A method for detecting the quality of multi-crystalline silicon based on multi-parameter coupling for zone melting

Abstract

The application discloses a kind of based on the zone melting of polycrystalline silicon quality detection method of multi-parameter coupling, comprising the following steps: S1, sample preparation, S2, sample surface treatment, S3, sample surface scanning, S4, scanning data processing, S5, raw material quality analysis.The advantages are: by argon ion polishing combined with EBSD electron backscattering diffractometer, break through the limitation of traditional surface etching, can three-dimensional analysis polycrystalline silicon internal grain size, distribution and grain boundary morphology, more accurately reflect the physical growth form and quality characteristics of polycrystalline silicon.Based on the raw material screening of grain uniformity standard, it provides scientific, reasonable and effective basis for effectively selecting qualified zone melting polycrystalline silicon raw material before drawing, can avoid dislocation problem caused by microcrystalline unmelted, thereby significantly improve the success rate of single crystal pulling, reduce raw material waste, reduce production cost;It can also provide a reference for the purchase and export price of polycrystalline silicon rod.

Description

Technical Field

[0001] This invention relates to the field of polycrystalline silicon production, and specifically to a method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling. Background Technology

[0002] In modern society, chips are becoming increasingly important. Silicon wafers, made from single-crystal silicon, are the most basic material for chip manufacturing. Currently, there are two methods for pulling single-crystal silicon: the Czochralski method and the zone pulling method. Because the zone pulling method avoids contact with the crucible and prevents contamination during the pulling process, single crystals pulled by the zone pulling method have advantages such as higher resistance and longer minority carrier lifetime compared to the Czochralski method. Therefore, single crystals pulled by the zone pulling method are widely used in power electronic devices, such as power diodes, thyristors, junction-gate bipolar transistors, and gate turn-off transistors.

[0003] During the zone melting process, the suspended region is easily stretched into an unmelted state, which eventually leads to the disappearance of the crystal lines of monocrystalline silicon. Under the conditions of stable crystal pulling equipment and personnel operation, stable crystal pulling process, and comparable impurity levels, the success rate of monocrystalline silicon pulling is directly affected by the physical growth morphology of the raw material (polycrystalline silicon for zone melting). Therefore, it is crucial to control the grain size and morphology inside polycrystalline silicon.

[0004] Chinese invention patent CN114034715B, entitled "A Method for Detecting Growth Defects in Zone-Melted Polycrystalline Silicon," discloses a method for detecting growth defects in zone-melted polycrystalline silicon. This method involves cutting a section of polycrystalline silicon from a zone-melted polycrystalline silicon rod; obtaining a polycrystalline silicon wafer from the rod; etching the wafer with an etchant; and capturing growth images of the etched wafer using a microscope, thereby detecting growth defects in the zone-melted polycrystalline silicon. However, this method can only roughly observe the polycrystalline silicon surface and macroscopic growth defects. It can only analyze defects generated during the pulling process after the pulling process is completed, allowing for adjustments to the process.

[0005] The selection of raw materials before crystal pulling, as well as the factors affecting the success rate of crystal pulling, such as the physicochemical properties, grain size, and grain size distribution of polycrystalline silicon rods, were not effectively analyzed.

[0006] Currently, there is still no reasonable and effective testing method to evaluate the quality of the physical growth morphology of polycrystalline silicon products for zone melting, making it difficult to effectively select qualified polycrystalline silicon raw materials for zone melting before pulling to improve the success rate of single crystal pulling. Summary of the Invention

[0007] To address the aforementioned problems, the present invention aims to provide a method for quality inspection of polycrystalline silicon used in zone melting based on multi-parameter coupling.

[0008] This invention is implemented by the following technical solution:

[0009] A method for quality inspection of polycrystalline silicon used in zone melting based on multi-parameter coupling includes the following steps:

[0010] S1. Sample preparation: Select a crack-free area on the surface of the polycrystalline silicon rod to be tested, drill a rod-shaped sample laterally using a hollow drill bit, cut the rod-shaped sample into a sheet-shaped sample, and then polish the surface of the sheet-shaped sample to obtain the sample.

