CN112561873A - CDSEM image virtual measurement method based on machine learning - Google Patents
CDSEM image virtual measurement method based on machine learning Download PDFInfo
- Publication number
- CN112561873A CN112561873A CN202011459003.3A CN202011459003A CN112561873A CN 112561873 A CN112561873 A CN 112561873A CN 202011459003 A CN202011459003 A CN 202011459003A CN 112561873 A CN112561873 A CN 112561873A
- Authority
- CN
- China
- Prior art keywords
- image
- cdsem
- photoetching
- neural network
- network model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
- G06T2207/10061—Microscopic image from scanning electron microscope
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Exposure And Positioning Against Photoresist Photosensitive Materials (AREA)
Abstract
A CDSEM image virtual measurement method based on machine learning comprises a training set and a verification set generation step, a photoetching space image map and a CDSEM image alignment step, a neural network model is adopted, the photoetching space image map is used as input, the CDSEM image corresponding to the photoetching space image map is used as target output, and N1 groups of photoetching space image map-CDSEM image data pairs in the training set are traversed to finish training of the neural network model; and traversing the N2 groups of the photoetching space image-CDSEM image data in the verification set to complete the verification of the neural network model. According to the method, the CDSEM image is generated by adopting machine learning through establishing the mapping between the post-OPC photomask pattern and the post-photoetching CDSEM image, and a verification model independent of an OPC model is obtained and used for confirming the quality of the photoetched pattern.
Description
Technical Field
The invention belongs to the field of semiconductor integrated circuit production and manufacturing, and relates to a CDSEM image virtual measurement method based on machine learning.
Background
In a photolithography process flow of semiconductor integrated circuit manufacturing, for a given pattern, a lithography Aerial Image (initial Image) of a photoresist on a wafer is determined under the condition of focusing of a lithography machine and dose determination, and a three-dimensional structure after development of the photoresist is determined under the condition of determination of the photoresist, at this time, a CDSEM Image taken by a Scanning Electron Microscope (SEM) is also determined, and the CDSEM Image is generally used for confirming the quality of the finally-photoetched pattern.
Generally, in a photolithography process, in order to improve the quality of a pattern of the photolithography process, an Optical Proximity Correction (OPC) is required to be performed on the pattern used for manufacturing a photomask. Model-based optical proximity correction it is important to establish an accurate OPC model that contains a series of parameters that require calibration of the OPC model by testing the experimental data of the pattern (e.g., CDSEM image), the more experimental data, the more accurate the model.
However, the test patterns may not cover all kinds, and the OPC model is not based entirely on physical principles, an experienced component in the model, so the accuracy of the OPC model cannot be guaranteed. Currently, the quality of the test pattern after OPC is determined by using the simulation result of the same OPC model as a verification means, and thus, the verification quality depends on the accuracy of the OPC model itself for all patterns.
It is clear to those skilled in the art that the OPC model itself is not guaranteed for the accuracy of all patterns. Therefore, there is a great need in the industry to ensure post-OPC graphics data quality for new verification means independent of other models of the OPC model.
Furthermore, lithography machines are classified into ultraviolet light sources (UV), deep ultraviolet light sources (DUV), extreme ultraviolet light sources (EUV). Random defects caused by random effects are a great problem in defect detection in lithography, especially in EUV lithography. Moreover, due to the limitation of resolution, the pattern after EUV lithography needs to be scanned by an electron beam scanner (E-beam) to determine whether the pattern after the lithography process on the wafer has defects, but due to the design constraint of the electron beam scanner, the electron beam scanner cannot detect the defects through a die-to-die inter-chip detection mode, and can only detect the defects through a die-to-database detection mode. That is, current EUV defect detection is achieved by comparing CDSEM images with OPC target geometry, which is not the best approach.
In the photoetching process, the stability of photoetching conditions plays a key role in stabilizing the image quality on the photoetched wafer. Specifically, in the photolithography process flow, for a given pattern, under the condition of focusing of a lithography machine and dose determination, a lithography Aerial Image (initial Image) of a photoresist on a wafer is also determined, and under the condition of photoresist determination, a three-dimensional structure after photoresist development is determined, and a CDSEM Image shot by a scanning electron microscope is also determined.
