JP2006228748A - Electron microscope - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently extract only a required field of view, by automatically deciding whether imaged field of view is suitable for searching a target form. <P>SOLUTION: The condition of the field of view is determined (S17) by a method of measuring the brightness (tone) of the line profile of an imaged field of view, a method of inspecting the coincidence (correlative function ) by finding the cross-correlation function of two transmissive images in the same field of view photographed at different electro-optical conditions , a method of inspecting the coincidence by finding the phase-only correlation function of two transmissive images photographed at different conditions, or the like; and if the condition of the field of view is not adequate for observation and searching, then the field of view moves to the following field of view. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an electron microscope, and more particularly to an electron microscope that automatically moves a visual field to search for a target form (structure) in a sample.

  In the conventional electron microscope, the measurement, analysis, and search of the field of view of the sample transmission image are performed directly by the operator through the sample stage operation and focus correction, and the desired field of view from the magnified sample transmission image projected on the fluorescent screen. The way to find out has been taken. This operation is very complicated and time consuming, and the worker is fatigued. As a result, a human error such as missing a required field of view is caused and the observation efficiency and accuracy are lowered. Therefore, as a method for automatically searching for a desired field of view, for example, Japanese Patent Laid-Open No. 8-298090 proposes a method for automatically driving a sample stage to display and store a sample image.

  In the method disclosed in Japanese Patent Laid-Open No. 8-298090, the field of view and the transmission image are automatically displayed and stored. However, the method is not completely automated, and the focus correction necessary for obtaining an electron microscope image is manually performed. Had to be done.

  As shown in FIG. 2, when observing a sample image or searching for a visual field, a thin film sample to be observed is placed on a sample holding mesh or microgrid and fixed to the sample holder. When the area to be searched is divided into nine as shown in FIG. 2, the transmission enlarged images of the sample are divided into three types as shown in FIGS. 2 (A), (B), and (C). When the observation visual field is the visual field (5) (FIG. 2B), a transmission enlarged image of the form 61 on the sample can be obtained. On the other hand, as shown in FIG. 2 (1), (4), and (7), when the entire field of view becomes dark (see FIG. 2). A)), or the field of view (9) is not suitable for observation when there is nothing in the field of view due to the tearing of the sample and it is completely white (FIG. 2C). is there. The conventional field-of-view observation method using an automatic stage drive in an electron microscope can intentionally and automatically remove a field unsuitable for observation, search, and analysis as shown in FIGS. There wasn't.

  Further, in the case of automatically searching for a form in the visual field required for automatic visual field observation in a conventional electron microscope, as shown in FIG. 3B, the ratio of the luminance between the form 61 and the background 65, that is, the contrast. When is small, there is a possibility that the search accuracy is remarkably lowered.

  In this way, in a conventional electron microscope that searches the target form by moving the field of view, when performing a search by automatically moving the field of view, a sample exists in the field of view and there is a field suitable for the search. There is a problem that the search is performed without distinguishing both of the unnecessary visual fields that are not suitable for the search that is an enlarged image of the sample holding mesh, and is inefficient. There is also a problem that complicated focus correction cannot be performed with high accuracy in a short time. In particular, when the search is performed over a wide area, the focus may be shifted for each field of view due to the curvature, contraction, and extension of the sample. Corrections had to be made and automation was impossible. Further, when focus correction for each field of view is not performed, there is a possibility that accurate search cannot be performed due to image blur due to defocus. In addition, when the contrast of the observation visual field is low, there is a problem that even if the visual field is suitable for the search, the target form cannot be searched or cannot be satisfactorily performed.

  Therefore, the present invention automatically determines whether the captured field of view is a field of view suitable for searching for the target form or an unsuitable field of view, and can efficiently extract only the necessary field of view. An object is to provide a microscope. It is another object of the present invention to provide an electron microscope that can automatically perform focus correction in a short time with high accuracy when a visual field is moved and a target form is searched. Furthermore, the present invention provides an electron microscope having a function of adjusting an electro-optical condition so as to satisfy a reference value when the automatically moved field of view does not satisfy a preset brightness or contrast reference value. For the purpose. It is another object of the present invention to provide an electron microscope having a function of selecting a pattern of a target form from the field of view and storing the number of forms, field-of-view coordinates where the target form exists, and an image where the target form exists.

  In order to achieve the above object, the electron microscope of the present invention is provided with means for determining the state of the field of view from the transmission magnified image of the imaged sample, and automatically moves to the next field if the field of view is not suitable for observation / search. Moving. A field of view that is not suitable for observation / retrieval is typically a field that is completely black or white and has no structure, and a field of view that is suitable for observation / retrieval typically has a change in brightness or a pattern within the field of view. Or a field of view having a structure. The state of the field of view is determined by creating a line profile of the captured field of view and measuring the brightness (gradation), obtaining the cross-correlation of two transmitted images of the same field of view taken by changing the electro-optic conditions, A method of viewing the degree of coincidence (correlation function) or a method of obtaining a phase-only correlation between two transmitted images taken under different conditions and viewing the degree of coincidence can be used.

  In addition, the electron microscope of the present invention includes a calculation device that calculates a defocus amount and a focus correction amount, and performs automatic focus correction by correcting the objective lens current value according to the calculated focus correction amount. The defocus amount can be calculated from a shift (amount of movement) between two enlarged specimen transmission images of the same field of view taken by changing the electro-optic conditions. In addition, the electron microscope of the present invention has a function of measuring the contrast of the visual field and adjusting the electro-optical condition so as to satisfy the reference value when it does not satisfy the preset reference value.

  That is, an electron microscope according to the present invention has an imaging device that supports a sample, a deflector that deflects an electron beam and enters the sample, an imaging unit that captures a sample transmission image of the electron beam transmitted through the sample, and Image calculation means for calculating a sample transmission image, means for setting visual field movement conditions, means for automatically moving the visual field based on the set visual field movement conditions, and imaging using the calculation result by the image calculation means A determination means for determining whether or not the field of view is suitable for observation, and displaying and / or storing a sample transmission image of the field of view determined to be suitable for observation by the determination means and coordinates of the field of view. And

  The field of view can be moved by changing the deflection amount of the deflector or by moving the sample support. Whether or not the imaged field of view is suitable for observation is determined by measuring the line profile of the field of view imaged by the image computing unit or by the image computing unit (electron beam incident on the sample). Can be performed by calculating the cross-correlation or phase-only correlation between two specimen transmission images of the same field of view taken with different deflection angles.

  An electron microscope according to another aspect of the present invention also includes a support base that supports the sample, a deflector that deflects the electron beam and enters the sample, and an imaging unit that captures a sample transmission image by the electron beam that has passed through the sample. Image calculation means for calculating an image of the captured sample transmission image, means for setting a visual field movement condition, means for automatically moving the visual field based on the set visual field movement condition, and a calculation result by the image calculation means Means for determining whether or not the field of view imaged using is suitable for observation, and means for imaging again the field of view determined to be unsuitable for observation by the determination unit by adjusting the electro-optic condition A sample transmission image of the field of view determined to be suitable for observation by the determination means and the coordinates of the field of view are displayed and / or stored.

  The adjustment of the electron optical condition includes adjustment of the electron source, adjustment of the lens condition of the irradiation lens, adjustment of the electron beam current, and adjustment of the hole diameter or position of the objective movable diaphragm. The electron microscope includes a means for registering a search target pattern, a search means for searching for a pattern that matches the search target pattern from a field of view determined to be suitable for observation by the determination means, and a search target pattern searched by the search means. Means may be provided for displaying and / or storing the number and field of view coordinates.

  An electron microscope according to another aspect of the present invention also includes a support base that supports the sample, a deflector that deflects the electron beam and enters the sample, and an imaging unit that captures a sample transmission image by the electron beam that has passed through the sample. A means for setting a field-of-view movement condition, a means for automatically moving the field of view based on the set field-of-view movement condition, a means for registering a search target pattern, and two of the same field of view by electron beams having different incident angles. A calculation means for obtaining two-dimensional discrete Fourier transform data for each sample transmission image and obtaining a cross-correlation of only phase information, and a field of view where the cross-correlation calculated by the calculation means has reached a predetermined threshold value are selected. Search for the target pattern registered from the field of view selected by the field of view selection means, the means for automatically correcting the focus from the cross-correlation results obtained by the computing means, and the field of view selected by the field of view selection means. A search unit that, characterized in that it comprises means for displaying and / or storing the number and the field coordinates of the search target pattern retrieved by the retrieval means.

