Method, device and computer readable storage medium for measuring thickness of thin film

Abstract

The application relates to a film thickness measurement method and device and a computer readable storage medium. The method comprises the following steps: performing spectrum collection on multiple point positions of a to-be-measured film sample to obtain multiple to-be-evaluated spectra; determining a target average spectrum matched with the type of the to-be-measured film sample and / or the type of the to-be-evaluated spectrum from a preset configuration file; for each to-be-evaluated spectrum, calculating the mean square error of the to-be-evaluated spectrum and the determined target average spectrum to obtain a corresponding to-be-evaluated mean square error value; for each to-be-evaluated spectrum, when the to-be-evaluated mean square error value corresponding to the to-be-evaluated spectrum is greater than a preset mean square error threshold, the to-be-evaluated spectrum is prompted as an abnormal spectrum, and the measurement of the thickness at the corresponding point position is ended; for each to-be-evaluated spectrum, when the to-be-evaluated mean square error value corresponding to the to-be-evaluated spectrum is less than or equal to the preset mean square error threshold, spectrum regression fitting is performed on the to-be-evaluated spectrum to obtain the thickness value at the corresponding point position.

Description

[0001] This application claims priority to Chinese Patent Application No. 202611068184.4, filed on July 17, 2026, entitled "Method, Apparatus and Computer-Readable Storage Medium for Measuring Thin Film Thickness", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of optical nanoscale precision measurement, specifically to a method, apparatus, and computer-readable storage medium for measuring thin film thickness. Background Technology

[0003] In the thin film nanometry industry, thickness measurement of unknown novel materials accounts for an increasingly large proportion. Thin film nanoellipsometry based on optical scattering measurement is essentially a model-based measurement method. First, a film thickness measurement device is used to acquire the measurement spectrum. Then, a thin film morphology model is established. Finally, the morphology model is used to fit and match the measurement spectrum, thereby extracting the parameters to be determined from the model, including thickness. When measuring thin film samples using film thickness measurement devices, especially in industrial production, it is often necessary to measure the thickness at multiple points on the thin film sample, i.e., mapping measurement.

[0004] However, regardless of whether the sample is regular (e.g., circular, square) or irregular, if the spectrum measured at a certain point is abnormal during the mapping measurement process, using this abnormal spectrum to measure the thickness at that point will lead to abnormal film thickness measurement results and increased measurement time. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus and computer-readable storage medium for measuring thin film thickness, so as to solve the problems in the prior art that the measurement results of thin film thickness are abnormal and the measurement time is long because some spectral anomalies measured at some points are ignored when measuring thin film sample mapping.

[0006] This application provides a method for measuring film thickness, comprising: acquiring spectra at multiple points on a film sample to be tested to obtain multiple spectra to be evaluated, wherein each of the multiple spectra to be evaluated corresponds one-to-one with the multiple points on the film sample to be tested; determining a target average spectrum from a preset configuration file that matches the type of the film sample to be tested and / or the type of the spectrum to be evaluated, wherein the preset configuration file stores the correspondence between the type of a standard film sample and the target average spectrum, and / or the correspondence between the type of a standard spectrum and the target average spectrum; and for each acquired spectrum to be evaluated, [the method is described in the original text]. The mean square error is calculated between the target average spectrum and the target mean square error to obtain the corresponding mean square error value to be evaluated. For each spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is greater than a preset mean square error threshold, the spectrum to be evaluated is indicated as an abnormal spectrum, and the measurement of the thickness of the thin film sample at the point corresponding to the spectrum to be evaluated is terminated. For each spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0007] The method for measuring film thickness further includes, before determining the target average spectrum that matches the type of the film sample to be tested and / or the type of the spectrum to be evaluated from a preset configuration file, the steps of which are: acquiring spectra at multiple points on a standard film sample to obtain multiple standard spectra; averaging the multiple standard spectra to obtain an average spectrum; calculating the mean square error between the multiple standard spectra and the average spectrum to obtain multiple standard mean square error values; filtering out abnormal standard spectra from the multiple standard spectra based on the multiple standard mean square error values; averaging all remaining standard spectra except for all abnormal standard spectra to obtain a target average spectrum; associating the target average spectrum with the type of the standard film sample and / or the type of the standard spectrum, and saving it to a preset configuration file.

[0008] The step of selecting abnormal standard spectra from the multiple standard spectra based on the multiple standard mean square error values ​​includes: selecting the n largest standard mean square error values ​​from the multiple standard mean square error values, and calculating the repeatability index value of all remaining standard spectra in the multiple standard spectra except for the n standard spectra corresponding to the n mean square error values; when the calculated repeatability index value is greater than a preset repeatability index threshold, the value of n is increased by 1, and then the step of selecting the n largest mean square error values ​​from the multiple standard error values ​​and calculating the repeatability index value of all remaining standard spectra in the multiple standard spectra except for the standard spectra corresponding to the n mean square error values ​​is returned, until the calculated repeatability index value is less than or equal to the preset repeatability index threshold; when the calculated repeatability index value is less than or equal to the preset repeatability index threshold, it is determined that the n standard spectra corresponding to the current n mean square error values ​​in the multiple standard spectra are all abnormal standard spectra.

[0009] The preset configuration file stores the correspondence between the type of standard spectrum and the target average spectrum, as well as the correspondence between the type of standard thin film sample and the type of standard spectrum. From the preset configuration file, the target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated is determined, including: from the preset configuration file, determining the target average spectrum corresponding to the type of standard spectrum that matches the type of the standard thin film sample to be tested, among all the types of standard spectra that match the type of the spectrum to be evaluated.

[0010] The step of performing spectral regression fitting on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the corresponding point of the spectrum to be evaluated includes: performing regression fitting on the spectrum to be evaluated using a preset thin film model, optimizing the model parameters of the thin film model until the thin film model converges, and then extracting the thickness parameter from the model parameters of the converged thin film model as the thickness value of the thin film sample at the corresponding point of the spectrum to be evaluated. The preset thin film model is pre-constructed based on the thin film sample to be evaluated.

