Measurement method for oxygen concentration of oxygen atom layer

Abstract

To provide a method for stably and easily measuring an oxygen concentration of an oxygen atom layer in an epitaxial wafer in a nondestructive manner.SOLUTION: Disclosed is a measurement method for an oxygen concentration of an oxygen atom layer in an epitaxial wafer in which the oxygen atom layer and a monocrystal silicon epitaxial layer on the oxygen atom layer are formed on a silicon monocrystal substrate. The measurement method for the oxygen concentration of the oxygen atom layer includes: previously creating a calibration curve between an oxygen concentration of an oxygen layer in an epitaxial wafer for a preliminary test and a band end emission intensity of the epitaxial wafer for the preliminary test according to a photo luminescence method or a cathode luminescence method and measuring an oxygen concentration of an oxygen atom layer in an epitaxial wafer of a measurement target from a measurement result of the band end emission intensity of the epitaxial wafer of the measurement target using the calibration curve.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method for measuring the oxygen concentration of an oxygen atomic layer of an epitaxial wafer. [Background technology]

[0002] Silicon substrates, which form semiconductor elements such as solid-state imaging devices and other transistors, are required to have the ability to getter away elements that disrupt device characteristics, such as heavy metals. Various gettering methods have been proposed and put into practical use, including providing a polycrystalline silicon (Poly-Si) layer on the backside of the silicon substrate, forming a damaged layer by blasting, using a silicon substrate with a high concentration of boron, and forming precipitates. Gettering by oxygen precipitation involves taking in metals that have a high ionization tendency (low electronegativity) against oxygen, which has a high electronegativity.

[0003] Another method proposed is proximity gettering, which involves forming a gettering layer near the active region of a device. For example, a substrate can be formed by epitaxially growing silicon on a substrate into which carbon has been ion-implanted. Gettering requires elements to diffuse to the gettering site (where the energy of the entire system is reduced by bonding or clustering at the site rather than by the metal existing as a single element). The diffusion coefficient of metal elements contained in silicon varies depending on the element, and in addition, the recent trend toward lower process temperatures has made it difficult for metals to diffuse to the gettering site, so the proximity gettering method has been proposed.

[0004] If oxygen can be used for proximity gettering, it is believed that a silicon substrate with a very effective gettering layer will be obtained. In particular, an epitaxial wafer with an oxygen atomic layer in the middle of the epitaxial layer can reliably getter metal impurities even in recent low-temperature processes. Therefore, in recent years, a structure has been proposed in which an oxygen atomic layer is formed on a silicon substrate and then a silicon epitaxial layer is formed.

[0005] The prior art of such a method for measuring the oxygen concentration in an oxygen atomic layer will be mentioned below. Patent Document 1 describes measuring the oxygen concentration in an epitaxial wafer having an oxygen atomic layer and a single crystal silicon epitaxial layer on a wafer such as silicon by SIMS (Secondary Ion Mass Spectrometry). Patent Document 2 describes measuring the thickness of an oxygen atomic layer on a substrate such as silicon having an oxygen atomic layer on its surface by ellipsometry.

[0006] Furthermore, Patent Documents 3 and 4 are cited as methods for measuring the oxygen concentration in silicon single crystals (silicon wafers), and describe methods for measuring the oxygen concentration in silicon substrates by photoluminescence or cathodoluminescence. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent Publication No. 2021-111696 [Patent Document 2] Japanese Patent Application Laid-Open No. 2014-165494 [Patent Document 3] Japanese Patent Application Publication No. 2019-160908 [Patent Document 4] Japanese Patent Application Publication No. 2020-085758 Summary of the Invention [Problem to be solved by the invention]

[0008] As mentioned above, oxygen concentration measurements of oxygen atomic layers have been performed in the past. However, while conventional techniques can measure oxygen concentration, they have problems such as the need to destroy the sample and fluctuations in measurement values ​​due to surface roughness. For example, the technique described in Patent Document 1 can measure the oxygen concentration with high accuracy, but has the problem of destroying the sample. Furthermore, in the technique described in Patent Document 2, since the oxygen atomic layer is thin, there is a problem in that the thickness of the oxygen atomic layer varies depending on the roughness of the surface of the silicon substrate.

