JP2012518278A - Defect detection on wafer - Google Patents

Abstract

  A method and system for detecting defects on a wafer is provided.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is a priority of US Provisional Patent Application No. 61 / 152,477, filed Feb. 13, 2009, entitled “Methods and Systems for Detecting Defects on a Wafer”. No. 61 / 152,477 is hereby incorporated by reference as if fully set forth herein.

  The present invention relates generally to defect detection on a wafer. Particular embodiments relate to assigning individual outputs in raw output to different segments for wafers generated by the inspection system.

  The following description and examples are related art and not prior art.

  Wafer inspection using optical or electron beam technology is an important technique in the semiconductor industry for debugging semiconductor manufacturing processes, monitoring process variations, and improving manufacturing yield. In addition to the increasingly complex manufacturing process, the difficulty of inspection is increasing as the scale of modern integrated circuits (ICs) continues to shrink.

  In each processing step performed on a semiconductor wafer, the same circuit pattern is printed on each die on the wafer. Most wafer inspection systems take advantage of this fact to detect defects on the wafer by making a relatively simple die-to-die comparison. However, each die's printed circuit may include multiple areas of patterned features that repeat in the x and y directions, such as DRAM, SRAM, or FLASH areas. This type of area is commonly referred to as an array area (other areas are called random areas or logical areas). In order to achieve better sensitivity, different strategies are used in advanced inspection systems to inspect array areas and random or logical areas.

  In order to set up a wafer inspection process for array inspection, in many currently used inspection systems, a user manually sets a region of interest (ROI) and defect detection in the same ROI. To detect the same set of parameters. However, this setting method is disadvantageous for several reasons. For example, as the design rules are reduced, the definition of the region becomes much more complex and the area becomes even smaller. Due to limitations in stage accuracy and inspection system resolution, manual setting of ROI will become increasingly impossible. On the other hand, if the distance between page breaks is larger than Fourier filtering can be performed, page breaks will not be suppressed in the array region.

  In other methods, intensity is used as a segmentation feature that classifies similarly intensity pixels into the same group. The same set of parameters is then applied to pixels in the same groove (based on intensity). However, this method also has some disadvantages. For example, an intensity-based segmentation algorithm can be used when the geometric features are evenly distributed. However, this is often not enough. Therefore, segmentation based on other attributes is required.

  Thus, a method for defect detection on a wafer that makes it possible to achieve better defect detection by taking advantage of the knowledge that the defect of interest and disturbance / noise belong geometrically to different segments And it would be advantageous to develop a system.

  The following description of various embodiments should not be construed as limiting the subject matter of the appended claims.

  One embodiment relates to a computer-implemented method for detecting defects on a wafer. The computer-implemented method includes obtaining raw output for a wafer generated by an inspection system. The computer-implemented method also includes identifying one or more characteristics of the raw output that correspond to one or more geometric characteristics of the patterned features formed on the wafer. . In addition, the computer-implemented method can identify the unprocessed output so that one or more geometric properties of the patterned features corresponding to each of the different segments are different. Or including assigning individual outputs in the raw output to different segments based on a plurality of characteristics. Further, the computer-implemented method includes assigning one or more defect detection parameters to different segments separately. The computer-implemented method also includes detecting defects on the wafer by applying the assigned defect detection parameter or parameters to individual outputs assigned to different segments.

  The steps of the computer-implemented method described above may be performed as described further herein. This computer-implemented method may include any other step (s) of any other method (s) described herein. The computer-implemented methods described above may be implemented using any system described herein.

  Another embodiment relates to a computer-readable medium that includes program instructions executable on a computer system that implements a method for detecting defects on a wafer. The method includes the computer-implemented method steps described above. The computer readable medium may be further configured as described herein. The method steps may be performed as further described herein. In addition, the method by which program instructions are executable may include any other step (s) of any other method (s) described herein.

  An additional embodiment relates to a system configured to detect defects on a wafer. The system includes an inspection subsystem configured to generate a raw output for a wafer by scanning the wafer. The system also includes a computer subsystem configured to obtain the raw output. The computer subsystem is configured to identify one or more characteristics of the raw output that correspond to one or more geometric characteristics of the patterned features formed on the wafer. In addition, the computer subsystem includes the identified one or more of the raw outputs so that the one or more geometric characteristics of the patterned features corresponding to each of the different segments are different. Based on the characteristics, the individual outputs in the raw output are configured to be assigned to different segments. The computer subsystem is further configured to separately assign one or more defect detection parameters to different segments. Further, the computer subsystem is configured to detect defects on the wafer by assigning the assigned defect detection parameter or parameters to individual outputs assigned to different segments. This system may be further configured as described herein.

  Other objects and features of the present invention will become apparent upon reading the following detailed description and upon reference to the accompanying drawings in which:

FIG. 6 is a block diagram illustrating one embodiment of a computer-readable medium that includes program instructions executable on a computer system that performs one or more method embodiments described herein. 1 is a side schematic diagram illustrating one embodiment of a system configured to detect defects on a wafer. FIG.

  While the invention is susceptible to various modifications and alternative forms, specific embodiments according to the invention are shown by way of example in the drawings and will be described in detail herein. However, the drawings and detailed description relating to the invention are not intended to limit the invention to the particular form disclosed, but rather to the spirit and scope of the invention as defined by the appended claims. It should be understood that it is intended to include all modifications, equivalents, and alternatives to which it belongs.