[0011] S2, Sample surface treatment: The sample prepared in S1, sample preparation is deeply polished to obtain a polished sample;

[0012] S3, Sample Surface Scanning: The polished sample obtained after surface treatment in S2 is scanned using an EBSD electron backscatter diffractometer to obtain scanning data;

[0013] S4. Scanning data processing: Based on the scanning data obtained in S3, sample surface scanning, calculate the characteristic parameters of the sample respectively. The characteristic parameters include melt consistency index (MCI), thermal stress resistance index (TSRI), and grain boundary stability index (GSI).

[0014] S5. Raw material quality analysis: Based on the characteristic parameters obtained in S4 and scanning data processing, the quality of the polycrystalline silicon rod to be tested is analyzed using the raw material quality analysis standard.

[0015] Furthermore, in S1, during the preparation of the test sample, the diameter of the hollow drill bit is φ10mm, the length of the rod-shaped sample is 20mm, and the thickness of the sheet-shaped sample is 0.1-3cm, whichever is convenient for instrument clamping. When polishing the surface of the sheet-shaped sample, metallographic sandpaper or silk velvet is used to polish it to make its roughness Ra < 0.6μm. This pre-polishing pretreatment prepares the sample for argon particle polishing. Unpolished unevenness will block light, and lines will affect the observation results. Moreover, argon particle polishing is costly; therefore, pre-polishing is necessary.

[0016] Furthermore, in S1, during the preparation of the test sample, when the surface of the sheet sample is polished, its roughness Ra is made to be less than 0.4 μm.

[0017] Furthermore, in step S2, the sample surface treatment involves polishing the sample prepared in step S1 twice using an argon ion polisher, wherein:

[0018] Step 1: Rough polishing: using a voltage of 7-9kV, an incident angle of 2.5-3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5-2h;

[0019] The second step is fine polishing: using a voltage of 3-5kV, an incident angle of 2.5-3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5-2h.

[0020] This step can completely remove machining marks from the sample preparation process, such as cutting lines and tool marks.

[0021] The aforementioned traces can affect the determination of some parameters, but argon particle polishing can largely shield these effects.

[0022] Furthermore, in S3, during the sample surface scanning, the aperture of the scanning electron microscope aperture is 110-120mm; the magnification is 800-1500X, and the step size is 0.12-0.16μm.

[0023] The scanning data includes: the average equivalent circle diameter, average roundness, size variation coefficient, maximum aspect ratio, maximum directional stretch, average roundness, elliptic angle concentration, effective grain density, spatial inhomogeneity, minimum Freret diameter, and maximum Freret diameter of all grains within the detection area on the surface of the polished sample.

[0024] The meanings of the parameters for each scan data point are as follows:

[0025] The equivalent circle diameter refers to the diameter of a circle with the same projected area as the grain, reflecting the average size of the grain; the average equivalent circle diameter is the average of the equivalent circle diameters of all grains, and the smaller this value is, the better.

[0026] Roundness = 4π × grain area / (grain perimeter) 2 The roundness reflects how close the grain shape is to a circle. The closer the roundness is to 1, the closer the grain shape is to a circle. The average roundness is the average of the roundness of all grains.

[0027] The coefficient of variation in size is calculated as the standard deviation of the equivalent circle diameter divided by the average equivalent circle diameter. It measures the degree of dispersion in grain size; the larger the coefficient, the greater the difference in grain size.

[0028] The maximum aspect ratio is the maximum aspect ratio of the fitted ellipse of all grains, and the maximum directional stretch is the maximum directional stretch of all grains.

[0029] Ellipse angle concentration is the standard deviation of the fitted ellipse angles of all grains within the range of 0-180°, representing the degree of concentration of grain orientation.

[0030] Effective grain density = total number of grains / detection area, which reflects the number of grains per unit area.

