Detection method and detection system for semiconductor process

Abstract

The invention relates to the technical field of semiconductor process detection, and discloses a semiconductor process detection method and system, and the method comprises the steps: carrying out the pretreatment of a semiconductor wafer, enabling the surface of the semiconductor wafer to be free of residual pollutants, and employing a chemical cleaning and drying technology to maintain the completeness of a crystal; the method comprises the following steps: selecting a plurality of isotopes which have the same chemical property as a measured target element but have obvious natural abundance difference as standard samples, and uniformly doping an isotope solution with known concentration into the plurality of standard samples; according to the method, target element isotope labels with known concentrations are introduced into a sample, then the SIMS is used for detecting the relative signal intensity of natural isotopes and doped isotopes at the same time, so that internal correction is achieved, the ionization efficiency difference caused by different matrixes is counteracted, and due to the fact that the chemical behaviors of different isotopes in the same matrix are basically consistent, the ionization efficiency is greatly improved. The matrix effect can be greatly reduced, so that the accuracy and repeatability of quantitative analysis are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor process detection, in particular to a semiconductor process detection method and system. BACKGROUND

[0002] Semiconductor detection refers to a series of processes of detecting and measuring wafer, thin film, device structure, interface and electrical properties in the process of semiconductor manufacturing and packaging to ensure that the physical, chemical and electrical parameters meet the process requirements. The purpose of semiconductor detection is to comprehensively understand and control various key parameters in the manufacturing process, and to discover small problems in the process in time, so as to prevent the accumulation of micro-defects from causing functional defects or performance failure of the final product. Detection not only helps to improve product yield and shorten research and development cycle, but also provides reliable data support to help engineers optimize and improve the process bottleneck.

[0003] Semiconductor detection usually includes online and offline detection, from material defects, pattern defects, surface roughness, doping uniformity to electrical performance parameters (such as leakage current, threshold voltage, response speed, etc.). At present, for the detection in the doping and ion implantation process, the detection means mainly depends on secondary ion mass spectrometry (SIMS), but SIMS is easily affected by the analysis matrix effect when performing quantitative analysis, and the signal response of the same dopant may be different in different matrices, so the calibration process becomes complex and prone to measurement errors. SUMMARY

[0004] In view of the problems existing in the prior art, the purpose of the present application is to provide a semiconductor process detection method and system, which can accurately determine whether the signal response in different matrices belongs to the same type of dopant when using SIMS for quantitative analysis, thereby achieving high accuracy in detecting doping quantitatively.

[0005] In order to achieve the above purpose, the present application provides the following technical scheme: a semiconductor process detection method, comprising: Pretreating the semiconductor wafer to make its surface free of residual contaminants, and using chemical cleaning and drying process to maintain the integrity of the crystal; Selecting a plurality of isotopes with the same chemical properties as the measured target element but with obvious natural abundance difference as standard samples, and uniformly doping the isotope solution with known concentration into the plurality of standard samples; Using the plurality of standard samples, measuring the ion signal ratio at different concentrations by SIMS, and constructing a response curve of the target element, which can reflect the proportional relationship between the internal standard isotope and the target isotope ion signal under the same matrix; The same SIMS measurement parameters as in constructing the response curve are used to perform multi-point scanning detection on the measurement target to obtain ion signal data of the target isotope and the internal standard isotope, and the signal ratio of the internal standard isotope and the target isotope is calculated from the collected ion signal data. The actual concentration of the target element in the sample to be measured is determined by the corresponding relationship between the signal ratio and the standard curve.

[0006] In some embodiments, the specific way of calculating the signal ratio of the internal standard isotope and the target isotope is to perform regression fitting with the known concentration of the target element as the independent variable and the signal ratio as the dependent variable to obtain a mathematical model relationship: Y = a × X + b, where Y represents the signal ratio, X represents the known concentration of the target element, and a and b are regression coefficients.