[0011] This application embodiment also provides a thin film thickness measurement device, which includes: a first spectral acquisition module, used to acquire spectra at multiple points of the thin film sample to be tested, obtaining multiple spectra to be evaluated, wherein the multiple spectra to be evaluated correspond one-to-one with the multiple points of the thin film sample to be tested; a determination module, used to determine a target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from a preset configuration file, wherein the preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum; and a first calculation module, used to calculate the value of each acquired spectrum to be evaluated. The mean square error (MSE) between the spectrum to be evaluated and the determined target average spectrum is calculated to obtain the corresponding MSE value. A first processing module is used to, for each acquired spectrum to be evaluated, indicate that the spectrum to be evaluated is an abnormal spectrum and terminate the measurement of the thickness of the thin film sample at the point corresponding to the spectrum to be evaluated when the MSE value corresponding to the spectrum to be evaluated is greater than a preset MSE threshold. A second processing module is used to, for each acquired spectrum to be evaluated, perform spectral regression fitting on the spectrum to be evaluated when the MSE value corresponding to the spectrum to be evaluated is less than or equal to the preset MSE threshold, to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0012] The film thickness measuring device further includes: a second spectral acquisition module for acquiring spectra at multiple points on a standard film sample to obtain multiple standard spectra; a second calculation module for averaging the multiple standard spectra to obtain an average spectrum; a third calculation module for calculating the mean square error between the multiple standard spectra and the average spectrum to obtain multiple standard mean square error values; a screening module for screening out abnormal standard spectra from the multiple standard spectra based on the multiple standard mean square error values; a fourth calculation module for averaging all remaining standard spectra except for all abnormal standard spectra to obtain a target average spectrum; and a storage module for associating the target average spectrum with the type of the standard film sample and / or the type of the standard spectrum, and saving it to a preset configuration file.

[0013] The screening module includes: a calculation unit, configured to select the largest n standard mean square error values ​​from the plurality of standard mean square error values, and calculate the repeatability index values ​​of all remaining standard spectra in the plurality of standard spectra except for the n standard spectra corresponding to the n mean square error values; a first processing unit, configured to increment the value of n by 1 when the calculated repeatability index value is greater than a preset repeatability index threshold, and then trigger the calculation unit to re-execute the operation of selecting the largest n standard mean square error values ​​from the plurality of standard mean square error values ​​and calculating the repeatability index values ​​of all remaining standard spectra in the plurality of standard spectra except for the n standard spectra corresponding to the n mean square error values, until the calculated repeatability index value is less than or equal to the preset repeatability index threshold; and a second processing unit, configured to determine that the n standard spectra corresponding to the current n mean square error values ​​in the plurality of standard spectra are all abnormal standard spectra when the calculated repeatability index value is less than or equal to the preset repeatability index threshold.

[0014] The preset configuration file stores the correspondence between the type of standard spectrum and the target average spectrum, as well as the correspondence between the type of standard thin film sample and the type of standard spectrum. The determination module is specifically used to: determine, from the preset configuration file, the target average spectrum corresponding to the type of standard spectrum that is consistent with the type of the standard thin film sample to be tested and is consistent with the type of the spectrum to be evaluated.

[0015] This application also provides a computer-readable storage medium storing a computer program adapted to be loaded by a processor to execute the film thickness measurement method described above.

[0016] The thin film thickness measurement method, apparatus, and computer-readable storage medium provided in this application acquire multiple spectra at multiple points on the thin film sample to be tested, obtaining multiple spectra to be evaluated. These multiple spectra correspond one-to-one with the multiple points on the thin film sample. Then, a target average spectrum matching the type of the thin film sample and / or the type of the spectrum to be evaluated is determined from a preset configuration file. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum. Next, for each acquired spectrum to be evaluated, the mean square error (MSE) is calculated between the spectrum to be evaluated and the determined target average spectrum to obtain the corresponding MSE value. Then, for each acquired spectrum to be evaluated, if the MSE value corresponding to the spectrum to be evaluated is greater than a preset MSE threshold, an indication is given that the spectrum to be evaluated is an abnormal spectrum, and the process for evaluating the thin film sample is terminated. The method involves measuring the thickness at the points corresponding to the spectrum to be evaluated. For each acquired spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is less than or equal to a preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated. This provides a method for screening abnormal points in advance during nanofilm mapping measurements. This method matches the target average spectrum from a preset configuration file and calculates the mean square error value between the spectrum measured at each point and the target average spectrum, thereby determining whether the spectrum measured at each point is abnormal. This enables rapid identification of abnormal spectra measured at abnormal points before spectral regression fitting, and avoids subsequent spectral regression fitting calculations for abnormal spectra. This saves the time of the entire mapping measurement and accurately predicts abnormal points in advance, significantly improving the efficiency and reliability of thin film thickness measurement. Attached Figure Description

[0017] The technical solution and other beneficial effects of this application will become apparent from the following detailed description of specific embodiments in conjunction with the accompanying drawings.

[0018] Figure 1 This is a schematic flowchart of the thin film thickness measurement method provided in the embodiments of this application; Figure 2 This is a flowchart illustrating the method for generating a target average spectrum provided in an embodiment of this application; Figure 3 This is another schematic flowchart of the method for generating a target average spectrum provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the thin film thickness measuring device provided in the embodiments of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0020] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication between two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0022] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature being directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0023] The following disclosure provides many different embodiments or examples for implementing different structures of this application. To simplify the disclosure, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the scope of this application. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, various specific examples of processes and materials are provided in this application, but those skilled in the art will recognize the application of other processes and / or the use of other materials.

[0024] In the prior art, when using film thickness measurement equipment to measure film samples, especially in industrial production, it is often necessary to measure the thickness at multiple points on the film sample, i.e., mapping measurement.

[0025] However, regardless of whether the sample is regular (e.g., circular, square) or irregular, during the mapping measurement process, material abnormalities or morphological anomalies at certain points can lead to abnormal spectra, resulting in inconsistent final regression algorithm results. In actual industrial production lines, controlling measurement time is crucial. If an abnormal spectrum is measured at a particular point, the subsequent regression algorithm calculation will take longer and may produce incorrect solutions. If the abnormal spectrum can be identified before the regression algorithm calculation, the subsequent calculation step can be eliminated, and the point can be pre-marked as an anomaly. This saves overall mapping measurement time and accurately predicts the location of the anomaly.

