacturing process, a method for manufacturing a silicon wafer, and a method for manufacturing a silicon single crystal ingot.

  In recent years, it has been studied to evaluate the carbon concentration of a silicon sample (for example, see Patent Document 1).

JP-A-2017-191800

  It is desired for a silicon wafer used as a semiconductor substrate to reduce impurity contamination that causes a decrease in device characteristics. In recent years, attention has been paid to carbon as an impurity contained in a silicon wafer, and reduction of carbon contamination of the silicon wafer has been studied.

  To reduce carbon contamination, evaluate the carbon concentration of the silicon sample, and based on the evaluation results, reduce the carbon mixed in the silicon wafer manufacturing process and the silicon single crystal ingot manufacturing process that cuts the silicon wafer in the manufacturing process. It is desirable to manage it. Finding a new method for evaluating the carbon concentration of a silicon sample is useful in performing such a process control.

  An object of one embodiment of the present invention is to provide a new method for evaluating the carbon concentration of a silicon sample.

One embodiment of the present invention provides
Introducing hydrogen atoms into the silicon sample to be evaluated,
The evaluation target silicon sample into which the hydrogen atoms have been introduced is subjected to an evaluation by an evaluation method for evaluating a trap level in a silicon band gap, and Ec (conduction band) is included in the evaluation results obtained by the evaluation. Energy at the bottom of the sample) -0.10 eV, Ec-0.13 eV and Ec-0.15 eV, based on the evaluation result of the density of at least one trap level selected from the group consisting of: To evaluate,
Including
The introduction of the hydrogen atoms is performed by bringing the silicon sample to be evaluated into contact with a solution, and the solution has a molar ratio (HNO ₃ / (HNO ₃ + HF)) in the range of 0.72 to 0.82 or 0. A method for evaluating the carbon concentration of a silicon sample, which is hydrofluoric nitric acid in the range of 0.87 to 0.91 (hereinafter also referred to as “carbon concentration evaluation method”);
About.

  In one embodiment, in the carbon concentration evaluation method, the evaluation target silicon sample into which the hydrogen atoms have been introduced can be subjected to the evaluation without performing the electron beam irradiation treatment.

  In one aspect, the evaluation of the carbon concentration of the silicon sample to be evaluated can be performed based on the evaluation result regarding the density of trap levels of Ec-0.15 eV among the evaluation results obtained by the above evaluation.

  In one aspect, the evaluation method can be a DLTS method (Deep-Level Transient Spectroscopy).

One embodiment of the present invention provides
Evaluating the carbon concentration of the silicon wafer manufactured in the silicon wafer manufacturing process to be evaluated by the above carbon concentration evaluation method, and evaluating the degree of carbon contamination in the silicon wafer manufacturing process to be evaluated based on the result of the above evaluation ,
Evaluation method of silicon wafer manufacturing process including (hereinafter also referred to as "manufacturing process evaluation method"),
About.

One embodiment of the present invention provides
Evaluating the silicon wafer manufacturing process by the above manufacturing process evaluation method, and, in the silicon wafer manufacturing process where the degree of carbon contamination is determined to be an allowable level as a result of the above evaluation, or as a result of the above evaluation, the degree of carbon contamination After performing a carbon contamination reduction process on a silicon wafer manufacturing process that is determined to exceed the allowable level, in this silicon wafer manufacturing process, manufacturing a silicon wafer,
A method for producing a silicon wafer, comprising:
About.

One embodiment of the present invention provides
Growing silicon single crystal ingots,
The carbon concentration of the silicon sample cut from the silicon single crystal ingot is evaluated by the carbon concentration evaluation method,
Based on the results of the above evaluation, determine the production conditions of the silicon single crystal ingot, and,
Growing a silicon single crystal ingot under determined manufacturing conditions,
A method for producing a silicon single crystal ingot containing
About.

  According to one embodiment of the present invention, a new method for evaluating the carbon concentration of a silicon sample can be provided.

It is explanatory drawing which shows the structure of the silicon single crystal pulling apparatus used in the Example. It is a graph which shows the trap level density calculated | required about the sample wet-processed using the hydrofluoric-nitric-acid of each molar ratio obtained by the Example and the comparative example. It is a graph which shows the trap level density calculated | required about the sample wet-processed using the hydrofluoric-nitric-acid of each molar ratio obtained by the Example and the comparative example.

[Method for evaluating carbon concentration of silicon sample]
In one embodiment of the present invention, a hydrogen atom is introduced into a silicon sample to be evaluated, and the silicon sample to be evaluated into which the hydrogen atom has been introduced is subjected to evaluation by an evaluation method for evaluating a trap level in a band gap of silicon. And among the evaluation results obtained by the above evaluation, the evaluation result regarding the density of at least one trap level selected from the group consisting of Ec-0.10 eV, Ec-0.13 eV, and Ec-0.15 eV And estimating the carbon concentration of the silicon sample to be evaluated based on the following conditions. The introduction of the hydrogen atoms is performed by bringing the silicon sample to be evaluated into contact with a solution, and the solution has a molar ratio of (HNO ₃ / (HNO ₃ + HF) )) Is the concentration of carbon in a silicon sample whose hydrofluoric nitric acid is in the range of 0.72 or more and 0.82 or less or 0.87 or more and 0.91 or less. On value method.

