Feature size parameter extraction method and system, terminal device and storage medium

Abstract

A feature size parameter extraction method and system, a terminal device, and a storage medium. The method comprises: providing a plurality of feature sizes of a feature to be measured and simulation spectrums corresponding to the feature sizes (1); using the feature sizes and the simulation spectrums as a data set of a neural network model to train the neural network model (2), until a training error of the neural network model is less than a preset threshold, and determining that the neural network model is trained; measuring a measurement spectrum of said feature by using a measurement device (3); predicting the feature sizes of said feature on the basis of the measurement spectrum and the trained neural network model, to obtain a predicted size range of the feature sizes (4); and fitting the measurement spectrum with corresponding simulation spectrums in the predicted size range to obtain a simulation spectrum that best fits the measurement spectrum, and extracting a theoretical feature size corresponding to the best-fit simulation spectrum as a final feature size (5). According to the method, feature sizes to be measured can be rapidly measured, and the needs of online measurement are met.

Description

Feature size parameter extraction method, system, terminal device and storage medium Technical Field

[0001] The present invention relates to the technical field of characteristic size detection, and in particular to a characteristic size parameter extraction method, system, terminal device and storage medium based on a neural network. Background Art

[0002] In the current semiconductor manufacturing field, it is necessary to measure the characteristic dimensions of nanostructures within semiconductors. For example, in the field of semiconductor nanostructure characteristic dimension measurement based on optical scatterometry, ellipsometers are a common characteristic dimension measurement device. The process of inverting and extracting the measured nanostructure parameters from the spectrum obtained by ellipsometer measurement is a typical inverse problem solution process. Solving the inverse problem in optical scatterometry is one of the key technologies. Traditional methods for solving inverse problems mainly rely on nonlinear fitting and library matching methods.

[0003] Generally, it is necessary to use numerical solution algorithms (such as complex simulation algorithms such as rigorous coupled wave analysis (RCWA)) to construct an optical property model for simulation. After obtaining the simulated spectrum, it is fitted with the measured spectrum. Each fitting requires the optical property model to be simulated once until the best fit result is obtained. The characteristic size is then obtained based on the fitting result. This takes a long time and requires a lot of calculations, and it cannot achieve online measurement of characteristic size. Technical issues

[0004] The main technical problem solved by the present invention is that the existing measurement method cannot realize online measurement of characteristic dimensions. Technical Solutions

[0005] According to the first aspect, an embodiment provides a method for extracting characteristic size parameters based on a neural network, comprising:

[0006] Providing multiple characteristic dimensions of the feature to be measured and the simulated spectra corresponding to the characteristic dimensions;

[0007] The characteristic size and the simulated spectrum are used as data sets of the neural network model to train the neural network model until the training error of the neural network model is less than a preset threshold, and the training of the neural network model is determined to be completed;

[0008] Measuring a measurement spectrum of the characteristic to be measured using a measurement device;

[0009] The characteristic size of the feature to be measured is predicted based on the measured spectrum and the trained neural network model to obtain a predicted size range of the characteristic size;

[0010] The measured spectrum is fitted with the corresponding simulated spectrum within the predicted size range to obtain the simulated spectrum that best fits the measured spectrum, and the theoretical characteristic size corresponding to the best-fitting simulated spectrum is extracted as the final characteristic size.

[0011] According to the second aspect, an embodiment provides a neural network-based feature size parameter extraction system, comprising:

[0012] A spectrum library module is used to provide multiple characteristic sizes of the feature to be measured and simulated spectra corresponding to the characteristic sizes;

[0013] A measuring device, used for measuring the spectrum of the characteristic to be measured to obtain a measured spectrum;

[0014] A neural network module is used to predict the characteristic size of the feature to be measured based on the measured spectrum and the trained neural network model to obtain a predicted size range of the characteristic size;

[0015] The fitting module fits the measured spectrum with the corresponding simulated spectrum within the predicted size range to obtain the simulated spectrum that best fits the measured spectrum, and extracts the theoretical characteristic size corresponding to the best-fitting simulated spectrum as the final characteristic size.

[0016] According to the third aspect, an embodiment provides a terminal device, including:

[0017] Memory, used to store programs;

[0018] A processor is configured to implement the method described in the first aspect by executing a program stored in a memory.

