A method and system for detecting the crystalline purity of graphite based on carbon phase separation

Abstract

The application provides a graphite crystalline purity detection method and system based on carbon phase separation, and relates to the technical field of graphite purity detection. In the prior art, the fixed carbon method cannot distinguish crystalline graphite carbon from amorphous carbon, residual carbon and other carbon phases, resulting in false high detection results, grading distortion and trade pricing deviation. The graphitization degree method only represents the lattice order degree and cannot reflect the overall proportion of crystalline carbon. Neither of them is an effective index of true purity, and there is no national standard or industry standard to define and measure the mass fraction of crystalline graphite carbon. The application realizes accurate separation and quantification of crystalline carbon and amorphous carbon by X-ray diffraction combined with peak fitting algorithm, solves the problems of false high detection and grading deviation, and has the advantages of no high-temperature calcination, short detection period, high precision and filling of the blank of standard methods in the field.

Description

Technical Field

[0001] This invention relates to the field of graphite purity detection technology, and in particular to a method and system for detecting the purity of graphite crystals based on carbon phase separation. Background Technology

[0002] In existing technologies, the industry-standard method for characterizing graphite purity mainly uses fixed carbon content to approximate graphite purity, based on the standard GB / T 1429-2025. This method calculates the fixed carbon content by subtracting the difference after measuring ash, moisture, and volatile matter. Another method, based on YB / T 5164, measures the degree of graphitization. However, fixed carbon content is essentially a macroscopic statistical measure of total solid carbon, assuming all carbon in the sample is graphitized. In reality, both natural and artificial graphite commonly contain multiple carbon phases, including crystalline graphitic carbon, amorphous carbon, residual carbon, and amorphous carbon. This method cannot distinguish between crystalline and non-graphitized carbon, leading to inflated test results. For samples with low graphitization, the measured value may be more than 20% higher than the actual crystalline purity, causing grading distortion and trade pricing discrepancies. Furthermore, the degree of graphitization only characterizes the degree of lattice order, reflecting the density of crystal defects in the graphitized portion, not the overall mass proportion of crystalline graphitic carbon. Therefore, it cannot replace crystalline purity as a true indicator of purity. Currently, there are no national or industry standards that specifically define the purity of graphite crystals, leaving a significant gap in the existing standard system.

[0003] To address the aforementioned shortcomings, this application provides a method and system for detecting the purity of graphite crystals based on carbon phase separation. Graphite crystal purity is defined as the percentage of crystalline graphite carbon mass in the total carbon mass of the sample. This aims to achieve accurate separation and quantification of crystalline graphite carbon from amorphous carbon and other carbon phases, fundamentally solving problems such as inflated results, grading distortion, and trade pricing deviations associated with fixed-carbon methods. The method employs an X-ray diffraction technique combined with a peak-splitting algorithm. By performing background subtraction and nonlinear least-squares iterative peak-splitting on the diffraction pattern, the spectral signal is split into characteristic peaks of crystalline graphite and diffuse broad peaks of amorphous carbon. A peak area-purity standard curve is established using high-purity standard samples to directly quantify the mass fraction of crystalline graphite carbon. This method eliminates the need for high-temperature calcination, enabling non-destructive testing. It offers significant advantages such as high accuracy, good repeatability, and a short testing cycle, filling the gap in standard methods for detecting the true purity of graphite and providing scientific and reliable technical support for the selection of high-end manufacturing materials and export control supervision. Summary of the Invention

