Theoretical spectral data optimization method and measurement method

Abstract

The invention provides a theoretical spectral data optimization method and a measurement method wherein the theoretical spectral data optimization method can obtain an optimized level pair set by obtaining the sensitivity corresponding to each convergence level pair under the condition of obtaining convergence level pair sets and screening the convergence level pair sets according to the sensitivity pairs. And obtaining corresponding theoretical spectral data according to the optimization level pair set. According to the method, the influence of the change quantity of the Fourier coefficient of the order pair on the change quantity of the theoretical spectral data is represented through the sensitivity, and the smaller the value of the sensitivity is, the smaller the influence of the change quantity of the Fourier coefficient of the corresponding order pair on the theoretical spectral data is; the corresponding level is more indispensable in the calculation process of theoretical spectral data. Due to the fact that the number of the order pairs in the optimization order pair set is smaller than that in the convergence order pair set, the calculation efficiency of theoretical spectrum data can be improved through the optimization order pair set.

Description

Technical Field

[0001] The present disclosure relates to the technical field of optical critical dimension measurement, and in particular to a theoretical spectrum data optimization method and a measurement method. Background Art

[0002] Optical Critical Dimension (OCD) technology estimates the specific morphological parameters of the sample by acquiring the scattering signal of the periodic structure in the measurement area on the sample and the model of the sample. It can meet the needs of fast and accurate measurement in new processes and new technologies, and is non-contact and non-destructive. It can be used in the semiconductor field to measure the morphological parameters of samples (such as wafers or masks). The principle of optical critical dimension measurement technology can be generally described as: first establish a theoretical spectral database corresponding to the structural model of the sample, and then obtain the scattering signal of the periodic structure in the measurement area on the sample and match it with the parameters of the theoretical spectral database to estimate the specific morphological parameters of the sample. In the process of establishing the theoretical spectral database corresponding to the structural model of the sample, the algorithm used to calculate the theoretical spectral data is the Rigorous Coupled Wave Analysis (RCWA) algorithm, hereinafter referred to as the RCWA algorithm. The process of the RCWA algorithm is to substitute the Fourier series expansion of the dielectric function of the sample, the incident area, the transmission area, and the electromagnetic field of the sample area into the Maxwell equations, and use the continuity condition of the electromagnetic field to solve the electric field in the incident area to obtain the reflection coefficient of the sample; by traversing each wavelength point and using the RCWA algorithm to calculate the reflection coefficient of the sample at the corresponding wavelength point, the theoretical spectrum data of the sample can be obtained.

[0003] Specifically, in the specific implementation of the RCWA algorithm, for a structural model of a sample having a periodic structure in both the first direction and the second direction, Fourier series expansion is generally performed on different optical characteristic parameters of the structural model in the first direction and the second direction, such as dielectric function, electric field, and magnetic field. For example, Figure 1 A schematic diagram of a structural model of a sample having a periodic structure in both the first direction and the second direction is shown. Figure 1 As shown, the sample can specifically be a wafer, comprising a substrate layer 001, a thin film layer 002, a second direction grating layer 003 extending along the second direction, and a first direction grating layer 004 extending along the first direction, which are stacked in sequence, wherein the first direction X and the second direction Y are two directions perpendicular to each other, and both are perpendicular to the height direction of the sample, and the periodic structure refers to that the sample has a grating structure periodically distributed in the first direction X and the second direction Y, including a plurality of light-transmitting areas and a plurality of opaque areas arranged according to preset rules, which is a conventional setting of those skilled in the art and will not be elaborated here.

[0004] It can be understood that the Fourier series expansion is from negative infinity to positive infinity, and in the actual application of the RCWA algorithm, the expansion order of the Fourier series needs to be truncated, where the maximum truncated order is called the truncation order; that is, in actual applications, the Fourier series expansion of the optical characteristic parameters only needs to be expanded to the truncation order. Figure 1 In the structural model shown, since there are two periodically distributed directions, the first direction X and the second direction Y, there are two truncated orders in the process of Fourier series expansion, namely Nx and Ny, where Nx corresponds to the truncated order in the first direction X, and Ny corresponds to the truncated order in the second direction Y; accordingly, the value of the order m in the first direction X can be an integer in the interval [-Nx, Nx], and the value of the order n in the second direction Y can be an integer in the interval [-Ny, Ny]. The combination of an order m in the first direction X and an order n in the second direction Y is called a order pair, denoted as (m, n), and the set of order pairs corresponding to the entire structural model can be denoted as:

[0005] {(m,n)|-Nx≤m≤Nx,-Ny≤n≤Ny}.

[0006] Based on the above-mentioned order pair set, the theoretical spectral data can be calculated by performing Fourier series expansion on different optical characteristic parameters and then applying the RCWA algorithm, that is, one order pair set can correspond to one theoretical spectral data. It can be understood that the more order pairs corresponding to the Fourier series expansion, the higher the accuracy of the theoretical spectral data calculated by the RCWA algorithm; however, the calculation efficiency of the theoretical spectral data is inversely proportional to the number of order pairs in the order pair set. Therefore, in order to ensure the calculation accuracy of the theoretical spectral data while improving the calculation efficiency, it is necessary to reasonably set the truncated orders in the first direction and the second direction.

[0007] Accordingly, the prior art proposes a solution for performing convergence analysis on the acquired level pair set to obtain a converged level pair set, which may specifically include:

[0008] First, a first level threshold Nxr corresponding to the cutoff level Nx in the first direction and a second level threshold Nyr corresponding to the cutoff level Ny in the second direction are set, wherein the first level threshold Nxr and the second level threshold Nyr are both positive integers. The set first level threshold Nxr and the second level threshold Nyr are respectively used as the cutoff levels in the first direction and the second direction, and the corresponding reference level pair set is obtained, and the first theoretical spectrum data corresponding to the reference level pair set is obtained based on the RCWA algorithm.

[0009] Secondly, according to the set starting point and step size, the truncated levels in the first direction and the second direction are changed so that the truncated levels in the first direction and the second direction increase toward the first level threshold Nxr and the second level threshold Nyr, respectively. In the process of increasing the truncated levels in the above-mentioned steps, a corresponding original level pair set is obtained according to each group of truncated levels in the first direction and the second direction, and the second theoretical spectrum data corresponding to the original level pair set is obtained based on the RCWA algorithm.