By the introduction of hydrogen atoms performed in the carbon concentration evaluation method, the trap level of Ec can be formed in the band gap of silicon. In this way, it is possible to obtain an evaluation result regarding the density of trap levels of Ec. As an example of such an evaluation result, a peak intensity (DLTS signal intensity) obtained by the evaluation by the DLTS method can be cited. Details of this point will be described later.
Regarding the trap level, the trap level of Ec in the band gap of silicon after the introduction of hydrogen atoms is a carbon-related level, and the density of this trap level correlates with the carbon concentration of the silicon sample. Therefore, the evaluation result regarding the trap level density of Ec obtained by the evaluation performed after the introduction of the hydrogen atom, that is, the evaluation result correlated with the trap level density correlates with the carbon concentration of the silicon sample. Furthermore, as a result of earnest studies by the present inventors, when hydrogen atoms are introduced into the silicon sample to be evaluated by contact with hydrofluoric nitric acid, the above-mentioned ratio is determined by the mixing ratio of HNO ₃ and HF, which are acid components in hydrofluoric nitric acid. The value of the density of the trap level of Ec evaluated by the evaluation method changes, and the density of the trap level is determined by using hydrofluoric nitric acid containing HNO ₃ and HF at a molar ratio in the range described above. Has been newly found to be possible. This is because the diffusion of hydrogen atoms in the silicon sample and the etching reaction of the silicon sample surface occur in parallel when the silicon sample is brought into contact with hydrofluoric nitric acid. Is presumed to be changed by the mixing ratio of HNO ₃ and HF of hydrofluoric nitric acid. As the density of trap levels increases, for example, in the DLTS method, the value of the measured DLTS signal intensity increases. For example, if a value of the trap level density can be obtained as a higher density value for a silicon sample having a certain carbon concentration, even a trace amount of carbon can be detected and evaluated with high sensitivity. That is, it is considered that the introduction of hydrogen atoms into the silicon sample to be evaluated using hydrofluoric-nitric acid containing HNO ₃ and HF in the above molar ratio contributes to improvement in sensitivity of carbon concentration evaluation.
Hereinafter, the carbon concentration evaluation method will be described in more detail.

<Evaluation silicon sample>
The silicon sample to be evaluated by the carbon concentration evaluation method can be, for example, a silicon sample cut out from a silicon single crystal ingot. For example, a sample obtained by further cutting a part of a sample cut into a wafer shape from a silicon single crystal ingot can be subjected to evaluation. Further, the silicon sample to be evaluated can be a silicon sample cut out from various silicon wafers (for example, a polished wafer, an epitaxial wafer, etc.) used as a semiconductor substrate. The silicon wafer may be a silicon wafer that has been subjected to various processings (for example, polishing, etching, cleaning, etc.) that are usually performed on the silicon wafer. The silicon sample may be n-type silicon or p-type silicon. The resistivity of the silicon sample can be, for example, about 1 to 1000 Ωcm, but is not particularly limited.

The concentration of interstitial oxygen Oi of the silicon sample to be evaluated (hereinafter, referred to as “oxygen concentration”) is not particularly limited. In one embodiment, the oxygen concentration of the silicon sample to be evaluated is, for example, 1.0 × 10 ¹⁷ atoms / cm ³ or more (for example, 1.0 × 10 ^{17 to} 27.5 × 10 ¹⁷ atoms / cm ³ ). it can. The oxygen concentration here is a value measured by the FT-IR method (Fourier Transform Infrared Spectroscopy). For example, a silicon sample derived from a silicon single crystal grown by the Czochralski method (CZ method) usually contains oxygen. On the other hand, as described in Patent Document 1 (Japanese Patent Application Laid-Open No. 2017-191800), in the luminescence method conventionally proposed as a method for evaluating the carbon concentration of a silicon sample, the quantified carbon concentration depends on the oxygen concentration. Would. This is because the conventionally proposed luminescence method requires electron beam irradiation. Therefore, in the luminescence method, the accuracy of the carbon concentration evaluation tends to decrease as the silicon sample has a higher oxygen concentration. On the other hand, by introducing hydrogen atoms, the above-mentioned carbon-related levels can be formed in an activated state without performing electron beam irradiation. As a result, the carbon concentration can be evaluated without depending on the oxygen concentration. Thus, the carbon concentration of a silicon sample having a relatively high oxygen concentration, for example, a silicon sample having an oxygen concentration in the above range can be evaluated with high accuracy. In the present invention and in the present specification, "do not perform electron beam irradiation treatment" means not performing a process of actively irradiating an electron beam to a silicon sample, and is inevitable under sunlight, lighting, or the like. It is assumed that the electron beam irradiation occurring in the above is allowed. An electron beam is a flow of electrons obtained by applying an accelerating voltage to electrons. The electron beam irradiation treatment has problems in that the lead time is long, large-scale equipment is required, the cost is increased, and in addition to the electron beam irradiation step, the production of a protective oxide film and the like are required, and the number of steps is increased. Therefore, it is preferable that the carbon concentration of the silicon sample can be evaluated without performing the electron beam irradiation treatment. However, the oxygen concentration of the silicon sample to be evaluated by the above carbon concentration evaluation method is not limited to the range exemplified above. In one embodiment of the carbon concentration evaluation method, electron beam irradiation can be performed by a known method.

