Apparatus and method for generating a deblurring model to deblur an image

Abstract

Described herein are methods and systems for deblurring an image of a patterned substrate (e.g., an SEM image) by training the deblurring model using a deblurring model and depth data associated with multiple layers of the patterned substrate. The method includes obtaining a simulated image of the substrate via a simulator using a target pattern as an input, the target pattern including a first target feature formed on a first layer and a second target feature formed on a second layer located below the first layer, determining edge range data for features of the substrate based on the depth data associated with the multiple layers of the substrate, and generating a deblurring model for a deblurred image of the captured image using the simulated image and the edge range data associated with the target pattern as training data to adjust parameters of a base model.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Patent Application No. 63 / 091126, filed October 13, 2020, the entire contents of which are incorporated herein by reference.

[0002]

[0002] The description herein relates generally to processing images acquired by inspection or metrology tools, and more particularly to deblurring images using machine learning. [Background technology]

[0003] Lithographic projection apparatuses may be used, for example, in the manufacture of integrated circuits (ICs). In such cases, a patterning device (e.g., a mask) may contain or be provided with a pattern (a "design layout") that corresponds to an individual layer of the IC, and this pattern may be transferred onto a target portion (e.g., comprising one or more dies) on a substrate (e.g., a silicon wafer) that is coated with a layer of radiation-sensitive material ("resist"), such as by irradiating the target portion through the patterning device. Typically, a single substrate will contain several adjacent target portions (one target portion at a time) onto which the pattern is successively transferred by the lithographic projection apparatus. In one type of lithographic projection apparatus, the pattern on the entire patterning device is transferred onto one target portion at a time; such an apparatus is commonly referred to as a stepper. In an alternative apparatus, commonly referred to as a step-and-scan apparatus, the projection beam is moved parallel to or anti-parallel to a given reference direction (the "scan" direction) in synchronization with scanning the patterning device in this reference direction. Different portions of a pattern on the patterning device are progressively transferred onto one target portion. In general, lithographic projection apparatuses have a de-magnification ratio M (e.g., 4), so that the speed F at which the substrate is moved is 1 / M times the speed at which the projection beam scans the patterning device. More information on lithographic devices as described herein can be found, for example, in U.S. Pat. No. 6,046,792, incorporated herein by reference.

[0004] Before transferring the pattern from the patterning device to the substrate, the substrate may undergo various procedures, such as priming, resist coating, and a soft bake. After exposure, the substrate may undergo other procedures ("post-exposure procedures"), such as a post-exposure bake (PEB), development, a hard bake, and measurement / inspection of the transferred pattern. This multitude of procedures is used as a basis for creating an individual layer of a device, e.g., an IC. The substrate may then undergo various processes, such as etching, ion implantation (doping), metallization, oxidation, chemical-mechanical polishing, etc., all intended to finish off an individual layer of the device. If several layers are required for a device, the entire procedure, or a variant thereof, is repeated for each layer. Eventually, a device is present on each target portion of the substrate. The devices are then separated from each other by techniques such as dicing or sawing, so that the individual devices can be mounted on a carrier, connected to pins, etc.

[0005]

[0005] Thus, manufacturing devices such as semiconductor devices typically involves processing a substrate (e.g., a semiconductor wafer) using multiple fabrication processes to form various features and multiple layers of the device. Such layers and features are typically produced and processed using, for example, deposition, lithography, etching, chemical-mechanical polishing, and ion implantation. Multiple devices may be fabricated on multiple dies on a substrate and then separated into individual devices. This device manufacturing process can be considered a patterning process. The patterning process includes a patterning step, such as optical and / or nanoimprint lithography, that uses a patterning device in a lithography apparatus to transfer a pattern on the patterning device to the substrate, and typically (but optionally) includes one or more associated pattern processing steps, such as developing the resist in a developer, baking the substrate using a bake tool, or etching using the pattern in an etcher. Summary of the Invention

[0006] According to one embodiment, a method for training an image deblurring model for processing an image is provided. The method includes obtaining a simulated image of a substrate corresponding to a target pattern via a simulator using a target pattern from which a pattern on the substrate is derived. The target pattern includes a first target feature formed on a first layer on the substrate and a second target feature formed on a second layer on the substrate, the second layer being located below the first layer on the substrate. The method further includes determining edge range data for features on the substrate corresponding to the first and second target features of the target pattern based on depth data associated with multiple layers of the substrate. The depth data characterizes edge blurring of features on the substrate as a function of the depth of each feature from a top layer of the substrate. The method further includes using the simulated image of the substrate and the edge range data as training data to adjust parameters of a base model and generate a deblurring model. The deblurring model is configured to generate a deblurred image of a captured image of the patterned substrate using the edge range data associated with the captured image.

[0007] In one embodiment, a method for deblurring a captured image of a patterned substrate is provided, the method including identifying features in the captured image based on target pattern and depth data associated with each layer of the patterned substrate, and deblurring the captured image by deblurring each of the features based on the target pattern and depth data.

[0008] In one embodiment, a method for deblurring a captured image of a patterned substrate is provided, the method comprising inputting the captured image of the patterned substrate associated with a target pattern and edge range data into a deblurring model.

[0009] In one embodiment, a system is provided that includes electron beam optics configured to capture an image of a patterned substrate and one or more processors configured to deblur the captured image by inputting the captured image of the patterned substrate and edge range data associated with a target pattern into a deblurring model, the deblurring model being trained based on edge range data associated with features of the target pattern at a particular depth, and running the deblurring model.

[0010] In one embodiment, one or more non-transitory computer-readable media are provided that include instructions corresponding to the processes of the methods herein. In one embodiment, the one or more non-transitory computer-readable media are for storing a deblurring model. In one embodiment, the one or more non-transitory computer-readable media are configured to generate a deblurred image according to the stored deblurring model. In particular, the one or more non-transitory computer-readable media store instructions that, when executed by one or more processors, provide the deblurring model. In one embodiment, the deblurring model is generated by the processes of the methods herein. For example, the process of generating the deblurring model includes obtaining a simulated image of a substrate corresponding to a target pattern via a simulator using a target pattern from which the pattern on the substrate is derived; determining edge range data for features on the substrate corresponding to first and second target features of the target pattern based on depth data associated with multiple layers of the substrate; and using the simulated image of the substrate and the edge range data as training data to adjust parameters of a basic model and generate the deblurring model. The deblurring model is configured to generate a deblurred image of the captured image of the patterned substrate using edge range data associated with the captured image.

[0011] According to embodiments of the present disclosure, a deblurring model is trained by using simulated images converted from design patterns through a simulator (e.g., a Monte Carlo-based simulator) and depth data associated with the design patterns. The training data, including the simulated images and depth data, can collectively cover significantly more patterns than images captured with an SEM. Improved pattern coverage can advantageously significantly improve the effectiveness and accuracy of the deblurring model. The need for retraining can be significantly reduced or even eliminated.

[0012]

[0012] The above aspects and other aspects and features will become apparent to those skilled in the art upon review of the following description of specific embodiments in conjunction with the accompanying drawings. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 shows a block diagram of various subsystems of a lithography system, according to one embodiment. [Figure 2A]

[0014] 1 is a pictorial representation of a target pattern including multiple features (e.g., contact holes), according to one embodiment. [Figure 2B]

[0015] 2B is a blurred image of a patterned substrate where the target pattern (of FIG. 2A) has been printed onto a first layer of the patterned substrate, according to one embodiment. [Figure 2C]

[0015] FIG. 2B is a blurred image of a patterned substrate with the target pattern (of FIG. 2A) printed on a second layer of the patterned substrate, according to one embodiment. [Figure 2D]

[0015] FIG. 2B is a blurred image of a patterned substrate where the target pattern (of FIG. 2A) has been printed on a third layer of the patterned substrate, according to one embodiment. [Figure 2E]

[0015] FIG. 2B is a blurred image of a patterned substrate where the target pattern (of FIG. 2A) has been printed on a fourth layer of the patterned substrate, according to one embodiment. [Figure 2F]

[0016] 2B is a blurred image of a patterned substrate with a target pattern (dotted circle corresponding to FIG. 2A) superimposed on the image, where the blurred image is the fourth layer of the substrate, according to one embodiment. [Figure 3]

[0017] 1 is a flowchart of a method for training a deblurring model, according to one embodiment. [Figure 4]

[0018] 1 is a flowchart of a method for deblurring a captured image of a patterned substrate, according to one embodiment. [Figure 5]

[0019] 10 is a flowchart of another method for deblurring a captured image of a patterned substrate, according to an embodiment. [Figure 6]

[0020] 10 is a flowchart of yet another method for deblurring a captured image of a patterned substrate, according to an embodiment. [Figure 7]

[0021] 1 illustrates the generation of a simulated image from a target pattern, according to one embodiment. [Figure 8A]

[0022] 1 is a two-dimensional image representation of a target pattern patterned onto multiple layers, each layer associated with a different target feature, according to one embodiment. [Figure 8B]

[0023] 8B is a two-dimensional image representation of edge coverage data determined based on depth data corresponding to each feature of the target pattern (of FIG. 8A), according to one embodiment. [Figure 9]

[0024] 1 illustrates the generation of a noisy simulated image according to an embodiment. [Figure 10]

[0025] 4 is a pictorial representation of the structure of a deblurring model constructed following the training in FIG. 3, according to one embodiment. [Figure 11]

[0026] 10 illustrates an example of generating a deblurred image of a simulated image via a deblurring model that uses depth data as input, according to one embodiment. [Figure 12]

[0027] 1 illustrates an example of deblurring a captured image, obtained via a scanning electron microscope, of a patterned multilayer substrate using a deblurring model and depth data, according to one embodiment. [Figure 13]

[0028] 1 schematically depicts an embodiment of a scanning electron microscope (SEM), according to one embodiment. [Figure 14]

[0029] 1 illustrates a schematic representation of an embodiment of an electron beam inspection apparatus, according to one embodiment. [Figure 15]

[0030] FIG. 1 is a block diagram of an exemplary computer system, according to one embodiment. [Figure 16]

[0031] 1 depicts a schematic diagram of a lithographic projection apparatus, according to one embodiment; [Figure 17]

[0032] 1 is a schematic diagram of another lithographic projection apparatus, according to an embodiment; [Figure 18]

[0033] FIG. 17 is a more detailed view of the device of FIG. 16, according to one embodiment. [Figure 19]

[0034] FIG. 19 is a more detailed diagram of the source collector module SO of the apparatus of FIGS. 17 and 18, according to one embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014]

[0035] Integrated circuit (IC) chips used in devices (e.g., phones, laptops, computer memory, etc.) contain complex circuit patterns. During the manufacture of such circuit patterns, it is desirable to capture an image of the circuit pattern printed on the chip (also referred to as a substrate) to determine whether the desired circuit pattern has been printed accurately. Often, the captured image is blurred, making it difficult to identify individual circuit features of the circuit pattern. In one example, deblurring of the captured image occurs because circuit features are formed on multiple layers of chips arranged one above the other. Therefore, the captured image is deblurred for better identification of individual circuit features. The captured image is deblurred by a deblurring model specially trained for such complex applications.

[0015]

[0036] One method for capturing an image of a printed circuit pattern is to use a capture device that projects an electron beam onto the circuit pattern. Electrons from the electron beam are diffracted differently depending on the characteristics of the layer on which the circuit features of the circuit pattern are formed. Herein, depth information for each layer is used to deblur the image. The depth information facilitates accurate modeling of the deblurring effect at different layer depths. For example, the depth information can indicate that features on deeper layers appear blurred compared to the top layer. Therefore, better deblurring of circuit pattern features located at different depths can be achieved, resulting in more accurate and less blurred images, which facilitates accurate identification of individual circuit features.

