Method of analyzing a semiconductor structure, analysis system, medium and electronic device

Abstract

The present disclosure relates to a semiconductor structure analysis method, an analysis system, a medium and an electronic device. The method comprises the following steps: establishing a database storing the spectrum and structure parameters of an experimental design wafer and the spectrum and structure parameters of a process record wafer; generating a data processing model based on the spectrum and structure parameters in the database; and analyzing the spectrum of an online wafer by using the data processing model. The spectrum of the online wafer to be detected is directly imported into the data processing model, and the prediction result can be obtained, thereby improving the work efficiency. The spectrum of the online wafer is also saved in the database during the analysis process, the data processing model is continuously updated, the accuracy of the data processing model is improved, and the spectrum of the online wafer that has been analyzed can be reused when detecting other online wafers, thereby improving the accuracy of the prediction result.

Description

Technical Field

[0001] This disclosure relates to the field of semiconductor technology, and more specifically, to a method, system, medium, and electronic device for analyzing semiconductor structures. Background Technology

[0002] As semiconductor integrated circuit manufacturing processes advance and critical dimensions shrink, monitoring the various semiconductor structures becomes increasingly difficult. Optical Critical Dimension (OCD) is a highly effective and precise measurement method that utilizes the principle of light scattering to analyze the spectrum of reflected light projected onto a fixed area on the wafer surface, thereby enabling the measurement of the size or contour of the tested pattern structure.

[0003] In online wafer inspection, the spectra of wafers recorded in the experimental designs and processes used during the initial database creation are typically retrieved manually and compared with the spectra of online wafers. This process is time-consuming and labor-intensive, resulting in poor efficiency. Furthermore, the analysis is largely based on the experience and judgment of engineers, leading to low accuracy.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this disclosure is to overcome the problems of poor efficiency and low accuracy of existing online wafer analysis methods, and to provide a method, system, medium and electronic device for analyzing semiconductor structures.

[0006] According to one aspect of this disclosure, a method for analyzing semiconductor structures is provided, comprising:

[0007] Establish a database storing spectral and structural parameters, including those of experimentally designed wafers and those of process-recorded wafers;

[0008] Data processing models are generated based on spectral and structural parameters;

[0009] The data processing model is used to analyze the spectrum of online wafers and provide prediction results, while the spectrum of online wafers is saved in the database.

[0010] In one embodiment of this disclosure, analyzing an online wafer based on a data processing model and providing prediction results includes: obtaining the current layer spectrum of the online wafer; obtaining the previous layer spectrum and previous layer structural parameters of the online wafer; comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structural parameters in the data processing model to obtain the current layer structural parameters; and calculating the matching degree between each of the current layer structural parameters and the current layer spectrum.

[0011] In one embodiment of this disclosure, the current layer structure parameters include current layer floating parameters and current layer fixed parameters; comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structure parameters in the data processing model to obtain the current layer structure parameters, includes: calculating the sensitivity of the previous layer structure parameters at each wavelength point on the current layer spectrum; fixing the previous layer structure parameters with sensitivity greater than a first preset value to the data processing model as current layer fixed parameters; individually floating the previous layer structure parameters with sensitivity less than the first preset value to obtain current layer floating parameters; and excluding the previous layer structure parameters with sensitivity less than a second preset value.

[0012] In one embodiment of this disclosure, the front layer structural parameters include the front layer critical dimension; calculating the sensitivity of the front layer structural parameters at each wavelength point in the current layer spectrum includes: analyzing the front layer spectrum and the current layer spectrum in a data processing model to obtain the sensitivity of the front layer critical dimension at each wavelength point in the current layer spectrum.

[0013] In one embodiment of this disclosure, the front-layer structural parameters with sensitivity less than a first preset value are individually floated to obtain the current-layer floating parameters, including: determining the corresponding wavelength point of the front-layer structural parameters with sensitivity less than the first preset value on the current-layer spectrum; individually floating the front-layer structural parameters with sensitivity less than the first preset value and calculating the sensitivity of the individually floated front-layer structural parameters at the corresponding wavelength point; and using the individually floated front-layer structural parameters with sensitivity greater than the first preset value as the current-layer floating parameters.