[0009] Furthermore, the techniques described in Patent Documents 3 and 4 require irradiation with high-energy particles to generate pairs of interstitial silicon and vacancies before measuring the oxygen concentration. Furthermore, the irradiation with high-energy particles also causes the problem of damaging the sample. Furthermore, these techniques are primarily intended to measure the oxygen concentration in a silicon single crystal, not the oxygen concentration in the oxygen atomic layer in an epitaxial wafer.

[0010] As described above, while some conventional techniques can measure the oxygen concentration in the oxygen atomic layer of an epitaxial wafer, they have problems such as the need to destroy the sample and difficulty in accurate measurement. Therefore, a method for easily and accurately measuring the oxygen concentration in the oxygen atomic layer is needed.

[0011] The present invention has been made in consideration of the above-mentioned problems of the conventional technology, and an object of the present invention is to provide a method for stably, easily and non-destructively measuring the oxygen concentration of an oxygen atomic layer in an epitaxial wafer. [Means for solving the problem]

[0012] In order to achieve the above object, the present invention provides a method for measuring an oxygen concentration in an oxygen atomic layer of an epitaxial wafer in which an oxygen atomic layer and a single crystal silicon epitaxial layer on the oxygen atomic layer are formed on a silicon single crystal substrate, the method comprising the steps of: a calibration curve is prepared in advance between the oxygen concentration of the oxygen atomic layer of the epitaxial wafer for a preliminary test and the band edge luminescence intensity of the epitaxial wafer for the preliminary test by a photoluminescence method or a cathodoluminescence method; The present invention provides a method for measuring the oxygen concentration of an oxygen atomic layer, which comprises using the calibration curve to measure the oxygen concentration of the oxygen atomic layer of an epitaxial wafer to be measured from the measurement results of the band edge emission intensity of the epitaxial wafer to be measured.

[0013] By using such an oxygen concentration measurement method, the oxygen concentration of the oxygen atomic layer in the silicon epitaxial wafer can be measured stably, easily, and non-destructively.

[0014] At this time, the epitaxial wafer for the preliminary test and the epitaxial wafer to be measured are: An epitaxial wafer can be prepared in which a plurality of oxygen atomic layers and single crystal silicon epitaxial layers are alternately formed on the silicon single crystal substrate.

[0015] In this way, the measurement method of the present invention is applicable to cases where a plurality of layers, such as the oxygen atomic layers and single crystal silicon epitaxial layers, are formed alternately, and is therefore highly versatile.

[0016] When preparing the calibration curve, the oxygen concentration of the oxygen atomic layer can be measured by SIMS.

[0017] In this way, the oxygen concentration in the oxygen atomic layer of the epitaxial wafer for the preliminary test can be measured. [Effects of the Invention]

[0018] According to the present invention, it is possible to provide a method for measuring the oxygen concentration of an oxygen atomic layer stably, easily, and non-destructively. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a flowchart showing an example of steps of a method for measuring the oxygen concentration of an oxygen atomic layer according to the present invention. [Figure 2] 1 is a graph showing a calibration curve illustrating the relationship between band edge emission intensity and oxygen concentration in an oxygen atomic layer in an example. [Figure 3] 10 is a graph showing a calibration curve illustrating the relationship between the film thickness of the oxygen atomic layer measured by spectroscopic ellipsometry and the oxygen concentration of the oxygen atomic layer in a comparative example. [Figure 4] 1 is a graph showing the relationship between the oxygen concentration of the oxygen atomic layer measured by SIMS and the oxygen concentration of the oxygen atomic layer estimated from calibration curves in Examples and Comparative Examples. DETAILED DESCRIPTION OF THE INVENTION