[0010] Finally, the mean square error between the second theoretical spectrum data and the first theoretical spectrum data is calculated, and the mean square error is compared with a preset threshold. Then, the original order pair set corresponding to the second theoretical spectrum data whose first mean square error is less than the preset threshold is taken as the converged order pair set, which can be expressed as {(m,n)|-Nxc≤m≤Nxc, -Nyc≤n≤Nyc}, where N xc is the convergence order in the first direction, N yc is the convergence level in the second direction. The level pair in the convergence level pair set is the convergence level pair, denoted as (m,n), where m and n are both levels.

[0011] It can be understood that, in the process of gradually increasing the truncation levels corresponding to the first direction and the second direction from the set starting point (for example, it can be 0) to the first level threshold Nxr and the second level threshold Nyr according to the set step size (for example, increasing by a step size of 1 in both the first direction and the second direction), the larger the truncation levels corresponding to the first direction and the second direction, the more level pairs there are in the corresponding original level pair set, and the smaller the mean square error between the obtained second theoretical spectrum data and the first theoretical spectrum data. The preset threshold corresponding to the mean square error can be determined according to the hardware equipment conditions used for optical critical dimension measurement. The value range of the preset threshold is, for example, 1e-7 to 1. If the mean square error is less than the preset threshold, it means that the second theoretical spectrum data obtained from the corresponding original level pair set can meet the calculation accuracy requirements. In general, the required converged level pair set can be found without increasing to the first level threshold Nxr and the second level threshold Nyr.

[0012] Theoretical spectral data are calculated based on the convergent order pair set, that is, the Fourier series terms corresponding to the convergent order pair set are retained when the optical characteristic parameters are expanded in Fourier series. The RCWA algorithm is used to calculate the theoretical spectral data, which can improve the calculation efficiency while ensuring accuracy. However, the convergent order pair set may still contain a large number of order pairs. In order to further improve the calculation efficiency of the theoretical spectral data, the convergent order pair set needs to be further optimized.

[0013] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0014] In view of the problems in the prior art, the purpose of the present invention is to provide a theoretical spectral data optimization method and a measurement method, which can screen out the optimized level pair set necessary for calculating the theoretical spectral data from the convergent level pair set, thereby improving the calculation efficiency of the theoretical spectral data.

[0015] Specifically, the first aspect of the present disclosure provides a method for optimizing theoretical spectral data, which may include the following steps:

[0016] Acquire a structural model of a sample having a periodic structure in both a first direction and a second direction;

[0017] Performing Fourier series expansion on the optical characteristic parameters corresponding to the structural model in the first direction and the second direction, obtaining a set of order pairs corresponding to the expanded Fourier series, and performing convergence analysis on the set of order pairs to obtain a converged set of order pairs;

[0018] Obtaining the perturbation value of the Fourier coefficient of the dielectric function corresponding to the convergent order pair set, and calculating the third theoretical spectrum data according to the convergent order pair set and the corresponding perturbation value;

[0019] The fourth theoretical spectrum data is obtained by calculating the set of convergent order pairs;

[0020] According to the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data and the perturbation value corresponding to the set of convergence order pairs, the sensitivity corresponding to each convergence order pair is obtained;

[0021] The convergence order pair set is screened according to preset conditions and sensitivity to obtain the optimized order pair set;

[0022] The corresponding theoretical spectral data are obtained according to the optimized order pair set.

[0023] In a possible implementation of the first aspect, the structural model has at least one layer in a third direction, the third direction is perpendicular to the first direction and the second direction, and the third direction is a height direction of the sample;

[0024] In the process of calculating the fourth theoretical spectrum data, the dielectric function is expanded by Fourier series. The calculation formula of the Fourier series of the dielectric function is:

[0025]

[0026] Where x is the coordinate in the first direction; y is the coordinate in the second direction; the structural model has Q layers in the third direction, Q is a positive integer greater than or equal to 1, and z is q is the coordinate of the qth layer in the third direction, 1≤q≤Q; λ is the wavelength; ε(x, y, z q ′λ) is the dielectric function; N xc is the convergence order in the first direction; N yc is the convergence order in the second direction; j is the imaginary unit; p x is the first direction period; p y is the second direction period; ε uv (z q ,λ) is the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function;

[0027] The calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function is:

[0028]

[0029] In the process of calculating the third theoretical spectrum data, the Fourier coefficients corresponding to the convergence order pair (m, n) are perturbed, and the calculation formula of the Fourier series of the dielectric function after perturbation is:

[0030]

[0031] Among them, ε′(x, y, z q ,λ) is the dielectric function after perturbation; Δε mn (z q ,λ) is the perturbation value corresponding to the convergence order pair (m, n), -N xc ≤m≤N xc , -N yc ≤n≤N yc , the modulus of the perturbation value satisfies: 0<|Δε mn (z q ,λ)|≤1;

[0032] The calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbation dielectric function is:

[0033] ε′ uv (z q ,λ)=ε uv (z q ,λ)-δ um δ vn Δε mn (z q ,λ);

[0034] Among them, when u=m, δ um=1; when u≠m, δ um =0; when v = n, δ vn =1; when u≠m, δ vn =0.

[0035] The above technical solution provides a specific implementation process for perturbing the Fourier coefficients corresponding to the convergence order pair (m, n). It can be understood that only the Fourier coefficients corresponding to the specific convergence order pair (m, n) change after the perturbation, and the Fourier coefficients corresponding to other convergence order pairs (u, v) of the dielectric function except the convergence order pair (m, n) remain unchanged.

[0036] In a possible implementation of the first aspect, for different wavelengths and different layers of the structural model in the third direction, the perturbation values ​​corresponding to the convergence levels are different;

[0037] Alternatively, for different wavelengths and / or different layering of the structural model in the third direction, the corresponding perturbation values ​​of the convergence order are the same.

[0038] The above technical scheme provides a variety of optional schemes for the specific setting of perturbation values, among which: the perturbation values ​​corresponding to the convergence order pairs of different layers and different wavelengths in the third direction are different, and the sensitivity results obtained in the subsequent calculation process will more accurately represent the influence of the change in the Fourier coefficients of the convergence order pairs (i.e., the perturbation values) on the change in the theoretical spectral data; and setting the perturbation values ​​corresponding to the convergence order pairs of different layers and different wavelengths in the third direction to be the same can simplify the calculation of sensitivity; and setting the perturbation values ​​to be associated only with different wavelengths can simplify the calculation of sensitivity and highlight the influence of wavelength on theoretical spectral data; and setting the perturbation values ​​to be associated only with different layers in the third direction can simplify the calculation of sensitivity and highlight the influence of the structure of each layer on the theoretical spectral data.

[0039] In a possible implementation of the first aspect above, in the process of calculating the third theoretical spectral data based on the set of convergence order pairs and the corresponding perturbation values, the perturbation value is added only to the Fourier coefficients corresponding to the layered convergence order pairs containing the preset morphological parameters to be measured.