<Introduction of hydrogen atoms into the silicon sample to be evaluated>
Hydrogen atoms are introduced into the silicon sample to be evaluated. By introducing a hydrogen atom, a trap level of Ec, which is a carbon-related level, can be formed. In the carbon density evaluation method, the introduction of hydrogen atoms is carried out by contact with a solution (wet processing), mixed acid of hydrofluoric nitric acid (HNO ₃ and HF containing a HNO ₃ and HF in a molar ratio of the above range as the solution Aqueous solution) is used. For example, when nitric acid (aqueous nitric acid solution) and hydrofluoric acid (aqueous hydrofluoric acid solution) are mixed to prepare hydrofluoric nitric acid, the molar ratio is determined from the number of moles calculated as follows.
The number of moles of HNO ₃ = liquid amount of nitric acid × density × concentration / molecular weight The number of moles of HF = liquid amount of hydrofluoric acid × density × concentration / molecular weight The unit of the liquid amount is, for example, mL, and the unit of the density is, for example, g / mL. It is. Hydrofluoric acid can be prepared, for example, by mixing nitric acid with an HNO ₃ concentration of 69% by mass and hydrofluoric acid with an HF concentration of 50% by mass. The molar ratio can be adjusted by the mixing ratio of nitric acid and hydrofluoric acid.

The molar ratio of hydrofluoric / nitric acid (HNO ₃ / (HNO ₃ + HF)) used for introducing hydrogen atoms into the silicon sample to be evaluated is in the range of 0.72 or more and 0.82 or less or 0.87 or more and 0.91 or less. Range. A high ratio of HF in hydrofluoric nitric acid (in other words, a low ratio of HNO ₃ ) may cause a decrease in the reproducibility of the evaluation result due to a decrease in the surface smoothness of the silicon sample to be evaluated. On the other hand, if the molar ratio is 0.72 or more, it is possible to suppress a decrease in surface smoothness due to contact with hydrofluoric nitric acid. When the molar ratio is in the range of 0.72 or more and 0.82 or less or 0.87 or more and 0.91 or less, the density of trap states evaluated by the above evaluation method can be increased. From the viewpoint of further increasing the density of the trap states, the molar ratio is preferably 0.80 or less, more preferably 0.78 or less, in the range of 0.72 or more and 0.82 or less. More preferably, it is 0.76 or less. Further, from the viewpoint of further increasing the trap level density, the molar ratio is preferably 0.90 or less, more preferably 0.89 or less, in the range of 0.87 or more and 0.91 or less. .

  For example, by immersing all or a part of the silicon sample to be evaluated in hydrofluoric nitric acid, the silicon sample to be evaluated can be brought into contact with the above hydrofluoric nitric acid. When a part of the silicon sample to be evaluated is immersed in hydrofluoric nitric acid, it is preferable to immerse at least the surface to be evaluated in hydrofluoric nitric acid in the evaluation by the above-described evaluation method. The contact time between the silicon sample to be evaluated and the above-mentioned hydrofluoric-nitric acid can be, for example, 1 to 10 minutes. Further, the liquid temperature of hydrofluoric nitric acid may not be controlled, and may be controlled (heating or cooling). In one embodiment, the hydrofluoric nitric acid can be brought into contact with the silicon sample to be evaluated at room temperature without controlling the liquid temperature of the hydrofluoric nitric acid. After the contact with the hydrofluoric-nitric acid, the silicon sample to be evaluated may be subjected to post-treatment such as washing with water and drying if necessary. The silicon sample to be evaluated into which hydrogen atoms have been introduced by contact with the hydrofluoric nitric acid is subjected to evaluation by an evaluation method for evaluating a trap level in a silicon band gap.

<Evaluation of silicon sample with hydrogen atoms>
In the carbon concentration evaluation method, a trap level of Ec-0.10 eV, Ec-0.13 eV, or Ec-0.15 eV is used as the carbon-related level. It is considered that trap levels of Ec-0.10 eV, Ec-0.13 eV and Ec-0.15 eV are formed in an activated state which can be detected by various evaluation methods by introducing hydrogen atoms. Thus, the carbon concentration can be evaluated based on the density of the trap levels (carbon-related levels). Further, by using hydrofluoric nitric acid containing HNO ₃ and HF at a molar ratio in the range described above as a solution used for the wet treatment for introducing hydrogen atoms, the density of the trap levels can be increased. The evaluation of the trap level density can be performed by various evaluation methods capable of evaluating the trap level in the band gap of silicon. Examples of such an evaluation method include a DLTS method, a lifetime method, an ICTS method (Isothermal Capacitance Transient Spectroscopy), a low-temperature photoluminescence (PL) method, a cathodoluminescence (CL) method, and the like. In the evaluation of carbon concentration by the conventional PL method and CL method (luminescence method), electron beam irradiation treatment was indispensable. On the other hand, according to the carbon concentration evaluation method, since the trap level of Ec is formed in an activated state by the introduction of hydrogen atoms, the density of the trap level can be increased without performing the electron beam irradiation treatment. , It is possible to evaluate the carbon concentration. Known techniques can be applied without any limitation to the measurement method using various evaluation methods.

