Gallium nitride single crystal substrate

Abstract

To provide a gallium nitride single crystal substrate without a processed altered layer. [Solution] A gallium nitride single crystal substrate having a diameter of 50 mm or more and two main surfaces facing opposite directions, wherein when light with a wavelength of 532 nm is incident perpendicularly to one of the main surfaces, the light diffusion transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light, satisfies Rd < 0.000791.

Description

[Technical Field]

[0001] This invention relates to a gallium nitride single crystal substrate. [Background technology]

[0002] When a semiconductor single crystal is processed into a substrate, its surface contains a surface layer (hereinafter referred to as the "processed altered layer") which has disordered areas of the single crystal, such as polycrystalline or amorphous layers introduced during slicing, grinding, or polishing, as well as crystal distortion, scratches, etc.

[0003] For example, Patent Document 1 discloses a method for evaluating a processed and altered layer of a semiconductor single crystal substrate, in which laser light is incident on the surface of the semiconductor single crystal substrate and the processed and altered layer of the semiconductor single crystal substrate is evaluated based on the intensity of the reflected and scattered light scattered inside the semiconductor single crystal substrate.

[0004] Furthermore, Patent Document 2 discloses a method for removing the processed altered layer by using the CL (cathodoluminescence) method to evaluate the processed altered layer and performing CMP (Chemical Mechanical Polishing) during grinding with a high-grit abrasive wheel. Hereafter, grinding using a high-grit abrasive wheel, as described in Patent Document 2, will be referred to as precision grinding. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 7344491 [Patent Document 2] Japanese Patent Publication No. 2023-032894 [Overview of the project] [Problems that the invention aims to solve]

[0006] The object of the present invention is to provide a gallium nitride single crystal substrate without a processed altered layer. [Means for solving the problem]

[0007] According to one aspect of the present invention, there is provided a gallium nitride single crystal substrate having a diameter of 50 mm or more and two main surfaces facing opposite directions to each other, wherein when light with a wavelength of 532 nm is perpendicularly incident on one of said main surfaces, the light diffuse transmittance Rd, which is a value obtained by dividing the intensity of diffuse transmitted light by the intensity of total transmitted light, satisfies Rd < 0.000791.

[0008] According to another aspect of the present invention, there is provided a gallium nitride single crystal substrate having a diameter of 50 mm or more and two main surfaces facing opposite directions to each other, wherein when light is perpendicularly incident on one of said main surfaces by a spectrophotometer, at least one wavelength λ of incident light satisfying the following formula (1) for the light diffuse transmittance Rd, which is a value obtained by dividing the intensity of diffuse transmitted light by the intensity of total transmitted light, exists in a range of 390 nm or more and 600 nm or less. Rd < a × b 2 ÷ (c × (λ - λ0) 2 + b 2 ) + d ···(1) [In formula (1), a=10, b=1.2, c=2.5, d=0.00065, λ is the wavelength of incident light (nm), and λ0=330.]

Effects of the Invention

[0009] According to the present invention, a gallium nitride single crystal substrate free of a process-affected layer can be provided.

Brief Description of the Drawings

[0010] [Figure 1] Fig. 1 is a flow chart showing an example of a method for manufacturing a nitride semiconductor substrate 100 according to a first embodiment of the present invention. [Figure 2] Fig. 2 is a schematic diagram for explaining a CMP processing step S102 according to the first embodiment of the present invention. [Figure 3] Fig. 3 is a schematic diagram for explaining a process-affected layer evaluation step S103 according to the first embodiment of the present invention. [Figure 4] Figure 4 is a schematic diagram of a spectrophotometer with an integrating sphere (with a reflector). [Figure 5] Figure 5 is a schematic diagram of a spectrophotometer with an integrating sphere (without a reflector). [Figure 6] Figure 6 is a schematic diagram illustrating the CMP machining process S102 according to another embodiment of the present invention. [Figure 7] Figure 7 shows a cathodoluminescence (CL) image of the main surface (before CMP processing) of Sample 1 according to an embodiment of the present invention. [Figure 8] Figure 8 is a CL image of the main surface (after CMP processing) of Sample 1 according to an embodiment of the present invention. [Figure 9] Figure 9 is a CL image of the main surface (after CMP processing) of Sample 2 according to an embodiment of the present invention. [Figure 10] Figure 10 is a CL image of the main surface (after CMP processing) of Sample 3 according to an embodiment of the present invention. [Figure 11] Figure 11 is a CL image of the main surface (after CMP processing) of sample 4 according to an embodiment of the present invention. [Figure 12] Figure 12 schematically shows the CL image (left side of the page) and the SEM backscattered electron image (right side of the page) of the main surface (before CMP processing) of Sample 1 according to an embodiment of the present invention. [Figure 13] Figure 13 schematically shows the CL image (left side of the page) and the SEM backscattered electron image (right side of the page) of the main surface (after CMP processing) of Sample 1 according to an embodiment of the present invention. [Figure 14] Figure 14 is a graph showing the results of calculating the light diffusion transmittance Rd for samples 1 to 4 according to the embodiment of the present invention using a spectrophotometer. [Figure 15] Figure 15 is a graph with the vertical axis of Figure 14 adjusted. [Figure 16] Figure 16 is a schematic diagram illustrating the configuration of a conventional CMP (Chemical Modulation) process. [Modes for carrying out the invention]

[0011] <Insights gained by the inventor> First, let me explain the findings the inventors have gained.

[0012] When attempting to evaluate the processed and altered layer of a gallium nitride single crystal substrate (hereinafter also referred to as a GaN substrate) by irradiating it with laser light from an oblique direction to the main surface of the substrate, as described in Patent Document 1, it was found that the evaluation is strongly affected by particles on the substrate surface with diameters ranging from submicron to micrometers, as well as surface irregularities, making it impossible to accurately evaluate the processed and altered layer.

