JP2002056798A - Spectrum data processing method of transmission type electron microscope using energy filter and its transmission type electron microscope - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide equipment which takes photos of element distribution and spectrum with high space resolution and high-energy resolution using a transmission type electron microscope having an energy filter. SOLUTION: When pictures are taken of element distribution of a sample and spectrum using the transmission type electron microscope of a energy filtering type, photography time of spectrum data concerning the element distribution or the spectrum pictured through a spectrometer, is time-shared so that it can respond to a drift of the sample. The plurality of data obtained by time sharing picturing are indexed as to their position and integrated and calculated by deducting background, to obtain the element distribution picture or spectrum by which the drift correction is carried out.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to spectral data processing of a transmission electron microscope of an energy filter system which divides the energy of an electron beam by a spectrometer and takes in the signal (spectral data on an element distribution image or spectrum of a sample). The present invention relates to a method and a transmission electron microscope, and more particularly, to a drift correction technique.

[0002]

2. Description of the Related Art In a transmission electron microscope using an energy filter (spectrometer), it is possible to measure not only the structure of a minute part of a sample but also an element distribution image, a chemical bond state, or a spectrum.

[0003] These analysis methods will be described sequentially with reference to FIG.

FIG. 3A shows a transmission electron microscope equipped with an energy filter capable of taking in an element distribution image of a sample or spectral data relating to its spectrum. The electron beam 3 emitted from the electron gun is converged by lenses 1, 2,
The sample is irradiated via the. When the electron beam 3 inelastically collides with an element contained in the sample, the electron beam 3 loses energy equal to the binding energy of the core electron inherent to the element, and excites the core electron.

[0005] The electron beam transmitted through the sample passes through the objective lens 4,
After passing through the intermediate lenses 5 and 6 and the projection lens 7, the magnetic field sector 1 constituting the spectrometer (energy filter) 103
The light is divided according to the energy by an energy slit 04, and is converted into an electric signal by a detector 105 via an energy slit 8, an image / spectrum selection lens 9, and a magnifying lens 10.

The spectrum at a certain energy position converging on the slit 8 converges on the slit 8 by acceleration voltage control corresponding to moving the slit, and the energy position is recognized from the relationship with the acceleration voltage control value.

The electric signal (spectral data) is subjected to data processing by an arithmetic unit of the electron microscope control system 11 such as a personal computer, and a spectrum or an element distribution image of an element of the sample component is obtained. This spectrum or element distribution image is displayed via the display device 12.

The electron beam 3 has a spectrum peak at an energy position equal to this energy loss, as shown in FIG. Therefore, by measuring the energy and intensity of the spectrum, it is possible to calculate the elements and their concentrations in the electron beam irradiation region in the sample, and to create an element distribution image as shown in FIG. is there.

As shown in FIG. 4 (a), the actual spectrum is observed on a background (spectral signal + BG0) where both the absolute value of the intensity and the intensity change are large, so that the measurement is performed as follows. It is performed as follows. That is, not only the intensity at the spectral peak position, but also the background intensity (BG1 and BG2)
) (Fig. 4. b, c). Based on this, if the intensity of the electron beam is I and the energy is E,

[0010]

(Equation 1)

The background intensity (BG.0) at the spectral position is calculated based on the power law equation (FIG. 4.D). The true spectral intensity can be obtained by subtracting the background intensity from the signal intensity at the spectral position (FIG. 4.E).

In a transmission electron microscope, by using an energy filter, as shown in FIG. 5, only an electron beam at a certain energy position can be taken into a two-dimensional image. The two signal intensities (element distribution + B.G.0) and the background intensities (B.G.1, B.G.2) of the two energy regions can be recorded as three two-dimensional element distribution image data. Yes (Fig. 5. In this case as well, the background intensity at the spectral position is calculated based on the power law equation as described above (FIG. 5.D), and the background intensity is subtracted from the signal intensity of the element distribution image at the spectral position to obtain a true value. (FIG. 5.D).

Normally, in a high-resolution image, there is a position shift due to a change in the sample position (drift). Therefore, as shown in FIG. Power-low operation,
Also, a difference operation from the signal strength is performed. In FIG. 2, the signal image (signal + B.G.
5B corresponds to the image shown in FIG. The same applies to the background.

[0014]

In the measurement of the element distribution image, signals (element images, spectra) and two
The capture time of each background image (B.G1, B.G2) needs to be 20 seconds to 100 seconds or more.

This time is one digit or more longer than the conventional time of 2 to 10 seconds for capturing a high-resolution structural image. on the other hand,
An electron microscope sample usually has a sample position fluctuation (drift) of 0.01 nm / sec. This drift is mainly caused by a temperature difference between the sample and the sample table when the sample is placed on the sample table, a thermal expansion when the sample is irradiated with an electron beam, and is also called a temperature drift.

Such a drift is a very minute variation, and even if it is not visually recognizable by the human eye, it becomes a variation that cannot be ignored in the above-described element distribution and spectrum measurement.

That is, in the electron microscope, if there is a positional shift due to the above-mentioned drift (mainly temperature drift), in the conventional observation of a high-resolution structural image, even if a resolution of 0.1 nm or less is guaranteed, a two-dimensional image is obtained. The resolution of the element distribution image and the spectrum is reduced to 1 nm because the imaging time is one digit longer than that of the high-resolution structure image.

If the image capturing time is shortened in order to reduce the influence of the sample drift, the signal intensity itself becomes insufficient and a proper image is not obtained. In addition, each of the three images in FIG. 5 already includes blurring of the image due to the sample drift, and improvement in resolution cannot be realized by simply performing precise alignment between the three images. The problem of insufficient signal strength is that the energy position of the spectrum cannot be determined accurately, that is, the energy resolution cannot be sufficiently obtained. As described above, in the element distribution image and the chemical bond state analysis using the conventional electron microscope, the spatial resolution, sensitivity, and energy resolution are all insufficient.

In the conventional image processing method, for example, as disclosed in JP-A-9-56707,
In the field of X-ray fluoroscopy equipment and the like, in order to suppress the effect of blurring due to body movement of the subject, the imaging time is
2. Description of the Related Art A technique for improving image quality by adding and synthesizing a plurality of images captured consecutively in time is known.

However, in the field of transmission electron microscopy, research on element distribution and spectral analysis using an energy filter is relatively new, so that a sufficient solution for how to correct sample drift is still presented. Had not been.

An object of the present invention is to obtain a sufficient signal intensity with high spatial resolution at the same time when acquiring an element distribution image or spectrum of a sample using a transmission electron microscope of an energy filter system. is there.

[0022]

Means for Solving the Problems The present invention basically comprises:
The configuration is as follows.

One is that when an element distribution or spectrum of a sample is photographed by using a transmission electron microscope of an energy filter type, the photographing time of spectral data relating to the element distribution or spectrum photographed through a spectrometer is shown. As shown in the principle diagram of FIG. 1, time division is performed so as to be able to cope with the drift of the sample, and a plurality of spectral data obtained by time division photographing are aligned and added (added) to calculate a background subtraction operation. As a result, a drift-corrected element distribution image or spectrum is obtained.

Here, as the spectral data to be time-divisionally photographed, for example, not only a signal image (signal + B.G.0) including a background at a spectral position, but also a back energy in two energy regions used in a power-low operation expression. There are ground images (BG.1 and BG.2)
A plurality of the images are taken in a time-division manner at short intervals.
These time-divided images (spectral data) are integrated after performing accurate positioning, and then power-low calculation and difference calculation are performed to obtain a high-resolution element distribution image with sufficient signal intensity. And spectra.

When aligning a plurality of the time-divided spectral data, in addition to the correlation method for obtaining the correlation between pixels of the data, the structure of the sample before and after the time-division imaging of the spectral data is determined. A method is proposed in which images are photographed, the structural images are compared, a drift correction amount in divided photographing is calculated, and spectral data is aligned based on the drift correction amount. That is, in the case of a structural image, there is an advantage that a clear image having a large SN ratio can be easily obtained in a short time, a signal alignment and a shift amount between structural images can be easily calculated, and a drift correction amount can be easily calculated.

The other is that, when imaging the element distribution or spectrum of the sample, the imaging time is time-divided so as to be able to correspond to the drift of the sample, and the drift based on the time-division is corrected to the sample stage. The present invention proposes a spectral data processing method including a step of performing while applying a mechanical force.
According to this method, by applying a mechanical force to the sample via the sample stage to resist the drift, the amount of shift due to the drift between the time-division photographed spectral data is reduced, and a plurality of images and spectra can be obtained. The drift correction amount in the case of alignment can be minimized, and a wide effective imaging region of an image or a spectrum can be secured.

Another method is to time-divide the imaging time of the spectral data relating to the element distribution or spectrum so as to correspond to the drift of the sample, and to photograph at least two structural images of the sample before the time-division imaging is started. Alternatively, a structural image is taken before and after each time-division imaging, a drift correction amount of the time-division imaging is calculated from the structural images, and the sample is prepared for each time-division imaging based on the drift correction amount. And proposes a method of performing time-division photographing by adding position correction to. This drift correction method also reduces the amount of drift caused by drift between spectral data taken in a time-division manner, minimizes the amount of drift correction when aligning multiple images and spectra, and reduces the effective imaging area of images and spectra. Can be widely secured.

[0028]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 6 is a system diagram of an energy-filtering transmission electron microscope according to one embodiment of the present invention.
Is a schematic diagram of the spectral data processing in the present embodiment, and FIG. 8 is a flowchart thereof.

In FIG. 6, the same symbols as those in FIG. 3 indicate the same elements.

The difference between this embodiment and the transmission electron microscope of FIG. 3 lies in the configuration of the control system 11. The system operation will be described with reference to the flowchart of FIG.

In this embodiment, the setting unit 115 inputs the imaging conditions of the spectral data, for example, the element distribution, the time-division imaging time for spectral imaging, and the number of times of imaging. When the measurement of the sample analysis is started, the imaging control device 114 controls the detector 105 (for example, a CCD camera) based on the above-mentioned imaging conditions, and controls the imaging time and the number of imaging of the spectral data. As described above, the type of the spectral data is the signal image of the element distribution or spectrum (signal + BG0) and 2
Background image of two areas (BG.1, BG.2)
It is.

The spectral data distributed by the spectrometer 103 is converted into an electric signal by the detector 105 and then stored in the storage device 113 via the input / output port 111.

In FIG. 8, for a sample (m), first, time-division photographing data serving as a signal image is taken in (n) times based on the condition setting (1). Next, condition setting (2),
In (3), the time-division data of the backgrounds BG.1 and BG.2 are fetched (n) times.

The arithmetic unit 112 converts the spectral data (time-division photographing data) into a signal image (element distribution image + BG0 in FIG. 7) and a background image (B .G.1 distribution image and B.G.2 distribution image) are aligned (superimposed) by obtaining a cross-correlation between pixels using a correlation method, and are added and stored in the storage device 113. . After that, B.G.0 is calculated by a power-low calculation, and the background B.G.0 is subtracted from the signal image to perform a calculation, thereby forming an element distribution image or a spectrum image and outputting it to the display device 12.

In the above embodiment, the image registration of the time-division photographing data is performed based on the correlation method. Instead, the correction drift amount is calculated to perform the image registration of the time-division photographing data (spectral data). It is also possible to do.

That is, as shown in FIG. 9, time-division photographing (measurement) of spectral data (element distribution image or spectrum).
Before and after (or before time-division imaging), two structural images (1) and (2) of the sample are captured with a time lag. Then, by comparing the structural images, the drift moving direction and the drift amount or the drift speed are calculated, the drift correction amount (including the directionality) is calculated based thereon, and a plurality of time divisions are calculated based on the drift correction amount. This is a method of aligning and integrating photographing data (spectral data). Although FIG. 9 shows only the element distribution image as the spectral data, the background images BG.1 and BG.2 are also aligned based on the drift correction amount obtained in the same manner as described above. . The reason why the structural image is used as the data of the drift correction amount is that the drift amount can be easily calculated as described above.

The subsequent data integration, power-low operation, and background subtraction operation are the same as described above.

When the embodiment shown in FIG. 9 is executed, the system shown in FIG.
Reference numeral 15 also inputs a photographing time and a timing of a structural image in addition to the time-division photographing time and the number of photographing times for elemental distribution and spectral photographing, and controls the detector 105 based on the conditions.

A specific example of the above embodiment will be described. As a sample, a TiN (50 nm) and Ru electrode (200 nm) were deposited on a silicon substrate, and a BaSrTiO3 film was deposited to a thickness of 50 nm by a sputtering method.

The laminated sample was cut, polished, and then subjected to ion milling to make the final sample thickness 0.1 μm.

The electron microscope used was: accelerating voltage: 200 kV, current density: 0.07 A / μm ² , opening angle of electron beam: 4 mrad, observation magnification: 1
It went at a magnification of 10,000. The observation energy position of the energy filter was Ba: 790 eV, Sr: 270 eV, Ti: 460 eV, and the energy window width was 30 eV in each case.

Two-dimensional detector 105 for recording images
Used a 1024 × 1024 pixel CCD camera. When two hours have elapsed after the supply of liquid nitrogen to the electron microscope main body, the drift amount of the sample can be suppressed to as small as 0.01 nm / sec.

Under these conditions, first, the energy window position was changed from 0 eV to 10 eV, and a structural image of the sample was obtained by exposing the inelastic scattered electrons by exposure for 1 second. This image is called a zero-loss image and corresponds to a conventional electron microscope image, but is extremely sharp because energy-loss electrons have been removed.

Next, the element distribution image of Ba was measured as follows. The energy window was set from 790 eV to 820 eV, and five signal images (time-division photographed data) were captured every 20 seconds. This is shown in (a) to (e) of FIG. The blur (resolution) of each image is 0.2 nm or less.

Next, 730 eV to 760 eV and 760 eV to 790 eV
At the two energy window positions, the background images at the two energy positions were captured in five-sheet increments of 20 seconds in the same manner as described above. This is shown in FIG. 14B and FIG.
To (e). Regarding the element distribution images of Sr and Ti, the observation energy positions were set to 270 eV and 460 eV, respectively, and the other operations were the same as described above. Finally, once again, the energy window position was changed from 0 eV to 10 eV, and a zero-loss (structural) image of the sample excluding inelastic scattered electrons was taken by exposure for 1 second. By comparing the two structural images before and after the element distribution image measurement, the direction and speed of the sample drift can be calculated. In this case, the drift direction was 0.01 nm / sec in the north-north-east direction as viewed from the observer, and it was confirmed that the drift direction and the speed were constant.

Next, a method of calculating the Ba measurement data will be described (Si and Ti are also calculated in the same manner as in Ba, so the description will be omitted).

The three types of five images are respectively shifted based on the measured drift direction (north-northeast) and the distance (0.2 nm / 20 s), aligned and integrated, and the three types of signal intensities are large. Images. This is shown in FIGS. In the image after the integration, the effective area is reduced by 3 nm and 1 nm in the vertical and horizontal directions, respectively, compared to the image before the integration. The energy positions of the two backgrounds, 745 eV and 775 eV, and
And the intensity of each pixel

[0049]

(Equation 2)

The values r and A are calculated for each pixel by substituting the values into I and E. Next, assuming that the signal energy position is 805 eV, the background intensity at the signal position for each pixel can be calculated. By subtracting the background intensity distribution thus calculated from the measured signal intensity distribution (including the background) and finally multiplying by the Ba scattering cross section (σ), a true Ba signal intensity distribution is obtained. This is shown in FIG.
(D). Image blur at this time (spatial resolution)
Is 0.2 nm.

The above results will be compared with the spectral data of FIG. 16 in which the photographing time is simply set to 100 seconds without time division.
FIGS. 16A to 16C show the signal and the background (two types), and FIG. 16D shows the Ba element distribution obtained from this. In this case, the spatial resolution is significantly reduced due to the direct influence of the sample drift, and the detailed distribution of Ba is not seen. As described above, each image is captured in a time-division manner, and then integrated to obtain a spatial resolution of 0.2 nm.
It can be understood that the two-dimensional element concentration distribution can be measured with such accuracy.

FIG. 10 is a system diagram of an energy-filtering transmission electron microscope according to another embodiment of the present invention. In the figure, the same reference numerals as those shown in FIG.
The same or common elements are shown.

In the embodiment described with reference to FIG. 9, the drift correction amount used for superimposing the time-division photographed data (element distribution image or spectrum image) is calculated from two structural images photographed at different times. In the present embodiment, the sample position is mechanically corrected via the sample position correction device 116 and its correction force applying mechanism (for example, a piezo element) 118 based on the drift correction amount.