[0020] The present invention will be described in detail below, but the present invention is not limited thereto. As described above, there has been a need for a method for stably, simply, and non-destructively measuring the oxygen concentration of an oxygen atomic layer in an epitaxial wafer. The present inventors have conducted extensive research to address the above-mentioned problem. As a result, they have discovered that, when measuring the oxygen concentration of an oxygen atomic layer in an epitaxial wafer (an oxygen atomic layer and a single-crystal silicon epitaxial layer formed on a silicon single-crystal substrate), a calibration curve between the oxygen concentration of the oxygen atomic layer of a preliminary test epitaxial wafer and the band-edge emission intensity of the preliminary test epitaxial wafer measured by a photoluminescence (PL) method or a cathodoluminescence (CL) method is prepared, and the calibration curve is used to measure the oxygen concentration of the oxygen atomic layer of the epitaxial wafer to be measured from the measurement results of the band-edge emission intensity of the epitaxial wafer to be measured. This finding led to the completion of the present invention. The details will be explained below with reference to the drawings.

[0021] FIG. 1 is a flow chart showing an example of the steps of the method of measuring the oxygen concentration of an oxygen atomic layer according to the present invention. As shown in Figure 1, the process is roughly divided into a preliminary test and a main test. The preliminary test is a step in which a preliminary test epitaxial wafer (hereinafter also referred to as preliminary test EPW) is used to create a calibration curve between the oxygen concentration in the oxygen atomic layer of the preliminary test EPW and the band-edge luminescence intensity of the preliminary test EPW measured by PL or CL. The main steps are: S11: Preparation of the preliminary test EPW; S12: Band-edge luminescence intensity measurement (PL or CL); S13: Oxygen concentration measurement (SIMS); and S14: Creation of the calibration curve. This test is the stage where the oxygen concentration of the oxygen atomic layer is measured (estimated) from the band edge emission intensity (PL or CL method) measurement results of the epitaxial wafer (main test EPW) to be measured using the calibration curve created in the preliminary test. It mainly consists of the following steps: S21: Preparation of the main test EPW, S22: Band edge emission intensity measurement (PL or CL method), S23: Oxygen concentration measurement. Each step in each stage will be described in detail below.

[0022] <Preliminary Exam> First, the preliminary test will be described. (S11: Preparation for preliminary examination EPW) The step S11 in FIG. 1 is a step of preparing an epitaxial wafer for a preliminary test. The preliminary test EPW is an epitaxial wafer in which an oxygen atomic layer and a single-crystal silicon epitaxial layer are formed on a silicon single crystal substrate (hereinafter also referred to as a silicon substrate). For example, if the conditions for forming the oxygen atomic layer of the main test EPW described below are known, an epitaxial wafer having an oxygen atomic layer formed under conditions close to those conditions can be used. Using such a preliminary test EPW enables more accurate measurements in the main test. The oxygen atomic layer may be formed in a plurality of layers. An epitaxial wafer including a plurality of oxygen atomic layers is a substrate in which oxygen atomic layers and single crystal silicon epitaxial layers are alternately and repeatedly formed (multiple layers are formed) on a single crystal silicon substrate. The measurement method of the present invention is applicable to such epitaxial wafers and is highly versatile.

[0023] The number of preliminary test EPWs to be prepared is not particularly limited, as long as it is sufficient to prepare a calibration curve in a subsequent process. For example, about 10 to 20 patterns of preliminary test EPWs with different conditions for forming the number of oxygen atoms can be prepared.

[0024] Here, the method for manufacturing the silicon substrate is not particularly limited. A substrate manufactured by the Czochralski method (hereinafter referred to as the CZ method) or a substrate manufactured by the floating zone method (hereinafter referred to as the FZ method) may be used. Also, a substrate in which single crystal silicon is epitaxially grown on a silicon substrate manufactured by the CZ method or the FZ method may be used.

[0025] An oxygen atomic layer refers to a layer in which oxygen is one atomic layer or less. In single-crystal silicon, oxygen atoms are stable at the bond center position between silicon atoms and the nearest silicon atoms. When one atomic layer of oxygen exists, the planar concentration of oxygen is 1.36 × 10 15 atoms / cm 2 This becomes: The method for forming the oxygen atomic layer is not particularly limited. For example, the oxygen atomic layer can be formed by removing a native oxide film from a silicon substrate and then oxidizing the silicon substrate by exposing it to an oxidizing gas or immersing it in a solution. The native oxide film can be removed by immersing the silicon substrate in a solution containing hydrofluoric acid or by heating the silicon substrate in a reducing atmosphere such as hydrogen.