[0014] In one embodiment of this disclosure, the analysis of online wafers based on the data processing model and the provision of prediction results further include: analyzing the matching degree to obtain a matching degree score, and outputting a list of matching degree scores as the prediction result.

[0015] In one embodiment of this disclosure, a cosine similarity algorithm is used to analyze the matching degree.

[0016] In one embodiment of this disclosure, establishing a database storing spectral and structural parameters includes: collecting spectral and structural parameters of experimentally designed wafers and integrating them into a first data file; importing the first data file into a pre-established empty database and saving it to form an initial database.

[0017] In one embodiment of this disclosure, establishing a database storing spectral and structural parameters further includes: collecting spectral and parameter data of process record wafers and integrating a second data file; integrating the second data file into an initial database to form a database.

[0018] In one embodiment of this disclosure, establishing a database storing spectral and structural parameters further includes integrating a second data file into the database.

[0019] In one embodiment of this disclosure, generating a data processing model based on spectral and structural parameters includes: establishing an initial model; generating training samples based on the spectral and structural parameters of multiple wafers in a database; and training the initial model based on the training samples to obtain a data processing model.

[0020] In one embodiment of this disclosure, the method further includes: importing new spectral and structural parameters; integrating the new spectral and structural parameters with those stored in the database into new training samples; and training a data processing model based on the new training samples.

[0021] In one embodiment of this disclosure, the structural parameters include at least one of key feature dimensions, thin film stacking, or semiconductor structure profile.

[0022] In one embodiment of this disclosure, the spectrum is stored in a database as a DAT file or a txt file.

[0023] According to another aspect of this disclosure, a semiconductor structure analysis system is provided, comprising: a storage module configured to establish a database storing spectral and structural parameters, the spectral and structural parameters including those of an experimentally designed wafer and those of a process-recorded wafer; a modeling module configured to generate a data processing model based on the spectral and structural parameters; and an analysis module configured to analyze the online wafer based on the data processing model and provide prediction results, while simultaneously saving the spectrum of the online wafer in the database.

[0024] According to another aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the method as described in any one aspect of this disclosure.

[0025] According to another aspect of this disclosure, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to perform the method as described in any one aspect of this disclosure.

[0026] This disclosed method for analyzing semiconductor structures establishes a database storing the spectral and structural parameters of experimentally designed wafers and process-recorded wafers. A data processing model is generated based on the spectral and structural parameters in the database, and this model is used to analyze the spectra of online wafers. The spectra of the online wafer to be tested can be directly imported into the data processing model to obtain prediction results, resulting in high efficiency. The analysis process also saves the spectra of the online wafers in the database, continuously updating the data processing model to improve its accuracy. When testing other online wafers, the spectra of previously analyzed online wafers can be reused, thereby improving the accuracy of the prediction results.

[0027] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0028] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0029] Figure 1 This is a flowchart of the database maintenance methods involved in the relevant technologies.

[0030] Figure 2 This is a flowchart of database optimization methods involved in related technologies.

[0031] Figure 3 This is a flowchart of a semiconductor structure analysis method according to an embodiment of this disclosure.

[0032] Figure 4 This is a top view of the front-end semiconductor structure involved in an embodiment of this disclosure.

[0033] Figure 5 This is a perspective view of the front-end semiconductor structure involved in an embodiment of this disclosure.

[0034] Figure 6 This is a top view of the current semiconductor structure involved in the embodiments of this disclosure.

[0035] Figure 7 This is a perspective view of the current semiconductor structure involved in an embodiment of this disclosure.

[0036] Figure 8 This is another perspective view of the current semiconductor structure involved in the embodiments of this disclosure.

[0037] Figure 9This is a spectral simulation fitting graph corresponding to the reference values ​​of the floating parameters involved in the embodiments of this disclosure.

[0038] Figure 10 This is a spectral simulation fitting graph corresponding to the reference value of the first floating parameter involved in the embodiments of this disclosure.

[0039] Figure 11 The spectral simulation fitting graph corresponds to the reference value of the second floating parameter involved in the embodiments of this disclosure.

[0040] Figure 12 This is a spectral simulation fitting graph corresponding to the reference value of the third floating parameter involved in the embodiments of this disclosure.

[0041] Figure 13 This is a spectral simulation fitting graph corresponding to the reference value of the fourth floating parameter involved in the embodiments of this disclosure.

