Computer, program, and method

Abstract

Provided is a technology that makes it possible to suitably manage the state of a charged particle beam device. In the present invention, a storage resource stores reference data that includes: imaging field coordinates of a charged particle beam device with respect to a sample which have been obtained by the charged particle beam device or by a simulation; generation time of a reference image of an observation region which corresponds to the imaging field coordinates; and a distribution of reference image quality evaluation values which have been obtained from the reference image by an image quality evaluation function for respective prescribed regions. A computer: (1) acquires an evaluation target image of an observation region which has been obtained by the charged particle beam device with respect to the sample during an evaluation period including evaluation time and which corresponds to the imaging field coordinates; (2) stores, in the storage resource, evaluation target data which includes the imaging field coordinates, the evaluation time, and a first distribution of image quality evaluation values that have been obtained from the evaluation target image by the image quality evaluation function for respective prescribed regions; and (3) compares the reference data and the evaluation target data and uses the image quality evaluation function to calculate the distribution of changes in the image quality evaluation values over time.

Description

Computer, program, and method

[0001] The present disclosure relates to a technology such as a computer for managing the state of a charged particle beam device.

[0002] An example of the prior art is JP 2022-546223 A (Patent Document 1), which states in claim 1, "A method for monitoring performance of a multi-beam system, the method including: receiving records of performance metrics of a plurality of beams of the multi-beam system operating during an imaging process, each record being associated with a beam; determining whether a beam anomaly has occurred based on a baseline value determined using a portion of the records; and providing an indication of the anomaly in response to the determination that the anomaly has occurred."

[0003] Special Publication No. 2022-546223

[0004] Charged particle beam devices have multiple adjustment points that affect the performance metrics of the charged particle beam (i.e., electron beam). Therefore, to keep the charged particle beam device in a normal condition, it is necessary to identify the adjustment points based on changes in the performance metrics.

[0005] For example, Patent Document 1 describes that it is possible to determine whether a beam abnormality has occurred based on a record of performance measurement criteria. However, with the technology described in Patent Document 1, there is room for further study and improvement in estimating the state of the charged particle beam device from changes in the performance measurement criteria. There is also room for further study and improvement in identifying adjustment points, in other words, correction points, for correcting an abnormal beam to a normal state.

[0006] An object of the present disclosure is to provide a technique that can suitably manage the state of a charged particle beam device.

[0007] A representative embodiment of the present disclosure has the following configuration: A computer according to one embodiment includes a processor and a storage resource storing a program, wherein the storage resource stores reference data including: image field coordinates of the charged particle beam device, obtained by a charged particle beam device or simulation, for a sample; a generation time of a reference image in which an observation region corresponding to the image field coordinates is imaged; and a distribution of reference image quality evaluation values ​​obtained from the reference image by an image quality evaluation function for each predetermined region; and the processor executing processing according to the program (1) acquires, during an evaluation period including an evaluation time, an evaluation target image in which an observation region corresponding to the image field coordinates is imaged, obtained by the charged particle beam device, for the sample; (2) stores in the storage resource evaluation target data including the image field coordinates, the evaluation time, and a first distribution of image quality evaluation values ​​obtained from the evaluation target image by the image quality evaluation function for each predetermined region; and (3) compares the reference data with the evaluation target data to calculate a distribution of temporal changes in image quality evaluation values ​​obtained from the image quality evaluation function.

[0008] According to the representative embodiments of the present disclosure, the state of a charged particle beam device can be appropriately managed. Problems, configurations, effects, etc. other than those described above will be described in the detailed description of the invention.

[0009] FIG. 1 is a diagram showing an example of the configuration of an observation system including a computer according to a first embodiment. FIG. 2 is a flow diagram of main processing by the computer according to the first embodiment. FIG. 3 is an explanatory diagram of an image quality evaluation method in an observation region and a predetermined region according to the first embodiment. FIG. 4 is an explanatory diagram showing another example of a predetermined region according to the first embodiment. FIG. 5 is an explanatory diagram of an image quality evaluation function according to the first embodiment. FIG. 6 is an explanatory diagram related to the distribution of a predetermined region with respect to the amount of change over time in image quality evaluation values ​​according to the first embodiment. FIG. 7 is an explanatory diagram related to estimation of an apparatus state using an apparatus state DB and estimation of a correction location using an apparatus correction DB according to the first embodiment. FIG. 8 is an explanatory diagram related to example data of an apparatus state DB and example data of an apparatus correction DB according to the first embodiment. FIG. 9 is a diagram showing an example of a screen including an input field as an example of a user interface according to the first embodiment. FIG. 10 is a diagram showing an example of a screen including an apparatus state monitor field as an example of a user interface according to the first embodiment. FIG. 11 is a diagram showing an example of the configuration of a computer according to the second embodiment. FIG. 12 is an explanatory diagram of learning of an apparatus state estimation engine according to the second embodiment. FIG. 13 is an explanatory diagram of estimation using an apparatus state estimation engine according to the second embodiment. FIG. 10 is an explanatory diagram of estimation using a correction location estimation engine in the second embodiment. FIG. 11 is a diagram showing an example of the configuration of a computer in the third embodiment. FIG. 12 is an explanatory diagram of estimation of an apparatus state using a multivariate distribution model in the third embodiment. FIG. 13 is an explanatory diagram of fitting using a multivariate distribution model in the third embodiment. FIG. 14 is a diagram showing an example of a case where the pattern shapes shown in an evaluation target image and a reference image are different in the fourth embodiment. FIG. 15 is an explanatory diagram of calculation of a distribution using a plurality of evaluation target data and a plurality of reference data in the fourth embodiment. FIG. 16 is a diagram showing another example of setting a plurality of predetermined regions for a plurality of images in the first embodiment. FIG. 17 is a diagram showing another example of displaying a distribution on a screen in the first embodiment. FIG. 18 is a diagram showing an example of scanning a charged particle beam over a plurality of observation regions in the first embodiment.

[0010] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the drawings, the same parts are generally designated by the same reference numerals, and repeated explanations will be omitted. In the drawings, the representation of components may not represent their actual positions, sizes, shapes, ranges, etc., in order to facilitate understanding of the invention.

[0011] For the purpose of explanation, when describing processing by a program, the program, functions, processing units, etc. may be described as the main components, but the main hardware components are the processor, or a controller, device, computer, system, etc. that is configured with the processor, etc. The computer executes processing according to the program read into memory using resources such as memory and communication interfaces as appropriate through the processor. This realizes predetermined functions, processing units, etc. The processor is configured, for example, with semiconductor devices such as a CPU / MPU or GPU. Processing is not limited to software program processing, and can also be implemented using dedicated circuits. Dedicated circuits such as FPGAs, ASICs, and CPLDs can be used.

[0012] The program may be pre-installed as data on the target computer, or may be distributed as data from a program source to the target computer. The program source may be a program distribution server on a communication network, or a non-transitory computer-readable storage medium, such as a memory card or disk. The program may be composed of multiple modules. The computer system may be composed of multiple devices. The computer system may be composed of a client-server system, a cloud computing system, an IoT system, etc. Various data and information may be composed of structures such as tables and lists, for example, but are not limited to these. Expressions such as identification information, identifiers, IDs, names, and numbers are interchangeable.

[0013] [Outline of the embodiment] In this embodiment, a computer and a method for managing the state of a charged particle beam device based on the distribution of image quality evaluation values ​​of images captured by the charged particle beam device are provided. The computer of one embodiment has a processor and a storage resource storing a device management program. The method of one embodiment is a method executed by the computer. The computer acquires images captured by the charged particle beam device or images generated by simulation. The charged particle beam device is, in other words, a microscope or an imaging device. The charged particle beam device is a device that enables imaging, observation, inspection, etc. of a sample.

[0014] The storage resource stores reference data including the image pickup field coordinates of the charged particle beam device obtained by a charged particle beam device or a simulation, an image of an observation area corresponding to the image pickup field coordinates (sometimes referred to as a reference image), the time of generation of the image, and a distribution of image quality evaluation values ​​(sometimes referred to as reference image quality evaluation values) obtained by an image quality evaluation function for each predetermined area from the reference image. The predetermined area is an area that serves as a unit for calculating the image quality evaluation value.

[0015] The processor of the computer executes the following processes in accordance with the device management program.

[0016] (1) The processor acquires an evaluation target image of the sample, which is obtained by the charged particle beam device and shows an observation area corresponding to the imaging field of view coordinates, during an evaluation period including the evaluation time.

[0017] (2) The processor stores in the memory resource evaluation target data including the imaging field of view coordinates of the charged particle beam device, the evaluation time, and a first distribution of image quality evaluation values ​​obtained by the image quality evaluation function for each specified region from the evaluation target image.

[0018] (3) The processor compares the reference data with the evaluation target data to calculate the distribution of temporal changes in the image quality evaluation value according to the image quality evaluation function. In other words, the processor compares the two-dimensional distribution of image quality evaluation values ​​for each predetermined region in the reference data at the generation time with the two-dimensional distribution of image quality evaluation values ​​for each predetermined region in the evaluation target data at the first evaluation time to calculate the temporal changes for each predetermined region, thereby obtaining the two-dimensional distribution of temporal changes in the image quality evaluation value.

[0019] (4) The processor estimates the state of the charged particle beam device from the distribution of the time change of the image quality evaluation value, and determines whether correction of the charged particle beam device is necessary based on the estimated state.If correction is necessary, it determines the areas to be corrected, etc.

[0020] The computer outputs to the user the following data and information regarding the above processing: Reference data Evaluation target data Image quality evaluation value, amount of change over time, distribution Estimated device state, whether correction is necessary, and correction location.

[0021] 1 to 8B , a method and a system according to a first embodiment of the present disclosure will be described. In the first embodiment, a system including a charged particle beam device and a computer will be referred to as an observation system. The method according to the first embodiment is a method executed by the system according to the first embodiment. In the first embodiment, an example will be described in which the state of the charged particle beam device is estimated and corrected based on the distribution of the amount of time change in the image quality evaluation value of an image in a predetermined region.

[0022] [Configuration of Observation System] FIG. 1 shows the configuration of an observation system including a computer according to the first embodiment. The observation system 1 in FIG. 1 includes a charged particle beam device 2 and a computer 3. The charged particle beam device 2 includes a main body 100 and a controller 200. The main body 100 includes components such as a stage serving as a sample holder and a drive circuit in a housing, in other words, a column or sample chamber. The controller 200 is a control system that drives and controls the main body 100 and can be implemented by a computer or circuit. The computer 3 is, in other words, a computer system. The observation system 1 includes components necessary for generating signal waveforms and images based on detection signals from the main body 100 of the charged particle beam device 2. An example of the charged particle beam device 2 is a scanning electron microscope (SEM).

[0023] The main body 100 of the charged particle beam device 2 outputs a detection signal s1 obtained based on the irradiation of the charged particle beam 103 onto a sample 109 on a stage 110. The controller 200 inputs and receives the detection signal s1, processes it, and generates and stores a signal such as an image as a detection signal of the charged particle beam device 2. The controller 200 outputs a signal s2 of the image or the like. The computer 3 inputs and receives the signal s2 of the image or the like and processes it.

[0024] In the main body 100 of the charged particle beam device 2, a charged particle beam 103 extracted from an electron source 101 by an extraction electrode 102 is accelerated by an acceleration electrode (not shown). The accelerated charged particle beam 103 is focused by a condenser lens 104, which is a type of converging lens. The focused charged particle beam 103 is corrected by an astigmatism corrector 105 to cancel out astigmatism generated by an objective lens 108. The charged particle beam 103 is deflected by an image-shift deflector 106 and a scanning deflector 107, thereby scanning one-dimensionally or two-dimensionally the surface of a sample 109 placed on a stage 110, which serves as a sample holder. The charged particle beam 103 is decelerated by a negative voltage applied to an electrode built into the stage 110, which serves as a sample holder, and focused by the lens action of the objective lens 108, and is irradiated onto the sample 109.

[0025] The stage 110 is a mechanism that can move, for example, in the horizontal direction shown in the figure and in the X and Y directions corresponding to the radial direction of the semiconductor wafer. However, the stage 110 may also be a mechanism that can move in the Z direction, which is perpendicular to the X and Y directions corresponding to the vertical direction, or a mechanism that can rotate and tilt in each axial direction. Furthermore, although the details of the implementation of the charged particle beam device 2 are not limited, a configuration that irradiates multiple beams may be used (see FIG. 21 described below). Although the charged particle beam device 2 is illustrated as including a single detector 121, the configuration is not limited thereto and may include multiple detectors. For example, the charged particle beam device 2 may include an SE detector that detects SEs and a BSE detector that detects BSEs. Furthermore, the charged particle beam device 2 may also be configured, for example, with multiple detectors installed at multiple positions. That is, the configuration for capturing an image may include multiple channels and multiple detection systems. Furthermore, when generating an image, a single image may be generated by processing, for example, integrating multiple images based on multiple signals repeatedly detected from the same region.

[0026] The sample 109 is, for example, a semiconductor wafer. When the charged particle beam 103 is irradiated onto the sample 109, electrons 120 such as secondary electrons (SE) and backscattered electrons (BSE) are emitted from inside the sample 109. The emitted electrons 120 are accelerated by an acceleration action based on a negative voltage applied to the sample 109, and are captured by a detector 121. A detection signal s1 output from the detector 121 is sent to a controller 200. The controller 200 receives the detection signal s1 via a communication device 201.

[0027] The controller 200 includes a communication device 201, a processor 202, a memory 203 such as a RAM or a nonvolatile storage device, a display device 204, and an input / output device 205. These components are interconnected via an architecture such as a bus. The controller 200 controls imaging by the main body 100 according to a set imaging recipe. The processor 202 generates an observation image, whose brightness is determined by the amount of electrons captured by the detector 121, based on a detection signal s1 obtained via the communication device 201. The processor 202 stores data such as the generated observation image in the memory 203. The processor 202 sends a signal s2 corresponding to the data such as the generated observation image to the computer 3 via the communication device 201. The computer 3 receives the signal s2 via the communication device 310. Note that a user U1 may use the controller 200 by operating the display device 204 and the input / output device 205.

[0028] The communication device 201 is a device equipped with a communication interface with the main body 100 and a communication interface with the computer 3. The communication interface may be, for example, a LAN, but is not limited to this. The input / output device 205 is an input device and an output device. The input device and the output device may be built-in or may be externally connected. Examples of input devices are a keyboard, a mouse, a microphone, etc. Examples of output devices are a display, a printer, a speaker, etc.

[0029] The computer 3 includes a communication device 310, a storage 320, a processor 330, a memory 331 such as a RAM or a non-volatile storage device, a display device 332, an input / output device 333, etc. These components are interconnected via an architecture such as a bus. The storage 320 is a memory with a relatively large storage area. The storage 320 stores a reference data DB 321, an evaluation target data DB 322, an equipment status DB 323, an equipment correction DB 324, and an equipment management program 325. The memory of the computer 3 also stores other necessary management information and databases (DBs). An external storage device, a server, etc. may be connected to the computer 3, and necessary data and information may be stored in the external storage device, server, etc., and the data and information may be read and written as appropriate.

[0030] The communication device 310 is a device that has a communication interface with the controller 200 implemented therein. An external device may be connected to the computer 3 via a network such as a LAN. The input / output device 333 is an input device and an output device. The input device and the output device may be built-in or may be externally connected. Examples of input devices are a keyboard, a mouse, a microphone, etc. Examples of output devices are a display, a printer, a speaker, etc.

[0031] A user U1 of the observation system 1 operates the display device 332, the input / output device 333, etc. to use the computer 3. In this way, the user U1 uses the observation system 1. The display screen of the display device 332 displays a screen that serves as the user interface of the observation system 1, in other words, a screen with a graphical user interface (GUI). The computer 3 may be configured as a client-server system. In this case, the user U1 operates a PC or the like that serves as a client terminal, and the PC or the like accesses the computer 3 that serves as a server via communication.

[0032] The computer 3 is a part that executes the processing that is characteristic of the first embodiment. The computer 3 is connected to the controller 200. The controller 200 may be configured as a part of the computer 3. The computer 3 may be configured as a part of the controller 200. The controller 200 and the computer 3 may be configured as an integrated computer system. The controller 200 may be configured to execute at least a part of the processing that is characteristic of the first embodiment. The entire system including the computer 3 may be the charged particle beam device 2.

[0033] The processor 330 may be configured with, for example, a CPU, ROM, RAM, etc. The memory 331 appropriately stores various data and information processed by the processor 330. The memory 331 stores, for example, program execution modules, image data, processing result information, etc. The device management program 325 also includes system setting information and user setting information. In addition to the illustrated DB, the storage 320 may also store processing result information, history information, screen data, etc.

[0034] The controller 200 is aware of the imaging field of view coordinate information of the main body 100. The controller 200 is also aware of the generation time, in other words, the image capture time, when an image (e.g., a reference image or an image to be evaluated) is generated based on the detection signal s1. The controller 200 stores data and information such as the imaging field of view coordinate information, the image, and the generation time in association with each other in the memory 203. When transmitting a signal s2 such as an image to the computer 3, the controller 200 transmits the signal s2 to the computer 3 as a series of associated data such as the imaging field of view coordinate information, the image, and the generation time. The computer 3 acquires and stores this series of associated data.

[0035] The imaging field of view coordinate information is information that configures the field of view when imaging is performed by irradiating the surface of the sample 109 on the stage 110 of the main body 100 with the charged particle beam 103. This field of view corresponds to the area of ​​the image to be captured.

[0036] The reference data DB 321 stores data such as the distribution of reference image quality evaluation values. The evaluation target data DB 322 stores data such as the distribution of calculated image quality evaluation values ​​of evaluation targets. The device status DB 323 stores data on the correspondence between the distribution of time-varying image quality evaluation values ​​and the device status. The device correction DB 324 stores data on the correspondence between the device status and correction locations and the correction locations.

[0037] The device management program 325 is a computer program that causes the calculator 3 to execute processing, and is a program that realizes a function of managing the state of the charged particle beam device 2, which is a characteristic function of the observation system 1. The processor 330 reads the device management program 325 from the storage 320 into the memory 331, and executes processing in accordance with the read program. This realizes the characteristic function and corresponding processing.

[0038] The observation system 1 and the computer system 3 are not limited to the configuration example of FIG. 1, and may be any system configured to include one or more processors and one or more memories.

[0039] In addition, when the computer 3 is configured as a client-server system, the operation is, for example, as follows: User U1 accesses the server computer 3 from a client terminal such as a PC via a network. The server provides a screen with a graphical user interface (GUI) to the client terminal. The server transmits GUI screen data (which may be, for example, a web page) for this purpose to the client terminal. The client terminal displays the GUI screen on a display based on the received screen data. User U1 views the GUI screen and inputs instructions, settings, etc. The client terminal transmits the input information to the server. The server executes processing according to the received input information. For example, the server evaluates the captured image and estimates the device status, stores the processing results, and transmits GUI screen data (which may be only update information) for displaying the processing results to the client terminal. The client terminal updates the display of the GUI screen based on the received screen data. User U1 can view the GUI screen to confirm the processing results, such as the device status estimation results.

[0040] 21 is a supplementary explanatory diagram showing an overview of imaging by scanning the charged particle beam 103 when a charged particle beam device 2, which is an SEM, captures multiple images of multiple observation regions 4 of a certain region 5 of a sample 109. First, in the case of an SEM that sequentially scans each observation region 4 with a single charged particle beam 103, for example, at a first time period, the charged particle beam 103 scans the observation region 410, thereby obtaining a detection signal corresponding to the observation region 410. Similarly, at a second time period, the charged particle beam 103 scans the observation region 411, thereby obtaining a detection signal corresponding to the observation region 411. At a third time period, a detection signal corresponding to the observation region 412 is obtained, and at a fourth time period, a detection signal corresponding to the observation region 413 is obtained. Respective images 6 are obtained from the respective detection signals.

[0041] On the other hand, when the charged particle beam device 2 is a multi-beam SEM, the following applies. Although not shown, the main body 100 in FIG. 1 is equipped with a mechanism for generating multiple charged particle beams from a single beam source, or a mechanism for generating multiple charged particle beams from multiple beam sources. The generated multiple charged particle beams are then simultaneously irradiated onto the sample 109. For example, four charged particle beams are simultaneously irradiated onto four observation regions 4 (410 to 413) in FIG. 21 while scanning. As a result, four detection signals corresponding to the four observation regions 4 (410 to 413) are obtained, and respective images 6 are obtained from the respective detection signals.

[0042] [Processing Flow of the Device Management Program] FIG. 2 shows the flow of a series of processes based on the device management program 325 by the computer 3 as the main processes of the observation system 1.

[0043] First, step S1 is a step of evaluating the image quality of the reference image and storing the reference data. In step S1, the main body 100 of the charged particle beam device 2 sets an imaging field of view based on drive control from the controller 200, irradiates the charged particle beam 103 so as to scan an observation area of ​​the sample 109 corresponding to the imaging field of view, and outputs a detection signal s1 captured by the detector 121 to the controller 200. At the same time, the main body 100 sends to the controller 200 imaging field of view coordinates that represent the relative positions of the charged particle beam device 2 and the observation area when the observation area corresponding to the imaging field of view is moved using deflection by the image shift deflector 106 without moving the stage 110. Alternatively, if the controller 200 knows such imaging field of view coordinates based on its own control, it does not need to receive the imaging field of view coordinates from the main body 100.

[0044] Based on the sent detection signal s1, the controller 200 generates, for example, a reference image as an observation image through processing by the processor 202. The controller 200 sends, to the computer 3 via the communication device 201, an image data signal s2 that associates the generated reference image with the imaging field of view coordinates and the generation time.

[0045] Based on the transmitted data signal s2, processor 330 of computer 3 calculates a reference image quality evaluation value for each predetermined region in the reference image. Processor 330 associates the reference image, the generation time, the imaging field of view coordinates, and the distribution of the calculated reference image quality evaluation values, and stores them as reference data in reference data DB 321.

[0046] Step S2 is a step of evaluating the image quality of the group of images to be evaluated and storing the evaluation target data. In step S2, the main body 100 of the charged particle beam device 2, under the control of the controller 200, irradiates different observation regions of the sample 109 corresponding to different imaging fields of view, i.e., each of the multiple observation regions, with the charged particle beam 103 so as to scan the beam, and sends a detection signal s1 captured by the detector 121 to the controller 200. The main body 100 also sends the imaging field of view coordinates of each observation region to the controller 200. Alternatively, if the controller 200 itself knows the imaging field of view coordinates, it does not need to obtain them from the main body 100.

[0047] Based on the detection signal s1 and the like, the controller 200 causes the processor 203 to generate a plurality of evaluation target images as observation images, and stores in the memory 203 an evaluation target image group including the generated plurality of evaluation target images in association with the imaging field of view coordinates corresponding to each evaluation target image and the evaluation target time, which is the time at which each image was generated. The controller 200 then sends a signal s2 containing data on the evaluation target image group, imaging field of view coordinates, and evaluation target time to the computer 3 via the communication device 201. The computer 3 receives this data and stores it in the memory 331. The processor 330 of the computer 3 calculates an image quality evaluation value for each predetermined region of each evaluation target image included in the evaluation target image group. The processor 330 associates each evaluation target image with the distribution of the calculated image quality evaluation values, the imaging field of view coordinates, and the evaluation target time, and stores them in the evaluation target data DB 322 as evaluation target data.

[0048] Step S3 is a step for estimating the state of the charged particle beam device 2. In step S3, the processor 330 of the computer 3 reads, from the evaluation target data DB 322, evaluation target data corresponding to an evaluation period including an evaluation time for estimating the state of the charged particle beam device 2. The processor 330 reads, from the reference data DB 321, reference data corresponding to the imaging field of view coordinates included in the read evaluation target data. The processor 330 compares the read evaluation target data with the reference data for each corresponding predetermined region included in the read evaluation target data to calculate the amount of change over time in the image quality evaluation value. The processor 330 inputs the distribution of the amount of change over time in the image quality evaluation value into the device status DB 323 and outputs an "device status" representing the state of the charged particle beam device 2 at the time the evaluation target image was generated.

[0049] Step S4 is a step for correcting the charged particle beam device 2. In step S4, the device status output in step S3 is input to the device correction DB 324, and correction points for the charged particle beam device 2 that reduce the amount of change over time in the image quality evaluation value relative to the reference data are output.

[0050] In step S5, the computer 3 checks whether the processing has ended, and if it has not ended and continues to monitor and evaluate the device status of the charged particle beam device 2 (NO), it returns to step S2 and repeats the processing in the same way, thereby continuing to monitor and evaluate the device status of the charged particle beam device 2.

[0051] In the following, from Figure 3 onwards, we will explain specific examples of estimating and correcting the device state of the charged particle beam device 2, assuming that the observation system 1 manages the device state of the charged particle beam device 2 while observing a semiconductor wafer as the sample 109.

[0052] [Method for Evaluating Image Quality of Observed Images] FIG. 3 is an explanatory diagram of a method for evaluating the image quality of an observed image in an observation region on the surface of a semiconductor wafer, which is the sample 109. The charged particle beam device 2 in FIG. 1 acquires multiple images 6 of multiple observation regions 4 in the observation region 5, which is a region 5 of the sample 109, by moving the scanning region of the charged particle beam 103, which corresponds to the imaging field of view, using an image shift deflector 106. The images 6 are, for example, images to be evaluated. For example, a group of images 6 to be evaluated is acquired. One imaging field of view corresponds to one observation region 4. One observation region 4 corresponds to one image 6. In the example of FIG. 3, the four observation regions 4 are a first observation region 410, a second observation region 411, a third observation region 412, and a fourth observation region 413. The group of images 6 to be evaluated may include one or more images 6. A center 400 indicates the center point of the observation region 5 in the x and y directions.

[0053] In the first embodiment, the multiple images 6 of the evaluation target image group 60 are acquired sequentially in time based on a single charged particle beam 103 ( FIG. 21 ), but this is not limiting. If the charged particle beam device 2 is a multi-beam device, the multiple images 6 of the evaluation target image group 60 may be acquired simultaneously by simultaneously scanning and irradiating the sample 109 with multiple charged particle beams.

[0054] The calculator 3 divides each observation region 4 corresponding to each image 6 into predetermined regions 7. In this example, the size of the predetermined regions 7 is smaller than the size of the observation region 4. In the example of Fig. 3, one observation region 4 is divided into multiple regions (e.g., three regions) in the horizontal direction, which is the x direction, and into multiple regions (e.g., three regions) in the vertical direction, which is the y direction. The shape of the predetermined regions 7 is a square of the same size in both the x and y directions.

[0055] The computer 3 divides the observation area 4 of the image 6 into a plurality of predetermined areas 7. The predetermined areas 7 are areas that serve as units for calculating image quality evaluation values. That is, one image quality evaluation value is calculated for one predetermined area 7. In order to obtain the distribution of image quality evaluation values ​​in two-dimensional space (i.e., the xy plane), it is essential to construct a distribution of the plurality of predetermined areas 7 in two-dimensional space.

[0056] The calculator 3 inputs each predetermined region 7 of each image 6 into the image quality evaluation function 430 to obtain an image quality evaluation value 440 for each predetermined region 7. An image quality evaluation result 450 shows an example of the image quality evaluation value 440 for the observation region 5.

[0057] Calculator 3 associates image quality evaluation value 440 of image quality evaluation result 450 with evaluation target image 6, which is the corresponding image 6, the imaging field of view coordinates, and the evaluation target time, which is the time at which image 6 was generated, and stores the data as evaluation target data in evaluation target data DB 322. Alternatively, Calculator 3 may store the data as reference data in reference data DB 321. Calculator 3 may store the data in at least one of the DBs. Calculator 3 may use image 6 as the reference image and store reference data including the reference image, or may use image 6 as the evaluation target image and store evaluation target data including the evaluation target image.

[0058] The imaging field of view coordinates are values ​​that represent the relative positions of the charged particle beam device 2 and the observation region when an observation image is acquired without moving the stage 110. The computer 3 may store evaluation target data acquired during operation of the observation system 1 as reference data in the reference data DB 321. When the above data is stored as reference data in the reference data DB 321, the image quality evaluation value 440 is not limited to calculation using an observation image generated from a signal acquired by the charged particle beam device 2, but may also be an image quality evaluation value obtained by simulation. Acquisition of an observation image refers to the controller 200 generating an observation image based on the signal s1 output by the detector 121, or the computer 3 receiving an observation image from the controller 200.

[0059] One observation region 4 corresponds to one imaging field of view. Depending on the configuration of the charged particle beam device 2, the imaging field of view and the observation region 4 can be moved by controlling the movement of the stage 110, or the imaging field of view and the observation region 4 can be moved by deflecting the charged particle beam 103 without moving the stage 110. Either method is acceptable.

[0060] In the first embodiment, a plurality of images 6 corresponding to a plurality of observation regions 4 are evaluated in order to cover a wide area on the surface of the sample 109. In other words, this is to cover a necessary and sufficient area to estimate the state of a component part (e.g., a beam irradiation system) of the charged particle beam device 2. As a modified example, only one observation image may be evaluated.

[0061] [Predetermined Region] FIG. 4 shows another example of the shape of the predetermined region 7, in other words, the divided region or image quality evaluation value calculation region. The shape of the observation region 4 when divided into the predetermined regions 7 is arbitrary. In the example of FIG. 3, the shape of the predetermined region 7 is square, in other words, rectangular, or lattice-shaped, but is not limited to this and may be, for example, as shown in the example of FIG. 4. In the example of FIG. 4, the shape of the predetermined region 7 is a vertical stripe shape in which the size in the vertical direction (y direction) is longer. In the example of FIG. 4, one observation region 4 is divided into multiple regions (e.g., four) in the x direction, but is not divided in the y direction. Other examples of the shape of the predetermined region 7 include a circle and a polygon. Various shapes can be set as long as a two-dimensional distribution of the predetermined region 7 can be obtained.

[0062] The number, size, and shape of the predetermined regions 7 may be specified according to the type of image quality evaluation value, in other words, the parameters. A shape suitable for calculation may be specified according to the image quality evaluation value. In other words, a suitable predetermined region 7 may be selected according to each image quality evaluation value. The method of dividing and setting the predetermined regions 7, i.e., the shape, size, number, etc. of the predetermined regions 7, may differ for each image quality evaluation value (for example, brightness, contrast, amount of distortion, and amount of blur in FIG. 5).

[0063] [Image Quality Evaluation Function] FIG. 5 is an explanatory diagram of the image quality evaluation function 430. Calculator 3 inputs an evaluation target image 501 (evaluation target image: I) to be subjected to image quality evaluation into image quality evaluation function 430. Evaluation target image 501 has the number of pixels px in the x direction and the number of pixels py in the y direction, and corresponds to observation region 4 or predetermined region 7 in FIG. 3. In this case, image quality evaluation function 430 calculates an image quality evaluation value for each image in each predetermined region 7 in FIG. 3. Calculator 3 uses at least one of the four functions included in image quality evaluation function 430, namely, brightness evaluation function 431, contrast evaluation function 432, distortion amount evaluation function 433, and blur amount evaluation function 434, to output image quality evaluation value 440. Image quality evaluation value 440 is at least one of brightness, contrast, distortion amount, and blur amount.

[0064] The brightness evaluation function 431 is a function that calculates the average brightness value of the evaluation target image. The contrast evaluation function 432 is a function that calculates the ratio between the difference between the maximum and minimum brightness values ​​of the evaluation target image and the sum of the maximum and minimum brightness values. The distortion amount evaluation function 433 is a function that calculates the distortion intensity at which the difference between the evaluation target image and an image obtained by adding distortion to a reference observation image containing the same observation area as the evaluation target image while changing the distortion intensity, as the distortion amount relative to the reference observation image. The blur amount evaluation function 434 is a function that calculates the blur intensity at which the difference between the evaluation target image and a blurred image created by changing the blur intensity of a process that blurs an image, such as Gaussian blur, is smallest, after a sharp image created by performing blur removal processing on the evaluation target image.

[0065] The calculation method for the four functions 431, 432, 433, and 434 included in the image quality evaluation function 430 is an example. This is not limiting, and other calculation methods may be used. Image quality evaluation function 430 may also be used to calculate a reference image quality evaluation value to be saved as reference data.

[0066] The blur evaluation function 434 may be a function that calculates the amount of blur in each of the horizontal direction (e.g., the x direction) and the vertical direction (e.g., the y direction) of the image. For example, in the charged particle beam 103 that passes off-axis of the objective lens 108 in FIG. 1 , astigmatism, which is a type of off-axis aberration in which the image formation distance differs between the horizontal and vertical directions, occurs, and the observed image becomes an image with different amounts of blur in each of the horizontal and vertical directions. By using the blur evaluation function 434 to calculate the amount of blur in each direction and comparing the amounts of blur in the two directions, the horizontal and vertical directions, it is possible to evaluate the amount of astigmatism and the direction of astigmatism.

[0067] [Distribution of Image Quality Evaluation Values ​​over Time in a Predetermined Region] FIG. 6 is an explanatory diagram of the distribution of image quality evaluation values ​​over time in a predetermined region. Calculator 3 references evaluation target data 601 generated during an evaluation period including an evaluation time for estimating the device state from evaluation target data DB 322, in other words, a distribution 8A of image quality evaluation values ​​over a predetermined region for the evaluation target image. Calculator 3 references reference data DB 321 for reference data 602 that uses the same division method for predetermined regions 7 as the referenced evaluation target data 601 and includes image quality evaluation values ​​used as standards for the device state, in other words, a distribution 8B of image quality evaluation values ​​over a predetermined region for the reference image. Calculator 3 compares the distribution 8A of image quality evaluation values ​​of the referenced evaluation target data 601 with the distribution 8B of reference image quality evaluation values ​​of the referenced reference data 602 to calculate the difference in process 610, thereby obtaining a distribution 9 of image quality evaluation values ​​over time in a predetermined region. Distribution 9 contains a value for the amount of change in image quality evaluation values ​​over time for each predetermined region 7. The calculator 3 compares the image quality evaluation value of a predetermined area 7 of the reference image with the image quality evaluation value of a predetermined area 7 of the image to be evaluated, which corresponds in position and size to the reference image, and calculates the amount of change in the image quality evaluation value over time, for example by performing a difference calculation.

[0068] In this example, in distribution 9, among the regions corresponding to observation region 5 in Figure 3, the value of the amount of change over time in the image quality evaluation value of each predetermined region 7 is 0.1 in observation regions 4B, 4C, and 4D at the top right, bottom left, and bottom right. On the other hand, in observation region 4A at the top left, the value of the amount of change over time in the image quality evaluation value of predetermined region 7 is 0.3 in the central predetermined region 7 of the 3 × 3, and 0.2 in the other predetermined regions 7.

[0069] The amount of change over time refers to the amount of change corresponding to the time difference between the generation time when the reference data was generated in the past and the evaluation time when the evaluation target data was generated. Since the amount of change over time in the image quality evaluation value is obtained for each predetermined region 7, a distribution of the amount of change over time in the image quality evaluation value can be obtained when viewed on a two-dimensional xy plane. During this period of change over time, the state of the charged particle beam device 2 can change due to various factors. Examples include misalignment of component parts, changes in electrical characteristics, and deterioration of components due to use time or vibration.

[0070] 7A and 7B are explanatory diagrams of estimation of the apparatus state using the apparatus state DB 323 from the distribution 9. Fig. 7A shows that the apparatus state DB 323 is referenced from the distribution 9 to obtain an output 701 of the apparatus state and whether correction is required, and the apparatus correction DB 324 is referenced from the output 701 to obtain an output 702 of the correction location. Fig. 7B shows an example of data 703 contained in the apparatus state DB 323 and an example of data 704 contained in the apparatus correction DB 324.

[0071] As shown in Fig. 7B, device status DB 323 has separate pages of data 703 for each of functions 431 to 434 included in image quality evaluation function 430, or for each combination of these functions. Each page stores, in association with one another, a distribution 731 of the amount of change in image quality evaluation value over time in a specified region, a label value 732 indicating the device status, and whether or not correction of the device status is required 733. Although Fig. 7B shows only the page for blur amount evaluation function 434, similar pages exist for the other functions.

[0072] Calculator 3 compares distribution 9 in a predetermined region for the amount of change over time in the image quality evaluation value with distribution 731 stored on the page of the corresponding image quality evaluation function in device status DB 423, thereby searching for distribution 731 with the smallest difference. Calculator 3 outputs label value 732 indicating the device status and correction necessity 733 associated with distribution 731 with the smallest difference as a result of the search. In the example of FIG. 7A , output 701 has a label value of 2 indicating the device status, indicating whether correction is necessary. If the output indication of whether correction is necessary is yes, Calculator 3 inputs the label value indicating the device status into device correction DB 324 and obtains, as an output, the correction location for reducing the amount of change over time in the image quality evaluation value.

[0073] 7B , apparatus correction DB 324 has a separate page for each of functions 431 to 434 included in image quality evaluation function 430, or for each combination of these functions. Each page stores a label value 741 indicating the apparatus state and a correction location 742 in association with each other. For example, the correction locations corresponding to an apparatus state with label value = 2 include the objective lens and the astigmatism corrector.

[0074] The device status DB 323 stores, as data / information, the correspondence between a distribution 731 of the amount of change in image quality evaluation value over time and a device status 732, etc. Therefore, for the input distribution 9, information such as the device status is obtained as output 701 from the device status DB 323. The device correction DB 324 stores, as data / information, the correspondence between a device status 741 and correction locations 742. Therefore, for the input information such as the device status, information such as the correction locations is obtained as output 702 from the correction DB 324. In the device correction DB 704 in FIG. 7B , correction locations for reducing the amount of change in image quality evaluation value over time, in other words, the difference between the reference image quality evaluation value and the image quality evaluation value of the evaluation target, or in other words, prioritized correction candidate locations, are described as correction locations 742 associated with the device status 741 and the associated distribution 731.

[0075] For example, if the blur amount for each direction described above is calculated as the image quality evaluation value using the blur amount evaluation function 434 in FIG. 5 , it is possible to estimate the change in the blur amount for each direction over time for each predetermined region 7. By comparing the blur amounts for each direction, Calculator 3 can determine the difference in the imaging distance between the horizontal and vertical directions of the objective lens 108 in FIG. 1 . Therefore, Calculator 3 can correct the parameters of the objective lens 108 in FIG. 1 or the parameters of the astigmatism corrector 105 so as to reduce the change over time from the time the reference data was generated. Calculator 3 may determine the amount of correction depending on the time difference between the time the reference data was generated and the time the evaluation target data was evaluated.

[0076] 7A and 7B are, in other words, classification values ​​when the distribution 9 of the amount of change over time of the image quality evaluation value and the state of the distribution 9 are understood as mathematical patterns, etc. In this example, the device state label values ​​are numbers such as 1, 2, and 3, but these are simply identification information and can be arbitrary, and do not represent the degree of change, etc.

[0077] The details of the device status DB 323 and the correction DB 324 are not limited, but as an implementation example, the designer / operator of the device management program 325 of the observation system 1 may set the contents of these DBs. Although the device status DB 323 and the device correction DB 324 are provided separately, this is not limiting, and they may be integrated into a single DB. For example, when the device status DB data 703 of FIG. 7A and the device correction DB data 704 of FIG. 7B are integrated into a single DB, the single DB stores the association between distribution, device status, whether correction is required, and the correction location.

[0078] 7B , Calculator 3 compares distribution 9 of a predetermined region for the amount of change in image quality evaluation value over time with each distribution 731 in device status DB 323, and searches for and extracts distribution 731 with the smallest difference as the closest distribution. Specific examples of how to calculate this difference, or in other words, the distribution difference, are as follows: Calculator 3 calculates the difference in value between each of the predetermined regions 7 that correspond in position between the distributions, and calculates the sum of the differences in value between each of the multiple predetermined regions 7. Calculator 3 determines that distribution 421 with the smallest sum is the closest distribution.

[0079] 8A and 8B show examples of a user interface, particularly a GUI screen, displayed on the display screen of the display device 204 of the controller 200 or the display device 332 of the computer 3 in FIG. 1. Fig. 8A shows the first half of the screen, and Fig. 8B shows the second half of the screen. This GUI screen may be provided, for example, as an application window or as a web page.

[0080] 8A , this GUI screen has an input field 801. The input field 801 is a field for inputting parameters for executing the device management program 325 in response to input operations by user U1 using the input / output device 205 or the input / output device 333 shown in FIG. 1 . The input field 801 also has an image quality evaluation function check box 820, a reference data DB file path input field 821, an evaluation target data DB file path input field 822, a device status DB file path input field 823, and a device correction DB file path input field 824. The image quality evaluation function check box 820 is a check box for selecting a function to be used from among the functions included in the image quality evaluation function 430. The reference data DB file path input field 821 is an input field for a file path indicating the location of the reference data DB 321 on storage. Similarly, the evaluation target data DB file path input field 822 is an input field for the file path of the evaluation data DB322, the device status DB file path input field 823 is an input field for the file path of the device status DB323, and the device correction DB file path input field 824 is an input field for the file path of the device correction DB324.

[0081] The input field 801 also includes a predetermined area setting field 810. In the predetermined area setting field 810, the aforementioned predetermined area 7 can be set for the observation area of ​​the image. In other words, the division method for the aforementioned predetermined area 7 can be set. This setting can be selected between manual setting by the user U1 and loading a setting file. In manual setting, the user U1 can select a shape and set the predetermined area 7 using the selected shape. For example, the user U1 can draw a line or enclose the area in a rectangle. Alternatively, although not shown, the user U1 may be able to input the vertical and horizontal widths and number of divisions of the predetermined area 7. The user U1 can set an appropriate predetermined area 7, for example, depending on the image quality evaluation value, including the examples in FIGS. 3 and 4 . In this way, the user can set the predetermined area 7 on the GUI screen for each image quality evaluation value, in other words, for each corresponding image quality evaluation function. It is also possible to set and apply a common predetermined area 7 regardless of the image quality evaluation value.

[0082] 8B, this GUI screen includes an apparatus status monitor field 802 that displays the execution results of the apparatus management program. The apparatus status monitor field 802 has an evaluation target image list field 803 and an apparatus status field 804. The evaluation target image list field 803 displays a list of evaluation target images 831 acquired by the charged particle beam apparatus 2, as well as file names 832 and generation times 833 of the images.

[0083] The device status column 804 displays a distribution 841 of a predetermined region for the amount of change over time, a generation time 842, a device status label value 843, whether correction is required 844, and a correction location 845 for each item (e.g., brightness, contrast, distortion, blur) selected in the image quality evaluation function check box 820 included in the input column 801. In this example, contrast and blur have been selected, so information is displayed in the contrast column 804A and the blur column 804B.

[0084] As a modified example, the display of the distribution 841 is not limited to numerical values, but may be expressed in a different way, such as by replacing numerical values ​​with colors. The device status label value 843 is not limited to numerical values, but may be displayed in natural language, for example. The correction necessity 844 is not limited to a binary value of "necessary" but may be a multi-stage value, and may display the time when correction should be made, the time when correction is recommended, etc. The correction location 845 is not limited to the name of a component part of the charged particle beam device 2, but may be displayed graphically, such as a configuration diagram, to show the location or position. Furthermore, a link may be attached to the correction location, so that when user U1 operates the link, the screen changes to a setting screen for the correction location, allowing detailed setting work for the correction location to be performed.

[0085] 20 shows an example of a modified example in which numerical values ​​are replaced with colors when displaying a distribution 841 on a GUI screen. In this distribution 841, the greater the amount of change over time in the image quality evaluation value, the darker the color displayed.

[0086] [Effects of the First Embodiment] As described above, the first embodiment allows the state of the charged particle beam device 2 to be appropriately managed. In particular, the user U1 can check on the screen the distribution of the amount of change over time in the image quality evaluation value, and whether or not correction is necessary depending on the state of the device, and the location of the correction. This allows the user U1 to perform correction work on the charged particle beam device 2 more easily and appropriately than in the past.

[0087] As an example, when the image quality evaluation function and parameter are blur, as shown in Figures 8A and 8B, a distribution 510 of the time change in the image quality evaluation value over a predetermined region is obtained, similar to that shown in Figure 6, and is displayed as distribution 841 in the blur column 804B. The device status label value is 2, the correction requirement is displayed as "needed," and the correction locations are displayed as the objective lens and astigmatism corrector. By viewing these, user U1 can see that the state of the charged particle beam device 2 requires correction, and specifically, that work should be performed to prioritize correction of the objective lens 108 and astigmatism corrector 105 shown in Figure 1. Furthermore, distribution 841 (similar to distribution 9 in Figure 6) shows that the change in blur is relatively large in the upper left region of the image. This corresponds to the possibility that, for example, the irradiation axis of the charged particle beam 103 is off-center. Therefore, it can be seen that the objective lens 108 and the astigmatism corrector 105 can be corrected so that the irradiation axis of the charged particle beam 103 does not deviate from the center by checking and adjusting them. This makes it possible to simplify and streamline the management and correction work by the user U1 to maintain the charged particle beam device 2 in a favorable state.

[0088] In the first embodiment, as shown in FIGS. 8A and 8B , a distribution 841, which is the distribution 9 of a predetermined region for the amount of change in image quality evaluation value over time in FIG. 6 , is output to the user U1. The user U1 can view and confirm the distribution 841 on the GUI screen. The GUI screen also outputs an apparatus status 843, a need for correction 844, and a correction location 845. The user U1 can view and confirm this information on the GUI screen. Based on this information, in other words, the support information, the user U1 can check the status of the charged particle beam apparatus 2 and perform correction work more easily and efficiently than before.

[0089] A feature of the first embodiment is that the state of the charged particle beam device 2 and correction locations are estimated from the aspect of the distribution 9 of the amount of change over time of the image quality evaluation value, in other words, the two-dimensional distribution corresponding to the predetermined region 7. The inventors have focused on the fact that there is a correspondence between the image quality evaluation value, the aspect of the distribution 9 (whether it is uniform or not uniform and where the center of change is biased), the device state, correction locations, etc., and have devised the above-mentioned new mechanism that estimates the device state, etc., by utilizing this correspondence.

[0090] The following modifications of the first embodiment are also possible. In the first embodiment, the reference image is stored in association with the reference data DB 321, and the evaluation target image is stored in association with the evaluation target data DB 322, and the images can be confirmed on the screen as needed. However, the present invention is not limited to this. At least data related to the image quality evaluation value needs to be stored, and the reference image and the evaluation target image may not be stored.

[0091] In the first embodiment, as described above, data related to the criteria is registered in the reference data DB 321, and data related to the evaluation target is registered in the evaluation target DB 322. However, it is also possible to register data related to the evaluation target that has already been evaluated in the reference data DB 321 so as to use it as new reference data. In other words, the registered data in the evaluation target DB 322 may be copied / migrated to the reference data DB 321 as reference data.

[0092] The reference image quality evaluation value of the reference image of the reference data is not limited to being generated based on an image captured by the charged particle beam device 2, and may be generated based on a simulation without capturing an image. For example, the computer 3 or any device in the observation system 1 may generate an ideal image quality evaluation value by simulation calculation or the like based on design data of the sample 109 and use the generated image quality evaluation value as the reference data. For example, the computer 3 may refer to the design data of the sample 109 from a manufacturing execution system (MES), server, or the like connected to the observation system 1, and generate an ideal image quality evaluation value from the design data by simulation calculation.

[0093] In the first embodiment, as described above, the predetermined area 7 for the observed image, in other words, the divided area or the area for calculating the image quality evaluation value, can be set in the predetermined area setting field 810 on the GUI screen. However, the present invention is not limited to this, and the predetermined area 7 corresponding to the image quality evaluation value or the like may be defined in advance as a design item in the system, device, method, device management program 325, etc., related to this embodiment.

[0094] The computer 3 also manages data on the method of dividing the image into the predetermined area 7, in other words, the method of setting the predetermined area 7. For example, the computer 3 may store data / information such as the method of dividing the image into the predetermined area 7 in association with the image or DB. Data / information such as the method of dividing the predetermined area 7 may be added as attribute information / metadata of the image. In the example of FIG. 3, an example of the method of dividing the predetermined area 7 may be to set the basic shape to a rectangle and specify the size of the rectangle in the x direction and the size of the rectangle in the y direction. Alternatively, the number of divisions in the x direction and the number of divisions in the y direction of the observation area 4 may be specified.

[0095] In the first embodiment, the computer 3 outputs up to the correction point 845 to the user U1 on the GUI screen. As a modified example, the computer 3 may automatically determine correction content information corresponding to the correction point 845 and output it on the GUI screen. For example, if the correction point 845 is the objective lens 108 in FIG. 1 , the correction content information is information regarding which parameters of the objective lens 108 should be corrected and how. For example, the correction content information is information indicating the amount of correction to be applied to the optical axis position as a parameter of the objective lens 108, in which direction and at what distance on the x-y plane. The correction content information may be correction instruction information or reference information serving as a guide for correction. In the example of distribution 9 in FIG. 6 or distribution 841 in the blur amount column 804B in FIG. 8B, since the value in the upper left region is relatively large overall, the correction amount may be adjusted to adjust the optical axis position of the objective lens 108 toward the lower right.

[0096] In the first embodiment, the explanation is based on the premise that the area of ​​the reference image to be compared and the area of ​​the image to be evaluated are the same, that is, the image capturing field coordinates, size, etc. of the images are the same, and the pattern structure captured in the observation area is the same, as shown in Figure 6. However, the present invention is not limited to this, and the features explained in the first embodiment can also be applied to cases where the area of ​​the reference image to be compared and the area of ​​the image to be evaluated are different, in other words, when there is a deviation in the pattern structure captured in the observation area. This will be described in the fourth embodiment below.

[0097] In the first embodiment, the description is based on the premise that the predetermined region 7 is smaller in size than the observation region 4 corresponding to the image 6, in other words, a single observation region 4 contains multiple predetermined regions 7, but this is not limiting. In a modified example, the predetermined region 7 may be larger than or equal to the observation region 4. Also, adjacent predetermined regions 7 may have overlapping portions. Even in such a case, it is sufficient to obtain the distribution of multiple predetermined regions 7 in two-dimensional space.

[0098] Fig. 19 shows an example of a predetermined region 7 in a modified example. In the example of Fig. 19 (1), there are 16 observation regions 4 in a 4 x 4 matrix in the image area 5, and there are corresponding images. In this example, one predetermined region 7 is set for four observation regions 4. The vertical and horizontal dimensions of the predetermined region 7 are larger than the vertical and horizontal dimensions of the observation region 4. This results in a distribution of four predetermined regions 7 in a 2 x 2 matrix.

[0099] 19(2), nine circular predetermined regions 7 are set for 16 observation regions 4 (4×4) in the image region 5. The circular predetermined regions 7 are set at the center of the image region 5, four positions shifted up, down, left, and right in the x-y directions from the center position, and four positions shifted to the upper left, upper right, lower left, and lower right from the center position. Adjacent predetermined regions 7 partially overlap each other.

[0100] Furthermore, after the correction points are presented as needing correction, the user U1 may manually correct and adjust the component parts, but this is not a limitation. If the charged particle beam device 2 has a function for automatically adjusting component parts, information such as the correction points may be transmitted from the computer 3 to the controller 200, and the controller 200 may automatically adjust the component parts using the automatic adjustment function based on the information. The user U1 checks the execution and results of the automatic adjustment on the GUI screen. As an example of the automatic adjustment function, in the case of adjusting the objective lens 108, the position of the objective lens 108 can be adjusted by controlling a drive circuit connected to the objective lens 108.

[0101] <Embodiment 2> Embodiment 2 will be described using Figure 9 and subsequent figures. The basic configuration of embodiment 2 etc. is the same as or common to embodiment 1, and the following description will mainly focus on the components of embodiment 2 etc. that are different from embodiment 1.

[0102] In the computer according to the second embodiment, a machine learning model is used to estimate the device state of the charged particle beam device 2, instead of the device state DB 323 described above. A specific example will be described below.

[0103] [Computer] Figure 9 shows the functional block configuration of the computer 3 of the observation system 1 in embodiment 2. Figure 9 differs from Figure 1 in that the storage 320 does not include the device status DB 323 and the device correction DB 324, but instead includes a learning model DB 360. The learning model DB 360 is a DB that stores learned parameters and the like that constitute a machine learning learning model. A specific example of a learning model is a convolutional neural network (CNN). The processor 330 also includes an equipment status estimation engine 350 as a function or processing unit implemented based on the processing of the equipment management program 325. The equipment status estimation engine 350 is a part that learns the learning model in the learning model DB 360 and performs the processing of evaluating and estimating the equipment status using the learning model. In other words, the equipment status estimation engine 350 is an equipment status estimation unit. The equipment management program 325 in embodiment 2 has implementations corresponding to the functions in embodiment 2.

[0104] Note that the implementation of the equipment state estimation engine 350 is not limited to the configuration shown in FIG. 9 . For example, in a configuration example of the observation system 1 according to the second embodiment, the equipment state estimation engine 350 and the learning model DB 360 may be implemented in an external computer system separate from the computer 3. For example, a company other than the company of the charged particle beam device 2 and the computer 3 may provide a machine learning service on cloud computing. The computer 3 may communicate with a server or the like provided by that company and use the equipment state estimation engine 350 or the like.

[0105] In the second embodiment, the device status DB 323 in Fig. 7A of the first embodiment is replaced with a device status estimation engine 350. Furthermore, in a modification of the second embodiment, the device status DB 323 and the device correction DB 324 in Fig. 7A of the first embodiment are replaced with a correction part estimation engine 370.

[0106] 10 is an explanatory diagram of learning by the apparatus state estimation engine 350 in the learning phase. The calculator 3 extracts at least one distribution 9 (the distribution of a predetermined region for the amount of time change in the image quality evaluation value described above) stored in the reference data DB 321 or the evaluation data DB 322, and creates a learning data set 1010. The learning data set 1010 is data for learning.

[0107] The device state estimation engine 350 receives each distribution of the learning data set 1010 as input, executes processing based on the internal parameters of the learning model, and outputs, as a processing result, an estimated value 1001. The output estimated value 1001 has the probability that the distribution 9 corresponds to each of the label values ​​indicating the device state described above, i.e., a probability value of the device state, and a probability value of whether correction is necessary.

[0108] The device state estimation engine 350 (and its corresponding learning model) learns the association between the distribution 9, the device state, and whether correction is required. To this end, the device state estimation engine 350 references attribute information 1002 associated with each distribution 9 in the learning data set 1010 from the reference data DB 321 or the evaluation data DB 322. The attribute information 1002 has a label value indicating the device state and whether correction is required. The device state estimation engine 350 updates the internal parameters of the learning model to reduce the error between the attribute information 1002 and the estimated value 2001. The updated internal parameters are stored as learned parameters in the learning model DB 360. The above-described processing is performed in the learning phase of the learning model.

[0109] FIG. 11 is an explanatory diagram of the estimation phase. The computer 3 uses the device state estimation engine 350 to estimate the device state and whether correction is required for evaluation target data 1110 including an evaluation target image referenced from the evaluation target data DB 322, based on the distribution 9 of a predetermined region for the amount of time change in image quality evaluation values. The device state estimation engine 350 reads and applies corresponding internal parameters from the learning model DB 360 so that the image quality evaluation function 430 used to calculate the distribution 9 is the same during learning and during device state estimation. The device state estimation engine 350 inputs the distribution 9 of the evaluation target data 1110 contained in the evaluation data DB 322, performs processing based on the internal parameters, and outputs the device state and whether correction is required at the time the evaluation target data 1110 was generated as an estimated value 1101 of the processing result.

[0110] In a modified example, the correction portion estimation is further realized by machine learning. For this purpose, a correction portion estimation engine 370 is used.

[0111] FIG. 12 is an explanatory diagram of the learning of the correction location estimation engine 370, which estimates correction locations from a distribution 9 of the time change amount of the image quality evaluation value in a predetermined region in a modified example. The computer 3 extracts one or more distributions 9 stored in the reference data DB 321 or the evaluation data DB 322 and creates a learning data set 1210. The correction location estimation engine 370 inputs each distribution 9 in the learning data set 1210, executes processing based on the internal parameters of the learning model, and outputs an estimated value 1201 as the processing result. The output estimated value 1201 has a probability value of the correction location of the charged particle beam device 2 corresponding to the distribution 9. If there are multiple correction locations, it has a probability value for each correction location. The correction location estimation engine 370 (and its corresponding learning model) learns the association between the distribution 9 and the correction location. To do this, the correction location estimation engine 370 references attribute information 1202 from the reference data DB 321, the evaluation target data DB 322, and the device correction DB 324. The attribute information 1202 includes correction points. The correction point estimation engine 370 updates the internal parameters of the learning model so as to reduce the error between the attribute information 1202 and the estimated value 1201. The updated internal parameters are stored as learned parameters in the learning model DB 360. The above-described processing is performed in the learning phase of the learning model.

[0112] Although not shown in FIG. 9, the device correction DB 324 in FIG. 12 is used during learning and is not used during estimation, and during estimation, the correction part estimation engine 370 performs the estimation.

[0113] Fig. 13 is an explanatory diagram of the estimation phase in a modified example. Fig. 13 particularly shows a case where a correction portion is estimated using a correction portion estimation engine 370 in combination with estimation by the apparatus state estimation engine 350 in Fig. 11, but the correction portion estimation engine 370 may be used alone.

[0114] [Effects of Second Embodiment, etc.] As described above, the second embodiment makes it possible to suitably manage the state of the charged particle beam device 2. In particular, the second embodiment uses machine learning to improve the accuracy of estimating the correspondence between the distribution 9 of the amount of change over time in the image quality evaluation value of an image and the device state, etc.

[0115] <Third Embodiment> A third embodiment will be described. In the third embodiment, the device state of the charged particle beam device 2 is estimated by fitting. In the third embodiment, an example will be described in which fitting of a multivariate distribution model is used to estimate the device state of the charged particle beam device 2, instead of the device state DB 323 described above.

[0116] Fig. 14 shows an example configuration of the computer 3 in the third embodiment. In the third embodiment, the device status DB 323 in Fig. 7A of the first embodiment is replaced with a multivariate distribution model, which is a mixture distribution model. The computer 3 has a multivariate distribution model fitting processing unit 380 as a processing unit realized by processing of the device management program 325 by the processor 330. The storage 320 stores a multivariate distribution model fitting result 390 and the like.

[0117] [Estimation of Device State Using a Multivariate Distribution Model] Figure 15 is an explanatory diagram of estimation of device state using a multivariate distribution model. The computer 3 performs fitting processing 1510 (see Figure 16, described below) with a multivariate distribution model, which is a mixed distribution model, from the distribution 9 of a predetermined region for the amount of time change in the image quality evaluation value of the evaluation target data using the multivariate distribution model fitting processing unit 380, obtains parameter search results as fitting results 1520, and stores them in memory (the multivariate distribution model fitting results 390 in Figure 14). The computer 3 outputs the fitting results 1520 to the user U1. Furthermore, the computer 3 outputs information 1530 to the user U1, which includes at least one of the device state, whether correction is necessary, and the correction location, based on the fitting results 1520.

[0118] [Fitting with a Multivariate Distribution Model] Figure 16 is an explanatory diagram of fitting with a multivariate distribution model. The calculator 3 sets two orthogonal axes for the distribution 9. Here, the aforementioned x direction (= x axis) and y direction (= y axis) are used as the two axes. In the example shown, the center point of the region of the distribution 9 is set as the origin, and the x axis and y axis are taken.

[0119] The computer 3 regards the amount of change over time in the image quality evaluation value for each predetermined region 7 in the distribution 9 as a probability density and approximates it using a bivariate normal distribution 1600. In other words, the computer 3 fits the distribution 9 to the bivariate normal distribution 1600. The bivariate normal distribution 1600 is an example of a multivariate distribution model. The bivariate normal distribution 1600 can be expressed by two parameters 1601, a mean matrix μ and a variance-covariance matrix Σ. The computer 3 searches for the parameters 1601 that best fit the distribution 510 using a steepest descent method or the like.

[0120] In fitting result 1520, which is the result of parameter search, parameter mean matrix μ indicates the center coordinates of distribution 9, and variance-covariance matrix Σ indicates the spread of distribution 9. From the searched parameters, calculator 3 can determine the variation center 1603 and spread of distribution 9 in a predetermined region for the amount of time change in image quality evaluation value. In the illustrated example, variation center 1603 corresponds to the region where the value in distribution 9 is 0.3. The portion corresponding to variation center 1603 has a relatively large amount of change, the ring-shaped portion surrounding it has a relatively medium amount of change, and the remaining portion has a relatively small amount of change. The center of the x-y plane in image region 1610 corresponding to distribution 9 is illustrated as point 1611. The variation center 1603 is located to the upper left of center point 1611.

[0121] The charged particle beam device 2 has multiple adjustment locations that affect a single image quality evaluation value, in other words, multiple correction candidate locations and corresponding component parts, which are closely related. Therefore, with conventional technology, it is difficult to select an adjustment location when a change occurs in the device state of the charged particle beam device 2. In contrast, in the third embodiment, the center of change and the spread of the image quality evaluation value can be identified, as in the parameter search result 1520. This makes it possible to narrow down the adjustment locations for correcting the image quality evaluation value that has changed as correction locations. In other words, it is easy to prioritize and narrow down suitable correction locations. Suitable correction locations can be selected in accordance with the locations with the largest temporal change in the image quality evaluation value and the spread, and presented to the user U1.

[0122] For example, if the amount of change over time in the amount of blur, which is one of the image quality evaluation values, is larger near the periphery of the image region 1610, in other words, if the center of change is located farther from the center toward the periphery, it can be estimated and determined that a change has occurred in the tendency of field curvature. Field curvature is a difference in the imaging distance between the charged particle beam 103 passing on and off the axis of the objective lens 108 in Figure 1. Therefore, based on this estimation and determination, it is possible to narrow down one of the priority correction points to the objective lens 108.

[0123] The spreading pattern may be, for example, uniform, radial, or local. The example in Fig. 16 shows a radial spreading pattern in concentric circles from a change center 1603. The correction location can be determined from the information on the change center and spreading pattern in the fitting result 1520. For example, the change center and spreading pattern correspond to the component parts of the charged particle beam device 2, and the correction location can be determined from this correspondence.

[0124] In the third embodiment, after obtaining the parameter search result 1520, the apparatus state, whether or not correction is necessary, the correction location, and the like may be estimated and output as subsequent processing, similar to the first embodiment. In the third embodiment, to determine the apparatus state, whether or not correction is necessary, and the correction location, for example, a configuration may be adopted in which the information on distribution 731 in the apparatus state DB 313 in Fig. 7B is replaced with the parameter search result 1520 (for example, the variation center and the spread mode) using a multivariate distribution model.

[0125] In the third embodiment, the parameter search result 1520 based on the distribution 9 can be said to be a visually easy-to-understand representation of the distribution of the amount of change in image quality over time in two-dimensional space, such as a graph or map. The calculator 3 displays such parameter search result 1520 on a GUI screen. By viewing the parameter search result 1520 on the GUI screen, the user U1 can easily recognize the center and spread of the change in the amount of change in image quality over time. This allows the user U1 to appropriately correct and adjust the objective lens 108, etc.

[0126] [Effects of the Third Embodiment, etc.] As described above, the third embodiment makes it possible to suitably manage the state of the charged particle beam device 2. In particular, by using a multivariate distribution model, it is possible to suitably and accurately grasp the center of change and the spread of the distribution of the amount of change over time of the image quality evaluation value, select suitable correction points, and improve the efficiency of the correction work by the user U1.

[0127] <Fourth Embodiment> A fourth embodiment will be described. The fourth embodiment is a modification of the method for calculating the distribution 9 of a predetermined region for the amount of change over time in the image quality evaluation value in FIG. 6 of the first embodiment. The fourth embodiment is also capable of dealing with cases where the observation regions are different between the evaluation target data and the reference data. In the fourth embodiment, an evaluation method will be described for cases where the shapes of the pattern structures included in the images of the evaluation target data and the reference data are different, in other words, where there is a deviation, when the calculator 3 calculates the distribution 9.

[0128] FIG. 17 shows an example in which the observation regions are different between the evaluation target image and the reference image. Region 1700 is an example of the same region on the surface of the sample 109. Evaluation target image 1701 is an image of region 1700 at a certain imaging position and imaging field of view. Reference images 1702a, 1702b, 1702c, and 1702d are images at respective imaging positions and imaging fields of view. These images are taken at slightly different imaging positions and imaging fields of view, and therefore the regions containing the shapes of the pattern structures shown are different. The region of evaluation target image 1701 includes a portion of each of the reference images 1702a, 1702b, 1702c, and 1702d.

[0129] 18 is an explanatory diagram of a method for calculating distribution 9 of a predetermined region for the amount of change over time in image quality evaluation values ​​when the evaluation target data and the reference data are observed at different positions and regions on the sample 109. Calculator 3 extracts, from evaluation target data DB 322, a plurality of evaluation target data 1801 (distribution 8A described above) within a certain evaluation period that includes the time at which the apparatus state is estimated, and calculates the average value of the image quality evaluation values ​​for each predetermined region 7 from the plurality of evaluation target data 1801, in other words, for each predetermined region 7 at a corresponding position in the x-y direction. The calculation result is evaluation target average value distribution 1811.

[0130] Furthermore, the calculator 3 extracts, from the reference data DB 321, a plurality of reference data 1802 (the above-described distribution 8B) including different observation regions, and calculates the average value of the image quality evaluation values ​​for each of the predetermined regions 7 from the plurality of reference data 1802. The calculation result is a reference average value distribution 1821.

[0131] Calculator 3 performs process 1830 in which the average values ​​for each predetermined region 7 are compared between evaluation object average value distribution 1811 calculated using a plurality of evaluation object data 1801 and reference average value distribution 1821 calculated using a plurality of reference data 1802, and the difference between them is calculated. As a result, calculator 3 obtains distribution 9 of the predetermined region for the amount of change over time in the image quality evaluation value as mean value difference distribution 1831, which is the calculation result.

[0132] Because image quality evaluation values ​​are affected by the pattern structure shape contained in the image to be evaluated, it is generally desirable to compare image quality evaluation values ​​between images containing the same pattern structure shape. Therefore, in the first embodiment, evaluation is performed by comparing an evaluation target image and a reference image having the same observation area. However, even if the evaluation target image and the reference image contain different pattern structure shapes, in other words, even if there is a difference in the observation area, the above-described statistical processing can be used to address this issue. In other words, by using the average image quality evaluation value based on multiple images taken over a certain period of time that contain various pattern structure shapes, it is possible to evaluate image quality taking into account the influence of differences in the pattern structure shapes between the evaluation target data and the reference data.

[0133] In the case of the third embodiment, it is not necessary to capture and manage the reference image and the evaluation target image so that the imaging position and imaging field of view are always the same. This also makes it possible to evaluate the image quality and to appropriately estimate and manage the device state even if there is no reference image of the reference data that has the same observation position and area on the sample 109 as the evaluation target image of the evaluation target data.

[0134] Although the embodiments of the present disclosure have been specifically described above, they are not limited to the above-described embodiments and various modifications are possible without departing from the spirit of the present disclosure. In each embodiment, components can be added, deleted, or replaced, except for essential components. Unless otherwise specified, each component may be singular or plural. Combinations of the embodiments and their variations are also possible. Some or all of the above-described configurations, functions, processing units, etc. may be realized by hardware, such as an integrated circuit design, or by software in which a processor interprets and executes a program. Data and information such as programs, tables, and files that realize each function can be stored in a storage device such as a memory, hard disk, or SSD, or in a storage medium such as an IC card, SD card, or DVD.

[0135] For example, the controller 200 may be configured to perform all or part of the processing of the computer 3 .

[0136] [Additional Note] The present embodiment can also be configured as follows.

[0137] In the computer of the embodiment, the processor acquires a plurality of evaluation target images over a certain period of time in the evaluation target data, calculates statistical values ​​of image quality evaluation values ​​for each specified area from the plurality of evaluation target images, acquires a plurality of reference images over a certain period of time in the reference data, calculates statistical values ​​of image quality evaluation values ​​for each specified area from the plurality of reference images, compares the distribution of statistical values ​​of image quality evaluation values ​​of the reference data with the distribution of statistical values ​​of image quality evaluation values ​​of the evaluation target data, and calculates the distribution of temporal change in image quality evaluation values.

[0138] In the computer of the embodiment, the processor moves the imaging field of view coordinates by deflecting the charged particle beam with the charged particle beam device, acquires an image group including at least a first image showing a first observation area and a second image showing a second observation area, and calculates a first distribution of image quality evaluation values ​​for each specified area of ​​the image area of ​​the image group.

[0139] In the computer of the embodiment, the processor acquires an image group including at least a first image showing a first observation area and a second image showing a second observation area by having the charged particle beam device simultaneously scan and irradiate multiple charged particle beams onto a sample, and calculates a first distribution of image quality evaluation values ​​for each specified area of ​​the image area of ​​the image group.

[0140] 1...observation system, 2...charged particle beam device, 3...computer, 4...observation area, 5...observation area, 6...image, 7...specified area, 8A, 8B...distribution, 9...distribution (distribution of specified area for time change in image quality evaluation value), 100...main body, 200...controller, 330...processor, 331...memory, 320...storage, 321...reference data DB, 322...evaluation target data DB, 323...device status DB, 324...device correction DB, 601...evaluation target data, 602...reference data.

Claims

1. A computer having a processor and a memory resource storing a program, wherein the memory resource stores reference data including the imaging field of view coordinates of the charged particle beam device obtained by a charged particle beam device or simulation for a sample, the generation time of a reference image in which an observation area corresponding to the imaging field of view coordinates is mapped, and a distribution of reference image quality evaluation values ​​obtained from the reference image by an image quality evaluation function for each specified area, and the processor that executes processing in accordance with the program: (1) acquires an evaluation target image of the sample, obtained by the charged particle beam device for an evaluation period including an evaluation time, the observation area corresponding to the imaging field of view coordinates is mapped, (2) stores in the memory resource evaluation target data including the imaging field of view coordinates, the evaluation time, and a first distribution of image quality evaluation values ​​obtained from the evaluation target image by the image quality evaluation function for each specified area, and (3) compares the reference data with the evaluation target data to calculate the distribution of time change in image quality evaluation values ​​using the image quality evaluation function.

2. A computer according to claim 1, wherein the processor estimates the state of the charged particle beam device at the evaluation time based on the relationship between the distribution of the amount of change over time of the image quality evaluation value and the state of the charged particle beam device.

3. A computer according to claim 2, wherein the processor determines whether or not correction is required for components of the charged particle beam device based on the distribution of the amount of change over time of the image quality evaluation value or the estimated result of the state of the charged particle beam device, and outputs the result of the determination as to whether or not correction is required.

4. A computer according to claim 2, wherein the processor determines the correction locations of component parts of the charged particle beam device based on the distribution of the time change in the image quality evaluation value or the estimated result of the state of the charged particle beam device, and outputs the correction locations as a result of the determination.

5. A computer according to claim 2, wherein the memory resource stores a database in which is stored the relationship between the distribution of the amount of change over time of the image quality evaluation value and the state of the charged particle beam device, and the processor inputs the distribution of the amount of change over time of the image quality evaluation value into the database and obtains, as an output, an estimate of the state of the charged particle beam device.

6. A computer according to claim 2, wherein the processor inputs the distribution of the amount of change over time of the image quality evaluation value into a model in which the relationship between the distribution of the amount of change over time of the image quality evaluation value and the state of the charged particle beam device has been learned, and obtains as an output an estimate of the state of the charged particle beam device.

7. A computer according to claim 2, wherein the processor fits a multivariate distribution model to the distribution of the time change of the image quality evaluation value, and obtains an estimate of the state of the charged particle beam device from parameters of the multivariate distribution model.

8. A computer according to claim 4, wherein the memory resource stores a database in which is stored the association between the state of the charged particle beam device and correction points of component parts of the charged particle beam device, and the processor inputs the estimated result of the state of the charged particle beam device to the database and obtains the correction points as an output.

9. A computer according to claim 4, wherein the processor inputs the distribution of the amount of change over time of the image quality evaluation value into a model in which the relationship between the distribution of the amount of change over time of the image quality evaluation value and the correction locations of the component parts of the charged particle beam device has been learned, and obtains as an output an estimate of the correction locations of the component parts of the charged particle beam device.

10. A computer according to claim 1, wherein a distribution of the amount of change over time in said image quality evaluation value is displayed on a screen provided to a user.

11. A computer as claimed in claim 1, wherein the specified area is an area obtained by dividing an image area of ​​one or more of the images to be evaluated into areas with a shape, size and number according to the image quality evaluation value to be calculated, and the specified area is set on a screen provided to a user based on the operation of the user, or preset setting information for the specified area is selected and applied.

12. A computer according to claim 1, wherein the image quality evaluation value obtained by the image quality evaluation function is at least one of brightness, contrast, blur, and distortion.

13. A computer according to claim 1, wherein the processor calculates the amount of blur for each direction in the image as the image quality evaluation value, and outputs an objective lens or an astigmatism corrector as a correction point as a component part of the charged particle beam device based on the distribution of the amount of blur for each direction.

14. The program of claim 1.

15. A method executed by a computer having a processor and a storage resource storing a program, wherein the storage resource stores reference data including: image field coordinates of the charged particle beam device obtained by a charged particle beam device or a simulation for a sample; a generation time of a reference image in which an observation area corresponding to the image field coordinates is mapped; and a distribution of reference image quality evaluation values ​​obtained from the reference image by an image quality evaluation function for each predetermined area; and the processor executing processing according to the program executes steps including: (1) a step of acquiring an evaluation target image of the sample, obtained by the charged particle beam device for an evaluation period including an evaluation time, in which an observation area corresponding to the image field coordinates is mapped, (2) a step of storing evaluation target data in the storage resource including: image field coordinates of the charged particle beam device, the evaluation time, and a first distribution of image quality evaluation values ​​obtained from the evaluation target image by the image quality evaluation function for each predetermined area; and (3) a step of comparing the reference data with the evaluation target data to calculate a distribution of time change in image quality evaluation values ​​by the image quality evaluation function. The method comprising: