JP2016162513A - Charged particle beam device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a charged particle beam device which uses a secondary electron image to align images or divide regions according to materials and the like.SOLUTION: To attain the objective described above, a charged particle beam device comprises: a first detector which detects secondary electrons emitted in a first direction from a sample; and a second detector which detects secondary electrons emitted in a second direction that is different from the first direction. A signal acquired by the first detector and a signal acquired by the second detector are subjected to a difference arithmetic on a pixel basis. In accordance with a feature amount obtained based on the difference arithmetic, an image is divided into regions or identification information is individually assigned to the regions.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a charged particle beam apparatus including an image processing apparatus, and more particularly, to a charged particle beam apparatus including an image processing apparatus that performs image processing based on detection signals obtained under a plurality of charged particle acquisition conditions. About.

  A charged particle beam apparatus used in a semiconductor inspection apparatus or the like detects a charged particle emitted from a sample or a charged particle generated when a charged particle emitted from a sample collides with another member, and uses the detection signal. Device for forming an image. In Patent Document 1, a method for specifying a desired position in an image obtained by a scanning electron microscope (SEM) used for semiconductor measurement or inspection is generated from design data of a semiconductor device. A pattern matching method using a template is described. Patent Document 2 describes a pattern inspection method for performing pattern matching using region information obtained based on detection of backscattered electrons (BSE).

Patent No. 4199939 (corresponding US Pat. No. 7,026,615) Patent No. 5313939 (corresponding US Pat. No. 8,653,456)

  Patent Document 1 describes that a template having a shape close to an actual pattern is created by deforming corner edges of a pattern obtained based on design data by smoothing processing (Smoothing). However, by simply comparing the deformed design data with the image to be inspected, it is possible to perform alignment at a position deviating from the correct position due to the deformation of the circuit pattern reflected in the image to be inspected or the density of the circuit pattern. There is sex. Further, as in Patent Document 2, when matching is performed after performing region division using information of backscattered electrons, the backscattered electrons are affected by the reflection characteristics of the material, and thus are originally secondary. In some cases, an excessive lower layer pattern that has not been observed with electrons is reflected in the image and fails to be aligned. In addition, alignment is a pretreatment for inspection, and it is expensive to acquire a backscattered electron image that is easy to align.

  In the following, a charged particle beam apparatus is proposed which aims to realize alignment between images and area division according to material using a secondary electron image.

  As one aspect for achieving the above object, a detector for detecting secondary electrons generated based on irradiation of a charged particle beam emitted from a charged particle source, and a detector generated based on the output of the detector will be described below. A charged particle beam apparatus including an image processing apparatus for processing a captured image, wherein secondary electrons emitted from the sample in the first direction or secondary electrons emitted in the first direction are transmitted to other members A first detector for detecting secondary electrons generated by the collision, and secondary electrons emitted in a second direction different from the first direction, or 2 emitted in the second direction. A second detector for detecting secondary electrons generated when the secondary electrons collide with another member; and the image processing apparatus includes a signal obtained by the first detector and the second detection. The signal obtained by the instrument Perform calculations, depending on the feature amount obtained based on the difference operation, the execution of the regions divided in the image, or to propose a charged particle beam apparatus for imparting identifying information respectively for each region.

  Further, as another aspect for achieving the above object, a detector for detecting secondary electrons generated based on irradiation of a charged particle beam emitted from a charged particle source, and generation based on the output of the detector A charged particle beam apparatus including an image processing apparatus that processes a captured image, wherein the edge in a first direction of a pattern formed on a sample is relatively relative to other edges The secondary electrons acquired under the first condition of high brightness and the edges in the second direction different from the first direction are relatively bright with respect to the edges in the first direction. The present invention proposes a charged particle beam apparatus that executes region division of two or more regions divided by the edge based on secondary electrons acquired under the condition 2, or gives identification information to each region.

  According to the above configuration, it is possible to realize alignment between images and area division according to the material using the secondary electron image.