[0004] Existing technologies, such as the fixed carbon method, cannot distinguish between crystalline graphitic carbon and various carbon phases including amorphous carbon, residual carbon, and non-crystalline carbon, leading to inflated test results and causing classification distortion and trade pricing deviations. The graphitization degree method only characterizes the degree of lattice order and cannot reflect the overall mass proportion of crystalline graphitic carbon; neither can serve as an effective indicator of the true crystalline purity of graphite. Furthermore, there are currently no national or industry standards specifically defining and measuring the mass fraction of crystalline graphitic carbon, leaving a significant gap in the existing standard system. Therefore, this invention first proposes a graphite crystal purity detection method based on carbon phase separation, specifically including the following steps: S1: Obtain the X-ray diffraction pattern of the graphite sample to be tested in the region covering the diffraction peaks of the graphite characteristic crystal plane and the adjacent amorphous carbon diffuse scattering signal region. The X-ray diffraction pattern shall contain at least the overlapping signal of the graphite characteristic crystal plane diffraction peak and the amorphous carbon diffuse scattering peak. 2θ represents the diffraction angle. S2: Perform background subtraction and intensity normalization preprocessing on the X-ray diffraction pattern. Use an iterative polynomial fitting algorithm to subtract the background scattering signal and normalize the pattern intensity to the same scale to obtain the preprocessed diffraction pattern. S3: A nonlinear least squares fitting method based on the Levenberg-Marquardt algorithm is used. A peak shape model for the characteristic peak components of crystalline graphite is constructed using the Pseudo-Voigt function, and a peak shape model for the broad peak components of amorphous carbon is constructed using the Gaussian function. Peak-separation iterative fitting is performed on the preprocessed X-ray diffraction pattern. The iteration convergence condition is set as the rate of change of the sum of squared residuals being less than 1 × 10⁻⁶. -6 Or, once the preset maximum number of iterations is reached, the integrated area of ​​the characteristic peaks of crystalline graphite and the integrated area of ​​the broad peaks of amorphous carbon after fitting will be output. S4: Substitute the integral area of ​​the characteristic peak of crystalline graphite into the pre-established standard working curve. The standard working curve is established by processing at least three high-purity graphite standard samples with known crystal purity gradient distribution under the same scanning conditions through steps S1 to S3 to obtain the integral area of ​​the characteristic peak, and then using the least squares linear regression algorithm to fit the linear relationship between the peak area and the crystal purity. Calculate the mass fraction of crystalline graphite carbon in the sample to be tested based on the regression equation of the standard working curve. S5: Calculate and output the crystal purity test result of the graphite sample to be tested based on the ratio of the mass fraction of carbon in the crystalline graphite to the total carbon mass of the sample.

[0005] Furthermore, in step S2, the background subtraction adopts an iterative polynomial fitting algorithm, specifically: in the X-ray diffraction pattern, a background region that does not contain characteristic diffraction peaks is selected, and the background points are fitted with least squares using third- to fifth-order polynomials. After multiple iterations, abnormal background points that deviate from the fitted curve by more than a preset threshold are removed and refitted until the background fitting curve is stable. The final background curve is then output and subtracted from the original pattern.

[0006] Furthermore, in step S3, the Pseudo-Voigt function is a linear combination of the Gaussian and Lorentzian functions, and its mathematical expression is: ; Where x represents the diffraction angle coordinate, V(x) represents the peak intensity value of the Pseudo-Voigt function at angle x, η represents the Lorentzian proportion coefficient with a value range of 0-1, L(x) represents the intensity value of the Lorentzian function at angle x, and G(x) represents the intensity value of the Gaussian function at angle x; the nonlinear least squares fitting simultaneously optimizes the peak position, peak height, full width at half maximum (FWHM), and η value of the characteristic peaks of crystalline graphite, as well as the peak position, peak height, and FWHM of the amorphous carbon peaks, a total of 7 free parameters in each iteration.

[0007] Furthermore, in step S3, physical constraints are applied to the peak-splitting iterative fitting: the peak position of the characteristic peak component of crystalline graphite is constrained within the allowable deviation range of the diffraction peak position of the characteristic crystal plane of graphite, the peak position of the broad peak component of amorphous carbon is constrained within the range of the diffuse scattering peak position of amorphous carbon, and the half-width at half maximum (WHM) of the characteristic peak component of crystalline graphite is constrained to be smaller than the WHM of the broad peak component of amorphous carbon. Iteration results that exceed the constraints are judged as invalid fitting and trigger the refitting process.

[0008] Furthermore, in step S3, the peak-splitting iterative fitting also includes a goodness-of-fit evaluation sub-step: calculating the fitting determination coefficient and the standardized residual distribution, wherein the fitting determination coefficient is used to characterize the degree of interpretation of the original data by the fitting model; when the fitting determination coefficient is lower than the preset threshold or the standardized residual shows a non-random distribution pattern, the fitting is deemed unqualified, and the initial parameters of the characteristic peaks of crystalline graphite and the broad peaks of amorphous carbon are automatically adjusted and iterative fitting is re-executed until the goodness of fit meets the preset standard.

[0009] Furthermore, in step S4, the establishment of the standard working curve includes: selecting multiple high-purity graphite standard samples with gradient distributions of crystal purity, processing them through steps S1 to S3 to obtain the integral area of ​​the characteristic peaks of crystalline graphite for each standard sample, using the integral area of ​​the peaks as the independent variable X and the known crystal purity as the dependent variable Y, and using the least squares linear regression to fit the equation Y = aX + b, where a represents the slope of the regression equation and b represents the intercept of the regression equation; and calculating the correlation coefficient r, where r represents the Pearson correlation coefficient, used to characterize the tightness of the linear relationship between peak area and crystal purity; when the correlation coefficient meets the preset threshold, the standard working curve is determined to be valid; otherwise, the standard sample data is re-collected.