[0026] Therefore, it is necessary to provide a method for pre-screening abnormal sites in nanofilm mapping measurements.

[0027] To address the above problems, this application provides a method, apparatus, and computer storage medium for measuring thin film thickness. The method involves acquiring spectra at multiple points on a thin film sample to be tested, obtaining multiple spectra to be evaluated. These multiple spectra correspond one-to-one with the multiple points on the thin film sample. Then, a target average spectrum matching the type of the thin film sample and / or the type of the spectrum to be evaluated is determined from a preset configuration file. The preset configuration file stores the correspondence between the type of a standard thin film sample and the target average spectrum, and / or the correspondence between the type of a standard spectrum and the target average spectrum. Next, for each acquired spectrum to be evaluated, the mean square error (MSE) is calculated between the spectrum to be evaluated and the determined target average spectrum to obtain the corresponding MSE value. Finally, for each acquired spectrum to be evaluated, if the MSE value corresponding to the spectrum to be evaluated exceeds a preset MSE threshold, the spectrum to be evaluated is flagged as an abnormal spectrum, and the process for measuring the thickness of the thin film sample is terminated. The thickness of the film sample at the point corresponding to the spectrum to be evaluated is measured. For each acquired spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is less than or equal to a preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the film sample at the point corresponding to the spectrum to be evaluated. This provides a method for screening abnormal points in advance in nanofilm mapping measurement. This method matches the target average spectrum from a preset configuration file and calculates the mean square error value between the spectrum measured at each point and the target average spectrum, thereby determining whether the spectrum measured at each point is abnormal. This achieves rapid identification of abnormal spectra measured at abnormal points before spectral regression fitting, and avoids subsequent spectral regression fitting calculations for abnormal spectra. This saves the time of the entire mapping measurement and accurately predicts abnormal points in advance, significantly improving the efficiency and reliability of film thickness measurement.

[0028] The following detailed description is based on specific embodiments. It should be noted that the embodiments of this application can be presented in various forms, and some examples will be described below.

[0029] Please see Figure 1 , Figure 1 This is a schematic flowchart of a method for measuring film thickness provided in an embodiment of this application. The specific process of the method for measuring film thickness is as follows: Step S11. By acquiring spectra at multiple points on the thin film sample to be tested, multiple spectra to be evaluated are obtained, and the multiple spectra to be evaluated correspond one-to-one with the multiple points on the thin film sample to be tested.

[0030] Specifically, a film thickness measurement device can be used to perform mapping multi-point spectral acquisition on the thin film sample to be tested (i.e., to acquire spectra at multiple different locations on the thin film sample to be tested), and the resulting multiple spectra are the aforementioned multiple spectra to be evaluated.

[0031] Specifically, among the multiple spectra to be evaluated collected above, each spectrum to be evaluated corresponds to the same band and the same spectral type. That is, each spectrum to be evaluated has the same band and the same type.

[0032] In some examples, among the multiple spectra to be evaluated collected above, the wavelength range of each spectrum to be evaluated can be 300 nm to 1000 nm, and the type of each spectrum to be evaluated can be PsiDelta spectrum, NCS spectrum, Mueller spectrum, or R reflectance spectrum, etc. Among them, PsiDelta spectrum is the combined spectrum of phase difference and amplitude ratio obtained by ellipsometer measurement, NCS spectrum is the normalized dispersion curve spectrum, Mueller spectrum is the full polarization information spectrum obtained based on the Mueller matrix, and R reflectance spectrum is the spectrum of reflectance as a function of wavelength measured by spectrophotometer.

[0033] Step S12. Determine the target average spectrum from the preset configuration file that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum.

[0034] In some embodiments, the preset configuration file may store the correspondence between the types of standard thin film samples and the target average spectrum. Accordingly, determining the target average spectrum matching the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from the preset configuration file may specifically include: determining the target average spectrum matching the type of the thin film sample to be tested from the preset configuration file. Specifically, the target average spectrum matching the type of the thin film sample to be tested in the preset configuration file may refer to the target average spectrum corresponding to the type of a standard thin film sample that is consistent with the type of the thin film sample to be tested in the preset configuration file.

[0035] In other embodiments, the preset configuration file may store the correspondence between the types of standard spectra and target average spectra. Accordingly, determining the target average spectrum matching the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from the preset configuration file may specifically include: determining the target average spectrum matching the type of the thin film sample to be tested from the preset configuration file. Specifically, the target average spectrum matching the type of the spectrum to be evaluated in the preset configuration file may refer to the target average spectrum corresponding to the type of standard spectrum in the preset configuration file that is consistent with the type of the spectrum to be evaluated.

[0036] In other embodiments, the aforementioned preset configuration file may store the correspondence between the type of standard spectrum and the target average spectrum, as well as the correspondence between the type of standard thin film sample and the type of standard spectrum. Accordingly, determining the target average spectrum matching the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from the preset configuration file may specifically include: determining the target average spectrum matching the type of the thin film sample to be tested and the type of the spectrum to be evaluated from the preset configuration file. Specifically, the target average spectrum matching the type of the thin film sample to be tested and the type of the spectrum to be evaluated in the aforementioned preset configuration file may refer to the target average spectrum corresponding to the type of standard spectrum among all standard spectrum types corresponding to the type of standard thin film sample consistent with the type of the thin film sample to be tested, and the type of standard spectrum consistent with the type of the spectrum to be evaluated.

[0037] In some specific embodiments, such as Figure 2 As shown, in order to obtain the aforementioned preset configuration file, before step S11 or step S12, the method for measuring the film thickness may further include: Step S21. Collect spectra at multiple points on the standard thin film sample to obtain multiple standard spectra.

[0038] Specifically, a film thickness measurement device can be used to perform mapping multi-point spectral acquisition on a standard thin film sample (i.e., to acquire spectra at multiple different locations on the standard thin film sample), and the resulting multiple spectra are the aforementioned multiple standard spectra.

[0039] The standard thin film sample can be circular or square, with a smooth surface, no significant material defects or surface damage, good thickness uniformity, and all performance parameters conforming to the preset standard range. It can be used as a reference standard to generate the target average spectrum.

