Test with previous step deletion

Abstract

An inspection system may generate first step images of a plurality of sample areas after a first process step and second step images of the sample areas after a second process step, wherein the second process step modifies a sample in at least one of the sample areas. The system may further identify one of the sample areas as a test area and at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image. The system may further generate a multi-step difference image by subtracting a combination of at least one of the second step comparison images and at least two of the first step images from the test image. The system may further identify defects in the test area associated with the second process step based on the multi-step difference image.

Description

[0001] CROSS REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 326,268, filed on March 31, 2022, under 35 USC §119(e), the entirety of which is incorporated herein by reference. Technical Field

[0003] The present disclosure relates generally to defect inspection, and more particularly, to compensating for noise during defect inspection of a current process step using images associated with a previous process step. Background Art

[0004] Inspection systems are often used during semiconductor manufacturing to identify defects in the manufacturing process that can lead to performance degradation or failure of the manufactured device. As the size of manufactured features continues to shrink, the size of manufacturing defects also shrinks. This results in weaker measurable signals associated with such defects and lower signal-to-noise ratios (SNRs) during defect inspection. Therefore, there is a need to develop systems and methods for addressing the above-mentioned defects. Summary of the invention

[0005] According to one or more illustrative embodiments, an inspection system is disclosed. In one illustrative embodiment, the system includes a controller to receive first step images of a plurality of sample areas after a first process step, and to receive second step images of the plurality of sample areas after a second process step, wherein the second process step modifies the sample in at least one of the plurality of sample areas. In another illustrative embodiment, the controller further identifies one of the plurality of sample areas as a test area and identifies at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image. In another illustrative embodiment, the controller generates a multi-step difference image from a weighted subtraction of at least one of the second step comparison images and at least two of the first step images in combination with the test image. In another illustrative embodiment, the controller identifies a defect in the test area associated with the second process step based on the multi-step difference image.

[0006] According to one or more illustrative embodiments, an inspection method is disclosed. In one illustrative embodiment, the method includes generating first step images of a plurality of sample areas after a first process step. In another illustrative embodiment, the method includes generating second step images of the plurality of sample areas after a second process step, wherein the second process step modifies the sample in at least one of the plurality of sample areas. In another illustrative embodiment, the method includes identifying one of the plurality of sample areas as a test area and identifying at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image. In another illustrative embodiment, the method includes generating a multi-step difference image from a weighted subtraction of at least one of the second step comparison images and at least two of the first step images in combination with the test image. In another illustrative embodiment, the method includes identifying defects in the test area associated with the second process step based on the multi-step difference image.

[0007] According to one or more illustrative embodiments, an inspection system is disclosed. In one illustrative embodiment, the system includes an imaging subsystem, the imaging subsystem including an illumination source and a detector configured to image a sample. In another illustrative embodiment, the system includes a controller. In another illustrative embodiment, the controller receives a first step image of a plurality of sample areas from the imaging subsystem after a first process step. In another illustrative embodiment, the controller receives a second step image of the plurality of sample areas after a second process step, wherein the second process step modifies the sample in at least one of the plurality of sample areas from the imaging subsystem. In another illustrative embodiment, the controller identifies one of the plurality of sample areas as a test area and identifies at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image. In another illustrative embodiment, the controller generates a multi-step difference image from a weighted subtraction of the test image by a combination of at least one of the second step comparison images and at least two of the first step images. In another illustrative embodiment, the controller identifies a defect in the test area associated with the second process step based on the multi-step difference image.

[0008] It should be understood that both the foregoing summary and the following detailed description are exemplary and explanatory only and do not necessarily limit the invention as claimed.The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the summary, serve to explain the principles of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Several advantages of the present disclosure may be better appreciated by those skilled in the art by referring to the accompanying drawings.

[0010] Figure 1A is a block diagram of an inspection system according to one or more embodiments of the present disclosure.

[0011] Figure 1B is a simplified schematic diagram of an optical imaging subsystem according to one or more embodiments of the present disclosure.

[0012] Figure 2 is a flow chart illustrating steps performed in an inspection method according to one or more embodiments of the present disclosure.

[0013] Figure 3 is a simplified schematic diagram of a portion of a sample depicting a series of sample regions that may be inspected for defects in accordance with one or more embodiments of the present disclosure.

[0014] Figure 4 is a diagram illustrating one or more embodiments of the present disclosure. Figure 2 A flow chart of the various steps of the method associated with a simulated image of a sample. DETAILED DESCRIPTION

[0015] Reference will now be made in detail to the disclosed subject matter illustrated in the accompanying drawings. The present disclosure has been particularly shown and described with respect to certain embodiments and specific features thereof. The embodiments set forth herein are to be considered illustrative rather than restrictive. It should be readily apparent to one of ordinary skill in the art that various changes and modifications in form and details may be made without departing from the spirit and scope of the present disclosure.

[0016] Embodiments of the present disclosure relate to systems and methods for performing defect inspection at a current process step, wherein a difference image representing noise associated with a previous process step is subtracted from a corresponding difference image of the current process step. In this way, the noise of the image to be analyzed for defects (referred to herein as the inspection difference image) may have a high signal-to-noise ratio (SNR).

[0017] Semiconductor devices are typically manufactured using a series of process steps, such as, but not limited to, deposition of process layers and / or photoresists, exposure of the photoresist using a desired pattern, etching of the photoresist and / or underlying process layers, polishing, etc. Variations in any of these processes, alone or in combination, may result in variations in the characteristics of the manufactured samples. In a general sense, some sample variations may be acceptable and considered upset points, while other sample variations may result in unacceptable performance degradation or device failure and are considered defects of interest (DOIs).

[0018] The inspection tool may be used after a selected process step to inspect the sample for defects. For example, the inspection tool may generate an image of a sample area of ​​interest (e.g., a die, a unit, etc.) after a particular process step of interest, which is referred to herein as a test image. The inspection tool may then identify defects in this sample area by comparing this test image with a reference image. As an illustration, a difference image may be generated based on subtracting a reference image from a test image (or vice versa) to generate a difference image, wherein features in the difference image correspond to sample variations. These sample variations may then be classified as DOIs or disturbing points. Such reference images may be generated based on additional sample areas (e.g., additional dies, units, etc. on the same or different samples) having a common design and / or design data. The generation of a difference image is generally described in U.S. Patent No. 11,270,430, issued on March 8, 2022, the entirety of which is incorporated herein by reference.