[0007] In some embodiments, the specific way of determining the actual concentration of the target element in the sample to be measured is to draw a standard curve, obtain a determination coefficient R2 by regression analysis, the R2 value ranges between 0 and 1, the closer the value is to 1, the higher the matching degree of the standard curve, and set a dopant identification threshold, compare the determination coefficient with the dopant identification threshold, and make a corresponding response according to the comparison result.

[0008] In some embodiments, when the determination coefficient is greater than or equal to the dopant identification threshold, it indicates that the matching degree of the measurement target and the standard sample is high, and it is determined that the signal response of the measurement target and the standard sample belongs to the same type of dopant; and when the determination coefficient is less than the dopant identification threshold, it indicates that the matching degree of the measurement target and the standard sample is low, and it is determined that the signal response of the measurement target and the standard sample does not belong to the same type of dopant.

[0009] In some embodiments, the specific way of drawing a standard curve is to use SIMS to measure the signal intensity of the target element and the internal standard element in each standard sample, calculate the sample signal ratio, and mark all points on the graph with the known concentration as the abscissa and the calculated signal ratio as the ordinate, and draw a line to make these points fall on this line, and set this line as the standard curve.

[0010] In some embodiments, the specific way of obtaining the determination coefficient R2 is to calculate the total fluctuation between all data points and their mean, and then compare the actual deviation between the data points and the regression straight line prediction value. The specific method is to first calculate the total deviation square sum, which measures the overall dispersion degree between all observation values and the sample mean; then calculate the residual square sum, which represents the difference between each actual data point and the regression prediction value, and use the least square method to make the value of the residual square sum reach the minimum, and express the relationship in a mathematical formula, that is, by comparing the proportion of the prediction error and the overall error, the determination coefficient R2 is obtained.

[0011] In some embodiments, a close determination threshold value is set which is less than the dopant determination threshold value, when the determination coefficient is less than the dopant determination threshold value, the determination coefficient is compared with the close determination threshold value, and when the determination coefficient is greater than the close determination threshold value, the ion beam parameter optimization strategy is executed.

[0012] In some embodiments, the ion beam parameter optimization strategy comprises adjusting the ion beam acceleration voltage, increasing or decreasing the initial parameter value by 5% to 10% respectively, performing group measurement, performing standard sample measurement on each group of adjustment data and calculating the respective standard curve R2, comparing the stability and linear range of the signal ion to the internal standard ion ratio under different energies, selecting the voltage condition that can significantly improve R2, and making a small range adjustment to the incident angle to observe its effect on the secondary ion generation efficiency, adjusting the beam spot size to reduce the difference of local effects on the sample surface, testing the standard curve under different incident angles and beam spot conditions respectively, and selecting the parameter combination that can make R2 further close to or even reach the dopant determination threshold value.

[0013] The application further provides the following technical solutions: a semiconductor process detection system comprises: A target processing module comprises pre-processing a semiconductor wafer so that the surface is free of residual contaminants, and using a chemical cleaning and drying process to maintain the integrity of the crystal; An internal standard incorporation module comprises selecting a plurality of isotopes with the same chemical properties as the target element to be measured but with significant natural abundance differences as standard samples, and uniformly incorporating a known concentration of isotope solution into the plurality of standard samples; A curve construction module comprises using the plurality of standard samples to obtain ion signal ratios at different concentrations through SIMS measurement respectively, and constructing a response curve of the target element, wherein the response curve can reflect the proportional relationship between the internal standard isotope and the target isotope ion signal under the same matrix; A target determination module comprises using the same SIMS measurement parameters as when the response curve is constructed to perform multi-point scanning detection on the measurement target to obtain ion signal data of the target isotope and the internal standard isotope, and calculating the signal ratio of the internal standard isotope to the target isotope from the collected ion signal data, and determining the actual concentration of the target element in the sample to be measured through the corresponding relationship between the signal ratio and the standard curve.

[0014] The application further provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the semiconductor process detection method.