  For example, the DLTS method is a preferable evaluation method from the viewpoint of enabling more sensitive carbon quantification. In the DLTS method, usually, a diode (sample element) manufactured by forming a semiconductor junction (Schottky junction or pn junction) and an ohmic layer on a measurement sample obtained by cutting out a part of a silicon sample to be evaluated is measured ( DLTS measurement). Generally, it is preferable that the surface of a sample subjected to DLTS measurement has high smoothness. Therefore, the silicon sample to be evaluated before cutting out the sample for measurement or the sample for measurement cut out from the silicon sample to be evaluated can be arbitrarily subjected to etching, polishing, or the like for improving the surface smoothness. The etching is preferably mirror etching. The polishing preferably includes a mirror polishing. For example, when the silicon sample to be evaluated is a silicon single crystal ingot or a part of an ingot, it is preferable to manufacture a sample element after polishing a measurement sample cut out from the silicon sample to be evaluated, and to perform a sample after mirror polishing. It is more preferable to manufacture an element. As the polishing process, a known polishing process applied to a silicon wafer such as a mirror polishing process can be performed. On the other hand, a silicon wafer is usually obtained through polishing such as mirror polishing. Therefore, when the silicon sample to be evaluated is a silicon wafer, the surface of the measurement sample cut from the silicon wafer usually has high smoothness without polishing. When the DLTS method is used as the evaluation method, the DLTS spectrum obtained as the sum of the respective peaks obtained by the DLTS method is subjected to fitting processing by a known method to obtain Ec-0.10 eV, Ec-0.13 eV or Ec-0. The DLTS spectrum of the trap level of 15 eV can be separated. For example, in the DLTS measurement at a frequency of 250 Hz, the trap level density of Ec-0.10 eV is a peak near 76 K, the trap level density of Ec-0.13 eV is a peak near 87 K, and the trap level of Ec-0.15 eV. As for the potential density, the carbon concentration can be determined based on the peak intensity (DLTS signal intensity) of the peak near 101K. The peak used to determine the carbon concentration is at least one of the above three peaks, and two or three peaks may be used. Normally, it can be determined that the higher the peak intensity value, the higher the carbon concentration. From the viewpoint of performing more accurate carbon concentration evaluation, it is preferable to obtain the carbon concentration of the silicon sample to be evaluated based on the evaluation results at Ec-0.13 eV and / or Ec-0.15 eV.

  DLTS measurement is usually performed by the following method. A semiconductor junction (Schottky junction or pn junction) is formed on one surface of a silicon sample, and an ohmic layer is formed on the other surface to produce a diode (sample element). The transient response of the capacitance (capacitance) of the sample element is measured by periodically applying a voltage while performing a temperature sweep. The application of the voltage is usually performed by alternately and periodically applying a reverse voltage for forming a depletion layer and a pulse voltage for filling a trap level in the depletion layer with carriers. The preferred position and width of the depletion layer formation region depend on the resistivity of the silicon sample. The depletion layer can be formed, for example, in a region of about 1 μm to 60 μm in depth from the surface of the silicon sample to be evaluated, in a width of about 1 to 50 μm, and preferably in a width of about 1 to 10 μm. . On the other hand, the thickness of the silicon sample to be evaluated can be, for example, about 100 to 1000 μm. However, it is not limited to this range. The position (measurement depth) of the measurement region can be controlled by a reverse voltage applied to form a depletion layer. Further, the width of the formed depletion layer can also be controlled by the reverse voltage. By plotting the DLTS signal against temperature, a DLTS spectrum can be obtained. By subjecting the DLTS spectrum obtained as the sum of each peak detected by the DLTS measurement to fitting processing by a known method, the DLTS spectrum of each trap level can be separated and the peak can be detected.

