Method for accurately detecting dislocation defect distribution of aluminum nitride single crystal

Abstract

According to the method for accurately detecting dislocation defect distribution of the aluminum nitride single crystal, the nitrogen surface and the aluminum surface of the aluminum nitride single crystal are subjected to etching and corrosion treatment respectively, image recognition analysis is combined, and distinguishing and distribution statistics of blade dislocation TED, screw dislocation TSD and base plane dislocation BPD are achieved. Compared with a traditional method which only depends on a single corrosion morphology, the double-sided comparison method in the application can reduce the misjudgment rate of the dislocation defect to 5% or below.

Description

Technical Field

[0001] The invention belongs to the technical field of semiconductor material defect analysis and relates to a method for accurately detecting the distribution of dislocation defects in an aluminum nitride single crystal. Background Art

[0002] Aluminum nitride single crystals, with their wide bandgap and high thermal conductivity, are ideal substrate materials for high-performance optoelectronic devices. However, during the growth process, dislocation defects such as edge dislocations (TEDs), screw dislocations (TSDs), and basal plane dislocations (BPDs) can occur, significantly degrading device performance.

[0003] Currently, transmission electron microscopy (TEM) is commonly used to detect dislocation defects in aluminum nitride single crystals. However, this detection is costly, inefficient, and difficult to achieve large-scale statistical analysis. Patent publication number CN107829144A discloses a wet etching process for aluminum nitride crystals. This method does not adequately clean the etching residue on the sample surface after etching, which can affect the identification of dislocations. Furthermore, this method only determines dislocations based on the size and shape of the etching pits in the image, making it difficult to distinguish between edge dislocations (TEDs) and screw dislocations (TSDs), leading to misidentification. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals, so as to solve the problem of misjudgment that is prone to occur in existing methods.

[0005] To achieve the above object, the present invention adopts the following technical solutions: The present application provides a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals, the method comprising: S01: cutting, grinding, polishing, and cleaning the aluminum nitride single crystal along the

[0001] direction to form the Al face and N face of the aluminum nitride single crystal.

[0006] A 2-4 inch diameter aluminum nitride single crystal is fixed on a metal or ceramic crystal fixing plate by solid wax or hot melt adhesive, and the crystal fixing plate is fixed on the workbench of a diamond single wire saw. Cut along the

[0001] direction of the aluminum nitride single crystal to obtain a wafer. The cut wafers are ground using diamond abrasive powder as the grinding liquid, and then polished using chemical mechanical polishing to make the surface roughness of the wafer ≤ 0.2nm, so as to reduce the influence of mechanical damage on the detection of dislocation defects.

[0007] The ground and polished wafers were ultrasonically cleaned in acetone for 10 minutes, in boiling anhydrous ethanol for 5 minutes, in a 50 vol% sulfuric acid solution at 70-80°C for 10 minutes, and finally in deionized water for 10 minutes to remove organic matter and particulate contaminants from the wafer surface, thereby obtaining the desired Al and N surfaces. The ultrasonic power was ≥100W and the ultrasonic frequency was 40kHz to avoid micro-damage to the Al and N surfaces caused by low-frequency ultrasound.

[0008] S02: At 355-400° C., the Al surface is corroded with a molten KOH-NaOH eutectic mixture to form a corroded Al surface.

[0009] Prepare a molten KOH-NaOH eutectic mixture by mixing KOH and NaOH in a 1:1 mass ratio. Completely melt the mixture at 355-400°C and use the molten KOH-NaOH eutectic mixture to etch the Al surface of the wafer for 1-5 minutes to prevent uneven etching due to partial solidification of the molten KOH-NaOH eutectic mixture, which could affect the etching results. During the etching process, a PID temperature control system is used to ensure that the temperature fluctuation of the molten KOH-NaOH eutectic mixture is ≤5°C to prevent variations in etch pit size due to temperature instability.

[0010] S03: etching the N surface with ammonia gas at 900-1100° C. to form an etched N surface.