Disclosure of Invention
In view of the above problem of the monitoring technology of the stability of the lithography conditions, the invention provides a CDSEM image virtual measurement method based on machine learning, which realizes generation of a CDSEM image by machine learning by establishing a mapping between a post-OPC photomask pattern and a post-lithography CDSEM image, and obtains a verification model independent of an OPC model for confirming the quality of a post-lithography pattern.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a CDSEM image virtual measurement method based on machine learning comprises the following steps:
step S1: generating a training set and a verification set; it includes:
step S11: providing a wafer, and presetting the number of times of a photoetching process as K times; wherein K is a positive integer greater than or equal to 1;
step S12: completing a photoetching process flow on the wafer once, and using a scanning electron microscope to perform photoetching on the wafer MiScanning at different coordinates to obtain MiA CDSEM image; wherein M isiIs a positive integer greater than or equal to 10, i is one value of 1,2,3 … K;
step S13: calculating a photoetching space image map with the same coordinate as the CDSEM image, forming a group of photoetching space image map-CDSEM image data pairs by one CDSEM image and the corresponding photoetching space image map, and finally obtaining M groups of photoetching space image map-CDSEM image data pairs, wherein the photoetching space image map comprises at least one two-dimensional image at different depths of the photoresist;
step S14: judging whether the number of the groups of the photoetching spatial image map-CDSEM image data pairs is equal to N, if not, executing the step S12; if yes, go to step S15; wherein:
step S15: dividing N groups of photoetching space image map-CDSEM image data pairs into a training set for model training and a verification set for verifying a model in proportion; wherein the set number ratio of the lithography aerial image-CDSEM image data pairs in the training set and the verification set is N1: N2, N is N1+ N2;
step S2: aligning the lithographic aerial image map and the CDSEM image;
step S3: adopting a neural network model, taking the photoetching aerial image map as an input, taking the CDSEM image corresponding to the photoetching aerial image map as a target output, and traversing N1 groups of photoetching aerial image map-CDSEM image data in the training set to finish the training of the neural network model; and traversing the N2 groups of the photoetching space image-CDSEM image data in the verification set to complete the verification of the neural network model.
Further, in the CDSEM image virtual measurement method based on machine learning, step S3 includes:
step S31: providing the neural network model;
step S32: taking the photoetching spatial image map in the training set as an input, taking the CDSEM image corresponding to the photoetching spatial image map as a target output, traversing the photoetching spatial image map-CDSEM image data pair in the training set, and training the neural network model;
step S33: traversing the photoetching space image-CDSEM image data pairs in the verification set, verifying the neural network model, and calculating a loss function of the verification set;
step S34: judging whether the loss function is smaller than a set value or not, if so, stopping training the neural network model to obtain a final neural network model; if not, repeatedly executing the steps S15 to S34; wherein the neural network model embodies a mapping between a lithography aerial image map and the CDSEM image.
Further, the neural network model is a deep convolution neural network DCNN model mainly based on convolution or a generative countermeasure network GAN model, and ReLU is used as an activation function; if the neural network model adopts the deep convolutional neural network DCNN model, the loss function is a mean square error loss function, and if the neural network model adopts the generative countermeasure network GAN model, the loss function is a cross entropy loss function.
Further, the number of sets N1 of the lithography aerial image map-CDSEM image data pairs in the training set is a multiple of 7, and the number of sets N2 of the lithography aerial image map-CDSEM image data pairs in the validation set is a multiple of 3.
Further, the CDSEM image virtual measurement method based on machine learning further includes:
step S4: based on the final neural network model, when a new lithography aerial image is input, the final neural network model generates a virtual CDSEM image corresponding to the new lithography aerial image.
Further, the CDSEM image virtual measurement method based on machine learning further includes step S5: and detecting the key size of the virtual CDSEM image generated by the final neural network model, and determining whether the OPC optical model needs calibration correction according to the key size.
Further, in the CDSEM image virtual measurement method based on machine learning, the step S5 includes:
s51: acquiring the outline of the virtual CDSEM image, and finding out the critical dimension through the outline;
s52: and judging whether the critical dimension meets the process requirement, and if not, calibrating and correcting the OPC optical model.
