Marking local maxima via clustered wafer features for metrology guided inspection

Abstract

Methods and systems are provided for generating information for use in setting a process performed on a sample. A method includes clustering dies on a sample based on colors assigned to the dies in response to predicted defect densities in the dies determined from measurements performed on the sample, thereby generating an initial cluster of dies. The method also includes analyzing the initial die clusters in a location space to determine whether any of the initial die clusters contains two or more die clusters. In addition, the method includes designating, as a final die cluster, the initial die cluster that does not include two or more die clusters and the two or more die clusters included in any of the initial die clusters. The method further includes storing information of the final die cluster for use in setting a process performed on the sample.

Description

Technical Field

[0001] The present invention generally relates to methods and systems for generating information for use in setting up a process to be performed on a sample. Certain embodiments relate to marking local maxima via clustering of wafer features for metrology guided inspection. Background Art

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

[0003] The manufacture of semiconductor devices, such as logic and memory devices, typically involves processing a substrate (e.g., a semiconductor wafer) using a number of semiconductor fabrication processes to form the various features and multiple layers of the semiconductor device. For example, photolithography is a semiconductor fabrication process that involves transferring a pattern from a reticle to a resist disposed on a semiconductor wafer. Additional examples of semiconductor fabrication processes include, but are not limited to, chemical mechanical polishing (CMP), etching, deposition, and ion implantation. Multiple semiconductor devices can be fabricated in an arrangement on a single 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, promoting 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 even more critical to the successful manufacture of acceptable semiconductor devices.

[0005] Inspection tools have various parameters that can be modified based on the sample being inspected. Modifiable parameters typically include imaging hardware parameters and image processing-related parameters. While changing tool parameters from sample to sample can be advantageous, determining the correct parameters for any given sample can be particularly difficult.

[0006] Many methods have been developed to determine not only which test parameters are appropriate for a given sample, but also which test parameters are appropriate for different regions on the same sample in the same test. For example, the variation across samples described herein may mean that a test parameter that is most suitable in one sample region may be suboptimal or even completely useless in another sample region.

[0007] Methods have been developed for selecting inspection parameters that vary from sample area to sample area based on several characteristics of the sample and the tool. Some of the sample characteristics that have been considered are primarily related to the design formed on the sample, including the characteristics of the patterned features in the design and the locations at which they are formed on the sample. Characteristics of the design that can be used include characteristics that are independent of the inspection tool and process, such as which patterned features in the design are most critical to device function and should therefore be inspected with maximum sensitivity. Characteristics of the design can also or alternatively include characteristics of the design that can have some impact on the inspection tool output and therefore can affect both the signal and noise in the inspection tool output, such as size, orientation, roughness, material properties, patterned features on underlying layers, etc.

[0008] Some other sample characteristics that have been considered relate to noise in the inspection tool output for different areas on the sample. This noise can be related to the design formed on the sample, as described above. Noise can also be related to other properties of the sample, such as variations in the sample caused by the processes performed on the sample. One obvious reason why inspection parameters may be based on noise is that noise can mask defect signals. For example, if the noise is too similar to the defect signal or even exceeds the defect signal, the defect signal may be missed and the inspection may fail.

[0009] While seemingly straightforward, these and other sample characteristics, along with the inspection tool configuration and its sample-dependent properties, make setting up the inspection process quite complex. For example, if we only design information for inspection setup, even that setup can be complex due to the complexity of the design and how it may affect the inspection tool output. In another example, it becomes increasingly difficult to predict the noise that will be present on the sample and how it may vary across samples, making noise-dependent inspection setup difficult. When considering a combination of factors in inspection process setup, the difficulty of setup can increase exponentially with the number of variables.

[0010] Other factors that can complicate inspection recipe setup include performance considerations and requirements. For example, the fastest inspection process typically involves the fewest different parameters per sample. Specifically, if an inspection process includes inspections with different parameter sets, it will typically take longer than if the same inspection could be performed using only one parameter set. Similarly, the inspection process will typically take longer, for example, if more parameters must be switched from area to area on the sample. Therefore, some inspection recipe setup processes have focused on how to group areas on the sample that are similar in the characteristics described above. Inspection recipe setup methods can also or alternatively focus on how to meet inspection goals (e.g., total defect capture rate, specific defect of interest (DOI) capture rate, disturbing point detection rate, etc.) in the minimum amount of time. These methods can exploit the possibility of trading processing power for performance.

[0011] While inspection recipe setup has made significant progress over the past few decades, many challenges remain. For example, as sample designs increase in complexity and their pattern sizes decrease, it has become increasingly difficult to predict how the sample itself will affect inspection tool output. Furthermore, defects of interest are only getting smaller, making their detection increasingly difficult, especially as most inspection tools are pushed to or used at the limits of their performance capabilities. These factors, along with the degree to which samples can vary across samples, also make creating an inspection recipe that performs adequately across the entire sample set particularly challenging.

[0012] Therefore, it would be advantageous to develop systems and methods for generating information for use in setting up a process to be performed on a sample without one or more of the disadvantages described above. Summary of the Invention

[0013] The following description of various embodiments should in no way be understood as limiting the subject matter of the appended technical solutions.

[0014] One embodiment relates to a system configured to generate information for use in setting up a process to be performed on a sample. The system includes one or more computer systems configured to cluster the die based on colors assigned to the die in response to predicted defect densities in the die on the sample determined from measurements performed on the sample, thereby generating initial die clusters. The one or more computer systems are further configured to analyze the initial die clusters in a location space to determine whether any of the initial die clusters contain two or more die clusters. Additionally, the one or more computer systems are configured to designate the initial die clusters that do not contain two or more die clusters and the two or more die clusters contained in any of the initial die clusters as final die clusters. The one or more computer systems are further configured to store information about the final die clusters for use in setting up a process to be performed on the sample. The system can be further configured as described herein.

[0015] Another embodiment relates to a computer-implemented method for generating information for use in setting up a process to be performed on a sample. The method includes the clustering, analyzing, specifying, and storing steps described above. The steps of the method are performed by one or more computer systems. The steps of the method may be further performed as described herein. The method may include any other steps of any other method described herein. The method may be performed by any system described herein.

[0016] Another embodiment relates to a non-transitory computer-readable medium storing program instructions executable on a computer system to perform a computer-implemented method for generating information used in setting up a process to be performed on a sample. The computer-implemented method includes the steps of the method described above. The computer-readable medium may be further configured as described herein. The steps of the computer-implemented method may be performed as further described herein. In addition, the computer-implemented method for which the program instructions may be executed may include any other steps of any other method described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Additional advantages of the present invention will become apparent to those skilled in the art upon having the benefit of the following detailed description of the preferred embodiments and after reference to the accompanying drawings, in which:

[0018] Figure 1 and Figure 2 is a schematic diagram illustrating a side view of an embodiment of a system configured as described herein;

[0019] Figure 3 is a schematic diagram illustrating an example of a die-level wafer map of defect density probabilities with four localized clusters;

[0020] Figure 4 is a schematic diagram illustrating an example of a die-level wafer map of clustering results generated by KMeans;

[0021] Figure 5 is a schematic diagram illustrating one example of a die-level wafer map of final die clustering that may be produced by embodiments described herein;

[0022] Figure 6 is a flow chart illustrating an embodiment of steps that may be performed to produce a final die cluster as described herein; and

[0023] Figure 7 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.

[0024] While the invention is susceptible to various modifications and alternative forms, specific embodiments of the invention are shown by way of example in the drawings and described in detail herein. The drawings may not be to scale. However, it should be understood that the drawings and detailed description thereof are not intended to limit the 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 invention as defined by the appended claims. DETAILED DESCRIPTION

[0025] Referring now to the drawings, it should be noted that the figures are not drawn to scale. In particular, the proportions of some elements of the figures are greatly exaggerated to emphasize the characteristics of the elements. It should also be noted that the figures are not drawn to the same scale. The same reference numerals have been used to indicate similarly configured elements shown in more than one figure. Unless otherwise noted herein, any elements described and shown may include any suitable commercially available components.

[0026] One embodiment relates to a system configured to generate information for use in setting up a process to be performed on a sample. In some embodiments, the sample is a wafer. The wafer may include any wafer known in the semiconductor art. Although some embodiments are described herein with respect to a wafer or wafers, the embodiments are not limited to the samples with which they may be used. For example, the embodiments described herein may be used with samples such as reticles, tablets, personal computer (PC) boards, and other semiconductor samples.

[0027] exist Figure 1 One embodiment of such a system is shown in . The system may include an imaging system comprising at least an energy source and a detector. The energy source is configured to generate energy that is directed toward a sample. The detector is configured to detect energy from the sample and generate an output in response to the detected energy.

[0028] In one embodiment, the energy directed to the sample comprises light, and the energy detected from the sample comprises light. Figure 1 , imaging system 10 includes an illumination subsystem configured to direct light to sample 14. The illumination subsystem includes at least one light source (e.g., 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. For example, Figure 1 , light from a light source 16 is directed through an optical element 18 and then through a lens 20 to a beam splitter 21, which directs the light at a normal angle of incidence to a sample 14. The angle of incidence may include any suitable angle of incidence, which may vary depending on, for example, the characteristics of the sample, the defects to be detected on the sample, the measurements to be performed on the sample, etc.

[0029] The illumination subsystem can be configured to direct light to the sample at different angles of incidence at different times. For example, the imaging system can be configured to modify one or more characteristics of one or more elements of the illumination subsystem so that the light can be directed to the sample at different angles of incidence. Figure 1 In one such example, the imaging system can be configured to move the light source 16, optical element 18, and lens 20 so that light is directed to the sample at different angles of incidence.

[0030] The imaging system can be configured to direct light to the sample at more than one angle of incidence at the same time. For example, the imaging system can include more than one illumination channel, one of which can include a plurality of illumination channels. Figure 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 components (such as those described further herein). If this light is directed to the sample simultaneously with 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, such that light originating from illuminating the sample at different angles of incidence can be distinguished from one another at a detector.

[0031] However, the lighting subsystem may include only one light source (e.g. Figure 1 ) and the light from the light source can 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 can then be directed to the sample. Multiple illumination channels can 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 can be configured to direct light with different characteristics to the sample at different times. For example, the optical element 18 can be configured as a spectral filter and the properties of the spectral filter can be changed in various ways (e.g., by swapping out the spectral filter) so that light of different wavelengths can be directed to the sample at different times. The illumination subsystem can have any other suitable configuration known in the art for directing light with different or the same characteristics to the sample, either sequentially or simultaneously, at different or the same angles of incidence.

[0032] Light source 16 may comprise a broadband plasma (BBP) light source. In this manner, the light generated by the light source and directed toward the sample may comprise broadband light. However, the light source may comprise any other suitable light source, such as any suitable laser known in the art that is configured to generate light of any suitable wavelength known in the art. Furthermore, the laser may be configured to generate monochromatic or near-monochromatic light. Thus, the laser may be a narrowband laser. The light source may also comprise a polychromatic light source that generates light at multiple discrete wavelengths or bands.

[0033] Light from the optical element 18 can be focused by the lens 20 onto the beam splitter 21. 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 elements onto the sample. Figure 1The illumination subsystem shown in FIG and described herein may include any other suitable optical elements (not shown). Examples of such optical elements include, but are not limited to, polarizing elements, 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 modify one or more elements of the illumination subsystem based on the type of illumination used for inspection, metrology, and the like.

[0034] The imaging system may also include a scanning subsystem configured to cause light to be scanned across the sample. For example, the imaging system may include a stage 22 on which the sample 14 is positioned during inspection, measurement, etc. The scanning subsystem may include any suitable mechanical and / or robotic assembly (including the stage 22) that can be configured to move the sample so that light can be scanned across the sample. Additionally or alternatively, the imaging system may be configured so that one or more optical elements of the imaging system perform a scan of light across the sample. The light can be scanned across the sample in any suitable manner.

[0035] The imaging system further includes one or more detection channels. At least one of the one or more detection channels includes a detector configured to detect light from the sample due to illumination of the sample by the imaging system and to generate an output in response to the detected light. For example, Figure 1 The imaging system shown in FIG 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, one detection channel is configured to detect specularly reflected light, and the other detection channel is configured to detect light that is not specularly reflected (e.g., scattered, diffracted, etc.) from the sample. However, two or more detection channels may be configured to detect the same type of light (e.g., specularly reflected light) from the sample. Although Figure 1 While an embodiment of an imaging system is shown that includes two detection channels, the imaging system may include a different number of detection channels (e.g., only one detection channel or two or more detection channels). Figure 1 Each light collector is shown in as a single refractive optical element, but each light collector may include one or more refractive optical elements and / or one or more reflective optical elements.

[0036] 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 detectors may also include non-imaging detectors or imaging detectors. If the detectors are non-imaging detectors, 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 position within the imaging plane. Thus, the output generated by each detector included in each detection channel may be a signal or data rather than an image signal or image data. In such examples, a computer system (e.g., computer system 36 of the system) may be configured to generate an image of the sample from the non-imaging output of the detectors. However, in other examples, the detectors may be configured as imaging detectors configured to generate imaging signals or image data. Thus, the system may be configured to generate images in a number of ways.

[0037] Provided in this article Figure 1 to generally illustrate the configuration of an imaging system that may be included in embodiments of the systems described herein. Obviously, the imaging system arrangements described herein may be modified to optimize the performance of the system as is typically performed when designing commercial inspection, metrology, and the like systems. In addition, the systems described herein may be implemented using existing inspection or metrology systems (e.g., by adding the functionality described herein to the existing inspection or metrology systems), such as the 29xx and 39xx series of tools, the SpectraShape series of tools, and the Archer series of tools commercially available from KLA Corp. of Milpitas, California. For some such systems, the embodiments described herein may be provided as optional functionality of the inspection or metrology system (e.g., in addition to other functionality of the inspection or metrology system). Alternatively, the imaging system described herein may be designed "from the ground up" to provide an entirely new inspection or metrology system.

[0038] The computer system 36 of the system can be coupled to the detector of the imaging system 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 system can receive the output generated by the detector during scanning of the sample. The computer system 36 can be configured to use the output of the detector to perform several functions as described herein and any other functions further described herein. This computer system can be further configured as described herein.

[0039] This computer system (and other computer systems described herein) may also be referred to herein as a computer subsystem. Each of the computer subsystems or systems described herein may take various forms, including personal computer systems, image computers, mainframe computer systems, workstations, network appliances, Internet appliances, 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. A computer subsystem or system may also include any suitable processor known in the art, such as a parallel processor. In addition, the computer subsystem or system may include a computer platform with high-speed processing and software (as a stand-alone tool or a networked tool).

[0040] If the system includes more than one computer system, the different computer systems may be coupled to each other so that images, data, information, instructions, etc. may be transmitted between the computer systems, as further described herein. For example, computer system 36 may be coupled to computer system 102 (e.g., by any suitable transmission medium, which may include any suitable wired and / or wireless transmission medium known in the art) via any suitable transmission medium. Figure 1 Two or more such computer systems may also be effectively coupled via a shared computer-readable storage medium (not shown).

[0041] Although the imaging system is described above as an optical or light-based system, the imaging system may be an electron beam-based system. For example, in one embodiment, the energy directed to the sample comprises electrons, and the energy detected from the sample comprises electrons. In this manner, the energy source may be an electron beam source. Figure 2 In one such embodiment shown in , the imaging system includes an electron column 122 coupled to a computer system 124 .

[0042] Also like Figure 2 , the electron column includes an electron beam source 126 configured to generate 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.

[0043] 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.

[0044] The electron column may include any other suitable elements known in the art. Furthermore, the electron column may be further configured as described in U.S. Patent No. 8,664,594 issued to Jiang et al. on April 4, 2014, U.S. Patent No. 8,692,204 issued to Kojima et al. on April 8, 2014, U.S. Patent No. 8,698,093 issued to Gubbens et al. on April 15, 2014, and U.S. Patent No. 8,716,662 issued to MacDonald et al. on May 6, 2014, which are incorporated herein by reference as if fully set forth.

[0045] Despite Figure 2 While the electron beam is shown as being configured so that electrons are directed at a sample at an oblique angle of incidence and scattered from the sample at another oblique angle, it should be understood that the electron beam can be directed at and scattered from the sample at any suitable angle. Furthermore, the electron beam system can be configured to generate images of the sample using multiple modes (e.g., with different illumination angles, collection angles, etc.). The multiple modes of the electron beam system can differ in any image generation parameter of the system.

[0046] The computer system 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. The electron beam image may include any suitable electron beam image. The computer system 124 may be configured to use the output of the detector and / or the electron beam image to perform any of the functions described herein. The computer system 124 may be configured to perform any additional steps described herein. The computer system 124 may be further configured to include, as described herein Figure 2 The imaging system shown in the figure.

[0047] Provided in this article Figure 2 The following generally illustrates the configuration of an electron beam-based imaging system that may be included in the embodiments described herein. As with the optical systems described above, the electron beam system arrangement described herein may be modified to optimize the system's performance, as is typically done when designing a commercial inspection or metrology system. Additionally, the systems described herein may be implemented using existing inspection, metrology, or high-resolution defect review systems, such as tools commercially available from KLA (e.g., by adding the functionality described herein to the existing inspection, metrology, or defect review system). For some such systems, the embodiments described herein may be provided as optional functionality of the system (e.g., in addition to the system's other functionality). Alternatively, the systems described herein may be designed "from the ground up" to provide an entirely new system.

[0048] Although the imaging system is described above as a light-based or electron-beam-based system, the imaging system may be an ion-beam-based system. Figure 2This imaging system is configured as shown in FIG, except that any suitable ion beam source known in the art may be used instead of the electron beam source. In addition, the imaging system may be any other suitable ion beam-based 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.

[0049] The imaging systems described herein can be configured to generate outputs (e.g., images) of a sample using multiple modes. In general, a "mode" is defined by the values of parameters of the imaging system used to generate an image of the sample (or an output for generating an image of the sample). Thus, different modes can differ in the value of at least one parameter of the imaging system (in addition to the location on the sample where the output is generated). In this way, an image can be generated by the imaging system using two or more different values of the parameters of the imaging system. For example, in an optical imaging system, different modes can be illuminated using light of different wavelengths. Modes can differ for different modes in terms of illumination wavelength, as further described herein (e.g., by using different light sources, different spectral filters, etc.). In addition, as mentioned above, an imaging system can include more than one illumination channel. Thus, different illumination channels can be used for different modes. Modes can differ in any one or more modifiable parameters of the imaging system (e.g., illumination polarization, angle, wavelength, etc., detection polarization, angle, wavelength, etc.).

[0050] Similarly, an image generated by an electron beam imaging system can include images generated by the electron beam imaging system using two or more different values of a parameter of the electron beam imaging system. Multiple modes of the electron beam imaging system can be defined by the values of the parameters of the electron beam imaging system used to generate an image of the sample. Thus, different modes can differ in the value of at least one of the electron beam parameters of the electron beam imaging system. For example, different modes can use different angles of incidence for illumination.

[0051] The imaging system embodiments described herein may be configured for inspection, metrology, defect detection, or another quality control related process performed on a sample. For example, the imaging system embodiments described herein and described in Figure 1 and 2 The embodiments of the imaging system shown in FIG. 5 provide different imaging capabilities depending on the application in which they are to be used. In one such example, Figure 1 The imaging system shown in can be configured to have higher resolution if it is to be used for defect review or metrology rather than for inspection. In other words, Figure 1 and 2The embodiments of the imaging systems shown in describe some common and various configurations of imaging systems that can be customized in a number of ways that will be apparent to those skilled in the art to produce imaging systems with different imaging capabilities that are more or less suitable for different applications.

[0052] As mentioned above, the imaging system can be configured to direct energy (e.g., light, electrons) to and / or scan the energy across a physical version of the sample, thereby producing an actual image of the physical version of the sample. In this way, the imaging system can be configured as a "real" imaging system rather than a "virtual" system. However, Figure 1 The storage medium (not shown) and computer system 102 shown in the figure can be configured as a "virtual" system. Specifically, the storage medium and computer system are not part of the imaging system 100 and do not have any capabilities for handling physical versions of the sample, but can be configured as a virtual inspector, a virtual metrology system, a virtual defect review tool, etc. that uses the stored detector output to perform inspection-like functions, a virtual metrology system, a virtual defect review tool, etc. Systems and methods configured as "virtual" systems 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 systems described herein may be further configured as described in these patents.

[0053] The system includes one or more computer systems, which may include any of the computer systems further described herein, the one or more computer systems configured to cluster the dies based on colors assigned to the dies in response to predicted defect densities in the dies on the sample determined from measurements performed on the sample, thereby generating initial clusters of dies. As used herein, the term "predicted" is synonymous with the terms "simulated," "estimated," and "approximately calculated," and "predicted defect density" is generally defined as a defect density likely to be detected by an imaging system on a sample and determined, for example, by modeling, from data unrelated to the defects on the sample.

[0054] In this manner, measurements can be performed on the sample by one of the imaging systems described herein or any other suitable metrology tool known in the art. In another embodiment, measurements are not performed for defect detection on the sample. For example, while any defects present on the sample may affect the measurements of the sample, measurements are not performed to detect defects on the sample. Instead, the measurements can include any suitable metrology type measurements, such as the thickness of one or more films formed on the sample, characteristics of patterned features (e.g., size, sidewall angle, roughness, shape, etc.), other roughness on the sample (e.g., film roughness), the superposition of features on one layer of the sample with features on another layer, and the variation of any such characteristics across the sample.

[0055] The measurements may include measurements of the layer of the sample on which the process is to be performed and / or measurements of another layer of the sample. For example, the measurements may include measurements of one or more structures formed on a layer underlying an overlying layer on which the process is to be performed, possibly in combination with measurements of one or more structures on the overlying layer.

[0056] Any of these measurements can be performed in any suitable manner known in the art and using any imaging system described herein or any other suitable metrology tool known in the art. Additionally, the measurements can be performed by the embodiments described herein. However, the measurements can be performed by another method or system and then acquired by the computer system of the embodiments described herein (e.g., from a shared computer-readable medium accessible by both the computer system and other computers or systems).

[0057] The measurements can then be used to predict the defect density in the die on the sample. Colors can then be assigned to the die based on the predicted defect density, and die clustering can then be performed for the die based on the assigned colors, thereby generating initial die clusters. Each of these steps can be performed as further described herein.

[0058] Although some steps may be described herein with respect to a die on a sample, the steps described herein may be performed for any other area on the sample. The area may correspond to a designed area on the sample other than the die, such as a field. However, the area may not have any correspondence with the design. For example, the wafer area may be artificially divided into a regular grid of tiles that may or may not be independent of the designed tiles formed on the sample. For example, the tiles may be larger than the repeating patterned area (e.g., the die) on the wafer, the tiles may be smaller than the repeating patterned area, the tiles may contain portions of two different repeating areas, etc. The tiles may also be non-overlapping (i.e., mutually exclusive) portions of the wafer. The steps described herein may then be performed for the tiles in the grid in the same manner as described herein with respect to the die. The terms "die," "tile," and "grid cell" are used interchangeably herein.

[0059] In one embodiment, a measurement is performed on a sample by a metrology tool, and the process comprises an inspection process. In this way, the embodiments described herein can use the metrology results to determine one or more parameters of the inspection. In other words, the embodiments described herein can be configured for metrology guided inspection (MGI). Figure 6 As shown in step 600 of , for example, the input to the computer system may be a grid (e.g., a die-level wafer map, e.g., Figure 3 ). Therefore, the grid may also be referred to as the MGI grid.

[0060] In one embodiment, the predicted defect density includes a defect density probability distribution. For example, metrology measurements may be performed at as many measurement points on the sample as desired. The metrology results may then be used to determine various distributions of metrology results across the sample, i.e., metrology measurements that vary spatially across the sample. In this manner, the metrology measurements may also be used to determine various distributions of defect density probabilities determined from metrology measurements in the spatial domain of the sample. For example, defect density probabilities may be determined at various locations on the sample from metrology measurements, and a distribution of defect density probabilities may then be determined based on the defect density probabilities that vary as a function of location. The predicted defect density probability distribution may then be used for clustering as described herein. Embodiments may combine the clustering and image processing steps further described herein to extract local maxima of the defect density probability distribution on the wafer.

[0061] In another embodiment, the predicted defect density is determined as a continuous value distribution across the sample. This is an important difference between the embodiments described herein and currently used methods and systems. For example, while there are several currently used clustering algorithms, they are not designed to extract local maxima from a continuous distribution (instead of clustering isolated points such as discrete locations of defects detected on a sample).

[0062] In some embodiments, the predicted defect density is generated by a machine learning (ML) model trained using training data, including ground-truth defect information for one or more training samples and measurements performed on the one or more training samples. For example, a defect probability distribution can be calculated using an ML model trained using inspection results as ground truth and one or more metrology data sources as features generated for one or more layers on one or more wafers, such as optical critical dimension (OCD) and patterned wafer geometry (PWG) data. In this way, after the ML model is trained using metrology measurements and ground-truth defect information for one or more training samples, the measurements described herein can be input into the ML model to generate a predicted defect density for the sample.

[0063] The ML model may have any suitable configuration known in the art and may be trained in any suitable manner known in the art. The computer system may or may not be configured to train the ML model. For example, the computer system or another method or system may be configured to train the ML model as described above. Training the ML model may include inputting metrology measurements of the training samples into the ML model and modifying one or more parameters of the ML model until the output of the ML model matches the ground truth defect density determined for the training samples (or exactly or within a predetermined tolerance). The modified parameters of the ML model may include any suitable parameters of the ML model, such as, for example, weights and biases of one or more convolutional layers.

[0064] The embodiments described herein may or may not include an ML model. For example, the ML model may or may not be included in one or more components (not shown) executed by a computer system. When the ML model is executed by the computer system, the computer system can generate a predicted defect density by inputting metrology results (such as those described further herein) into the ML model.

[0065] In this manner, embodiments described herein can generate predicted defect densities. However, embodiments described herein can simply obtain or receive predicted defect densities from another method or system that generates predicted defect densities. In other words, one method or system can be configured to generate predicted defect densities, and another method or system can be configured to use the predicted defect densities, as described herein.

[0066] The embodiments described herein may also be performed using predicted defect densities generated in any other manner. In other words, the embodiments described herein are not limited in the manner in which the predicted defect density is generated or obtained. For example, as long as the predicted defect density is correct, the embodiments described herein may be used to enhance the results by identifying the locations of local maxima in the predicted defect density.

[0067] In an additional embodiment, the computer system is configured to assign colors to the die by: determining a predicted defect density per die; determining an initial color for the die based on the predicted defect density per die; and determining a final color for the die through histogram equalization. For example, the computer system may use RGBA colors to plot the defect density and then use histogram equalization to enhance contrast. RGBA=RBGAlpha, where the alpha parameter is a number between 0.0 (fully transparent) and 1.0 (fully opaque). As shown in step 602, for example, the computer system may be configured to calculate a die-level prediction. Step 602 may include aggregating the die-level prediction by calculating the mean of all values that land on the same die. In this way, the input to this step may be predictions initially calculated in a dense grid (thousands of values per die), and after this step, each die may have a single summary value (e.g., mean). The computer system may also be configured to calculate a color (e.g., RGBA) vector from the histogram-equalized wafer map, as shown in step 604. Histogram equalization may be performed in any suitable manner known in the art.

[0068] Figure 3 The wafer map 300 shown in FIG is a grayscale version of a color die-level wafer map with defect density probabilities for four localized clusters (302, 304, 306, and 308) that are easily detected by the human eye. For example, it is difficult to convey in a black and white image how easily these clusters can be detected by a user in a color wafer map. Therefore, to give an example of how these clusters can be presented to a user in a color map, cluster 302 can be a bright yellow to yellow-green color, cluster 304 can be a yellow-green color and greener than cluster 302, cluster 306 can be a bright light green color, and cluster 308 can be a darker light green color than cluster 306. The remaining dies in the map can have a relatively dark blue color, with some dies surrounding clusters 302 and 304 having a slightly greenish shade of blue.

[0069] No clustering algorithm can be run based on die / grid location, so the clustering step of the embodiments described herein focuses on the color space (or defect density probability value) of each die in the grid. As further described herein, embodiments can individually identify Figure 3 All four clusters shown in . Figure 3 The wafer map shown in FIG and other wafer maps described herein are included merely as examples of wafer maps to aid in understanding the embodiments described herein. Obviously, the wafer map generated by any embodiment of the systems and methods described herein may vary depending on the sample for which the steps described herein are performed.

[0070] In one embodiment, clustering is not performed based on the location of the die on the sample. For example, the clustering step may include obtaining a die-level map of defect density probabilities and assigning each die to a cluster based on a color assigned to it and not any location information associated with the die. Each cluster identifies regions with substantially equal levels of predicted defect density (i.e., predicted defect density values that differ from each other by no more than a predetermined value). In another embodiment, clustering includes applying a K-means method to the colors assigned to the die. For example, Figure 6 As shown in step 606 of , the computer system can be configured to run KMeans using the color vector as a feature. The K-means (or KMeans) method can include any suitable K-means method known in the art that is performed in any suitable manner known in the art.

[0071] In another embodiment, clustering includes applying a K-means method to the colors assigned to the die and selecting the result of the K-means method with the highest silhouette score as one of the initial die clusters. The "silhouette score" is a measure of the "quality" of the clusters generated by the algorithm. A set of "good quality" clusters will include well-separated clusters, and the components (in this case, dies) included in each of the clusters will be relatively close together. In this way, the higher the silhouette score, the better the clustering generated by the algorithm. For example, the computer system can use KMeans on this color space to identify the cluster with the highest silhouette score. In one such example, as shown in step 608, the computer system can be configured to select the KMeans number cluster with the highest silhouette score.

[0072] As shown in step 610, the computer system can be configured to loop through steps 606 and 608 for a number of clusters. The number of clusters can be a predetermined number, for example, between 2 and 6 clusters. The predetermined number of clusters can be a target number of clusters, which can be determined or set by the user or in any other manner known in the art. For example, K-means supports specification of a cluster count (i.e., a target cluster count and / or a maximum cluster count). The predetermined number of clusters can be determined based on the expected number of clusters versus the time allowed to find them (more clusters = more time involved), and this number can be modified based on the data as needed.

[0073] Unlike currently used clustering algorithms based on distance, clustering in color space may not be able to resolve spatially separated clusters. This can be explained by the KMeans algorithm in Figure 4This can be seen in the clusters identified in wafer map 400 shown in

[15] . The Kmeans method has identified two clusters in the wafer map, illustrated by the differently shaded portions of the die. For example, one cluster has been identified for the die contained in shaded portions 402 and 404, and the other cluster has been identified for the die contained in shaded portions 406, 408, and 410. Thus, as can be seen from this figure, color-based clustering can result in some clusters containing die that are spatially separated from one another in the map. The cluster containing portions 406, 408, and 410 does not correspond to similar colors (probability) in the original wafer map, but should be identified as a different cluster if the goal is to identify local maxima for diversity sampling. Similarly, the cluster containing portions 402 and 404 in color space is actually two clusters in position space.

[0074] The computer system is also configured to analyze the initial die clusters in the location space to determine whether any of the initial die clusters contains two or more die clusters. In this way, each initial die cluster can be analyzed separately to determine whether it contains more than one die cluster. For example, Figure 6 As shown in step 612 of , the computer system can be configured to filter the grid cells (or dies) one cluster at a time by KMeans cluster ID. This step can be performed only to separate the grid cells into their initial die clusters for analysis purposes. In addition, the computer system can be configured to subsample the filtered grid cells, as shown in step 614. Subsampling can be performed randomly and can be performed when time is a factor. For example, if the algorithm used for the analysis step does not scale linearly and there is a time limit (seconds versus minutes of runtime per iteration), the subsampling step can be performed to help the analysis step be fast enough. Therefore, the subsampling step is optional.

[0075] In one embodiment, the analysis step includes applying a density-based spatial clustering with noise (DBScan) method to the initial die clustering. In this way, the computer system can use DBScan in the location space to identify clusters within each previously found cluster. For example, as shown in step 616, the computer system can be configured to run DBScan using grid positions (e.g., x and y positions) as features. Thus, the embodiments described herein may use KMeans in color space (which may be contrast enhanced) and then use DBScan in location space. The DBScan method may include any suitable data clustering algorithm or method known in the art. In addition, while DBScan may be a particularly advantageous way to perform the analysis step, analyzing the initial die clusters may be performed using any other location-space-based clustering algorithm or method.

[0076] In another embodiment, the analysis step includes: applying the DBScan method multiple times to the initial die clusters using different parameters of the DBScan method; identifying the results of the DBScan method applied using each of the different parameters with the highest silhouette score; and determining whether any of the initial die clusters contains two or more die clusters based on the identified results. For example, the computer system may select the DBScan parameters with the highest silhouette, as shown in step 618. As shown in step 620, the computer system may also loop through the DBScan parameters to thereby perform steps 616 and 618 using different DBScan parameters. These steps may be performed to brute force find the parameters of the DBScan clustering algorithm that give the best cluster (the cluster with the highest silhouette score) by iterating on the parameters of the DBScan clustering algorithm. Thus, for each parameter set, DBScan may be run, and the quality of the clusters may be calculated. The results generated for each parameter set may then be used to select the parameters with the best score, and the analysis step may then be performed using the results generated using the selected parameters.

[0077] Once the DBScan steps are performed for one cluster, the computer system may loop through the KMeans clustering, as shown in step 622, such that steps 612, 614, 616, 618, and 620 are performed for the next cluster identified by KMeans. In this way, steps 612, 614, 616, 618, and 620 may be performed independently for each cluster found by KMeans or other initial clustering performed in the color space. For example, if 6 clusters are found by running KMeans, the loop may be performed 6 times. In this way, these steps are performed for each KMeans cluster found in the color space so that each KMeans cluster can be segmented by DBScan in the position space as needed. Figure 4 In the example shown in , kidney-shaped cluster 402 and shrimp-shaped cluster 404 may be in the same cluster based on color, but in position space, they are far apart and therefore end up in two different clusters. Therefore, they may be in the same cluster after Kmeans and in different clusters after DBScan.

[0078] In one such embodiment, prior to the determining step, analyzing includes eliminating any of the two or more die clusters having a number of dies below a predetermined die count from the identification results of any of the initial die clusters. Figure 6As shown in step 624 of , the computer system may filter out clusters with relatively low grid cell counts. This step may be performed to avoid "background clustering." The predetermined die count or relatively low grid cell count may be determined in any suitable manner, such as based on input from a user or dynamically determined, for example, based on the number of dies or grid cells in the initial die cluster. In this manner, the results of the DBScan method with the highest silhouette score for the initial die cluster may be examined to determine whether any of the clusters in the results have a number of dies or grid cells below the predetermined die or grid cell count. Any of the clusters may then be eliminated from the die clusters in the DBScan results for the initial die cluster. In this manner, any relatively small clusters identified by the DBScan results may not be separated from the initial die cluster into a different die cluster or may be assigned to a background cluster or group that is not designated as a final die cluster. Eliminating substantially small clusters identified by DBScan may simplify the inspection process and improve its processing capacity with relatively little or no adverse effect on its defect detection performance.

[0079] In one embodiment, the clustering and analysis steps (e.g., KMeans and DBScan) are performed such that two or more die clusters contained in any of the initial die clusters and the initial die clusters correspond to local maxima in the predicted defect density. For example, the embodiments described herein adapt the clustering algorithm to find the local maximum cluster in the defect density probability distribution of the wafer. The embodiments described herein allow a computer system to resolve local maxima, such as Figure 5 Specifically, each original cluster is correctly identified. More specifically, the final die clusters 502, 504, 506, and 508 shown in the wafer map 500 correspond to Figure 3 300. Additionally, the final die clustering may include cluster 510, which is not obvious in wafer map 300 but is more obvious in the colored version of the defect density probability die-level map and which may be identified by the clustering and analysis steps described herein.

[0080] The computer system may be further configured to designate an initial die cluster that does not contain two or more die clusters and two or more die clusters contained in any of the initial die clusters as a final die cluster. In other words, if DBScan determines that an initial die cluster does not contain two or more die clusters, then the initial die cluster may be identified as a final die cluster. If DBScan determines that an initial die cluster contains two or more die clusters, then the two or more die clusters are identified as the final die cluster and the original initial die cluster is not identified as the final die cluster. Figure 6As shown in step 626 of , the computer system can assign cluster IDs to all grid cells. In this manner, the designation step can include assigning cluster IDs to all grid cells regardless of whether they are in their original initial die cluster or in a different die cluster identified by DBScan. Additionally, any non-clustered grid cells in the wafer map can also be assigned a cluster ID indicating that they are non-clustered or background cells. The cluster ID can be of any suitable form (e.g., numeric, alphanumeric, etc.) and format known in the art.

[0081] The computer system may also generate an output, as shown in step 628, which may include the final cluster IDs of the grid cells and any other information of the final die clusters generated by the embodiments described herein. The generated output may be in any suitable form or format known in the art that enables the output to be used when setting up a process, as further described herein. The computer system is also configured to store the information of the final die clusters for use when setting up a process to be performed on the sample (which may be performed as further described herein).

[0082] In one embodiment, the computer system is configured to determine the median predicted defect density of the final cluster of dies and eliminate any of the final cluster of dies that have a median predicted defect density below a predetermined threshold. For example, the nature of the clustering algorithms described herein means that each die can be part of a cluster regardless of whether it belongs to a local maximum (e.g., Figure 5 The clusters shown in Figure 5 In the embodiments described herein, one or more clusters, such as cluster 510 and the cluster of unshaded or patterned dies, are easily discarded by considering the median value of each cluster and discarding clusters with median values below a user-defined threshold.

[0083] In another embodiment, the computer system is configured to determine the median predicted defect density of the final die clusters and to sort the final die clusters based on the median predicted defect density. For example, the computer system may be configured to rank the final clusters by median cluster probability, e.g., from highest median cluster probability to lowest median cluster probability. The ranking step may be performed after assigning the final cluster ID. Thus, the final cluster ID may be ranked by the median predicted defect density. Figure 6 Perform this step after all other steps presented in . This ranking can be used in the setup process described further in this article.

[0084] In additional embodiments, the computer system is configured to set up the process by determining sampling for the process based on information from the final die clusters. For example, local maximum clusters identified by the embodiments described herein can be used to assist in diversity sampling. Thus, one advantage of the embodiments described herein is that they can be used to find local maxima in the defect density probability distribution on a wafer, which can be used to help find defects that are not part of the global maximum. In this way, the tool user can bias the inspection to explore areas that would otherwise be ignored due to stronger signals in other areas.

[0085] Thus, the diversity sampling that can be performed using the embodiments described herein differs fundamentally from previously used so-called diversity sampling. For example, previously developed diversity sampling methods operate on a population of defects to find the defects that differ most from one another. In this way, diversity sampling aims to find the most interesting and most distinct defects. This diversity sampling can be useful for several applications, such as defect discovery.

[0086] In contrast, the diversity sampling performed by or using the embodiments described herein is based on the diversity in predicted defect densities in different regions of the wafer. For example, a final die cluster may correspond to a local maximum in the predicted defect density, which is a primary objective of the embodiments described herein. Sampling for each of the final die clusters can then be independently determined based on its respective predicted defect density. For example, for a final die cluster expected to have a substantially high defect density, a relatively dense inspection sampling plan can be selected so that the final die cluster can be more thoroughly examined, for example, for diagnostic purposes. In contrast, for a final die cluster expected to have a substantially low defect density, a substantially dense inspection sampling plan can be selected because a less dense sampling plan may miss all or too many relatively sparse defects. In this way, the sampling plan independently selected or generated for each final die cluster can vary depending, at least in part, on the objective of the inspection (e.g., establishing a robust process versus defect discovery for a less robust process for potential process excursion monitoring).

[0087] The embodiments described herein can be executed as part of the MGI analysis flow during post-metrology processing. Additionally, the process can be configured to consume cluster IDs and use them for inspection sampling. In this manner, a primary application where the embodiments described herein can be used is for calculating defect probability distributions (from previous metrology step measurements) and combining them with inspection tool sampling planning capabilities to help devise better dynamic inspection plans.

[0088] Thus, the final die cluster can be used as a region of interest (CA) on the sample, either instead of or in addition to other CAs on the sample. For example, the parameters of the process used in different regions on the wafer corresponding to the final die cluster can be different in the same way that CA parameters can be different. In this way, one or more computer systems can be configured to designate some or all of the final die clusters as CAs for a process performed on the sample using the imaging system. Designating some or all of the final die clusters as CAs can include generating some CA identifiers for the final die clusters, which can include any alphanumeric or other suitable identifier known in the art. The output of this step can include the CA name and any other information generated for the CA, such as sample location, size, etc.

[0089] Instead of (or in addition to) sampling, one or more other parameters of the process may be independently selected for each final die cluster. For example, the parameters may include an indication of the final die cluster for which the process is to be performed, e.g., the final die cluster for which final die cluster inspection is to be performed versus the final die cluster for which final die cluster inspection is not to be performed. Other parameter values independently selected for each final die cluster may include different sensitivities for detecting defects on the sample in the inspection recipe. The different sensitivities may be controlled, for example, by one or more thresholds used to separate the image into pixels corresponding to defects versus pixels not corresponding to defects. Additionally or alternatively, the computer system may be configured to optimize the thresholds per final die cluster to achieve a desired number of defects. This sensitivity may be set using preset capture rates ("caprates") for different final die clusters. For example, an inspection target (e.g., a desired disturbed point rate + a desired defect capture rate per defect type) may be specified per final die cluster.

[0090] Although selecting inspection recipe parameters can be performed separately for different final die clusters, each of the inspection recipe parameters selected separately and independently for the final die clusters can be combined into a single inspection recipe that is executed on the sample. In addition, selecting inspection recipe parameters can be performed simultaneously for multiple final die clusters to thereby optimize the inspection recipe parameters for the multiple final die clusters.

[0091] In another embodiment, measurements and processes are performed on a sample using different tools. For example, in general, the systems described herein are configured for either metrology or inspection, but not both. Thus, measurements input to the steps described herein may be generated by one tool configured for metrology, and processes performed using the results of the steps described herein may be performed by a different tool configured for inspection. The steps described herein may be performed by any or none of these tools. For example, the embodiments described herein may be implemented by one or more computer systems (e.g., a computer system) contained in or coupled to a metrology tool and / or an inspection tool. Figure 1 The computer system 36 shown in Figure 2 ) is executed on a tool or by a computer system not included in or directly coupled to any of such tools (e.g., the computer system 124 shown in FIG. Figure 1 The computer system 102 shown in FIG. 1 is executed outside the tool.

[0092] In this manner, the systems described herein may or may not include an imaging system, e.g. Figure 1 and Figure 2

[0014] The imaging system shown in

[0014] is used. If the system includes an imaging system, the same computer system that generates the final die clustering information and, possibly, sets up a process based on that information can also be configured to subsequently perform the setup process on the sample. If the system does not include an imaging system, another system or method can be configured to perform the setup process on the sample. Additionally, one system can be configured to generate and store the final die clustering information as described herein, another system can be configured to set up a process using the stored information, and yet another system can be configured to perform the setup process.

[0093] The computer system may be configured to store information about the final die clusters for use when setting up a process to be performed on a sample (e.g., an inspection of the sample) and / or for performing the process on the sample once the process has been set up. For example, the computer system may be configured to store the information in a recipe or by generating a recipe for a process that will use the final die clusters. As used herein, the term "recipe" can generally be defined as a set of instructions that can be used by a tool to perform a process on a sample. In this manner, generating a recipe can include generating information about how to perform a process, which can then be used to generate instructions for performing the process. The information about the final die clusters stored by the computer system can include any information that can be used to identify and / or use the final die clusters (e.g., such as a file name and its stored location, and the file can include information about the final die clusters, such as a final die cluster ID, a final die cluster location, etc.).

[0094] The computer system may be configured to store the final die clustering information in any suitable computer-readable storage medium. The information may be stored along with any of the results described herein and may be stored in any manner known in the art. The storage medium may include any of the storage media described herein or any other suitable storage medium known in the art. After the information has been stored, it may be accessed from the storage medium and used by any of the method or system embodiments described herein, formatted for display to a user, used by another software module, method, or system, etc. For example, the embodiments described herein may generate an inspection recipe as described above. The inspection recipe may then be stored and used by the system or method (or another system or method) to inspect a sample to thereby generate information (e.g., defect information) about the sample.

[0095] The computer system and / or imaging system can be configured to use the results of one or more steps described herein to perform an inspection process on the sample. This inspection process can generate results for any defects detected on the sample, such as information about the bounding box of the detected defect (e.g., location, etc.), a detection score, information about the defect classification (e.g., a class label or ID, etc.), or any other suitable information known in the art. The defect results can be generated by the computer system and / or imaging system in any suitable manner. The defect results can be in any suitable form or format, such as a standard file type. The computer system and / or imaging system can generate and store the results so that the results can be used by the computer system and / or another system or method to perform one or more functions on the sample or another sample of the same type. For example, the information can be used by the computer system or another system or method to sample the defects for defect inspection or other analysis to determine the root cause of the defects, etc.

[0096] Functions that can be performed using this information also include, but are not limited to, modifying 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, the computer system can be configured to determine one or more changes to a process performed on a sample inspected as described herein and / or to a process that will be performed on the sample based on the detected defect. A change in a process can include any suitable change to one or more parameters of the process. The computer system preferably determines the change so that defects can be reduced or prevented on other samples on which the revised process is performed, 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, and so on. The computer system can determine such changes in any suitable manner known in the art. Such changes can also be determined using the results of other processes described herein.

[0097] 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 imaging system described herein may be coupled to the semiconductor manufacturing system, for example, via one or more common elements, 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.

[0098] Thus, as described herein, embodiments can be used to set up new processes or recipes. Embodiments can also be used to modify existing processes or recipes, whether they are used for a sample or a process or recipe generated for one sample and adapted for another sample.

[0099] Each embodiment of each system described above may be combined together into a single embodiment.

[0100] Another embodiment relates to a computer-implemented method for generating information for use in setting up a process to be performed on a sample. The method includes the clustering, analyzing, specifying, and storing steps described above. Each step of the method can be performed as further described herein. The method can also include any other steps that can be performed by the imaging system and / or computer system described herein. The clustering, analyzing, specifying, and storing steps are performed by one or more computer systems that can be configured according to any embodiment described herein. In addition, the method described above can be performed by any system embodiment described herein.

[0101] 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 generating information for use in setting up a process to be performed on a sample. Figure 7 One such embodiment is shown in FIG. Figure 7 , non-transitory computer-readable medium 700 includes program instructions 702 executable on a computer system 704. The computer-implemented method may include any of the steps of any method described herein.

[0102] Program instructions 702 implementing methods such as those described herein may be stored on a computer-readable medium 700. 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.

[0103] The program instructions may be implemented in any of a variety of ways, including procedural, component-based, 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), or other technologies or methodologies, as desired.

[0104] Computer system 704 may be configured according to any of the embodiments described herein.

[0105] In view of this description, it will be understood by those skilled in the art that other modifications and alternative embodiments of various aspects of the present invention are possible. For example, a method and system for generating information used when a process performed on a sample is provided. Therefore, this description should only be interpreted as illustrative and is for the purpose of teaching those skilled in the art to implement the general manner of the present invention. It should be understood that the form of the present invention shown and described herein will be considered as a current preferred embodiment. As will be understood by those skilled in the art after benefiting from this description of the present invention, elements and materials can replace the elements and materials illustrated and described herein, parts and processes can be reversed, and the specific features of the present invention can 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 generate information for use in setting up a process to be performed on a sample, comprising: One or more computer systems configured to: clustering the dies based on colors assigned to the dies in response to predicted defect densities in dies on the sample determined from measurements performed on the sample, thereby generating initial clusters of dies; analyzing the initial die clusters in location space to determine whether any of the initial die clusters contain two or more die clusters; designating the initial die cluster that does not contain two or more die clusters and the two or more die clusters contained in any of the initial die clusters as final die clusters; and Information of the final die clustering is stored for use in setting up a process to be performed on the sample. 2 . The system of claim 1 , wherein the predicted defect density comprises a defect density probability distribution. 3 . The system of claim 1 , wherein the predicted defect density is determined as a continuous value distribution across the sample. 4 . The system of claim 1 , wherein the clustering and analyzing are performed such that the initial die clustering and the two or more die clusters contained in any of the initial die clusterings correspond to local maxima in the predicted defect density.

5. The system of claim 1, wherein the measurements are performed on the sample by a metrology tool, and wherein the process comprises an inspection process. The system of claim 1 , wherein the measuring is not performed for defect detection on the sample.

7. The system of claim 1 , wherein the predicted defect density is generated by a machine learning model trained using training data, the training data comprising ground truth defect information for one or more training samples and the measurements performed on the one or more training samples.

8. The system of claim 1 , wherein the one or more computer systems are further configured to assign the colors to the dies by: determining a predicted defect density per die; determining an initial color for the die based on the predicted defect density per die; and determining a final color for the die by histogram equalization.

9. The system of claim 1, wherein the clustering comprises applying a K-means method to the colors assigned to the dies.

10. The system of claim 1, wherein the clustering comprises applying a K-means method to the colors assigned to the dies and selecting a result of the K-means method having a highest silhouette score as one of the initial die clusters.

11. The system of claim 1, wherein the clustering is not performed based on locations of the dies on the sample.

12. The system of claim 1, wherein the analyzing comprises applying a density-based spatial clustering of applications with noise (DBScan) method to the initial die clustering.

13. The system of claim 1, wherein the analyzing comprises: applying a density-based spatial clustering with application of noise (DBScan) method to the initial die clusters multiple times using different parameters of the DBScan method; A result of the DBScan method applied using each of the different parameters having a highest silhouette score is identified; and based on the identified results, a determination is made as to whether any of the initial die clusters contains two or more die clusters. 14 . The system of claim 13 , wherein prior to the determining, the analyzing further comprises eliminating any of the two or more die clusters having a number of dies below a predetermined die count from the identification result of any of the initial die clusters.

15. The system of claim 1, wherein the one or more computer systems are further configured to determine a median predicted defect density for the final cluster of dies and eliminate any of the final cluster of dies having the median predicted defect density below a predetermined threshold.

16. The system of claim 1, wherein the one or more computer systems are further configured to determine a median predicted defect density for the final clusters of dies and to rank the final clusters of dies based on the median predicted defect density.

17. The system of claim 1, wherein the one or more computer systems are further configured for setting up the process by determining a sampling of the process based on the information of the final die clustering.

18. The system of claim 1, wherein the measuring and the process are performed on the sample using different tools.

19. A non-transitory computer-readable medium storing program instructions executable on a computer system to perform a computer-implemented method for generating information for use in setting up a process to be performed on a sample, wherein the computer-implemented method comprises: clustering the dies based on colors assigned to the dies in response to predicted defect densities in dies on the sample determined from measurements performed on the sample, thereby generating initial clusters of dies; analyzing the initial die clusters in location space to determine whether any of the initial die clusters contain two or more die clusters; designating the initial die cluster that does not contain two or more die clusters and the two or more die clusters contained in any of the initial die clusters as final die clusters; and Information of the final die clustering is stored for use in setting up a process to be performed on the sample.

20. A computer-implemented method for generating information for use in setting up a process to be performed on a sample, comprising: clustering the dies based on colors assigned to the dies in response to predicted defect densities in dies on the sample determined from measurements performed on the sample, thereby generating initial clusters of dies; analyzing the initial die clusters in location space to determine whether any of the initial die clusters contain two or more die clusters; designating the initial die cluster that does not contain two or more die clusters and the two or more die clusters contained in any of the initial die clusters as final die clusters; and Information of the final die clustering is stored for use in setting up a process to be performed on the samples, wherein the clustering, analyzing, assigning, and storing are performed by one or more computer systems.

Images