[0040] In this embodiment, among the multiple standard spectra collected, each standard spectrum corresponds to the same band and the same spectral type. That is, each standard spectrum has the same band and the same type.

[0041] In some examples, the wavelength range of each of the above-mentioned standard spectra can be from 300 nm to 1000 nm, and the type of each standard spectrum can be PsiDelta spectrum, NCS spectrum, Mueller spectrum, or R reflectance spectrum, etc. Among them, PsiDelta spectrum is the combined spectrum of phase difference and amplitude ratio obtained by ellipsometer measurement, NCS spectrum is the normalized dispersion curve spectrum, Mueller spectrum is the full polarization information spectrum obtained based on the Mueller matrix, and R reflectance spectrum is the spectrum of reflectance as a function of wavelength measured by spectrophotometer.

[0042] Step S22. Calculate the average value of multiple standard spectra to obtain an average spectrum.

[0043] Specifically, after obtaining the aforementioned multiple standard spectra, the average spectrum can be generated by calculating the numerical mean of these standard spectra at each wavelength point. This average spectrum can reflect the overall trend of the spectral characteristics of the standard thin film sample at different locations, effectively eliminating the influence of local noise and random errors, thus serving as a benchmark for subsequent spectral comparison and analysis.

[0044] Step S23. Calculate the mean square error of the multiple standard spectra with the average spectrum to obtain multiple standard mean square error values.

[0045] Specifically, after obtaining the average spectrum, for each of the multiple standard spectra, the standard mean square error value corresponding to that standard spectrum can be calculated using the first calculation formula, wherein the first calculation formula is as follows: , In the formula, MSE represents the standard mean square error value corresponding to the standard spectrum, and m is the number of wavelength points in the standard spectrum. Let be the spectral intensity value of the standard spectrum at the i-th wavelength point. Let be the spectral intensity value of the average spectrum at the i-th wavelength point, and let the standard spectrum, the 1st wavelength point, and the m-th wavelength point be the starting and ending wavelength points of the band of the standard spectrum, respectively.

[0046] After calculating the standard mean square error (SSE) values ​​for all standard spectra mentioned above, the aforementioned SSE values ​​can be obtained. Furthermore, for each standard spectrum, the SSE value reflects the degree of deviation between the standard spectrum and the average spectrum. The smaller the SSE value, the higher the similarity between the standard spectrum and the average spectrum, and the more stable the spectral characteristics of the standard spectrum; conversely, a larger SSE value indicates a greater difference between the standard spectrum and the average spectrum, suggesting potential anomalies in the standard spectrum.

[0047] Step S24. Based on multiple standard mean square error values, screen out abnormal standard spectra from multiple standard spectra.

[0048] Among these, abnormal standard spectra refer to those standard spectra that deviate relatively significantly from the average spectrum, i.e., those standard spectra with relatively large standard mean square errors. Such spectra are considered invalid data due to interference during acquisition or local defects in the sample and should be discarded to ensure the representativeness and reliability of the standard spectra used in subsequent analyses.

[0049] Specifically, the screening process for the aforementioned abnormal standard spectra may include: sorting the obtained standard mean square error values ​​in descending order, and selecting a predetermined number of standard spectra corresponding to the top-ranked predetermined number of standard mean square error values ​​as abnormal standard spectra; or, comparing each of the obtained standard mean square error values ​​with a predetermined threshold, and determining that the corresponding standard spectrum is an abnormal standard spectrum when the standard mean square error value is greater than the predetermined threshold. The predetermined number and predetermined threshold can be set based on empirical values ​​or preliminary understanding of the standard thin film samples, with the aim of initially eliminating spectral data that significantly deviate from the overall trend.

[0050] In some specific embodiments, such as Figure 3 As shown, step S24 above may include: Step S241. Select the largest n standard mean square error values ​​from multiple standard mean square error values, and calculate the repeatability index value of all remaining standard spectra in multiple standard spectra except for the n standard spectra corresponding to the n mean square error values ​​respectively.

[0051] Specifically, after obtaining the aforementioned standard mean squared error (MSE) values, these values ​​can be sorted in descending order. The top n largest values ​​are selected as the n standard MSE values. Each of these n MSE values ​​corresponds to an abnormal standard spectrum, which should be discarded. Subsequently, the repeatability index value is calculated for all remaining standard spectra excluding the currently identified abnormal standard spectra. The repeatability index value measures the degree of consistency among all remaining standard spectra. A smaller repeatability index value indicates less fluctuation among the remaining standard spectra and better data repeatability. Calculating the repeatability index value helps to further identify potential abnormal standard spectra among all remaining standard spectra, thereby optimizing the screening results.

[0052] Furthermore, in specific implementation, the repeatability index value of all remaining standard spectra excluding all currently determined abnormal standard spectra (i.e., the n standard spectra corresponding to the aforementioned n standard mean square error values) can be calculated using the second calculation formula. The second calculation formula is as follows: , In the formula, Repeatability represents the repeatability index value. Where m is the number of remaining standard spectra, and yi represents the i-th remaining standard spectrum. This represents the average spectrum of all remaining standard spectra, with 5.15 being the coverage factor at a 99% confidence level.

[0053] Thus, by calculating the repeatability index value, the consistency level of the remaining spectral data can be quantitatively assessed, providing a basis for further identification and removal of abnormal standard spectra.

[0054] Step S242. When the calculated repeatability index value is greater than the preset repeatability index threshold, the value of n is increased by 1, and then the above step S241 is returned to be executed until the calculated repeatability index value is less than or equal to the preset repeatability index threshold.

[0055] Where n is a positive integer, and its initial value can be determined based on 10% to 15% of the number of standard spectra collected, to ensure that the exclusion range is reasonable and that effective data is not over-filtered. In some examples, the number of standard spectra collected can be greater than or equal to 20, and the initial value of n can be 2.

[0056] It is understandable that, since the spectral data of different types of thin film samples vary significantly, and the spectral data of different types also vary considerably, the value of the aforementioned preset repeatability threshold can be determined according to the actual situation. In some examples, the aforementioned preset repeatability threshold can be 0.01.