[0019] Such test and reference images associated with any particular process step may be generated using any suitable technique and may correspond to independently generated images or portions of one or more larger images (e.g., may correspond to sub-images of one or more larger images). As an illustration, a test image may be associated with a portion of a larger image representing multiple dies on a sample. In this manner, a test image may be a sub-image representing an area of ​​interest (e.g., a particular die) for defect inspection at a particular time. As another illustration, a reference image may be generated based on one or more different portions of a larger image associated with different areas (e.g., additional dies) on the sample. For example, a reference image may correspond to an average of sub-images associated with one or more additional dies on the sample. In this configuration, a difference image at a particular process step may correspond to a difference between an inspection sub-image and a reference image (e.g., a combination of sub-images).

[0020] Furthermore, the definition of a test image or reference image may depend on the particular region of interest and may change over time. For example, an image of a particular die may be used as a test image when identifying defects on that particular die and may be used as a reference image (or as one of several images combined to form a reference image) when identifying defects on a different die.

[0021] Various types of inspection tools have been developed, including optical inspection tools that generate images based on illumination using a light source or particle beam inspection tools that generate images based on illumination using a particle beam (e.g., an electron beam (e-beam), an ion beam, etc.). Optical inspection tools are often used for in-line inspection due to relatively high processing power. In addition, many inspection tools generate images by scanning a beam (e.g., a light beam or a particle beam) across a sample in multiple scan bands and creating an image of the sample based entirely or partially on the scan bands.

[0022] Various conditions can reduce the signal-to-noise ratio (SNR) of defect signals during inspection and thus limit accuracy or reliability. For example, as the size of fabricated device features shrinks, the size of associated process defects that lead to degraded performance or device failure also shrinks. Even under ideal conditions, this results in reduced SNR when feature and / or defect sizes approach the resolution limit of the inspection equipment. In addition, various process layers and / or photoresist layers on the sample are typically at least partially transparent to the inspection tool, especially the optical inspection tool. In other words, the inspection tool can typically be imaged at least partially into the volume of the sample. This transparency is beneficial for detecting sub-surface defects.

[0023] However, it is contemplated herein that signals from manufactured features associated with previous process steps may contribute to noise or disturbing point signals during inspection of the current process step. For example, an optical inspection tool may be configured to provide a shallow depth of field, where features associated with the current process step under inspection are in focus and have relatively high contrast, while features associated with any number of previous focus steps may be blurred and may be seen as noise or disturbing point signals in the test image. The presence of such signals may further limit the SNR associated with small and / or weak defects and thus limit the accuracy and / or reliability of the inspection.

[0024] In some embodiments, the inspection SNR is increased by performing defect inspection on a multi-step difference image, where the multi-step difference image is generated based on subtracting reference images associated with both the current and previous layers from the test image. In this context, the term reference image is used to describe any image (or portion thereof) associated with a region different from the region being considered as the test image at a particular time.

[0025] It is contemplated herein that the multi-step difference image may be constructed in different ways within the spirit and scope of the present disclosure. In some embodiments, the multi-step difference image is formed by subtracting multiple images of other regions under both the current process step and the previous process step from a test image associated with the region of interest (e.g., die) under the current process step. In some embodiments, the multi-step difference image is formed by subtracting the weighted difference of images of other regions under both the current process step and the previous process step from a test image associated with the region of interest (e.g., die) under the current process step. In this configuration, the weights may be determined by an optimization algorithm (e.g., designed to minimize the pixel values ​​of the multi-step difference image).

[0026] It is further contemplated herein that the systems and methods disclosed herein can more effectively reduce image noise and improve detection sensitivity than prior art techniques utilizing data from prior process layers. Wafer noise reduction by cross-layer image pruning is generally described, for example, in U.S. Pat. No. 10,504,213, issued Dec. 10, 2019, which is incorporated herein by reference in its entirety.

[0027] It is contemplated herein that the systems and methods disclosed herein may be particularly advantageous for, but not limited to, applications where the effect of the current process step is a relatively mild optical perturbation as measured by an inspection tool. In other words, the systems and methods disclosed herein may be particularly advantageous for, but not limited to, applications where the test images at the current and previous steps (e.g., the original test images) are relatively unchanged except for the manufactured features of interest. For example, the systems and methods disclosed herein may be particularly advantageous for, but not limited to, inspection of vias, mask etches (e.g., hard mask etches), punch defects, hole features, line features, scratches, pits, or residual material.

[0028] Some embodiments of the present disclosure relate to methods for performing defect inspection based on subtracting noise associated with previous layer features. Some embodiments of the present disclosure relate to inspection systems configured to perform defect inspection based on subtracting noise associated with previous layer features.

[0029] Some embodiments of the present disclosure relate to an inspection recipe for defect inspection based on subtracting noise associated with previous layer features. For example, the inspection recipe may define various parameters associated with imaging a sample and / or process steps associated with identifying defects on a sample based on one or more images.

[0030] Reference Figures 1A to 4 , a system and method for defect inspection are described in more detail according to one or more embodiments of the present disclosure.

[0031] Figure 1A is a block diagram of an inspection system 100 according to one or more embodiments of the present disclosure.

[0032] In one embodiment, the inspection system 100 includes an imaging subsystem 102 configured to generate one or more images of a sample 104. For example, the imaging subsystem 102 may include an illumination subsystem 106 configured to illuminate the sample 104 using illumination 108 (e.g., an illumination beam) from an illumination source 110 and a collection subsystem 112 configured to generate an image of the sample 104 using radiation emitted from the sample (e.g., sample radiation 114) using a detector 116.

[0033] In some embodiments, the sample 104 may include a substrate (e.g., a wafer, etc.) formed of a semiconductor or non-semiconductor material. For example, the semiconductor or non-semiconductor material may include, but is not limited to, single crystal silicon, gallium arsenide, and indium phosphide. The sample 104 may further include one or more layers placed on the substrate. For example, such layers may include, but are not limited to, resists, dielectric materials, conductive materials, and / or semiconductive materials. Many different types of such layers are known in the art, and as used herein, the term sample is intended to cover samples on which all types of such layers can be formed. The one or more layers formed on the sample may be patterned or unpatterned. For example, the sample may include multiple dies each having repeatable patterned features. The formation and processing of such material layers may ultimately lead to a completed device. Many different types of devices may be formed on the sample, and as used herein, the term sample is intended to cover samples on which any type of device known in the manufacturing technology is manufactured.

[0034] The imaging subsystem 102 may generate one or more images of the sample 104 using any technique known in the art. In some embodiments, the imaging subsystem 102 is an optical imaging subsystem 102, wherein the illumination source 110 is an optical source configured to generate the illumination 108 in the form of light, and wherein the light collection subsystem 112 images the sample 104 based on light emitted from the sample 104. In some embodiments, the imaging subsystem 102 is a particle imaging subsystem 102, wherein the illumination source 110 is a particle source configured to generate the illumination 108 in the form of particles. For example, the particle illumination 108 may be in the form of an electron beam (e-beam), an ion beam (e.g., a focused ion beam), or a neutral particle beam. In addition, the light collection subsystem 112 may image the sample 104 based on particles emitted from the sample 104 (e.g., backscattered electrons, etc.). In some cases, the particle inspection system 100 may also image the sample 104 based on light emitted from the sample 104 in response to the incident particle illumination 108 (e.g., based on photoluminescence, etc.).

[0035] Figure 1B is a simplified schematic diagram of optical imaging subsystem 102 according to one or more embodiments of the present disclosure.

[0036] The illumination source 110 may include any type of illumination source known in the art suitable for producing the optical illumination 108, which may be in the form of one or more illumination beams. Furthermore, the illumination 108 may have any spectrum, such as, but not limited to, extreme ultraviolet (EUV) wavelengths, ultraviolet (UV) wavelengths, visible wavelengths, or infrared (IR) wavelengths. Furthermore, the illumination source 110 may be a broadband source, a narrowband source, and / or a tunable source.

[0037] In some embodiments, the illumination source 110 includes a broadband plasma (BBP) illumination source. In this regard, the illumination 108 may include radiation emitted by the plasma. For example, the BBP illumination source 110 may include (but does not necessarily include) one or more pump sources (e.g., one or more lasers) configured to focus into a gas volume, thereby causing energy to be absorbed by the gas so as to create or maintain a plasma suitable for emitting radiation. Furthermore, at least a portion of the plasma radiation may be used as the illumination 108.

[0038] In another embodiment, illumination source 110 may include one or more lasers. For example, illumination source 110 may include any laser system known in the art capable of emitting radiation in the infrared, visible, or ultraviolet portions of the electromagnetic spectrum.

[0039] The illumination source 110 may further generate illumination 108 having any temporal profile. For example, the illumination source 110 may generate continuous wave (CW) illumination 108, pulsed illumination 108, or modulated illumination 108. Additionally, the illumination 108 may be delivered from the illumination source 110 via free space propagation or guided light (e.g., optical fiber, light pipe, etc.).

[0040] The illumination subsystem 106 may include various components for directing the illumination 108 to the sample 104, such as, but not limited to, a lens 118, a mirror, etc. In addition, such components may be reflective elements or transmissive elements. In this manner, Figure 1B The depiction of lens 118 as a transmissive element in FIG. 1 is merely illustrative and non-limiting. The illumination subsystem 106 may further include one or more optical elements 120 for modifying and / or conditioning light in an associated optical path, such as, but not limited to, one or more polarizers, one or more filters, one or more beam splitters, one or more diffusers, one or more homogenizers, one or more apodizers, or one or more beam shapers.

[0041] In some embodiments, the inspection system 100 includes a translation stage 122 for fixing and / or positioning the sample 104 during imaging. For example, the translation stage 122 may include any combination of linear actuators, rotary actuators, or angular actuators to position the sample 104 using any number of degrees of freedom.

[0042] The imaging subsystem 102 may include various components for collecting at least a portion of the sample radiation 114 radiation (e.g., sample light in the case of the optical imaging subsystem 102) emitted from the sample 104 and directing at least a portion of the sample light to a detector 116 for producing an image. The image produced by the inspection system 100 may be any type of image known in the art, such as, but not limited to, a bright field image, a dark field image, a phase contrast image, etc.

[0043] The imaging subsystem 102 may include various components for directing the sample radiation 114 to the detector 116, such as, but not limited to, lenses 124, mirrors, etc. Furthermore, such components may be reflective elements or transmissive elements. In this manner, Figure 1B The depiction of lens 118 as a transmissive element in FIG. 1 is merely illustrative and non-limiting. The imaging subsystem 102 may further include one or more optical elements 126 for modifying and / or conditioning light in an associated optical path, such as, but not limited to, one or more polarizers, one or more filters, one or more beam splitters, one or more diffusers, one or more homogenizers, one or more apodizers, or one or more beam shapers.

[0044] The detector 116 may include any type of sensor known in the art suitable for measuring sample light. For example, the detector 116 may include a multi-pixel sensor such as, but not limited to, a charge coupled device (CCD), a complementary metal oxide semiconductor (CMOS) device, a line sensor, or a time delay integration (TDI) sensor. As another example, the detector 116 may include two or more single-pixel sensors such as, but not limited to, a photodiode, a collapsed photodiode, a photomultiplier tube, or a single photon detector.

[0045] The illumination subsystem 106 and the imaging subsystem 102 can be configured in various ways within the spirit and scope of the present disclosure. Figure 1B , the inspection system 100 includes at least one beam splitter 128 that is common to the optical paths of the illumination subsystem 106 and the imaging subsystem 102. In this way, both the illumination subsystem 106 and the imaging subsystem 102 can share a common objective lens 130 and can both utilize the full available pupil or numerical aperture (NA) provided by the objective lens 130. In some embodiments, although not shown, the illumination subsystem 106 and the imaging subsystem 102 can have separate optical paths without common elements.