[0010] Furthermore, in step S5, the total carbon mass of the sample is determined by one of the following methods: Method 1, the sum of the integral area of ​​the characteristic peak of crystalline graphite and the integral area of ​​the broad peak of amorphous carbon obtained in step S3 is converted into the total carbon mass through a normalization calibration factor; Method 2, the total carbon mass of the sample is determined by independent thermogravimetric analysis or elemental analysis, and the total carbon mass is input into the quantitative calculation module to participate in the calculation of crystal purity.

[0011] According to another aspect of the present invention, a graphite crystal purity detection system based on carbon phase separation is also proposed, applied to the above-mentioned graphite crystal purity detection method based on carbon phase separation, comprising: The X-ray diffraction pattern acquisition module is used to perform X-ray diffraction scanning on the graphite sample to be tested, and to obtain a diffraction pattern that includes at least the diffraction peaks of the characteristic crystal planes of graphite and the diffuse scattering peaks of amorphous carbon. The spectrum preprocessing module is connected to the X-ray diffraction pattern acquisition module. It is used to perform background subtraction and intensity normalization preprocessing on the diffraction pattern. The background subtraction adopts an iterative polynomial fitting algorithm and outputs the preprocessed diffraction pattern. The peak fitting module is connected to the spectrum preprocessing module. It has a built-in nonlinear least squares fitting engine and uses the Levenberg-Marquardt algorithm. It constructs peak shape models of characteristic peaks of crystalline graphite and broad peaks of amorphous carbon using the Pseudo-Voigt function and Gaussian function, respectively. It performs iterative peak fitting with physical constraints on the preprocessed diffraction spectrum, sets the convergence threshold, and outputs the integral area of ​​each component. The quantitative calculation module is connected to the peak fitting processing module. It is used to store the pre-established standard working curve and substitute the integral area of ​​the characteristic peak of crystalline graphite output by the peak fitting processing module into the regression equation of the standard working curve to calculate the mass fraction of crystalline graphite carbon in the sample to be tested. The results output module, connected to the quantitative calculation module, is used to calculate and output the graphite crystal purity test results based on the ratio of the mass fraction of carbon in crystalline graphite to the total carbon mass of the sample.

[0012] Furthermore, the peak-splitting fitting module also includes an automatic fitting quality evaluation submodule, which is used to calculate the determination coefficient R for each fitting. 2 The sum of squared residuals and the standardized residual distribution, where R 2 This represents the coefficient of determination for fitting; when the fitting quality does not meet the preset threshold, the initial fitting parameters are automatically adjusted and refitting is triggered, while the trajectory of parameter changes in the iteration history is recorded for algorithm debugging and result tracking.

[0013] Furthermore, it also includes a standard curve dynamic maintenance module. This module is used to periodically input newly prepared standard sample data, dynamically update the regression coefficients of the standard working curve using a cumulative least squares recursive algorithm, and perform anomaly detection on the standard sample data based on Cook distance or leverage value statistics. Cook distance is used to measure the statistical influence of a single standard sample data point on the overall fit of the regression model, while leverage value is used to measure the degree of outlier status of the standard sample data point in the independent variable space. It automatically marks and removes abnormal standard sample points that have a significant impact on the regression model, ensuring the long-term stability and accuracy of the standard working curve.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: Firstly, this invention defines the purity of graphite crystallization, which is the percentage of the mass of crystalline graphite carbon in the total carbon mass of the sample. By combining X-ray diffraction with peak fitting algorithm, it achieves accurate separation and quantification of crystalline graphite carbon from carbon phases such as amorphous carbon, residual carbon, and amorphous carbon. This fundamentally solves the long-standing industry problems such as the inability of the fixed carbon method to distinguish carbon phases, which leads to falsely high detection results, distorted classification, and deviations in trade pricing. Secondly, this invention uses X-ray diffraction non-destructive testing technology, which does not require high-temperature burning to destroy the sample. Combined with nonlinear least square iterative peak fitting algorithm and standard curve external standard quantitative model, the detection can be completed in about 30 to 60 minutes. It has significant advantages such as high detection accuracy, good repeatability and convenient operation, which greatly improves the detection efficiency. Third, this invention fills the gap in the lack of a standard method for defining and measuring the carbon mass fraction of crystalline graphite as the true purity of graphite, both domestically and internationally. It provides a scientific and quantitative technical basis for the enforcement and supervision of the export of high-energy-consuming, high-polluting, and high-emission graphite, and at the same time provides reliable data support for the selection and quality control of graphite materials in high-end manufacturing fields such as photovoltaics, semiconductors, and nuclear engineering. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 The flowchart shows a method for determining the purity of graphite crystals based on carbon phase separation. Figure 2 This is a schematic diagram of an embodiment; Figure 3 This is a schematic diagram of an embodiment; Figure 4 This is a schematic diagram of a graphite crystal purity detection system based on carbon phase separation. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0018] The specific embodiments of the present invention will be described below.