[0057] Step S243. When the calculated repeatability index value is less than or equal to the preset repeatability index threshold, determine the n standard spectra that correspond to the current n mean square error values ​​from the multiple standard spectra as abnormal standard spectra.

[0058] Specifically, after calculating the repeatability index value, it can be determined whether the calculated repeatability index value is greater than the preset repeatability index threshold. If the calculated repeatability index value is less than or equal to the preset repeatability index threshold, it indicates that the consistency among all the remaining standard spectra currently determined is good, the data fluctuation is within an acceptable range, and no further elimination is needed; otherwise, it indicates that there are still abnormal standard spectra among all the remaining standard spectra currently determined, and further screening is required.

[0059] Furthermore, in specific implementation, if the calculated repeatability index value is greater than the preset repeatability index threshold, the value of n is increased by 1, and the above step S241 is returned to be executed until the calculated repeatability index value is less than or equal to the preset repeatability index threshold, or the value of n reaches the preset maximum number of eliminations.

[0060] Specifically, if the currently calculated repeatability index value does not meet the requirements, i.e., it is greater than the aforementioned preset repeatability index threshold, it means that there are still abnormal standard spectra among all the currently determined remaining standard spectra that cause a decrease in overall consistency. At this time, the value of n is increased by 1, that is, one more remaining standard spectrum with the largest standard mean square error value is removed from all the currently determined remaining standard spectra. Then, the repeatability index value of all remaining standard spectra after removing this remaining standard spectrum is recalculated, and the judgment is made again. This process is repeated until the calculated repeatability index value is less than or equal to the aforementioned preset repeatability index threshold. At this time, it is considered that all the currently determined remaining standard spectra have good consistency. Alternatively, when the value of n increases to the preset maximum number of removals (for example, set to 10% of the total number of standard spectra or a certain fixed value), even if the repeatability index value still does not meet the standard, the screening process is stopped to avoid infinite loops or excessive removal leading to insufficient samples.

[0061] Step S25. Calculate the average value of all remaining standard spectra except for all anomalous standard spectra among the multiple standard spectra to obtain the target average spectrum.

[0062] Specifically, after removing outlier standard spectra, the average spectrum of the remaining standard spectra with high consistency is recalculated to obtain the target average spectrum. Compared with the unscreened average spectrum, this target average spectrum is more representative of the true spectral characteristics of the standard thin film sample, reducing the interference of outliers on the reference spectrum, thus providing a more reliable and stable reference standard for the subsequent anomaly judgment of the spectrum to be evaluated.

[0063] Step S26. Associate the target average spectrum with the type of standard thin film sample and / or the type of standard spectrum, and save it to a preset configuration file.

[0064] Specifically, in the above embodiment where the correspondence between the type of standard thin film sample and the target average spectrum is stored in the preset configuration file, step S26 may specifically include: associating the target average spectrum with the type of standard thin film sample and saving it to the preset configuration file, so as to facilitate the rapid retrieval of the corresponding target average spectrum according to the type of standard thin film sample in the future.

[0065] In the above embodiment where the correspondence between the type of standard spectrum and the target average spectrum is stored in the preset configuration file, step S26 may specifically include: associating the target average spectrum with the type of standard spectrum and saving it to the preset configuration file, so as to facilitate the quick retrieval of the corresponding target average spectrum according to the type of standard spectrum in the future.

[0066] In the above-mentioned embodiment, the correspondence between the type of standard spectrum and the target average spectrum, and the correspondence between the type of standard thin film sample and the type of standard spectrum can be stored in the preset configuration file. In this embodiment, step S26 can specifically include: associating the type of standard thin film sample and the type of standard spectrum with the target average spectrum respectively, and saving them to the preset configuration file, so as to facilitate the rapid retrieval of the corresponding target average spectrum according to the type of standard thin film sample and the type of standard spectrum.

[0067] The correspondence between the types of standard thin film samples and the target average spectrum can include: multiple types of standard thin film samples and multiple target average spectra that correspond one-to-one with the types of standard thin film samples.

[0068] The correspondence between the types of standard spectra and the target average spectra can include: multiple types of standard spectra, and multiple target average spectra that correspond one-to-one with the types of standard spectra.

[0069] The correspondence between the types of standard thin film samples and the types of standard spectra can include: multiple types of standard thin film samples, and at least one type of standard spectrum corresponding to each type of standard thin film sample among the multiple types of standard thin film samples.

[0070] Furthermore, in specific implementation, for each type of standard thin film sample, steps S21 to S26 can be executed sequentially to obtain the target average spectrum corresponding to a type of standard spectrum (i.e., the type of standard spectrum acquired in step S21). Thus, for each type of standard thin film sample, by executing steps S21 to S26 multiple times, multiple target average spectra corresponding to various standard spectral types for that type of standard thin film sample can be obtained.

[0071] Step S13. For each spectrum to be evaluated that has been acquired, calculate the mean square error between the spectrum to be evaluated and the determined target average spectrum to obtain the corresponding mean square error value.

[0072] Specifically, for each spectrum to be evaluated that is acquired, the mean square error between the spectrum to be evaluated and the determined target average spectrum is calculated to obtain the mean square error value corresponding to the spectrum to be evaluated.

[0073] Furthermore, in specific implementation, for each spectrum to be evaluated acquired, the mean square error between the spectrum to be evaluated and the determined target average spectrum is calculated to obtain the mean square error value corresponding to the spectrum to be evaluated. The specific implementation method for calculating the standard mean square error corresponding to each standard spectrum among the above multiple standard spectra using the first calculation formula for each acquired spectrum to be evaluated is not repeated here.

[0074] Step S14. For each spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is greater than the preset mean square error threshold, the spectrum to be evaluated is indicated as an abnormal spectrum, and the measurement of the thickness of the thin film sample to be tested at the point corresponding to the spectrum to be evaluated is terminated.