One of them is that, under the control of the imaging control unit 114, at least two structural images of the sample are taken before the time-divisional imaging is started, and the arithmetic unit 112 calculates the drift direction and drift of the sample from those structural images. The speed is obtained, and an electric signal corresponding to a mechanical force for correcting the drift is calculated based on the drift speed and the drift direction. Then, the electric signal is applied to the piezo element 118 via the sample position correcting device 116, so that the time-division photographing of the spectral data is performed while applying the mechanical force to the sample stage.

By doing so, drift correction can be suppressed. In this embodiment, further, in order to improve image quality, spectral data obtained by time-division photography after time-division photography is used. The positions were integrated by the correlation method and integrated. Thereafter, the background is subtracted to obtain a drift-corrected element distribution image or spectrum.

According to this embodiment, it is possible to obtain an element distribution image or a spectrum image having a sufficient signal intensity at a high spatial resolution at the same time, and it is possible to suppress a shift amount of spectral data upon alignment. There is an advantage that the effective area of the integrated image can be secured as much as possible.

FIGS. 11 to 12 show another embodiment in which drift correction (sample position correction) is performed using the apparatus shown in FIG.

FIG. 11 shows that at least two structural images of a sample are taken before the time-division imaging of spectral data, and the drift direction and speed of the sample are calculated from the structural images (1) and (2). Then, a drift correction amount of the time-division imaging of the spectral data is calculated, and a position correction is applied to the sample for each time-division imaging based on the drift correction amount to perform the time-division imaging. This sample position correction is executed not only for the element distribution image or the time-division photographing of the spectrum but also for the backgrounds BG.1 and BG.2, which is shown as a flowchart in FIG.
This is an example.

In the transmission electron microscope system shown in FIG.
The calculation of the drift correction amount is performed by the arithmetic unit 112,
An electrical signal for correcting the sample position is converted to a sample position correcting device 11
6 is applied to the piezo element 118. Thereby,
The sample stage is returned to the direction in which drift is corrected by using a time interval between each time-division imaging.

In the embodiment shown in FIG. 13, a structural image is photographed before and after each photographing by time division, and the drift correction amount of the time division photographing (spectral data photographing) is calculated from the structural images before and after the photographing. The position correction is performed on the sample for each time-division imaging based on the drift correction amount. Such position correction is performed in addition to the background distribution B.G.1 and the background distribution imaging in addition to the time-division imaging of the element distribution image or the spectral image. This is also executed for BG2 shooting. Such imaging conditions can also be executed using the setting unit 115, the imaging control device 114, the sample position correction device 116, and the correction force applying mechanism 118 in FIG.

Also in this case, in order to improve the image quality, spectral data obtained by time-division photographing was integrated and aligned by the correlation method every time photographing by time-division was performed. Thereafter, the background is subtracted to obtain a drift-corrected element distribution image or spectrum.

[0062]

According to the present invention, when an element distribution image or spectrum is photographed (measured) using a transmission electron microscope equipped with an energy filter, the data of the spectral data signal and the background are measured in a time-division manner. After that, by performing accurate alignment and then integrating, measurement with high spatial resolution and sufficient signal strength at the same time becomes possible, and finally, high spatial resolution and high energy resolution Measurement can be realized.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of spectral data processing according to the present invention.

FIG. 2 is a diagram illustrating the principle of a conventional method of spectral data processing.

FIG. 3 is a system diagram of a conventional transmission electron microscope of an energy filter type, and an explanatory diagram of spectral data obtained thereby.

FIG. 4 is an explanatory diagram showing a conventional spectrum processing method.

FIG. 5 is an explanatory view showing a conventional element distribution meter processing method.

FIG. 6 is a system configuration diagram of an energy-filtering transmission electron microscope according to one embodiment of the present invention.

FIG. 7 is a diagram illustrating the principle of a spectral data processing method according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram showing a flowchart of a spectral data processing method according to one embodiment of the present invention.

FIG. 9 is a diagram illustrating the principle of a spectral data processing method according to another embodiment of the present invention.

FIG. 10 is a system configuration diagram of an energy-filtering transmission electron microscope according to another embodiment of the present invention.

FIG. 11 is a diagram illustrating the principle of a spectral data processing method according to another embodiment of the present invention.

FIG. 12 is a diagram illustrating the principle of a spectral data processing method according to another embodiment of the present invention.

FIG. 13 is a diagram illustrating the principle of a spectral data processing method according to another embodiment of the present invention.

FIG. 14 is an explanatory view showing an image of time-division photographic data in spectral data processing according to the present invention.

FIG. 15 is an explanatory diagram showing an image of time-division photographing data in spectral data processing according to the present invention.

FIG. 16 is an explanatory diagram showing an image of time-division photographing data in spectral data processing according to the present invention.

[Explanation of symbols]

1, 2 ... converging lens, 3 ... electron beam trajectory, 4 ... objective lens, 5, 6 ... intermediate lens, 7 ... projection lens, 8 ... energy slit, 9 ... image / spectrum selection lens, 10
... magnifying lens, 11 ... control system, 103 ... spectrometer, 104
... Magnetic field sector, 118... Correction force applying mechanism (piezo element).

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Fumiko Yano 5--20-1, Kamimizuhonmachi, Kodaira-shi, Tokyo Within the Semiconductor Group, Hitachi, Ltd. No. 20-1 F-term in Hitachi Semiconductor Co., Ltd. F-term (reference) 2G001 AA03 BA11 CA03 EA05 FA06 FA09 GA01 HA01 HA07 HA13 JA02 JA11 KA01 5C001 AA03 CC03 DD02 5C033 SS02 SS04 SS08

Claims (18)

[Claims] 1. An element photographed through the spectrometer when an element distribution or spectrum of the sample is photographed using an energy-filtering transmission electron microscope in which the energy of an electron beam transmitted through the sample is distributed by a spectrometer. Time division of the imaging time of the spectral data related to the distribution or spectrum so as to correspond to the drift of the sample, alignment of the plurality of spectral data obtained by time division imaging, integration, and subtraction of the background A spectral data processing method for a transmission electron microscope, wherein a drift-corrected element distribution image or spectrum is obtained. 2. The spectral data is a signal image including a background at a spectral position and a background image in two energy regions. These signal images and the background image are respectively time-divisionally photographed, aligned, and integrated. Then, the background intensity of the signal image at the spectral position is calculated by applying a power-low operation formula to the background integrated value of the two energy regions, and the background intensity and the integrated value of the signal image are calculated. 2. The spectral data processing method for a transmission electron microscope according to claim 1, wherein the element distribution image or spectrum is obtained by performing a subtraction operation of: 3. The spectral data processing method for a transmission electron microscope according to claim 1, wherein the unit time of the time division is set such that a resolution by drift correction is one decimal place or less of nm. 4. A method for capturing a structural image of a sample before and after the time-division imaging, calculating a drift correction amount in the time-division imaging by comparing the structural images, and obtaining a plurality of drift correction amounts obtained by the time-division imaging. 4. The spectral data processing method for a transmission electron microscope according to claim 1, wherein the spectral data is aligned based on the drift correction amount. 5. When capturing an element distribution or a spectrum of a sample using an energy-filtering transmission electron microscope in which the energy of an electron beam transmitted through the sample is distributed by a spectrometer, the imaging time is reduced to the drift of the sample. A spectral data processing method for a transmission electron microscope, comprising a step of performing time division so as to be able to cope with the time division, and performing the time division imaging while applying a mechanical force to the sample stage to correct the drift. 6. At least two structural images of a sample are taken before the time-division imaging, and a drift direction and a drift speed of the sample are obtained from the structural images, and the drift speed and the drift direction are determined based on the drift speed and the drift direction. 6. The spectral data processing method for a transmission electron microscope according to claim 5, wherein an electric signal of a mechanical force for compensating is calculated. 7. An element distribution or an element distribution photographed through the spectrometer when an element distribution or a spectrum of the sample is photographed using an energy-filtering transmission electron microscope in which the energy of an electron beam transmitted through the sample is distributed by a spectrometer. Alternatively, the imaging time of the spectral data related to the spectrum is time-divided so as to correspond to the drift of the sample, and at least two structural images of the sample are captured before the time-division imaging is started. Calculating a drift correction amount for imaging, and performing a time-division imaging by applying a position correction to the sample for each time-division imaging based on the drift correction amount. Spectral data processing method. 8. When an element distribution or a spectrum of a sample is photographed using an energy-filtering transmission electron microscope in which an energy of an electron beam transmitted through the sample is distributed by a spectrometer, an element photographed through the spectrometer is used. The imaging time of the spectral data related to the distribution or spectrum is time-divided so as to correspond to the drift of the sample, and a structural image is captured before and after each imaging by the time-division, and the time-division imaging is performed based on the structural images before and after the imaging. Calculate the drift correction amount of
A method for processing spectral data of a transmission electron microscope, comprising a step of performing position correction on the sample for each time-division imaging based on the drift correction amount. 9. After the time-division photographing, the spectral data obtained by the time-division photographing are aligned, integrated, and the background is subtracted to obtain a drift-corrected element distribution image or spectrum. The method for processing spectral data of a transmission electron microscope according to any one of claims 5 to 8. 10. In a transmission electron microscope in which the energy of an electron beam transmitted through a sample is distributed by a spectrometer and the element distribution or spectrum of the sample is photographed, spectral data relating to the element distribution or spectrum of the sample corresponds to the drift of the sample. A time-division shooting device,
A storage device for storing a plurality of spectral data obtained by the time-division photographing, and an arithmetic device for creating an element distribution image or a spectrum by aligning and integrating the plurality of spectral data and subtracting a background to perform an arithmetic operation. A transmission electron microscope comprising: 11. A transmission electron microscope in which the energy of an electron beam transmitted through a sample is distributed by a spectrometer and an element distribution or spectrum of the sample is photographed. An imaging device for performing time-divisional imaging so as to be able to cope with the drift of the sample; a storage device for storing the structural image and a plurality of spectral data obtained by the time-divisional imaging; An element distribution image or spectrum is calculated by calculating a drift correction amount in the time-division imaging from the structural images of the sheets, and performing a position-based integration of the spectral data obtained by the time-division imaging based on the drift correction amount, subtracting the background, and calculating the background. A transmission electron microscope, comprising: an arithmetic unit for generating the data. 12. A transmission electron microscope in which the energy of an electron beam transmitted through a sample is distributed by a spectrometer to photograph the element distribution or spectrum of the sample. A transmission-type electronic device comprising: an imaging device having a function of performing time-division imaging so as to obtain; and a sample position correction device for applying a mechanical force against the drift to the sample stage during imaging by the time-division imaging. microscope. 13. A sample position correcting apparatus, comprising: a storage device for storing a structural image of a sample and a plurality of the spectral data obtained by time-division imaging; and a sample device based on at least two structural images taken at staggered times. A calculating device that calculates an electrical signal corresponding to a mechanical force against the drift by calculating a drift direction and a drift speed of the device; and 13. The transmission electron microscope according to claim 12, wherein the transmission electron microscope comprises a piezo element for applying a force to generate a force. 14. A transmission electron microscope for distributing the energy of an electron beam transmitted through a sample by a spectrometer and photographing the element distribution or spectrum of the sample. An imaging device for performing time-divisional photography for spectral data, a storage device for storing the structural image and a plurality of spectral data obtained by the time-divisional imaging, and at least two images taken at staggered times A position correction device for calculating a drift correction amount in the time-division imaging from the structural image of the above, and correcting the position of the sample for each time-division imaging based on the drift correction amount. microscope. 15. The transmission electron microscope according to claim 13, wherein the timing of photographing at least two structural images with a time lag is before photographing the spectral data in a time-division manner. 16. The transmission electron microscope according to claim 11, wherein the timing for photographing at least two structural images with a time lag is before the time-division photographing of the spectral data or before or after the time-division photographing. 17. A timing for photographing at least two structural images with a time lag, before or after time-division photographing of the spectral data, or before or after each time-division photographing.
15. The transmission electron microscope according to 3 or 14. 18. The spectral data is a signal image including a background at a spectral position and a background image in two energy regions, and the image capturing apparatus captures each of the signal image and the background image in a time-division manner. A transmission electron microscope according to any one of claims 10 to 17.