[0019] Existing technologies, such as the fixed carbon method, cannot distinguish between crystalline graphitic carbon and various carbon phases including amorphous carbon, residual carbon, and non-crystalline carbon, leading to inflated test results and causing classification distortion and trade pricing deviations. The graphitization degree method only characterizes the degree of lattice order and cannot reflect the overall mass proportion of crystalline graphitic carbon; neither can serve as an effective indicator of the true crystalline purity of graphite. Furthermore, there are currently no national or industry standards specifically defining and measuring the mass fraction of crystalline graphitic carbon, leaving a significant gap in the existing standard system. This invention achieves precise separation and quantification of crystalline graphitic carbon and other carbon phases, such as amorphous carbon, through X-ray diffraction combined with peak fitting algorithms. This fundamentally solves the industry problems of inflated test results, classification distortion, and trade pricing deviations caused by the fixed carbon method. Moreover, it eliminates the need for high-temperature sample burning, has a short testing cycle, and high accuracy, filling the gap in standard testing methods for the true crystalline purity of graphite and providing reliable technical support for material selection in high-end manufacturing and export control supervision.

[0020] Example 1 like Figure 1 , Figure 2 and Figure 3 As shown, this invention proposes a method for detecting the purity of graphite crystals based on carbon phase separation. Specifically, it includes the following steps: S1: Obtain the X-ray diffraction pattern of the graphite sample under test in the region covering the diffraction peaks of the characteristic crystal planes of graphite and the diffuse scattering signal of adjacent amorphous carbon. The X-ray diffraction pattern within the scanning range must contain at least the overlapping signal of the diffraction peaks of the characteristic crystal planes of graphite and the diffuse scattering peaks of amorphous carbon, where 2θ represents the diffraction angle, and 2θ is set between 20° and 30°. Specifically, the graphite sample under test is ground through a 200-mesh standard sieve and then pressed into a pellet. The pellet is placed on the sample stage of the X-ray diffractometer, and the X-ray diffraction pattern of the sample is acquired within the set scanning range. This pattern contains the overlapping signal of the diffraction peaks of the characteristic crystal planes of graphite located near 26.4° with the diffuse scattering peak of amorphous carbon.

[0021] S2: Perform background subtraction and intensity normalization preprocessing on the X-ray diffraction pattern. An iterative polynomial fitting algorithm is used to subtract the background scattering signal, and the pattern intensity is normalized to the same scale to obtain the preprocessed diffraction pattern. Background subtraction uses an iterative polynomial fitting algorithm, specifically: Background regions without characteristic diffraction peaks are selected in the X-ray diffraction pattern; third- to fifth-order polynomials are used to perform least-squares fitting on the background points; after multiple iterations, abnormal background points deviating from the fitted curve by more than a preset threshold are removed, and the process is repeated until the background fitting curve stabilizes. The final background curve is then output and subtracted from the original pattern.

[0022] Specifically, background subtraction and intensity normalization were performed on the acquired X-ray diffraction patterns. An iterative polynomial fitting algorithm was used to subtract the background scattering signal. Specifically, background regions without characteristic diffraction peaks were selected in the diffraction pattern, and third- to fifth-order polynomials were used to perform least-squares fitting on the background points. After multiple iterations to remove abnormal background points, the final background curve was output and subtracted from the original pattern. The intensity of the background-subtracted pattern was then normalized to the same scale to obtain the preprocessed diffraction pattern.

[0023] S3: A nonlinear least squares fitting method based on the Levenberg-Marquardt algorithm is used. A peak shape model for the characteristic peak components of crystalline graphite is constructed using the Pseudo-Voigt function, and a peak shape model for the broad peak components of amorphous carbon is constructed using the Gaussian function. Peak-separation iterative fitting is performed on the preprocessed X-ray diffraction pattern. The iteration convergence condition is set as the rate of change of the sum of squared residuals being less than 1 × 10⁻⁶. -6 Alternatively, after reaching the preset maximum number of iterations, the integrated area of ​​the characteristic peaks of crystalline graphite and the integrated area of ​​the broad peaks of amorphous carbon are output after fitting.

[0024] Specifically, the Pseudo-Voigt function is a linear combination of the Gaussian and Lorentzian functions, and its mathematical expression is: ; Where x represents the diffraction angle coordinate, V(x) represents the peak intensity value of the Pseudo-Voigt function at angle x, η represents the Lorentzian proportion coefficient with a value range of 0-1, L(x) represents the intensity value of the Lorentzian function at angle x, and G(x) represents the intensity value of the Gaussian function at angle x; the nonlinear least squares fitting simultaneously optimizes the peak position, peak height, full width at half maximum (FWHM), and η value of the characteristic peaks of crystalline graphite, as well as the peak position, peak height, and FWHM of the amorphous carbon peaks, a total of 7 free parameters in each iteration.

[0025] A nonlinear least squares fitting method based on the Levenberg-Marquardt algorithm was employed. A peak shape model for the characteristic peak components of crystalline graphite was constructed using the Pseudo-Voigt function, and a peak shape model for the amorphous carbon peak components was constructed using the Gaussian function. Peak-separation iterative fitting was then performed on the preprocessed diffraction pattern. Physical constraints were also imposed on the peak-separation iterative fitting: the peak positions of the crystalline graphite characteristic peak components were constrained to the range of 26.4° ± 0.3°, and the peak positions of the amorphous carbon peak components were constrained to the range of 20°–25°. Furthermore, the full width at half maximum (FWHM) constraint for the crystalline graphite characteristic peak components was no greater than 1.5°, and the FWHM constraint for the amorphous carbon peak components was within the range of 3°–8°. Iteration results exceeding these constraints were deemed invalid fittings, triggering a refit process. The iteration convergence condition was set as the rate of change of the sum of squared residuals being less than 1 × 10⁻⁶. -6 Alternatively, after reaching the preset maximum number of iterations, the integrated area of ​​the characteristic peaks of crystalline graphite and the integrated area of ​​the broad peaks of amorphous carbon are output after fitting. After fitting, the coefficient of determination R² is calculated. When R² < 0.99, the initial parameters are automatically adjusted and refitted until the goodness-of-fit requirement is met. Peak-splitting iterative fitting also includes a goodness-of-fit evaluation sub-step: calculating the fitting coefficient of determination R². 2 and standardized residual distribution, where R 2 R represents the coefficient of determination, used to characterize the degree to which the fitted model interprets the original data; when R... 2 If the value is less than 0.99 or the standardized residuals show a non-random distribution pattern, the fitting is deemed unqualified. The initial parameters of the characteristic peaks of crystalline graphite and the broad peaks of amorphous carbon are automatically adjusted, and the iterative fitting is re-executed until the goodness of fit meets the preset standard.

[0026] S4: Substitute the integral area of ​​the characteristic peak of crystalline graphite into the pre-established standard working curve. The standard working curve is established by processing at least three high-purity graphite standard samples with known crystal purity gradient distributions under the same scanning conditions through steps S1 to S3 to obtain the integral area of ​​the characteristic peaks, and then using the least squares linear regression algorithm to fit the linear relationship between the peak area and the crystal purity. Calculate the mass fraction of crystalline graphite carbon in the sample to be tested based on the regression equation of the standard working curve.

[0027] The establishment of the standard working curve includes: selecting 5-8 high-purity graphite standard samples with a crystal purity gradient distribution within the range of 60%-100%, and obtaining the integral area of ​​the characteristic peaks of crystalline graphite for each standard sample through steps S1 to S3. Using the integral area as the independent variable X and the known crystal purity as the dependent variable Y, a least-squares linear regression equation Y = aX + b is used to fit the curve, where a represents the slope of the regression equation and b represents the intercept. The correlation coefficient r is calculated, where r represents the Pearson correlation coefficient, used to characterize the strength of the linear relationship between peak area and crystal purity. The standard working curve is considered valid when r ≥ 0.995; otherwise, standard sample data are re-collected. The mass fraction of crystalline graphite carbon in the sample to be tested is calculated based on this regression equation.

[0028] S5: Based on the ratio of the mass fraction of carbon in crystalline graphite to the total carbon mass of the sample, calculate and output the crystallinity purity test result of the graphite sample to be tested. The total carbon mass of the sample is determined by one of the following methods: Method 1, convert the sum of the integral area of ​​the characteristic peak of crystalline graphite and the integral area of ​​the broad peak of amorphous carbon obtained in step S3 into the total carbon mass through a normalization calibration factor; Method 2, determine the total carbon mass of the sample by independent thermogravimetric analysis or elemental analysis, and input the total carbon mass into the quantitative calculation module to participate in the crystallinity purity calculation. Specifically, the graphite crystallinity purity is calculated based on the ratio of the mass fraction of carbon in crystalline graphite to the total carbon mass of the sample, such as: mass of carbon in crystalline graphite / total carbon mass × 100%, and the test result is output.

