Multi-mode defect detection

Abstract

Methods and systems are provided for detecting defects on a sample. A method includes generating first and second modality test, reference, and difference images of a sample for first and second modalities of an inspection subsystem, respectively. The method also includes combining the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as inputs for defect detection. In addition, the method includes detecting a defect on the sample based at least on the first and second pattern difference images.

Description

Technical Field

[0001] The present invention generally relates to methods and systems for detecting defects on a sample using multiple modes of an inspection subsystem. Background Art

[0002] The following description and examples are not admitted to be prior art by inclusion in this section.

[0003] The manufacture of semiconductor devices such as logic and memory devices typically includes processing a sample (e.g., a semiconductor wafer) using a number of semiconductor manufacturing processes to form various features and multiple levels of the semiconductor device. For example, photolithography is a semiconductor manufacturing process that typically involves transferring a pattern to a photoresist disposed on a semiconductor wafer. Additional examples of semiconductor manufacturing processes include, but are not limited to, chemical mechanical polishing, etching, deposition, and ion implantation. Multiple semiconductor devices may be arranged and fabricated on a semiconductor wafer and then separated into individual semiconductor devices.

[0004] Inspection processes are used at various steps during the semiconductor manufacturing process to detect defects on the wafer to promote higher yields and therefore higher profits in the manufacturing process. Inspection has always been an important part of manufacturing semiconductor devices. However, as the size of semiconductor devices decreases, inspection becomes more important to the successful manufacture of acceptable semiconductor devices because smaller defects may cause device failure.

[0005] Many inspection tools have adjustable parameters for many of the output (e.g., image) generating elements of the tool. The parameters of one or more elements (e.g., energy sources, polarizers, lenses, detectors, and the like) may be altered depending on the type of sample being inspected and the characteristics of the defect of interest (DOI) on the sample. For example, different types of samples may have significantly different characteristics, which may cause the same tool with the same parameters to image the sample in significantly different ways. Additionally, because different types of DOIs may have significantly different characteristics, inspection system parameters that are suitable for detecting one type of DOI may not be suitable for detecting another type of DOI. Furthermore, different types of samples may have different noise sources, which may interfere with the detection of DOIs on the sample in different ways.

[0006] The development of inspection tools with adjustable parameters has also led to the increased use of inspection processes that involve scanning a sample using more than one combination of parameter values ​​(otherwise referred to as "modes") for a number of reasons, such as detecting more defects, detecting different defect types using different modes, disturber filtering, etc. For example, multi-mode defect detection utilizes scans from multiple modes to improve detection performance. The information obtained from two or more modes provides independent information to allow the detection algorithm to strengthen weakly detected defects and / or suppress or separate undesirable disturber anomalies.

[0007] Existing multi-mode methods may use a single mode scan to detect defects on a sample and then revisit and collect information about the resulting defects using a different mode. These methods may then apply rules to the attributes from the two scans to determine whether to keep or filter out the defect. Another less complex method uses the results of two independent inspection scans performed using different modes to apply rules to the attributes of the common defect to filter or keep it. Such methods may generally be referred to as "result fusion."

[0008] Such currently used multi-mode inspection methods and systems have several important disadvantages. For example, such methods typically use essentially thermal scans that detect a very large number of defects in a primary mode, followed by a secondary mode where additional defect attributes are collected at the defect locations detected in the primary mode. Such methods may also utilize rule-based filters on defect attributes rather than the image itself. This rule-based filtering may result in subpar performance when one of the defect attributes of the mode performs poorly. Such methods may also have poor detectability because the primary optical mode can detect a limited number of target defects. In the case of result fusion, uncommon defects are not processed, resulting in no multi-mode benefit for these defects. Again, when multiple modes are actually considered, the process should be cost-effective in terms of processing power, storage, and data I / O rates.

[0009] Therefore, it would be advantageous to develop systems and methods for detecting defects on a sample without one or more of the disadvantages described above. Summary of the invention

[0010] The following description of various embodiments should in no way be construed as limiting the subject matter of the appended claims.

[0011] One embodiment relates to a system for detecting defects on a sample. The system includes an inspection subsystem configured to generate first and second mode test images of the sample using first and second modes of the inspection subsystem, respectively. The system also includes a computer system configured to generate first and second mode reference images from the first and second mode test images, respectively. The computer system is also configured to subtract the first and second mode reference images from the first and second mode test images, respectively, to thereby generate first and second mode difference images, respectively. In addition, the computer system is configured to combine the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as inputs for defect detection. The system is further configured to detect defects on the sample based on at least the first and second mode difference images in the inputs. The system may be further configured as described herein.

[0012] Another embodiment relates to a method for detecting defects on a sample. The method includes acquiring first and second mode test images of the sample generated using first and second modes of an inspection subsystem, respectively. The method also includes the steps of generating first and second mode reference images, subtracting, combining, and detecting described above, which are performed by a computer system. Each step of the method described above may be further performed as further described herein. In addition, the method described above may include any other steps of any other method described herein. In addition, the method described above may be performed by any system described herein.

[0013] Another embodiment relates to a non-transitory computer-readable medium storing program instructions that can be executed on a computer system to perform a computer-implemented method for detecting defects on a sample. The computer-implemented method includes the steps of the method described above. The computer-readable medium can be further configured as described herein. The steps of the computer-implemented method can be performed as further described herein. In addition, the computer-implemented method for which the program instructions can be executed can include any other steps of any other method described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Further advantages of the present invention will become apparent to those skilled in the art with the benefit of the following detailed description of the preferred embodiments and after reference to the accompanying drawings, in which:

[0015] Figure 1 and 2 is a schematic diagram illustrating a side view of an embodiment of a system configured as described herein;

[0016] Figure 3 and 4 is a flow chart illustrating an embodiment of steps that may be performed to detect defects on a sample; and

[0017] Figure 5 is a block diagram illustrating one embodiment of a non-transitory computer-readable medium storing program instructions for causing a computer system to perform the computer-implemented methods described herein.

[0018] Although the present invention is susceptible to various modifications and alternative forms, specific embodiments of the present invention are shown by way of example in the drawings and described in detail herein. The drawings may not be drawn to scale. However, it should be understood that the drawings and detailed description thereof are not intended to limit the present invention to the specific forms disclosed, but on the contrary, are intended to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims. DETAILED DESCRIPTION

[0019] As used herein, the term "disturbing point" (which may sometimes be used interchangeably with "disturbing point defect") is generally defined as a defect that the user does not care about and / or an event that is detected on a sample but is not actually an actual defect on the sample. Disturbing points that are not actual defects may be detected as events due to non-defect noise sources on the sample (e.g., die in a metal line on the sample, signals from underlying layers or materials on the sample, line edge roughness (LER), relatively small critical dimension (CD) variations in patterning properties, thickness variations, etc.) and / or due to marginalities in the inspection system itself or its configuration for inspection.

[0020] As used herein, the term "defect of interest (DOI)" is defined as a defect that is detected on a sample and is actually an actual defect on the sample. Therefore, DOI is of interest to users because users are generally concerned about the number and type of actual defects on the inspected sample. In some contexts, the term "DOI" is used to refer to a subset of all actual defects on a sample that only includes actual defects that the user is concerned about. For example, there may be multiple types of DOI on any given sample, and the user may be more concerned about one or more of them than one or more other types. However, in the context of the embodiments described herein, the term "DOI" is used to refer to any and all real defects on a sample.

[0021] The terms "design", "design data", and "design information" as used interchangeably herein generally refer to the physical design (layout) of an IC or other semiconductor device and data derived from the physical design by complex simulations or simple geometric and Boolean operations. The design may include any other design data or design data proxy described in co-owned U.S. Pat. No. 7,570,796 issued to Zafar et al. on Aug. 4, 2009 and co-owned U.S. Pat. No. 7,676,077 issued to Kulkarni et al. on Mar. 9, 2010, both of which are incorporated herein by reference as if fully set forth. In addition, the design data may be standard cell library data, integrated layout data, design data for one or more layers, derivatives of the design data, and full or partial chip design data. Furthermore, the "design", "design data", and "design information" described herein refer to information and data generated by a semiconductor device designer during the design process and thus can be used in the embodiments described herein well before the design is printed on any physical samples (e.g., reticles and wafers).

[0022] The terms “first” and “second” are used herein to merely indicate two things different from each other and are not used to indicate any temporal, spatial, preference or other characteristics of the components referred to herein as “first” and “second”.

[0023] Referring now to the drawings, it should be noted that the drawings are not drawn to scale. In particular, the proportions of some elements of the drawings are greatly exaggerated to emphasize the characteristics of the elements. It should also be noted that the drawings are not drawn to the same scale. The same reference numerals have been used to indicate similarly configurable elements shown in more than one figure. Unless otherwise mentioned herein, any element described and shown may include any suitable commercially available element.

[0024] In general, the embodiments described herein are configured for detecting defects on a sample using images generated by multiple modes of an inspection subsystem. In this way, the embodiments are configured for multi-mode inspection of a sample. The embodiments described herein are particularly advantageous for multi-mode inspection applications such as optical wafer inspection but may be configured for any other type of inspection described herein.

[0025] As the DOI size becomes smaller, the DOI signal becomes more sensitive to the inspection tool mode. A certain DOI signal is not significant enough to be detected using one mode but can be strongly detected using another mode. Generally speaking, different inspection tool modes produce different disturbing point signatures. If multi-mode information is properly utilized, more DOIs can be detected while suppressing disturbing points.

[0026] Existing multi-mode detection methods have several weaknesses, including those described further above. In addition, one potential weakness of most existing multi-mode detection methods is that they are based on the following assumptions: 1) all DOIs are detected using a primary mode with a large number of disturbing points, and 2) there is a second mode that can be used to determine some properties that can be used to maintain all or most DOIs but significantly suppress disturbing points. The problem is that in many cases, finding two modes that satisfy the assumptions can be substantially challenging.

[0027] DOI and disturbing points may be common or unique to the inspection tool mode. In order to suppress disturbing points when detecting DOI, it is necessary to effectively utilize dual-mode information at the detection level. The problem then is how to effectively prepare and use multi-mode data sets. The embodiments described herein address this problem in at least two ways. One way is to minimize the processing power or system cost impact by reducing system overhead when collecting multi-mode data. The second way is to directly utilize multi-mode data at the time of detection.

[0028] In some embodiments, the sample is a wafer. The wafer may include any wafer known in semiconductor technology. Although some embodiments are described herein with respect to a wafer or a plurality of wafers, the embodiments are not limited to using samples thereof. For example, the embodiments described herein may be used for samples such as reticles, flat panels, personal computer (PC) boards, and other semiconductor samples.

[0029] One embodiment relates to a system for detecting defects on a sample. Figure 1One embodiment of such a system is shown in . The system includes an inspection subsystem 100 configured for generating first and second mode test images of a sample using first and second modes of the inspection subsystem, respectively. The inspection subsystem is coupled to one or more computer systems 102. However, in general, the system includes a computer system, which may include any computer subsystem or any configuration of systems described herein.

[0030] In some embodiments, the inspection subsystem is configured to generate first and second pattern test images using light. Figure 1 In the embodiment shown in , the inspection subsystem is configured as a light-based inspection subsystem. However, in other embodiments described herein, the inspection subsystem is configured as an electron beam or charged particle beam inspection subsystem. In this way, in other embodiments, the inspection subsystem is configured to generate the first and second mode test images using electrons.

[0031] In general, the inspection subsystem described herein includes at least one energy source, a detector, and a scanning subsystem. The energy source is configured to generate energy directed to the sample by the inspection subsystem. The detector is configured to detect energy from the sample and generate an output in response to the detected energy. The scanning subsystem is configured to change the location on the sample to which energy is directed and from which energy is detected.

[0032] In the light-based inspection subsystems described herein, the energy directed to the sample comprises light, and the energy detected from the sample comprises light. Figure 1 In an embodiment of the system shown in FIG. 1 , the inspection subsystem includes an illumination subsystem configured to direct light to the sample 14. The illumination subsystem includes at least one light source. For example, Figure 1 , the illumination subsystem includes a light source 16. The illumination subsystem is configured to direct light to the sample at one or more incident angles, which may include one or more oblique angles and / or one or more normal angles. Figure 1 , light from light source 16 is directed at an oblique angle of incidence through optical element 18 and then through lens 20 to sample 14. The oblique angle of incidence may include any suitable oblique angle of incidence, which may vary depending on, for example, the characteristics of the sample and the process to be performed on the sample.

[0033] The illumination subsystem may be configured to direct light to the sample at different incident angles at different times. For example, the inspection subsystem may be configured to modify one or more characteristics of one or more elements of the illumination subsystem so that light can be directed to the sample at different incident angles. Figure 1 In one such example, the inspection subsystem may be configured to move the light source 16, optical element 18, and lens 20 so that light is directed to the sample at different oblique angles of incidence or normal (or near-normal) angles of incidence.

[0034] In some examples, the inspection subsystem may be configured to direct light to the sample at more than one incident angle at the same time. For example, the illumination subsystem may include more than one illumination channel, one of which may include, for example, Figure 1 1, and another of the illumination channels (not shown) may include similar elements that may be configured differently or identically or may include at least one light source and possibly one or more other elements (such as those described further herein). If this light is directed to the sample at the same time as the other light, one or more characteristics (e.g., wavelength, polarization, etc.) of the light directed to the sample at different angles of incidence may be different, so that light originating from illuminating the sample at different angles of incidence can be distinguished from one another at the detector.

[0035] In another example, the illumination subsystem may include only one light source (e.g., Figure 1 ) and the light from the light source may be separated into different optical paths by one or more optical elements (not shown) of the illumination subsystem (e.g., based on wavelength, polarization, etc.). The light in each of the different optical paths may then be directed to the sample. Multiple illumination channels may be configured to direct light to the sample at the same time or at different times (e.g., when different illumination channels are used to sequentially illuminate the sample). In another example, the same illumination channel may be configured to direct light with different characteristics to the sample at different times. For example, in some examples, the optical element 18 may be configured as a spectral filter and the properties of the spectral filter may be changed in a variety of different ways (e.g., by swapping out one spectral filter with another spectral filter) so that light of different wavelengths may be directed to the sample at different times. The illumination subsystem may have any other suitable configuration known in the art for directing light with different or the same characteristics to the sample sequentially or simultaneously at different or the same angles of incidence.

[0036] The light source 16 may include a broadband plasma (BBP) light source. In this manner, the light generated by the light source and directed to the sample may include broadband light. However, the light source may include any other suitable light source, such as any suitable laser known in the art, and may be configured to generate light of any suitable wavelength known in the art. Additionally, the laser may be configured to generate monochromatic or near-monochromatic light. In this manner, the laser may be a narrowband laser. The light source may also include a polychromatic light source that generates light of multiple discrete wavelengths or bands.

[0037] Light from the optical element 18 can be focused onto the sample 14 by the lens 20. Figure 1 Lens 20 is shown in FIG. 2 as a single refractive optical element, but in reality, lens 20 may include several refractive and / or reflective optical elements that combine to focus light from the optical element onto the sample. Figure 1The illumination subsystem shown in and described herein may include any other suitable optical elements (not shown). Examples of such optical elements include, but are not limited to, polarizing components, spectral filters, spatial filters, reflective optical elements, apodizers, beam splitters, apertures, and the like, which may include any such suitable optical elements known in the art. Additionally, the system may be configured to alter one or more elements of the illumination subsystem based on the type of illumination used for inspection.

[0038] The inspection subsystem may also include a scanning subsystem configured to change the location on the sample to which light is directed and from which light is detected, and possibly cause the light to scan over the sample. For example, the inspection subsystem may include a stage 22 on which the sample 14 is placed during inspection. The scanning subsystem may include any suitable mechanical and / or robotic assembly (including the stage 22) that may be configured to move the sample so that light can be directed to and detected from different locations on the sample. Additionally or alternatively, the inspection subsystem may be configured so that one or more optical elements of the inspection subsystem perform a certain scan of light over the sample so that light can be directed to and detected from different locations on the sample. In examples where light is scanned over the sample, the light may be caused to scan over the sample in any suitable manner, such as in a serpentine path or in a spiral path.

[0039] The inspection subsystem further includes one or more detection channels. At least one of the detection channels includes a detector configured to detect light from the sample as a result of illumination of the sample by the system and to generate an output in response to the detected light. For example, Figure 1 The inspection subsystem shown in FIG. 1 includes two detection channels, one detection channel formed by a light collector 24, an element 26 and a detector 28 and the other detection channel formed by a light collector 30, an element 32 and a detector 34. Figure 1 , two detection channels are configured to collect and detect light at different collection angles. In some examples, the two detection channels are configured to detect scattered light, and the detection channels are configured to detect light scattered from the sample at different angles. However, one or more detection channels may be configured to detect another type of light (e.g., reflected light) from the sample.

[0040] like Figure 1As further shown in FIG, two detection channels are shown as being positioned in the plane of the paper and the illumination subsystem is also shown as being positioned in the plane of the paper. Thus, in this embodiment, the two detection channels are positioned (e.g., centered) in the plane of incidence. However, one or more detection channels may be positioned outside the plane of incidence. For example, the detection channel formed by the light collector 30, the element 32, and the detector 34 may be configured to collect and detect light scattered from the plane of incidence. Thus, this detection channel may generally be referred to as a "side" channel, and this side channel may be centered in a plane substantially perpendicular to the plane of incidence.

[0041] Although Figure 1 An embodiment of an inspection subsystem including two detection channels is shown, but the inspection subsystem may include a different number of detection channels (e.g., only one detection channel or two or more detection channels). In one such example, the detection channel formed by the light collector 30, the element 32, and the detector 34 may form a side channel as described above, and the inspection subsystem may include an additional detection channel (not shown) formed as another side channel positioned on the opposite side of the incident plane. Thus, the inspection subsystem may include a detection channel that includes the light collector 24, the element 26, and the detector 28 and is centered in the incident plane and is configured to collect and detect light at a scattering angle normal to or close to normal to the sample surface. Therefore, this detection channel may generally be referred to as a "top" channel, and the inspection subsystem may also include two or more side channels configured as described above. Thus, the inspection subsystem may include at least three channels (i.e., one top channel and two side channels), and each of the at least three channels has its own light collector, each light collector being configured to collect light at a scattering angle different from each other light collector.

[0042] As further described above, each detection channel included in the inspection subsystem can be configured to detect scattered light. Figure 1 The inspection subsystem shown in can be configured for dark field (DF) inspection of the sample. However, the inspection subsystem may also or alternatively include a detection channel configured for bright field (BF) inspection of the sample. In other words, the inspection subsystem may include at least one detection channel configured to detect light specularly reflected from the sample. Thus, the inspection subsystem described herein may be configured for only DF inspection, only BF inspection, or both DF inspection and BF inspection. Although in Figure 1 Each light collector is shown as a single refractive optical element in , but each light collector may include one or more refractive optical elements and / or one or more reflective optical elements.

[0043] One or more detection channels may include any suitable detector known in the art, such as a photomultiplier tube (PMT), a charge coupled device (CCD), and a time-delay integration (TDI) camera. The detector may also include a non-imaging detector or an imaging detector. If the detector is a non-imaging detector, each detector may be configured to detect certain characteristics of the scattered light (e.g., intensity) but may not be configured to detect such characteristics that vary depending on the position within the imaging plane. Thus, the output generated by each detector in each detection channel included in the inspection subsystem may be a signal or data rather than an image signal or image data. In such examples, a computer system (e.g., computer subsystem 36 of the inspection subsystem) may be configured to generate an image of the sample from the non-imaging output of the detector. However, in other examples, the detector may be configured as an imaging detector configured to generate an image signal or image data. Thus, the inspection subsystem may be configured to generate an image in a number of ways.

[0044] Provided in this article Figure 1 to generally illustrate configurations of inspection subsystems that may be included in embodiments of the systems described herein. Obviously, the inspection subsystem configurations described herein may be altered to optimize the performance of the inspection subsystem as is typically performed when designing commercial inspection systems. Additionally, the systems described herein may be implemented using existing systems such as the 29xx / 39xx series of tools commercially available from KLA (e.g., by adding the functionality described herein to the existing inspection system). For some such systems, the methods described herein may be provided as optional functionality of the system (e.g., in addition to other functionality of the system). Alternatively, the systems described herein may be designed "from the ground up" to provide an entirely new system.

[0045] The computer subsystem 36 may be coupled to the detector of the inspection subsystem in any suitable manner (e.g., via one or more transmission media, which may include "wired" and / or "wireless" transmission media) so that the computer subsystem can receive the output generated by the detector. The computer subsystem 36 may be configured to use the output of the detector to perform a number of functions, including the steps further described herein.

[0046] The computer subsystem of the inspection subsystem may be further configured as described herein. For example, the computer subsystem 36 may be part of or may be configured as a computer system described herein. In particular, the computer subsystem 36 may be configured to perform the steps described herein. Thus, the steps described herein may be performed "on the tool" by a computer system or subsystem that is part of the inspection subsystem.

[0047] The computer subsystem of the inspection subsystem (as well as other computer subsystems described herein) may also be referred to herein as a computer system. Each of the computer subsystems or systems described herein may take various forms, including a personal computer system, an image computer, a mainframe computer system, a workstation, a network appliance, an Internet appliance, or other devices. In general, the term "computer system" may be broadly defined to encompass any device having one or more processors that execute instructions from a memory medium. The computer subsystem or system may also include any suitable processor known in the art (e.g., a parallel processor). In addition, the computer subsystem or the system may include a computer platform (as a stand-alone tool or a network-linked tool) with high-speed processing and software.

[0048] If the system includes more than one computer subsystem, the different computer subsystems may be coupled to each other so that images, data, information, instructions, etc. may be sent between the computer subsystems. For example, the computer subsystem 36 may be coupled to the computer subsystem 36 via any suitable transmission media, such as a wireless transmission media, which may include any suitable wired and / or wireless transmission media known in the art. Figure 1 The dashed line in is shown as being coupled to computer system 102. Two or more of these computer subsystems may also be effectively coupled via a shared computer-readable storage medium (not shown).

[0049] In an electron beam inspection subsystem, the energy directed to the sample comprises electrons, and the energy detected from the sample comprises electrons. Figure 2 In one such embodiment shown in FIG. 1 , the inspection subsystem includes an electron column 122, and the system includes a computer subsystem 124 coupled to the inspection subsystem. The computer subsystem 124 may be configured as described above. In addition, the inspection subsystem may be configured as described above and in Figure 1 can be coupled to another or more computer systems in the same manner as shown in FIG.

[0050] Likewise Figure 2 , the electron column includes an electron beam source 126 configured to produce electrons that are focused by one or more elements 130 to a sample 128. The electron beam source may include, for example, a cathode source or emitter tip, and the one or more elements 130 may include, for example, a gun lens, an anode, a beam limiting aperture, a gate valve, a beam current selection aperture, an objective lens, and a scanning subsystem, all of which may include any such suitable elements known in the art.

[0051] Electrons returning from the sample (eg, secondary electrons) may be focused by one or more elements 132 onto a detector 134. One or more elements 132 may include, for example, a scanning subsystem, which may be the same scanning subsystem included in element 130.

[0052] The electron column may include any other suitable elements known in the art. In addition, the electron column may be further configured as described in U.S. Pat. No. 8,664,594 issued to Jiang et al. on April 4, 2014, U.S. Pat. No. 8,692,204 issued to Kojima et al. on April 8, 2014, U.S. Pat. No. 8,698,093 issued to Gubbens et al. on April 15, 2014, and U.S. Pat. No. 8,716,662 issued to MacDonald et al. on May 6, 2014, which are incorporated herein by reference as if fully set forth.

[0053] Although in Figure 2 , the electron beam is shown as being configured so that electrons are directed to the sample at an oblique angle of incidence and scattered from the sample at another oblique angle, but the electron beam may be directed to the sample and scattered from the sample at any suitable angle. Additionally, the electron beam inspection subsystem may be configured to generate an image of the sample using multiple modes (e.g., using different illumination angles, collection angles, etc.), as further described herein. The multiple modes of the electron beam inspection subsystem may differ in any output generation parameter of the inspection subsystem.

[0054] The computer subsystem 124 may be coupled to a detector 134, as described above. The detector may detect electrons returning from the surface of the sample, thereby forming an electron beam image of the sample (or other output for the sample). The electron beam image may include any suitable electron beam image. The computer subsystem 124 may be configured to use the output generated by the detector 134 to perform the steps further described herein. The computer subsystem 124 may be further configured as described herein to include Figure 2 The system of inspection subsystems shown in .

[0055] Provided in this article Figure 2 to generally illustrate the configuration of an electron beam inspection subsystem that may be included in the embodiments described herein. As with the optical inspection subsystem described above, the electron beam inspection subsystem may be modified to optimize the performance of the inspection subsystem as is typically performed when designing a commercial system. In addition, the systems described herein may be implemented using existing systems such as tools commercially available from KLA (e.g., by adding the functionality described herein to the existing system). For some such systems, the methods described herein may be provided as optional functionality of the system (e.g., in addition to other functionality of the system). Alternatively, the systems described herein may be designed "from the ground up" to provide an entirely new system.

[0056] Although the inspection subsystem is described above as a light or electron beam inspection subsystem, the inspection subsystem may be an ion beam inspection subsystem. Figure 2This inspection subsystem is configured as shown in FIG. 1 , except that any suitable ion beam source known in the art may be used in place of the electron beam source. Additionally, the inspection subsystem may include any other suitable ion beam imaging system, such as those included in commercially available focused ion beam (FIB) systems, helium ion microscope (HIM) systems, and secondary ion mass spectrometer (SIMS) systems.

[0057] As further mentioned above, the inspection subsystem is configured to have multiple modes. In general, a "mode" can be defined by the value of a parameter of the inspection subsystem used to generate an output (e.g., an image) of the sample. Thus, the modes can differ in the value of at least one of the optical or electron beam parameters of the inspection subsystem (in addition to the location on the sample where the output is generated). For example, for a light-based inspection subsystem, different modes can use different wavelengths of light and / or different focus offsets, polarizers, apertures, optical pixels, stage speeds, etc. The modes can differ in the wavelength of light directed to the sample (e.g., by using different light sources, different spectral filters, etc. for different modes), as further described herein. In another embodiment, different modes can use different illumination channels. For example, as mentioned above, the inspection subsystem can include more than one illumination channel. Thus, different illumination channels can be used for different modes.

[0058] The multiple modes may also differ in illumination and / or collection / detection. For example, as further described above, the inspection subsystem may include multiple detectors. One detector may be used for one mode and another detector may be used for another mode. The modes may also differ from each other in more than one of the ways described herein (e.g., different modes may have one or more different illumination parameters and one or more different detection parameters). For example, depending on the ability to scan a sample simultaneously using multiple modes, the inspection subsystem may be configured to scan a sample using different modes in the same scan or in different scans.

[0059] As mentioned above, the inspection subsystem may be configured to direct energy (e.g., light, electrons) to and / or scan energy over a physical version of the sample, thereby generating an actual image of the physical version of the sample. In this manner, the inspection subsystem may be configured as a "real" inspection system rather than a "virtual" system. However, Figure 1The storage medium (not shown) and computer subsystem 102 shown in the can be configured as a "virtual" system. In particular, the storage medium and computer subsystem are not part of the inspection subsystem 100 and do not have any ability to handle a physical version of the sample but can be configured as a virtual inspector that performs inspection-like functions using the stored detector output. Systems and methods configured as a "virtual" system are described in the following commonly assigned patents: U.S. Patent No. 8,126,255 issued to Bhaskar et al. on February 28, 2012; U.S. Patent No. 9,222,895 issued to Duffy et al. on December 29, 2015; and U.S. Patent No. 9,816,939 issued to Duffy et al. on November 14, 2017, which are incorporated herein by reference as if fully set forth. The embodiments described herein can be further configured as described in these patents. For example, the computer system described herein can be further configured as described in these patents.

[0060] As further described above, the inspection subsystem generates first and second mode test images of the sample using the first and second modes of the inspection subsystem, respectively. The inspection subsystem may be an actual inspection subsystem that generates the first and second mode test images by scanning a physical version of the sample using at least the first and second modes (M1, M2, ..., MN). The inspection subsystem may also or alternatively be configured as a virtual inspection subsystem that plays a stored image of a physical version of the sample generated and stored as described above. In either case, the scanning may be performed as further described herein.

[0061] Although embodiments may be described herein with respect to first and second modes, embodiments may be used with more than two modes. Figure 3 , the inspection subsystem may perform a Mode 1 scan 200 and a Mode 2 scan 318. The Mode 1 scan may include scanning one or more scan bands across the sample. Figure 3 As shown in FIG. 1 , a mode 1 scan 300 may include a cross-sample ( Figure 3 306 on the sample). In a similar manner, the Mode 2 scan may include scanning corresponding scan bands across the rows of dies 302, 304, and 306 on the sample. The steps described herein may be performed separately and independently for each of the different scan bands (or each set of corresponding scan bands) to reduce the complexity of the embodiment, and the defect detection results generated for any scan band may be combined into a single inspection result for the sample. Scanning any scan band using the first mode may generate an M1 frame 308, and scanning any scan band using the second mode may generate an M2 frame 326.

[0062] Generally speaking, as used in this article, T ij=Test image for pattern j and job j. A job is a collection of frames that can be processed together for the functions described herein (e.g., defect detection). Therefore, the configuration of jobs and frames can be determined by the configuration of various hardware elements described herein. Jobs can also be design-location specific, meaning that the frames in any job are all generated at locations within the design where the same patterned features with the same patterned feature characteristics are desired to be formed. In other words, different jobs can be defined for different specific design locations. However, depending on the number of frames generated for any specific design location (e.g., when not all frames can be processed together due to size), there can be multiple jobs for a single design location. For example, depending on how the pattern repeats within a die, the frames in any job can also be generated within a single die or multiple dies. For example, there can be multiple locations within a die with the same patterned features with the same patterned feature characteristics, and the image frames generated at these locations can be grouped in one job. Image frames generated at corresponding locations in one or more other dies can also be grouped in the same job.

[0063] Test images generated using the same pattern can be aligned to each other, for example, so that test images generated at the same design location on the sample can be jointly identified and used for inspection. The computer system can be configured to align test images generated using a single pattern to each other by individually aligning the test images to the design of the sample or another common reference (e.g., an alignment target or mark). This image alignment can be a pixel-to-design (PDA)-based alignment. PDA is a way to align images to a design in design coordinates. After this alignment, images sharing common design coordinates can be jointly identified and processed. In this way, intra-pattern image alignment can utilize an alignment-to-design algorithm (e.g., an alignment-to-design algorithm available from KLA and described in the patent of Kulkarni cited above) to align images generated using a single pattern to a design coordinate system. Aligning the test image to the design may also be performed as described in U.S. Pat. No. 9,830,421, issued to Bhattacharyya et al. on Nov. 27, 2017, and U.S. Pat. No. 10,698,325, issued to Brauer et al. on Jun. 30, 2020, both of which are incorporated herein by reference as if fully set forth. The embodiments described herein may be further configured as described in these patents. This step may also be performed as further described herein.

[0064] All frame images in one job should have exactly the same pattern, but they may have other variations, such as color variations, that may be problematic for defect detection. The computer system may perform a similarity metric job (SMJ) to calculate similarity values ​​and determine the job containing the most similar image frames. In another embodiment, the computer system is configured to calculate first and second similarity metrics for first and second mode test images, respectively, and separate the first and second mode test images into first and second detection jobs based on the first and second similarity metrics, respectively, and the generation of first and second mode reference images, subtraction, combination, and detection further described herein are performed separately for each of two or more of the first and second detection jobs.

[0065] In one such example, a scan (e.g., a partial MCATx scan) may be performed, and SMJ may be performed to determine the best reference image for each target image. The term "partial MCATx scan" as used herein is defined as a path scanned over a sample using multiple modes. In this way, the scan is "partial" in that it does not cover the entire sample being inspected. The purpose of this partial scan is to minimize system overhead such as mode switching and maximize processing power by performing fast calculations using a relatively minimal data set. In general, SMJ is a job that calculates similarities between images and determines a group of the most similar images for a detection job. The detection job then processes together the images that have been determined to be most similar to each other.

[0066] SMJ can be performed separately for different modes in the same scan band. For example, SMJ can be used to determine the best reference image set that is most similar to each other. Since the first and second modes can be completely different modes that produce substantially different images, different SMJs are performed separately for different modes. For example, if Figure 3 As shown in FIG. 4 , the computer system may perform SMJ 310 for M1 frame 308 and perform SMJ 328 for M2 frame 326 .

[0067] In this way, for one swath of data generated for a sample, the images in the swath can be grouped into more than one job based on image similarity, i.e., images in job 1 are most similar to each other, images in job 2 are most similar to each other, etc. Thus, images in any job can be mutually exclusive with images in any other job, meaning that any image can be included in only one job. Then, the steps described herein can be performed separately and independently for each job.

[0068] The computer system is configured to generate first and second mode reference images from first and second mode test images, respectively, and to subtract the first and second mode reference images from the first and second mode test images, respectively, to thereby generate first and second mode difference images, respectively. ij= reference image for mode i and job j. For each job of images generated by SMJ, a reference image may be generated from the images in that job. The reference image may be used only to generate a difference image for the test images in the same job. The difference image may be used for defect detection as described herein. The defect detection results generated for each job may be combined into one single defect detection result for the scanned strip.

[0069] In one such example, Figure 3 In the example, SMJ 310 can be identified as M1(T 1j ) of job j outputted from test images 314 of the first mode. These test images may include all test images that are most similar to each other generated using the first mode. By separating the mode 1 images based on similarity, the SMJ step may also separate the mode 1 images based on patterned features formed at the location where the images are generated. However, some substantially different patterned features may be similarly presented in the test images (e.g., depending on the imaging capabilities of the inspection subsystem). Therefore, in some examples, SMJ may be performed only on test images generated at substantially the same patterned features in the design of the sample. For example, prior to SMJ, the computer system may separate the test images based on the design to thereby generate different groups of test images corresponding to different patterned features in the design, respectively. SMJ may be performed separately for each of these groups of test images. In any case, SMJ will separate the test images based on other changes in the test images, so that test images with similar noise (and possibly other) characteristics can be processed together and independently of other test images.

[0070] M1(T 1j ) may include all Model 1 test images for job j identified by SMJ, and one or more of these test images may be used to generate M1(R 1j ). Reference image 312 may be generated from test image 314 in any suitable manner. For example, singular value decomposition (SVD) is one method that may be used to generate a reference image for the embodiments described herein. Another method that may be used to generate a reference image is a linear combination of test images that may be performed, for example, by an MCAT algorithm that is used by some inspection tools commercially available from KLA. Next, M1(T) may be generated by subtracting reference image 312 from test image 314. 1j- R 1j ) of the image 316. This subtraction can be performed in any suitable manner known in the art.

[0071] In a similar manner, M2(T 2j) may include all test images for job j identified by SMJ 328, and one or more of these test images may be used to generate M2(R 2j ) of reference image 330. Reference image 330 may be generated from test image 332 in any suitable manner, such as SVD or linear combination. In addition, the methods used by the computer system to generate reference images 312 and 330 may be the same or different. M2(T) may be generated by subtracting reference image 330 from test image 332 2j -R 2j ) of the reference image 334. This subtraction can be performed in any suitable manner known in the art. Generating reference images and difference images as described above can also be performed for any other operations that can be predetermined in the inspection recipe and / or determined by SMJ.

[0072] In one such embodiment, generating first and second mode reference images performed for at least one of the first and second inspection operations includes selecting one or more of the first and second mode test images used to generate the first and second mode reference images, respectively, independently of the location on the sample at which the first and second test mode images are generated. In other words, the images used for reference image generation can be selected independently of their spatial relationship to the test images. For example, images from anywhere in the scan band can be selected for reference image generation, rather than using only images adjacent to the test images as reference images. By relaxing the spatial constraints on images that can be used as reference images, the embodiments described herein can generate better reference images (i.e., reference images with fewer non-defect differences from the test images) that can achieve more sensitive inspections (e.g., detecting more DOIs and / or smaller defects). In other words, the embodiments described herein can generate better references and thereby better sensitivity because the references are generated from images selected based on similarity-based metrics rather than proximity-based images.

[0073] The multimodal detection embodiments described herein may differ in this respect from currently used multimodal detection systems and methods. For example, the primary detection steps of the embodiments described herein may utilize advanced methods for computing optimal reference images. Some examples of such methods are described in U.S. Patent Application Publication No. 2021 / 0090229 published by Brauer et al. on March 25, 2021, which is incorporated herein by reference as if fully set forth. The embodiments described herein may be further configured as described in such a publication. When a difference image is used for processing, the MCATx detection example described herein may or may not use the same optimal reference image.

[0074] The computer system is further configured to combine the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as inputs for defect detection and detect defects on the sample based at least on the first and second mode difference images in the inputs. For example, images generated using the first and second modes (original T, R, and difference image TR) may be combined to perform final defect detection at locations scanned using the first and second modes. Figure 3 As shown in FIG. 3 , for example, the computer subsystem may combine images in step 336 for detecting (T 1j , R 1j , T 1j -R 1j , T 2j , R 2j , T 2j -R 2j ). The defect detection step 354 may be performed based at least on the difference image in the combined image. The key is to utilize data from both modes, a) each reference / test image set, b) each difference image set, or both a) and b) together. The images used for defect detection and the defect detection method used may vary depending on the actual input data. The defect detection method may also vary depending on the defect type, for example, different defect types may be detected more or less using different defect detection methods as further described herein.

[0075] Test images generated using different modes can be aligned with each other so that test images generated at the same design location on the sample can be used together for inspection. In one embodiment, the computer system is configured to align the first and second mode test images with each other by individually aligning the first and second mode test images to the design of the sample. Figure 3 , the computer system may perform alignment in step 338. Specifically, alignment step 338 may include performing alignment 344 of M1 image 340 to M2 image 342 such that image pixel (m, n) 346 in the first mode image is aligned to corresponding image pixel (m, n) 348 in the second mode image. In this way, the image T at (m, n) 1j , R 1j 350 can be aligned to the image T at (m, n) in the second mode image 2j , R 2j 352. Thus, when image datasets from multiple modalities are fully aligned, e.g. Figure 3 As shown in the figure, at each image pixel (m,n), the image information (T 1j , R 1j , T 2j , R 2j ) and for example T 1j -R 1j and T2j -R 2j Any or all of this information can be used directly in the detection step.

[0076] When processing two different mode images simultaneously or together, inter-mode alignment can be a challenge. The inter-mode image alignment performed by the embodiments described herein can be a PDA-based alignment, which can be performed as further described herein. Although the images from the two modes are different, once each image is aligned to the design by the PDA, the images share common design coordinates. In this way, inter-mode image alignment can utilize an alignment to design algorithm (such as the alignment to design algorithm described above) to align images of multiple modes to the design coordinate system. Aligning the first and second mode images to the design can also be performed as described in the following patents: U.S. Patent No. 9,830,421 issued to Bhattacharyya et al. on November 27, 2017 and U.S. Patent No. 11,580,650 issued to Brauer et al. on February 14, 2023, which are incorporated herein by reference as if set forth in full. The embodiments described herein can be further configured as described in these patents.

[0077] As described above and in Figure 3 As shown in FIG. 1 , steps can be performed for job j and multiple modes. The same steps can be performed for other jobs (e.g., jobs 1, ..., j) determined in the inspection recipe or by SMJ. In this way, steps can be performed separately for each job. Figure 4 It further shows how steps can be performed for multiple operations. The steps of each operation shown in this figure can be performed sequentially or simultaneously. Figure 4 The steps are shown being performed for two jobs, but the same steps may be performed sequentially or simultaneously for any number of jobs.

[0078] like Figure 4 , the inspection subsystem may perform a Mode 1 scan 400 and a Mode 2 scan 424, both of which may be performed as described herein. For example, a Mode 1 scan 400 may include scanning across a sample ( Figure 4 The mode 2 scan 424 may include scanning a scan stripe across the rows of dies 402, 404, and 406 on the sample (not shown) to thereby generate an M1 frame 408, and the mode 2 scan 424 may include scanning a scan stripe across the rows of dies 402, 404, and 406 on the sample to thereby generate an M2 frame 432. The SMJ step 410 may be performed on the M1 frame, and the SMJ step 434 may be performed on the M2 frame. The SMJ step 410 may separate the M1 frame into two jobs, M1, and the test image (T) of job 1. 11 )412 and M1, the test image of operation 2 (T 12) 418. SMJ step 434 can divide the M2 frame into two test images (T 21 )436 and M2, the test image of operation 2 (T 22 ) 442. Then, a different reference image may be generated for each operation of each mode of the test image. For example, the computer system may generate a reference image (R ) for M1, operation 1, for the test image T11. 11 )414, for the test image T 12 M1, the reference image of job 2 (R 12 )420, for the test image T 21 M2, the reference image of job 1 (R 21 )438 and for the test image T 22 M2, the reference image of job 2 (R 22 ) 444. Each of these reference images may be generated as further described herein.

[0079] The computer system also generates a difference image for each set of reference and test images. For example, the computer system may generate a difference image (T 11 -R 11 )416, M1, the difference image of operation 2 (T 12 -R 12 )422, M2, the difference image of operation 1 (T 21 -R 21 )440 and M2, the difference image of operation 2 (T 22 -R 22 ) 446. Each of these difference images may be generated as further described herein. The computer system may then combine the images for defect detection, as shown in step 448, and the combined image may include any of the test, reference, and difference images generated as described above. In some examples, the computer system may perform image alignment 450 to align an image from one mode to an image from another mode, which may be performed as further described herein. Additionally, the computer system may perform defect detection in step 452 using any of the images in the combined input. Defect detection may be performed in step 452 as further described herein.

[0080] The detection step is one of the novel aspects of multi-modal detection described herein. In particular, the information processed from the modes is not limited to attributes. Instead, the input to defect detection contains a difference image for each detected defect so that more advanced algorithms (such as but not limited to difference image cross-correlation) can be used to combine the images of the two modes and better filter out or preserve defects. The use of raw images in defect detection also allows additional new attribute calculations that are not feasible using only individual scan results. In this way, the defect detection described herein can be a type of multi-modal fusion detection in which a computer system fuses data between modes using MCATx scan strip images.

[0081] Defect detection using the combined input can be performed in one of a variety of different ways, such as MDAT or nonlinear detectors (such as machine learning (ML) or deep learning (DL)), and detectors based on one-dimensional (1D) or two-dimensional (2D) models (such as 1D or 2D cloud detection). In addition, the defect detection step may include a fused combined difference image using cross-correlation or HLAT-like. MDAT and HLAT are defect detection algorithms used by some inspection tools commercially available from KLA. For example, some examples of MDAT are also described in U.S. Patent No. 10,923,317 issued to Huang et al. on February 16, 2021, which is incorporated herein by reference as if set forth in its entirety. The embodiments described herein may be further configured as described in such patent.

[0082] The inspection step may include a fused cloud approach, which is based on a noise cloud. The noise cloud may be a distribution of difference image values ​​(like TR). Defect inspection is performed to find outliers in the distribution of difference values ​​(or noise cloud). Currently, inspection uses only one noise distribution from one optical mode. Even in currently used multi-mode inspections, inspection may still be based on the noise distribution in the primary optical mode, and disturber suppression is based on image properties calculated at the same location as the image of the secondary optical mode. In contrast, the embodiments described herein may perform a noise cloud approach based on two inspection tool modes simultaneously.

[0083] In some embodiments, the detection step comprises a 1D noise cloud detection method. Thus, the detection step may comprise a multimodal 1D detection. The 1D noise cloud method (or "1st order cloud method") may generate a scalar value from two difference images. In one such example, the 1D noise cloud method may comprise converting the two noise clouds into one normalized scalar value (e.g. Where σ is derived from (T ij -R ij) estimate) and detect outliers. If more than one reference is generated for any one test image, for example as can be done in dual detection, the 1D noise cloud method can also be used in the same way for a second detection using a second reference image. In addition, this 1D noise cloud method can be performed separately for each job of the images, each of which can be generated, for example, based on the similarity of the images in the scanned swath to each other.

[0084] In another embodiment, the detecting step comprises a 2D noise cloud detection method. Thus, the detecting step may comprise a multimodal 2D cloud detection method. The 2D noise cloud method generates a mixed value from two difference images. For example, the 2D noise cloud method may comprise performing a modal difference cross-correlation (e.g. |T 11 -R 11 |⊙|T 21 -R 21 |) and detecting outliers. The 2D noise cloud method may include, for example, mapping the two individual noise distributions into a 2D distribution map and detecting outliers. If more than one reference is generated for any one test image, for example if this can be done in a dual detection, the 2D noise cloud method may also be used in the same manner for a second detection using a second reference image. Additionally, this 2D noise cloud method may be performed separately for each job of the images, each of which may be generated, for example, based on the similarity of the images in the scan swath to each other.

[0085] Embodiments may also generate noise clouds with more than two dimensions. For example, a 2D noise cloud approach may be suitable for some defect types, but other defect types may require a more complex, higher dimensional nonlinear map. For such defect types, a three-dimensional (3D) noise cloud may be useful for defect detection if it is not too sparse. One such noise cloud can be plotted on the x-axis by plotting |T 11 -R 11 |, plot |T on the y-axis 21 -R 21 | and draw on the z-axis In this way, the 3D noise cloud method can be a type of extension of the 2D distribution map to a scalar value (Z direction). Higher dimensions expand the problem domain, sometimes allowing the determination of a linear plane that separates two data sets. For example, two data sets at X and Y coordinates each have their own distribution. The individual distributions can be mapped in 2D based on common coordinates. Using the 2D map, a threshold can be determined and used to find outliers. However, some defects may be inside the 2D cloud and therefore cannot be detected by the threshold. In this case, hyperdimensions (e.g., 3D) can help with separation, but may encounter data sparsity (or insufficiency) problems, which can cause overfitting.

[0086] In further embodiments, the detection step includes DL-based defect detection. In some such examples, ML or DL ​​may be specialized to expand feature dimensions and manage the overfitting problems described above that may be experienced using higher dimensional noise cloud methods. Some examples of defect classifiers that may also perform defect detection and are based on ML are described in U.S. Pat. No. 10,607,119, issued to He et al. on Mar. 31, 2020, which is incorporated herein by reference as if fully set forth. Some examples of ML-based defect detectors are described in the following patents: U.S. Patent No. 10,186,026 issued to Karsenti et al. on January 22, 2019, U.S. Patent No. 10,599,951 issued to Bhaskar et al. on March 24, 2020, U.S. Patent No. 10,713,769 issued to Zhang et al. on July 14, 2020, and U.S. Patent No. 11,551,348 issued to Zhang et al. on January 10, 2023, which are incorporated herein by reference as if fully set forth. The embodiments described herein may be further configured as described in these references.

[0087] The detection step may also be performed as described in U.S. Patent Application No. 18 / 128,125 filed by Virk et al. on March 29, 2023, which is incorporated herein by reference as if fully set forth. The embodiments described herein may be further configured as described in such patent application.

[0088] In one embodiment, the inspection subsystem is configured to generate first and second mode test images by scanning a scan band across a row of dies on a sample using the first and second modes, respectively, the scan band having a length equal to or greater than the width of the row of dies, and the computer system is configured to perform the steps of generating first and second mode reference images, subtracting, combining, and detecting using only the first and second mode test images generated by scanning. In this way, one scan band across a row of dies on a sample can be scanned sequentially or simultaneously using the first and second modes. The entire width of a row of dies on a sample can be scanned using both modes, although this is not necessary. In addition, the scan band may or may not cover the entire height of the row of dies, and multiple scan bands can be scanned over the row of dies using the same mode, as described herein. After the entire row of dies has been scanned using both modes, the steps described herein can be performed (although some steps can be performed only after the first mode scan has been performed). In this way, the steps described herein can be performed separately and independently for each row of dies scanned on the sample. If more than two scan bands are inspected on a sample using the embodiments described herein, the inspection results generated for the sample may include all defects detected in all scan bands, for example, by simply combining the results of the different scan bands. However, more complex ways of combining the inspection results of multiple scan bands may be used and / or the inspection results of multiple scan bands may be used to perform additional steps, such as detecting systematic defects, determining spatial distribution characteristics of defects, etc.

[0089] The embodiments described above may be particularly useful when multi-mode data is only available sequentially. In such examples, embodiments may retain the mode data in a permanent storage (such as a hard disk) or a temporary storage (such as a memory) until the second mode data is available. The more data stored, the higher the data rate and the larger the storage space required. In addition, since changing the inspection tool mode takes substantial time, it is advantageous to minimize the frequency of mode changes.

[0090] The scanning tape and data processing strategy described above, which may also be referred to herein as the MCATx scanning tape strategy, can overcome or at least minimize many of the problems identified above. The scanning tape strategy is unique in that it scans the same scanning tape area at the row of dies (which may also be referred to as "one path") over the entire sample. The scanning tape strategy maintains the entire scanning tape information for one path. This scanning tape and data processing strategy may also provide higher sensitivity than other methods because so much data may be maintained and processed together, which may be performed as further described herein. Although this scanning tape and data processing strategy provides advantages for a single-mode inspection process, the embodiments described herein provide additional advantages, such as providing enhanced detection sensitivity when this scanning tape and data processing strategy is used with multi-mode information.

[0091] In some embodiments, the detection step does not include determining a first mode attribute of a first mode candidate defect only from one or more of the first mode test, reference, and difference images or determining a second mode attribute of a second mode candidate defect only from one or more of the second mode test, reference, and difference images. In other words, the embodiments described herein do not determine separate attributes (e.g., first defect attributes from the first mode image, second defect attributes from the second mode image, etc.) from images generated for or using different modes of the inspection subsystem and then use the attributes to determine whether to report or filter the defect as a disturbing point, such as using a rule-based approach. In addition, the embodiments described herein do not use the results of two independent detection scans from different modes to apply rules to the attributes of a common defect to filter or retain the defect. Instead, the embodiments described herein jointly utilize the images from the different modes themselves for defect detection, which can be beneficial in several ways further described herein.

[0092] In another embodiment, the detection step includes determining an attribute of a candidate defect based on at least one of a first mode test, a reference, and a difference image in combination with at least one of a second mode test, a reference, and a difference image. In this way, the embodiments described herein enable the determination of defect attributes from multi-mode images. For example, determining the attributes of a candidate defect may include determining an attribute of a candidate defect that varies according to the first and second difference images. In this way, a computer system may calculate attributes based on multi-mode images. The defect attributes determined from the multi-mode images may include any defect attributes known in the art, such as defect size, defect shape, defect roughness, defect orientation, and the like. In addition, since the defect attributes may be determined based on at least one image generated using the first mode and at least one image generated using the second mode, the defect attributes may be more accurate than when the defect attributes are determined separately from different mode images. The embodiments described herein also make multi-mode defect attributes or defect attributes that cannot be determined from images generated using a single mode feasible. In other words, the embodiments described herein may provide a new way to describe the attributes of a defect using images generated in more than one mode.

[0093] The embodiments described herein may then use the multi-mode defect attributes for defect detection, and these defect attributes may be used for defect detection in the same manner as other defect attributes. For example, defect detection may include applying a threshold to the multi-mode defect attribute and identifying only portions of an image with the multi-mode defect attribute above the threshold as defects. In another example, defect detection may include applying one or more rules to the multi-mode defect attribute to detect defects in an image. In this manner, multi-mode defect attributes may be determined for portions of an image that are not known to be defects and the attributes may then be used for defect detection. However, such multi-mode defect attributes may also or alternatively be used after defect detection for functions such as disturber point filtering, defect classification, and the like. In one such embodiment, a computer system is configured to determine an attribute of at least one of the detected defects based on at least one of a first mode test, a reference, and a difference image in combination with at least one of a second mode test, a reference, and a difference image. In this manner, the multi-mode defect attributes may be used in a post-defect detection manner. In such examples, defects may be detected in any suitable manner as described herein, and then for any potential defects, the multi-mode defect attributes may be determined and used for one or more of the functions described above.

[0094] In a further embodiment, a computer system is configured to detect first mode candidate defects on a sample based on a first mode difference image, and a second mode reference image is generated, the second mode reference image is subtracted from the second mode test image, the combining step and the detecting step are performed only for the location of the first mode candidate defect. In this way, the computer system can determine candidates in the event that are potential DOIs. For example, a similar MCATx detection scan can be performed using the second mode and the location provided by the first mode. In other words, the first mode can be used to scan the sample in a first scan band, and defect detection can be performed using only the images generated in the scan as described herein. Then, the same scan band can be scanned using the second mode to obtain target and reference image candidates at the location provided by the first mode defect detection using the second mode. Then, the inspection performed using the second mode image can be performed as further described herein. In some examples, the reference image generated using the first mode can also be used for the second mode. Whether the same reference is used for both modes can depend on the similarity of the images generated by the first and second modes. Then, the first and second mode images (original T, R or difference image TR) can be combined to perform final defect detection at each candidate location.

[0095] Thus, when the system alternates between modes for scanning a sample (meaning the modes are used sequentially to scan the sample), the embodiments described herein can coordinate scanning and data processing in several different ways. In one way, the entire area is scanned using two modes and then data processing begins after the scan is complete. In this way, each scanned position can be checked using both modes. In other words, defect detection will only occur when both the first and second mode data sets are fully available. This approach is relatively simple when the two modes can be used to scan the sample simultaneously but becomes more complex when the system alternates sequentially between modes for scanning. For example, this approach can be advantageous except for the case where it causes issues with data storage and data rates where it is preferred to write / read the entire data without delaying any scan. More specifically, the goal of some currently used inspection methods is to complete the inspection when the scan is complete, meaning that defect detection and scanning are almost real-time. Current inspection scan data rates can be tens of GB / s, which can require a lot of storage technology or cost to keep up with the speed. To alleviate this from a system perspective, embodiments can use test images as they become available to perform defect detection, even if not all mode images are available. In this way, detection can begin immediately when one path of both modes is completed, since one path data can be completely independent of the other paths.

[0096] In this manner, the scanning strategy used by the embodiments described herein may include revisiting potential candidate defects detected using one mode to scan them using another mode. However, the pattern scanning strategy used by the embodiments described herein may also include using independent scanning of the modes, meaning that the area on the sample scanned using one mode is independent of any results produced using the other mode. For example, the inspection subsystem described herein may be configured to scan an entire swath on the sample using both modes and then jointly process the data after the scanning of the swath has been completed.

[0097] The inspection subsystem may alternate between the two modes until the complete inspection area is inspected. The inspection area of ​​a partial scan for each mode is a trade-off between memory and hardware performance, since the mode change requires significant additional consumption of switching time. The impact of mode switching on the inspection subsystem hardware may also be considered.

[0098] In one such embodiment, detecting a first mode candidate defect includes applying a thermal threshold to the first mode difference image. The term "thermal" threshold as used herein is generally defined as a threshold set at, near, or even within the noise floor of the image produced by the inspection subsystem. For example, the computer system may detect a first mode candidate defect with a substantially low threshold that may be set at the noise floor. In this way, the first mode candidate defect may include a large number of disturbing points, but may be filtered out using the second mode image. In addition, using this low threshold to detect candidate defects using the first mode increases the probability that all defect candidates on the sample will be detected using the first mode, even if not all defect candidates will be detected using the normal threshold. In other words, using this low threshold reduces the probability that a defect candidate will be missed by the first mode image processing and therefore not considered when performing the second mode image processing. Applying the thermal threshold may be performed in any suitable manner, for example, only identifying portions of the image above the thermal threshold as defect candidates.

[0099] In one embodiment, at least one of the defects is detectable only in the first or second mode. For example, some defect types are sensitive to optical modes, and therefore, there may be a very weak signal or no signal for a defect in one mode using a particular optical mode but not for a defect in another mode. Thus, this embodiment differs from some currently used multi-mode inspection processes in which defect candidates are detected separately in two or more modes and then defects that are common to the two or more modes are identified as actual defects. Instead, by considering multi-mode images together for defect detection, for example, based on one of the noise cloud methods described herein, defects that can be detected using only one mode may be detected. Thus, the embodiments described herein may advantageously increase the probability that all defects on a sample are detected by inspection.

[0100] In additional embodiments, the computer system is configured to perform single-mode defect detection for a sample based only on the first or second test image and to generate an inspection result for the sample based on the single-mode defect detection and the results of the detection step. For example, the embodiments described herein may contain additional processing channels that maintain an existing single-mode detection of one or both modes. In this way, a single-mode detection may be performed using a first mode, a single-mode detection may also or alternatively be performed using a second mode, and so on. The results of the single-mode defect detection and the multi-mode defect detection described herein may be combined in any suitable manner into a single inspection result for the sample.

[0101] In some embodiments, the computer system is configured to perform additional multi-pattern defect detection for the sample based on the first and second pattern attributes determined for the defect candidates from the first and second test patterns, respectively, and to generate an inspection result for the sample based on the additional multi-pattern defect detection and the results of the detection steps. For example, the embodiments described herein may contain additional processing channels that maintain the currently used or other multi-pattern processing methods (such as the multi-pattern processing methods further described above). The results of different multi-pattern defect detections can be combined into a single inspection result for the sample in any suitable manner.

[0102] Any computer system described herein may generate inspection results, which may include the results of any of the steps described herein. Inspection results may include information of detected defects, such as defect ID, location of bounding boxes of detected defects, etc., size, detection score, information about defect classification (such as class tags or IDs, etc.), or any such suitable information known in the art. Results of defects may be generated by a computer system in any suitable manner. Results of defects may have any suitable form or format, such as a standard file type. A computer system may generate results and store the results so that the results may be used by the computer system and / or another system or method to perform one or more functions of a sample or another sample of the same type.

[0103] The computer system may be configured to store information of detected defects in any suitable computer-readable storage medium. The information may be stored with any results described herein and may be stored in any manner known in the art. The storage medium may include any storage medium described herein or any other suitable storage medium known in the art. After the information has been stored, the information may be accessed in the storage medium and used by any method or system embodiment described herein, formatted for display to a user, used by another software module, method or system, etc.

[0104] The results and information generated by performing inspections on samples can be used in various ways through the embodiments and / or other systems and methods described herein. Such functions include (but are not limited to): changing a process, such as a manufacturing process or step that has been or will be performed on the inspected sample or another sample in a feedback or feedforward manner. For example, a computer system may be configured to determine one or more changes to a process that has been or will be performed on a sample inspected as described herein based on detected defects. The change in the process may include any suitable change in one or more parameters of the process. The computer system preferably determines the change so that defects on other samples on which the revised process is performed can be reduced or prevented, defects on the sample can be corrected or eliminated in another process performed on the sample, defects can be compensated for in another process performed on the sample, etc. The computer system can determine such changes in any suitable manner known in the art.

[0105] The changes may then be sent to a semiconductor manufacturing system (not shown) or a storage medium (not shown) accessible to the computer system and the semiconductor manufacturing system. The semiconductor manufacturing system may or may not be part of the system embodiments described herein. For example, the computer system and / or inspection subsystem described herein may be coupled to the semiconductor manufacturing system, for example, via one or more common components such as a housing, a power supply, a sample handling device or mechanism, etc. The semiconductor manufacturing system may include any semiconductor manufacturing system known in the art, such as a lithography tool, an etching tool, a chemical-mechanical polishing (CMP) tool, a deposition tool, and the like.

[0106] Each embodiment of the system described above can be combined together into a single embodiment. In other words, unless otherwise mentioned herein, a system embodiment is not mutually exclusive with any other system embodiment.

[0107] The embodiments described herein have several important advantages over other multi-mode inspection methods and systems. For example, the embodiments described herein provide improved multi-mode defect detection by combining raw images between modes. The embodiments described herein also provide improved multi-mode defect detection performance by using MCATx scan bands in each mode to calculate the best reference image in each mode. The embodiments described herein also provide improved defect detection performance on currently used inspection tools, which can directly lead to additional authorization using existing inspection tool hardware.

[0108] The improvements and advantages provided by the embodiments described herein are achieved by several new features, including, but not limited to, the use of MCATx in multimodal inspection and multimodal MCATx image fusion for detection.

[0109] Another embodiment relates to a method for detecting defects on a sample. The method includes acquiring first and second mode test images of the sample generated using first and second modes of an inspection subsystem, respectively. Acquiring the image can be performed using the inspection subsystem described herein (e.g., by directing a light or electron beam to the sample and detecting the light or electron beam from the sample, respectively). In this way, acquiring the image can be performed using the physical sample itself and some type of imaging hardware. However, acquiring the image does not necessarily include imaging the sample using imaging hardware. For example, another system and / or method may generate an image and may store the generated image in one or more storage media (e.g., a virtual inspection system as described herein or another storage medium described herein). Therefore, acquiring the image may include acquiring the image from a storage medium that has stored the image.

[0110] The method also includes the steps described above of generating first and second mode reference images, subtracting the reference images from the test images on a mode-to-mode basis, combining the multi-mode images as input for defect detection, and detecting defects. The steps are performed by a computer system that can be configured according to any embodiment described herein.

[0111] Each step of the method may be performed as further described herein. The method may also include any other steps that may be performed by the inspection subsystem and / or computer system described herein. In addition, the method described above may be performed by any system embodiment described herein.

[0112] Additional embodiments relate to a non-transitory computer-readable medium storing program instructions executable on a computer system to perform a computer-implemented method for detecting defects on a sample. Figure 5 One such embodiment is shown in FIG. Figure 5 , non-transitory computer readable medium 500 includes program instructions 502 executable on a computer system 504. The computer-implemented method may include any of the steps of any of the methods described herein.

[0113] Program instructions 502 implementing methods such as those described herein may be stored on computer readable medium 500. The computer readable medium may be a storage medium such as a magnetic or optical disk, tape, or any other suitable non-transitory computer readable medium known in the art.

[0114] The program instructions may be implemented in any of a variety of ways, including process-based techniques, component-based techniques, and / or object-oriented techniques, etc. For example, the program instructions may be implemented using ActiveX controls, C++ objects, JavaBeans, Microsoft Foundation Classes ("MFC"), SSE (Streaming SIMD Extensions), GPU programming packages (e.g., Tensorflow, CUDA, etc.), or other techniques or methodologies as desired.

[0115] Computer system 504 may be configured according to any of the embodiments described herein.

[0116] In view of this description, further modifications and alternative embodiments of various aspects of the present invention will be apparent to those skilled in the art. For example, methods and systems for detecting defects on a sample are provided. Therefore, this description should be interpreted only as illustrative and for the purpose of teaching those skilled in the art the general manner of implementing the present invention. It should be understood that the form of the present invention shown and described herein will be considered as the current preferred embodiment. As will be understood by those skilled in the art after having the benefit of this description of the present invention, elements and materials may be substituted for the elements and materials illustrated and described herein, parts and processes may be reversed, and the specific attributes of the present invention may be utilized independently. Changes may be made to the elements described herein without departing from the spirit and scope of the present invention as described in the appended claims.

Claims

1. A system configured to detect defects on a sample, comprising: an inspection subsystem configured to generate first and second mode test images of a sample using first and second modes of the inspection subsystem, respectively; and A computer system configured to: generating first and second mode reference images from the first and second mode test images, respectively; subtracting the first and second mode reference images from the first and second mode test images, respectively, to thereby generate first and second mode difference images, respectively; combining the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as input for defect detection; and Defects on the sample are detected based on at least the first and second mode difference images in the input.

2. The system of claim 1, wherein the computer system is further configured to align the first and second pattern test images to each other by individually aligning the first and second pattern test images to a design of the sample.

3. A system according to claim 1, wherein the inspection subsystem is further configured to generate the first and second mode test images by scanning a scanning band across the row of die on the sample using the first and second modes, respectively, wherein the scanning band has a length equal to or greater than the width of the row of die, and wherein the computer system is further configured to perform the generating the first and second mode reference images, subtracting, combining and detecting using only the first and second mode test images generated by the scanning.

4. A system according to claim 1, wherein the computer system is further configured to calculate first and second similarity metrics for the first and second mode test images, respectively, and to separate the first and second mode test images into first and second detection operations based on the first and second similarity metrics, respectively, and wherein the generating of the first and second mode reference images, subtraction, combination and detection are performed separately for each of two or more of the first and second detection operations.

5. A system according to claim 4, wherein the generating of the first and second mode reference images performed for at least one of the first and second detection operations includes selecting one or more of the first and second mode test images for the generating of the first and second mode reference images independently of the locations on the sample where the first and second test mode images are generated.

6. A system according to claim 1, wherein the detection does not include determining a first pattern attribute of a first pattern candidate defect only from one or more of the first pattern test, reference, and difference images or determining a second pattern attribute of a second pattern candidate defect only from one or more of the second pattern test, reference, and difference images.

7. The system of claim 1, wherein the detecting comprises determining an attribute of a candidate defect based on at least one of the first mode test, a reference, and a difference image in combination with at least one of the second mode test, a reference, and a difference image.

8. The system of claim 1, wherein the computer system is further configured to determine an attribute of at least one of the detected defects based on at least one of the first pattern test, reference, and difference images in combination with at least one of the second pattern test, reference, and difference images.

9. The system of claim 1 , wherein the computer system is further configured for detecting a first mode candidate defect on the sample based on the first mode difference image, and wherein the generating of the second mode reference image, the subtracting of the second mode reference image from the second mode test image, the combining, and the detecting are performed only for the location of the first mode candidate defect. 10 . The system of claim 9 , wherein detecting the first pattern candidate defect comprises applying a thermal threshold to the first pattern difference image.

11. The system of claim 1, wherein at least one of the defects is detectable only in the first or second mode.

12. The system of claim 1, wherein the detecting comprises a one-dimensional noise cloud detection method.

13. The system of claim 1, wherein the detecting comprises a two-dimensional noise cloud detection method.

14. The system of claim 1, wherein the detecting comprises deep learning based defect detection.

15. The system of claim 1, wherein the computer system is further configured for performing single pattern defect detection for the sample based on only the first or second test image and for generating an inspection result of the sample based on the single pattern defect detection and a result of the detection.

16. The system of claim 1, wherein the computer system is further configured for performing additional multi-mode defect detection for the sample based on first and second pattern attributes determined for defect candidates from the first and second test images, respectively, and for generating inspection results for the sample based on the additional multi-mode defect detection and results of the detection.

17. The system of claim 1, wherein the inspection subsystem is further configured for generating the first and second pattern test images using light.

18. The system of claim 1, wherein the inspection subsystem is further configured for electronically generating the first and second pattern test images.

19. A non-transitory computer readable medium storing program instructions executable on a computer system to perform a computer implemented method for detecting defects on a sample, wherein the computer implemented method comprises: acquiring first and second mode test images of the sample generated using first and second modes of the inspection subsystem, respectively; generating first and second mode reference images from the first and second mode test images, respectively; subtracting the first and second mode reference images from the first and second mode test images, respectively, to thereby generate first and second mode difference images, respectively; combining the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as input for defect detection; and Defects on the sample are detected based at least on the first and second mode difference images, wherein the generating, subtracting, combining, and detecting are performed by the computer system.

20. A computer-implemented method for detecting defects on a sample, comprising: acquiring first and second mode test images of the sample generated using first and second modes of the inspection subsystem, respectively; generating first and second mode reference images from the first and second mode test images, respectively; subtracting the first and second mode reference images from the first and second mode test images, respectively, to thereby generate first and second mode difference images, respectively; combining the first and second mode test images, the first and second mode reference images, and the first and second mode difference images as input for defect detection; and Defects on the sample are detected based at least on the first and second mode difference images, wherein the generating, subtracting, combining and detecting are performed by a computer system.

Images