[0046] The inspection system 100 may further image the sample 104 using any technique known in the art. In some embodiments, the inspection system 100 generates an image of the sample 104 in a scanning mode by focusing the illumination 108 onto the sample 104 as a point or line of light; capturing a point or line image; and scanning the sample 104 to establish a two-dimensional image. For example, the inspection system 100 may scan the illumination 108 across the sample 104 in various scan bands that may partially overlap. In this configuration, the scanning may be achieved by moving the sample 104 relative to the illumination 108 (e.g., using a translation stage 122), by moving the illumination 108 relative to the sample 104 (e.g., using an actuable mirror, etc.), or a combination thereof. In some embodiments, the inspection system 100 generates an image of the sample 104 in a static mode by directing the illumination 108 to the sample 104 in a two-dimensional field of view and directly capturing the two-dimensional image using the detector 116.

[0047] In some embodiments, various alignment and / or stitching operations are performed on the data generated by the imaging subsystem 102 to form one or more images of the sample 104. As an example, in the case of a scanning system, data associated with multiple scan swaths or portions thereof may be aligned and / or stitched to form an image of the entire sample 104 or only a portion thereof. In this manner, the term image is used herein to broadly describe any array of pixels representing a portion of the sample 104 and is not intended to confer limitations associated with imaging techniques. Thus, an image may correspond to a complete data set provided by the imaging subsystem 102, may correspond to a subset of this data (e.g., a sub-image), or may correspond to multiple data sets that are stitched (and appropriately aligned).

[0048] As previously described herein, images of sample 104 or portions thereof may be used in various ways. For example, an image of sample 104 corresponding to a particular region (e.g., a particular die, cell, etc.) may be used as a test image when identifying defects in this particular region and may be used as a reference image (or used to generate a reference image along with images of other regions) when identifying defects in other regions of sample 104.

[0049] In some embodiments, the inspection system 100 includes a controller 132 that may be communicatively coupled to the imaging subsystem 102 or any component therein. The controller 132 may include one or more processors 134 configured to execute program instructions maintained on a memory 136 (e.g., a storage medium). In this regard, the one or more processors 134 of the controller 132 may be configured to perform any of the various process steps described throughout the present disclosure. For example, the controller 132 may perform steps such as, but not limited to, developing and / or implementing an inspection recipe; receiving data associated with an image of the sample 104 or a portion thereof for any number of processing steps from the imaging subsystem 102; performing alignment and / or stitching operations; generating or identifying an image associated with a selected portion of the sample 104 as a test image or reference image; generating a difference image between the inspection and reference images; identifying defects on the sample 104 based on any combination of inspection, reference, or difference images associated with any number of process steps; or classifying identified defects.

[0050] The one or more processors 134 of the controller 132 may include any processing element known in the art. In this sense, the one or more processors 134 may include any microprocessor type device configured to execute algorithms and / or instructions. In one embodiment, the one or more processors 134 may be composed of a desktop computer, a large computer system, a workstation, an image computer, a parallel processor, or any other computer system (e.g., a networked computer) configured to execute a program (which is configured to operate the inspection system 100), as described throughout this disclosure. It should be further recognized that the term "processor" may be broadly defined to cover any device having one or more processing elements that execute program instructions from the non-transitory memory 136.

[0051] The memory 136 may include any storage medium known in the art that is suitable for storing program instructions that can be executed by the associated one or more processors 134. For example, the memory 136 may include a non-transitory storage medium. By way of another example, the memory 136 may include, but is not limited to, a read-only memory, a random access memory, a magnetic or optical storage device (e.g., a disk), a tape, a solid-state drive, etc. It should be further noted that the memory 136 may be housed in a common controller housing with the one or more processors 134. In one embodiment, the memory 136 may be remotely located relative to the physical location of the one or more processors 134 and the controller 132. For example, the one or more processors 134 of the controller 132 may access a remote memory (e.g., a server) that is accessible via a network (e.g., the Internet, an intranet, etc.). Therefore, the above description should not be interpreted as a limitation of the present invention but is merely illustrative.

[0052] The inspection system 100 may be further configured according to an inspection recipe, which may be implemented during run time (e.g., by the controller 132, the illumination subsystem 106, the imaging subsystem 102, or any other component). In some embodiments, the inspection recipe includes parameters for configuring the inspection system 100 to image the sample 104, such as, but not limited to, parameters for controlling the illumination 108, such as, but not limited to, wavelength, polarization, spot size, angle of incidence, or pattern of swaths used to produce an image. In some embodiments, the inspection recipe includes parameters for controlling the collection of sample radiation 114, such as, but not limited to, wavelength, polarization, collection angle, or operating parameters of the detector 116. In some embodiments, the inspection recipe includes directions for imaging the sample 104 after selected process steps. In some embodiments, the inspection recipe includes various processing steps associated with identifying defects based on images.

[0053] Reference Figures 2 to 4 , defect inspection based on pruning signals from previous process steps is described in more detail according to one or more embodiments of the present disclosure.

[0054] Figure 2 is a flow chart illustrating steps performed in an inspection method 200 according to one or more embodiments of the present disclosure. Applicants note that the embodiments and implementation techniques previously described herein in the context of the inspection system 100 should be interpreted as extending to the method 200. However, it should be further noted that the method 200 is not limited to the architecture of the inspection system 100. In some embodiments, the method 200 or its steps are part of an inspection recipe (e.g., for the inspection system 100).

[0055] Figure 2 Reference is made to various process steps associated with manufacturing a device in a multi-step process. In particular, Figure 2 Reference is made to a first process step and a second process step. In this framework, the second process step may correspond to a current process step, wherein a defect of interest is detected after applying this current process step to the sample 104, and wherein the first process step may refer to any process step preceding the current process step. In some embodiments, the first and second process steps are performed sequentially. In some embodiments, one or more intermediate process steps may be performed between the first and second process steps. In this manner, the terms first process step and second process step are used only to indicate the relative order of operations, but do not otherwise limit the method 200. Furthermore, the first and second process steps may be associated with any type of process steps, such as, but not limited to, deposition of a process layer and / or photoresist, exposure of the photoresist using a desired pattern, etching of the photoresist and / or underlying process layer, polishing, and the like.

[0056] Figure 2Reference is also made to various images. It should be understood that such images may be images directly generated by the inspection system 100 or portions thereof in one step, or composite images formed by stitching together sub-images generated by the inspection system 100. In addition, the images may be cropped or otherwise modified as desired. In this manner, references to images herein are merely illustrative and do not limit the types of images that may be generated by the inspection system 100.

[0057] In some embodiments, the method 200 includes a step 202 of generating a first step image of a plurality of sample areas 302 of the sample 104 after a first process step (e.g., a previous process step). The sample areas 302 may include any area on the sample 104 that is to be inspected for defects. For example, the sample areas 302 may include a die or a portion (e.g., a cell) of a die. In some embodiments, the method 200 includes a step 204 of generating a second step image of the plurality of sample areas 302 of the sample 104 after a second process step (e.g., a current process step).

[0058] For example, steps 202 and 204 may include generating an image of the same sample area 302 (e.g., the same die, cell, etc.) after two process steps (e.g., a first or previous process step and a second or current process step). The first and second process steps (e.g., previous and current process steps) may correspond to any process steps, such as, but not limited to, a photolithography step or an etching step. Furthermore, the second process step may modify the sample 104 in at least some of the sample area 302.

[0059] The first step image and the second step image may be generated using any technique known in the art. For example, the first step and second step images may (but need not) be associated with data from multiple scan strips (or portions thereof) from the scanning imaging subsystem 102 after associated process steps of alignment and / or stitching (e.g., by the controller 132).

[0060] In some embodiments, method 200 includes step 206 of identifying one of the plurality of sample areas 302 as a test area and identifying at least some of the remaining sample areas 302 as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image.

[0061] Figure 3 is a simplified schematic diagram of a portion of a sample 104 depicting a series of sample regions 302 that may be inspected for defects in accordance with one or more embodiments of the present disclosure. Figure 3In the method 200, one sample area 302' is identified as a test area, while the remaining sample areas are identified as comparison areas. In this way, the test area may correspond to one of the sample areas 302 inspected for defects at a particular time, while the comparison area may correspond to an additional sample area 302 of the same or similar design. In some embodiments, the comparison area is used to generate a reference image. As previously described herein, any sample area 302 may be identified as a test area in any iteration of the method 200 and further identified as a comparison area in another iteration of the method 200. In this way, the method 200 (or portions thereof) may be repeated to provide defect identification in multiple sample areas 302.

[0062] In some embodiments, method 200 includes a step 208 of generating a multi-step difference image from a weighted subtraction of a test image by a combination of at least one of the second step comparison images and at least two of the first step images.

[0063] It is contemplated herein that the first step image may include or may have data associated with noise or disturbing point signals in the second step image. For example, it may be the case that certain intensity variations in the second step image may be associated with structures manufactured prior to the second process step. In some cases, certain intensity variations in the second step image may be associated with out-of-focus features located outside the depth of focus of the imaging subsystem 102. It should be further contemplated herein that similar intensity variations may also be present in the first step image. The extent to which the intensity variations in the first step and second step images are correlated may depend on the effect of the second process step on the sample 104. In many manufacturing processes, the second process step provides a localized isolated modification of the sample (e.g., etching of selected features, etc.) or a uniform modification of the sample (e.g., deposition of one or more films, etc.).

[0064] Thus, subtracting the first step image from the second step image may beneficially remove common intensity variations that may be due to noise or disturber point signals, which may improve the SNR associated with the DOI.

[0065] In some embodiments, the method 200 includes a step 210 of identifying defects on the sample 104 associated with the second process step based on the multi-step difference image. Any suitable technique may be used to identify defects from the multi-step difference image. In addition, any type of defect (or different types of defects) may be identified using the systems and methods disclosed herein. For example, a defect may include a deviation in at least one of the shape, size, or orientation of a feature manufactured by the second process step. As another example, a defect may include the absence of a feature that is desired to be manufactured by the second process step. As another example, a defect may include at least one of a scratch, a pit, or residual material in the inspection area after the second process step. As another example, a defect may include at least one of an unexpected bridge between two features manufactured by the second process step or an unexpected break in a feature manufactured by the second process step.

[0066] In some embodiments, step 210 includes identifying defects based on applying a threshold to the multi-step difference image. For example, pixel values ​​greater than or equal to a selected threshold may be characterized as defects, while pixel values ​​below the threshold may be characterized as non-defects (e.g., background, distractors, noise, etc.).

[0067] In some embodiments, step 210 includes detecting defects using machine learning techniques, which may be unsupervised, supervised, or a combination thereof. For example, the multi-step difference image may be provided as input to a supervised or unsupervised machine learning algorithm suitable for identifying defects. As an illustration, the machine learning algorithm may provide a hard classification of a selected portion of the sample as a disturbing point or a defect (in any number of defect categories) or a soft classification with a probability that the selected portion of the sample may be classified as a disturbing point or a defect (in any number of defect categories). As another example, a supervised machine learning algorithm may be trained using a multi-step difference image with known (e.g., labeled) defects. This trained machine learning algorithm may then be used during runtime to identify defects in a new sample (e.g., a test sample) based on its multi-step difference image. In some embodiments, step 210 may include utilizing a multi-die automatic threshold (MDAT) technique. MDAT technology is generally described in the following: U.S. Patent No. 10,535,131; U.S. Patent No. 10,395,358; U.S. Patent No. 10,557,802; U.S. Patent No. 10,801,968; U.S. Patent No. 9,518,934; U.S. Patent No. 11,120,546; U.S. Patent No. 11,416,982; U.S. Patent No. 11,494,924; U.S. Patent No. 10,943,838; U.S. Patent No. 9, No. 613,411; No. 9,619,876; No. 10,599,944; No. 10,535,131; No. 11,551,348; No. 11,550,309; No. 9,518,934; No. 10,923,317 and U.S. Patent Publication No. 2021 / 0010945, all of which are incorporated herein by reference in their entirety. As an illustration, the MDAT technique may include clustering analysis based on a two-dimensional point cloud generated based on a multi-step difference image and a test image.

[0068] In some cases, the defects identified in step 210 may include any combination of DOIs or disturbing points. Thus, additional steps or techniques may be utilized in addition to or in combination with method 200.

[0069] In some embodiments, method 200 further includes the step of classifying at least some of the defects into one or more categories. For example, the categories may include, but are not limited to, defects of different types or defects having different effects on device performance.

[0070] Step 208 of generating a multi-step difference function will now be described in more detail according to one or more embodiments of the present disclosure.