  Although embodiments relating to wafers are disclosed herein, it should be noted that these embodiments can be used to detect defects on other samples, such as a reticle, which can be commonly referred to as a mask or photomask. Should be understood. Many different types of reticles are known in the art, and the terms “reticle”, “mask”, and “photomask” as used herein are intended to include all of the reticles known in the art. Should be understood.

  One embodiment relates to a computer-implemented method for detecting defects on a wafer. This computer-implemented method includes obtaining raw output for a wafer generated by an inspection system. Obtaining the raw output for the wafer may be performed using an inspection system. For example, the acquisition of raw output may be performed by scanning light on a wafer using an inspection system and reacting to light scattered and / or reflected from the wafer detected by the inspection system while performing the scan. Generating a processing output. In this way, obtaining the raw output may include scanning the wafer. However, obtaining the raw output does not necessarily include scanning the wafer. For example, obtaining raw output may include obtaining raw output from a storage medium in which raw output is stored (eg, by an inspection system). Obtaining raw output from a storage medium may be implemented in any suitable manner, and the storage medium from which output is obtained may include any storage medium described herein. In any case, the method includes collecting raw output (eg, raw data).

  In one embodiment, the raw output is responsive to light scattered from the wafer. In particular, the raw output may be responsive to light scattered from the wafer and detected by the inspection system. Alternatively, the raw output may be responsive to light reflected from the wafer and detected by the inspection system. The raw output may include any suitable raw output and may vary depending on the configuration of the inspection system. For example, the raw output may include signals, data, image data, etc. In addition, the raw output may be generally defined as the output for at least a portion (eg, a plurality of pixels) of the total output generated for the wafer by the inspection system. Furthermore, the raw output includes all of the raw output generated for the entire wafer by the inspection system, whether the raw output corresponds to a defect on the wafer, or the wafer scanned by the inspection system. It may include all of the raw output generated for all parts, all of the raw output generated for a wafer by one channel of the inspection system, and others.

  In contrast, an individual output may be generally defined as an output for individual pixels of the total output generated for a wafer by an inspection system. Thus, the raw output may include multiple individual outputs. In other words, the individual outputs may be outputs generated separately for different locations on the wafer. For example, the individual outputs can include individual separate outputs generated for different locations on the wafer. In particular, the different locations may correspond to different “inspection points” on the wafer. In other words, the different locations may correspond to locations on the wafer whose outputs are generated separately by the inspection system. In this way, different positions may correspond to each position on the wafer where a “measurement” has been performed by the inspection system. Accordingly, the different locations may vary depending on the configuration of the inspection system (eg, how the inspection system generates output for the wafer). Individual outputs include individual outputs that correspond to defects on the wafer and individual outputs that do not correspond.

  The inspection system may be configured as described herein. For example, the inspection system may be configured for a dark field (DF) inspection of a wafer. Thus, the inspection system may include a DF inspection system. The inspection system may be configured as further described herein. In another example, the inspection system may be configured for bright field (BF) inspection of a wafer. Thus, the inspection system may include a BF inspection system. The BF inspection system may have any suitable configuration known in the art. The inspection system may be configured for BF or DF inspection. Further, the inspection system may be configured as a scanning electron microscopy (SEM) inspection and observation system, and such inspection system may have any suitable configuration known in the art. . In addition, the inspection system may be configured for inspection of patterned wafers, or may be configured for inspection of unpatterned wafers.

  The computer-implemented method also includes identifying one or more characteristics of the raw output that correspond to one or more geometric characteristics of the patterned features formed on the wafer. . In one embodiment, the identified one or more characteristics of the raw output include a projection along a line in the raw output. A projection may be generally defined as a group, group, or group of individual outputs that have some pattern within the raw output. For example, projections along unprocessed output horizontal lines and circular lines may be collected. In this way, the x and y projections in the raw output that define or correspond to one or more geometric properties of the patterned features can be identified. As such, identifying one or more characteristics of the raw output may include performing a two-dimensional (2D) projection of the raw output. However, one or more characteristics of the raw output corresponding to one or more geometric characteristics of the patterned features formed on the wafer may be any other characteristic (single) of the raw output. Or plural). Identifying one or more characteristics of the raw output as described above may be performed in any suitable manner using any suitable method and / or algorithm.

  In one embodiment, the one or more geometric properties of the patterned feature are edges, shapes, textures, mathematical operations that define the geometric shape of the patterned features, or these Contains any combination. For example, properties that can be used for segmentation based on geometric shapes can be implemented as further described herein, and any mathematical operations that define edges, shapes, textures, geometric shapes / Conversion, or some combination of these. All patterned features formed on the wafer have some roughness and thus some sort of “texture”, but the texture is generally the overall texture of the patterned features Roughness refers to and describes only the roughness on the periphery of the patterned features, while referring to (eg, so designed or not so designed). In this sense, texture and roughness are different. One example of a mathematical operation / transformation that can be used to define the geometric shape of a patterned feature is a Fourier filter algorithm, which can be used between geometric shape and light scattering. Describe relationships. For example, a Fourier filter algorithm can be used to predict a projection in the raw output that will correspond to one or more geometric characteristics of the patterned features.

  In one embodiment, identifying the one or more characteristics of the raw output is based on how the design layout of the patterned features affects the one or more characteristics of the raw output. Implemented. For example, a property used for segmentation that can be implemented as described herein is the design layout. In particular, the design layout may be used to identify one or more geometric characteristics of the patterned features in the design layout. Then one or more characteristics (eg, projections) of the raw output that will correspond to one or more identified geometric characteristics (eg, empirically, theoretically, other Can be determined). In this way, one or more expected characteristics of the raw output that will correspond to one or more geometric characteristics of the patterned features may be determined. These one or more expected characteristics are then compared in any suitable manner with the one or more characteristics of the raw output, thereby providing one or more of the patterned features. One or more characteristics of the raw output corresponding to the geometric feature are identified. The design layout used in this step may be obtained in any suitable way and may have any suitable form.

  In other embodiments, identifying one or more characteristics of the raw output is performed while acquisition of the raw output is performed. In this way, identifying one or more characteristics of the raw output may be performed while a wafer is being scanned by the inspection system. For example, identifying one or more features of the raw output may be performed using a reference raw output for the wafer. The reference unprocessed output is compared to the unprocessed output to detect defects on the wafer, and the reference unprocessed output is acquired for the wafer in the same scan as the unprocessed output. The reference raw output may include any of the criteria described herein. As such, other steps described herein (eg, segmentation) performed using one or more specified characteristics of the raw output may also be used to obtain the raw output for the wafer. It may be performed during execution.

  The computer-implemented method is based on the identified one or more characteristics of the raw output such that one or more characteristics of the patterned features corresponding to each of the different segments are different. , Including assigning individual outputs in the raw output to different segments. Thus, the embodiments described herein are configured for segmentation based on geometric shapes. More specifically, the embodiments described herein describe how the geometrical characteristic (s) (eg, shape) of a wafer pattern can affect raw output and have different effects on raw output. We use how to separate the pattern that gives In other words, the embodiments described herein show how the geometrical characteristic (s) (eg, shape) of the pattern on the wafer affects the raw output, and the individual in the raw output Use how the output is separated into different segments. For example, patterned features having one or more different geometric characteristics can have different effects on the light scattered from the wafer, resulting in different on the raw output generated for the wafer May have an impact. These patterned features can be effectively separated into different segments according to the embodiments described herein. Assigning the individual outputs in the raw output to different segments as described herein may be performed in any suitable manner using any suitable method and / or algorithm.

  A “segment” may be generally defined as a different part of the entire range of possible values for an individual output. Segments can be defined based on values for different characteristics of the individual output depending on the defect detection algorithm using the segment. For example, in a multiple die auto-thresholding (MDAT) algorithm, the value for the characteristic of the individual output used to define the segment may include a median intensity. In one such illustrative and non-limiting example, if the entire intensity median range is 0 to 255, the first segment may include an intensity median from 0 to 100 and the second segment is 101 To median intensity from 255 to 255. Thus, the first segment corresponds to a darker area of unprocessed output and the second segment corresponds to a brighter area of unprocessed output. In some examples, a segment can be defined using a single wafer, and a predefined segment can be used for a wafer that requires the same geometric shape as the single wafer described above.

  In one embodiment, identifying the aforementioned one or more characteristics of the raw output and assigning the individual outputs to different segments is done automatically without user input. For example, the embodiments described herein automatically take advantage of the geometrical characteristic (s) (eg, shape) and projection of the pattern on the wafer and project the individual outputs in the raw output into different segments. Can be separated. Thus, unlike methods that involve manually setting a region of interest (ROI) and applying the same set of parameters for defect detection in the same ROI, the design rule is reduced and the wafer is segmented As the different areas above become smaller, segmentation will not become more complicated by using the embodiments described herein. In addition, unlike manual methods, automatically identifying one or more characteristics of raw output and assigning individual outputs to different segments, without user input, can lead to inspection system stage accuracy and Not affected by resolution limitations. Thus, by using the segmentation embodiments described herein, the stage accuracy and resolution limitations of the inspection system will not make segmentation impossible.

  In other embodiments, assigning individual outputs to different segments is performed regardless of design data regarding the patterned features. For example, the design layout may be used as described above to determine one or more expected geometric properties that will correspond to one or more geometric features of the patterned features. Even if it is done, segmentation is not performed based on the design data itself. In other words, segmentation is based on how one or more geometric features of the patterned feature affects the raw output, and can be one of the patterned features or It is not based on multiple geometric features themselves. Thus, unlike other methods and systems that segment raw output based on design data about patterned features, how one or more geometric characteristics of the patterned features are Different criticality for the performance of devices formed with different design data, different electrical functions, different electrical characteristics, patterned features by performing segmentation based on whether it affects raw output , Etc., the patterned features are assigned to the same segment as long as these patterned features affect the raw output in a similar manner. For example, a segment based on how the raw output geometric characteristic (s) affects the raw output characteristic (s) (eg, strength) rather than the geometry itself By implementing, the patterned features that generate significant noise in the raw output can be assigned to the same segment regardless of the design data for these pattern features and can be ignored in the raw output Other patterned features that generate noise can again be assigned to different segments, regardless of design data regarding these other patterned features. In this way, high noise patterned features are also segmented, and low noise patterned features are also segmented.

  In additional embodiments, assigning individual outputs to different segments is performed without considering the strength of the individual outputs. In other words, the segmentation is performed based on one or more specified characteristics of the raw output, which are specified based on the strength of the multiple individual outputs in the raw output, while the segmentation is individual It is not implemented based on the output intensity itself. For example, projections along a line in the raw output can include individual outputs that can have a variety of values, and in some cases include dramatically different intensities. Nevertheless, all of the individual outputs may correspond to the same one or more geometric characteristics of the patterned features, such as page breaks. As such, all of the individual outputs corresponding to the same one or more geometric characteristics of the patterned features are assigned to the same segment, even though all of the individual outputs have dramatically different intensities. Can be. Thus, unlike methods that perform segmentation based on the intensity of individual pixels, the segmentation performed by the embodiments described herein is affected by non-uniform dispersion from the patterned features. Will not be.

  In some embodiments, assigning individual outputs to different segments includes analyzing the identified one or more characteristics of the raw output and applying a threshold to the individual outputs. For example, as described above, projections along the horizontal and vertical lines in the raw output may be collected. The projection can then be analyzed and a threshold is set to separate the individual outputs in the raw output into different areas of interest (segments). Analyzing one or more characteristics of the raw output and applying thresholds to the individual outputs may reduce the number of individual outputs corresponding to the boundary region from being inappropriately assigned to the aforementioned segments.

  In one embodiment, the one or more geometric characteristics corresponding to one of the different segments may include one or more geometric characteristics of the page break, and the other of the different segments. The one or more geometric characteristics corresponding to one segment of the array include one or more geometric characteristics of the array area. A page break is generally defined in the art as a region of a die that separates a substantially contiguous region of physical memory. Each of the contiguous regions of physical memory can generally be referred to as a page frame. When segmentation is performed as described herein, one or more characteristics (eg, x-projection and / or y-projection) of the raw output that define the geometry for page breaks in the array area are: Can be identified and used to assign individual outputs corresponding to page breaks to one segment and to assign individual outputs corresponding to array areas to different segments.

  In other embodiments, one or more characteristics of the raw output corresponding to one or more geometric features of some of the patterned features may be suppressed by a Fourier filter. Absent. For example, unlike some methods for segmentation, page breaks can be suppressed in the array area even though the distance between page breaks is larger than a Fourier filter is feasible. In one such example, for some inspection systems, manual setting of the ROI is not possible if the page break width is about 5 micrometers and the interval between page breaks is about 5 micrometers. While not impractical, the Fourier filter becomes impractical if not impossible. Thus, the signal (noise) generated in the raw output by the page break cannot be suppressed and thus can reduce the defect detection sensitivity that can be achieved using the raw output. However, with the embodiments described herein, an individual output corresponding to a page break can be identified (eg, based on a projection in the raw output) and corresponds to the page break described above. Individual outputs to be assigned to one segment, and other individual outputs can be assigned to other segments. As a result, as described further herein, different sensitivities can be used to detect different defects in different segments.

  The computer-implemented method further includes assigning one or more defect detection parameters to different segments separately. One or more defect detection parameters may be assigned separately to all of the different segments. Thus, some of the individual outputs cannot be ignored for defect detection. Instead, defects can be detected using individual outputs assigned to all of the different segments. In other words, defects can be detected using all segments of raw output. In this way, different segments can be processed differently using different detection methods. Different detection methods may be different in defect detection algorithms assigned to different segments. Alternatively, different detection methods may differ in one or more parameters of the same defect detection algorithm assigned to different segments. The defect detection algorithm assigned to different segments, or one or more parameters assigned to different segments, may include any suitable defect detection algorithm. For example, the defect detection algorithm may be a segmented auto-thresholding (SAT) algorithm or an MDAT algorithm. Such a detection algorithm is particularly suitable for BF inspection. However, the defect detection algorithm may be a defect detection algorithm suitable for DF inspection. For example, the defect detection algorithm may be a FAST algorithm or an HLAT algorithm.

  Different detection methods may differ in one or more optical parameters of the inspection system used to obtain the raw output for the wafer. For example, in multi-pass inspection, different passes may be performed using different values for at least one optical parameter (eg, polarization, wavelength, illumination angle, convergence angle, etc.) of the inspection system. The generated raw output may be used to detect defects in different regions of the wafer where patterned features having one or more different geometric characteristics are formed. In this way, regions of the wafer containing patterned features having one or more different geometric characteristics are generated in different passes of a multi-pass inspection performed using one or more different optical parameters. Can be examined using the processed raw output.

  In one embodiment, the one or more defect detection parameters include a threshold applied to the difference between the individual output and the reference. In this way, different thresholds can be applied for the difference between the individual output and the reference, depending on the segment to which the individual output is assigned. For example, a reference (eg, an 8-bit reference image) can be subtracted from an individual output (such as an 8-bit test image) in the raw output, regardless of the segment to which the individual output is assigned. The reference is an individual output corresponding to a die on the wafer that is different from the die from which the individual output for which the reference is drawn is generated, and an individual output corresponding to a cell on the wafer that is different from the cell from which the individual output for which the reference is drawn is generated. Any suitable criteria such as, etc. may be included. Any individual output that has a difference that exceeds the assigned threshold may be identified as a defect. In this way, defects can be detected using different thresholds depending on the segment to which the individual output is assigned.

  In other embodiments, assigning one or more defect detection parameters separately to different segments is performed so that defects are detected using individual outputs assigned to different segments having different sensitivities. . Thus, the embodiments described herein achieve better defect detection by utilizing the knowledge that defects of interest (DOI) and jamming / noise exist in geometrically different segments. Is possible. For example, different geometric shapes may represent different types of defects. In one such example, in the array pattern region, the raw output may include a linear pattern with alternating relatively bright individual outputs and relatively dark individual outputs. In some such examples, the DOI may be placed in a portion of the raw output that includes a relatively bright individual output, while the jamming defect may be placed in a region of the raw output that contains a relatively dark individual output. Thus, the sensitivity of the detection algorithm improves the sensitivity in the array area by segmentation using the characteristic (s) that define the geometry (eg, x-projection or y-projection for page breaks in the array area). And can be set differently to reduce interference from page breaks. Thus, the embodiments described herein advantageously allow an automatic method of separating different geometric patterns of a wafer into different segments. This segmentation allows these areas to be treated differently, so that better sensitivity can be achieved. Different geometric shapes also scatter light in different ways. In this way, some geometric shapes may have relatively large raw output noise, and other geometric shapes may have relatively small raw output noise. However, if only the intensity of the individual output is used for segmentation, the individual output corresponding to the region with relatively high noise and the region with relatively low noise in the raw output (for example, the boundary definition is insufficient Can be grouped as well). In contrast, in the embodiments described herein, higher due to defects placed in areas of the wafer having one or more geometric characteristics corresponding to less noise in the raw output. Sensitivity can be achieved. In addition, in narrowband inspection systems, defects can often be buried in noise because the pattern also disperses a significant amount of light. However, the embodiments described herein allow the detection of defects attenuated by noise from adjacent patterns.

  The computer-implemented method includes applying the assigned defect detection parameter or parameters to individual outputs assigned to different segments, thereby detecting defects on the wafer. As described above, different segments can be processed differently using different inspection methods. In this manner, applying the assigned defect detection parameter or parameters to the individual outputs may include inspecting using different methods, thereby detecting defects on the wafer. For example, a segment that is assigned an individual output can be used to determine a threshold that is applied to the difference between the individual output and the reference. After determining the segments to which the individual outputs are assigned and assigning one or more defect detection parameters to different segments, the assigned one or more defect detection parameters are assigned to the different segments as is normally performed. Can be assigned to assigned individual outputs.

  In one embodiment, obtaining the raw output is performed in one of the multiple-pass inspections of the wafer, and the computer-implemented method is an unprocessed acquisition in the other passes of the multiple-pass inspection. It is not implemented for processing output. Thus, the segmentation described herein can only be performed on one pass of a multi-pass inspection. The raw output obtained in other passes can be used for other purposes. For example, multi-pass inspection fits the purpose of segmentation using one pass with the best signal for the defect and another pass that provides segmentation based on geometry. In particular, one or more different defect detection parameters and / or one or more different optics so that the different paths of the multi-pass inspection will have different raw outputs and / or defect detection results for different paths. It may be implemented using parameters. In one such example, one optical mode used in one pass of a multi-pass inspection may allow segmentation, while another optical mode of the inspection system used in other passes of the multi-pass inspection is DOI. Can provide the highest sensitivity.

  In other embodiments, additional defects are detected by using raw output obtained in other passes, and the method includes generating inspection results for the wafer by combining the defects and the additional defects. Including. For example, as described above, one path of the multi-pass inspection may be used for segmentation and the other path of the multi-path inspection may be used to detect the DOI with the optimal signal. Thus, different paths in a multi-pass inspection can detect different types of defects. In this way, the results of different passes of the multiple pass inspection may be combined to produce an overall inspection result for the wafer. The results of defects detected using raw outputs acquired in different passes may be combined after defect detection is performed using raw outputs generated in all of the different passes. Alternatively, defect detection results generated using raw output acquired in different passes may be combined during execution or while some of the raw output is still being acquired.

  In an additional embodiment, the method detects additional defects on the wafer by applying one or more predetermined defect detection parameters to the raw output and combines the aforementioned defects with the additional defects. Generating an inspection result for the wafer. For example, a reference (eg, an 8-bit reference image) can be subtracted from an individual output (such as an 8-bit test image) in the raw output, regardless of the segment to which the individual output is assigned. The criteria may include any suitable reference as described above. In addition, the same criteria detect defects by applying one or more assigned defect detection parameters to the individual outputs and applying one or more predetermined defect detection parameters to the raw output. Can be used to The result of the subtraction can be an absolute difference. A predetermined direct difference threshold may then be applied to the absolute difference and any individual output having an absolute difference that exceeds the direct difference threshold may be identified as a defect. In addition, the same predetermined direct difference threshold may be applied to the absolute difference regardless of the segment to which the individual output is assigned. The defects thus detected can then be combined with the detected defects by applying one or more assigned defect detection parameters to the individual outputs to produce a final inspection result for the wafer. Good. For example, a defect mask may be generated separately for all defects detected in any manner. A region “growth” is performed from both different images, so that a final mask for all defects can be generated.

  As described above, detecting defects in different ways may provide defect rediscovery. This can be advantageous for several reasons. For example, segmentation based on automatic 2D projection and geometry provides robust defect redetection and improved usability for defect redetection. In addition, the segmentation described herein provides a dynamic method for mapping defects and reference images. For example, if the segment is noisy, the difference can be attenuated. In contrast, if the segment noise is small, the difference can be increased. In addition, the double detection described above reduces the possibility of false alarms from either detection method.

  The method also includes storing the results of any step of the method on a storage medium. This result can include any result described herein and can be stored in any manner known in the art. For example, a segment assigned an individual output and / or one or more defect detection parameters assigned to different segments may be a data structure such as a lookup table stored on a storage medium connected to the inspection system. May be used to generate The storage medium may include any suitable storage medium known in the art. After the results are stored, the results are accessed on the storage medium, formatted for display to the user, and used by other software modules, methods or systems, etc. as described herein. Can be used. Furthermore, the results can be stored “permanently”, “semi-permanently”, “temporarily” or for some period of time. For example, the storage medium can be a random access memory (RAM) and the results are not always stored in the storage medium. Storing the results may be performed as described in Applicant's US Patent Application Serial No. 12 / 234,201 filed September 19, 2008 by Bhaskar et al. This U.S. Patent Application Serial No. 12 / 234,201 was published as U.S. Patent Application Publication No. 2009/0080759 on March 26, 2009, and as if fully described herein. Which is incorporated herein by reference.

  Referring now to the drawings, it should be noted that the scale factor is not constant in the drawings. In particular, the scale of some of the elements in the drawing is greatly exaggerated to emphasize the characteristics of the elements. It should be noted that not all drawings are drawn to the same scale. Elements having similar configurations shown in multiple figures are indicated using the same reference numerals.

  Another embodiment relates to a computer-readable medium that includes program instructions executable on a computer system that performs a method for detecting defects on a wafer (ie, a computer-implemented method). One such example is shown in FIG. For example, as shown in FIG. 1, a computer readable medium 10 includes program instructions 12 executable on a computer system 14 that performs a method for detecting defects on a wafer as described above. A computer-implemented method capable of executing the aforementioned program instructions may include any other step (s) of any other method (s) described herein.

  Program instructions 12 implementing methods such as those described herein may be stored on computer readable medium 10. The computer readable medium may be a computer readable medium, such as read only memory, RAM, magnetic or optical disk, magnetic tape, or any other computer readable medium known in the art.

  Program instructions may 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, program instructions may use Matlab, Visual Basic, ActiveX controls, C, C ++ objects, C #, Java Beans, Microsoft Foundation Classes (“MFC”), or other techniques or methodologies as desired. Can be implemented.

  The computer system 14 may take various forms including a personal computer system, mainframe computer system, workstation, system computer, imaging computer, programmable imaging computer, parallel processor, or any other device known in the art. It can take. In general, the term “computer system” may be broadly defined as including any device having one or more processors and executing instructions from a memory medium.

  An additional embodiment relates to a system configured to detect defects on a wafer. One such system is shown in FIG. As shown in FIG. 2, the system 16 includes an inspection subsystem 18 and a computer subsystem 20. The inspection subsystem is configured to generate raw output for the wafer by scanning the wafer. For example, as shown in FIG. 2, the inspection subsystem includes a light source 22 such as a laser. The light source 22 is configured to guide light to the polarizing component 24. In addition, the inspection subsystem comprises one or more polarizing components (not shown), each of which may be independently placed in the light path from the light source. Each of the polarizing components may be configured to change the polarization of light from the light source in different ways. The inspection subsystem moves the polarizing component to enter or leave the optical path from the light source, depending on which polarization setting is selected to illuminate the wafer during the scan. Configured in any suitable manner. Polarization settings used for wafer illumination during scanning can include P-polarized light (P), S-polarized light (S), or circularly polarized light (C).

  Light exiting the polarizing component 24 is directed to the wafer 26 at an acute incident angle, which may include any suitable acute incident angle. The inspection subsystem may include one or more optical elements (not shown) configured to direct light from the light source 22 to the polarizing component 24 or from the polarizing component 24 to the wafer 26. The optical component may include any suitable optical component known in the art, including but not limited to reflective optical components. In addition, the light source, polarizing component, and / or one or more optical components may emit light at one or more angles of incidence (eg, acute angles of incidence, or substantially perpendicular angles of incidence). It can be configured to lead to a wafer. The inspection subsystem may be configured to perform the scan by scanning light over the wafer in any suitable manner.

  Light scattered from the wafer 26 can be collected and detected by multiple channels of the inspection subsystem during the scan. For example, light scattered from the wafer 26 at a relatively near right angle can be collected by the lens 28. The lens 28 may include a refractive optical element shown in FIG. In addition, the lens 28 may comprise one or more refractive optical elements and / or one or more reflective optical elements. The light collected by the lens 28 can be guided to the polarizing component 30. The polarizing component 30 may include any suitable polarizing component known in the art. In addition, the inspection subsystem comprises two or more polarizing components (not shown), each of which may be independently placed in the optical path from the light source collected by the lens. Each of the polarizing components may be configured to change the polarization of the light collected by the lens in different ways. The inspection subsystem may enter or leave the optical path collected by the lens, depending on which polarization setting is selected to detect the light collected by the lens 28 during the scan. As such, it can be configured in any suitable manner to move the polarizing component. The polarization settings used to detect the light collected by lens 28 during the scan can be any of the polarization settings described herein (eg, P-polarized, S-polarized, and unpolarized (N)). May be included.

  Light exiting the polarizing component 30 is guided to the detector 32. The detector 32 may include any suitable detector known in the art, such as a charge coupled device (CCD) or other type of imaging detector. The detector 32 is configured to generate an unprocessed output that is responsive to the scattered light collected by the lens 28 and transmitted by the polarizing component 30 when placed in the optical path of the collected scattered light. Accordingly, the lens 28 and the polarizing component 30 and the detector 32 when arranged in the optical path condensed by the lens 28 form one channel of the inspection subsystem. This channel of the inspection subsystem may include any other suitable optical component (not shown) known in the art, such as a Fourier filter component.

  Light scattered from the wafer 26 at different angles can be collected by the lens 34. The lens 34 may be configured as described above. The light collected by the lens 34 may be guided to the polarization component 36. The polarizing component 36 may include any suitable polarizing component known in the art. In addition, the inspection subsystem comprises two or more polarizing components (not shown), each of which may be independently placed in the optical path from the light source collected by the lens. Each of the polarizing components may be configured to change the polarization of the light collected by the lens in different ways. The inspection subsystem may enter or leave the optical path collected by the lens, depending on which polarization setting is selected to detect the light collected by the lens 34 during the scan. As such, it can be configured in any suitable manner to move the polarizing component. The polarization setting used to detect the light collected by lens 34 may include P, S, or N.

  The light emitted from the polarization component 36 is guided to the detector 38. The detector 38 may be configured as described above. The detector 38 is also configured to produce a raw output that is responsive to the collected scattered light passing through the polarizing component 36 when placed in the optical path of the scattered light. Accordingly, the lens 34, the polarizing component 36 and the detector 38 when placed in the optical path collected by the lens 34 form another channel of the inspection subsystem. This channel may include any other optical components (not shown) described above. In some embodiments, lens 34 may be configured to collect scattered light from a polar angle wafer of about 20 degrees to about 70 degrees. In addition, the lens 34 may be configured as a reflective optic (not shown) configured to collect light scattered from the wafer at an azimuth angle of about 360 degrees.

  The inspection subsystem shown in FIG. 2 may include one or more other channels (not shown). For example, the inspection subsystem includes an additional channel that includes any lens, one or more polarizing components, and a detector configured as a side channel. Parts can be provided. The lens, one or more polarizing components, and the reflector may be further configured as described herein. In one such example, the side channel may be configured to collect and detect light scattered away from the entrance surface (eg, the side channel is configured to include a lens and a reflector, the lens being , Located in the center of a plane substantially perpendicular to the entrance plane, the detector being configured to detect the light collected by the lens).

  If inspection of the wafer includes one or more passes, the values of any optical parameters of the inspection subsystem may be changed between passes in any suitable manner, if necessary. For example, the polarization component 24 may be removed and / or replaced with a different polarization component, as described herein, to change the illumination change state between passes. In other examples, the position of the light source and / or any other optical component (eg, polarizing component 24) used to direct light to the wafer to change the illumination angle between passes is any suitable May be changed between paths in a manner.

  The computer subsystem 20 is configured to obtain the raw output generated by the inspection subsystem. For example, the raw output generated by the detector during the scan may be provided to the computer subsystem 20. In particular, the computer subsystem may include any suitable transmission medium known in the art (eg, any known in the art) so that the computer subsystem can receive the raw output generated by the detector. , By one or more transmission media indicated by dotted lines in FIG. The computer subsystem may be connected to each of the detectors in any suitable manner. The raw output generated by the detector while scanning the wafer may include any raw output described herein.

  The computer subsystem may provide one or more raw outputs corresponding to one or more geometric characteristics of the patterned features formed on the wafer in accordance with any embodiment described herein. Configured to identify characteristics of The one or more characteristics of the raw output may include any such characteristics described herein. The one or more geometric characteristics may include any such characteristics described herein. Patterned features may include any of the patterned features described herein.

  In addition, the computer subsystem includes the identified one or more of the raw outputs so that one or more geometric characteristics of the patterned features corresponding to each of the different segments are different. The individual outputs in the raw output are configured to be assigned to different segments based on the characteristics of The computer subsystem may be configured to assign individual outputs to different segments in accordance with any embodiment described herein. An individual output may include any individual output described herein. Different segments may be configured as described herein. The identified one or more characteristics of the raw output may include any such characteristics described herein.

  The computer subsystem is further configured to separately assign one or more defect detection parameters to different segments in accordance with any embodiment described herein. The one or more defect detection parameters may include any defect detection parameters described herein. The computer subsystem is also configured to detect defects on the wafer by assigning the assigned defect detection parameter or parameters to individual outputs assigned to different segments. Note that this detection can be performed according to any embodiment described herein. The assigned defect detection parameter or parameters may include any such parameters described herein.

  The computer subsystem may be configured to perform any other step (s) of any method embodiment (s) described herein. The computer subsystem, inspection subsystem, and system may be further configured as described herein.

  It should be noted that FIG. 2 is provided herein to generally illustrate one configuration of an inspection subsystem that may be included in the system embodiments described herein. Obviously, the inspection subsystem configuration described herein may be modified to optimize the performance of the inspection subsystem, as is usually done when designing a commercial inspection system. In addition, the system described herein uses existing inspection systems, such as the Puma 90xx, 91xx, and 93xx series instruments that are commercially available from KLA-Tencor, Milpitas, Calif. ( For example, it may be implemented by adding the functionality described herein to an existing inspection system. For some such systems, the methods described herein may be provided as an optional feature of the system (eg, added to other features of the system). Alternatively, the system described herein may be designed from a “blank state” to provide a completely new system.

  Further modifications and alternative embodiments of the various aspects of the present invention will become apparent to those skilled in the art in view of this description. For example, methods and systems are provided for detecting defects on a wafer. Accordingly, this 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 invention. It should be understood that the form of the invention shown and described herein is the presently preferred embodiment. After having the convenience of describing the present invention, elements and materials may be replaced with those illustrated and described herein, and parts and processes may be reversed, as will be apparent to those skilled in the art. Certain features of the invention may be utilized independently. Changes in the elements described herein may be made without departing from the spirit and scope of the invention as described in the following claims.

Claims (19)

Obtain raw output for wafers generated by the inspection system,
Identifying one or more characteristics of the raw output corresponding to one or more geometric characteristics of the patterned features formed on the wafer;
Based on the identified one or more characteristics of the raw output such that the one or more geometric characteristics of the patterned features corresponding to each of the different segments are different, Assign individual outputs in the raw output to different segments;
Assigning one or more defect detection parameters to the different segments separately;
Detecting defects on the wafer by assigning the one or more defect detection parameters to the individual outputs assigned to the different segments;
A computer implemented method for detecting defects on a wafer.   The method of claim 1, wherein the raw output is responsive to light scattered from the wafer.   The method of claim 1, wherein the identified one or more characteristics of the raw output include a projection along a line in the raw output.   The one or more geometric characteristics of the patterned feature may be edges, shapes, textures, mathematical operations that define the geometric shape of the patterned features, or some combination thereof. The method of claim 1 comprising.   The method of claim 1, wherein the identification is performed based on how a design layout of the patterned features affects the one or more characteristics of the raw output.   The method of claim 1, wherein the identification is performed while the acquisition is performed.   The method of claim 1, wherein the identification and the assignment of the individual outputs is performed automatically without user input.   The method of claim 1, wherein the assignment of the individual outputs is performed regardless of design data relating to the patterned features.   The method of claim 1, wherein the assignment of the individual outputs is performed regardless of the strength of the individual outputs.   The method of claim 1, wherein the assignment of the individual outputs comprises analyzing the identified one or more characteristics of the raw output and applying a threshold to the individual outputs.   The one or more geometric characteristics corresponding to one of the different segments include one or more geometric characteristics of a page break and correspond to the other one of the different segments The method of claim 1, wherein the one or more geometric characteristics include one or more geometric characteristics of an array area.   The one or more characteristics of the raw output corresponding to the one or more geometric characteristics of some of the patterned features cannot be suppressed by a Fourier filter. The method according to 1.   The method of claim 1, wherein the one or more defect detection parameters include a threshold applied to a difference between the individual output and a reference.   2. The separately assigning the one or more defect detection parameters is implemented such that defects are detected using the individual outputs assigned to the different segments having different sensitivities. the method of.   The acquisition is performed in one pass of a multi-pass inspection of the wafer, and the computer-implemented method is not performed on raw output acquired in another pass of the multi-pass inspection. Item 2. The method according to Item 1.   The acquisition is performed in one pass of the multi-pass inspection of the wafer, and the computer-implemented method is not performed on raw output acquired in other passes of the multi-pass inspection, Additional defects are detected using the raw output acquired in the other pass, and the method generates an inspection result for the wafer by combining the defects and the additional defects. The method of claim 1, further comprising:   Applying one or more predetermined defect detection parameters to the raw output to detect additional defects on the wafer and generate inspection results for the wafer by combining the defects and the additional defects The method of claim 1, further comprising: A computer readable medium comprising program instructions executable on a computer system for performing a method for detecting defects on a wafer, the method comprising:
Obtain raw output for wafers generated by the inspection system,
Identifying one or more characteristics of the raw output corresponding to one or more geometric characteristics of the patterned features formed on the wafer;
Based on the identified one or more characteristics of the raw output such that the one or more geometric characteristics of the patterned features corresponding to each of the different segments are different, Assign individual outputs in the raw output to different segments;
Assigning one or more defect detection parameters to the different segments separately;
Detecting defects on the wafer by assigning the one or more defect detection parameters to the individual outputs assigned to the different segments;
A computer-readable medium including: An inspection subsystem configured to generate raw output for the wafer by scanning the wafer;
Obtain the raw output,
Identifying one or more characteristics of the raw output corresponding to one or more geometric characteristics of the patterned features formed on the wafer;
Based on the identified one or more characteristics of the raw output such that the one or more geometric characteristics of the patterned features corresponding to each of the different segments are different, Assign individual outputs in the raw output to different segments;
Assigning one or more defect detection parameters to the different segments separately;
A computer subsystem configured to detect defects on the wafer by assigning the one or more defect detection parameters to the individual outputs assigned to the different segments;
A system configured to detect defects on the wafer.

Images