[0026] Examples of oxidizing gases that can be used include oxygen molecules, ozone, nitric oxide, nitrous oxide, and nitrogen dioxide. These gases may be used alone or diluted with an inert gas such as nitrogen, argon, helium, neon, krypton, or xenon. Oxidation with a gas may be carried out at room temperature or by heating.

[0027] Examples of liquids that can be used include pure water, SC1 solution, hydrogen peroxide, and ozone water. SC1 solution is a mixture of ammonia, hydrogen peroxide, and water. The hydrogen peroxide oxidizes the surface of the silicon substrate, while the ammonia etches the oxide film, thereby lifting off and removing particles adhering to the surface of the single-crystal silicon substrate. Therefore, if an SC1 solution is used in the cleaning process of single-crystal silicon substrates, an oxide film will be formed on the surface of the silicon substrate. Furthermore, oxidation using a liquid can be performed at a temperature of, for example, 10°C or higher and lower than 100°C.

[0028] Furthermore, the formation of a single crystal silicon epitaxial layer, i.e., the epitaxial growth of single crystal silicon, can be performed at a temperature of, for example, 450°C or higher and 800°C or lower. By performing epitaxial growth at such a temperature, the formation of dislocations and stacking faults in the epitaxial layer to be formed can be prevented, and a high-quality epitaxial layer can be formed more stably. Since the higher the temperature, the higher the epitaxial growth rate, a thick epitaxial layer can be formed in a short time by forming the film at a high temperature. On the other hand, if a thin epitaxial layer is desired, film formation can be performed at a low temperature.

[0029] In this way, the growth temperature can be changed depending on the desired thickness of the epitaxial layer. Furthermore, within this temperature range, it is possible to more reliably prevent oxygen from diffusing from the oxygen atomic layer during growth of the epitaxial layer, thereby preventing the oxygen atomic layer from disappearing. Furthermore, the deposition time can be adjusted to adjust the thickness of the epitaxial layer. Furthermore, when depositing at high temperatures, shortening the deposition time can prevent oxygen from diffusing outward from the single-crystal silicon substrate, thereby preventing a decrease in the heat resistance of the oxygen atomic layer. The gases used for growth may be monosilane and disilane, and nitrogen and hydrogen may be used as carrier gases.

[0030] (S12: Band-edge emission intensity measurement (PL or CL method)) S12 in FIG. 1 is a step of measuring the band edge emission intensity. When a silicon single crystal substrate is irradiated with high-energy particles above the band gap, some of the electrons in the valence band are excited into the conduction band, generating excess electrons in the conduction band and excess holes in the valence band, which emit light when they recombine (the excited electrons return to the ground state). The luminescence that occurs when light is used as the high-energy particle is called photoluminescence, and the luminescence that occurs when electrons are used is called cathodoluminescence. The energy of the excited electrons varies depending on the energy of the incident light or electrons, but the excited electrons emit heat and their energy immediately drops to the conduction band edge. Therefore, either method emits light at an energy lower than the band gap. Essentially, there is no difference in the luminescent species that can be detected by photoluminescence and cathodoluminescence. Therefore, either photoluminescence or cathodoluminescence can be used.

[0031] In PL, the higher the energy of the incident light (the shorter the wavelength), the shallower the penetration depth. In CL, the higher the energy of the incident electrons, the deeper the penetration depth. This means that the measurement depth can be changed by adjusting the energy of the incident light or electrons. For example, in PL, the wavelength of the incident light can be set to 300 nm or more and 1100 nm or less, and in CL, the acceleration voltage can be set to 1 keV or more and 50 keV or less. Under these conditions, silicon atoms in lattice positions are more reliably ejected, preventing the generation of pairs of interstitial silicon atoms and their vacancies, and thus preventing damage to the sample.