[0042] Figure 14 This is a spectral simulation fitting graph corresponding to the reference value of the fifth floating parameter involved in the embodiments of this disclosure.

[0043] Figure 15 This is a spectral simulation fitting graph corresponding to the reference value of the sixth floating parameter involved in the embodiments of this disclosure.

[0044] Figure 16 This is a schematic diagram of the semiconductor structure analysis system according to an embodiment of the present disclosure.

[0045] Figure 17 This is a schematic diagram of the structure of an electronic device according to an embodiment of this disclosure. Detailed Implementation

[0046] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar structures, and therefore detailed descriptions of them will be omitted. Furthermore, the drawings are merely illustrative of this disclosure and are not necessarily drawn to scale.

[0047] Although relative terms such as "up" and "down" are used in this specification to describe the relative relationship of one component of an icon to another, these terms are used only for convenience, such as according to the orientation of the examples shown in the accompanying drawings. It is understood that if the device of the icon is flipped upside down, the component described as "up" will become the component described as "down." When a structure is "up" of another structure, it may mean that the structure is integrally formed on the other structure, or that the structure is "directly" mounted on the other structure, or that the structure is "indirectly" mounted on the other structure through another structure.

[0048] The terms “a,” “one,” “the,” “the,” and “at least one” are used to indicate the presence of one or more elements / components / etc.; the terms “including” and “having” are used to indicate an open-ended inclusion and to mean that there may be other elements / components / etc. in addition to the listed elements / components / etc.; the terms “first,” “second,” and “third,” etc., are used only as markers and are not a limitation on the number of objects.

[0049] In related technologies, such as Figure 1 As shown, database maintenance involves checking wafers with abnormal online measurement values ​​under stable process conditions. The maintenance process includes: Step S101, manually retrieving the spectra of wafers used for experimental design and process recording. Step S102, importing the spectra of wafers used for experimental design and process recording. Step S103, importing the online wafer spectra. Step S104, comparing and analyzing the spectra of the online wafers with those of the wafers used for experimental design and process recording.

[0050] Problems encountered: When performing online inspection, it's necessary to manually retrieve the spectra of wafers from the Design of Experiments (DOE) and verified Process Records (PORs) used during the initial database construction. The retrieval results are then imported into the data processing model and compared with the spectra of the online wafers for analysis. Therefore, this process is time-consuming and labor-intensive in practice. While the data processing model is generated based on a large amount of data and closely approximates the actual structure, current online wafer analysis largely relies on engineers' experience and judgment, lacking intelligent analysis and prediction methods to accelerate the efficiency of analysis and judgment.

[0051] like Figure 2As shown, the database optimization addresses the discrepancy between the current database and the online process records, rendering it unusable. The optimization process includes: Step S201, detecting the need for optimization of the existing database; Step S202, collecting the spectrum of the new wafer and its corresponding reference data; Step S203, acquiring the spectrum and reference data of wafers from the existing training samples; Step S204, merging the spectrum and reference data of the new wafer with those from the existing training samples; and Step S25, generating a new database based on the merged spectrum and reference data.

[0052] Problems encountered: During maintenance, it was discovered that the data processing model needed to be rebuilt. This model requires new spectra from Design of Experiments (DOE) wafers. However, OCD databases, similar to machine learning (ML) algorithms, have certain requirements regarding the amount of spectral data. Therefore, in the initial stages of development, the training samples need to be expanded in stages. The training samples include spectra and their corresponding structural parameters. Each training iteration must include all the spectra and structural parameters to be used, all within a single file. Database optimization requires a large amount of training data, and existing optimization methods are inefficient.

[0053] Based on this, the present disclosure provides a method for analyzing semiconductor structures. For example... Figure 3 As shown, the method includes:

[0054] Step S10: Establish a database storing spectral and structural parameters, including the spectral and structural parameters of the experimental design wafer and the spectral and structural parameters of the process record wafer.

[0055] Step S20: Generate a data processing model based on spectral and structural parameters.

[0056] Step S30: Analyze the spectrum of the online wafer based on the data processing model and give the prediction results, while saving the spectrum of the online wafer in the database.

[0057] The spectra of the online wafers to be inspected can be directly imported into the data processing model to obtain prediction results, resulting in high work efficiency. The analysis process also saves the spectra of the online wafers in the database, continuously updating the data processing model to improve its accuracy. When inspecting other online wafers, the spectra of previously analyzed online wafers can be reused, thereby improving the precision of the prediction results.

[0058] It should be noted that the wafer spectra in the database can correspond to the spectra of multiple semiconductor structures. The spectra can be those of regions with repeating periodic structures, and are stored in the database as DAT or txt files. The spectra can be saved in the database or exported and saved as a new database file. The structural parameters corresponding to the spectra include at least one of the following: key feature dimensions, thin film stacking, or semiconductor structure profile. These structural parameters can be obtained from scanned images of the wafer.

[0059] Establishing a database storing spectral and structural parameters involves: collecting the spectral and structural parameters of the experimentally designed wafers and integrating them into a first data file. The first data file is then imported into a pre-established empty database and saved to form an initial database. Next, spectral and parameter data from the process record wafers are collected and integrated into a second data file. This second data file is then integrated into the initial database to form the final database. If more spectral and structural parameters are needed, the second data file is integrated into the database, updating the database accordingly.

[0060] The data processing model is generated based on spectral and structural parameters, including: establishing an initial model; generating training samples based on the spectral and structural parameters of multiple wafers in the database; and training the initial model based on the training samples to obtain the data processing model. For the first online wafer analysis, the spectral and structural parameters of multiple wafers in the database include those of the experimentally designed wafer and those of the process-recorded wafer. For subsequent online wafer analyses, the spectral and structural parameters of multiple wafers in the database include not only those of the experimentally designed wafer and the process-recorded wafer, but also those of the previously analyzed online wafers.

[0061] When a discrepancy is detected between the existing database and the condition of the online wafers, the existing database becomes unusable and requires optimization. The optimization process includes: importing new spectral and structural parameters; integrating these new parameters with those stored in the database to create new training samples; and training a data processing model based on these new training samples. This ensures that the newly generated data processing model matches the condition of the online wafers.

[0062] The analysis and prediction results of online wafer fabrication based on the data processing model include:

[0063] The process involves acquiring the current layer spectrum of the online wafer; acquiring the previous layer spectrum and previous layer structural parameters of the online wafer; the previous layer structural parameters include the previous layer critical dimensions. The previous layer and current layer spectra are analyzed in a data processing model to obtain the sensitivity of the previous layer critical dimensions at each wavelength point in the current layer spectrum. The corresponding wavelength points in the current layer spectrum for previous layer structural parameters with sensitivities less than a first preset value are determined. These previous layer structural parameters with sensitivities less than the first preset value are individually floated, and the sensitivity of each individually floated previous layer structural parameter at the corresponding wavelength point is calculated. Previous layer structural parameters with sensitivities greater than the first preset value are used as current layer floating parameters. Previous layer structural parameters with sensitivities greater than the first preset value are fixed in the data processing model as current layer fixed parameters. Previous layer structural parameters in the current layer spectrum with sensitivities less than a second preset value are excluded. It is understood that obtaining current layer structural parameters in this way can reduce measurement time.

[0064] like Figure 4 and Figure 5 As shown, the front-layer structural parameters include the first front-layer key dimension a, the second front-layer key dimension b, the third front-layer key dimension c, the fourth front-layer key dimension d, the fifth front-layer key dimension e, the sixth front-layer key dimension f, the seventh front-layer key dimension g, the eighth front-layer key dimension h, and the ninth front-layer key dimension i. Among them, the sensitivity of the first front-layer key dimension a, the second front-layer key dimension b, the fourth front-layer key dimension d, the fifth front-layer key dimension e, and the sixth front-layer key dimension f at each wavelength point in the current layer spectrum is greater than a first preset value. Therefore, the first front-layer key dimension a, the second front-layer key dimension b, the fourth front-layer key dimension d, the fifth front-layer key dimension e, and the sixth front-layer key dimension f are fixed to the data processing model as fixed parameters of the current layer.

[0065] like Figures 6 to 8 As shown, the fixed parameters for each layer include a1, b1, c1, d1, and e1.

[0066] The sensitivity of the third, seventh, and eighth front-layer key dimensions c, g, and h at each wavelength point in the current layer spectrum is less than the first preset value. Therefore, it is necessary to determine the corresponding wavelength points of the third, seventh, and eighth front-layer key dimensions c, g, and h in the current layer spectrum. Individual floating of the third, seventh, and eighth front-layer key dimensions c, g, and h is performed to obtain the first, second, and third front-layer floating key dimensions. The sensitivity of the first, second, and third front-layer floating key dimensions at the corresponding wavelength points is calculated. The sensitivity of the first and second front-layer floating key dimensions is greater than the first preset value. Therefore, the first front-layer floating key dimension is used as the first current layer floating parameter f1, and the second front-layer floating key dimension g1 is used as the second current layer floating parameter.

[0067] The sensitivity of the ninth front layer critical dimension at each wavelength point in the current layer spectrum is less than the second preset value. Therefore, the ninth front layer critical dimension is excluded and is not used as a structural parameter of the current layer.

[0068] It should be noted that the first preset value can be 80%, and the second preset value can be 20%. This is only an illustrative example and is not a specific limitation on the first and second preset values. They can also be adjusted according to actual needs.

[0069] In addition to the first and second current layer floating parameters, a third current layer floating parameter h1, a fourth current layer floating parameter i1, a fifth current layer floating parameter j1, and a sixth current layer floating parameter k1 are added as current layer floating parameters.

[0070] The in-layer floating parameters are floated in the data processing model to match structural parameters with the same changing trend. Specifically, the first in-layer floating parameter f1 is floated in the data processing model to obtain three different first reference values ​​of 20 nm, 40 nm, and 60 nm; the second in-layer floating parameter g1 is floated in the data processing model to obtain three different second reference values ​​of 11 nm, 5.5 nm, and 0 nm; the third in-layer floating parameter h1 is floated in the data processing model to obtain three different third reference values ​​of 25 nm, 15 nm, and 5 nm; the fourth in-layer floating parameter i1 is floated in the data processing model to obtain three different fourth reference values ​​of 0 nm, 5 nm, and 15 nm; the fifth in-layer floating parameter j1 is floated in the data processing model to obtain three different fifth reference values ​​of 20 nm, 13 nm, and 5 nm; and the sixth in-layer floating parameter k1 is floated in the data processing model to obtain four different sixth reference values ​​of 120 nm, 135 nm, 150 nm, and 165 nm.

[0071] like Figure 9As shown, a machine learning algorithm is used to simulate and fit the spectrum 2 corresponding to the reference value of the floating parameter with the spectrum 1 of the current layer, thereby obtaining the matching degree of the first current layer floating parameter. Figures 10 to 15 As shown, a machine learning algorithm is used to simulate and fit the spectra 112, 122, and 132 corresponding to three different first reference values ​​with the three current layer spectra 111, 121, and 131, respectively, to obtain the matching degree of the first current layer floating parameters. Similarly, a machine learning algorithm is used to simulate and fit the spectra 212, 222, and 232 corresponding to three different second reference values ​​with the three current layer spectra 211, 221, and 231, respectively, to obtain the matching degree of the second current layer floating parameters. The same algorithm is used to simulate and fit the spectra 312, 322, and 332 corresponding to three different third reference values ​​with the three current layer spectra 311, 321, and 331, respectively, to obtain the matching degree of the third current layer floating parameters. Finally, a machine learning algorithm is used to simulate and fit the spectra 412, 422, and 432 corresponding to three different fourth reference values ​​with the three current layer spectra 411, 421, and 431, respectively, to obtain the matching degree of the fourth current layer floating parameters. Machine learning algorithms were used to simulate and fit the spectra 512, 522, and 532 corresponding to three different fifth reference values ​​with the three current layer spectra 511, 521, and 531, respectively, to obtain the matching degree of the fifth current layer floating parameter. Similarly, machine learning algorithms were used to simulate and fit the spectra 612, 622, 632, and 642 corresponding to four different sixth reference values ​​with the four current layer spectra 611, 621, 631, and 641, respectively, to obtain the matching degree of the sixth current layer floating parameter.

[0072] The matching degree was analyzed using a cosine similarity algorithm to obtain a matching degree score, and a list of matching degree scores was output as the prediction result. Table 1 shows the list of analysis conclusions for the matching degree.

[0073] Table 1. List of analysis conclusions on matching degree

[0074] Serial Number When the layer float parameter Match score 1 First layer floating parameter 0.82 2 Second layer floating parameter 0.35 3 Third layer floating parameter 0.41 4 Fourth layer floating parameters 0.32 5 Fifth layer floating parameters 0.22 6 Sixth layer floating parameters 0.18

[0075] It can be seen that the reference value of the first layer floating parameter f1 is relatively close to the layer structural parameters, while the reference values ​​of the second layer floating parameter g1, the third layer floating parameter h1, the fourth layer floating parameter i1, the fifth layer floating parameter j1, and the sixth layer floating parameter k1 differ significantly from the actual values ​​of the layer structural parameters. Therefore, it can be concluded that the layer structural characteristic parameter corresponding to the first layer floating parameter f1, such as at least one of the key feature dimensions, thin film stacking, or semiconductor structure profile, is the feature that substantially causes the difference between the current layer and the previous layer states of the in-line wafer.

[0076] like Figure 16As shown, this disclosure also provides a semiconductor structure analysis system 100 for performing the above-described analysis method. The analysis system may include: a storage module 110 configured to establish a database storing spectral and structural parameters, including the spectral and structural parameters of experimentally designed wafers and process-recorded wafers; a modeling module 120 configured to generate a data processing model based on the spectral and structural parameters; and an analysis module 130 configured to analyze the online wafers based on the data processing model and provide prediction results, while simultaneously saving the spectra of the online wafers in the database.

[0077] In one embodiment, the storage module may include a first collection unit and a first integration unit. The first collection unit is configured to collect the spectral and structural parameters of the experimental design wafer and integrate a first data file. The first integration unit is configured to import the first data file into a pre-established empty database and save it to form an initial database.

[0078] In one embodiment, the storage module may further include a second collection unit and a second integration unit, wherein the second collection unit is configured to collect spectral and parameter data of the process record wafer and integrate a second data file; and the second integration unit is configured to integrate the second data file into an initial database to form a database.

[0079] In one implementation, the second integration unit is further configured to integrate a second data file into the database.

[0080] In one implementation, the modeling module includes a modeling unit, a generation unit, and a training unit. The modeling unit is configured to establish an initial model; the generation unit is configured to generate training samples based on the spectral and structural parameters of multiple wafers in a database; and the training unit is configured to train the initial model based on the training samples to obtain a data processing model.

[0081] In one embodiment, the analysis module includes a first acquisition unit, a second acquisition unit, and a comparison unit. The first acquisition unit is configured to acquire the current layer spectrum of the online wafer; the second acquisition unit is configured to acquire the previous layer spectrum and previous layer structural parameters of the online wafer; the comparison unit is configured to compare the current layer spectrum and the previous layer spectrum, perform separate floating of the previous layer structural parameters in the data processing model to obtain the current layer structural parameters, and calculate the matching degree between each of the current layer structural parameters and the current layer spectrum.

[0082] In one embodiment, the current layer structure parameters include current layer floating parameters and current layer fixed parameters. The comparison unit includes a calculation subunit and a comparison subunit. The calculation subunit is configured to calculate the sensitivity of the previous layer structure parameters at each wavelength point in the current layer spectrum. The comparison subunit is configured to fix the previous layer structure parameters with a sensitivity greater than a first preset value to the data processing model as current layer fixed parameters; to float the previous layer structure parameters with a sensitivity less than the first preset value individually to obtain current layer floating parameters; and to exclude the previous layer structure parameters with a sensitivity less than a second preset value.

[0083] In one implementation, the front-layer structural parameters include the front-layer critical dimension, and the computational subunit is specifically configured to analyze the front-layer spectrum and the current-layer spectrum in a data processing model to obtain the sensitivity of the front-layer critical dimension at each wavelength point in the current-layer spectrum.

[0084] In one embodiment, the comparison subunit is specifically configured to determine the corresponding wavelength point of the current layer spectrum for the front layer structural parameters whose sensitivity is less than a first preset value; to float the front layer structural parameters whose sensitivity is less than the first preset value individually, and to calculate the sensitivity of the floated front layer structural parameters at the corresponding wavelength point; and to use the floated front layer structural parameters whose sensitivity is greater than the first preset value as the floated parameters of the current layer.