<Evaluation of carbon concentration>
Whichever method is used as the evaluation method, the carbon concentration based on the evaluation result regarding the density of at least one trap level selected from the group consisting of Ec-0.10 eV, Ec-0.13 eV, and Ec-0.15 eV Can be evaluated using a calibration curve or without a calibration curve. When the calibration curve is not used, for example, the carbon concentration can be evaluated by a relative criterion for determining that the larger the value obtained as the evaluation result is, the higher the carbon concentration is. For example, it can be determined that the higher the value of the DLTS spectrum peak intensity (DLTS signal intensity), the higher the carbon concentration. When a calibration curve is used, the calibration curve may be, for example, a correlation between the density of the trap level obtained from the evaluation result (eg, DLTS signal intensity) obtained for the silicon sample to be evaluated and the known carbon concentration. It is preferable to create a calibration curve as shown. A relational expression for obtaining the density of trap levels from various evaluation results is known. Further, the above-mentioned known carbon concentration can be obtained by measuring by a method other than the evaluation method used for evaluating the silicon sample to be evaluated. For example, when the silicon sample to be evaluated is evaluated by the DLTS method, the above-mentioned known carbon concentration can be obtained by, for example, the SIMS method or the FT-IR method. Relational expressions for obtaining the carbon concentration from the evaluation results obtained by these methods are also known. The silicon sample to be evaluated by the same evaluation method as the silicon sample to be evaluated to prepare the calibration curve (silicon sample for preparing the calibration curve) and the silicon sample for obtaining the known carbon concentration are the same silicon sample (for example, It is preferably a silicon sample cut out from the same ingot, the same wafer, or the like) or a silicon sample that has gone through the same manufacturing process. Regarding the preparation of the calibration curve, reference can also be made to paragraphs 0038 to 0040 of Patent Document 1 (JP-A-2017-191800). The calibration curve creation silicon sample is preferably a silicon sample that has been subjected to various processes such as a hydrogen atom introduction process in the same manner as the silicon sample to be evaluated. For example, it is preferable to introduce hydrogen atoms into the calibration curve-producing silicon sample using hydrofluoric nitric acid having the same molar ratio as hydrofluoric nitric acid used for introducing hydrogen atoms into the silicon sample to be evaluated.

[Evaluation method of silicon wafer manufacturing process and silicon wafer manufacturing method]
One aspect of the present invention is to evaluate the carbon concentration of a silicon wafer manufactured in a silicon wafer manufacturing process to be evaluated by the carbon concentration evaluation method, and in the silicon wafer manufacturing process to be evaluated based on the result of the evaluation. The present invention relates to a method for evaluating a silicon wafer manufacturing process including evaluating a degree of carbon contamination.

  Further, one embodiment of the present invention is to evaluate a silicon wafer manufacturing process by the above-described method for evaluating a silicon wafer manufacturing process, and as a result of the evaluation, a silicon wafer manufacturing process in which the degree of carbon contamination is determined to be an allowable level In, or, as a result of the above evaluation, after performing a carbon contamination reduction process on the silicon wafer manufacturing process determined that the degree of carbon contamination exceeds the allowable level, in this silicon wafer manufacturing process, to manufacture a silicon wafer, The present invention relates to a method for manufacturing a silicon wafer including:

  The silicon wafer manufacturing process to be evaluated in the above-described manufacturing process evaluation method may be a part or all of the processes for manufacturing a product silicon wafer. The production process of a product silicon wafer is generally performed by cutting a wafer from a silicon single crystal ingot (slicing), surface treatment such as polishing and etching, a cleaning process, and a post-process (epitaxial layer) performed as necessary according to the use of the wafer. Formation, etc.). Each of these steps and each process is known.

  In the silicon wafer manufacturing process, carbon contamination may occur in the silicon wafer due to contact between a member used in the manufacturing process and the silicon wafer. By assessing the degree of carbon contamination by evaluating the carbon concentration of the silicon wafers manufactured in the manufacturing process to be evaluated, the tendency of carbon contamination to occur in the product silicon wafer due to the silicon wafer manufacturing process to be evaluated is evaluated. You can figure out. That is, it can be determined that the higher the carbon concentration of the silicon wafer manufactured in the manufacturing process to be evaluated, the more likely it is that carbon contamination occurs in the manufacturing process to be evaluated. Therefore, for example, the allowable level of the carbon concentration is set in advance, and if the carbon concentration obtained for the silicon wafer manufactured in the silicon wafer manufacturing process to be evaluated exceeds the allowable level, the manufacturing process to be evaluated is Therefore, it can be determined that carbon is not likely to be used in the production process of a product silicon wafer because of a high tendency to generate carbon contamination. It is preferable that the silicon wafer manufacturing process to be evaluated, which is determined in this way, be used for manufacturing a product silicon wafer after performing a carbon contamination reduction process. Details of this point will be described later.

  The carbon concentration of the silicon wafer manufactured in the silicon wafer manufacturing process to be evaluated is obtained by the above-described carbon concentration evaluation method according to one embodiment of the present invention. The details of the carbon concentration evaluation method are as described in detail above. The silicon wafer to be subjected to the carbon concentration evaluation is at least one silicon wafer manufactured in the silicon wafer manufacturing process to be evaluated, and may be two or more silicon wafers. When the carbon concentration of two or more silicon wafers is obtained, for example, the average value, the maximum value, and the like of the obtained carbon concentrations can be used for evaluating the silicon wafer manufacturing process to be evaluated. Further, the silicon wafer may be subjected to carbon concentration evaluation as it is, or a part thereof may be cut out and subjected to carbon concentration evaluation. When two or more samples are cut out from one silicon wafer and subjected to carbon concentration evaluation, the average value, maximum value, and the like of the carbon concentrations obtained for the two or more samples are determined as the carbon concentration of the silicon wafer. be able to.