[0031] The method for calculating spatial non-uniformity is as follows: the sample in the detection area is divided into a 1mm×1mm grid, and then the standard deviation of the grain density of each grid is calculated; this parameter reflects the uniformity of the grain distribution in space.

[0032] The minimum and maximum Freret diameters are the minimum and maximum distances of all grain projections in different directions, respectively.

[0033] The scan data described above were directly exported from the EBSD electron backscatter diffractometer software, requiring no manual calculations. This scan data provides a comprehensive and detailed understanding of the actual condition of the internal grains of polycrystalline silicon.

[0034] Furthermore, in step S4, during the scanning data processing, the calculation method for the melt flow index (MCI) is as follows:

[0035]

[0036] This index combines grain size and shape characteristics to predict the melting synchronization of the polycrystalline silicon rod being tested; the higher the index, the better the melting synchronization.

[0037] The physical meaning of this formula is as follows:

[0038] Part 1 – (1 / Average equivalent circle diameter × 50): Emphasizing the advantages of small grains (the smaller the diameter, the larger the value).

[0039] Part Two – (1 + average roundness / (size variation coefficient + 0.1)): This index is amplified when the grain shape is more rounded and the size is more uniform.

[0040] Furthermore, in step S4, during the scanning data processing, the method for calculating the Thermal Stress Resistance Index (TSRI) is as follows:

[0041]

[0042] This index predicts the risk of thermal stress concentration in polycrystalline silicon rods by measuring grain shape anisotropy; the higher the index, the stronger the resistance to thermal cracking.

[0043] The physical meaning of this formula is: to characterize the risk of thermal stress concentration by penalizing slender grains (aspect ratio) and strong directionality (directional extension), as well as penalizing highly consistent grain orientation (which easily generates synergistic thermal stress).

[0044] Furthermore, in step S4, during the scanning data processing, the method for calculating the grain boundary stability index GSI is as follows:

[0045]

[0046] Predicting grain boundary failure risk by spatial distribution; higher values ​​indicate more stable grain boundaries.

[0047] Physical meaning:

[0048] High density and uniform distribution improve stability

[0049] Minimum / maximum diameter ratio avoids extreme size differences

[0050] Furthermore, in S5, the raw material quality analysis standard is as follows:

[0051] If the melt flow index (MCI) is greater than or equal to 75, the thermal stress resistance index (TSRI) is greater than or equal to 85, and the grain boundary stability index (GSI) is greater than or equal to 1.5, then the polycrystalline silicon rod to be tested is judged to be a high-quality material.

[0052] If the melt flow index (MCI) is less than 60, the thermal stress resistance index (TSRI) is less than 70, or the grain boundary stability index (GSI) is less than 1.0, then the polycrystalline silicon rod to be tested is determined to be unqualified.

[0053] In other cases, the polycrystalline silicon rod to be tested is determined to be a qualified material.

[0054] Advantages of this invention:

[0055] 1. Precise characterization of grain structure and prediction of raw materials before crystal pulling: By combining argon ion polishing with EBSD electron backscatter diffraction, the limitations of traditional surface etching are overcome, and the size, distribution and grain boundary morphology of polycrystalline silicon can be analyzed in three dimensions, which can more accurately reflect the physical growth morphology and quality characteristics of polycrystalline silicon.

[0056] 2. Significantly improves crystal pulling success rate and reduces raw material waste: Raw material screening based on grain uniformity standards provides a scientific, reasonable, and effective basis for selecting qualified polycrystalline silicon raw materials for zone melting before pulling. This avoids dislocation problems caused by unmelted microcrystals, thereby significantly improving the success rate of single crystal pulling, reducing raw material waste, and lowering production costs. It also provides a reference for the procurement and export pricing of polycrystalline silicon rods. 3. Fills a domestic testing gap: It proposes argon ion polishing parameter combinations, EBSD scanning parameters, and raw material quality analysis standards, establishing a standardized testing method for the entire process, with broad application prospects and significant industry value. Attached Figure Description

[0057] Figure 1 These are metallographic microscope images of the polished sample after surface treatment in Example 1.