Further, the CDSEM image virtual measurement method based on machine learning further includes step S5': and judging whether the random effect of the primary photoetching process is acceptable or not.
Further, the step S5' in the CDSEM image virtual measurement method based on machine learning includes:
s51', using the new photoetching spatial image corresponding photoetching process condition to carry out a photoetching process flow, and measuring to obtain an actual CDSEM image;
s52', the virtual CDSEM image generated by the final neural network model is compared with the actual CDSEM image, and if the mean square error of the pixel values of the virtual CDSEM image and the actual CDSEM image meets the precision requirement, the random effect of the photoetching process is judged to be acceptable.
Further, the image size and resolution of the lithographic aerial image and the CDSEM image are the same.
From the technical scheme, the CDSEM image virtual measurement method based on machine learning has the advantages that mapping of the post-OPC photomask pattern and the post-photoetching SEM image is established, the SEM image is generated through machine learning, and accordingly the purpose of confirming the quality of the photo-etched image is achieved.
Drawings
FIG. 1 is a schematic flow chart illustrating a CDSEM image virtual measurement method based on machine learning according to an embodiment of the present invention
FIG. 2 is a block diagram of an architecture for creating a CDSEM image after a photolithography process based on machine learning according to an embodiment of the present invention
FIG. 3 is a schematic diagram of a lithography aerial image calculated by a rigorous optical model for a post-OPC reticle pattern according to an embodiment of the present invention
FIG. 4 is a schematic diagram of an experimental CDSEM image after photolithography of a post-OPC reticle pattern in an embodiment of the present invention
FIG. 5 is a schematic diagram of a virtual CDSEM image generated by a deep learning model after learning according to an embodiment of the present invention
Detailed Description
The following description of the present invention will be made in detail with reference to the accompanying drawings 1 to 5.
It should be noted that, in the photolithography process of the CDSEM image virtual measurement method based on machine learning disclosed in the present invention, the photolithography machine maps the pattern on the reticle onto the Wafer (Wafer) coated with the photoresist through EUV or UV, etc., in the actual photolithography process flow, after the photolithography process parameters (depth of focus and dose) in the photolithography machine are determined, the pattern on the Wafer is correspondingly determined, at this time, the CDSEM image captured by the scanning electron microscope is also determined. Therefore, in the case of a process flow determination, there is a certain correspondence between the CDSEM image and the lithographic process parameters.
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a CDSEM image virtual measurement method based on machine learning according to an embodiment of the present invention. As shown in fig. 1, the CDSEM image virtual measurement method based on machine learning may include the following steps:
step S1: generating a training set and a verification set; it includes:
step S11: providing a wafer, and presetting the number of times of a photoetching process as K times; wherein K is a positive integer greater than or equal to 1;
step S12: completing a photoetching process flow on the wafer once, and using a scanning electron microscope to perform photoetching on the wafer MiScanning at different coordinates to obtain MiA CDSEM image; wherein M isiIs a positive integer greater than or equal to 10, i is one value of 1,2,3 … K;
step S13: calculating a photoetching space image map with the same coordinate as the CDSEM image, forming a group of photoetching space image map-CDSEM image data pairs by one CDSEM image and the corresponding photoetching space image map, and finally obtaining M groups of photoetching space image map-CDSEM image data pairs, wherein the photoetching space image map comprises at least one two-dimensional image at different depths of the photoresist;
step S14: judging whether the number of the groups of the photoetching spatial image map-CDSEM image data pairs is equal to N, if not, executing the step S12; if yes, go to step S15; wherein:
step S15: dividing N groups of photoetching space image map-CDSEM image data pairs into a training set for model training and a verification set for verifying a model in proportion; wherein, the ratio of the number of groups of the photoetching aerial image-CDSEM image data pairs in the training set and the verification set is N1: N2, and N is N1+ N2.
Step S2: aligning the lithography aerial image map and the CDSEM image prior to model training;
step S3: adopting a neural network model, taking the photoetching aerial image map as an input, taking the CDSEM image corresponding to the photoetching aerial image map as a target output, and traversing N1 groups of photoetching aerial image map-CDSEM image data in the training set to finish the training of the neural network model; and traversing the N2 groups of the photoetching space image-CDSEM image data in the verification set to complete the verification of the neural network model.