  In one aspect of the method for manufacturing a silicon wafer, the silicon wafer manufacturing process is evaluated by the manufacturing process evaluation method, and as a result of the evaluation, the silicon wafer is manufactured in the silicon wafer manufacturing process in which the degree of carbon contamination is determined to be an allowable level. I do. This makes it possible to ship a high-quality silicon wafer having a low carbon contamination level as a product wafer. In another aspect of the method of manufacturing a silicon wafer, a silicon wafer manufacturing process is evaluated by the manufacturing process evaluation method, and as a result of the evaluation, a silicon wafer manufacturing process in which the degree of carbon contamination is determined to exceed an allowable level is determined. After performing the carbon contamination reduction process in the process, a silicon wafer is manufactured in this silicon wafer manufacturing process. Thus, carbon contamination due to the manufacturing process can be reduced, so that a high-quality silicon wafer having a low carbon contamination level can be shipped as a product wafer. The above-mentioned tolerance level can be appropriately set according to the quality required for the product wafer. In addition, the carbon contamination reduction processing includes replacement and cleaning of members included in a silicon wafer manufacturing process. As an example, in the case where a susceptor made of SiC is used as a susceptor that is a member on which a silicon wafer is mounted in a silicon wafer manufacturing process, a portion of a contact with the susceptor may be carbon-contaminated due to deterioration of a susceptor used repeatedly. . In such a case, for example, by replacing the susceptor, carbon contamination caused by the susceptor can be reduced.

[Method of manufacturing silicon single crystal ingot]
One embodiment of the present invention is to grow a silicon single crystal ingot, to evaluate the carbon concentration of a silicon sample cut out from the silicon single crystal ingot by the carbon concentration evaluation method, based on the result of the evaluation, silicon The present invention relates to a method for manufacturing a silicon single crystal ingot, including determining manufacturing conditions for a single crystal ingot, and growing a silicon single crystal ingot under the determined manufacturing conditions.

  The silicon single crystal ingot can be grown by a known method such as a CZ method (Czochralski method) and an FZ method (floating zone melting method). For example, carbon may be mixed into a silicon single crystal ingot grown by the CZ method due to carbon mixed in the source polysilicon, CO gas generated during the growth, and the like. It is preferable to evaluate such a mixed carbon concentration and determine the manufacturing conditions based on the evaluation result in order to manufacture a silicon single crystal ingot in which the mixed carbon is suppressed. Therefore, the above-described method for evaluating the concentration of carbon according to one embodiment of the present invention is suitable as a method for evaluating the concentration of mixed carbon.

For details such as the shape of the silicon sample cut out from the silicon single crystal ingot, the above description regarding the silicon sample to be evaluated by the above carbon concentration evaluation method can be referred to. The number of silicon samples subjected to the carbon concentration evaluation is at least one, and may be two or more. When the carbon concentrations of two or more silicon samples are obtained, for example, the average value, the maximum value, and the like of the obtained carbon concentrations can be used for determining the manufacturing conditions of the silicon single crystal ingot. For example, when the obtained carbon concentration is at a predetermined allowable level, a silicon single crystal ingot is grown under the production conditions when a silicon single crystal ingot obtained by cutting out a silicon sample whose carbon concentration has been evaluated is grown. As a result, a silicon single crystal ingot with less carbon contamination can be manufactured. On the other hand, for example, when the obtained carbon concentration exceeds the allowable level, the carbon single crystal ingot is grown under the determined production conditions by adopting a means for reducing the carbon concentration, thereby reducing the carbon content. It is possible to manufacture a silicon single crystal ingot with less contamination. As means for reducing carbon contamination, for example, for the CZ method, one or more of the following means (1) to (3) can be adopted. Further, for example, for the FZ method, one or more of the following means (4) to (6) can be adopted.
(1) Use a high-grade product with less carbon contamination as the raw material polysilicon.
(2) The pulling speed and / or the flow rate of argon (Ar) gas at the time of crystal pulling are appropriately adjusted to suppress the dissolution of CO in the polysilicon melt.
(3) To change the design of the carbon member included in the lifting device, change the mounting position, and the like.
(4) Use high-grade products with less carbon contamination as silicon raw materials.
(5) To suppress the incorporation of carbon from the atmospheric gas by increasing the flow rate of the gas introduced into the single crystal manufacturing apparatus.
(6) Replacement of a member made of a carbon-containing material included in the single crystal manufacturing apparatus, change in the design of the member, change in the mounting position, and the like.

  Thus, according to one embodiment of the present invention, a silicon single crystal ingot and a silicon wafer having a low carbon concentration can be provided.

  Hereinafter, the present invention will be further described based on examples. However, the present invention is not limited to the embodiments shown in the examples. The following treatments and operations were performed at room temperature in an air atmosphere unless otherwise specified.