The flowchart which shows the process process which segments the area | region in an image using a some secondary electron image. The figure which shows an example of the secondary electron image of a line and space pattern. Sectional drawing of a line and space pattern. The figure which shows an example of the secondary electron image of the line pattern in which a different pattern exists in a lower layer. The figure which shows an example of a scanning electron microscope. The figure explaining the discharge | release direction of a secondary electron. The figure which shows the example of an image produced | generated based on the output of an upper stage detector and a lower stage detector. The flowchart which shows the process of performing the area | region division in an image using the some image acquired on multiple conditions. The figure which shows the example which calculated | required the brightness | luminance difference for every corresponding pixel of an Upper image (image based on the output of an upper side detector), and a Lower image (image based on the output of a lower side detector), and imaged it. Sectional drawing of a multilayer circuit pattern. The figure which shows the track | orbit of a secondary electron when a secondary electron is deflected by the aligner which deflects a secondary electron to the detector side. The figure which shows the image produced | generated based on the output of an upper stage detector, and the image produced | generated based on the output of a lower stage detector. The flowchart which shows the area | region division process of an image. The figure which shows the example of an image used in order to extract a pattern part from an image. The figure which shows the some image which made the edge of a specific direction high-intensity, and the synthesized image of the said image. The figure which shows an example of a scanning electron microscope. The figure which shows the example of the image acquired on the conditions from which the left edge becomes high brightness, and the image acquired on the conditions from which the right edge becomes high brightness. The flowchart which shows the area | region division process of an image. The figure which shows an example of a scanning electron microscope. The figure which shows the example of an image acquired by the detector arrange | positioned in multiple directions. The figure which shows the positional relationship of the discharge | release direction of an electron, and a detector. The figure which shows the image acquired by the detector arrange | positioned in multiple directions, and each detector. The figure which shows an example of the arithmetic unit which performs an image process. The flowchart which shows the process of selecting appropriate aligner conditions. The figure which shows the high-intensity edge extracted from the image.

  Embodiments described below relate to an image processing apparatus that processes an image signal obtained by an appearance inspection apparatus, a pattern measurement apparatus that measures a pattern dimension, a defect inspection apparatus that inspects a defect on a sample, and the like. The present invention relates to an image processing apparatus that performs pattern matching using a template generated based on device design data, and a charged particle beam apparatus equipped with the image processing apparatus.

  In the embodiment described below, the field of view (FOV) is divided into regions mainly using information of two or more secondary electron images having different shooting conditions, and the position is determined based on the region division. The image processing apparatus and charged particle beam apparatus which perform a matching are illustrated.

  According to the embodiment described below, it is possible to perform a robust alignment that is not affected by a layout in which a high-density circuit pattern exists, deformation, and occurrence of a defect.

  The embodiment described below can be realized by image processing using a general-purpose computer. An image processing apparatus that realizes the image processing includes basic computer components such as an input / output interface (display, keyboard, mouse, LAN port, USB port, etc.), CPU (arithmetic unit), and memory (storage unit). . The following embodiment is also an explanation of a program that causes a computer to execute pattern matching processing and the like.

  The outline of the present embodiment will be described with reference to FIG. FIG. 1 is a flowchart showing processing steps for dividing a region in an image using a plurality of secondary electron (SE) images. The flowchart illustrated in FIG. 1 includes a step 101 for reading two or more SE images with different imaging conditions, a step 102 for extracting an edge region from each SE image, and a flat region by combining the edge regions of each image. And step 104 of acquiring layer information of the flat region from the flat region, the edge region, and the photographing conditions.

  The imaging conditions in step 101 include the three-dimensional position and positional relationship of the secondary electron detector of the inspection apparatus, the angle and direction of electrons acquired by the detector, the set value of the energy amount of electrons acquired, and an aligner described later. This is a value obtained by mathematically calculating the degree of deflection. The edge region in step 102 is the white band 201 included in the SE image as shown in FIG. The edge portion of the pattern tends to be relatively brighter than the flat portion due to the edge effect, and the edge portion has higher brightness than the pattern portion 202 and the space portion 203 in the example of FIG. . The flat area segmentation performed in step 103 is a process of segmenting the edge portion, the pattern portion 202, and the space portion 203 in the image using the edge region information extracted in step 102.

  The layer information in step 104 is identification information indicating which layer each region on the two-dimensional image belongs to. Specifically, information indicating whether each region illustrated in FIG. 2 is a region corresponding to the convex portion (upper portion) 301 or the concave portion (lower portion) 302 in the cross-sectional view illustrated in FIG. It is. In the SE image, as illustrated in FIG. 4, the upper layer pattern portion 401 and a lower layer pattern 402 that is located in the lower layer of the upper layer pattern portion 401 and part of which is hidden in the upper layer pattern portion 401 are 2 A circuit pattern of more than one layer may be reflected. The space part 403 is an area without a pattern.