  An electron microscope according to still another aspect of the present invention includes a support base that supports a sample, a deflector that deflects an electron beam and enters the sample, and an imaging unit that captures a sample transmission image by the electron beam transmitted through the sample. , Means for setting visual field movement conditions, means for automatically moving the visual field based on the set visual field movement conditions, means for registering a search target pattern, and contrast measuring means for measuring the contrast of the picked-up visual field And means for adjusting the electro-optical conditions so that the contrast is increased when the contrast measured by the contrast measuring means is smaller than a predetermined value, and the contrast measured by the contrast measuring means is predetermined. Search means for searching for a search target pattern registered from a field of view larger than the measured value, and the search means Characterized in that it comprises a means for displaying and / or storing the number and the field coordinates of the search target pattern. The contrast of the visual field can be measured by creating a line profile of the signal intensity (brightness) of the photographed sample transmission image.

  An electron microscope according to another aspect of the present invention includes a support that supports a sample, a deflector that deflects an electron beam and enters the sample, an imaging unit that captures a sample transmission image of the electron beam that has passed through the sample, Means for setting a field-of-view movement condition; means for automatically moving the field of view based on the set field-of-view movement condition; means for registering a search target pattern; and two pieces of the same field of view by electron beams having different incident angles Calculation means for generating respective two-dimensional discrete Fourier transform data for the sample transmission image and obtaining the cross-correlation of only the phase information, and irradiation electron lens system when the cross-correlation calculated by the calculation means does not reach a predetermined threshold value Alternatively, the imaging electron lens system is adjusted and two sample transmission images are taken again to obtain the cross-correlation of only the phase information, and the cross-correlation calculated by the calculation means is a predetermined threshold. Visual field selection means for selecting the visual field that has reached the point, means for automatically correcting the focus from the result of cross-correlation obtained by the calculation means, contrast measuring means for measuring the contrast of the visual field after the focus correction, and contrast A means for adjusting the electro-optical condition so that the contrast is increased when the contrast measured by the measuring means is smaller than a predetermined value, and a search target pattern registered from a field of view where the contrast is larger than a predetermined value. And a means for displaying and / or storing the number of search target patterns searched by the search means and the coordinates of the field of view.

The electron microscope of each aspect of the present invention has the following characteristics.
(1) The visual field can be automatically moved and selected, and it can be automatically determined whether or not the selected visual field has a brightness suitable for the search.
(2) It is possible to automatically move and select the field of view, automatically determine whether or not the selected field of view is suitable for search, extract only the necessary field of view, and record an image.
(3) Automatically select and move the field of view, automatically determine whether the selected field of view is suitable for search, extract only the necessary field of view, record the coordinate information of the extracted field of view, and re-view the field of view Make observation easier.
(4) The visual field can be automatically moved and selected, and the focus correction can be automatically performed in the selected visual field.
(5) It is possible to automatically measure the contrast between the background and form of the automatically selected visual field.
(6) The contrast between the background of the automatically selected field of view and the form is automatically measured, the contrast is automatically adjusted in a short time, and the search accuracy according to the target form can be improved.

  According to the present invention, it is possible to automatically perform a sample search operation using an electron microscope, which has been complicated and requires much labor and considerable time, with high accuracy and in a short time.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic functional block diagram of an example of a transmission electron microscope according to the present invention. The number of stages of the electron beam deflection coil is not limited, but here, an example where a total of four stages of electron beam deflection coils are used, two stages at the top of the sample and two stages at the bottom of the sample will be described. Further, all the embodiments described below will be described on the assumption that the transmission electron microscope shown in FIG. 1 is used.

  The electron beam 73 emitted from the electron gun 1 and accelerated passes through the front magnetic field of the first irradiation lens coil 2, the second irradiation lens coil 3, and the objective lens coil 4, and reaches the sample 14 held on the sample stage 13. Irradiated. The electron beam 73 that has passed through the sample 14 is magnified by the first intermediate lens coil 5 and the second intermediate lens coil 6, and then further magnified by the first projection lens coil 7 and the second projection lens coil 8, on the scintillator 16. An enlarged transmission image 59 of the sample is formed. Here, a coil applying an electromagnetic field force is used as a magnifying lens for electron beam deflection and sample transmission image, but electrostatic deflection and electrostatic lens using electrostatic force are used for electron beam deflection and sample transmission image magnification. May be used.

  A sample transmission enlarged image 59 converted into an optical image by the scintillator 16 is captured by an imaging device, for example, a TV camera 17. The video signal from the TV camera is captured and processed by the microprocessor 46 via the TV camera control unit 33 and the image capturing interface 34, and then displayed as an image on the CRT 50 controlled by the CRT controller 49. Here, the scintillator 16 and the TV camera are used when the transmission magnified image is taken into the microprocessor 46, but a detection element such as an MCP (microchannel plate) that can directly convert an electron beam into an electric signal is used. Also good.

  The microprocessor 46 is connected to the first irradiation lens coil 2, the second irradiation lens coil 3, and the objective lens coil 4 of the electron microscope via DACs (digital-analog converters) 35, 36, 37, 38, 39, 40, 41. The excitation power supplies 18, 19, 20, 21, 22, 23, 24 for supplying power to the first intermediate lens coil 5, the second intermediate lens coil 6, the first projection lens coil 7, and the second projection lens coil 8 are controlled. Excitation power sources 25, 26, 27, and 28 for the sample upper first deflection coil 9, second deflection coil 10, sample lower first deflection coil 11, and second deflection coil 12 are also supplied to the DACs 42, 43, 44, 45 by the microprocessor 46. Are controlled in the same way.

  The microprocessor 46 is connected to an external storage device 47 such as a hard disk, an arithmetic device 48, a magnification switching rotary encoder 53, a keyboard 55, a RAM 57, a ROM 58, and the like via a bus. The magnification switching rotary encoder 53 is connected to the bus via an I / F (interface) 51. The sample stage 13 is driven by a fine drive motor 29 for driving the stage connected to the microprocessor 46 via the motor driver 30.

  Next, as an example of image calculation according to the present invention, the principle of obtaining the degree of coincidence between two images by the phase-only correlation method and the principle of performing automatic focus correction will be described. A transmission image 1 which is a reference specimen transmission enlarged image as shown in FIG. 3A is recorded as f1 (m, n) in the storage device with the number of pixels of M × N. Next, a current is applied to the two upper electron beam deflection coils, and an appropriate tilt deflection angle α is applied to the electron beam applied to the sample. Is recorded in the storage device as f2 (m, n). Here, m = 0, 1, 2,..., M−1; n = 0, 1, 2,.

  The discrete Fourier images F1 (u, v) and F2 (u, v) of the transmission images f1 (m, n) and f2 (m, n) are defined by the following [Equation 1] and [Equation 2], respectively. Here, u = 0, 1, 2,..., M−1; v = 0, 1, 2,..., N−1, A (u, v) and B (u, v) are amplitude spectra, α (u, v) and β (u, v) are phase spectra.

  In phase-only correlation, if there is a parallel movement of an image between two images, the position of the correlation peak is shifted by the amount of movement. A method for deriving the movement amount will be described below. First, assuming that the transmitted image f2 (m, n) has moved by r ′ in the m direction, f3 (m, n) = f2 (m + r ′, n). A discrete Fourier image F3 (u, v) of f3 (m, n) is obtained from [Equation 2] to [Equation 3].

  If the amplitude spectrum B (u, v) is set as a constant, a phase image independent of the contrast and brightness of the image is obtained. The phase image F3 ′ (u, v) of f3 is expressed by the following [Equation 4]. Similarly, the phase image F1 '(u, v) of f1 is expressed by the following [Equation 5].