[0019] According to a fourth aspect, an embodiment provides a computer-readable storage medium, on which a program is stored. The program can be executed by a processor to implement the method described in the first aspect. Beneficial effects

[0020] According to the characteristic dimension parameter extraction method, system, terminal device, and storage medium of the above-described embodiment, a spectral library is first established using preset characteristic dimensions and corresponding simulated spectra. A neural network model is then trained using the data from the spectral library. Finally, the trained neural network model is used to predict the characteristic dimensions and fit the actual measured spectrum to obtain a simulated spectrum that best fits the measured spectrum. The theoretical characteristic dimensions corresponding to the best-fitting simulated spectrum are then extracted as the final characteristic dimensions. Therefore, using the neural network model for prediction and fitting allows for rapid measurement of the characteristic dimensions to be measured, meeting the requirements of online measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] FIG1 is a flow chart (I) of a method for extracting characteristic size parameters provided by an embodiment of the present application;

[0022] FIG2 is a flow chart (II) of a characteristic size parameter extraction method provided by an embodiment of the present application;

[0023] FIG3 is a schematic structural diagram of a characteristic size parameter extraction system provided by an embodiment of the present application.

[0024] Reference numerals: 10 - spectral library module; 20 - measurement device; 30 - neural network module; 40 - fitting module. Modes for Carrying Out the Invention

[0025] The present invention will be further described in detail below by means of specific embodiments in conjunction with the accompanying drawings. Similar elements in different embodiments are numbered with associated similar elements. In the following embodiments, many detailed descriptions are provided to enable the present application to be better understood. However, those skilled in the art will readily appreciate that some of the features may be omitted in different circumstances, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the present application are not shown or described in the specification. This is to avoid the core portion of the present application being overwhelmed by excessive descriptions, and for those skilled in the art, it is not necessary to describe these related operations in detail. They will fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0026] In addition, the features, operations, or characteristics described in the specification may be combined in any appropriate manner to form various embodiments. Furthermore, the steps or actions in the method description may be reordered or adjusted in a manner readily apparent to those skilled in the art. Therefore, the various sequences in the specification and drawings are provided solely for the purpose of clearly describing a particular embodiment and are not intended to be mandatory, unless otherwise specified.

[0027] Component numbers used herein, such as "first" and "second," are used solely to distinguish the components being described and do not convey any sequential or technical meaning. References to "connection" and "coupling" herein, unless otherwise specified, include both direct and indirect connections (couplings).

[0028] In the field of inspection, the workpiece to be tested needs to be inspected in multiple processing steps. Among them, the detection of the characteristic dimensions of the nanostructure (i.e., the features to be tested) in the workpiece to be tested is a necessary inspection process to ensure processing accuracy, and the inspection can be performed after each processing step.

[0029] In semiconductor manufacturing, the part to be tested generally refers to the wafer. In this case, the feature to be tested can refer to any structure in the processed wafer pattern that needs to be detected. The feature to be tested is not limited to a specific size. Since the current semiconductor processing accuracy is already 3nm (such as TSMC technology), the conventional processing accuracy is at the nanometer or micron level. Therefore, the feature to be tested in the wafer is generally called a nanostructure, or a micro-nanostructure.

[0030] The DUT can also refer to a display panel, such as one based on micro-LED technology. In this case, the nanostructure is the display unit inside the display panel or the structure inside the display unit. Of course, the DUT can also refer to other products that need to be tested.

[0031] Therefore, the present application does not limit the specific forms of the test piece and the nanostructure, and may also be products such as a glass panel and a mask.

[0032] In existing detection methods, in order to calculate the spectral response and performance of nanostructures, complex simulation algorithms such as the finite element method (FEM), finite-difference time-domain method (FDTD) and rigorous coupled wave analysis (RCWA) can be used to simulate the spectrum. These optical characteristic model algorithms require a long calculation time.

[0033] RCWA is an algorithm based on electromagnetic field theory that can accurately calculate the diffraction efficiency of periodic grating structures. To extract the parameters of a nanostructure, the optical property model established by RCWA is first used to calculate the simulated spectrum of the nanostructure. The measured spectrum is then fitted to the simulated spectrum in real time using methods such as the Levenberg-Marquarelt (LM) nonlinear regression algorithm to obtain the parameter value corresponding to the best-fit simulated spectrum.

[0034] However, since fitting algorithms such as the LM algorithm need to call optical characteristic models such as RCWA to calculate the simulated spectrum of the nanostructure in each iteration process, calculations based on the optical characteristic models require a lot of time, making its parameter extraction process very time-consuming and difficult to meet the needs of online measurement.

[0035] To address the slow real-time fitting speed, the industry has proposed a library matching method for extracting the parameters to be measured. This method first uses an established optical property model to pre-build a spectral library of the characteristic dimensions of the feature to be measured. The measured spectrum is then matched with the simulated spectra in the spectral library to find the simulated spectrum that best matches the measured spectrum. The parameter value corresponding to this simulated spectrum is then used as the parameter value of the nanostructure to be measured. However, when the nanostructure has a large number of parameters to be measured, the amount of library data required for construction increases dramatically, occupying a large amount of storage space. Furthermore, the library matching search process also takes up a lot of time, making it difficult to meet the needs of online measurement.

