JP2010091425A - Device and method for inspecting defect - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain three images which are to be compared in double detection through main scanning toward the same direction, and to perform main scanning in both the normal and reverse directions, in defect inspection for scanning a sample surface on which a repeated pattern where a plurality of basic patterns are repeated is formed, and comparing each image on corresponding parts of the plurality of respective basic patterns. <P>SOLUTION: This defect inspection device 1 includes an imaging means 13 for imaging the sample surface; a moving means 11 for moving relatively the imaging means 13 and the sample; a scanning control means 40 for two-dimensionally scanning the sample by the imaging means 13 by relative movement; an image storage means 22 for storing an image obtained by the main scanning of twice or more, including the main scanning in a reverse direction performed on the three or more basic patterns, a selection means 23 for selecting an image obrtained by the main scanning toward the same direction of the corresponding part of the plurality of basic patterns in the stored image; a defect candidate detection means 24 for detecting a defect candidate by comparing selected images; and a determination means 25 for determining in which basic pattern the defect candidate is located. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  In the present invention, the surface of a sample on which a repetitive pattern in which a plurality of basic patterns repeat is formed is scanned two-dimensionally with an imaging means, and images of corresponding portions of the plurality of basic patterns are compared with each other, The present invention relates to a defect inspection apparatus and a defect inspection method for detecting defects.

  Manufacture of semiconductor wafers, liquid crystal display panels, and the like is composed of many man-hours, and it is important for improving the yield to inspect the occurrence of defects in the final and intermediate processes and feed back to the manufacturing process. In order to detect defects during the manufacturing process, pattern defect inspection is widely performed to detect defects existing on the sample surface by imaging the pattern formed on the surface of the sample and inspecting the resulting image. It has been broken.

  A defect detection method based on image comparison is widely used as one of pattern defect inspections. In the defect detection method based on image comparison, a plurality of basic patterns are assumed on the premise that the same defect is likely to occur in the same place between the basic patterns that are the same pattern repeatedly arranged on the sample. Defects are detected by comparing each other and obtaining the difference. For example, a plurality of dies are formed on a semiconductor wafer. For this reason, it is possible to detect a defect by die-to-die comparison in which images of a plurality of dies are compared using a single die pattern as a basic pattern.

  FIG. 1 is an explanatory diagram of die-to-die comparison of patterns formed on a wafer. In this example, image comparison is performed between the die 111 and the die 112 formed on the wafer 100, and between the die 112 and the die 113, and a different portion is detected as a defect. The reason why the die 112 is compared not only with the die 111 but also with the die 113 is that if a difference is detected in the comparison between the dies 111 and 112, it is not known which of the dies 111 and 112 is caused by the difference. Because. The image comparison process in which one die is compared with two or more other dies in this way is called double detection.

JP 2004-177397 A JP-A-2-210249

  In the die arrangement example shown in FIG. 1, dies 111 to 119 are arranged on a wafer 100 in 3 rows and 3 columns. Therefore, when double detection processing is performed on the die 112 as compared with the adjacent dies 111 and 113 on both sides, it is necessary for the double detection processing if the main scanning is performed on the wafer 100 along the path indicated by reference numeral 120. The various dies 111 to 113 can be scanned by one main scan.

  However, in the actual defect inspection, there are cases where the three basic patterns necessary for the double detection process cannot be scanned by one main scan. As one of them, it can be considered that the number of basic patterns arranged along the main scanning direction is less than three. For example, when large-sized dies are formed on the wafer 100, the number of dies arranged along the main scanning direction may be less than three.

  In addition, defect detection by image comparison is used not only for inspection of products such as semiconductor wafers and semi-finished products, but also for inspection of reticles and masks used to transfer patterns to such products and semi-finished products. . Since the reticle pattern is reduced and transferred onto a wafer or the like, the size of the die on the reticle is larger than the pattern actually transferred onto the wafer or the like. For this reason, as compared with the wafer 100, in the case of a reticle, the number of dies arranged along the main scanning direction is often less than three.

  FIG. 2 is an explanatory diagram of die-to-die comparison of patterns formed on a reticle. In this example, dies 211 to 214 in 2 rows and 2 columns are arranged on the reticle 200. Since there are no three dies in the direction of the paths 220 and 230 in which the main scanning is performed, in this arrangement, in addition to the dies 211 and 212 belonging to the same column, the dies 214 belonging to different columns are scanned. A double detection process is performed using the acquired image.