[0032] Light emission can be broadly divided into two types: band-edge emission inherent to silicon, and light emission derived from impurities and defects in silicon. Because silicon is an indirect transition semiconductor, its luminescence efficiency is lower than that of direct transition semiconductors, but it emits light equivalent to the band gap via phonons. This is called band edge emission. The band gap increases as the temperature decreases, but at room temperature it is 1.12 eV (for silicon). This corresponds to a wavelength of 1100 nm. When silicon is cooled to 4.2 K, the band edge emission that appeared as a single light at room temperature splits into several due to differences in the energy of the phonon modes. Generally, the emission from TO phonons is stronger than other phonons such as TA, so band-edge emission is sometimes called TO line.

[0033] Luminescence derived from impurities in silicon includes luminescence from dopants such as boron and phosphorus, interstitial carbon, interstitial carbon-substituted carbon pairs, interstitial carbon-interstitial oxygen pairs, etc., while luminescence derived from defects includes luminescence from interstitial silicon clusters, dislocations, etc. Since the luminescence wavelength (luminescence wavelength via impurity levels or defect levels) differs depending on the impurity or defect, the amount of impurity or defect can be quantified individually.

[0034] As mentioned above, some of the excited carriers are consumed by recombination accompanied by light emission, but the rest are consumed by non-radiative recombination due to defects and surface states other than those mentioned above, where carriers recombine without light emission and release heat. In other words, the excited carriers are distributed among band-edge emission, light emission derived from impurities and defects, and heat due to non-radiative recombination.

[0035] Broadly speaking, excited carriers are distributed between silicon itself and impurities and defects. As mentioned above, silicon is an indirect transition semiconductor, so the interband radiative recombination rate is slow and it is easily affected by impurities and defects. For this reason, the band-edge luminescence intensity decreases as the number of impurities and defects increases. Therefore, when a particular impurity or defect is present in a large amount, its concentration can be measured from the band edge emission intensity. This method does not require the generation of interstitial silicon-vacancy pairs by high-energy electron beams or ion implantation to convert non-luminescent defect species into luminescent defect species, so the concentration can be measured without any additional pretreatment or damage to the epitaxial wafer.

[0036] Measurements can be performed at room temperature or cooled, which reduces the effects of non-radiative recombination and allows for more accurate measurements. The band edge emission intensity can be the average value of the band edge emission intensity over the entire surface of the epitaxial wafer, or the band edge emission intensity obtained by point measurement. If the oxygen concentration of the oxygen atomic layer varies within the epitaxial wafer surface, point measurement will be more accurate, but if the oxygen concentration of the oxygen atomic layer is uniform within the epitaxial wafer surface, there is no problem in using the average value of the band edge emission intensity over the entire surface of the epitaxial wafer.

[0037] (S13: Oxygen concentration measurement (SIMS)) The step S13 in FIG. 1 is a step of measuring the oxygen concentration (planar concentration of oxygen) of the oxygen atomic layer of the preliminary test EPW by a conventionally known measurement method (for example, SIMS (Secondary Ion Mass Spectrometry)). When silicon containing an oxygen atomic layer is measured by SIMS, a peak appears at the depth where the oxygen atomic layer has formed. The planar concentration can be calculated by integrating the product of the volume concentration and depth from a single sputtering step near the peak.

[0038] (S14: Creating a calibration curve) S14 in FIG. 1 is a step of creating a calibration curve between the oxygen concentration of the oxygen atomic layer measured by SIMS in S13 and the band edge emission intensity measured in S12. There are no particular restrictions on the function used for the calibration curve, but a linear function can be used, for example. The band-edge emission intensity decreases as the oxygen concentration in the oxygen atomic layer increases, so the oxygen concentration can be estimated from the band-edge emission intensity. As the oxygen concentration in the oxygen atomic layer increases, defects and oxygen (oxygen atoms can become non-radiative recombination centers) in the epitaxial layer of single-crystal silicon above the oxygen atomic layer increase, and it is thought that the band-edge emission intensity decreases as excited carriers are consumed by them. In this way, the correlation between the oxygen concentration in the oxygen atomic layer and the band edge emission intensity is determined in advance using the preliminary test EPW.