[0085] In one implementation, the analysis module further includes an analysis unit configured to analyze the matching degree and output a list of analysis conclusions on the matching degree as a prediction result.

[0086] In one implementation, the analysis module uses a cosine similarity algorithm to analyze the matching degree.

[0087] In one implementation, the analysis system further includes an optimization module configured to import new spectral and structural parameters; integrate the new spectral and structural parameters with those stored in the database to form new training samples; and train a data processing model based on the new training samples.

[0088] In one embodiment, the structural parameters include at least one of key feature dimensions, thin film stacking, or semiconductor structure profile.

[0089] In one embodiment, the spectrum is a DAT file or a txt file in the database.

[0090] Exemplary embodiments of this disclosure also provide a computer-readable storage medium that can be implemented as a program product including program code, which, when run on an electronic device, causes the electronic device to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure. In one embodiment, the program product can be implemented as a portable compact disc read-only memory (CD-ROM) including program code and can run on an electronic device, such as a personal computer. However, the program product of this disclosure is not limited thereto. In this document, the readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0091] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0092] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0093] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0094] Program code for performing the operations of this disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing devices can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0095] Exemplary embodiments of this disclosure also provide an electronic device, which may be a back-end server for a semiconductor structure analysis system. References below... Figure 17 This electronic device will be described. It should be understood that... Figure 17 The electronic device 1000 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0096] like Figure 17 As shown, the electronic device 1000 is presented in the form of a general-purpose computing device. The components of the electronic device 1000 may include, but are not limited to: at least one processing unit 1010, at least one storage unit 1020, and a bus 1030 connecting different system components (including storage unit 1020 and processing unit 1010).

[0097] The storage unit stores program code, which can be executed by the processing unit 1010 to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 1010 can perform, as follows: Figure 3 The methods and steps shown are as follows.

[0098] Storage unit 1020 may include volatile storage units, such as random access memory (RAM) 1021 and / or cache memory 1022, and may further include read-only memory (ROM) 1023.

[0099] Storage unit 1020 may also include a program / utility 1024 having a set (at least one) program module 1025, such program module 1025 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0100] Bus 1030 may include a data bus, an address bus, and a control bus.

[0101] Electronic device 1000 can also communicate with one or more external devices 1100 (e.g., keyboard, pointing device, Bluetooth device, etc.) via input / output (I / O) interface 1040. Electronic device 1000 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public network, such as the Internet) via network adapter 1050. As shown, network adapter 1050 communicates with other modules of electronic device 1000 via bus 1030. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0102] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to exemplary embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0103] Those skilled in the art will understand that various aspects of this disclosure can be implemented as systems, methods, or program products. Therefore, various aspects of this disclosure can be embodied in entirely hardware implementations, entirely software implementations (including firmware, microcode, etc.), or implementations combining hardware and software aspects, collectively referred to herein as “circuit,” “module,” or “system.” Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0104] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is defined only by the appended claims.

Claims

1. A method for analyzing semiconductor structures, characterized in that, include: Establish a database storing spectral and structural parameters, including the spectral and structural parameters of experimentally designed wafers and the spectral and structural parameters of process-recorded wafers; A data processing model is generated based on the spectrum and the structural parameters; The data processing model is used to analyze the spectrum of the online wafer and provide prediction results, while the spectrum of the online wafer is saved in the database. The analysis and prediction results of online wafer fabrication based on the data processing model include: Obtain the current layer spectrum of the online wafer; Obtain the front-layer spectrum and front-layer structure parameters of the online wafer; By comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structural parameters in the data processing model, the current layer structural parameters are obtained. Calculate the matching degree between each of the layer structure parameters and the layer spectrum; The current layer structure parameters include current layer floating parameters and current layer fixed parameters; comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structure parameters in the data processing model to obtain the current layer structure parameters, includes: Calculate the sensitivity of the preceding layer structural parameters at each wavelength point in the current layer spectrum; The preceding layer structure parameters whose sensitivity is greater than the first preset value are fixed to the data processing model and used as the current layer fixed parameters; The current layer floating parameters are obtained by individually floating the front layer structural parameters whose sensitivity is less than a first preset value. The front-layer structural parameters whose sensitivity is less than the second preset value are excluded.