In such a case, the layer information acquired in step 104 is acquired for the upper layer pattern portion 401, the lower layer pattern portion 402, and the space portion 403. Further, it may be simply divided into a pattern portion and a space portion. The layer information is not limited to the two-layer or three-layer information described above. Further, after step 104, alignment using the layer information is performed.
Details of each of the above processes will be described below. FIG. 5 is a diagram illustrating a charged particle beam apparatus that performs image region division processing and alignment processing. In this embodiment, a scanning electron microscope will be described as an aspect of the charged particle beam apparatus. As an apparatus for acquiring an image, an ion beam irradiation apparatus that detects charged particles obtained by irradiating a sample with an ion beam is used. You may make it adopt. The scanning electron microscope is a device that irradiates a sample with an electron beam (electron beam 503) extracted from an electron source 501 by an extraction electrode 502. An electron beam 503 accelerated by an accelerating electrode (not shown) is focused by a condenser lens 504 which is a form of a focusing lens, and then one-dimensionally on a sample (semiconductor wafer) 509 by a scanning deflector (not shown) or Scanned two-dimensionally.

  The electron beam 503 is decelerated by a negative voltage applied to an electrode built in the sample stage 508 and is focused by the lens action of the objective lens 506 and irradiated onto the sample 509. When the sample 509 is irradiated with the electron beam 503, secondary electrons and electrons 510 such as backscattered electrons are emitted from the irradiated portion. The emitted electrons 510 are accelerated in the direction of the electron source by the acceleration action based on the negative voltage applied to the sample, collide with the conversion electrode 512, and generate secondary electrons 511. The secondary electrons 511 emitted from the conversion electrode 512 are captured by the detector 513, and the output I of the detector 513 changes depending on the amount of captured secondary electrons. Depending on the output I, the brightness of a display device (not shown) changes. For example, when a two-dimensional image is formed, an image of the scanning region is formed by synchronizing the deflection signal to the scanning deflector 505 and the output I of the detector 513. The scanning electron microscope illustrated in FIG. 5 includes a deflector (not shown) that moves the scanning region of the electron beam. The control device 514 controls each component of the scanning electron microscope, and forms a pattern on the sample based on the function of forming an image based on detected electrons and the intensity distribution of detected electrons called a line profile. It has a function to measure the pattern width.

  In this embodiment, a method is adopted in which secondary electrons emitted from the sample collide with the conversion electrode (other member) and the secondary electrons generated from the collision are detected. However, the present invention is not limited to this. For example, it is possible to employ a detection system in which the detection surface of the detector is arranged on the trajectory of the secondary electrons in order to directly detect the secondary electrons emitted from the sample.

  As illustrated in FIG. 23, the control device 514 includes basic computer components such as a CPU 2101, an image memory 2102, an LSI 2103, an output device 2104, and a storage device 2105. In order to receive design data and the like, it is connected to the design system 2108 and its control device 2107 via a network such as a LAN. The control device 514 has a built-in arithmetic device (image processing device) for processing the obtained image, and executes image region division processing and alignment processing described later.

  Next, the behavior of secondary electrons will be described. When electrons are irradiated from the electron source 501 toward the semiconductor wafer 509, secondary electrons 601 are generated as shown in FIG. The secondary electrons emitted from the side wall portion 602 of the pattern include an electron 603 emitted on the sample surface and an electron 604 that collides with a structure in the sample. The electrons 603 can be detected by appropriately guiding them to a detector or a conversion electrode, but the electrons 604 cannot be detected. The electron 604 is an electron having a structure such as a pattern on its emission trajectory.

  Since there is no circuit pattern to be shielded around the upper part 605 of the circuit pattern, even an electron having a relatively large relative angle between the electron emission direction and the optical axis direction of the electron beam can be detected. However, in the space portion 606 of the circuit pattern, electrons are blocked by the circuit pattern, so that only electrons having a relatively small relative angle to the beam optical axis can be detected. Further, the electrons emitted from the side wall 607 are blocked by the side wall 607 while the electrons directed to the left side are blocked, but since there are no obstacles on the right side, they are emitted to the optical axis direction of the beam and to the right side of the side wall 607. The detected electrons can be detected.

  FIG. 7 illustrates an image generated based on the detection signals of the upper detector 513a and the lower detector 513b of the scanning electron microscope illustrated in FIG. FIG. 7 illustrates a line and space (Line & Space: L & S) pattern and a hole pattern as circuit pattern shapes. The L & S pattern is a pattern in which wiring portions and space portions of a circuit pattern are alternately arranged. The hole pattern is an opening connecting the lower layer and the upper surface of the upper layer.