[0039] <Actual exam> This test will be described below. (S21: Preparation for the actual exam EPW) The step S21 in FIG. 1 is a step of preparing an epitaxial wafer to be measured (main test EPW). The silicon single crystal substrate, oxygen atomic layer, and single crystal silicon epitaxial layer for this test EPW are the same as those for the preliminary test EPW. An epitaxial wafer with an oxygen atomic layer under specified formation conditions can be newly manufactured and prepared, or it can be, for example, an epitaxial wafer provided by a client or a ready-made product with the above structure.

[0040] (S22: Band-edge emission intensity measurement (PL or CL method)) S22 in FIG. 1 is a step of measuring the band edge emission intensity. In the same manner as in step S12 of the preliminary test, the band edge emission intensity of the main test EPW is measured by the PL method or the CL method.

[0041] (S23: Oxygen concentration measurement) S23 in Figure 1 is the process of converting the band-edge emission intensity to oxygen concentration. Using a calibration curve between the band-edge emission intensity and the oxygen concentration in the oxygen atomic layer, which was previously created from the measurement results of the preliminary test EPW in S14, the oxygen concentration in the oxygen atomic layer in the main test EPW is estimated and measured from the band-edge emission intensity measured in S22.

[0042] The method for measuring the oxygen concentration of an oxygen atomic layer in an epitaxial wafer according to the present invention as described above is extremely useful because it allows the oxygen concentration of the oxygen atomic layer to be measured stably, easily, and non-destructively.

[0043] The upper limit of the measurable oxygen concentration varies depending on the depth of the oxygen atomic layer, the energy of the incident light or electrons, the temperature, etc. These conditions can be determined appropriately each time. Furthermore, it is preferable to create a calibration curve according to the number of oxygen atomic layers. In the case of multiple oxygen atomic layers (as described above, when multiple oxygen atomic layers and single-crystal silicon epitaxial layers are alternately formed), a calibration curve can be created between the sum of the oxygen concentrations of each oxygen atomic layer and the band-edge emission intensity. By using the sum of the oxygen concentrations in this way, a calibration curve can be created even when the oxygen concentrations of individual layers cannot be measured separately, i.e., when the spacing between oxygen atomic layers is narrow. [Example]

[0044] The present invention will be explained in more detail below by showing examples of the present invention, but the present invention is not limited to these examples. (Example) The method for measuring the oxygen concentration of the oxygen atomic layer according to the present invention shown in FIG. 1 was carried out by carrying out the following preliminary test and main test. <Preliminary Exam> First, a single crystal silicon substrate was prepared for the preliminary EPW test. The conductivity type, diameter, crystal plane orientation, and oxygen concentration of the prepared single crystal silicon substrate were as follows: Substrate conductivity type: p-type Diameter: 300mm Crystal plane orientation: (100) Oxygen concentration: 14 ppma (JEITA)

[0045] Next, the prepared single-crystal silicon substrate was immersed in hydrofluoric acid to remove the native oxide film, and then immersed in 13 levels of solution (hydrogen peroxide solution or pure water (H2O2 concentration: 0 (ppm))) with varying solution temperatures, immersion times, and H2O2 concentrations as shown in Table 1 to form an oxygen atomic layer.

[0046] [Table 1]

[0047] Thereafter, the thickness of the oxygen atomic layer at one point in the center was measured by spectroscopic ellipsometry as a parameter to be used in the comparative example described later. The device used was the M-2000 manufactured by JA Woollam Co., Ltd. The measurement was performed using white light as the light source. It should be noted that the time from the formation of the oxygen atomic layer to the measurement by spectroscopic ellipsometry and the measurement time for spectroscopic ellipsometry are short, so it has been confirmed that there is no increase in the oxygen concentration in the oxygen atomic layer due to the measurement by spectroscopic ellipsometry.

[0048] Next, the single-crystal silicon substrate with the oxygen atomic layer formed on its surface as described above was transferred into a single-crystal silicon epitaxial furnace, where single-crystal silicon was epitaxially grown using monosilane gas. The temperature was set to 700°C, the partial pressure of the monosilane gas was set to 60 Pa, and the film thickness was set to 60 nm. Hydrogen was used as the carrier gas. In this way, 13 levels of pre-test EPW were prepared.