2. The method for analyzing semiconductor structures according to claim 1, characterized in that, The front-layer structural parameters include the critical dimensions of the front layer; calculating the sensitivity of the front-layer structural parameters at each wavelength point in the current layer spectrum includes: The front layer spectrum and the current layer spectrum are analyzed in the data processing model to obtain the sensitivity of the front layer key dimension at each wavelength point in the current layer spectrum.

3. The method for analyzing semiconductor structures according to claim 1, characterized in that, Individually floating the parameters of the preceding layer structure with sensitivity less than a first preset value to obtain the current layer floating parameters include: Determine the corresponding wavelength point on the current layer spectrum for the front layer structural parameter whose sensitivity is less than a first preset value; The parameters of the front-layer structure with sensitivity less than a first preset value are individually floated, and the sensitivity of the individually floated front-layer structure parameters at the corresponding wavelength points is calculated. The structural parameters of the previous layer after individual floating, where the sensitivity is greater than the first preset value, are used as the floating parameters of the current layer.

4. The method for analyzing semiconductor structures according to claim 1, characterized in that, The analysis and prediction results based on the data processing model for online wafer fabrication also include: The matching degree is analyzed to obtain the matching degree score; Output a list of the matching scores as the prediction results.

5. The method for analyzing semiconductor structures according to claim 4, characterized in that, The matching degree is analyzed using a cosine similarity algorithm.

6. The method for analyzing semiconductor structures according to claim 1, characterized in that, The establishment of a database storing spectral and structural parameters includes: Collect the spectral and structural parameters of the experimentally designed wafer and integrate them into the first data file; The first data file is imported into a pre-established empty database and saved to form the initial database.

7. The method for analyzing semiconductor structures according to claim 6, characterized in that, Establishing a database storing spectral and structural parameters also includes: Collect the spectral and structural parameters of the wafer recorded in the process and integrate them into a second data file; The second data file is integrated into the initial database to form the database.

8. The method for analyzing semiconductor structures according to claim 1, characterized in that, A data processing model is generated based on the spectrum and the structural parameters, including: Establish an initial model; Training samples are generated based on the spectra and structural parameters of multiple wafers in the database; The initial model is trained based on the training samples to obtain the data processing model.

9. The method for analyzing semiconductor structures according to claim 8, characterized in that, The method further includes: New spectral and structural parameters were imported; The new spectral and structural parameters are integrated with the spectral and structural parameters stored in the database to form new training samples, and the data processing model is trained based on the new training samples.

10. The method for analyzing semiconductor structures according to claim 1, characterized in that, The structural parameters include at least one of key feature dimensions, thin film stacking, or semiconductor structure profile.

11. The method for analyzing semiconductor structures according to claim 1, characterized in that, The spectra are in the database as DAT or txt files.

12. A semiconductor structure analysis system, using the semiconductor structure analysis method as described in any one of claims 1-11, characterized in that, include: The storage module is configured to establish a database storing spectral and structural parameters, including the spectral and structural parameters of experimental design wafers and process record wafers. The modeling module is configured to generate a data processing model based on the spectrum and the structural parameters; The analysis module is configured to analyze the online wafer based on a data processing model and provide prediction results, while simultaneously saving the spectrum of the online wafer in the database; wherein, analyzing the online wafer based on the data processing model and providing prediction results includes: Obtain the current layer spectrum of the online wafer; Obtain the front-layer spectrum and front-layer structure parameters of the online wafer; By comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structural parameters in the data processing model, the current layer structural parameters are obtained. Calculate the matching degree between each of the layer structure parameters and the layer spectrum; The current layer structure parameters include current layer floating parameters and current layer fixed parameters; comparing the current layer spectrum and the previous layer spectrum, and individually floating the previous layer structure parameters in the data processing model to obtain the current layer structure parameters, includes: Calculate the sensitivity of the preceding layer structural parameters at each wavelength point in the current layer spectrum; The preceding layer structure parameters whose sensitivity is greater than the first preset value are fixed to the data processing model and used as the current layer fixed parameters; The current layer floating parameters are obtained by individually floating the front layer structural parameters whose sensitivity is less than a first preset value. The front-layer structural parameters whose sensitivity is less than the second preset value are excluded.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 11.

14. An electronic device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 11.

Images