[0058] Figure 2 This is a scanning electron microscope image obtained by EBSD electron backscatter diffraction in Example 1;

[0059] Figure 3 The images are from Examples 1, 2, and 3. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1

[0062] like Figure 1-3 The method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling, as shown, includes the following steps:

[0063] S1. Sample preparation: Select a crack-free area on the surface of the polycrystalline silicon rod to be tested, and use a φ10mm hollow drill bit to drill a 20mm long rod-shaped sample laterally. Cut the rod-shaped sample into a 2mm thick sheet sample, and then polish the sheet sample with 800-grit metallographic sandpaper to a surface roughness Ra=0.55μm. Then polish it with silk velvet and diamond polishing paste to Ra=0.38μm to obtain the sample.

[0064] S2. Sample Surface Treatment: The sample prepared in S1, sample preparation, is subjected to deep polishing in two steps using an argon ion polishing machine to obtain a polished sample; wherein:

[0065] Step 1: Rough polishing: using a voltage of 8kV, an incident angle of 2.8°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.8h;

[0066] The second step is fine polishing: using a voltage of 4kV, an incident angle of 2.8°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5h.

[0067] S3. Sample surface scanning: The polished sample obtained after surface treatment in S2 was scanned using an EBSD electron backscatter diffractometer to obtain scanning data. The aperture of the scanning electron microscope was set to 115 mm, and a high beam current mode was used. The magnification was 1200X, the step size was 0.156 μm, and the polished surface was scanned in its entirety.

[0068] The scanning data includes: the average equivalent circular diameter of all grains on the surface of the polished sample is 52 μm, the average roundness is 0.89, the coefficient of variation is 0.15, the maximum aspect ratio is 1.6, the maximum directional stretch is 0.18, the elliptic angle concentration is 175°, and the effective grain density is 135 grains / mm². 2 Spatial inhomogeneity 12, minimum Freret diameter 45 μm and maximum Freret diameter 58 μm;

[0069] S4. Scanning data processing: Based on the scanning data obtained in S3, sample surface scanning, the characteristic parameters of the sample are calculated respectively. Among them, the calculated melt consistency index MCI is 78.2, the thermal stress resistance index TSRI is 88.4, and the grain boundary stability index GSI is 1.68.

[0070] S5. Raw material quality analysis: Based on the characteristic parameters obtained in S4. Scanning data processing, the raw material quality analysis standard is used to analyze the quality of the polycrystalline silicon rod to be tested in this embodiment. The quality of the raw material is high-quality.

[0071] When the polycrystalline silicon rod, the raw material to be tested in this embodiment, was used for zone melting crystal pulling, the silicon rod did not crack, and the crystal pulling was successful.

[0072] Example 2

[0073] A method for quality inspection of polycrystalline silicon used in zone melting based on multi-parameter coupling includes the following steps:

[0074] S1. Sample preparation: Select a crack-free area on the surface of the polycrystalline silicon rod to be tested, and use an 8mm hollow drill bit to drill a 15mm long rod sample laterally. Cut the rod sample into a 1.5mm thick sheet sample, and then polish the sheet sample with 2000-grit metallographic sandpaper until the surface roughness Ra=0.30μm. Finally, polish it with nanodiamond suspension to obtain the sample.

[0075] S2. Sample Surface Treatment: The sample prepared in S1, sample preparation, is subjected to deep polishing in two steps using an argon ion polishing machine to obtain a polished sample; wherein:

[0076] Step 1: rough polishing: using a voltage of 7kV, an incident angle of 3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 2 hours;

[0077] The second step is fine polishing: using a voltage of 3kV, an incident angle of 2.5°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 2 hours.

[0078] S3. Sample surface scanning: The polished sample obtained after surface treatment in S2 was scanned using an EBSD electron backscatter diffractometer to obtain scanning data. The aperture of the scanning electron microscope was set to 120 mm, and a high beam current mode was used. The magnification was 1500X, the step size was 0.12 μm, and the polished surface was scanned in its entirety.