[0049] After deposition, the band-edge emission intensity was measured by photoluminescence at room temperature over the entire surface of the preliminary test EPW. The equipment used was a SiPHER manufactured by Nanometrics. A laser with a wavelength of 532 nm was used as the excitation light source.

[0050] The planar concentration of oxygen in the oxygen atomic layer was then measured by SIMS (these SIMS measurements were also used in the comparative examples described later). The equipment used was a PHI ADEPT-1010 manufactured by ULVAC-PHI. Cs was used as the primary ion.+ The measurement was performed at an acceleration voltage of 3.0 kV using the above method. The measurement values ​​are also shown in Table 1 above.

[0051] A calibration curve was created from the above-mentioned band-edge emission intensity measurements and the oxygen concentration measurements by SIMS. The created calibration curve is shown in Figure 2. The horizontal axis is the band-edge emission intensity measured by PL spectroscopy, and the vertical axis is the oxygen concentration of the oxygen atomic layer measured by SIMS. The black circles represent the measured values, and the solid line is a line fitted using the least squares method. The relationship between the band-edge emission intensity and the oxygen concentration in the oxygen atomic layer (solid line in Fig. 2) is expressed as [oxygen concentration (atoms / cm 2 )]=-5.00×10 13 × [Band edge emission intensity] + 1.51 × 10 15 and can be fitted using the least squares method. The coefficient of determination of this equation, R 2 was 0.91.

[0052] Here, the band-edge emission intensity was the average value of the band-edge emission intensity over the entire surface of the epitaxial wafer. If there is variation in the oxygen concentration of the oxygen atomic layer within the epitaxial wafer surface, point measurements can create a more accurate calibration curve. However, since the oxygen concentration of the oxygen atomic layer was uniform over the entire surface of the epitaxial wafer for the sample prepared by the above method, it was possible to create a calibration curve using the average value.

[0053] <Actual exam> Next, the epitaxial wafers to be measured were fabricated (main test EPW) by forming an oxygen atomic layer and a single-crystal silicon epitaxial layer on a single-crystal silicon substrate under the same conditions as in the preliminary test, except for the formation of the oxygen atomic layer. Specifically, the oxygen atomic layer was formed by immersing the wafers in six levels of pure water (H2O2 concentration: 0 (ppm)) with the solution temperatures and immersion times shown in Table 2 varied.

[0054] [Table 2]

[0055] After the single crystal silicon epitaxial layer was formed, the band edge emission intensity was measured by photoluminescence at room temperature in the same manner as in the preliminary test. The oxygen concentration in the oxygen atomic layer of the EPW in this test was estimated and measured from the measured band edge emission intensity and the calibration curve shown in Figure 2, which was obtained in advance in a preliminary test.

[0056] After that, for verification, the oxygen concentration of the oxygen atomic layer of the test EPW was measured by SIMS under the same measurement conditions as in the preliminary test.

[0057] (Comparative Example) The oxygen concentration was measured from the thickness of the oxygen atomic layer as follows. A calibration curve between the thickness of the oxygen atomic layer and the oxygen concentration was calculated from the film thickness measured by spectroscopic ellipsometry and the oxygen concentration (planar concentration) of the oxygen atomic layer measured by SIMS at the stage when an oxygen atomic layer was formed on the surface of a single crystal silicon substrate under the conditions in Table 1, as determined in the preliminary test of the above-mentioned example. The calibration curve is shown in Figure 3. The horizontal axis is the film thickness of the oxygen atomic layer measured by spectroscopic ellipsometry, and the vertical axis is the oxygen concentration of the oxygen atomic layer measured by SIMS. The black circles represent the measured values, and the solid line is a line fitted by the least squares method. The relationship between the thickness of the oxygen atomic layer and the oxygen concentration of the oxygen atomic layer is expressed as [oxygen concentration (atoms / cm 2 )]=2.06×10 14 ×[Film thickness (Å)]-5.51×10 14 and can be fitted using the least squares method. The coefficient of determination of this equation, R 2 was 0.29.