[0015] Compared with the prior art, the technical solutions provided by the application have the following beneficial effects: One, the present application realizes internal correction by introducing a known concentration of isotope-labeled target element into the sample, and then detecting the relative signal intensity of natural isotopes and incorporated isotopes at the same time by SIMS, so as to offset the ionization efficiency difference caused by different matrices. Since the chemical behavior of different isotopes in the same matrix is basically the same, the matrix effect can be greatly reduced, thereby improving the accuracy and repeatability of quantitative analysis, so as to accurately determine whether the signal response in different matrices belongs to the same type of dopant.

[0016] Secondly, the present application realizes overall optimization of the ion beam parameters by executing an ion beam parameter optimization strategy and repeatedly confirming the influence of each adjustment on the determination coefficient by using a closed-loop feedback mechanism, so as to ensure that the matching degree of the signal response of the target element in the sample to be measured and the standard sample reaches the expected standard, and then reliably determine the concentration of the target element, so as to remove the influence of ion beam energy on the matrix effect and avoid interference with the judgment of whether the signal response belongs to the same type of dopant. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The logical diagram of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0019] It can be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of one element can be one, and in another embodiment, the number of the element can be multiple, and the term "one" cannot be understood as a limitation on the number.

[0020] The present application provides a semiconductor process detection method, as shown in Figure 1 The method comprises the following steps: Step one, pretreat the semiconductor wafer to make its surface free of residual contaminants, and use chemical cleaning and drying process to maintain the integrity of the crystal; Step two, select a plurality of isotopes with the same chemical properties as the measured target element but with obvious natural abundance difference as standard samples, and uniformly incorporate a known concentration of isotope solution (or powder) into the plurality of standard samples, specifically by using chemical deposition method or ion implantation method to make it fully mixed with the original matrix. It should be noted that the amount of incorporation needs to be accurately controlled to ensure that the internal standard and the measured element are uniformly distributed in the whole sample. Step three, using multiple standard samples, the ion signal ratio at different concentrations is obtained by SIMS measurement respectively, and the response curve of the target element is constructed, which can reflect the proportional relationship between the internal standard isotope and the target isotope ion signal under the same matrix; Step four, using the same SIMS measurement parameters as when constructing the response curve, the measurement target is scanned and detected at multiple points to obtain the ion signal data of the target isotope and the internal standard isotope, and the signal ratio of the internal standard isotope and the target isotope is calculated from the collected ion signal data. Through the corresponding relationship between the signal ratio and the standard curve, the actual concentration of the target element in the sample to be measured is determined.

[0021] In the isotope dilution method, the ion signal intensities of the target element and the internal standard isotope are measured simultaneously, and the specific way to calculate the signal ratio of the internal standard isotope and the target isotope is: the known concentration of the target element is taken as the independent variable, and the signal ratio is taken as the dependent variable for regression fitting to obtain the mathematical model relationship: Y=a×X+b, where Y represents the signal ratio, X represents the known concentration of the target element, and a and b are regression coefficients.

[0022] The specific way to determine the actual concentration of the target element in the sample to be measured is: draw the standard curve, obtain the determination coefficient R2 through regression analysis, the R2 value ranges from 0 to 1, the closer the value is to 1, the higher the matching degree of the standard curve, and set a dopant identification threshold. Compare the determination coefficient with the dopant identification threshold, and make a corresponding response according to the comparison result. When the determination coefficient is greater than or equal to the dopant identification threshold, it indicates that the matching degree of the measurement target and the standard sample is high, and since the chemical behavior of different isotopes in the same matrix is basically consistent, the matrix effect can be greatly reduced, so that it can be judged that the signal response of the measurement target and the standard sample belongs to the same type of dopant. When the determination coefficient is less than the dopant identification threshold, it indicates that the matching degree of the measurement target and the standard sample is low, and it is judged that the signal response of the measurement target and the standard sample does not belong to the same type of dopant.

[0023] The specific way to draw the standard curve is to use SIMS to measure the signal intensity of the target element and the internal standard element in each standard sample, calculate the sample signal ratio, and take the known concentration as the abscissa and the calculated signal ratio as the ordinate. Mark all points on the graph, and use a line (straight line or suitable curve) to make these points fall on this line, and set this line as the standard curve.

