JP6207893B2 - Template creation device for sample observation equipment - Google Patents

Description

  The present invention relates to a template creation device for a sample observation device.

  2. Description of the Related Art Conventionally, sample observation apparatuses for inspecting or observing observation objects such as semiconductor wafers, photomasks, magnetic devices, liquid crystal substrates, and the like have been used. For example, in semiconductor manufacturing process management, an electron beam automatic length measuring device is used as a sample observation device (inspection / observation system) in order to measure a pattern dimension with high accuracy.

  In recent years, the miniaturization of semiconductors has progressed due to higher functionality of digital home appliances, and patterns in integrated circuits have become more complicated. Along with this, the length measurement points in the semiconductor manufacturing process are also increasing, and high throughput and length measurement accuracy are required for the electron beam automatic length measurement apparatus. Furthermore, in terms of manufacturing, high-efficiency production by automation is required for the purpose of cost reduction, short TAT (Turn-around time) and high quality.

  In order to automatically measure an observation object with an electron beam automatic measuring device, it is necessary to create a recipe file. This recipe file is a condition setting file in which various parameters for performing inspection or observation processing such as sample size, device pattern size information, defect detection, line width measurement, and the like are set. A recipe file is created for each device type or manufacturing process and stored on the system. By using this recipe file, inspection or observation processing such as defect detection and line width measurement can be automatically performed by the sample observation apparatus.

  In order to detect the pattern position with high accuracy during the execution of the recipe file, the positioning by the template matching method is mainly used at present. The template matching method will be described with reference to FIG. FIG. 12 is a schematic diagram of template matching.

  In template matching, an image on a known coordinate on the sample coordinate system or an image whose positional relationship is clear relative to the alignment coordinate is registered in advance as a template 1200. The template forms one frame of an image with 256 to 2048 electron beam scans, and in recent years only 4096 captured images. When the contrast of the pattern is low, a clearer image can be obtained by increasing the number of scanning lines or the number of frames.

  A sample as an observation object is positioned and placed on a stage provided in the sample observation apparatus. Then, a captured image 1201 on a known coordinate on the sample coordinate system or a captured image 1201 whose relative positional relationship is clear with respect to the alignment coordinate is acquired. Then, the template 1200 and the acquired captured image 1201 are overlapped by raster scan to identify the most similar part. Thereby, the position on the stage coordinate system of the sample observation apparatus is specified. In order to improve matching accuracy, the method described in Patent Document 1 is also used.

  By the way, when the coordinates are specified by template matching using the template 1202, there are cases where accurate position detection cannot be performed. This is because there are a plurality of locations on the captured image 1201 that are similar to the pattern in the template 1202, and a similar portion or a different portion on the captured image 1201 cannot be determined.

  In addition, when the recipe file is actually automatically executed, the similarity or difference between the image registered as the template and the captured image may decrease due to a process variation or the like. If the degree of similarity or difference between the template and the captured image decreases, the number of cases where a desired pattern position cannot be detected and an error occurs increases. As a result, such an increase in errors prevents automatic execution by the recipe file.

JP 2006-190201 A

S. M. Smith and J. M. Brady, "SUSAN-a new approach to low level image processing", Int. J. Compt. Vis., Vol.23, no.1, pp.45-78, May1997.

  In general, since the recipe file is created at the device development stage, the appearance of the template image is slightly different at the device mass production stage. Therefore, when a recipe file once created is used for a long period of time, the appearance of the registered template image and the captured image may differ with time due to the influence of the process or the like. As a result, when the template matching is performed, the score is lowered, and a desired pattern position cannot be detected and an error occurs.

  The present invention has been made in view of these problems associated with template matching, and an object thereof is to provide a template creation apparatus for a sample observation apparatus for supporting automation by a recipe file.

  In order to solve the above problems, for example, the configuration described in the claims is adopted. The present application includes a plurality of means for solving the above-mentioned problems. To give an example, a template creation apparatus for a sample observation apparatus that scans a charged particle beam on a sample to obtain a captured image of the sample is provided. Is done. The template creation device includes a storage unit that stores parameter information when performing observation by the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device; An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images.

  According to the present invention, an image can be automatically selected from captured images, and an average image can be generated as a template image from the selected captured images. By registering this template image in the recipe file, it is possible to support the automation process using the recipe file.

  Further features related to the present invention will become apparent from the description of the present specification and the accompanying drawings. Further, problems, configurations and effects other than those described above will be clarified by the description of the following examples.

It is a schematic block diagram of a scanning electron microscope (SEM). 1 is a system configuration diagram of an inspection / observation system including an SEM. It is a flowchart explaining the process which selects the image for template creation for creating a template image automatically. It is a flowchart explaining the process performed after the process of FIG. 3A. It is a flowchart explaining the process performed at the time of review of imaging conditions. It is an example of a score list. This is an example in which score data is displayed for each date. This is an example in which score data is displayed for each process number (PNo). FIG. 6 is an example of a correlation diagram between a focus difference ΔOBJ and a score. It is the figure which showed the detection position (FOV: * 200k) of the captured image preserve | saved after performing template matching. It is an example of a GUI screen when creating recipe file information. It is an example of a template image generation screen. It is an example of a wafer map. It is the figure which showed the automatic generation of the template image for every item (Average, ChipNo, SelectChip). It is an example of the screen for setting a template image creation function to a recipe file. It is a schematic diagram of template matching.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. The accompanying drawings show specific embodiments in accordance with the principle of the present invention, but these are for the understanding of the present invention, and are never used to interpret the present invention in a limited manner. is not.

  As a sample observation apparatus, there is a charged particle beam apparatus that uses a secondary generated electron by scanning a charged particle beam (for example, electrons) on a sample surface. A charged particle beam apparatus is an apparatus that accelerates particles (charged particles) having a charge such as electrons and cations with an electric field and irradiates a sample. A charged particle beam apparatus performs observation, analysis, processing, and the like of a sample by utilizing an interaction between the sample and charged particles. The present invention can also be applied to a scanning electron microscope, a scanning transmission electron microscope, a transmission electron microscope, an ion microscope, a focused ion beam device, a combined device of these and a sample processing device, or an inspection / observation device using these devices. .

  Hereinafter, a scanning electron microscope (SEM) will be described as a representative example of the charged particle beam apparatus. A scanning electron microscope obtains a captured image by scanning a fine pattern on a sample with an electron beam and detecting secondary electrons generated from the sample with a detector. The scanning electron microscope can detect a position of a desired pattern by executing a template matching process between the acquired captured image and the template image.

  According to the demand for forming a fine pattern such as an ArF resist, the resist pattern used in recent years may shrink (shrink) depending on the electron beam irradiation conditions. When creating a recipe file, if a shrunken pattern is registered as a template image, the score at the time of pattern matching is reduced to a threshold value or less, and accurate position information cannot be obtained. Further, even when the process conditions are changed in the previous process, a difference in appearance between the captured image and the template image occurs in the same manner, resulting in a pattern matching error. Therefore, hereinafter, an apparatus for automatically creating and registering a template image when a template matching error occurs will be described.

  FIG. 1 is a diagram illustrating an outline of an SEM that is an embodiment of a sample observation apparatus. The scanning electron microscope includes an electron optical system that irradiates a sample with an electron beam emitted from an electron source, and a vacuum exhaust structure for exhausting the electron optical system. In a scanning electron microscope, a sample is observed by irradiating a sample with a primary electron beam irradiated from an electron source in the lens body and detecting a secondary electron beam generated thereby with a detector.

  Specifically, an electron beam 103 extracted from the electron source 101 in the mirror body 100 by the extraction electrode 102 and accelerated by an acceleration electrode (not shown) is narrowed by a condenser lens 104 which is a form of a focusing lens. The narrowed electron beam 103 is scanned one-dimensionally or two-dimensionally on the sample 106 in the sample chamber 114 by the scanning deflector 105.

  The electron beam 103 is decelerated by a negative voltage applied to an electrode (not shown) built in the sample stage 107, is focused by the lens action of the objective lens 108, and is irradiated onto the sample 106.

  When the sample 106 is irradiated with the electron beam 103, electrons such as secondary electrons 109 or backscattered electrons are emitted from the irradiated portion. The emitted electrons are accelerated in the direction of the electron source by the acceleration action based on the negative voltage applied to the sample 106, collide with the conversion electrode 110, and generate secondary electrons 109.

  The secondary electrons 109 emitted from the conversion electrode 110 are captured by the detector 111, and the output of the detector 111 changes depending on the amount of captured secondary electrons. The control device 210 converts the electronic output into an electrical signal, and outputs an image displaying black and white shading shown in multi-level tone. 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 105 and the output of the detector 111.

  The scanning electron microscope illustrated in FIG. 1 includes a deflector (not shown) that moves the scanning region of the electron beam 103. This deflector is used to form an image of a pattern having the same shape existing at different positions. This deflector is also called an image shift deflector, and enables movement of the field of view of the electron microscope without moving the sample by the sample stage 107 or the like. Note that the image shift deflector and the scanning deflector 105 may be configured as a common deflector, and the image shift signal and the scanning signal may be superimposed and supplied to the deflector.

  The mirror body 100 may include other lenses, electrodes, and detectors other than those described above, or some of them may be different from the above, and the configuration of the electron optical system is not limited to this. In the example of FIG. 1, an example is described in which electrons emitted from the sample 106 are converted by the conversion electrode 110 and detected. However, the present invention is not limited to such a configuration. It is possible to adopt a configuration in which the detection surface of the electron multiplier tube or the detector is arranged on the orbit.

  The sample 106 is carried into the load lock chamber 113 through the mini-en 112 and then carried into the sample chamber 114. The load lock chamber 113 is provided with gate valves 115 and 116 so that the load lock chamber 113 can be evacuated.

  FIG. 2 is a detailed explanatory diagram of the inspection / observation system including the SEM. The system includes an SEM main body 201, a control device 210 for controlling the SEM main body 201, and an arithmetic processing device 220. The SEM body 201 is the one shown in FIG. The control device 210 controls each component of the SEM main body 201. Further, the control device 210 measures the pattern width of the pattern formed on the sample 106 based on the intensity distribution of detected electrons called a line profile and an image forming unit that forms an image based on the detected electrons. May be provided.

  The arithmetic processing unit 220 may be realized using a general-purpose computer. For example, the arithmetic processing device 220 includes a central arithmetic processing device, an auxiliary storage device, and a main storage device. For example, the central processing unit is constituted by a CPU (Central Processing Unit) or image processing hardware such as an ASIC or FPGA. For example, the auxiliary storage device is a hard disk, and the main storage device is a memory.

  The processing unit of the arithmetic processing device 220 described below may be realized as a function of a program executed on a computer. That is, processing described below may be stored in a memory as program code, and the CPU may execute each program code. Further, the processing unit described below may be realized by hardware, for example, by designing with an integrated circuit.

  The arithmetic processing device 220 includes an arithmetic processing unit 230 that supplies a predetermined control signal to the control device 210, a memory 240 that stores image information (captured image) and recipe file information acquired by the SEM body 201, and an input / output device. 250. Here, the recipe file information includes information such as sample size, device pattern size information, various parameters when performing inspection or observation processing such as defect detection and line width measurement. The recipe file information may be created for each device type or manufacturing process and stored in the memory 240.

  Electrons emitted from the sample 106 are captured by the detector 111 and converted into a digital signal by an A / D converter built in the control device 210. Then, image processing hardware such as a CPU, ASIC, and FPGA built in the arithmetic processing unit 220 executes image processing according to the purpose.

  The arithmetic processing unit 230 includes a stage control unit 231, a matching execution unit 232, a visual field movement control unit 233, a dimension measurement unit 234, and an automatic template generation unit 235.

  The stage control unit 231 sets a moving condition of the sample stage 107 on which the sample 106 is arranged. The matching execution unit 232 executes template matching between the acquired captured image and the template image. This template image is registered in advance in the recipe file information. Template matching is a technique for specifying a location where a captured image to be aligned and a template image match based on similarity determination using a normalized correlation method or the like. The matching execution unit 232 outputs the similarity as a matching score. Note that template matching is not limited to this method, and other existing methods may be used.

The matching execution unit 232 specifies a desired position of the captured image based on the similarity determination. For example, when the recipe file is executed, the position of the captured image is specified three times as follows.
(1) When performing SEM alignment in which the rotation and magnification of the shot matrix formed on the base of the sample 106 are calculated and adjusted to the coordinate system of the sample stage 107 (2) To recognize the approximate position of the measurement position When pattern matching is performed at a position several μm away from the measurement position at a field magnification of the electron microscope x 10k (hereinafter, addressing)
(3) When pattern matching is performed at or near the measurement position at a field magnification of the electron microscope × 200k times in order to recognize an accurate measurement position

  The visual field movement control unit 233 sets the deflection condition of the image shift deflector. Both the stage control unit 231 and the visual field movement control unit 233 are for determining the scanning position of the electron beam 103, and the control device 210 is based on the setting information set by these control units 231 and 233. Control the scanning position.

  The dimension measuring unit 234 measures a pattern dimension on the sample 106 based on the obtained detection signal (for example, a line profile). The dimension measuring unit 234 can calculate an accurate pattern dimension.

  The template automatic generation unit 235 automatically selects a captured image (template creation image) for generating a template image from a plurality of captured images stored in the memory 240. Furthermore, after selecting a captured image, the template automatic generation unit 235 creates an average image as a template from the selected plurality of captured images. Thereafter, the template automatic generation unit 235 stores the template image in the memory 240.

  During execution of the recipe file, when a pattern matching error occurs, the template automatic generation unit 235 creates an average image as a template image from a plurality of selected captured images, and the template image is created as a recipe file. Automatically register online. Then, the inspection / observation system executes the recipe file again.

  Note that the template automatic generation unit 235 selects a captured image according to conditions set on a template image setting screen (see FIGS. 9A and 9B) described later, and creates a template image from the selected plurality of captured images.

  The input / output device 250 includes an input unit such as a keyboard and a pointing device (such as a mouse) and an output unit such as a display, a LAN port, and a USB port. A display (display unit) provided in the input / output device 250 displays a GUI (Graphical User Interface) for displaying images, inspection results, and the like to the operator. Further, the input / output device 250 is an imaging recipe file creation device for creating a recipe file including the coordinates of an electronic device required for inspection, a pattern matching template used for positioning, imaging conditions, and the like. Also works.

  Next, a process for automatically generating a template image to be registered in the recipe file will be described. In the embodiment described below, a process for automatically selecting a template creation image for creating a template image when an error occurs due to a decrease in the matching score between the template image and the captured image will be described. To do. A process for creating a template image from a template creation image and registering the template image in a recipe file online will also be described.

  FIG. 3A is a flowchart for describing processing for automatically selecting a template creation image for creating a template image. The subject of the processing in FIG. 3A is the arithmetic processing unit 220. The present embodiment is a flow performed during length measurement using a scanning electron microscope, but the following example can be applied to SEM alignment and addressing in which position detection is performed by template matching. is there.

  When the template image is automatically generated, it is desirable that the image is more robust without causing an error even when the pattern size or contrast is slightly changed. In addition, it must be stable for a long time without error. Therefore, it is necessary to actually execute the recipe file in a certain period and create a highly accurate template image that matches the actual situation from the captured image.

  In step S301, recipe file information stored in the memory 240 is executed in order to perform length measurement by SEM. The recipe file information includes parameters such as the size of the sample that is the imaging target, the pitch of the device pattern formed on the sample, arrangement information, and the measurement position. Recipe file information is created for each product type or process, and recipe file information corresponding to each product or process is executed. When executing the recipe file information, each control unit of the arithmetic processing unit 230 transmits a control signal to the control device 210 according to various parameters of the recipe file information. The control device 210 controls each component of the SEM main body 201 according to the control signal.

  In step S302, various types of information obtained by executing the recipe file information are stored in the memory 240. Specifically, the arithmetic processing unit 220 stores in the memory 240 an MSR file in which the captured image of the sample 106 that is the imaging target, the length measurement result, the focus result, and the like are stored. Here, the MSR file is a file including the imaging conditions of the SEM body 201, the captured image of the sample 106, and the length measurement result. Note that product or process information is added to the captured image as supplementary information.

  In step S303, the matching execution unit 232 of the arithmetic processing device 220 executes template matching on the captured image obtained for each product type or process using a template image registered in advance in the recipe file. The matching score between the captured image and the template image is stored in the memory 240 in association with the captured image.

  Hereinafter, a process of selecting a template creation image from a plurality of captured images stored in the memory 240 will be described. In step S304, the template automatic generation unit 235 determines whether the score obtained in step S303 is equal to or greater than a threshold value. If the score is less than the threshold value, the process proceeds to step S311 to review the imaging conditions. On the other hand, if the score is greater than or equal to the threshold, the process proceeds to step S305.

  In step S305, the template automatic generation unit 235 determines whether the uniqueness of the template image is greater than or equal to a threshold value. The reason for making such a determination is that the higher the uniqueness, the higher the reliability of the score. As a method for calculating the uniqueness, the similarity (score) or difference between the captured image that is closest to the template image on the captured image and the captured image that is second closest to the template image on the captured image There is a method that uses a ratio of such indicators. The uniqueness calculation method is not limited to this method, and other existing methods may be used. If the uniqueness is less than the threshold value, the process advances to step S311 to review the imaging conditions. On the other hand, if the uniqueness is greater than or equal to the threshold, the process proceeds to step S306. Although the score and uniqueness are used in this example, these threshold values can be arbitrarily set.

  In step S306, the template automatic generation unit 235 compares scores between a plurality of captured images stored in the memory 240, and determines whether the difference in scores is equal to or less than a threshold value. The comparison of the scores is performed on captured images having the same pattern or similar patterns.

  FIG. 5 is an example of a list of scores stored in the memory 240. The memory 240 of the arithmetic processing unit 220 stores captured images for each product type or process. The captured image data is stored for the number of times the recipe file information has been executed. The captured image is stored in association with the date and score when the captured image is acquired as supplementary information. The score list in FIG. 5 has five measurement lengths for a sample (measurement number 5 points / sample), and two captured images are acquired for each measurement point in one day (2 sheets / sample). In this example, captured images are acquired over a period of 3 days. The score list includes items such as product types and processes. As described above, the score may be stored in association with the product type, the process, or the like.

  The score list data in FIG. 5 can be visually confirmed on the display of the input / output device 250. Note that the score list can also be displayed on the display of the input / output device 250 in the format of FIGS. 6A and 6B. FIG. 6A is an example in which all stored score data is displayed for each date. FIG. 6B shows an example of display for each process number (PNo) assigned to each length measurement pattern when measuring a plurality of locations in the chip. Note that the period for acquiring the captured image can be arbitrarily set on the GUI screen on the input / output device 250.

  As shown in FIG. 5, since the pattern matching score is accumulated for the same or similar captured images, the difference in score can be determined using the accumulated information. The reason for making such a determination means that the greater the difference in the score, the lower the similarity with the captured image that has been accumulated so far, and as a template creation image candidate for creating a template image. This is because it may not be appropriate. Thereby, an inappropriate captured image can be excluded from the template creation image candidates, and a highly accurate template image candidate can be selected. In step S306, if the score difference is equal to or smaller than the threshold value, the process proceeds to step S307. On the other hand, if the score difference is larger than the threshold value, the process advances to step S312 to exclude the captured image from the template creation image candidates. It is possible to arbitrarily set the difference threshold.

In step S307, the template automatic generation unit 235 determines the correlation between the focus difference ΔOBJ of the objective lens 108 and the score. In step S308, the template automatic generation unit 235 determines the correlation between the retarding voltage focus difference ΔOBJ and the score. Usually, focusing is performed near the length measurement pattern before and after performing template matching. There are the following two focusing methods.
(1) Method of exciting the objective lens 108 (2) Method of applying a voltage (retarding voltage) to the sample 106 Here, the focus difference ΔOBJ is a difference in focus performed before and after template matching.

  According to the sample 106, the mode (1) or (2) is selected, and focusing is performed with higher accuracy. In FIG. 3A, both steps S307 and S308 are described, but either step may be executed according to the selected mode.

  The focus values before and after template matching are stored in captured images and MSR files in which the length measurement results are stored. The template automatic generation unit 235 calculates the focus difference ΔOBJ using the stored focus value. The correlation between the score and the focus difference ΔOBJ is output on the display of the input / output device 250. Thereby, the correlation between the focus difference ΔOBJ performed before and after the template matching and the score can be confirmed on the GUI.

  FIG. 7 shows an example of a correlation diagram between the score and the focus difference ΔOBJ. In the graph of FIG. 7, the horizontal axis is score, and the vertical axis is ΔOBJ. When ΔOBJ is large, the score tends to be low. Conversely, when ΔOBJ is small, the score tends to be large. In order to create a template image with higher accuracy, an image in which the correlation between the score and the focus difference ΔOBJ falls within a predetermined range (the dotted line range in FIG. 7) is left as a template creation image. The score threshold and the ΔOBJ threshold can be arbitrarily set so that an image having a small ΔOBJ and a high score can be selected.

  In steps S307 and S308, the template automatic generation unit 235 proceeds to step S309 when the correlation between ΔOBJ and the score falls within a predetermined range. If not, the process proceeds to step S312, and the captured image is excluded from the template creation image candidates.

  Next, in step S309, the template automatic generation unit 235 determines whether or not the pattern detection position is within the field of view (FOV) of the electron microscope with respect to the captured image. This determination will be described with reference to FIG. FIG. 8 shows a pattern detection position when pattern matching is performed with an FOV of × 200k. In FIG. 8, there is one portion where pattern detection is performed outside the range of the FOV. It is considered that an image outside the range of the FOV is inappropriate as a template creation image. Therefore, it is excluded from the template creation image candidates. Thereby, a template image with higher accuracy can be created. If the pattern detection position is within the FOV, the process proceeds to step S310. On the other hand, if not, the process proceeds to step S312, and the captured image is excluded from the template creation image candidates.

  In step S310, the template automatic generation unit 235 stores the captured image that has not been excluded in the above steps in the memory 240 as a template creation image. The template automatic generation unit 235 aligns each image of the template creation image, and creates an average image by image averaging processing. Different template images are created according to the settings. Details of the creation of the template image will be described later with reference to FIGS. 10A and 10B. The template automatic generation unit 235 stores the created image in the memory 240 as a template image.

Note that the period for acquiring the captured image can be arbitrarily specified, and a highly accurate template image is generated and updated every day. A template image is generated and stored for each PNo.
Here, PNo is a process number assigned for each length measurement pattern when measuring a plurality of locations in the chip. It should be noted that the picked-up image selected as the template creation image may be arbitrarily specified as to whether all points in the sample 106 or in units of chips. In addition, when the captured image is different in some chip locations of the sample 106 compared to other chip locations due to the influence of the previous process, the template automatic generation unit 235 uses the average image of the specific location of the sample 106 and other locations. The average image can be generated.

  Next, processing when the process proceeds to step S311 in FIG. 3A will be described. FIG. 4 is a flowchart for explaining processing executed when the imaging condition is reviewed. The subject of the processing in FIG. 4 is the arithmetic processing unit 220.

  In step S401, it is determined whether the number of frames when scanning the length measurement pattern on the sample 106 with the electron beam 103 is 64 or more. For example, when the pattern contrast is low, increasing the number of frames increases the contrast information, and a clearer captured image can be obtained. However, when the number of pixels when acquiring an image is as large as 1024, 2048, and 4096, the number of scanning the electron beam 103 with respect to the length measurement pattern increases, and the dose per unit area increases. This increases the damage to the pattern. Furthermore, the processing time increases. Taking these into account, the number of frames is normally set in the range of 8 to 64.

  In step S401, if the number of frames is 64 or less, the process proceeds to step S402. On the other hand, if the number of frames is greater than 64, the process proceeds to step S405. When the process proceeds to step S405, the template image or the captured image pattern is changed, and the process returns to step S301 in FIG. 3A again.

  When the process proceeds to step S402, 8 is added to the number of frames currently set in the arithmetic processing unit 220. Note that the number of frames to be added can be arbitrarily specified. Next, in step S <b> 403, the recipe file information is executed again so as to acquire the captured image with the number of frames set in step S <b> 402, and a signal is transmitted to the control device 210. Thereby, a captured image is acquired again.

  Next, in step S404, the matching execution unit 232 of the arithmetic processing device 220 executes template matching using the template image for the captured image acquired in step S403, and determines whether the score is equal to or greater than a certain threshold value. judge. When the score is smaller than the threshold value, the process returns to step S401 again, and S401 to S404 are repeatedly executed. In this case, the number of frames is added within a range of 64 or less, and the recipe file is repeatedly executed until the score becomes equal to or greater than the threshold value. On the other hand, if the score is equal to or greater than the threshold, the process proceeds to step S406, and the process returns again to step S301 in FIG. 3A.

  FIG. 3B is a flowchart illustrating a process executed after the process of FIG. 3A. The subject of the processing in FIG. 3B is the arithmetic processing unit 220. In step S321, SEM observation is executed using the recipe file information.

  Next, in step S322, the matching execution unit 232 of the arithmetic processing device 220 executes template matching using the template image, and determines whether the score is equal to or greater than a certain threshold value. When the score is smaller than the threshold value, the process proceeds to step S323. On the other hand, if the score is equal to or greater than the threshold value, the process proceeds to step S324, and the process proceeds to length measurement using the next length measurement line profile.

  Next, when it progresses to step S323, a new template image is registered on a recipe file online. Thereby, while the recipe file is being executed, the template image can be updated online and the recipe file can be automatically completed. The type of template image to be registered will be described later with reference to FIGS. 10A and 10B. After the template image is newly updated, the process returns to step S321 to execute the recipe file information again. The process of FIG. 3B is repeatedly executed until the score becomes equal to or higher than the threshold value.

  FIG. 9A shows an example of a GUI screen when creating recipe file information. FIG. 9A is a recipe file sequence setting screen (Sequence Editor), which is displayed on the display of the input / output device 250. In the following, only items relating to the generation of template images will be described in particular.

  The screen 901 includes an item (TP Image) 902. By turning on the item (TP Image) 902 on the screen 901, the captured image is stored in the memory 240 of the arithmetic processing unit 220. The captured image is stored in the memory 240 for each product type or process. Further, by pressing a button 903 next to the item (TP Image) 902, the template image generation screen of FIG. 9B is displayed.

  FIG. 9B is an example of a template image generation screen. The template image generation screen 911 includes a first item (Date) 912 for setting a period for generating a template image, a second item (ΔOBJ vs Score) 913 for setting a threshold value of ΔOBJ and score, A third item (Image) 914 for setting a template image creation method and a fourth item (Image Change) 915 for setting a target for creating a template image are provided.

  In the first item 912, an arbitrary period such as Week or Month can be set as a period for generating a template image according to the number of processed lots. In the second item 913, a focus difference ΔOBJ threshold (ΔOBJ TH), a score threshold (Score TH), and the like can be set. This makes it possible to perform determination by template matching and determination of the correlation between the focus difference ΔOBJ and the score. Although not shown here, an item for setting a uniqueness threshold value or the like may be provided. The threshold may be set for each product type and process.

  In the third item 914, an arbitrary method can be set as a template image creation method from “Average”, “ChipNo”, and “SelectChip”. When “Average” is selected, a template image is generated for each PNo from all stored captured images. When the variation in contrast is small at the measurement location in the sample, a highly accurate template image with high robustness can be generated.

  When “ChipNo” is selected, a template image is created from a plurality of captured images stored within a period for each PNo in units of chips. Template image registration increases depending on the number of chips to be measured and the PNo. However, matching with the template image generated for each chip is performed, so matching errors are generated as much as possible even when there is large variation in contrast at the measurement points in the sample. Can be suppressed.

  In addition, when “SelectChip” is selected, a template image is created from a plurality of captured images stored in a period for each PNo on a chip in which an error has occurred, and other chips are imaged for other chips. Create an average template image from the images. When the pattern shape, contrast, and the like are different in a specific chip, it is possible to generate a highly accurate template image by individually creating the template images.

  In a fourth item 915, a target for creating a template image is set. The fourth item 915 can set a target for creating a template image related to the presence or absence of an error. When “ALLPNo” is designated, template images are registered for all PNos regardless of whether an error has occurred. When “Error PNo” is designated, the template image is registered only with the PNo in which an error has occurred.

  FIG. 10A is an example of a semiconductor wafer map. FIG. 10A shows an example in which the shot matrix is 7 × 7, there are three chips to be measured in the sample, and there are two patterns to be measured in the chip. The coordinates of the chip in the sample are (1, 4), (4, 4), and (4, 1), respectively. It is assumed that there are two length measurement patterns (PNo1, PNo2) in the chip.

  FIG. 10B is a diagram illustrating automatic generation of a template image for each item (Average, ChipNo, SelectChip). When “Average” is selected on the template image generation screen 911, two average template images are created from the captured images selected for each PNo for all stored captured images (P1, P2). ).

  When “ChipNo” is selected on the template image generation screen 911, six template images are created for each PNo from the captured images stored for each (1, 4) (4, 4) (4, 1) chip. (C_P1_1-4 to C_P1_4-1 and C_P2_1-4 to C_P2_4-1). Here, a chip No. for identifying which chip was generated is added to each template image as attached information.

  For example, assume that a template matching error has occurred in the (4, 4) chip. In this case, when “SelectChip” is selected on the template image generation screen 911, two template images are created for each PNo from the captured images stored in the chip (4, 4) in which the template matching error has occurred (S_P1_4). -4, S_P2_4-4). In addition, two average template images are created for each PNo from captured images obtained with the other chips (1, 4) and (4, 1) (S_P1, S_P2).

  In this example, the location where an error has occurred is one chip, but if it occurs on multiple chips, a template screen is created from each captured image stored in each chip, and no error has occurred. Create an average image with the chip. As in the case where “ChipNo” is selected, a chip number for identifying which chip is generated is added to the template image as attached information.

  In step S310 of FIG. 3A, different template images are created according to the settings (“Average”, “ChipNo”, “SelectChip”). Further, in step S323 of FIG. 3B, the method of registering the template image in the recipe file differs for each of “Average”, “ChipNo”, and “SelectChip”.

  When “Average” is set, the following processing is performed. Even when a template matching error occurs, an average template image is created from the picked-up images selected for each PNo for all stored picked-up images of the PNo template image in which the error has occurred ( P1, P2). Then, the template images (P1, P2) are registered in the recipe file.

  When “ChipNo” is set, the following processing is performed. A template image is created for each PNo from captured images stored for each (1, 4) (4, 4) (4, 1) chip. Then, template images (C_P1_1-4 to C_P1_4-1 and C_P2_1-4 to C_P2_4-1) are registered in the recipe file.

  When “SelectChip” is set, the following processing is performed. When an error occurs in the (4, 4) chip, a template image is created for each PNo from the captured image saved in the (4, 4) chip in which the template matching error has occurred (S_P1_4-4, S_P2_4-4). ). Further, an average template image is created for each PNo from captured images obtained with the other (1, 4) (4, 1) chips (S_P1, S_P2). Then, template images (S_P1_4-4, S_P2_4-4, S_P1, S_P2) are registered in the recipe file.

  In all modes, the template image registration process can be changed in relation to the presence or absence of an error. When “ALLPNo” is designated on the template image generation screen 911, the template images are registered for all PNos regardless of whether or not an error has occurred. When “Error PNo” is designated, the template image is registered only with the PNo in which an error has occurred.

  FIG. 11 shows an example of a screen (Additional Setting) for setting the above-described template image creation and registration function in a recipe file. The screen 1101 includes an item (Retry Pattern Recognition with TP Image) 1102 for setting a template image creation and registration function. If a template matching error occurs by selecting the item 1102, the processing of FIGS. 3A and 3B can be performed.

  As described above, according to the present embodiment, when a recipe file is executed and a similarity is lowered due to template matching and an error occurs, a template creation image is automatically selected from the captured images that have already been saved and selected. An average image can be automatically created as a template image from a plurality of template creation images and registered in a recipe file. Thereby, a template image can be updated online and a recipe file can be completed automatically.

  In addition, the present embodiment is characterized in that a captured image at the time of executing a recipe file can be stored for an arbitrary period. The stored captured image is stored for each product type or process. A threshold such as a score of a captured image can be set for each type or process.

  Further, according to the present embodiment, as a mode for creating a template image from a captured image stored within a certain target period, (1) an average of all the chips of the target chip (chip from which the captured image is acquired) Any one of an image, (2) an average image for each chip number, and (3) an average image for each specific chip can be arbitrarily selected.

  Further, as a method of detecting the normal focal position of the electron beam irradiated on the sample before template matching, there is a method of detecting the normal focal point by detecting the height with a Z sensor or irradiating the electron beam on the pattern. . Focusing is performed by applying excitation to the objective lens coil with the detected value. When pattern position detection is performed with high accuracy, focus adjustment is performed using a pattern near the measurement position immediately before template matching. The present embodiment has a function of automatically selecting a template creation image using the correlation between the difference ΔOBJ before and after the focus adjustment and the score.

  The present embodiment also has a function of excluding from the template creation image when the position detected by template matching is out of the field of view (FOV) of the electron microscope.

  The effect of the present embodiment will be described. The recipe file needs to be registered for each type of device that performs defect detection and line width measurement, and for each manufacturing process. Due to the increasing complexity of devices in recent years, the number of processes for performing inspection / observation processing is as many as 100 in one type. Therefore, it takes a lot of time to create a recipe file, which places a heavy burden on the operator. In order to simplify the creation of recipe files, there is a mechanism that allows template images to be saved and copied to recipe files for other device products. ing.

  Conventionally, there is an average template creation function that generates an average image from a plurality of stored captured images and handles variations between samples and between lots. However, the processing is registered in the recipe file as a template image after performing the alignment by arbitrarily selecting the captured image offline, and the sample and the apparatus are occupied, resulting in a decrease in the apparatus operating rate.

  According to this embodiment, even when a template matching error occurs, a template creation image is automatically selected from the stored captured image, and an average image can be created as a template image. In addition, since the recipe file is executed again after registering the created template image, the automation rate can be improved and the burden on the operator can be reduced. In addition, the apparatus operating rate is improved and the cost can be reduced.

  In addition, this invention is not limited to the Example mentioned above, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. In addition, a part of the configuration of one embodiment may be replaced with the configuration of another embodiment, and the configuration of another embodiment may be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  For example, in terms of automatically selecting an image for creating a template, this apparatus only needs to include at least one process of steps S304 to S309 in FIG. 3A. In this embodiment, steps S304 to S309 in FIG. 3A are provided, but this is a more preferable form, and the present invention can be configured by deleting a part of the processing in FIG. 3A.

  Further, as described above, each processing unit of the arithmetic processing unit 230 may be realized by a program code of software that realizes the function of the embodiment. In this case, a storage medium in which the program code is recorded is provided to the information processing apparatus, and the information processing apparatus (or CPU) reads the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the program code itself and the storage medium storing it constitute the present invention. As a storage medium for supplying such program code, for example, a flexible disk, CD-ROM, DVD-ROM, hard disk, optical disk, magneto-optical disk, CD-R, magnetic tape, nonvolatile memory card, ROM Etc. are used. In addition, each processing unit of the arithmetic processing unit 230 may be realized by hardware by designing a part or all of them, for example, with an integrated circuit.

  Further, the control lines and information lines in the drawings are those that are considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. All the components may be connected to each other.

DESCRIPTION OF SYMBOLS 100: Mirror body 101: Electron source 102: Extraction electrode 103: Electron beam 104: Condenser lens 105: Scanning deflector 106: Sample 107: Sample stage 108: Objective lens 109: Secondary electron 110: Conversion electrode 111: Detector 112 : Minien 113: Load lock chamber 114: Sample chambers 115 and 116: Gate valve 201: SEM main body 210: Controller 220: Arithmetic processor 230: Arithmetic processor 231: Stage controller 232: Matching execution unit 233: Visual field movement control Unit 234: Dimension measurement unit 235: Template automatic generation unit 240: Memory (storage unit)
250: Input / output device

Claims (12)

A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
Equipped with a,
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
The arithmetic processing unit
If the score obtained from the matching process is equal to or greater than a first threshold and the uniqueness of the first template image is equal to or greater than a second threshold, the captured image is selected as the template creation image. A template creation device. The template creation device according to claim 1 ,
The uniqueness includes the first score by the matching process between the first template image and the first captured image, and the matching process between the first template image and the second captured image. A template creation device, characterized in that the ratio is a ratio with the second score. The template creation device according to claim 1 ,
The storage unit stores the score obtained from the matching process in association with the captured image,
The arithmetic processing unit obtains the difference of the scores between captured images of the same or similar pattern in the plurality of captured images,
The said calculation process part selects the said captured image as said template preparation image, when the difference of the said score is below a 3rd threshold value, The template preparation apparatus characterized by the above-mentioned. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
The arithmetic processing unit has a score obtained from the matching process equal to or greater than a first threshold value, and a correlation between the score and a focus difference in the sample observation apparatus before and after the matching process is within a predetermined range. In the case of the above, the captured image is selected as the template creation image. The template creation device according to claim 4 ,
The template creation apparatus characterized in that the focus difference is a focus difference when excitation is applied to an electron optical system of the sample observation apparatus or a focus difference when a voltage is applied to the sample. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
When the score obtained from the matching process is equal to or greater than a first threshold and the detection position obtained by the matching process is within the field of view of the sample observation device, the arithmetic processing unit converts the captured image into the template. A template creation apparatus characterized by being selected as a creation image. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
The sample is a semiconductor wafer;
The storage unit stores the plurality of captured images for each chip on the semiconductor wafer,
When the score obtained from the matching process is equal to or greater than a first threshold, the arithmetic processing unit selects the captured image as the template creation image,
The arithmetic processing unit includes:
By averaging the captured images of all chips of interest, the process of creating one of the second template image with respect to all the chips to be the target,
A process of creating the second template image for each chip by averaging captured images for each chip , and an image obtained by averaging captured images of chips in which an error has occurred in the matching process; A template creation apparatus that executes any one of processes for creating an image obtained by averaging captured images of a chip in which no error has occurred as the second template image. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The said calculation processing part produces the said 2nd template image online, performing the observation of the said sample observation apparatus using the said recipe file information, The template creation apparatus characterized by the above-mentioned. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
After the arithmetic processing unit creates the second template image,
Executing a matching process between the first template image and the captured image;
When the score obtained from the matching process is smaller than a threshold , the second template image is registered online in the recipe file information. The template creation device according to claim 9 ,
The arithmetic processing unit includes:
Register the second template image in the recipe file information for the part on the sample where an error has occurred in the matching process, or
It is possible to set whether to register the second template image in the recipe file information regardless of an error in the matching process. A template creation device for a sample observation device that scans a charged particle beam on a sample to obtain a captured image of the sample,
A storage unit that stores parameter information when observing with the sample observation device and recipe file information including the first template image, and a plurality of captured images acquired from the sample observation device;
An arithmetic processing unit that selects a template creation image from the plurality of captured images and creates a second template image obtained by averaging the template creation images;
With
The arithmetic processing unit includes a matching execution unit that executes a matching process between the first template image and the captured image.
The arithmetic processing unit includes:
If the score obtained from the matching process is greater than or equal to a first threshold, select the captured image as the template creation image;
When the score is smaller than the first threshold, the number of frames of the captured image is increased, and the captured image is acquired again. The template creation device according to claim 11 ,
The template processing apparatus, wherein the arithmetic processing unit increases the number of frames between 8 and 64.

Images