[0058] Next, we fabricated epitaxial wafers similar to those used in the present test EPW of the example. Specifically, we formed an oxygen atomic layer by immersing the wafer in six levels of pure water (H2O2 concentration: 0 (ppm)) at different solution temperatures and immersion times, as shown in Table 2. The thickness of the oxygen atomic layer at one point in the center was measured using spectroscopic ellipsometry, and the oxygen concentration of the oxygen atomic layer of the epitaxial wafer was estimated and measured based on the measured thickness and the calibration curve shown in Figure 3 obtained earlier.

[0059] The correlation between the oxygen concentration in the oxygen atomic layer and the measured value by SIMS ("oxygen concentration in the oxygen atomic layer") is shown on the horizontal axis, and the measured value using the calibration curve (Example and Comparative Example) ("estimated oxygen concentration in the oxygen atomic layer") is shown on the vertical axis in Figure 4. The black circles represent data from the Example, and the triangles represent data from the Comparative Example. The dashed line indicates the case where the two (SIMS measurements and measurements using the calibration curve) match. The closer the data is to the dashed line, the more stable and accurate the measurement method is.

[0060] As a result of the example, the values ​​obtained using the calibration curve (measurements according to the present invention) were almost the same as the SIMS measurements, confirming the validity of the created calibration curve. On the other hand, as can be seen from the data of the comparative example, the calibration curve using film thickness deviates significantly from the SIMS measurements.

[0061] As can be seen from FIG. 4, it is clear that the accuracy is higher when the oxygen concentration of the oxygen atomic layer is estimated from the band edge emission intensity in the example. The reason why the accuracy of the comparative example, in which the oxygen concentration is estimated from the film thickness measured by spectroscopic ellipsometry, is lower is thought to be that the film thickness of the oxygen atomic layer varies because the surface roughness varies depending on the conditions for forming the oxygen atomic layer and because the surface roughness of the single-crystal silicon substrate before the oxygen atomic layer is formed varies. From these facts, it is better to use the band edge emission intensity in the example to obtain a coefficient of determination R 2 It was found that the accuracy of estimating the oxygen concentration in the monolayer using this calibration curve was higher.

[0062] The present invention is not limited to the above-described embodiments. The above-described embodiments are merely examples, and any configuration that is substantially identical to the technical idea described in the claims of the present invention and that provides similar effects is included within the technical scope of the present invention.

Claims

1. 1. A method for measuring an oxygen concentration of an oxygen atomic layer of an epitaxial wafer in which an oxygen atomic layer and a single-crystal silicon epitaxial layer on the oxygen atomic layer are formed on a silicon single-crystal substrate, the method comprising: a calibration curve is prepared in advance between the oxygen concentration of the oxygen atomic layer of the epitaxial wafer for a preliminary test and the band edge luminescence intensity of the epitaxial wafer for the preliminary test by a photoluminescence method or a cathodoluminescence method; A method for measuring the oxygen concentration of an oxygen atomic layer of an epitaxial wafer to be measured, comprising using the calibration curve to measure the oxygen concentration of the oxygen atomic layer of the epitaxial wafer to be measured from the measurement results of the band edge emission intensity of the epitaxial wafer to be measured.

2. The epitaxial wafer for the preliminary test and the epitaxial wafer to be measured are:

2. The method for measuring the oxygen concentration of an oxygen atomic layer according to claim 1, further comprising the step of preparing an epitaxial wafer in which a plurality of oxygen atomic layers and a plurality of single crystal silicon epitaxial layers are alternately formed on the silicon single crystal substrate.

3. 2. The method for measuring the oxygen concentration of the oxygen atomic layer according to claim 1, wherein the oxygen concentration of the oxygen atomic layer is measured by SIMS when the calibration curve is created.

4. 3. The method for measuring the oxygen concentration of the oxygen atomic layer according to claim 2, wherein the oxygen concentration of the oxygen atomic layer is measured by SIMS when the calibration curve is created.

Images