[0024] The specific way to get the determination coefficient R2 is to calculate the total fluctuation between all data points and their mean value, and then compare the actual deviation between the data points and the predicted value of the regression straight line. The specific method is to first calculate the total deviation square sum, which measures the overall dispersion between all observations and the sample mean; then calculate the residual square sum, which indicates the difference between each actual data point and the regression predicted value, and use the least squares method to minimize the value of the residual square sum. The relationship between the two is expressed in mathematical formula, that is, by comparing the proportion of prediction error and overall error, the determination coefficient R2 is obtained.

[0025] The above method realizes internal correction by introducing a known concentration of isotope-labeled target element into the sample, and then detecting the relative signal intensity of natural isotopes and doped isotopes at the same time by SIMS, so as to offset the ionization efficiency difference caused by different matrixes. Since the chemical behavior of different isotopes in the same matrix is basically the same, the matrix effect can be greatly reduced, thereby improving the accuracy and repeatability of quantitative analysis.

[0026] For example, five standard samples with known concentrations are prepared, and the concentrations (in ppm) are 10, 20, 30, 40 and 50 respectively. After measurement by SIMS method, the ion signal data of the target element and the internal standard element are obtained, and the corresponding signal ratio is calculated. For example, after multiple measurements, the signal ratio data obtained are 10.1, 17.9, 26.3, 33.8 and 41.9 respectively. Taking the concentration as the independent variable (X) and the signal ratio as the dependent variable (Y), the data points (10, 10.1), (20, 17.9), (30, 26.3), (40, 33.8), (50, 41.9) are input into the regression analysis software (such as MATLAB or Python). The software will automatically use the least squares method to fit these points, and the best fitting straight line can be represented by the mathematical model: Y = a × X + b. In this example, after calculation, the slope a and the intercept b of the fitting straight line are approximately 0.8 and 2 respectively, and the fitting model is: Y = 0.8X + 2. To evaluate the goodness of fit of the model, the determination coefficient R2 is calculated. The significance of the determination coefficient is to measure the degree to which the regression model can explain the total data variation, and the calculation process is as follows: First, calculate the mean value of the observations For the data in this example, = (10.1 + 17.9 + 26.3 + 33.8 + 41.9) / 5 ≈ 26.0; Calculate the total deviation square sum (SS_tot), also known as the total variation, which is the sum of the squares of the differences between each observation and the mean value: SS_tot = (10.1 - 26.0)2+ (17.9 - 26.0)2+ (26.3 - 26.0)2+ (33.8 - 26.0)2+ (41.9 - 26.0)2 ≈ ( - 15.9)2+ ( - 8.1)2+ (0.3)2+ (7.8)2+ (15.9)2 ≈ 252.8 + 65.6 + 0.09 + 60.8 + 252.8 = 631.09 The predicted value Y_pred is calculated for each X using the fitted linear model, and the sum of squares of the residuals (SS_res) between the predicted value and the actual observed value is calculated: For X = 10, Y_pred = 0.8 * 10 + 2 = 10; residual = 10.1 - 10 = 0.1, residual square = 0.01.

[0027] For X = 20, Y_pred = 0.8 * 20 + 2 = 18; residual = 17.9 - 18 = -0.1, residual square = 0.01.

[0028] For X = 30, Y_pred = 0.8 * 30 + 2 = 26; residual = 26.3 - 26 = 0.3, residual square = 0.09.

[0029] For X = 40, Y_pred = 0.8 * 40 + 2 = 34; residual = 33.8 - 34 = -0.2, residual square = 0.04.

[0030] For X = 50, Y_pred = 0.8 * 50 + 2 = 42; residual = 41.9 - 42 = -0.1, residual square = 0.01.

[0031] In summary, SS_res = 0.01 + 0.01 + 0.09 + 0.04 + 0.01 = 0.16. And according to the formula R2= 1 – (SS_res / SS_tot), substituting the above values, R2= 1–(0.16 / 631.09) ≈ 1–0.000253 ≈ 0.99975. In practical applications, the higher R2is (closer to 1), the better the standard curve fits the experimental data. In order to determine the concentration of the target element in the sample to be tested, in step four, the same SIMS measurement parameters are used to scan multiple points on the sample to be tested, and the signal ratio of the sample to be tested is calculated. Substitute the measured ratio into the linear model Y = 0.8X + 2 obtained above, and solve for the concentration X. In addition, the calculated R2needs to be compared with the pre-set dopant identification threshold. When R2is greater than or equal to the threshold, it can be confirmed that the sample to be tested and the standard sample have a high degree of matching in terms of chemical matrix and signal response, indicating that the ion signal response belongs to the same type of dopant, and can be used to accurately determine the concentration of the target element; if R2is lower than the threshold, it indicates that the matching degree is insufficient, and thus the standard curve may need to be recalibrated or other possible influencing factors between samples need to be considered.

[0032] Further, set a near identification threshold less than the dopant identification threshold, for example, set the dopant identification threshold to 0.9, and set the near identification threshold to 0.8. When the determination coefficient is less than the dopant identification threshold, compare the determination coefficient with the near identification threshold. When the determination coefficient is greater than the near identification threshold, it indicates that although the determination coefficient does not meet the standard for being identified as the same type of dopant, it is close to the standard, and thus the ion beam parameter optimization strategy is executed. The ion beam parameter optimization strategy includes adjusting the ion beam acceleration voltage, increasing or decreasing the initial parameter value by 5% to 10% respectively, performing grouped measurement, measuring the standard sample for each group of adjusted data and calculating the respective standard curve R2, comparing the stability and linear range of the signal ion to the internal standard ion ratio under different energies, selecting the voltage condition that can significantly improve R2, and adjusting the incident angle in a small range (e.g. ±5°) to observe its effect on the secondary ion generation efficiency, adjusting the beam spot size (e.g. by modifying the focusing mirror settings or using different apertures) to reduce the difference in local effects on the sample surface, testing the standard curve under different incident angles and beam spot conditions, and selecting the parameter combination that can make R2further approach or even reach 0.9 (dopant identification threshold).

[0033] The reason for performing the above strategy is that the sputtering process of SIMS is affected by parameters such as ion beam energy, incident angle and beam current density, which may also be modulated by matrix effects. By optimizing the ion beam energy (for example, using a lower energy to reduce sputtering-related matrix dependence), precisely controlling the beam current density, and adjusting the incident angle, the ion generation efficiency in the sample to be measured can be improved, and the interference of different matrices on the ionization process can be reduced. By systematically adjusting the ion beam energy, beam current, incident angle, beam spot size and scanning speed, and experimentally verifying the combination of each parameter, the measurement conditions are gradually optimized. The influence of each adjustment on R2 is repeatedly confirmed by using a closed-loop feedback mechanism, and the overall optimization of the ion beam parameters is finally achieved, so as to ensure that the signal response of the target element in the sample to be measured matches the standard sample to the expected standard, and the concentration of the target element can be reliably determined.

[0034] In summary, the present application aims to design a semiconductor process detection method, which is susceptible to the influence of analysis matrix effects when SIMS is used for quantitative analysis, so that the signal responses of the same dopant in different matrices are different. The present application introduces a known concentration of isotopically labeled target element into the sample, and then uses SIMS to detect the relative signal intensity of the natural isotope and the doped isotope, thereby achieving internal correction and offsetting the ionization efficiency difference caused by different matrices. Since the chemical behavior of different isotopes in the same matrix is basically the same, the matrix effect can be greatly reduced, thereby improving the accuracy and repeatability of quantitative analysis, so as to accurately determine whether the signal response in different matrices belongs to the same type of dopant. Furthermore, by performing an ion beam parameter optimization strategy, the influence of each adjustment on the determination coefficient is repeatedly confirmed by using a closed-loop feedback mechanism, and the overall optimization of the ion beam parameters is finally achieved, so as to ensure that the signal response of the target element in the sample to be measured matches the standard sample to the expected standard, and the concentration of the target element can be reliably determined, so as to remove the influence of ion beam energy on matrix effects and avoid interference with the judgment of the signal response belonging to the same type of dopant.

