Real-time recipe tuning for inspection systems

Abstract

The present disclosure provides a method for real-time parameter tuning of a wafer inspection system. The method includes acquiring one or more input images streamed from a wafer inspection system; applying a plurality of image enhancement parameters to the one or more input images to generate a plurality of images having different characteristics; identifying defects from the plurality of images by applying the plurality of defect detection parameters; and determining a parameter combination of the plurality of image enhancement parameters and the plurality of defect detection parameters to be used for detecting the first defect type based on the defects identified on the plurality of images.

Description

Cross-reference to related applications

[0001] This application claims priority to U.S. Application 63 / 538,268, filed on September 13, 2023, which is incorporated herein by reference in its entirety. Technical Field

[0002] The embodiments provided herein generally relate to check-recipe tuning techniques, and more particularly to real-time check-recipe tuning techniques utilizing offline computing platforms. Background Technology

[0003] In the manufacturing process of integrated circuits (ICs), incomplete or completed circuit components are inspected to ensure they are manufactured according to the design and free of defects. Inspection systems utilizing optical microscopes or charged particle (e.g., electron) beam microscopes (such as scanning electron microscopy (SEM)) can be employed. As the physical dimensions of IC components continue to shrink, the accuracy and throughput of defect detection become increasingly important. Therefore, precise control of inspection tools such as SCPM tools is crucial for accurate defect detection and measurement. Because the process record (POR) recipes for inspection tools typically contain thousands of parameters, some of which are interdependent, recipe tuning for inspection tools is labor-intensive and time-consuming, resulting in long production cycles. There is a need in the industry to provide a recipe tuning technique that does not reduce the throughput of inspection systems. Summary of the Invention

[0004] Some embodiments of this disclosure may provide a method for real-time parameter tuning of a wafer inspection system. The method includes acquiring one or more input images streaming from the wafer inspection system; applying a plurality of image enhancement parameters to the one or more input images to generate a plurality of images with different characteristics; identifying defects from the plurality of images by applying a plurality of defect detection parameters; and determining a parameter combination of the plurality of image enhancement parameters and the plurality of defect detection parameters, based on the defects identified on the plurality of images, for detecting a first defect type.

[0005] Some embodiments of this disclosure provide a method for real-time parameter tuning of a wafer inspection system. The method includes acquiring one or more input images streaming from the wafer inspection system; applying a first parameter combination to the one or more input images and identifying a first set of defects in the one or more input images; applying a second parameter combination to the one or more input images and identifying a second set of defects in the one or more input images; determining a first key performance indicator (KPI) value for the first parameter combination based on the first set of defects; determining a second KPI value for the second parameter combination based on the second set of defects; and ranking the first parameter combination and the second parameter combination based on the first KPI value and the second KPI value.

[0006] Some embodiments of this disclosure may provide an apparatus for real-time parameter tuning of a wafer inspection system. The apparatus includes a memory storing an instruction set; and at least one processor configured to execute the instruction set to cause the apparatus to perform operations including: acquiring one or more input images streaming from the wafer inspection system; applying a plurality of image enhancement parameters to the one or more input images to generate a plurality of images with different characteristics; identifying defects from the plurality of images by applying a plurality of defect detection parameters; and determining a combination of the plurality of image enhancement parameters and the plurality of defect detection parameters, based on the defects identified on the plurality of images, for detecting a first defect type.

[0007] Other advantages of the embodiments of this disclosure will become apparent from the following description taken in conjunction with the accompanying drawings, in which certain embodiments of the invention are illustrated and exemplified. Attached Figure Description

[0008] The above and other aspects of this disclosure will become more apparent from the description of exemplary embodiments in conjunction with the accompanying drawings.

[0009] Figure 1 This is a schematic diagram illustrating an example charged particle beam inspection system consistent with embodiments of the present disclosure.

[0010] Figure 2A The illustration shows embodiments consistent with this disclosure. Figure 1 A schematic diagram of an example multi-beam tool as part of an exemplary charged particle beam inspection system.

[0011] Figure 2B The illustration shows embodiments consistent with this disclosure. Figure 1 A schematic diagram of an example single-beam tool as part of an exemplary charged particle beam inspection system.

[0012] Figure 3 This is a flowchart illustrating an example real-time recipe tuning method for an inspection tool, consistent with embodiments of this disclosure.

[0013] Figure 4A This is an example parameter structure for parameter setting consistent with embodiments of this disclosure.

[0014] Figure 4B The illustration shows example results after applying various image enhancement parameters to an input image, consistent with embodiments of this disclosure.

[0015] Figure 4C The illustration shows an example synthesized image after various image enhancement parameters have been applied to an input image, consistent with embodiments of this disclosure.

[0016] Figure 5AThe illustration shows example parameter combinations resulting from applying various detection parameters to an enhanced image, consistent with embodiments of this disclosure.

[0017] Figure 5B An example composite image suitable for critical size measurement, consistent with embodiments of this disclosure, is illustrated.

[0018] Figure 5C The illustration shows an example composite image suitable for detecting bridging defect types, consistent with embodiments of this disclosure.

[0019] Figure 5D The illustration shows an example composite image suitable for detecting damage and defect types, consistent with embodiments of this disclosure.

[0020] Figure 6A The illustration shows example parameter combinations resulting from applying various threshold parameters to a first defect type, consistent with embodiments of this disclosure.

[0021] Figure 6B The illustration shows the effect of threshold parameter variations consistent with embodiments of this disclosure on defect detection.

[0022] Figure 7A The illustration shows a first example analysis result of a combination of setting parameters consistent with embodiments of this disclosure.

[0023] Figure 7B The illustration shows a second example analysis result of a combination of setting parameters consistent with embodiments of this disclosure.

[0024] Figure 7C An example KPI chart for each parameter combination consistent with embodiments of this disclosure is illustrated.

[0025] Figure 8 This is a flowchart illustrating an example method for real-time recipe tuning for an online inspection system, consistent with embodiments of this disclosure.

[0026] Figure 9 This is a block diagram of an example computer system consistent with embodiments of the present disclosure. Detailed Implementation

[0027] Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. The following description refers to the accompanying drawings, wherein, unless otherwise indicated, the same numerals in the different drawings denote the same or similar elements. The implementations set forth in the following description of the exemplary embodiments do not represent all implementations. Rather, they are merely examples of apparatuses and methods consistent with various aspects of the disclosed embodiments as set forth in the appended claims. For example, although some embodiments are described in the context of the use of electron beams, this disclosure is not limited thereto. Other types of charged particle beams (e.g., including protons, ions, mesons, or any other charged particles) can be similarly applied. Furthermore, other imaging systems can be used, such as optical imaging, photoelectric detection, X-ray detection, ion detection, etc.

[0028] Electronic devices consist of circuits formed on a semiconductor material called a substrate. Semiconductor materials can include, for example, silicon, gallium arsenide, indium phosphide, or silicon germanium. Many circuits can be formed together on the same silicon wafer and are called integrated circuits or ICs. The size of these circuits has drastically decreased, allowing more circuits to be mounted on the substrate. For example, the IC chip in a smartphone can be as small as a fingernail but can include more than 2 billion transistors, each less than 1 / 1000th the size of a human hair.

[0029] Manufacturing these ICs, which have extremely small structures or components, is a complex, time-consuming, and expensive process, typically involving hundreds of individual steps. Even an error in one step can result in a defective IC that renders it unusable. Therefore, one goal of the manufacturing process is to avoid such defects in order to maximize the number of functional ICs manufactured in the process; that is, to increase the overall yield of the process.

[0030] A component of increasing yield is monitoring the chip manufacturing process to ensure it produces a sufficient number of functional integrated circuits. One way to monitor this process is to inspect it at various stages of chip circuit structure formation. This inspection can be performed using a scanning charged particle microscope (“SCPM”). For example, SCPM could be a scanning electron microscope (SEM). SCPM can be used to image these extremely small structures, essentially taking “photographs” of the wafer structure. This image can be used to determine if the structure is correctly formed in the correct location. If a defect is found, the process can be adjusted to make the defect less likely to recur.

[0031] As the physical dimensions of IC components continue to shrink, the accuracy and throughput of defect detection become increasingly critical. Therefore, precise control of inspection tools, such as SCPM tools, is essential for accurate defect detection and measurement. Tuning inspection tool recipes is labor-intensive and time-consuming, resulting in long production cycles, as the recorded process (POR) recipes for inspection tools typically contain thousands of parameters, some of which are interdependent. Current recipe tuning methods are highly dependent on user experience, leading to inconsistent recipe tuning performance and potentially unrobust synthesized recipes. Furthermore, current inspection tools used for in-line inspection have limited computational resources and cannot support additional computationally intensive processing, such as recipe tuning, since image processing for defect detection and measurement is already a bottleneck in high-volume manufacturing (HVM) wafer processes. Allocating the computational resources of inspection tools for recipe tuning mid-production consumes wafer production cycles, thus reducing the throughput of HVM wafer processes. As discussed, there is a trade-off between inspection sensitivity and throughput in wafer production cycles. Therefore, HVM production recipes are typically set to capture specific defect types under certain detection thresholds and image enhancement conditions and maintained throughout production. However, HVM wafer processes often undergo process changes over time, leading to missing defects due to fixed production recipes. Furthermore, because HVM processes are configured to use the same threshold for specific defect types throughout the entire inspection period for all HVM wafers, it may be impossible to capture new defects or new defect types once production begins without increasing wafer production cycles. A recipe tuning method is needed in the industry without reducing the throughput of the inspection system.

[0032] According to some embodiments of this disclosure, a real-time recipe tuning technique for an inspection system is provided. According to some embodiments of this disclosure, an offline computing platform, separate from the online inspection tools used in production, is used for real-time recipe tuning of the online inspection tools. According to some embodiments of this disclosure, comprehensive and computationally intensive recipe tuning for image inspection or metrology can be performed in parallel with inline production without increasing wafer production cycles. According to some embodiments of this disclosure, parameter combinations that can improve defect capture rates, including new defect detection or new defect type detection, can be provided. According to some embodiments of this disclosure, weaknesses in subsequent inspections can be predicted.

[0033] For clarity, the relative dimensions of the components in the accompanying drawings may be exaggerated. Throughout the following description of the drawings, the same or similar reference numerals refer to the same or similar components or entities, and only differences relative to the various embodiments are described. Other objects and advantages of this disclosure can be realized through the elements and combinations set forth in the embodiments discussed herein. However, embodiments of this disclosure are not necessarily required to achieve such exemplary objects or advantages, and some embodiments may not achieve any of the described objects or advantages.

[0034] Without limiting the scope of this disclosure, some embodiments can be described in the context of providing scanning deflection systems and methods in systems utilizing electron beams (“e-beams”). Some scanning deflection systems may use an electric field to influence a charged particle beam. However, this disclosure is not limited thereto. Other types of charged particle beams may be applied similarly. For example, systems and methods may be applicable to optical, photonic, X-ray, and ion beams. Deflection can be used to scan a beam on a surface in, for example, a cathode ray tube (CRT), a lithography machine, a scanning charged particle microscope (SCPM), or other analytical instruments. While some embodiments have been discussed with reference to deflection systems that use an electric field to influence a beam, deflection can also be achieved using, for example, a magnetic field.

[0035] As used herein, unless otherwise specified, the term "or" covers all possible combinations except where impractical. For example, if a descriptive component includes A or B, then, unless otherwise specified or impractical, the component may include A, B, or A and B. As a second example, if a descriptive component includes A, B, or C, then, unless otherwise specified or impractical, the component may include A, or B, or C, or A and B, or A and C, or B and C, or A, B, and C. Expressions such as "at least one of..." do not necessarily modify the entire list below, and do not necessarily modify each member of the list, such that "at least one of A, B, and C" should be understood to include only one A, only one B, only one C, or any combination of A, B, and C. The phrase "one of A and B" or "any one of A and B" should be interpreted in the broadest sense as including one A or one B.

[0036] Figure 1 An example electron beam inspection (EBI) system 100 consistent with embodiments of this disclosure is illustrated. The EBI system 100 can be used for imaging. Figure 1As shown, the EBI system 100 includes a main chamber 101, a loading / locking chamber 102, a beam tool 104, and an equipment front-end module (EFEM) 106. The beam tool 104 is located within the main chamber 101. The EFEM 106 includes a first loading port 106a and a second loading port 106b. The EFEM 106 may include multiple additional loading ports. The first loading port 106a and the second loading port 106b receive wafer front-opening transfer cassettes (FOUPs) containing wafers to be inspected (e.g., one or more semiconductor wafers made of (multiple) other materials) or samples (wafers and samples are interchangeable). A “batch” is a plurality of wafers that can be processed as a single load.

[0037] One or more robotic arms (not shown) in EFEM 106 can transport the wafer to loading / locking chamber 102. Loading / locking chamber 102 is connected to a loading / locking vacuum pump system (not shown), which removes gas molecules from loading / locking chamber 102 to achieve a first pressure below atmospheric pressure. After reaching the first pressure, one or more robotic arms (not shown) can transport the wafer from loading / locking chamber 102 to main chamber 101. Main chamber 101 is connected to a main chamber vacuum pump system (not shown), which removes gas molecules from main chamber 101 to achieve a second pressure below the first pressure. After reaching the second pressure, the wafer is inspected by a beam tool 104. Beam tool 104 can be a single-beam system or a multi-beam system.

[0038] The controller 109 is electronically connected to the beam tool 104. The controller 109 may be a computer configured to perform various controls of the EBI system 100. Although in Figure 1 The controller 109 is shown outside the structure including the main chamber 101, the loading / locking chamber 102 and the EFEM 106, but it should be understood that the controller 109 may be part of the structure.

[0039] In some embodiments, controller 109 may include one or more processors (not shown). A processor may be a general-purpose or specific electronic device capable of manipulating or processing information. For example, a processor may include any number of central processing units (or “CPU”), graphics processing units (or “GPU”), optical processors, programmable logic controllers, microcontrollers, microprocessors, digital signal processors, intellectual property (IP) cores, programmable logic arrays (PLAs), programmable array logic (PALs), general-purpose array logic (GALs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), system-on-a-chip (SoCs), application-specific integrated circuits (ASICs), and any combination of any type of circuitry capable of data processing. A processor may also be a virtual processor, comprising one or more processors distributed across multiple machines or devices coupled via a network.

[0040] In some embodiments, controller 109 may also include one or more memories (not shown). The memories can be general-purpose or specific electronic devices capable of storing processor-accessible (e.g., via a bus) code and data. For example, the memories can include any number of random access memory (RAM), read-only memory (ROM), optical discs, magnetic disks, hard disks, solid-state drives, flash drives, secure digital cards (SD cards), memory sticks, compact flash (CF) cards, or any combination of any type of storage device. The code and data can include an operating system (OS) and one or more applications (or “applications”) for a specific task. The memories can also be virtual memory, which includes one or more memories distributed across multiple machines or devices coupled via a network.

[0041] Figure 2A The illustration shows an example multi-beam tool 104A (also referred to herein as apparatus 104A) and an image processing system 290 consistent with embodiments of the present disclosure, which image processing system 290 can be configured for use in EBI system 100 ( Figure 1 ).

[0042] The beam tool 104A includes a charged particle source 202, a gun aperture 204, a converging lens 206, a primary charged particle beam 210 emitted from the charged particle source 202, a source conversion unit 212, multiple beams 214, 216, and 218 of the primary charged particle beam 210, a primary projection optics system 220, a motorized wafer stage 280, a wafer holder 282, multiple secondary charged particle beams 236, 238, and 240, a secondary optics system 242, and a charged particle detection device 244. The primary projection optics system 220 may include a beam splitter 222, a deflection scanning unit 226, and an objective lens 228. The charged particle detection device 244 may include detection sub-regions 246, 248, and 250.

[0043] The charged particle source 202, the gun aperture 204, the converging lens 206, the source conversion unit 212, the beam splitter 222, the deflection scanning unit 226, and the objective lens 228 can be aligned with the primary optical axis 260 of the device 104A. The secondary optical system 242 and the charged particle detection device 244 can be aligned with the secondary optical axis 252 of the device 104A.

[0044] Charged particle source 202 can emit one or more charged particles, such as electrons, protons, ions, muons, or any other charged particles. In some embodiments, charged particle source 202 can be an electron source. For example, charged particle source 202 can include a cathode, extractor, or anode, wherein primary electrons can be emitted from the cathode and extracted or accelerated to form a primary charged particle beam 210 (in this case, a primary electron beam) with a cross (virtual or real) 208. For ease of explanation and without ambiguity, electrons are used as examples in some descriptions herein. However, it should be noted that any charged particle can be used in any embodiment of this disclosure, not limited to electrons. The primary charged particle beam 210 can be visualized as being emitted from the cross 208. The aperture 204 can block peripheral charged particles of the primary charged particle beam 210 to reduce the Coulomb effect. The Coulomb effect can lead to an increase in the probe spot size.

[0045] Source conversion unit 212 may include an image forming element array and a beam-limiting aperture array. The image forming element array may include an array of micro-deflectors or microlenses. The image forming element array can form multiple parallel images (virtual or real) at the intersection 208 of multiple beam waves 214, 216, and 218 of the primary charged particle beam 210. The beam-limiting aperture array can limit the multiple beam waves 214, 216, and 218. Although in Figure 2A Three beams 214, 216, and 218 are shown, but embodiments of this disclosure are not limited thereto. For example, in some embodiments, device 104A may be configured to generate a first number of beams. In some embodiments, the first number of beams may be in the range of 1 to 1000. In some embodiments, the first number of beams may be in the range of 200 to 500. In some embodiments, device 104A may generate 400 beams.

[0046] The converging lens 206 can focus the primary charged particle beam 210. The currents of the beams 214, 216, and 218 downstream of the source conversion unit 212 can be changed by adjusting the focusing capability of the converging lens 206 or by changing the radial dimensions of the corresponding beam-limiting apertures within the beam-limiting aperture array. The objective lens 228 can focus the beams 214, 216, and 218 onto the wafer 230 for imaging, and can form multiple probe spots 270, 272, and 274 on the surface of the wafer 230.

[0047] Beam splitter 222 can be a Wien filter type beam splitter that generates electrostatic dipole fields and magnetic dipole fields. In some embodiments, if they are applied, the force exerted by the electrostatic dipole field on the charged particles (e.g., electrons) of beam waves 214, 216, and 218 on the charged particles can be substantially equal in amplitude and opposite in direction to the force exerted by the magnetic dipole field on the charged particles. Therefore, beam waves 214, 216, and 218 can pass directly through beam splitter 222 with zero deflection angle. However, the total dispersion of beam waves 214, 216, and 218 generated by beam splitter 222 can also be non-zero. Beam splitter 222 can separate secondary charged particle beams 236, 238, and 240 from beam waves 214, 216, and 218 and guide secondary charged particle beams 236, 238, and 240 toward secondary optical system 242.

[0048] The deflection scanning unit 226 can deflect beams 214, 216, and 218 to scan probe spots 270, 272, and 274 on the surface region of wafer 230. In response to the incident beams 214, 216, and 218 at probe spots 270, 272, and 274, secondary charged particle beams 236, 238, and 240 can be emitted from wafer 230. The secondary charged particle beams 236, 238, and 240 can include charged particles (e.g., electrons) with an energy distribution. For example, the secondary charged particle beams 236, 238, and 240 can be a secondary electron beam comprising secondary electrons (energy ≤ 50 eV) and backscattered electrons (energy between 50 eV and the landing energy of beams 214, 216, and 218). The secondary optical system 242 can focus the secondary charged particle beams 236, 238, and 240 onto the detection sub-regions 246, 248, and 250 of the charged particle detection device 244. The detection sub-regions 246, 248, and 250 can be configured to detect the corresponding secondary charged particle beams 236, 238, and 240, and generate corresponding signals (e.g., voltage, current, etc.) for reconstructing an inspection image of the structure on or below the surface region of the wafer 230.

[0049] The generated signals can represent the intensity of the secondary charged particle beams 236, 238, and 240, and can be provided to an image processing system 290 that communicates with the charged particle detection device 244, the primary projection optics system 220, and the motorized wafer stage 280. The movement speed of the motorized wafer stage 280 can be synchronized and coordinated with the beam deflection controlled by the deflection scanning unit 226, so that the movement of the scanning probe spots (e.g., scanning probe spots 270, 272, and 274) can orderly cover the region of interest on the wafer 230. The parameters of this synchronization and coordination can be adjusted to accommodate different materials of the wafer 230. For example, different materials of the wafer 230 may have different resistivity-capacitance characteristics, which may result in different signal sensitivities to the movement of the scanning probe spots.

[0050] The intensities of the secondary charged particle beams 236, 238, and 240 can vary depending on the external or internal structure of the wafer 230, thus indicating whether the wafer 230 contains defects. Furthermore, as discussed above, beams 214, 216, and 218 can be projected onto different locations on the top surface of the wafer 230 or different sides of a local structure of the wafer 230 to generate secondary charged particle beams 236, 238, and 240 with varying intensities. Therefore, by mapping the intensities of the secondary charged particle beams 236, 238, and 240 to regions of the wafer 230, the image processing system 290 can reconstruct an image reflecting the characteristics of the internal or external structure of the wafer 230.

[0051] In some embodiments, the image processing system 290 may include an image acquirer 292, a storage device 294, and a controller 296. The image acquirer 292 may include one or more processors. For example, the image acquirer 292 may include a computer, server, mainframe, terminal, personal computer, any kind of mobile computing device, or a combination thereof. The image acquirer 292 may be communicatively coupled to the charged particle detection device 244 of the beam tool 104A via a medium such as an electrical conductor, fiber optic cable, portable storage medium, IR, Bluetooth, Internet, wireless network, radio, or a combination thereof. In some embodiments, the image acquirer 292 may receive signals from the charged particle detection device 244 and may construct an image. The image acquirer 292 may thus acquire an inspection image of the wafer 230. The image acquirer 292 may also perform various post-processing functions, such as generating contours, overlaying indicators on the acquired image, etc. The image acquirer 292 may be configured to perform adjustments to the brightness and contrast of the acquired image. In some embodiments, the storage device 294 may be a storage medium such as a hard disk, flash drive, cloud storage device, random access memory (RAM), or other types of computer-readable storage. Storage device 294 can be coupled to image acquirer 292 and can be used to store scanned original image data as the original image and post-processed images. Image acquirer 292 and storage device 294 can be connected to controller 296. In some embodiments, image acquirer 292, storage device 294 and controller 296 can be integrated into a single control unit.

[0052] In some embodiments, image acquisition unit 292 may acquire one or more inspection images of a wafer based on imaging signals received from charged particle detection device 244. The imaging signals may correspond to a scanning operation for performing charged particle imaging. The acquired image may be a single image comprising multiple imaging regions. The single image may be stored in storage device 294. The single image may be an original image that can be divided into multiple regions. Each region may include an imaging region containing features of wafer 230. The acquired images may include multiple images of a single imaging region of wafer 230 sampled multiple times over a time series. Multiple images may be stored in storage device 294. In some embodiments, image processing system 290 may be configured to perform image processing steps on multiple images of the same location on wafer 230. In some embodiments, image processing system 290 may be configured to detect or enhance defects on an image according to a POR recipe. In some embodiments, the POR recipe may set a defect type for detection and set a threshold for the defect type. In some embodiments, the POR recipe may set multiple image enhancement parameters to be applied to the images acquired by image acquisition unit 292. In some embodiments, the detected defects and their corresponding locations on the wafer or mask can be stored in the storage device 294.

[0053] In some embodiments, the image processing system 290 may include measurement circuitry (e.g., an analog-to-digital converter) to obtain the distribution of detected secondary charged particles (e.g., secondary electrons). Charged particle distribution data collected during the detection time window can be combined with corresponding scan path data of beams 214, 216, and 218 incident on the wafer surface to reconstruct an image of the inspected wafer structure. The reconstructed image can be used to reveal various features of the internal or external structure of the wafer 230, and thus can be used to reveal any defects that may exist in the wafer.

[0054] In some embodiments, the charged particles can be electrons. When electrons from the primary charged particle beam 210 are projected onto the surface of the wafer 230 (e.g., probe spots 270, 272, and 274), the electrons from the primary charged particle beam 210 can penetrate the surface of the wafer 230 to a certain depth and interact with the particles of the wafer 230. Some electrons from the primary charged particle beam 210 can elastically interact with the material of the wafer 230 (e.g., in the form of elastic scattering or collision) and can be reflected or bounced off the surface of the wafer 230. Elastic interaction conserves the total kinetic energy of the interacting subjects (e.g., electrons from the primary charged particle beam 210), where the kinetic energy of the interacting subjects is not converted into other forms of energy (e.g., heat, electromagnetic energy, etc.). Such reflected electrons generated by elastic interaction can be called backscattered electrons (BSE). Some electrons from the primary charged particle beam 210 can inelastically interact with the material of the wafer 230 (e.g., in the form of inelastic scattering or collision). Inelastic interaction does not conserve the total kinetic energy of the interacting subjects, where some or all of the kinetic energy of the interacting subjects is converted into other forms of energy. For example, through inelastic interactions, the kinetic energy of some electrons in the primary charged particle beam 210 may lead to electronic excitation and transitions between material atoms. This inelastic interaction can also generate electrons that leave the surface of the wafer 230; these electrons can be referred to as secondary electrons (SEs). The yield or emission rate of BSEs and SEs depends, for example, on the material being examined and the landing energy of the electrons from the primary charged particle beam 210 on the material surface. The energy of the electrons in the primary charged particle beam 210 can be partly determined by its accelerating voltage (e.g., ...). Figure 2A The accelerating voltage between the anode and cathode of the charged particle source 202 is imparted. The amounts of BSE and SE can be more or less (or even the same) than the injected electrons of the primary charged particle beam 210.

[0055] Now refer to Figure 2B Another example of a charged particle beam device is discussed. Beam tool 104B (also referred to herein as device 104B) can be an example of beam tool 104 and can be similar to... Figure 2A The beam tool 104A is shown. However, unlike the device 104A, the device 104B can be a single-beam tool, which uses only one primary electron beam to scan one position on the wafer at a time.

[0056] like Figure 2BAs shown, the apparatus 104B includes a wafer holder 136 supported by a motorized stage 134 for holding a wafer 150 to be inspected. The beam tool 104B includes an electron emitter that may include a cathode 103, an anode 121, and a gun aperture 122. The beam tool 104B also includes a beam-limiting aperture 125, a converging lens 126, a cylindrical aperture 135, an objective lens assembly 132, and a detector 144. In some embodiments, the objective lens assembly 132 may be a modified SORIL lens, including a pole piece 132a, a control electrode 132b, a deflector unit 132c, and an excitation coil 132d. In the inspection or imaging process, an electron beam 161 emitted from the tip of the cathode 103 can be accelerated by the voltage of the anode 121, pass through the gun aperture 122, the beam-limiting aperture 125, and the converging lens 126, and is focused by the modified SORIL lens into a probe spot 170 and impacts the surface of the wafer 150. The probe spot 170 can be scanned on the surface of the wafer 150 by a deflector (such as deflector unit 132c in a SORIL lens or other deflectors). The detector 144 can collect secondary particles or scattered particles (such as secondary electrons emitted from the wafer surface or scattered primary electrons) to determine the beam intensity and make it possible to reconstruct an image of the region of interest on the wafer 150.

[0057] An image processing system 199 may also be provided, including an image acquirer 120, a storage device 130, and a controller 109. The image acquirer 120 may include one or more processors. For example, the image acquirer 120 may include a computer, server, mainframe, terminal, personal computer, any kind of mobile computing device, or a combination thereof. The image acquirer 120 may be connected to the detector 144 of the beam tool 104B via a medium such as an electrical conductor, fiber optic cable, portable storage medium, IR, Bluetooth, Internet, wireless network, radio, or a combination thereof. The image acquirer 120 may receive signals from the detector 144 and may construct an image. The image acquirer 120 may thus acquire an image of the wafer 150. The image acquirer 120 may also perform various post-processing functions, such as image averaging, contour generation, and overlaying indicators on the acquired image. The image acquirer 120 may be configured to perform adjustments such as brightness and contrast of the acquired image. The storage device 130 may be a storage medium such as a hard disk, random access memory (RAM), cloud storage device, or other types of computer-readable storage. Storage device 130 may be coupled to image acquirer 120 and may be used to store original image data of scans as original images and post-processed images. Image acquirer 120 and storage device 130 may be connected to controller 109. In some embodiments, image acquirer 120, storage device 130 and controller 109 may be integrated into a single electronic control unit. In some embodiments, image processing system 199 may be configured to detect defects in an image or enhance an image according to a POR recipe. In some embodiments, the POR recipe may set the defect type for detection and set a threshold for the defect type. In some embodiments, the POR recipe may set multiple image enhancement parameters to be applied to the image(s) acquired by image acquirer 120. In some embodiments, detected defects and their corresponding locations on a wafer or mask may be stored in storage device 130.

[0058] In some embodiments, the image acquirer 120 may acquire one or more images of a sample based on imaging signals received from the detector 144. The imaging signals may correspond to scanning operations for imaging charged particles. The acquired image may be a single image comprising multiple imaging regions, which may contain various features of the wafer 150. The single image may be stored in the storage device 130. Imaging may be performed based on imaging frames.

[0059] The converging and illumination optics of an electron beam tool may include or be supplemented by an electromagnetic quadrupole electron lens. For example, such as Figure 2BAs shown, the electron beam tool 104B may include a first quadrupole lens 148 and a second quadrupole lens 158. In some embodiments, the quadrupole lens can be used to control the electron beam. For example, the first quadrupole lens 148 can be controlled to adjust the beam current, and the second quadrupole lens 158 can be controlled to adjust the beam spot size and beam shape.

[0060] Figure 2B The illustration depicts a charged particle beam device that can use a single primary beam configured to generate secondary electrons through interaction with a wafer 150. A detector 144 can be positioned along an optical axis 105, as shown. Figure 2B As in the illustrated embodiment, the primary electron beam can be configured to travel along the optical axis 105. Therefore, the detector 144 can include a hole at its center, allowing the primary electron beam to pass through to reach the wafer 150. Figure 2B An example of a detector 144 with an opening at its center is shown. However, some embodiments may use a detector positioned off-axis relative to the optical axis along which the primary electron beam travels. For example, as discussed above. Figure 2B As in the illustrated embodiment, a beam splitter 222 can be provided to guide the secondary electron beam toward an off-axis detector. Figure 2A As shown, beam splitter 222 can be configured to direct the secondary electron beam toward electron detection device 244.

[0061] Images generated by SCPM can be used for defect inspection. For example, a generated image capturing a test equipment area of ​​a wafer can be compared with a reference image capturing the same test equipment area. The reference image can be predetermined (e.g., by simulation) and does not include known defects. If the difference between the generated image and the reference image exceeds a tolerance level, a potential defect can be identified. For another example, SCPM can scan multiple areas of a wafer, each including a test equipment area designed to be identical, and generate multiple images capturing these manufactured test equipment areas. These multiple images can be compared with each other. If the difference between the multiple images exceeds a tolerance level, a potential defect can be identified.

[0062] In some embodiments, machine learning can be used to generate inspection images, reference images, or other images associated with apparatus 100, 104A, or 104B. For example, in some embodiments, the machine learning system can be used with, for example... Figures 1 to 2B The controller 109 or 296, image processing system 199 or 290, image acquirer 120 or 292, or storage unit 130 or 294 operate in association. In some embodiments, machine learning may be associated with, for example... Figure 3 The offline computing platform 301 is associated with real-time recipe tuning methods, for example... Figure 3Method 300 or Figure 8 The 800, which will be described below. In some embodiments, the machine learning system may include a discriminative model. In some embodiments, the machine learning system may include a generative model. For example, learning can have two types of mechanisms: discriminative learning, which can be used to create classification and detection algorithms, and generative learning, which can be used to actually create models, in extreme cases, which can render images. For example, as further described below, a generative model can be configured to generate images from design clips that resemble corresponding locations on a wafer in an SEM image. This can be performed by 1) training the generative model with the design clips and associated actual SEM images from these locations on the wafer; and 2) feeding the model design clips in inference mode at locations where simulated SEM images are needed. Such simulated images can be used as reference images, for example, in die-to-database inspections.

[0063] If the model(s) include one or more discriminative models, then the discriminative models(s) can have any suitable architecture or configuration known in the art. A discriminative model (also called a conditional model) is a class of models in machine learning used to model the dependency of an unobserved variable “y” on an observed variable “x”. Within a probabilistic framework, this is achieved by modeling a conditional probability distribution P(y|x), which can be used to predict y based on x. Unlike generative models, discriminative models may not allow the generation of samples from the joint distribution of x and y. However, for tasks such as classification and regression that do not require a joint distribution, discriminative models may yield superior performance. On the other hand, generative models are generally more flexible than discriminative models in expressing dependencies in complex learning tasks. Furthermore, most discriminative models are inherently supervised and cannot be easily extended to unsupervised learning. The specific details of the application ultimately determine the suitability of choosing between discriminative and generative models.

[0064] Generative models can generally be defined as models that are inherently probabilistic. In other words, a "generative" model is not a model that performs forward simulation or a rule-based approach; therefore, it may not be necessary to model the physics of the process involved in generating the actual image or output (which is generating the simulated image or output). Instead, a generative model can be learned based on a suitable training dataset (because its parameters can be learned). Such a generative model can offer many advantages for the embodiments described herein. Furthermore, generative models can be configured with deep learning architectures because they can include multiple layers that can perform various algorithms or transformations. The number of layers included in a generative model can depend on the specific use case. For practical purposes, a suitable range of layers is from two to dozens.

[0065] Deep learning is a type of machine learning. Machine learning can often be defined as a form of artificial intelligence (AI) that gives computers the ability to learn without explicit programming. Machine learning focuses on developing computer programs that can learn, grow, and change when exposed to new data. Machine learning explores the research and construction of algorithms that can learn from and predict data—making data-driven predictions or decisions by building models from sample inputs. These algorithms overcome the limitations of strictly adhering to static program instructions.

[0066] The machine learning described herein can be further implemented as described in Sugiyama's "Introduction to Statistical Machine Learning" (Morgan Kaufman, 2016, p. 534); Jebara's "Discriminative, Generative, and Imitative Learning" (MIT, 2002, p. 212); and Hand et al.'s "Principles of Data Mining (Adaptive Computation and Machine Learning)" (MIT, 2001, p. 578), which are incorporated herein by reference as if fully stated herein. The embodiments described herein can also be configured as described in those references.

[0067] In some embodiments, a machine learning system may include a neural network. For example, the model may be a deep neural network with a set of weights that models the world based on data fed to it for training. Neural networks can generally be defined as a computational approach that loosely models how the biological brain solves problems by axonally connected clusters of biological neurons based on a relatively large set of neural units. Each neural unit is connected to many other neural units, and the influence of these links on the activation state of the connected neural units can be either coercive or inhibitory. These systems are self-learning and trained, rather than explicitly programmed, and excel in domains where solutions or feature detection are difficult to express in traditional computer programs.

[0068] Neural networks typically consist of multiple layers, with signal paths traversing from front to back. The goal of neural networks is to solve problems in the same way as the human brain, although some neural networks are far more abstract. Modern neural network projects often use thousands to millions of neurons and millions of connections to function. Neural networks can have any suitable architecture or configuration known in the art.

[0069] In yet another embodiment, the model may include convolutional and deconvolutional neural networks. For example, the embodiments described herein can leverage learning concepts such as convolutional and deconvolutional neural networks to solve representation transformation problems (e.g., rendering) that are typically difficult to solve. The model may have any convolutional and deconvolutional neural network configuration or architecture known in the art.

[0070] Now for reference Figure 3 , Figure 3 This is a flowchart of an example real-time recipe tuning method 300 for an inspection tool, consistent with embodiments of this disclosure. According to some embodiments of this disclosure, the real-time recipe tuning method 300 can be executed on an offline computing platform 301, which is separate from the inline tool 1001. In some embodiments, the inline tool 1001 can be an inspection tool, such as… Figure 1 Inspection system 100 Figure 2A Multi-beam tool 104A or Figure 2B The single-beam tool 104B. In some embodiments, the inline tool 1001 may be one or more online systems used in the HVM wafer process. For example, when multiple inspection systems in the HVM process use the same design data, the offline computing platform 301 may perform a real-time recipe tuning method 300 on these multiple inspection systems. In some embodiments, the offline computing platform 301 has the capability to perform image processing and defect detection to simulate the inspections and measurements performed in the inline tool 1001. Because the inline tool 1001 can continue to perform its operations while the offline computing platform 301 performs the recipe tuning method 300, recipe tuning of the inline tool 1001 can be performed without any impact on the throughput of the inline tool 1001.

[0071] According to some embodiments of this disclosure, the real-time recipe tuning method 300 can obtain... Figure 3 One or more input images 3001 are shown. In some embodiments, the input image 3001 can be acquired from the inline tool 1001. In some embodiments, the input image 3001 is streamed from the inline tool 1001 in real time as it is acquired. Although Figure 3 Although not illustrated, the input image 3001 can be stored on the offline computing platform 301, such as a database on the offline computing platform 301. According to some embodiments of this disclosure, the input image 3001 can be, for example, generated by... Figure 1 EBI system 100, Figure 2A Multi-beam tool 104A or Figure 2BThe inspection image is generated by the single-beam tool 104B. In some embodiments, the inspection image is an SCPM image of a sample or wafer. In some embodiments, the input image 3001 may be the original image scanned by the inspection tool. Although the input image 3001 is a single form, it should be understood that the input image 3001 may refer to a set of input images streamed from the inline tool 1001.

[0072] According to some embodiments of this disclosure, when an input image 3001 is streamed to an offline computing platform 301, the inline tool 1001 can stream a Proof-of-Order (POR) recipe, which includes parameter information for generating the input image 3001. The POR recipe can also be stored in the offline computing platform 301. In some embodiments, the inline tool 1001 can also stream its image inspection and measurement results of the input image 3001 along with the POR parameter recipe for performing image inspection and measurement on the input image 3001. For example, the POR parameter recipe for performing image inspection and measurement may include image enhancement parameters, defect type parameters, thresholds for determining defect types, etc. In some embodiments where the inline tool 1001 uses multiple bundles, the inline tool 1001 can also stream information about which of the multiple bundles is associated with the input image 3001. In some embodiments utilizing multiple online systems, the inline tool 1001 can also stream information about which of the multiple online systems is associated with the input image 3001.

[0073] According to some embodiments of this disclosure, parameter formulations can be set for experiments in a Design of Experiment (DOE) process 310. During the DOE process 310, parameter values ​​for a set of parameters of interest can be set for the experiment. In some embodiments, the parameter set may include at least one parameter in inline tool 1001 for inspecting and measuring inspection images. In some embodiments, the DOR process 310 can be iterated to traverse various combinations of parameters of interest. In some embodiments, when an input image 3001 is streamed to an offline computing platform 301, an experiment with a pre-designed or predetermined formulation(s) combination can be automatically triggered. According to some embodiments of this disclosure, the input image 3001 is used as input for image inspection for defect detection or measurement using a predetermined formulation(s) combination.

[0074] In some embodiments, the offline computing platform 301 may provide an interface that displays parameters organized in a hierarchical structure 400, which may facilitate the parameter value setting process. Figure 4A This is an example parameter structure for parameter setting consistent with embodiments of this disclosure. For example... Figure 4A As shown, parameter structure 400 can be configured to have multiple levels. Figure 4AIn this framework, each level can include multiple parameter groups belonging to different categories. For example, the first-level parameter L1 can include a first group of parameters PG1 related to region selection for inspection / measurement, a second group of parameters PG2 related to image enhancement, and a third group of parameters PG3 related to defect type. The second-level parameter L2 can include subgroups of parameters belonging to each parameter group PG1, PG2, or PG3 in the first-level L1. In some embodiments, subgroups of parameters belonging to a specific parameter group in the first-level L1 can be configured to be displayed on the interface after selecting the corresponding parameter group. Figure 4A In the diagram, parameter subgroups belonging to the third group PG3 are shown as selected, indicated by gray. For example, the second level L2 may include a first subgroup of parameter PSG31 associated with the first defect type, a second subgroup of parameter PSG32 associated with the second defect type, and a third subgroup of parameter PSG33 associated with the third defect type. Similarly, the third level parameter L3 may include parameter subgroups belonging to each of the upper parameter subgroups PSG31, PSG32, or PSG33 in the second level L2. Figure 4A In the diagram, the parameter subgroup belonging to the second subgroup PSG32 is shown as selected, indicated by a darker gray. For example, the third level L3 may include a first subgroup of parameters PSG321 related to a first threshold for determining a second defect type, a second subgroup of parameters PSG322 related to a second threshold for determining a second defect type, and a third subgroup of parameters PSG323 related to a third threshold for determining a second defect type.

[0075] Although the parameter structure 400 is illustrated to have three levels, each with three groups, it should be understood that any level with any number of parameter groups / parameters can be used in embodiments of this disclosure. It should also be understood that any parameter that can be used for inspection / measurement in the inline tool 1001 can be set and tested for optimization. In some embodiments, the parameter structure 400 may include additional parameters not used in the inline tool 1001. For example, when the inline tool 1001 uses a POR recipe to detect a first defect type and a second defect type, the parameter structure 400 may have parameters for detecting a third defect type.

[0076] Return to reference Figure 3 According to some embodiments of this disclosure, image enhancement (IH) parameters can be set and applied during DOE process 310. During IH process 3110, various image enhancement parameters can be set and applied to the input image 3001. Figure 4B The illustration shows an example enhanced image 3002 after various image enhancement parameters have been applied to an input image 3001, consistent with embodiments of this disclosure. For example... Figure 4BAs shown, the first set of images indicates the result after applying the first image enhancement parameter E1 to a set of input images 3001. Similarly, the second set of images indicates the result after applying the second image enhancement parameter E2 to the same set of input images 3001. Figure 4B The figure also illustrates the Nth set of images after the Nth image enhancement parameter EN is applied to the input image 3001.

[0077] According to some embodiments of this disclosure, image ranges with different characteristics can be filled by applying various image enhancement parameters to the input image 3001. Figure 4C An example enhanced image 3002, consistent with embodiments of this disclosure, is illustrated after applying various image enhancement parameters (e.g., E1 to EN) to an input image 3001. For example, enhanced image 401 can be obtained by applying a first image enhancement parameter to the input image 3001, and enhanced image 402 can be obtained by applying an Nth image enhancement parameter to the input image 3001. Figure 4C In this context, input image 3001 is also included as the original image without any image enhancement. (As will be discussed later...) Figure 5B , Figure 5C and Figure 5D As discussed, each enhanced image 3002 exhibits different characteristics that can make the enhanced image more suitable for certain defect inspections or measurements.

[0078] Return to reference Figure 3 According to some embodiments of this disclosure, during DOE process 310, inspection (DD) parameters can be set and applied. During DD process 3120, various inspection parameters can be set, and defect detection or measurement can be performed on the input image 3001 or the enhanced image 3002 according to the set inspection parameters.

[0079] Figure 5A The illustration shows an example detection result 3003 after applying various detection parameters to an enhanced image, consistent with embodiments of this disclosure. Figure 5A In the diagram, the first column C1 indicates that the first detection parameter D1 is applied to the enhanced image (e.g., ...). Figure 4B The result after enhancing the image (3002), and the second column C2 indicates the application of the second detection parameter D2. Figure 4B The result after enhancing image 3002. Similarly, the Mth column CM represents the result after applying the Mth detection parameter DM. Figure 4B The result after enhancing image 3002. According to some embodiments of this disclosure, Figure 5A The detection result 3003 can be used to evaluate the performance of each combination of two parameters: image enhancement parameters (e.g., E1 to EN) and detection parameters (e.g., D1 to DM).

[0080] According to some embodiments of this disclosure, each enhanced image 3002 having different image enhancement parameters than an original input image 3001 can exhibit different characteristics. Figure 5B , 5C The 5D illustrations show example enhanced images that are more suitable for certain defect detection or measurement. Figure 5B Example images more suitable for critical size measurement, consistent with embodiments of this disclosure, are illustrated. Figure 5B In this context, image 510 can be the original image without any image enhancement parameters applied, such as input image 3001. Figure 5B As shown, image 510 may have properties suitable for measuring the critical dimension 5001 of a structure with a specific pattern. It should be understood that image 510 and its critical dimension 5001 measurement are provided as an example for illustration; any enhancements can be applied to the input image to fit any other type of critical dimension. Figure 5C Example images suitable for bridging defect detection, consistent with embodiments of this disclosure, are illustrated. Figure 5C In this context, image 520 can be an enhanced image after applying specific image enhancement parameters (such as higher contrast) to input image 3001. For example... Figure 5C As shown, image 520 may have characteristics suitable for detecting bridging defects 5002. Figure 5D Example images suitable for detecting damage and defects, consistent with embodiments of this disclosure, are illustrated. Figure 5D In this context, after applying specific image enhancement parameters (such as lower brightness) to the input image 3001, image 530 can be an enhanced image. For example... Figure 5D As shown, image 530 may have characteristics suitable for detecting defects 5003.

[0081] Return to reference Figure 3 According to some embodiments of this disclosure, in the DD process 3120, one or more additional detection parameters can be set, and defect detection or measurement can be performed on the input image 3001 or the enhanced image 3002 according to the set detection parameters. Figure 6A The illustration shows example results after applying various threshold parameters to a first defect, consistent with embodiments of this disclosure. Figure 6A In the diagram, the first column C11 indicates that the first threshold T1 of the first detection parameter D1 is applied to the enhanced image (e.g., ...). Figure 4B The result after enhancing the image (3002), and the second column C12 indicates the second threshold T2 applied to the first detection parameter D1. Figure 4B The result after enhancing image 3002. Similarly, column K, C1K, represents the result after applying the Kth threshold TK of the first detection parameter D1. Figure 4BThe result after enhancing image 3002. According to some embodiments of this disclosure, Figure 6A The results can be used to evaluate the performance of each combination of the three parameters: image enhancement parameters (e.g., E1), first detection parameters (e.g., D1), and various thresholds (e.g., T1 to TK).

[0082] According to some embodiments of this disclosure, a defect can be identified when a certain measurement exceeds a set threshold. In some embodiments, the threshold for detecting a particular defect type can be determined based on the most stable critical size suitable for a particular pattern. Figure 6B The illustration shows the effect of threshold parameter variations, consistent with embodiments of this disclosure, on defect detection. Figure 6B The diagram illustrates two bridging defect candidates, 6004 and 6005. Depending on the applied threshold, these two candidates can be identified as either defects or non-defects. For example, a bridging defect can be identified when the distance between two adjacent structures is less than the threshold. When the threshold is set too low, neither candidate 6004 nor 6005 can be identified as a bridging defect. Conversely, when the threshold is set too high, both candidates 6004 and 6005 can be identified as bridging defects. When the threshold is set in between, the first candidate 6004 may not be identified as a bridging defect, while the second candidate 6005 may be identified as one. Setting an appropriate threshold for a particular defect type can improve the defect detection rate and reduce the nuisance rate. Although embodiments have been illustrated with a finite number of parameter combinations including a finite number of parameters, it is to be understood that in DOE process 310, any number of parameter combinations including any number of parameters can be used to generate various experimental results, based on which the performance of setting parameter combinations can be evaluated.

[0083] Return to reference Figure 3According to some embodiments of this disclosure, detection analysis can be performed based on detection results 3003. In some embodiments, detection analysis can be performed for each parameter combination. In some embodiments, defects found from detection results 3003 can be classified as defects during analysis process 320. In some embodiments, defects found from detection results 3003 can be associated with corresponding locations or design data on the wafer. In some embodiments, an interference rate can be determined from detection results 3003 during analysis process 320. For example, the ratio of the number of disruptive defects to the total number of defects found from detection results 3003 can be determined. In some embodiments, defects found from detection results 3003 that are not disruptive defects can be classified as captured defects. A capture rate can also be determined for each parameter combination during analysis process 320. In some embodiments, captured defects can be grouped according to their defect type. In some embodiments, the process time for defect detection or measurement can be determined for each parameter combination.

[0084] In some embodiments, defects or their corresponding locations can be identified by comparing the detection result 3003 with the corresponding design data. In some embodiments, the design data can be a layout file of a wafer design, such as a Gold Image or Graphical Database System (GDS) format, Graphical Database System II (GDS II) format, Open Art System Exchange Standard (OASIS) format, Caltech Intermediate Format (CIF), etc. The wafer design can include patterns or structures for inclusion on the wafer. The patterns or structures can be mask patterns used to transfer features from a lithographic mask or stencil to the wafer. In some embodiments, the layout in GDS or OASIS formats can include feature information stored in a binary file format, representing planar geometry, text, and other information related to the wafer design. In some embodiments, design data can also be obtained from the inline tool 1001 when the corresponding input image 3001 is streamed from the inline tool 1001.

[0085] In some embodiments, captured defects can be compared with results from inline tool 1001. In some embodiments, when a new defect is found in detection result 3003 that has not yet been found by inline tool 1001, the new defect is indicated as a newly discovered defect. In some embodiments, when a new defect type that has not been found by inline tool 1001 is found in detection result 3003, the new defect type is indicated as a newly discovered defect type. In some embodiments, when a defect found by inline tool 1001 is not found in detection result 3003, the defect is indicated as a missing defect.

[0086] Figure 7A The illustration shows a first example analysis result of a combination of setting parameters consistent with embodiments of this disclosure. Figure 7A Figure 710 illustrates the defect detected at the corresponding location on wafer 711. Figure 7A In this context, defect 7001 indicates a defect already found by inline tool 1001, and defect 7002 indicates a newly discovered defect that offline computing platform 301 has combined with setting parameters. In some embodiments, offline computing platform 301 may not detect existing defect 7001, which has been found by inline tool 1001 and may be referred to as a missing defect in this disclosure. In some embodiments, although Figure 7A Although not illustrated in Figure 710, missing defects can also be indicated in Figure 710. While Figure 710 illustrates a defect of the first defect type, it is important to understand that a similar analysis can be performed on any defect type. Figure 7B The illustration shows a second example analysis result of a combination of setting parameters consistent with embodiments of the present disclosure. Graph 720 shows the number of new defects 7002 compared to the number of existing defects 7001 of the first defect type D1 and the second defect type D2. In some embodiments, graph 720 may show the ratio of the number of new defects 7002 to the number of existing defects 7001.

[0087] According to some embodiments of this disclosure, refer to Figure 3 It can generate analysis results 3004 (such as) for each parameter combination including two or more parameters or for each defect type. Figure 7A and Figure 7B The results shown are as follows. In some embodiments, the analysis results 3004 may be provided in any other form that may facilitate qualitative or quantitative evaluation of the parameter combinations. According to some embodiments of this disclosure, the analysis results 3004 may also include the interference rate for each parameter combination or each defect type. According to some embodiments of this disclosure, the analysis results 3004 may have a capture rate or interference rate for each defect type or each parameter combination. According to some embodiments of this disclosure, the analysis results 3004 may include throughput information, such as process time, for each defect type or each parameter combination. According to some embodiments of this disclosure, the results may be evaluated by a single-beam inspection system (e.g., Figure 4B The input image 3001 generated by the single-beam tool 104B or by a multi-beam inspection system (e.g., Figure 4A The output image 3001 generated by each beam of the multi-beam tool 104A is analyzed using process 320. Furthermore, in some embodiments, analysis can be performed on each beam channel (e.g., ...). Figure 4B The single-beam tool 104B has a single beam or Figure 4AEach beam of the multi-beam tool 104A generates analysis results 3004. According to some embodiments of this disclosure, deep learning algorithms can be applied to the analysis process 320. In some embodiments, deep learning algorithms can be used to group defects into multiple groups according to defect type to determine capture rates or interference rates, etc. Because the offline computing platform 301, separate from the inline tool 1001, performs the recipe tuning of the inline tool 1001, large amounts of data can be analyzed in real time using a computationally intensive deep learning model without increasing the wafer production cycle.

[0088] Return to reference Figure 3 According to some embodiments of this disclosure, parameter optimization can be performed based on the analysis results 3004. In some embodiments, during the optimization process 330, a combination of parameters optimized for each defect type detection can be found. According to some embodiments of this disclosure, key performance indicators (KPIs) can be set to evaluate the performance of the parameter combination used to detect a specific defect type. For example, a KPI for a specific defect type can indicate a range of capture rate, throughput, etc. In some embodiments, the KPI can vary depending on the defect type, implementation method, requirements, etc. For example, the KPI can be determined based on various criteria, such as defect threshold, defect intensity, grayscale histogram, image sharpness, etc.

[0089] Figure 7C Example KPI charts for each parameter combination consistent with embodiments of this disclosure are illustrated. Figure 7C In the graph, the X-axis represents the value of the first parameter P1, and the Y-axis represents the value of the second parameter P2. This graph illustrates the capture rate as a KPI based on the combination of the first parameter P1 and the second parameter P2. For example, in Figure 7C In the graph, the capture rate is 60% to 80% under the values ​​of the first parameter P1 (p11) and the second parameter P2 (p21). Similarly, the graph can indicate the capture rate for any combination of the first parameter P1 and the second parameter P2.

[0090] According to some embodiments of this disclosure, parameter combinations can be automatically ranked based on their KPI values ​​during process optimization 330. For example, parameter combinations can be ranked according to their capture rate for a specific defect type, such that the parameter combination with the highest capture rate can be ranked higher. In some embodiments where two parameter combinations have the same capture rate, the parameter combinations can be ranked according to additional criteria. For example, the parameter combination that provides more stable defect detection results can be ranked higher. Figure 7CIn this example, parameter combinations falling within the first region A1 and the second region A2 have the same capture rate of 60% to 80%. In this example, a first parameter combination with the value p11 of the first parameter P1 and the value p21 of the second parameter P2 can rank higher than a second parameter combination with the values ​​p12 of the first parameter P1 and p22 of the second parameter P2 because the first parameter combination is located in a wider region (i.e., A1) than the second parameter combination in the second region A2. In this case, when using the first parameter combination, defect detection results can be more stable even if the parameters are not precisely calibrated. It should be understood that various ranking strategies can be applied to some embodiments of this disclosure. In some embodiments, parameter combinations that only meet specific criteria can be automatically ranked. For example, the criterion could be a defect capture rate equal to or greater than 60%. In this example, parameter combinations with a capture rate equal to or greater than 60% can be ranked for inline production.

[0091] According to some embodiments of this disclosure, deep learning algorithms can be applied to optimize process 330. In some embodiments, deep learning algorithms can be trained to determine optimized parameter combinations based on analysis results 3004 and set criteria. In some embodiments, the deep learning algorithm can automatically rank parameter combinations according to its set criteria and KPIs. Because the offline computing platform 301, separate from the inline tool 1001, performs the recipe tuning of the inline tool 1001, large amounts of data can be analyzed in real time using a computationally intensive deep learning model without increasing the wafer production cycle. According to some embodiments of this disclosure, optimization results, including optimized parameter combinations 3005 for specific defect types or ranked parameter combinations, can be stored in the offline computing platform 301, for example, for future reference. In some embodiments, parameter combinations ranked using different KPIs or criteria (e.g., throughput) can also be stored in the offline computing platform 301. In some embodiments, only a certain percentage or a certain number of top-ranked parameter combinations can be stored in the offline computing platform 301.

[0092] According to some embodiments of this disclosure, a parameter combination 3005, determined to be an optimized combination of parameters for a specific defect type, can be provided to the inline tool 1001. In some embodiments, the parameter combination 3005 can be used in the real-time inline tool 1001 to detect defects of the defect type. In some embodiments, the inline tool 1001 can update the parameters of the POR recipe corresponding to the parameter combination 3005 and continue to perform checks using the updated parameters. According to some embodiments of this disclosure, the parameter combination 3005, determined to be an optimized combination of parameters for a specific defect type, can also be used in the DOE process 310. For example, during the DOE process 310, the parameter range for experiments used to detect a specific defect type can be reduced based on the provided parameter combination 3005. Therefore, more accurate experimental results can be obtained with less experimental time.

[0093] Return to reference Figure 3 According to some embodiments of this disclosure, weakness prediction can be performed based on the analysis results 3004. In some embodiments, during the prediction process 340, pattern analysis can be performed to determine patterns(s) of defects(s) more likely to have a specific defect type based on the patterns of the found defects and design data. Patterns of defects(s) likely to have a specific defect type can be determined based on the analysis results 3004. In some embodiments, pattern search can be applied to find additional locations on a mask or mask pattern that may have a specific defect type. In some embodiments, pattern search can be performed to find locations(s) with patterns(s) on the design data that are similar to or identical to the patterns of defects(s) having a specific defect type.

[0094] According to some embodiments of this disclosure, weakness data 3006 may be provided to inline tool 1001. In some embodiments, inline tool 1001 may utilize weakness data 3006 to inspect wafer portions corresponding to weakness locations that may have defects of a specific defect type. In some embodiments, inline tool 1001 may use parameter combination 3005 to inspect wafer portions corresponding to weakness locations of a specific defect type, the parameter combination 3005 being determined as an optimized parameter combination for that specific defect type.

[0095] According to some embodiments of this disclosure, because the inspection performance of specific parameter combinations is monitored during real-time recipe tuning 300, HVM process variations or inspection tool drift can be noticed and fixed at an early stage. Furthermore, according to some embodiments of this disclosure, because all tuned parameter combinations are preserved, tool-to-tool matching can also be provided to achieve better accuracy and consistency in defect inspection or measurement. For example, by comparing the performance of different runs, the POR recipe of an inspection tool can be compared and verified between different runs on the same tool. In some embodiments using multiple bundles, the POR recipe of each bundle can be compared and verified between two different bundles. In some embodiments using two different systems, the POR recipe can be compared and verified between two different tools.

[0096] Figure 8 This is a flowchart illustrating an example method for real-time recipe tuning for an online inspection system, consistent with embodiments of this disclosure. The steps of method 800 can be performed by an offline computing platform (e.g., Figure 3 The offline computing platform 301) is used for execution. It should be understood that the illustrated method 800 can be modified to change the order of the steps and include additional steps.

[0097] In step S810, one or more input images can be acquired from a separate wafer inspection system. In some embodiments, the wafer inspection system can be an online inspection tool, such as... Figure 3 The inline tool 1001. In some embodiments, the input image may be... Figure 3 The input image 3001. In some embodiments, the input image 3001 is streamed from the inline tool 1001 in real time when acquired. According to some embodiments of this disclosure, when the input image 3001 is acquired, a POR recipe including parameter information for generating the input image 3001 can also be acquired from the inline tool 1001. In some embodiments, the image inspection and measurement results of the inline tool for the input image 3001 can also be acquired together with the POR parameter recipe for performing image inspection and measurement on the input image 3001.

[0098] In step S820, a parameter formula can be set, and detection results based on the set parameter formula can be generated. In some embodiments, the experimental results may be... Figure 3The detection result is 3003. In step S820, parameter values ​​for a set of parameters of interest can be set for the experiment. In some embodiments, the parameter set may include at least one parameter in the inline tool 1001 for inspecting and measuring the inspected image. In some embodiments, step S820 may be iterated to traverse various combinations of parameters of interest. In some embodiments, when the input image 3001 is streamed to the offline computing platform 301, an experiment with a pre-designed or predetermined combination of recipes may be automatically triggered. According to some embodiments of this disclosure, the input image 3001 is used as input for image inspection for defect detection or measurement using a combination of predetermined recipes.

[0099] In some embodiments, in step S820, image enhancement (IH) parameters may be set and applied. In step S820, various image enhancement parameters may be set and applied to the input image 3001. According to some embodiments of this disclosure, image ranges with different characteristics can be filled by applying various image enhancement parameters to the input image 3001. In some embodiments, in step S820, detection parameters may be set and applied. In step S820, various detection parameters may be set, and defect detection or measurement can be performed on the input image 3001 or the enhanced image 3002 according to the set detection parameters. In some embodiments, the detection parameters may include size measurement criterion parameters, defect type parameters, threshold parameters for detecting specific defect types, etc. In step S820, the detection result may include critical size measurements or defects found on the image according to the set parameter formula.

[0100] In step S830, a detection analysis can be performed based on the detection results. In some embodiments, the detection analysis can be performed for each parameter combination. In some embodiments, in step S830, defects found from the detection results 3003 can be listed as defects. In some embodiments, defects found from the detection results 3003 can be associated with corresponding locations or design data on the wafer. In some embodiments, in step S830, an interference rate can be determined from the detection results 3003. In some embodiments, defects found from the detection results 3003 that are not obstructive defects can be listed as captured defects. During the analysis process 320, a capture rate can also be determined for each parameter combination. In some embodiments, captured defects can be grouped according to their defect type. In some embodiments, the process time for defect detection or measurement can be determined for each parameter combination. In some embodiments, captured defects can be compared with results from the inline tool 1001. In some embodiments, when a new defect is found from the detection results 3003, if the new defect has not yet been found by the inline tool 1001, the new defect is indicated as a newly discovered defect. In some embodiments, when a new defect type not detected by the inline tool 1001 is found in the detection result 3003, the new defect type is indicated as a newly discovered defect type. In some embodiments, when a defect detected by the inline tool 1001 is not found in the detection result 3003, the defect is indicated as a missing defect.

[0101] In step S840, parameter optimization can be performed based on the analysis results from step S830. In some embodiments, the analysis results may be... Figure 3 The analysis result 3004 is provided. In some embodiments, parameter combinations optimized for each defect type detection can be found in step S840. According to some embodiments of this disclosure, key performance indicators (KPIs) can be set to evaluate the performance of parameter combinations used to detect a specific defect type. For example, the KPI for a specific defect type can indicate a range of capture rate, throughput, etc. In some embodiments, the KPI can vary depending on the defect type, implementation method, requirements, etc. In some embodiments, in step S840, parameter combinations can be automatically ranked according to their KPI values. For example, parameter combinations can be ranked according to the capture rate of a specific defect type, so that the parameter combination with the highest capture rate can be ranked higher. It should be understood that various ranking strategies can be applied to some embodiments of this disclosure. In some embodiments, parameter combinations that only meet specific criteria can be automatically ranked.

[0102] In step S860, a combination of parameters determined to be an optimized combination of parameters for a specific defect type can be provided to the inline tool 1001. In some embodiments, the parameter combination may be... Figure 3The parameter combination 3005. In some embodiments, the parameter combination 3005 can be used in the real-time inline tool 1001 to detect defects of a defect type. In some embodiments, the inline tool 1001 can update the parameters of the POR recipe corresponding to the parameter combination 3005 and continue to perform checks using the updated parameters. According to some embodiments of this disclosure, the parameter combination 3005, which is determined to be an optimized parameter combination for a specific defect type, can also be used in step S820. For example, in step S820, the parameter range for the experiment used to detect a specific defect type can be reduced based on the provided parameter combination 3005.

[0103] In step S850, weakness prediction may be performed. In some embodiments, in step S850, pattern analysis may be performed to determine (multiple) patterns of defects more likely to have a specific defect type based on the patterns of the found defects and design data. Patterns of defects likely to have a specific defect type may be determined based on analysis results 3004. In some embodiments, pattern search may be applied to find additional locations on a mask or mask pattern that may have a specific defect type. In some embodiments, pattern search may be performed to find (multiple) locations on the design data that have patterns similar to or identical to the patterns of defects having a specific defect type.

[0104] In step S860, weakness data may be provided to the inline tool 1001. In some embodiments, the inline tool 1001 may use the weakness data to examine wafer portions corresponding to weakness locations that may have a specific defect type. In some embodiments, the inline tool 1001 may use a parameter combination 3005 to examine wafer portions corresponding to weakness locations of a specific defect type, the parameter combination 3005 being determined as an optimized parameter combination for that specific defect type.

[0105] Figure 9 This is a block diagram illustrating an example computer system 900 that can assist in implementing the methods, processes, or apparatus disclosed herein. In some embodiments, the computer system 900 can assist in implementing... Figure 1 Controller 109 Figure 2A Image processing system 290 or Figure 2B At least a portion of the image processing system 199. In some embodiments, the computer system 900 may assist in implementing... Figure 3 The process or offline computing platform 301 or Figure 8The method 800 is at least a part of the computer system 900. The computer system 900 includes a bus 902 or other communication mechanism for transmitting information and a processor 904 (or multiple processors 904 and 905, which may include hardware accelerators) coupled to the bus 902 for processing information. The computer system 900 also includes a main memory 906, such as random access memory (RAM) or other dynamic storage device, coupled to the bus 902 for storing information and instructions to be executed by the processor 904. The main memory 906 may also be used to store temporary variables or other intermediate information during the execution of instructions to be executed by the processor 904. The computer system 900 also includes a read-only memory (ROM) 908 or other static storage device coupled to the bus 902 for storing static information and instructions for the processor 904. A storage device 910, such as a magnetic disk or optical disk, is provided and coupled to the bus 902 for storing information and instructions.

[0106] Computer system 900 can be coupled to display 912, such as a cathode ray tube (CRT) or flat panel or touchpad display, via bus 902 for displaying information to the computer user. Input device 914, including alphanumeric keys and other keys, is coupled to bus 902 for transmitting information and command selections to processor 904. Another type of user input device is cursor control 916, such as a mouse, trackball, or cursor arrow keys, for transmitting directional information and command selections to processor 904 and controlling cursor movement on display 912. This input device typically has two degrees of freedom on two axes (a first axis, e.g., x) and a second axis, e.g., y), allowing the device to specify its position in a plane. Touchpad (screen) displays can also be used as input devices.

[0107] According to some embodiments, portions of one or more methods described herein can be executed by computer system 900 in response to processor 904 executing one or more sequences of one or more instructions contained in main memory 906. Such instructions may be read into main memory 906 from another computer-readable medium, such as storage device 910. Execution of the instruction sequence contained in main memory 906 causes processor 904 to perform the process steps described herein. One or more processors in a multiprocessor arrangement may also be employed to execute the instruction sequence contained in main memory 906. In alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions. Therefore, the description herein is not limited to any specific combination of hardware circuitry and software.

[0108] As used herein, the term "computer-readable medium" refers to any medium that participates in providing instructions to the processor 904 for execution. Computer-readable media includes non-transitory computer media. Computer-readable media can store information for use with a controller (e.g., Figure 1 The processor of the controller 109) Figure 2A The image processing system 290's processor, Figure 2B Image processing system 199 processor or Figure 3 The instructions of the processor in the offline computing platform 301 are used to perform image inspection, image acquisition, stage positioning, beam focusing, electric field adjustment, beam bending, converging lens adjustment, activation of charged particle sources, and beam deflection. Figure 3 Method 300 and Figure 8 Method 800, etc. This medium can take many forms, including but not limited to non-volatile media, volatile media, and transmission media. Non-volatile media include, for example, optical discs or magnetic disks, such as storage devices 910. Volatile media include dynamic memory, such as main memory 906. Transmission media include coaxial cables, copper wires, and optical fibers, including the wires forming bus 902. Transmission media can also take the form of sound waves or light waves, such as those generated during radio frequency (RF) and infrared (IR) data communication. Common forms of computer-readable media include, for example, floppy disks, flexible disks, solid-state drives, hard disks, magnetic tape or any other magnetic data storage media, compact disc read-only memory (CD-ROM), DVDs, any other optical data storage media, punched cards, paper tape, any other physical media with a perforated pattern, random access memory (ROM), programmable read-only memory (PROM) and erasable programmable read-only memory (EPROM), flash memory EPROM, any other flash memory, non-volatile random access memory (NVRAM), cache, registers, any other memory chip or cassette tape, the carrier waves described below, or any other media from which a computer can read, and their networked versions.

[0109] Various forms of computer-readable media may involve carrying one or more sequences of one or more instructions to processor 904 for execution. For example, the instructions may initially be carried on a disk of a remote computer. The remote computer may load the instructions into its dynamic memory and transmit the instructions over a telephone line using a modem. A modem local to computer system 900 may receive data over the telephone line and convert the data into an infrared signal using an infrared transmitter. An infrared detector coupled to bus 902 may receive the data carried in the infrared signal and place the data on bus 902. Bus 902 carries the data to main memory 906, from which processor 904 retrieves and executes the instructions. Instructions received by main memory 906 may optionally be stored on storage device 910 before or after execution by processor 904.

[0110] Computer system 900 may also include a communication interface 918 coupled to bus 902. Communication interface 918 provides bidirectional data communication coupling to network link 922, which is connected to local network 922. For example, communication interface 918 may be an Integrated Services Digital Network (ISDN) card or modem to provide data communication connectivity with a corresponding type of telephone line. As another example, communication interface 918 may be a Local Area Network (LAN) card to provide data communication connectivity with a compatible LAN. A wireless link may also be implemented. In any such implementation, communication interface 918 transmits and receives electrical, electromagnetic, or optical signals carrying digital data streams representing various types of information.

[0111] Network link 920 typically provides data communication to other data devices via one or more networks. For example, network link 920 may provide a connection to host computer 924 or data equipment operated by Internet Service Provider (ISP) 926 via local network 922. ISP 926, in turn, provides data communication services via the Global Packet Data Network (now generally referred to as "Internet" 928). Both local network 922 and Internet 928 use electrical, electromagnetic, or optical signals carrying digital data streams. Signals through various networks, as well as signals on network link 920 and through communication interface 918 (carrying digital data to and from computer system 900), are exemplary forms of carrier waves for transmitting information.

[0112] Computer system 900 can send messages and receive data, including program code, via networks(s), network link 920, and communication interface 918. In the Internet example, server 930 can send request codes for an application via the Internet 928, ISP 926, local network 922, and communication interface 918. For example, such a download application can provide all or part of the methods described herein. The received code can be executed by processor 904 upon receipt and / or stored in storage device 910 or other non-volatile storage device for later execution. In this way, computer system 900 can obtain application code in carrier form.

[0113] The embodiments may be further described using the following terms: 1. A method for real-time parameter tuning of a wafer inspection system, comprising: Acquire one or more input images streaming from the wafer inspection system; Multiple image enhancement parameters are applied to one or more input images to generate multiple images with different properties; Defects are identified from multiple images by applying multiple defect detection parameters; and Based on defects identified on multiple images, a combination of multiple image enhancement parameters and multiple defect detection parameters is determined for use in detecting a first defect type. 2. The method according to Clause 1 also includes: To identify newly discovered defects that were not previously detected by the wafer inspection system; and The defect has been detected by the wafer inspection system and was not previously identified from multiple images. 3. The method according to Clause 1 or 2, wherein identifying defects from multiple images includes: Identify the defects and their corresponding locations. 4. The method according to any one of clauses 1 to 3 also includes: Defects are grouped by defect type. 5. According to the method in Clause 4, determining the combination of parameters includes: The parameter combination is determined by grouping defects by defect type. 6. The method according to any one of clauses 1 to 5, wherein determining the combination of parameters includes: Deep learning algorithms are used to determine the parameter combination. 7. The method according to any one of clauses 1 to 6 also includes: The determined parameter combination is provided to the wafer inspection system, which then updates the recoded process formula in real time based on the determined parameter combination used for inline production. 8. The method according to any one of clauses 1 to 7 also includes: Weaknesses in design data are predicted based on defects identified on multiple images. 9. According to the method in Clause 8, the weaknesses predicted in the design data include: Weaknesses in design data are predicted by applying pattern search based on pattern data of defects identified on multiple images. 10. The method pursuant to Clause 8 or 9 also includes: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 11. The method according to any one of clauses 1 to 10 also includes: Set key performance indicators (KPIs); and The various combinations of multiple image enhancement parameters and multiple defect detection parameters are ranked based on the corresponding KPI values ​​determined by defects identified on multiple images. 12. The method according to Clause 11, wherein the KPI is at least defect capture rate or throughput. 13. The method according to any one of clauses 1 to 12, wherein the method is performed in parallel while the wafer inspection system is in production. 14. The method according to any one of Clauses 1 to 13, wherein the wafer inspection system is a separate wafer inspection system. 15. A method for real-time parameter tuning of a wafer inspection system, comprising: Acquire one or more input images streaming from the wafer inspection system; The first parameter combination is applied to one or more input images, and identifies a first set of defects in one or more input images; The second parameter combination is applied to one or more input images, and identifies a second set of defects in one or more input images; The first key performance indicator (KPI) value is determined based on the first defect set and the first parameter combination. The second KPI value is determined based on the second defect set, and the second parameter combination is used to determine the second KPI value; and The first parameter combination and the second parameter combination are ranked based on the first KPI value and the second KPI value. 16. The method pursuant to Clause 15 also includes: Newly discovered defects that were not previously detected by the wafer inspection system are identified in the first defect set; and The identifier is a missing defect that has been detected by the wafer inspection system and was not previously identified by the first defect set. 17. The method according to Clause 15 or 16, wherein identifying the first set of defects includes: Identify the first set of defects and their corresponding locations. 18. The method pursuant to any one of clauses 15 to 17 further includes: The first defect set is grouped by defect type. 19. The method according to Clause 18, wherein determining the combination of parameters includes: The parameter combination is determined by grouping defects by defect type. 20. The method according to Clause 19, wherein the first KPI value and the second KPI value are associated with the first defect type. 21. The method according to any one of clauses 15 to 20, wherein ranking the first combination of parameters and the second combination of parameters includes: A deep learning algorithm is used to rank the first and second parameter combinations. 22. The method pursuant to any one of clauses 15 to 21 further includes: The parameter combination that ranks higher between the first and second parameter combinations is provided to the wafer inspection system, enabling the wafer inspection system to update the recoded process in real time based on the parameter combination provided for inline production. 23. The method pursuant to any one of clauses 15 to 22 further includes: Weaknesses in the design data are predicted based on the first set of defects. 24. According to the method of Clause 23, the weaknesses in the predicted design data include: Weaknesses in design data are predicted by applying pattern search based on pattern data from a first set of defects. 25. The method pursuant to Clause 23 or 24 also includes: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 26. The method according to any one of clauses 15 to 25, wherein the KPI is at least defect capture rate or throughput. 27. The method according to any one of Clauses 15 to 26, wherein the wafer inspection system is a separate wafer inspection system. 28. An apparatus for real-time parameter tuning of a wafer inspection system, comprising: Memory, storing instruction sets; and At least one processor is configured to execute a set of instructions to cause the device to perform operations, including: Acquire one or more input images streaming from the wafer inspection system; Multiple image enhancement parameters are applied to one or more input images to generate multiple images with different properties; Defects are identified from multiple images by applying multiple defect detection parameters; and Based on defects identified on multiple images, a combination of multiple image enhancement parameters and multiple defect detection parameters is determined for use in detecting a first defect type. 29. The apparatus pursuant to Clause 28, wherein operation further includes: To identify newly discovered defects that were not previously detected by the wafer inspection system; and The defect has been detected by the wafer inspection system and was not previously identified from multiple images. 30. The apparatus according to clause 28 or 29, wherein, when identifying defects from multiple images, at least one processor is configured to execute a set of instructions to cause the apparatus to perform further actions: Identify the defects and their corresponding locations. 31. The apparatus according to any one of clauses 28 to 30, wherein operation further includes: Defects are grouped by defect type. 32. The apparatus according to clause 31, wherein, when determining the combination of parameters, at least one processor is configured to execute a set of instructions to cause the apparatus to further perform: The parameter combination is determined by grouping defects by defect type. 33. An apparatus according to any one of clauses 28 to 32, wherein, when determining the combination of parameters, at least one processor is configured to execute a set of instructions to cause the apparatus to further perform: Deep learning algorithms are used to determine the parameter combination. 34. The apparatus according to any one of clauses 28 to 33, wherein operation further includes: The determined parameter combination is provided to the wafer inspection system, which then updates the recoded process formula in real time based on the determined parameter combination used for inline production. 35. The apparatus according to any one of clauses 28 to 34, wherein operation further includes: Weaknesses in design data are predicted based on defects identified on multiple images. 36. The apparatus according to clause 35, wherein, when predicting weaknesses in design data, at least one processor is configured to execute a set of instructions to cause the apparatus to perform further actions: Weaknesses in design data are predicted by applying pattern search based on pattern data of defects identified on multiple images. 37. The apparatus pursuant to clause 35 or 36, wherein operation further includes: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 38. The apparatus according to any one of clauses 28 to 37, wherein operation further includes: Set key performance indicators (KPIs); and The various combinations of multiple image enhancement parameters and multiple defect detection parameters are ranked based on the corresponding KPI values ​​determined by defects identified on multiple images. 39. The apparatus pursuant to Clause 38, wherein the KPI is at least defect capture rate or throughput. 40. An apparatus according to any one of clauses 28 to 39, wherein at least one processor is configured to execute a set of instructions to cause the apparatus to perform operations in parallel when the wafer inspection system is in production. 41. An apparatus according to any one of clauses 28 to 40, wherein the chip inspection system is separate from the apparatus. 42. An apparatus for real-time parameter tuning of a wafer inspection system, comprising: Memory, storing instruction sets; and At least one processor is configured to execute a set of instructions to cause the device to perform operations, including: Acquire one or more input images streaming from the wafer inspection system; The first parameter combination is applied to one or more input images, and identifies a first set of defects in one or more input images; The second parameter combination is applied to one or more input images, and identifies a second set of defects in one or more input images; The first key performance indicator (KPI) value is determined based on the first defect set and the first parameter combination. The second KPI value is determined based on the second defect set, and the second parameter combination is used to determine the second KPI value; and The first parameter combination and the second parameter combination are ranked based on the first KPI value and the second KPI value. 43. The apparatus pursuant to Clause 42, wherein operation further includes: Newly discovered defects that were not previously detected by the wafer inspection system are identified in the first defect set; and The identifier is a missing defect that has been detected by the wafer inspection system and was not previously identified by the first defect set. 44. An apparatus pursuant to clause 42 or 43, wherein, upon identifying the first defect set, at least one processor is configured to execute a set of instructions to cause the apparatus to further perform: Identify the first set of defects and their corresponding locations. 45. The apparatus according to any one of clauses 42 to 44, wherein operation further includes: The first defect set is grouped by defect type. 46. ​​The apparatus according to clause 45, wherein, when determining the combination of parameters, at least one processor is configured to execute a set of instructions to cause the apparatus to further perform: The parameter combination is determined by grouping defects by defect type. 47. The apparatus according to Clause 46, wherein the first KPI value and the second KPI value are associated with a first defect type. 48. An apparatus according to any one of clauses 42 to 47, wherein, in ranking the first combination of parameters and the second combination of parameters, at least one processor is configured to execute a set of instructions to cause the apparatus to further perform: A deep learning algorithm is used to rank the first and second parameter combinations. 49. The apparatus according to any one of clauses 42 to 48, wherein operation further includes: The parameter combination that ranks higher between the first and second parameter combinations is provided to the wafer inspection system, enabling the wafer inspection system to update the recoded process in real time based on the parameter combination provided for inline production. 50. The apparatus according to any one of clauses 42 to 49, wherein operation further includes: Weaknesses in the design data are predicted based on the first set of defects. 51. The apparatus according to clause 50, wherein, when predicting weaknesses in design data, at least one processor is configured to execute a set of instructions to cause the apparatus to perform further actions: Weaknesses in design data are predicted by applying pattern search based on pattern data from a first set of defects. 52. The apparatus pursuant to clause 50 or 51, wherein operation further includes: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 53. An apparatus according to any one of clauses 42 to 52, wherein the KPI is at least defect capture rate or throughput. 54. An apparatus according to any one of clauses 42 to 53, wherein the wafer inspection system is separate from the apparatus. 55. A non-transient computer-readable medium storing an instruction set executable by at least one processor of a computing device to cause the computing device to perform a method for real-time parameter tuning of a wafer inspection system, the method comprising: Acquire one or more input images streaming from the wafer inspection system; Multiple image enhancement parameters are applied to one or more input images to generate multiple images with different properties; Defects are identified from multiple images by applying multiple defect detection parameters; and Based on defects identified on multiple images, a combination of multiple image enhancement parameters and multiple defect detection parameters is determined for use in detecting a first defect type. 56. A computer-readable medium pursuant to Clause 55, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: To identify newly discovered defects that were not previously detected by the wafer inspection system; and The defect has been detected by the wafer inspection system and was not previously identified from multiple images. 57. A computer-readable medium pursuant to clause 55 or 56, wherein, in identifying defects from multiple images, a set of instructions executable by at least one processor of a computing device causes the computing device to perform: Identify the defects and their corresponding locations. 58. A computer-readable medium according to any one of clauses 55 to 57, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Defects are grouped by defect type. 59. A computer-readable medium pursuant to Clause 58, wherein, in determining the combination of parameters, a set of instructions executable by at least one processor of the computing device causes the computing device to perform: The parameter combination is determined by grouping defects by defect type. 60. A computer-readable medium according to any one of clauses 55 to 59, wherein, in determining the combination of parameters, a set of instructions executable by at least one processor of the computing device causes the computing device to perform: Deep learning algorithms are used to determine the parameter combination. 61. A computer-readable medium according to any one of clauses 55 to 60, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: The determined parameter combination is provided to the wafer inspection system, which then updates the recoded process formula in real time based on the determined parameter combination used for inline production. 62. A computer-readable medium according to any one of clauses 55 to 61, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Weaknesses in design data are predicted based on defects identified on multiple images. 63. A computer-readable medium pursuant to Clause 62, wherein, in predicting weaknesses in design data, a set of instructions executable by at least one processor of the computing device causes the computing device to further perform: Weaknesses in design data are predicted by applying pattern search based on pattern data of defects identified on multiple images. 64. A computer-readable medium pursuant to clause 62 or 63, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 65. A computer-readable medium according to any one of clauses 55 to 64, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Set key performance indicators (KPIs); and The various combinations of multiple image enhancement parameters and multiple defect detection parameters are ranked based on the corresponding KPI values ​​determined by defects identified on multiple images. 66. A computer-readable medium pursuant to Clause 65, wherein the KPI is at least the defect capture rate or throughput. 67. A computer-readable medium according to any one of clauses 55 to 66, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to perform the method in parallel when the wafer inspection system is in production. 68. A computer-readable medium according to any one of clauses 55 to 67, wherein the wafer inspection system is a separate wafer inspection system. 69. A non-transient computer-readable medium storing an instruction set executable by at least one processor of a computing device to cause the computing device to perform a method for real-time parameter tuning of a wafer inspection system, the method comprising: Acquire one or more input images streaming from the wafer inspection system; The first parameter combination is applied to one or more input images, and identifies a first set of defects in one or more input images; The second parameter combination is applied to one or more input images, and identifies a second set of defects in one or more input images; The first key performance indicator (KPI) value is determined based on the first defect set and the first parameter combination. The second KPI value is determined based on the second defect set, and the second parameter combination is used to determine the second KPI value; and The first parameter combination and the second parameter combination are ranked based on the first KPI value and the second KPI value. 70. A computer-readable medium pursuant to Clause 69, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Newly discovered defects that were not previously detected by the wafer inspection system are identified in the first defect set; and The identifier is a missing defect that has been detected by the wafer inspection system and was not previously identified by the first defect set. 71. A computer-readable medium pursuant to Clause 69 or 70, wherein, in identifying the first set of defects, a set of instructions executable by at least one processor of the computing device causes the computing device to perform: Identify the first set of defects and their corresponding locations. 72. A computer-readable medium according to any one of clauses 69 to 71, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: The first defect set is grouped by defect type. 73. A computer-readable medium pursuant to Clause 72, wherein, in determining the combination of parameters, a set of instructions executable by at least one processor of the computing device causes the computing device to perform: The parameter combination is determined by grouping defects by defect type. 74. A computer-readable medium pursuant to Clause 73, wherein a first KPI value and a second KPI value are associated with a first defect type. 75. A computer-readable medium according to any one of clauses 69 to 74, wherein, in ranking the first combination of parameters and the second combination of parameters, a set of instructions executable by at least one processor of the computing device causes the computing device to perform: A deep learning algorithm is used to rank the first and second parameter combinations. 76. A computer-readable medium according to any one of clauses 69 to 75, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: The parameter combination that ranks higher between the first and second parameter combinations is provided to the wafer inspection system, enabling the wafer inspection system to update the recoded process in real time based on the parameter combination provided for inline production. 77. A computer-readable medium according to any one of clauses 69 to 76, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Weaknesses in the design data are predicted based on the first set of defects. 78. A computer-readable medium pursuant to Clause 77, wherein, in predicting weaknesses in design data, a set of instructions executable by at least one processor of a computing device causes the computing device to perform: Weaknesses in design data are predicted by applying pattern search based on pattern data from a first set of defects. 79. A computer-readable medium pursuant to clause 78 or 79, wherein a set of instructions executable by at least one processor of a computing device causes the computing device to further perform: Information about weaknesses is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the wafer portion corresponding to the weakness. 80. A computer-readable medium pursuant to any of clauses 69 to 79, wherein the KPI is at least defect capture rate or throughput. 81. A computer-readable medium pursuant to any of Clauses 69 to 80, wherein the wafer inspection system is a separate wafer inspection system.

[0114] The block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer hardware or software products according to various exemplary embodiments of the present disclosure. In this regard, each block in the diagrams may represent certain arithmetic or logical operations that can be implemented using hardware such as electronic circuits. A block may also represent a module, segment, or code portion, including one or more executable instructions for implementing a specified logical function. It should be understood that in some alternative implementations, the functions indicated in the blocks may not occur in the order mentioned in the drawings. For example, two blocks shown consecutively may be executed or implemented substantially concurrently, or the two blocks may sometimes be executed in reverse order, depending on the functionality involved. Some blocks may also be omitted. It should also be understood that each block of the block diagram, and combinations of blocks, may be implemented by a dedicated hardware-based system (performing the specified function or action) or a combination of dedicated hardware and computer instructions.

[0115] It should be understood that the embodiments of this disclosure are not limited to the precise constructions described above and illustrated in the accompanying drawings, and various modifications and changes can be made without departing from its scope. This disclosure has been described in conjunction with various embodiments, and other embodiments of the invention will be apparent to those skilled in the art in light of the specification and practice of the invention disclosed herein. The specification and examples are intended to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

Claims

1. A method for real-time parameter tuning of a wafer inspection system, the method comprising: Acquire one or more input images streaming from the wafer inspection system; Multiple image enhancement parameters are applied to the one or more input images to generate multiple images with different characteristics; Defects are identified from the multiple images by applying multiple defect detection parameters; as well as Based on the defects identified on the plurality of images, a combination of parameters to be used to detect a first defect type is determined from the plurality of image enhancement parameters and the plurality of defect detection parameters.

2. The method according to claim 1, further comprising: Identify newly discovered defects that were not previously detected by the wafer inspection system; as well as The identification marks are missing defects that have been detected by the wafer inspection system and were not previously identified from the plurality of images.

3. The method according to claim 1, wherein determining the parameter combination comprises: The parameter combination is determined using a deep learning algorithm.

4. The method according to claim 1, further comprising: The determined parameter combination is provided to the wafer inspection system, which then updates the recoded recipe process in real time based on the determined parameter combination used for inline production.

5. The method according to claim 1, further comprising: Weaknesses in the design data are predicted based on the defects identified on the plurality of images.

6. The method of claim 5, wherein the predicted weakness in the design data comprises: The weaknesses in the design data are predicted by applying pattern search based on pattern data of the defects identified on the plurality of images.

7. The method according to claim 5, further comprising: The information about the weakness is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the portion of the wafer corresponding to the weakness.

8. The method according to claim 1, further comprising: Set key performance indicators (KPIs); as well as Based on the corresponding KPI values ​​determined according to the defects identified on the plurality of images, various parameter combinations of the plurality of image enhancement parameters and the plurality of defect detection parameters are ranked.

9. A method for real-time parameter tuning of a wafer inspection system, the method comprising: Acquire one or more input images streaming from the wafer inspection system; The first parameter combination is applied to the one or more input images, and a first set of defects in the one or more input images is identified; The second parameter combination is applied to the one or more input images, and a second set of defects is identified in the one or more input images; Based on the first defect set, determine the first key performance indicator value, i.e., the first KPI value, of the first parameter combination; Based on the second defect set, determine the second KPI value of the second parameter combination; as well as The first parameter combination and the second parameter combination are ranked based on the first KPI value and the second KPI value.

10. The method of claim 9, further comprising: Newly discovered defects that were not previously detected by the wafer inspection system are identified in the first defect set; as well as The identifier is a missing defect that has been detected by the wafer inspection system and was not previously identified by the first defect set.

11. The method of claim 9, wherein ranking the first parameter combination and the second parameter combination comprises: The first parameter combination and the second parameter combination are ranked using a deep learning algorithm.

12. The method according to claim 9, further comprising: The higher-ranking parameter combination between the first and second parameter combinations is provided to the wafer inspection system, enabling the wafer inspection system to update the recoded recipe process in real time based on the parameter combination provided for inline production.

13. The method of claim 9, further comprising: Weaknesses in the design data are predicted based on the first set of defects.

14. The method of claim 13, further comprising: The information about the weakness is provided to the wafer inspection system, which then updates the recoded process in real time to inspect the portion of the wafer corresponding to the weakness.

15. An apparatus for real-time parameter tuning of a wafer inspection system, the apparatus comprising: Memory stores instruction sets; as well as At least one processor is configured to execute the instruction set to cause the device to perform operations, the operations including: Acquire one or more input images streaming from the wafer inspection system; Multiple image enhancement parameters are applied to the one or more input images to generate multiple images with different characteristics; Defects are identified from the multiple images by applying multiple defect detection parameters; and Based on the defects identified on the plurality of images, a combination of parameters to be used for detecting a first defect type is determined, comprising the plurality of image enhancement parameters and the plurality of defect detection parameters.

Images