Inspection system and inspection method

Abstract

An inspection system includes a controller comprising one or more processors configured to execute program instructions. The program instructions cause the one or more processors to receive at least a first portion of a first set of repeating scan bands of a first scan path of a sample. The program instructions cause the one or more processors to generate an image by averaging the first portion of the first set of repeating scan bands. Averaging the first portion of the first set of repeating scan bands reduces noise in the image. The program instructions cause the one or more processors to use the image to detect one or more defects in a detection area of ​​the sample.

Description

Technical Field

[0001] This invention relates generally to defect detection, and more specifically to reducing shot noise in defect detection. Prior Art

[0002] Inspection systems are typically used in semiconductor manufacturing processes to identify manufacturing defects that can lead to performance degradation or failure of a manufactured device. As manufacturing feature sizes continue to shrink, the size of manufacturing defects also decreases. This results in weaker measurable signals associated with these defects and lower signal-to-noise ratios (SNR) during defect detection. Therefore, there is a need to develop systems and methods to address these defects. Summary of the Invention

[0003] According to one or more embodiments of the present invention, a detection system is described. In some embodiments, the detection system includes a controller comprising one or more processors configured to execute program instructions. In some embodiments, the program instructions cause the one or more processors to receive at least a first portion of a first set of repeating scan bands of a first scan path of a sample. In some embodiments, the program instructions cause the one or more processors to generate an image by averaging the first portion of the first set of repeating scan bands. Averaging the first portion of the first set of repeating scan bands reduces noise in the image. In some embodiments, the program instructions cause the one or more processors to use the image to detect one or more defects in a detection area of ​​the sample.

[0004] In some embodiments, the image is an average detection image of one of the detection areas or the image is a reference image.

[0005] In some embodiments, the image is the averaged detection image. The program instructions cause one or more processors to receive a plurality of detection images. The plurality of detection images are repeated images of the detection region. The plurality of detection images include a first plurality of signal-to-noise ratios (SNRs). Each of the plurality of detection images includes noise contributing to the first plurality of SNRs. The program instructions cause one or more processors to average the plurality of detection images to generate the averaged detection image. The averaged detection image includes a second SNR. The second SNR is greater than each of the first plurality of SNRs.

[0006] In some embodiments, the detection system generates the plurality of detection images from the first portion of the first set of repeating scan bands in a scanning mode.

[0007] In some embodiments, the program instructions cause one or more processors to subtract the reference image from the average detected image to generate a difference image.

[0008] In some embodiments, the program instructions cause one or more processors to detect one or more defects in the detection area based on the difference image.

[0009] In some embodiments, the program instructions cause one or more processors to apply a difference filter to the difference image to generate a filtered difference image, and detect one or more defects in the detection area based on the filtered difference image.

[0010] In some embodiments, the program instructions cause one or more processors to classify the one or more defects as at least one of a disturbance point or a defect of concern.

[0011] In some embodiments, the program instructions cause one or more processors to align the first portion of the first set of repeating scan bands before averaging the first portion.

[0012] In some embodiments, these instructions cause one or more processors to align the first portion using synchronous interpolation.

[0013] In some embodiments, the detection system includes one or more optical detection subsystems. The one or more optical detection subsystems generate the first set of repeating scan bands of the first scan path. The controller is communicatively coupled to the one or more optical detection subsystems.

[0014] In some embodiments, the detection system includes a stage. The stage is translated, thereby causing one or more optical detection subsystems to generate the first set of repeating scan bands of the first scan path.

[0015] In some embodiments, the detection system includes one or more encoders or interferometers. The encoder or interferometers are configured to monitor the translation of the stage. Programming instructions cause one or more processors to control the translation of the stage to align it with the first set of repeating scan bands of the first scan path.

[0016] In some embodiments, the one or more processors receive at least the first portion of the first set of repeating scan bands of the first scan path from the one or more optical detection subsystems.

[0017] In some embodiments, the one or more optical detection subsystems include two or more optical detection subsystems.

[0018] In some embodiments, the one or more optical detection subsystems include a single optical detection subsystem.

[0019] In some embodiments, the one or more optical detection subsystems include at least one first optical mode; wherein the one or more optical detection subsystems use the first optical mode to generate the first set of repeating scan bands.

[0020] In some embodiments, the first optical mode includes a first illumination wavelength, a first illumination polarization, and a first illumination angle of one or more optical detection subsystems.

[0021] In some embodiments, the image is the average detection image of the detection area. The image also serves as a reference image for subsequent detection.

[0022] In some embodiments, the first set of repeating scan bands includes at least two repeating scan bands. In some embodiments, the first set of repeating scan bands includes between two and four repeating scan bands.

[0023] In some embodiments, the first set of repeating scan bands includes at least four repeating scan bands.

[0024] In some embodiments, the image is one of a plurality of images used to generate a reference image. The program instructions cause one or more processors to generate the reference image from the plurality of images. The single image is generated by scan averaging. In some embodiments, the detection region includes one or more patterned features. The plurality of images are derived from each of the plurality of reference regions including the one or more patterned features.

[0025] In some embodiments, the program instructions cause one or more processors to receive a second portion of a second set of repeating scan bands in a second scan path of the sample and to average the second portion of the second set of repeating scan bands; wherein the image includes at least the first portion and the second portion.

[0026] In some embodiments, the noise includes a readout noise or a shot noise.

[0027] According to one or more embodiments of the present invention, a detection system is described. In some embodiments, the detection system includes one or more optical detection subsystems. In some embodiments, the detection system includes a controller communicatively coupled to one or more optical detection subsystems. The controller includes one or more processors for executing program instructions. The program instructions cause the one or more processors to receive at least a first portion of a first set of repeating scan bands of a first scan path of a sample. The program instructions cause the one or more processors to generate an image by averaging the first portion of the first set of repeating scan bands. Averaging the first portion of the first set of repeating scan bands reduces noise in the image. The program instructions cause the one or more processors to use the image to detect one or more defects in a detection area of ​​the sample.

[0028] In some embodiments, the image is an average detection image of one of the detection areas; the image is a reference image; or the image is one of a plurality of images used by the processors to generate the reference image.

[0029] In some embodiments, the image is the averaged detection image. The program instructions cause one or more processors to receive a plurality of detection images. The plurality of detection images are repeated images of the detection region. The plurality of detection images include a first plurality of signal-to-noise ratios (SNRs). Each of the plurality of detection images includes noise contributing to the first plurality of SNRs. The program instructions cause one or more processors to average the plurality of detection images to generate the averaged detection image. The averaged detection image includes a second SNR. The second SNR is greater than each of the first plurality of SNRs.

[0030] In some embodiments, the image is one of a plurality of images. The program instructions cause one or more processors to generate the reference image from the plurality of images.

[0031] In some embodiments, the detection region includes one or more patterned features, wherein the plurality of images are derived from a plurality of reference regions each including the one or more patterned features.

[0032] According to one or more embodiments of the present invention, a method is described. In some embodiments, the method includes receiving at least a first portion of a first set of repeating scan bands of a first scan path of a sample. In some embodiments, the method includes generating an image by averaging the first portion of the first set of repeating scan bands. Averaging the first portion of the first set of repeating scan bands reduces noise in the image. In some embodiments, the method includes using the image to detect one or more defects in a detection region of the sample.

[0033] It should be understood that the foregoing overview and the following detailed description are merely illustrative and not intended to limit the invention as claimed. The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments of the invention and, together with the overview, serve to explain the principles of the invention. Simple diagram description

[0034] Those skilled in the art can better understand the many advantages of this invention by referring to the accompanying drawings, among which:

[0035] Figure 1A is a block diagram of a detection system according to one or more embodiments of the present invention.

[0036] Figure 1B is a simplified schematic diagram of one of the optical detection systems according to one or more embodiments of the present invention.

[0037] Figure 2A is a flowchart illustrating one of the steps performed in a detection method according to one or more embodiments of the present invention.

[0038] Figure 2B is a simplified block diagram of a flowchart depicting a method in Figure 2A according to one or more embodiments of the present invention.

[0039] Figure 3 depicts an example of generating an average image from a set of four images according to one or more embodiments of the present invention.

[0040] Figures 4A and 4B depict graphs illustrating improved defect detection in differential images generated by self-scanning averaging according to one or more embodiments of the present invention.

[0041] Figure 5 depicts a sample with one of the scanning paths according to one or more embodiments of the present invention. Implementation Method

[0042] Cross-reference to related applications

[0043] This application asserts the rights of U.S. Provisional Application No. 63 / 326,268, filed March 31, 2022, pursuant to 35 USC § 119(e), the entire contents of which are incorporated herein by reference.

[0044] The invention has been particularly shown and described with respect to certain embodiments and specific features thereof. The embodiments set forth herein are to be regarded as illustrative rather than restrictive. It will be readily understood by those skilled in the art that various changes and modifications in form and detail can be made without departing from the spirit and scope of the invention. Detailed reference will now be made to the disclosed subject matter illustrated in the accompanying drawings.

[0045] Embodiments of this invention relate to a system and method for defect detection using multiple images. The multiple images may be inspection images and / or reference images. The multiple images are repeated images from a region of a sample. The region may be a inspection region and / or a reference region. The region may also be referred to as a scan band, a portion of a scan band, multiple scan bands, or portions of multiple scan bands. The multiple images are averaged to produce an average image of the region. Image averaging may also be referred to as scan averaging, frame averaging, N-pass averaging, and / or signal averaging. Averaging may occur by adding the intensity values ​​of each common pixel location and then dividing by the number of images. The average image may be an average inspection image. A difference image is generated by subtracting a reference image from the average inspection image. The difference image can be analyzed to detect defects in the inspection region of the sample.

[0046] Semiconductor devices are typically manufactured using a series of process steps, such as (but not limited to) deposition of process layers and / or photoresist, exposure of a photoresist to be patterned, etching of the photoresist and / or underlying process layers, polishing, or similar processes. Variations in any of these processes, individually or in combination, can result in variations in the characteristics of the manufactured sample. In a general sense, some sample variations may be acceptable and considered as points of disturbance, while others may lead to unacceptable performance degradation or device failure and are considered defects of concern (DOI).

[0047] Inspection tools can be used to inspect a sample for defects after selected procedural steps. For example, an inspection tool can generate an inspection image of a sample after a specific procedural step. This inspection image can also be called a test image. This inspection image is then typically compared to a reference image to identify defects, where defects are related to differences between the inspection and reference images. For example, a difference image can be generated by subtracting the reference image from the inspection image, where features in the difference image correspond to sample variations. These sample variations can then be classified as DOIs or scrambling points. Although various types of inspection tools have been developed, optical inspection tools that generate an image using illumination from a single light source are commonly used for online inspection due to their relatively high processing power.

[0048] Various conditions can reduce the signal-to-noise ratio (SNR) of defect signals during inspection, thus limiting accuracy or reliability. For example, as the size of features in manufacturing equipment decreases, the size of associated process defects leading to degradation or equipment failure also decreases. Even under ideal conditions, this results in a reduced SNR when the size of features and / or defects approaches the resolution limit of an inspection device. A reduced SNR can cause various noise sources to dominate or obscure the defect signal.

[0049] Shot noise is one source of noise that contributes to the signal-to-noise ratio (SNR) of a defect signal during detection. Shot noise can be attributed to photon absorption in the sensor well, random variations in absorbed photons, random electrical fluctuations in the sensor, and the like. Shot noise can limit the SNR associated with small and / or weak defect signals, and thus limit the accuracy and / or reliability of detection. For example, shot noise can overwhelm the SNR of a defect signal due to the size of a feature of the manufacturing device. Therefore, it is desirable to suppress or reduce shot noise to detect defect signals.

[0050] Some embodiments of the present invention relate to methods for reducing shot noise in images generated during optical inspection. It has been carefully considered herein that reducing shot noise in images can provide superior performance (e.g., differentiation between defects and background signals) compared to techniques that do not reduce shot noise. Reduction of shot noise in images can increase the signal-to-noise ratio (SNR). Patterned features and / or defects can then be distinguished on the reduced shot noise.

[0051] The average image can be any one or more of the images used to generate a difference image. For example, the average image can be an averaged detection image. The difference image can be generated by subtracting a reference image from the averaged detection image. The average image then introduces less noise into the difference image. Therefore, defects and shot noise can be distinguished in the difference image. Thus, the average image can improve a defect capture rate.

[0052] U.S. Patent No. 10,533,848, filed August 7, 2018, entitled "Metrology and control of overlay and edge placement errors"; U.S. Patent No. 11,415,526, filed December 29, 2020, entitled "Multi-controller inspection system"; U.S. Patent No. 10,535,131, filed November 14, 2016, entitled "Systems and methods for region-adaptive defect detection"; U.S. Patent No. 10,395,358, filed November 6, 2017, entitled "High sensitivity repeater defect detection"; U.S. Patent No. 10,557,802, filed August 13, 2018, entitled "Capture of repeater defects on a semiconductor wafer"; and U.S. Patent No. 10,557,802, filed April 19, 2019, entitled "Algorithm selector based on image U.S. Patent No. 10,801,968, entitled "frames"; U.S. Patent No. 9,518,934, entitled "Wafer defect discovery," filed November 3, 2015; U.S. Patent No. 11,120,546, entitled "Unsupervised learning-based reference selection for enhanced defect inspection sensitivity," filed September 4, 2020; U.S. Patent No. 11,416,982, entitled "Controlling a process for inspection of a specimen," filed September 23, 2020; U.S. Patent Publication No. 2021 / 0010945, entitled "Systems and methods for selecting defect detection methods for inspection of a specimen," filed June 23, 2020; and U.S. Patent No. 11,494,924, entitled "Image alignment for noisy images," filed November 10, 2020.U.S. Patent No. 10,943,838, filed June 24, 2018, entitled "Measurement of overlay error using device inspection system"; U.S. Patent No. 9,613,411, filed October 2, 2014, entitled "Creating defect classifiers and nuisance filters"; U.S. Patent No. 9,619,876, filed March 12, 2013, entitled "Detecting defects on wafers based on 2D scatter plots of values ​​determined for output generated using different optics modes"; U.S. Patent No. 10,599,944, filed November 27, 2012, entitled "Visual feedback for inspection algorithms and filters"; U.S. Patent No. 10,535,131, filed November 14, 2016, entitled "Systems and methods for region-adaptive defect detection"; and U.S. Patent No. 10,535,131, filed April 2, 2020, entitled "Learnable defect detection for semiconductor". The following U.S. patents are incorporated herein by reference in their entirety: U.S. Patent No. 11,551,348 entitled "Applications"; U.S. Patent No. 11,550,309 entitled "Unsupervised defect segmentation" filed January 8, 2019; U.S. Patent No. 9,518,934 entitled "Wafer defect discovery" filed November 3, 2015; U.S. Patent No. 10,923,317 entitled "Detecting defects in a logic region on a wafer" filed August 18, 2019; and U.S. Patent No. 10,801,968 entitled "Algorithm selector based on image frames" filed April 19, 2019.

[0053] Referring now to Figures 1A to 4B, a system and method for optical detection will be described in more detail according to one or more embodiments of the present invention.

[0054] Figure 1A is a block diagram of a detection system 100 according to one or more embodiments of the present invention.

[0055] In one embodiment, the detection system 100 includes one or more optical imaging subsystems 102 (e.g., optical imaging tools) configured to generate one or more images of a sample 104, wherein the one or more optical imaging subsystems 102 can be configured to image the sample 104 using multiple optical detection modes. For example, an optical imaging subsystem 102 may include an illumination subsystem 106 configured to illuminate one of the sample 104 using illumination 108 from an illumination source 110 and a light-gathering subsystem 112 configured to generate an image of the sample 104 using a detector 116 in response to light emitted from the sample (e.g., sample light 114) and illumination 108.

[0056] Sample 104 may comprise a substrate (e.g., a wafer or the like) formed of a semiconductor or non-semiconductor material. For example, a semiconductor or non-semiconductor material may comprise (but is not limited to) single-crystal silicon, gallium arsenide, and indium phosphide. Sample 104 may further comprise one or more layers disposed on the substrate. For example, such layers may comprise (but are not limited to) a photoresist, a dielectric material, a conductive material, and a semi-conductive material. Many different types of such layers are known in the art, and the terminology used herein is intended to cover a sample on which all types of such layers can be formed. The one or more layers formed on a sample may be patterned or unpatterned. For example, a sample may comprise a plurality of grains, each having repeatable patterned features. The formation and processing of these material layers can ultimately result in a completed device. Many different types of devices may be formed on a sample, and the terminology used herein is intended to cover a sample on which any type of device known in the art is manufactured.

[0057] Optical imaging subsystem 102 can generate one or more images of sample 104 using any technique known in the art. In some embodiments, optical imaging subsystem 102 is an optical imaging subsystem 102 in which illumination source 110 is configured to generate illumination 108 in a light-dependent form, and in which light-collecting subsystem 112 images sample 104 based on light emitted from sample 104. In some embodiments, imaging subsystem 102 is a particle imaging subsystem 102 in which illumination source 110 is configured to generate illumination 108 in a particle-dependent form. For example, particle illumination 108 may be in the form of an electron beam, an ion beam (e.g., a focused ion beam), or a neutral particle beam. Furthermore, light-collecting subsystem 112 may image sample 104 based on particles emitted from sample 104 (e.g., backscattered electrons or the like). In some cases, a particle detection system 100 may also image the sample 104 based on light emitted from the sample 104 in response to incident particle illumination 108 (e.g., based on photoluminescence or the like).

[0058] Figure 1B is a simplified schematic diagram of an optical detection system 100 according to one or more embodiments of the present invention.

[0059] The illumination source 110 may comprise any type of illumination source known in the art suitable for generating one or more forms of optical illumination 108. Furthermore, the illumination 108 may have any spectrum, such as (but not limited to) extreme ultraviolet (EUV) wavelengths, ultraviolet (UV) wavelengths, visible light wavelengths, or infrared (IR) wavelengths. Additionally, the illumination source 110 may be a broadband source, a narrowband source, and / or a tunable source.

[0060] In some embodiments, the illumination source 110 includes a broadband plasma (BBP) illumination source. In this respect, illumination 108 may include radiation emitted by a plasma. For example, a BBP illumination source 110 may include (but does not need to include) one or more pump sources (e.g., one or more lasers) configured to be focused into a gas volume, thereby causing energy to be absorbed by the gas to generate or sustain a plasma suitable for emitting radiation. Furthermore, at least a portion of the plasma radiation may be used as illumination 108.

[0061] In another embodiment, the illumination source 110 may include one or more lasers. For example, the 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.

[0062] The illumination source 110 can further generate illumination 108 with any time profile. For example, the illumination source 110 can generate continuous wave (CW) illumination 108, pulsed illumination 108, or modulated illumination 108. In addition, illumination 108 can be delivered from the illumination source 110 via free space propagation or guided light (e.g., an optical fiber, a light guide, or the like).

[0063] The illumination subsystem 106 and / or the optical imaging subsystem 102 may include various components for directing illumination 108 to sample 104, such as (but not limited to) a lens 118, a mirror, or the like. Furthermore, these components may be reflective or transmissive elements. In this way, the depiction of lens 118 as a transmissive element in FIG1B is merely illustrative and not limiting. The illumination subsystem 106 may further include one or more optical elements 120 for modifying and / or adjusting light in the 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. In an embodiment, the illumination subsystem 106 and / or the optical imaging subsystem 102 include a beam splitter 130 oriented to simultaneously direct illumination 108 to sample 104 and collect sample light 114 emitted from sample 104. In this respect, illumination 108 and sample light 114 can share the same path between beam splitter 130 and sample 104.

[0064] In some embodiments, the detection system 100 includes a translation stage 122 for fixing and / or positioning one of the samples 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.

[0065] The optical imaging subsystem 102 may include various components for collecting at least a portion of the sample light 114 radiation emitted from the sample 104 (e.g., sample light in the case of an optical imaging subsystem 102) and directing at least a portion of the sample light 114 to a detector 116 for generating an image.

[0066] The detection system 100 can further image the sample 104 using any technique known in this art. In some embodiments, the detection system 100 generates an image of the sample 104 in a scanning mode by the following steps: focusing illumination 108 onto the sample 104 as a spot or line; capturing a spot or line image; and scanning the sample 104 to establish a two-dimensional image. In this configuration, scanning can be achieved by moving the sample 104 relative to illumination 108 (e.g., using a translation stage 122), by moving illumination 108 relative to the sample 104 (e.g., using an actuated mirror or the like), or a combination thereof. Scanning may include scanning the sample 104 along a scanning path to generate a scan band of the scanning path. In some embodiments, the detection system 100 generates an image of the sample 104 in a static mode by guiding illumination 108 onto the sample 104 in a two-dimensional field of view and directly capturing a two-dimensional image using detector 116.

[0067] An image generated by the detection system 100 can 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, or the like. In some embodiments, the image can be a raw image from an optical imaging subsystem 102. In this configuration, the detection image can include various patterned features on the sample. Furthermore, images can be stitched together to form a composite image of sample 104 or a portion thereof, but this is not intended to be a limitation of the invention. Although the image has been described as containing patterned features, this is not intended to be a limitation of the invention. It should be further considered that the image may originate from a sample or wafer without patterned features.

[0068] Detector 116 may include any type of sensor known in the art suitable for measuring sample light 114. For example, a 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 one-line sensor, or a time-delay integration (TDI) sensor. As another example, a detector 116 may include two or more single-pixel sensors, such as (but not limited to) a photodiode, a collapsing photodiode, a photomultiplier tube, or a single-photon detector. In some embodiments, detector 116 may include a TDI sensor. The TDI sensor may include multiple pixel columns and a readout column. The TDI sensor may include a clock signal that continuously moves charge from one pixel column to the next until the charge reaches the readout column, thereby generating a series of images. By synchronizing the charge transfer (e.g., based on the clock signal) with the movement of the sample along the scan direction, the charge can continue to accumulate across the pixel columns to provide a signal-to-noise ratio relatively higher than that of a one-line sensor.

[0069] In some embodiments, the detection system 100 includes a controller 124. The controller 124 may include one or more processors 126 configured to execute program instructions held on a memory 128 (e.g., a memory medium). In this respect, one or more processors 126 of the controller 124 may execute any of the various program steps described herein. The controller 124 may be communicatively coupled to one or more optical detection subsystems 102. The controller 124 may receive images from the optical detection subsystems 102.

[0070] One or more processors 126 of controller 124 may comprise any processing element known in the art. In this sense, one or more processors 126 may comprise any microprocessor type device configured to execute algorithms and / or instructions. In one embodiment, one or more processors 126 may comprise a desktop computer, host computer system, workstation, video computer, parallel processor, or any other computer system (e.g., a network-connected computer) configured to execute a program (configured to operate detection system 100), as described throughout the invention. It should be further appreciated that the term "processor" may be broadly defined to cover any device having one or more processing elements that execute program instructions from a non-transitory memory 128.

[0071] Memory 128 may comprise any storage medium known in the art suitable for storing program instructions executable by one or more associated processors 126. For example, memory 128 may comprise a non-transitory memory medium. By another example, memory 128 may comprise (but is not limited to) a read-only memory, a random access memory, a magnetic or optical memory device (e.g., a magnetic disk), a magnetic tape, a solid-state drive, and the like. It should be further noted that memory 128 may be housed together with one or more processors 126 in a common controller housing. In one embodiment, memory 128 may be remotely located relative to the physical location of one or more processors 126 and controller 124. For example, one or more processors 126 of controller 124 may access a remote memory (e.g., a server) accessible via a network (e.g., the Internet, an intranet, and the like). Therefore, the above description should not be construed as limiting the invention but is merely illustrative.

[0072] An optical detection mode may correspond to any combination of parameters used to generate an image of one of the samples 104, including (but not limited to) the nature of the illumination directed to the sample 104 or the light collected from the sample 104. Furthermore, imaging using different optical detection modes can typically be performed using any number of optical imaging subsystems 102.

[0073] An optical detection mode provided by the optical imaging subsystem 102 can be controlled based on the control of any combination of components in the illumination subsystem 106 or the light-gathering subsystem 112. For example, control of the illumination 108 directed to the sample 104 can be provided directly by the illumination source 110 and / or by optical elements 120, such as (but not limited to) a spectral filter for controlling a wavelength of the illumination 108, a polarizer for controlling a polarized beam of the illumination 108, or an apodizer (e.g., in an illumination pupil plane) for controlling an angular distribution of the illumination 108 on the sample 104. As another example, the control of the sample light 114 collected from sample 104 and transmitted to detector 116 can be provided by optical element 120, such as (but not limited to) a spectral filter for controlling one wavelength of the sample light 114 transmitted to detector 116, a polarizer for controlling one polarization of the sample light 114 transmitted to detector 116, or an apodizer (e.g., in a focusing pupil plane) for controlling one angular distribution of the sample light 114 transmitted to detector 116.

[0074] As an illustration, a particular optical detection mode may correspond to illumination 108 having a selected spectrum (e.g., described by a bandwidth and / or a center wavelength) and being directed to a selected polarized light on the sample at a selected incident angle (e.g., defined by an illumination aperture or apodizer). The optical detection mode may further correspond to a specific spectrum and polarized light of sample light 114 directed to detector 116 (both of which may be the same as or different from the illumination 108 incident on sample 104).

[0075] Furthermore, either the illumination source 110 and / or the optical element 120 can be adjusted so that the detection system 100 can be configured to provide different optical detection modes. For example, either of the optical elements 120 can be directly tuned and / or controlled by an actuator to provide different optical detection modes. In some embodiments, the controller 124 generates drive signals for either the illumination source 110 and / or the optical element 120 to selectively provide different optical detection modes.

[0076] In embodiments, the optical imaging subsystem 102 may generate one or more images. The images may include a detection image and a reference image, as will be further described herein. In embodiments, the optical imaging subsystem 102 may use one or more optical modes to generate one or more images.

[0077] While much of this invention describes images in the context of inspection and reference images, this is not intended to limit the invention. There is not necessarily a distinction between inspection and reference images. For example, such inspection and reference images may correspond to independently generated images or portions of one or more larger images (e.g., sub-images of one or more larger images). As an illustration, an inspection image may be associated with a portion of a larger image representing one of multiple grains on a sample. In this way, an inspection image may be a sub-image representing a region of interest (e.g., a specific grain) for defect detection at a particular time. As another illustration, a reference image may be generated based on one or more different portions of larger images associated with different regions (e.g., additional grains) on the sample. For example, a reference image may correspond to an average of one or more sub-images associated with one or more additional grains on the sample. In this configuration, a difference image at a particular process step may correspond to a difference between the image of the inspection sub-image and the reference image (e.g., a combination of sub-images).

[0078] In this document, it has been carefully considered that defects on the same sample can respond differently to imaging using different optical modes, thus allowing for improved defect analysis or identification by taking into account the images produced using different optical modes. Additional embodiments of the invention relate to inspection systems suitable for performing multi-mode optical inspection. In this document, it has been carefully considered that multi-mode inspection can provide substantially superior performance (e.g., differentiation between defects and background signals) compared to single-mode inspection techniques. Furthermore, increasing the number of inspection modes generally increases inspection performance. However, it should be further recognized that, especially in cases where multi-mode inspection requires multiple imaging scans and is therefore more time-consuming to perform, it is desirable to balance the number of optical inspection modes used during inspection.

[0079] In some embodiments, a single optical imaging subsystem 102 may be configured to image the sample 104 simultaneously or sequentially using multiple optical detection modes. In some embodiments, different optical imaging subsystems 102 are used to provide at least some different optical detection modes.

[0080] In some embodiments, the detection system 100 sequentially provides images using different optical detection modes. For example, the detection system 100 may sequentially switch between different optical imaging subsystems 102 and / or adjust the parameters of an optical imaging subsystem 102 to provide sequential images of the sample 104 using different optical detection modes.

[0081] In some embodiments, the optical imaging subsystem 102 can be configured to simultaneously provide two or more images using different optical detection modes. As an illustration, an optical imaging subsystem 102 may include two or more collection channels, each having a separate detector 116. The optical imaging subsystem 102 may then have additional optical elements 120 (e.g., separate spectral filters, polarizers, or the like) for splitting sample light 114 into one or more of the various channels and / or for providing separate control over the properties of the sample light 114 directed to the associated detectors 116 in each channel.

[0082] Referring now to Figures 2A and 2B, a flowchart illustrating one of the steps performed in a method 200 is described according to one or more embodiments of the present invention. The embodiments and implementation techniques previously described in the background of the detection system 100 should be interpreted as extensions to method 200. For example, memory 128 may hold program instructions that cause processor 126 to execute one or more steps of method 200. However, it should be further noted that method 200 is not limited to the architecture of the detection system 100.

[0083] Figures 2A and 2B also reference various images. It should be understood that these images may be images directly generated by the detection system 100 or a portion thereof in a single step, or composite images formed by stitching together sub-images generated by the detection system 100. Furthermore, images may be cropped or otherwise modified as needed. Therefore, the references to images herein are illustrative only and do not limit the types of images that can be generated by the detection system 100.

[0084] In some embodiments, method 200 includes a step 210 of receiving a reference image 212. Program instructions may cause a processor to generate the reference image 212 from a plurality of images, each comprising a plurality of regions with a patterned feature. The reference image 212 is generated from the plurality of images. The plurality of images are derived from a plurality of reference regions each comprising a matching patterned feature. The plurality of reference regions may be derived from sample 104 and / or from one or more additional samples. The reference regions may comprise a design. For example, the reference region may comprise a design having one or more patterned features. In some embodiments, the plurality of images used to generate the reference image 212 are averaged to reduce shot noise. In this respect, the reference image 212 may be generated from the averaged images.

[0085] In this embodiment, the reference image 212 may be received by one or more processors 126. The detection system 100 may generate the reference image 212. The reference image 212 may be received by one or more processors 126 from the optical imaging subsystem 102. The reference image 212 received by one or more processors 126 may then be maintained in memory 128. The reference image 212 received by one or more processors 126 may also be received from memory 128.

[0086] Reference image 212 may include a pixel array 214. Each pixel in pixel array 214 may include an intensity 216. Intensity 216 may also be referred to as a signal or a grayscale. Intensity 216 may be associated with patterned features in the reference area, defects in the reference area, and signals associated with shot noise. Reference image 212 may represent an ideal or reference. It may be assumed that the reference area is defect-free such that intensity 216 is not associated with any defects, but this is not intended to be limiting.

[0087] In some embodiments, method 200 includes step 220 of receiving a detection image 222 of a detection area of ​​a sample 104. The detection area may include any area on the sample 104 to be detected for defects. The detection area may include a design. For example, the detection area may include a design having one or more patterned features. The detection image 222 is a repeating image of the detection area of ​​the sample 104. The repeating images may be acquired from the same detection area using the same optical mode. In this respect, the variable factor between the detection images 222 may be shot noise.

[0088] In this embodiment, the detection image 222 may be received by one or more processors 126. The detection system 100 may generate the detection image 222. The detection image 222 may be received from the optical imaging subsystem 102 by one or more processors 126. The detection image 222 received by one or more processors 126 may then be maintained in memory 128. The reference image 212 received by one or more processors 126 may also be received from memory 128.

[0089] The inspection image 222 may include a pixel array 224. Each pixel in the pixel array 224 may include an intensity 226. The intensity 226 may be associated with patterned features in the inspection area, defects in the inspection area, and signals associated with shot noise. The inspection image 222 may represent a deviation from the reference of an original image of an area of ​​interest due to defects. The inspection area may be inspected without assuming the absence of defects, but rather with respect to defects. The signal associated with shot noise may be relatively high compared to the signal associated with patterned features and / or the signal associated with defects. In this respect, shot noise can introduce noise into patterned features and / or defects. The inspection image 222 may each include a signal-to-noise ratio (SNR). Shot noise can contribute to the SNR. Therefore, the inspection image 222 may contain a relatively low signal-to-noise ratio (SNR). The relatively low SNR of the inspection image 222 can cause defects to be hidden by shot noise.

[0090] The detection images 222 can be aligned, registered, and / or scaled so that they all have a common number of aligned pixels. Pixels in pixel array 224 can correspond to positions on the detection area of ​​the sample. In this respect, each detection image 222 may contain a common pixel position in pixel array 224 corresponding to a common physical position on the sample. In some embodiments, step 220 further includes registering and / or scaling the detection images 222. In this way, the detection images 222 may have a common number of pixels, and each pixel of all detection images 222 can correspond to a common position on the sample. Registration and / or scaling may also be referred to as aligning pixels in pixel array 224. In embodiments, the images can be aligned by processor 126. It should be further considered that the detection images 222 may be pre-aligned before receiving the detection images 222. Alignment can occur by any suitable alignment method, as will be further described herein.

[0091] In some embodiments, method 200 includes a step 230 of generating an average inspection image 232. The average inspection image 232 may be generated by averaging an inspection image 222 of an inspection area. The average inspection image 232 may include signals associated with patterned features, signals associated with defects, and signals associated with shot noise.

[0092] The averaged detection image 232 may include an average pixel array 234. The average pixel array 234 can be found by averaging each of the pixel arrays 224 of the detection image 222. Each pixel in the average pixel array 234 may include an average intensity 236. The average pixel 236 can be found by averaging the intensity 236 between the pixels of the detection image 222. The average intensity 236 may refer to adding the intensity value of each common pixel location and then dividing by the number of images. The averaged detection image 232 may include a signal-to-noise ratio (SNR). Shot noise can contribute to the SNR. However, the shot noise in the SNR of the averaged detection image 232 can be averaged from each of the detection images 222. Averaging can reduce the shot noise in the averaged detection image 232 and similarly increase the SNR. Therefore, the averaged detection image 232 may include an SNR that is relatively high or greater than the SNR of the detection image 222. The relatively high SNR of the average detected image (232) prevents defects from being hidden by shot noise. However, defects can be hidden by patterned features.

[0093] In some embodiments, method 200 includes a step 240 of generating a difference image 242. The difference image 242 can be generated by subtracting a reference image 212 from an average detection image 232. The detection region and one or more reference regions have a common design. For example, the detection region and one or more reference regions may each include a patterned feature. The patterned feature can then be subtracted from the detection region using an image from the reference region. Pixels in the pixel array 214 of the reference image 212 can be aligned with pixels in the average pixel array 234 of the average detection image 232. In this respect, the pixel array 214 and the average pixel array 234 can be aligned such that each pixel corresponds to a common patterned feature. The common patterned feature can be aligned between the reference image 212 and the average pixel array 234 by aligning the reference image 212 and the average detection image 232. In some embodiments, the reference image 212 and the average detection image 232 are pre-aligned. Alignment can occur by any suitable alignment method, as will be further described herein.

[0094] The difference image 242 may include a difference pixel array 244. The difference pixel array 244 can be found by subtracting the pixel array 214 of the reference image 212 from the average pixel array 234 of the average detection image 232. Each pixel in the difference pixel array 244 may include a difference intensity 246. The difference intensity 246 can be found by subtracting the intensity 216 between pixels of the average pixel array 224 from the average intensity 236 between pixels of the average pixel array 234. The difference image 242 may include a signal-to-noise ratio (SNR). The signal in the difference image 242 may refer to the intensity of defects and / or patterned features, and the noise may refer to shot noise. However, shot noise and patterned features can contribute minimally to the SNR. For example, shot noise can contribute minimally to the SNR by averaging. By another example, patterned features can contribute minimally to the SNR by subtraction to find differences. Subtraction can remove common features (e.g., the intensity associated with patterned features). Therefore, the difference image 242 may contain a relatively high signal-to-noise ratio (SNR). The relatively high SNR of the difference image 242 prevents defects from being hidden by shot noise and / or patterned features. Improving the SNR can be expected to improve the detection rate or defect capture rate of one of the defects in the difference image 242.

[0095] In this embodiment, the SNR of the reference image 212, the average detection image 232, and / or the difference image 242 can be defined by the following equation.

[0096] SNR = (Defect Contrast Ratio) *

[0097] N well, v can refer to the well depth in the photons of detector 116. The number of repetitions can refer to the number of averaged repeating images. Performing scan averaging involves multiplying the N well, v term in the above equation by the number of averaged repeating images. In this respect, the number of repeating images can reduce shot noise by the square root of the number of repeating images.

[0098] Defect contrast can refer to the contrast between a pixel associated with a defect and the background pixels surrounding the pixel associated with the defect. Defect contrast can be defined by the following equation:

[0099]

[0100] Image dynamic range refers to the ratio between the maximum and minimum intensity of an image. Image dynamic range can be defined by the following equation:

[0101]

[0102] In some embodiments, method 200 includes step 250 of detecting one or more defects. Defects may be detected based on difference image 242. Defects may be detected in a detection area based on the difference image. In some embodiments, in step 250, defects may be detected directly from the difference image 242.

[0103] In some embodiments, step 250 further includes detecting defects from a filtered difference image. In an embodiment, controller 124 may be configured to apply a difference filter to the difference image. The difference image is filtered by one or more difference filters. The difference filter may be applied to difference image 242 to produce a filtered difference image. The difference filter may include (but is not limited to) high-pass, low-pass, mixed, narrow-band, vertical, horizontal, other selected-direction filters, a convolution filter, a digital convolution filter, and the like. In an embodiment, the difference filter may be customized to filter the difference image determined by self-scanning or frame averaging. For example, the number of scans used to generate multiple images may be customized to use a difference filter. Defects can then be detected in the detection area based on the filtered difference image.

[0104] Defects can be detected using any technique. In an embodiment, the controller 124 can be configured to apply a detection algorithm to process filtered and / or unfiltered difference images for defect detection. The detection algorithm may include any detection algorithm, such as (but not limited to) a multi-grain adaptive criterion (MDAT) and the like.

[0105] In some embodiments, method 200 may include a step of classifying one or more defects. Defects may typically include defects of interest (DOI) and / or disruptive defects. Defects of interest may affect device performance, degrade system performance, and / or be fatal defects. Disruptive defects typically do not affect device performance. The step may include classifying a defect as at least one of a disruptive point, a defect of interest (DOI), or the like.

[0106] Although step 220 is depicted after step 210 in the flowchart of method 200, this is not intended to be a limitation of the invention. It should be further considered that step 220 may be performed before and / or simultaneously with step 210. In this regard, a detection image 222 may be generated and then a reference image 212 may be generated; a reference image 212 may be generated and then a detection image 222 may be generated; or a reference image 212 and a detection image 222 may be generated simultaneously. Generally speaking, step 220 may be performed after step 210. Generally speaking, step 230 may be performed after step 220.

[0107] In some embodiments, the detection system 100 generates a reference image 212 and / or a detection image 222 in a scanning mode. The detection system 100 can repeatedly scan the reference area to generate the reference image 212 and / or repeatedly scan the reference area to generate the detection image 222.

[0108] In some embodiments, one or more steps of method 200 may be repeated for multiple optical modes. An optical mode may refer to an optical mode of optical detection subsystem 102. An optical detection mode may include any unique combination of optical imaging parameters of one or more optical imaging subsystems 102. For example, an optical detection mode may include a unique combination of parameters (such as, but not limited to, illumination wavelength, polarization, or angle) associated with illumination 108 and / or sample light 114 used to generate an image.

[0109] The optical inspection subsystem 102 can generate a set of inspection images of the inspection area of ​​the sample and / or a set of reference images of the reference area using optical inspection modes. Method 200 can be performed on each plurality or set of reference images 212 and inspection images 222 acquired using different optical modes. Images acquired in separate modes can be averaged separately. Then, reference images and averaged inspection images can be generated for each optical mode. Then, difference images for each optical mode can be generated by subtracting the reference images and averaged inspection images. The difference images can then depict the intensity of defects in each optical mode within the inspection area.

[0110] For example, the optical inspection subsystem 102 can be configured to generate a first plurality of inspection images 222 and / or a first plurality of reference images 212 using a first optical mode. For example, the optical inspection subsystem 102 may have a first illumination wavelength, a first illumination polarization, and a first illumination angle. The processor 126 can receive the first plurality of inspection images and the first reference images, generate a first average inspection image by averaging the first plurality of inspection images 222, and subtract the first reference image from the first average inspection image to generate a first difference image.

[0111] By another example, the optical inspection subsystem 102 can be configured to generate a second plurality of inspection images 222 and / or a second plurality of reference images 212 using a second optical mode. The processor 126 can generate a second average inspection image by averaging the second plurality of inspection images and subtract the second reference image from the second average inspection image to generate a second difference image.

[0112] By another example, the optical inspection subsystem 102 can also be configured to generate any Nth plurality of inspection images 222 and / or Nth plurality of reference images 212 using an Nth optical mode for any integer N.

[0113] In some embodiments, one or more multi-mode inspection techniques may be performed on the difference image. Multi-mode inspection can be designed to reject wafer noise when shot noise is negligible. Multi-mode inspection techniques may complement method 200 when defect contrast is low and shot noise constitutes a large portion of the total noise.

[0114] Although method 200 has been described as being repeated for any number of optical modes, this is not intended to be limiting. When implementing method 200, time requirements can effectively limit the number of optical modes.

[0115] As previously described herein, some optical imaging subsystems 102 may be able to generate images simultaneously using multiple optical detection modes (e.g., via different collection channels), while others may require multiple separate measurements (e.g., scans) to acquire images using different optical detection modes. In this regard, multiple sets of reference images 212 and / or multiple sets of detection images 222 can be generated and / or received simultaneously in multiple optical modes. For example, the optical imaging subsystem 102 may be configured to simultaneously generate a first plurality of detection images acquired in a first optical mode and a second plurality of detection images acquired in a second optical mode. Method 200 can then generate differential images 242 for each optical mode.

[0116] In some embodiments, sample 104 includes a plurality of detection regions. One or more steps of method 200 may be repeated for each detection region of sample 104. For example, steps 220 to 250 may be repeated for each detection region to detect defects in the detection region. Step 210 may be repeated together with steps 220 to 250 as appropriate. For example, if the current detection region contains the same patterned features as one or more reference regions in memory 128, step 210 may or may not be repeated. If the detection region contains the same patterned features as the reference regions maintained in memory 128, reference image 212 may be received from memory 128.

[0117] Referring now to Figure 3, a non-limiting example of the application of method 200 is described. As depicted, images 302a to 302d are received. Images 302a to 302d are averaged to produce an average image 304. Images 302a to 302d and the average image 304 each contain a pixel array. Each pixel in the pixel array contains an intensity value. In the example shown, the intensity value is a gray level of the pixel.

[0118] As depicted, images 302a to 302d and the average image 304 may contain shot noise. Shot noise may be random across the image system. For example, shot noise may follow a Poisson distribution or vary randomly within a region. In some embodiments, images 302a to 302d and the average image 304 may each contain a standard deviation of one intensity value. The average deviation of the intensity values ​​may indicate the contribution of shot noise to the image's SNR.

[0119] The SNR of the average image 304 is substantially higher than that of images 302a through 302d. Image 302a is depicted with a standard deviation of 8.5147 gray levels, image 302b with a standard deviation of 8.5483 gray levels, image 302c with a standard deviation of 8.5721 gray levels, image 302d with a standard deviation of 8.5634 gray levels, and image 304 with a standard deviation of 4.324 gray levels. The standard deviation of the average image 304 has been significantly reduced to be lower than that of images 302a through 302d. For example, the standard deviation of the average image 304 may be about half that of images 302a through 302d, because the four images were averaged to produce the average image 304.

[0120] It should be noted that the intensity values ​​of patterned features and defects are not present in images 302a to 302d and not present in image 304. The absence of intensity values ​​of patterned features and defects in images 302a to 302d and 304 clearly illustrates the shot noise reduction provided by averaging. Although not depicted, the images may contain intensity values ​​indicating patterned features and / or defects.

[0121] In some embodiments, images 302a to 302d may be instances of detection image 222. Similarly, average image 304 may be an instance of average detection image 232.

[0122] In this example, four images 302 are received. Further consideration should be given that method 200 can be applied to any number of images. For example, a detection image 222 in a given optical mode may contain two or more images. By another example, a detection image 222 in a given optical mode may contain three or more images. By another example, a detection image 222 in a given optical mode may contain four or more images. By another example, a detection image 222 in a given optical mode may contain between four and ten images. By another example, a detection image 222 in a given optical mode may contain ten or more images. In this regard, any number of repeating images can be averaged to produce an average detection image 232. The number of repeating images selected can be reduced based on the amount of shot noise weighted relative to the process time required to produce the repeating images.

[0123] Referring now to Figures 4A and 4B, Figures 402 and 404 are described according to one or more embodiments of the present invention. Figure 402 depicts experimental results of a difference image formed by subtracting a reference image 212 from a detection image 222. In this respect, the difference image referenced in Figure 402 is not formed by subtracting the average detection image 242 and the reference image 212. Figure 404 depicts experimental results of a difference image formed by subtracting the reference image 212 from the average detection image 232. For example, the average detection image 232 may be averaged from ten images or scans.

[0124] Figures 402 and 404 are obtained from inspection images within the same inspection area. The inspection area contains known defects. Known defects are marked using defect identifiers. For example, defect IDs include 003, 103, 105, 114, 213, 214, 301, 313, 314, 413, 414, 505, 602, 605, 613, 703, and 714.

[0125] The horizontal axis of Figures 402 and 404 refers to a grayscale level of a differential image. Specifically, the horizontal axis of Figure 402 refers to a grayscale level of a differential image that has not yet been averaged from multiple scans. The horizontal axis of Figure 402 refers to a grayscale level of a differential image that has not yet been averaged from ten scans.

[0126] In Figures 402 and 404, the vertical axis represents the minimum value between the cumulative density function (CDF) and a value minus the CDF (e.g., MIN(CDF, ​​1-CDF)). The vertical axis indicates the probability that background noise exceeds the difference gray level (horizontal axis) in magnitude. The CDF can be a function of the background noise.

[0127] N-pass averaging can reduce the background reception rate in one array region as measured by the cumulative density function (CDF) of background noise. Ten repetitions or fewer can meaningfully improve the defect capture rate. For example, Figure 402 depicts three defects that are detectable, while the remaining defects are undetectable. By another example, Figure 404 depicts fourteen defects that are detectable, while the remaining defects are undetectable. Thus, in this example, N-pass averaging has improved the defect capture rate from three to fourteen.

[0128] Referring now to FIG5, a scanned sample 104 is described according to one or more embodiments of the present invention. In some embodiments, images (e.g., detection image 222 and / or reference image 212) may be acquired by swathing. Swathing may be necessary to allow the detector 116 to scan across the sample 104. The detector 116 may scan across the sample 104 by translating the stage. The detector 116 may scan across the sample 104 in a scan path 502. As depicted, the scan path 502 may be horizontal. The detector may also step in a vertical direction and scan in a new scan path 502.

[0129] A scan band may include an image capturing a portion of a row of dies on a semiconductor wafer. The scan band may be received and maintained in memory 128. Images of each row of dies may be formed using multiple scan bands. For example, one scan band may capture a portion of a region of that row of dies. Subsequently, multiple scan bands may capture multiple portions of that region of the row of dies. The scan bands of the row of dies may be aligned to capture images of the dies. The scan bands and / or portions of the scan bands may be aligned before averaging. The scan bands and / or portions of the scan bands may be partitioned into images. For example, a portion of a scan band may define an image. The image may be a detection image. Defects may be detected in the detection image. The image may also be one of a plurality of images combined to generate a reference image.

[0130] In an embodiment, the detection system 100 may generate repeating scan bands along the scan path 502 of the sample 104. For example, the detection system 100 may generate a first set of repeating scan bands along a first scan path 502a of the sample 104, a second set of repeating scan bands along a second scan path 502b of the sample 104, and so on. In some embodiments, the number of sets of repeating scan bands may be selected to fully image the patterned features.

[0131] In some embodiments, repeated scan bands of a scan path are received and then averaged immediately before moving to the next scan path. Immediate averaging of repeated scan bands can help reduce the amount of scan bands maintained in memory 128. Immediate averaging of repeated scan bands of a scan path can be advantageous when the translation stage 122 can repeat the scan path with desired accuracy. In some cases, one or more sensors (e.g., interferometers, encoders, etc.) are used to track the movement of the translation stage 122 to achieve the desired accuracy.

[0132] In some embodiments, all repeating scan bands of the scan path are maintained in memory. A portion of the repeating scan bands used to generate any of the various images can then be received and averaged. The averaged scan bands can then be stitched together to form an image. Maintaining all repeating scan bands of the scan path in memory can be advantageous when there are variations in scan band alignment. For example, alignment can vary within repeating scan bands of the same scan path and / or across scan bands of multiple scan paths. Any of the various scan bands can be maintained in memory and then aligned. Maintaining repeating scan bands in memory may require additional memory and / or processing resources.

[0133] Processor 126 may receive one or more portions of repeating scan bands of scan path 502 of sample 104. These portions may be used to generate an image. For example, processor 126 receives at least a first portion of a first set of repeating scan bands of a first scan path 502a and at least a second portion of a second set of repeating scan bands of a second scan path 502b. Processor 126 generates an image by averaging the first portion of the first set of repeating scan bands and averaging the second portion of the second set of repeating scan bands. Averaging at least the first portion of the first set of repeating scan bands and averaging at least the second portion of the second set of repeating scan bands reduces shot noise in the image. Processor 126 then uses the image to detect one or more defects in a detection area. For example, the image may be one of multiple images used to generate reference image 212. By another example, the image may be an averaged detection image 232.

[0134] Referring again generally to Figures 1A through 5, a sensor can generate an image. A pixel of the sensor can absorb a limited number of photons before pixel saturation. Pixel saturation is related to the potential well depth of the sensor within the sensor. Once the potential well depth is reached, more bits, more photons, and / or more integration time do not reduce shot noise generated by the sensor in the image. Defect contrast of less than a given percentage can be limited by the sensor well depth. For example, shot noise can fill the well depth, thereby limiting defect contrast. Method 200 can improve defect contrast by reducing shot noise.

[0135] While much of the invention has been described in the context of averaging images to reduce shot noise, this is not intended to be a limitation of the invention. It should be further considered that averaging images can reduce various sources of random noise, such as (but not limited to) readout noise, shot noise, and the like. Averaging can generally reduce random noise that is not originating from wafer variations.

[0136] As used throughout this invention, one or more images can be aligned. The images can be detection images aligned before averaging. The images can be a reference image and an averaged detection image. The reference image and the averaged detection image can be aligned before subtraction. Any suitable method can be used to align the images to the same coordinate system or origin across detections. For example, the trajectory of stage 122 can be monitored by an encoder and / or interferometer. The trajectory of stage 122 can then be controlled by a feedback system to ensure that the trajectories of repeating images are identical. In another example, residual alignment errors can be corrected by an algorithm. The algorithm can shift a first image by an offset using synchronous interpolation based on the Nyquist theorem. A minimization algorithm can determine the offset that minimizes the sum of the squares of the differences between the gray values ​​of the shifted first image and a second image at the pixel level. Preferably, alignment is performed by an image computer at subpixel resolution.

[0137] As used throughout this invention, the term "sample" generally refers to a substrate (e.g., thin-film glass or the like) formed of a semiconductor or non-semiconductor material. For example, a semiconductor or non-semiconductor material may include (but is not limited to) single-crystal silicon, gallium arsenide, indium phosphide, or a glass material. A sample may include one or more layers. For example, such layers may include (but are not limited to) a resist (including a photoresist), a dielectric material, a conductive material, and a semi-conductive material. Many different types of such layers are known in this art, and the term "sample" as used herein is intended to cover a sample on which all types of such layers can be formed. One or more layers formed on a sample may be patterned or unpatterned. For example, a sample may include a plurality of grains, each having repeatable patterned features. The formation and processing of these material layers can ultimately result in a completed device. Many different types of devices can be formed on a sample, and the term "sample" as used herein is intended to cover a sample on which any type of device known in this art is manufactured. Furthermore, for the purposes of this invention, the terms "sample" and "wafer" should be interpreted as interchangeable. Additionally, for the purposes of this invention, the terms "patterning device," "mask," and "reducing mask" should be interpreted as interchangeable.

[0138] In some embodiments, sample 104 may include one or more regions. Sample 104 may include one or more detection regions and one or more reference regions on sample 104. In some examples, the one or more reference regions may be regions on sample 104 that have a common design with the detection regions. For example, the detection regions and reference regions may each include patterned features. In some embodiments, the detection regions and reference regions may be adjacent regions on sample 104, but this is not intended to be a limitation of the invention.

[0139] Although the detection area and reference area are described as being on sample 104, this is not intended to be a limitation of the invention. In other examples, one or more reference areas may be areas from an additional sample (not depicted) that have a common design with the detection area. In this respect, the detection area may be on sample 104 and the reference area may be on an additional sample.

[0140] All methods described herein may include storing the results of one or more steps of the method embodiments in memory. The results may include any results described herein and may be stored in any manner known in the art. The memory may include any memory described herein or any other suitable storage medium known in the art. After the results have been stored, they may be accessed in memory and used by any method or system embodiment described herein, formatted for display to a user, used by another software module, method, or system, and the like. Furthermore, the results may be stored "permanently," "semi-permanently," "temporarily," or for a period of time. For example, the memory may be random access memory (RAM), and the results may not necessarily be stored in memory indefinitely.

[0141] Further careful consideration should be given to the fact that each embodiment of the methods described above may include any other step of any other method(s) described herein. Furthermore, each embodiment of the methods described above may be performed using any system described herein.

[0142] Those skilled in this art will recognize that, for clarity of concept, the component operations, devices, objects, and accompanying discussions described herein are used as examples, and various configuration modifications are anticipated. Therefore, the specific examples and accompanying discussions used herein are intended to represent their more general categories. In general, the use of any specific example intended to represent its category, and the omission of specific component operations, devices, and objects, should not be considered limiting.

[0143] As used herein, directional terms such as “top,” “bottom,” “above,” “below,” “up,” “down,” “under,” and “down” are intended to provide relative positions for descriptive purposes and are not intended to specify an absolute frame of reference. Those skilled in the art will understand that various modifications to the described embodiments are possible, and that the general principles defined herein can be applied to other embodiments.

[0144] Regarding any substantially plural and / or singular terms used in this document, those skilled in the art can convert plural to singular and / or singular to plural depending on the context and / or application. For clarity, various singular / plural arrangements are not explicitly described in this document.

[0145] The objects described herein sometimes illustrate different components contained within or connected to other components. It should be understood that the architectures depicted are merely illustrative, and many other architectures can in fact be implemented to achieve the same functionality. In a conceptual sense, any configuration of components used to achieve the same functionality is effectively "associated" to achieve the desired functionality. Therefore, any two components combined herein to achieve a particular functionality can be considered "associated" with each other to achieve the desired functionality, regardless of the architecture or intermediate components. Similarly, any two such associated components can also be considered "connected" or "coupled" to each other to achieve the desired functionality, and any two components that can be suchly associated can also be considered "coupled" to each other to achieve the desired functionality. Specific examples of couplerability include (but are not limited to) components that can physically cooperate and / or physically interact and / or wirelessly interact and / or logically interact and / or logically interact.

[0146] Furthermore, it should be understood that this invention is defined by the appended claims. Those skilled in the art will understand that, generally speaking, the terminology used herein, and especially in the appended claims (e.g., the body of the appended claims), is generally intended to be "open-ended" (e.g., the term "including" should be interpreted as "including (but not limited to)", the term "having" should be interpreted as "at least having", the term "includes" should be interpreted as "including but not limited to", and the like). Those skilled in the art will further understand that if a particular number of claims is desired, this intention will be explicitly stated in the claims, and if such a statement is lacking, this intention does not exist. For example, as an aid to understanding, the appended claims may contain the use of the introductory phrases "at least one" and "one or more" to introduce claims. However, the use of such phrases should not be construed as implying that a claim statement introduced by the indefinite article "a / an" limits any particular claim containing this introductory claim statement to inventions containing only that one statement, even if the same claim contains the introductory phrases "one or more" or "at least one" and indefinite articles such as "a / an" (e.g., "a / an" should generally be interpreted as meaning "at least one" or "one or more"); the same applies to the use of definite articles used to introduce claim statements. Furthermore, even if a specific number of claims is explicitly stated, those skilled in the art will recognize that this statement should generally be interpreted as meaning at least that number of statements (e.g., the basic statement of "two statements" (without other modifiers) generally means at least two statements or two or more statements). Furthermore, in examples where a conventional expression similar to "at least one of A, B, and C and similar" is used, this construction generally refers to the conventional meaning that a person skilled in the art would understand (e.g., "a system having at least one of A, B, and C" would include (but is not limited to) systems having only A, only B, only C, both A and B, both A and C, both B and C, and / or both A, B, and C, etc.). In examples where a conventional expression similar to "at least one of A, B, or C and similar" is used, this construction generally refers to the conventional meaning that a person skilled in the art would understand (e.g., "a system having at least one of A, B, or C" would include (but is not limited to) systems having only A, only B, only C, both A and B, both A and C, both B and C, and / or both A, B, and C, etc.). Those skilled in the art should further understand that any conjunction and / or phrase presenting two or more alternatives, whether in the implementation, the scope of the invention application, or the drawings, should generally be understood as a deliberate consideration of the following possibilities: including one of the alternatives, any one of the alternatives, or both of the alternatives.For example, the phrase "A or B" will generally be understood as including the possibility of "A" or "B" or "A and B".

[0147] It is believed that the invention and many of its accompanying advantages will be understood from the foregoing description, and it will be appreciated that various changes can be made to the form, construction, and configuration of the components without departing from the disclosed subject matter or sacrificing all its material advantages. The forms described are merely illustrative, and the following claims are intended to cover and include such changes. Furthermore, it should be understood that the invention is defined by the appended claims.

[0148] 100: Detection System / Particle Detection System / Optical Detection System 102: Optical Imaging Subsystem 104: Sample 106: Lighting Subsystem 108:Lighting 110: Lighting source 112: Photon Collection System 114: Sample Light 116: Detector 118: Lens 120: Optical components 122: Translation Platform 124: Controller 126: Processor 128: Memory 130: Beam Splitter 200: Method 210: Steps 212: Reference Image 214: pixel array 216: Strength 220: Steps 222: Detection Image 224: pixel array 226: Strength 230: Steps 232: Average Detection Image 234: Average pixel array 236: Average Strength 240: Steps 242: Differential Images 244: Differential pixel array 246: Difference Strength 250: Steps 302a to 302d: Images 304: Average Image 402: Chart 404: Chart 502a: First scan path 502b: Second scan path

Claims

1. A detection system comprising: A controller comprising one or more processors configured to execute program instructions, thereby causing the one or more processors to: receive at least a first portion of a first set of repeat swaths of a first scan path of a sample; align and scale the first portion of the first set of repeat swaths such that each pixel of the first portion of the first set of repeat swaths has a common number of pixels and a common pixel position; generate an image by averaging the first portion of the first set of repeat swaths; wherein averaging the first portion of the first set of repeat swaths reduces noise in the image; wherein the first portion of the first set of repeat swaths is aligned before averaging; and use the image to detect one or more defects in a detection area of ​​the sample.

2. The detection system of Request 1, wherein one of the following is true: the image is an average detection image of the detection area; or the image is a reference image.

3. The detection system of claim 2, wherein the image is the averaged detection image; wherein the program instructions cause the one or more processors to: receive a plurality of detection images; wherein the plurality of detection images are repeated images of the detection area; wherein the plurality of detection images include a first plurality of signal-to-noise ratios; wherein each of the plurality of detection images includes noise contributing to the first plurality of signal-to-noise ratios; and average the plurality of detection images to generate the averaged detection image; wherein the averaged detection image includes a second signal-to-noise ratio; wherein the second signal-to-noise ratio is greater than each of the first plurality of signal-to-noise ratios.

4. The detection system of claim 3, wherein the detection system generates the plurality of detection images from the first portion of the first set of repeated scan bands in a scanning mode.

5. The detection system of claim 2, wherein the program instructions cause one or more processors to subtract the reference image from the average detection image to generate a difference image.

6. The detection system of claim 5, wherein the program instructions cause one or more processors to detect one or more defects in the detection area based on the difference image.

7. The detection system of claim 5, wherein the program instructions cause the one or more processors to: apply a difference filter to the difference image to generate a filtered difference image; and detect one or more defects in the detection area based on the filtered difference image.

8. The detection system of claim 5, wherein the program instructions cause one or more processors to classify the one or more defects as at least one of a nuisance or a defect of concern.

9. The detection system as requested in item 1, wherein the program instructions cause one or more processors to align the first part using synchronous interpolation.

10. The detection system of claim 2, comprising one or more optical detection subsystems; wherein the one or more optical detection subsystems generate the first set of repeating scan bands of the first scan path; wherein the controller is communicatively coupled to the one or more optical detection subsystems.

11. The detection system of claim 10, comprising a stage; wherein the stage is translated such that the one or more optical detection subsystems generate the first set of repeating scan bands of the first scan path.

12. The detection system of claim 11, comprising one or more encoders or interferometers; wherein the encoders or interferometers are configured to monitor the translation of the stage; wherein the program instructions cause the one or more processors to control the translation of the stage to align with the first set of repeating scan bands of the first scan path.

13. The detection system of claim 10, wherein the one or more processors receive at least the first portion of the first set of repeating scan bands of the first scan path from the one or more optical detection subsystems.

14. The detection system of claim 10, wherein the one or more optical detection subsystems comprises two or more optical detection subsystems.

15. The detection system of claim 10, wherein the one or more optical detection subsystems include a single optical detection subsystem.

16. The detection system of claim 10, wherein the one or more optical detection subsystems include at least one first optical mode; wherein the one or more optical detection subsystems use the first optical mode to generate the first set of repeating scan bands.

17. The detection system of claim 16, wherein the first optical mode includes a first illumination wavelength, a first illumination polarization, and a first illumination angle of one or more optical detection subsystems.

18. The detection system of claim 2, wherein the image is the average detection image of the detection area; wherein the image is the reference image in a subsequent detection.

19. The detection system of claim 1, wherein the first set of repeating scan bands includes at least two repeating scan bands.

20. The detection system of claim 19, wherein the first set of repeating scan bands includes between two and four repeating scan bands.

21. The detection system of claim 1, wherein the first set of repeating scan bands includes at least four repeating scan bands.

22. The detection system of claim 1, wherein the image is one of a plurality of images of a reference image used by the processors to generate the reference image; wherein the program instructions cause the one or more processors to generate the reference image from the plurality of images; wherein the one image is generated by scan averaging.

23. The detection system of claim 22, wherein the detection area includes one or more patterned features; wherein the plurality of images are derived from a plurality of reference areas each including the one or more patterned features.

24. The detection system of claim 1, wherein the program instructions cause one or more processors to receive a second portion of a second set of repeating scan bands of a second scan path of the sample and to average the second portion of the second set of repeating scan bands; wherein the image includes at least the first portion and the second portion.

25. The detection system of claim 1, wherein the noise includes at least one of a readout noise or a shot noise.

26. A detection system comprising: One or more optical detection subsystems; A controller communicatively coupled to one or more optical inspection subsystems, the controller including one or more processors to execute program instructions that cause the one or more processors to: receive at least a first portion of a first set of repeating scan bands of a first scan path of a sample; align and scale the first portion of the first set of repeating scan bands such that each pixel of the first portion of the first set of repeating scan bands has a common number of pixels and a common pixel position; generate an image by averaging the first portion of the first set of repeating scan bands; wherein averaging the first portion of the first set of repeating scan bands reduces noise in the image; wherein the first portion of the first set of repeating scan bands is aligned before averaging; and use the image to detect one or more defects in a detection area of ​​the sample.

27. The detection system of claim 26, wherein at least one of the following is true: the image is an average detection image of the detection area; the image is a reference image; or the image is one of a plurality of images used by the processors to generate the reference image.

28. The detection system of claim 27, wherein the image is the averaged detection image; wherein the program instructions cause one or more processors to: receive a plurality of detection images; wherein the plurality of detection images are repeated images of the detection area; wherein the plurality of detection images include a first plurality of signal-to-noise ratios; wherein each of the plurality of detection images includes noise contributing to the first plurality of signal-to-noise ratios; and average the plurality of detection images to generate the averaged detection image; wherein the averaged detection image includes a second signal-to-noise ratio; wherein the second signal-to-noise ratio is greater than each of the first plurality of signal-to-noise ratios.

29. The detection system of claim 27, wherein the image is one of a plurality of images; wherein the program instructions cause one or more processors to generate the reference image from the plurality of images.

30. The detection system of claim 29, wherein the detection area includes one or more patterned features; wherein the plurality of images are derived from a plurality of reference areas each including the one or more patterned features.

31. A detection method, comprising: Receive at least a first portion of a first set of repeating scan bands in a first scan path of a sample; align and scale the first portion of the first set of repeating scan bands such that each pixel of the first portion of the first set of repeating scan bands has a common number of pixels and a common pixel position; generate an image by averaging the first portion of the first set of repeating scan bands; wherein averaging the first portion of the first set of repeating scan bands reduces noise in the image; wherein the first portion of the first set of repeating scan bands is aligned before averaging; and use the image to detect one or more defects in a detection area of ​​the sample.

Images