[0079] The scanning data includes: the average equivalent circular diameter of all grains on the surface of the polished sample is 58 μm, the average roundness is 0.82, the coefficient of variation of size is 0.22, the maximum aspect ratio is 1.9, the maximum directional stretch is 0.32, the elliptic angle concentration is 168°, and the effective grain density is 115 grains / mm². 2Spatial inhomogeneity 18, minimum Freret diameter 38 μm and maximum Freret diameter 72 μm;

[0080] S4. Scanning data processing: Based on the scanning data obtained in S3, sample surface scanning, the characteristic parameters of the sample are calculated respectively. Among them, the calculated melt consistency index MCI is 67.5, the thermal stress resistance index TSRI is 73.6, and the grain boundary stability index GSI is 1.15.

[0081] S5. Raw material quality analysis: Based on the characteristic parameters obtained in S4. Scanning data processing, the raw material quality analysis standard is used to analyze the quality of the polycrystalline silicon rod to be tested in this embodiment. The quality of the raw material is qualified.

[0082] When the polycrystalline silicon rod, the raw material to be tested in this embodiment, was used for zone melting crystal pulling, the silicon rod did not crack, and the crystal pulling was successful.

[0083] Example 3

[0084] A method for quality inspection of polycrystalline silicon used in zone melting based on multi-parameter coupling includes the following steps:

[0085] S1. Sample preparation: Select a crack-free area on the surface of the polycrystalline silicon rod to be tested, and use a φ5mm hollow drill bit to drill a 10mm long rod-shaped sample laterally. Cut the rod-shaped sample into a sheet-shaped sample with a thickness of 1.5mm. Then, polish the sheet-shaped sample with 800-grit metallographic sandpaper to a surface roughness Ra=0.55μm, and then polish it with silk velvet and diamond polishing paste to Ra=0.38μm to obtain the sample.

[0086] S2. Sample Surface Treatment: The sample prepared in S1, sample preparation, is subjected to deep polishing in two steps using an argon ion polishing machine to obtain a polished sample; wherein:

[0087] Step 1: Rough polishing: using a voltage of 9kV, an incident angle of 2.5°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5h;

[0088] The second step is fine polishing: using a voltage of 5kV, an incident angle of 3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.8h.

[0089] S3. Sample surface scanning: The polished sample obtained after surface treatment in S2 was scanned using an EBSD electron backscatter diffractometer to obtain scanning data. The aperture of the scanning electron microscope was set to 110 mm, and a high beam current mode was used. The magnification was 1500X, the step size was 0.12 μm, and the polished surface was scanned in its entirety.

[0090] The scanning data includes: average equivalent circular diameter of all grains on the polished sample surface (62 μm), average roundness (0.78), coefficient of variation (0.28), maximum aspect ratio (2.1), maximum directional stretch (0.38), elliptic angle concentration (162°), and effective grain density (95 grains / mm²). 2 Spatial inhomogeneity 24, minimum Freret diameter 32 μm and maximum Freret diameter 85 μm;

[0091] S4. Scanning data processing: Based on the scanning data obtained in S3, sample surface scanning, the characteristic parameters of the sample are calculated respectively. Among them, the calculated melt consistency index MCI is 61.8, the thermal stress resistance index TSRI is 71.2, and the grain boundary stability index GSI is 1.03.

[0092] S5. Raw material quality analysis: Based on the characteristic parameters obtained in S4. Scanning data processing, the raw material quality analysis standard is used to analyze the quality of the polycrystalline silicon rod to be tested in this embodiment. The quality of the raw material is qualified.

[0093] When the polycrystalline silicon rod, the raw material to be tested in this embodiment, was used for zone melting crystal pulling, the silicon rod did not crack, and the crystal pulling was successful.

[0094] By applying the method of this invention to test and analyze the quality of the raw polycrystalline silicon rods, the actual crystal pulling success rate increased from 65% when the raw material quality was not tested to 92%. Since the cost of each breakage during crystal pulling is very high, the method of this invention significantly reduces the costs associated with breakage.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for quality inspection of polycrystalline silicon used in zone melting based on multi-parameter coupling, characterized in that, Includes the following steps: S1. Sample preparation: Select a crack-free area on the surface of the polycrystalline silicon rod to be tested, use a hollow drill bit to drill a rod-shaped sample laterally, cut the rod-shaped sample into a sheet sample, and then polish the surface of the sheet sample to obtain the sample. S2. Sample surface treatment: The sample prepared in S1, sample preparation, is polished twice using an argon ion polishing machine to obtain a polished sample; wherein: Step 1: Rough polishing: using a voltage of 7-9kV, an incident angle of 2.5-3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5-2h; The second step is fine polishing: using a voltage of 3-5kV, an incident angle of 2.5-3°, liquid nitrogen cooling protection temperature of 0℃, and a polishing time of 1.5-2h; S3. Sample Surface Scanning: The polished sample obtained after surface treatment in S2 is scanned using an EBSD electron backscatter diffractometer to obtain scanning data. The scanning data includes: the average equivalent circle diameter, area, circumference, maximum Feretta diameter, major axis / minor axis of the fitted ellipse, angle of the fitted ellipse, aspect ratio of the fitted ellipse, grain aspect ratio, centroid X / centroid Y, and average orientation difference within the grain. The following parameters can be calculated from the above scanning parameters: ① average roundness, ② size variation coefficient, ③ elliptic angle concentration, ④ effective grain density, and ⑤ spatial inhomogeneity. S4. Scanning data processing: Based on the scanning data obtained in S3, sample surface scanning, calculate the characteristic parameters of the sample respectively. The characteristic parameters include melt consistency index MCI, thermal stress resistance index TSRI, and grain boundary stability index GSI. The method for calculating the melt flow uniformity index is as follows: Wherein, MCI is the melt flow index; The method for calculating the thermal stress resistance index is as follows: Wherein, TSRI is the thermal stress resistance index; The method for calculating the grain boundary stability index is as follows: Wherein, GSI is the grain boundary stability index; S5. Raw material quality analysis: Based on the characteristic parameters obtained in S4. Scanning data processing, the quality of the polycrystalline silicon rod to be tested is analyzed using the raw material quality analysis standard.

2. The method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling according to claim 1, characterized in that, In step S1, during the preparation of the test sample, the diameter of the hollow drill bit is φ5-10mm, the length of the rod-shaped sample is 10-20mm, and the thickness of the sheet-shaped sample is 0.1-3cm; when the sheet-shaped sample is surface-polished, its roughness Ra is less than 0.6μm.

3. The method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling according to claim 2, characterized in that, In the preparation of the test sample in S1, the surface roughness Ra of the sheet sample is made to be less than 0.4 μm when it is polished.

4. The method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling according to claim 1, characterized in that, In the S3 sample surface scanning, the aperture of the scanning electron microscope aperture is 110-120mm; the magnification is 800-1500X, and the step size is 0.12-0.16μm.

5. The method for quality inspection of polycrystalline silicon for zone melting based on multi-parameter coupling according to claim 4, characterized in that, In the S5 section, the raw material quality analysis standard is as follows: If the melt flow index (MCI) is greater than or equal to 75, the thermal stress resistance index (TSRI) is greater than or equal to 85, and the grain boundary stability index (GSI) is greater than or equal to 1.5, then the polycrystalline silicon rod to be tested is judged to be a high-quality material. If the melt flow index (MCI) is less than 60, the thermal stress resistance index (TSRI) is less than 70, or the grain boundary stability index (GSI) is less than 1.0, then the polycrystalline silicon rod to be tested is determined to be unqualified. In other cases, the polycrystalline silicon rod to be tested is determined to be a qualified material.

Images