[0029] In this embodiment, the purity of graphite crystallization, i.e., the percentage of crystalline graphite carbon mass in the total carbon mass of the sample, is clearly defined. Accurate separation and quantification of crystalline graphite carbon from amorphous carbon, residual carbon, and other carbon phases are achieved through X-ray diffraction combined with a peak fitting algorithm. This fundamentally solves the long-standing industry problem of inflated detection results, distorted classification, and trade pricing deviations caused by the inability of the fixed carbon method to distinguish carbon phases. Employing non-destructive X-ray diffraction technology eliminates the need for high-temperature burning to destroy the sample. Combined with a nonlinear least squares iterative peak fitting algorithm and a standard curve external standard quantitative model, detection can be completed in approximately 30 to 60 minutes. It boasts significant advantages such as high detection accuracy, good repeatability, and convenient operation, greatly improving detection efficiency.

[0030] Example 2 like Figure 4 As shown, this invention also proposes a graphite crystal purity detection system based on carbon phase separation, using a graphite crystal purity detection method based on carbon phase separation as described in Example 1, including the following: The X-ray diffraction pattern acquisition module is used to perform X-ray diffraction scanning on the graphite sample to be tested, and to obtain a diffraction pattern that includes at least the diffraction peaks of the characteristic crystal planes of graphite and the diffuse scattering peaks of amorphous carbon. The spectrum preprocessing module is connected to the X-ray diffraction pattern acquisition module. It is used to perform background subtraction and intensity normalization preprocessing on the diffraction pattern. The background subtraction adopts an iterative polynomial fitting algorithm and outputs the preprocessed diffraction pattern. The peak fitting module is connected to the spectrum preprocessing module. It has a built-in nonlinear least squares fitting engine and uses the Levenberg-Marquardt algorithm. It constructs peak shape models of characteristic peaks of crystalline graphite and broad peaks of amorphous carbon using the Pseudo-Voigt function and Gaussian function, respectively. It performs iterative peak fitting with physical constraints on the preprocessed diffraction spectrum, sets the convergence threshold, and outputs the integral area of ​​each component. The quantitative calculation module is connected to the peak fitting processing module. It is used to store the pre-established standard working curve and substitute the integral area of ​​the characteristic peak of crystalline graphite output by the peak fitting processing module into the regression equation of the standard working curve to calculate the mass fraction of crystalline graphite carbon in the sample to be tested. The results output module, connected to the quantitative calculation module, is used to calculate and output the graphite crystal purity test results based on the ratio of the mass fraction of carbon in crystalline graphite to the total carbon mass of the sample.

[0031] In this embodiment, the peak-splitting fitting processing module further includes an automatic fitting quality evaluation submodule, which is used to calculate the determination coefficient R for each fitting. 2 The sum of squared residuals and the standardized residual distribution, where R 2 This represents the coefficient of determination for fitting; when the fitting quality does not meet the preset threshold, the initial fitting parameters are automatically adjusted and refitting is triggered, while the trajectory of parameter changes in the iteration history is recorded for algorithm debugging and result tracking.

[0032] It also includes a standard curve dynamic maintenance module, which is used to periodically input newly prepared standard sample data, dynamically update the regression coefficients of the standard working curve through a cumulative least squares recursive algorithm, and perform anomaly detection on the standard sample data based on Cook distance or leverage value statistics. Cook distance is used to measure the statistical influence of a single standard sample data point on the overall fit of the regression model, and leverage value is used to measure the degree of outlier of the standard sample data point in the independent variable space. It automatically marks and removes abnormal standard sample points that have a significant impact on the regression model, ensuring the long-term stability and accuracy of the standard working curve.

[0033] This study fills the gap in the lack of a standardized method for determining the carbon mass fraction of crystalline graphite as the true purity of graphite, both domestically and internationally. It provides a scientific and quantitative technical basis for the enforcement and supervision of the export of high-energy-consuming, high-polluting, and high-efficiency graphite, and at the same time provides reliable data support for the selection and quality control of graphite materials in high-end manufacturing fields such as photovoltaics, semiconductors, and nuclear engineering.