That is, referring to fig. 2, fig. 2 is a functional block diagram illustrating stability control of an offline photolithography process based on images according to an embodiment of the present invention. As shown in fig. 2, the training set for model training and the verification set for model verification are obtained from multiple actual photolithography processes (for example, 5 times of photolithography are performed, and the wafer coordinates of each scan are 200, 300, 50, 150 and 300, respectively, so that 1000 CDSEM images are finally obtained, that is, N is 1000). N groups of photoetching spatial image diagram-CDSEM image data pairs are proportionally divided into a training set for model training and a verification set for verifying a model; the ratio of the training set to the validation set is N1: N2, N1+ N2. Preferably, the method can be performed according to a ratio of 7:3 between a training set and a verification set, wherein the training set comprises 700 sets of photoetching aerial image-CDSEM image data pairs, and the verification set comprises 300 sets of photoetching aerial image-CDSEM image data pairs.
In the embodiment of the present invention, after the lithography spatial image-SEM image data pair is provided, the mapping relationship between the two may be derived by a Deep Convolutional Neural Network (DCNN) or a Generative Adaptive Network (GAN).
Referring to fig. 3, 4 and 5, fig. 3 is a schematic diagram illustrating a lithography aerial image obtained by calculating a strict optical model for a post-OPC reticle pattern according to an embodiment of the present invention, fig. 4 is a schematic diagram illustrating an experimental CDSEM image of the post-OPC reticle pattern after lithography according to the embodiment of the present invention, and fig. 5 is a schematic diagram illustrating a virtual CDSEM image generated by a deep learning model after learning according to the embodiment of the present invention.
In the embodiment of the present invention, since there may be a deviation between the coordinates of the actual pattern after lithography and the corresponding pattern coordinates on the reticle, before performing model training, step S2 is further executed: the lithographic aerial image map and the CDSEM image are aligned and, preferably, the image size and resolution of the lithographic aerial image map and the CDSEM image are the same. The image size is determined by the specific case, and may be 512 × 512 in this example.
Step S3: adopting a neural network model, taking the photoetching aerial image map as an input, taking the CDSEM image corresponding to the photoetching aerial image map as a target output, and traversing N1 groups of photoetching aerial image map-CDSEM image data in the training set to finish the training of the neural network model; and traversing the N2 groups of the photoetching space image-CDSEM image data in the verification set to complete the verification of the neural network model.
Specifically, a method from an Image To an Image (Image To Image) is mainly used for generating a corresponding CDSEM Image based on a photoetching aerial Image in exposed photoresist, taking the photoetching aerial Image as the input of a neural network model, taking a corresponding CDSEM Image as the target output of the neural network model, continuously training and verifying the neural network model, and adjusting parameters of the neural network model To finally complete the mapping from the photoetching aerial Image To the CDSEM Image.
In an embodiment of the present invention, step S3 of the CDSEM image virtual measurement method based on machine learning specifically includes:
step S31: providing the neural network model;
step S32: taking the photoetching spatial image map in the training set as an input, taking the CDSEM image corresponding to the photoetching spatial image map as a target output, traversing the photoetching spatial image map-CDSEM image data pair in the training set, and training the neural network model;
step S33: traversing the photoetching space image-CDSEM image data pairs in the verification set, verifying the neural network model, and calculating a loss function of the verification set;
step S34: judging whether the loss function is smaller than a set value or not, if so, stopping training the neural network model to obtain a final neural network model; if not, repeatedly executing the steps S15 to S34; wherein the neural network model embodies a mapping between a lithography aerial image map and the CDSEM image.
Further, the neural network model is a deep convolution neural network DCNN model mainly based on convolution or a generative countermeasure network GAN model, and ReLU is used as an activation function; if the neural network model adopts the deep convolutional neural network DCNN model, the loss function is a mean square error loss function, and if the neural network model adopts the generative countermeasure network GAN model, the loss function is a cross entropy loss function.
Further, the DCNN model may include an input layer, 13 intermediate layers, and an output layer, where the intermediate layers have the same structure, the convolution kernel size is 3 × 3, the width of each layer is 64 or 128 feature maps, each convolution layer is followed by batch normalization, the input layer performs only convolution and activation operations, and the output layer performs only convolution operations.
