CN115795871A - Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask - Google Patents
Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask Download PDFInfo
- Publication number
- CN115795871A CN115795871A CN202211525221.1A CN202211525221A CN115795871A CN 115795871 A CN115795871 A CN 115795871A CN 202211525221 A CN202211525221 A CN 202211525221A CN 115795871 A CN115795871 A CN 115795871A
- Authority
- CN
- China
- Prior art keywords
- defect
- multilayer film
- bot
- space image
- mask
- 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.)
- Pending
Links
Images
Landscapes
- Preparing Plates And Mask In Photomechanical Process (AREA)
Abstract
A reconstruction method for the defect appearance of a multilayer film of an extreme ultraviolet photoetching blank mask. The invention adopts the blank mask multilayer film defect space images in a plurality of illumination directions to represent phase information of image deletion, and constructs the relation between the multilayer film defect space images and the appearance parameters of the multilayer film defects through a trained convolutional neural network model. The method includes firstly simulating a space image set of multilayer film defects in different illumination directions to construct training set data, and then inputting the training set data into a convolutional neural network model for training. And inputting the actually measured mask multilayer film defect space image set into the trained model to obtain defect shape parameters. The method is based on the relation between the direct construction of the multilayer film defect space image in multiple illumination directions and the defect appearance parameters, and has the characteristics of rapid process, strong generalization and high reconstruction accuracy.
Description
Technical Field
The invention relates to a photoetching mask, in particular to a method for reconstructing the defect appearance of a multilayer film of an extreme ultraviolet photoetching mask.
Background
Photolithography is a critical technique for the manufacture of very large scale integrated circuits. Mask defects are one of the major problems with extreme ultraviolet (hereinafter abbreviated EUV) lithography. Multilayer film defects are defects peculiar to EUV lithography masks, and refer to deformation of multilayer films caused by projections, depressions on a substrate and particles falling during deposition, which affect both the amplitude and phase of light reflected by the multilayer films. Since the exposure wavelength of EUV lithography is very short, the reflected light can generate obvious phase change only by the multilayer film defect with nanometer-sized three-dimensional morphology, so that the reflected light can obviously affect the imaging quality of EUV lithography.
The surface morphology of the multilayer film defect can be measured by using the existing detection equipment (such as an Atomic Force Microscope (AFM)), but the measurement of the surface morphology only can hardly meet the requirements of defect simulation analysis and compensation. The three-dimensional morphology of the multilayer film defects is difficult to measure directly with a non-destructive measurement mode. In order to solve the problem, researchers have proposed various methods for indirectly obtaining the three-dimensional shape of the defect of the multilayer film.
Luminescennt Technologies, inc. proposed a phase-type defect topography reconstruction method (see prior art 1.D.G.Stearns, P.B.Mirkarimi, and E.Spiller, "Localized defects in multilayer coatings," Thin Solid Films 446,37-49 (2004)); upadhhyaya et al propose a defect reconstruction method that takes into account the effect of deposition conditions on the growth model (see prior art 2.M. Upadhhyaya, v. Jindal, a. Baseplaippa, h. Herbol, j. Harrisjones, i. -y. Jang, k. A. Goldberg, i. Mochi, s. Marokkey, w. Demerle, t. V. Pistor, and g. Debeaux, "Evaluating printing reliability of deposited natural mask phase defects a modification and deposition approach," proc. Spie 9422,94220q (2015)). The methods are based on the growth model, need to consider the influence of different deposition tools and deposition conditions, and have low universality. Xu et al, the institute for Fraunhofer, germany, proposed a method for reconstructing defect morphology based on aerial images (see prior art 3.D. Xu, P. Evanschictzky, and A. Erdmann, "Extreme ultrasound multilayer defect analysis and geometry reconstruction," J.micro/nanolith. MEMS MOEMS 15,014002 (2016)). The method adopts an intensity transmission equation to recover the phase of the space image, adopts a Principal Component Analysis (PCA) method to extract the characteristic value of the space image, and utilizes an artificial neural network to reconstruct defect parameters. But the height and the full width at half maximum of the bottom of the defect are set to the same value in the modeling process of the method. In practical situations, the two are generally unequal, and the method has obvious defects on the characterization of the three-dimensional shape of the defect. Chen et al propose an improved defect profile characterization method based on blank mask image, which optimizes neural network model and improves reconstruction accuracy using a circular consistency technique (see prior art 4. Wei et al propose an extreme ultraviolet phase defect characterization method based on aerial image complex amplitude, which reconstructs the aerial image complex amplitude of a defect-containing blank mask using Fourier stack imaging (FP). The relationship between the aerial image and the defect bottom topography is established by a deep learning method (see the prior art 5.W. Cheng, S. Li, X. Wang, and Z. Zhang, ". Extreme ultrasound phase defect reconstruction based on complex algorithms of the analytical images," Applied Optics 60,5208-5219 (2021)).