  By multiplying the phase image F1 ′ (u, v) by the complex conjugate of F3 ′ (u, v), a synthesized phase image H13 (u, v) represented by the following [Equation 6] can be obtained. A correlation strength image, that is, a correlation index (degree of coincidence between two images) g13 (r, s) is expressed by the following [Equation 7] by performing inverse Fourier transform on the synthesized image H13 (u, v).

When the correlation strength image obtained by [Equation 7] is normalized, if the obtained value is 0, the two images can be recognized as completely different images, while if 100, the two images are exactly the same image. Can be recognized.
When there is a positional shift amount r ′ in the m direction between the two images according to [Equation 7], the correlation peak position of the correlation strength image is shifted by −r ′. In this way, by using the phase-only correlation method, the degree of coincidence between the two images of the transmission image 1 and the transmission image 2 without depending on the contrast and brightness of the image, and the image shift between the two images can be reduced. Can be sought.

When a peak occurs at a position shifted by ΔG [pixel] in the correlation intensity image as a result of the calculation processing of [Equation 1] to [Equation 7] between the two specimen transmission images, ΔG [pixel] This corresponds to the amount of movement on the light-receiving surface of a detector such as a TV camera, and ΔG is converted to the amount of movement Δx on the sample surface. If the diameter of the light receiving surface to be detected is L [m], the magnification of the electron microscope on the light receiving surface is M, and the number of pixels of the detector is L _d [pixel], the amount of movement on the sample surface between the two images. Δx is calculated by the following [Equation 8]. However, [Equation 8] includes the amount of movement δ of the image due to the spherical aberration of the electron lens, and the true amount of movement Δx _t of visual field deviation is obtained by subtracting δ from Δx. Δ on the sample surface is expressed as [Equation 9] by the spherical aberration Cs and the electron beam deflection angle α. From the above, the movement amount Δx _t of the image generated between the two specimen transmission enlarged images is expressed by [Equation 10].

Between the movement amount [Delta] x _t and the defocus amount Δf of the image, a relationship of the following [Equation 11]. By using this relationship, it is possible to calculate the defocus amount Δf from the movement amount [Delta] x _t of the image.

An objective current correction value is calculated from the defocus amount Δf calculated in [Equation 11]. There is a relationship of [Equation 12] between the focal length f of the electron lens and the objective lens current I. Here, N is the number of turns of the electron lens coil, E ^* is the relativistic corrected acceleration voltage, and I is the objective lens current value. Therefore, focusing can be achieved by adding the objective current correction value obtained from the relationship of [Equation 12] to the objective current value.

Hereinafter, embodiments of the present invention will be specifically described.
[Embodiment 1]
As a first embodiment, referring to the flowchart of FIG. 4, the field of view is automatically moved and selected using the transmission electron microscope shown in FIG. 1, and the field of view is not suitable for observation or search. A method for determining whether or not only the appropriate visual field is efficiently observed or searched will be described.

  In step 11 of FIG. 4, an enlargement magnification is set to obtain an arbitrary sample transmission image using a transmission electron microscope. The magnification input of the sample transmission image is performed using a rotary encoder 53 for magnification switching. The pulse wave generated by the rotary encoder 53 is converted into a digital signal by the I / F 51. The microprocessor 46 refers to the magnification display data preset in the ROM 58 based on the digital signal input from the I / F 51 and causes the CRT 50 to display the corresponding magnification. At the same time, the first irradiation lens coil 2, the second irradiation lens coil 3, the objective lens coil 4, the first intermediate lens coil 5, the second intermediate lens coil 6, the first projection lens coil 7, Data of each lens of the dual projection lens coil 8 is output to the DACs 35, 36, 37, 38, 39, 40, and 41, and the data of the lens system mechanism is converted into an analog signal. The DAC outputs analog signals to the excitation power supplies 18, 19, 20, 21, 22, 23, and 24, and outputs current to the lens coils of each lens system mechanism.

Next, in step 12, a condition for automatically moving the visual field is set. The setting of the visual field movement speed and the search range is input using the keyboard 55 or the mouse 56, processed by the microprocessor, and stored in the storage device.
In step 13, the origin of visual field movement is set. For example, as shown in FIG. 2, the visual field movement origin 64 determines the angle of the visual field displayed on the display device (CRT) 50 using an input device such as a mouse. The coordinate position processed by the microprocessor 46 and determined as the origin is stored in the storage device 47.

  In step 14, according to the visual field movement conditions set in step 12 and step 13, the visual field is moved, for example, as (1) → (2) → (3) →. However, the movement of the observation visual field 62 is not limited to the method of sequentially performing the visual field (1) to the visual field (9) as illustrated in FIG. 2, and the observation visual field 62 may be moved to a randomly extracted visual field. The field of view can be moved electromagnetically using electron beam deflection coils placed at the top and bottom of the sample, mechanically using a sample stage driving device, or a stage using a piezoelectric element such as a piezo element. It can be performed by a drive mechanism or the like.

  FIG. 5A is a schematic explanatory diagram of a method for electromagnetically moving the visual field. Under the command of the microprocessor 46, two electron beam deflection coils (the sample upper first deflection coil 9 and the sample upper second deflection coil 10) arranged on the upper part of the sample in accordance with the field movement conditions set in step 12 Thus, the electron beam 73 is translated from the electron beam optical axis 72 passing through the visual field 70 at the center of the sample like the electron beam 67 at the time of deflection to irradiate the sample 14. The deflection electron beam 67 irradiates a visual field 71 that is a distance d away from the center of the sample. The electron beam after passing through the sample is returned to the electron beam optical axis 72 by the electron beam deflection coils arranged at the lower part of the sample, that is, the sample lower first deflection coil 11 and the sample lower second deflection coil 12. As a result, an enlarged sample transmission image after the visual field movement is obtained.

  FIG. 5B is a schematic explanatory diagram of a method for mechanically moving the visual field. In this case, the microprocessor 46 controls the motor driver 30 in accordance with the visual field movement condition set in step 12 and drives the fine movement mechanism 13 of the sample stage by the stage drive motor 29 to perform fine movement of the sample. In the visual field movement method using a piezoelectric element, the stage drive motor is replaced with a piezoelectric element, and the sample is finely moved.

  In step 15 of FIG. 4, an operation for stopping the visual field movement is performed. The condition for stopping the field movement is determined by the field movement condition set in step 12 and the size per field determined by the magnification input in step 11, and image overlap occurs between the field to be observed next. Move and stop so that there is no.

  In step 16, a sample transmission image is taken with the field of view stopped. The electron beam that has passed through the sample 14 passes through the objective lens 4, the first and second intermediate lenses 5 and 6, the first and second projection lenses 7 and 8, and forms an enlarged sample transmission image 59 on the scintillator 16. . An image projected on the scintillator 16 is picked up by the TV camera 17, and an enlarged image is registered as a transmission image 1 in the storage device by the image capturing interface 34.

  In step 17, it is determined whether or not the transmission image 1 imaged in step 16 is suitable for observation. Whether the field of view of the transmitted image 1 has a brightness (gradation) suitable for observation is determined by creating a line profile of the captured field of view and measuring the brightness (gradation), and changing the electro-optical conditions. The cross-correlation between two transmitted images of the same field of view is obtained, and a method of judging based on the degree of coincidence (correlation function), or the phase-only correlation of two transmitted images taken under different conditions is obtained, and the degree of coincidence It can be performed by a method of determining by the above. As a result of determining whether or not the field of view is suitable for observation according to these determination methods, if it is determined that the field of view is inappropriate, the process returns to step 14. On the other hand, if it is determined that it is suitable for observation, the process proceeds to step 18.

  FIG. 6 is an explanatory diagram of a method for determining the visual field state based on the captured visual field line profile. As shown in FIG. 6R> 6 (A), the x and y coordinates are determined, the measurement lines 74 as shown in the figure are drawn in the x direction and / or the y direction, and the change in luminance on each measurement line is represented as a line profile. measure. 6B, 6C, and 6D show examples of measurement results in the x direction, respectively. In the case of a 256-tone black-and-white image, as shown in FIGS. 6B and 6D, when the line profile shows gradation 256 or 0 over all pixels, This is a field of view in which an unsampled region is imaged, and is a field that is inappropriate for observation. FIG. 6C shows an example of a line profile of a visual field suitable for observation. If the form 61 exists in the visual field 62 as shown in FIG. 6R> 6 (A), the form appears as a change in contrast when the electron beam passes through the sample. Therefore, the line profile as shown in FIG. be painted. In FIG. 6A, only one measurement line 74 is drawn in the x direction, but the number of measurement lines is arbitrary, and an entire scan may be used. Further, it is preferable to draw a plurality of measurement lines not only in the x direction but also in the y direction and examine the line profile on the measurement line.

  FIG. 7 is a flowchart for explaining a method of determining the visual field state by the phase only correlation method. A series of operations up to step 15 is as described with reference to FIG. Proceeding to step 16 ′, a transmission enlarged image projected on the scintillator 16 is captured using the TV camera 17. This transmission enlarged image is stored as a transmission image 1 in a storage device. Next, in step 17 ′, the electron beam applied to the sample is made incident by being inclined by the deflection angle α, and in the subsequent step 18 ′, the enlarged transmission image projected on the scintillator 16 is stored as a transmission image 2 in the storage device.

  In step 19 ′, the transmission image 1 and the transmission image 2 are called from the storage device, the discrete Fourier transform data of each image is created by the arithmetic unit 48, and the phase-only correlation described in the above [Equation 1] to [Equation 7]. The degree of coincidence between the transmission image 1 and the transmission image 2 is calculated by the method.

  In step 20 ′, based on the degree of coincidence between the transmitted image 1 and the transmitted image 2 obtained in step 19 ′, whether the current field of view is an image that is inappropriate for field-of-view observation and search, or is appropriate. Judge whether the field of view is correct. If the degree of coincidence between the transmission image 1 and the transmission image 2 is 0, the visual field cannot be measured, and the process returns to step 14 to search for the next visual field. If the degree of coincidence between the transmitted image 1 and the transmitted image 2 is neither 0 nor 100, the process proceeds to step 21 '(corresponding to step 18 in FIG. 4) as an appropriate measurable area. If the degree of coincidence between the transmission image 1 and the transmission image 2 is 100, the current visual field is an image captured on the mesh (FIG. 2A), or there is no form in the visual field due to a tearing of the sample ( 2 (C)), and the process returns to step 14. Theoretically, it is assumed that the two images are completely equal at a coincidence degree of 100. However, as a result of the experiment, the two images having the coincidence degree of 100 are black images as shown in FIG. It is known that the image is a pure white image as shown in (C). That is, an image in which the field of view becomes black indicates that the sample holding mesh 6 is being imaged, and an image in which the field of view becomes pure white indicates that there is no form in the field of view due to the tearing of the sample. . Note that the time required to calculate and determine one field of view based on the above-described phase-only correlation method is approximately 0.9 to 1 second in the current apparatus.

  Here, it will be described that the degree of coincidence between two images showing uniform gradation over all the pixels of M × N pixels is 0 or 100. The function of image 1 is defined as f1 (m, n), and the function of image 2 is defined as f2 (m, n). Image 1 and image 2 have constant brightness (gradation) over all pixels. Further, m = 0, 1, 2,..., M−1; n = 0, 1, 2,. From the above conditions, f2 (m, n) is given by [Equation 13]. Therefore, [Equation 14] and [Equation 15] are obtained from [Equation 1] and [Equation 2].

  Considering the phase-only correlation, [Equation 14] and [Equation 15] can be obtained by making the amplitude component A (u, v) constant as 1 and F1 ′ (u, v) and F2 ′ (u, v), respectively. [Equation 16] and [Equation 17].

  A composite phase image H (u, v) obtained by multiplying [Equation 16] by the complex conjugate of [Equation 17] is expressed by [Equation 18].

  Next, the correlation index (correlation strength image) g (r, s) is obtained by performing inverse Fourier transform on [Equation 18] to obtain [Equation 19].

If the obtained value MN is normalized, the correlation index is 100 or 0.
As an example of a method for determining the visual field state, cross-correlation can be used. The operation flow is the same as Step 16 ′ to Step 20 ′ of FIG. 7 using the phase only correlation method, but the calculation principle is different. A method for calculating the degree of image coincidence using cross-correlation will be described below.

  The sample transmission magnified image is recorded as transmission image 1 and recorded as f1 (m, n) in the storage device with the number of pixels of M × N. Next, a sample transmission enlarged image having the same field of view as the transmission image 1 obtained by applying an electric current to the two upper electron beam deflection coils 9 and 10 and applying an inclined deflection angle α in the electron beam 73 irradiated to the sample 14. As a transmission image 2 and recorded in the storage device as f2 (m, n) with the number of pixels of M × N. Here, m = 0, 1, 2,..., M−1; n = 0, 1, 2,.

  The discrete Fourier images F1 (u, v) and F2 (u, v) of the transmission images f1 (m, n) and f2 (m, n) are defined by the above-described [Equation 1] and [Equation 2], respectively. By multiplying the discrete Fourier transform image F1 (u, v) of the transmission image 1 by the complex conjugate of the discrete Fourier transform image F2 (u, v) of the transmission image 2, the composite image H12 represented by the following [Equation 20]. (u, v) can be obtained. The correlation strength image, that is, the correlation index (the degree of coincidence between the two images) g12 (r, s) is obtained by performing the inverse Fourier transform on the composite image H12 (u, v) as shown in [Expression 21].

  When the correlation strength image obtained in [Equation 21] is normalized, if the obtained value is 0, the two images can be recognized as completely different images, whereas if they are 100, the two images are exactly the same image. It can be recognized that there is. Even in the method using cross-correlation, if the degree of coincidence is obtained as 0 or 100, it is determined that the current field of view is inappropriate for observation and retrieval, and the process returns to step 14.

  Returning to FIG. 4, in step 18 (step 21 ′ in FIG. 7), using the TV camera 17, the sample transmission enlarged image 59 projected on the scintillator 16 when it is determined that it is suitable for observation and retrieval. An image is taken and stored as image data in the storage device 47 and displayed on the CRT 50 via the CRT driver 49 or composition analysis is performed. Further, the coordinates of the selected visual field are stored in the recording device 47.

Finally, in step 19 in FIG. 4 (step 22 ′ in FIG. 7), it is determined whether or not to end the flow based on the visual field movement condition set in step 12, and if not, the process returns to step 14 to search for the next visual field again. Return.
By the above operation, it is possible to automatically move and select the visual field, determine whether or not the selected visual field is suitable for observation, and efficiently observe only the visual field suitable for observation.

  As an actual measurement example, assuming that the sample size is 2 mm in diameter, the search is performed by enlarging the sample to a diameter of 100 mm at a transmission image magnification of 10,000 times. At this time, the size per field of view corresponds to about 10 μm on the sample surface. When the entire region is observed, 40000 images are required. However, since the sample is actually supported by the sample holding mesh, there are a field of view where nothing is captured when the mesh is photographed as shown in FIG. 2, and a field of view where no sample exists. It can be estimated that the mesh area in the entire search area having a diameter of 2 mm is about ½, and the field in which the sample exists is 1/10 of the whole area. Therefore, the visual field suitable for observation in all the search areas is about 1/20 of 40,000 photographing visual fields. According to the present embodiment, it is possible to automatically select only the field of view suitable for observation, to reduce the number of imaging cases to 1/20, and to shorten the time required for search to about 1/20 as a result.

[Embodiment 2]
As a second embodiment, referring to the flowchart of FIG. 8, the visual field is automatically moved and selected, it is determined whether or not the visual field is appropriate for observation or search, and only the appropriate visual field is efficiently used. A method for automatically adjusting the electro-optical condition when the field of view is determined to be unsuitable for observation after observation or retrieval will be described.

  In FIG. 8, the magnification input in step 21, the visual field movement condition setting in step 22, the movement origin setting in step 23, the visual field movement in step 24, and the visual field movement stop in step 25 are the same as steps 11 to 15 in FIG. Therefore, the overlapping description is omitted.

  In step 26, the sample transmission magnified image 59 projected on the scintillator 16 is imaged using the TV camera 17 or the like, and the current field of view is obtained using the line profile method, the phase-only correlation method, or the cross-correlation method described above. It is determined whether or not is suitable for observation and measurement. If it is determined in step 26 that the current field of view is suitable for observation, the process proceeds to step 28 to measure, observe, and analyze the sample transmission magnified image, but it is determined that the current field of view is unsuitable for observation. If yes, go to Step 27.

  In step 27, an item which is considered to be a cause of the field of view being unsuitable for observation is automatically investigated, and the electro-optic condition is adjusted. Here, the following four items are investigated and adjusted. (1) Investigation and adjustment of electron beams. (2) Investigation and adjustment of the lens conditions of the irradiation lens. (3) Investigation and adjustment of electron beam current. (4) Investigation / adjustment of the hole diameter or position of the objective movable diaphragm. (1) investigates the possibility that an electron beam is not emitted when the entire field of view is black. (2) investigates the possibility that the transmission magnified image cannot be detected by the TV camera sensitivity, depending on the irradiation lens condition that makes the sample irradiation electron beam density very low. (3) investigates whether the emission current or filament current is insufficient. (4) is that the hole diameter of the objective movable aperture is selected to be smaller than necessary and the field of view becomes dark, making it impossible to detect with the TV camera, and the position of the aperture of the movable aperture and the electron beam optical axis. Investigate the possibility of misalignment.

  These four items are compared with values set in advance, and when it is determined that the set values are not satisfied, adjustment is performed so that the set values are satisfied for each item. If it is determined that the current field of view is not suitable for observation as a result of re-imaging and determination despite satisfying the conditions in accordance with the above items, the process returns to step 24 to search for the next field of view. . On the other hand, as a result of the adjustment, if it is determined that the current visual field satisfies the set value and is suitable for imaging, the process proceeds to step 28, and observation, storage, measurement, and analysis of the transmission image of the current visual field are performed. Do. Further, the coordinates of the selected visual field are stored in the recording device 47. Finally, the process proceeds to step 29, where it is determined whether or not to end from the visual field movement condition set in step 22 to step 23, and when the measurement is performed again based on the same flow, the process returns to step 24.

  Through the above operation, the field of view is automatically moved and selected, whether or not the selected field of view is suitable for observation is determined, and the determination is made again by adjusting the electron optical conditions in the inappropriate field of view, so that there is no leakage. Automatic observation is possible. According to the present embodiment, it is possible to minimize a search omission due to insufficient adjustment of the electron optical conditions and to prevent an artificial error such as no electron beam being emitted.

[Embodiment 3]
As the third embodiment, referring to the flowchart of FIG. 9, the visual field is automatically moved and selected, and an arbitrary inspection target pattern equivalent to the target form is set in advance, and the visual field is observed or searched. It is judged whether the brightness (gradation) is inappropriate, and only the appropriate field of view is efficiently searched. If it is determined to be inappropriate, the electro-optic condition of the device is automatically adjusted and the inspection is performed in the field of view. A method of searching for a form having a pattern equal to the target pattern, measuring the number of searched forms, and displaying / saving the form will be described.

  In this embodiment, for example, the triangular shape shown in FIG. 10A is set as a search target pattern, and the field of view is automatically moved as a result of automatic movement of the field of view as shown in FIG. Suppose that it was obtained. Forms having the same pattern as the triangle selected as the search target pattern are automatically recognized and marked, and the number of search target forms in the field of view is output and displayed.

  In step 31 of FIG. 9, the magnification for setting the sample transmission image is set. The magnification of the sample transmission image is set, and a lens current corresponding to the magnification is output to each lens coil. In step 32, a pattern (search target pattern) having the same shape as the search target form is set using the keyboard 55 or the mouse 56 or the like. The search target pattern can be set according to conditions such as the angle range formed by the sides, the ellipticity, and the ratio of the length between the major axis and the minor axis. The search target pattern may be set by calling a shape previously stored in the storage device.

  The field movement condition setting in step 33, the field movement origin setting in step 34, the field movement movement in step 35, the field movement stop in step 36, and the specimen transmission magnified image capturing in step 37 are the same as steps 12 to 16 in FIG. Yes, duplicate explanation is omitted. In step 38, it is determined whether or not the current visual field is suitable for observation and retrieval. The determination can be performed by the above-described line profile method, phase-only correlation method, or cross-correlation method. If it is determined that the field of view is suitable for observation, the process proceeds to step 40. If it is determined that it is not suitable for observation, the routine proceeds to step 39.

  In step 39, as in step 27 of FIG. 8, the cause for determining that the field of view is not suitable for observation is automatically investigated, and the electro-optical conditions are adjusted. As described above, if the current field of view is determined to be unsuitable for observation as a result of re-imaging and determination in spite of satisfying the conditions according to each survey item, the process returns to step 35 and the next Search the field of view. On the other hand, as a result of the adjustment, if it is determined that the current visual field satisfies the set value and is suitable for imaging, the process proceeds to step 40.

  In step 40, the form 61 determined to be the same as the search target pattern set in step 32 is extracted for the visual field determined to be appropriate for measurement and observation. Next, the process proceeds to step 41, where the number of search target forms extracted in the field of view photographed in step 37 is measured by the microprocessor 46, and the measurement result is stored in the storage device 47, or the field image is stored in the storage device. Or analysis of the detection target form and photography. Further, the coordinates of the selected visual field are stored in the recording device 47.

  Further, the field of view in which the search target form is detected is displayed on the display device 50 so as to be differentiated from the field of view in which the search target form has not been detected. For example, the field of view in which the search target form exists is displayed in red, and the field of view in which the search target form does not exist is displayed in gray.

  FIG. 16 is a schematic diagram illustrating a display example of the display device 50. In this example, an operation state display unit 81 and a visual field display unit 82 for displaying an enlarged image of the current observation visual field are provided side by side on the display screen 80 of the display device 50. The operation state display unit 81 displays a schematic diagram 83 of the sample holding mesh 6 and also schematically displays the positional relationship of the individual observation fields with respect to the sample holding mesh. In addition, the visual field arranged in the operation state display unit 81 displays the visual field in which the search target form exists and the visual field in which the search target form does not exist as distinguished from each other by color coding or the like as described above.

  Returning to FIG. 9, it is finally determined whether the automatic search operation is to be repeated or terminated in accordance with the visual field movement conditions set in steps 33 and 34 (step 42). When the automatic search operation is repeated, the process returns to step 35 and the series of operations so far are performed.

  With the above operation, the field of view is automatically moved, and the selected electron microscope searches only the field of view suitable for observation. It is possible to adjust the optical conditions so as to be used for observation, and to automatically search for the form to be searched. Also, the search is not performed on the mesh where the sample cannot be found clearly in the field of view or on the portion where the sample is broken.

  According to the present embodiment, similar to the above-described embodiment, the search operation of the target form that has been conventionally performed after taking a photograph by the hand of an operator can be performed instantaneously simultaneously with the imaging. Moreover, since only the visual field suitable for visual field observation is automatically extracted, it is more efficient than the method of capturing the entire region of the sample and searching for the target form.

  When a sample with a diameter of 2 mm is enlarged to a diameter of 100 mm at a transmission image magnification of 10000 times, all fields of view on the sample holding mesh are photographed, and the retrieval operation of the target form is manually performed from the photographed photos. In this case, it takes about 670 hours since a total of 40,000 images are required for one minute as the time required for searching the form of one photo. According to the present embodiment, it is possible to automatically extract only the field of view of the sample that is not torn or on the mesh and search for the target structure form. The field of view suitable for observation is about 1/20 of the entire area, and the retrieval time of the form in one field of view is about 1 second with the current apparatus, so the retrieval time of the form is dramatically shortened.

  In addition, by adjusting the electro-optical conditions, it is possible to minimize search leaks due to insufficient adjustment of the electro-optical conditions, to prevent human error such as the electron beam not being emitted, and for a long time. Thus, it is possible to keep the observation conditions stable by suppressing the change of the electro-optic conditions over time, which may be generated by the automatic observation operation. Furthermore, the field of view in which the target form exists is clarified by color and the coordinates are saved, thereby facilitating observation and analysis work again after the search operation is completed. When re-observation is performed, by selecting a displayed field of view, the sample stage 13 is driven based on coordinates stored in relation to the field of view, or a target field of view is used using an electron beam deflector. , And a transmission enlarged image of a desired visual field can be instantaneously obtained.

[Embodiment 4]
As a fourth embodiment, referring to the flowchart of FIG. 11, the field of view is automatically moved and selected, it is determined whether the current field of view is suitable for observation and search, and automatic focus correction is performed, and a preset search is performed. A method for automatically searching for a desired shape pattern will be described.

  In FIG. 11, the setting of the enlargement magnification in step 51, the setting of the search target pattern in step 52, the condition setting for automatically moving the visual field in step 53, the origin setting for the visual field movement in step 54, The visual field movement and the visual field movement stop in step 56 are the same as those in step 31 to step 36 in FIG.

  In step 57 to step 60, the correlation by the discrete Fourier transform of only the phase components of the two images is calculated. First, in step 57, the sample transmission enlarged image 59 by the electron beam 73 incident perpendicularly to the sample along the electron beam optical axis 72 is projected onto the scintillator 16 and imaged using the TV camera 17 or the like. This transmission image is referred to as a transmission image 1. Next, in step 58, the electron beam incident on the sample is incident on the electron beam optical axis with an arbitrary tilt deflection angle using the two electron beam deflection coils on the upper part of the sample. A sample transmission enlarged image having the same field of view as the transmission image 1 on the dropped scintillator is captured as a transmission image 2 using a TV camera or the like. With respect to these transmission image 1 and transmission image 2, in step 60, a correlation index (degree of coincidence) of discrete Fourier transform using only the phase component is obtained. The principle of calculation is as shown in [Formula 1] to [Formula 7].

  Based on the correlation index calculated in step 60, the process proceeds to step 61, where it is determined whether the current field of view is bright enough or inappropriate. This determination result is given four branches according to the value of the correlation index.

(1) If the correlation index is 0, it is determined that the current visual field cannot be searched and measured, and the flow returns to step 55 to search for another visual field.
(2) When the correlation index is 100, it is determined that the current field of view cannot be appropriately obtained as a sample transmission magnified image, such as a field of view where there is no sample due to photographing on the mesh of the sample holding mesh or sample breakage. Then, the process returns to step 55 to search for another visual field.
(3) If the correlation index is equal to or greater than a preset reference value and is not 100, it is determined that the current visual field is a visual field suitable for measurement and search, and the process proceeds to the next step 62. The reference value of the correlation index is set in advance as a threshold necessary for the automatic focus correction of the specimen transmission magnified image to operate sufficiently satisfactorily. The correlation index changes depending on the amount of shift between the two transmission images, the contrast between the transmission image form and the background, the S / N of the image, and the like. In the flowchart of FIG. 11, the threshold value is set to 5 from an experimentally obtained value. In step 62, automatic focus correction is performed according to the method shown in [Equation 8] to [Equation 12]. After the automatic focus correction is completed, the process proceeds to step 63.
(4) If the correlation index is less than the reference value and not 0, it is determined that the current visual field is suitable for measurement and retrieval, but the complete operation of automatic focus correction cannot be guaranteed. Move to 63. In step 63, the search target pattern having an arbitrary shape set in step 52 is called from the storage device, and the microprocessor searches for a desired shape pattern in the current visual field. In step 64, the number of forms determined as the same form pattern as a result of the search, image registration, display, analysis, and the like are performed. Further, the coordinates of the selected visual field are stored in the recording device 47.

  Further, the field of view in which the form targeted for the search is detected is displayed on the display device, differentiating from the field of view in which the form to be searched is not searched. For example, the field of view in which the search target form exists is displayed in red, and the field of view in which no search target form exists is displayed in gray. Further, the coordinates of each field of view in which the search target form exists are displayed on the display device and simultaneously stored in the recording device. Finally, in step 65, it is determined whether or not the automatic visual field search is to be continued. If the search is executed again, the process returns to step 55 and the series of operations is repeated. If the search is not executed again, the process ends.

  According to the present embodiment, only the visual field suitable for observation can be retrieved in the same manner as in the third embodiment with the electron microscope that automatically moves the visual field and retrieves the target structure. The conventional automatic search could not separate the field of view suitable for observation and the field of vision unsuitable for observation, but by searching the target structure only for the field of view suitable for observation as in this embodiment, Significant time savings can be achieved. In addition, there is a possibility that the lens focal length may change in each field of view due to sample breakage or warping, but automatic focus correction is performed when moving the field of view, thus preventing deterioration in pattern matching accuracy due to image blurring due to defocusing. Can do. Furthermore, when the search is performed once and the field of view where the target structure is present is observed again, the sample stage 13 is driven by selecting the stored coordinates, or the target stage using the electron beam deflector is selected. It is also possible to obtain a transmission enlarged image of the visual field instantly.

[Embodiment 5]
As a fifth embodiment, referring to the flowchart of FIG. 12, the search of automatic search is automatically performed in an electron microscope that automatically moves and selects a field of view and automatically searches for a form that matches a search target pattern having an arbitrary shape. Shows how to improve accuracy and reliability.

  In FIG. 12, the setting of the enlargement magnification at step 71, the setting of the search target pattern at step 72, the condition setting for automatically moving the visual field at step 73, the origin setting for visual field movement at step 74, and the visual field at step 75 The movement and stop of the visual field movement in step 76 are the same as those in step 31 to step 36 in FIG. 9 or step 51 to step 56 in FIG.

  In step 77, the magnified sample transmission image 59 projected on the scintillator 16 of the selected visual field is photographed using the TV camera 17 or the like. The photographed transmission image is stored in a storage device. In contrast measurement in step 78, the captured image is loaded from the storage device, and a line profile of signal intensity (brightness) is created by the microprocessor. FIG. 13 shows a line profile corresponding to the sample transmission image.

In the sample transmission image (left side) in FIG. 13, the ratio between the background 65 and the signal intensity (brightness) of form 61 in the sample is represented as contrast C. When a measurement line 74 that crosses the form in the field of view is drawn and the signal intensity distribution on this measurement line is graphed, a line profile as shown on the right side of FIG. 13 is drawn (the horizontal axis is the pixel [pixel]. , Vertical axis is signal intensity (luminance) [arb.u]). Assuming that the background signal intensity is I ₀ and the form signal intensity is I ₁ , the contrast C is defined by the following [Equation 22].

  FIG. 13 is an example of a sample transmission image in which the contrast is experimentally changed, and there are a total of 34 search object forms. The numerical value described in the transmission image indicates the form 75 that is determined to be the same as the search target pattern as a result of the search. FIG. 13A shows the case where the contrast C = 1.1, and the form cannot be detected at all. FIG. 13 (2) shows a case where the contrast C = 1.2, and 25 out of 34 shapes could be detected. FIG. 13 (3) shows a case where the contrast is C = 1.6, and all 34 forms were detected. Thus, the number of searchable forms changes depending on the contrast of the sample transmission image. The number of detections is summarized in Table 1.

  For each contrast, the percentage of the number of detections with respect to the total number of forms of 34 is defined as detection accuracy and graphed as shown in FIG. As can be seen from FIG. 14, the detection accuracy increases as the contrast increases. In an image with a low contrast, the difference between the noise and statistical fluctuation in the background signal and the signal intensity of the form is small and cannot be separated, so that the search accuracy is lowered.

  Therefore, Step 78 to Step 80 are performed in order to improve detection accuracy and improve measurement reliability. The relationship between the contrast and the detection accuracy based on the experimental result as shown in FIG. 14 is stored in the ROM in advance, and it is determined in step 79 whether or not the measured contrast is a contrast that makes the detection accuracy equal to 100. . In the flowchart of FIG. 12R> 2, C> 1.4 is used as a criterion for determination. In step 79, if C> 1.4, the process proceeds to step 81 because 100% of the detection accuracy can be secured. If C ≦ 1.4, the detection accuracy is insufficient, and the process proceeds to contrast adjustment in step 80.

  In step 80, the electro-optical conditions for increasing the contrast are adjusted. In the transmission electron microscope, there are the following four methods (1) to (4) as methods for increasing the contrast. (1) Reduce the hole diameter of the objective movable aperture. (2) Give an appropriate amount of insufficient focus. (3) The captured image is processed using a microprocessor. (4) Lower the acceleration voltage. Here, the methods (1) and (2) will be described, and the detailed description of the methods (3) and (4) will be omitted.

  First, the method (1) uses the principle that if the hole diameter of the objective movable aperture is reduced to prevent unnecessary electron beam scattering, the image of pure information of the sample is formed and the contrast is increased. The movable diaphragm drive mechanism 31 connected to the objective movable diaphragm 15 is operated from the microprocessor 46 via the drive mechanism driver 32, and the contrast is increased by changing to a hole diameter smaller than the current hole diameter.

  The method (2) uses the principle of Fresnel diffraction. In an electron microscopic image, fringes due to Fresnel diffraction (Fresnel fringes) occur when there is a focus shift. Fresnel stripes increase the contrast between form and background. A moderately underfocused image is commonly used as an electron micrograph technique.

  In step 80, contrast improvement processing is performed by any one of the methods (1) to (4). For example, the contrast is increased by providing an appropriate insufficient focal amount corresponding to the magnification. When the contrast is adjusted in step 80, the process returns to step 77 to determine again whether or not the contrast is appropriate.

  The transmission image determined to have an appropriate contrast in step 79 is searched for a pattern form that matches the search target pattern set in advance in step 72 in the field of view. Next, the number of forms determined to be the same pattern in step 82, image registration, display, analysis, etc. are performed. Further, the coordinates of the selected visual field are stored in the recording device 47. Finally, in step 83, it is determined whether or not the automatic visual field search is continued. When the search operation is executed again, the process returns to step 75 to repeat the series of operations. If the search is not executed again, the process ends.

  As described with reference to FIG. 13, the search accuracy of pattern matching for searching for a field of view where a target structural pattern exists with an electron microscope that automatically moves and selects the field of view depends on the contrast of the specimen transmission magnified image. As in the present embodiment, by measuring and automatically adjusting the contrast of the transmission image, it is possible to improve the search accuracy of the target structure. Even in the same sample, the contrast is not always constant in all fields of view due to non-uniform staining during sample preparation. Therefore, the search accuracy can be improved by measuring the contrast in each field of view and automatically adjusting the contrast.

[Embodiment 6]
As a sixth embodiment, referring to the flowchart of FIG. 15, in an electron microscope that automatically moves and selects a field of view and automatically searches for a form that matches a search target pattern of an arbitrary shape, automatic focus correction and sample A method for searching for a desired form with high detection efficiency and high accuracy by performing automatic contrast adjustment of a transmission image will be described. The image calculation of this embodiment uses a calculation method for obtaining a cross-correlation limited to a phase component from a discrete Fourier transform of a transmission image. In FIG. 15, the operations from step 91 to step 96 are the same as the operations from step 31 to step 36 shown in FIG.

  Next, in steps 97 to 101, the same operations as in steps 57 to 61 in FIG. In step 97, an enlarged sample transmission image by an electron beam incident perpendicularly to the sample along the electron beam optical axis is captured and recorded as a transmission image 1. In step 98, the electron beam incident on the sample is incident on the sample by the two electron beam deflection coils at an upper part of the sample with a certain tilt deflection angle with respect to the electron beam optical axis, and the transmission image provided by the deflected electron beam is provided. A sample transmission enlarged image having the same field of view as 1 is captured and recorded as a transmission image 2. In step 100, a correlation index (degree of coincidence) of discrete Fourier transform using only the phase component is obtained for these transmission image 1 and transmission image 2. The principle of calculation is as shown in [Formula 1] to [Formula 7]. Using the calculation result of step 100, in step 101, it is determined whether the current visual field is a visual field suitable for search or not. The decision result is given three branches according to the value of the correlation index.

(1) The correlation index is 0 or does not satisfy a predetermined threshold. In this case, the process proceeds to step 102. In this embodiment, the threshold value obtained experimentally is 5.
(2) When the correlation index is 100. In this case, it is determined that the current field of view is either the image of the mesh portion of the sample holding mesh, or the field of view where there is no sample due to sample breakage, etc., and the process returns to step 95 to search for another field of view. To do.
(3) If the correlation index is not less than the arbitrarily set threshold value and not 100, the process proceeds to step 104.

  In step 102, the cause that the field of view is determined to be unsuitable is automatically investigated and the electro-optical conditions are adjusted. As in step 27 of the flowchart of FIG. 8 described in the second embodiment, (1) Investigation / adjustment of electron beam emission, (2) Investigation / adjustment of lens condition of irradiation lens, (3) Investigation / adjustment of electron beam current (4) Investigate and adjust the objective movable aperture hole diameter and hole position. Thereafter, the process proceeds to step 103, and the degree of coincidence of images obtained from the transmission image and the phase component of the discrete Fourier image is calculated in the same manner as in step 97 to step 100. If the coincidence is 0, it is determined that the field of view is not suitable for observation, and the process returns to step 95. If the coincidence is not 0, the field of view is determined to be observable and the process proceeds to step 104.

  In step 104, the focus correction is automatically performed by adjusting the objective lens current or the height of the sample stage from the result of calculation of the amount of positional deviation accompanying the image calculation. In step 105, the transmitted image 3 is projected onto the scintillator 16. The sample transmission enlarged image 59 is captured using the TV camera 17 or the like and stored in the storage device 47. This transmitted image 3 is a sample-transmitted image with a focus corrected and a normal focus.

  As described with reference to FIGS. 13 and 14, in order to automatically search for a form, if the contrast of the magnified specimen transmission image is not sufficiently secured, the search accuracy is lowered. Therefore, by the operation from step 106 to step 108, whether or not the transmission image 3 taken in step 105 has a contrast satisfying the search accuracy is measured and judged, and the contrast is automatically adjusted. In step 106, the transmission image 3 is called from the storage device, and a line profile of signal intensity (brightness) is created by the microprocessor 46. Calculate the contrast from the line profile. Next, it is determined whether or not the contrast is such that the detection accuracy is 100 from the relationship between the contrast and the detection accuracy of FIG. In the present embodiment, if the contrast is 1.4 or more, the detection accuracy reaches 100, so the determination threshold is set to 1.4. If the contrast is 1.4 or more in Step 107, the process proceeds to Step 109. On the other hand, if it has not reached 1.4, the routine proceeds to step 108.

  In step 108, the contrast is adjusted in the same manner as in step 80 of FIG. The methods for increasing the contrast include (1) a method of reducing the hole diameter of the objective movable diaphragm, (2) a method of giving an appropriate insufficient focus, (3) a method of image processing of a photographed image using a microprocessor, (4 There are four types of methods for lowering the acceleration voltage. Here, a method of increasing the contrast by giving an appropriate insufficient focal amount according to the observation magnification for the normal focus state automatically corrected in step 104 is used. In step 108, the insufficient focus data set in advance according to the observation magnification and stored in the ROM 58 is called, lens data to be applied to the objective lens coil 4 is output to the DAC 37, an analog signal is applied to the lens excitation power source 20, and the lens current is calculated. Output. In step 108, an appropriate insufficient focal amount is given, and the process returns to step 105 again, and a sample transmission enlarged image with high contrast is taken and stored in the storage device.

  If it is determined that the contrast of the signal intensity between the background and the form of the transmission image 3 satisfies the condition of 1.4 or more, the process proceeds to step 109 and matches the search target pattern set in step 92 in the field of view. Automatically search for the presence of a pattern form. The search target pattern is called from the storage device, and a search operation is performed by the microprocessor for the presence of the same pattern form in the field of view. The forms determined to be the same pattern are marked, the number of the forms is counted, the image is registered in the storage device, and the transmission image is displayed on the display device. Further, the composition analysis is performed on the form to be searched if necessary (step 110). Further, the coordinates of the selected visual field are stored in the recording device 47.

  Finally, in step 111, it is determined whether or not the automatic visual field search is continued. When the search operation is executed again and the next visual field search is performed, the flow returns to step 95 and the series of steps after step 95 is repeated. If the search is not executed again, the process ends.

  The time required to search for a target structure existing in one sample with an electron microscope that automatically moves and selects the field of view takes several hours or more, depending on the magnification of the electron microscope to be observed. If the search operation is performed unattended for several hours, there is a possibility that the brightness of the transmission enlarged image cannot be obtained satisfactorily due to, for example, the life of the filament of the electron gun or the abnormal current. In addition, it is impossible to make adjustments through the hands of an operator during the automatic search operation in order to achieve unmanned operation and speedup.

  Therefore, in the present embodiment, all the features of the first to fifth embodiments are combined. In the conventional electron microscope that automatically moves and selects the field of view, the structure of the target structure is searched by taking a picture of a broken sample or a field of view on the mesh where the structure is not clearly present. Since the electron microscope omits a structure pattern search operation in a field of view where a structure cannot exist or cannot be observed, it can be completed in about 1/20 of the conventional time. Furthermore, it is possible to automatically adjust the time-dependent change in the electro-optical conditions accompanying the repeated search operation for a long time. In addition, this electron microscope, which automatically searches for the visual field, can automatically correct for changes in the focal length of the objective lens due to part of the sample being broken or warped, and automatically changes the contrast due to uneven sample staining. Therefore, the target structure can be searched with high accuracy and high efficiency. In addition, since the coordinates and images where the target structure existed are recorded and stored in the recording device, when you want to perform observation, photography, and analysis of the sample transmission magnified image again after the sample search operation is completed. Can be obtained instantaneously at the target sample position.

1 is a schematic functional block diagram of an example of an electron microscope according to the present invention. The schematic diagram explaining the automatic visual field movement observation and search of a sample. The figure which shows the contrast with the sample transmission image example, background, and form. 6 is a flowchart for explaining a method for automatically moving and selecting a field of view and determining whether the field of view is suitable for observation or search. Schematic explaining the method of visual field movement. The figure which shows the line profile of the sample transmission image from which a visual field state differs. It is a flowchart explaining the method using a correlation method in order to determine whether a visual field is suitable for observation or a search, moving and selecting a visual field automatically using an electron microscope. It automatically moves and selects the field of view, determines whether the field of view is adequately bright for observation or search, and efficiently observes only the appropriate field of view. The flowchart explaining the method of automatic investigation and adjustment. It automatically moves and selects the field of view, determines whether the field of view is adequately bright for observation or search, and efficiently observes only the appropriate field of view. The flowchart explaining the method of measuring, displaying, and saving the same number of patterns as the search target pattern determined in advance by automatic investigation and adjustment. The schematic diagram which shows the example of a search object pattern and the search object pattern extraction in an imaging visual field. Explains how to automatically move and select the field of view, determine whether the field of view is bright enough for observation or search based on the cross-correlation of only the phase information, perform automatic focus correction, and search for a preset form pattern Flowchart. 6 is a flowchart for explaining a method for automatically moving and selecting a field of view and improving the accuracy of searching for a form that matches a search target pattern and improving the reliability. The schematic diagram which shows the outline of the contrast and signal intensity distribution of a sample transmission image. The graph which shows the experimental result of the change of the detection accuracy of the target search target pattern with respect to the change of contrast. It automatically moves and selects the field of view, determines whether the field of view is suitable for observation and search, and if so, performs automatic focus correction and searches for a form that matches the preset search pattern with high accuracy. The flowchart explaining the method of performing automatic adjustment to the state suitable for a search, when it determines with not being suitable. FIG. 6 shows a display example of a display device.

Explanation of symbols

1: electron gun, 2: first irradiation lens coil, 3: second irradiation lens coil, 4: objective lens coil, 5: first intermediate lens coil, 6: second intermediate lens coil, 7: first projection lens coil 8: second projection lens coil, 9: sample upper first deflection coil, 10: sample upper second deflection coil, 11: sample lower first deflection coil, 12: sample lower second deflection coil, 13: sample stage, 14: Sample, 15: Objective movable diaphragm, 16: Scintillator, 17: TV camera, 18, 19, 20, 21, 22, 23, 24: Lens coil excitation power supply, 25, 26, 27, 28: Electron beam deflection coil Excitation power source, 29: fine motion motor, 30: motor driver, 31: movable aperture drive mechanism, 32: drive mechanism driver, 33: TV camera control unit, 34: image capture interface, 35, 36 37, 38, 39, 40, 41, 42, 43, 44, 45: D / A converter (DAC), 46: microprocessor, 47: storage device, 48: arithmetic device, 49: display driver, 50: display device 51, 52: Interface, 53, 54: Rotary encoder, 55: Keyboard, 56: Mouse, 57: RAM, 58: ROM, 59: Sample transmission enlarged image, 60: Sample holding mesh, 61: Form, 62: Observation Field of view, 63: mesh portion, 64: origin, 65: background, 66: electron beam before deflection, 67: electron beam during deflection, 68: sample transmission electron beam before field movement, 69: sample transmission electron beam after field movement, 70: Field of view at the center of the sample, 71: Field of view at a distance d from the center of the sample, 72: Optical axis of the electron beam, 73: Electron beam, 74: Measurement line, 75: Detected shape , 76: form that has not been detected

Claims (8)

A support for supporting the sample;
Adjusting means for adjusting the optical conditions of the electron beam;
An imaging means for capturing a sample transmission image by an electron beam transmitted through the sample;
The degree of coincidence of two transmitted images taken by changing the optical conditions by the adjusting means is obtained, and when the two transmitted images coincide or the coincidence reaches a predetermined value, the transmitted image An electron microscope characterized by stopping the search of the field of view obtained or performing a search of a different field of view in place of the search of the field of view.   2. The electron microscope according to claim 1, wherein the degree of coincidence is determined by comparing a line profile formed based on luminance of the two images, a cross-correlation between the two transmitted images, or a phase-only correlation. An electron microscope characterized by being performed. A support for supporting the sample;
Adjusting means for adjusting the optical conditions of the electron beam;
An imaging means for capturing a sample transmission image by an electron beam transmitted through the sample;
The degree of coincidence between two transmitted images taken by changing the optical conditions by the adjusting means is obtained, and the two transmitted images do not coincide, or the coincidence degree is less than a predetermined value and a preset reference value When it is above, the transmission microscope and / or coordinate of the visual field which acquired the transmission image are displayed and / or preserve | saved, The electron microscope characterized by the above-mentioned.   4. The electron microscope according to claim 3, wherein the degree of coincidence is determined by comparing a line profile formed based on luminance of the two images, a cross-correlation between the two transmitted images, or a phase-only correlation. An electron microscope characterized by being performed. Adjusting means for adjusting the optical conditions of the transmitted electrons that pass through the sample;
Visual field moving means for sequentially moving the visual field of the transmission electron beam;
In an electron microscope provided with search means for searching for a search target pattern included in the field of view,
When two images obtained when the optical condition is changed are matched, the search of the pattern to be searched for the field of view is stopped and the next field of view is searched. microscope. Adjusting means for adjusting the optical conditions of the transmitted electrons that pass through the sample;
Visual field moving means for sequentially moving the visual field of the transmission electron beam;
In an electron microscope provided with search means for searching for a search target pattern included in the field of view,
An electron microscope characterized in that, when two images obtained when the optical condition is changed do not match, a search target pattern for the visual field is searched. Visual field moving means for sequentially moving the visual field of the transmission electron beam;
In an electron microscope provided with search means for searching for a search target pattern included in the field of view,
Determine whether the field of view is suitable for observation;
An electron microscope characterized by determining whether or not to search for a search target pattern included in the field of view based on the determination. In an electron microscope equipped with a lens that adjusts the focus of the electron beam,
An electron microscope characterized by determining whether or not a field of view of an electron microscope is suitable for observation based on two images obtained by focus adjustment by the lens.

Images