[0035] The present application provides a semiconductor process detection method and a detection system, comprising: The target processing module includes pre-treating the semiconductor wafer to remove residual contaminants from its surface, and using a chemical cleaning and drying process to maintain the integrity of the crystal; The internal standard doping module includes selecting a plurality of isotopes with the same chemical properties as the target element to be measured but with significant natural abundance differences as standard samples, and uniformly doping a known concentration of isotope solution into the plurality of standard samples; The curve construction module comprises the following steps: using a plurality of standard samples, respectively obtaining ion signal ratios at different concentrations by SIMS measurement, and constructing a response curve of a target element, wherein the response curve can reflect the proportional relationship between the internal standard isotope and the target isotope ion signal under the same matrix; The target determination module comprises the following steps: using the same SIMS measurement parameters as those used in constructing the response curve, performing multi-point scanning detection on the measurement target to obtain ion signal data of the target isotope and the internal standard isotope, calculating the signal ratio of the internal standard isotope and the target isotope from the collected ion signal data, and determining the actual concentration of the target element in the sample to be measured by the corresponding relationship between the signal ratio and the standard curve.

[0036] The processes described above with reference to the flowcharts can be implemented as a computer software program according to the embodiments of the present application. The embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, and the computer program comprises program codes for executing the method shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a detachable medium. When the computer program is executed by a central processing unit, the above-mentioned functions defined in the method of the present application are executed. It should be noted that the computer readable medium of the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more conductive segments, a portable computer diskette, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, wherein the computer readable program code is carried. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, device or apparatus. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to a wireless segment, a wire segment, an optical cable, an RF cable, etc., or any suitable combination of the above.

[0037] The computer program product of the present application can be a computer program product, which is a machine-readable medium (or media) having stored therein some code (i.e., some computer code or software) that, when executed by a machine, causes the machine to perform any of the functions disclosed herein. Note that the computer program product can be a non-transitory computer program product. The term "non-transitory" is used herein to exclude only those types of computer-readable media that store data temporally (e.g., cache and server memory). Excluded from the term "non-transitory" are, for example, signals, waves, and other forms of electromagnetic radiation, whether modulated or unmodulated, that do not themselves represent a result of a process (e.g., a machine learning model) applied to data, and that do not consist of at least one repeated use of a data signal. Note also that, while the computer program product can be a machine-readable medium, it need not be, and the described functionality can be fully implemented without a machine-readable medium.

[0038] Those skilled in the art should understand that the above description is only a specific implementation of the present application, and the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for semiconductor process inspection, characterized in that, include: Semiconductor wafers are pretreated to remove residual contaminants from their surfaces, and chemical cleaning and drying processes are used to maintain the integrity of the crystals. Multiple isotopes with the same chemical properties as the target element but with significant differences in natural abundance were selected as standard samples, and isotope solutions of known concentrations were uniformly incorporated into the multiple standard samples. Using multiple standard samples, the ratio of ion signals at different concentrations was obtained by SIMS measurement, and the response curve of the target element was constructed. The response curve can reflect the ratio of the ion signals of the internal standard isotope and the target isotope under the same matrix. Using the same SIMS measurement parameters as when constructing the response curve, multi-point scanning detection was performed on the target to obtain ion signal data of the target isotope and the internal standard isotope. The signal ratio of the internal standard isotope to the target isotope was calculated from the collected ion signal data. The actual concentration of the target element in the sample was determined by the correspondence between the signal ratio and the standard curve.

2. The semiconductor process inspection method according to claim 1, characterized in that, The specific method for calculating the signal ratio between the internal standard isotope and the target isotope is as follows: using the known concentration of the target element as the independent variable and the signal ratio as the dependent variable, a regression fitting is performed to obtain the mathematical model relationship: Y = a × X + b, where Y represents the signal ratio, X represents the known concentration of the target element, and a and b are regression coefficients.

