Evaluation method for silicon distribution uniformity in CVD silicon-carbon composite material

Abstract

The invention relates to a method for evaluating distribution uniformity of silicon in a CVD (chemical vapor deposition) silicon-carbon composite material. The method comprises the following steps: (S1) acquiring a scanning electron microscope backscattered electron image and a silicon element surface distribution diagram of a sample; (S2) carrying out binarization processing on the silicon element surface distribution diagram through image analysis software, and counting the area proportion sigma Si of the deposited silicon particles; (S3) combining the carbon mass fraction c% of the sample and the correction coefficient delta, and calculating the mass fraction w% of undeposited silicon particles; the correction coefficient delta is obtained through calibration: under the same imaging condition, selecting a region in which all particles are completely covered by a silicon signal and the number of complete particles is not less than 50 from a silicon element surface distribution diagram as a'complete silicon deposition region '; and carrying out the same binarization processing on the silicon particles, wherein the area ratio sigma Si0 of the silicon particles is a correction coefficient delta. The evaluation method provided by the invention has high accuracy, wide applicability and excellent measurement repeatability, and is a scientific and reliable quantitative analysis means.

Description

Technical Field

[0001] This invention belongs to the field of lithium-ion battery technology, specifically relating to a method for evaluating the uniformity of silicon distribution in CVD silicon-carbon composite materials. Background Technology

[0002] Silicon-carbon composites, as anode materials for high-energy-density lithium-ion batteries, rely heavily on the uniformity of silicon distribution within the carbon matrix for their performance. Chemical vapor deposition (CVD) has become a crucial method for preparing high-performance silicon-carbon composite anodes due to its relatively low equipment requirements, controllable process, and ease of industrial production. This method introduces silicon by depositing silanes within the pores of porous carbon, resulting in CVD silicon-carbon composites. The core process lies in achieving uniform silicon deposition, which directly impacts the material's initial coulombic efficiency, cycle stability, and volume expansion. Currently, the main methods used in the industry to evaluate silicon deposition uniformity include TEM, elemental mapping, and small-angle X-ray scattering (SAXS) diffraction, but these methods all have significant limitations.

[0003] TEM method: This method can directly observe the silicon distribution in local areas, but the field of view is extremely small, and the results are limited by whether the selected area is representative. It is difficult to reflect the overall uniformity of the material and cannot achieve quantitative, high-throughput analysis.

[0004] Elemental Mapping: Currently, the commonly used methods for assessing silicon deposition uniformity in the industry are mainly based on surface scanning of Si using energy dispersive spectroscopy (EDS) or electron energy loss spectroscopy (EELS). This involves statistically analyzing the proportion of undeposited silicon particles within the field of view at a specific magnification to determine the uniformity of the deposition. However, this method has several limitations: First, the statistical results are limited by the selected field of view. Due to the inherent particle size distribution of porous carbon, the statistical results differ across different regions, resulting in poor repeatability. Second, the process is time-consuming and difficult to implement high-throughput analysis. More importantly, this method can only provide a rough characterization based on the field of view and cannot quantitatively describe the amount of silicon deposited, i.e., the true proportion of silicon deposited, thus failing to reflect the overall silicon deposition uniformity level of the material.

[0005] Small-angle X-ray scattering (SAXS): It can non-destructively and statistically analyze the filling of pores within a large number of particles. However, SAXS provides scattering intensity curves I(q) vs. q, rather than direct images. Mathematical models must be used to invert the actual structure from the scattering data. The appropriateness of the model selection directly affects the accuracy of the results, and there is a lack of uniqueness in the interpretation. At the same time, it requires a deep theoretical knowledge and a thorough understanding of scattering theory to correctly build the model and interpret the data. For non-experts, the threshold is high, so it is not suitable for rapid detection and process control in production lines. Furthermore, it cannot distinguish whether the scattering signal comes from silicon or other high electron density impurities (such as metal catalyst particles), and must be used in conjunction with compositional analysis methods such as EDS.

[0006] CN118777355A discloses a method for measuring the uniformity of silicon distribution in silicon-carbon composite materials, including the following steps: S1, Sample preparation: The silicon-carbon composite anode material to be tested is pre-packed into a PE self-sealing bag. A puncture is made at the bottom of the self-sealing bag with a fine needle. The material to be tested is sprayed onto conductive adhesive in the sample stage through the puncture. An air blowing device is used to remove excess powder not adhering to the conductive adhesive. The sample stage is then placed in the scanning electron microscope (SEM) testing area; S2, Observation: A backscattered electron microscope (SEM) is used to capture a microscopic image; S3, Analysis: The uniformity of grayscale distribution in the backscattered electron image is measured using measurement software. This invention introduces grayscale analysis indicators to quantitatively analyze the uniformity of silicon distribution. However, its evaluation indicators lack a direct and quantitative correlation with the material's quality properties, thus limiting its application in guiding process optimization and predicting electrochemical performance.

[0007] Therefore, there is an urgent need in this field for a new evaluation method that can accurately, quantitatively, rapidly, and economically evaluate the uniformity of silicon deposition in silicon-carbon composite materials prepared by CVD, thereby providing a direct basis for production process optimization and product quality control. Summary of the Invention

[0008] In view of the problems in evaluating the uniformity of silicon deposition in silicon-carbon composites in the prior art, the present invention provides a method for evaluating the uniformity of silicon distribution in CVD silicon-carbon composites. This method is based on scanning electron microscopy element mapping and image analysis software, and uses an analytical model to establish a calculation formula for the mass fraction w% of undeposited silicon particles. It can accurately and quantitatively evaluate the uniformity of silicon deposition in CVD silicon-carbon composites, and can also evaluate the influence of different amounts of silicon deposition on the uniformity of silicon deposition.

[0009] To achieve the above objectives, the present invention adopts the following technical solution:

[0010] A method for evaluating the uniformity of silicon distribution in CVD silicon-carbon composite materials includes the following steps:

[0011] (S1) Obtain the scanning electron microscope backscattered electron image of the CVD silicon-carbon composite material to be tested, and simultaneously acquire the silicon element surface distribution map (Si-mapping) in the same field of view.

[0012] (S2) The silicon element surface distribution map is binarized using image analysis software to distinguish between silicon signal regions and silicon-free signal regions, resulting in a binary image. Based on this binary image, the proportion of the area of ​​the silicon signal region to the total particle projection area is calculated using image analysis software and denoted as the area ratio ΣSi of the silicon-deposited particles.

[0013] (S3) The mass fraction w% of unprecipitated silicon particles in the CVD silicon-carbon composite material is calculated using Equation 1:

[0014] w% = c% × (1 (Equation 1)

[0015] Where c% is the carbon mass fraction of the CVD silicon-carbon composite material, and δ is the correction factor;

[0016] The δ is obtained through the following calibration steps: Under the same imaging conditions as in step (S1), in the silicon elemental distribution map, select a continuous region in which all particles are completely covered by silicon signals, and the number of such complete particles in the region is not less than 50. Define this region as a "completely silicon-deposited region". Perform the same binarization processing as in step (S2) on the "completely silicon-deposited region" as on the sample to be tested. Calculate the area ratio of silicon-deposited particles in this region using image analysis software, denoted as ΣSi0. Then, the correction coefficient δ = ΣSi0.

[0017] The term "completely silicon-deposited region" refers to a region in which each particle is completely covered by silicon signals, and the total number of such complete particles is at least 50.

[0018] The core of the silicon distribution uniformity evaluation method proposed in this invention lies in combining the area ratio of deposited silicon particles (ΣSi) obtained from image analysis with the carbon mass fraction (c%) measured by chemical analysis, and calculating the mass fraction of undeposited silicon particles (w%) using Equation 1. This w% can evaluate the silicon deposition uniformity of CVD silicon-carbon composite materials. The lower the w% value, the fewer carbon particles in the material that are not effectively deposited by silicon, that is, the better the silicon distribution uniformity.

[0019] In practice, due to limitations such as image brightness, probe response, sample surface morphology differences, and threshold settings, even for areas theoretically completely covered by silicon, the area of ​​the "silicon signal region" identified by image analysis software will deviate from the actual physical area. To eliminate this systematic error and ensure the accuracy and comparability of results obtained from different batches and under different testing conditions, this invention introduces a correction coefficient δ. From Equations 1 and 5, it can be deduced that when a sample is in an ideal fully silicon-deposited state (i.e., w% = 0), δ = ΣSi0. It should be noted that δ is not obtained based on a physically ideal fully silicon-deposited sample, but rather on establishing a relatively stable internal standard based on the image analysis process for subsequent calculation and calibration of all similar samples. That is, the physical meaning of the correction coefficient δ is: it represents the theoretical limit of the percentage of silicon-deposited area that can be identified when performing image analysis on an artificially selected "fully silicon-deposited" region under the current testing system. The essence of δ is a normalization calibration factor, making the calculated mass fraction of undeposited silicon particles w% closer to the true mass fraction.

[0020] Further, step (S1) specifically involves: dispersing the CVD silicon-carbon composite material to be tested on conductive adhesive in the sample stage, blowing away excess powder on the conductive adhesive; after gold sputtering, placing it in the scanning electron microscope test area, capturing backscattered electron images, and simultaneously acquiring silicon element surface distribution maps (Si-mapping) in the same field of view.

[0021] Furthermore, the magnification of the scanning electron microscope backscattered electron image in step (S1) is selected based on the median particle size D50 of the CVD silicon-carbon composite material to be tested. Specifically, when D50 < 5 μm, the magnification is 1000~2000 times; when 5 μm ≤ D50 ≤ 10 μm, the magnification is 500~1000 times; and when D50 > 10 μm, the magnification is 300~500 times.

[0022] Furthermore, step (S2) specifically includes the following sub-steps:

[0023] (S21) Import the silicon element surface distribution map into the image analysis software, select a representative field of view containing 200 to 500 complete particle groups as the processing area, and perform image size calibration in the image analysis software according to the scale marked on the scanning electron microscope image to establish the conversion relationship between pixels and actual physical size.

[0024] (S22) Set a threshold based on the grayscale histogram and optimize and calibrate the threshold in combination with chromaticity features to binarize the image and distinguish between silicon signal regions and silicon-free signal regions.

[0025] (S23) Based on the binary image, use image analysis software to calculate the area A of the silicon signal region. SiArea A of the total particle projection region total The area ratio of silicon particles in the sample to be tested, ΣSi=A, was calculated. Si / A total .

[0026] Furthermore, the chromaticity features mentioned in step (S22) are color information parameters extracted based on the differences in color composition of different pixels in the image. For example, they can be converted to the L*a*b* color space to analyze the differences in color component (such as a*, b* channels) distribution between the silicon element signal region and the background region. These differences are used to correct and optimize the threshold determined based on the gray-level histogram, thereby more accurately distinguishing the silicon signal region from the non-silicon signal region, and finally obtaining a binary image with better separation effect.

[0027] Furthermore, Equation 1 in step (S3) is based on the following assumptions and model, and is derived through image analysis results, as follows:

[0028] (1) Establishing hypotheses:

[0029] Assumption 1: The porous carbon particles in the raw material can be equivalent to uniform small spheres with a median particle size D50;

[0030] Hypothesis 2: The silicon deposition process mainly takes place in the internal pores and surface of porous carbon particles. After silicon deposition, the total volume and spherical diameter of the particles remain approximately unchanged, while the particle size of the undeposited particles is randomly distributed.

[0031] Assumption 3: Based on the signal intensity of the silicon element surface distribution map, in the binary image of step (S2), if the entire region of a particle belongs to the region without silicon signal, then the particle is determined to be a "non-silicon deposited particle".

[0032] (2) Establish a calculation model based on the above assumptions:

[0033] The total number of particles N in a CVD silicon-carbon composite material with mass m is:

[0034] N (Equation 2)

[0035] Where c% is the carbon mass fraction of the CVD silicon-carbon composite material, and m c The mass of the equivalent porous carbon microsphere;

[0036] The number of unprecipitated silicon particles N1 in a CVD silicon-carbon composite material with mass m is:

[0037] N1 (Equation 3)

[0038] Wherein, w% is the mass fraction of unprecipitated silicon particles in the CVD silicon-carbon composite material;

[0039] Then, the number N2 of silicon particles in a CVD silicon-carbon composite material with mass m is:

[0040] N2 = N N1= (Equation 4)

[0041] The area percentage ΣSi of the silicon particles in the silicon-mapping image is:

[0042] ΣSi = ×δ = × δ = × δ (Equation 5)

[0043] Where S is the maximum projected area of ​​a single equivalent porous carbon microsphere, and δ is the correction coefficient;

[0044] (3) Equation 5 is transformed to obtain Equation 1.

[0045] Ideally, the area ratio ΣSi of the silicon deposited particles obtained from image analysis should be equal to the number ratio of the silicon deposited particles. However, due to factors such as image brightness, probe response, and differences in particle surface morphology, the measured values ​​deviate from the theoretical values. Therefore, a correction coefficient δ is introduced in Equation 5. δ is a parameter related to instrument imaging conditions, detector response, and particle morphology characteristics, which can be obtained through the above calibration steps and used as the normalization benchmark for all subsequent sample calculations. A schematic diagram of the principle model of the analytical method described in this invention is shown below. Figure 1 As shown in the diagram. A schematic diagram of the calibration process for the correction factor δ is shown below. Figure 2 As shown.

[0046] Preferably, the number of intact particles in step (S3) is 50 to 100.

[0047] Furthermore, the carbon mass fraction (c%) of the CVD silicon-carbon composite material can be determined by a carbon-sulfur analyzer or an elemental analyzer.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0049] 1. The evaluation method of this invention combines the area ratio of silicon-deposited particles (ΣSi) obtained from image analysis with the carbon mass fraction (c%) obtained from chemical analysis, and outputs the intuitive indicator of the mass fraction w% of undeposited silicon particles. This enables accurate quantitative analysis of CVD silicon-carbon composite materials. w% can be directly correlated with the CVD process and electrochemical performance, providing a key basis for process optimization and product quality improvement.

[0050] 2. By introducing a correction coefficient δ, systematic errors are eliminated, and the accuracy and repeatability of the evaluation method are improved.

[0051] 3. The evaluation method of this invention is based on conventional SEM and image analysis software, which is easy to apply in R&D and quality control. It can provide clear guidance for process optimization and can be correlated with key performance indicators. It is an effective tool for connecting material preparation and performance evaluation. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the principle model of the analysis method described in this invention.

[0053] Figure 2 This is a schematic diagram of the calibration process for the correction coefficient δ.

[0054] Figure 3 To verify the superimposed image of the backscattered electron image of the scanning electron microscope and the surface distribution of silicon element in different samples in Example 1. Detailed Implementation

[0055] The present invention will be further described below with reference to specific embodiments, but the present invention is not limited to the following embodiments.

[0056] Unless otherwise specified, the experimental methods described in the following examples are conventional methods; unless otherwise specified, the reagents, materials and equipment are commercially available.

[0057] Scanning electron microscope, model Axia, purchased from Thermo Fisher Scientific.

[0058] ImageJ, an image analysis software, is provided by the National Institutes of Health (NIH).

[0059] The self-made CVD silicon-carbon composite material has a median particle size D50 of 6.1 μm and a carbon mass fraction (c%) of 41.67% as measured by a carbon-sulfur analyzer.

[0060] Porous carbon, self-made, with a median particle size D50 of 6.0 μm.

[0061] Example

[0062] (S1) Sample preparation and image acquisition: Take about 5 mg of CVD silicon-carbon composite material and spread it evenly on the conductive adhesive in the sample stage. Blow away the excess powder on the conductive adhesive. After sputtering with gold (10 nm thickness), place it in the testing area of ​​the scanning electron microscope. Use a scanning electron microscope (10 kV voltage, 800x magnification) to take backscattered electron images and simultaneously acquire (2048×2048 sampling pixels) the same field of view of silicon element surface distribution map (Si-mapping).

[0063] (S2) Image binarization and calculation of the area ratio ΣSi of silicon deposited particles, including the following sub-steps:

[0064] (S21) Import the Si-mapping image into ImageJ software, select a representative field of view containing about 350 complete particle groups as the processing area, and perform image size calibration in ImageJ software according to the scale marked on the scanning electron microscope image to establish the conversion relationship between pixels and actual physical size (μm).

[0065] (S22) The threshold is initially selected using the "Triangle" automatic thresholding algorithm of the grayscale histogram, and the threshold of the b component is set to 105 in the L*a*b* color space for optimization and calibration. The image is binarized to clearly distinguish the silicon signal area and the silicon-free signal area.

[0066] (S23) Statistically measure the area A of the region containing silicon signals in the binary graph. Si Area A of the total particle projection region total Calculate the area ratio of silicon-deposited particles ΣSi=A in CVD silicon-carbon composite materials. Si / A total =89.71%;

[0067] (S3) Calculation of correction coefficient δ and mass fraction w% of undeposited silicon particles: During image acquisition in step (S1) of CVD silicon-carbon composite material, a continuous region where all particles are completely covered by silicon signals is selected from the Si-mapping images of the same batch. Approximately 60 such complete particles are found in this region, which is defined as the "completely silicon-deposited region". The "completely silicon-deposited region" undergoes the same image analysis process as in steps (S22) and (S23) above, yielding an area ratio of silicon-deposited particles ΣSi0 = 92.67%, i.e., correction coefficient δ = ΣSi0 = 92.67%. Substituting ΣSi = 89.71%, δ = 92.67%, and c% = 41.67% into w% = c% × (1 In the calculation, the mass fraction of unprecipitated silicon particles in the CVD silicon-carbon composite material was w%=1.33%.

[0068] Verification Example 1

[0069] By doping porous carbon particles (undeposited silicon particles) with different proportions into CVD silicon-carbon composite materials, silicon-carbon composite material samples with known undeposited silicon uniformity are artificially constructed. Then, using the method of this invention, the mass fraction w% of undeposited particles in the sample is calculated and compared with the actual mass fraction w1% of undeposited particles obtained by determining the doping ratio, thereby verifying the accuracy and applicability of the evaluation method of this invention. The specific process is as follows:

[0070] (1) Weigh 1.00g of CVD silicon-carbon composite material (as a reference sample, the mass fraction of unprecipitated silicon particles was found to be 1.33% in Example 1), and place it in anhydrous ethanol with 0.05g, 0.10g, 0.20g, and 0.30g of porous carbon (unprecipitated silicon particles) respectively. Disperse the mixture ultrasonically for 15min and stir it under magnetic stirring for 30min to obtain samples A, B, C, and D. The carbon mass fraction c% in samples A, B, C, and D were tested by a carbon-sulfur analyzer and found to be 43.73%, 45.89%, 48.87%, and 53.23%, respectively.

[0071] (2) Repeat steps (S1) and (S2) of the embodiment for samples A, B, C, and D respectively, and find that the area ratio ΣSi of the silicon particles in each sample is 79.77%, 71.75%, 59.62%, and 50.88% respectively (the scanning electron microscope backscattered electron images and silicon element surface distribution superimposed images of different samples are shown in the figure). Figure 3 (as shown)

[0072] (3) Substitute the correction coefficient δ=92.67% determined in the examples and the ΣSi and c% of each sample into w%= c% × (1- In the calculation, the mass fraction of unprecipitated silicon particles w% for each sample was obtained; the specific data are shown in Table 1.

[0073] (4) Calculate the actual mass fraction w1% of unprecipitated silicon particles according to the following formula 6, and compare it with w%;

[0074] w1% = × 100% (Equation 6)

[0075] Where m1 is the mass (g) of the doped porous carbon.

[0076] Equation 6 essentially calculates the actual mass fraction w1% of unprecipitated silicon particles based on the doping amount. This calculation includes both the unprecipitated silicon particles in the reference sample itself and the porous carbon containing the additional unprecipitated silicon. w% is the mass fraction of unprecipitated silicon particles in each sample measured using the method of this invention. The smaller the difference between w1% and w%, the more accurate the method of this invention. Specific results are shown in Table 1.

[0077] Table 1 Comparison of mass fraction of unprecipitated silica particles in different samples

[0078]

[0079] As can be seen from Table 1, for a series of samples with varying degrees of inhomogeneity, the mass fraction w% of unsettled silicon particles obtained by the method of the present invention is in high agreement with the actual mass fraction w1% of unsettled silicon particles calculated by the doping amount, with a very small difference. This fully demonstrates the accuracy and wide applicability of the evaluation method of the present invention.

[0080] Verification Example 2

[0081] To verify the repeatability and stability (consistency of multiple measurements of the same sample) of the evaluation method of this invention, sample A from verification example 1 was selected as the test object, and the following tests were performed:

[0082] (1) From the sample A that has been prepared and mixed evenly in verification example 1, 6 test samples (approximately 5 mg each) are randomly and independently weighed.

[0083] (2) For the above 6 test samples, the complete process from (S1) to (S3) in the embodiment was performed completely independently, and the mass fraction w of unprecipitated silicon particles for each test sample was calculated.

[0084] The specific test results are shown in Table 2.

[0085] Table 2 Repeatability test results for the same sample

[0086]

[0087] As shown in Table 2, the mass fraction (w%) of unprecipitated silicon particles obtained from six complete and independent measurements of the same sample A is very close, falling within a narrow range of 6.02% to 6.11%. The relative standard deviation of the six measurements is only 0.59%, indicating that the evaluation method of this invention has very high measurement repeatability and stability. Furthermore, the average value of the six measurements, 6.05%, is very close to the actual mass fraction of unprecipitated silicon particles, 6.03%, further verifying the accuracy of the method of this invention.

[0088] In summary, the evaluation method provided by this invention has both high accuracy and wide applicability, as well as excellent measurement repeatability and stability, making it a scientific and reliable quantitative analysis method.

[0089] Application examples

[0090] The evaluation method of this invention has application significance in research and development and quality control. For example, to optimize the CVD deposition process, four silicon-carbon composite material samples were prepared by changing only the deposition temperature (T1, T2, T3, and T4) while keeping other process parameters unchanged. Each sample was evaluated according to the method described in the examples, and the calculated undeposited silicon particle mass fraction (w%) was 6.67%, 8.18%, 2.49%, and 1.59%, respectively. A lower w% indicates better silicon deposition uniformity. The results show that the samples prepared under process conditions T3 and T4 have significantly lower w% values ​​than those under T1 and T2, clearly indicating that T3 / T4 is a better process window. Furthermore, a correlation analysis between the w% value and the initial coulombic efficiency data of the battery revealed a significant negative correlation between the w% value and the initial coulombic efficiency (lower w% indicates higher efficiency), quantitatively verifying the key influence of silicon deposition uniformity on electrochemical performance. This application example demonstrates that the w% index obtained by the evaluation method of the present invention can provide clear guidance for process optimization and can be correlated with key performance indicators, making it an effective tool for linking material preparation and performance evaluation.

Claims

1. A method for evaluating uniformity of silicon distribution in a CVD silicon-carbon composite material, characterized by, Includes the following steps: (S1) Obtain the scanning electron microscope backscattered electron image of the CVD silicon-carbon composite material to be tested, and simultaneously acquire the silicon element surface distribution map of the same field of view. (S2) The silicon element surface distribution map is binarized using image analysis software to distinguish between silicon signal regions and silicon-free signal regions, resulting in a binary image. Based on this binary image, the proportion of the area of ​​the silicon signal region to the total particle projection area is calculated using image analysis software and denoted as the area ratio ΣSi of the silicon-deposited particles. (S3) The mass fraction w% of unprecipitated silicon particles in the CVD silicon-carbon composite material is calculated using Equation 1: w% = c% × (1 ) (Equation 1) Where c% is the carbon mass fraction of the CVD silicon-carbon composite material, and δ is the correction factor; The δ is obtained through the following calibration steps: Under the same imaging conditions as in step (S1), in the silicon elemental distribution map, a continuous region in which all particles are completely covered by silicon signals is selected, and the number of such complete particles in the region is not less than 50. This region is defined as the "completely silicon-deposited region". The "completely silicon-deposited region" is subjected to the same binarization processing as the sample to be tested in step (S2). The area ratio of silicon-deposited particles in this region is statistically analyzed by image analysis software and denoted as ΣSi0. Then the correction coefficient δ = ΣSi0.

2. The evaluation method according to claim 1, characterized by Step (S1) specifically involves: dispersing the CVD silicon-carbon composite material to be tested on conductive adhesive in the sample stage, blowing away excess powder on the conductive adhesive; after gold sputtering, placing it in the scanning electron microscope test area, capturing backscattered electron images, and simultaneously acquiring silicon element surface distribution maps in the same field of view.

3. The evaluation method according to claim 1, characterized in that, Step (S1): The magnification of the scanning electron microscope backscattered electron image is selected based on the median particle size D50 of the CVD silicon-carbon composite material to be tested. Specifically, when D50 < 5 μm, the magnification is 1000~2000 times; when 5 μm ≤ D50 ≤ 10 μm, the magnification is 500~1000 times; when D50 > 10 μm, the magnification is 300~500 times.

4. The evaluation method according to claim 1, characterized in that, Step (S2) specifically includes the following sub-steps: (S21) Import the silicon element surface distribution map into the image analysis software, select a representative field of view containing 200 to 500 complete particle groups as the processing area, and perform image size calibration in the image analysis software according to the scale marked on the scanning electron microscope image to establish the conversion relationship between pixels and actual physical size. (S22) Set a threshold based on the grayscale histogram and optimize and calibrate the threshold in combination with chromaticity features to binarize the image and distinguish between silicon signal regions and silicon-free signal regions. (S23) Based on the binary image, use image analysis software to calculate the area A of the silicon signal region. Si Area A of the total particle projection region total The area ratio of silicon particles in the sample to be tested, ΣSi=A, was calculated. Si / A total .

5. The evaluation method according to claim 4, characterized in that, The chromaticity feature mentioned in step (S22) is a color information parameter extracted based on the differences in color composition of different pixels in the image.

6. The evaluation method according to claim 1, characterized in that, Equation 1 in step (S3) is based on the following assumptions and models, and is derived from the results of image analysis. The specific process is as follows: (1) Establishing hypotheses: Assumption 1: The porous carbon particles in the raw material can be equivalent to uniform small spheres with a median particle size D50; Hypothesis 2: The silicon deposition process mainly takes place in the internal pores and surface of porous carbon particles. After silicon deposition, the total volume and spherical diameter of the particles remain approximately unchanged, while the particle size of the undeposited particles is randomly distributed. Assumption 3: Based on the signal intensity of the silicon element surface distribution map, in the binary image of step (S2), if the entire region of a particle belongs to the silicon-free signal region, then the particle is determined to be an "unsettled silicon particle". (2) Establish a calculation model based on the above assumptions: The total number of particles N in a CVD silicon-carbon composite material with mass m is: N (Equation 2) wherein c is the carbon mass fraction of the CVD silicon-carbon composite material, m c is the mass of the equivalent porous carbon sphere; The number of unprecipitated silicon particles N1 in a CVD silicon-carbon composite material with mass m is: N1 (Equation 3) Wherein, w% is the mass fraction of unprecipitated silicon particles in the CVD silicon-carbon composite material; Then, the number N2 of silicon particles in a CVD silicon-carbon composite material with mass m is: N2 = N N1= (Equation 4) The area percentage ΣSi of the silicon particles in the silicon-mapping image is: ΣSi = ×δ = × δ = × δ (Equation 5) Where S is the maximum projected area of ​​a single equivalent porous carbon microsphere, and δ is the correction coefficient; (3) Equation 5 is transformed to obtain Equation 1.

7. The evaluation method according to claim 1, characterized in that, The number of complete particles mentioned in step (S3) is 50 to 100.

8. The evaluation method according to claim 1, characterized in that, The carbon mass fraction (c%) of the CVD silicon-carbon composite material can be determined by a carbon-sulfur analyzer or an elemental analyzer.

Images