[0016]

[0037] While specific reference may be made herein to the manufacture of ICs, it should be expressly understood that the description herein has many other possible applications. For example, the description herein may be used in the manufacture of integrated optical systems, guidance and detection patterns for magnetic domain memories, liquid crystal display panels, thin-film magnetic heads, and the like. Those skilled in the art will understand that, in the context of such alternative applications, the use of the terms “reticle,” “wafer,” or “die” herein should be considered interchangeable with the more general terms “mask,” “substrate,” and “target portion,” respectively. Substrates referred to herein may be processed, before or after exposure, in, for example, a track (e.g., a tool that typically applies a layer of resist to a substrate and develops the exposed resist) or a metrology or inspection tool. Where applicable, the disclosure herein may be applied to such and other substrate processing tools. Furthermore, substrates may be processed multiple times, for example to create multi-layer ICs, and the term substrate as used herein may also refer to a substrate that already includes multiple processed layers.

[0017]

[0038] The critical dimension (CD) of a device refers to the smallest width of a line or hole, or the smallest space between two lines or two holes. The CD therefore determines the overall size and density of the designed device. Of course, one of the goals of device fabrication is to faithfully reproduce the original design intent on the substrate (via the patterning device).

[0018]

[0039] In this document, the terms "radiation" and "beam" may be used to encompass all types of electromagnetic radiation, including ultraviolet radiation (e.g., having a wavelength of 365, 248, 193, 157, or 126 nm) and EUV (extreme ultraviolet radiation, e.g., having a wavelength in the range of about 5-100 nm).

[0019]

[0040] The terms "mask" or "patterning device" as used herein may be broadly interpreted as referring to any general patterning device that can be used to impart an incoming radiation beam with a patterned cross section that corresponds to the pattern to be produced in a target portion of a substrate, and the term "light valve" may also be used in this context. In addition to traditional masks (transmissive or reflective; binary, phase-shifting, hybrid, etc.), examples of other such patterning devices include: - a programmable mirror array. An example of such a device is a matrix-addressable surface having a viscoelastic control layer and a reflective surface. The basic principle behind such an apparatus is that (for example) addressed areas of the reflective surface will reflect incident radiation as diffracted radiation, while unaddressed areas will reflect the incident radiation as undiffracted radiation. Using an appropriate filter, the said undiffracted radiation can be filtered out of the reflected beam, leaving only the diffracted radiation behind; in this way the beam is patterned according to the addressing pattern of the matrix-addressable surface. The required matrix addressing can be performed using suitable electronic means. - A programmable LCD array. An example of such a construction is given by US Patent No. 5,229,872, which is incorporated herein.

[0020]

[0041] 1 shows an exemplary lithographic projection apparatus 10A, the main components of which are a radiation source 12A (as discussed herein, the lithographic projection apparatus itself need not have a radiation source), which may be a deep ultraviolet excimer laser source or other type of source, including an extreme ultraviolet (EUV) source, illumination optics, which may include optics 14A, 16Aa, and 16Ab that shape the radiation from source 12A, e.g., to define the partial coherence (denoted as sigma), a patterning device 18A, and transmission optics 16Ac that project an image of the patterning device pattern onto a substrate surface 22A. An adjustable filter or aperture 20A at the pupil plane of the projection optical system may limit the range of beam angles that can impinge on the substrate surface 22A, where the maximum possible angle defines the numerical aperture of the projection optical system, NA=n sin(Θmax), where n is the refractive index of the medium between the substrate and the last element of the projection optical system, and Θmax is the maximum angle of the beam exiting the projection optical system that can still impinge on the substrate surface 22A.

[0021]

[0042] In a lithographic projection apparatus, a source provides illumination (i.e., radiation) to a patterning device, and a projection optical system guides and shapes the illumination through the patterning device and onto a substrate. The projection optical system may include at least some of components 14A, 16Aa, 16Ab, and 16Ac. The aerial image (AI) is the radiation intensity distribution at substrate level. A resist layer on the substrate is exposed, and the aerial image is transferred to the resist layer as a latent "resist image" (RI). The resist image (RI) can be defined as the spatial distribution of resist solubility in the resist layer. A resist model can be used to calculate the resist image from the aerial image, examples of which can be found in U.S. Patent Application Publication No. 2009 / 0157360, the disclosure of which is incorporated herein by reference in its entirety. The resist model is only concerned with the properties of the resist layer (e.g., the effects of chemical processes occurring during exposure, PEB, and development). The optical properties of a lithographic projection apparatus (e.g. properties of the source, patterning device, and projection optics) determine the aerial image. Because the patterning devices used in a lithographic projection apparatus can be varied, it may be desirable to decouple the optical properties of the patterning device from the optical properties of the rest of the lithographic projection apparatus, which includes at least the source and projection optics.

[0022]

[0043] In chip manufacturing (also referred to as a patterning process or semiconductor manufacturing), for example, a target pattern (e.g., a circuit pattern) for a chip may be printed onto multiple layers on a substrate. Each layer has specific features (e.g., lines, contact holes, bars, etc.) of the target pattern (e.g., the circuit pattern). These features, when connected to each other, provide the desired electrical or logical functionality of the chip or circuit. The features may be connected to each other through multiple layers to form the target pattern for the chip.

[0023]

[0044] Patterned substrates are often inspected or measured using a scanning electron microscope (SEM) to examine the printing performance of the target pattern. The SEM image is inspected to ensure that the target pattern of chips is printed on the substrate within acceptable threshold limits for physical properties associated with the features. If the printing performance is not within acceptable threshold limits, the yield of chip manufacturing is adversely affected. Based on the inspection, one or more processes of chip manufacturing can be adjusted to improve the printing performance and the yield of chip manufacturing.

[0024]

[0045] Typically, a high landing energy (HLE) setting on an SEM tool is used to measure features at high depths (e.g., in the third or fourth layer located vertically below the top layer) for inspection and metrology purposes. The HLE SEM image signal is primarily derived from backscattered electrons (BSE). The BSE signal causes a lot of blurring in the SEM image. For example, based on the diffraction properties of BSE, features on the bottom layer (e.g., at the greatest depth from the top surface of the substrate) have the greatest diffraction effect in the SEM image, which can be seen in the form of edge blurring. Therefore, a deblurring algorithm can be used to obtain a clearer SEM image. For example, a deblurred image is an image in which some features or feature boundaries are blurred. On the other hand, after deblurring the image, the feature boundaries have sharper or more defined edges.

[0025]

[0046] 2A-2F, examples of blurring due to BSE diffraction effects on different layers of a substrate are shown. FIG. 2A shows a target pattern TP1 (e.g., in GDS format) including multiple features, such as contact holes CH (only a few holes are labeled for illustrative purposes). FIG. 2B shows a first pattern L1 corresponding to the target pattern TP1. In this example, the first pattern L1 is printed on a first layer (e.g., the top layer) of the substrate. The first pattern L1 may be formed at a first depth (e.g., 0 nm) in the substrate. The boundaries of the contact holes in the first pattern L1 are sharp. For example, when the contact holes in TP1 are overlaid on the first pattern L1, the intensity of the pixels (within L1) at the boundaries of the holes shows minimal or no change. Therefore, the contact holes in the first pattern L1 can be accurately identified. However, as the depth of layers on the substrate increases (e.g., vertically below the top layer), the edges of features (e.g., contact holes) become increasingly blurred. For example, pixel intensity at the edges of features varies significantly, making it difficult to accurately discern the boundaries of features in deeper layers. Exemplary blurring of deep features is shown in Figures 2C, 2D, and 2E.

[0026]

[0047] FIG. 2C shows a second pattern L2 formed on a second layer L2 at a second depth (e.g., 2 nm vertically below the first layer). It can be seen that the edges of the features (black circles) are blurred compared to the edges of the features in the first pattern L1. FIG. 2D shows a third pattern L3 formed on a third layer L3 at a third depth (e.g., 5 nm vertically below the first layer). It can be seen that the edges of the features (black circles) are blurred compared to the edges of the features in the second pattern L2. Similarly, FIG. 2D shows a fourth pattern L4 formed on a fourth layer L3 at a fourth depth (e.g., 8 nm vertically below the first layer). It can be seen that the edges of the features (black circles) are blurred compared to the edges of the features in the third pattern L3. Therefore, referring to the image L4-TP1 in FIG. 2E, when the target pattern TP1 is superimposed on the fourth pattern L4, it is unclear where the edges of the features in L4 actually are.

[0027]

[0048] Currently, methods for deblurring and denoising HLE SEM images have several challenges. For example, feature boundaries in deblurred HLE SEM images are unclear and inaccurate. Another challenge is aliasing, which refers to image distortion that occurs when the signal reconstructed from the sample differs from the original signal, resulting in inaccurate edge determination. In SEM images of patterned multilayer substrates, BSE causes diffraction at a first depth in a first layer that differs from diffraction at a second depth in a second layer, which leads to more complex edge blurring. Traditional Monte Carlo simulations used to analyze diffraction effects are time-consuming. Because it is difficult to obtain clean and accurate HLE SEM images, ground truth images cannot be used to improve one or more aspects of the patterning process, such as to increase the accuracy of process models related to determining feature CD.

[0028]

[0049] The present disclosure provides a method for deblurring a captured image based on depth data that accounts for the deblurring of the captured image. In one embodiment, a deblurring model is trained using the effect of feature depth from a top layer of a patterned substrate. For example, the effect of feature depth can be seen at the edge of the feature. As the depth of a feature from the top layer increases, the edge of the feature becomes more blurred due to the weaker BSE signal compared to the BSE signal from the top layer.

[0029]

[0050] In one embodiment, the trained deblurring model can be used in one or more lithography and metrology applications. For example, a captured image of an actual patterned substrate having memory circuit features can be deblurred, for example, via the deblurring model. From the deblurred image, the memory circuit features can be identified and accurately extracted. Such extracted features can be further provided to train more accurate process models related to lithography (e.g., resist models, optical proximity correction (OPC)-related models, etc.). In another application, the accuracy of metrology data can be improved because more accurate feature geometries (e.g., shape and size) are extracted from the deblurred image of the patterned substrate. Furthermore, the deblurred image can improve the reliability of inspection of the patterned substrate.

[0030]

[0051] 3 is a flowchart of a method 300 for training a deblurring model configured to deblur a captured image of a patterned substrate. To better understand the method, this discussion uses feature edges as exemplary characteristics that are affected due to, for example, the depth of a feature from a top layer. Deblurring the captured image allows for more accurate identification of feature characteristics (e.g., feature edges). In one embodiment, the process of training a deblurring model is discussed in further detail with respect to steps P301, P303, and P305 below.

[0031]

[0052] Process P301 includes obtaining a simulated image 302 of a substrate corresponding to a target pattern TP via a simulator using the target pattern TP from which the pattern on the substrate is derived. In one embodiment, the target pattern TP may be formed on multiple layers of the substrate arranged vertically one above the other, each layer having one or more features corresponding to target features of the target pattern. For example, a first target feature is formed on a first layer of the substrate, and a second target feature is formed on a second layer located below the first layer of the substrate.

[0032]

[0053] In one embodiment, the target pattern TP is presented in the form of a polygon-based hierarchical data format. For example, the polygon-based data format may be a graphic data system (GDS) format, a color image, a vector format, or other data representation. In one embodiment, the target pattern TP includes geometric data associated with a first target feature and a second target feature. The geometric data may be, for example, a desired shape of the first target feature and the second target feature, respectively, and a target position of the edge of the first target feature and the second target feature within the target pattern TP. In one embodiment, each target feature is associated with a layer of a substrate on which the each target feature may be formed.

[0033]

[0054] In one embodiment, the simulated image 302 includes noise, such as variations in the brightness or color information of the image. In one embodiment, obtaining the simulated image 302 further includes extracting noise from a captured image of the patterned substrate and adding noise to the simulated image 302. The noise in the captured image can be, for example, inherent to the capture device.

[0034]

[0055] In one embodiment, the simulator includes a model that is a function of the physical properties of the substrate on which the target pattern TP is printed. The simulator is configured to generate the simulated image 302 by adjusting the BSE characteristics of the electron beam of a capture device (e.g., SEM). Adjusting the BSE characteristics can take into account different diffraction effects associated with the target pattern, thereby enabling feature characteristics (edges, size, grayscale values, etc.) in each layer of the substrate to be determined. In one embodiment, the physical properties of the substrate can be the material of a particular layer, the thickness of a particular layer, resist process parameters, etch process parameters, or a combination thereof. The model used in the simulator can be a physics-based model that models the physical properties of diffraction, electron absorption / diffraction by a particular layer, etc. The model can be a statistical model fitted based on BSE data.

[0035]

[0056] In one embodiment, obtaining the simulated image 302 includes simulating a model to generate the simulated image 302 through a Monte Carlo simulation process. During the Monte Carlo simulation process, one or more BSE characteristics, layer depths, etc., can be adjusted to generate the simulated image 302. The simulated image 302 can represent a real image of the patterned substrate captured via a capture device (e.g., an SEM). In one embodiment, obtaining the simulated image 302 includes simulating a bi-Gaussian model configured to generate the simulated image 302 based on the depth data.

[0036]

[0057] 7 shows an example of obtaining a simulated image from a simulator using a target pattern. An exemplary target pattern 702 (e.g., a DRAM circuit pattern) includes multiple features formed on different layers on a substrate. The target pattern 702 includes a first target feature F1 (e.g., a horizontal line) formed on the first layer, a second target feature F2 (e.g., a vertical line) formed on the second layer, a third feature F3 (e.g., a slanted bar) formed on the third layer, and a fourth feature F4 (e.g., a circle) formed on the fourth layer.

[0037]

[0058] In one embodiment, the target pattern 702 is input to a simulator 704 configured to generate a simulated image that mimics an actual patterned substrate. For example, the simulator 704 includes a Monte Carlo simulation for a model. The model generates a simulated image of the substrate as a function of diffraction effects at different layers of the substrate. For example, in one embodiment, the model can be configured to assign / vary the number of electrons projected onto the substrate. For example, the number of electrons can be about 2000, about 2500, about 3000, etc. The model predicts that when the electrons hit a feature in the first layer, there can be about 140 backscattered electrons. Additionally, the model can be configured to input the electron landing energy (e.g., 10 keV, 20 keV, etc.) and the electron density (e.g., about 2.4). Beyond these exemplary settings, the simulator 704 can be configured to include additional characteristics that affect the diffraction of electrons after hitting the target feature.

[0038]

[0059] In one embodiment, simulator 704 outputs a simulated image 706 that mimics an image of an actual patterned substrate. In one embodiment, simulated image 706 is a pixelated grayscale image, with each pixel intensity being a result of simulator 704 predicting a signal, for example, from a BSE. Because the simulated image is used as a training data set, the deblurring model can be trained to cover a large number of patterns, for example, compared to a training data set including actual SEM images. Capturing a large number of SEM images can damage the patterned substrate, affecting the yield of the patterning process and slowing down semiconductor manufacturing.

[0039]

[0060] FIG. 9 illustrates another example of generating a simulated image. In this example, a simulated image 706 (generated through a simulator (e.g., Monte Carlo simulation), as discussed above) is modified by adding image noise 906. For example, the image noise 906 can be extracted from the actual SEM image or statically generated as random noise. In one embodiment, the noise 906 can be extracted by applying a noise filter to the actual SEM image. Adding noise to the simulated image generates a noisy simulated image 706′. In one embodiment, the noisy simulated image 706′ may also be referred to as the simulated image 706. In this disclosure, the noisy simulated image 706′ may be interchangeably referred to as the simulated image 706.

[0040]

[0061] In one embodiment, the use of simulated images 706′ (which contain noise) as training data improves the robustness of the trained deblurring model. For example, the trained deblurring model produces reliable results even when the input image (e.g., a real SEM image) is significantly noisy. For example, the noisy image may be due to poor acquisition conditions, measurement settings, etc.

[0041]

[0062] A process P303 determines edge range data ERD for features on the substrate corresponding to the first and second target features of the target pattern TP based on depth data DEP associated with each layer of the substrate, the depth data DEP characterizing the blurring of the edges of features on the substrate as a function of the depth of the respective feature from the top layer of the substrate.

[0042]

[0063] In one embodiment, the depth data DEP comprises at least one of the depth of the feature from a top layer of the substrate, a material used in a particular layer of the substrate, and a diffraction behavior of the feature at a given depth. For example, the depth data DEP may comprise a first depth of a first layer located vertically downward from the top layer of the substrate and a second depth of a second layer located vertically downward from the top layer, material data associated with the first and second layers of the substrate, or diffraction behavior associated with first and second features on the substrate corresponding to first and second target features, respectively, of the target pattern TP.

[0043]

[0064] In one embodiment, the edge range data ERD includes a range of edge positions for features on the substrate corresponding to each feature in the target pattern TP. In one embodiment, determining the edge range data ERD includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern TP to be printed on the substrate. In one embodiment, the edge range data ERD is represented as an image of the target pattern TP complemented with the edge range data ERD of each feature of the target pattern TP.

[0044]

[0065] 8A and 8B illustrate an exemplary target pattern and corresponding edge coverage data associated with the depth of features within the target pattern. Referring to FIG. 8A, an exemplary target pattern 702 (e.g., a DRAM circuit pattern) includes multiple features formed on different layers on a substrate. As previously discussed, the target pattern 702 includes a first target feature F1 (e.g., a horizontal line) formed on a first layer of the substrate, a second target feature F2 (e.g., a vertical line) formed on a second layer of the substrate, a third feature F3 (e.g., a slanted bar) formed on a third layer of the substrate, and a fourth feature F4 (e.g., a circle) formed on a fourth layer of the substrate.

[0045]

[0066] In this example, the first layer may be considered the top layer. The second layer is a layer located at a depth (e.g., 5 nm) vertically below the first layer. The third layer is a layer located at a depth (e.g., 10 nm) vertically below the first layer. The fourth layer is a layer located at a depth (e.g., 15 nm) vertically below the first layer. Thus, features F1, F2, F3, and F4 are printed on respective layers of the substrate. When an image of the patterned substrate is captured, the image of the bottom layer (e.g., having feature F4 at the greatest depth) will have the greatest diffraction effect (e.g., the edge of feature F4 will be more blurred) due to the characteristics of BSE backscattered electrons. In one embodiment, the change in characteristics is related to such depth data, for example, based on the depth of each feature (e.g., F1, F2, F3, F4). Based on the depth data of each feature, edge range data for each feature may be determined. The edge range data characterizes the range over which the edge of the feature may be located on the substrate.

[0046]

[0067] 8B is a pictorial representation of edge range data 702ER generated for target pattern 702. In edge range data 702ER, the thickness of each feature represents the extent of that feature. For example, the thicknesses of line E1 (corresponding to feature F1), line E2 (corresponding to feature F2), line E3 (corresponding to feature F3), and line E4 (corresponding to feature F4) indicate the amount that diffraction at the edges of features F1, F2, F3, and F4 is affected due to depth data. For example, line E4 is associated with the largest change in the BSE diffraction signal. If such edge range data E1, E2, E3, and E4 are used as a training data set, more accurate determination of feature edges can be made in highly blurred areas.

[0047]

[0068] In one embodiment, the edge range data 702ER may be determined using a model (e.g., similar to that used in simulator 704 discussed with respect to FIG. 7 ). The model is configured to determine edge range data for features on a substrate based on depth data, such as material data for each layer, diffraction properties associated with the material and depth, etc. For example, the material of feature F1 is chromium, the material of feature F2 is copper, the material of feature F3 is tungsten, and the material of feature F4 is a polyresin material. The materials discussed herein are merely exemplary and do not limit the scope of the present disclosure. The layer materials may be SiO , Si, or other materials used in semiconductor manufacturing. In one embodiment, the model may be a function (or mapping) between edge range, layer depth, and layer material. Such a function (or mapping) may be established, for example, by correlating actual data of features, materials, and depths of a patterned substrate.

[0048]

[0069] A process P305 uses the simulated image 302 and edge range data ERD for features on the substrate corresponding to the target pattern TP as training data to adjust parameters of the base model to generate a deblurring model DBM. The deblurring model DBM uses the edge range data ERD associated with the captured image to generate a deblurred image of the captured image of the patterned substrate. This disclosure is not limited to a particular type of model, as long as the model is configured to receive the edge range data and the captured image as inputs and generate a deblurred image of the captured image.

[0049]

[0070] In one embodiment, the base model or deblurring model DBM is a machine learning model that includes weights and biases as model parameters. During the training process, the weights and biases of the base model are continuously updated based on training data. At the end of training, the base model is referred to as a deblurring model. In one embodiment, the deblurring model DBM is a convolutional neural network (e.g., CNN) or a deep convolutional network (e.g., DCNN). The model parameters include weights and biases of one or more layers of the deep convolutional network.

[0050]

[0071] In one embodiment, the deblurring model DBM is a neural network that includes a feature vector containing values ​​indicative of features of the target pattern TP, and an attention vector containing values ​​related to depths associated with the features to be printed on the substrate, the attention vector being configured to operate on the feature vector. In one embodiment, the attention vector includes any value in the range 0 to 1, or the binary values ​​0 or 1.

[0051]

[0072] In one embodiment, the machine learning model structure is configured to multiply the attention vector by the feature vector such that a depth-related weight is assigned by the attention vector to edges of particular features in the feature vector, e.g., a first target feature associated with a first layer of the substrate is multiplied with a higher attention value compared to an attention value multiplied with a second target feature associated with a second layer of the substrate.

[0052]

[0073] 10 is a pictorial representation of an exemplary structure of the deblurring model DBM configured to receive an image (e.g., an SEM image or a simulated image of a target pattern TP) and depth-related data (e.g., edge range data ERD). As shown, each input may be represented as a feature vector FV. For example, the feature vector may be a vector representation of a captured image (e.g., a simulated image or an actual SEM image). In one embodiment, the feature vector FV includes information such as pixel location and intensity. In one embodiment, the intensity values ​​indicate features of the target pattern (e.g., TP). Optionally, another layer is provided that includes statistical information of the feature vectors.

[0053]

[0074] In one embodiment, the deblurring model also includes an attention vector AV, which includes values ​​related to depths associated with features to be printed on the substrate. In this example, for a layer, the attention vector includes any value ranging from 0 to 1 (or other values, e.g., 1 to 100). For example, in a top-down view, the first depth is assigned a value of 0.95, the second depth is assigned a value of 0.8, the third depth is assigned a value of 0.6, and the fourth depth is assigned a value of 0.5. In one embodiment, edge range data (e.g., ERD) is used to generate attention values ​​(or weights), which are then applied to the feature maps (e.g., the transformed SEM images in a particular layer of the CNN).

[0054]

[0075] In one embodiment, a feature vector of a simulated or captured image is computed with a focus vector AV (e.g., generated from edge range data ERD). The feature vector FV is then transformed into another vector FVX. As a result, the transformed feature vector FVX has the described depth data.

[0055]

[0076] 11 shows an example of training the deblurring model DBM using a simulated image 706′ (which is noisy) and edge extent data 702ER associated with a target pattern TP. The deblurring model DBM generates a deblurred image 1101 with clearly defined boundaries of features in the target pattern TP (not shown in FIG. 11). Because the boundaries of the features can be clearly identified, the deblurring model DBM is considered trained.

[0056]

[0077] 4 is a flowchart of a method 400 for deblurring a captured image of a patterned substrate. For example, a metrology tool (e.g., an SEM imager) may be used to capture the image of the patterned substrate. An exemplary implementation of the method includes processes P401 and P403. The processes may be executed via one or more processors of a computer system discussed herein. In one embodiment, the processes are implemented in a processor of a metrology tool (e.g., an SEM tool).

[0057]

[0078] Process P401 includes inputting a captured image CI of a patterned substrate and edge range data ERD associated with a target pattern into a deblurring model DBM. As discussed above, the deblurring model DBM is trained based on edge range data ERD associated with features of the target pattern at a particular depth.

[0058]

[0079] As discussed above, edge range data ERD for features on the substrate corresponding to features of the target pattern TP is determined based on depth data associated with the target pattern and the substrate, including a first depth of a first layer located vertically downward from a top layer of the substrate and a second depth of a second layer located vertically downward from the top layer, material data associated with the first and second layers of the substrate, diffraction behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern TP, respectively, or a combination thereof.

[0059]

[0080] In one embodiment, the edge range data ERD includes a range of edge positions for features on the substrate corresponding to each feature in the target pattern. In one embodiment, determining the edge range data ERD includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. In one embodiment, the edge range data ERD is represented as an image of the target pattern complemented with the edge range data ERD of each feature of the target pattern.

[0060]

[0081] Process P403 includes deblurring the captured image CI by running a deblurring model, and the resulting deblurred image DBCI has sharper edges of features even in deeper layers (e.g., 10 nm, 20 nm, 30 nm deep from the top of the substrate).

[0061]

[0082] 12 illustrates an example method for deblurring a captured actual SEM image. Deblurring involves inputting a captured image SEM1 of a patterned substrate associated with a target pattern (e.g., using a TP to pattern the substrate) and edge range data 702ER into a deblurring model DBM. As discussed above, the deblurring model DBM is trained based on edge range data ERD associated with features of the target pattern at a particular depth. The captured image is then deblurred by running the deblurring model DBM. The deblurring model DBM outputs a deblurred image dbSEM1.

[0062]

[0083] 5 is another flow chart of a method 500 for deblurring a captured image of a patterned substrate. In this embodiment, the method includes processes P501, P503, P505, and P507, as discussed below.

[0063]

[0084] Process P501 includes receiving a captured image CI of a patterned substrate. For example, the captured image CI can be received via a metrology tool (e.g., FIGS. 13 and 14). The patterned substrate can be a patterned multilayer substrate. Therefore, a high landing energy can be used to capture an image that shows the multilayer pattern in the image. As discussed herein, the captured image CI is blurred because backscattered electrons have a weak signal for deep layers (e.g., 5 nm, 10 nm, 20 nm deep, the bottom layer, etc.). For example, features on deep layers appear blurred in the captured image CI. Therefore, deblurring of the captured image CI is performed.

[0064]

[0085] Process P503 includes inputting the captured image CI into a deblurring model. The deblurring model is trained by processes P301, P303, and P305 (of FIG. 3). Process P505 includes receiving output from the deblurring model using the captured image CI and edge range data for features of the patterned substrate. The output is a deblurred captured image CI. Process P507 includes generating a deblurred version of the captured image CI on a user interface for display. In one embodiment, the deblurred version of the captured image CI may be generated as a vector that can be input into other models related to the patterning process. Thus, the deblurred version of the captured image CI can be used to improve one or more aspects of the patterning process, such as improving OPC, dose, focus, etc.

[0065]

[0086] 6 is another flow chart of a method 600 for deblurring a captured image of a patterned substrate. In this embodiment, the method includes processes P601 and P603, as discussed below.

[0066]

[0087] Process P601 includes identifying features in the captured image CI based on the target pattern TP6 and depth data DD associated with each layer of the substrate. In one embodiment, the depth data includes data such as depth, material, and diffraction characteristics at a particular depth in the substrate. In one embodiment, the depth data DD includes feature characteristics derived from the depth data DD. For example, the feature characteristics may be the edge extent of each feature in a particular layer. Process P603 includes deblurring the captured image CI by deblurring each of the features based on the target pattern TP6 and the depth data DD6. After deblurring, a deblurred image DBCI of the captured image CI is obtained. For example, FIG. 12 shows an example of a deblurred image dbSEM1.

[0067]

[0088] As discussed herein, an example of a deblurring model is a machine learning model. Both unsupervised and supervised machine learning models can be used to generate deblurred images from input noisy images, such as SEM images of patterned substrates. Without limiting the scope of the present invention, the application of supervised machine learning algorithms is described below.

[0068]

[0089] Supervised learning is the machine learning task of inferring a function from labeled training data. The training data includes a set of training examples. In supervised learning, each example is a pair of an input object (typically a vector) and a desired output value (also called a supervisory signal). A supervised learning algorithm analyzes the training data and generates an inferred function that can be used to map new examples. The optimal scenario would allow the algorithm to correctly determine the class label of unseen instances. This requires the learning algorithm to generalize from the training data to unseen situations in a "reasonable way."

[0069]

[0090] x i is the feature vector of the i-th example, and y i is its label (i.e., class) {(x1,y1), (x2,y2), ..., (x N ,y NGiven a set of N training examples of the form {\displaystyle \mathbb {X}}, a learning algorithm finds a function g:X→Y, where X is the input space and Y is the output space. A feature vector is an n-dimensional vector of numerical features that represent some object. Many algorithms in machine learning require a numerical representation of an object because it facilitates processing and statistical analysis. When representing an image, the feature values ​​might correspond to the pixels of the image; when representing text, the feature values ​​might correspond perhaps to term occurrence frequencies. The vector space associated with these vectors is often called the feature space. The function g is an element of some space of possible functions G, usually called the hypothesis space. g is the function that finds the highest score:

number

number

[0070]

[0091] Although G and F can be any space of functions, many learning algorithms are probabilistic models where g takes the form of a conditional probability model g(x) = P(y|x) or f takes the form of a joint probability model f(x,y) = P(x,y). For example, Naive Bayes and Linear Discriminant Analysis are joint probability models, and logistic regression is a conditional probability model.

[0071]

[0092] There are two basic approaches to selecting f or g: empirical risk minimization and structural risk minimization. Empirical risk minimization finds the function that best fits the training data. Structural risk minimization includes a penalty function that controls the bias / variance tradeoff.

[0072]

[0093] In both cases, the training set consists of independent and identical distribution pairs (x i ,y i ) samples. To measure how well the function fits the training data, we use a loss function L:

number

number

number

[0073]

[0094] The risk R(g) of a function g is defined as the expected loss of g.

number

[0074]

[0095] Exemplary models of supervised learning include decision trees, ensemble methods (bagging, boosting, random forests), k-NN, linear regression, naive Bayes, neural networks, logistic regression, perception, support vector machines (SVM), relevance vector machines (RVM), and deep learning.

[0075]

[0096] SVM is an example of a supervised learning model that can analyze data, recognize patterns, and be used for classification and regression analysis. Given a set of training examples, each marked as belonging to one of two categories, an SVM training algorithm builds a model that assigns new examples to one or the other category, making it a non-probabilistic binary linear classifier. An SVM model is a representation of examples as points in space, mapped so that examples from different categories are separated by as wide an apparent gap as possible. New examples are then mapped into the same space and predicted to belong to a category based on which side of the gap they fall on.

[0076]

[0097] In addition to performing linear classification, SVMs can efficiently perform non-linear classification using so-called kernel methods to implicitly map their inputs into a high-dimensional feature space.

[0077]

[0098] Kernel methods require a user-specified kernel, i.e., a similarity function for pairs of data points in the raw representation. Kernel methods get their name from the use of kernel functions, which allow them to operate in a high-dimensional implicit feature space without ever computing the coordinates of the data in that space, but simply by computing the dot product between the images of all pairs of data in the feature space. This operation is often computationally cheaper than explicitly computing the coordinates. This technique is called the "kernel trick."

[0078]

[0099] The effectiveness of an SVM depends on the choice of kernel, the kernel parameters, and the soft margin parameter C. A common choice is a Gaussian kernel with a single parameter γ. The best combination of C and γ is often an exponentially increasing sequence of C and γ (e.g., C∈{2 -5 , 2 -4 , …, 2 15 , 2 16};γ∈{2 -15 , 2 -14 , …, 24 , 2 5}) by grid search (also known as "parameter sweep").

[0079]

[0100] Grid search is an exhaustive search through a manually specified subset of a learning algorithm's hyperparameter space. Grid search algorithms are guided by some performance metric, typically measured by cross-validation on a training set or evaluation on a provided validation set.

[0080]

[0101] Each combination of parameter selection may be checked using cross-validation, and the parameters with the best cross-validation accuracy are chosen.

[0081]

[0102] Cross-validation, sometimes called rotational estimation, is a model validation technique that evaluates how the results of a statistical analysis generalize to an independent dataset. It is primarily used in situations where the goal is prediction and one wants to estimate how accurately a predictive model will perform in practice. In prediction problems, a model is typically given a dataset of known data (a training dataset) on which training is performed and a dataset of unknown (or unseen) data (a test dataset) on which the model is tested. The goal of cross-validation is to define a dataset (i.e., a validation dataset) to "test" the model during the training phase, to limit problems like overfitting and provide insight into how the model generalizes to an independent dataset (i.e., an unknown dataset from a real problem, for example). One round of cross-validation involves partitioning data samples into complementary subsets, conducting an analysis on one subset (called the training set), and validating the analysis on the other subset (called the validation or test set). To reduce variability, multiple rounds of cross-validation are performed using different partitions, and the validation results across the rounds are averaged.

[0082]

[0103] A final model that can be used for testing and to classify new data is then trained on the entire training set using the selected parameters.

[0083]

[0104] Another example of supervised learning is regression. Regression infers the relationship between a dependent variable and one or more independent variables from a set of values ​​of the dependent variable and corresponding values ​​of the independent variables. Regression may estimate the conditional expectation of the dependent variable given the independent variables. The inferred relationship is sometimes called a regression function. The inferred relationship may be probabilistic.

[0084]

[0105] In one embodiment, a system is provided that can generate a deblurred image using the model DBM after the system captures an image of a patterned substrate. In one embodiment, the system can be, for example, the SEM tool of FIG. 13 or the inspection tool of FIG. 14 configured to include the model DBM discussed herein. For example, the metrology tool includes an electron beam generator for capturing an image of the patterned substrate and one or more processors that include the deblurring model. The one or more processors are configured to execute the trained model using the captured image and depth data (e.g., edge coverage data) of the target pattern as inputs to generate a deblurred image of the captured image. As previously mentioned, the deblurring model DBM can be a convolutional neural network.

[0085]

[0106] Additionally, in one embodiment, the one or more processors may be further configured to update the deblurring model based on the captured image of the patterned substrate, hi one embodiment, updating the deblurring model includes running the deblurring model using the captured image to generate a deblurred image, and updating one or more parameters of the deblurring model based on a comparison of the deblurred image with a reference deblurred image.

[0086]

[0107] In one embodiment, the deblurred image can be used to improve the patterning process. The deblurred image can be used in a simulation of the patterning process, for example, to predict contours, CDs, edge placement (e.g., edge placement errors), etc. in the resist and / or etched image. The goal of the simulation is to accurately predict, for example, the edge placement and / or aerial image intensity gradient and / or CD of the printed pattern. These values ​​can be compared to the intended design, for example, to correct the patterning process, identify where defects are expected to occur, etc. The intended design is typically defined as a pre-OPC design layout and can be provided in a standard digital file format such as GDSII or OASIS, or other file format.

[0087]

[0108] In some embodiments, the inspection or metrology tool may be a scanning electron microscope (SEM) that obtains an image of the exposed or transferred structure on the substrate (e.g., some or all of the structure of a device). Figure 13 depicts an embodiment of an SEM tool. A primary electron beam EBP emitted from an electron source ESO is focused by a condenser lens CL and then passes through a beam deflector EBD1, an ExB deflector EBD2, and an objective lens OL to illuminate at a focal point a substrate PSub on a substrate table ST.

[0088]

[0109] When the substrate PSub is irradiated with the electron beam EBP, secondary electrons are generated from the substrate PSub. The secondary electrons are deflected by the E×B deflector EBD2 and detected by the secondary electron detector SED. For example, a two-dimensional electron beam image can be obtained by detecting the electrons generated from the sample in synchronization with two-dimensional scanning of the electron beam by the beam deflector EBD1, or with repeated scanning of the electron beam EBP by the beam deflector EBD1 in the X or Y direction together with continuous movement of the substrate PSub by the substrate table ST in the other of the X or Y directions.

[0089]

[0110] The signal detected by the secondary electron detector SED is converted into a digital signal by an analog-to-digital (A / D) converter ADC, and the digital signal is sent to an image processing system IPU. In one embodiment, the image processing system IPU may have a memory MEM for storing all or part of the digital image for processing by the processing unit PU. The processing unit PU (e.g., specially designed hardware or a combination of hardware and software) is configured to convert or process the digital image into a data set representative of the digital image. Furthermore, the image processing system IPU may have a storage medium STOR configured to store the digital image and the corresponding data set in a reference database. A display device DIS may be connected to the image processing system IPU, so that an operator can perform the required operations of the equipment with the aid of a graphical user interface.

[0090]

[0111] As described above, SEM images can be processed to extract contours that depict the edges of objects in the image, representing device structures. These contours are quantified using metrics such as CD. Therefore, images of device structures are typically compared and quantified using simple metrics, such as edge-to-edge distance (CD) or simple pixel differences between images. Typical contour models for detecting object edges in images to measure CD use image gradients. Indeed, these models rely on strong image gradients. However, in reality, images typically have noisy and discontinuous boundaries. Techniques such as smoothing, adaptive thresholding, edge detection, erosion, and dilation can be used to process the results of image gradient contour models to address noisy and discontinuous images, but these techniques ultimately result in low-resolution quantification of high-resolution images. Therefore, mathematical manipulation of device structure images to reduce noise and automate edge detection often leads to a loss of image resolution, thereby resulting in a loss of information. The result, therefore, is low-resolution quantification, resulting in a simplified representation of complex, high-resolution structures.

[0091]

[0112] Thus, it is desirable to have a mathematical representation of a structure (e.g., a circuit feature, alignment mark, or metrology target portion (e.g., grating feature), etc.) that is produced or expected to be produced using a patterning process, whether the structure is in a latent resist image, in a developed resist image, or transferred, e.g., by etching, to a layer on a substrate that maintains resolution and yet can represent the general shape of the structure. In the context of lithography or other patterning processes, the structure may be a device or portion thereof under fabrication, and the image may be an SEM image of the structure. In some cases, the structure may be a feature of a semiconductor device, e.g., an integrated circuit. In this case, the structure may be referred to as a pattern including multiple features of the semiconductor device, or as a desired pattern. In some cases, the structure may be an alignment mark, or a portion thereof (e.g., a grating of an alignment mark), used in an alignment measurement process to determine the alignment of an object (e.g., a substrate) with another object (e.g., a patterning device), or may be a metrology target, or a portion thereof (e.g., a grating of a metrology target), used to measure a parameter of the patterning process (e.g., overlay, focus, dose, etc.) In one embodiment, the metrology target is a diffraction grating, for example used to measure overlay.

[0092]

[0113] 14 schematically illustrates a further embodiment of an inspection apparatus. The system is used to inspect a sample 90 (such as a substrate) on a sample stage 88 and includes a charged particle beam generator 81, a condenser lens module 82, a probe-forming objective lens module 83, a charged particle beam deflection module 84, a secondary charged particle detector module 85, and an image formation module 86.

[0093]

[0114] The charged particle beam generator 81 generates a primary charged particle beam 91. The condenser lens module 82 collects the generated primary charged particle beam 91. The probe forming objective lens module 83 focuses the collected primary charged particle beam into a charged particle beam probe 92. The charged particle beam deflection module 84 scans the formed charged particle beam probe 92 across the surface of an area of ​​interest on a sample 90 fixed on a sample stage 88. In one embodiment, the charged particle beam generator 81, the condenser lens module 82, and the probe forming objective lens module 83, or equivalent designs, alternatives, or any combination thereof, together form a charged particle beam probe generator that generates the scanning charged particle beam probe 92.

[0094]

[0115] The secondary charged particle detector module 85 detects secondary charged particles 93 emitted from the sample surface (possibly along with other charged particles reflected or scattered from the sample surface) and generates a secondary charged particle detection signal 94 upon being struck by the charged particle beam probe 92. The imaging module 86 (e.g., a computing device) is coupled to the secondary charged particle detector module 85 to receive the secondary charged particle detection signal 94 from the secondary charged particle detector module 85 and, in response, form at least one scanned image. In one embodiment, the secondary charged particle detector module 85 and the imaging module 86, or equivalent designs, alternatives, or any combination thereof, together form an imaging device that forms a scanned image from the detected secondary charged particles emitted from the sample 90 struck by the charged particle beam probe 92.

[0095]

[0116] In one embodiment, the monitoring module 87 is coupled to the imaging module 86 of the imaging device to use scanned images of the sample 90 received from the imaging module 86 to monitor and control the patterning process and / or to derive parameters for designing, controlling, monitoring, etc., the patterning process. As such, in one embodiment, the monitoring module 87 is configured or programmed to perform the methods described herein. In one embodiment, the monitoring module 87 comprises a computing device. In one embodiment, the monitoring module 87 comprises a computer program for providing the functionality described herein, the computer program encoded on a computer-readable medium forming or disposed within the monitoring module 87.

[0096]

[0117] In one embodiment, similar to the electron beam inspection tool of Figure 13 that uses a probe to inspect a substrate, the electron current in the system of Figure 14 is significantly larger than, for example, a CD-SEM as depicted in Figure 13, so the probe spot may be sufficiently large and, as a result, the inspection speed may be fast. However, the resolution may not be as high as a CD-SEM due to the large probe spot. In one embodiment, the inspection apparatus discussed above may be a single beam or a multi-beam apparatus without limiting the scope of the present disclosure.

[0097]

[0118] For example, SEM images from the systems of Figures 13 and / or 14 can be processed to extract contours that delineate the edges of objects in the images, representing device structures. These contours are then typically quantified using metrics such as CD at user-defined cut lines. Images of device structures are therefore typically compared and quantified using metrics such as edge-to-edge distance (CD) measured on the extracted contours or simple pixel differences between images.

[0098]

[0119] In one embodiment, one or more procedures of processes 300, 400, 500, and / or 600 may be implemented as instructions (e.g., program code) in a processor of a computer system (e.g., process 104 of computer system 100). In one embodiment, the procedures may be distributed across multiple processors (e.g., parallel computing) to increase computational efficiency. In one embodiment, a computer program product including a non-transitory computer-readable medium has instructions stored on the non-transitory computer-readable medium, which, when executed by a computer hardware system, perform the methods described herein.

[0099]

[0120] According to the present disclosure, combinations and subcombinations of the disclosed elements constitute separate embodiments. For example, a first combination may include determining a deblurring model based on depth data related to a design pattern. A subcombination may include determining a deblurred image using the deblurring model. In another combination, the deblurred image may be used in an inspection process to determine OPC or SMO based on variance data generated with the model. In another example, the combination may include determining process adjustments to a lithography process, a resist process, or an etch process based on inspection data based on the deblurred image to improve yield of the patterning process.

[0100]

[0121] 15 is a block diagram illustrating a computer system 100 that can assist in implementing the methods, flows, or apparatuses disclosed herein. The computer system 100 includes a bus 102 or other communication mechanism for communicating information and a processor 104 (or multiple processors 104 and 105) coupled with the bus 102 for processing information. The computer system 100 also includes a main memory 106, such as a random access memory (RAM) or other dynamic storage device, coupled to the bus 102 for storing information and instructions executed by the processor 104. The main memory 106 may also be used for storing temporary variables or other intermediate information during execution of instructions executed by the processor 104. The computer system 100 further includes a read-only memory (ROM) 108 or other static storage device coupled to the bus 102 for storing static information and instructions for the processor 104. A storage device 110, such as a magnetic or optical disk, is provided and coupled to the bus 102 for storing information and instructions.

[0101]

[0122] Computer system 100 may be coupled via bus 102 to a display 112, such as a cathode ray tube (CRT), flat panel, or touch panel display, for displaying information to a computer user. An input device 114, including alphanumeric and other keys, is coupled to bus 102 for communicating information and command selections to processor 104. Another type of user input device is a cursor control 116, such as a mouse, trackball, or cursor direction keys, for communicating directional information and command selections to processor 104 and for controlling cursor movement on display 112. This input device typically has two degrees of freedom, allowing the device to be positioned in two axes—a first axis (e.g., x) and a second axis (e.g., y)—within a plane. A touch panel (screen) display may also be used as an input device.

[0102]

[0123] According to some embodiments, portions of one or more methods herein may be performed by computer system 100 in response to processor 104 executing one or more sequences of one or more instructions contained in main memory 106. Such instructions may be read into main memory 106 from another computer-readable medium, such as storage device 110. Execution of the sequences of instructions contained in main memory 106 causes processor 104 to perform the process steps described herein. One or more processors in a multi-processing arrangement may be used to execute the sequences of instructions contained in main memory 106. In some alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions. Thus, the description herein is not limited to any specific combination of hardware circuitry and software.

[0103]

[0124] The term "computer-readable medium," as used herein, refers to any medium that participates in providing instructions to processor 104 for execution. Such media may take many forms, including, but not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media include, for example, optical or magnetic disks, such as storage device 110. Volatile media include dynamic memory, such as main memory 106. Transmission media include coaxial cables, copper wire, and fiber optics (including the wires that comprise bus 102). Transmission media may also take the form of acoustic or light waves, such as those generated during radio frequency (RF) and infrared (IR) data communications. Common forms of computer-readable media include, for example, floppy disks, flexible disks, hard disks, magnetic tape, other magnetic media, CD-ROMs, DVDs, other optical media, punch cards, paper tape, other physical media with patterns of holes, RAM, PROMs, and EPROMs, FLASH-EPROMs, other memory chips or cartridges, carrier waves, as described below, or other computer-readable media.

[0104]

[0125] Various forms of computer-readable media may be involved in carrying one or more sequences of one or more instructions to processor 104 for execution. For example, the instructions may initially reside on a magnetic disk of a remote computer. The remote computer may load the instructions into its dynamic memory and send the instructions over a telephone line using a modem. A modem local to computer system 100 may receive the data on the telephone line and use an infrared transmitter to convert the data to an infrared signal. An infrared detector coupled to bus 102 may receive the data carried in the infrared signal and place the data on bus 102. Bus 102 carries the data to main memory 106, from which processor 104 retrieves and executes the instructions. The instructions received by main memory 106 may optionally be stored on storage device 110 either before or after execution by processor 104.

[0105]

[0126] Computer system 100 may also include a communication interface 118 coupled to bus 102. The communication interface 118 provides a two-way data communication coupling to a network link 120 that is connected to a local network 122. For example, communication interface 118 may be an Integrated Services Digital Network (ISDN) card or a modem to provide a data communication connection to a corresponding type of telephone line. As another example, communication interface 118 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN. A wireless link may also be implemented. In such an implementation, communication interface 118 sends and receives electrical, electromagnetic, or optical signals that carry digital data streams representing various types of information.

[0106]

[0127] Network link 120 typically provides data communication through one or more networks to other data devices. For example, network link 120 may provide a connection through local network 122 to a host computer 124 or to data equipment operated by an Internet Service Provider (ISP) 126. ISP 126 in turn provides data communication services through the world wide packet data communication network (now commonly referred to as the "Internet" 128). Local network 122 and Internet 128 both use electrical, electromagnetic, or optical signals that carry digital data streams. The signals through the various networks and on network link 120 and through communication interface 118, which carry the digital data to and from computer system 100, are exemplary forms of carrier waves transporting the information.

[0107]

[0128] Computer system 100 can send messages and receive data, including program code, through one or more networks, network link 120, and communication interface 118. In the Internet example, server 130 might send a requested code for an application program through Internet 128, ISP 126, local network 122, and communication interface 118. Such a downloaded application may provide all or a portion of the methods herein. The received code may be executed by processor 104 as received, and / or stored in storage device 110, or other non-volatile storage for later execution. In this manner, computer system 100 may obtain the application code in the form of a carrier wave.

[0108]

[0129] 16 schematically depicts an exemplary lithographic projection apparatus that can be used in conjunction with the techniques described herein. The apparatus includes: - an illumination system IL for conditioning the radiation beam B. In this particular case, the illumination system also includes a radiation source SO; - a first object table (e.g., patterning device table) MT comprising a patterning device holder for holding a patterning device MA (e.g., a reticle) and connected to a first positioner for accurately positioning the patterning device relative to the item PS; - a second object table (substrate table) WT comprising a substrate holder for holding a substrate W (e.g. a resist-coated silicon wafer) and connected to a second positioner for accurately positioning the substrate relative to the item PS; a projection system ("lens") PS (e.g. a refractive, reflective or catadioptric optical system) that images an irradiated portion of the patterning device MA onto a target portion C (e.g. comprising one or more dies) of the substrate W.

[0109]

[0130] As depicted herein, the apparatus is of a transmissive type (i.e. has a transmissive patterning device). In general, however, it may be of a reflective type (e.g. have a reflective patterning device). The apparatus may use other types of patterning device than a conventional mask; examples include a programmable mirror array or an LCD matrix.

[0110]

[0131] A source SO (e.g. a mercury lamp or excimer laser, LPP (Laser Produced Plasma) EUV source) produces a radiation beam. This beam is fed to an illumination system (illuminator) IL, either directly or after having traversed conditioning means, such as a beam expander Ex. The illuminator IL may comprise conditioning means AD for setting the outer and / or inner radial extent (commonly referred to as σ-outer and σ-inner, respectively) of the intensity distribution of the beam. Furthermore, it will generally comprise various other components, such as an integrator IN and a condenser CO. In this way, the beam B impinging on the patterning device MA has a desired uniformity and intensity distribution in its cross-section.

[0111]

[0132] It should be noted, with regard to Figure 12, that the source SO may be located within the housing of the lithographic projection apparatus (as is often the case when the source SO is, for example, a mercury lamp), but it may also be remote from the lithographic projection apparatus and the radiation beam that it produces may be directed into the apparatus (for example by the use of appropriate directing mirrors). This latter scenario is often the case when the source SO is an excimer laser (for example based on KrF, ArF or F2 lasing).

[0112]

[0133] Subsequently, beam PB intersects the patterning device MA, which is held on a patterning device table MT. After traversing the patterning device MA, beam B passes through a lens PL, which focuses beam B onto a target portion C of the substrate W. Using the second positioning means (and interferometric measurement means IF), the substrate table WT can be precisely moved, e.g., to position a different target portion C in the path of beam PB. Similarly, the first positioning means can be used to precisely position the patterning device MA with respect to the path of beam B, e.g., after mechanical retrieval of the patterning device MA from a patterning device library, or during a scan. In general, movement of the object table MT, WT, is realized using a long-stroke module (coarse positioning) and a short-stroke module (fine positioning), which are not explicitly depicted in FIG. 16 . However, in the case of a stepper (as opposed to a step-and-scan tool), the patterning device table MT may be connected to a short-stroke actuator only, or may be fixed.

[0113]

[0134] The depicted tool can be used in two different modes: - in step mode, the patterning device table MT remains essentially stationary and the entire patterning device image is projected onto the target portion C in one go (i.e. in a single "flash"), and the substrate table WT is then shifted in the x and / or y directions so that a different target portion C can be irradiated by the beam PB; In scan mode, essentially the same scenario applies, except that a given target portion C is not exposed in a single "flash". Instead, the patterning device table MT is movable in a given direction (the so-called "scan direction", e.g. the y direction) with a speed v, such that the projection beam B is caused to scan over the patterning device image. In parallel, the substrate table WT is simultaneously moved in the same or opposite direction with a speed V = Mv, in which M is the magnification factor of the lens PL (typically, M = 1 / 4 or 1 / 5). In this way, a relatively large target portion C can be exposed, without having to compromise on resolution.

[0114]

[0135] FIG. 17 schematically depicts another exemplary lithographic projection apparatus LA that can be used in conjunction with the techniques described herein.

[0115]

[0136] The lithographic projection apparatus LA includes: - Source Collector Module SO an illumination system (illuminator) IL configured to condition a radiation beam B (e.g., EUV radiation); a support structure (e.g., patterning device table) MT constructed to support a patterning device (e.g., a mask or reticle) MA and connected to a first positioner PM configured to accurately position the patterning device; a substrate table (e.g., wafer table) WT constructed to hold a substrate (e.g., a resist-coated wafer) W and connected to a second positioner PW configured to accurately position the substrate; and a projection system (e.g. a reflective projection system) PS configured to project a pattern imparted to the radiation beam B by the patterning device MA onto a target portion C (e.g. comprising one or more dies) of the substrate W.

[0116]

[0137] As depicted here, the apparatus LA is reflective (e.g., uses a reflective patterning device). Note that because most materials are absorptive in the EUV wavelength range, the patterning device can have a multilayer reflector including, for example, a multistack of molybdenum and silicon. In one example, a multistack reflector has 40 layer pairs of molybdenum and silicon, with each layer being a quarter wavelength thick. Even smaller wavelengths can be produced using x-ray lithography. Because most materials are absorptive at EUV and x-ray wavelengths, a thin strip of patterned absorbing material (e.g., TaN absorber on a multilayer reflector) on the patterning device topography defines where features will or will not print (positive resist) or print (negative resist).

[0117]

[0138] Referring to FIG. 17 , the illuminator IL receives an extreme ultraviolet radiation (EUV) beam from a source collector module SO. Methods for generating EUV radiation include, but are not limited to, converting a material into a plasma state having at least one element (e.g., xenon, lithium, or tin) with one or more emission lines in the EUV range. In one such method, often referred to as laser-produced plasma (“LPP”), the plasma can be generated by irradiating a fuel, such as droplets, streams, or clusters of material having a line-emitting element, with a laser beam. The source collector module SO may be part of an EUV radiation system that includes a laser (not shown in FIG. 17 ) that provides a laser beam that excites the fuel. The resulting plasma emits output radiation (e.g., EUV radiation), which is collected using a radiation collector disposed in the source collector module. The laser and source collector module may be separate entities, for example, when a CO laser is used to provide the laser beam for fuel excitation.

[0118]

[0139] In such cases, the laser is not considered to form part of the lithographic apparatus, and the radiation beam is passed from the laser to the source collector module using a beam delivery system, for example comprising appropriate directing mirrors and / or beam expanders. In other cases, the source may be an integral part of the source collector module, for example when the source is a discharge produced plasma EUV generator, often referred to as a DPP source.

[0119]

[0140] The illuminator IL may include an adjuster for adjusting the angular intensity distribution of the radiation beam. Generally, at least the outer and / or inner radial extent (commonly referred to as σ-outer and σ-inner, respectively) of the intensity distribution in a pupil plane of the illuminator may be adjusted. In addition, the illuminator IL may include various other components, such as faceted field and pupil mirror devices. The illuminator may be used to condition the radiation beam to have a desired uniformity and intensity distribution in its cross-section.

[0120]

[0141] The radiation beam B is incident on a patterning device (e.g., mask) MA, which is held on a support structure (e.g., patterning device table) MT, and is patterned by the patterning device. After reflecting from the patterning device (e.g., mask) MA, the radiation beam B passes through a projection system PS, which focuses the beam onto a target portion C of a substrate W. The substrate table WT can be accurately moved using a second positioner PW and a position sensor PS2 (e.g., an interferometric device, a linear encoder, or a capacitive sensor), for example, to position a different target portion C in the path of the radiation beam B. Similarly, the first positioner PM and another position sensor PS1 can be used to accurately position the patterning device (e.g., mask) MA with respect to the path of the radiation beam B. The patterning device (e.g., mask) MA and substrate W may be aligned using patterning device alignment marks M1, M2 and substrate alignment marks P1, P2.

[0121]

[0142] The depicted apparatus LA could be used in at least one of the following modes:

[0143] 1. In step mode, the support structure (e.g. patterning device table) MT and the substrate table WT remain essentially stationary while an entire pattern imparted to the radiation beam is projected onto a target portion C in one go (i.e. a single static exposure), where the substrate table WT is then shifted in the X and / or Y directions so that a different target portion C can be exposed.

[0144] 2. In scan mode, the support structure (e.g. patterning device table) MT and the substrate table WT are scanned synchronously (i.e. single dynamic exposure) while a pattern imparted to the radiation beam is projected onto a target portion C. The velocity and direction of the substrate table WT relative to the support structure (e.g. patterning device table) MT may be determined by the de-magnification and image reversal characteristics of the projection system PS.

[0145] 3. In another mode, the support structure (e.g. patterning device table) MT holds a programmable patterning device and remains essentially stationary, and the substrate table WT is moved or scanned, while a pattern imparted to the radiation beam is projected onto a target portion C. In this mode, a pulsed radiation source is generally used, and the programmable patterning device is updated as required after each movement of the substrate table WT, or between successive radiation pulses during a scan. This mode of operation is readily adaptable to maskless lithography, employing a programmable patterning device such as a programmable mirror array of the type referred to above.

[0122]

[0146] FIG. 18 shows in more detail the apparatus LA, including the source collector module SO, the illumination system IL, and the projection system PS. The source collector module SO is constructed and arranged such that a vacuum environment can be maintained within the enclosure 220 of the source collector module SO. The EUV radiation-emitting plasma 210 can be formed by a discharge-produced plasma source. The EUV radiation can be generated by a gas or vapor (e.g., Xe gas, Li vapor, or Sn vapor, in which a very hot plasma 210 is created to emit radiation in the EUV range of the electromagnetic spectrum). The very hot plasma 210 is created, for example, by a discharge that produces an at least partially ionized plasma. A partial pressure of, for example, 10 Pa of Xe, Li, Sn vapor, or any other suitable gas or vapor may be required for efficient generation of radiation. In an embodiment, a plasma of excited tin (Sn) is provided to generate the EUV radiation.

[0123]

[0147] Radiation emitted by the high-temperature plasma 210 passes from the source chamber 211 into the collector chamber 212 through an optional gas barrier or contaminant trap 230 (sometimes also referred to as a contaminant barrier or foil trap) located in or behind the opening of the source chamber 211. The contaminant trap 230 may include a channel structure. The contaminant trap 230 may also include a gas barrier or a combination of a gas barrier and a channel structure. The contaminant trap or contaminant barrier 230 further described herein includes at least a channel structure, as known in the art.

[0124]

[0148] The collector chamber 211 may include a radiation collector CO, which may be a so-called grazing incidence collector. The radiation collector CO has an upstream radiation collector side 251 and a downstream radiation collector side 252. Radiation traversing the collector CO may be reflected off a grating spectral filter 240 to be focused to a virtual source point IF along the optical axis indicated by the dash-dotted line "O". The virtual source point IF is commonly called the intermediate focus, and the source collector module is positioned such that the intermediate focus IF is located at or near the opening 221 of the enclosure structure 220. The virtual source point IF is an image of the radiation-emitting plasma 210.

[0125]

[0149] The radiation subsequently traverses an illumination system IL, which may include a faceted field mirror device 22 and a facetted pupil mirror device 24 arranged to provide a desired angular distribution of the radiation beam 21 at the patterning device MA, and a desired uniformity of the radiation intensity at the patterning device MA. Upon reflection of the radiation beam 21 off the patterning device MA, which is held by a support structure MT, a patterned beam 26 is formed, and the patterned beam 26 is imaged by the projection system PS via reflective elements 28, 30 onto a substrate W held by a substrate table WT.

[0126]

[0150] Generally, more elements than shown may be present in illumination optics unit IL and projection system PS. A grating spectral filter 240 may optionally be present, depending on the type of lithographic apparatus. Furthermore, more mirrors than shown in the figures may be present, for example, 1 to 6 additional reflective elements may be present in projection system PS than shown in Figure 18.

[0127]

[0151] 18 is depicted as a nested collector with grazing incidence reflectors 253, 254, and 255, just as an example of a collector (or collector mirror). The grazing incidence reflectors 253, 254, and 255 are arranged axisymmetrically about the optical axis O, and this type of collector system CO can be used in combination with a discharge produced plasma source, often referred to as a DPP source.

[0128]

[0152] Alternatively, the source collector module SO may be part of an LPP radiation system, as shown in Figure 19. A laser LA is arranged to deposit laser energy into a fuel such as xenon (Xe), tin (Sn), or lithium (Li) to create a highly ionized plasma 210 with an electron temperature of tens of eV. Energetic radiation produced during de-excitation and recombination of these ions is emitted from the plasma and collected by a near-normal incidence collector system CO and focused onto an opening 221 in an enclosure structure 220.

[0129]

[0153] The concepts disclosed herein can be used to simulate or mathematically model general imaging systems for imaging subwavelength features and can be particularly useful for new imaging technologies capable of producing shorter wavelengths. New technologies already in use include EUV (extreme ultraviolet), DUV lithography, which can produce wavelengths of 193 nm using ArF lasers and even 157 nm using fluorine lasers. EUV lithography can also produce wavelengths within this range by using synchrotrons or by bombarding materials (solids or plasmas) with high-energy electrons to generate photons within the 20-5 nm range.

[0130]

[0154] Embodiments of the present disclosure may be further described using the following clauses. 1. A non-transitory computer-readable medium for storing a deblurring model configured to deblur a captured image of a patterned substrate, the non-transitory computer-readable medium comprising stored instructions, The instructions, when executed by one or more processors, obtaining a simulated image of the substrate corresponding to the target pattern via a simulator using a target pattern from which the pattern on the substrate is derived, the target pattern including a first target feature formed on a first layer on the substrate and a second target feature formed on a second layer on the substrate, the second layer being located below the first layer on the substrate; determining edge extent data for features on the substrate corresponding to the first target feature and the second target feature of the target pattern based on depth data associated with a plurality of layers of the substrate, the depth data characterizing edge blur of the features on the substrate as a function of depth of the respective features from a top layer of the substrate; using the simulated image of the substrate and the edge range data as training data to adjust parameters of the base model to generate a deblurring model, the deblurring model generating a deblurred image of the captured image of the patterned substrate using the edge range data associated with the captured image; A non-transitory computer-readable medium for causing operations to occur, including: 2. The medium of clause 1, wherein the target pattern includes geometric data associated with a first target feature and a second target feature. 3. The medium described in clause 2, wherein the geometric data includes desired shapes of the first target feature and the second target feature and target positions of the edges of the first target feature and the second target feature, respectively, within the target pattern. 4. Depth data is a first depth of a first layer located vertically downward from the top layer, and a second depth of a second layer located vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; The medium according to any one of clauses 1 to 3, comprising at least one of the following: 5. The medium of any one of clauses 1-4, wherein the edge range data includes a range of edge positions on the substrate corresponding to each feature in the target pattern. 6. The medium of any one of clauses 1-5, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 7. The medium of any one of clauses 1 to 6, wherein the edge range data is represented as an image of the target pattern complemented with edge range data for each feature to be printed on the substrate. 8. Obtaining a simulated image extracting noise from the captured image of the patterned substrate; adding noise to the simulated image; 8. The medium of any one of clauses 1 to 7, further comprising: 9. The medium of any one of clauses 1 to 8, wherein the deblurring model is a machine learning model. 10. The medium of any one of clauses 1-9, wherein the deblurring model is a deep convolutional network and the model parameters include weights and biases of one or more layers of the deep convolutional network. 11. The deblurring model a feature vector containing values ​​indicating features of the target pattern; an attention vector including depth-related values ​​associated with features to be printed on the substrate, the attention vector configured to be computed with the feature vector; 11. The medium of any one of clauses 1 to 10, which is a neural network comprising: 12. The medium of clause 11, wherein the attention vector comprises a value in the range of 0 to 1, or the value 0 or 1. 13. The medium of clause 12, wherein the machine learning model is configured to multiply the attention vector with the feature vector such that a depth-related weight is assigned by the attention vector to edges of particular features in the feature vector. 14. The medium of clause 13, wherein a first target feature associated with a first layer is multiplied by a higher attention value compared to an attention value multiplied by a second target feature associated with a second layer. 15. The medium of any one of clauses 1 to 14, wherein the simulator includes a model that is a function of the physical properties of the substrate on which the target pattern is printed, and that generates a simulated image, thereby enabling determination of feature characteristics in each layer of the substrate. 16. The medium of clause 15, wherein the physical characteristics of the substrate include at least one of the material of a particular layer, the thickness of a particular layer, a resist process parameter, or an etching process parameter. 17. The medium of clause 16, wherein obtaining the simulated image includes simulating the model to generate the simulated image through a Monte Carlo simulation process. 18. The medium of clause 16, wherein obtaining the simulated image includes simulating a double Gaussian model configured to generate the simulated image based on the depth data. 19. Capturing an image of the patterned multilayer substrate using a given target pattern via a capture device; determining edge range data for features on the substrate corresponding to each feature of a given target pattern; inputting the edge coverage data and the captured image into a deblurring model to generate a deblurred image of the captured image; 19. The medium of any one of clauses 1 to 18, further comprising: 20. The medium of clause 19, wherein the captured image is a scanning electron beam (SEM) image obtained by an SEM at a high energy setting used to capture an image of the patterned multilayer substrate. 21. The medium of any one of clauses 1-20, wherein the target pattern is in the form of a polygon-based hierarchical data format. 22. The medium according to clause 21, wherein the polygon-based data format includes the GDS format. 23. A non-transitory computer-readable medium for deblurring an image obtained from a scanning electron microscope (SEM) by using data related to layer depths of a patterned substrate, the non-transitory computer-readable medium comprising stored instructions, The instructions, when executed by one or more processors, receiving a captured image of the patterned substrate; inputting the captured image into a deblurring model, the deblurring model comprising: obtaining a simulated image of the substrate corresponding to a target pattern via a simulator using a target pattern from which a pattern on the substrate is derived, the target pattern including a first target feature formed on a first layer on the substrate and a second target feature formed on a second layer on the substrate, the second layer being located below the first layer on the substrate; determining edge extent data for features on the substrate corresponding to the first target feature and the second target feature of the target pattern based on depth data associated with a plurality of layers of the substrate, the depth data characterizing edge blur of the features on the substrate as a function of depth of the respective features from a top layer of the substrate; using the simulated image of the substrate and edge coverage data as training data to adjust parameters of the base model to generate a deblurring model; and be trained by receiving an output from a deblurring model using the captured image and edge range data associated with features of the patterned substrate; generating a deblurred captured image on a user interface for display; A non-transitory computer-readable medium for causing operations to occur, including: 24. The medium of clause 23, wherein the target pattern includes geometric data associated with a first target feature and a second target feature. 25. The medium of clause 24, wherein the geometric data includes desired shapes of the first target feature and the second target feature and target positions of the edges of the first target feature and the second target feature, respectively, within the target pattern. 26. Geometric data is a first depth of the first layer measured vertically downward from the top layer, and a second depth of the second layer measured vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 26. The medium according to any one of clauses 23 to 25, comprising at least one of the following: 27. The medium of any one of clauses 23-26, wherein the edge range data includes a range of edge positions of features on the substrate corresponding to each feature in the target pattern. 28. The medium of any one of clauses 23-27, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 29. The medium of any one of clauses 23-28, wherein the edge range data is represented as an image of the target pattern complemented with edge range data for each feature of the target pattern. 30. The medium of any one of clauses 23 to 29, wherein the deblurring model is a machine learning model. 31. The medium of any one of clauses 23-30, wherein the deblurring model is a deep convolutional network and the model parameters include weights and biases of one or more layers of the deep convolutional network. 32. A medium described in any one of clauses 23 to 31, wherein the simulator includes a model that is a function of the physical properties of the substrate on which the target pattern is printed, and that generates a simulated image, thereby enabling the determination of feature characteristics in each layer of the substrate. 33. The medium of any one of clauses 23-32, wherein the captured image is a scanning electron beam (SEM) image obtained by an SEM at a high energy setting used to capture an image of the patterned multilayer substrate. 34. A non-transitory computer-readable medium for deblurring a captured image of a patterned substrate, comprising stored instructions, The instructions, when executed by one or more processors, inputting a captured image of the patterned substrate and edge range data associated with a target pattern into a deblurring model, the deblurring model being trained based on the edge range data associated with features of the target pattern at a particular depth; deblurring the captured image by implementing a deblurring model; A non-transitory computer-readable medium for causing operations to occur, including: 35. The medium of clause 34, wherein edge range data for features on the substrate corresponding to features of the target pattern is determined based on depth data associated with the target pattern and the substrate. 36. Depth data is a first depth of the first layer measured vertically downward from the top layer, and a second depth of the second layer measured vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 36. The medium of clause 35, comprising at least one of: 37. The medium of any one of clauses 34-36, wherein the edge range data includes a range of edge positions for features on the substrate corresponding to each feature in the target pattern. 38. The medium of clause 34, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 39. The medium of any one of clauses 34-38, wherein the edge range data is represented as an image of the target pattern supplemented with edge range data corresponding to each feature of the target pattern. 40. A method for deblurring a captured image of a patterned substrate, comprising: Identifying features in the captured image based on target patterns and depth data associated with multiple layers of the patterned substrate; deblurring the captured image by deblurring each of the features based on the target pattern and depth data; A method comprising: 41. Depth data is a first depth of the first layer from a top layer of the patterned substrate, and a second depth of the second layer from the top layer; material data for the first and second layers of the patterned substrate, or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 41. The method of claim 40, comprising at least one of: 42. The method of clause 41, wherein identifying features in the captured image includes determining edge range data for features on the patterned substrate based on depth data associated with multiple layers of the patterned substrate. 43. The method of clause 42, wherein the edge range data comprises a range of edge positions of features on the substrate corresponding to each feature in the target pattern. 44. The method of clause 43, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 45. A method according to any one of clauses 42 to 44, wherein the edge range data is represented as an image of the target pattern complemented with edge range data corresponding to each feature of the target pattern. 46. ​​A method for generating a deblurring model to deblur a captured image of a patterned substrate, comprising: obtaining a simulated image of the substrate corresponding to the target pattern via a simulator using a target pattern from which the pattern on the substrate is derived, the target pattern including a first target feature formed on a first layer on the substrate and a second target feature formed on a second layer on the substrate, the second layer being located below the first layer on the substrate; determining edge extent data for features on the substrate corresponding to the first target feature and the second target feature of the target pattern based on depth data associated with a plurality of layers of the substrate, the depth data characterizing edge blur of the features on the substrate as a function of depth of the respective features from a top layer of the substrate; using the simulated image of the substrate and the edge range data as training data to adjust parameters of the base model to generate a deblurring model, the deblurring model generating a deblurred image of the captured image of the patterned substrate using the edge range data associated with the captured image; A method comprising: 47. The method of clause 46, wherein the target pattern includes geometric data associated with a first target feature and a second target feature. 48. The method of clause 47, wherein the geometric data includes desired shapes of the first target feature and the second target feature and target positions of edges of the first target feature and the second target feature, respectively, within the target pattern. 49. Depth data is a first depth of a first layer located vertically downward from the top layer, and a second depth of a second layer located vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 49. The method of any one of clauses 46 to 48, comprising at least one of: 50. The method of any one of clauses 46-49, wherein the edge range data comprises a range of edge positions for features on the substrate corresponding to each feature in the target pattern. 51. The method of any one of clauses 46 to 50, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 52. A method according to any one of clauses 46 to 51, wherein the edge range data is represented as an image of the target pattern complemented with edge range data corresponding to each feature of the target pattern. 53. Obtaining a simulated image extracting noise from the captured image of the patterned substrate; adding noise to the simulated image; 53. The method of any one of clauses 46 to 52, further comprising: 54. The method of any one of clauses 46 to 53, wherein the deblurring model is a machine learning model. 55. The method of any one of clauses 46 to 54, wherein the deblurring model is a deep convolutional network and the model parameters include weights and biases of one or more layers of the deep convolutional network. 56. The blur removal model is a feature vector containing values ​​indicating features of the target pattern; an attention vector including depth-related values ​​associated with features to be printed on the substrate, the attention vector configured to be computed with the feature vector; 56. The method of any one of clauses 46 to 55, wherein the neural network comprises: 57. The method of clause 56, wherein the interest vector comprises a value in the range 0 to 1, or the value 0 or 1. 58. The method of clause 57, wherein the machine learning model is configured to multiply the attention vector with the feature vector such that a depth-related weight is assigned by the attention vector to edges of particular features in the feature vector. 59. The method of clause 58, wherein a first target feature associated with a first layer of the substrate is multiplied by a higher attention value compared to an attention value multiplied by a second target feature associated with a second layer of the substrate. 60. A method according to any one of clauses 46 to 59, wherein the simulator comprises a model that is a function of the physical properties of the substrate on which the target pattern is printed, and that generates a simulated image, thereby enabling determination of feature characteristics in each layer of the substrate. 61. The method of clause 60, wherein the physical properties of the substrate include at least one of the material of a particular layer, the thickness of a particular layer, a resist process parameter, or an etching process parameter. 62. The method of clause 61, wherein obtaining the simulated image includes simulating a model to generate the simulated image through a Monte Carlo simulation process. 63. The method of clause 62, wherein obtaining the simulated image includes simulating a bi-Gaussian model configured to generate the simulated image based on the depth data. 64. Capturing an image of the patterned multilayer substrate using a given target pattern via a capture device; determining edge range data for features on the substrate corresponding to each feature of a given target pattern; inputting the edge coverage data and the captured image into a deblurring model to generate a deblurred image of the captured image; 64. The method of any one of clauses 46 to 63, further comprising: 65. The method of clause 64, wherein the captured image is a scanning electron beam (SEM) image obtained by an SEM at a high energy setting used to capture an image of the patterned multilayer substrate. 66. The method of any one of clauses 46 to 65, wherein the target pattern is in the form of a polygon-based hierarchical data format. 67. The method of clause 66, wherein the polygon-based data format includes a GDS format. 68. A method for deblurring a captured image of a patterned substrate, comprising: inputting a captured image of the patterned substrate and edge range data associated with a target pattern into a deblurring model, the deblurring model being trained based on the edge range data associated with features of the target pattern at a particular depth; deblurring the captured image by implementing a deblurring model; A method comprising: 69. The method of clause 68, wherein edge range data for features on the patterned substrate that correspond to features of the target pattern is determined based on depth data associated with the target pattern and the substrate. 70. Depth data is a first depth of a first layer located vertically downward from the top layer, and a second depth of a second layer located vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 69. The method of claim 69, comprising at least one of: 71. The method of any one of clauses 68 to 70, wherein the edge range data comprises a range of edge positions for features on the substrate corresponding to each feature in the target pattern. 72. The method of clause 68, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 73. A method according to any one of clauses 68 to 72, wherein the edge range data is represented as an image of the target pattern complemented with edge range data corresponding to each feature of the target pattern. 74. A system for deblurring a captured image of a patterned substrate, comprising: an electron beam optical system configured to capture an image of a patterned substrate patterned based on a target pattern; one or more processors configured to input the captured image of the patterned substrate and edge range data associated with the target pattern into a deblurring model, the deblurring model being trained based on edge range data associated with features of the target pattern at a particular depth, and to deblur the captured image by running the deblurring model; Including, the system. 75. The system of clause 74, wherein edge range data for features on the patterned substrate corresponding to features of the target pattern is determined based on depth data associated with the target pattern and the substrate. 76. Depth data is a first depth of a first layer located vertically downward from the top layer, and a second depth of a second layer located vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffractive behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; 76. The system of claim 75, comprising at least one of: 77. The system of any one of clauses 74-76, wherein the edge range data includes a range of edge positions for features on the substrate corresponding to each feature in the target pattern. 78. The system of clause 77, wherein determining edge range data includes executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern. 79. A system according to any one of clauses 74 to 78, wherein the edge range data is represented as an image of the target pattern complemented with edge range data corresponding to each feature of the target pattern.

[0131]

[0155] Although the concepts disclosed herein may be used for imaging on substrates such as silicon wafers, it will be understood that the disclosed concepts may be used with any type of lithographic imaging system, for example, one used for imaging on substrates other than silicon wafers.

[0132]

[0156] As used herein, unless otherwise specified, the term "or" includes all possible combinations unless it is not feasible. For example, if it is specified that a database can include A or B, then the database can include A, or B, or A and B, unless otherwise specified or not feasible. As a second example, if it is specified that a database can include A, B, or C, then the database can include A, or B, or C, or A and B, or A and C, or B and C, or A and B and C, unless otherwise specified or not feasible.

[0133]

[0157] The above description is intended to be illustrative, not limiting. Thus, it will be apparent to one skilled in the art that modifications may be made as described without departing from the scope of the claims set out below.

Claims

1. 1. A non-transitory computer-readable medium for storing a deblurring model configured to deblur a captured image of a patterned substrate, the non-transitory computer-readable medium comprising stored instructions, The instructions, when executed by one or more processors, obtaining a simulated image of the substrate corresponding to a target pattern via a simulator using a target pattern from which a pattern on the substrate is derived, the target pattern including a first target feature formed on a first layer on the substrate and a second target feature formed on a second layer on the substrate, the second layer being located below the first layer on the substrate; determining edge extent data for features on the substrate corresponding to the first target feature and the second target feature of the target pattern based on depth data associated with multiple layers of the substrate, the depth data characterizing edge blur of the features on the substrate as a function of depth of each feature from a top layer of the substrate; using the simulated image of the substrate and the edge range data as training data to adjust parameters of a base model to generate the deblurring model, the deblurring model generating a deblurred image of the captured image of the patterned substrate using edge range data associated with the captured image; A non-transitory computer-readable medium for causing operations to occur, including:

2. The medium of claim 1 , wherein the target pattern includes geometric data associated with the first target feature and the second target feature.

3. 3. The medium of claim 2, wherein the geometric data includes a desired shape of the first target feature and the second target feature and a target position of the edge of the first target feature and the second target feature, respectively, within the target pattern.

4. The depth data is a first depth of the first layer located vertically downward from the top layer, and a second depth of the second layer located vertically downward from the top layer of the substrate; material data associated with the first and second layers of the substrate; or diffraction behavior associated with first and second features on the substrate corresponding to the first and second target features of the target pattern, respectively; The medium of claim 1 , comprising at least one of:

5. The medium of claim 1 , wherein the edge range data comprises a range of edge positions on the substrate corresponding to each feature in the target pattern.

6. Determining the edge extent data includes: The medium of claim 1 , further comprising executing a model configured to calculate edge variations of features on the substrate corresponding to each feature of the target pattern.

7. The medium of claim 1 , wherein the edge range data is represented as an image of the target pattern complemented with the edge range data for each feature to be printed on the substrate.

8. said obtaining said simulated image further comprising: extracting noise from the captured image of the patterned substrate; adding the noise to the simulated image; The medium of claim 1 further comprising:

9. The medium of claim 1 , wherein the deblurring model is a machine learning model.

10. The medium of claim 1 , wherein the deblurring model is a deep convolutional network and the model parameters include weights and biases of one or more layers of the deep convolutional network.

11. The deblurring model is a feature vector containing values ​​indicative of features of the target pattern; an attention vector including depth-related values ​​associated with the features to be printed on the substrate, the attention vector being configured to be computed with the feature vector; The medium of claim 1 , wherein the medium is a neural network comprising:

12. The medium of claim 11 , wherein the attention vector comprises a value ranging from 0 to 1, or a value of 0 or 1.

13. 13. The medium of claim 12, wherein the deblurring model is configured to multiply the attention vector with the feature vector such that a depth-related weight is assigned by the attention vector to an edge of a particular feature of the feature vector.

14. The medium of claim 13 , wherein the first target feature associated with the first layer is multiplied with a higher attention value compared to an attention value multiplied with the second target feature associated with the second layer.

15. 1. A method for deblurring a captured image of a patterned substrate, comprising: obtaining, via a simulator, a simulated image corresponding to a target pattern associated with a plurality of layers of the patterned substrate; determining edge coverage data for features on the patterned substrate corresponding to the target pattern based on depth data associated with the plurality of layers; generating a deblurring model that generates a deblurred image of the captured image of the patterned substrate using edge range data associated with the captured image by adjusting parameters of a base model using the simulated image and the edge range data as training data; deblurring the captured image by deblurring each of the features using the deblurring model; A method comprising:

Images