  In order to efficiently perform relative movement between the reticle 200 and the imaging means for scanning the reticle 200, the main scanning direction for scanning the dies 211 and 212 and the main scanning direction for scanning the die 214 are reversed. Has been done. In this case, the image of the die 214 to be compared with the image of the die 211 and / or the die 212 is an image acquired by scanning in the opposite direction to the scan for acquiring the image of the die 211 and the die 212.

  If the two images to be compared are acquired by scanning in the opposite directions, the result of the image comparison may be affected due to the difference in imaging conditions. For example, when a sensor whose characteristics change depending on the scanning direction is used as an imaging sensor that scans a sample, a difference in characteristics depending on the scanning direction may appear as a gray level difference, which may reduce the defect detection sensitivity. There is a TDI camera as such an image sensor. Further, it is conceivable that the behavior of a relative movement mechanism between the imaging means and the sample, such as a stage for moving the sample, also varies depending on the scanning direction.

  For this reason, from the viewpoint of preventing the influence of the difference in the main scanning direction on the captured image, it is desirable that the two images compared in the defect inspection are images obtained by the main scanning proceeding in the same direction. . FIG. 3 is an explanatory diagram of a scanning path for acquiring partial images to be compared with each other by a main scanning that proceeds in the same direction between a plurality of dies belonging to different columns. Scan paths 401 to 404 indicate main scan scan paths, and these main scans are scans that proceed in the forward direction indicated by an arrow 400.

  In one main scan, a band-like image corresponding to the width of the image sensor used for the scan is acquired. The belt-like region 211-1 of the die 211, the belt-like region 212-1 of the die 212, the belt-like region 213-1 of the die 213, and the belt-like region 214-1 of the die 214 are between the dies 211 to 214 which are basic patterns. Are areas corresponding to each other. Similarly, the band-like region 211-2 of the die 211, the band-like region 212-2 of the die 212, the band-like region 213-2 of the die 213, and the band-like region 214-2 of the die 214 are the basic pattern of the die 211. This is a region corresponding to each other in a range of ~ 214.

  Then, images of the region 211-1 and the region 212-1 are acquired by the main scanning 401. Further, the images of the region 213-1 and the region 214-1 are acquired by the main scanning 402. Images of the region 211-2 and the region 212-2 are acquired by the main scanning 403. Further, the images of the region 213-2 and the region 214-2 are acquired by the main scanning 404.

  According to this scanning path, the images of the regions 211-1 to 214-1 in the corresponding regions of the dies 211 to 214 can be acquired by main scanning in the same direction, and the regions 211-2 to 214-in the corresponding regions. Two images can be acquired by main scanning in the same direction.

  When this scanning path is used, it is opposite to the main scanning direction as indicated by arrows 411, 412 and 413 between the completion of each main scanning 401 to 403 and the start of the next main scanning 402 to 404. It is necessary to move the scanning position in the direction. Since the dies 211 to 214 are not scanned during the movement in the reverse direction, the total amount of movement between the reticle and the imaging means required for defect inspection is increased, resulting in a decrease in inspection speed.

  In addition, as the case where the three dies required for the double detection process cannot be scanned by one main scanning, the following cases can be considered. For example, when the number of dies arranged along the main scanning direction in the peripheral part of the substrate or wafer is less than 3 due to the reason that the substrate or wafer on which the dies are formed is not rectangular, the peripheral part has 1 Three dies cannot be scanned in one main scan.

  In addition, when the defect detection process is performed for each die and the algorithm for performing double detection in comparison with the adjacent dies on both sides of the die is always used in the defect detection process of each die, When this algorithm is applied to the dies arranged in the outer row, the adjacent die exists only on one side, so that the die required for the double detection process cannot be scanned in one main scan. As the die size increases and the number of dies arranged on one wafer or reticle decreases, the ratio of the number of dies on the outer edge to the total number of dies increases. For this reason, the proportion of dies that cannot be subjected to double detection processing in one main scan becomes larger.