[0013] Furthermore, Patent Document 2 does not describe how to completely eliminate the processed altered layer, that is, how to reduce it using precision grinding and CMP, even though it is possible to reduce it. When CMP is performed on a GaN substrate to remove the processed altered layer, a new processed altered layer is generated by the CMP, making it difficult to completely remove the processed altered layer. In particular, when polishing the c-plane of a GaN substrate, there is a problem in that the processed altered layer is easily introduced.

[0014] The inventor conducted intensive research on the aforementioned problem and discovered that by reducing the contact area between the CMP pad and the GaN substrate, the processed and altered layer of the GaN substrate could be completely removed.

[0015] Typically, in CMP (Chemical Polymer Processing), the entire surface of the crystal to be processed (the main surface) is brought into contact with a CMP pad attached to a flat platen, and the CMP process is carried out by the rotation of both the platen and the crystal. In other words, the contact area between the CMP pad and the substrate is equal to the area of ​​the main surface. At this time, it has been found that if there are multiple dislocations (penetrating dislocations) that penetrate the main surface of the crystal, new crystal defects will occur between these penetrating dislocations.

[0016] Figure 16 is a schematic diagram illustrating the configuration of a conventional CMP machining process. As shown in Figure 16, in a conventional CMP machining process, the main surface of the workpiece 10 and the upper surface of the workpiece platen 20 are parallel, so the entire surface of the main surface is in contact with the CMP pad 21. With this configuration, through dislocations exceed a predetermined density (for example, if the through dislocation density is 10 6 cm -2 When CMP processing is performed on the main surface of a GaN crystal (to a certain extent), new crystal defects are generated between the threading dislocations.

[0017] In contrast, by reducing the contact area between the CMP pad and the GaN substrate according to the density of through-dislocations in the crystal (for example, by making it smaller than the area of ​​the main surface of the GaN substrate), CMP processing can be performed without generating crystal defects between through-dislocations. This processing method makes it possible to realize a GaN substrate without a processed altered layer.

[0018] [Details of the Embodiments of the Invention] Next, one embodiment of the present invention will be described below with reference to the drawings. However, the present invention is not limited to these examples and is intended to include all modifications within the meaning and scope equivalent to the claims as shown in the claims.

[0019] <First Embodiment of the Invention> (1) Method for manufacturing nitride semiconductor substrate 100 First, the manufacturing method of the nitride semiconductor substrate 100 (also called GaN substrate 100) of this embodiment will be described. Figure 1 is a flowchart showing an example of the manufacturing method of the nitride semiconductor substrate 100 of this embodiment. As shown in Figure 1, the manufacturing method of the nitride semiconductor substrate 100 of this embodiment includes, for example, a workpiece preparation step S101, a CMP processing step S102, and a processed altered layer evaluation step S103.

[0020] (Workpiece preparation step S101) The workpiece preparation step S101 is a step of preparing a workpiece 10 made of a nitride semiconductor such as gallium nitride. In this embodiment, we will describe the case where the workpiece 10 is a gallium nitride single crystal intermediate for manufacturing a gallium nitride single crystal substrate (GaN substrate). Specifically, for example, gallium nitride can be epitaxially grown by HVPE (Hydride Vapor Phase Epitaxy) to obtain a bulk crystal of gallium nitride, and then the workpiece 10 can be prepared by performing appropriate slicing, grinding, polishing, etc. The diameter of the workpiece 10 is, for example, 2 inches or more and 8 inches or less (50 mm or more and 200 mm or less). In this embodiment, the main surface of the workpiece 10 to be processed by CMP is, for example, the c-plane. Here, the c-plane is {0001} as the index plane and the c-axis is

[0001] .

[0021] The main surface of the workpiece 10 is machined flat by mechanical polishing or precision grinding, and the surface roughness Ra of the main surface is, for example, 10 nm or less. However, a processed altered layer is introduced to the main surface of the workpiece 10 during slicing or grinding / polishing, and in order to manufacture a high-performance device, it is necessary to completely remove the processed altered layer.

[0022] (CMP processing step S102) The CMP process S102 is a process in which chemical mechanical polishing (CMP) is performed on the main surface of the workpiece 10, for example. This process makes it possible to completely remove the processed altered layer introduced on the main surface of the workpiece 10.

[0023] FIG. 2 is a schematic diagram for explaining the CMP processing step S102 of the present embodiment. As shown in FIG. 2, in this step, for example, the workpiece 10 is rotated about the center of the main surface of the workpiece 10 as a rotation axis, and is brought into contact with a chemical mechanical polishing pad (CMP pad 21) mounted on a surface plate 20 of a chemical mechanical polishing apparatus (CMP apparatus). Any known materials for the CMP pad 21 and chemical mechanical polishing slurry used herein can be used. The slurry referred to herein is a substance in which abrasive grains are dispersed in a liquid called a lubricant. When the GaN crystal serving as the workpiece 10 comes into contact with the CMP pad 21 via the slurry, the lubricant oxidizes the surface of the GaN crystal through a chemical reaction to form gallium oxide, and the abrasive grains mechanically polish and remove the gallium oxide. In this case, by selecting a material for the abrasive grains that is harder than gallium oxide and softer than GaN, only gallium oxide is mechanically polished and removed, so that the damaged layer can be removed while preventing the GaN crystal itself from being polished. That is, it is preferable to select abrasive grains such that the hardness relationship satisfies gallium oxide < abrasive grains < GaN crystal, and colloidal silica or the like can be used as a specific abrasive grain material.

[0024] As shown in Fig. 2, in the CMP processing step S102 of the present embodiment, for example, a conical surface plate 20 is used, and the CMP pad 21 also has a conical shape conforming to the surface plate 20. A work holder 11 that supports the workpiece 10 is, for example, in the shape of a flat disk, the surface of the work holder 11 is in contact with the surface opposite to the main surface of the workpiece 10, and the main surface and the surface of the work holder 11 are substantially parallel. A work rotation shaft 12 is provided substantially at the center of the surface of the work holder 11 opposite to the surface on which the workpiece 10 is mounted, the rotation shaft extends in the normal direction to the surface of the work holder 11, and is substantially parallel to the normal direction of the conical surface on which the CMP pad 21 is mounted. A surface plate rotation shaft 22 is provided substantially at the center of the lower surface of the conical surface plate 20. With such a configuration, when the main surface of the workpiece 10 is brought into contact with the CMP pad 21, the contact area is not the entire main surface of the workpiece 10, but is smaller than the area of the main surface. Note that the entire main surface can be processed by CMP by rotating the workpiece 10. Performing CMP processing in this state can suppress the generation of a new damaged layer (crystal defects) during CMP processing, so that the damaged layer can be completely removed.

[0025] The contact area S between the main surface of the workpiece 10 and the CMP pad 21 c can be adjusted by the shape, size, and material of the CMP pad 21, the pressure with which the workpiece 10 is pressed against the CMP pad 21, and other factors. In the CMP processing step S102, it is preferable to control the contact area S c according to the threading dislocation density (TDD) on the main surface of the workpiece 10. The higher the threading dislocation density TDD is, the more easily new crystal defects are generated between threading dislocations, so it is preferable to perform CMP processing while controlling the contact area S c to be small. This further suppresses the generation of a new damaged layer (crystal defects) during CMP processing.

[0026] Specifically, for the contact area S between the main surface of the workpiece 10 and the CMP pad 21 c , when the threading dislocation density TDD on the main surface is 1×10 7 cm -2In the above cases, the area of ​​the main surface S0 shall be 10% or less, and the TDD shall be 1 × 10 6 cm -2 The above 1 x 10 7 cm -2 If it is less than 10% of S0 and 30% or less, TDD is 1 × 10 5 cm -2 The above 1 x 10 6 cm -2 If it is less than 30% of S0 and 50% or less, TDD is 1 × 10 5 cm -2 If it is less than 50%, it is preferable to set it to more than 50% but less than 100% of S0. In this way, the contact area S depends on the through-dislocation density TDD of the main surface. c By controlling this, the processed altered layer can be removed uniformly in the plane, improving the in-plane uniformity of the GaN substrate 100.

[0027] The CMP machining process S102 completely removes the processed altered layer from the main surface of the workpiece 10 and improves the flatness of the main surface. Specifically, after the CMP machining process S102, the main surface of the workpiece 10 (GaN substrate 100) is mirror-polished to a surface roughness Ra of 1 nm or less.

[0028] (Processing and alteration layer evaluation process S103) The processing alteration layer evaluation step S103 is a step to evaluate the processing alteration layer on the main surface of the workpiece 10 (GaN substrate 100), for example. In other words, it can be rephrased as a step to confirm whether the processing alteration layer on the main surface of the GaN substrate 100 has been completely removed.

[0029] Paragraph 0002 of Patent Document 1 defines the processed altered layer as "a surface layer having crystal distortions, scratches, etc., introduced during slicing, grinding, or polishing." Of these, the processed altered layer that is particularly problematic in GaN crystals is the part other than the elastic deformation region of the crystal lattice, where crystal distortion is considered the elastic deformation region; it is the disorder of single crystal structure introduced by the crystal processing process, i.e., the plastic deformation region of the crystal lattice. Furthermore, the uneven shape of the substrate surface and the processed altered layer are different phenomena, and even if there is stress (strain) inherent in the crystal or unevenness as a geometric shape, if the plastic deformation region of the crystal lattice introduced by the crystal processing process is removed, there will be no problem in subsequent processes. However, with the laser light reflection scattering evaluation method of Patent Document 1, it is difficult to separate the uneven shape of the substrate surface from the processed altered layer.

[0030] Furthermore, Patent Document 2 discloses that the processed altered layer can be removed by performing CMP after precision grinding using a high-grit grinding wheel, and states that by appropriately selecting the abrasive particle size of the grinding wheel during precision grinding, it is possible to shorten the processing time, including the CMP process. Here, CL is used to evaluate the processed altered layer, and it is stated that the processed altered layer consists of a main altered layer and latent defects. It is not described how the main altered layer is observed with CL, but it is explained that latent defects can be observed as black lines (dark lines). As mentioned above, Patent Document 2 states that the processing time can be shortened by grinding with an appropriate abrasive particle size and CMP, but it does not describe the process until the latent defects disappear, meaning that the processed altered layer is not completely removed. Furthermore, paragraph 0042 of Patent Document 2 states that it is difficult to guarantee that there are no latent defects over a wide area across the entire substrate with CL. In addition, Patent Document 2 states that latent defects constituting the processed altered layer can be detected with CL, but in reality, there are cases where they cannot be detected. CL utilizes the property that GaN crystals emit light when irradiated with an electron beam. If there are linear or planar crystal defects such as dislocations or stacking faults (including crystal defects in the processed layer introduced by the crystal processing process), those regions will not emit light and will be observed as dark lines or dark areas. However, if the processed layer is extremely fine, or if it is located in a very shallow region of the crystal surface compared to the electron beam penetration depth, the dark lines or dark areas may be obscured by the emission from the matrix phase of the crystal surrounding the processed layer or in regions deeper than the processed layer, resulting in a difference in contrast that makes detection impossible. In such cases, nothing may be found in the CL surface observation even though a processed layer is present. When the electron beam acceleration voltage during CL observation is 5, 10, or 15 kV, the electron penetration depth into the GaN crystal is 150, 470, or 900 nm, respectively.As observed in the examples described later, processed alteration layers that are equal to or deeper than the electron penetration depth can be detected by surface observation of CL. However, for the reasons mentioned above, it is difficult to observe processed alteration layers in regions that are significantly shallower. Therefore, even with an acceleration voltage of 5kV, which is the shallowest electron penetration depth, it is difficult to observe non-luminescent components (processed alteration layers, etc.) located at depths shallower than 150nm from the surface. Furthermore, if the acceleration voltage is too low (for example, less than 5kV), the resolution decreases, making CL observation itself difficult.

[0031] Another problem is that when CL observation is performed after CMP to evaluate the processed and altered layer, electron beam irradiation causes mainly carbon-based deposits to form on the crystal surface, requiring cleaning to remove them. In other words, although Patent Document 2 describes cleaning after CMP, if CL observation is performed afterward, further cleaning to remove carbon-based deposits becomes necessary.

[0032] Paragraphs 0030 to 0032 of Patent Document 1 disclose methods for removing particles adhering to the substrate surface by various cleaning processes. However, these cleaning methods do not always completely remove particles adhering to the GaN crystal surface generated during the GaN crystal processing process, or carbon-based deposits generated during CL observation. In particular, cleaning using ammonia water or hydrogen peroxide water as disclosed in paragraph 0032 of Patent Document 1 results in etching of the nitrogen polar surface (N-surface) on the back side, opposite to the c-surface which is generally used as the main surface, causing the desired shape to be distorted, thus requiring mirror polishing of the back side again.

[0033] The objective of this invention is to provide a gallium nitride single crystal substrate without a processed alteration layer. This involves establishing a technology that enables the evaluation of the presence or absence of a processed alteration layer over a wide area with high depth resolution, which has been difficult in the past, while simultaneously achieving complete removal of the processed alteration layer.

[0034] As a result of the inventor's diligent research, it was discovered that the processed and altered layer of a GaN substrate can be accurately evaluated by using a spectrophotometer to illuminate the substrate with light (e.g., white light) from approximately the normal direction to the main surface of the GaN substrate, and measuring the light diffusion transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light.

[0035] As mentioned above, in the case of reflection and scattering evaluation using laser light, there is a problem in that particles attached to the surface and the surface irregularities are also acquired as data. In contrast, in the measurement by spectral analysis of transmitted light using white light in this method, these influences are small, and it has been found to be suitable for evaluating the processed altered layer. Furthermore, it has been found that the processed altered layer can also be evaluated by irradiating with light of a specific wavelength other than white light and calculating the light diffusion transmittance Rd. In addition, the processing altered layer evaluation of the present invention has the advantage that it does not require cleaning of the gallium nitride single crystal substrate after evaluation, and a clean surface can be maintained.

[0036] Figure 3 is a schematic diagram illustrating the processing and alteration layer evaluation process S103. As shown in Figure 3, in the processing and alteration layer evaluation process S103 of this embodiment, light is incident perpendicular to the main surface of the GaN substrate 100 using a spectrophotometer. The irradiation area of ​​the light is preferably, for example, φ10 mm or more. The light that is transmitted through the GaN substrate 100 without diffusion is defined as forward transmitted light, and the light that is transmitted through the GaN substrate 100 after diffusion is defined as diffuse transmitted light. The light diffusion transmittance Rd is calculated by dividing the intensity of the diffuse transmitted light by the intensity of the total transmitted light (forward transmitted light + diffuse transmitted light). That is, if there are crystal defects in the crystal that are of a size that causes scattering with respect to the wavelength of the incident light, the light diffusion transmittance Rd obtained using this spectrophotometer will be large. In this specification, perpendicular to the main surface of the GaN substrate 100 includes not only being perfectly perpendicular to the main surface, but also being slightly off by an angle (for example, 5 degrees or less) from perpendicular.

[0037] The method for measuring the light diffusion transmittance Rd will be explained in more detail. Figures 4 and 5 are schematic diagrams of a spectrophotometer having an integrating sphere. As shown in Figure 4, when the exit of the integrating sphere 110 is blocked by the reflector 120, the total transmitted light that has passed through the GaN substrate 100 is detected after multiple reflections within the integrating sphere 110, so the intensity of the total transmitted light can be measured. On the other hand, as shown in Figure 5, when the reflector 120 is absent, only the positive transmitted light is emitted from the integrating sphere 110, so the intensity of the diffuse transmitted light can be measured. The value obtained by dividing the intensity of the diffuse transmitted light by the intensity of the total transmitted light is the light diffusion transmittance Rd.

[0038] The more processed and altered layers there are, the higher the proportion of diffusely transmitted light. Therefore, a low light diffusion transmittance Rd indicates a substrate with few processed and altered layers. Through diligent research by the inventor, it has been found that if the light diffusion transmittance Rd satisfies the following equation (1), the processed and altered layer on the main surface can be considered to have been completely removed. Rd <a×b 2 ÷(c × (λ - λ0) 2 +b 2 )+d ···(1) However, in equation (1), a=10, b=1.2, c=2.5, d=0.00065, λ is the wavelength of the incident light (nm), and λ0=330.

[0039] In other words, if there are scatterers (crystal defects or inclusions) within the crystal that are about 1 / 10th the size of the incident light wavelength λ, Rd will increase, and the size of these scatterers in the processed altered layer can be said to be around tens to 100 nm. To put it another way, if equation (1) is satisfied, it can be said that there are no scatterers larger than this size, and in fact, when a diode device was fabricated by epitaxial growth on this crystal, a good device with low leakage current was obtained. On the other hand, when the same fabrication was performed on a crystal that did not satisfy equation (1), the device had a large leakage current.

[0040] In the processed altered layer evaluation step S103, for example, it is sufficient to confirm that there is at least one incident light wavelength λ in the range of 390 nm to 600 nm such that the light diffusion transmittance Rd satisfies equation (1) above. Furthermore, when the incident light wavelength λ (nm) is an integer value, it is preferable that the light diffusion transmittance Rd satisfies equation (1) above for more than half of the λ values ​​in the range of 390 nm to 600 nm. Moreover, it is even more preferable that the light diffusion transmittance Rd satisfies equation (1) above for the entire range of incident light wavelength λ from 390 nm to 600 nm. If the light diffusion transmittance Rd does not satisfy equation (1) above, the CMP processing step S102 should be repeated, and CMP processing should be performed under appropriate conditions until the processed altered layer is completely removed.

[0041] Furthermore, the status of the remaining or removed processed altered layer on the main surface can also be confirmed, for example, by performing CL observation on a cross-section of the GaN substrate 100. When a cross-section of the GaN substrate 100, which has been cleaved in a way that does not damage the main surface, is observed using CL, if a processed altered layer is present, a layered dark area will be observed near the main surface. In other words, if a dark area is not present when observing a cross-section of the GaN substrate 100 using CL, it can be said that the processed altered layer has been removed. With cross-sectional CL observation, the aforementioned problem of electron penetration depth is not present, and the degree of removal of the processed altered layer can be evaluated by whether or not a layered dark area is present near the surface (main surface). However, these cross-sectional CL observations are destructive tests, and furthermore, the size of the area that can be observed is limited by the magnification of the scanning electron microscope (SEM), making it difficult to evaluate the entire GaN crystal in a short time. In addition, when the processed altered layer is evaluated using CL observation, there is the disadvantage that cleaning to remove carbon-based deposits is required after the evaluation. In contrast, the evaluation of the processed altered layer using the light diffusion transmittance of the present invention has a wide evaluation area, allows for non-destructive evaluation of the entire substrate surface, eliminates problems with depth resolution, and avoids surface contamination.

[0042] The processed and altered layer may be evaluated at any multiple points on the main surface. If the deviation ((maximum value - minimum value) ÷ average value) when measuring the light diffusion transmittance Rd at any multiple points on the main surface is 5% or less, it can be said that the processed and altered layer has been uniformly removed within the plane. Note that the above deviation may be calculated using the average value of the light diffusion transmittance Rd in the range of incident light wavelength λ from 390 nm to 600 nm, or it may be calculated using the light diffusion transmittance Rd corresponding to any λ in the range of 390 nm to 600 nm.

[0043] Through the above process, a GaN substrate 100 from which the processed and altered layer has been completely removed can be manufactured. The GaN substrate 100 manufactured by the manufacturing method of this embodiment is particularly suitable for manufacturing high-performance devices.

[0044] (2) Nitride semiconductor substrate 100 Next, the nitride semiconductor substrate 100 (GaN substrate 100) of this embodiment will be described. The GaN substrate 100 is, for example, a gallium nitride single crystal substrate having a diameter of 50 mm or more and two main surfaces facing opposite directions. The two main surfaces of the GaN substrate 100 are mirror-polished to a surface roughness Ra of 1 nm or less by the CMP processing step S102.

[0045] The GaN substrate 100 is in a state where the processed altered layer on the main surface has been completely removed by the CMP processing step S102. To confirm that the processed altered layer has been completely removed, it is sufficient to confirm in the processed altered layer evaluation step S103 that the light diffusion transmittance Rd satisfies the above-mentioned conditions. Furthermore, the fact that the light diffusion transmittance Rd satisfies the above-mentioned conditions also means that there are no inclusions inside the crystal. That is, the crystal interior of the GaN substrate 100 does not contain any inclusions that can be detected by transmission microscopy observation using at least one of ultraviolet light, visible light, and infrared light with a wavelength longer than the band edge wavelength of the crystal. When confirming the absence of inclusions by transmission microscopy observation, it is preferable to observe with a field of view of φ10 mm or larger.

[0046] Furthermore, the complete removal of the processed altered layer may be confirmed by cross-sectional CL observation. For example, when observing the cross-section obtained by cleaving the GaN substrate 100 with CL, it may be confirmed that no dark area corresponding to the processed altered layer is observed in the region from the main surface up to 150 nm.

[0047] <Other embodiments of the present invention> Although embodiments of the present invention have been specifically described above, the present invention is not limited to the embodiments described above, and various modifications are possible without departing from the spirit of the invention.

[0048] For example, in the above embodiment, the case in the processed altered layer evaluation step S103 was described in which the light diffusion transmittance Rd is measured by changing the wavelength λ of the incident light in the range of 390 nm to 600 nm. However, in the processed altered layer evaluation step S103, the light diffusion transmittance Rd may be measured for a specific wavelength, and the processed altered layer may be evaluated based on whether or not it satisfies the above equation (1). Specifically, when light with a wavelength of 532 nm is incident perpendicularly to the main surface of the GaN substrate 100, if the light diffusion transmittance Rd satisfies Rd < 0.000791, it may be considered that the processed altered layer on the main surface has been completely removed. Using this method makes it possible to evaluate the processed altered layer more simply.

[0049] Figure 6 is a schematic diagram illustrating a CMP machining process S102 according to another embodiment of the present invention. As shown in Figure 6, in the CMP machining process S102, the CMP pad 21 may be attached to the surface (side) of the cylinder (or column) of the cylindrical (or column) shaped surface plate 20. The work holder 11 is a flat disc shape, similar to the first embodiment described above, and the work rotation axis 12 is substantially parallel to the direction normal to the side surface of the cylinder (or column) on which the CMP pad 21 is attached. With this configuration, similar to the first embodiment, the area (contact area) when the main surface of the workpiece 10 is in contact with the CMP pad 21 can be made smaller than the area of ​​the main surface, thereby suppressing the generation of a new processed altered layer (crystal defect) during CMP machining and making it possible to completely remove the processed altered layer.

[0050] Furthermore, although the above-described embodiment described the case in which the CMP processing step S102 is performed on both main surfaces of the workpiece 10 (the Ga surface of the +c surface and the N surface of the -c surface), the N surface of the GaN crystal is easily etched by chemical etching, so even if polishing is performed by conventionally known methods (for example, CMP processing using an alkaline slurry with a conventional CMP apparatus as shown in Figure 16), it is possible to remove the processed altered layer. Therefore, the GaN substrate 100 according to the present invention also includes cases in which, even if the N surface is in an uneven, matte state, it is mirror-polished by conventionally known methods so that the surface roughness Ra is 1 nm or less, and as a result, in the processed altered layer evaluation step S103, the light diffusion transmittance Rd satisfies the above-described conditions.

[0051] Furthermore, the processed altered layer evaluation step S103 described in the above embodiment can also be applied as an evaluation method for a GaN substrate to evaluate the presence or absence of a processed altered layer. [Examples]

[0052] Next, embodiments of the present invention will be described. These embodiments are examples of the present invention, and the present invention is not limited to these embodiments.

[0053] <Example 1> First, the diameter is 50 mm and the penetration dislocation density on the main surface c is 3 × 10 6 cm -2 Four workpieces made of gallium nitride single crystals were prepared and designated as Samples 1-4. Samples 1-4 were precision ground so that the surface roughness Ra of the main surface was 10 nm or less. Figure 7 shows the cathodoluminescence (CL) image of the main surface of Sample 1 (before CMP processing). CL observation was performed using a scanning electron microscope (SU5000) manufactured by Hitachi High-Tech Corporation. As shown in Figure 7, dark areas and dark lines were present on the main surface of the workpiece, confirming the introduction of a processed altered layer.

[0054] Next, CMP machining was performed on both main surfaces of samples 1 to 4 until the surface roughness Ra was 1 nm or less. For sample 1, CMP machining was performed with the configuration shown in Figure 2 as described in the first embodiment above, and the contact area between the main surface of the workpiece and the CMP pad was controlled to be 20% of the area of ​​the main surface. For samples 2 to 4, CMP machining was performed with the configuration shown in Figure 16 as described as a conventional example, that is, a configuration in which the entire main surface of the workpiece is in contact with the CMP pad.

[0055] CL observation was performed on the main surfaces of samples 1-4 after CMP processing. Figure 8 shows the CL image of the main surface of sample 1 (after CMP processing), Figure 9 shows the CL image of the main surface of sample 2 (after CMP processing), Figure 10 shows the CL image of the main surface of sample 3 (after CMP processing), and Figure 11 shows the CL image of the main surface of sample 4 (after CMP processing). As shown in Figure 8, no dark areas or dark lines were observed on the main surface of sample 1, which was processed using CMP while controlling the contact area between the main surface of the workpiece and the CMP pad to be smaller than the area of ​​the main surface. On the other hand, as shown in Figures 9-11, dark areas and dark lines were present on the main surfaces of samples 2-4, which were processed using a conventional configuration, indicating that the processed altered layer was not completely removed. Note that the dark spots (scotoma) observed in Figures 8-11 are non-luminescent regions corresponding to penetrating dislocations. The more processed altered layer there is, i.e., the more dark lines and dark areas there are, the more difficult it becomes to obtain contrast between these dark spots and luminescent regions in the CL image. Therefore, even with crystals having a similar penetration dislocation density, dark spots are not observed in Figure 7 before CMP processing. The contrast between the dark spots and the luminescent regions in Figures 8 to 11, i.e., the ease of observing the dark spots, is correlated with the light diffusion transmittance in the present invention, which will be discussed later.

[0056] Furthermore, CL observation and SEM observation were performed on the cross-section of Sample 1 before and after CMP processing, using the same field of view. However, the cross-sectional sample before CMP processing was a separate sample that had undergone the same pre-processing (precision grinding) as Sample 1 for the CMP process, and it was confirmed in advance that the occurrence of the processed altered layer after this pre-processing was reproducible. Figure 12 schematically shows the CL image (left) and SEM image (right) of a separate sample similar to Sample 1 before CMP processing, and Figure 13 schematically shows the CL image (left) and SEM image (right) of Sample 1 after CMP processing. The SEM image shows the physical boundary region, and the CL image shows the boundary region between the luminescent region without processing-induced crystal defects and the non-luminescent region (processed altered layer) with processing-induced crystal defects. Therefore, the difference in the boundary regions between the CL image and the SEM image is the thickness of the processed altered layer. As shown in Figure 12, in the separate sample similar to Sample 1 before CMP processing, the difference in the boundary regions between the CL image and the SEM image was 0.3 μm. In contrast, as shown in Figure 13, the boundary regions of the CL image and the SEM image coincided in Sample 1 after CMP processing. Based on the results of surface CL observation and cross-sectional CL observation, it can be said that the processed altered layer was completely removed in Sample 1 after CMP processing.

[0057] The light diffusion transmittance Rd was calculated using a spectrophotometer for Sample 1 before CMP processing and for Samples 1-4 after CMP processing, and the processed altered layer was evaluated. A Hitachi High-Tech UV-Vis-Near-Infrared Spectrophotometer (UH4150) was used. The measurement wavelength was 200-2500 nm, scanning from long wavelength to short wavelength, with a scan speed of 600 nm / min and a sampling interval of 1 nm. The results are shown in Figures 14 and 15. Figures 14 and 15 also show graphs illustrating the right-hand side of the following equation (1). Rd <a×b 2 ÷(c × (λ - λ0) 2 +b 2 )+d ···(1) However, in equation (1), a=10, b=1.2, c=2.5, d=0.00065, λ is the wavelength of the incident light (nm), and λ0=330.

[0058] Figures 14 and 15 show the light diffusion transmittance Rd for the measured wavelength range of 350 to 750 nm. In Figure 14, the vertical axis is adjusted to show the wavelength dependence of the light diffusion transmittance Rd of sample 1 before CMP processing, while in Figure 15, the light diffusion transmittance Rd of sample 1 before CMP processing is not shown, but the vertical axis is adjusted to show the differences between samples 1 to 4 after CMP processing. As shown in Figure 15, sample 1, which underwent CMP processing by controlling the contact area between the main surface of the workpiece and the CMP pad to be smaller than the area of ​​the main surface, satisfied equation (1) above for the light diffusion transmittance Rd in the entire range of incident light wavelength λ from 390 nm to 600 nm. On the other hand, samples 2 to 4, which underwent CMP processing using a conventional configuration, did not satisfy equation (1) above for the light diffusion transmittance Rd in the entire (or most) range of incident light wavelength λ from 390 nm to 600 nm. Furthermore, only sample 1 satisfied equation (1) above for the light diffusion transmittance Rd for light at a wavelength of 532 nm.

[0059] From the above, we confirmed that the processed altered layer can be completely removed by controlling the contact area between the main surface of the workpiece and the CMP pad to be smaller than the area of ​​the main surface during CMP processing. Furthermore, we confirmed that the processed altered layer of the GaN substrate can be accurately evaluated by calculating the light diffusion transmittance Rd using a spectrophotometer. This method of evaluating the processed altered layer using light diffusion transmittance is non-destructive and does not involve contact with the processed main surface of the workpiece. Therefore, it is possible to evaluate the layer without the effects of surface deposits caused by scattering using laser light, as in the conventional techniques mentioned above, or problems such as surface damage and carbon deposits associated with electron beam irradiation in cathodoluminescence.

[0060] <Example 2> First, the diameter is 100 mm and the penetration dislocation density on the main surface c is 3 × 10 6 cm -2 Two workpieces made of gallium nitride single crystals were prepared and designated as Sample 5 and Sample 6. Samples 5 and Sample 6 were precision ground so that the surface roughness Ra of both main surfaces was 10 nm or less.

[0061] Next, CMP machining was performed on both main surfaces of samples 5 and 6 until the surface roughness Ra of the main surface was 1 nm or less. For sample 5, CMP machining was performed with the configuration shown in Figure 2 as described in the first embodiment above, and the contact area between the main surface of the workpiece and the CMP pad was controlled to be 20% of the area of ​​the main surface. For sample 6, CMP machining was performed with the configuration shown in Figure 16 as described as a conventional example.

[0062] For samples 5 and 6 after CMP processing, the light diffusion transmittance Rd was calculated using a spectrophotometer to evaluate the processed and altered layer. The light diffusion transmittance Rd was measured at five locations in a plan view: the center of the substrate (measurement point a), 20 mm to the right of the center of the substrate (measurement point b), 20 mm to the left of the center of the substrate (measurement point c), 20 mm above the center of the substrate (measurement point d), and 20 mm below the center of the substrate (measurement point e). Table 1 shows the light diffusion transmittance Rd at each measurement point when the incident light wavelength λ is 532 nm.

[0063] [Table 1]

[0064] As shown in Table 1, Sample 5, which underwent CMP processing with the contact area between the main surface of the workpiece and the CMP pad controlled to be 20% of the area of ​​the main surface, had a deviation of 5% or less in the light diffusion transmittance Rd ((maximum value - minimum value) ÷ average value). On the other hand, Sample 6, which underwent CMP processing using a conventional configuration, had a deviation of 5% or more in the light diffusion transmittance Rd ((maximum value - minimum value) ÷ average value). Also, the right-hand side of equation (1) for λ = 532 nm is 7.91 × 10⁻⁶. -4 Therefore, it can be seen that the relationship between the magnitudes of the light diffusion transmittance Rd is Sample 5 < Equation (1) < Sample 6.

[0065] Based on the above, we confirmed that by appropriately controlling the contact area between the main surface of the workpiece and the CMP pad during CMP processing, the deviation of the light diffusion transmittance Rd ((maximum value - minimum value) ÷ average value) can be reduced to 5% or less. This indicates that the processed altered layer is being removed uniformly within the surface.

[0066] <Preferred Embodiments of the Invention> Preferred embodiments of the present invention are described below.

[0067] (Note 1) A gallium nitride single crystal substrate having a diameter of 50 mm or more and having two main surfaces facing opposite directions, A gallium nitride single crystal substrate is provided, in which, when light with a wavelength of 532 nm is incident perpendicularly to one of the main surfaces, the light diffusion transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light, satisfies Rd < 0.000791.