  The material of the region including the L & S pattern is titanium nitride TiN in the upper part (pattern part) and silicon Si in the lower part (space part). The material of the region including the hole pattern is silicon nitride SiN at the top and silicon Si at the bottom. The lower part of FIG. 7 is an example of an image acquired by the Upper detector 513a, and the upper part is an example of acquisition by the Lower detector 513b. As described with reference to FIG. 6, the electrons detected by the lower detector 513b include many electrons having a relatively large relative angle between the beam optical axis direction and the electron emission direction. Is blocked by a structure such as a pattern, but this is not the case in the pattern portion. For this reason, the space portion 701a tends to be darker than the pattern upper portion 702 as illustrated in the upper part of FIG. On the other hand, the electrons detected by the upper detector 513a include many electrons having a relatively small relative angle between the beam optical axis direction and the electron emission direction, so that the amount of electrons blocked by the structure on the sample is detected by the lower detection. The brightness of the space portion 703a is relatively bright. In addition, the space portion of the L & S pattern indicates that the brightness of the image generated based on the electrons detected by the Upper detector 513a is greater than the brightness of the image generated based on the electrons detected by the Lower detector 513b. Is expensive.

  However, this relationship may not be established. Even if the electron beam irradiation conditions are the same, the amount of secondary electron emission differs depending on the material.For example, when a material with a large amount of secondary electron emission exists in the lower layer, it is detected by the detector. Even in an image based on electrons, the lower layer may be brighter than the upper layer.

  In this embodiment, an image processing apparatus and a charged particle beam apparatus that can perform region division according to the type of pattern based on luminance information of an image regardless of changes in the material of the sample and the like will be described. In this embodiment, in particular, a processing technique for dividing the uneven area of the circuit pattern based on an image obtained by a scanning electron microscope including a plurality of detectors will be described.

  FIG. 8 is a flowchart showing a process of dividing a region in an image using a plurality of images obtained under a plurality of conditions. Step 101 for reading an image and step 102 for extracting an edge region are the same as in FIG. Here, the edge extraction in step 102 for extracting the edge region uses an edge extraction filter such as a Sobel filter or a Laplacian filter. Further, since the edge portion appears in the image as a white band having a higher luminance than the flat portion, it can be extracted using binarization or the like. Further, more advanced edge region extraction may be performed by combining a filter and binarization.

  In step 801 for synthesizing the edge region, the edge region obtained from the upper image (the image generated based on the detection signal detected by the upper detector) and the lower image (the detection signal detected by the lower detector) are combined. A combined edge region is created by combining edge regions obtained from an image generated based on the image). For example, the composite edge region may be given as a binary image in which the edge portion is 1 and the non-edge portion is 0. The form of the composite edge region is not limited to this. Further, the combination method may be a process of obtaining a sum and binarizing. In the case of a multilayer image, the edge region obtained from the Upper image may not exist in the edge region obtained from the Lower image. In such a case, the area may be ignored and divided into the upper layer and the other layers, or the area may be divided into three or more types in consideration of the lower layer. A method for considering will be described later. The method for synthesizing the edge region is not limited to the above-described method using the sum.

  In the mask step 802 of the edge portion, the edge region of the Upper image and the edge region of the Lower image are masked using the combined edge region. For example, processing such as setting the pixel value to a negative value may be performed. The mask method is not limited to this. Further, when performing the processing described later, it may be determined each time whether or not it is a composite edge region, and whether or not the processing for each image is to be performed may be determined. In the luminance difference calculation step 803, the area is divided by using the edge area masked portion as a divided area. For example, if adjacent pixels are not negative, processing such as attaching the same label is performed. For both the upper image and the lower image, the average pixel value is calculated within the same label, and the difference is calculated. Alternatively, the average of the differences may be obtained after taking a difference between each pixel of the upper image and the lower image (after performing the difference calculation).

  FIG. 9 is a diagram illustrating an example in which a feature amount such as a luminance difference is obtained for each corresponding pixel of the Upper image and the Lower image and is imaged. FIG. 9 is a difference image generated based on the L & S pattern image illustrated in FIG. A gray area 901 surrounded by a broken line is an area having a large luminance difference between the upper image and the lower image, and a black area 902 is an area having a small luminance difference. Although it cannot be visually recognized as an image, the vicinity of the boundary between the gray region 901 and the black region 902 is a masked insensitive region. In a region classification step 804, each region is classified using the luminance difference obtained in the luminance difference calculation 803 as an index value. As the method, for example, when each luminance difference is plotted on one dimension, two classes may be classified between points having the longest distance, or an algorithm such as a K-average method may be used. The classification method is not limited to this. In the layer information conversion step 805, the classified class is assigned to each area. For example, the image may be possessed in an image format in which the edge portion is a negative value as a separate image and the class number arbitrarily determined in each flat region is a luminance value. The description method of the layer information is not limited to this.

  Next, a case where the circuit pattern is a multilayer will be described. FIG. 10 illustrates a cross-sectional view of a multilayer circuit pattern. Some secondary electrons 1001 emitted based on the irradiation of the lower layer with the electron beam 503 are blocked by the upper side wall 1002. Compared with the case where there is no upper side wall 1002, the amount of detected electrons is reduced. However, this amount is larger than the secondary electrons 1003 emitted based on the irradiation of the electron beam to the space portion. Therefore, as in the multilayer image shown in FIG. 4, in the lower image, the luminance of the flat area of the circuit pattern is basically upper layer 401> lower layer 402> space part 403. However, the above formula may not hold depending on the material.

  However, the characteristics of the upper image and the lower image (relative relationship between the upper image and the lower image) do not change regardless of the presence or absence of the lower layer. The upper layer area is established. Therefore, in the area classification step 804, division for the included layers may be performed. If it is unclear how many layers of wiring will appear in the image, the number of divisions in the K-Means method may be automatically determined using index values such as the Akaike information criterion and the Bayes information criterion.

  If the number of layers to be captured as input parameters and design data used for alignment are given, the number of divisions may be determined based on the information. The method for determining the number of divisions is not limited to this. If the number of divisions can be determined, then layer information can be formed by the same method as in the layer information step 805. When performing region division, the luminance difference between the upper detector and the lower detector corresponding to the type of sample is stored in a database in advance, and identification information is added to each region by referring to the information. good.

  Next, a case where secondary electrons at an arbitrary angle are observed using the aligner 505 will be described. FIG. 11 is a diagram showing the behavior of electrons when secondary electrons are deflected by the aligner 505. Secondary electrons 1101 emitted from the semiconductor wafer 509 (secondary electrons emitted in the first direction) are deflected by the aligner 505 and pass through the gap between the conversion electrodes 512 (passage opening of the electron beam 503). The secondary electrons 1101 that have passed through the gap between the conversion electrodes 512 are further deflected by the aligner 1103 and collide with the conversion electrode 1104. When secondary electrons 1101 collide with the conversion electrode 1104, secondary electrons (tertiary electrons) are generated therefrom, and the electrons are captured by the upper detector. The condition of the aligner 505 in this embodiment is adjusted so that the secondary electrons 1101 pass through the gap between the conversion electrodes 512.

  Electrons 1102 emitted at angles other than the specified angle (secondary electrons emitted in the second direction) are deflected by the aligner 505, collide with the lower conversion electrode 512, and are detected by the lower detector. Strictly speaking, electrons emitted from the conversion electrode 512 are captured by the Lower detector.

  FIG. 12 is a diagram (left diagram) showing an example of an image acquired under the condition that electrons 1101 emitted mainly in the left direction are guided by the aligner 505 to the upper detector. In the image based on the electrons detected by the upper detector, the luminance of the left edge (white band 1201) is higher than that of the right edge (shadow region 1202) because of the deflection condition of the aligner 505. A region where there is no shielding member (in this case, a pattern) for secondary electrons, like the space portion 1203, has a higher luminance than the shadow region 1202.

  In the Lower detection image (the right diagram in FIG. 12), the edge portion other than the left side surface of the pattern is relatively brighter (white band 1204) than the other portions. On the other hand, the deflection condition is such that most of the secondary electrons emitted from the left side surface portion 1205 are guided to the upper detector, and therefore the left side surface portion 1205 has a relatively low luminance with respect to the white band 1204. However, because of the edge effect, a larger amount of secondary electrons are emitted as compared with the flat portion, so that the luminance is higher than that of the flat portion.

  FIG. 13 is a flowchart showing the process of dividing the image as shown in FIG. The steps up to the edge extraction step (step 102) are the same as those in FIG. In this example, since the electrons emitted in the left direction are observed by the upper detector, the region on the right side of the white band of the upper detector image is a pattern portion. In the provisional region determination step (step 1301), a provisional determination region 1401 is determined as illustrated in FIG. A classification process using the white band edge 1402 obtained from the Lower detection image is executed for the provisional determination area 1401 (step 1302). The white band edge 1402 is selected based on, for example, a process of extracting a high-luminance region. When the white band of the Upper detector image exists inside the provisional determination area, the area division is also performed there. It is divided into a closed region 1403 surrounded by edges and an outer region 1404 of the edges. Of the divided areas, the area in the same direction as the provisional determination area is created, in this example, the edge outer area 1404 is deleted. A classification process (step 1304) is performed with the temporary area that has not been deleted as a pattern part and an area that is not included in the temporary area as a space part, and a layer information process (step 805) is performed.

  Note that an appropriate aligner condition for extracting an edge in each direction may be set through a process illustrated in FIG. FIG. 24 is a flowchart illustrating a process of determining an appropriate aligner condition based on an image obtained when the aligner condition is changed. First, the aligner is set as an initial condition, and the output images of the upper and lower detectors (Upper detector, Lower detector) obtained at that time are evaluated (steps 2401 and 4022). If any of the upper and lower detector images satisfies the predetermined condition, the aligner condition is determined and the information is registered as an imaging recipe for the scanning electron microscope (steps 2403 and 2404). If the output image of the vertical detector does not satisfy the predetermined condition, the change of the aligner condition and the image evaluation are repeated based on the determination that the aligner condition is inappropriate. By performing such processing, appropriate aligner conditions can be found. If the aligner condition is determined, the same aligner condition is considered to be valid if the pattern is the same. Therefore, if the aligner condition is registered as an imaging recipe, it is possible to set conditions for measuring the same pattern.

  The image evaluation in step 2402 can be performed, for example, by performing step 101 to step 1303 in FIG. 13 and then determining whether or not an appropriate closed figure is formed by the arithmetic unit. In addition, when the proper aligner conditions are satisfied, the edges of the pattern included in the image generated from the output of the upper and lower detectors are in a bright state, and therefore a line having a predetermined value or a luminance (or contrast) equal to or higher than the predetermined value. The longest minute (that is, a state in which all edges have a predetermined luminance) may be set as an appropriate aligner condition. In this case, since it is possible that noise is included in the image, the aligner condition is selected so that the length of the line segment is a predetermined length (peripheral length of an ideal closed figure ± predetermined value). May be. In addition, the aligner condition may be determined based on a determination as to whether or not a combined figure of the figures included in the output image of each detector has a predetermined shape.

  Note that if the aligner condition is known in advance so that the upper detector can acquire an image with a high luminance at the left edge, the pattern is positioned on the left side of the edge included in the image acquired under the aligner condition. It is possible to limit the range of the tentatively determined area, search for other edges in the area, and realize the efficiency of the pattern portion specifying process.

  In addition, the range of the provisional determination area may be specified according to the pattern shape obtained in the initial aligner state. For example, in the case of a cross pattern as illustrated in FIG. 12, when a high-brightness edge 2501 having a shape as illustrated in FIG. 25A is extracted, an upper detector has more electrons on the upper edge than other edges. It can be seen that the aligner condition leads to In addition, when the high-luminance edge 2502 having the shape illustrated in FIG. 25B is extracted, it can be seen that the condition is such that more electrons on the right edge are guided to the upper detector than other edges. . In this way, the direction of the provisional determination region (lower side in the case of FIG. 25A, left side in the case of FIG. 25B) is determined according to the high-luminance edge shape acquired by one detector. Anyway.

  Since the shape of the pattern can be known by referring to the design data, the pattern formation direction may be specified by associating the layout data obtained based on the design data with the high luminance edge.

  In addition, when the aligner 505 is used, it is possible to create an image 1501 in which an edge in a specific direction is selectively bright as shown in FIG. If the white band of the composite image 1502 is used, it is not necessary to use the white band of the Upper detector image in the process of FIG. As the synthesized image, for example, a method of synthesizing a portion where the luminance of each of the high-luminance images 1501a to 1501d in each direction becomes maximum is conceivable. The method for creating a composite image and the combination of images are not limited to this. Moreover, although the example of 4 directions was shown in FIG. 15, this method is not limited to it.

  FIG. 16 illustrates another configuration of a scanning electron microscope that acquires an image to be measured or inspected. The difference from FIG. 5 is that two Lower detectors 513b in FIG. 5 are installed. These two detectors are installed in opposite directions orthogonal to the beam optical axis.

  As shown in the figure, the left is the Left detector 1613a, and the right is the Right detector 1613b. The two are collectively called an LR detector. In FIG. 16, the upper detector is omitted. The upper detector is not indispensable when performing region division in the inspection apparatus configuration of FIG.

  FIG. 17 illustrates images obtained from the Left detector 1613a and the Right detector 1613b. These sets of images are referred to as LR images in the following description. The LR image is similar to the image 1501b and the image 1501c illustrated in FIG. As illustrated in FIG. 21, the electrons emitted on the left side are detected by the Left detector 2101, and the electrons emitted on the right side are detected by the Right detector 2102.

  FIG. 18 shows a region dividing process using an image obtained by a scanning electron microscope having the apparatus configuration illustrated in FIG. In the image reading process (step 101), an LR image is read. In the edge region extraction process (step 1801), white bands are extracted from the LR image and synthesized to obtain an edge region. Alternatively, the edge region may be obtained from the white band of the composite image by combining the LR images. As an image synthesis method, for example, a method of taking a value of an image having a high luminance can be considered. For extraction of the white band, the same method as used in step 102 in FIG. 1 can be used.

  In the pattern calculation process (step 1802), the pattern area can be obtained by performing the provisional area setting process described with reference to FIG. 14 for either the L image or the R image. Moreover, you may carry out about both L image and R image. Thereby, it is possible to classify into a pattern part, an edge part, and other parts.

  Layer information processing (step 805) is performed using this classification information. Further, for each of the L image and the R image, the temporarily determined area described with reference to FIG. Each process is not limited to the method described above.

  FIG. 19 illustrates still another configuration of a scanning electron microscope that acquires an image to be measured or inspected. The difference from FIG. 16 is that a total of four Lower detectors 513b in FIG. 5 are installed in all directions. Hereinafter, each detector will be referred to as an X +, X-, Y +, Y-detector. The electrons detected by each detector are detected according to the position of the detector as described in the principle with reference to FIG. In the example of FIG. 19, two detectors are added in the Y-axis direction (perpendicular to the paper surface) that are arranged mirror-symmetrically about the beam optical axis.

  An image generated based on the electrons detected by each detector is illustrated in FIG. The image obtained from each detector has characteristics similar to an image in which the alignment condition of the aligner is changed and the edge in each direction is enhanced, and can be used for region classification based on edge extraction as described above. it can.

  Next, a method for dividing an image of a multilayer pattern in which a plurality of patterns are stacked will be described. FIG. 22 is an example of a multilayer pattern image generated based on electrons detected by detectors arranged in different directions. In the upper detector image, the lower layer pattern 2201 is clearly visible, but the lower layer pattern may not be displayed in the image based on electrons obtained by the X + and X− detectors corresponding to the LR detector of FIG. This is because the distance between adjacent patterns is short and the pattern density is high, so electrons with a large relative angle to the beam optical axis emitted from the lower layer pattern are blocked by the upper layer pattern and are not detected by the X + and X− detectors. It is. Such a situation can also occur in the LR detector. However, since the Y + and Y− detectors have no shielding object, electrons emitted from the lower layer 2102 can be detected. For this reason, first, using the X + and X− detector images, the region division processing using the processing method illustrated in FIG. 18 is performed on the method described so far, for example, the image acquired by the scanning electron microscope of FIG. As a result, the upper layer pattern 2103 and the other region 2104 are divided into regions. Further, an image in which the upper layer pattern 2103 and its edge region 2105 are masked with respect to the Y +, Y− detector image is created, and the lower layer pattern 2102 and the space portion are input in the same manner as the X +, X− detector image. 2106 is divided into regions. This is layer information. As described above, it is possible to divide a multi-layered image by combining the methods described above. The processing for the four-direction multilayer pattern image is not limited to this.

  Even in the case of a single detector, the function of adjusting the amount of electron detection such as an energy filter and the direction of electrons detected by the aligner are controlled, and multiple parameters are created to create an upper image. It is possible to acquire an image having the same characteristics as the image obtained by the Lower detector or the LR detector. Even for such an image, it is possible to divide the region using the processing described so far.

  By performing a matching process that matches the areas divided as described above, it is possible to perform alignment that is not easily affected by pattern deformation or the like. Specifically, identification is performed when a search is performed in an image that has been subjected to the above-described region segmentation processing using a template in which identification information of each pattern formed based on design data or the like is given in advance. The matching position is a relative position where the information is matched and the identification information matches. In particular, when there are a plurality of patterns and the arrangement state of the plurality of patterns is unique, proper alignment can be performed even when the pattern shape is greatly deformed.

501 Electron source 502 Extraction electrode 503 Electron beam 504 Condenser lens 505 Deflector (aligner)
506 Objective lens 507 Vacuum sample chamber 508 Sample stage 509 Sample 510 Electron 511 Secondary electron 513 Detector 514 Control device 515 Z sensor (light source)
516 Z sensor (light receiving part)

Claims (12)

Charged particles comprising a detector that detects secondary electrons generated based on irradiation of a charged particle beam emitted from a charged particle source, and an image processing device that processes an image generated based on the output of the detector In the wire device,
A first detector that detects secondary electrons emitted from the sample in a first direction or secondary electrons emitted when the secondary electrons emitted in the first direction collide with other members;
A second electron that detects secondary electrons emitted in a second direction different from the first direction, or secondary electrons generated when the secondary electrons emitted in the second direction collide with other members. With two detectors,
The image processing device performs a difference operation on a pixel basis for the signal obtained by the first detector and the signal obtained by the second detector, and a feature amount obtained based on the difference operation The charged particle beam apparatus according to claim 1, wherein the image is divided into regions or identification information is assigned to each region. In claim 1,
The charged particle beam apparatus according to claim 1, wherein the image processing device executes the region division according to a luminance difference of each pixel obtained by the difference calculation. In claim 1,
The image processing apparatus extracts an edge region included in an image based on outputs of the first detector and the second detector, and performs mask processing on the edge region. Wire device. In claim 3,
The image processing apparatus includes a first edge region extracted based on the output of the first detector and a combined edge region of the second edge region extracted based on the output of the second detector. A charged particle beam apparatus characterized by extracting In claim 3,
The charged particle beam apparatus, wherein the image processing device calculates a luminance difference for each region separated by the edge region. In claim 1,
The secondary electrons emitted from the edge in the first direction of the pattern formed on the sample or the secondary electrons generated by the secondary electrons emitted from the edge in the first direction are From the detector, secondary electrons emitted from the edge in the second direction other than the first direction, or emitted from the edge in the second direction so as to be detected more by the first detector. A deflector for deflecting secondary electrons emitted from the sample so that more secondary electrons generated by the secondary electrons are detected by the second detector than by the first detector. A charged particle beam apparatus characterized by that. Charged particles comprising a detector that detects secondary electrons generated based on irradiation of a charged particle beam emitted from a charged particle source, and an image processing device that processes an image generated based on the output of the detector In the wire device,
The image processing apparatus includes secondary electrons acquired under a first condition in which an edge in a first direction of a pattern formed on a sample has a relatively high brightness with respect to other edges, and the first An edge in a second direction different from the direction of is distinguished by the edge based on secondary electrons acquired under a second condition in which the edge is relatively brighter than the edge in the first direction. A charged particle beam apparatus characterized in that two or more areas are divided or identification information is assigned to each area. In claim 7,
The image processing apparatus performs a difference calculation on a pixel basis for the secondary electron signal acquired under the first condition and the secondary electron signal acquired under the second condition, and based on the difference calculation A charged particle beam apparatus characterized in that the image is divided into regions or identification information is given to each region according to the obtained feature amount. In claim 7,
The secondary electrons acquired under the first condition are generated when the secondary electrons emitted from the sample in the first direction or the secondary electrons emitted in the first direction collide with other members. Secondary electrons acquired under the second condition are emitted in a second direction different from the first direction, or emitted in the second direction. A charged particle beam device characterized in that the secondary electrons are secondary electrons generated by colliding with other members. In claim 9,
A secondary electron that is arranged between the charged particle source and the sample and emitted in the first direction or a secondary electron emitted in the first direction collides with another member. A second detector that detects the secondary electrons, the secondary detector disposed between the first detector and the sample and emitted in the second direction, or the second direction. A charged particle beam apparatus comprising: a second detector that detects secondary electrons generated when the secondary electrons emitted from the battery collide with another member. In claim 9,
The first direction and the second direction are opposite directions orthogonal to the charged particle beam optical axis, and secondary electrons emitted in the first direction, or emitted in the first direction. A first detector for detecting secondary electrons generated when the secondary electrons collided with other members, and secondary electrons emitted in the second direction, or emitted in the second direction. And a second detector for detecting secondary electrons generated when the secondary electrons collide with other members. In claim 11,
The image processing apparatus extracts a first edge of a pattern from a first image generated based on secondary electrons acquired by the first detector, and is acquired by the second detector. Extracting a second edge from the second image generated based on the secondary electrons, setting a pattern candidate area in a predetermined direction starting from the first edge, and A charged particle beam apparatus characterized by dividing using a second edge.