[0075] Step S15. For each spectrum to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0076] Specifically, for each calculated mean square error value corresponding to a spectrum to be evaluated, it can be determined whether the mean square error value of the spectrum to be evaluated is greater than the preset mean square error threshold. If it is greater, the spectrum to be evaluated is determined to be an abnormal spectrum, and a direct prompt is given, and no further regression algorithm calculation is performed on the spectrum to be evaluated, thereby avoiding wasting computational resources and time and ensuring measurement efficiency and accuracy. If it is not greater, the spectrum to be evaluated is determined to be a normal spectrum, and spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0077] Understandably, the aforementioned preset mean square error threshold can be determined based on the specific thin film sample. In industrial production, this value varies depending on different precision requirements. Furthermore, this value also differs for thin film samples with different morphologies.

[0078] In some embodiments, the aforementioned notification that the spectrum to be evaluated is an abnormal spectrum may specifically include: notifying the upper-level system that the spectrum to be evaluated is abnormal, thereby triggering the upper-level system to execute the corresponding abnormality handling process, such as marking the point as an invalid measurement point, adding a new measurement point for retesting, or prompting the operator to perform a re-inspection, to ensure the integrity and reliability of the overall measurement data. By promptly notifying the abnormal spectrum and terminating the thickness measurement at the corresponding point, it is possible to effectively avoid obtaining incorrect thickness results based on invalid data, while also saving computing resources and measurement time, and improving the overall efficiency of the measurement process. In addition, the location information of the point corresponding to the abnormal spectrum can be recorded in the log information, which facilitates subsequent traceability and analysis, and provides data support for process optimization.

[0079] In some embodiments, step S15 may specifically include: for each acquired spectrum to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, a preset thin film model is used to perform regression fitting on the spectrum to be evaluated, the model parameters of the thin film model are optimized until the thin film model converges, and then the thickness parameter in the model parameters of the converged thin film model is extracted as the thickness value of the thin film sample to be tested at the point corresponding to the spectrum to be evaluated. The preset thin film model is pre-established based on the thin film sample to be tested. For example, it can be pre-constructed based on the physical properties of the thin film sample to be tested (such as refractive index, extinction coefficient, etc.) and the optical principles of spectral acquisition (such as reflection spectrum, transmission spectrum, etc.). The model parameters include thin film thickness, substrate refractive index, and dispersion parameters of the thin film material. During the regression fitting process, the model parameters are adjusted to minimize the deviation (e.g., mean square error) between the theoretical spectrum simulated by the thin film model and the spectrum to be evaluated. When the deviation reaches a preset convergence condition (e.g., the number of iterations reaches the upper limit or the deviation is less than the convergence threshold), optimization stops, and the thin film thickness value in the current model parameters is output as the desired measurement value. In practical applications, to improve fitting efficiency and accuracy, initial parameter estimation can be introduced. That is, initial values ​​are set for the model parameters based on experience or approximate information from the sample to shorten the number of iterations required for model convergence. Furthermore, for different types of thin films (e.g., single-layer films, multilayer films, anisotropic films, etc.), appropriate thin film models should be selected for fitting to ensure the accuracy of the thickness calculation.

[0080] As can be seen from the above, the thin film thickness measurement method provided in this embodiment involves acquiring multiple spectra at multiple points on the thin film sample to be tested, obtaining multiple spectra to be evaluated. These multiple spectra to be evaluated correspond one-to-one with the multiple points on the thin film sample to be tested. Then, a target average spectrum matching the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated is determined from a preset configuration file. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum. Next, for each acquired spectrum to be evaluated, the mean square error is calculated between the spectrum to be evaluated and the determined target average spectrum to obtain the corresponding mean square error value. Then, for each acquired spectrum to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is greater than a preset mean square error threshold, the spectrum to be evaluated is indicated as an abnormal spectrum, and the evaluation of the thin film sample under the target light is terminated. This method measures the thickness at the corresponding points in the spectrum. For each acquired spectrum to be evaluated, if the mean square error value of the spectrum to be evaluated is less than or equal to a preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the corresponding point in the spectrum to be evaluated. This provides a method for screening abnormal points in advance in nanofilm mapping measurements. This method matches the target average spectrum from a preset configuration file and calculates the mean square error value between the spectrum measured at each point and the target average spectrum, thereby determining whether the spectrum measured at each point is abnormal. This achieves rapid identification of abnormal spectra measured at abnormal points before spectral regression fitting, and avoids subsequent spectral regression fitting calculations for abnormal spectra. This saves the time of the entire mapping measurement and accurately predicts abnormal points in advance, significantly improving the efficiency and reliability of thin film thickness measurement.

[0081] Based on the methods described in the above embodiments, this embodiment will further describe the method from the perspective of a film thickness measuring device. Please refer to [link / reference needed]. Figure 4 , Figure 4 The present application provides a specific description of a thin film thickness measuring device, which may include: a first spectral acquisition module 301, a determination module 302, a first calculation module 303, a first processing module 304, and a second processing module 305, wherein: (1) First spectral acquisition module 301 The first spectral acquisition module 301 is used to acquire spectra at multiple points on the thin film sample to be tested, and obtain multiple spectra to be evaluated. These multiple spectra to be evaluated correspond one-to-one with the multiple points on the thin film sample to be tested.

[0082] (2) Determine module 302 The determination module 302 is used to determine, from a preset configuration file, a target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum.

[0083] (3) First calculation module 303 The first calculation module 303 is used to calculate the mean square error between the acquired spectrum to be evaluated and the determined target average spectrum for each acquired spectrum to be evaluated, and obtain the corresponding mean square error value to be evaluated.

[0084] (4) First processing module 304 The first processing module 304 is used to, for each acquired spectrum to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is greater than the preset mean square error threshold, indicate that the spectrum to be evaluated is an abnormal spectrum and end the measurement of the thickness of the thin film sample to be tested at the point corresponding to the spectrum to be evaluated.

[0085] (5) Second processing module 305 The second processing module 305 is used to perform spectral regression fitting on each acquired spectrum to be evaluated when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, so as to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0086] Specifically, the second processing module 305 is used for: A preset thin film model is used to perform regression fitting on the spectrum to be evaluated. The model parameters of the thin film model are optimized until the thin film model converges. Then, the thickness parameter in the model parameters of the converged thin film model is extracted as the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0087] The preset thin film model is constructed in advance based on the thin film sample to be tested.