[0036] This application proposes a neural network-based method for extracting characteristic size parameters. This method, using a neural network model, can rapidly obtain characteristic dimensions and their corresponding ideal simulated spectra, meeting the high-precision and high-yield requirements of online measurement. This method can quickly and efficiently process large amounts of data and exhibits good generalization performance.

[0037] As shown in FIG1 , an embodiment of the present application provides a method for extracting characteristic size parameters based on a neural network, which may include:

[0038] Step 1: Provide multiple characteristic dimensions of the feature to be measured and the corresponding simulated spectra. This can be done by using data from a pre-matched spectral library or by performing a one-to-one matching using an optical property model to construct a spectral library that matches the current device under test and the measurement equipment. It's understandable that if the wafer being tested is a batch product, a one-to-one matching will be required when initially constructing the spectral library, but subsequent tests can continue to use the pre-matched spectral library.

[0039] In some embodiments, step 1 above may include:

[0040] Step 110 : Determine measurement parameters of the measuring device according to a theoretical range of characteristic dimensions of the feature to be measured; the characteristic dimensions may include at least two parameters to be measured.

[0041] For example, characteristic dimensions can include at least two of the following parameters: diameter (or width), length, and thickness; that is, there are multiple characteristic dimensions. Spectral characteristics can include at least one of reflectance, ellipsometric parameters, Mueller matrix, central wavelength, peak amplitude, half-width, Bayesian peak area, and wavelength shift. Based on the product's machining accuracy and the designed process, the theoretical range of any characteristic dimension of the characteristic to be measured can be determined, and appropriate measurement equipment and measurement parameters can be selected based on this theoretical range.

[0042] For example, the measurement device can be an ellipsometer or some other measurement device based on optical scattering measurement methods, such as a UV-visible absorption spectrometer, a laser Raman spectrometer, a Fourier transform infrared spectrometer, or a radial diffractometer. The measurement parameters of the measurement device can include the measurement wavelength, the angle of incidence, and the azimuth angle. These measurement parameters are used to provide parameters for the optical characteristic model simulation.

[0043] Step 120: construct an optical characteristic model based on the measured parameters, and use the optical characteristic model to simulate and obtain corresponding simulated spectra based on characteristic sizes within the theoretical range; wherein each set of characteristic sizes corresponds to a simulated spectrum, and all simulated spectra constitute a spectrum library.

[0044] In some embodiments, in step 120, constructing the optical characteristic model based on the measurement parameters may include:

[0045] The optical characteristic model is constructed based on the measured parameters and a numerical solution algorithm. In this step, how to construct the optical characteristic model can refer to existing implementation methods and is not limited by this application.

[0046] For example, the numerical solution algorithm may be a rigorous coupled wave analysis algorithm, a finite element method, a finite difference time domain method, or a finite difference frequency domain method.

[0047] In the simulation calculation of nanostructure spectra, there are many numerical simulation methods to choose from, such as RCWA, FEM, FDTD, FDFD, etc.

[0048] Rigorous Coupled Wave Analysis (RCWA), a Fourier-based rigorous coupled wave analysis method, is a very common method for calculating the reflection and transmission spectra of periodic optical structures. This method is capable of simulating nanostructures with complex periodic morphologies and is widely used in semiconductor devices, solar cells, nanobiosensors, and other fields.

[0049] FEM (Finite Element Method) and FDTD (Finite Difference Time-Domain Method) are numerical methods for solving electromagnetic wave equations. Both are used in solving spectral problems. Generally speaking, FEM is more suitable for structures with complex shapes, while FDTD is more suitable for time-domain problems.

[0050] Compared to RCWA methods, FEM and FDTD methods generally require higher computational resources. Therefore, when dealing with large or complex structures, runtime and computational complexity may be increased. Furthermore, the selection of a specific numerical simulation method also requires consideration of the algorithm's accuracy and stability. Therefore, selecting the appropriate numerical simulation method requires consideration of multiple factors, including model accuracy, solution time, and computational resources.

[0051] In the above step 120, an optical characteristic model is constructed by using a suitable numerical solution algorithm, and the characteristic to be measured is simulated to obtain an accurate simulation spectrum.

[0052] For example, the theoretical range of a parameter to be measured in the characteristic size is 9nm-15nm. The data in this range is simulated to obtain the corresponding spectrum library.

[0053] Step 2: Use the characteristic size and the simulated spectrum as data sets for the neural network model, and train the neural network model until the training error of the neural network model is less than a preset threshold, and determine that the neural network model has completed training.

[0054] In some embodiments, the neural network model may be a feedforward neural network model, a recursive neural network model, or a convolutional neural network model. The loss function of the neural network model may be a mean square error loss function, a mean absolute error loss function, or a divergence loss function.

[0055] Feedforward Neural Networks (FNNs) are one of the most basic neural network models, consisting of an input layer, several hidden layers, and an output layer. FNNs can automatically identify nonlinear mapping relationships between data sets and make predictions by learning from a given set of sample data. The advantage of FNNs is that they can be trained and optimized end-to-end, facilitating batch processing and prediction.

[0056] Recurrent Neural Networks (RNNs): RNNs are primarily used to process sequential data, such as time series or text data. Neurons in an RNN receive not only the current input but also the previous output, preserving previous state information. For semiconductor nanostructure spectral prediction, RNN variants such as LSTM (Long Short-Term Memory) can be used.

[0057] Convolutional Neural Networks (CNNs): CNNs are primarily used to process spatially structured data, such as images and audio. Through convolution and pooling operations, CNNs automatically extract features at different scales and utilize fully connected layers for final prediction. In the task of semiconductor nanostructure spectral prediction, spectral data can be treated as a two-dimensional image, and convolutional neural networks are used for feature extraction and prediction.

[0058] When training a neural network, it's important to choose an appropriate loss function to measure the difference between predicted and actual results, allowing the backpropagation algorithm to optimize the neural network's weight parameters and bias. For nanostructure characterization problems, the following loss function can be used to train the neural network.

[0059] The mean squared error (MSE) loss function is commonly used in regression problems to calculate the squared error between the predicted and actual values. For this problem, the output of the neural network is compared with the actual spectral reflectance calculated by the RCWA model to assess the error between the predicted and actual values.

[0060] Mean Absolute Error (MAE): This loss function is also commonly used in regression problems and calculates the absolute error between the predicted value and the actual value. Compared to MSE, MAE pays more attention to the prediction accuracy of small errors and gives less weight to large errors.

[0061] Kullback-Leibler Divergence (KL): This loss function is commonly used in classification and probability distribution fitting problems to measure the distance and difference between two probability distributions. For this problem, KL divergence can measure the difference between the predicted spectral distribution and the actual distribution and can be used to optimize the prediction accuracy of the neural network.

[0062] Therefore, the embodiments of the present application do not limit the specific neural network model and the loss function used for training, and can be adjusted according to the actual training results.

[0063] In some embodiments, the simulated spectrum may include spectral features in multiple dimensions; in step 2 above, training the neural network model may include:

[0064] Step 210: Perform dimensionality reduction processing on the multiple spectral features to screen out spectral features whose eigenvalues ​​are greater than a preset condition.

[0065] In some embodiments, in step 210, performing dimensionality reduction processing on the plurality of spectral features may include:

[0066] Step 211: Use principal component analysis or linear discriminant analysis to perform dimensionality reduction processing on multiple spectral features.

[0067] Dimensionality reduction can effectively improve the calculation speed, while reducing the dimension and complexity of the data, making it easier to analyze and process the subsequent data.

[0068] In the spectroscopic characterization of semiconductor nanostructures, dimensionality reduction methods include principal component analysis (PCA) and linear discriminant analysis (LDA). These methods linearly combine multiple highly correlated measurement parameters to generate a set of uncorrelated new variables, called principal components. By selecting an appropriate number of principal components, the dimensionality of the original data can be reduced to a lower dimension, thereby reducing computational effort and optimizing training.

[0069] When using neural networks for spectral prediction, the principal components after dimensionality reduction can be used as input features to achieve more efficient calculations. At the same time, using dimensionality reduction methods can also reduce the interference of noise and redundant information, improving the accuracy and reliability of measurements.

[0070] In some embodiments, in step 210, screening out spectral features having characteristic values ​​greater than a preset condition may include:

[0071] Step 212: When principal component analysis is used, spectral features with a variance resolution greater than a first preset threshold are selected; when linear discriminant analysis is used, spectral features with a data distribution discrimination greater than a second preset threshold are selected. The preset threshold is adjusted based on actual calculation conditions and the training results of the neural network model.

[0072] Identifying the feature with the largest eigenvalue can be accomplished through dimensionality reduction methods such as principal component analysis (PCA) or linear discriminant analysis (LDA). In these methods, the eigenvalue represents the variance of the principal components and serves as an indicator of each principal component's contribution to the explanation of the original data. Principal components with larger eigenvalues ​​contribute more to the explanation of the data.

[0073] In PCA, methods such as singular value decomposition (SVD) or eigenvector decomposition (EVD) can be used to calculate the principal components and corresponding eigenvalues ​​of the data. During the calculation process, the resulting eigenvalues ​​can be sorted by size, and the principal component with the largest eigenvalue is selected as the representation of the primary information in the data. Alternatively, the eigenvectors of the covariance matrix can be calculated. The covariance matrix describes the relationships between different variables and reflects the variance and correlation of the data. The eigenvectors and eigenvalues ​​of the covariance matrix are key to identifying the principal components. The eigenvalues ​​represent the variance of each principal component, while the eigenvectors indicate the direction of the corresponding principal component. By calculating the eigenvalues ​​and eigenvectors of the covariance matrix, the principal components of the data and their corresponding variance contributions can be obtained. In principal component analysis, principal components with higher variance contributions are generally selected to describe the primary information in the data.

[0074] In LDA, you can use the between-class scattering matrix and the within-class scattering matrix to calculate the linear discriminant direction and the corresponding eigenvalue. The within-class scattering matrix measures the differences within the data, while the between-class scattering matrix measures the differences between the data. By calculating the ratio of the within-class scattering matrix to the between-class scattering matrix, you can obtain the eigenvalue of the linear discriminant analysis. The linear discriminant direction with the largest eigenvalue corresponds to the most discriminative data distribution.

[0075] Step 3: Use a measuring device to measure the measurement spectrum of the feature to be measured.

[0076] For example, in an embodiment of the present application, an ellipsometer may be used for measurement to obtain an actual measured spectrum for subsequent fitting.

[0077] Step 4: Predict the characteristic size of the feature to be measured based on the measured spectrum and the trained neural network model to obtain a predicted size range of the characteristic size.

[0078] For example, based on a trained neural network model, the output can be predicted to obtain a predicted range of the output. For example, if the characteristic dimension of the feature to be measured is its thickness and the spectral characteristic is reflectivity, and the input (characteristic dimension) of the neural network model is 10nm-12nm, and the reflectivity range corresponding to the spectral characteristics of the theoretical simulation spectrum is 1.1-1.3, then when the reflectivity of the measured spectrum is 1.2, it can be predicted that the predicted size range of the characteristic dimension will fall within 10nm-12nm. In other words, the neural network model can be used to determine the predicted size range of the characteristic dimension corresponding to the measured spectrum.

[0079] Step 5: Fit the measured spectrum to the corresponding simulated spectrum within the predicted size range to obtain the simulated spectrum that best fits the measured spectrum, and extract the theoretical characteristic size corresponding to the best-fitting simulated spectrum as the final characteristic size.

[0080] In the above step 5, based on the predicted size range of the characteristic size, a characteristic size can be input into the neural network model starting from the middle value or the maximum value of this range to obtain a corresponding simulation spectrum, and the measured spectrum is fitted with this simulation spectrum. The input amount is adjusted according to the measurement results until the best fitting simulation spectrum is obtained.

[0081] For example, when the predicted size range of the feature size is 10nm-12nm, 11nm is input into the neural network model to obtain the simulation spectrum corresponding to 11nm, and the input amount is adjusted according to the fitting result. It can be adjusted to 11±0.1nm until the input is adjusted to the best fit.

[0082] In some embodiments, in step 5 above, fitting the measured spectrum to the corresponding simulated spectrum within the predicted size range may include:

[0083] A curve fitting algorithm is used to fit the measured spectrum to the corresponding simulated spectrum within the predicted size range.

[0084] For example, the curve fitting algorithm may be a nonlinear regression algorithm, a least squares algorithm, a polynomial fitting algorithm, or a kernel function regression algorithm.

[0085] For example, during the fitting process, the nonlinear regression algorithm (LM algorithm) is an iterative optimization algorithm. Each iteration requires calculating the model's predictions based on the current parameters and evaluating the loss function, thereby updating the parameters. The RCWA optical model is relatively time-consuming to calculate, so the LM algorithm needs to call RCWA to calculate the simulated spectrum at each iteration, which may result in high computational costs in practical applications. However, in the embodiments of the present application, the LM algorithm calls a neural network model for calculations at each iteration, which is fast and meets the requirements of online measurement.

[0086] In summary, the method provided in the embodiments of the present application can solve the problems of slow speed, complex calculation, and large storage space occupation in the inverse problem solving process of characteristic size measurement of nanostructures. The present application proposes a characteristic size parameter extraction method based on a neural network, which can quickly and accurately extract nanostructure parameters to meet the high precision and high yield requirements of online measurement. Compared with traditional nonlinear fitting and library matching methods, the method proposed in the present application has significant advantages in real-time fitting speed and computational efficiency. At the same time, the method can process large amounts of data and has good generalization performance. Therefore, the present application has significant beneficial effects in improving the characteristic size measurement accuracy and efficiency of nanostructures, while also reducing the large amount of storage space required for the library matching method.

[0087] As shown in FIG2 , a specific embodiment is provided below for further explanation.

[0088] The first step is to determine the range of values ​​for each parameter within the characteristic dimensions of the feature to be measured. This includes determining the range of values ​​for the measurement equipment's measurement parameters, such as the wavelength, angle of incidence, and azimuth. Using RCWA, an optical property model is established to calculate simulated spectra for each combination of the characteristic dimensions of the feature to be measured, and a spectral library corresponding to the characteristic dimensions of the feature to be measured is established. The simulated spectra can be represented using reflectivity, ellipsometric parameters, and the Mueller matrix.

[0089] In the second step, the established spectral library is used as the data set of the neural network. Based on the characteristic size parameters of the features to be measured and the characteristic distribution of the simulated spectrum, a fully connected neural network model that can map nonlinear functions is established, and the mean square error (MSE) is used as the loss function for training.

[0090] Before training, the dataset needs to be preprocessed, performing data cleaning and normalization to remove any outliers and ensure that the data is within a consistent range. Due to the large dimensionality of spectral features, principal component analysis (PCA) can be used to transform the dataset, compressing the spectral features into a lower feature space without losing too much information. By calculating the covariance matrix of the spectral matrix and then obtaining the eigenvalues ​​and eigenvectors of the covariance matrix, a matrix consisting of the eigenvectors corresponding to the k features with the largest eigenvalues ​​(i.e., the largest variance) is selected. The spectral matrix is ​​converted to the new space to achieve dimensionality reduction of the spectral features. The number of spectral data features is determined by the variance interpretation. Generally, k features with a variance interpretation of (1-10-7)*100% are selected to prevent large errors in the data after dimensionality reduction.

[0091] By employing an exhaustive grid search method to fine-tune hyperparameters such as the number of neurons, activation function, and optimizer, and comparing the mean square error (MSE) of different neural network models, model performance can be optimized. When the dataset is small or the prediction error is large, the neural network-based feature size parameter extraction method proposed in this application can also be trained and predicted using deep learning network models such as CNN (convolutional neural network) or RNN (recurrent neural network) to obtain a preset size range for the feature size corresponding to the current measured spectrum. These deep learning network models have the ability to process complex data and better capture the relationship between the measured parameters and spectral data, thereby improving model accuracy and reliability. For example, a CNN model can gradually extract local features from spectral data through convolutional and pooling layers, and then combine these features in fully connected layers for tasks such as classification or regression. An RNN model can take into account time series information and utilize a recurrent structure to model spectral data, preserving data correlations and thus enabling better prediction of spectral data.

[0092] The third step is to first use the trained neural network model to predict the spectrum of the characteristic size of the feature to be measured, and after obtaining the prediction result, use it as the spectral data of the LM algorithm iteration for nonlinear fitting to extract the parameters to be measured. The LM algorithm is a commonly used nonlinear least squares optimization algorithm with high accuracy and robustness when estimating parameters. Using the LM algorithm as the core algorithm for inverse problem solving can quickly and accurately extract the parameters to be measured. When the parameters to be measured of the nanostructure are large, the traditional library matching method needs to establish a large amount of spectral data, which takes up a lot of storage space, and the search time will also increase, making it difficult to meet the needs of online measurement. In contrast, the neural network model takes up less memory, and the speed of using the neural network to predict a set of spectra is faster, usually completing in about 1ms. Therefore, in the characteristic size parameter extraction method proposed in the present application, the method of using the neural network prediction can significantly improve the computational efficiency and accuracy, thereby meeting the requirements of online measurement.

[0093] In summary, through the above method, when using scattering measurement equipment such as ellipsometers to detect semiconductor nanostructures, selecting appropriate optical feature models and fitting algorithms can obtain more accurate simulated spectra and fitted structures. For example, the RCWA algorithm is used to establish a spectral library of the nanostructure to be measured as a neural network data set; a neural network model is constructed with the characteristic dimensions of the feature to be measured as the mapping input and the theoretical simulated spectrum as the output, establishing a nonlinear relationship between the nanostructure parameters and the spectral data. A dimensionality reduction method based on principal component analysis is used to compress the spectral features, perform parameter tuning, and complete rapid training of the neural network. The trained neural network model is used to predict the simulated spectrum of the parameter to be measured, and the spectrum used as the LM algorithm iteration is nonlinearly fitted. The parameter to be measured is extracted from the best-fit simulated spectrum. This method can achieve fast and accurate measurement and meet the high-precision and high-yield requirements of online measurement.

[0094] For example, the RCWA algorithm is computationally intensive, especially for large structures or high-resolution calculations. This leads to very high computational time and memory requirements, making online computation difficult. Therefore, replacing the RCWA algorithm with a pre-trained neural network model, such as the CNN+LM algorithm, can significantly reduce computational time and memory requirements, enabling faster online computation.

[0095] The above is a description of the characteristic size parameter extraction method based on a neural network provided in an embodiment of the present application. The following describes a characteristic size parameter extraction system based on a neural network provided in an embodiment of the present application.

[0096] As shown in FIG3 , an embodiment of the present application provides a neural network-based feature size parameter extraction system for implementing the method described in the above embodiment. The system may include:

[0097] The spectrum library module 10 is used to provide multiple characteristic sizes of the features to be measured and simulated spectra corresponding to the characteristic sizes.

[0098] The measuring device 20 is used to measure the spectrum of the characteristic to be measured to obtain a measured spectrum.

[0099] The neural network module 30 is used to predict the characteristic size of the feature to be measured based on the measured spectrum and the trained neural network model to obtain a predicted size range of the characteristic size.

[0100] The fitting module 40 fits the measured spectrum with the corresponding simulated spectrum within the predicted size range to obtain a simulated spectrum that best fits the measured spectrum, and extracts the theoretical characteristic size corresponding to the best-fitting simulated spectrum as the final characteristic size.

[0101] The spectral library module 10, neural network module, and fitting module 40 can be implemented using one or more processors. The measurement device can be an ellipsometer or some measurement device based on optical scattering measurement, such as a UV-visible absorption spectrometer, a laser Raman spectrometer, a Fourier transform infrared spectrometer, or a radial diffractometer.

[0102] In the above system, the neural network module 30 is also used to use the characteristic size and the simulated spectrum as the data set of the neural network model to train the neural network model until the training error of the neural network model is less than a preset threshold, and it is determined that the neural network model has completed training.

[0103] The above system is used to implement the above characteristic size parameter extraction method, and has corresponding technical effects, which will not be repeated here.

[0104] The characteristic size parameter extraction method provided in this application is implemented by a terminal device, which may include a memory and a processor. For example, the terminal device may be a computer, a server, or other device with computing and data processing capabilities.

[0105] The memory is used to store a program. The processor is used to implement the feature size parameter extraction method described in the above embodiment by executing the program stored in the memory.

[0106] Those skilled in the art will appreciate that all or part of the functions of the various methods in the above embodiments can be implemented by hardware or by computer program. When all or part of the functions in the above embodiments are implemented by computer program, the program can be stored in a computer-readable storage medium, and the storage medium can include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to implement the above functions. For example, the program is stored in the memory of the device, and when the program in the memory is executed by the processor, all or part of the above functions can be implemented. In addition, when all or part of the functions in the above embodiments are implemented by computer program, the program can also be stored in a storage medium such as a server, another computer, disk, optical disk, flash disk or mobile hard disk, and saved in the memory of the local device by downloading or copying, or the system of the local device is updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be implemented.

[0107] This document is described with reference to various exemplary embodiments. However, those skilled in the art will recognize that changes and modifications may be made to the exemplary embodiments without departing from the scope of this document. For example, the various operational steps and components used to perform the operational steps may be implemented in different ways (e.g., one or more steps may be deleted, modified, or incorporated into other steps) depending on the specific application or any number of cost functions associated with the operation of the system.

[0108] Although the principles of this invention have been shown in various embodiments, many modifications of structure, arrangement, proportion, elements, materials and components that are particularly suitable for specific environments and operational requirements can be used without departing from the principles and scope of this invention. The above modifications and other changes or amendments are intended to be included within the scope of this invention.

[0109] The foregoing detailed description has been described with reference to various embodiments. However, those skilled in the art will recognize that various modifications and changes can be made without departing from the scope of this disclosure. Therefore, the present disclosure will be considered in an illustrative rather than a restrictive sense, and all such modifications will be included within its scope. Similarly, the advantages, other advantages and solutions to the problems of the various embodiments have been described above. However, the benefits, advantages, solutions to the problems and any elements that can produce these, or make them more specific, should not be interpreted as critical, required or necessary. The term "comprising" and any other variants used in this article are all non-exclusive inclusions, so that a process, method, article or device that includes a list of elements includes not only these elements, but also other elements that are not explicitly listed or do not belong to the process, method, system, article or device. In addition, the term "coupled" and any other variants used in this article refer to physical connections, electrical connections, magnetic connections, optical connections, communication connections, functional connections and / or any other connections.

[0110] Those skilled in the art will recognize that many changes can be made to the details of the above embodiments without departing from the basic principles of the invention. Therefore, the scope of the present invention should be determined solely by the claims.

Claims

1. A method for extracting characteristic size parameters based on neural network, characterized in that: include: Providing multiple characteristic dimensions of the characteristic to be measured and the simulation spectrum corresponding to the characteristic dimensions; Using the characteristic size and the simulated spectrum as data sets of a neural network model, training the neural network model until a training error of the neural network model is less than a preset threshold, and determining that the training of the neural network model is completed; Measuring the measurement spectrum of the feature to be measured using a measuring device; Predicting the characteristic size of the feature to be measured according to the measured spectrum and the trained neural network model to obtain a predicted size range of the characteristic size; The measured spectrum is fitted with the simulation spectrum corresponding to the predicted size range to obtain a simulation spectrum that best fits the measured spectrum, and a theoretical characteristic size corresponding to the simulation spectrum that best fits is extracted as the final characteristic size.

2. The method according to claim 1, characterized in that The providing of a plurality of characteristic sizes of the feature to be measured and the simulated spectra corresponding to the characteristic sizes comprises: Determine the measurement parameters of the measuring device according to the theoretical range of the characteristic size of the feature to be measured; the characteristic size includes at least two parameters to be measured; An optical characteristic model is constructed according to the measurement parameters, and the corresponding simulated spectrum is obtained by simulating using the optical characteristic model according to the characteristic size within the theoretical range; wherein each group of the characteristic sizes corresponds to a simulated spectrum, and all the simulated spectra constitute a spectrum library.

3. The method according to claim 2, characterized in that The step of constructing an optical characteristic model according to the measurement parameters comprises: An optical characteristic model is constructed according to the measurement parameters and using a numerical solution algorithm.

4. The method according to claim 3, characterized in that The numerical solution algorithm is a rigorous coupled wave analysis algorithm, a finite element method, a finite difference time domain method or a finite difference frequency domain method.

5. The method according to claim 1, characterized in that The fitting of the measured spectrum to the corresponding simulated spectrum within the predicted size range comprises: A curve fitting algorithm is used to fit the measured spectrum to the corresponding simulated spectrum within the predicted size range.

6. The method according to claim 5, characterized in that The curve fitting algorithm is a nonlinear regression algorithm, a least squares algorithm, a polynomial fitting algorithm or a kernel function regression algorithm.

7. The method according to claim 1, characterized in that The neural network model is a feedforward neural network model, a recursive neural network model or a convolutional neural network model.

8. The method according to claim 7, characterized in that The loss function of the neural network model is a mean square error loss function, a mean absolute error loss function or a divergence loss function.

9. The method according to claim 1, characterized in that The simulated spectrum includes spectral features in multiple dimensions; The training of the neural network model comprises: A dimension reduction process is performed on the plurality of spectral features to screen out the spectral features whose characteristic values ​​are greater than a preset condition.

10. The method according to claim 9, characterized in that The performing dimensionality reduction processing on the plurality of spectral features comprises: The principal component analysis method or the linear discriminant analysis method is used to perform dimension reduction processing on the plurality of spectral features.

11. The method according to claim 10, characterized in that The step of screening out the spectral features whose characteristic values ​​are greater than a preset condition comprises: When the principal component analysis method is used, the spectral features having a variance resolution greater than a first preset threshold are selected; When the linear discriminant analysis method is adopted, the spectral feature whose data distribution discrimination is greater than the second preset threshold is selected.

12. The method according to claim 9, characterized in that The characteristic size includes at least two of the diameter, length and thickness of the feature to be measured; The spectral characteristics include at least one of reflectivity, ellipsometric parameters, Mueller matrix, central wavelength, peak amplitude, half-width, Bayesian peak area and wavelength shift.

13. A characteristic size parameter extraction system based on neural network, characterized in that: include: A spectrum library module, used to provide multiple characteristic sizes of the characteristics to be measured and simulated spectra corresponding to the characteristic sizes; A measuring device, used to measure the spectrum of the feature to be measured to obtain a measured spectrum; A neural network module, used for predicting the characteristic size of the feature to be measured according to the measured spectrum and the trained neural network model, and obtaining a predicted size range of the characteristic size; The fitting module fits the measured spectrum with the corresponding simulated spectrum within the predicted size range to obtain a simulated spectrum that best fits the measured spectrum, and extracts a theoretical characteristic size corresponding to the best-fitting simulated spectrum as a final characteristic size.

14. A terminal device, characterized in that: include: Memory, used to store programs; A processor, configured to implement the method according to any one of claims 1 to 12 by executing the program stored in the memory.

15. A computer-readable storage medium, characterized in that: The medium stores a program, which can be executed by a processor to implement the method according to any one of claims 1 to 12.

Images