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a method for reconstructing the defect appearance of a mask multilayer film of an extreme ultraviolet lithography, which constructs a direct mapping relation between defect space images of the mask multilayer film in multiple illumination directions and defect appearance parameters, avoids the need of high aperture overlapping rate to ensure the improvement of algorithm convergence by using Fourier laminated imaging, shortens the acquisition time and reduces the calculated amount; the convolutional neural network is constructed by adopting the residual network structure expansion and the antagonistic network structure generation, so that the accuracy of the defect reconstruction of the multilayer film of the EUV mask is improved. The method has the characteristics of quick process, strong generalization and high reconstruction accuracy.
The technical solution of the invention is as follows:
a reconstruction method for the defect appearance of a multilayer film of an extreme ultraviolet lithography mask comprises the following steps:
1) Setting simulation conditions:
(1) setting imaging simulation conditions:
setting the illumination condition as conventional illumination but not limited to the conventional illumination, wherein the partial coherence factor of the conventional illumination is sigma, and the wavelength of the light wave is lambda; setting the numerical aperture NA range of the projection objective to be 0.33-0.55; setting the number of the layers of the molybdenum/silicon double-layer film as m; the thickness of the molybdenum is set to be Mo thick The thickness of silicon is Si thick (ii) a Setting the refractive index of molybdenum as Mo ref The refractive index of silicon is Si ref (ii) a Setting the substrate SiO 2 The thickness is d; setting the magnification of an optical system to be alpha; the acquisition lengths of the object plane in the directions of the transverse axis and the longitudinal axis are respectively X 0 And Y 0 At a sampling interval dX 0 And dY 0 (ii) a Setting the acquisition lengths of the horizontal axis and the longitudinal axis of the space image as X and Y, and the sampling intervals as dX and dY;
(2) setting defect shape parameters of a blank mask containing multilayer film defects in a training set:
characterizing the appearance of the multi-layer film defect of the mask by Gaussian defect parameters, and setting the surface height of the multi-layer film defect as h top The surface half-height width of the multilayer film defect is omega top The height of the defect bottom of the multilayer film is h bot And the half height width at the bottom of the multilayer film defect is omega bot (ii) a Coupled as a morphological parameter Δ (Δ = (h) for multi-layer film defects top 、ω top 、h bot 、ω bot ) ); setting multilayer film defect top height h top In the range of 0.5-5nm, the value interval is dh top Full width at half maximum omega at the top of the defect top Within the range of 35-70nm, the value interval is d omega top Height of defect bottom h bot In the range of 6-20nm, the value interval is dh bot Full width at half maximum of defect bottom omega bot Within the range of 15-50nm, the value interval is d omega bot 。
2) The intensity of a blank mask space image containing the multilayer film defects in different illumination directions is obtained through simulation:
setting n lighting directions, named as l 1 ，l 2 …l n The incident angle of the EUV conventional illumination chief ray is θ =6 °, wherein，θ r The included angle between the main ray of the traditional illumination and the main ray of the new illumination direction is phi, the included angle is an azimuth angle relative to the main ray of the traditional illumination, the photoetching simulation software is utilized to carry out imaging simulation on the blank mask containing the multilayer film defects, and the illumination angle is l 1 ，l 2 …l n Intensity of blank mask aerial image containing multilayer film defect I lr1 ，I lr2 ，…，I lrn And setting the intensity map of the multilayer film defect blank mask space image as a training set sample. Reducing the value interval dh of the top height of the multilayer film defect in the step (1) top Full width at half maximum spacing d ω from the top of the defect top The height of the bottom of the defect is taken as the interval dh bot Full width at half maximum of defect bottom value interval d omega bot Forming defect appearance parameters of another group of multilayer film defect blank masks, and setting an intensity graph of space images of the multilayer film defect blank masks obtained through photoetching simulation software as a test set sample;
3) Training a convolution neural network with the shape parameter delta of the multilayer film defect as output:
simulating an amplitude image I of a space image of a multi-layer film defect blank mask in the step (2) lr1 ，I lr2 ，…，I lrn As the input of the convolutional neural network, the appearance parameter delta of the corresponding mask multilayer film defect is used as the output of the convolutional neural network, and a defect appearance parameter reconstruction model is constructed by adopting an expanded residual network structure; then, the shape parameter delta of the mask multilayer film defect is used as the input of the neural network, and the amplitude image I of the blank mask space image of the multilayer film defect lr1 ，I lr2 ，…，I lrm And as the output of the neural network, constructing a plurality of generation countermeasure networks according to different outputs to form a plurality of identification models, and changing the loss function of the defect appearance parameter reconstruction model through the identification models so as to improve the accuracy of the defect appearance parameter reconstruction model. The number of identification models affects the improvement of the defect appearance parameter reconstruction model. Setting the number of identification models as M, setting a loss function by MSE (mean square error), preprocessing a training set and a test set, training the constructed models by the processed training set, and obtaining the models by the test setAn optimal trained model;
4) Collecting an actually measured space image:
setting the illumination and numerical aperture of the detection system according to the same parameters in the step 1), loading a blank mask to be detected by the detection system, and completing the acquisition of a space image by using a space image sensor;
5) Detecting the defect shape parameters:
inputting the aerial image measured in the step 4) into a model to obtain a defect morphology parameter delta.
Compared with the prior art, the invention has the following advantages:
1. the direct mapping relation is constructed between the mask multilayer film defect space images in multiple illumination directions and defect appearance parameters, the condition that the algorithm convergence is improved by using Fourier laminated imaging with high aperture overlapping rate is avoided, the acquisition time is shortened, and the calculation amount is reduced.
2. The convolutional neural network is constructed by adopting the residual network structure expansion and the antagonistic network structure generation, so that the accuracy of the defect reconstruction of the multilayer film of the EUV mask is improved.
Drawings
FIG. 1 is a schematic view of the illumination mode employed in the present invention
FIG. 2 is a schematic view of the direction of illumination employed in the present invention
FIG. 3 is a schematic diagram of a mask multi-layer film defect
FIG. 4 is a regression graph of accuracy for reconstructing defect morphology parameters of a multi-layer film of a blank mask by using the technical scheme of the invention
Detailed Description
The present invention will be further described with reference to the following examples and drawings, but the scope of the present invention should not be limited thereto.
The illumination mode is shown in fig. 1, and includes, but is not limited to, the conventional illumination of fig. 1. This example uses conventional illumination, in which part of the coherence factor is σ =0.5.
The illumination directions are shown in fig. 2, including but not limited to the nine directions of fig. 2.
The defect profile of the mask multilayer film is shown in FIG. 3, including but not limited to that of FIG. 1Bottom-layer sunken multilayer film defect surface height h top Surface half-height width omega top Bottom height h bot And bottom half-height width omega bot 。
Examples
The method for reconstructing the multilayer film defect appearance of the extreme ultraviolet lithography mask comprises the following four steps:
1) The simulation conditions were set as follows:
the illumination conditions were conventional, the partial coherence factor σ =0.5, and the numerical aperture of the projection objective NA =0.33. The number of the layers of the molybdenum/silicon double-layer film is m =40, and the thickness of the molybdenum is Mo thick =2.78nm and the thickness of silicon is Si thick =4.17nm. Molybdenum has a refractive index of Mo ref =0.9238-0.006435j, refractive index of silicon is Si ref =0.9990-0.0018265j. SiO substrate 2 The thickness is d =20nm. Magnification α =900. Collecting length X of object plane in horizontal axis and vertical axis directions 0 And Y 0 380nm, sample interval dX 0 And dY 0 Is 3.8nm; the length of acquisition X and Y of the aerial image in the horizontal and vertical axis directions was 342 μm, and the sampling intervals dX and dY were 3.6 μm. H of multilayer film defect top Sampling interval dh top ＝0.5nm、ω top Sampling interval d omega top ＝5nm、h bot Sampling interval dh bot ＝2nm、ω bot Sampling interval d omega bot ＝5nm；
2) The intensity of a blank mask space image containing the multilayer film defects in different illumination directions is obtained through simulation:
5 illumination angles l in the table 1 are obtained by utilizing photoetching simulation software simulation 1 ，l 2 …l 5 Intensity I of blank mask aerial image containing multilayer film defect with lower different defect morphology parameters lr1 ，I lr2 …I lr5 Setting an intensity graph of the multilayer film defect blank mask space image as a training set sample; illumination angle l 1 ，l 2 …l 5 Set as shown in Table 1, where θ r Representing the angle between the illumination light and the main incident light and phi the azimuth angle. Changing the value interval of the defect in the simulation condition, and setting the multilayer film defect omega top Sampling interval d omega top ＝1nm、h bot Sampling interval dh bot ＝2nm、ω bot Sampling interval d omega bot The method comprises the following steps of (1) forming defect appearance parameters of another group of multilayer film defect blank masks, and setting an intensity graph of a space image of the multilayer film defect blank masks obtained through photoetching simulation software as a test set sample;
TABLE 1 illumination light Angle settings
3) Training a convolutional neural network with the shape parameter delta of the multilayer film defect as output:
simulating the amplitude image I of the space image of the multi-layer film defect blank mask lr1 ，I lr2 …I lr5 And as the input of the convolutional neural network, the appearance parameter delta of the corresponding mask multilayer film defect is used as the output of the convolutional neural network. And constructing a defect appearance parameter reconstruction model by adopting the expanded residual network structure. The shape parameter delta of the mask multilayer film defect is used as the input of the neural network, and the amplitude image I of the blank mask space image of the multilayer film defect lr1 ，I lr2 And as the output of the neural network, constructing a plurality of generation countermeasure networks according to the difference of the output to form a plurality of identification models. The accuracy of the defect appearance parameter reconstruction model is improved by changing the loss function of the defect appearance parameter reconstruction model through the identification model.
The neural network takes MSE (mean square error) as a loss function, and the formula is as follows:
wherein k is the input sample size; reconstructing the model p for the defect profile parameters rec For the predicted defect profile parameter, p def The defect shape parameters of the label are obtained; for the authentication model p rec For predicted amplitude images I lr1 ，I lr2 ，p def As amplitude image I of the label lr1 ，I lr2 . Passing through the notchAnd adding the loss function of the sunken appearance parameter reconstruction model and the average loss function of the plurality of identification models to form the loss function of the integral model.
And preprocessing the training set and the test set, training the convolutional neural network through the processed training set, and obtaining the optimal trained convolutional neural network through the test set.
4) Collecting an actually measured space image:
and (2) setting the illumination and numerical aperture of the detection system according to the same parameters in the step 1), loading the blank mask to be detected by the detection system, and completing the acquisition of a space image by using a space image sensor.
In this embodiment, the accuracy of the present invention is illustrated by reconstructing the defect morphology parameters of the multilayer film by a simulation method. Setting multilayer film defect h top Sampling interval dh top ＝0.5nm，ω top Sampling interval d omega top ＝1nm、h bot Sampling interval dh bot ＝1nm、ω bot Sampling interval d omega bot And (5) randomly extracting defect morphology parameters for simulation, wherein the defect morphology parameters are 1 nm. 80 space images of the multilayer film defects with different appearances are obtained and serve as actual measurement space images, and the space images form an actual measurement space image set.
5) Detecting the defect appearance parameters:
inputting the measured space image set into the trained model to obtain a measured defect parameter set p rec 。
This example was verified with simulation accuracy. Will calculate the measured parameter set p rec And a simulation set value p def Average Relative Error (ARE):
the mean relative error for the model reconstruction was found to be 1.51%. The model regression result is shown in fig. 4, which shows that the model has high accuracy for reconstructing the defect morphology parameters.
Experiments show that the direct mapping relation is established between the mask multilayer film defect space images in multiple illumination directions and defect morphology parameters, the problem that the algorithm convergence is improved due to the fact that the Fourier laminated imaging needs high aperture overlapping rate is avoided, the acquisition time is shortened, and the calculated amount is reduced. The method adopts the residual network structure expansion and the antagonistic network structure generation to construct the convolutional neural network, thereby improving the accuracy of the defect reconstruction of the EUV mask multilayer film.
Claims (1)
1. A reconstruction method for the defect appearance of a multilayer film of an extreme ultraviolet lithography mask comprises the following steps:
1) Setting simulation conditions:
(1) the imaging simulation conditions were set as follows:
the illumination condition is traditional illumination but is not limited to traditional illumination, the partial coherence factor of the traditional illumination is sigma, and the wavelength of light wave is lambda; the numerical aperture NA range of the projection objective is 0.33-0.55; the number of the layers of the molybdenum/silicon double-layer film is m; the thickness of the molybdenum is Mo thick The thickness of silicon is Si thick (ii) a Molybdenum has a refractive index of Mo ref The refractive index of silicon is Si ref (ii) a SiO substrate 2 The thickness is d; the magnification of the optical system is alpha; the acquisition lengths of the object plane in the directions of the transverse axis and the longitudinal axis are respectively X 0 And Y 0 At a sampling interval dX 0 And dY 0 (ii) a The acquisition length of the space image in the direction of the transverse axis and the acquisition length of the space image in the direction of the longitudinal axis are X and Y, and the sampling interval is dX and dY;
(2) the defect shape parameters of the photomask blank containing the multilayer film defects in the training set are set as follows:
characterizing the appearance of the mask multilayer film defect by Gaussian defect parameters, and setting the surface height of the multilayer film defect as h top The surface half-height width of the multilayer film defect is omega top The height of the defect bottom of the multilayer film is h bot And the half height width at the bottom of the multilayer film defect is omega bot (ii) a Namely the morphological parameter delta (delta = (h) of the multilayer film defect top 、ω top 、h bot 、ω bot ) ); multi-layer film defect top height h top In the range of 0.5-5nm, the value interval is dh top Full width at half maximum omega at the top of the defect top In the range of 35-70nm, the value interval is d omega top Height of defect bottom h bot In the range of 6-20nm, the value interval is dh bot Full width at half maximum of defect bottom omega bot Within the range of 15-50nm, the value interval is d omega bot ；
2) The intensity of a blank mask space image containing the multilayer film defects in different illumination directions is obtained through simulation:
setting n lighting directions, named as l 1 ，l 2 …l n The incident angle of the EUV traditional illumination main ray is theta =6 degrees and theta r The included angle between the main ray of the traditional illumination and the main ray of the new illumination direction is phi, the included angle is an azimuth angle relative to the main ray of the traditional illumination, the photoetching simulation software is utilized to carry out imaging simulation on the blank mask containing the multilayer film defects, and the illumination angles are respectively obtained 1 ，l 2 …l n The intensity of the blank mask space image containing the multilayer film defect is I lr1 ，I lr2 ，…，I lrn Setting an intensity graph of the multilayer film defect blank mask space image as a training set sample; reducing the value interval dh of the top height of the multilayer film defect in the step 1) top Full width at half maximum spacing d ω from the top of the defect top The height of the bottom of the defect is taken as the interval dh bot Full width at half maximum spacing d ω from the bottom of the defect bot Forming defect appearance parameters of another group of multilayer film defect blank masks, and setting an intensity graph of space images of the multilayer film defect blank masks obtained through photoetching simulation software as a test set sample;
3) Training a convolutional neural network with the shape parameter delta of the multilayer film defect as output:
simulating an amplitude image I of a space image of a multi-layer film defect blank mask in the step (2) lr1 ，I lr2 ，…，I lrn The shape parameter delta of the corresponding mask multilayer film defect is used as the output of the convolutional neural network, and a defect shape parameter reconstruction model is constructed by adopting an expanded residual network structure; then, the shape parameter delta of the mask multilayer film defect is used as the input of the neural network, and the amplitude image I of the blank mask space image of the multilayer film defect lr1 ，I lr2 ，…，I lrm As the output of the neural network, a plurality of networks are constructed according to different outputsThe generated countermeasure network forms a plurality of identification models, and the loss function of the defect appearance parameter reconstruction model is changed through the identification models, so that the accuracy of the defect appearance parameter reconstruction model is improved; the number of the identification models influences the improvement of the identification models on the defect appearance parameter reconstruction models. Setting the number of identification models as M, setting a loss function by MSE (mean square error), preprocessing a training set and a test set, training the constructed models by the processed training set, and obtaining the optimal trained models by the test set;
4) Collecting an actually measured space image:
setting the illumination and numerical aperture of the detection system according to the same parameters in the step 1), loading a blank mask to be detected by the detection system, and completing the acquisition of a space image by using a space image sensor;
5) Detecting the defect shape parameters:
inputting the aerial image measured in the step 4) into a model to obtain a defect morphology parameter delta.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211525221.1A CN115795871A (en) | 2022-11-30 | 2022-11-30 | Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211525221.1A CN115795871A (en) | 2022-11-30 | 2022-11-30 | Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115795871A true CN115795871A (en) | 2023-03-14 |
Family
ID=85444161
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211525221.1A Pending CN115795871A (en) | 2022-11-30 | 2022-11-30 | Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115795871A (en) |
-
2022
- 2022-11-30 CN CN202211525221.1A patent/CN115795871A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6342436B2 (en) | Method and apparatus for analyzing and removing defects in EUV photomasks | |
| US11257207B2 (en) | Inspection of reticles using machine learning | |
| CN109031894B (en) | Method for detecting bottom morphology of phase type defect of multilayer film of extreme ultraviolet lithography mask | |
| CN110297401B (en) | Phase type defect detection method for multilayer film of extreme ultraviolet lithography mask | |
| CN107148596B (en) | Method for producing mask in extreme ultraviolet wavelength range, mask and device | |
| TW202026859A (en) | Method of calibrating a plurality of metrology apparatuses, method of determining a parameter of interest, and metrology apparatus | |
| TWI618976B (en) | Qualifying patterns for microlithography | |
| TWI704410B (en) | Method and appliance for predicting the imaging result obtained with a mask when a lithography process is carried out | |
| JP2006250845A (en) | Method and apparatus for inspecting defect of pattern | |
| TW202012919A (en) | Method and apparatus for evaluating an unknown effect of defects of an element of a photolithography process | |
| US9870612B2 (en) | Method for repairing a mask | |
| TW202244607A (en) | Apparatus and method for analysing an element of a photolithography process with the aid of a transformation model | |
| JP6792901B2 (en) | Manufacturing method of reflective mask blank, reflective mask blank, manufacturing method of reflective mask, reflective mask, and manufacturing method of semiconductor device | |
| CN114341741A (en) | Method and apparatus for lithographic process performance determination | |
| CN115795871A (en) | Method for reconstructing multilayer film defect morphology of extreme ultraviolet lithography mask | |
| KR102464254B1 (en) | Method of evaluating surface condition of photomask blank-associated substrate | |
| Pang et al. | EUV multilayer defect compensation (MDC) by absorber pattern modification: from theory to wafer validation | |
| Wang et al. | Preliminary study of photomask pattern inspection by beam-shaped knife-edge interferometry | |
| CN114509921B (en) | Photoetching defect diffraction light intensity difference detection method | |
| WO2024141484A1 (en) | Computer implemented method, computer-readable medium, computer program product and corresponding systems for generating aerial images of photolithography masks | |
| CN111553064A (en) | Feature selection method suitable for optical scattering measurement | |
| Pang et al. | Expanding the applications of computational lithography and inspection (CLI) in mask inspection, metrology, review, and repair |
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 |