[0071] It is contemplated herein that merely subtracting the first step image from the second step image may not be sufficient to remove undesired signals and increase the SNR of the DOI. Rather, it may be desirable to utilize multiple images associated with the current and previous process steps (e.g., multiple first step and second step images) in combination to increase the SNR of the DOI. Furthermore, different combinations of first step and second step images may be utilized to generate multi-step difference images within the spirit and scope of the present disclosure.

[0072] In some embodiments, a multi-step difference image may be generated based on a combination of multiple images from both the current and previous steps. For example, the multi-step difference image may correspond to a weighted difference (e.g., a weighted subtraction) of images from multiple sample regions associated with the first and second process steps:

[0073] I MSDIFE,i =I SecondStep,i -∑ j≠i α j I SecondStep,j -∑ j β j I FirstStep,j (1)

[0074] where indices i and j refer to sample locations (e.g., die, cell, etc.), and where α j and β j In addition, the subscripts FirstStep and SecondStep refer to the first step image (e.g., generated in step 202) and the second step image (e.g., generated in step 204), respectively. In this configuration, the test area may correspond to the sample area i so that the test image (e.g., identified in step 206) may correspond to item i. SecondStep,i The remaining sample area can then be characterized as a comparison area so that the term ∑ j≠i α j I SecondStep,j It may be referred to as the second step image of the comparison area or the second step comparison image.

[0075] In some embodiments, the number of first-step and second-step images used to generate the multi-step difference image is equal. For example, the value of j appears in the term ∑ associated with the second-step image of the comparison region. j≠i α j I SecondStep,j and the term ∑ associated with the first step image j β j I FirstStep,j In addition, since the test image I SecondStep,i is the second step image, so the number of second step images (e.g., with item I SecondStep,i and∑ j≠i α j ISecondStep,j The number of images associated with the combination) is equal to the number of images in the first step (e.g., with the term ∑ i α j I FirstStep,j In other words, the number of first-step images used to generate the multi-step difference image (e.g., related to the term ∑ j β j I FirstStep,j The number of associated images) may be one greater than the number of second step comparison images used to generate the multi-step difference image.

[0076] The weights α and β may be determined using any suitable technique. In some embodiments, the weights are determined based on a metric associated with the sample 104 in the associated sample position and / or an image quality metric associated with the associated image, such as, but not limited to, exposure, contrast, or noise. In some embodiments, the weights are determined by an optimization technique. For example, the weights may be determined based on a fitting technique to minimize the pixel intensity (I MSDIFF For example, the weights may be determined based on regression techniques to minimize the pixel intensity (I MSDIFF ). As another example, the weights may be determined using a defect detection algorithm such as, but not limited to, a multi-color adaptive threshold (MCAT) technique. MCAT techniques are generally described in: U.S. Pat. No. 11,270,430; U.S. Pat. No. 11,120,546; U.S. Pat. No. 11,416,982; U.S. Pat. No. 11,494,924; U.S. Pat. No. 10,801,968; and U.S. Published Patent Application No. 2021 / 0010945, all of which are incorporated herein by reference in their entirety. As another example, the weights may be determined using machine learning techniques (e.g., unsupervised and / or supervised machine learning techniques).

[0077] Equation (1) may be characterized and / or implemented in various ways (e.g., by controller 132). In some embodiments, a multi-step difference image is generated based on equation (1) in a single step using various images stored in a memory (e.g., memory 136). In some embodiments, one or more intermediate images are generated. For example, the term ∑ j≠i α j I SecondStep,j The second step reference image may correspond to an image of an additional sample area (e.g., die, cell, etc.) generated after the second process step using an image of an additional sample area (e.g., die, cell, etc.) in addition to the test area. SecondStep,i -∑ j≠i α j I SecondStep,j It can correspond to the second step difference image. In addition, the term ∑ j βj I FirstStep,j May correspond to a first step reference image generated after the first process step using an image of any suitable sample area.

[0078] It is contemplated herein that when identifying defects created by a second process step, this first step reference image may optionally include an image of the test area after the first process step (e.g., ∑ j β j I FirstStep,j , including examples where j=i), because the first step is before the second step and the associated sample areas can be considered equivalent in this first process step. In addition, in some embodiments, equation (1) can be adjusted to provide a first-step difference image generated in a manner similar to the second-step difference image. In this configuration, the first-step reference image can be generated based on various sample areas other than the test area (e.g., for j≠i, I FirstStep,j ) and deleted from the first step image of the test area (e.g., I FirstStep,i ):

[0079] I MSDIFE,i =I SecondStep,i -∑ j≠i α j I SecondStep,j -[I FirstStep,i -∑ j≠i β j I FirstStep,j ] (2)

[0080] Furthermore, additional weights may be applied to such difference images. For example, a multi-step difference image may be generated by:

[0081] I MSDIFF,i =I SecondStep,i -∑ j≠i α j I SecondStep,j -γ[I FirstStep,i -∑ j≠iβj I FirstStep,j ] (3)

[0082] Where γ is an additional weight applied to the first step reference image. In a general sense, the weight γ may be determined using any suitable technique, including but not limited to the techniques disclosed above for determining weights α and / or β. In some embodiments, weights α and β are determined separately and then weight γ is determined. For example, weights α and β may be determined in a first step (e.g., by an optimization technique), followed by an additional step of determining γ (e.g., by another optimization technique). It is contemplated herein that this method may be relatively simpler, more efficient, and / or utilize fewer computational resources than some other methods. In some embodiments, weights α, β, and γ are determined in a single step (e.g., a single optimization step).

[0083] In addition, although not explicitly shown, additional weight is given to term I FirstStep,i (E.g., the first step image corresponding to the test region.) More generally, any of the weights (e.g., α, β, γ, etc.) can have any value or any sign (e.g., positive or negative) for any region.

[0084] Additional variations of the multi-step difference image are now described.

[0085] In some embodiments, after both the first and second process steps, all sample areas 302 have a common design. For example, sample areas 302 may correspond to different dies having the same design and, therefore, ideally, the same fabricated features. Thus, each sample area 302 may be used as a test area or comparison area (e.g., for generating a reference image), as described above. Furthermore, the multi-step difference image generated in this configuration may not include any intentionally fabricated features.

[0086] In some embodiments, after the first process step but not after the second process step, sample areas 302 other than the test area (e.g., using the nomenclature of equations (1)-(3), sample areas 302j≠i) may have a common design with the test area (e.g., i). For example, the second process step may produce features (e.g., vias, etched patterns, fabricated structures, etc.) in the test area (i) but not in the additional sample areas 302j≠i. In this configuration, the multi-step difference image may include fabricated features. This configuration may be suitable for, but is not limited to, comparing fabricated features to design specifications (e.g., to identify missing features, improperly fabricated features, etc.).

[0087] Furthermore, referring generally to equations (1) to (3), the first step image and the second step image used to generate the multi-step difference image may generally have the same or different sample areas 302 .

[0088] Reference Figure 4 , describing a non-limiting example of an application of method 200 . Figure 42 is a flow chart depicting a simulated image of a sample 104 associated with various steps of method 200 according to one or more embodiments of the present disclosure. In particular, Figure 4 It includes a first step difference image 402 , a second step difference image 404 and a multi-step difference image 406 .

[0089] For example, a second-step difference image 404 may be generated based on a difference between a second-step test image of a sample area of ​​interest (e.g., a test area) after a second process step and a reference image, wherein the second-step difference image may be generated based on one or more images of additional sample areas. Using the nomenclature from equations (1) to (3), the second-step difference image 404 may be characterized as I SecondStep,i -∑ j≠i α j I SecondStep,j In this simulation, the second step difference image 404 includes a simulated defect 408 buried within the noise in the image generated during the second process step. For example, based on the signal amplitude of 121 and the noise amplitude of 134, the SNR of the second step difference image 404 is 0.9.

[0090] A first-step difference image 402 may similarly be generated based on the difference between the first-step test image of the test area and the first-step reference image, where the first-step reference image may be generated based on one or more images of the additional sample area. Again using the nomenclature from equations (1) to (3), the first-step difference image 402 may be characterized as I FirsrStep,i -∑ j≠i β j I FirstStep,j . The first step difference image 402 does not include the simulated defect 408 because this simulated defect 408 is generated in the subsequent second process step. In addition, this simulated image has a noise level of N=103 that is lower than the noise level (121) of the subsequent second step difference image 404, but maintains a similar distribution. As previously described herein, the common distribution of noise in the first step difference image 402 and the second step difference image 404 (or more generally, between the first step and second step images) can be based on a common physical structure on the sample 104 that can be observed using the inspection system 100.

[0091] The multi-step difference image 406 may then correspond to the difference between the second step difference image 404 and the first step difference image 402. Continuing with the nomenclature of equations (1) to (3), this multi-step difference image 406 may thus be characterized by equation (2): MSDIFF,i =I SecondStep,i -∑ j≠i α j I SecondStep,j -[I FirstStep,i-∑ j≠i β j I FirstStep,j ]. It is clearly observed that the multi-step difference image 406 exhibits substantially less noise and thus substantially greater SNR than the second-step difference image 404, such that the simulated defect 408 can be clearly observed. Specifically, the multi-step difference image 406 has a noise magnitude of N=19.4 and an SNR of 6.2, compared to the noise magnitude of N=134 and an SNR of 0.9 of the second-step difference image 404.

[0092] The subject matter described herein sometimes illustrates different components contained in or connected to other components. It should be understood that such depicted architectures are merely exemplary, and in fact many other architectures that realize the same functionality can be implemented. In a conceptual sense, any arrangement of components for realizing the same functionality is effectively "associated" so as to realize the desired functionality. Therefore, any two components combined in this article to realize a specific functionality can be regarded as "associated" with each other so as to realize the desired functionality without regard to the architecture or intermediate components. Similarly, any two components so associated can also be regarded as "connected" or "coupled" to each other to realize the desired functionality, and any two components that can be so associated can also be regarded as "coupleable" to each other to realize the desired functionality. Specific examples that can be coupled include but are not limited to physically interactive and / or physically interactive components and / or wirelessly interactive and / or wirelessly interactive components and / or logically interactive and / or logically interactive components.

[0093] It is believed that the present disclosure and many of its attendant advantages will be understood from the foregoing description, and it will be appreciated that various changes may be made to the form, construction, and arrangement of components without departing from the disclosed subject matter or sacrificing all of its material advantages. The forms described are merely illustrative, and it is intended that such changes be covered and included by the appended claims. Furthermore, it should be understood that the present disclosure is defined by the appended claims.

Claims

1. A testing system, comprising: A controller comprising one or more processors configured to execute program instructions to cause the one or more processors to implement a test recipe by: receiving a first step image of a plurality of sample areas after a first process step; receiving a second step image of the plurality of sample areas after a second process step, wherein the second process step modifies a sample in at least one of the plurality of sample areas; identifying one of the plurality of sample areas as a test area and identifying at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a second step comparison image; generating a multi-step difference image from the test image by weighted subtraction of at least one of the second-step comparison images and a combination of at least two of the first-step images; and A defect in the test area associated with the second process step is identified based on the multi-step difference image.

2. The inspection system according to claim 1, wherein the number of the first step images used to generate the multi-step difference image is greater than the number of the second step comparison images by 1 so as to generate the multi-step difference image based on the same number of the first step images and the second step images.

3. The inspection system of claim 2, wherein the number of first-step images used to generate the multi-step difference image is at least three.

4. The inspection system according to claim 1, further comprising: At least some of the defects are classified as at least one of disturbers or defects of interest.

5. The inspection system of claim 1, wherein the weights associated with the weighted pruning are determined by a fitting technique.

6. The inspection system of claim 1, wherein the weights associated with the weighted pruning are determined by regression techniques.

7. The inspection system of claim 1, wherein the weights associated with the weighted pruning are determined by a multi-color adaptive thresholding (MCAT) technique.

8. The inspection system of claim 1, wherein the weights associated with the weighted pruning are determined by machine learning techniques.

9. The inspection system of claim 1 , wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -∑ j β j I FirstStep，j Generate the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and ∑ j β j I FirstStep，j Corresponding to the at least two of the first step images, where subscripts i and j correspond to the sample regions, and where α and β correspond to weights.

10. The inspection system of claim 1, wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -[I FirstStep，i -∑ j≠i β j I FirstStep，j ] generates the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and [I FirstStep，i -∑ j≠i β j I FirstStep，j ] corresponds to at least one of the first step images, where subscripts i and j correspond to the sample areas, and where α and β correspond to weights.

11. The inspection system of claim 1 , wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -γ[I FirstStep，i -∑ j≠i β j I FirstStep，j ] generates the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and [I FirstStep，i -∑ j≠i β j I FirstStep，j ] corresponds to the at least one of the first step images, where subscripts i and j correspond to the sample areas, where α, β and γ correspond to weights.

12. The inspection system of claim 1 , wherein detecting defects on the sample based on the multi-step difference image comprises: providing the inspection reference image to the machine learning algorithm; and Defects on the sample are detected based on the output of the machine learning algorithm.

13. The inspection system of claim 1 , wherein detecting defects on the sample based on the multi-step difference image comprises: Defects on the sample are detected based on the multi-step difference image using a multi-die adaptive threshold (MDAT) technique.

14. The inspection system of claim 1, wherein the defect comprises: A deviation in at least one of shape, size, or orientation of a feature produced by said second process step.

15. The inspection system of claim 1, wherein the defect comprises: The lack of features produced by the second process step is desired.

16. The inspection system of claim 1, wherein the defect comprises: At least one of scratches, pits, or residual material in the inspection area after the second process step.

17. The inspection system of claim 1, wherein the defect comprises: At least one of an unintended bridging between two features produced by the second process step or an unintended break in a feature produced by the second process step.

18. A testing method, comprising: generating a first step image of a plurality of sample areas after the first process step; generating a second step image of the plurality of sample areas after a second process step, wherein the second process step modifies a sample in at least one of the plurality of sample areas; identifying one of the plurality of sample areas as a test area and identifying at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a comparison image; generating a multi-step difference image from the test image by weighted subtraction of at least one of the second step comparison images and a combination of at least two of the first step images; and A defect in the test area associated with the second process step is identified based on the multi-step difference image.

19. The inspection method according to claim 18, wherein the number of the first step images used to generate the multi-step difference image is greater than the number of the second step comparison images by 1 so as to generate the multi-step difference image based on the same number of the first step images and the second step images.

20. The inspection method according to claim 19, wherein the number of first-step images used to generate the multi-step difference image is at least three.

21. The testing method according to claim 18, further comprising: At least some of the defects are classified as at least one of disturbers or defects of interest.

22. The inspection method of claim 18, wherein the weights associated with the weighted pruning are determined by a fitting technique.

23. The inspection method of claim 18, wherein the weights associated with the weighted pruning are determined by regression techniques.

24. The inspection method of claim 18, wherein the weights associated with the weighted pruning are determined by a multi-color adaptive thresholding (MCAT) technique.

25. The inspection method of claim 18, wherein the weights associated with the weighted pruning are determined by machine learning techniques.

26. The inspection method of claim 18, wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -∑ j β j I FirstStep，j Generate the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and ∑ j β j I FirstStep，j Corresponding to the at least two of the first step images, where subscripts i and j correspond to the sample regions, and where α and β correspond to weights.

27. The inspection method of claim 18, wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -[I FirstStep，i -∑ j≠i β j I FirstStep，j ] generates the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and [I FirstStep，i -∑ j≠i β j I FirstStep，j ] corresponds to at least one of the first step images, where subscripts i and j correspond to the sample areas, and where α and β correspond to weights.

28. The inspection method of claim 18, wherein generating the multi-step difference image from the weighted subtraction of the test image by the combination of at least one of the second-step comparison images and at least two of the first-step images comprises: Based on Equation I MSDIFF，i =I SecondStep，i -∑ j≠i α j I SecondStep，j -γ[I FirstStep，i -∑ j≠i β j I FirstStep，j ] generates the multi-step difference image (I MSDIFF，i ), where I SecondStep，i Corresponding to the test image, ∑ j≠i α j I SecondStep，j Corresponding to the at least one of the second step comparison images, and [I FirstStep，i -∑ j≠i β jIFirstStep，j ] corresponds to the at least one of the first step images, where subscripts i and j correspond to the sample areas, where α, β and γ correspond to weights.

29. The inspection method of claim 18, wherein detecting defects on the sample based on the multi-step difference image comprises: providing the inspection reference image to the machine learning algorithm; and Defects on the sample are detected based on the output of the machine learning algorithm.

30. The inspection method of claim 18, wherein detecting defects on the sample based on the multi-step difference image comprises: Defects on the sample are detected based on the multi-step difference image using a multi-die adaptive threshold (MDAT) technique.

31. The inspection method according to claim 18, wherein the defect comprises: A deviation in at least one of shape, size, or orientation of a feature produced by said second process step.

32. The inspection method according to claim 18, wherein the defect comprises: The lack of features produced by the second process step is desired.

33. The inspection method according to claim 18, wherein the defect comprises: At least one of scratches, pits, or residual material in the inspection area after the second process step.

34. The inspection method according to claim 18, wherein the defect comprises: At least one of an unintended bridging between two features produced by the second process step or an unintended break in a feature produced by the second process step.

35. A testing system comprising: an imaging subsystem comprising an illumination source and a detector configured to image the sample; and A controller communicatively coupled to the imaging subsystem, the controller comprising one or more processors configured to execute program instructions to cause the one or more processors to implement an inspection recipe by: receiving a first step image of a plurality of sample areas from the imaging subsystem after a first process step; receiving a second step image of the plurality of sample areas after a second process step, wherein the second process step modifies the sample in at least one of the plurality of sample areas from the imaging subsystem; identifying one of the plurality of sample areas as a test area and identifying at least some of the remaining sample areas as comparison areas, wherein the second step image of the test area is a test image and the second step image of the comparison area is a second step comparison image; generating a multi-step difference image from the test image by weighted subtraction of at least one of the second-step comparison images and a combination of at least two of the first-step images; and A defect in the test area associated with the second process step is identified based on the multi-step difference image.

36. The inspection system of claim 35, wherein the number of the first step images used to generate the multi-step difference image is greater than the number of the second step comparison images by 1 so as to generate the multi-step difference image based on the same number of the first step images and the second step images.

37. The inspection system of claim 35, wherein the imaging subsystem comprises: Brightfield imaging subsystem.

38. The inspection system of claim 35, wherein the imaging subsystem comprises: Dark field imaging subsystem.

Images