3. The semiconductor process inspection method according to claim 2, characterized in that, The specific method for determining the actual concentration of the target element in the sample to be tested is as follows: plot a standard curve, obtain the coefficient of determination R2 through regression analysis. The R2 value ranges from 0 to 1. The closer the value is to 1, the higher the matching degree of the standard curve. Set a dopant identification threshold, compare the coefficient of determination with the dopant identification threshold, and make corresponding responses based on the comparison results.

4. The semiconductor process inspection method according to claim 3, characterized in that, When the coefficient of determination is greater than or equal to the dopant identification threshold, it indicates that the measurement target and the standard sample have a high degree of matching, and it is determined that the signal responses of the measurement target and the standard sample belong to the same type of dopant. When the coefficient of determination is less than the dopant identification threshold, it indicates that the measurement target and the standard sample have a low degree of matching, and it is determined that the signal responses of the measurement target and the standard sample do not belong to the same type of dopant.

5. A method for semiconductor process inspection according to claim 4, characterized in that, The specific method for plotting a standard curve is to use SIMS to measure the signal intensity of the target element and the internal standard element in each standard sample, calculate the sample signal ratio, use the known concentration as the abscissa and the calculated signal ratio as the ordinate, mark all points on the graph, draw a line so that these points fall on this line, and set this line as the standard curve.

6. A method for semiconductor process inspection according to claim 5, characterized in that, The specific method for obtaining the coefficient of determination R² is as follows: calculate the total fluctuation between all data points and their mean, and then compare the actual deviations between the data points and the predicted values ​​of the regression line. Specifically, first calculate the total sum of squared deviations of the data, which measures the overall dispersion between all observed values ​​and the sample mean; then calculate the sum of squared residuals, which represents the difference between each actual data point and the regression prediction value. The least squares method is used to minimize the sum of squared residuals. The relationship between the two is expressed mathematically as the ratio of the prediction error to the overall error, thus obtaining the coefficient of determination R².

7. A method for semiconductor process inspection according to claim 6, characterized in that, Set a near-identification threshold that is less than the dopant identification threshold. When the coefficient of determination is less than the dopant identification threshold, compare the coefficient of determination with the near-identification threshold. When the coefficient of determination is greater than the near-identification threshold, execute the ion beam parameter optimization strategy.

8. A method for semiconductor process inspection according to claim 7, characterized in that, The ion beam parameter optimization strategy includes adjusting the ion beam accelerating voltage, increasing or decreasing the initial parameter values ​​by 5% to 10%, performing group measurements, measuring standard samples for each group of adjusted data and calculating their respective standard curves R2, comparing the stability and linear range of the ratio of signal ions to internal standard ions at different energies, selecting voltage conditions that can significantly improve R2, making small-range adjustments to the incident angle and observing its impact on secondary ion generation efficiency, adjusting the beam spot size to reduce the difference in local effects on the sample surface, testing standard curves under different incident angles and beam spot conditions, and selecting parameter combinations that can make R2 further approach or even reach the dopant identification threshold.

9. A semiconductor process inspection system, characterized in that, The method for semiconductor process inspection according to any one of claims 1-8 comprises: The target processing module includes pre-processing the semiconductor wafer to remove residual contaminants from its surface, and employing chemical cleaning and drying processes to maintain the integrity of the crystal. The internal standard doping module includes selecting multiple isotopes with the same chemical properties as the target element but with significant differences in natural abundance as standard samples, and uniformly doping multiple standard samples with isotope solutions of known concentrations. The curve construction module includes using multiple standard samples to obtain the ratio of ion signals at different concentrations through SIMS measurement, and constructing the response curve of the target element. The corresponding curve can reflect the ratio of the internal standard isotope ion signals to the target isotope ion signals under the same matrix. The target determination module includes using the same SIMS measurement parameters as when constructing the response curve to perform multi-point scanning detection on the target to obtain ion signal data of the target isotope and the internal standard isotope. It also calculates the signal ratio of the internal standard isotope to the target isotope based on the collected ion signal data, and determines the actual concentration of the target element in the sample by the correspondence between the signal ratio and the standard curve.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to implement a semiconductor process inspection method according to any one of claims 1-8.

Images