[0040] The above technical solution only adds perturbation values ​​to the corresponding Fourier coefficients for the convergence order of the layer containing the morphological parameters to be measured, and does not add perturbation values ​​to the layer not containing the morphological parameters to be measured, which can highlight the influence of the morphological parameters to be measured on the theoretical spectral data.

[0041] In a possible implementation of the first aspect above, the sensitivity is the ratio of the mean square error between the third theoretical spectral data and the fourth theoretical spectral data to the modulus of the average value of the perturbation values ​​corresponding to all convergence order pairs in the convergence order pair set.

[0042] The above technical solution provides a calculation method for obtaining the sensitivity value, using the ratio of the mean square error to the modulus of the average value of the perturbation value as the sensitivity, to avoid the situation where the sensitivity is smaller due to a larger perturbation value corresponding to a certain layer and a certain wavelength, or the sensitivity is larger due to a smaller perturbation value corresponding to a certain layer and a certain wavelength, so that the obtained sensitivity corresponding to the convergence order pair (m, n) is consistent for all layers and wavelengths.

[0043] In a possible implementation of the first aspect above, the sensitivity is the ratio of the mean square error between the third theoretical spectral data and the fourth theoretical spectral data to the modulus of the maximum value of the perturbation values ​​corresponding to all convergence order pairs in the convergence order pair set.

[0044] The above technical scheme provides another calculation method for obtaining the sensitivity value, using the ratio of the mean square error to the modulus of the maximum value in the perturbation value as the sensitivity, which can minimize the sensitivity result corresponding to the convergence order pair (m, n), so that the sensitivity corresponding to the convergence order pair (m, n) obtained can characterize the influence of the change of the Fourier coefficient of the convergence order pair on the change of the theoretical spectral data to the greatest extent.

[0045] In a possible implementation of the first aspect, screening the convergence level pair set according to preset conditions and sensitivity to obtain the optimized level pair set includes the following steps:

[0046] According to the preset first threshold and second threshold, the convergence level pair set is traversed, the convergence level pairs whose sensitivity is less than or equal to the first threshold are retained, the theoretical spectral data of all retained convergence level pairs are obtained, and when the mean square error between the theoretical spectral data and the fourth theoretical spectral data is less than or equal to the second threshold, all currently retained convergence level pairs are used as the optimized level pair set.

[0047] The above technical scheme provides a specific solution for obtaining the optimized level pair set. By setting and adjusting the first threshold and / or the second threshold, the screened optimized level pair set can meet the accuracy requirements of theoretical spectral data calculation, while improving the computational efficiency of theoretical spectral data acquisition.

[0048] In a possible implementation of the first aspect, screening the convergence level pair set according to preset conditions and sensitivity to obtain the optimized level pair set includes the following steps:

[0049] Sort the convergence order pairs according to the sensitivity;

[0050] According to the preset screening ratio, the convergence level pairs with smaller sensitivity values ​​are retained, and the theoretical spectral data of all retained convergence level pairs are obtained. When the mean square error between the theoretical spectral data and the fourth theoretical spectral data is less than or equal to the preset third threshold, all the currently retained convergence level pairs are used as the optimized level pair set.

[0051] The above technical solution provides another specific solution for obtaining the optimized level pair set. By setting and adjusting the screening ratio and / or the third threshold, the screened optimized level pair set can also meet the accuracy requirements of the theoretical spectral data calculation, while improving the computational efficiency of theoretical spectral data acquisition.

[0052] In a possible implementation of the first aspect above, the optical characteristic parameters include dielectric function, electric field and magnetic field; the third theoretical spectral data are calculated according to the RCWA algorithm, the convergent order pair set and the corresponding perturbation value; the fourth theoretical spectral data are calculated according to the RCWA algorithm and the convergent order pair set; the corresponding theoretical spectral data are obtained according to the RCWA algorithm and the optimized order pair set.

[0053] The above technical solution obtains theoretical spectral data by applying the RCWA algorithm to the set based on the selected optimization level, which can significantly improve the calculation efficiency of theoretical spectral data.

[0054] A second aspect of the present disclosure provides a measurement method, which may specifically include the following steps:

[0055] Establishing a theoretical spectrum database corresponding to the sample according to the theoretical spectrum data optimization method provided in the first aspect, wherein the theoretical spectrum database includes morphological parameters of the sample and theoretical spectrum data corresponding to the morphological parameters;

[0056] Obtaining measurement spectrum data of a measurement area of ​​a sample to be measured;

[0057] The morphological parameters of the measurement area of ​​the sample to be measured are determined based on the measured spectral data and the theoretical spectral database.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] Through the technical solution provided by the present disclosure, when obtaining a convergent order pair set, it is possible to obtain the sensitivity corresponding to each convergent order pair, and screen the convergent order pair set according to the sensitivity to obtain an optimized order pair set, wherein the sensitivity characterizes the influence of the change in the Fourier coefficient of the order pair (i.e., the perturbation value) on the change in the theoretical spectral data, and the smaller the value of the sensitivity, the smaller the influence of the change in the Fourier coefficient of the corresponding order pair on the theoretical spectral data, and the more indispensable the corresponding order pair is in the calculation process of the theoretical spectral data. Since the number of order pairs in the optimized order pair set is less than the number of order pairs in the convergent order pair set, the calculation efficiency of the theoretical spectral data can be improved by applying the RCWA algorithm to the optimized order pair set to calculate the theoretical spectral data. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Other features, objects and advantages of the present disclosure will become more apparent from a reading of the detailed description of non-limiting embodiments made with reference to the following accompanying drawings.

[0061] Figure 1 According to the prior art, a schematic diagram of a structural model of a sample having a periodic structure in both a first direction and a second direction is provided.

[0062] Figure 2 A schematic flow chart of a theoretical spectral data optimization method is provided according to an embodiment of the present disclosure.

[0063] Figure 3 According to an embodiment of the present disclosure, a schematic diagram of sensitivity distribution of each level pair corresponding to a convergence level pair set is provided.

[0064] Figure 4 According to an embodiment of the present disclosure, a schematic diagram of screening a set of optimization level pairs is provided.

[0065] Figure 5 According to an embodiment of the present disclosure, another schematic diagram of screening a set of optimization levels is provided.

[0066] Figure 6 A flow chart of a measurement method is provided according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0067] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in the present disclosure. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in the present disclosure can also be modified or changed in various ways according to different viewpoints and application systems without departing from the spirit of the present disclosure. It should be noted that the embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0068] The following is a detailed description of the embodiments of the present disclosure with reference to the accompanying drawings so that those skilled in the art can easily implement the present disclosure. The present disclosure can be embodied in many different forms and is not limited to the embodiments described herein.

[0069] In the representations of the present disclosure, the reference terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" and the like mean that the specific features, structures, materials, or characteristics represented in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. Moreover, the specific features, structures, materials, or characteristics represented may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples represented in the present disclosure and the features of different embodiments or examples, unless they are mutually contradictory.

[0070] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the representation of the present disclosure, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0071] In order to clearly describe the present disclosure, components not related to the description are omitted, and the same reference numerals are given to the same or similar components throughout the specification.

[0072] Throughout the specification, when a device is said to be "connected" to another device, this includes not only the case of "direct connection" but also the case of "indirect connection" by placing other elements therebetween. In addition, when a device is said to "include" a certain component, unless otherwise stated, it does not exclude other components, but means that other components may be included.

[0073] When a device is said to be "on" another device, it may be directly on the other device, but there may also be other devices between it. In contrast, when a device is said to be "directly" on another device, there are no other devices between it.

[0074] Although the terms first, second, etc. are used to represent various elements in some examples, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, the first interface and the second interface, etc. are represented. Moreover, as used in this article, the singular forms "one", "one" and "the" are intended to also include plural forms, unless there is an opposite indication in the context. It should be further understood that the terms "comprising" and "including" indicate the existence of features, steps, operations, elements, components, projects, kinds, and / or groups, but do not exclude the existence, occurrence or addition of one or more other features, steps, operations, elements, components, projects, kinds, and / or groups. The terms "or" and "and / or" used herein are interpreted as inclusive, or mean any one or any combination. Therefore, "A, B or C" or "A, B and / or C" means "any of the following: A; B; C; A and B; A and C; B and C; A, B and C". Only when the combination of elements, functions, steps or operations is inherently mutually exclusive in some way, will there be an exception to this definition.

[0075] The technical terms used herein are only used to refer to specific embodiments and are not intended to limit the present disclosure. The singular form used herein also includes the plural form unless the sentence clearly indicates the contrary meaning. The meaning of "including" used in the specification is to specify specific characteristics, regions, integers, steps, operations, elements and / or components, and does not exclude the existence or addition of other characteristics, regions, integers, steps, operations, elements and / or components.

[0076] Although not defined differently, all terms, including technical and scientific terms used herein, have the same meaning as those generally understood by those skilled in the art to which the present disclosure belongs. Terms defined in commonly used dictionaries are additionally interpreted as having meanings consistent with the content of relevant technical literature and current disclosures, and unless defined, shall not be overly interpreted as ideal or very formal meanings.

[0077] According to the relevant explanations and descriptions in the prior art, although the existing convergence analysis can improve the calculation efficiency while ensuring the calculation accuracy, in order to further improve the calculation efficiency, it is necessary to further optimize the set of convergence level pairs, and further reduce the number of level pairs required while ensuring the calculation accuracy.

[0078] In order to solve the above problems, the technical solution provided by the present disclosure provides a theoretical spectral data optimization method and a measurement method, which can screen out the optimization level pair set necessary for calculating the theoretical spectral data from the convergent level pair set according to the sensitivity, thereby improving the calculation efficiency of the theoretical spectral data. The specific implementation of this optimization method will be specifically explained below.

[0079] Specifically, in some embodiments provided in the present disclosure, Figure 2 A schematic flow chart of a theoretical spectral data optimization method is provided. Figure 2 As shown, the following steps may be specifically included:

[0080] Step 100: Obtain a structural model of a sample having a periodic structure in both the first direction and the second direction. The structural model can be as follows: Figure 1 As shown, no further elaboration is given here.

[0081] Step 200: Perform Fourier series expansion on the optical characteristic parameters corresponding to the structural model in the first direction and the second direction, obtain a set of level pairs corresponding to the expanded Fourier series, and perform convergence analysis on the set of level pairs to obtain a converged set of level pairs. For obtaining the converged set of level pairs, reference can be made to the relevant technical solutions provided in the background technology section, which will not be elaborated here.

[0082] Step 300: Obtain the perturbation value of the Fourier coefficient of the dielectric function corresponding to the convergence order pair set, and calculate the third theoretical spectrum data according to the convergence order pair set and the corresponding perturbation value. In the embodiment of the present disclosure, the setting of the perturbation value and the specific acquisition of the third theoretical spectrum data will be described later and will not be repeated here.

[0083] Step 400: Obtain fourth theoretical spectrum data by calculating the set according to the convergence level. The specific acquisition of the fourth theoretical spectrum data will be described later and will not be repeated here. In addition, the execution of the above steps 300 and 400 can be performed synchronously or asynchronously, and the execution order is not limited.

[0084] Step 500: Obtain the sensitivity corresponding to each convergent order pair according to the mean square error between the third theoretical spectral data and the fourth theoretical spectral data and the perturbation value corresponding to the convergent order pair set. The sensitivity is used to characterize the influence of the change in the Fourier coefficient of the order pair (i.e., the perturbation value) on the change in the theoretical spectral data. The lower the value of the sensitivity, the smaller the influence of the error in the Fourier coefficient of the corresponding order pair on the theoretical spectral data in the process of calculating the theoretical spectral data, which means that the corresponding order pair is necessary for calculating the theoretical spectral data. Therefore, the sensitivity can be used to further screen the order pairs in the convergent order pair set (i.e., the convergent order pairs). In the embodiments disclosed herein, the specific calculation method of the sensitivity will be described later and will not be repeated here.

[0085] Step 600: Screen the convergence level pair set according to the preset conditions and sensitivity to obtain the optimized level pair set. In the embodiment of the present disclosure, the specific preset conditions and screening process will be described later and will not be repeated here.

[0086] Step 700: Obtain corresponding theoretical spectral data according to the set of optimization levels.

[0087] It can be understood that, through the above steps 100 to 700, the sensitivity corresponding to each convergent order pair can be obtained by increasing the perturbation value, and the convergent order pair set can be further screened according to the sensitivity, so as to obtain an optimized order pair set with fewer order pairs. Each order pair in the optimized order pair set is necessary for the calculation of theoretical spectral data. Therefore, using the optimized order pair set to obtain theoretical spectral data can improve the calculation efficiency of theoretical spectral data while ensuring the accuracy of theoretical spectral data.

[0088] The specific implementation of the above steps 100 to 700 will be further described below.

[0089] In the above embodiment, specifically, the optical characteristic parameters may include dielectric function, electric field and magnetic field. Accordingly, in the process of obtaining the third theoretical spectrum data, it can be calculated according to the RCWA algorithm, the convergence order pair set and the corresponding perturbation value to obtain; similarly, in the process of obtaining the fourth theoretical spectrum data, it can be calculated according to the RCWA algorithm and the convergence order pair set to obtain; in the process of obtaining the theoretical spectrum data, it can also be calculated according to the RCWA algorithm and the optimization order pair set to obtain.

[0090] The following will take the specific acquisition process of the fourth theoretical spectrum data as an example for exemplary description:

[0091] In one embodiment of the present disclosure, in addition to having a periodic structure in the first direction X and the second direction Y, the sample also has at least one layer in the third direction Z. The third direction Z is perpendicular to the first direction X and the second direction Y. Figure 1 In the structural model shown, the third direction Z is the height direction of the sample. On this basis, in the process of calculating the fourth theoretical spectrum data, the dielectric function in the optical characteristic parameter is expanded by Fourier series, where the calculation formula of the Fourier series of the dielectric function can be expressed as:

[0092]

[0093] Where x is the coordinate in the first direction; y is the coordinate in the second direction; the structural model has Q layers in the third direction, Q is a positive integer greater than or equal to 1, and z is qis the coordinate of the qth layer in the third direction, 1≤q≤Q; λ is the wavelength; ε(x, y, z q ,λ) is the dielectric function; N xc is the convergence order in the first direction; N yc is the convergence order in the second direction; j is the imaginary unit; p x is the first direction period; p y is the second direction period; ε uv (z q ,λ) is the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function. It can be understood that the Fourier coefficient ε uv (z q ,λ) is at wavelength λ and the qth layer coordinate is z q The Fourier coefficient corresponding to the order pair (u, v) in the Fourier series of the dielectric function at the time uv (z q ,λ) is generally different for different wavelengths and different layers.

[0094] Among them, the Fourier order pair (u, v) is equivalent to the convergence order pair (m, n) in the background technology. This is because the orders m and n in the background technology are variables. In the embodiment of the present disclosure, the letters u and v are also variables. The purpose of using these variables is to implement traversal within the range of the corresponding convergence order, as shown in the above formula, -N xc ≤u≤N xc , -N yc ≤v≤N yc , the letters u and v are used here instead of m and n because: in this embodiment, it is necessary to distinguish a specific level pair (m, n) from the other level pairs except the specific level pair (m, n) in the convergent level pair set, so different letters are used to indicate. In addition, the convergent level pair (m, n) can also be called a Fourier level pair, but in the disclosed embodiment, the level pair (m, n) is called a convergent level pair, and the level pair (u, v) is called a Fourier level pair, which is just a naming of the level pairs.

[0095] By q The dielectric function ε(x, y, z q ,λ) is transformed to obtain the Fourier coefficient ε uv (z q , λ), its calculation formula can be expressed as:

[0096]

[0097] It represents the wavelength λ and the qth layer coordinate z qThe Fourier coefficient corresponding to the order pair (u, v) in the Fourier series of the dielectric function at the time is the qth layer coordinate z q and wavelength λ.

[0098] In the process of calculating the fourth theoretical spectrum data, it also includes performing Fourier series expansion on the electric field and magnetic field in the optical characteristic parameters, and substituting the expansion results and the expansion results of the dielectric function into the Maxwell equations, and then using the relevant process of the RCWA algorithm to obtain the corresponding fourth theoretical spectrum data. This is an existing technical solution that can be applied by technicians in this field and will not be elaborated here.

[0099] In the above embodiments, it can be understood that the technical solution provided by the present disclosure further obtains the influence of the change in the Fourier coefficient of the convergence order pair on the change in the theoretical spectral data by adding a perturbation value to the Fourier coefficient corresponding to the convergence order pair.

[0100] In order to introduce the specific implementation process of the above step 300 in detail, the process of perturbing the Fourier coefficients corresponding to the convergence order (m, n) is specifically described below:

[0101] In the process of calculating the third theoretical spectrum data, the Fourier coefficients corresponding to the convergence order (m, n) are perturbed, and the calculation formula of the Fourier series of the dielectric function after perturbation is:

[0102]

[0103] Among them, ε′(x, y, z q ,λ) is the dielectric function after perturbation; Δε mn (z q ,λ) is the perturbation value corresponding to the convergence order pair (m, n), -N xc ≤m≤N xc , -N yc ≤n≤N yc , the modulus of the perturbation value satisfies: 0<|Δε mn (z q ,λ)|≤1;

[0104] The calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbation dielectric function is:

[0105] ε′ uv (z q ,λ)=ε uv (z q ,λ)-δ um δ vn Δε mn (z q ,λ);

[0106] Among them, when u=m, δ um =1; when u≠m, δ um =0; when v = n, δ vn =1; when u≠m, δ vn =0.

[0107] It can be understood that, in the set of Fourier order pairs of dielectric functions (i.e., the set of convergent order pairs), only the Fourier coefficients corresponding to a specific convergent order pair (m, n) change after the perturbation, and the Fourier coefficients corresponding to other convergent order pairs in the set of Fourier order pairs of dielectric functions except the convergent order pair (m, n) remain unchanged. In a specific example, if N is set xc =1,N yc = 1, then for the original (i.e., non-perturbative) dielectric function ε(x, y, z q The Fourier coefficient of the Fourier series of uv (z q ,λ), there is ε -1，-1 (z q ,λ),ε -1，0 (z q ,λ),ε -1，1 (z q ,λ),ε 0，-1 (z q ,λ),ε 0，0 (z q ,λ),ε 0，1 (z q ,λ),ε 1，-1 (z q ,λ),ε 1，0 (z q ,λ),ε 1，1 (z q ,λ) are the nine original Fourier coefficients. If we further set m=1 and n=1, then the perturbation order pair is (1,1). At this time, the perturbed dielectric function ε′(x,y,z q The Fourier coefficient of the Fourier series of uv (z q ,λ), there exists ε′ ′1，-1 (z q ,α),ε′ -1，0 (z q ,λ),ε′ -1，1 (z q ,λ),ε′ 0，-1 (z q ,λ),ε′ 0，0 (z q ,λ),ε′ 0，1(z q ,λ),ε′ 1，-1 (z q ,λ),ε′ 1，0 (z q ,λ),ε′ 1，1 (z q ,λ) There are nine perturbed Fourier coefficients, of which only ε 1，1 (z q ,λ)≠ε1′ ，1 (z q ,λ), the remaining eight perturbed Fourier coefficients are equal to the corresponding original Fourier coefficients.

[0108] In one embodiment of the present disclosure, the specific setting of the perturbation value is further described below:

[0109] In an optional implementation of the above embodiment, for different wavelengths and different layers of the structural model in the third direction, the perturbation values ​​corresponding to the convergence order can be set to be different. It can be understood that, as shown above, Δε mn (z q ,λ) is the perturbation value corresponding to the convergence order pair (m,n), where the variable contains the qth layer coordinate z q As well as the wavelength λ, different perturbation values ​​are set for different layers and different wavelengths. The sensitivity results obtained in the subsequent calculation process will more accurately represent the impact of the change in the Fourier coefficients of the convergence order pair on the change in the theoretical spectral data.

[0110] In another optional implementation of the above embodiment, for different wavelengths and different layers of the structural model in the third direction, the perturbation values ​​corresponding to the convergence order can be set to be the same, that is, Δε mn (z q ,λ)=α,0<|α|≤1, the above setting can simplify the calculation of sensitivity. Correspondingly, the calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbation dielectric function can be simplified as follows:

[0111] ε′ uv (z q ,λ)=ε uv (z q ,λ)-δ um δ vn α.

[0112] In another optional implementation of the above embodiment, the perturbation value may be set to be associated only with different wavelengths, that is, the perturbation values ​​for different layers of the structural model in the third direction are the same, but the perturbation values ​​for different wavelengths are different, that is, the perturbation value Δε mn (z q,λ) is simplified to Δε mn (λ), which can simplify the calculation of sensitivity and highlight the influence of wavelength on theoretical spectral data. Correspondingly, the calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbation dielectric function can be simplified to:

[0113] ε′ uv (z q ,λ)=ε uv (z q ,λ)-δ um δ vn Δε mn (λ).

[0114] In another optional implementation of the above embodiment, the perturbation value may be set to be associated only with different layers of the structural model in the third direction, that is, the perturbation values ​​for different layers of the structural model in the third direction are different, but the perturbation values ​​for different wavelengths are the same, that is, the perturbation value Δε mn (z q ,λ) is simplified to Δε mn (z q ), which can simplify the calculation of sensitivity and highlight the influence of each layer structure on the theoretical spectral data. Accordingly, the calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbed dielectric function can be simplified to:

[0115] ε′ uv (z q ,λ)=ε uv (z q ,λ)-δ um δ vn Δε mn (z q ).

[0116] In another optional implementation of the above embodiment, the perturbation value is added only to the Fourier coefficients corresponding to the convergence order pair of the layer containing the morphology parameter to be measured, and the perturbation value is not added to the layer not containing the morphology parameter to be measured. Accordingly, the calculation formula of the Fourier coefficients corresponding to the Fourier order pair (u, v) of the dielectric function after perturbation can be classified and expressed as:

[0117] ε′ uv (z i ,λ)=ε uv (z i ,λ)-δ um δ vn Δε mn (z i ,λ);

[0118] ε′ uv (zc ,λ)=ε uv (z c ,λ);

[0119] Among them, z i represents the coordinates of the i-th layer containing the morphological parameters to be measured, z c represents the coordinates of the cth layer that does not contain the morphological parameters to be measured, where 1≤i≤Q, 1≤c≤Q. Figure 1 In the structural model shown, the coordinate plane XOY corresponding to the first direction X and the second direction Y is coplanar with the top layer of the first direction grating layer 004 in the structural model. The height of the first direction grating layer 004 in the third direction Z is 5nm, the height of the second direction grating layer 003 in the third direction Z is 5nm, the height of the thin film layer 002 in the third direction Z is 1nm, and the height of the substrate layer 001 in the third direction Z is 775um. If the thin film layer 002, the second direction grating layer 003 and the first direction grating layer 004 are equally divided into layers in units of 1nm, and the substrate layer 001 is regarded as an independent layer, then the coordinates of each layer in the third direction Z are z1=1, z2=2, z3=3, z4=4, z5=5, z6=6, z7=7, z8=8, z9=9, ... 10 =10, z 11 =11, z 12 =775011. If the measured morphology parameter is specifically the height of the first directional grating layer 004, then the coordinates of the layer containing the measured morphology parameter are z1=1, z2=2, z3=3, z4=4, z5=5. Correspondingly, the calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the perturbation dielectric function can be classified and expressed as:

[0120] ε′ uv (z i ,λ)=ε uv (z i ,λ)-δ um δ vn Δε mn (z i , λ), i=1, 2, 3, 4, 5;

[0121] ε′ uv (z c ,λ)=ε uv (z c , λ), c=6, 7, 8, 9, 10, 11, 12.

[0122] It can be understood that the above settings can simplify the calculation of sensitivity and highlight the influence of the measured morphological parameters on the theoretical spectral data.

[0123] In the above embodiment, in the specific process of obtaining the sensitivity, the sensitivity can be set as the ratio of the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data to the modulus of the average value of the perturbation values ​​corresponding to all the convergence order pairs in the convergence order pair set, and its mathematical expression can be as follows:

[0124]

[0125] Wherein, sensitivity(m,n) is the sensitivity of the convergence order to (m,n), mse is the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data, is the average value of the perturbation values ​​corresponding to all the convergence order pairs. It can be understood that the ratio of the mean square error to the modulus of the perturbation value average is used as the sensitivity to avoid the small sensitivity caused by the large perturbation value corresponding to a certain layer and a certain wavelength or the large sensitivity caused by the small perturbation value corresponding to a certain layer and a certain wavelength, so that the obtained sensitivity corresponding to the convergence order pair (m, n) is consistent for all layers and wavelengths.

[0126] In the above embodiment, in the specific process of obtaining the sensitivity, the sensitivity can also be set as the ratio of the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data to the modulus of the maximum value of the perturbation values ​​corresponding to all the convergence order pairs in the convergence order pair set, and its mathematical expression can be as follows:

[0127]

[0128] Among them, sensitivity(m,n) is the sensitivity of the convergence order to (m,n), mse is the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data, max(Δε mn (z q , λ)) is the maximum value of the perturbation value corresponding to all the convergence order pairs. It can be understood that the ratio of the mean square error to the modulus of the maximum perturbation value is used as the sensitivity, which can minimize the sensitivity result corresponding to the obtained convergence order pair (m, n), so that the obtained sensitivity corresponding to the convergence order pair (m, n) can characterize the influence of the change of the Fourier coefficient of the convergence order pair on the change of the theoretical spectrum data to the greatest extent.

[0129] It can be understood that in the embodiments provided in the present disclosure, sensitivity is used to characterize the influence of the change in the Fourier coefficients of the order pair on the change in the theoretical spectral data. The lower the sensitivity value, the smaller the influence of the error in the Fourier coefficients of the corresponding order pair on the theoretical spectral data in the process of calculating the theoretical spectral data, which means that the corresponding order pair is necessary for calculating the theoretical spectral data. Therefore, in the subsequent screening process, the order pairs with smaller sensitivity need to be retained as the optimized order pair set.

[0130] Specifically, in an optional implementation of the above embodiment, the first threshold and the second threshold can be pre-set as needed, and the convergence level pair set can be traversed according to the first threshold and the second threshold, and the convergence level pairs with a sensitivity less than or equal to the first threshold are first retained; then the theoretical spectrum data of all the retained convergence level pairs are obtained, and when the mean square error between the theoretical spectrum data and the fourth theoretical spectrum data is less than or equal to the second threshold, all the currently retained convergence level pairs are used as the optimized level pair set. The value range of the first threshold can be set to [0,1], and the value range of the second threshold can be set to [1e-10, 1e-4]. It can be understood that the setting of the first threshold is adjustable within its value range, and the setting of the first threshold can be based on the mean square error between the theoretical spectrum data obtained by the level pair corresponding to the first threshold and the fourth theoretical spectrum data being less than or equal to the second threshold; when the above conditions are met, the smaller the first threshold, the fewer the number of level pairs in the optimized level pair set, and the higher the calculation efficiency of the corresponding theoretical spectrum data. Those skilled in the art will appreciate that the first threshold and / or the second threshold can be set and adjusted so that the screened set of optimized level pairs meets the accuracy requirements of theoretical spectral data calculation, while improving the computational efficiency of theoretical spectral data acquisition.

[0131] For example, Figure 3 A schematic diagram of the sensitivity distribution of each order pair in a convergence order pair set is shown, wherein the convergence order pair set is represented by {(m,n)|-2≤m≤2,-2≤n≤2}. In the process of obtaining the sensitivity, the perturbation value Δε corresponding to the convergence order pairs of different layers and different wavelengths in the third direction of the structural model is set. mn (z q ,λ) are the same, and Δε mn (z q , the value of λ) is 1, so as to calculate Figure 3The sensitivity distribution diagram shown in FIG. 1 shows the sensitivity distribution diagram, in which the horizontal direction represents the order m in the first direction X, and the vertical direction represents the order n in the second direction Y. The figure also shows the sensitivity of each convergence order pair (m, n). For example, the sensitivity of the convergence order pair (-2, -2) is 6.68e-11, which is indicated by 6.68E-11 in the figure. Using the screening method in the above optional implementation, setting the first threshold to 0 and the second threshold to 1e-4, the retained convergence order pairs are as follows: Figure 4 The convergence level pairs covered by the medium and dark areas, that is, the convergence level pairs covered by the dark areas, constitute the set of optimized level pairs. Figure 3 and Figure 4 It can be seen that the convergent order pair set contains 25 convergent order pairs, and the optimized order pair set contains 9 convergent order pairs. The number of order pairs in the optimized order pair set obtained by the optimization method provided by the present invention is smaller, thereby improving the efficiency of subsequent calculation of theoretical spectral data using the RCWA algorithm.

[0132] In another optional implementation of the above embodiment, the convergence order pairs may be sorted according to the sensitivity first; and then the convergence order pairs with smaller sensitivity values ​​may be retained according to the preset screening ratio; and then the theoretical spectral data of all the retained convergence order pairs are obtained, and when the mean square error between the theoretical spectral data and the fourth theoretical spectral data is less than or equal to the preset third threshold, all the currently retained convergence order pairs are used as the optimized order pair set. The value range of the third threshold can be set to [1e-10, 1e-4], and the value range of the screening ratio is (0, 1). Similarly, the screening ratio in the above optional implementation is adjustable, and the setting of the screening ratio can be based on the mean square error between the theoretical spectral data obtained from all the retained convergence order pairs corresponding to the screening ratio and the fourth theoretical spectral data being less than or equal to the third threshold; when the above conditions are met, the smaller the screening ratio, the fewer the number of order pairs in the optimized order pair set, and the higher the calculation efficiency of the corresponding theoretical spectral data. Those skilled in the art can understand that, by setting and adjusting the screening ratio and / or the third threshold, the set of optimized level pairs obtained by screening can also meet the accuracy requirements of theoretical spectral data calculation, while improving the computational efficiency of theoretical spectral data acquisition.

[0133] For example, Figure 3The schematic diagram of the sensitivity distribution of the order pairs shown in the figure adopts the screening method in the above optional implementation, and sets the screening ratio to 52%, so as to retain the convergent order pairs corresponding to the 52% sensitivities with smaller sensitivities. Taking the sorting of the sensitivities from small to large as an example, the convergent order pairs corresponding to the first 52% sensitivities are retained. Taking the sorting of the sensitivities from large to small as an example, the convergent order pairs corresponding to the last 52% sensitivities are retained. The third threshold is 1e-4, so the retained convergent order pairs are as follows: Figure 5 The convergence level pairs covered by the medium and dark areas, that is, the convergence level pairs covered by the dark areas, constitute the set of optimized level pairs. Figure 3 and Figure 5 It can be seen that the convergent order pair set contains 25 convergent order pairs, and the optimized order pair set contains 13 convergent order pairs. The number of order pairs in the optimized order pair set obtained by the optimization method provided by the present invention is smaller, which can also improve the efficiency of subsequent calculation of theoretical spectral data using the RCWA algorithm.

[0134] In some embodiments of the present disclosure, Figure 6 A flow chart of a measurement method is also shown. Figure 6 As shown, this measurement method may specifically include the following steps:

[0135] Step 601: Establish a theoretical spectrum database corresponding to the sample according to a theoretical spectrum data optimization method. The theoretical spectrum data optimization method is the method described in the above embodiment and will not be described in detail here; the theoretical spectrum database includes the morphological parameters of the sample and the theoretical spectrum data corresponding to the morphological parameters.

[0136] Step 602: Obtaining measurement spectrum data of a measurement area of ​​a sample to be measured.

[0137] Step 603: Determine the morphological parameters of the measurement area of ​​the sample to be measured according to the measured spectrum data and the theoretical spectrum database.

[0138] In summary, through the technical solution provided by the present disclosure, when obtaining a convergent order pair set, it is possible to obtain the sensitivity corresponding to each convergent order pair, and screen the convergent order pair set according to the sensitivity to obtain an optimized order pair set, wherein the sensitivity characterizes the influence of the change in the Fourier coefficient of the order pair (i.e., the perturbation value) on the change in the theoretical spectral data. The smaller the value of the sensitivity, the smaller the influence of the change in the Fourier coefficient of the corresponding order pair on the change in the theoretical spectral data, and the more indispensable the corresponding order pair is in the calculation process of the theoretical spectral data. Since the number of order pairs in the optimized order pair set is less than that in the convergent order pair set, the calculation efficiency of the theoretical spectral data can be improved by applying the RCWA algorithm to the optimized order pair set to calculate the theoretical spectral data.

[0139] The above content is a further detailed description of the present disclosure in combination with specific preferred implementation modes, and it cannot be determined that the specific implementation of the present disclosure is limited to these descriptions. For ordinary technicians in the technical field to which the present disclosure belongs, several simple deductions or substitutions can be made without departing from the concept of the present disclosure, which should be regarded as falling within the scope of protection of the present disclosure.

Claims

1. A method for optimizing theoretical spectral data, characterized in that: The steps include: Acquire a structural model of a sample having a periodic structure in both a first direction and a second direction; Performing Fourier series expansion on the optical characteristic parameters corresponding to the structural model in the first direction and the second direction, obtaining a set of order pairs corresponding to the expanded Fourier series, and performing convergence analysis on the set of order pairs to obtain a converged set of order pairs; Obtaining a perturbation value of the Fourier coefficient of the dielectric function corresponding to the convergence order pair set, and calculating and obtaining third theoretical spectrum data according to the convergence order pair set and the corresponding perturbation value; Obtain fourth theoretical spectrum data by calculating the set of convergence order pairs; Obtaining the sensitivity corresponding to each of the convergence order pairs according to the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data and the perturbation value corresponding to the set of convergence order pairs; Screening the convergence order pair set according to the preset condition and the sensitivity to obtain an optimized order pair set; The corresponding theoretical spectrum data is obtained according to the set of optimization levels.

2. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: The structural model has at least one layer in a third direction, the third direction is perpendicular to the first direction and the second direction, and the third direction is the height direction of the sample; In the process of calculating the fourth theoretical spectrum data, the dielectric function is expanded by Fourier series. The calculation formula of the Fourier series of the dielectric function is: Wherein, x is the coordinate in the first direction; y is the coordinate in the second direction; the structural model has Q layers in the third direction, Q is a positive integer greater than or equal to 1, and z q is the coordinate of the qth layer in the third direction, 1≤q≤Q; λ is the wavelength; ε(x, y, z q ,λ) is the dielectric function; N xc is the convergence order in the first direction; N yc is the convergence order in the second direction; j is the imaginary unit; p x is the first direction period; p y is the second direction period; ε uv (z q , λ) is the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function; The calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function is: In the process of calculating the third theoretical spectrum data, the Fourier coefficients corresponding to the convergence order pair (m, n) are perturbed, and the calculation formula of the Fourier series of the dielectric function after perturbation is: Among them, ε′(x, y, z q ,λ) is the dielectric function after perturbation; Δε mn (z q ,λ) is the perturbation value corresponding to the convergence order pair (m, n), -N xc ≤m≤N xc , -N yc ≤n≤N yc , the modulus of the perturbation value satisfies: 0<|Δε mm (z q ,λ)|≤1; The calculation formula of the Fourier coefficient corresponding to the Fourier order pair (u, v) of the dielectric function after perturbation is: e' uv (z q ,λ)=ε uv (z q ,l)-d um d vn No mn (z q ,l); Among them, when u=m, δ um =1; when u≠m, δ um =0; when v = n, δ vn =1; when u≠m, δ vn =0.

3. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: For different wavelengths and different layers of the structural model in the third direction, the perturbation values ​​corresponding to the convergence order are different; Alternatively, for different wavelengths and / or different layers of the structural model in the third direction, the perturbation values ​​corresponding to the convergence orders are the same.

4. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: In the process of calculating the third theoretical spectrum data according to the set of convergence order pairs and the corresponding perturbation values, the perturbation value is added only to the Fourier coefficients corresponding to the layered convergence order pairs containing the preset morphological parameters to be measured.

5. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: The sensitivity is the ratio of the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data to the modulus of the average value of the perturbation values ​​corresponding to all the convergence order pairs in the convergence order pair set.

6. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: The sensitivity is the ratio of the mean square error between the third theoretical spectrum data and the fourth theoretical spectrum data to the modulus of the maximum value of the perturbation values ​​corresponding to all the convergence order pairs in the convergence order pair set.

7. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: The step of screening the convergence level pair set according to the preset condition and the sensitivity to obtain the optimized level pair set comprises the following steps: The convergence level pair set is traversed according to the preset first threshold and second threshold, the convergence level pairs whose sensitivity is less than or equal to the first threshold are retained, the theoretical spectral data of all the retained convergence level pairs are obtained, and when the mean square error between the theoretical spectral data and the fourth theoretical spectral data is less than or equal to the second threshold, all the currently retained convergence level pairs are used as the optimized level pair set.

8. The method for optimizing theoretical spectrum data according to claim 1, characterized in that: The step of screening the convergence level pair set according to the preset condition and the sensitivity to obtain the optimized level pair set comprises the following steps: Sorting the convergence order pairs according to the sensitivity; The convergence level pairs with smaller sensitivity values ​​are retained according to a preset screening ratio, and theoretical spectral data of all retained convergence level pairs are obtained. When the mean square error between the theoretical spectral data and the fourth theoretical spectral data is less than or equal to a preset third threshold, all the currently retained convergence level pairs are used as the optimized level pair set.

9. The method for optimizing theoretical spectrum data according to any one of claims 1 to 8, characterized in that: The optical characteristic parameters include dielectric function, electric field and magnetic field; the third theoretical spectral data are calculated according to the RCWA algorithm, the convergence order pair set and the corresponding perturbation value; the fourth theoretical spectral data are calculated according to the RCWA algorithm and the convergence order pair set; the corresponding theoretical spectral data are obtained according to the RCWA algorithm and the optimized order pair set.

10. A measurement method, characterized in that: The steps include: Establishing a theoretical spectrum database corresponding to the sample according to the theoretical spectrum data optimization method according to any one of claims 1 to 9, wherein the theoretical spectrum database includes morphological parameters of the sample and theoretical spectrum data corresponding to the morphological parameters; Obtaining measurement spectrum data of a measurement area of ​​a sample to be measured; The morphological parameters of the measurement area of ​​the sample to be measured are determined according to the measured spectral data and the theoretical spectral database.

Images