With the neural network model described above, step S4 can be performed, i.e., the final neural network model generates a corresponding virtual CDSEM image when a new lithography aerial image is input.
Step S4: based on the trained neural network model, when a new photoetching spatial image is input, the neural network model generates a virtual CDSEM image corresponding to the new photoetching spatial image; the photoetching aerial image is a photoetching aerial image at the position with the same coordinate as the CDSEM image, which is calculated according to the pattern of an EUV mask and process parameters, or the photoetching aerial image is a photoetching aerial image at the position with the same coordinate as the CDSEM image, which is obtained by calculating the wafer through an OPC optical model, wherein the position with the same coordinate as the CDSEM image is the position of the post-OPC photomask pattern.
In an embodiment of the present invention, when the new lithography aerial image is the lithography aerial image of the wafer at the same coordinates as the CDSEM image obtained by the optical model calculation for the post-OPC reticle pattern, the method further includes step S5: and detecting the key size of the virtual CDSEM image generated by the final neural network model, and determining whether the OPC optical model needs calibration correction according to the key size.
Further, the step S5 may include:
and detecting the key size of the virtual CDSEM image generated by the final neural network model, and determining whether the OPC optical model needs calibration correction according to the key size.
In another embodiment of the present invention, when the new lithography aerial image is the lithography aerial image at the same coordinates as the CDSEM image calculated according to the EUV reticle pattern and the process parameters, the method further includes step S5': generating a virtual CDSEM image by using the generated neural network model for the photoetching space image of the EUV mask as a reference CDSEM image of electron beam defect scanning, and comparing the CDSEM image after EUV photoetching with the reference CDSEM image; when the mean square error of the pixel values of the CDSEM image after photoetching and the reference CDSEM image meets the precision requirement, judging that the geometric data of the photoetching pattern generated after EUV photoetching does not have a problem, otherwise, judging that the geometric data of the photoetching pattern generated after EUV photoetching has random defects.
Specifically, the step S5' may include:
s51', using the new photoetching spatial image corresponding photoetching process condition to carry out a photoetching process flow, and measuring to obtain an actual CDSEM image;
s52', the virtual CDSEM image generated by the final neural network model is compared with the actual CDSEM image, and if the mean square error of the pixel values of the virtual CDSEM image and the actual CDSEM image meets the precision requirement, the random effect of the photoetching process is judged to be acceptable.
That is, in the stage of model application, a lithography Aerial image (initial image) is input, the model outputs a corresponding CDSEM virtual image after lithography, the CDSEM virtual image can be used as a standard independent of the OPC model, and compared with the image after lithography obtained by the simulation of the OPC model, the OPC model whose mean square error is smaller than the preset accuracy obtained by the comparison is an acceptable OPC model, otherwise, the OPC model is recalibrated using more experimental data.
The above description is only for the preferred embodiment of the present invention, and the embodiment is not intended to limit the scope of the present invention, so that all the equivalent structural changes made by using the contents of the description and the drawings of the present invention should be included in the scope of the present invention.
Claims (10)
1. A CDSEM image virtual measurement method based on machine learning is characterized by comprising the following steps:
step S1: generating a training set and a verification set; it includes:
step S11: providing a wafer, and presetting the number of times of a photoetching process as K times; wherein K is a positive integer greater than or equal to 1;
step S12: completing a photoetching process flow on the wafer once, and using a scanning electron microscope to perform photoetching on the wafer MiScanning at different coordinates to obtain MiA CDSEM image; wherein M isiIs a positive integer greater than or equal to 10, i is one value of 1,2,3 … K;
step S13: calculating a photoetching space image map with the same coordinate as the CDSEM image, forming a group of photoetching space image map-CDSEM image data pairs by one CDSEM image and the corresponding photoetching space image map, and finally obtaining M groups of photoetching space image map-CDSEM image data pairs, wherein the photoetching space image map comprises at least one two-dimensional image at different depths of the photoresist;
step S14: judging whether the number of the groups of the photoetching spatial image map-CDSEM image data pairs is equal to N, if not, executing the step S12; if yes, go to step S15; wherein:
step S15: dividing N groups of photoetching space image map-CDSEM image data pairs into a training set for model training and a verification set for verifying a model in proportion; wherein the set number ratio of the lithography aerial image-CDSEM image data pairs in the training set and the verification set is N1: N2, N is N1+ N2;
step S2: aligning the lithographic aerial image map and the CDSEM image;
step S3: adopting a neural network model, taking the photoetching aerial image map as an input, taking the CDSEM image corresponding to the photoetching aerial image map as a target output, and traversing N1 groups of photoetching aerial image map-CDSEM image data in the training set to finish the training of the neural network model; and traversing the N2 groups of the photoetching space image-CDSEM image data in the verification set to complete the verification of the neural network model.
2. The machine learning-based CDSEM image virtual measurement method according to claim 1, wherein step S3 includes:
step S31: providing the neural network model;
step S32: taking the photoetching spatial image map in the training set as an input, taking the CDSEM image corresponding to the photoetching spatial image map as a target output, traversing the photoetching spatial image map-CDSEM image data pair in the training set, and training the neural network model;
step S33: traversing the photoetching space image-CDSEM image data pairs in the verification set, verifying the neural network model, and calculating a loss function of the verification set;
step S34: judging whether the loss function is smaller than a set value or not, if so, stopping training the neural network model to obtain a final neural network model; if not, repeatedly executing the steps S15 to S34; wherein the neural network model embodies a mapping between a lithography aerial image map and the CDSEM image.
3. The machine-learning-based CDSEM image virtual measurement method according to claim 2, wherein the neural network model is a convolution-based Deep Convolution Neural Network (DCNN) model or a generative countermeasure network (GAN) model, and ReLU is used as an activation function; if the neural network model adopts the deep convolutional neural network DCNN model, the loss function is a mean square error loss function, and if the neural network model adopts the generative countermeasure network GAN model, the loss function is a cross entropy loss function.
4. The machine-learning-based CDSEM image virtual measurement method of claim 1, wherein the number of sets N1 of the lithography aerial image map-CDSEM image data pairs in the training set is a multiple of 7, and the number of sets N2 of the lithography aerial image map-CDSEM image data pairs in the verification set is a multiple of 3.
5. The machine-learning-based CDSEM image virtual measurement method according to claim 1, further comprising:
step S4: based on the final neural network model, when a new lithography aerial image is input, the final neural network model generates a virtual CDSEM image corresponding to the new lithography aerial image.
6. The machine-learning-based CDSEM image virtual measurement method according to claim 5, further comprising step S5: and detecting the key size of the virtual CDSEM image generated by the final neural network model, and determining whether the OPC optical model needs calibration correction according to the key size.
7. The machine learning-based CDSEM image virtual measurement method according to claim 6, wherein the step S5 comprises:
s51: acquiring the outline of the virtual CDSEM image, and finding out the critical dimension through the outline;
s52: and judging whether the critical dimension meets the process requirement, and if not, calibrating and correcting the OPC optical model.
8. The machine-learning-based CDSEM image virtual measurement method according to claim 5, further comprising step S5': and judging whether the random effect of the primary photoetching process is acceptable or not.
9. The machine learning-based CDSEM image virtual measurement method according to claim 8, wherein the step S5' comprises:
s51', using the new photoetching spatial image corresponding photoetching process condition to carry out a photoetching process flow, and measuring to obtain an actual CDSEM image;
s52', the virtual CDSEM image generated by the final neural network model is compared with the actual CDSEM image, and if the mean square error of the pixel values of the virtual CDSEM image and the actual CDSEM image meets the precision requirement, the random effect of the photoetching process is judged to be acceptable.
10. The machine-learning-based CDSEM image virtual measurement method of claim 1, wherein the image size and resolution of the lithography aerial image map and CDSEM image are the same.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011459003.3A CN112561873B (en) | 2020-12-11 | 2020-12-11 | CDSEM image virtual measurement method based on machine learning |
| PCT/CN2021/134560 WO2022121736A1 (en) | 2020-12-11 | 2021-11-30 | Cdsem image virtual measurement method based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011459003.3A CN112561873B (en) | 2020-12-11 | 2020-12-11 | CDSEM image virtual measurement method based on machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112561873A true CN112561873A (en) | 2021-03-26 |
| CN112561873B CN112561873B (en) | 2022-11-25 |
Family
ID=75062324
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011459003.3A Active CN112561873B (en) | 2020-12-11 | 2020-12-11 | CDSEM image virtual measurement method based on machine learning |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112561873B (en) |
| WO (1) | WO2022121736A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022121736A1 (en) * | 2020-12-11 | 2022-06-16 | 上海集成电路装备材料产业创新中心有限公司 | Cdsem image virtual measurement method based on machine learning |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114841378B (en) * | 2022-07-04 | 2022-10-11 | 埃克斯工业(广东)有限公司 | Wafer characteristic parameter prediction method and device, electronic equipment and readable storage medium |
| CN117669473B (en) * | 2024-01-29 | 2024-04-19 | 全智芯(上海)技术有限公司 | Method for model calibration, electronic device and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1862383A (en) * | 2006-04-04 | 2006-11-15 | 上海微电子装备有限公司 | Aberration field measuring method for imaging optical system of photoetching apparatus |
| CN101042526A (en) * | 2004-02-23 | 2007-09-26 | 株式会社东芝 | Mask data correction method, photomask and optical image prediction method |
| CN101877016A (en) * | 2009-04-30 | 2010-11-03 | 新思科技有限公司 | Modeling critical-dimension (CD) scanning-electron-microscopy (CD-sem) CD extraction |
| CN107004060A (en) * | 2014-11-25 | 2017-08-01 | 流动马赛克公司 | Improved process control technology for semiconductor fabrication process |
| CN108228981A (en) * | 2017-12-19 | 2018-06-29 | 上海集成电路研发中心有限公司 | The Forecasting Methodology of OPC model generation method and experimental pattern based on neural network |
| CN108475417A (en) * | 2016-01-04 | 2018-08-31 | 科磊股份有限公司 | It is applied for semiconductor and high-definition picture is generated by low-resolution image |
| US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
| CN111310407A (en) * | 2020-02-10 | 2020-06-19 | 上海集成电路研发中心有限公司 | Method for designing optimal feature vector of reverse photoetching based on machine learning |
| CN111386500A (en) * | 2017-11-22 | 2020-07-07 | 卡尔蔡司Smt有限责任公司 | Method for authenticating a mask for microlithography |
| CN111985611A (en) * | 2020-07-21 | 2020-11-24 | 上海集成电路研发中心有限公司 | Computing method based on physical characteristic diagram and DCNN machine learning reverse photoetching solution |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017171891A1 (en) * | 2016-04-02 | 2017-10-05 | Intel Corporation | Systems, methods, and apparatuses for modeling reticle compensation for post lithography processing using machine learning algorithms |
| DE102018207880A1 (en) * | 2018-05-18 | 2019-11-21 | Carl Zeiss Smt Gmbh | Method and apparatus for evaluating an unknown effect of defects of an element of a photolithography process |
| WO2019238372A1 (en) * | 2018-06-15 | 2019-12-19 | Asml Netherlands B.V. | Machine learning based inverse optical proximity correction and process model calibration |
| KR20230175346A (en) * | 2018-12-28 | 2023-12-29 | 에이에스엠엘 네델란즈 비.브이. | Determining pattern ranking based on measurement feedback from printed substrate |
| US11061318B2 (en) * | 2019-02-28 | 2021-07-13 | Taiwan Semiconductor Manufacturing Co., Ltd. | Lithography model calibration |
| CN112561873B (en) * | 2020-12-11 | 2022-11-25 | 上海集成电路装备材料产业创新中心有限公司 | CDSEM image virtual measurement method based on machine learning |
-
2020
- 2020-12-11 CN CN202011459003.3A patent/CN112561873B/en active Active
-
2021
- 2021-11-30 WO PCT/CN2021/134560 patent/WO2022121736A1/en unknown
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101042526A (en) * | 2004-02-23 | 2007-09-26 | 株式会社东芝 | Mask data correction method, photomask and optical image prediction method |
| CN1862383A (en) * | 2006-04-04 | 2006-11-15 | 上海微电子装备有限公司 | Aberration field measuring method for imaging optical system of photoetching apparatus |
| CN101877016A (en) * | 2009-04-30 | 2010-11-03 | 新思科技有限公司 | Modeling critical-dimension (CD) scanning-electron-microscopy (CD-sem) CD extraction |
| CN107004060A (en) * | 2014-11-25 | 2017-08-01 | 流动马赛克公司 | Improved process control technology for semiconductor fabrication process |
| CN108475417A (en) * | 2016-01-04 | 2018-08-31 | 科磊股份有限公司 | It is applied for semiconductor and high-definition picture is generated by low-resolution image |
| CN111386500A (en) * | 2017-11-22 | 2020-07-07 | 卡尔蔡司Smt有限责任公司 | Method for authenticating a mask for microlithography |
| CN108228981A (en) * | 2017-12-19 | 2018-06-29 | 上海集成电路研发中心有限公司 | The Forecasting Methodology of OPC model generation method and experimental pattern based on neural network |
| US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
| CN111310407A (en) * | 2020-02-10 | 2020-06-19 | 上海集成电路研发中心有限公司 | Method for designing optimal feature vector of reverse photoetching based on machine learning |
| CN111985611A (en) * | 2020-07-21 | 2020-11-24 | 上海集成电路研发中心有限公司 | Computing method based on physical characteristic diagram and DCNN machine learning reverse photoetching solution |
Non-Patent Citations (3)
| Title |
|---|
| ANOOP CHERIAN: "Sem-GAN: Semantically-Consistent Image-to-Image Translation", 《ARXIV》 * |
| 张逸: "基于卷积神经网络的复材射线图像孔缺陷识别", 《光学与光电技术》 * |
| 邡鑫，史峥: "基于卷积神经网络的晶圆缺陷检测与分类算法", 《计算机工程》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022121736A1 (en) * | 2020-12-11 | 2022-06-16 | 上海集成电路装备材料产业创新中心有限公司 | Cdsem image virtual measurement method based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112561873B (en) | 2022-11-25 |
| WO2022121736A1 (en) | 2022-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112561873B (en) | CDSEM image virtual measurement method based on machine learning | |
| KR100596760B1 (en) | Visual inspection and verification system | |
| US7518110B2 (en) | Pattern measuring method and pattern measuring device | |
| KR102441582B1 (en) | MPC(Mask Process Correction) verification method, and method for fabricating mask comprising the MPC verification method | |
| US20040225488A1 (en) | System and method for examining mask pattern fidelity | |
| KR20200015708A (en) | Measurement method and device | |
| TW201721280A (en) | Method and apparatus for pattern correction and verification | |
| US20180364589A1 (en) | Improvements in gauge pattern selection | |
| CN111430261A (en) | Process detection method and device | |
| US7930654B2 (en) | System and method of correcting errors in SEM-measurements | |
| KR102294369B1 (en) | Methods for measuring fluctuations, inspection systems, computer programs, and computer systems | |
| TWI780470B (en) | Method and apparatus for lithographic process performance determination | |
| CN111771167B (en) | Alignment mark positioning in lithographic processes | |
| CN112541545A (en) | Method for predicting CDSEM image after etching process based on machine learning | |
| Weisbuch et al. | Improving ORC methods and hotspot detection with the usage of aerial images metrology | |
| KR20210046459A (en) | OPC(Optical Proximity Correction) method using multi-OPC model, and methods for manufacturing mask using the OPC method | |
| JP2005250360A (en) | Verification apparatus and verification method for mask pattern | |
| TW201931003A (en) | Microlithographic mask, method for determining edge positions of the images of the structures of such a mask and system for carrying out such a method | |
| US20230132893A1 (en) | Mask layout correction methods based on machine learning, and mask manufacturing methods including the correction methods | |
| TWI788783B (en) | Method for measuring photomasks | |
| US20220283496A1 (en) | Photomask and method for inspecting photomask | |
| CN112099310B (en) | Method for acquiring light intensity threshold and method for detecting development condition of auxiliary pattern | |
| JP7492389B2 (en) | Image inspection device and image inspection method | |
| TW202407638A (en) | Mask inspection for semiconductor specimen fabrication | |
| TW202331644A (en) | Mask inspection for semiconductor specimen fabrication |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |