JP2017027651A - Charged particle beam device and program storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve such a problem that a template image cannot be created if there is no design data, when creating a template image for use in detection of an alignment pattern in pattern matching by global alignment for correcting the coordinate.SOLUTION: A charged particle beam device includes a charged particle beam optical system for irradiating a sample with a charged particle beam, a sample stage for holding the sample, a controller for controlling the charged particle beam optical system and sample stage, an arithmetic unit for creating the image of the sample from signals of secondary charges particles obtained from the sample by irradiation with the charged particle beam, and a storage device for storing the image of the sample. The arithmetic unit sets a pattern specified on a composite image, created by superposing multiple imaged captured previously, as a template pattern, and detects the position of a pattern matching the template pattern.SELECTED DRAWING: Figure 4

Description

  The present disclosure relates to a charged particle beam apparatus and can be applied to a charged particle beam apparatus having a global alignment function.

  In the design and manufacture of semiconductor devices, it is important to manage dust generation in manufacturing equipment such as exposure and etching equipment, and to evaluate the shape of circuit patterns formed on wafers. Scanning Electron Microscope (SEM) and scanning Inspection and measurement are performed by charged particle beam devices such as a scanning ion microscope (SIM) and a scanning transmission electron microscope (STEM). Examples of the SEM type charged particle beam apparatus include a scanning electron microscope for measuring length (Critical Dimension Scanning Electron Microscope: CD-SEM) and a scanning electron microscope for defect review (Defect Review Scanning Electron Microscope: DR-SEM). . In a charged particle beam apparatus typified by a CD-SEM, it is necessary to perform global alignment of a wafer (detection of wafer misalignment / rotation) in order to image and evaluate a fine circuit pattern. In global alignment, images of several positions on the wafer with known coordinates are imaged as alignment patterns. By matching this captured image with an image of an alignment pattern prepared in advance (hereinafter referred to as “template image for global alignment” or simply “template image”), the positional deviation or rotation of the wafer is detected. Template images for global alignment are disclosed in Japanese Patent Application Laid-Open No. 2011-135522 (Patent Document 1), Japanese Patent Application Laid-Open No. 2012-14475 (Patent Document 2), and International Publication No. 2012/070549 (Patent Document 3). In addition, it is created based on the design data of the circuit pattern formed on the wafer.

JP 2011-135032 A JP 2012-14475 A International Publication No. 2012/070549

  However, the design data is required to synthesize and create the template image for global alignment from the design data of the circuit pattern formed on the wafer described in Patent Document 1, but an apparatus for creating inspection information In order to use the design data above, a function for reading the design data into the apparatus and creating a template image, or a dedicated apparatus is required. Even if a template image for global alignment is created from design data, the circuit pattern formed on the sample through the semiconductor manufacturing process does not necessarily have a shape that matches the circuit pattern of the design data. The decline in pattern matching accuracy cannot be denied. It also causes a decrease in throughput. Patent Document 2 and Patent Document 3 also explain that global alignment pattern matching is performed using a template image created based on design data. However, if there is no base design data, a template image for global alignment is created. It is not possible.

Accordingly, an object of the present disclosure is to provide a charged particle beam apparatus that creates a template image for global alignment without using design data.
Other problems and novel features will become apparent from the description of the present disclosure and the accompanying drawings.

The outline of a representative one of the present disclosure will be briefly described as follows.
That is, the charged particle beam device includes a charged particle optical system that irradiates a sample with a charged particle beam, a sample stage that holds the sample, a control device that controls the charged particle optical system and the sample stage, and the charged particle An arithmetic device that generates an image of the sample from a signal of secondary charged particles obtained from the sample by irradiation of a line, and a storage device that stores the image of the sample. The arithmetic unit sets a pattern designated on a composite image created by superimposing a plurality of images captured in advance as a template pattern, and detects a position of a pattern that matches the template pattern.

  According to the charged particle beam apparatus, a template image for global alignment can be created without using design data.

It is a figure which shows schematic structure of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the test | inspection sequence of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the example of the optical microscope image for global alignment of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the position shift on a wafer when pattern matching fails by global alignment of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the picked-up image at the time of pattern matching failing with the optical microscope image for global alignment and the charged particle microscope image of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the preparation method of the template image for global alignment of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the global alignment image of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the global alignment image of the charged particle beam apparatus which concerns on an Example, (a) is the synthesized image created by synthesize | combining based on a global alignment image, (b) is the synthesized image which binarized (a). It is. It is a figure which shows the synthesized image of the charged particle beam apparatus which concerns on an Example, (a) shows the binarized image which deleted the unnecessary part from the synthesized image, (b) is from the binarized image of (a). The range which cuts out the template image for low magnification is shown. It is a figure which shows the template image for high magnification created from the template image for low magnification of the charged particle beam apparatus which concerns on an Example. It is a figure which shows the template image creation screen in the charged particle beam apparatus which concerns on an Example.

  The charged particle beam apparatus of the embodiment can be obtained by imaging when inspection in the same device or manufacturing process has already been performed in the creation of inspection information (recipe) for creating a template image for global alignment. A template image is created by synthesizing the alignment pattern images based on the alignment pattern images that are present.

More specifically, the charged particle beam device includes a charged particle optical system that irradiates a sample with a charged particle beam, a sample stage that holds the sample, a control device that controls the charged particle optical system and the sample stage, And an arithmetic unit that generates an image of the sample from a signal of secondary charged particles obtained from the sample by irradiation of the charged particle beam, and a storage device that stores the image of the sample. The arithmetic unit sets a pattern designated on a composite image created by superimposing a plurality of images captured in advance as a template pattern, and detects a position of a pattern that matches the template pattern.
The pre-captured image is an image of a sample of the same device or manufacturing process as the sample held on the sample stage.
The composite image is created by superimposing images with high similarity among a plurality of images captured in advance.
The image having a high similarity is determined by performing pattern matching of a plurality of images captured in advance and calculating the similarity for each combination of images.
A plurality of the composite images are generated, and the template pattern is set on the composite image selected from the plurality of the composite images.
The charged particle beam apparatus includes a display device that displays an image, a display device that displays an image obtained by binarizing the composite image, an input device that designates and inputs the template pattern on the binarized image, Is provided.
The charged particle beam apparatus selects and cuts out a necessary range from a template image that is an image including the template pattern, and creates a template image for low magnification.
The charged particle beam apparatus creates a high-magnification template image based on the low-magnification template image.
The charged particle beam apparatus includes an optical microscope, and the image captured in advance is captured by the optical microscope.
The charged particle beam apparatus corrects the positional deviation of the sample based on the position of the pattern that matches the template pattern.

  According to the charged particle beam apparatus of the present embodiment, a sample and design data are required in creating a template image for global alignment for performing pattern matching with an alignment pattern formed on a sample surface such as a semiconductor wafer. In addition, since it is created on the basis of an alignment pattern image that has already been captured, an improvement in pattern matching accuracy can be expected from a template image for global alignment created from design data.

  Furthermore, since no sample or design data is required on the device, there is no need for a function to read the design data on the device and create a template image for global alignment based on the design data, or a dedicated device. If the inspection of the manufacturing process has already been performed, a template image for global alignment can be created at any time, and the trouble of waiting for the sample to be transported to the sample stage in the apparatus at the time of creating the recipe is eliminated. Since no sample is required, it is not necessary to irradiate the sample surface with an unnecessary electron beam to capture a template image for global alignment, and contamination and damage to the sample surface can be suppressed. .

  Hereinafter, examples and modifications will be described with reference to the drawings. However, in the following description, the same components are denoted by the same reference numerals, and repeated description is omitted. In the following embodiments, a charged particle beam apparatus (DR-SEM) for defect review will be described as an example. However, the present invention is not limited to this, and a charged particle beam apparatus having a global alignment function such as a CD-SEM is used. Needless to say, this is applicable.

  FIG. 1 shows a schematic configuration of a charged particle beam apparatus according to an embodiment. FIG. 1 shows a state in which a sample 103 is carried into the charged particle beam apparatus 100. The charged particle beam device 100 includes a charged particle optical system 101 including a charged particle source 102, a sample stage 106 to which a sample 103 is fixed, a control device 141 for controlling them, input to the control device 141, display, and the like. And a defect review device 143 that performs automatic defect inspection and classification sequence. Each of the control device 141 and the host device 142 is also referred to as an arithmetic device. For example, electrons or ions are used as the charged particles.

  Based on the inspection information input via the operator operation monitor 132 or the external interface 133, the apparatus control unit 131 controls the transport control unit 135, the stage control unit 134, the charged particle optical system control unit 130, and the image processing control unit 127. Control is performed to detect defects by performing global alignment automatically to correct coordinate deviation and rotational deviation between the sample 103 and the sample stage 106. The device control unit 131 and the operator operation monitor 132 include a central processing unit (CPU), a storage device (memory) for storing programs and data, and the like. In addition, various programs and data are stored in the auxiliary storage device 138 constituted by a hard disk or the like. Note that a program stored in a non-temporary and tangible storage medium such as a DVD is input to the auxiliary storage device 138 via the external interface 133.

  When an inspection request is input, the apparatus control unit 131 executes an inspection sequence and issues a sample transfer instruction to the transfer control unit 135, and the sample 103 is transferred to the sample stage 105 in the sample chamber 105 via the sample exchange chamber 104. Carry up and hold. The apparatus control unit 131 reads recipe information such as the coordinate position of each defect candidate and charged particle optical system conditions from the external interface 133 and the auxiliary storage device 138, and the stage control unit 134 reads the coordinate position of the defect candidate from the apparatus control unit 131. The coordinate of defect candidates is adjusted so that the charged particle beam is accurately scanned by controlling the sample stage 106 in consideration of the correction amount by the global alignment based on the above. The charged particle optical system control unit 130 is based on charged particle optical system conditions such as acceleration voltage, retarding voltage, and imaging magnification from the apparatus control unit 131, a high voltage control unit 108, a retarding voltage control unit 109, a first condenser lens. The control unit 110, the second condenser lens control unit 111, the alignment control unit 112, the deflection current control unit 113, and the objective lens control unit 114 are controlled so that an optimum charged particle beam can be scanned. By controlling the extraction electrode 115 via the high voltage control unit 108, the primary charged particle beam 116 is extracted from the charged particle source 102, passes through the first condenser lens 117 and the second condenser lens 118, and is aligned by the alignment coil 119. Adjustment is performed, and the primary charged particle beam 116 that has passed through the deflection coil 137 and the objective lens 120 is converged by the action of each optical lens and scanned around the coordinate position of the defect candidate on the sample 103. When the primary charged particle beam 116 is scanned on the sample 103, the secondary charged particle beam 121 generated on the surface of the sample 103 is captured by the secondary charged particle detector 122 and amplified by the amplifier 107 as an electric signal.

  The image processing control unit 127 converts the electrical signal amplified by the amplifier 107 into luminance information, stores it as a captured image in the image memory 126, performs luminance correction processing by the image correction control unit 128, and then transfers it to the image display unit 129. Then, the captured secondary charged particle image is displayed.

  Further, when a captured image of a defect candidate is stored in the image memory 126, the defect detection control unit 123 automatically determines whether the defect is a defect based on the captured image, and the captured image that is truly detected as a defect is The data is automatically transferred to the automatic defect classification control unit 124, and the detected defects are classified and analyzed, and the result is displayed on the display monitor 125. In addition, an optical microscope 136 is disposed above the sample chamber 105 for use in capturing an optical microscope image. The image captured by the optical microscope 136 is sent to the image processing control unit 127 via the amplifier 107.

  The charged particle beam apparatus 100, which is a DR-SEM, uses a charged particle beam to optically detect defects (for example, pattern formation abnormalities) scattered at unspecified coordinate positions on a sample 103 (for example, a semiconductor wafer) in advance. Or, based on inspection results such as position information obtained by inspection with a charged particle beam type defect inspection apparatus, defects are automatically detected, and shapes are observed and classified. In order to detect a defect, first, a defect candidate to be inspected is sampled from all defect candidates scattered at unspecified coordinate positions on the sample 103, and the accurate information on the sample 103 is obtained based on the position information of the defect candidate. In order to recognize the coordinate position, an accurate position is detected by pattern matching an alignment pattern existing at a specific part on the sample 103 with a template image for global alignment. By correcting the amount, defect candidates are captured in the imaging field after the stage is moved. Then, a reference image (hereinafter referred to as a low-magnification reference image) and a defect candidate image (hereinafter referred to as a low-magnification defect image) for scanning a primary charged particle beam (hereinafter referred to as a charged particle beam) at a low magnification and comparing with a defect candidate. Image. Then, a difference image between the low-magnification reference image obtained by imaging and the low-magnification defect image is generated, the exact coordinate position of the difference portion on the difference image is specified, and the difference portion is truly a defect. Switching to a high magnification that makes it easy to determine whether or not there is, scans the charged particle beam to capture a defect image (hereinafter referred to as a high-magnification defect image), and automatically recognizes the defect (Automatic Through Defect Review (ADR), it is determined whether the defect candidate is truly a defect or a false report. At the same time, a process of automatically classifying defects (Automatic Defect Classification: ADC) performs various classifications according to shapes and the like based on images obtained by imaging.

Next, an inspection sequence in which global alignment is performed will be described with reference to FIGS. 2 and 3A to 2C.
FIG. 2 is a diagram illustrating an inspection sequence of the charged particle beam apparatus according to the embodiment. FIG. 3A is a diagram illustrating an example of an optical microscope image for global alignment of the charged particle beam apparatus according to the embodiment. FIG. 3B is a diagram illustrating a positional deviation on the wafer when pattern matching fails in global alignment of the charged particle beam apparatus according to the embodiment. FIG. 3C is a diagram illustrating a captured image when pattern matching fails between an optical microscope image for global alignment and a charged particle microscope image of the charged particle beam apparatus according to the embodiment.

  First, when an inspection request is input from the operator operation monitor 132 to the apparatus controller 131, the sample 103 is transported onto the sample stage 106 in step 201. During this time, the recipe information such as the coordinates and imaging conditions of each point to be globally aligned in step 208 is read, and the charged particle optical system conditions such as the acceleration voltage and the probe current are set in step 209 according to the read recipe information. Good. After completion of the conveyance of the sample 103, it is necessary to correct and match the coordinate deviation and rotational deviation between the stage 106 and the sample 103 which are caused by placing the sample 103 on the sample stage 106. Correction is performed by global alignment in steps 202 and 203. Here, in order to succeed in global alignment without retrying, it is necessary to capture the alignment pattern formed scattered on the surface of the sample 103 within the imaging field of view. First, in step 202, imaging is performed with the optical microscope 136. Coarse coordinate correction is performed by taking an optical microscope image with a low magnification. A pattern matching is performed using the alignment pattern included in the captured optical microscope image and the alignment pattern (template pattern) included in the optical microscope template image to calculate the amount of coordinate deviation and determine the correction value. The captured optical microscope image is stored in the auxiliary storage device 138 as global alignment image information. Next, in step 203, in order to capture the alignment pattern in the imaging field of view with a charged particle microscope having a high magnification, it is necessary to correct the amount of rotational deviation of the sample 103 itself, and at least 2 using an optical microscope 136 with a low magnification. It is also necessary to obtain a sample center in the stage XY coordinate system by performing global alignment above the point. After completing the global alignment at low magnification, switch to the setting to capture a high-magnification charged particle microscope image and perform highly accurate coordinate correction. A pattern matching is performed between the alignment pattern included in the captured charged particle microscope image and the alignment pattern (template pattern) included in the template image for charged particle microscope to calculate the amount of deviation of the coordinates, and the correction value is determined. Here, if there is a large error in the correction values for coordinate deviation and rotation deviation obtained at low magnification, the alignment pattern does not enter the field of view of the captured image when imaged at high magnification. This requires additional processing such as performing a process, and it takes time to detect the alignment pattern, resulting in a decrease in throughput.

  3A, 3B, and 3C as an example, when the display area is 135 × 135 mm, the actual display area (Field of View: FOV) of the low-magnification optical microscope image 303 is 675 μm, and the high-magnification charged particle microscope image 309 is used. Assuming that the FOV is 13.5 μm, the magnification ratio of the optical microscope image / charged particle microscope image is 50 times. When the L-shaped alignment patterns 305 and 306 formed on the wafer 302 on the sample stage 301 are corrected on the basis of the optical microscope images 303 and 304 captured by the optical microscope 136, the stage X and Y coordinate shifts. If the error of the correction value for the amount of rotation deviation is small, when switching to a high-magnification charged particle microscope image, an alignment pattern region 308 enters the imaging field. Here, the alignment pattern region 308 indicates a region of the alignment pattern for the charged particle microscope image in the optical microscope image. However, if the error is large, there is no pattern shape in the field of view of the captured image 309 in the charged particle microscope as shown in FIG. 3B, and the area 310 outside the field of view of the captured image 309 must be searched to detect the alignment pattern region 310. I must. Here, the alignment pattern region 310 indicates a region of the alignment pattern for the charged particle microscope image in the charged particle microscope image. In addition, by matching the adjacent line pattern 307 on the right side of 85 μm instead of the alignment pattern 306 at the time of pattern matching with the alignment pattern 306 shown in the optical microscope image 304 of FIG. The area where the alignment pattern region 310 enters the imaging field of view becomes a charged particle microscope image 313 where the alignment pattern region 310 does not enter the imaging field of the charged particle microscope image. For this reason, the alignment pattern (template pattern) included in the template image approximates or is the same as the actual circuit pattern in order to perform global alignment pattern matching with high precision using an optical microscope image at a low magnification. Needless to say that is good.

  After global alignment, automatic defect inspection and classification sequence (step 204 to step 207) of each defect candidate are executed. The correction value of coordinate deviation and rotation deviation obtained by global alignment is used for the coordinates of each defect candidate in the recipe information. The correction amount may be taken into account when the stage is moved, or the coordinates may be recalculated at once.

Next, creation of a template image used in global alignment will be described with reference to FIGS.
FIG. 4 is a diagram illustrating a method for creating a template image for global alignment of the charged particle beam apparatus according to the embodiment. FIG. 5 is a diagram illustrating a global alignment image of the charged particle beam apparatus according to the embodiment. FIG. 6A is a diagram showing a composite image created based on a global alignment image of the charged particle beam apparatus according to the embodiment, and FIG. 6B is a binary image of the charged particle beam apparatus according to the embodiment. FIG. FIG. 7A is a diagram illustrating a binarized image obtained by deleting unnecessary portions from the synthesized image of the charged particle beam apparatus according to the embodiment, and FIG. 7B is a binary image of the charged particle beam apparatus according to the embodiment. It is a figure which shows the range which cuts out the template image for low magnification from a digitized image. FIG. 8 is a diagram illustrating a high-magnification template image created from a low-magnification template image of the charged particle beam apparatus according to the embodiment.

  When creating an initial recipe (recipe information such as coordinates and imaging conditions for each point for global alignment) of a sample manufactured with a new device or process, the information obtained by executing the inspection recipe with the same device or process is completely Therefore, it does not have alignment image information. Therefore, the template image for global alignment is created by a conventional method. For example, when there is a sample, the alignment pattern portion may be cut out from the image of the peripheral portion of the alignment pattern obtained by imaging. When there is no sample, the alignment pattern portion may be generated by handwriting on the operator operation monitor 132, or a pattern image may be created by using a commercially available image processing tool or the like.

  Although it is possible to generate an alignment pattern if there is design data, it goes without saying that the alignment pattern generated from the design data does not match the length and thickness of the line segment of the actual circuit pattern generated on the sample. A decrease in accuracy during pattern matching is inevitable.

  Here, when the inspection recipe is executed, coordinate correction by global alignment is performed, so that an alignment pattern image of an actual circuit pattern imaged for pattern matching can be accumulated. With the same device and manufacturing process, it is possible to generate an alignment pattern image by overlaying and synthesizing images as described later, based on the alignment pattern image obtained by executing the recipe without design data. It is. In addition, the template image generated by synthesizing from the alignment pattern image of the actual circuit pattern that has been accumulated even compared with the template image generated from the design data should be generated in a line segment shape that approximates the actual circuit pattern. It is possible to prevent a decrease in pattern matching accuracy.

  An example of a template image creation method will be described with reference to FIG. Each time an inspection is executed, global alignment is executed, and alignment image information in an actual circuit pattern obtained by imaging on the sample is accumulated in the apparatus. Since the template image is created based on the alignment pattern image information accumulated in the apparatus, first, from the alignment pattern image information accumulated in step 401, the device name or manufacturing process name to be created is used as a search condition. All alignment pattern image information that matches the conditions is extracted. In step 402, it is determined whether the alignment pattern image information has been extracted. If the alignment pattern image information has been extracted, the process proceeds to step 403. If the alignment pattern image information has not been extracted, a template image is created by the conventional method of step 409. To do.

  Next, in step 403, image similarity is calculated as a round robin with respect to the extracted alignment pattern image. The similarity is calculated by performing pattern matching using a known image processing algorithm such as SSD (Sum of Squared Difference) or SAD (Sum of Absolute Difference). Based on the combination of images with high similarity obtained in step 404, a combined image is created by superimposing pattern line segment positions with high similarity between the images. Since the captured image of the actual circuit pattern is used, the alignment pattern may have rubbing or distortion. In addition, it is impossible for the captured images of a plurality of actual circuit patterns to be the same image. Therefore, an image suitable for the template image can be obtained by combining several points of the most similar captured images. By synthesizing images by superimposing pattern line segment positions with high similarity, a composite image having line segment formation that approximates the alignment pattern in the actual circuit pattern can be created.

  Assuming that the extracted alignment pattern images are four of FIGS. 5A, 5B, 5C, and 5D, when the images are combined, the alignment pattern 601 and the pattern matching as shown in FIG. Sometimes an image including a line pattern line segment 602 that is erroneously recognized as an alignment pattern may be obtained. As shown in FIG. 6B, when the composite image is binarized, a pattern line segment 602 that is erroneously recognized is easily visually identified. In this way, in the generation of a composite image, if all images extracted by matching are used for synthesis, depending on the image processing algorithm used for synthesis, the gap between the adjacent line segments is filled with an intermediate color, which is different from the actual one. Since it may be a line, it may be better to compose with about three images. Since the alignment pattern image information in the actual circuit pattern accumulated every time the inspection recipe is executed increases, many composite images can be obtained. Therefore, in step 405, an image that seems to be optimal is selected from the several composite images created in step 404, and a template image candidate is determined. There may be a plurality of template image candidates to be selected. In that case, it is preferable to add the priority of the template image used at the time of pattern matching to the image information.

  In step 406, depending on the image processing algorithm used at the time of pattern matching, the pattern matching accuracy may be reduced due to the influence of the characteristic line segment pattern. A binarized image 701 as shown in FIG. 7A is created by deleting an unnecessary line segment shape that seems to affect the image quality. In step 407, as shown in FIG. 7B, a necessary range 702 as a template image is selected from the binarized image 701 and cut out to create a template image 703 for an optical microscope image at a low magnification. An alignment pattern 601 in FIG. 7C is a template pattern included in the template image 703. Here, in order to improve the accuracy of the pattern matching, an idea such as thickening the line segment pattern may be added.

  When the creation of the template image for the optical microscope image at the low magnification is completed, in step 408, the template image 802 for the charged particle microscope image at the high magnification is created based on the template image 703 at the low magnification. As shown in FIG. 8, first, a point 801 at which the line pattern line segment 601 of the alignment pattern serving as the center of the visual field at the high magnification in the template image 703 at a low magnification intersects with the imaging magnification at a right angle is enlarged. Next, a range is selected from the enlarged image and cut out to create a template image 802 for a charged particle microscope image. An alignment pattern 801 in FIG. 8 is a template pattern included in the template image 802. Regarding the template image for the charged particle microscope image, since there is a template image for the optical microscope image that has already been created by synthesizing the alignment pattern image, it can be easily created automatically.

  FIG. 9 is a diagram showing a template image creation screen in the charged particle beam apparatus according to the embodiment.

  Creation of a template image for global alignment is performed by a GUI (Graphical User Interface) as shown in FIG. First, in order to create a template image for global alignment, a device name input area 902 or a process that is a parameter input area of the template image creation screen 901 is used on the operator operation monitor 132 or the external interface 133 (display device and input device). A name is entered in the name input area 903 and a search 904 is executed.

  A list of alignment pattern images that match the input search condition is displayed in the alignment pattern image list 905. Next, an execution condition of an image processing algorithm for combining images is selected from a pull-down menu of condition 906. The image processing algorithm here is, for example, filters (for example, Gaussian or smoothing) performed on the image before edge detection. Then, by clicking on the image composition execution 907 with the mouse, pattern matching is executed for all the images displayed in the list, and 2 from the left of the composite image list 908 in the descending order of the matching score value in the synthesized image. Display the valuated composite image. If there is a line segment that is considered unnecessary for pattern matching, an image is individually selected by clicking on the combined image list 908 with the mouse to select it for enlargement, and the range is selected and deleted with the mouse. Repeat the above operation until there are no more unnecessary line segments. Here, since the amount of image information increases as the alignment pattern image is accumulated, if the amount of information is large, pattern matching is executed after arbitrarily selecting the image information displayed in the alignment pattern image list 905. Also good.

  Click and select a composite image that seems to be optimal in the low-magnification composite image list 908 with the mouse, and register it in the recipe information as a low-magnification template image. To do. By executing the low-magnification image registration 909, a high-magnification image is automatically created and the execution result is displayed in the high-magnification composite image list 910.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiments, examples, and modifications. However, the present invention is not limited to the above-described embodiments, examples, and modifications. It goes without saying that various changes can be made.

101 ... charged particle optical system 102 ... charged particle source 103 ... sample 104 ... sample exchange chamber 105 ... sample chamber 106 ... sample stage 107 ... amplifier 108 ... high voltage Control unit 109 ... retarding voltage control unit 110 ... first condenser lens control unit 111 ... second condenser lens control unit 112 ... alignment control unit 113 ... deflection current control unit 114 ... Objective lens control unit 115 ... extraction electrode 116 ... primary charged particle beam 117 ... first condenser lens 118 ... second condenser lens 119 ... alignment coil 120 ... objective lens 121 ... Secondary charged particle beam 122 ... secondary charged particle detector 123 ... defect detection control unit 124 ... automatic defect classification control unit 125 ... display monitor 1 6 ... Image memory 127 ... Image processing control unit 128 ... Image correction control unit 129 ... Image display unit 130 ... Charged particle optical system control unit 131 ... Device control unit 132 ... Monitor for operator operation 133... External interface 134... Stage control unit 135... Transport control unit 136.

Claims (15)

A charged particle optical system for irradiating a sample with a charged particle beam;
A sample stage for holding the sample;
A controller for controlling the charged particle optical system and the sample stage;
An arithmetic unit for generating an image of the sample from a signal of secondary charged particles obtained from the sample by irradiation of the charged particle beam;
A storage device for storing an image of the sample;
With
The arithmetic device sets a pattern specified on a composite image created by superimposing a plurality of pre-captured images as a template pattern, and detects a position of a pattern that matches the template pattern. A charged particle beam apparatus characterized by: In claim 1,
The charged particle beam apparatus according to claim 1, wherein the pre-captured image is an image of a sample of the same device or manufacturing process as the sample held on the sample stage. In claim 1,
The charged particle beam apparatus, wherein the composite image is created by superimposing images having high similarity among a plurality of images captured in advance. In claim 3,
The charged particle beam apparatus according to claim 1, wherein the image having a high similarity is determined by performing pattern matching on a plurality of images captured in advance and calculating a similarity for each combination of images. In claim 1,
A plurality of the composite images are generated,
The charged particle beam apparatus, wherein the template pattern is set on a composite image selected from a plurality of the composite images. In claim 1,
A display device for displaying an image obtained by binarizing the composite image;
A charged particle beam device comprising: an input device for designating and inputting the template pattern on the binarized image. In claim 1,
A charged particle beam apparatus, wherein a necessary range is selected from a template image that is an image including the template pattern and cut out to create a template image for low magnification. In claim 7,
A charged particle beam apparatus, wherein a template image for high magnification is created based on the template image for low magnification. In claim 1,
Equipped with an optical microscope,
The charged particle beam apparatus according to claim 1, wherein the pre-captured image is captured by the optical microscope. In claim 1,
A charged particle beam apparatus that corrects a positional deviation of the sample based on a position of a pattern that matches the template pattern. A charged particle optical system for irradiating a sample with a charged particle beam, a sample stage for holding the sample, a control device for controlling the charged particle optical system and the sample stage, and a sample obtained by irradiation with the charged particle beam. A charged particle beam device comprising: an arithmetic device that generates an image of the sample from a signal of secondary charged particles that is generated; and a storage device that stores the image of the sample;
A pattern specified on a composite image created by superimposing a plurality of pre-captured images is set as a template pattern, and a position of a pattern that matches the template pattern is detected. Non-transitory and tangible program storage medium. In claim 11,
The non-transitory and tangible program storage medium, wherein the composite image is created by superimposing images having high similarity among a plurality of images captured in advance. In claim 12,
The non-temporary and tangible program storage is characterized in that the image having a high similarity is determined by performing pattern matching on a plurality of images captured in advance and calculating a similarity for each combination of images. Medium. In claim 11,
A plurality of the composite images are generated,
A non-transitory and tangible program storage medium, wherein the template pattern is set on a composite image selected from a plurality of the composite images. In claim 11,
A non-transitory and tangible program recording medium characterized in that a positional deviation of the sample is corrected based on a position of a pattern matching the template pattern.

Images