[0011] Place the wafer in a high-temperature tubular furnace and, at atmospheric or low pressure, etch the N-side with ammonia at a temperature of 900-1100°C and a flow rate of 100-300 sccm for 20-40 minutes to form an etched N-side. During the etching process, samples are taken every 5 minutes to observe the dynamic evolution of the etch pit morphology.

[0012] S04: After cooling, cleaning, drying and dust removal, the corroded Al surface and the etched N surface are subjected to image recognition to obtain dislocation defect distribution.

[0013] The etched Al and N surfaces were cooled to room temperature in a dry N2 stream to prevent condensation. The surfaces were then cleaned in boiling anhydrous ethanol for 5 minutes, then in a 50 vol% sulfuric acid solution at 70-80°C for 10 minutes, and finally ultrasonically cleaned in deionized water three times for 10 minutes each to remove organic matter and particulate contaminants. After cleaning, the surfaces were immediately dried with high-purity N2 and dried in a vacuum oven at 80-85°C for 30 minutes.

[0014] The dried Al and N surfaces were sprayed with high-purity N2 at a pressure of 8 atm to remove dust and prevent mechanical contact damage. After high-purity N2 spraying, image recognition was performed. Etching the N surface revealed only the TSD of hexagonal pits. Therefore, the TSD density of the N surface could be determined based on image recognition of the N surface. Etching the Al surface revealed conchoidal and hexagonal pits. The BPD appeared as conchoidal pits, while the TSD and TED both appeared as hexagonal pits of varying sizes. Therefore, the BPD density, TSD, and TED density of the Al surface could be determined based on image recognition of the Al surface. By comparing the TSD density within the same dislocation region on the Al and N surfaces, the TSD and TED densities of the Al surface were determined. The TED density of the Al surface is calculated as: the total number of hexagonal pits on the Al surface minus the TSD density of the N surface.

[0015] The present invention has the following beneficial effects: This application uses etching and corrosion treatment on both the nitrogen and aluminum sides of aluminum nitride single crystals, combined with image recognition and analysis, to distinguish and analyze the distribution of edge dislocations (TEDs), screw dislocations (TSDs), and basal plane dislocations (BPDs). Compared to traditional methods that rely solely on a single etched morphology, the dual-side comparison method in this application can reduce the misjudgment rate of dislocation defects to less than 5%. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] Please refer to the attached Figure 1 , attached Figure 1 A flow chart of a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals provided in an embodiment of the present application is shown. The technical solution of the present invention is further explained and illustrated through specific embodiments below.

[0018] Example 1 The present invention provides a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals, the method comprising: S101: Fix a 2-inch diameter aluminum nitride single crystal on a metal crystal fixing plate with solid wax, and fix the crystal fixing plate on the workbench of a diamond single wire saw. Wire cut the aluminum nitride single crystal along the

[0001] direction to obtain a wafer. Use 10, 5, 2.5, 1 The cut wafers were ground with a slurry containing diamond abrasive powder, followed by chemical mechanical polishing to a surface roughness of ≤0.2 nm. The ground and polished wafers were then cleaned in acetone for 10 minutes, boiling anhydrous ethanol for 5 minutes, and 50 vol% sulfuric acid at 75°C for 10 minutes. Finally, they were ultrasonically cleaned in deionized water for 10 minutes at a power of 200 W and a frequency of 40 kHz.

[0019] S102: Prepare a molten KOH-NaOH eutectic mixture by mixing KOH and NaOH in a mass ratio of 1:1. Completely melt the mixture at 355±1°C and use the molten KOH-NaOH eutectic mixture to etch the Al electrode surface for 1-5 minutes to avoid uneven corrosion caused by partial solidification of the molten KOH-NaOH eutectic mixture, which would affect the corrosion effect. During the etching process, a PID temperature control system was used to ensure that the temperature fluctuation of the molten KOH-NaOH eutectic mixture was ≤5°C.

[0020] S103: Place the wafer in a tubular high-temperature furnace and etch the N-side for 40 minutes using ammonia gas at 900°C and a flow rate of 200 sccm at normal pressure or low pressure to form an etched N-side. During the etching process, samples are taken every 5 minutes to observe the dynamic evolution of the etch pit morphology.

[0021] S104: The etched Al and N surfaces are cooled to room temperature in a dry N2 stream to prevent condensation. They are then cleaned in boiling anhydrous ethanol for 5 minutes, then in a 50 vol% sulfuric acid solution at 75°C for 10 minutes, and finally ultrasonically cleaned in deionized water three times for 10 minutes each to remove organic matter and particulate contaminants. After cleaning, they are immediately dried with high-purity N2 and dried in a vacuum oven at 80°C for 30 minutes. The dried Al and N surfaces are then dedused using a high-purity N2 jet at 8 atm to prevent mechanical damage. After dedusing with high-purity N2, image recognition is performed.

[0022] After etching, only the hexagonal pit TSD is exposed on the N side. Based on the image recognition of the N side, the TSD density of the N side can be determined to be 1.2×10 4 cm -2 After etching, the Al surface shows shell-shaped pits and hexagonal pits. Among them, BPD is shell-shaped pits, while TSD and TED are hexagonal pits, but with different sizes. Therefore, the BPD density of the Al surface can be determined to be 4.8×10 3 cm -2 Since both TSD and TED appear as hexagonal pits and are difficult to distinguish, the hexagonal pits on the Al surface are counted together, and the total density is 3×10 4 cm -2, that is, the total density of Al surface TSD and TED is 3×10 4 cm -2 Comparing the TSD density in the same dislocation region on the Al and N surfaces, the TED density is obtained by subtraction, that is, the TED density of the Al surface = the total number of hexagonal pits on the Al surface 3×10 4 cm -2 -TSD density on the N side is 1.2×10 4 cm -2 =1.7×10 4 cm -2 .

[0023] Example 2 The present embodiment provides a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals. The steps of the method are the same as those in Example 1, except that the etching temperature of the Al surface is 380±1°C and the etching time is 3 minutes. The TSD density of the hexagonal pits on the N surface after etching is 1.3×10 4 cm -2 The BPD density of shell-shaped pits after Al surface corrosion is 4.8×10 3 cm -2 , the hexagonal pits on the Al surface are counted together, and the total density is 3.1×10 4 cm -2 Comparing the TSD density in the same dislocation region on the Al and N surfaces, the TED density is obtained by subtraction, that is, the TED density of the Al surface = the total number of hexagonal pits on the Al surface 3.1×10 4 cm -2 -TSD density on the N side is 1.3×10 4 cm -2 =1.8×10 4 cm -2 .

[0024] Example 3 The present embodiment provides a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals. The steps of the method are the same as those in Example 1, except that the etching temperature of the N surface is 1000°C and the etching time is 30 minutes. The TSD density of the hexagonal pits after etching the N surface is 1.5×10 4 cm -2 The BPD density of shell-shaped pits after Al surface corrosion is 4.9×10 3 cm -2 , the hexagonal pits on the Al surface are counted together, and the total density is 3.2×10 4 cm -2 Comparing the TSD density in the same dislocation region on the Al and N surfaces, the TED density is obtained by subtraction, that is, the TED density of the Al surface = the total number of hexagonal pits on the Al surface 3.2×10 4 cm-2 -TSD density on the N side is 1.5×10 4 cm -2 =1.7×10 4 cm -2 .

[0025] Example 4 The present embodiment provides a method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals. The steps of this method are the same as those in Example 1, except that the etching temperature of the Al surface is 400±1°C and the etching time is 1 minute; the etching temperature of the N surface is 1100°C and the etching time is 20 minutes. The TSD density of the hexagonal pits on the N surface after etching is 0.9×10 4 cm -2 The BPD density of shell-shaped pits after Al surface corrosion is 3.5×10 3 cm -2 , the hexagonal pits on the Al surface are counted together, and the total density is 2.7×10 4 cm -2 Comparing the TSD density in the same dislocation region on the Al and N surfaces, the TED density is obtained by subtraction, that is, the TED density of the Al surface = the total number of hexagonal pits on the Al surface 2.7×10 4 cm -2 -TSD density on the N side is 0.9×10 4 cm -2 =1.8×10 4 cm -2 .

[0026] Comparative Example 1 This comparative example provides a method for detecting the distribution of dislocation defects in aluminum nitride single crystals. The steps are exactly the same as those in Example 3, except that the ammonia flow rate is adjusted to 50 sccm. The density of hexagonal pits (TSDs) exposed after etching on the N side is 0.5×10 4 cm -2 , which is significantly lower than 1.5×10 4 cm -2 The density of shell-shaped pits (BPD) in the Al surface is 4.9×10 3 cm -2 , which is consistent with Example 3; the total density of the hexagonal pits on the Al surface is 3.0×10 4 cm -2 The TED density was calculated by subtraction: the TED density of the Al surface = the total number of hexagonal pits on the Al surface 3.0×10 4 cm -2 -TSD density on the N side is 0.5×10 4 cm -2 =2.5×10 4 cm -2, significantly higher than 1.7×10 4 cm -2 .

[0027] Due to insufficient ammonia flow, the N-side etching was incomplete, resulting in some TSDs not being fully exposed and etched, leading to a lower TSD density detected on the N-side. In subsequent comparisons of the same area on both the Al and N-sides, a large number of TSDs not identified on the N-side were incorrectly counted as TEDs, resulting in an inflated TED density of approximately 47%. This demonstrates that the double-side comparison mechanism fails when the ammonia flow rate is below 100 sccm, reproducing the misidentification issue described in the background technology regarding the difficulty in distinguishing TSDs from TEDs.

[0028] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals, characterized in that: include: S01: cutting, grinding, polishing, and cleaning the aluminum nitride single crystal along the [0001] direction to form the Al face and N face of the aluminum nitride single crystal; S02: etching the Al surface with a molten KOH-NaOH eutectic mixture at 355-400° C. to form an etched Al surface; S03: etching the N surface using ammonia gas at 900-1100° C. to form an etched N surface; S04: After cooling, cleaning, drying and dust removal, the corroded Al surface and the etched N surface are subjected to image recognition to obtain dislocation defect distribution.

2. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In the S01, the grinding powder used for grinding is 0.5-10 of diamond abrasive powder.

3. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In S01, the cleaning includes: sequentially using acetone ultrasound, boiling anhydrous ethanol, 70-80°C sulfuric acid solution and deionized water ultrasound for cleaning, wherein the ultrasonic power is ≥100W and the ultrasonic frequency is 40kHz.

4. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In the S02, the mass ratio of KOH to NaOH in the molten KOH-NaOH eutectic mixture is 1:

1.

5. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In S03, the flow rate of the ammonia gas is 100-300 sccm, and the etching time is 20-40 minutes.

6. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In the step S04, the corroded Al surface and the etched N surface are cooled to room temperature in a dry N2 gas flow.

7. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In S04, the cleaning includes: sequentially using boiling anhydrous ethanol, 70-80° C. sulfuric acid solution and deionized water for three ultrasonic cleanings.

8. The method for accurately detecting the dislocation defect distribution of aluminum nitride single crystal according to claim 1, characterized in that: In the step S04, the drying includes drying with N2 and then drying at 80-85°C.

9. The method for accurately detecting the distribution of dislocation defects in aluminum nitride single crystals according to claim 1, characterized in that: In the S04, obtaining the dislocation defect distribution through image recognition includes: Determining the TSD density of the N-side according to the image recognition of the etched N-side; determining the BPD density, TSD and TED density of the Al surface according to the image recognition of the corroded Al surface; The TSD density in the same dislocation region of the corroded Al surface and the etched N surface is compared to determine the TSD density and TED density of the Al surface.

Images