  According to the present invention, the surface of a sample on which a repetitive pattern in which a plurality of basic patterns repeats is formed is two-dimensionally scanned with an imaging unit, and images of corresponding portions of the plurality of basic patterns are compared with each other. In defect inspection for detecting defects, an object is to acquire three images to be compared by double detection by main scanning in the same direction, and to acquire images in main scanning in either forward or reverse direction.

  A defect inspection apparatus according to an embodiment of the present invention is a defect inspection apparatus for a sample in which a repetitive pattern in which a plurality of basic patterns repeat at least in the column direction among the row direction and the column direction is formed on the surface. Imaging means for imaging, moving means for relatively moving the imaging means and the sample in at least a two-dimensional direction, and controlling the relative movement of the imaging means and the sample by the moving means so that the row direction and the column direction are respectively in the main scanning and sub scanning directions. An image that simultaneously stores images obtained by two or more main scans that include main scans in mutually opposite directions and that are performed on three or more basic patterns, respectively. An image obtained by performing main scanning in the same direction on mutually corresponding portions of a plurality of basic patterns from among the storage means and the images stored in the image storage means A selection means for selecting each, a defect candidate detection means for performing image comparison between the images selected by the selection means and detecting a different portion as a defect candidate, and a defect candidate being any defect of a plurality of basic patterns Determining means for determining whether or not there is.

  FIG. 4 is a diagram for explaining which range of captured images is stored simultaneously in the embodiment of the present invention. A plurality of basic patterns 511 to 513 are formed on the sample 500. The band-shaped area 511-1 of the basic pattern 511, the band-shaped area 512-1 of the basic pattern 512, and the band-shaped area 513-1 of the basic pattern 513 are areas corresponding to each other between the basic patterns 511 to 513. Similarly, the band-like region 511-2 of the basic pattern 511, the band-like region 512-2 of the basic pattern 512, and the band-like region 513-2 of the basic pattern 513 are mutually in the basic patterns 511 to 513. The corresponding area.

  The images of the region 511-1 and the region 512-1 are acquired by the main scanning 401 directed in the forward direction indicated by the arrow 400. Further, the images of the region 511-2 and the region 512-2 are acquired by the main scanning 402 directed in the reverse direction indicated by the arrow 400. Further, an image of the region 513-1 is acquired by the main scanning 403 toward the forward direction. In addition, an image of the region 513-2 is acquired by the main scanning 404 directed in the reverse direction.

  According to the embodiment of the present invention, for each of the basic patterns 511, 512, and 513, which are three or more basic patterns, an image obtained by two or more main scans including the main scan in the opposite direction, that is, the region 511- Images of 1 and 511-2, areas 512-1 and 512-2, and areas 513-1 and 513-2 are simultaneously stored in the storage unit.

  Therefore, double detection can be performed using the images of the regions 511-1 to 513-1 when the basic patterns 511, 512, and 513 are scanned in the forward direction only.

  On the other hand, in order to store the images of the regions 511-1 to 513-1 and the regions 511-2 to 513-2 scanned by two or more main scans including scanning in the opposite direction in the storage means, While performing the main scanning in the forward direction, it is possible to store the image acquired by performing the main scanning in the reverse direction. Alternatively, the image obtained by performing the main scanning in the forward direction can be stored while performing the main scanning in the reverse direction twice. For this reason, an image of a basic pattern can be acquired in main scanning in either the forward or reverse direction.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. FIG. 5 is an overall configuration diagram of the defect inspection apparatus according to the first embodiment of the present invention. The defect inspection apparatus 1 includes a stage 11, a sample stage 12, an imaging device 13, an image processing unit 20, and a stage control unit 40.

  The stage 11 can move the sample stage 12 in a three-dimensional direction consisting of a horizontal XY plane and a Z-axis direction perpendicular thereto. The sample stage 12 is provided on the upper surface of the stage 11 and includes a chuck mechanism for holding a reticle 200 as a sample. When the stage 11 moves while the sample stage 12 holds the reticle 200, relative movement between the reticle 200 and the imaging device 13 is realized. Alternatively, the reticle 200 and the imaging device 13 may be moved relative to each other by moving the imaging device 13 side, or both the reticle 200 and the imaging device 13 may be moved.

  The imaging device 13 is provided above the sample stage 12 and captures an optical image of the surface of the reticle 200. As the imaging device 13, an image sensor such as a one-dimensional TDI camera is used, and an optical image of the surface of the reticle 200 formed on the light receiving surface thereof is converted into an electric signal. In this configuration example, a one-dimensional TDI camera is used as the imaging device 13. As the optical system of the imaging device 13, either a bright field illumination optical system or a dark field illumination optical system may be used.

  The imaging device 13 outputs a so-called swath image having a width corresponding to the sensor width. The image signal output from the imaging device 13 is input to the image processing unit 20 after being converted into a multi-value digital signal (gray level signal).

  The stage control unit 40 drives the stage 11 to scan the surface of the reticle 200 with the imaging device 13 along a predetermined scanning path. The scanning path information 51 for specifying the scanning path of the reticle 200 by the imaging device 13 is given in advance by an operator, for example. Further, the scanning path information 51 is automatically determined by the stage control unit 40 or other calculation means by calculation according to the dimension information of the reticle 200 and the dimension information and coordinate information of the die formed on the reticle 200. It may be information.

  The image processing unit 20 processes the image output from the imaging device 13 to detect a defect existing on the surface of the reticle 200, and generates defect information including the coordinates of the detected defect. The image processing unit 20 includes an image data unit generation unit 21, an image storage unit 22, a selection unit 23, a defect candidate detection unit 24, and a determination unit 25.

  The image data unit generation unit 21 extracts image data of a portion imaged by each die from a band-shaped swath image sequentially output from the imaging device 13, and adds information on the acquisition position to the extracted image data. As a result, an image data unit is generated. FIG. 6 is a diagram illustrating an example of a data structure of an image data unit generated by the image data unit generation unit.

  The image data unit includes die position information designating a die from which the image data has been imaged, position information in the die from which the image data has been imaged, and a direction in the main scanning direction when the image data has been imaged. And image data. This example is a data structure example when a plurality of dies are two-dimensionally arranged on the reticle 200, and the row number and column number of the die on the reticle 200 are used as an example of die position information.

  The image data unit generation unit 21 is a reticle including position information and speed information of the stage 11 supplied from the stage control unit 40, dimension information of the reticle 200, and dimension information and coordinate information of a die formed on the reticle 200. Based on the pattern coordinate information 50, die position information, in-die position information, and a scanning direction are determined.

  Note that image data units that store image data corresponding to each other between different dies have different die position information and the same in-die position information. The image data unit generation unit 21 stores the created image data unit in the image storage unit 22.

  Refer to FIG. 5 again. The selection unit 23 selects a combination of image data units whose in-die position information and scanning direction match from among a plurality of image data units stored in the image storage unit 22, and provides the combination to the defect candidate detection unit 24. To do. The defect candidate detection unit 24 detects one of the combinations of the provided image data units as an inspection image, the other image data as a reference image, compares these image data, and detects a different portion as a defect candidate. .

  FIG. 7 is a block diagram illustrating a configuration example of the defect candidate detection unit 24 illustrated in FIG. The defect candidate detection unit 24 includes a difference detection unit 30, a defect threshold value determination unit 31, and a detection unit 32.

  The difference detection unit 30 calculates a gray level difference signal between pixels at corresponding positions between the inspection image and the reference image.

  The detection threshold value determination unit 31 determines a detection threshold value by a predetermined calculation process based on the statistics of the gray level difference detected by the difference detection unit 30 and outputs the detection threshold value to the detection unit 32.

The detection unit 32 compares the gray level difference input from the difference detection unit 30 with the detection threshold determined by the detection threshold determination unit 31, and detects a defect candidate included in the inspection image. When the gray level difference signal exceeds the detection threshold, the detection unit 32 determines that one of the inspection image and the reference image includes a defect at the pixel position where such a gray level difference is detected. To do.
The detecting unit 32 generates and outputs comparison result information including die identification information that specifies a combination of dies in which defect candidates are detected, in-die position information that specifies the positions of defect candidates in the die, and defect candidate information. FIG. 8 shows an example of the data format of the comparison result information. The defect candidate information may include information such as the size of the detected defect candidate, the gray level difference between the inspection image and the reference image, and the gray level value of these images.

  Refer to FIG. 5 again. The determination unit 25 receives the comparison result information output from the detection unit 32 of the defect candidate detection unit 24. The determination unit 25 performs a double detection process for each of the image data units acquired by each die using the comparison result information created by comparison with the image data units acquired by at least two other dies, It is determined which die has the defect candidate indicated by the comparison result information. The details of the double detection process by the determination unit 25 will be described later.

  FIG. 9 is an overall flowchart of the defect inspection method according to the embodiment of the present invention, and FIG. 10 is a diagram showing an arrangement example of the reticle 200 used for explaining the method of FIG. The arrangement of the dies 211 to 214 in the reticle 200 shown in FIG. 10, the positional relationship between the regions 211-1 to 214-1 and the regions 211-2 to 214-2 in the dies 211 to 214, the main scanning direction, and the sub-scanning. The direction is the same as that of FIG.

  In the reticle 200, dies 211 to 214 in 2 rows and 2 columns are formed. The die in the first row and the first column is the die 213, the die in the second row and the first column is the die 214, and the first A die in the second row and the second column is referred to as a die 211, and a die in the second row and the second column is referred to as a die 212.

  In step S1, the stage control unit 40 shown in FIG. 5 drives the stage 11 according to the scanning path information 51 to move the imaging device 13 relative to the reticle 200, and reference numerals 401, 411, 402 shown in FIG. The surface of the reticle 200 is two-dimensionally scanned by the imaging device 13 along a scanning path in the order of 412, 403, 413, and 404.

  Paths 401 and 403 are main scanning paths in the forward direction indicated by arrows 400, and paths 402 and 404 are main scanning paths in the reverse direction indicated by arrows 400, and paths 411, 412, and 413 are provided. Is a sub-scanning path.

  Images of the region 211-1 and the region 212-1 are acquired by the main scanning 401. Further, the images of the region 211-2 and the region 212-2 are acquired by the main scanning 402. Images of the area 213-1 and the area 214-1 are acquired by the main scanning 403. Further, the images of the region 213-2 and the region 214-2 are acquired by the main scanning 404.

  When the swath images obtained by capturing the surface of the reticle 200 by the scanning by the imaging device 13 are sequentially generated, in step S2, the image data unit generation unit 21 generates the regions 211-1 to 214-1 and 211- of each die from the swath image. The image data of each part imaged at 22-14-2 is taken out, and information on the acquisition position is added to each piece of the taken out image data to generate an image data unit. The image data unit generation unit 21 sequentially stores the generated image data units in the image storage unit 22.

  The image storage unit 22 has a capacity capable of storing a swath image acquired by one main scanning in each of the forward direction and the reverse direction for a die of 2 rows and 2 columns. For example, in the case of the die arrangement shown in FIG. 10, image data units including the images of the regions 211-1 to 214-1 and 211-21 to 214-2 of the dies 211 to 214 for 2 rows and 2 columns are simultaneously held. can do.

  Steps S <b> 1 to S <b> 2 are repeated until the swath images acquired by the main scanning in the forward direction and the reverse direction for the die for 2 rows and 2 columns are stored in the image storage unit 22 (step S <b> 3). The image data units are sequentially stored. When the swath images acquired by the main scanning in the forward direction and the backward direction for the die of 2 rows and 2 columns are stored in the image storage unit 22, the process proceeds to step S4 (step S3). By completing the scanning in all of the areas 211-1 to 214-1 and 211-2 to 214-2, the image storage unit 22 stores a set of image data units indicated by arrows 52 in FIG. .

  In step S4, the selection unit 23 selects a combination of image data units having the same in-die position information and the scanning direction from the plurality of image data units stored in the image storage unit 22, and selects a defect candidate detection unit. 24.

  In the case of the image data unit shown in FIG. 11, double detection processing is performed by comparing the images of the areas corresponding to the area 211-1 between the die 211 and any other two dies. As a combination for obtaining a comparison result used in the above, a combination C1 of image data units related to the areas 211-1 and 212-1 and a combination C2 of image data units related to the areas 211-1 and 213-1 can be considered.

  Moreover, as a combination that obtains a comparison result used for the double detection process by comparing the images of the areas corresponding to the area 212-1 between the die 212 and any other two dies. A combination C4 of image data units relating to the areas 211-1 and 212-1 and a combination C3 of image data units relating to the areas 212-1 and 213-1 are conceivable. Here, the combination C1 and the combination C4 are the same.

  The selection unit 23 selects only C1 to C3 excluding C4 as a combination for comparing the images of the three regions 211-1, 212-1 and 213-1 and uses these combinations C1 to C3 to select images. The data units are combined and provided to the defect candidate detection unit 24.

  In step S5, the defect candidate detection unit 24 compares one image data of the provided combination of image data units as an inspection image and the other image data as a reference image, compares these image data, and determines a difference portion as a defect candidate. And the comparison result information is output.

  In step S <b> 6, when the determination unit 25 describes the three dies as A, B, and C, the images of the areas corresponding to each other are compared between the dies A and B and between the dies A and C, respectively. A double detection process is performed using the comparison result information obtained in this way, and it is determined which of these three dies the defect candidate detected as a result of the comparison exists. The determination unit 25 selects comparison result information to be used for double detection processing in each die according to die identification information and in-die position information included in each comparison result information.

  FIG. 12 is an explanatory diagram of a determination method for determining a die in which a defect exists by double detection processing. This example considers the case of dies 211, 212, and 213 as three dies. The determination unit 25 is obtained by comparing the images of the regions 211-1, 212-1 and 213-1 at the same in-die position between the die 211 and the die 212 and between the die 211 and the die 213. Based on the comparison result information, it is determined which of the regions 211-1 to 213-1 has a defect.

  In the table shown in FIG. 12, the value “◯” indicates a case where there is no difference as a result of comparison between the regions 211-1 and 212-1 or as a result of comparison between the regions 211-1 and 213-1. The value “x” indicates a case where there is a difference. Where there is no difference between the comparison between the regions 211-1 and 212-1 and the comparison between the regions 211-1 and 213-1, any of the regions 211-1, 212-1, and 213-1 It is determined that there is no defect.

  It is determined that there is a defect in the region 213-1 at a location that is different in the comparison between the regions 211-1 and 213-1 and that is not different in the comparison between the regions 211-1 and 212-1. That is, if a defect candidate is detected in the comparison between the regions 211-1 and 213-1 but this defect candidate is not detected in the comparison between the regions 211-1 and 212-1, this defect is detected. The candidate is determined to be due to a defect in the region 213-1.

  It is determined that there is a defect in the region 212-1 at a portion that is different in the comparison between the regions 211-1 and 212-1 and that is not different in the comparison between the regions 211-1 and 213-1. If the comparison between the regions 211-1 and 212-1 and the comparison between the regions 211-1 and 213-1 are different, it is determined that the region 211-1 is defective.

  As described above, when the selection unit 23 determines a combination for comparing the images of the three areas 211-1, 212-1 and 213-1, the combination C1 of the areas 211-1 and 212-1 and the area 211- The combination C2 of 1 and 213-1 and the combination C3 of the regions 212-1 and 213-1 are selected, and the image data units are combined with these combinations C1 to C3 and provided to the defect candidate detection unit 24. The defect candidate detection unit 24 performs image comparison once for each of these combinations C1 to C3.

  The determination unit 25 compares the image of the region 211-1 with the image of the region 212-1 and the image of the region 213-1 respectively, and the image of the region 212-1 as the image of the region 211-1. When the double detection processing performed by comparing with the images of the region 213-1 is performed, one image between the regions 211-1 and 212-1 in these two double detection processings. Use the result of the comparison in common. For this reason, the process in the defect candidate detection part 24 can be reduced.

  In step S7, it is determined whether or not the inspection has been completed for all inspection regions on the reticle 200. If all the inspection areas have been inspected, the process ends. If an area that has not yet been inspected remains, the imaging device 13 is positioned at the scanning start position of the next inspection area in step S8, and the process returns to step S1.

  FIG. 13 is a diagram for explaining a reticle drawing method. In the illustrated reticle 200, dies 211 to 213 in 3 rows and 2 columns are arranged. In the case of such a reticle, since three dies are arranged along the row direction, die-to-die comparison is performed in the same row by performing main scanning along the row direction.

  On the other hand, a pattern drawing apparatus for drawing a transfer pattern on a reticle always moves the irradiation position of an electron beam in a certain direction as indicated by arrows 411 to 413 in the drawing in order to maintain the accuracy of the drawing pattern. A transfer pattern is drawn on top. For this reason, pattern errors may repeatedly occur in the portion below the dies 211, 212, and 213 at the beginning of drawing due to a defect in the pattern drawing apparatus.

  Therefore, when the main scanning direction for scanning the reticle 200 and the drawing direction (411 to 413) when the pattern is drawn on the reticle are different, the defect inspection method of this embodiment causes the direction to be different from the main scanning direction. Perform die-to-die comparison between arranged dies and compare the dies 211, 212, and 213 at the beginning of drawing with dies in other columns, thereby improving the detection accuracy of defects due to defects in the pattern drawing apparatus. Can do.

  When the drawing direction (411 to 413) when the pattern is drawn on the reticle is unknown, as described with reference to FIGS. 10 to 12, dies (211, 212 and 213) belonging to different rows and different columns are used. ) By performing die-to-die comparison between them, it is possible to improve the detection accuracy of defects caused by the defects of the pattern drawing apparatus.

  FIG. 14 is an overall configuration diagram of the defect inspection apparatus according to the second embodiment of the present invention. The defect inspection apparatus 1 according to the present embodiment includes a plurality (n) of defect candidate detection units 24 and determination units 25 of the embodiment shown in FIG. That is, the defect inspection apparatus 1 shown in FIG. 14 includes n image processing post-stage units 27-1 to 27-n, and each image process post-stage unit 27-i (i = 1 to n) is a defect candidate detection unit. 24-i and determination unit 25-i.

  Further, the defect inspection apparatus 1 includes a data distribution unit 26 that distributes the image data unit selected by the selection unit 23 to each of the image processing post-stage units 27-1 to 27-n. In defect inspection by image comparison, processing after image comparison may take longer than time required for image collection. For this reason, by providing the plurality of image processing post-stage units 27-1 to 27-n, it is possible to reduce a decrease in throughput due to the image comparison process.

  In the present invention, the surface of a sample on which a repetitive pattern in which a plurality of basic patterns repeat is formed is scanned two-dimensionally with an imaging means, and images of corresponding portions of the plurality of basic patterns are compared with each other, The present invention can be applied to a defect inspection apparatus and a defect inspection method for detecting defects.

It is explanatory drawing of the die-to-die comparison of the pattern formed on the wafer. It is explanatory drawing of the die-to-die comparison of the pattern formed on the reticle. It is explanatory drawing of the scanning path | route in the conventional defect inspection method. It is a figure explaining which range of the captured image is stored simultaneously in the Example of this invention. 1 is an overall configuration diagram of a defect inspection apparatus according to a first embodiment of the present invention. It is a figure which shows the example of the data structure of the image data unit produced | generated by the image data unit production | generation part. It is a block diagram which shows the structural example of the defect candidate detection part shown in FIG. It is a figure which shows the example of the data format of comparison result information. 3 is an overall flowchart of a defect inspection method according to an embodiment of the present invention. It is a figure which shows the example of arrangement | positioning of the reticle used for description of the method of FIG. It is a figure which shows the example of the data of the image data unit stored in an image memory | storage part. It is explanatory drawing of the determination method which determines the die | dye which a defect exists by a double detection process. It is a figure explaining the drawing method of a reticle. It is a whole block diagram of the defect inspection apparatus by 2nd Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Defect inspection apparatus 200 Reticle 211-214 Die

Claims (12)

A defect inspection apparatus for a sample in which a repetitive pattern in which a plurality of basic patterns repeat in at least a column direction among a row direction and a column direction is formed on a surface,
Imaging means for imaging the surface of the sample;
Moving means for relatively moving the imaging means and the sample in at least a two-dimensional direction;
Scanning control means for controlling the relative movement of the imaging means and the sample by the moving means to scan the repetitive pattern with the imaging means with the row direction and the column direction as main scanning and sub scanning directions, respectively;
Image storage means for simultaneously storing images obtained by two or more main scans including main scans in opposite directions, each performed on three or more basic patterns;
Selecting means for respectively selecting images obtained by performing main scanning in the same direction on mutually corresponding portions of the plurality of basic patterns from the images stored in the image storage means;
Defect candidate detection means for performing image comparison between the images selected by the selection means and detecting a different portion as a defect candidate;
Determining means for determining which defect of the plurality of basic patterns is the defect candidate;
A defect inspection apparatus comprising:   The defect inspection apparatus according to claim 1, wherein the three or more basic patterns include a plurality of basic patterns arranged in different rows.   The scanning control means scans the first region in one of the two basic patterns arranged in different rows while proceeding in the first scanning direction along the main scanning direction, and then the two basic patterns. The surface of the sample while moving in the second scanning direction that is opposite to the first scanning direction until the second region in the other basic pattern of the pattern is scanned while moving in the first scanning direction. The defect inspection apparatus according to claim 2, wherein main scanning is performed on any one of the areas.   The image obtained by two or more main scans including the main scans in opposite directions is one of the images of the first area and the second area scanned while proceeding in the first scan direction, and the second scan direction. The defect inspection apparatus according to claim 3, comprising an image of any one of the regions scanned while proceeding to step 4. When the three basic patterns are first to third basic patterns,
The determination means is performed three times between the first basic pattern and the second basic pattern, between the first basic pattern and the third basic pattern, and between the second basic pattern and the third basic pattern. A double detection process for comparing the first basic pattern with the second and third basic patterns, and a double detection process for comparing the second basic pattern with the first and third basic patterns, using the image comparison results of The defect inspection apparatus according to any one of claims 1 to 4, wherein: The defect candidate detection means includes information on the detected defect candidate and identification information indicating which of the basic patterns on the sample is the basic pattern on which the image comparison in which the defect candidate is detected is performed. Generate comparison result information including
6. The defect inspection apparatus according to claim 5, wherein the determination unit selects the comparison result information used for the double detection process according to the identification information. A defect inspection method for a sample in which a repetitive pattern in which a plurality of basic patterns repeat in at least a column direction among a row direction and a column direction is formed on a surface,
With the row direction and the column direction as the main scanning direction and the sub-scanning direction, respectively, the imaging unit that images the surface of the sample, and the sample unit scans the repetitive pattern by moving the sample relative to each other. A scanning step of sequentially storing the obtained images in the storage means;
A plurality of images obtained by two or more main scans including main scans in opposite directions performed on each of three or more basic patterns, which are simultaneously stored in the storage unit, A selection step of selecting images obtained by main scanning the corresponding portions of the basic pattern in the same direction;
A defect candidate detection step of performing an image comparison between the images selected in the selection step and detecting a difference portion as a defect candidate;
A determination step of determining which defect of the plurality of basic patterns is the defect candidate;
Including defect inspection method.   The defect inspection method according to claim 7, wherein the three or more basic patterns include a plurality of basic patterns arranged in different rows.   In the scanning step, a first region in one of the two basic patterns arranged in different rows is scanned while proceeding in the first scanning direction along the main scanning direction, and then the two basic patterns are scanned. Until the second region in the other basic pattern is scanned while proceeding in the first scanning direction, while proceeding in the second scanning direction that is opposite to the first scanning direction. The defect inspection method according to claim 8, wherein main scanning is performed on any region.   The image obtained by two or more main scans including the main scans in opposite directions is one of the images of the first area and the second area scanned while proceeding in the first scan direction, and the second scan direction. The defect inspection method according to claim 9, comprising an image of any one of the regions scanned while proceeding to step 1. When the three basic patterns are first to third basic patterns,
In the determination step, three times are performed between the first basic pattern and the second basic pattern, between the first basic pattern and the third basic pattern, and between the second basic pattern and the third basic pattern, respectively. A double detection process for comparing the first basic pattern with the second and third basic patterns, and a double detection process for comparing the second basic pattern with the first and third basic patterns, using the image comparison results of The defect inspection method according to claim 7, wherein the defect inspection method is performed. The defect candidate detection step includes information on the detected defect candidate and identification information indicating which of the basic patterns on the sample is the basic pattern on which the image comparison in which the defect candidate is detected is performed. Generate comparison result information including
12. The defect inspection method according to claim 11, wherein the determining step selects the comparison result information used for the double detection process according to the identification information.

Images