[0068] (Note 2) A gallium nitride single crystal substrate having a diameter of 50 mm or more and having two main surfaces facing opposite directions, A gallium nitride single crystal substrate is provided, wherein, when light is incident perpendicularly to one of the main surfaces using a spectrophotometer, at least one point exists in the range of 390 nm to 600 nm where the optical diffuse transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light, satisfies the following equation (1). Rd <a×b 2 ÷(c × (λ - λ0) 2 +b 2 )+d ···(1) [In equation (1), a=10, b=1.2, c=2.5, d=0.00065, λ is the wavelength of the incident light (nm), and λ0=330.]

[0069] (Note 3) The gallium nitride single crystal substrate described in Appendix 2, When the wavelength λ (nm) of the incident light is an integer value, the light diffusion transmittance Rd satisfies equation (1) for more than half of the λ values ​​in the range of 390 nm to 600 nm.

[0070] (Note 4) The gallium nitride single crystal substrate described in Appendix 2, In the entire range of the incident light wavelength λ from 390 nm to 600 nm, the light diffusion transmittance Rd satisfies formula (1).

[0071] (Note 5) A gallium nitride single crystal substrate as described in Appendix 1 or Appendix 2, For any multiple points on the main surface, the deviation ((maximum value - minimum value) ÷ average value) when measuring the light diffusion transmittance Rd is 5% or less.

[0072] (Note 6) A step to prepare a gallium nitride single crystal intermediate having a surface roughness Ra of 10 nm or less and a main surface in which a processed altered layer exists, The process includes a step of performing chemical mechanical polishing on the main surface, A method for manufacturing a gallium nitride single crystal substrate is provided, in which the chemical mechanical polishing step is made smaller than the area of ​​the main surface. Preferably, the contact area S depends on the through-dislocation density TDD of the main surface. c Control. Preferably, the contact area S c The penetrating dislocation density (TDD) is 1 × 10 7 cm -2 In the above cases, the area S0 of the main surface shall be 10% or less, and the TDD shall be 1 × 10 6 cm -2 The above 1 x 10 7 cm -2 If it is less than 10% of S0 and 30% or less, TDD is 1 × 10 5 cm -2 The above 1 x 10 6 cm -2 If it is less than 30% of S0 and 50% or less, TDD is 1 × 10 5 cm -2 If it is less than 50%, it will be considered to be more than 50% but less than 100% of S0.

[0073] (Note 7) A method for evaluating a gallium nitride single crystal substrate is provided, which evaluates the presence or absence of a processed altered layer by determining whether the optical diffuse transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light when light with a wavelength of 532 nm is incident perpendicularly to the main surface of the gallium nitride single crystal substrate, satisfies Rd < 0.000791. Preferably, the presence or absence of a processed altered layer is evaluated by determining whether there is at least one point in the range of 390 nm to 600 nm where the wavelength λ of incident light satisfies the following equation (1) for the light diffusion transmittance Rd. Rd <a×b 2 ÷(c × (λ - λ0) 2 +b 2 )+d ···(1) [In equation (1), a=10, b=1.2, c=2.5, d=0.00065, λ is the wavelength of the incident light (nm), and λ0=330.] Preferably, when the wavelength λ (nm) of the incident light is an integer value, the presence or absence of a processed altered layer is evaluated by determining whether the light diffusion transmittance Rd satisfies formula (1) for more than half of the λ values ​​in the range of 390 nm to 600 nm. Preferably, the presence or absence of a processed altered layer is evaluated by determining whether the light diffusion transmittance Rd satisfies formula (1) in the entire range of incident light wavelength λ from 390 nm to 600 nm.

[0074] (Note 8) A gallium nitride single crystal substrate having a diameter of 50 mm or more and having two main surfaces facing opposite directions, A substrate is provided in which, when the cross-section obtained by cleaving the gallium nitride single crystal substrate is observed with CL, no dark area corresponding to the processed altered layer is observed in the region from the main surface up to 150 nm. [Explanation of symbols]

[0075] 10 Workpiece 11 Work holder 12 Workpiece rotation axis 20 Surface plate 21 CMP pads 22. Rotating shaft of the surface plate 100 Nitride semiconductor substrates (GaN substrates) 110 Integrating Sphere 120 Reflector S101 Workpiece preparation process S102 CMP processing process S103 Processed and altered layer evaluation process

Claims

1. A gallium nitride single crystal substrate having a diameter of 50 mm or more and having two main surfaces facing opposite directions, A gallium nitride single crystal substrate in which, when light with a wavelength of 532 nm is incident perpendicularly to one of the main surfaces, the light diffusion transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light, satisfies Rd < 0.000791.

2. A gallium nitride single crystal substrate having a diameter of 50 mm or more and having two main surfaces facing opposite directions, A gallium nitride single crystal substrate in which, when light is incident perpendicularly to one of the main surfaces using a spectrophotometer, there is at least one point in the range of 390 nm to 600 nm where the light diffusion transmittance Rd, which is the value obtained by dividing the intensity of diffusely transmitted light by the intensity of total transmitted light, satisfies the following equation (1). Red<a×b ２ ÷(c×(λ-λ ０ ) ２ +b ２ )+O・・・(1) [In equation (1), a = 10, b = 1.2, c = 2.5, d = 0.00065, λ is the wavelength of the incident light (nm), λ ０ Let's assume it equals 330.

3. The gallium nitride single crystal substrate according to claim 2, wherein the light diffusion transmittance Rd satisfies formula (1) in the entire range of the incident light wavelength λ from 390 nm to 600 nm.

4. The gallium nitride single crystal substrate according to claim 1 or claim 2, wherein the deviation ((maximum value - minimum value) ÷ average value) when measuring the light diffusion transmittance Rd at a plurality of arbitrary points on the main surface is 5% or less.

Images