Quantitative linear independent vector based method (QLIVBM) for image alignment

Abstract

Methods and systems for image alignment are provided. In one embodiment, a method includes individually aligning candidate alignment target images in an image generated for a sample to corresponding set images of set alignment targets selected to have mutually linearly independent vectors between a position of the set alignment targets and a reference position. The method also includes, for any of the candidate images successfully aligned to their corresponding set images, individually determining coordinates of the reference position from coordinates of the aligned candidate images and their corresponding mutually linearly independent vectors. In addition, the method includes determining final coordinates of the reference position in the image generated for the sample from the individually determined coordinates of the reference position.

Description

TECHNICAL FIELD

[0001] The present disclosure relates generally to methods and systems for image alignment to reference locations of a specimen, such as a wafer and reticle. Particular embodiments described herein relate to a quantitative linear independent vector based method (QLIVBM) for image alignment that is particularly suitable for non-optimal sites and avoids false alignment. BACKGROUND

[0002] The following description and examples do not admit to being prior art merely by virtue of their inclusion in this section.

[0003] Inspection processes are used during various steps in a semiconductor manufacturing process to detect defects on wafers to promote higher yield in the manufacturing process and thus higher profits. Inspection has always been an important part of fabricating semiconductor devices. However, as semiconductor devices shrink, inspection becomes even more important to successfully fabricating acceptable semiconductor devices because smaller defects can cause device failure.

[0004] One important step that, when performed incorrectly and inaccurately, can cause inspection failures is the alignment of test and reference images to each other. For example, many inspection processes subtract a reference image from a test image and then apply some defect detection method to the resulting image, often referred to as a difference image. If the images are not properly aligned to each other before subtraction, many differences between the images caused by poor alignment can be falsely detected as defects.

[0005] Accordingly, much research and development in inspection processes has been directed to developing accurate and efficient methods for aligning test and reference images to each other. One widely used existing method involves selecting a single alignment site for wafer alignment in a setup phase. The setup phase also includes recording the alignment site kernel image in a recipe for the inspection process.

[0006] During run time, such methods include aligning the single alignment site kernel and the run time alignment site image stream by one of the one or more alignment algorithms. The method can also include determining whether the kernel image matches the run time alignment site image. If the kernel and run time alignment site images do not match, the method can include determining whether all alignment algorithms have been tried. If not, another alignment algorithm can be used to perform alignment on the same image. If the kernel and run time images still do not match and the method determines that all alignment algorithms have been tried, the method can abort the alignment attempt. If the kernel and run time images match, the method can determine that alignment has been achieved.

[0007] However, there are several disadvantages to many current methods used for image alignment. For example, in the above-described methods, false alignments and low die corner accuracy determinations are possible. Some attempts to prevent this include trying to make the alignment sites of the on-line run-time wafer look more like the alignment sites saved in the kernel during setup. Algorithm development can also involve trying to make the algorithm less sensitive (where the algorithm does more) or more sensitive (so the algorithm does less). For example, an "do more" algorithm can do more image processing (such as highlighting certain patterns, increasing contrast, filtering out certain frequencies, extracting certain patterns, etc.) to make the run-time alignment site images look more like the setup alignment site images, thereby making the tool less sensitive to differences between the run-time and setup images and increasing the chances of achieving alignment. However, a less sensitive algorithm can result in false alignments and pre-map and pixel-to-design alignment (PDA) aborts. In one such example, all of the above-described image processing designed to make the algorithm less sensitive to changes in the alignment site images can actually change the run-time alignment site images so much that they align to the wrong setup alignment site images, thereby causing false alignments and pre-map and PDA aborts. In the same way, a more sensitive algorithm can cause failed alignment aborts due to changes between the run-time and setup alignment target images that are not corrected by the algorithm. Thus, current algorithm development has a dilemma with respect to false alignments and failed alignments.

[0008] In summary, if the alignment algorithm is less sensitive, then the alignment sites tend to match to the wrong locations, which causes false alignments and subsequent pre-map and PDA aborts. However, if the alignment algorithm is too sensitive, then the algorithm can frequently fail due to process changes, which causes immediate alignment aborts. There is also currently no solution for aligning run-time wafers with missing or damaged alignment sites.

[0009] Therefore, it would be advantageous to develop systems and methods for aligning images of a sample to reference locations in an image that do not have one or more of the above-described disadvantages. SUMMARY

[0010] The following description of various embodiments should in no way be interpreted as a limitation of the scope of the accompanying claims.

[0011] One embodiment relates to a system configured for image alignment. The system includes an imaging subsystem configured to generate images of a sample. The system also includes a computer subsystem configured for individually aligning candidate alignment targets in the images generated for the sample to corresponding setup images of setup alignment targets selected to have a mutual linearly independent vector between a position of the setup alignment target and a reference position. For each of one or more of the candidate alignment target images that successfully align to one of their corresponding setup images, the computer subsystem is configured for individually determining a coordinate for the reference position in the images generated for the sample from a coordinate of each of the one or more candidate alignment target images and its corresponding one of the mutual linearly independent vector. In addition, the computer subsystem is configured for determining a final coordinate for the reference position in the images generated for the sample from the individually determined coordinates of the reference position. The system can be further configured as described herein.

[0012] Another embodiment relates to a computer-implemented method for image alignment. The method includes the above-described individually aligning, individually determining coordinates, and determining a final coordinate steps, which are performed by a computer subsystem. Each of the steps of the method can be performed as further described herein. The method can include any other steps of any other method described herein. The method can be performed by any of the systems described herein.

[0013] Another embodiment relates to a non-transitory computer-readable medium storing program instructions executable on a computer system for performing a computer-implemented method for image alignment. The computer-implemented method includes the steps of the above-described method. 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 are executable can include any other steps of any other method described herein. BRIEF DESCRIPTION OF DRAWINGS

[0014] Those skilled in the art will appreciate the additional advantages of the application upon reading the following detailed description of the preferred embodiments, taken in conjunction with the accompanying drawings, in which:

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

[0016] Figure 3 is a flowchart of an embodiment of steps that can be performed for a setup phase of image alignment;

[0017] Figure 4 is a flowchart illustrating an embodiment of steps that can be performed for image alignment of a run-time phase;

[0018] Figure 5 is a schematic diagram of a plan view illustrating an embodiment of setting alignment targets selected to have a mutual linear independent vector between a position of the setting alignment target and a reference position;

[0019] Figure 6 is a table illustrating an embodiment of data that can be recorded for setting alignment targets during a setting phase;

[0020] Figure 7 is a schematic diagram of a plan view illustrating an embodiment of candidate alignment target images in images produced for a sample that can be individually aligned to Figure 5 setting alignment targets shown in FIG. 1 1 ;

[0021] Figure 8 is a table illustrating an embodiment of data that can be recorded for candidate alignment target images during a run-time phase;

[0022] Figure 9 is a schematic diagram illustrating some advantages of the embodiments described herein for algorithm development and alignment success; and

[0023] Figure 10 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 application can be susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and are herein described in detail. The drawings can not be to scale. It should be understood that the drawings and detailed description thereto are not intended to limit the application to the particular form disclosed, but on the contrary, the intention is to cover all modifications, equivalents and alternatives falling within the spirit and scope of the application as defined by the appended claims. DETAILED DESCRIPTION

[0025] Turning now to the drawings, it is noted that the figures are not drawn to scale. In particular, the proportions of some of the elements of the figures are greatly exaggerated to emphasize the characteristics of the elements. It is also noted that the figures are not drawn to the same scale. Elements shown in more than one figure that can be similarly configured have been indicated using the same element number. Unless otherwise indicated herein, any element described and shown can comprise any suitable commercially available element.

[0026] In general, the embodiments described herein are configured for image alignment. Certain embodiments relate to a method based on quantized linear independent vectors (QLIVBM), which is particularly useful for non-optimal site alignment solutions and avoiding false alignments.

[0027] The measurement of the average number of wafers run between interruptions (MWBI) value is used to assess the reliability and availability of the inspection platform during run time, which can have significant financial impact. Alignment failures are the major root cause of most interruptions and they can severely impact the MWBI value. Therefore, it becomes less likely that the inspection platform will be used for high volume research and development fabs utilization and global high volume manufacturing applications, as the probability of alignment failures increases.

[0028] In current usage methods of current wafer inspection tools, if the alignment algorithm is not very sensitive, then the alignment sites tend to match to the wrong locations, which causes false alignments and subsequent pre-mapping and pixel to design alignment (PDA) interruptions. Additionally, if the alignment algorithm is more sensitive, then the alignment frequently fails due to process changes, which causes the run time images to change from the set images, thereby causing immediate alignment interruptions. There also does not exist a solution to align run time wafers with missing or damaged alignment sites.

[0029] The QLIVBM embodiments described herein have several important advantages over current usage methods. One such advantage is that they can reduce and even eliminate false alignments during run time on a wafer inspection platform. Another advantage is that they can accomplish a substantially high confidence wafer alignment and die corner determination during run time on a wafer inspection platform, even if the alignment sites are not optimal, including severely damaged or even completely missing sites.

[0030] In some embodiments, the sample is a wafer. The wafer can include any wafer known in semiconductor technology. Although some embodiments can be described herein with respect to one or several wafers, embodiments are not limited to the samples with which they can be used. For example, the embodiments described herein can be used for samples such as reticles, flat panels, personal computer (PC) boards, and other semiconductor samples.

[0031] One embodiment relates to a system configured for image alignment. Figure 1 One embodiment of this system is shown in FIG. 1. The system includes an imaging subsystem 100 configured to generate images of a sample. The imaging subsystem includes and / or is coupled to a computer subsystem, such as computer subsystem 36 and / or one or more computer systems 102.

[0032] The terms "imaging system" and "imaging subsystem" are used interchangeably herein and generally refer to hardware configured to produce an image of a sample. Generally, the imaging subsystems described herein include at least an energy source and a detector. The energy source is configured to produce energy that is directed to the sample. The detector is configured to detect energy from the sample and produce an output in response to the detected energy.

[0033] In optical-based imaging subsystems, the energy directed to the sample includes light, and the energy detected from the sample includes light. For example, as shown in Figure 1 The imaging subsystem includes an illumination subsystem configured to direct light to the sample 14. The illumination subsystem includes at least one light source, such as the light source 16. The illumination subsystem is configured to direct light to the sample at one or more angles of incidence, which can include one or more oblique angles and / or one or more normal angles. For example, as shown in Figure 1 The light from the light source 16 is directed through the optical element 18 and then the lens 20 to the sample 14 at an oblique angle of incidence. The oblique angle of incidence can include any suitable oblique angle of incidence, which can vary, for example, depending on the characteristics of the sample and the defects to be detected on the sample, the characteristics of the sample to be measured.

[0034] The illumination subsystem can be configured to direct light to the sample at different angles of incidence. For example, the imaging subsystem can be configured to alter one or more parameters of one or more elements of the illumination subsystem so that light can be directed to the sample at angles of incidence that are different from the angles of incidence shown in Figure 1 In one such example, the imaging subsystem can be configured to move the light source 16, the optical element 18, and the lens 20 so that light is directed to the sample at different oblique angles of incidence or normal (or near normal) angles of incidence. The illumination subsystem can have any other suitable configuration known in the art for sequentially or simultaneously directing light to the sample at one or more angles of incidence.

[0035] The illumination subsystem can also be configured to direct light having different characteristics to the sample. For example, the optical element 18 can be configured as a spectral filter and the properties of the spectral filter can be varied in various different ways (e.g., by swapping one spectral filter out with another spectral filter) so that different wavelengths of light can be directed to the sample at different times.

[0036] The light source 16 can include a broadband plasma (BBP) light source. In this way, the light produced by the light source and directed to the sample can include broadband light. However, the light source can include any other suitable light source, such as any suitable laser known in the art configured to produce light at any suitable wavelength. Additionally, the laser can be configured to produce monochromatic or near-monochromatic light. In this way, the laser can be a narrowband laser. The light source can also include a polychromatic light source that produces light at multiple discrete wavelengths or wavebands.

[0037] Light from optical element 18 can be focused onto sample 14 by lens 20. Although lens 20 is shown in Figure 1 as a single refractive optical element, in practice, lens 20 can include several refractive and / or reflective optical elements that combine to focus light from the optical element to the sample. Figure 1 The illumination subsystem shown in and described herein can 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 can include any such suitable optical elements known in the art. Additionally, the system can be configured to alter one or more elements of the illumination subsystem based on the type of illumination used for imaging.

[0038] The imaging subsystem can also include a scanning subsystem configured to change the position on the sample at which light is directed to and detected from and can cause light to be scanned over the sample. For example, the imaging subsystem can include a stage 22 on which sample 14 is disposed during imaging. The scanning subsystem can include any suitable mechanical and / or robotic assembly, including stage 22, which can be configured to move the sample so that light can be directed to and detected from different positions on the sample. Additionally or alternatively, the imaging subsystem can be configured so that one or more optical elements of the imaging subsystem perform some light scanning over the sample so that light can be directed to and detected from different positions on the sample. The light can be scanned over the sample in any suitable manner, such as along a serpentine path or along a spiral path.

[0039] The imaging subsystem includes one or more detection channels. At least one of the detection channels includes a detector configured to detect light from the sample due to the system illuminating the sample and generate an output in response to the detected light. Figure 1 The imaging subsystem shown in includes two detection channels, one formed by collector 24, element 26, and detector 28, and the other formed by collector 30, element 32, and detector 34. The 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 of the detection channels can be configured to detect another type of light from the sample, such as reflected light.

[0040] In Figure 1In this embodiment, the two detection channels are shown positioned within the plane of the paper, and the illumination subsystem is also shown positioned within the plane of the paper. Therefore, in this embodiment, the two detection channels are positioned (e.g., centered) within the incident plane. However, one or more of the detection channels may be positioned outside the incident plane. For example, the detection channel formed by the light collector 30, element 32, and detector 34 may be configured to collect and detect light scattered from the incident plane. Therefore, 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 incident plane.

[0041] although Figure 1 An embodiment of an imaging subsystem comprising two detection channels is shown, but the imaging subsystem may comprise a different number of detection channels (e.g., only one detection channel or two or more detection channels). The detection channel formed by the light collector 30, element 32, and detector 34 may form one of the side channels described above, and the imaging subsystem may include an additional detection channel (not shown) formed as another side channel positioned on the opposite side of the incident surface. Thus, the imaging subsystem may include a detection channel comprising the light collector 24, element 26, and detector 28 and centered within the incident surface and configured to collect and detect light at a scattering angle oriented toward or near normal to the sample surface. Therefore, this detection channel may be commonly referred to as the "top" channel, and the imaging subsystem may also include two or more side channels configured as described above. Thus, the imaging subsystem may include at least three channels (i.e., one top channel and two side channels), and each of the at least three channels is configured to collect light at a scattering angle different from that of each of the other light collectors.

[0042] As further described above, one or more detection channels can be configured to detect scattered light. Therefore, Figure 1 The imaging subsystem shown herein can be configured for dark-field (DF) imaging. However, the imaging subsystem may also, or alternatively, include a detection channel configured for bright-field (BF) imaging. Therefore, the imaging subsystem described herein can be configured for DF-only, BF-only, or both DF and BF imaging. Although each condenser... Figure 1 It is shown as a single refractive optical element, but each light collector may contain refractive optical elements and / or reflective optical elements.

[0043] The one or more detection channels can include any suitable detector known in the art, such as a photomultiplier tube (PMT), a charge-coupled device (CDD), and a time delay integration (TDI) camera. The detectors can also include non-imaging detectors or imaging detectors. If the detectors are non-imaging detectors, each detector can be configured to detect a particular characteristic of light (e.g., intensity), but not configured to detect certain characteristics that vary with position within an imaging plane. Thus, the output generated by each detector in each detection channel can be a signal or data, but not an image signal or image data. In such examples, the computer system can be configured to generate an image of the sample from the non-imaging output of the detectors. However, in other examples, the detectors can be configured as imaging detectors configured to generate imaging signals or image data. Accordingly, the imaging subsystem can be configured to generate images in several ways.

[0044] The computer subsystem 36 can be coupled to the detectors of the imaging subsystem in any suitable manner (e.g., via one or more transmission media, which can include “wired” and / or “wireless” transmission media) such that the computer subsystem can receive the output generated by the detectors. The computer subsystem 36 can be configured to perform several functions using the output of the detectors, as described further herein. The computer subsystem 36 can be further configured as described herein.

[0045] The computer subsystem 36 (and other computer subsystems described herein) can also be referred to herein as a computer system. Each of the computer subsystems or systems described herein can take various forms, including a personal computer system, image computer, mainframe computer system, workstation, network appliance, Internet appliance, or other device. In general, the term “computer system” can be broadly defined to encompass any device having one or more processors, which executes instructions from a memory medium. The computer subsystem or system can also include any suitable processor known in the art, such as a parallel processor. Additionally, the computer subsystem or system can include a computer platform with high speed processing and software as a standalone or networked appliance.

[0046] If the system includes more than one computer system, the different computer systems can be coupled to each other such that images, data, information, instructions, etc. can be sent between the computer systems. For example, the computer subsystem 36 can be coupled to the computer system 102 by any suitable transmission medium (which can include any suitable wired and / or wireless transmission medium known in the art) as shown by the dashed line in FIG. 1. Two or more such computer systems can also be operatively coupled by a shared computer-readable storage medium (not shown). Figure 1

[0047] In an electron beam imaging subsystem, the energy directed to the sample includes electrons, and the energy detected from the sample includes electrons. In​Figure 2 In one embodiment shown, the imaging subsystem includes an electron column 122, and the system includes a computer subsystem 124 coupled to the imaging subsystem. The computer subsystem 124 can be configured as described above. Additionally, this imaging subsystem can be configured as described above and... Figure 1 It is coupled to one or more computer systems in the same way as shown in the diagram.

[0048] For example Figure 2 As shown, the electron column includes an electron beam source 126 configured to generate electrons focused onto the sample 128 by one or more elements 130. The electron beam source may include, for example, a cathode source or an emitter tip, and the one or more elements 130 may include, for example, a gun lens, an anode, a beam confinement aperture, a gate valve, a beam selection aperture, an objective lens, and a scanning subsystem, all of which may include any such suitable elements known in the art.

[0049] Electrons returning from the sample (e.g., secondary electrons) can be focused onto detector 134 by one or more elements 132. One or more elements 132 may include, for example, a scanning subsystem, which may be the same scanning subsystem included in element 130.

[0050] The electron column may include any other suitable element known in the art. Additionally, the electron column may be further configured as described below: U.S. Patent No. 8,664,594, issued April 4, 2014 by Jiang et al.; U.S. Patent No. 8,692,204, issued April 8, 2014 by Kojima et al.; U.S. Patent No. 8,698,093, issued April 15, 2014 by Gubbens et al.; and U.S. Patent No. 8,716,662, issued May 6, 2014 by MacDonald et al., all of which are incorporated herein by reference as if stated herein in their entirety.

[0051] Despite the electron column Figure 2 The diagram illustrates a configuration where electrons are guided to the sample at an oblique angle of incidence and scattered from the sample at another oblique angle. However, the electron beam can be guided to and scattered from the sample at any suitable angle. Furthermore, the electron beam imaging subsystem can be configured to use multiple modes to produce the sample output, as further described herein (e.g., using different illumination angles, collection angles, etc.). These multiple modes of the electron beam imaging subsystem can differ in the output parameters produced by any of the imaging subsystem's modes.

[0052] The computer subsystem 124 can be coupled to the detector 134, as described above. The detector can detect electrons returning from the surface of the sample, thereby forming an electron beam image (or other output) of the sample. The electron beam image can include any suitable electron beam image. The computer subsystem 124 can be configured to perform any of the steps described herein. Including Figure 2 The system of the imaging subsystem shown in FIG. 1 can be further configured as described herein.

[0053] Provided herein Figure 1 and 2 To generally describe the configuration of the imaging subsystem that can be included in the system embodiments described herein. It is clear that the imaging subsystem arrangements described herein can be altered to optimize the performance of the imaging subsystem that is normally performed when designing a commercial imaging system. In addition, the systems described herein can use existing imaging subsystem implementations (e.g., by adding the functionality described herein to an existing inspection system), such as tools available from KLA Corporation (Milpitas, CA). For some such systems, the methods described herein can be provided as an optional function of the imaging subsystem (e.g., in addition to other functions of the imaging system). Alternatively, the imaging subsystem described herein can be designed "from scratch" to provide a completely new system.

[0054] Although the imaging subsystem is described above as a light or electron beam imaging subsystem, the imaging subsystem can be an ion beam imaging subsystem. Such an imaging subsystem can be configured as shown in FIG. 1, except that the electron beam source can be replaced by any suitable ion beam source known in the art. In addition, the imaging subsystem can include any other suitable ion beam system, such as ion beam systems included in commercially available focused ion beam (FIB) systems, helium ion microscope (HIM) systems, and secondary ion mass spectrometers (SIMS) systems. Figure 2

[0055] The imaging subsystem can be configured to use multiple modes to generate an output, such as an image, of the sample. Generally, a "mode" is defined by the values of the parameters of the imaging subsystem used to generate an image of the sample (or to generate an output of an image of the sample). Thus, modes can differ in the values of at least one parameter of the imaging subsystem (in addition to the location on the sample where the output is generated). For example, modes can differ in any one or more alterable parameters of the imaging subsystem (such as illumination polarization, angle, wavelength, etc., detection polarization, angle, wavelength, etc.). The imaging subsystem can be configured to use different modes to scan the sample in the same scan or in different scans, e.g., depending on the ability to simultaneously use multiple modes to scan the sample.

[0056] ​In a similar manner, the electron beam subsystem can be configured to produce images using two or more modes, the two or more modes being defined by values of parameters of the electron beam subsystem used to produce the images of the specimen. Thus, the modes can differ in values of at least one electron beam parameter of the electron beam subsystem. For example, different modes can use different angles of incidence for illumination.

[0057] The imaging subsystems described herein can be configured as inspection systems, metrology systems, and / or defect review systems. For example, Figure 1 and 2 Embodiments of the imaging subsystems shown in FIGS. 1-3 can be modified by one or more parameters to provide different imaging capabilities depending on the application for which they will be used. In one such example, the imaging subsystem can be configured to have a higher resolution when it is used for metrology rather than for inspection. In other words, Figure 1 and 2 Embodiments of the imaging subsystems shown in FIGS. 1-3 describe some general and various configurations of the imaging subsystems, which can be adapted in several ways as will be apparent to those skilled in the art to produce systems having different imaging capabilities that are more or less suitable for different applications.

[0058] In this manner, the imaging subsystems can be configured for producing outputs that are suitable for detecting or re-detecting defects on a specimen (in the case of an inspection system or defect review system, respectively) and suitable for measuring one or more characteristics of a specimen (in the case of a metrology system). In an inspection system, Figure 1 The computer subsystem 36 shown in FIG. 1 can be configured for detecting defects on the specimen 14 by applying a defect detection method or algorithm to the output produced by the detector 28. In a defect review system, Figure 2 The computer subsystem 124 shown in FIG. 3 can be configured for re-detecting defects on the specimen 128 by a defect re-detection method applied to the output produced by the detector 134 and can determine additional information for the re-detected defects using the output produced by the detector. In a metrology system, Figure 1 The computer subsystem 36 shown in FIG. 1 can be configured for determining one or more characteristics of the specimen 14 using the output produced by the detector 28 and / or 34. The system can be further configured for detecting or re-detecting defects on the specimen, determining characteristics of the specimen, determining other information for the specimen, etc., as described further herein.

[0059] As mentioned above, the imaging subsystems are configured for scanning energy (e.g., light, electrons, etc.) on a physical version of a specimen, thereby producing an output of the physical version of the specimen. In this manner, the imaging subsystems can be configured as "real" subsystems rather than "virtual" subsystems. However, the storage media (not shown) and Figure 1The computer system 102 shown in the middle can be configured as a "virtual" system. In particular, the storage media and computer system can be configured as the "virtual" imaging system described in commonly-assigned U.S. Patent No. 8,126,255 to Bhaskar et al., issued February 28, 2012, and commonly-assigned U.S. Patent No. 9,222,895 to Duffy et al., issued December 29, 2015, which are incorporated by reference as if fully set forth herein. The embodiments described herein can be further configured as described in these patents.

[0060] The system includes a computer subsystem, which can include any of the above computer subsystems or any configuration of the system. The computer subsystem is configured for individually aligning a candidate alignment target image (also referred to herein simply as a "candidate image") in an image generated for a specimen to a corresponding setup image in which a setup alignment target is disposed, the setup alignment target selected to have a mutual linearly independent vector between a position of the setup alignment target and a reference position. In this way, the QLIVBM utilizes a group of unique kernel images and their corresponding linearly independent vectors determined during alignment setup and applies them during run-time alignment to greatly reduce the likelihood of misalignment and failed alignment sites on a wafer inspection platform (and other platforms described herein).

[0061] The computer subsystem can be configured for selecting the setup alignment target, as shown in step 300 of Figure 3 Alternatively, another method or system can select the setup alignment target, and the computer subsystem can acquire information for the setup alignment target from a storage media in which the information has been stored. The information can include the setup image of the setup alignment target, its corresponding mutual linearly independent vector, and any other information generated for the setup alignment target, such as the information shown in step 300 of Figure 6 The computer subsystem or other method or system can select the setup alignment target as further described herein or in any other manner known in the art.

[0062] In one embodiment, the set alignment targets include at least two alignment targets selected to include unique patterns relative to each other. For example, the set 300 can include kernel pre-slices targeted to have unique candidates of mutual linearly independent vectors between the location of the candidate and the reference location. The kernel pre-slices can generally include separating the set image into N unique candidate kernel images or N candidates. The terms "candidate alignment target" and "candidate" can be used interchangeably herein. In this way, the set alignment targets described herein are preferably unique relative to each other both in the pattern they contain and their linearly independent vectors to the reference location. Such set alignment targets can be identified from the design of the specimen (e.g., a design image or other design data), which can include any design information known in the art for the specimen described herein. Additionally or alternatively, the set alignment targets can be selected from one or more specimen images generated by the imaging subsystem for the specimen. The patterns can be unique relative to each other in any suitable manner (e.g., by pattern shape, orientation, size, number, periodicity, etc.).

[0063] The set stage can also include collecting set images (also referred to herein as "kernel images" and "kernels") of the set alignment targets, as shown in step 302. In this step, a set image can be collected for each set alignment target. Collecting the set image can include extracting an image of the set alignment target from a specimen image generated for the specimen. When the set alignment target is selected from a specimen image generated by the imaging subsystem, the same image can be used to extract the set alignment target image. When the set alignment target is selected from design information for the specimen, collecting the set image can include capturing an image of the specimen using the imaging subsystem at a location where the set alignment target is formed on the specimen. The location of the set alignment target on the specimen can be determined in any suitable manner known in the art. The size and shape of the candidate kernel can vary for several reasons (e.g., the size can be increased to increase the uniqueness of the candidate), and the set images described herein can have any suitable size and shape. Collecting the set image can also include storing the set image in a computer-readable storage medium (as described further herein), such that the set image can be accessed and used during a process performed on the specimen.

[0064] The set stage can further include calculating a mutual linearly independent vector for each selected set alignment target (as shown in step 304), which can be performed in any suitable manner known in the art. Generally, the mutual linearly independent vector can be determined from the center of the set alignment target to the reference location. The mutual linearly independent vector (as well as the location of the candidate image and the reference location) can be determined in any suitable coordinates (e.g., coordinates relative to the specimen, coordinates relative to the image, coordinates relative to the imaging subsystem, etc.). Such coordinates can also be used to translate the vectors and locations to other coordinate systems, such as design coordinates relative to the design of the specimen.

[0065] In another embodiment, each set alignment target is positioned within a die on the sample, and the reference location is a corner of the die. For example, as shown in Figure 5 Die 500 includes a pattern 502 formed therein, which is a pattern formed on the sample. Pattern 502 is shown here merely for further understanding of the present invention and does not represent any actual pattern that can be formed on the sample described herein. Die 500 also includes a die corner 504, which can be used as a reference location in the embodiments described herein. In this way, the die corner can not be positioned along the outer periphery of the die, but can be an interior corner, as shown in Figure 5 as further shown in Figure 5 In die 500, eight set alignment targets 506 have been selected, each having a mutual linear independent vector 508 between the center thereof and the reference location (die corner 504). In other words, as shown in Figure 5 the mutual linear independent vector between the set alignment targets and the reference location is different for each set alignment target. The mutual linear independent vector can be determined by the computer subsystem in any suitable manner known in the art or can be a user input during setup. As shown in Figure 5 Each set alignment target contains a unique pattern compared to the pattern in each other set alignment target, as shown in

[0066] Although eight set alignment targets are shown in Figure 5 the computer subsystem can select and / or use any suitable number of set alignment targets. For example, the minimum number of set alignment targets can be two, and the ideal number of set alignment targets can be between five and ten. The number of set alignment targets selected can vary, for example, depending on the number of unique patterns that can be used in the sample design, the expected degree of information of the patterns formed on the sample in the manufacturing process, any handling considerations that can make fewer set alignment targets more attractive, and the like.

[0067] A die can be any suitable die on a specimen, and although embodiments can be described herein with respect to reference locations being corners of a die, the reference locations can be any other suitable reference locations in any suitable area on a specimen. For example, the reference locations can be reference locations of a field of a design of a specimen. Additionally, embodiments described herein can be used for any type of alignment that can be performed using the reference locations in one of the processes described herein. For example, the alignment can be performed to align to a specimen location (instead of a design of a specimen). However, the alignment can also be performed for alignment to a design-like application, such as a pixel-to-design alignment (PDA), where a specimen image is aligned to design coordinates of a design formed on the specimen. The alignment can also be performed for a single image, multiple images, a single swath or multiple swaths of an image, an image of only a portion of a specimen, an image generated for an entire specimen (or an entire imaged area of a specimen), and so forth.

[0068] The setup stage can include outputting information of the selected set alignment targets, so that the information can be used during run time. As shown in Figure 6 , the information of the selected set alignment targets can be stored and output in a table, but the results can be stored in any other suitable file or format. The information can be output and stored in a computer readable storage medium, as further described herein.

[0069] The setup can also be performed once using a setup wafer (e.g., a first wafer that can be used for a given design and process step (or level)), and then the results of the setup stage can be reused for all other wafers formed using the same design and process step (or level). If there is any change in the design (including any materials used in the design), manufacturing conditions, or anything else that will change how the wafer appears in images produced by the imaging subsystem, the setup stage will be performed again for another wafer. In this way, the setup stage can be performed once for a specimen and then the same setup information can be reused for all specimens of the same type, unless or until there is a change in the specimen. In other words, the setup stage described herein does not have to be performed for every specimen for which alignment will be performed in run time.

[0070] In the table shown in Figure 6 , different set alignment targets can be assigned a candidate number (as shown in the first column), which can be used as their set alignment target ID. The candidate numbers shown in the first column of the table correspond to the mutually linearly independent vector indices of the set alignment targets shown in Figure 5 , which are also shown in the last column in Figure 6 . In other words, candidate 1 in Figure 6 is the last column in Figure 6 , shown for candidate 1. In other words, Figure 5 ​the set-up alignment targets shown in Figure 6 Candidate 2 in Figure 6 the mutually linearly independent vectors shown in the last column of Figure 5 the set-up alignment targets shown in

[0071] The results of the set-up phase can also include candidate kernels set-up alignment targets, as shown in the second column of the table in Figure 6 These candidate kernel images can be extracted from the sample images in any suitable manner known in the art. Additionally, the candidate kernel images can have any suitable size known in the art.

[0072] The results of the set-up phase can further include the coordinates of the kernel centers of the candidates determined during set-up, as shown in the third column of the table in Figure 6 In other words, (Xs1, Ys1) is the coordinate of the center of the kernel image of set-up alignment target candidate 1, and the mutually linearly independent vectors are the vectors between that coordinate and the coordinates of the die corners; (Xs2, Ys2) is the coordinate of the center of the kernel image of set-up alignment target candidate 2, and the mutually linearly independent vectors are the vectors between that coordinate and the coordinates of the die corners; and so on.

[0073] The computer subsystem can individually align the candidate images in the images generated for the sample to the corresponding set-up images of the set-up alignment targets in the run-time phase, as shown in Figure 4 For example, as shown in step 400, the imaging subsystem can generate the run-time sample images, which can be performed as further described herein. In step 402, the computer subsystem can retrieve the candidate alignment target images (from the run-time sample images) and the set-up images of the current candidate (the location where they are stored). The current candidate can be any set-up alignment target, such as one of the set-up alignment targets identified to be used first.

[0074] In step 404, the computer subsystem can perform image alignment on the current candidate using a current alignment method. If there is only one alignment method available for use in the process, the current alignment method can be this one alignment method. If there are multiple alignment methods available for use, the current alignment method can be any alignment method (e.g., one of the alignment methods identified to be used first, one of the alignment methods randomly selected by the computer subsystem, etc.). In this way, step 404 can include inputting the core candidate image and the runtime candidate image into one of a plurality of alignment methods or algorithms, which can include any suitable alignment method or algorithm known in the art, such as a matrix image library (MIL) pattern matching function, edge detection matching, context-based algorithms, etc.

[0075] In step 406, the computer subsystem can determine whether the candidate and the current candidate's setup image have been successfully aligned using the current alignment method. Determining whether the candidate and setup images have been aligned can be performed in any suitable manner. For example, the computer subsystem can compare any results of the alignment step to one or more predetermined criteria to determine whether the images have been successfully aligned. The computer subsystem can also or instead generate a score responsive to how well the images have been aligned to each other. Any such results or characteristics of the alignment step results can be used to determine which alignment results are used in other steps described herein. For example, the computer subsystem can remove alignment but relatively low score alignment sites from use in other steps described herein. The computer subsystem can also or instead attempt another alignment method for such a low-score alignment site to see whether a better alignment score can be achieved using a different alignment method. The computer subsystem can then determine which steps to perform next based on whether alignment of the current candidate has been achieved using the current alignment method.

[0076] In some embodiments, for at least one of the candidate images, the computer subsystem is configured to use two or more alignment methods to individually align the candidate image to its corresponding setup image until the candidate image is successfully aligned to its corresponding setup image. Thus, the computer subsystem can perform image alignment on a candidate using different alignment methods until at least one successful image alignment has been achieved for that candidate. The alignment steps can be performed in this way in different ways further described herein. In particular, for any one candidate, the alignment steps can be performed using one alignment method, then another alignment method (and so on) until all alignment methods have been tried or until a successful alignment has been achieved. Alternatively, for any one alignment method, the alignment steps can be performed on one candidate, then another candidate (and so on) until all candidates have been tried or a sufficiently successful alignment has been achieved. If all candidates face failure or a sufficiently successful alignment is not achieved using one alignment method, the computer subsystem can try the next alignment method.

[0077] In some examples, when the computer subsystem determines that the current candidate does not align the images using the current alignment method, the computer subsystem can attempt image alignment for other candidates. For example, when the computer subsystem determines in step 406 that the current candidate does not successfully align the images, the computer subsystem can determine in step 422 whether there are unattempted candidates. When the computer subsystem determines that there are unattempted candidates, the computer subsystem can select one of the unattempted candidates as a new candidate in step 424. The computer subsystem can then retrieve in step 402 the candidate alignment target image and the set image of the new (and now current) candidate and re-perform image alignment step 404 for the new (and now current) candidate. In this way, when the run-time and set images of one candidate do not match or do not successfully align, the computer subsystem can move to the next candidate until candidate N has been attempted (where N = the total number of set alignment targets).

[0078] In another embodiment, the computer subsystem is configured to perform individual alignment steps using a first alignment method, and when no candidate alignment target image successfully aligns with its corresponding set image, the computer subsystem is configured to perform individual alignment steps using a second alignment method. For example, as described above, the computer subsystem can move from candidate to candidate using one alignment method until all candidates have been attempted using that alignment method. When none of the candidates are successfully aligned using the current alignment method, the computer subsystem can then continue to attempt one or more other alignment methods, for example, for any candidates that were not successfully aligned, or even attempt alignment of all candidates again using a new method. For example, when the computer subsystem determines in step 406 that the current candidate does not successfully align the images using the current alignment method and when the computer subsystem determines in step 422 that all candidates have been attempted, the computer subsystem can then determine whether all alignment methods have been attempted. If the computer subsystem determines that there is one or more unattempted alignment methods, the computer subsystem can select one of the unattempted alignment methods to become the current alignment method. The image alignment and other steps described herein can then be performed for any candidates whose images have not been successfully aligned using the new alignment method and / or for all candidates. These steps can be performed for any or all alignment methods.

[0079] In another embodiment, the individual alignment step is performed using the first alignment method, and when no candidate image successfully aligns with its corresponding setup image, the computer subsystem is configured to repeat the individual alignment step using one or more additional alignment methods, which can be performed as described herein. In this embodiment, when no candidate image successfully aligns with its corresponding setup image using the one or more additional alignment methods, the computer subsystem is configured to abort the step of the process. For example, when the computer subsystem determines in step 434 that there are no more alignment methods to try, the computer subsystem can proceed to step 438 to determine whether any (or how many) images have successfully aligned using any of the alignment methods. If no images have successfully aligned using any of the alignment methods, the computer subsystem can abort the process in step 442.

[0080] Thus, as described above, the computer subsystem can try one alignment method for all candidates, and then try one or more other alignment methods for some or all of the candidates. However, for any one candidate, the computer subsystem can try different alignment methods (all of them or until a successful alignment has been found) and then move to the next candidate and repeat the same steps for the next candidate. In one such example, when the computer subsystem determines that the current candidate alignment image has not been aligned using the current alignment method, the computer subsystem can try other alignment methods for the current candidate, for example, until the current candidate is successfully aligned. In other words, if the computer subsystem determines in step 406 that the current candidate has not been successfully aligned using the current alignment method, the computer subsystem can determine whether there are any untried alignment methods for the current candidate. If the computer subsystem determines that there are any untried alignment methods, the computer subsystem can select a new alignment method from the untried ones. The newly selected alignment method then becomes the current alignment method, and the computer subsystem can perform image alignment in step 404 using the new current alignment method.

[0081] In this manner, for any one candidate, image alignment can be performed using different alignment methods until the candidate's image is aligned or when there are no more alignment methods. When the computer subsystem determines in step 406 that the image is not aligned and determines that there are no more alignment methods, the computer subsystem can determine whether there are any more untried candidates. The computer subsystem can select one of the untried candidates as the next candidate and then re-perform step 402 to retrieve the candidate alignment target and the setup image of the next (and now current) candidate and re-perform step 404 to align the image of the next (and now current) candidate using the current alignment method. In this manner, the computer subsystem can perform the same steps (performing image alignment using different methods until the image is successfully aligned) for any untried candidate. In the same manner as described above, if the computer subsystem has tried all candidates and all alignment methods, the computer subsystem can proceed to step 438 to determine whether any image has been aligned. If the computer subsystem determines in step 438 that no image has been aligned, the computer subsystem can abort the method in step 442. In this manner, when no candidate has been successfully aligned (no candidate has been aligned or only erroneous alignment has occurred) and all alignment methods have been tried, the computer subsystem can abort the alignment attempt.

[0082] For each of the one or more candidate images that are successfully aligned to their corresponding setup images, the computer subsystem is configured to separately determine, from each of the one or more candidate images and the coordinates of their corresponding mutually linearly independent vectors, coordinates of a reference location in the generated image for the sample. For example, when the computer subsystem determines in step 406 that the image of the current candidate is also successfully aligned, the computer subsystem can determine in step 408 the coordinates of the reference location of the current candidate. Determining the reference location of any of the candidates that have been successfully aligned can include vector addition operations and recording the die corner locations of the one candidate, which can be performed as further described herein.

[0083] When the computer subsystem determines the reference location coordinates of any candidate whose run-time image has been successfully aligned to its setup image, the computer subsystem can determine in step 422 whether there are any untried candidates. When there are untried candidates, the computer subsystem can select a next candidate in step 424. The computer subsystem can then retrieve in step 402 the candidate alignment target and the setup image of the new candidate and perform image alignment on the new candidate in step 404. The other steps described above can also be performed on the new candidate. In this manner, when the images of the candidates have been successfully aligned to each other, the computed reference location coordinates can be recorded and the computer subsystem can move to the next candidate until all candidates have been tried (up to candidate #N).

[0084] The computer subsystem is also configured to determine the final coordinates of the reference position in the image generated for the sample from the individually determined coordinates of the reference position. Determining the final coordinates of the reference position may involve first determining how many candidate images have been successfully aligned to their corresponding set image. For example, as further described herein, the final coordinates of the reference position may be determined in different ways depending on how many candidate images have been successfully aligned.

[0085] Determining how to calculate the final reference position coordinates based on the number of candidate images successfully aligned to their corresponding set image is important for several reasons. For example, this step can eliminate individual reference position coordinates determined from poorly aligned or completely misaligned alignment target images. For instance, when defects exist in the alignment targets formed on the sample, this unsuccessful alignment of the set alignment target image can occur during runtime.

[0086] In one of these examples, such as Figure 7 As shown, the bare wafer 700 can be formed on the sample and has a reference position 704, which is the same as... Figure 5 The setup stage is shown in the diagram. However, defects (or particles, foreign materials, etc.) indicated by the dark gray ellipse 710 can also form (or exist) on the sample. Specifically, defects exist in the candidate images of the setup alignment targets 712, 714, 716, 718, 720, and 722. These defects can interfere with the alignment of the candidate images of these setup alignment targets to their corresponding setup images. Therefore, Figure 4 Step 406 shown in the diagram can determine that these candidate images are misaligned, and these candidates can be eliminated when determining the final coordinates of the reference position.

[0087] In comparison, the defect does not exist. Figure 7 The candidate images for setting alignment targets 706 and 708 are shown in the diagram. Therefore, the alignment of the candidate images for setting alignment targets 706 and 708 should be successful (if there are no other interfering factors), and Figure 4 Step 406 shown in the figure should determine the alignment of these candidate images.

[0088] During or after the runtime alignment phase, the computer subsystem may produce results that set the alignment target, such as... Figure 8 As shown in the table. The first two columns shown in this table are the same as those shown in the table. Figure 6 The first two columns shown in the image. However, Figure 8 The third column in the table displays the coordinates of the candidate kernel center during runtime, rather than the set coordinates of the candidate kernel center. The coordinates of the center of the candidate kernel image can be determined in any suitable manner known in the relevant field. Figure 8 The vector from the kernel center to the corner of the die, shown in the last column of the table, is the same as... Figure 6 The vectors shown in the image.

[0089] Figure 8 It may also include indications of which candidate images have been successfully aligned to their corresponding setup images and therefore which candidates can be used to determine the final coordinates of the reference position. For example, Figure 8 The box 800 shown can be used to indicate only Figure 7 The alignment targets 706 and 708 shown are used to determine the final coordinates of the reference position.

[0090] like Figure 8 As shown, information generated during runtime for the selected alignment target can be stored and output in a table, but the results can be stored in any other suitable file or format. Information can be output and stored on computer-readable storage media, as further described herein.

[0091] In one embodiment, when two or more candidate images are successfully aligned to their respective corresponding set images, the computer subsystem is configured to determine whether the coordinates of individually determined reference positions for each of the two or more candidate images converge to within a predetermined distance from each other. For example, when the computer subsystem determines in step 422 that there are no more untried candidates, the computer subsystem may determine in step 426 whether more than one candidate reference position coordinate has been determined. If the computer subsystem determines that more than one candidate reference position coordinate has been determined, then the computer subsystem may determine in step 428 whether at least some of the reference position coordinates have converged. Determining whether the reference position coordinates have converged may include determining whether the individually determined reference position coordinates are within a radius of, for example, 500 nm (or some other predetermined distance) from each other.

[0092] When two or more candidate images are successfully aligned to their corresponding set image, determining whether the coordinates of the reference positions individually determined for each of the two or more candidate images converge can advantageously be performed before runtime die corner determination to avoid misalignment. In other words, this step can identify any candidate whose image appears to be successfully aligned to the set image but is actually misaligned, based on whether the die coordinates determined for each candidate are within a specific distance from each other.

[0093] To further illustrate this, it is assumed that it has been determined Figure 5 The candidate images for candidates 2, 3, and N shown were successfully aligned with their corresponding set images during runtime. The computer subsystem can then determine the DC position of candidate 2 as follows:

[0094]

[0095] The DC position of candidate 3 is determined as follows:

[0096]

[0097] and the DC position of candidate N is determined to be:

[0098]

[0099] Because being linearly independent from each other, the computer subsystem will determine that the individually determined calculated die corner positions converge if and only if:

[0100] a predetermined distance, and

[0101] a predetermined distance, and

[0102] a predetermined distance

[0103] (where the predetermined distance is, for example, 500 nm or a distance determined by the positional accuracy of the alignment method), then the computer subsystem will determine that all of these individually determined die corner positions should be used for the final coordinate determination.

[0104] In one such embodiment, the computer subsystem is configured to determine the final coordinates using only the individually determined coordinates of reference positions that are determined to converge to within a predetermined distance of each other. For example, when the computer subsystem determines in step 428 that at least some of the reference position coordinates have converged, the computer subsystem can eliminate in step 430 the reference position coordinates that have not converged. The computer subsystem can then determine in step 432 the final coordinates of the reference positions from the remaining reference position coordinates, as described further herein. In this way, the reference position coordinates that are not within a predetermined distance of each other can be eliminated from the final reference position coordinate calculation, which advantageously enables true high confidence alignment and enables determination of substantially high accuracy final reference position coordinates.

[0105] In another such embodiment, the computer subsystem is configured to determine the final coordinates using all of the individually determined coordinates of reference positions that are determined to converge to within a predetermined distance of each other. For example, one way in which the embodiments described herein increase the robustness of the final coordinate determination of the reference positions is by using as many of the individually determined reference position coordinates as possible that are determined to converge. For example, if in the above example the computer subsystem determines that the die corner position coordinates determined for candidates 2, 3, and N converge to within a predetermined distance, then the computer subsystem can determine the final coordinates of the die corner (DC) to be:

[0106]

[0107] where A is the number of candidates whose alignment to their corresponding setup image and whose reference locations individually determined coordinates converge to within a predetermined distance of each other. In this way, only candidates whose alignment has been completed (and confirmed by applying the convergence criteria) can be used to determine the precise die coordinates.

[0108] In another such embodiment, the individual alignment step is performed using a first alignment method, and when the coordinates of the reference locations individually determined for each of the two or more candidate images are determined not to converge to within a predetermined distance of each other, the computer subsystem is configured for repeating the individual alignment using a second alignment method. For example, if the reference location coordinates individually determined for a candidate do not converge to within a predetermined distance of each other, it can indicate that only a false alignment has occurred up to this point. Thus, when the computer subsystem determines in step 428 that no reference location coordinates have converged, the computer subsystem can proceed to step 434 to determine whether there are any untried alignment methods. If there are any untried alignment methods, the computer subsystem can select one of the untried alignment methods in step 436 as the next alignment method. The computer subsystem can then repeat the above steps using the next alignment method (e.g., by repeating at least steps 404 using the next alignment method).

[0109] When alignment of more than one candidate is implemented (as determined in 426) and when at least some of the reference location coordinate determinations have not converged (in step 428), the computer subsystem can determine whether there are any untried alignment methods (in step 434). When there is one or more untried alignment methods, the computer subsystem can select a next alignment method (as shown in step 436) and can perform the steps performed up to this point using the next alignment method. However, when there are no more untried alignment methods, the computer subsystem can determine the number of aligned images in step 438. When the computer subsystem determines in step 438 that the number of aligned images is greater than one but their reference location coordinates do not converge and all alignment methods have been tried, the computer subsystem can determine that this is a false alignment situation and abort the process in step 442. When no candidate has implemented alignment and all alignment methods have been tried, the computer subsystem determines in step 438 that the number of aligned images is zero and the computer subsystem can abort the process in step 442.

[0110] In some such embodiments, when two or more of the candidate images are successfully aligned to their respective corresponding setup images using the second alignment method, the computer subsystem is configured to individually determine additional coordinates of the reference position from the coordinates of each of the two or more candidate images that were successfully aligned using the second alignment method and their corresponding mutually linearly independent vectors, and determine whether the individually determined additional coordinates of the reference position converge within a predetermined distance of each other. In this way, if the computer subsystem determines that there are two or more successfully aligned candidate images using the second alignment method, the computer subsystem can individually determine the additional coordinates of the reference position for each of the candidate images. The computer subsystem can then also determine whether the coordinates determined for the candidate images converge. These steps can be performed as described above, but only for the reference position coordinates resulting from successful alignment using the second alignment method. In other words, any results resulting from the first alignment method can be discarded or otherwise not used with the results resulting from the second alignment method. Then, only the reference position coordinates resulting from successful alignment using the second alignment method can be used as described herein to determine the final coordinates of the reference position.

[0111] In another embodiment, when the individually determined additional coordinates of the reference position are determined not to converge within a predetermined distance of each other, the computer subsystem is configured to determine whether a third alignment method exists. In other words, if the individually determined reference position coordinates determined using the first alignment method do not converge with each other and if the individually determined reference position coordinates determined using the second alignment method also do not converge with each other, the computer subsystem can determine whether any other alignment methods exist that have not been tried. If there are more than two untried methods, the computer subsystem can select one of them as the third alignment method. These steps can be performed as described above and as many times or until there are no more alignment methods to try, as necessary.

[0112] In one such embodiment, when a third alignment method exists, the computer subsystem is configured to repeat the individual alignment steps using the third alignment method. For example, when a third alignment method exists, the computer subsystem can repeat at least steps 404, 406, and 408 for the third alignment method. The computer subsystem can also perform additional steps for the third alignment method, such as determining whether any of the individually determined reference position coordinates of the candidate images that were successfully aligned using the third alignment method converge, and if they do, determining the final reference position coordinates from the converging position coordinates.

[0113] In another such embodiment, the computer subsystem is configured to abort the process when no third alignment method exists. In other words, if the reference locations individually determined for candidate images successfully aligned using the first alignment method do not converge, if the reference locations individually determined for candidate images successfully aligned using the second alignment method do not converge, and if no other untried alignment method exists, the computer subsystem can abort the process. In this way, the embodiments described herein can avoid using a false alignment result to determine final coordinates for the reference locations, which, although non-ideal, is better than using a false alignment result for other steps described herein. Additionally, providing feedback to the user regarding the process results in such a scenario can enable the user to take other actions, such as ensuring that the correct wafer is loaded on the tool or inspecting the wafer for a serious defect that can have interfered with all alignment attempts made up to this point.

[0114] In the above-described steps, the computer subsystem can move from alignment method to alignment method until it finds an alignment method that can successfully align a candidate image whose determined reference locations converge. In such embodiments, only reference locations determined by a single one of the plurality of alignment methods can be used. However, in other embodiments, reference locations determined from aligning candidate images to their set-up images using different alignment methods can be used in combination to determine whether the individually determined reference locations converge. If the reference location coordinates individually determined from different alignment methods converge, they can then be used in combination to determine the final reference location coordinates.

[0115] In another embodiment, the computer subsystem is configured to perform the individual alignment steps for all candidate images using the first alignment method; and for any candidate alignment target images that are not successfully aligned to their corresponding set-up images using the first alignment method, the computer subsystem is configured to perform the individual alignment step using the second alignment method and the individual determination step for each of the one or more candidate images that are successfully aligned to their corresponding set-up images using the second alignment method. In other words, for a run-time die corner determination in which false alignments are avoided, if a candidate is not aligned using the current alignment method, the computer subsystem can move to the next alignment method and re-try. For example, as described further above, Figure 4Steps 402, 404, and 406 shown in the middle can be performed using one alignment method for all candidate images. After all candidate alignments have been attempted using one alignment method, the computer subsystem can determine whether there are any unattempted alignment methods. When there are unattempted alignment methods, the computer subsystem can select the next alignment method. Steps 402, 404, and 406 can then be performed using the next alignment method for all candidates or only those candidates that have not been successfully aligned. When any candidate is successfully aligned using the first or second alignment method, the computer subsystem can determine the coordinates of the reference position for that candidate using the results of the first or second alignment method, respectively, in step 408.

[0116] As further described herein, other embodiments, if no candidate has been successfully aligned, the computer subsystem can select the next alignment method and reattempt alignment for each candidate. If all algorithms have been attempted and no candidate has been successfully aligned, the computer subsystem can abort alignment, as Figure 4 shown in step 442.

[0117] In one such embodiment, when at least a first one of the candidate images is successfully aligned to its corresponding setup image using the first alignment method and at least a second one of the candidate images is successfully aligned to its corresponding setup image using the second alignment method, the computer subsystem is configured for: determining whether the separately determined coordinates of the reference positions determined for at least the first and second candidate images converge to within a predetermined distance of each other; and performing the determining final coordinates step using only the separately determined coordinates of the reference positions determined to converge to within the predetermined distance of each other. In other words, when successful image alignment of any candidate image has been achieved, the coordinates of the reference position for that candidate image can be determined, regardless of whether a different alignment method was used to achieve successful alignment. In this way, the embodiments described herein can combine results from different alignment methods for other steps, such as determining final coordinates of the reference position.

[0118] In one such example, if only alignment method 1 is used to successfully align the candidate image for setup alignment target 1 to its corresponding setup image and only alignment method 2 is used to successfully align the candidate images for setup alignment targets 2, 3, and n to their corresponding setup images, the computer subsystem can determine the coordinates of the reference position for each of candidates 1, 2, 3, and n separately using the results of the different alignment methods as described further herein. The separately determined reference position coordinates can then be input together to one or more other steps described herein. For example, in another such embodiment, the computer subsystem is configured for determining the final coordinates of the reference position using all of the separately determined coordinates that converge to within a predetermined distance of each other. In the above example, the computer subsystem can determine whether the separately determined coordinates of the reference position for each of candidates 1, 2, 3, and n converge, which can be performed as described herein. The computer subsystem can then determine the final coordinates of the reference position using only the separately determined coordinates of the reference position that converge, which can also be performed as described further herein.

[0119] In some embodiments, when only one candidate image is successfully aligned to its corresponding setup image, the computer subsystem is configured for determining the final coordinates of the reference position from the separately determined coordinates of the reference position determined for only one candidate image. In this way, if only one candidate has been successfully aligned to its corresponding image, alignment can be accomplished, but with a greater likelihood of false alignment than other embodiments described herein (e.g., when two or more candidate images are successfully aligned to their corresponding corresponding setup images). In such a scenario, the accuracy of the final coordinates of the die corner position can be the same as can be achieved by the alignment method currently being used.

[0120] In one such embodiment, as shown in step 426, when the computer subsystem determines that the reference coordinates for more than one candidate are not determined, and as shown in step 434, when the computer subsystem determines that there are no more alignment methods to try, the computer subsystem can determine in step 438 whether the reference position coordinates for one candidate have been determined. When the computer subsystem determines that the reference position coordinates for one candidate have been determined, the computer subsystem determines the final coordinates of the reference position from the one alignment candidate in step 440. In this case, if the only candidate that has been successfully aligned is candidate N, the computer subsystem can determine the final coordinates of the die corner position to be:

[0121]

[0122] In this way, if the computer subsystem can only find one successfully aligned candidate across all alignment methods and all candidates available to align the wafer during run time, that one successfully aligned candidate can be used as the best choice for determining the final reference coordinates.

[0123] Any die corner determination steps described herein may be modified in one or more ways. For example, instead of using all alignment candidates (or all aligned and converged candidates), embodiments may use only a majority of all alignment candidates (or a majority of all aligned and converged candidates) to determine the die corner coordinates. In another instance, a computer subsystem may be configured to determine the die corner coordinates from the alignment candidates using partial die corner calculations.

[0124] The embodiments described herein offer several advantages in addition to those already described. For example, the embodiments described herein can avoid misalignment on wafer inspection platforms during runtime. Furthermore, the embodiments described herein can perform generally high-confidence wafer alignment and die corner determination on wafer inspection platforms during runtime, even if the alignment points are not optimal, including severely damaged or even completely missing points. In this way, the embodiments described herein can reduce operational interruptions for all inspection platforms with alignment components for users in both large-scale research and development facilities and large-scale manufacturing facilities.

[0125] The embodiments described herein can also advantageously be adapted to several matching candidates (i.e., candidate images that have been successfully aligned to their corresponding set images). Specifically, the embodiments determine the final coordinates of the reference position differently depending on whether only one candidate matches or whether two or more of the candidates match. Furthermore, the alignment confidence of the embodiments increases with the number of successfully matched candidate images.

[0126] Another advantage of the embodiments described herein is that they can employ all currently available alignment methods, while also being able to identify only alignment methods such as actively assisted alignment, pre-mapping, and PDA alignment (e.g., by determining whether the alignment was successful and whether the individually determined reference position coordinates converged). Furthermore, the embodiments described herein will not have a significant impact on processing power. For example, during both setup and runtime, a single ripple can be used to capture and process N candidate images.

[0127] Another advantage of the embodiments described herein is that they require no design intellectual property, which increases the scenarios in which the embodiments described herein can be used. The embodiments also require no machine learning capabilities or databases, which increases the simplicity and adoptability of the embodiments described herein. Furthermore, the embodiments described herein can generate generally high-confidence data that can be used to accurately locate operational errors (e.g., faulty inspection chips, etc.).

[0128] Figure 9Advantages of the embodiments described herein are also illustrated. For example, the algorithm development of the described embodiments can vary across the dimensions 900 from less sensitive to more sensitive. When the algorithm development is on the less sensitive side, a scenario 902 can arise where the probability of misalignment can tend to 0, which is different from the current usage method where misalignment, pre-mapping, and PDA interruption can occur when the algorithm is less sensitive. When the algorithm development is on the more sensitive side, a scenario 904 can arise where the probability of misalignment is reduced to (0 / N+1 / N), which is different from the current usage method where misalignment interruption can occur when the algorithm is more sensitive.

[0129] During run-time, the computer subsystems can be configured to align different images to each other based on the final coordinates of the reference locations. For example, as shown in Figure 5 As shown in FIG. 6, there can be N different set alignment targets within one die that are unique with respect to each other and have a vector pointing to a die corner on one die by themselves. One die used during the set and run-time can be a reference die. The inspection recipe can generally include information specifying the reference die. Each die printed on the wafer is the same and spaced apart from each other by a fixed die width and height. Thus, all die coordinates on all dies can be easily translated back to their coordinates on the reference die using the known die width and height. In this way, images of different dies on a sample can be aligned to each other and then used for inspection or another process described herein.

[0130] Different images can be aligned to each other in the same way using other reference locations described herein. Image-to-image alignment can also be performed on all or only some of the images generated for a sample in a process in run-time. In addition, the image-to-image alignment based on the final coordinates of the reference locations can be performed in image or sample coordinates or other coordinates such as design coordinates.

[0131] In some embodiments, after the final coordinates of the reference locations are used to align the different images to one another during runtime, the computer subsystem is configured to detect defects by subtracting one image (i.e., a test image) from another image to which it is aligned (i.e., a reference image). The result of subtracting the reference image from the test image is commonly referred to in the art as a "difference image." Detecting defects can also include applying a certain threshold to the difference image, and the threshold can be commonly referred to in the art as a defect detection threshold. Only the image signals or data in the difference image that have values above the threshold can be identified by the computer subsystem as defects or potential defects. Of course, this can be the simplest version of how defect detection can be performed using the result of subtracting the reference image from the test image. The embodiments described herein are not limited to the defect detection methods or algorithms that can be used with the images aligned to one another described herein. For example, once the reference image is subtracted from the test image, there are many different types of defect detection methods or algorithms that can be used to detect defects or potential defects in the resulting difference image. Any or all such defect detection methods or algorithms can be used in the embodiments described herein, and the appropriate method or algorithm can be selected based on information about the sample.

[0132] The computer subsystem can be configured to generate results for the sample, which can include any of the results or information described herein. The results for the sample can be generated by the computer subsystem in any suitable manner. All of the embodiments described herein can be configured to store the results of one or more steps of the embodiments in a computer-readable storage medium. The results can include any of the results described herein and can be stored in any manner known in the art. The results for the sample can be in any suitable form or format, such as a standard file type. The storage medium can include any of the storage media described herein or any other suitable storage medium known in the art.

[0133] After the results are stored, the results can be accessed in 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 to perform one or more functions for the sample or another sample of the same type, etc. For example, the results generated by the computer subsystem described herein can include information for any defects detected on the sample, such as the locations of bounding boxes for the detected defects, etc., a detection score, information about a defect classification (such as a class label or ID), any defect attributes determined from any of the images, etc., or any such suitable information known in the art. The information can be used by the computer subsystem or another system or method to perform additional functions for the sample and / or the detected defects, such as determining a defect classification for the defects, filtering the defects to remove nuisance and / or uninteresting defects, sampling the defects for defect review or other analysis, determining root causes for the defects, etc.

[0134] In addition to the above functionality, such functionality further includes, but is not limited to, altering a process, such as a manufacturing process or step that has been or will be performed on the specimen, etc., in a feedback or feedforward manner. For example, the computer subsystem can be configured to determine, based on the detected defects, one or more changes to a process that has been performed on the specimen and / or will be performed on the specimen. The changes to the process can include any suitable change to one or more parameters of the process. The computer subsystem preferably determines the changes such that the detected defects on the specimen can be reduced or prevented on other specimens on which the revised process is performed, can be corrected or eliminated in another process performed on the specimen, can be compensated for in another process performed on the specimen, etc. The computer subsystem can determine such changes in any suitable manner known in the art.

[0135] Those changes can then be sent to a semiconductor manufacturing system (not shown) or to a storage medium (not shown) accessible by both the computer subsystem and the semiconductor manufacturing system. The semiconductor manufacturing system can or can not be part of the system embodiments described herein. For example, the imaging subsystem and / or computer subsystem described herein can be coupled to the semiconductor manufacturing system, e.g., via one or more common elements such as a housing, power supply, specimen handling devices or mechanisms, etc. The semiconductor manufacturing system can include any semiconductor manufacturing system known in the art, such as a lithography tool, an etch tool, a chemical mechanical polishing (CMP) tool, a deposition tool, and the like.

[0136] The embodiments described herein can enable inspection processes and tools to detect defects with increased sensitivity due to the relatively high accuracy image alignment achieved by the embodiments described herein. This increased sensitivity to defects allows users to improve their ability to make correct process decisions.

[0137] The embodiments described herein can also be used to set up image alignment and perform run-time image alignment in image-based processes, such as inspection, defect review, metrology, etc. The improved image alignment capabilities achieved by the embodiments described herein provide significant advantages to any such processes, such as reduced time to results, increased robustness, better performance of the resulting process, etc.

[0138] Each of the above embodiments can be combined together into a single embodiment. In other words, no embodiment is mutually exclusive with any other embodiment unless otherwise stated herein.

[0139] Another embodiment relates to a computer-implemented method for image alignment. The method includes the above-described individual alignment, individual determination of coordinates, and determination of final coordinates steps. The steps are performed by a computer subsystem that can be configured according to any of the embodiments described herein.

[0140] 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 subsystem and / or computer subsystem described herein. In addition, the above-described method can be performed by any of the system embodiments described herein.

[0141] Additional embodiments relate to a non-transitory computer-readable medium storing program instructions executable on a computer system for performing a computer-implemented method for image alignment. One such embodiment is shown in Figure 10 Figure 10 In particular, as shown in

[0142] The program instructions 1002 implementing a method, such as the methods described herein, can be stored on the computer-readable medium 1000. The computer-readable medium can be a storage medium such as a magnetic or optical disk, magnetic tape or any other suitable non-transitory computer-readable medium known in the art.

[0143] The program instructions can be implemented in any of a variety of ways, including procedure-based techniques, component-based techniques, and / or object-oriented techniques, among others. For example, the program instructions can be implemented using ActiveX controls, C++ objects, JavaBeans, Microsoft Foundation Classes ("MFC"), SSE (Streaming SIMD Extensions), or other technologies or methodologies, as desired.

[0144] The computer system 1004 can be configured according to any of the embodiments described herein.

[0145] Those skilled in the art will appreciate that yet other modifications and alternative embodiments can be apparent to those skilled in the art in view of the foregoing description. For example, methods and systems for image alignment are provided. Accordingly, the present description is to be construed as illustrative only and is for the purpose of teaching those skilled in the art the general manner of carrying out the present invention. It is to be understood that the forms of the invention shown and described herein are to be taken as the presently preferred embodiments. Elements and materials can be substituted for those illustrated and described herein, parts and processes can be reversed, and certain attributes can be utilized independently, all without departure from the spirit and scope of the present invention. Changes can be made to the elements described herein without departing from the spirit and scope of the application as described in the following claims.​

Claims

1. A system configured for image alignment, comprising: an imaging subsystem configured to generate images of a specimen; and a computer subsystem configured for: separately aligning candidate alignment target images in the images generated for the specimen to corresponding setup images of setup alignment targets selected to have a mutual linearly independent vector between a position of the setup alignment target and a reference position; for each of one or more of the candidate alignment target images that successfully align to one of their corresponding setup images, separately determining a coordinate of the reference position in the images generated for the specimen from a coordinate of the each of the one or more candidate alignment target images and its corresponding one of the mutual linearly independent vector; and determining a final coordinate of the reference position in the images generated for the specimen from the separately determined coordinates of the reference position.

2. The system of claim 1, wherein the setup alignment targets comprise at least two alignment targets selected to include patterns that are unique relative to each other.

3. The system of claim 1, wherein each of the setup alignment targets is located on a die on the specimen, and wherein the reference position is a corner of the die.

4. The system of claim 1, wherein when two or more of the candidate alignment target images successfully align to their respective ones of the corresponding setup images, the computer subsystem is further configured for determining whether the separately determined coordinates of the reference position for each of the two or more of the candidate alignment target images converge to within a predetermined distance of each other.

5. The system of claim 4, wherein the computer subsystem is further configured for determining the final coordinate using only the separately determined coordinates of the reference position that are determined to converge to within the predetermined distance of each other.

6. The system of claim 5, wherein the computer subsystem is further configured for determining the final coordinate using all of the separately determined coordinates of the reference position that are determined to converge to within the predetermined distance of each other.

7. The system of claim 4, wherein the separate alignments are performed using a first alignment method, and wherein when the separately determined coordinates of the reference position for the each of the two or more of the candidate alignment target images are determined not to converge to within the predetermined distance of each other, the computer subsystem is further configured for repeating the separate alignments using a second alignment method.

8. The system of claim 7, wherein when two or more of the candidate alignment target images are successfully aligned to their respective ones of the corresponding setup images using the second alignment method, the computer subsystem is further configured for individually determining additional coordinates of the reference location from the coordinates of each of the two or more of the candidate alignment target images that are successfully aligned using the second alignment method and their respective ones of the mutually linearly independent vectors; and determining whether the individually determined additional coordinates of the reference location converge within the predetermined distance of each other.

9. The system of claim 8, wherein when the individually determined additional coordinates of the reference location are determined not to converge within the predetermined distance of each other, the computer subsystem is further configured for determining whether a third alignment method exists; and when the third alignment method exists, the computer subsystem is further configured for repeating the individual alignments using the third alignment method.

10. The system of claim 8, wherein when the individually determined additional coordinates of the reference location are determined not to converge within the predetermined distance of each other, the computer subsystem is further configured for determining whether a third alignment method exists; and when the third alignment method does not exist, the computer subsystem is further configured for aborting the individual alignments, individually determining the coordinates, and determining the final coordinates step.

11. The system of claim 1, wherein when only one of the candidate alignment target images is successfully aligned to one of its corresponding setup images, the computer subsystem is further configured for determining the final coordinates of the reference location from the individually determined coordinates of the reference location determined for only the one of the candidate alignment target images.

12. The system of claim 1, wherein for at least one of the candidate alignment target images, the computer subsystem is further configured for performing the individually aligning the candidate alignment target images to the corresponding setup images using two or more alignment methods until the at least one of the candidate alignment target images is successfully aligned to one of its corresponding setup images.

13. The system of claim 1, wherein the computer subsystem is further configured for performing the individual alignments using a first alignment method for all of the candidate alignment target images; and for any of the candidate alignment target images that are not successfully aligned to one of their corresponding setup images using the first alignment method, performing the individual alignments using a second alignment method and the individual determinations for each of one or more of the candidate alignment target images that are successfully aligned to one of their corresponding setup images using the second alignment method.

14. The system of claim 13, wherein when at least a first one of the candidate alignment target images is successfully aligned to one of its corresponding setup images using the first alignment method and at least a second one of the candidate alignment target images is successfully aligned to one of its corresponding setup images using the second alignment method, the computer subsystem is further configured for: determining whether the separately determined coordinates of the reference position determined for the at least first and second ones of the candidate alignment target images converge within a predetermined distance of each other; and performing the determining the final coordinates using only the separately determined coordinates of the reference position that are determined to converge within the predetermined distance of each other.

15. The system of claim 14, wherein the computer subsystem is further configured for performing the determining the final coordinates using all of the separately determined coordinates of the reference position that are determined to converge within the predetermined distance of each other.

16. The system of claim 1, wherein the computer subsystem is further configured for performing the separate alignments using a first alignment method, and wherein when none of the candidate alignment target images is successfully aligned to one of its corresponding setup images, the computer subsystem is further configured for performing the separate alignments using a second alignment method.

17. The system of claim 1, wherein the separate alignments are performed using a first alignment method, wherein when none of the candidate alignment target images is successfully aligned to one of its corresponding setup images, the computer subsystem is further configured for repeating the separate alignments using one or more additional alignment methods, and wherein when none of the candidate alignment target images is successfully aligned to one of its corresponding setup images using the one or more additional alignment methods, the computer subsystem is further configured for aborting the separate alignment, separately determining the coordinates, and determining the final coordinates steps.

18. The system of claim 1, wherein the imaging subsystem is further configured as an inspection subsystem.

19. A non-transitory computer-readable medium storing program instructions executable on a computer system for performing a computer-implemented method for image alignment, wherein the computer-implemented method comprises: separately aligning candidate alignment target images of images produced by an imaging subsystem for a sample to corresponding setup images of setup alignment targets selected to have a mutual linearly independent vector between a position of the setup alignment target and a reference position; for each of one or more of the candidate alignment target images that are successfully aligned to one of its corresponding setup images, separately determining a coordinate of the reference position in the images produced for the sample from the coordinate of the each of the one or more of the candidate alignment target images and its corresponding one of the mutual linearly independent vector; and determining, from the individually determined coordinates of the reference location, final coordinates of the reference location in the image produced for the sample.

20. A computer-implemented method for image alignment, comprising: individually aligning candidate alignment targets in an image produced by an imaging subsystem for a sample to corresponding setup images in which setup alignment targets are disposed, the setup alignment targets being selected to have a mutual linearly independent vector between a location of the setup alignment target and a reference location; for each of one or more of the candidate alignment target images that successfully align to one of their corresponding setup images, individually determining coordinates of the reference location in the image produced for the sample from the coordinates of the each of the one or more of the candidate alignment target images and their corresponding one of the mutual linearly independent vector; and determining final coordinates of the reference location in the image produced for the sample from the individually determined coordinates of the reference location, wherein the individually aligning, the individually determining coordinates, and the individually determining final coordinates are performed by a computer subsystem.

21. A computer-implemented method for image alignment, comprising: individually aligning candidate alignment targets in an image produced by an imaging subsystem for a sample to corresponding setup images in which setup alignment targets are disposed, the setup alignment targets being selected to have a mutual linearly independent vector between a location of the setup alignment target and a reference location; for each of one or more of the candidate alignment target images that successfully align to one of their corresponding setup images, individually determining coordinates of the reference location in the image produced for the sample from the coordinates of the each of the one or more of the candidate alignment target images and their corresponding one of the mutual linearly independent vector; and determining final coordinates of the reference location in the image produced for the sample from the individually determined coordinates of the reference location, wherein the individually aligning, the individually determining coordinates, and the individually determining final coordinates are performed by a computer subsystem.

22. A computer-implemented method for image alignment, comprising: individually aligning candidate alignment targets in an image produced by an imaging subsystem for a sample to corresponding setup images in which setup alignment targets are disposed, the setup alignment targets being selected to have a mutual linearly independent vector between a location of the setup alignment target and a reference location; for each of one or more of the candidate alignment target images that successfully align to one of their corresponding setup images, individually determining coordinates of the reference location in the image produced for the sample from the coordinates of the each of the one or more of the candidate alignment target images and their corresponding one of the mutual linearly independent vector; and determining final coordinates of the reference location in the image produced for the sample from the individually determined coordinates of the reference location, wherein the individually aligning, the individually determining coordinates, and the individually determining final coordinates are performed

Images