[0088] In some embodiments, the above-mentioned film thickness measuring device may further include: (6) Second Spectrum Acquisition Module The second spectral acquisition module is used to acquire spectra at multiple points on the standard thin film sample to obtain multiple standard spectra.

[0089] (7) Second Calculation Module The second calculation module is used to calculate the average value of multiple standard spectra to obtain an average spectrum.

[0090] (8) Third Calculation Module The third calculation module is used to calculate the mean square error of multiple standard spectra with the average spectrum, and obtain multiple standard mean square error values.

[0091] (9) Filtering module The filtering module is used to filter out abnormal standard spectra from multiple standard spectra based on multiple standard mean square error values.

[0092] (10) Fourth Calculation Module The fourth calculation module is used to calculate the average of all remaining standard spectra in multiple standard spectra except for all anomalous standard spectra, to obtain the target average spectrum.

[0093] (11) Save module The save module is used to associate the target average spectrum with the type of standard thin film sample and / or the type of standard spectrum, and save it to a preset configuration file.

[0094] In some specific embodiments, the above-mentioned filtering module may include: The calculation unit is used to select the largest n standard mean square error values ​​from multiple standard mean square error values, and to calculate the repeatability index value of all remaining standard spectra in multiple standard spectra except for the n standard spectra corresponding to the n mean square error values ​​respectively; The first processing unit is used to increment the value of n by 1 when the calculated repeatability index value is greater than the preset repeatability index threshold. Then, it triggers the above calculation unit to re-execute the above operation of selecting the largest n standard mean square error values ​​from multiple standard mean square error values ​​and calculating the repeatability index values ​​of all remaining standard spectra in multiple standard spectra except for the n standard spectra corresponding to the n mean square error values, until the calculated repeatability index value is less than or equal to the preset repeatability index threshold. The second processing unit is used to determine that when the calculated repeatability index value is less than or equal to the preset repeatability index threshold, the n standard spectra that correspond to the current n mean square error values ​​are all abnormal standard spectra.

[0095] It should be noted that, in specific implementation, the above units and modules can be implemented as independent entities, or they can be arbitrarily combined and implemented as the same or several entities. For the specific implementation of the above units and modules, please refer to the previous method embodiments, which will not be repeated here.

[0096] As can be seen from the above, the semiconductor device measurement device provided in this embodiment includes a first spectral acquisition module, used to acquire spectra at multiple points of the thin film sample to be tested, obtaining multiple spectra to be evaluated, each of which corresponds one-to-one with the multiple points of the thin film sample; a determination module, used to determine a target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from a preset configuration file, wherein the preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum; a first calculation module, used to calculate the mean square error between the acquired spectrum and the determined target average spectrum for each acquired spectrum to be evaluated, obtaining the corresponding mean square error value; and a first processing module, used to indicate that the spectrum to be evaluated is an abnormal spectrum and terminate the process when the mean square error value corresponding to the acquired spectrum is greater than a preset mean square error threshold. The first module measures the thickness of the thin film sample at the point corresponding to the spectrum to be evaluated. The second module is used to perform spectral regression fitting on each acquired spectrum to be evaluated when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to a preset mean square error threshold. This yields the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated, thus providing a method for screening abnormal points in advance during nanofilm mapping measurements. This method matches the target average spectrum from a preset configuration file and calculates the mean square error value between the spectrum measured at each point and the target average spectrum. This allows for the determination of whether the spectrum measured at each point is abnormal. This enables the rapid identification of abnormal spectra measured at abnormal points before spectral regression fitting, and avoids subsequent spectral regression fitting calculations for abnormal spectra. This saves the time of the entire mapping measurement and accurately predicts abnormal points in advance, significantly improving the efficiency and reliability of thin film thickness measurement.

[0097] Accordingly, embodiments of this application also provide a computer-readable storage medium storing a computer program that can be loaded by a processor to execute the steps in any of the thin film thickness measurement methods provided in embodiments of this application. For example, the computer program can execute the following steps: Spectra are collected at multiple points on the thin film sample to be tested to obtain multiple spectra to be evaluated. These multiple spectra correspond one-to-one with the multiple points on the thin film sample to be tested. From the preset configuration file, determine the target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum. For each spectrum to be evaluated that is collected, the mean square error between the spectrum to be evaluated and the determined target average spectrum is calculated to obtain the corresponding mean square error value of the spectrum to be evaluated. For each spectrum to be evaluated, if the mean square error value corresponding to the spectrum to be evaluated is greater than the preset mean square error threshold, the spectrum to be evaluated is indicated as an abnormal spectrum, and the measurement of the thickness of the thin film sample to be tested at the point corresponding to the spectrum to be evaluated is terminated. For each spectrum to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

[0098] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0099] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0100] Since the computer program stored in the computer-readable storage medium can execute the steps in any of the thin film thickness measurement methods provided in the embodiments of this application, the beneficial effects that any of the thin film thickness measurement methods provided in the embodiments of this application can achieve can be realized, as detailed in the preceding embodiments, and will not be repeated here.

[0101] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0102] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for measuring thin film thickness, characterized in that, include: Spectra are collected at multiple points on the thin film sample to be tested to obtain multiple spectra to be evaluated. Each of the multiple spectra to be evaluated corresponds one-to-one with the multiple points on the thin film sample to be tested. From a preset configuration file, a target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated is determined. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum. For each spectrum to be evaluated that has been collected, the mean square error between the spectrum to be evaluated and the determined target average spectrum is calculated to obtain the corresponding mean square error value of the spectrum to be evaluated. For each spectrum to be evaluated that is acquired, if the mean square error value corresponding to the spectrum to be evaluated is greater than the preset mean square error threshold, the spectrum to be evaluated is indicated as an abnormal spectrum, and the measurement of the thickness of the thin film sample at the point corresponding to the spectrum to be evaluated is terminated. For each spectrum to be evaluated that is acquired, when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, spectral regression fitting is performed on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

2. The method for measuring film thickness according to claim 1, characterized in that, Before determining, from a preset configuration file, a target average spectrum matching the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated, is determined, the method further includes: Spectroscopic data were collected at multiple locations on the standard thin film sample to obtain multiple standard spectra; The average of the multiple standard spectra is calculated to obtain an average spectrum. The mean square error of the multiple standard spectra is calculated by comparing each of the multiple standard spectra with the average spectrum, resulting in multiple standard mean square error values. Based on the multiple standard mean square error values, abnormal standard spectra are screened from the multiple standard spectra; The target average spectrum is obtained by averaging all remaining standard spectra except for all anomalous standard spectra among the multiple standard spectra. The target average spectrum is associated with the type of the standard thin film sample and / or the type of the standard spectrum, and saved to a preset configuration file.

3. The method for measuring film thickness according to claim 2, characterized in that, The step of filtering out abnormal standard spectra from the multiple standard spectra based on the multiple standard mean square error values ​​includes: Select the n largest standard mean square error values ​​from the plurality of standard mean square error values, and calculate the repeatability index value of all remaining standard spectra in the plurality of standard spectra except for the n standard spectra that correspond to the n mean square error values ​​respectively; When the calculated repeatability index value is greater than the preset repeatability index threshold, the value of n is increased by 1, and then the process returns to the step of selecting the largest n mean square error values ​​from the plurality of mean square error values ​​and calculating the repeatability index values ​​of all remaining standard spectra in the plurality of standard spectra except for the standard spectra corresponding to the n mean square error values, until the calculated repeatability index value is less than or equal to the preset repeatability index threshold. When the calculated repeatability index value is less than or equal to the preset repeatability index threshold, it is determined that the n standard spectra among the multiple standard spectra corresponding to the current n mean square error values ​​are all abnormal standard spectra.

4. The method for measuring film thickness according to claim 1, characterized in that, The preset configuration file stores the correspondence between the type of standard spectrum and the target average spectrum, as well as the correspondence between the type of standard thin film sample and the type of standard spectrum; The step of determining the target average spectrum from a preset configuration file that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated includes: From the preset configuration file, determine the target average spectrum corresponding to the type of standard spectrum that matches the type of standard thin film sample to be tested, and the type of standard spectrum that matches the type of spectrum to be evaluated.

5. The method for measuring film thickness according to claim 1, characterized in that, The step of performing spectral regression fitting on the spectrum to be evaluated to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated includes: A pre-defined thin film model is used to perform regression fitting on the spectrum to be evaluated. The model parameters of the thin film model are optimized until the model converges. Then, the thickness parameter in the converged thin film model is extracted as the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated. The pre-defined thin film model is constructed in advance based on the thin film sample to be evaluated.

6. A device for measuring thin film thickness, characterized in that, include: The first spectral acquisition module is used to acquire spectra at multiple points on the thin film sample to be tested, thereby obtaining multiple spectra to be evaluated. The multiple spectra to be evaluated correspond one-to-one with the multiple points on the thin film sample to be tested. The determination module is used to determine a target average spectrum that matches the type of the thin film sample to be tested and / or the type of the spectrum to be evaluated from a preset configuration file. The preset configuration file stores the correspondence between the type of standard thin film sample and the target average spectrum, and / or the correspondence between the type of standard spectrum and the target average spectrum. The first calculation module is used to calculate the mean square error between the acquired spectrum to be evaluated and the determined target average spectrum for each acquired spectrum to be evaluated, and obtain the corresponding mean square error value to be evaluated. The first processing module is used to, for each of the acquired spectra to be evaluated, when the mean square error value corresponding to the spectrum to be evaluated is greater than a preset mean square error threshold, indicate that the spectrum to be evaluated is an abnormal spectrum and end the measurement of the thickness of the thin film sample at the point corresponding to the spectrum to be evaluated. The second processing module is used to perform spectral regression fitting on each of the acquired spectra to be evaluated when the mean square error value corresponding to the spectrum to be evaluated is less than or equal to the preset mean square error threshold, so as to obtain the thickness value of the thin film sample at the point corresponding to the spectrum to be evaluated.

7. The thin film thickness measuring device according to claim 6, characterized in that, The device further includes: The second spectral acquisition module is used to acquire spectra at multiple points on the standard thin film sample to obtain multiple standard spectra. The second calculation module is used to calculate the average value of the multiple standard spectra to obtain an average spectrum; The third calculation module is used to calculate the mean square error between the multiple standard spectra and the average spectrum to obtain multiple standard mean square error values. The filtering module is used to filter out abnormal standard spectra from the multiple standard spectra based on the multiple standard mean square error values; The fourth calculation module is used to calculate the average value of all remaining standard spectra in the multiple standard spectra except for all abnormal standard spectra, so as to obtain the target average spectrum; The save module is used to associate the target average spectrum with the type of the standard thin film sample and / or the type of the standard spectrum, and save it to a preset configuration file.

8. The thin film thickness measuring device according to claim 7, characterized in that, The filtering module includes: The calculation unit is used to select the n largest standard mean square error values ​​from the plurality of standard mean square error values, and to calculate the repeatability index value of all remaining standard spectra in the plurality of standard spectra except for the n standard spectra that correspond to the n mean square error values ​​respectively; The first processing unit is configured to increment the value of n by 1 when the calculated repeatability index value is greater than the preset repeatability index threshold, and then trigger the calculation unit to re-execute the operation of selecting the largest n standard mean square error values ​​from the plurality of standard mean square error values ​​and calculating the repeatability index values ​​of all remaining standard spectra in the plurality of standard spectra except for the n standard spectra corresponding to the n mean square error values ​​respectively, until the calculated repeatability index value is less than or equal to the preset repeatability index threshold. The second processing unit is used to determine that when the calculated repeatability index value is less than or equal to the preset repeatability index threshold, the n standard spectra that correspond to the current n mean square error values ​​among the multiple standard spectra are all abnormal standard spectra.

9. The thin film thickness measuring device according to claim 6, characterized in that, The preset configuration file stores the correspondence between the type of standard spectrum and the target average spectrum, as well as the correspondence between the type of standard thin film sample and the type of standard spectrum; The determining module is specifically used for: From the preset configuration file, determine the target average spectrum corresponding to the type of standard spectrum that matches the type of standard thin film sample to be tested, and the type of standard spectrum that matches the type of spectrum to be evaluated.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted to be loaded by a processor to perform the method for measuring film thickness according to any one of claims 1-5.

Images