[0034] Example 3 An electronic device, comprising: Processor and memory; The processor executes the steps of a graphite crystal purity detection method based on carbon phase separation, as described in any of Examples 1, by calling programs or instructions stored in memory.

[0035] Example 4 A computer-readable storage medium includes computer program instructions that cause a computer to perform the steps of a graphite crystal purity detection method based on carbon phase separation, as described in any of Embodiment 1.

[0036] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting the purity of graphite crystals based on carbon phase separation, characterized in that, Includes the following steps: S1: Obtain the X-ray diffraction pattern of the graphite sample to be tested in the region covering the diffraction peaks of the graphite characteristic crystal plane and the adjacent amorphous carbon diffuse scattering signal region. The X-ray diffraction pattern contains at least the overlapping signal of the graphite characteristic crystal plane diffraction peaks and the amorphous carbon diffuse scattering peaks. 2θ represents the diffraction angle. S2: Perform background subtraction and intensity normalization preprocessing on the X-ray diffraction pattern. Use an iterative polynomial fitting algorithm to subtract the background scattering signal and normalize the pattern intensity to the same scale to obtain the preprocessed diffraction pattern. S3: A nonlinear least squares fitting method based on the Levenberg-Marquardt algorithm is used. A peak shape model for the characteristic peak components of crystalline graphite is constructed using the Pseudo-Voigt function, and a peak shape model for the broad peak components of amorphous carbon is constructed using the Gaussian function. The preprocessed X-ray diffraction pattern is then subjected to peak-separation iterative fitting. The iterative convergence condition is set as the rate of change of the sum of squared residuals being less than 1 × 10⁻⁶. -6 Or, once the preset maximum number of iterations is reached, the integrated area of ​​the characteristic peaks of crystalline graphite and the integrated area of ​​the broad peaks of amorphous carbon after fitting will be output. S4: Substitute the integral area of ​​the characteristic peak of the crystalline graphite into the pre-established standard working curve. The standard working curve is established by at least three high-purity graphite standard samples with known crystal purity gradient distributions, which are processed by the steps S1 to S3 under the same scanning conditions to obtain the integral area of ​​the characteristic peaks, and then the least squares linear regression algorithm is used to fit the linear relationship between the peak area and the crystal purity. The mass fraction of crystalline graphite carbon in the sample to be tested is calculated according to the regression equation of the standard working curve. S5: Calculate and output the crystal purity test result of the graphite sample to be tested based on the ratio of the mass fraction of carbon in the crystalline graphite to the total carbon mass of the sample.

2. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 1, characterized in that, In step S2, the background subtraction adopts an iterative polynomial fitting algorithm, specifically: a background region without characteristic diffraction peaks is selected in the X-ray diffraction pattern, and the background points are fitted with least squares using third- to fifth-order polynomials. After multiple iterations, abnormal background points that deviate from the fitted curve by more than a preset threshold are removed and refitted until the background fitting curve is stable. The final background curve is then output and subtracted from the X-ray diffraction pattern.

3. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 1, characterized in that, In step S3, the Pseudo-Voigt function is a linear combination of the Gaussian and Lorentzian functions, and its mathematical expression is: ； Where x represents the diffraction angle coordinate, V(x) represents the peak intensity value of the Pseudo-Voigt function at angle x, η represents the Lorentzian proportion coefficient with a value range of 0-1, L(x) represents the intensity value of the Lorentzian function at angle x, and G(x) represents the intensity value of the Gaussian function at angle x; the nonlinear least squares fitting simultaneously optimizes the peak position, peak height, full width at half maximum (FWHM), and η value of the characteristic peaks of crystalline graphite, as well as the peak position, peak height, and FWHM of the amorphous carbon peaks, a total of 7 free parameters in each iteration.

4. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 3, characterized in that, In step S3, the peak-splitting iterative fitting also applies physical constraints: the peak position of the characteristic peak component of crystalline graphite is constrained within the allowable deviation range of the diffraction peak position of the characteristic crystal plane of graphite, the peak position of the broad peak component of amorphous carbon is constrained within the range of the diffuse scattering peak position of amorphous carbon, and the half-width at half maximum (WHM) of the characteristic peak component of crystalline graphite is constrained to be smaller than the WHM of the broad peak component of amorphous carbon. Iteration results that exceed the constraints are judged as invalid fitting and trigger the refitting process.

5. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 4, characterized in that, In step S3, the peak-splitting iterative fitting further includes a goodness-of-fit evaluation sub-step: calculating the coefficient of determination and the standardized residual distribution, wherein the coefficient of determination is used to characterize the degree of interpretation of the original data by the fitting model; when the coefficient of determination is lower than a preset threshold or the standardized residuals show a non-random distribution pattern, the fitting is deemed unqualified, and the initial parameters of the characteristic peaks of crystalline graphite and the broad peaks of amorphous carbon are automatically adjusted before iterative fitting is re-executed until the goodness of fit meets the preset standard.

6. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 1, characterized in that, In step S4, the establishment of the standard working curve includes: selecting multiple high-purity graphite standard samples with gradient distributions of crystal purity, processing them through steps S1 to S3 to obtain the integral area of ​​the characteristic peaks of crystalline graphite for each standard sample, using the integral area of ​​the peaks as the independent variable X and the known crystal purity as the dependent variable Y, and using the least squares linear regression to fit the equation Y = aX + b, where a represents the slope of the regression equation and b represents the intercept of the regression equation; and calculating the correlation coefficient r, where r represents the Pearson correlation coefficient, used to characterize the tightness of the linear relationship between peak area and crystal purity; when the correlation coefficient meets a preset threshold, the standard working curve is determined to be valid; otherwise, the standard sample data is re-collected.

7. The method for detecting the purity of graphite crystals based on carbon phase separation according to claim 5, characterized in that, In step S5, the total carbon mass of the sample is determined by one of the following methods: Method 1, the sum of the integral area of ​​the characteristic peak of crystalline graphite and the integral area of ​​the broad peak of amorphous carbon obtained in step S3 is converted into the total carbon mass through a normalization calibration factor; Method 2, the total carbon mass of the sample is determined by independent thermogravimetric analysis or elemental analysis, and the total carbon mass is input into the quantitative calculation module to participate in the calculation of crystal purity.

8. A graphite crystal purity detection system based on carbon phase separation, applicable to the graphite crystal purity detection method based on carbon phase separation as described in any one of claims 1 to 7, characterized in that, include: The X-ray diffraction pattern acquisition module is used to perform X-ray diffraction scanning on the graphite sample to be tested, and to obtain a diffraction pattern that includes at least the diffraction peaks of the characteristic crystal planes of graphite and the diffuse scattering peaks of amorphous carbon. The spectrum preprocessing module is connected to the X-ray diffraction pattern acquisition module and is used to perform background subtraction and intensity normalization preprocessing on the diffraction pattern. The background subtraction adopts an iterative polynomial fitting algorithm and outputs the preprocessed diffraction pattern. The peak fitting module is connected to the spectrum preprocessing module. It has a built-in nonlinear least squares fitting engine and uses the Levenberg-Marquardt algorithm. It constructs peak shape models of crystalline graphite characteristic peaks and amorphous carbon broad peaks with Pseudo-Voigt and Gaussian functions, respectively. It performs iterative peak fitting with physical constraints on the preprocessed diffraction spectrum, sets the convergence threshold, and outputs the integral area of ​​each component. The quantitative calculation module is connected to the peak fitting processing module. It is used to store the pre-established standard working curve and substitute the integral area of ​​the characteristic peak of crystalline graphite output by the peak fitting processing module into the regression equation of the standard working curve to calculate the mass fraction of crystalline graphite carbon in the sample to be tested. The result output module, connected to the quantitative calculation module, is used to calculate and output the graphite crystal purity test result based on the ratio of the mass fraction of carbon in crystalline graphite to the total carbon mass of the sample.

9. The graphite crystallization purity detection system based on carbon phase separation according to claim 8, characterized in that, The peak-splitting fitting module also includes an automatic fitting quality evaluation submodule, which is used to calculate the determination coefficient R for each fitting. 2 The sum of squared residuals and the standardized residual distribution, where R 2 This represents the coefficient of determination for fitting; when the fitting quality does not meet the preset threshold, the initial fitting parameters are automatically adjusted and refitting is triggered, while the trajectory of parameter changes in the iteration history is recorded for algorithm debugging and result tracking.

10. The graphite crystallization purity detection system based on carbon phase separation according to claim 8, characterized in that, It also includes a standard curve dynamic maintenance module, which is used to periodically input newly prepared standard sample data, dynamically update the regression coefficients of the standard working curve through a cumulative least squares recursive algorithm, and perform anomaly detection on the standard sample data based on Cook distance or leverage value statistics. The Cook distance is used to measure the statistical influence of a single standard sample data point on the overall fit of the regression model, and the leverage value is used to measure the degree of outlier of the standard sample data point in the independent variable space. It automatically marks and removes abnormal standard sample points that have a significant impact on the regression model, ensuring the long-term stability and accuracy of the standard working curve.

Images