1. Growing Silicon Single Crystal Ingot by CZ Method A silicon single crystal ingot (n-type silicon) was grown using a silicon single crystal pulling apparatus having the configuration shown in FIG.
Hereinafter, the silicon single crystal pulling apparatus shown in FIG. 1 will be described in detail.
A silicon single crystal pulling apparatus 10 shown in FIG. 1 includes a chamber 11, a support rotation shaft 12 penetrating vertically through a bottom center of the chamber 11, and a graphite susceptor fixed to an upper end of the support rotation shaft 12. 13, a quartz crucible 14 accommodated in the graphite susceptor 13, a heater 15 provided around the graphite susceptor 13, a support shaft driving mechanism 16 for raising and lowering and rotating the support rotation shaft 12, and a seed crystal. The holding of the seed chuck 17, the pulling wire 18 for suspending the seed chuck 17, the wire winding mechanism 19 for winding the pulling wire 18, and the silicon single crystal ingot 20 by the radiant heat from the heater 15 and the quartz crucible 14. It prevents heating and suppresses temperature fluctuation of the silicon melt 21. And the heat shield member 22, and a control unit 23 that controls each unit.
A gas inlet 24 for introducing Ar gas into the chamber 11 is provided at an upper portion of the chamber 11. Ar gas is introduced into the chamber 11 from the gas inlet 24 through the gas pipe 25, and the amount of Ar gas is controlled by the conductance valve 26.
A gas outlet 27 for exhausting Ar gas in the chamber 11 is provided at the bottom of the chamber 11. Ar gas in the sealed chamber 11 is discharged from the gas discharge port 27 to the outside via the exhaust gas pipe 28. A conductance valve 29 and a vacuum pump 30 are provided in the middle of the exhaust gas pipe 28, and the vacuum pump 30 sucks Ar gas in the chamber 11 while controlling the flow rate of the Ar gas in the chamber 11 to reduce the pressure in the chamber 11. The state is maintained.
Further, a magnetic field supply device 31 for applying a magnetic field to the silicon melt 21 is provided outside the chamber 11. The magnetic field supplied from the magnetic field supply device 31 may be a horizontal magnetic field or a cusp magnetic field.
In the above method, two or more silicon single crystal ingots having different carbon concentrations were grown by changing one or more manufacturing conditions selected from the group consisting of a raw material polysilicon grade, a pulling apparatus, and growing conditions.

2. Cutting out a silicon sample A wafer-shaped sample was cut out from the silicon single crystal ingot grown in the above and processed into a silicon wafer by processing such as mirror polishing. The resistivity was 10 to 13 Ωcm. From this silicon wafer, a silicon sample for SIMS measurement, a silicon sample for oxygen concentration measurement, and a plurality of silicon samples for DLTS measurement were obtained. Hereinafter, a sample obtained from one of two ingots having different carbon concentrations is referred to as “level 1”, and a sample obtained from the other ingot is referred to as “level 2”.

3. Carbon concentration measurement by SIMS method and oxygen concentration measurement by FT-IR method The carbon concentration of the above silicon sample for SIMS measurement was evaluated by SIMS method (raster change method). The carbon concentration determined for the sample at 1.10 × 10 ¹⁵ atms / cm ³ and level 2 was 8.60 × 10 ¹⁴ atms / cm ³ .
The oxygen concentration of the silicon sample for oxygen concentration measurement obtained by the FT-IR method was in the range of 2.0 × 10 ^{17 to} 12.0 × 10 ¹⁷ atoms / cm ³ at both Level 1 and Level 2.

4. Measurement by DLTS Method The following processes (A), (B) and (C) were sequentially performed on the above-mentioned silicon sample for DLTS measurement. By the following process (A) (wet process), hydrogen atoms were introduced into the silicon sample for DLTS measurement. For the above-mentioned plurality of silicon samples for DLTS measurement, in the treatment of the following (A), hydrofluoric nitric acid having a different molar ratio (HNO ₃ / (HNO ₃ + HF)) was used. Hydrofluoric acid was prepared by mixing nitric acid with an HNO ₃ concentration of 69% by mass and hydrofluoric acid with an HF concentration of 50% by mass, and the molar ratio was adjusted by the mixing ratio of nitric acid and hydrofluoric acid. After the following process (A), a Schottky junction is formed on one surface of the silicon sample by the following process (B), and an ohmic layer (Ga layer) is formed on the other surface by the following process (C). Thus, a diode was manufactured.
(A) The whole is immersed in hydrofluoric-nitric acid for 5 minutes and then washed with water for 10 minutes (B) Formation of Schottky electrode (Au electrode) by vacuum deposition (C) Formation of back ohmic layer by gallium rubbing

In the Schottky junction of the diode fabricated above, a reverse voltage for forming a depletion layer having a width of 6 μm in a region at a depth of 2 μm from the surface of the silicon sample and a pulse voltage for capturing carriers in the depletion layer are alternately and alternately applied. It was applied periodically. The transient response of the capacitance (capacitance) of the diode generated corresponding to the above voltage was measured.
The above-described voltage application and capacitance measurement were performed while sweeping the sample temperature within a predetermined temperature range. DLTS signal strength ΔC was plotted against temperature to obtain a DLTS spectrum. The measurement frequency was 250 Hz.
The obtained DLTS spectrum was subjected to fitting processing (True shape fitting processing) using a program manufactured by SEMILAB, and separated into DLTS spectra having a trap level of Ec-0.15 eV (peak position: temperature 101K). From the DLTS signal intensity at this peak position, the trap level density was determined by a known relational expression. Table 1 and FIGS. 2 (level 1) and FIG. 3 (level 2) show the trap level densities obtained for the samples wet-treated using hydrofluoric nitric acid at each molar ratio.

As shown in Table 1 and FIGS. 2 and 3, the molar ratio (HNO ₃ / (HNO ₃ + HF)) ranges from 0.72 to 0.82 or from 0.87 to 0.91. In Examples 1 to 4 in which a hydrogen atom was introduced using hydrofluoric nitric acid, hydrogen fluoride and nitric acid (molar ratio: 1.00) having a molar ratio outside the above range were used for a sample cut from the same ingot. Higher trap level density values were obtained as compared with Comparative Examples 1 and 2 in which atoms were introduced. If the trap level density can be increased, the carbon concentration can be evaluated with higher sensitivity.
An example of the carbon concentration evaluation is as follows.
For example, in the CZ method, a plurality of silicon single crystal ingots having different carbon concentrations are manufactured by changing at least one manufacturing condition selected from the group consisting of a raw material polysilicon grade, a pulling device, and a growing condition. The silicon samples cut out from each silicon single crystal ingot were subjected to the above-described processes (A) to (C) and DLTS measurement in the same manner as in the above example, and Ec-0.10 eV, Ec-0.13 eV and Ec For one or more trap levels selected from the group consisting of -0.15 eV, the DLTS signal strength at the peak position is determined. The carbon concentration of the silicon sample can be evaluated by a relative criterion for determining that the larger the value of the DLTS signal strength thus obtained is, the higher the carbon concentration is.
Alternatively, for example, the same processes (A) to (C) and the DLTS measurement as in the above-described embodiment are performed on a plurality of silicon samples having different carbon concentrations. By plotting the trap level density determined in this way with respect to the carbon concentration determined by SIMS of a silicon sample cut out from the same silicon single crystal ingot as each of the plurality of silicon samples having different carbon concentrations, A calibration curve can be created. The calibration curve thus created can be used to evaluate the carbon concentration of a silicon sample whose carbon concentration is unknown.
In the above example, in the evaluation of the carbon concentration, a trap level of Ec-0.15 eV was used as the trap level. However, as described in Patent Document 1 (Japanese Patent Application Laid-Open No. 2017-191800), Ec Trap levels of −0.10 eV and Ec−0.13 eV can also be used for carbon concentration evaluation.

  INDUSTRIAL APPLICATION This invention is useful in the technical field of a silicon single crystal ingot and a silicon wafer.

Claims (7)

Introducing hydrogen atoms into the silicon sample to be evaluated,
The evaluation target silicon sample into which the hydrogen atoms have been introduced is subjected to an evaluation by an evaluation method for evaluating a trap level in a band gap of silicon, and among the evaluation results obtained by the evaluation, Ec-0. Evaluating the carbon concentration of the silicon sample to be evaluated based on the evaluation result regarding the density of at least one trap level selected from the group consisting of 10 eV, Ec-0.13 eV, and Ec-0.15 eV;
Including
The introduction of the hydrogen atoms is performed by bringing the silicon sample to be evaluated into contact with a solution, and the solution has a molar ratio (HNO ₃ / (HNO ₃ + HF)) in the range of 0.72 to 0.82 or 0. A method for evaluating the carbon concentration of a silicon sample, which is hydrofluoric nitric acid in the range of 87 to 0.91. 2. The method for evaluating a carbon concentration of a silicon sample according to claim 1, wherein the evaluation target silicon sample into which the hydrogen atoms have been introduced is subjected to the evaluation without performing electron beam irradiation treatment. The evaluation of the carbon concentration of the silicon sample to be evaluated is performed based on the evaluation result regarding the trap level density of Ec-0.15 eV among the evaluation results obtained by the evaluation. Method for evaluating carbon concentration of silicon sample. The method for evaluating the carbon concentration of a silicon sample according to any one of claims 1 to 3, wherein the evaluation method is a DLTS method. The carbon concentration of the silicon wafer manufactured in the silicon wafer manufacturing process to be evaluated is evaluated by the method according to any one of claims 1 to 4, and the silicon wafer manufacturing process to be evaluated is based on the result of the evaluation. Assessing the extent of carbon pollution in
And a method for evaluating a silicon wafer manufacturing process. An evaluation of the silicon wafer manufacturing process by the evaluation method according to claim 5, and in the silicon wafer manufacturing process in which the degree of carbon contamination is determined to be an allowable level as a result of the evaluation, or as a result of the evaluation, After performing a carbon contamination reduction process on the silicon wafer manufacturing process determined that the degree of contamination exceeds the allowable level, in the silicon wafer manufacturing process, manufacturing a silicon wafer,
A method for manufacturing a silicon wafer, comprising: Growing silicon single crystal ingots,
Evaluating the carbon concentration of a silicon sample cut out from the silicon single crystal ingot by the method according to any one of claims 1 to 4,
Based on the result of the evaluation, determining the production conditions of the silicon single crystal ingot, and,
Growing a silicon single crystal ingot under determined manufacturing conditions,
A method for producing a silicon single crystal ingot, comprising: