Method for calibrating, predicting and controlling random defects in EUV lithography

Abstract

An initial probability of occurrence of a random defect within an inspection region of a workpiece is received. And sorting all positions of the random defect according to the initial occurrence probability. A cumulative expected defect count is determined and normalized to a fraction of a total expected defect count. A number of defect locations for capturing potential random defects above a threshold of total random defects is determined.

Description

[0001] Cross-reference of related applications

[0002] This application claims priority to provisional patent application No. 63 / 540,605, filed and assigned on September 26, 2023, the disclosure of which is hereby incorporated by reference. Technical Field

[0003] This disclosure relates to the measurement of random defects during semiconductor manufacturing. Background Technology

[0004] The evolution of the semiconductor manufacturing industry places increasingly higher demands on yield management, and specifically on metrology and inspection systems. Critical dimensions continue to shrink, but the industry needs to reduce the time spent achieving high-yield, high-value production. Minimizing the total time from detecting a yield problem to resolving it maximizes the semiconductor manufacturer's return on investment.

[0005] Semiconductor devices, such as logic and memory devices, typically involve processing semiconductor wafers using numerous manufacturing processes to form the various features and multiple layers of the semiconductor device. For example, photolithography is a semiconductor manufacturing process involving transferring a pattern from a photomask to a photoresist disposed on a workpiece, such as a semiconductor wafer. Additional examples of semiconductor manufacturing processes include, but are not limited to, chemical mechanical polishing (CMP), etching, deposition, and ion implantation. An arrangement of multiple semiconductor devices fabricated on a single semiconductor wafer can be separated into individual semiconductor devices.

[0006] Inspection processes are used at various stages of semiconductor manufacturing to detect defects on wafers, thereby promoting higher yields and, consequently, higher profits. Inspection has always been a crucial part of manufacturing semiconductor devices such as integrated circuits (ICs). However, as the size of semiconductor devices shrinks, inspection becomes even more critical for the successful manufacture of acceptable semiconductor devices because even small defects can cause device failure. For example, as the size of semiconductor devices decreases, the detection of smaller defects has become necessary because even relatively small defects can still cause undesirable aberrations in the semiconductor device.

[0007] Defect re-inspection typically involves re-inspecting defects detected during the inspection process and using high-magnification optical systems or scanning electron microscopy (SEM) to generate additional information about the defects at higher resolution. Defect re-inspection is usually performed at discrete locations on samples that have already passed inspection and been found to have defects. The higher-resolution data of defects generated by defect re-inspection is better suited for determining defect properties such as profile, roughness, or more accurate size information.

[0008] Photolithography can have defects driven by the quantized nature of light and materials. For example, light is quantized into photons, and the chemical reactants in photoresist are discrete molecules. These are often called shot noise defects or random defects. These random defects can be prevalent in extreme ultraviolet (EUV) photolithography, but can occur at the exposure wavelengths used in other photolithography processes such as ArF immersion. "Random" means that the average behavior can be within a desired specification (e.g., photoresist width, line end tip-to-tip measurement, or photoresist thickness), while exhibiting fluctuations that cause patterns to fail (e.g., bridging or breaking of line / space patterns) with a non-zero probability. Given that a workpiece contains billions of transistors, even a small probability of failure can result in a huge yield loss.

[0009] Random defects can present multiple challenges in a manufacturing environment. Typically, defects can be assumed to be deterministic, such that when manufacturing is performed according to a known production recipe that includes a pattern of elements to be manufactured on a sample and exposure parameters, the known defects will always be present. For example, process window qualification (PWQ) typically identifies process-limited defects that always occur when exposure conditions are outside of a process window. In an example, the process window can define limits on defocus associated with a position of a sample along an optical axis of a lithography tool (e.g., a focus position of the sample) or energy dose from an illumination source incident on the sample during exposure.

[0010] EUV photolithography processes used in high-end semiconductor device manufacturing can result in defects on a workpiece that tend to be random in nature. This can occur immediately after development of a photoresist image, or further downstream (e.g., during an etch process). Mathematical models have been used to predict the behavior of these random defects to aid in manufacturing process optimization. Generalized empirical models have previously been built primarily based on experimental data, and sometimes with physical constraints. These generalized empirical models rely on a large amount of defect rate data from optical inspection and SEM re-inspection to build and calibrate the model.

[0011] Empirical models have also been built using experimental SEM data. These typically build models using less data by exploiting geometric symmetries on designed features (e.g., 1D line / space features or contact hole arrays). However, these empirical models tend to be applicable only to specific designed features, and thus, are impractical to use in a general way across complex design layouts.

[0012] Accelerated rigorous models are built to simulate the results of full 3D rigorous physical simulators. These models typically require less experimental data to build. Typically, lower fidelity / higher precision experimental data is sufficient to build these models, rather than high resolution SEM data. Additional experimental data is typically used to calibrate / tune the models to align them with observed defect rates.

[0013] Empirical models often apply to specific design features, which are impractical to use across complex design layouts. Collecting the experimental data sets needed to build and calibrate these models is time consuming and expensive. Calibration of these models can require extensive curation of experimental data sets to improve prediction quality, which makes it difficult to implement. It can also be challenging or impractical to build large curated data sets that enable calibration of models to predict low probability hotspots across a sufficient variety of complex designs. Techniques that rely on experimental observation of defects under non- nominal operating conditions and extrapolation to nominal operating conditions can be potentially erroneous and can risk under / over predicting the defect free process window. Improved systems and techniques are needed. SUMMARY

[0014] In a first embodiment, a method is provided. The method includes receiving, at a processor, an initial occurrence probability of random defects within an inspection area of a workpiece. The initial occurrence probability of random defects is generated using a model. Using the processor, all locations of the random defects are sorted by the initial occurrence probability. Using the processor, a cumulative expected defect count is determined. Using the processor, the cumulative expected defect count is normalized as a fraction of a total expected defect count, thereby determining a normalized cumulative expected defect count. Using the processor, a number of defect locations for capturing potential random defects above a threshold of total random defects is determined.

[0015] The method can further include associating the normalized cumulative expected defect count with a total count of the inspection area.

[0016] The method can further include associating the normalized cumulative expected defect count with the inspection area.

[0017] The method can further include using the processor to determine the initial occurrence probability of random defects using the model.

[0018] The method can further include using the processor to select a subset of locations above the threshold, using the processor to group the subset of locations by pattern shape, thereby forming pattern shape groups, and using the processor to sort the pattern shape groups by expected defect count. The cumulative expected defect count is based on the pattern shape groups.

[0019] The method can further include using the processor to select a subset of locations by probability. The cumulative expected defect count is based on the subset of locations, and wherein the threshold is a fraction of a defect rate.

[0020] The method may further include using the processor to select a subset of locations according to probability. The cumulative expected defect count is based on the subset of locations. The threshold may be the number of locations.

[0021] The method may further include using the processor to select a subset of locations according to probability. The cumulative expected defect count is based on the subset of locations. The threshold may be a probability value.

[0022] The method may further include: using the processor to select one or more of the defect locations above the threshold, thereby generating selected defect locations; using the processor to group the selected defect locations according to pattern shapes on the workpiece; using the processor to determine pattern sensitivity based on at least one geometric distance in the pattern shapes; and using the processor to determine the expected defect count for each of the pattern shapes. Optionally, a subset of the pattern shapes may be examined.

[0023] The method may further include: using the processor to select one or more of the defect locations above the threshold, thereby generating selected defect locations; using the processor to group the selected defect locations according to pattern shapes on the workpiece; using the processor to determine pattern sensitivity based on at least one topological description of the pattern shapes; and using the processor to determine the expected defect count for each of the pattern shapes. Optionally, a subset of the pattern shapes may be examined.

[0024] The non-transitory computer-readable medium of the stored procedure may be configured to instruct a processor to execute embodiments of the method.

[0025] In a second embodiment, a system is provided. The system includes an inspection tool configured to image a workpiece and a processor in electronic communication with the inspection tool. The processor is configured to: receive an initial probability of occurrence of random defects within an inspection area of ​​the workpiece; sort all locations of the random defects according to the initial probability of occurrence; determine a cumulative expected defect count; normalize the cumulative expected defect count to a fraction of the total expected defect count, thereby determining a normalized cumulative expected defect count; and determine the number of defect locations for capturing potential random defects above a threshold of the total random defects. The initial probability of occurrence of random defects is generated using a model.

[0026] The processor may be further configured to associate the normalized cumulative expected defect count with the total count of the inspection area or with the inspection area.

[0027] The processor can be further configured to use the model to determine the initial probability of occurrence of the random defect.

[0028] The processor can be further configured to select a subset of locations above the threshold, group the subset of locations by pattern shape, thereby forming a pattern shape group, and sort the pattern shape groups by expected defect count. The cumulative expected defect count is based on the pattern shape groups.

[0029] The processor can be further configured to select a subset of locations by probability using the processor. The cumulative expected defect count is based on the subset of locations. The threshold can be a fraction of defect rate, a number of locations, or a probability value.

[0030] The processor can be further configured to select one or more of the defect locations above the threshold, thereby generating selected defect locations, group the selected defect locations by pattern shape on the workpiece, determine a pattern sensitivity based on at least one geometric distance in the pattern shape, and determine an expected defect count for each of the pattern shapes.

[0031] The processor can be further configured to select one or more of the defect locations above the threshold, thereby generating selected defect locations, group the selected defect locations by pattern shape on the workpiece, determine a pattern sensitivity based on at least one topological description in the pattern shape, and determine an expected defect count for each of the pattern shapes. BRIEF DESCRIPTION OF DRAWINGS

[0032] For a more complete understanding of the nature and objects of this disclosure, reference should be made to the following detailed description taken in connection with the accompanying drawings in which:

[0033] Figure 1 An embodiment of a method according to the present disclosure is illustrated;

[0034] Figure 2 FIG. 1 is a diagram of an exemplary optical tool;

[0035] Figure 3 FIG. 2 is a diagram of an exemplary e-beam tool;

[0036] Figure 4 FIG. 3 is an exemplary plot showing expected defects versus pattern shape count;

[0037] Figure 5 FIG. 4 is an exemplary plot showing expected defects versus location count; and

[0038] Figure 6 FIG. 5 shows exemplary minimum exclusion zones and fields of view. DETAILED DESCRIPTION

[0039] Although the claimed subject matter will be described in terms of certain embodiments, other embodiments (including embodiments that do not provide all of the benefits and features set forth herein) are also within the scope of the claims. Various structural, logic, process steps, and electronic changes can be made without departing from the scope of the claims. The claims should not be limited to the embodiments set forth herein for implementing various aspects of the technology.

[0040] Random failures are typically found with inspection tools and then characterized by top-down SEM. Inspection and characterization can occur after lithography (after development inspection (ADI)) or after subsequent etch and clean steps (after etch inspection (AEI) or after clean inspection (ACI)). AEI / ACI results can be most relevant to yield and depend on the remaining photoresist thickness under ADI. Embodiments disclosed herein can calibrate a random defect prediction model, such as an accelerated rigorous model that requires fine-tuning. Calibration can be performed with less experimental data and can avoid geometric grouping of images. Instead, embodiments can utilize grouping of hot spot locations by predicted probability of defects. Predicted probabilities across a layout can be processed in a manner that best suits downstream applications such as hot spot inspection / monitoring or hot spot repair.

[0041] Figure 1 An embodiment of the method 100 is shown. A processor can be used to perform at least some steps of the method 100. At 101, initial occurrence probabilities of random defects within an inspection area of a workpiece are received. This can be based on one or more images of the workpiece. The workpiece can be, for example, a semiconductor wafer or another type of workpiece. A model is used to generate the initial occurrence probabilities of random defects. The model can have been previously calibrated. At 102, all locations of the random defects are sorted by the initial occurrence probabilities. The probabilities can be percentages that are sorted. The probabilities are related to particular locations on the workpiece. Figure 1 In the figure, the inspection area is represented by a square illustrated with dashed lines. While only a portion of the workpiece is shown, the entire workpiece surface can be analyzed.

[0042] The model can include a rigorous random lithography simulation model and / or a transformed model (e.g., by applying machine learning) that can empirically reproduce the random defect probability predictions of the rigorous model while improving computational speed at the cost of a slight decrease in accuracy for certain input samples. The model can be tuned for a desired tradeoff between input space coverage, speed, and accuracy. An exemplary model is described in U.S. Publication No. 2022 / 0129775, which is incorporated by reference in its entirety. Tunable parameters can belong to the model. While the model contains several parameters that are learned during a training phase, these tunable parameters generally affect global characteristics of the output of the model, such as the number of detected hot spots and the overall magnitude of the predicted defect probabilities.

[0043] At 102, all locations of the random defects are sorted by the initial occurrence probabilities. The probabilities can be percentages that are sorted. The probabilities are related to particular locations on the workpiece.

[0044] At 103, a cumulative expected defect count is determined. The expected defect count within a region can be determined as the sum of the probabilities of all hotspots detected within the region. It is possible that the image is swept from 0% of the region to 100% of the region (e.g., in priority order). Then, the cumulative expected defect count at a given region percentage can be determined as the sum of the expected defect counts within all zones up to the particular percentage of the region. At 100% region, the cumulative expected defect count is the total number of defects expected to be observed across the entire region.

[0045] At 104, the cumulative expected defect count is normalized as a fraction of the total expected defect count. This can determine a normalized cumulative expected defect count. Normalization can be performed by dividing each cumulative defect count value by the cumulative defect count at 100% region.

[0046] An optimal amount of data (e.g., number of defect locations) for capturing potential defects can be determined. For example, the amount of data can be determined such that 50% of the total defects are expected to be observed in the top 5% of locations. In another example, the amount of data can be determined such that 80% of the total defects are expected to be observed in the top 20% of locations. In another example, the amount of data can be determined such that 95% of the total defects are expected to be observed in the top 60% of locations. Other parameters are possible and these are merely examples.

[0047] At 105, the number of defect locations for capturing potential random defects above a threshold of the total random defects can be determined. For example, the threshold can be selected based on a magnitude of defect probability that a semiconductor manufacturer can tolerate. The thresholding can be based on probability. Another approach is to set a threshold that selects a number of locations that are within a test budget. Then, the defect locations can be imaged using a test tool. The test model can be further validated based on this test and optionally calibrated.

[0048] The normalized cumulative expected defect count can be associated with a total count of a test region or the test region itself. Figure 1 An example of a graph showing an exemplary defect versus location count is shown in FIG. 1.

[0049] In an example, a model can be used to determine an initial occurrence probability of random defects. A desired pattern on a reticle can be an input to the model. In addition to the mask pattern, the input to the model can include process parameters such as exposure dose, focus, source shape, photoresist thickness, etch resistance, etc.

[0050] In an embodiment, a subset of locations above a threshold is selected. The subset of locations is grouped by pattern shape, thereby forming pattern shape groups. A pattern can be formed by shapes within a specified distance around a given location. Patterns can be grouped together based on geometric similarity of the patterns. A set of locations can be considered a group if the difference between the shapes forming each pattern does not exceed a predefined limit. The pattern shape groups are ordered by expected defect count. The cumulative expected defect count is based on the pattern shape groups.

[0051] An optimal amount of data (e.g., pattern shape count) for capturing potential defects can be determined. For example, the amount of data can be determined such that 50% of the total defects are expected to be observed in the top 70 pattern shapes. In another example, the amount of data can be determined such that 85% of the total defects are expected to be observed in the top 200 pattern shapes. In another example, the amount of data can be determined such that 99% of the total defects are expected to be observed in the top 500 pattern shapes. Other parameters are possible and these are merely examples. Figure 4 An example of a graph showing exemplary defects versus pattern shape count is shown in FIG. 5.

[0052] In an embodiment, a subset of locations is selected by probability of defect occurrence. The cumulative expected defect count can be based on the subset of locations. The threshold can be a fraction of defect rate, number of locations, or probability value.

[0053] An optimal amount of data (e.g., defect rate) for capturing potential defects can be determined. For example, the amount of data can be determined such that the impact on defect rate reaches a percentile (e.g., 80% of defect rate). In another example, the amount of data can be determined such that it is within a review budget (e.g., maximum 100,000 locations within review budget). In another example, the amount of data can be determined such that the focus is on the highest probability of defects (e.g., > 3 log-P). Other parameters are possible and these are merely examples. Figure 5 An example of a graph showing exemplary defects versus location count is shown in FIG. 6.

[0054] In an embodiment, one or more of the defect locations above a threshold can be selected, thereby resulting in selected defect locations. The selected defect locations are grouped by pattern shape on the workpiece. A pattern sensitivity can be determined based on at least one geometric distance in the pattern shape. The pattern sensitivity can be determined for each pattern shape that indicates the pattern sensitivity. The sensitivity (sensitivity to defect formation) of a pattern shape or pattern topology can be known from historical process data for a number of candidates or can also be known from historical simulation data. For example, the geometric distance can be the distance between a center hotspot location and its nearest polygon. Thus, design, pattern, or structure significance can be a factor in the analysis. An expected defect count for each of the pattern shapes can be determined. Optionally, a subset of the pattern shapes can be inspected using an inspection system.

[0055] In an embodiment, one or more of the defect locations above the threshold can be selected, thereby generating selected defect locations. The selected defect locations can be grouped by pattern shape on the workpiece. A pattern sensitivity can be determined based on at least one topological description in the pattern shape. A pattern topological identification can be determined by the shape closest to the center of the pattern. These patterns can include, for example, tip-to-tip, elbow, elbow-to-elbow, or two bars. An expected defect count for each of the pattern shapes can be determined. The defect probability and the number of instances of the pattern shape within the layout can be used as inputs to the expected defect count. Optionally, a subset of the pattern shapes can be inspected using an inspection system. Optionally, a subset of the pattern shapes can be inspected using an inspection system.

[0056] In an example, the pattern shapes are sorted by expected defect count and thresholded to a desired number of locations. In another example, the pattern shapes are sorted by expected defect count and thresholded to a desired number of locations such that the maximum number of locations per shape is selected. In another example, the pattern shapes are sorted by expected defect count. A cumulative expected defect count is determined and thresholded to a desired fraction of the defect rate. In another example, the pattern shapes are sorted by probability and thresholded to a desired number of pattern shapes or a desired probability value. In another example, the pattern shapes are sorted by geometric distance and thresholded by distance value. In another example, the pattern shapes are sorted by expected defect count and thresholded by geometric distance (e.g., design criticality, pattern importance, structure importance). Then, the pattern shapes are thresholded by number of locations, number of patterns, or cumulative defect count.

[0057] In an embodiment, the defect locations are processed to determine the best region of interest (ROI) on the workpiece for inspection. This can apply inspection system specific constraints. For example, the ROI can be equal to or smaller than the field of view (FOV) imaged with the inspection system. In other examples, positioning errors of the inspection system or minimum exclusion bands between inspection sites can be compensated. Figure 6 An exemplary minimum exclusion band and field of view is shown in the center.

[0058] In an embodiment, after grouping by pattern shape, topological signature information (e.g., elbow, two bars, or end-to-end, etc.) indicative of pattern sensitivity is included for each pattern shape. For each topographical signature, computational geometry information (e.g., end-to-end width, etc.) can be determined. An expected defect count for each pattern shape can be estimated. Various techniques can then be used to determine a subset of pattern shapes for inspection. For example, the pattern shapes can be ranked by expected defect count. A cumulative expected defect count can be determined and thresholded to a desired fraction of defectivity. In another example, the pattern shapes can be ranked by probability. The pattern shapes can be thresholded to a desired number of pattern shapes or a desired probability value. In another example, the pattern shapes can be ranked by topological signature information such as end-to-end width and then thresholded. In another example, the pattern shapes can be ranked by probability and expected defect count. This thresholds by topological information (e.g., end-to-end width) and then by probability, expected defect count, or cumulative defect count.

[0059] Figure 2 One embodiment of a system 200 is shown. The system 200 includes an optical-based subsystem 201. In general, the optical-based subsystem 201 is configured for generating an optical-based output of a workpiece 202 from light of the workpiece 202 by directing light to the workpiece 202 (or scanning light over the workpiece 202) and detecting the light from the workpiece 202. In one embodiment, the workpiece 202 includes a wafer. The wafer can include any wafer known in the art. In another embodiment, the workpiece 202 includes a reticle. The reticle can include any reticle known in the art.

[0060] In Figure 2 In the embodiment of the system 200 shown, the optical-based subsystem 201 includes an illumination subsystem configured to direct light to the workpiece 202. The illumination subsystem includes at least one light source. For example, as shown in Figure 2 In one embodiment, the illumination subsystem is configured to direct light to the workpiece 202 at one or more angles of incidence that can include one or more oblique angles and / or one or more normal angles. For example, as shown in Figure 2 In one embodiment, the illumination subsystem is configured to direct light to the workpiece 202 at one or more angles of incidence that can include one or more oblique angles and / or one or more normal angles. For example, as shown in

[0061] The optical-based subsystem 201 can be configured to direct light to the workpiece 202 at different angles of incidence at different times. For example, the optical-based subsystem 201 can be configured to alter one or more properties of one or more elements of the illumination subsystem such that light can be directed to the workpiece 202 at different angles of incidence at different times. For example, the optical-based subsystem 201 can be configured to alter one or more properties of one or more elements of the illumination subsystem such that light can be directed to the workpiece 202 at different angles of incidence at different times. Figure 2The incident angle shown guides the light to the workpiece 202. In one example, the optical subsystem 201 can be configured to move the light source 203, optical element 204, and lens 205 such that the light is guided to the workpiece 202 at different tilting incident angles or normal (or near-normal) incident angles.

[0062] In some examples, the optical-based subsystem 201 can be configured to guide light to the workpiece 202 simultaneously at more than one incident angle. For example, the illumination subsystem may include more than one illumination channel, one of which may include, for example, Figure 2 The light source 203, optical element 204, and lens 205 shown herein, and another of the illumination channels (not shown), may contain similar elements in different or identical configurations, or may contain at least one light source and possibly one or more other components (such as those further described herein). If this light and another light are simultaneously directed to the sample, then one or more characteristics (e.g., wavelength, polarization, etc.) of the light directed to the workpiece 202 at different incident angles may be different, such that the light generated by illuminating the workpiece 202 at different incident angles can be distinguished from each other at the detector.

[0063] In another example, the lighting subsystem may contain only one light source (e.g., Figure 2 The light source 203 shown in the diagram can be split into different optical paths (e.g., based on wavelength, polarization, etc.) by one or more optical elements (not shown) of the illumination subsystem. The light from each of these different optical paths can then be directed to the workpiece 202. Multiple illumination channels can be configured to direct light to the workpiece 202 simultaneously or at different times (e.g., when using different illumination channels to sequentially illuminate a sample). In another example, the same illumination channel can be configured to direct light with different characteristics to the workpiece 202 at different times. For example, in some examples, optical element 204 can be configured as a spectral filter, and the properties of the spectral filter can be changed in various different ways (e.g., by swapping out the spectral filter) so that light of different wavelengths can be directed to the workpiece 202 at different times. The illumination subsystem can have any other suitable configuration known in the art for sequentially or simultaneously directing light with different or the same characteristics to the workpiece 202 at different or the same incident angles.

[0064] In one embodiment, the light source 203 can include a broadband plasma (BBP) source. In this manner, light generated by the light source 203 and directed to the workpiece 202 can include broadband light. However, the light source can include any other suitable light source, such as a laser. The laser can include any suitable laser known in the art, and can be configured to generate light at any one or several suitable wavelengths known in the art. Additionally, the laser can be configured to generate light that is monochromatic or nearly monochromatic. In this manner, the laser can be a narrowband laser. The light source 203 can also include a polychromatic light source that generates light at a plurality of discrete wavelengths or wavebands.

[0065] Light from the optical element 204 can be focused onto the workpiece 202 by the lens 205. Although the lens 205 is shown in Figure 2 as a single refractive optical element, it should be understood that, in practice, the lens 205 can include several refractive and / or reflective optical elements that, in combination, focus light from the optical element to the sample. In Figure 2 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 (such as the beam splitter 213), apertures, and the like, which can include any such suitable optical elements known in the art. Additionally, the optics-based subsystem 201 can be configured to alter one or more of the elements of the illumination subsystem based on the type of illumination to be used to generate the optics-based output.

[0066] The optics-based subsystem 201 can also include a scanning subsystem configured to cause light to be scanned across the workpiece 202. For example, the optics-based subsystem 201 can include a stage 206 on which the workpiece 202 is disposed during generation of the optics-based output. The scanning subsystem can include any suitable mechanical and / or robotic assembly (including the stage 206) that can be configured to move the workpiece 202 so that light can be scanned across the workpiece 202. Additionally or alternatively, the optics-based subsystem 201 can be configured so that one or more optical elements of the optics-based subsystem 201 perform the scanning of light across the workpiece 202. The light can be scanned across the workpiece 202 in any suitable manner, such as in a serpentine path or in a spiral path.

[0067] The optics-based subsystem 201 further includes one or more detection channels. At least one of the one or more detection channels includes a detector configured to detect light from the workpiece 202 as a result of illuminating the workpiece 202 by the subsystem, and generate an output in response to the detected light. For example, Figure 2The optical-based subsystem 201 shown in FIG. 2 includes two detection channels, a detection channel formed by collector 207, element 208, and detector 209 and another detection channel formed by collector 210, element 211, and detector 212. As Figure 2 As shown in FIG. 2, 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 workpiece 202 at different angles. However, one or more of the detection channels can be configured to detect another type of light (e.g., reflected light) from workpiece 202.

[0068] As Figure 2 As further shown in FIG. 2, the two detection channels are shown positioned in the plane of the paper, and the illumination subsystem is also shown positioned in the plane of the paper. Thus, in this embodiment, the two detection channels are positioned in (e.g., centered in) the plane of incidence. However, one or more of the detection channels can be positioned out of the plane of incidence. For example, the detection channel formed by collector 210, element 211, and detector 212 can be configured to collect and detect light scattered out of the plane of incidence. Thus, this detection channel can be generally referred to as a "side" channel, and this side channel can be centered in a plane that is substantially perpendicular to the plane of incidence.

[0069] Although Figure 2 While an embodiment of optical-based subsystem 201 including two detection channels is shown, optical-based subsystem 201 can include a different number of detection channels (e.g., only one detection channel or two or more detection channels). In one such example, the detection channel formed by collector 210, element 211, and detector 212 can form one side channel as described above, and optical-based subsystem 201 can include an additional detection channel (not shown) formed as another side channel positioned on an opposite side of the plane of incidence. Thus, optical-based subsystem 201 can include a detection channel including collector 207, element 208, and detector 209 and centered in the plane of incidence and configured to collect and detect light at or near a scattering angle normal to the surface of workpiece 202. Thus, this detection channel can be generally referred to as a "top" channel, and optical-based subsystem 201 can also include two or more side channels configured as described above. As such, optical-based subsystem 201 can include at least three channels (i.e., one top channel and two side channels), and each of the at least three channels has its own collector, each of which is configured to collect light at a different scattering angle than each of the other collectors.

[0070] As further described above, each of the detection channels included in optical-based subsystem 201 can be configured to detect scattered light. Thus,Figure 2 The optical-based subsystem 201 illustrated in FIG. 1 can be configured for dark field (DF) output generation of the specimen 202. However, the optical-based subsystem 201 can also or instead include detection channels configured for bright field (BF) output generation of the specimen 202. In other words, the optical-based subsystem 201 can include at least one detection channel configured to detect light specularly reflected from the workpiece 202. Thus, the optical-based subsystem 201 described herein can be configured for DF only, BF only, or both DF and BF imaging. Although each of the collectors is illustrated in FIG. 1 as a single refractive optical element, it should be understood that each of the collectors can include one or more refractive optical dies and / or one or more reflective optical elements. Figure 3

[0071] The one or more detection channels can include any suitable detector known in the art. For example, the detectors can include photomultiplier tubes (PMTs), charge-coupled devices (CCDs), time delay integration (TDI) cameras, and any other suitable detector known in the art. The detectors can also include non-imaging detectors or imaging detectors. In this way, if the detectors are non-imaging detectors, each of the detectors can be configured to detect a particular characteristic of the scattered light (e.g., intensity), but can not be configured to detect such a characteristic as a function of position within an imaging plane. As such, the output generated by each of the detectors included in each of the detection channels of the optical-based subsystem can be a signal or data, rather than an image signal or image data. In such examples, a processor (e.g., the processor 214) can be configured to generate an image of the workpiece 202 from the non-imaging output of the detectors. However, in other examples, the detectors can be configured as imaging detectors configured to generate an imaging signal or image data. Thus, the optical-based subsystem can be configured to generate the optical images or other optical-based outputs described herein in several ways.

[0072] It should be noted that the configurations of the optical-based subsystem 201 provided herein Figure 3 The configurations of the optical-based subsystem 201 described herein can be altered to optimize the performance of the optical-based subsystem 201 as is typically performed when designing a commercial output acquisition system. Additionally, the systems described herein can be implemented using existing systems (e.g., by adding the functionality described herein to an existing system). For some such systems, the methods described herein can be provided as optional functionality of the system (e.g., in addition to other functionality of the system). Alternatively, the systems described herein can be designed as a completely new system.

[0073] ​Figure 3 is a block diagram of an embodiment of a system 300. The system 300 includes a wafer inspection tool (which includes an electron column 301) configured to generate an image of a workpiece 304.

[0074] The wafer inspection tool includes an output acquisition subsystem that includes at least an energy source and a detector. The output acquisition subsystem can be an electron beam-based output acquisition subsystem. For example, in one embodiment, the energy directed to the workpiece 304 includes electrons, and the energy detected from the workpiece 304 includes electrons. In this way, the energy source can be an electron beam source. In Figure 3 One such embodiment, shown in

[0075] Also as shown in Figure 3 The electron column 301 includes an electron beam source 303 configured to generate electrons that are focused through one or more components 305 to the workpiece 304, as shown in

[0076] Electrons (e.g., secondary electrons) returned from the workpiece 304 can be focused through one or more components 306 to a detector 307. The one or more components 306 can include, for example, a scanning subsystem, which can be the same scanning subsystem included in the components 305.

[0077] The electron column 301 can also include any other suitable components known in the art.

[0078] Although the electron column 301 is shown in Figure 3 as being configured such that the electrons are directed to the workpiece 304 at an oblique angle of incidence and scattered from the workpiece 304 at another oblique angle, the electron beam can be directed to the workpiece 304 and scattered from the workpiece 304 at any suitable angles. In addition, the electron beam-based output acquisition subsystem can be configured to generate an image of the workpiece 304 using a variety of modes (e.g., employing different angles of illumination, angles of collection, etc.). The variety of modes of the electron beam-based output acquisition subsystem can differ in any image generation parameter of the output acquisition subsystem.

[0079] It should be noted that the embodiments provided herein ​To generally illustrate a configuration of an electron beam-based output acquisition subsystem that can be used in the embodiments described herein. The electron beam-based output acquisition subsystem configuration described herein can be altered to optimize the performance of the output acquisition subsystem as is commonly performed when designing a commercial output acquisition system. Additionally, the systems described herein can be implemented using existing systems (e.g., by adding the functionality described herein to an existing system). For some such systems, the methods described herein can be provided as optional functionality of the system (e.g., in addition to other functionality of the system). Alternatively, the systems described herein can be designed as a new system.

[0080] Although the output acquisition subsystem is described above as an electron beam-based output acquisition subsystem, the output acquisition subsystem can be an ion beam-based output acquisition subsystem. Such an output acquisition subsystem can be configured as shown in ​ with the exception that the electron beam source can be replaced by any suitable ion beam source known in the art. Additionally, the output acquisition subsystem can be any other suitable ion beam-based output acquisition subsystem, such as the output acquisition subsystems included in commercially available focused ion beam (FIB) systems, helium ion microscopy (HIM) systems, and secondary ion mass spectrometers (SIMS) systems.

[0081] Embodiments of the system 200 and the system 300 include a processor and an electronic data storage unit, such as the processor 214 or the processor 308 and the electronic data storage unit 215 or the electronic data storage unit 309. The processor 214 or the processor 308 can include a microprocessor, a microcontroller, or other device. In an example, the processor 309 is part of a computer subsystem, such as the computer subsystem 302.

[0082] The processor 214 or the processor 308, other systems, or other subsystems described herein can be part of various systems, including a personal computer system, image computer, mainframe computer system, workstation, network appliance, internet appliance, or other device. The subsystems or systems can also include any suitable processor known in the art, such as a parallel processor. Additionally, the subsystems or systems can include a platform having high speed processing and software as an independent or networked linked tool.

[0083] The processor 214 or the processor 308 can be coupled to elements of the system 200 or the system 300 in any suitable manner (e.g., via one or more transmission media, which can include wired and / or wireless transmission media) so that the processor 214 or the processor 308 can receive outputs. The processor 214 or the processor 308 can be configured to perform a number of functions using the outputs. The wafer inspection tool can receive instructions or other information from the processor 214 or the processor 308. The processor 214 or the processor 308 can optionally be in electronic communication with another wafer inspection tool, wafer metrology tool, or wafer review tool (not illustrated) to receive additional information or send instructions.

[0084] The processor 214 or the processor 308 is in electronic communication with a wafer inspection tool (e.g., the detector 209, the detector 212, or the detector 307). The processor 214 or the processor 308 can be configured to process images generated using measurements from the detector or perform other functions. For example, the processor can perform embodiments of the method 100 or the method 200.

[0085] The processor 214 or the processor 308 and its associated electronic data storage units can be disposed in or otherwise part of the system 200 or the system 300 or can be part of another device. In an example, the processor 214 or the processor 308 and its associated electronic data storage units can be part of a standalone control unit or in a centralized quality control unit. Multiple processors 214 or processors 308 or electronic data storage units can be used.

[0086] The processor 214 or the processor 308 can be implemented in practice through any combination of hardware, software, and firmware. Moreover, its functions as described herein can be performed by one unit, or divided up among different components, each of which can in turn be implemented by any combination of hardware, software, and firmware. Program code or instructions for the processor 214 or the processor 308 to implement various methods and functions can be stored in a readable storage medium, such as the electronic data storage unit 215, the memory in the electronic data storage unit 309, or other memory.

[0087] The processor 214 or the processor 308 can be configured to perform a number of functions using outputs of the system 200, the system 300, or other outputs. For example, the processor 214 or the processor 308 can be configured to send outputs to its respective electronic data storage unit or another storage medium. The processor 214 or the processor 308 can be further configured as described herein.

[0088] The processor 214 or the processor 308 can be configured according to any of the embodiments described herein. For example, the processor 214 or the processor 308 can be configured to perform embodiments of the method 100. The processor 214 or the processor 308 can also be configured to perform other functions or additional steps using the output of the system 200, the system 300, or using images or data from other sources.

[0089] If the system 200 or the system 300 includes more than one processor 214 or processor 308, the different subsystems can be coupled to each other such that images, data, information, instructions, etc. can be sent between the subsystems directly. For example, one subsystem can be coupled to an additional subsystem by any suitable transmission medium that can include any suitable wired and / or wireless transmission medium known in the art. Two or more such subsystems can also be effectively coupled by sharing a computer-readable storage medium (not shown).

[0090] A processor (such as the processor 214 or the processor 308) for performing embodiments disclosed herein can be a single computer or part of a network of computers. The processor can include one or more CPU processors or can include a combination of CPU and GPU processors. In an example, there are 15 computer nodes in the network and each node includes 44 CPUs with 385 GB of RAM and 500 GB of hard disk space and 2 GPUs.

[0091] The various steps, functions, and / or operations of the system 200, the system 300, or methods disclosed herein are performed by one or more of the following: electronic circuitry, logic gates, multiplexers, programmable logic devices, ASICs, analog or digital controls / switches, microcontrollers, or computing systems. Program instructions implementing a method such as the methods described herein can be transmitted over or stored on carrier media. The carrier media can include a storage medium such as a read-only memory, a random access memory, a magnetic or optical disk, a non-volatile memory, a solid state memory, a magnetic tape, and the like. The carrier media can include a transmission medium such as a wire, cable, or wireless transmission link. Accordingly, the above description should not be interpreted as a limitation on the scope of the disclosure but merely as an illustration.

[0092] Additional embodiments relate to a non-transitory computer readable medium storing program instructions executable on a controller for performing a computer-implemented method as disclosed herein. In particular, the electronic data storage unit 215, the electronic data storage unit 309, or other storage medium can contain a non-transitory computer readable medium including program instructions executable on a processor. The computer-implemented method can include any steps of any method described herein, including embodiments of the method 100.

[0093] Program instructions can be implemented in any of various 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++ object-oriented programming, JavaBeans, Microsoft Foundation Classes (MFC), Flow- Streaming SIMD Extensions (SSE), or other technologies or methodologies, as desired.

[0094] Each of the steps of the method can be performed as described herein. The method can also include any other steps that can be performed by the processor and / or computer subsystems or systems described herein. The steps can be performed by one or more computer systems that can be configured according to any of the embodiments described herein. Additionally, the methods described above can be performed by any of the system embodiments described herein.

[0095] While the present disclosure has been described with respect to one or more particular embodiments, it is understood that other embodiments of the present disclosure can be made without departing from the scope of the present disclosure. Accordingly, the present disclosure is deemed to be limited only by the appended claims and their reasonable interpretation.

Claims

1. A method comprising: receiving, at a processor, an initial occurrence probability of random defects within an inspection area of a workpiece, wherein the initial occurrence probability of random defects is generated using a model; ranking, using the processor, all locations of the random defects by the initial occurrence probability; determining, using the processor, a cumulative expected defect count; normalizing, using the processor, the cumulative expected defect count to a fraction of a total expected defect count, thereby determining a normalized cumulative expected defect count; and determining, using the processor, a number of defect locations for capturing potential random defects above a threshold of total random defects.

2. The method of claim 1, further comprising associating the normalized cumulative expected defect count with a total count of the inspection area.

3. The method of claim 1, further comprising associating the normalized cumulative expected defect count with the inspection area.

4. The method of claim 1, further comprising determining, using the processor, the initial occurrence probability of random defects using the model.

5. The method of claim 1, further comprising: selecting, using the processor, a subset of locations above the threshold; grouping, using the processor, the subset of locations by pattern shape, thereby forming pattern shape groups; and ranking, using the processor, the pattern shape groups by expected defect count; wherein the cumulative expected defect count is based on the pattern shape groups.

6. The method of claim 1, further comprising selecting, using the processor, a subset of locations by probability, wherein the cumulative expected defect count is based on the subset of locations, and wherein the threshold is a fraction of a defect rate.

7. The method of claim 1, further comprising selecting, using the processor, a subset of locations by probability, wherein the cumulative expected defect count is based on the subset of locations, and wherein the threshold is a number of locations.

8. The method of claim 1, further comprising selecting, using the processor, a subset of locations by probability, wherein the cumulative expected defect count is based on the subset of locations, and wherein the threshold is a probability value.

9. The method of claim 1, further comprising: selecting, using the processor, one or more of the defect locations above the threshold, thereby generating selected defect locations; grouping, using the processor, the selected defect locations by pattern shapes on the workpiece; determining, using the processor, a pattern sensitivity based on at least one geometric distance in the pattern shapes; and determining, using the processor, an expected defect count for each of the pattern shapes.

10. The method of claim 9, further comprising inspecting a subset of the pattern shapes.

11. The method of claim 1, further comprising: selecting, using the processor, one or more of the defect locations above the threshold, thereby generating selected defect locations; grouping, using the processor, the selected defect locations by pattern shapes on the workpiece; determining, using the processor, a pattern sensitivity based on at least one topological description in the pattern shapes; and determining, using the processor, an expected defect count for each of the pattern shapes.

12. The method of claim 11, further comprising inspecting a subset of the pattern shapes.

13. A non-transitory computer readable medium storing a program configured to instruct a processor to perform the method of claim 1.

14. A system comprising: an inspection tool configured to image a workpiece; and a processor in electronic communication with the metrology tool, wherein the processor is configured to: receive an initial occurrence probability of random defects within an inspection area of a workpiece, wherein the initial occurrence probability of random defects is generated using a model; sort all locations of the random defects by the initial occurrence probability; determine a cumulative expected defect count; normalize the cumulative expected defect count to a fraction of a total expected defect count, thereby determining a normalized cumulative expected defect count; and determine a number of defect locations for capturing potential random defects above a threshold of total random defects.

15. The system of claim 14, wherein the processor is further configured to correlate the normalized cumulative expected defect count to a total count of the inspection area or to the inspection area.

16. The system of claim 14, wherein the processor is further configured to employ the model to determine the initial occurrence probability of random defects.

17. The system of claim 14, wherein the processor is further configured to: select a subset of locations above the threshold; group the subset of locations by pattern shapes, thereby forming pattern shape groups; and sort the pattern shape groups by expected defect count; wherein the cumulative expected defect count is based on the pattern shape groups.

18. The system of claim 14, wherein the processor is further configured to select a subset of locations by probability using the processor, wherein the cumulative expected defect count is based on the subset of locations, and wherein the threshold is a fraction of a defect rate, a number of locations, or a probability value.

19. The system of claim 14, wherein the processor is further configured to: select one or more of the defect locations above the threshold, thereby generating selected defect locations; group the selected defect locations by pattern shapes on the workpiece; determine a pattern sensitivity based on at least one geometric distance in the pattern shapes; and determine an expected defect count for each of the pattern shapes.

20. The system of claim 14, wherein the processor is further configured to: select one or more of the defect locations above the threshold, thereby generating selected defect locations; group the selected defect locations by pattern shapes on the workpiece; determining a pattern sensitivity based on at least one topological description in the pattern shape; and determining an expected defect count for each of the pattern shapes.

Images